X /<^. /^.^:S^
Digitized by the Internet Archive
in 2009 with funding from
Boston Library Consortium IVIember Libraries
http://www.archive.org/details/titanotheresofan01osbo
Department of the Interior
Ray Lyman Wilbur, Secretary
S Oi
U. S. GEOLOGICAL SURVEY
George Otis Smith, Director
Monograph 55
THE TITANOTHERES OF ANCIENT WYOMING,
DAKOTA, AND NEBRASKA
BY
HENRY FAIRFIELD OSBORN
VOLUME 1
\ ) i
DocuMmrs department
RECEIVED
JAM 2 m?
Wilbui Cross Library
Univeisity ci Connecticut
UNITED STATES
GOVERNMENT PRINTING OFFICE
WASHINGTON; 1929
Note. — Monograph 55 is issued in two volumes. Volume 1 contains Chapters
I- VII and Plates I-XLII; volume 2 contains Chapters VIII-XI, Plates XLIII-
CCXXXVI, an appendix, and the index to both volumes.
xi^ii
CONTENTS
Page
Letter of transmittal xix
Preface i xxi
Vertebrate paleontology in the national surveys xxi
Preparation of the present monograph xxi
Work by the author, 1878-1919 xxii
Research and collaboration xxii
Cooperation of museums xxiii
Work on text and illustrations xxiii
Summary of geologic and anatomic principles xxiii
Chapter I. Introduction to mammalian paleontology 1
Section 1. Exploration and research made in the preparation of this monograph 1
Section 2. Preliminary survey of the monograph and the conclusions presented 2
Range of the titan otheres in geologic time 2
Hay den's subdivisions of the Eocene and the Oligocene 5
Discovery of the titanotheres of the plains 6
Discovery of the mountain-basin environment of the titanotheres 6
Discovery and delimitation of periods of sedimentation and of life zones 8
Principle of local and continental adaptive radiation 10
Comparison of the four life phases of Europe and North America during Eocene and early Oligocene time 12
Old and new systems of classification 13
Old terminology retained 13
Linnaean methods of defining species, genera, and phyla of titanotheres 14
Recognition of many lines of descent; polyphyly the key to interpretation of the family 14
Relation of the phylogenetic classification to the Linnaean classification 15
Comparison between zoologic and paleontologic species 18
Proportions of the skull in bears and in titanotheres 19
Features distinguishing phyla of titanotheres - 19
Mutations of Waagen 19
Zoologic and paleontologic nomenclature 20
Summary of differences between old and new systems 22
Study of the evolution of single characters 22
Phylogeny of the nine typical families of the Perissodactyla 23
Wide geographic distribution of the Perissodactyla 24
Causes of evolution 27
Adaptive evolution and overs volution of the form of skull, tooth, and foot 27
Phyletic divergence in the evolution of new proportions in horses and in titanotheres 28
Evolution of the limbs and feet of the titanotheres 33
Origin of new characters as distinguished from changes in proportion 34
Velocity in the development of characters and in phylogeny 39
Summary of the evolution of the titanotheres 41
Section 3. Bibliography of literature cited or consulted in the preparation of Chapter I 42
Chapter II. Environment of the titanotheres and effect of adaptive radiation on their variation 43
Section 1 . Geology and geography 43
Correlation of early Tertiary events in the Rocky Mountain region with those in western Europe 43
Late Cretaceous and early Tertiary climates 45
Eocene geography of western North America and its relation to f aunal migrations 47
Geographic divisions and their bearing on migration 47
Character of the mountain-basin, plateau, and plains regions 51
Eocene topography in the Rocky Mountain region 51
Contrast in phj-siographic conditions east and west of the Rocky Mountain Front Range 53
Lateral and main river systems in the mountain-basin region 54
Section 2. Eocene and lower Oligocene formations and f aunal zones 56
First f aunal phase (basal Eocene) 56
Seventeen life zones 56
Basal Eocene time in Montana and New Mexico 60
Summary of faunal events of basal Eocene time 60
Basal Eocene faunal zones 63
Zones 1 and 2: Ectoconus and Polymastodon zones (Puerco fauna; part of Thanetian of Europe) 63
Zones 3 and 4: Deltatherium and Pantolamhda zones (Torrejon and Fort Union faunas; part of Thanetian
of Europe) 64
IV CONTENTS
Chaptek II — Continued.
Section 2 — Continued. Page
Second f aunal phase (lower Eocene) 64
Transitional basal Eocene faunas 64
Zone 5: Phenacodus-Nothodectes-Coryphodon zone (base of Wasatch formation of Big Horn Basin, first
Wasatch Ufe zone, Big Horn A; Cernaysian of Europe) 64
Early Eocene time 65
Lower Eocene faunal zones 68
Zone 6: Eohippus-Coryphodon zone (second Wasatch life zone, Big Horn B; lower Sparnacian of
Europe) 68
Zone 7: Systemodon-Coryphodon-Eohippus zone (third Wasatch life zone, Big Horn C; upper
Sparnacian of Europe) 69
Zone 8: Heptodon-Coryphodon-Eohippics zone (fourth Wasatch life zpne, Big Horn D and Wind
River A; lower Ypresian of Europe) 69
Zone 9: Lambdotherium-Eotitanops-Coryphodon zc*ne (fifth Wasatch life zone. Big Horn E, Wind
River B, and Huerfano A; upper Ypresian of Europe) 69
Transitional lower to middle Eocene deposits 74
Huerfano formation of Colorado 74
Wind River beds and their fauna 74
Third faunal phase (middle and upper Eocene) 77
Correlation of American zones with European stages 77
Typical Bridger formation 78
Zone 10: Eometarhinus-Trogosus-Palaeosyops fontinalis zone (Bridger A and Huerfano B; lower
Lutetian of Europe) ^ 82
Zone 11: Palaeosyops paludosus-Orohippus zone (Bridger B; upper Lutetian of Europe) 84
Zone 12: Uintatherium-Manteoceras-Mesatirhinus zone (Bridger C and D, Washakie A, and Uinta
A; part of Bartonian of Europe) 84
Washakie Basin, Wyo 85
Stratigraphy of the basin 85
Zones 13 and 14: Metarlmius zone and Eobasileus-Dolichorhinus zone (Uinta B 1 and Washakie B 1;
Uinta B 2) 89
Uinta Basin, Utah 91
Physiographic, climatic, and volcanic conditions in the Uinta Basin during middle (?) and later Eocene
time 91
Geologic horizons in the Uinta Basin 91
Uinta B 1 (Metar/iinus zone = zone 13) 94
Uinta B 2 {Eobasileus-Dolichorhinus zone = zone 14) 94
Zone 15: Diplacodon-Protiianotherium-Epihippus zone (Uinta C 1; Ludian of Europe) 94
Summary of faunas of Uinta B and C 97
Adaptive radiation of the titanotheres in the Uinta Basin 97
Genera and species represented 97
Adaptive radiation of phyla 98
Fauna unrepresented 99
Zone 16: Theoretic Uinta C 2 99
Composite Eocene and lower Oligocene section at Beaver Divide 99
Fourth faunal phase (lower Oligocene) 101
Lower Oligocene mammals 101
Correlation of European and American forms 101
Zone 17: Titanotheriuni-Mesohippus zone (Chadron A, B, and C; Sannoisian of Europe) 101
Oligocene flood-plain sedimentation in the western Great Plains region 103
Conditions of deposition 103
South Dakota in Titanotherium time : 106
Rapid fluviatile sedimentation in the Cypress Hills, Saskatchewan 109
Slow sedimentation in South Dakota 109
Geographic distribution of the Chadron formation 110
Comparison of basins in western United States with the flood plain of the Nile 112
Faunal divisions in the Chadron formation 113
Three f aunistic levels determined 113
Stratigraphic distribution of species of Oligocene titanotheres 113
Hatcher's coUections, 1886-1888 115
Sources of error in determining stratigraphic levels 116
Mammalian life of the lower Oligocene Titanotherium zone 117
Notes on the habitat of the fauna of the clay and sandstone as a whole 120
Section 3. Adaptive radiation, primary and secondary, through change of environment a cause of diversification of the
titanotheres 121
Habitat of the ungulates 121
Polyphyly among hoofed mammals ' ; 121
The titanotheres and other extinct forms 121
The existing African antelopes 124
CONTENTS V
Chapter II — Continued. Page
Section 3 — Continued.
Continental adaptive radiation of the African antelopes 125
Adaptive radiation in the feeding habits of antelopes 126
Causes of variation and polyphyly among quadrupeds 127
Habits of the rhinoceroses parallel to those of the Oligocene titanotheres 128
Habits of the existing tapirs parallel to those of the Eocene titanotheres 128
Vertical geographic range of quadrupeds -- 129
Vertical geographic range of the titanotheres 129
Ten chief habitat zones of mammals _ 129
Conclusions as to habitats of the titanotheres 132
Section 4. Bibliography for Chapter II 132
Chapter III. Discovery of the titanotheres and original descriptions of the types 141
Section 1. History of discovery 141
The Oligocene titanotheres 141
The pioneer period: Prout, Owen, Evans, Leidy (1846-1873).-^ 141
Taxonomic arrangement and comparison 144
Work of Marsh and Cope (1870-1887) 144
Summary of Marsh's contributions 145
Summary of Cope's contributions 146
Reinterpretation and phylogenetio study (Osborn, 1887-1919) 146
Study of certain features 146
Geologic levels and succession of t3'pes (Hatcher, 1886-1893) 147
First European notice (Toula, 1892) 148
Distinctions of sex (Osborn and Wortman, 1895) 148
Monoph3'Ietic interpretation (Osborn, 1896) 148
Polyphyletic interpretation (Osborn, 1902-1919) 148
Recent discoveries by Lull, Lambe, and others 149
The Eocene titanotheres 149
Pioneer discoveries 149
Work in the Bridger, Washakie, and Uinta Basins by Leidy, Marsh, Cope, Scott, Osborn, and others
(1870-1889) 149
Discovery in Hungary 150
Princeton and Cope-Wortman expeditions 150
First systematic and evolutionary revision (Earle, 1889-1891) 150
American Museum and other explorations of the Eocene basins (1891-1895) 151
Investigations and explorations made in preparation for the present monograph (1900-1919) 152
Section 2. Original descriptions of types of Eocene titanotheres 153
Five rules for determining the names of titanotheres 153
The genera and species of Eocene titanotheres 155
Descriptions of the species- 157
Section 3. Original descriptions of types of Oligocene titanotheres 201
List of genera and species 201
Prout's descriptions of a fragmentary lower jaw, the first titano there made known to science 202
Pomel's genus Menodus, based on Prout's description and figure 204
Early notices by Leidy and others, 1850-1870 205
Species described by Marsh and Cope in 1873-1876 209
First notice of Canadian titanotheres by Cope, 1886 219
Species described by Scott and Osborn in 1887 219
Species described by Marsh in 1887 222
Canadian species described by Cope in 1889 225
Last five species described by Marsh, 1890-91 227
Last species described by Cope, 1891 229
First European Oligocene species, described by Toula, 1892 230
Species described by Osborn in 1896 and 1902 231
Species described by Lull in 1905 234
Species described by Osborn in 1908 235
Canadian species described by Lambe in 1908 235
Second European Oligocene species, described by Kiernik, 1913 240
Final Oligocene species described by Osborn in 1916-1919 241
Chapter IV. Systematic classification of the titanotheres 243
Section 1. Phyletic versus Linnaean S3'stem of classification 243
Neo-Linnaean systematic divisions (zoologic) and evolutionary phyla (paleontologic) 243
Superfamily names proposed by Osborn (1898) and Hay (1902) 243
Family names proposed or adopted by Marsh (1873), Flower (1875), Cope (1879-1889), and Osborn (1889) 243
Subfamily names and phsda proposed by Steinmann and Doderlein (1890), Earle (1892), and Riggs (1912) 245
Division of the Oligocene titanotheres into four contemporary phyla, Osborn (1902) 245
Reclassification of the Eocene and Oligocene subfamilies by Osborn (1914) 246
Species wrongly referred to the titanotheres 246
Section 2. Classification of the titanotheres adopted in this monograph 247
VI CONTENTS
Page
Chapter V. Evolution of the skull and teeth of Eocene titanotheres 251
Section 1. General principles of the study of the characters of the skuU and teeth 251
Proportion characters and tendencies of evolution distinguished by analysis and synthesis 251
Distinctions between proportion characters and new rectigradation characters 251
Steps in transformation of characters 252
Proportion and flexures of the skull 254
Summary as to craniometrj' 255
Changing proportions of the cranium and face 256
Cyptocephaly , or f aciocranial flexure 256
Dolichocephaly, brachycephaly, and correlation 257
Zygomatic cephalic indices in the titanotheres and other perissodactyls 259
Relative values of indices 259
Indices of skulls of Eocene and Oligocene titanotheres 259
Differences in terminology of skull proportions in titanotheres and in man 260
Contrast in features of brachy cephaUc and dolichocephalic skulls and teeth 261
List of abbreviations used in illustrations of skulls -262
Terminology of the upper molar teeth 263
Section 2. Introduction to the anatomy of the skuU and teeth of the Eocene titanotheres ._- 264
Tj-pes of skull of Eocene titanotheres 264
Feeding habits of broad-headed and long-headed titanotheres 264
Origin and structure of the "horns" in titanotheres 266
Proportion and rectigradation in the grinding teeth of Eocene titanotheres 267
Mechanism of the titanothere grinding teeth 269
Molarization of the premolars 270
Correlation of dimensions of upper and lower teeth 272
Geologic succession and geographic distribution of the Eocene titanotheres 272
Section 3. The lower Eocene titanotheres 273
Ancestral titanotheres of the Lambdolherium zone of Wyoming at the end of lower Eocene time 273
Physiographic environment at the end of lower Eocene time 273
Contrasts and resemblances between Lambdotherium and Eotitanops 276
Explorations and discoveries 279
Systematic descriptions of the lower Eocene titanotheres 279
Section 4. The middle and upper Eocene titanotheres 297
Phyla distinguished 297
Species of Palaeosj'opinae and Dolichorhininae from the upper Huerfano {Trogosus zone) 297
S3'stematic descriptions of the middle and upper Eocene titanotheres 297
The palaeosy opine group 297
The Manteoceras-DolichorJiinus group 357
Successors to the Manteoceras-Dolichorhinus group 434
Chapter VI. Evolution of the skull and dentition of OHgocene titanotheres 443
Section 1. Review of the environment, geologic succession, and geographic distribution of the lower Oligocene titano-
theres 443
Section 2. Introduction to the anatomy of the skull and the dentition of the Oligocene titanotheres 444
Horns : transformation, elongation 444
Nasals : expansion, abbreviation 446
Zygomatic arches: expansion, buccal plates 446
Occipital pillars : auditory meatus 446
Sexual characters common to all phyla 448
Teeth : distinctive features and evolution 448
Development of the skull and dentition 451
Summary of the replacement of the teeth in OUgocene titanotheres 455
Stages of wear of the adult grinding teeth 456
Age and other characters common to both sexes of titanotheres of all stratigraphic levels 456
Section 3. Division of the Oligocene titanotheres into groups and subfamilies 457
Characters of the skuU and teeth of the menodontine and brontotheriine groups 457
Characters and relations of the subfamilies 465
Possible Eocene ancestors of the brontotheriine group — 468
Section 4. Oligocene genera accepted as vahd in this monograph 469
Section 5. The menodontine group 470
Subfamily Brontopinae, including the phyla Manteoceras, Protilanotherium, Teleodus, Brontops, and Diploclonus... 470
Stratigraphic level and distinguishing features 470
Subfamily characters of Teleodus, Brontops, and Diploclonus 471
Comparisons and contrasts 471
Conspectus of characters of the subfamily ' 477
Conspectus of characters of species 478
Measurements of the Brontops series 479
Systematic descriptions of genera and species in the Brontops-Diploclonus phylum ; 481
CONTENTS VII
Chaptjsr VI — Continued.
Section 5 — Continued. Page
Subfamily Menodontinae 505
Systematic descriptions of genera and species in the Alloys phylum 506
The Menodus monophjdum 518
Systematic descriptions of genera and species in the Menodus phylum 522
Section 6. The brontotheriine group 538
Group characters 538
Sexual characters 540
Subfamily Megaceropinae 540
Systematic descriptions of genera and species in the Megacerops phjdum 541
Subfamily Brontotheriinae 550
Systematic descriptions of genera and species in the Broniotherium phylum 555
Chapter VII. Evolution of the skeleton of Eocene and Oligocene titanotheres 583
Section 1. Methods by which the titanothere skeleton has been studied 583
Principles of the evolution of the limbs of hoofed animals 583
Size and proportions of Eocene titanotheres 584
Divergence and convergence in the skeleton of polyphyletic series 586
Diverse adaptive types of limb structure ^ 586
Terms used in describing the skeleton of the titanotheres 588
Section 2. The postcranial skeleton of lower Eocene titanotheres 590
Subfamily Lambdotheriinae 590
Subfamily Eotitanopinae 59 1
Section 3. Middle Eocene groups and phyla J : 598
Double parallelism in the palaeosyopine and Manteoceras-DoKchorhinus groups 598
Family and subfamily characters of skeletal parts in middle Eocene titanotheres 599
Systematic descriptions of middle Eocene titanotheres 612
Subfamily Palaeosyopinae 612
Subfamily Manteoceratinae 631
Section 4. The postcranial skeleton of upper Eocene titanotheres 636
Subfamily Dolichorhininae 636
Subfamilies Telmatheriinae, Brontopinae?, and Diplacodontinae 652
Section 5. The postcranial skeleton of Oligocene titanotheres 662
Subfamily Brontopinae 664
Subfamily Menodontinae 678
Subfamily Brontotheriinae 689
Bibliography for Chapters III- VII 698
ILLUSTRATIONS
Plate
Frontispiece. Herd of Brontotherium platyceras. Page
I. .4, Eruption of the crater of Taal, Philippine Islands; B, Flooded area 140
II. A, Qjo Alamo, San Juan County, N. M.ex., looking north; B, Base of Puerco formation resting on eroded
surface of Ojo Alamo sandstone 140
III. A, Upper Torrejon beds, Torrejon Arroyo, Sandoval County, N. Mex.; B, Exposures of Puerco formation
east of Ojo Alamo, N. Mex 140
IV. A, Eohippus-Coryphodon zone. Little Sand Coulee, Clark Fork Basin, Wyo.; B, Phenacodus-Nothodecles-
Coryphodon zone, Clark Fork Basin, north of Ralston, Wyo 140
V. A, Typical "Lysite" locality, at Cottonwood Draw, north of Lost Cabin, Wyo.; B, Typical "Gray Bull"
locality, south of Otto, Big Horn Basin, Wyo 140
VI. A, A typical Huerfano locality, west of Gardner, Huerfano Basin, Colo.; B, A typical "Lost Cabin" locality,
east of Lost Cabin, Wind River Basin, Wyo 140
VII. A, Henrys Fork Table, looking northward across Henrys Fork, Bridger Basin, Wyo.; B, Grizzly Buttes, south
of Mountain View, Uinta County, Wyo 140
VIII. A, Northwest point of Haystack Mountain, head of Bitter Creek, Sweetwater County, Wyo.; B, View
southeastward from Laclede station, Sweetwater County, Wyo 140
IX. A, Columnar sandstones, LTinta A, White River Canyon, Uinta Basin, Utah; B, Panoramic view. White
River Canyon below Wagonhound Bend, Uinta Basin, Utah 140
X. A, Northern boundary of Coyote Basin, Uinta Basin, Utah; B, Divide between White River Canyon and
Coyote Basin, Uinta Basin, Utah 140
XL A, North face of Beaver Divide, Wind River Basin, Wyo.; B, Exposures at Wagonbed Spring, Beaver
Divide, Fremont County, Wyo 140
XII. A, Contact between Titanotherium zone and Pierre shale. Cedar Creek, Big Badlands, S. Dak.; B, Badlands
south of White River, Utah, showing the Diplacodon zone 140
XIII. " Mauvaises Terres" or Big Badlands of South Dakota 140
XIV. Exposures at Quinn Draw, Big Badlands, S. Dak., showing summit of Chadron formation 140
XV. A, South end of Sheep Mountain, near head of Corral Draw, Big Badlands, S. Dak.; B, Cedar Creek, Big
Badlands, S. Dak., showing the Oreodon zone overlying the Titanotherium zone 140
VIII CONTENTS
Plate Page
XVI. The region of the horn swelling in Palaeosyops, Manteoceras, and Telmatherium 266
XVII. The region of the horn swelling in Manteoceras, Mesatirhinus, and Dolichorhinus 267
XVIII. Restorations of the heads of four genera of Oligocene titanotheres 582
XIX. Incisors and canines of Brontotherium and Teleodus 582
XX. Upper and lower canines of Oligocene titanotheres 582
XXI. Left upper premolars of Oligocene titanotheres 582
XXII. Third left lower molar in Menodus and Brontotherium 582
XXIII. Juvenile jaw referred by Marsh to Brontops 582
XXIV. Juvenile jaws and teeth of Oligocene titanotheres 582
XXV. Superior deciduous and permanent grinding teeth of Menodus giganteus 582
XXVI. Type skeleton of Eotitanops princeps 702
XXVII. Mounted skeleton of Palaeosyops leidyi 702
XXVIII. Restoration of Palaeosyops of the Bridger Basin, Wyo 702
XXIX. Restoration of Manteoceras and Dolichorhinus of the Uinta Basin, Utah 702
XXX. Restoration of the skeleton of Dolichorhinus longiceps 702
XXXI. Manus and pes of Dolichorhinus longiceps 702
XXXII. Skeleton of Dolichorhinus longiceps 702
XXXIII. Mounted skeletons of Brontops dispar and Brontops robustus (type) 702
XXXIV. Mounted skeleton of Brontops robustus (type) , oblique front and side views 702
XXXV. Mounted skeleton referred to Brontops robustus? 702
XXXVI. Vertebral column of Brontop srobustus • 702
XXXVII. Manus and hind limb of Diploclonus tyleri . 702
XXXVIII. Mounted skeleton of Allops marshi L 702
XXXIX. Mounted skeleton of Brontotherium hatcheri, left side view ___ 702
XL. Mounted skeleton of Brontotherium hatcheri, right side view 702
XLI. Mounted skeleton of Brontotherium hatcheri, front view 702
XLII. Mounted skeleton of Brontotherium hatcheri, back view 702
Figure
1. "Fragment of the inferior maxillary of the left side" of Front's "gigantic Palaeotherium" 1
2. Type of Palaeotherium? proutii 1
3. Geologic ages and orogenic periods in North America 2
4. Successive and overlapping Oligocene and early Eocene formations of the Rocky Mountains 3
5. Map showing areas throughout the world in which remains of titanotheres have been found and areas in which titano-
theres were probably in migration during Eocene and Oligocene time 4
6. The Meek and Hayden Tertiary section of 1862 5
7. Panoramic section of the Big Badlands of South Dakota, looking southeastward across Cheyenne and White Rivers
to Porcupine Butte 6
8. Map showing the type locality of the Titanotherium zone on Bear Creek, S. Dak 7
9. Map showing cluster of typical lower, middle, and upper Eocene sedimentary basins in the Rocky Mountain region. _ 8
10. Restorations of Eotitanops borealis and Brontotherium platyceras 10
1 1 . Ambly poda : Skeletons and restorations of an ancestral and a specialized form 11
12. Diagram showing the gradual extinction of archaic mammals and their replacement by modernized mammals 14
13. Phenacodus and Coryphodon drawn to the same scale 15
14. Contrast between the Linnaean and phylogenetic systems of classification 16
15. The family tree of the titanotheres 17
16. Theoretic descent of existing members of the dog family from a common ancestor 19
17. Successive invasion of nine families of perissodactyls in North America and western Europe 23
18. Outlines of the body form of the perissodactyls, drawn to the same scale 25
19. The family tree of the Perissodactyla 26
20. Periods of expansion and extinction of the perissodactyls and contemporary forms 27
21. Phyletie divergence in the evolution of new proportions in horses and in titanotheres 29
22. Contours of the head and of parts of the mouth in browsing and grazing perissodactyls 30
23. Heads of lower Eocene and modern perissodactyls, showing changes of proportion and of the lip structure 31
24. Restorations of the heads of some of the principal types of titanotheres 32
25. Lower jaws of the first and the last of the titanotheres 33
26. Structure of the feet in extinct and living odd-toed ungulates 34
27. Restorations of nine species of titanotheres 35
28. Evolution of the skeleton of the titanotheres 36
29. Evolution of the skull and molar teeth in the titanotheres 37
30. Adaptive radiation in the evolution of the upper molar teeth in the perissodactyls 38
31. Three types of teeth of members of nine typical famihes of perissodactyls 39
32. The family tree of the perissodactyls, showing adaptive radiation of the nine families and thirty-five subfamilies 40
33. Outlines of the bodies of titanotheres at different stages of evolution 44
34. Map showing the known areas and the hypothetical areas of titanothere migration and habitat 45
35. General geologic sketch map of the Rocky Mountain region, showing existing topography and drainage areas and their
relation to areas of Eocene and lower Oligocene sedimentation 46
36. Map of western North America showing supposed routes of migration of animals 49
37. Map showing the orogeny of the western mountain and plateau region 50
CONTENTS IX
Figure Page
38. Geologic map of the Uinta Range, showing the Tertiary sediments of the Bridger Basin, Wye, at the north, and of the
Uinta Basin, Utah, at the south 52
39. Chronologic relations of formations in the mountain-basin region 54
40. Section of deposits near Barrel Springs, Washakie Basin, Wyo 55
41. Eocene and lower Oligocene mammalian life zones in eleven typical correlated areas in New Mexico, Colorado, Utah,
Wyoming, South Dakota, and Montana 59
42. Section of Upper Cretaceous and basal Eocene (Fort Union) deposits in Sweet Grass County, Mont 61
43. Section of Eocene deposits in the San Juan Basin, N. Mex 62
44. Columnar section of Cretaceous and Eocene sediments exposed along Bear River, Wyo., showing the typical Wasatch
group of Hayden 66
45. Generalized section through Upper Cretaceous and basal and lower Eocene deposits near Pumpkin Buttes, Powder
River Valley, Wyo 68
46. Composite section of the Eocene deposits of the Big Horn and Clark Fork Basins, Wyo 70
47. A typical "Lost Cabin" locality. Alkali Creek, Wind River Basin, Wyo 71
48. Section through the Wind River formation (lower Eocene) near Lost Cabin, Wyo 72
49. Map showing cluster of lower, middle, and upper Eocene sedimentary basins in southwestern Wyoming and northern
Utah, exhibiting parts of areas of the Wasatch, Wind River, Bridger, and Uinta formations 73
50. Sketch map of the region of the Huerfano and Cuchara formations in southern Colorado 74
51. Section of the Huerfano formation in southeastern Colorado 75
52. Section of exposures from lower Eocene to lower Oligocene at Green Cove, on Beaver Divide, Wind River Basin, Wyo_- 76
53. Section across Wind River Basin, Wyo., from Hudson to top of Beaver Divide 77
54. Map showing the Eocene sediments encircling the Uinta Mountains of southwestern Wyoming and northern Utah 78
55. Geologic section of the Bridger formation in the Bridger Basin, Wyo 80
56. Map of the Bridger Basin, Wyo., and section of the Bridger formation 82
57. Section of the lower part of the Bridger formation in the Bridger Basin, Wyo 83
58. Section of the upper part of the Bridger formation in the Bridger Basin, Wyo 86
59. Section of deposits near Barrel Springs, Washakie Basin, southern Wyoming 87
60. Section of the Washakie Basin, Wyo., from north to south 88
61. Sketch map of the Washakie Basin region, Wyo 88
62. Columnar section of Washakie Basin, Wyo., showing life zones 90
63. Section of the Uinta formation exposed in the north wall of White River Canyon, Utah 91
64. Section of the Uinta formation from Kennedy's Basin to White River Canyon, Utah 1 92
65. Section of the Eobasileus-Dolichorhinus and Metarhinus zones in the Uinta Basin, Utah 93
66. Badlands near mouth of White River, Uinta Basin, Utah 95
67. Section of deposits at Green Cove, Beaver Divide, Wyo 100
68. Section across the Wind River Basin, Wyo., from Hudson to top of Beaver Divide 101
69. Map showing exposures originally described as the "White River group" by Meek and Hayden 102
70. Facsimile of the Meek and Hayden Tertiary section of 1862 103
71. Map showing tributaries of Chej'enne River, S. Dak., and the type locality of the " Titanotherium beds" of Hayden. _ 104
72. Type locality of the " Titanotherium beds" of Hayden, on Bear Creek, S. Dak 105
73. Panoramic section of the Big Badlands of South Dakota 106
74. Section of the Big Badlands of South Dakota, showing the chief faunal zones of the OUgocene (White River group)
and the Miocene 107
75. Map showing principal exposures of the Chadron formation 108
76. Section showing the results of stratigraphic leveling in the Chadron formation (Titanotherium zone) in the badlands
of White River, S. Dak ■ 115
77. The family tree of the Perissodactyla 116
78. Geographic cross section showing the nature of the habitats of the larger existing ungulates and of the titanotheres as
illustrating adaptive radiation 122
79. Original radiation of the unguligrade Herbivora, Carnivora, and Insectivora, with adaptations to environment 123
80. Adaptations in the structure of the skull and teeth of Herbivora to diverse habits of feeding 125
81. Convergent adaptations in the structure of the limbs and feet of ungulates 125
82. Adaptive radiation in the feeding habits of antelopes 126
83. Mauvaises Terres, Nebraska 142
84. "Vertical view of the posterior tooth belonging to the lower jaw of Mr. Front's Palaeotherium" 143
85. Original figures of Front's "gigantic Palaeotherium" 143
86. Osborn's first restoration of Palaeosyops paludosus Leidy 151
87. Four stages in the origin and evolution of the horns in titanotheres 152
88. Leidy's cotypes of Palaeosyops paludosus 157
89. Leidy's type (holotype) of Palaeosyops major 158
90. Leidy's type of Palaeosyops humilis 159
91. Leidy's cotypes of Palaeosyops Junius 159
92. Marsh's type of Palaeosyops laticeps 160
93. Marsh's type of Telmatherium validus 161
94. Marsh's type of Limnohyus robustus 161
95. Cope's cotypes of Palaeosyops vallidens 162
96. Cope's cotypes of Limnohyops laevidens 163
97. Cope's type (holotype) of Limnohyus fontinalis ■ — 164
X CONTENTS
Figure Page
98. Cope's type (holotype) o{ Palaeosyops diaconus 1(55
99. Marsh's type of Diylacndon elatus 166
100. Type (holotype) lower jaw of Brachydiastematherium transilvanicum 167
101. T3'pe (holot^'pe) oi Leurocephalus cultridens 168
102. Type (holotype) of Palaeosyops borealis 168
103. Type (holotype) of Lambdotherium popoagicum 169
104. Cope's type of Lambdotherium brownianum 1 170
105. Type (holotype) of Palaeosyops hyognathus 170
106. Type (holotype) of skull of Palaeosyops megarhinus 171
107. Earle's cotypes of Palaeosyops minor 172
108. Earle's type of Palaeosyops longirostris 173
109. Type (holotype) of Telmatotherium diploconum 173
110. Type (holotype) of Telmatotherium cornutum 174
111. Type (holotype) of Sphenocoelus uintensis 175
112. T\'pe (holotype) of Diplacodon emarginatus 176
113. Cotypes of Manleoceras manieoceras {Telmatotherium vallidens) 179'
114. T^'pe (holotype) of Lambdotherium primaevum 180
115. Type (holotype) of Limnohyops prisons 180
116. Type (holotype) skull oi Limnohyops matthewi 180
117. Type (holotype) skull oi Limnohyops 7nonoconus _' 180
118. Type (holotype) skull of Palaeosyops leidyi 181
119. Type (holotype) of Palaeosyops grangeri 181
120. Type (holotype) of Palaeosyops copei 182
121. Type (holotype) skull of Manteoceras washakiensis 182
122. Type (holotype) skull of Mesatirhinus petersoni 183
123. Type (holotype) skull of Metarhinus fltwiatilis 183
124. Type (holotype) skull of Metarhinus earlei 183
125. Type (holotype) skull oi Dolichorhinus intermedius 184
126. Type (holotype) skull of Telmatherium ullimum 184
127. Type (holotype) of Telmatherium? altidens 185
128. Type (holotype) of Protitanotherium superbum 185
129. Type (holotype) skull of Telmatherium? incisivum 186
130. Type (holotype) of Telmatherium? incisivum -- 187
131. Type (holotype) skull of Manieoceras uintensis -- 187
132. Type (holotype) of Manteoceras uintensis 187
133. Type (holotype) skull of Dolichorhinus heterodon 188
134. Type (holotype) of Dolichorhinus heterodon 188
135. Type (holot.ype) skull of Dolichorhinus longiceps 188
136. Type (holotype) of Dolichorhinus longiceps 189
137. Tj'pe (holotype) skull of Mesatirhinus superior 190
138. Type (holotype) skull of Metarhinus riparius 191
139. Type (holotype) skull of Metarhinus cristatus 191
140. Type (holot3'pe) skull oi Dolichorhinus fluminalis . 192
141. Type (holotype) skull oi Rhadinorhinus abbolti 193
142. Type (holotype) teeth of Eotitanops gregoryi 193
143. Lower jaws of Lambdotherium and Eotitanops 194
144. Type (holotype) of Eotitanops princeps 195
145. Type (holotype) of Eotitanops major 195
146. Type (holotype) of Lambdotherium priscum 195
147. Type (holotype) of Lambdotherium progressum 196
148. Type of Diploceras oshorni 196
149. Type of Diploceras osborni 197
1.50. Type (holotype) skeleton of H eterotitanops parvus 198
151. Type (holotj'pe) skull of H eterotitanops parvus 198
152. Type (holotype) of H eterotitanops parvus 198
153. Cotypes of Telmatherium? birmanicum : 198
154. Type (holotype) of Lambdotherium magnum 199
155. Type (holotype) of Eotitanops minimus 199
156. Type (holotype) skull of Eometarhinus huerfanensis 200
157. "Vertical view of the posterior tooth belonging to the lower jaw of Mr. Prout's Palaeotherium" 203
158. Original figures of Prout's "gigantic Palaeotherium" 203
159. Type of Menodus giganteus 204
160. Owen's specimens of Palaeotherium? proutii ,-- 205
161. Type (holotype) of Palaeotherium maximum 206
162. Cotypes of Rhinoceros americanus 206
163. Cotypes of Palaeotherium giganteum 207
164. Type (holotype) of Megacerops coloradensis. 208
165. Type (lectotype) of Brontotherium gigas 210
CONTENTS XI
Figure Page
166. Type (lectotype) jaw of Symborodon ioruus ^ 211
167. Type (holotype) skull of Megaceratops acer 212
168. Type (holotype) skull of Megaceratops heloceras 213
169. Type (lectotj'pe) skull of Symborodon bucco 214
170. Type skulls of Symborodon altirostris, S. bucco, and Megaceratops acer 215
171. Type (holotype) skull of Symborodon altirostris 216
172. Type (holotype) skull of Symborodon trigonoceras 217
173. Type (holotype) skull of Brontolherium ingens 218
174. Type (lectotype) of Symborodon hypoceras 218
175. Type (holotype) of Anisacodon montanus 219
176. Cope's cotypes of Menodus angustigenis 220
177. Anterior part of skulls of " Megacerops coloradensis," Menodus iichoceras, and Menodus dolichoceras 221
178. Type (holotype) horns of Menodus platyceras 222
179. Type (holotype) skeleton of Brontops robustus 222
180. Type (holotype) lower jaw of Brontops dispar 223
181. Type (holotype) skull of Menops varians 223
182. Type (holotype) skull of Titanops curtus 224
183. Type (holotype) skull of Titanops elatus 224
184. Type (holotype) skull of Allops serotinus 225
185. Type of Menodus selwynianus 225
186. Type of Menodus syceras 226
187. Type skull of Diploclonus amplus 227
188. Type of Teleodus avus , 228
189. Type skull of Allops crassicornis 229
190. Type (holotype) skull of Brontops validus 230
191. Type (holotype) skull of Titanops medius ■ 231
192. Type (holotype) nasofrontal shield of Menodus peltoceras 232
193. Cotypes of Menodus? rumelicus 232
194. Type (holotype) skull of Titanotherium ramosum 232
195. Type skull of Megacerops hrachycephalus 233
196. Type (holotype) skull and lower jaw of Megacerops bicornutus 234
197. Type skull of Megacerops marshi 234
198. Type (holotype) skull of Brontolherium leidyi 235
199. Upper premolars of type skull of Brontolherium leidyi 235
200. Type (holotype) skull of Megacerops lyleri 236
201. Right manus and right hind limb of the type of Megacerops iyleri 237
202. Type (holotype) skull of Brontolherium haicheri 238
203. Type (holotype) skull of Symborodon copei 238
204. Type (holotype) jaw of Megacerops primilivus 239
205. Type (holotype) jaw of Megacerops assiniboiensis 239
206. Type of Titanotherium bohemicum 240
207. Type (holotype) skull of Allops walcotli 241
208. Type (holotype) jaw of Megacerops riggsi 242
209. Characteristic basal sections of horns of Oligocene titanotheres 245
210. Skulls showing different numerical and proportional characters in five separate phyla of titanotheres 253
211. Standard measurements of Eocene titanothere skulls 255
212. Unequal elongation of face and cranium in titanotheres and horses 256
213. Faciocranial flexure, or cyptocephaly 256
214. Faciocranial flexure in Patoeosyops and Dolichorhinus 256
215. Cranial proportions of Eocene titanotheres — Palaeosyops, Manteoceras, and Dolichorhinus 257
216. Cranial proportions in man and in the titanotheres 258
217. Natural and artificial brachycephaly and dolichocephaly ■ 258
218. Contra.sting forms of upper teeth in Eocene titanotheres 264
219. Skulls of Eocene titanotheres of the principal genera 265
220. Heads of Eocene titanotheres of four phyla 266
221. Upper and lower molars of bunoselenodont pattern 268
222. Upper and lower molar patterns of Hyracotherium 268
223. Bunoselenodont patterns of upper and lower molars in Tertiary perissodactyls 268
224. Relations of upper and lower molars in Telmatherium cultridens 269
225. Dental mechanism of titanotheres 269
226. Grinding teeth of a titanothere and an insectivore 270
227. Contrast of braohyodont and semihypsodont molars in titanotheres 270
228. Cross sections through second upper and lower molars of Lambdotherium and Menodus 270
229. Upper premolar-molar teeth of the earliest and latest known titanotheres 271
230. Reconstructed skeletons and restorations of Lambdotherium popoagicum and Eotitanops borealis 277
231. Lower jaws of Lambdotherium, Eotitanops, and Tapirus 278
232. Restored contours of skulls of Lambdotherium and Eotitanops . 278
233. Skull of Lambdotherium popoagicum, reconstructed , 281
XII CONTENTS
Figure Page
234. Lower premolars of three "species" or mutations of Lambdotherium 282
235. Upper and lower grinding teeth of Lambdotherium 283
236. Lower jaws and teeth of Lambdotherium popoagicum 284
237. Lower jaws and teeth of Lambdotherium popoagicum, side view 285
238. Front part of type lower jaw of Lambdotherium priscum 286
239. Incomplete lower jaw of Lambdotherium priscum 286
240. Jaws and teeth of Lambdotherium priscum and L. magnum 287
241. Lower jaw and teeth of Lambdotherium progressum • 288
242. Upper teeth of Lambdotherium progressum 288
243. Restoration of Eotitanops borealis 289
244. Skulls of the oldest known titanotheres, Lambdotherium popoagicum and Eotitanops borealis 290
245. Model of skull of Eotitanops gregoryi 291
246. Lower premolars and molars of Eotitanops 291
247. Lower jaws of Eotitanops gregoryi and E. brownianus 292
248. Lower jaw of Eotitanops borealis 293
249. Lower teeth of Eotitanops borealis 294
250. Skull of Eotitanops borealis, palatal and side views 294
251. Skull of Eotitanops borealis, top and occipital views 294
252. Lower jaw of Eotitanops princeps 296
253. Lower grinding teeth of three species of Eotitanops from the upper Huerfano formation 296
254. Skull sections of brachycephalic and dolichocephalic Eocene titanotheres 299
255. Cross sections of the skull m middle Eocene titanotheres 300
256. Three skulls typical of the palaeosyopine group 301
257. Distribution of Palaeosyops and associated fauna in the Bridger formation 301
258. Anterior part of skull of Limnohyops laevidens 306
259. Skull of Limnohyops priscus 307
260. Back part of skull of Limnohyops priscus 308
261. SkuUs of three species of Limnohyops 309
262. SkuU of Limnohyops matthemi 309
263. Skull of Limnohyops monoconus 310
264. Skull of Limnohyops laticeps 311
265. Third right upper molar of Limnohyops laticeps 311
266. Lower jaws of Limnohyops and Palaeosyops 314
267. Lower jaws of Palaeosyops 314
268. Lower jaws of three species of Palaeosyops 316
269. Young skull of Palaeosyops fontinalis 317
270. Upper molars of Palaeosyops fontinalis 318
271. Teeth of Palaeosyops fontinalis 1 318
272. Skull of Palaeosyops major 319
273. SkuU and head of Palaeosyops leidyi 324
274. Incisors and canines ot Limnohyops a.nd Palaeosyops 325
275. Skull of Palaeosyops leidyi 326
276. Type skull of Palaeosyops leidyi 327
277. Type skull of Palaeosyops leidyi, top view 328
278. Type skull of Palaeosyops leidyi, palatal view 328
279. Skulls of Palaeosyops major and P. leidyi 329
280. Lower jaws of Palaeosyops leidyi : 330
281. Skulls of Palaeosyops leidyi and P. copei? (aff. P. robustus) 331
282. Jaws and deciduous teeth of Palaeosyops leidyi? 332
283. Deciduous cheek teeth of Palaeosyops leidyi? 332
284. Fragments of jaws of Palaeosyops 333
285. Skull of Palaeosyops robustus 333
286. Hyperbrachycephalic old male skull of Palaeosyops robustus 334
287. Basicranial region of Palaeosyops robustus 334
288. Nasals of Palaeosyops robustus 335
289. Progressive hypsodonty of the molars in Telmatherium 341
290. Upper jaw of Telmatherium cuUridens 342
291. Upper and lower teeth of Telmatherium cultridens 343
292. Upper and lower teeth of Telmatherium cultridens, interlocked 343
293. Lower jaw of Telmatherium cultridens 344
294. Type skull and lower jaw of Telmatherium ultimum 346
295. Type skull of Telmatherium ultimum, side, front, and occipital views 347
296. Type skull of Telmatherium ultimum, palatal and top views 348
297. Paratype skull of Telmatherium ultimum 349
298. Lower jaw of Telmatherium ultimum 350
299. Hypothetical reconstruction of the skull of Telmatherium altidens 352
300. Lower jaws of Telmatherium ultimum and T. altidens 353
301. Type skull of Sthenodectes incisivus . — 356
CONTENTS XIII
Figure Page
302. Skulls of titanotheres of the Manteoceras-DoKchorhinus group 359
303. Skulls of Manteoceras manieoceras 363
304. Type skull of Manieoceras manteoceras 366
305. Skulls of Manteoceras manteoceras and Palaeosyops leidyi 367
306. Skulls of Manteoceras manteoceras and M. washakiensis 367
307. Skull of Manteoceras manteoceras, side view 368
308. Skull of Manteoceras manteoceras, anterior half 368
309. Incisors and canines of Manteoceras manteoceras 369
310. Lower jaw of Manteoceras 370
311. Skulls of Manteoceras manteoceras and M. washakiensis 371
312. Type skull of Manteoceras uintensis •- 373
313. Upper canines and incisors of Manteoceras uintensis 374
314. Restoration of Protitanotherium emarginatum 374
315. Lower jaws of Protitanotherium and Brachydiastematherium 375
316. Type skull of Protitanotherium emarginatum; reconstruction, side view 376
317. Type skull of Protitanotherium emarginatum, front and side views 376
318. Nasal region in three specimens of Protitanotherium 377
319. Sections of the nasals and horns of Protitanotherium emarginatum 377
320. Lower jaw of Protitanotherium emarginatum 378
321. Lower jaw of Protitanotherium superbum 381
322. Phylogenetic relations of the species of Metarhinus, Mesatirhinus, Dolichorhinus, and Rhadinorhinus 383
323. Top view of the skull in the Manteoceras-Dolichorhinus group 385
324. Palatal view of the skull in the Manteoceras-Dolichorhinus group 385
325. Leidy's cotypes of Palaeosyops Junius 386
326. Type skull of Mesatirhinus megarhinus 389
327. Type skull of Mesatirhinus petersoni 390
328. Skull of Mesatirhinus petersoni, side, top, and palatal views 391
329. Skulls of Mesatirhinus petersoni, front and occipital views 392
330. Incisors, canines, and premaxillae of Mesatirhinus 392
331 . Lower jaws of Mesatirhinus 394
332. Lower jaw of Mesatirhinus sp. with deciduous dentition 395
333. Imperfect cranium of Mesatirhinus petersoni? 396
334. Geologic section of the Bridger formation in the Washakie Basin 397
335. Restoration of Dolichorhinus longiceps ■ 398
336. Skull and lower jaw of Dolichorhinus hyognathus 398
337. Skulls of Dolichorhinus hyognathus and modern horse 399
338. Geologic section of the Eobasileus-Dolichorhinus and Metarhinus zones in the Uinta Basin 400
339. Skulls showing progressive dolichoceplialy in the Mesatirhinus-Dolichorhinus phylum, side view 401
340. Skulls showing progressive dolichocephaly in the Mesatirhinus-Dolichorhinus phylum, top and palatal views 402
341. Upper premolars of Mesatirhinus, Dolichorhinus, and Metarhinus 403
342. Skull of Dolichorhinus intermedins 406
343. Skulls of Dolichorhinus intermedins, D. heterodon, and D. longiceps 408
344. Skull referred to Dolichorhinus longiceps? 409
345. Hyoid apparatus of Dolichorhinus longiceps? compared with that of a modern tapir 410
346. Skulls of Dolichorhinus 411
347. Skull of Dolichorhinus hyognathus, palatal view 412
348. Skulls of Dolichorhinus hyognathus, front and occipital views 413
349. Skull of Dolichorhinus hyognathus, side view 413
350. Upper incisors and canines of Dolichorhinus hyognathus : 414
351. Lower incisors and canines of Dolichorhinus hyognathus 414
352. Left upper canine of Dolichorhinus hyognathus 414
353. Lower jaws of Dolichorhinus 415
354. Skull of Sphenocoelus uintensis 418
355. Type skull of Eometarhinus huerfanensis 419
356. Skull of Metarhinus fluviatilis 423
357. Right lower premolars of Metarhinus fluviatilis 424
358. Lower jaws of Metarhinus 425
359. liOwer jaw o{ Metarhinus? (Rhadinorhinus?) sp 426
360. Skull and deciduous teeth of type of lieterotitanops parvus 426
361. Type skull of Metarhinus earlei ; 427
362. Type skull of Rhadinorhinus diploconus, side and top views 432
363. Type skull of Rhadinorhinus diploconus, top and palatal views 433
364. Type skull of Rhadinorhinus diploconus, side, front, and occipital views 434
365. Skulls of Eotitanotherium osborni , 436
366. Nasals and horn swellings of Eotitanotherium osborni 437
367. Two upper raolara of Eotitanotherium (" Diploceras") osborni 438
368. Type skull of Diplacodon elatus, partial reconstruction, palatal view 439
369. Type skull of Diplacodon elatus, upper jaw and zygoma 440
XIV CONTENTS
Figure Page
370. Third and fourth upper premolars of Diplacodon elatus 440
371. Upper molars of Diplacodon and Proiitanotherium compared 441
372. Facial region of Eotitanoiherium osborni and Bronlotherium leidyi 441
373. Map showing areas in which remains of titanotheres have been found 443
374. Comparison of upper Eocene and lower OUgocene titanotheres 444
375. Sections at base of horn in the six chief generic types of Oligocene titanotheres 445
376. Position of the standard sections and contours of Oligocene titanotheres skulls 445
377. Male and female skulls of Bronlotherium gigas 446
378. Occipital view of skulls in different phyla of OHgocene titanotheres 447
379. Influence of progressive brachycephaly on the auditory region of perissodactyls 447
380. Inferior aspect of chin in Manleoceras • 449
381. Upper molars of Menodus giganteus and Allops marshi 450
382. Extreme dolichocephaUc and brachycephahc types of upper premolar-molar series in Oligocene titanotheres 450
383. Third left lower molar of Bronlotherium leidyi 451
384. Development of jaws and teeth, stage 4 452
385. Development of jaws and teeth, stage 6 453
386. Occiput of young skull of Brontops? brachycephalus 454
387. Stages of wear in the adult upper grinding teeth of Ohgooene titanotheres 455
388. Skull contours showing extreme divergence between Menodus giganteus and Bronlotherium platyceras 456
389. Skulls of the menodontine group, side view 459
390. Skulls of the bronototheriine group, side view 460
391. Skulls of the menodontine group, top view 461
392. Skulls of the brontotheriine group, top view 462
393. Skulls of the menodontine and brontotheriine groups, palatal view 463
394. Skulls of the menodontine and brontotheriine groups, front view 464
395. Lower jaws of the Bronlotherium phylum 465
396. Lower jaws of the Brontops and Menodus phyla 466
397. Lower jaws of the Diplodonus and Allops phyla 466
398. Heads of Oligocene titanotheres 466
399. Sections at base of horn in five principal lower Ohgocene phyla of titanotheres 468
400. Restorations of lower OUgocene titanotheres of the four principal genera 469
401. Skulls of Rhadinorhinus and Bronlotherium, palatal view 470
402. Skulls of Rhadinorhinus and Bronlotherium, side view 471
403. Skulls of Rhadinorhinus and Bronlotherium, top view 472
404. Lower jaws of Metarhinus fluvialilis and Bronlotherium halcheri 473
405. Progressive evolution of the upper premolars in Bronlotherium. and its predecessors 474
406. Progressive evolution of the upper premolars in Menodus and Brontops and their predecessors 474
407. Progressive evolution of the lower premolars in Bronlotherium and its predecessors 475
408. Progressive evolution of the lower premolars in Brontops and its predecessors 475
409. Phyla of the Brontopinae and Menodontinae 477
410. Evolution of the horns in the Brontops phylum 477
411. Basal section of the horns in the Brontops phylum 477
412. Progressive broadening of the nasals in the Brontops phylum 478
413. Lower jaws of Teleodus primitivus, Brontops brachycephalus, and Allops walcotti? _ 478
414. Sections and contours of skulls of Brontops brachycephalus 483
415. Upper canines and incisors of Brontops brachycephalus 484
416. Reconstruction of crushed skull of Brontops brachycephalus 485
417. Left upper grinding teeth of Brontops brachycephalus 486
418. Skull and horn region of Brontops brachycephalus? 486
419. Sections and contours of skulls of Brontops brachycephalus and B. dispar 487
420. Restoration of Brontops robuslus 492
421. Sections and contours of skull of Brontops robuslus 493
422. Skull of Brontops robuslus 495
423. Lovi'er jaws of Brontops dispar and B. robuslus 496
424. Sections and contours of skull of Brontops sp 497
425. Hyoid bones of Brontops compared with those of the tapir, black rhinoceros, and horse 497
426. Sections and contours of skulls of Diplodonus hicornutus and D. tyleri 498
427. Sections and contours of skull of Diplodonus ampins 499
428. Lower jaws of Diplodonus bicornutus and D. tyleri 503
429. Sections and contours of skulls of Allops walcotti and A. marshi 610
430. Upper teeth of Allops walcotti 511
431. Skull of Allops marshi 513
432. Lower jaws of Allops marshi and Allops? sp 514
433. Sections and contours of skulls of Allops serotinus and A. crassicornis 516
434. Coossified nasals and proximal part of horns of Allops? serotinus? 517
435. Sections and contours of skulls of Menodus heloceras and M. trigonoceras 519
436. Skull of Menodus heloceras 526
437. Lower jaws of Menodus (Symhorodon) lorvus and M. trigonoceras 527
CONTENTS XV
Figure Page
438. Upper teeth of Menodus ■proutii 528
439. Skull of Menodus trigonoceras 629
440. Skull of Menodus trigonoceras belonging with the mounted skeleton in the Colorado Museum, Denver 530
441. Restoration of Menodus giganteus 531
442. Skull of Menodus giganteus, front view 532
443. Skull of Menodus giganteus, palatal view 532
444. Sections and contours of skulls of Menodus giganteus and M. various 534
445. Sections and contours of skull of Menodus giganteus 535
446. Lower jaws of Menodus giganteus 536
447. Teeth and nasals of Menodus montanus i 538
448. Sections and contours of nasals and horns of Megacerops coloradensis 544
449. Sections and contours of skull of Megacerops bucco 545
450. Sections and contours of skulls of Megacerops copei and M. acer 546
451. Upper part of occiput of Megacerops acer 547
452. Skull of Megacerops acer, side and top views 548
453. Skull of Megacerops acer, palatal view 549
454. Restoration of Megacerops copei 549
455. Lower jaws of Megacerops assiniboiensis and M. riggsi 550
456. Sections and contours of skull of Megacerops? syceras 550
457. Composite sections showing the evolution of the horns and reduction of the free nasals in the Brontotherium phylum. _ 551
458. Basal sections of the horns in the Brontotherium phylum 552
459. Skulls of male and female brontotheres 552
460. Contrast in contour of horns and nasals between male and female brontotheres 552
461. Sections and contours of skulls of Brontotherium leidyi and B. hypoceras 558
462. Lower jaws of Brontotherium leidyi ' 559
463. Two lower molars and symphyseal region of Brontotherium? rumelicum 560
464. Lower jaws of Brontotherium rumelicum? and B. leidyi 561
465. Sections and contours of skull of Brontotherium? hatcheri 563
466. Skull of Brontotherium hatcheri, side view j 564
467. Skull of Brontotherium hatcheri, front view 564
468. Lower jaws of Brontotherium hatcheri and B. gigas 566
469. Sections and contours of skull of Brontotherium? tichoceras 567
470. Sections and contours of skulls of Brontotherium hatcheri and B. gigas 568
471. Lower jaws of Brontotherium gigas and B. medium 569
472. Sections and contours of skull of Brontotherium gigas? 572
473. Sections and contours of skull of Brontotherium dolichoceras 572
474. Skull of Brontotherium dolichoceras 572
475. Sections and contours of skulls of Brontotherium medium and B. curtum 674
476. Horns of Brontotherium curtum 576
477. Sections and contours of skull of Brontotherium curtum 576
478. Left horn and nasals of Brontotherium curtum? 577
479. Sections and contours of skull of Brontotherium ramosum 578
480. Restoration of Brontotherium platyceras 579
481. Sections and contours of skull of Brontotherium platyceras 580
482. Evolution of the skeleton in titanotheres 584
483. Estimated height at shoulder of Eocene and Oligocene titanotheres and tapir 585
484. The phyla of Eocene titanotheres, as represented by the manus 587
485. Progressive broadening of the magnum in Eocene titanotheres 587
486. Reconstructed skeleton and restoration of Lambdotherium popoagicum 591
487. Atlas and scapula of Lambdotherium popoagicum 591
488. Fore limb of Lambdotherium popoagicum 592
489. Forearm and manus of Lambdotherium popoagicum 592
490. Left manus of Lambdotherium and Eotitanops 592
491. Astragalus of Lambdotherium popoagicum 593
492. Restorations of Lambdotherium popoagicum, Eotitanops princeps, and E. gregoryi 593
493. Metatarsals of Eotitanops 593
494. Reconstructed skeleton and restoration of Eotitanops borealis 594
495. Atlas of Eotitanops borealis 595
496. Vertebrae of Eotitanops princeps 595
497. Radius of Eotitanops borealis 595
498. Lunars of Eotitanops 595
499. Manus of Eotitanops princeps 595
500. Humerus and femur of Eotitanops princeps 596
501. Pelvis of Eotitanops borealis -• 596
502. Left pes of cursorial and subeursorial Eocene Perissodactyla 597
503. Astragalus and calcaneum of cursorial and submediportal Eocene Perissodactyla 598
504. Astragalocalcaneal facets in lower Eocene Perissodactyla , 598
505. Left astragalus and calcaneum of Eotitanops sp 599
XVI CONTENTS
Figure Page
506. Metatarsal and tibia of Eotitanops major 599
507. Restoration of Eoiitmiops horealis 600
608. Atlas of Eocene titanotheres 601
509. Types of scapula in middle Eocene titanotheres 602
510. Tj'pes of fore limb in Eocene and Oligocene titanotheres 603
511. Characteristic details of radius and ulna in middle and upper Eocene titanotheres 604
512. Manus of lower and middle Eocene titanotheres 605
513. Comparison of the riglit scaphoid in middle Eocene titanotheres 605
514. Terminal phalanges of the manus in middle Eocene titanotheres and amyuodonts 605
515. Progressive graviportal adaptation in the pelvis of Eocene and Oligocene titanotheres ■ 606
516. Femora and tibiae of middle Eocene titanotheres 609
517. Distal end of the femur in Manteoceras and Amynodon 610
518. Angulation of the knee joint: relation of patellar facet to long axis of femur * 611
519. Inner side view of left fibula of Palaeosyops, Limnohyops, and Brontotherium 611
520. Comparison of pes in four species of middle Eocene titanotheres 613
521. Astragali of Eocene titanotheres 614
622. Calcanea of Eocene titanotheres 615
523. Left ectocuneiform tarsi of lower and middle Eocene titanotheres j 615
524. Principal measurements of the carpus and tarsus 615
525. Humerus, radius, and ulna of Limnohyops monoconus? 615
526. Left manus, radius, and ulna of Mesatirhinus petersoni 616
527. Manus, radius, and ulna of Limnohyops monoconus 616
528. Right scaphoid of Palaeosyops sp. and Limnohyops monoconus 617
529. Left hind limb of Limnohyops monoconus 618
530. Right pes of Limnohyops monoconus? 618
631. Ventral surface of sacrum of Limnohyops laticeps 618
632. Right OS innominatum of Limnohyops laticeps 619
533. Pelvis of Palaeosyops major 619
534. Right femur and tibia of Palaeosyops major 620
635. Astragalus and calcaneum of Palaeosyops major 620
536. Composite mounted skeleton of Palaeosyops leidyi 621
537. Manus of Palaeosyops leidyi 622
538. Pelvis of Limnohyops 624
539. Pelvis of Palaeosyops cf. P. leidyi 624
640. Left pes of Palaeosyops leidyi 626
641. Relations of facets of the astragalus and calcaneum in Palaeosyops 626
542. Atlas of Palaeosyops robustus 627
543. Atlas and axis of Palaeosyops leidyi? 627
544. Cervicals and dorsals of Palaeosyops robustus 627
645. Left scapula of Palaeosyops robustus 627
546. Bones of forearm of Palaeosyops 628
647. Left astragalus of Palaeosyops copei? 629
648. Fore limb of Palaeosyops copei? 629
649. Left manus of Palaeosyops copei? 629
550. Right hind limbs of Palaeosyops major and P. copei? 630
551. Atlas of Manteoceras manteoceras 632
562. Seventh cervical vertebra of Manteoceras manteoceras compared with that of Palaeosyops leidyi 633
553. Left humerus of Manteoceras manteoceras 633
554. Right manus of Manteoceras manteoceras 633
655. Pelvis of Manteoceras? 1 634
556. Femora and tibiae of Manteoceras manteoceras 635
557. Left astragalus of Manteoceras manteoceras 635
558. Restoration of the skeleton of Mesatirhinus petersoni 637
559. Restorations of Mesatirhinus petersoni and Palaeosyops leidyi 637
660. Atlas of Mesatirhinus megarhinus * 638
561. Humerus of Mesatirhinus megarhinus 638
562. Radius and ulna of Mesatirhinus petersoni 638
563. Left forearm and manus of Mesatirhinus petersoni? 639
564. Right manus and fragments of radius and ulna of Mesatirhinus petersoni 639
565. Left manus, radius, and ulna of Mesatirhinus petersoni? 639
566. Right scaphoid of Mesatirhinus and Manteoceras 639
567. Right manus of Mesatirhinus petersoni? 641
568. Left femur and tibia of Mesatirhinus petersoni? 642
569. Left pes of Mesatirhinus petersoni? 642
570. Left astragali of Mesatirhinus petersoni? 642
571. Left entocuneiform tarsi of Palaeosyops and Mesatirhinus 642
572. Pes referred to Mesatirhinus 644
573 Pes of Meiarhinus cf. M. earlei 644
CONTENTS XVII
Figure i'ags
574. Astragalus, calcaneum, and navicular of Metarhinus cf. M. earlei 644
575. Astragalus of Metarhinus cf. M. earlei 644
576. Left scapula of Metarhinus? sp 645
577. Left radius and ulna of Metarhinus earlei'? 645
578. Skeleton of a newly born animal, provisionally identified as Metarhinus sp 646
579. Provisional restoration of the skeleton of Dolichorhinus hyognathus 646
580. Vertebral column of Dolichorhinus hyognathus 647
581. Atlas referred to Dolichorhinus sp 647
582. Left scapula of Dolichorhinus? hyognathus 649
583. Humerus of Dolichorhinus hyognathus 649
584. Radius and ulna of Dolichorhinus hyognathus 649
585. Metatarsals of Dolichorhinus hyognathus 649
586. Manus of AmynodoJi and Mesatirhinus compared 650
587. Left fore limb of the amphibious rhinoceros Amynodon intermediusf 650
588. Left astragali of Dolichorhinus and allied types 651
589. Cervical vertebrae of Dolichorhinus longiceps? 651
590. Right fore limb of Dolichorhinus longiceps? 652
591. Manus of Dolichorhinus longiceps? 652
592. Hind limb referred to Telmatherium ultimum 653
593. Pes of Tehnatheriumf ultimum? 653
594. Vertebrae and fore limb of Diplacodon or Protitanotherium 654
595. Astragalus and calcaneum of Diplacodon or Protitanotherium 655
596. Left astragalus of Protitanotherium superbu7n 655
597. Incomplete ilium and ischium of Diplacodon elatus 656
598. Atlas and axis of Eotitanotherium osborni : 656
599. Vertebrae of Eotitanotherium osborni 657
600. Scapula of Eotitanotherium osborni 657
601. Humerus, radius, and ulna of Eotitanotherium osborni , 657
602. Manus of Eotitanotherium osborni 658
603. Femur, tibia, and pelvis of Eotitanotherium osborni 658
604. Pes of Eotitanotherium osborni 658
605. Restoration of skeleton of Eotitanotherium osborni 659
606. Mounted skeleton of Brontops ^ 670
607. Three views of mounted skeleton of Brontops -. 671
608. Scapulae of Oligocene titanotheres 673
609. Manus of Brontops? sp. and B. dispar? 674
610. Mounted skeleton of Brontops brachyeephalus? 676
611. Mounted skeleton of Brontops brachyeephalus?, oblique front view 677
612. Parts of skeleton of Allops crassicornis? 680
613. Pes of Menodus trigonoceras, referred, and M. heloceras 681
614. Manus of Menodus trigonoceras? 682
615. Restorations of Menodus trigonoceras and Allops marshi 683
616. Mounted skeletons of Brontops dispar? and Menodus trigonoceras. 684
617. Left astragalus of Menodus giganteus - 685
618. Cervical and first four dorsal vertebrae of Brontops robustus and Menodus giganteus 686
619. Manus referred to Menodus giganteus 687
620. Restorations of Brontotherium leidyi and B. platyceras 688
621. Atlas and axis of Brontotherium leidyi 689
622. Vertebrae of Brontops robustus and Brontotherium gigas 689
623. Scapulae of Oligocene titanotheres 690
624. Humeri of Brontops robustus and Brontotherium leidyi 691
625. Humeri of Megacerops? acer? and Brontotherium gigas? 691
626. Radii of Brontops robustus, Brontotherium leidyi, and Brontotherium gigas 691
627. Radius and ulna of Brontotherium 692
628. Ulnae of Brontops robustus, Brontotherium leidyi, and Brontotherium gigas 692
629. Olecrana of Brontotherium and Megacerops? 692
630. Manus of Oligocene titanotheres . 693
631. Manus and pes referred to Brontotherium gigas hatcheri 694
632. Manus and pes referred to Brontotherium hatcheri? 695
633. Manus referred to Brontotherium gigas, as restored 695
634. Pelvis of Brontotherium gigas hatcheri 696
635. Femora of Brontops robustus and Brontotherium leidyi 696
636. Tibiae of Brontops robustus and Brontotherium leidyi 696
637. Tibia and fibula of Brontotherium leidyi 696
638. Femora of Megacerops? and Brontotherium? 696
639. Pes of Oligocene titanotheres 697
101959— 29— VOL 1 2
LETTER OF TRANSMITTAL
Dr. George Otis Smith,
Director United States Geological Survey,
Washington, D. C.
Dear Sir: I have the honor to transmit herewith
a monograph on the evolution of a pecuharly American
family of quadrupeds known as the titanotheres.
This designation was given to them in 1852 by Joseph
Leidy while he was employed as vertebrate paleon-
tologist in David Dale Owen's survey of a part of
the ancient territory of Nebraska. This family is
one of a group of vertebrate animals whose fossil
remains, found in the western United States, were
long studied by Othniel Charles Marsh, my distin-
guished predecessor in this work in the United States
Geological Survey. Early in the eighties Professor
Marsh projected a monograph on the Brontotheridae
(here called the titanotheres), and subsequently he
made the largest and most valuable contributions to
our knowledge of this family and of its evolution. He
planned the monumental field work of John Bell
Hatcher, by which the great collection for the United
States National Museum was made, and he super-
vised the preparation of sixty lithographic plates,
which are here reproduced. Unfortunately he died
before he had even begun to prepare the manuscript.
The duty of continuing his work was intrusted to me
June 30, 1900, by your predecessor, Charles D. Wal-
cott. During this period of nearly 20 years I have
supervised the preparation of the monograph on the
Ceratopsia by Hatcher and Lull and have half com-
pleted the monograph on the Sauropoda. The mono-
graph on the Stegosauria has not yet been prepared.
The task of preparing the present monograph has
been long and difficult. First, it proved necessary
to reexplore the entire Eocene and lower Oligocene
series of rocks in Wyoming, Colorado, and South
Dakota, where the fossilized remains of titanotheres
are found, both to determine precisely their geologic
succession and to close up gaps in the stages of evolu-
tion; second, it proved necessary to examine and com-
pare the titanotheres of these geologic epochs in all
the museums of this country and in several museums
abroad; third, it pi-oved necessary, in order thoroughly
to understand the titanotheres, to discover and to
follow many side lines of investigation that have not
hitherto been followed in vertebrate paleontology.
This work has been done with the aid of many
specialists, foremost among whom is my junior col-
league Prof. William K. Gregory, without whose in-
telligent and unremitting cooperation the monograph
could never have been completed.
It is perhaps not too much to say that this work
has transformed our knowledge of the early Tertiary
geology of the Rocky Mountain basin region. First,
the six life periods recognized by Marsh and his no
less distinguished contemporary Edward Drinker Cope
may now be replaced by sixteen life periods, which may
be clearly defined and separated and certain of which
may be more or less precisely correlated with life
periods established for western Europe. Second, a
much clearer notion has been gained of the changing
geographic, physiographic, climatic, and volcanic con-
ditions in Wyoming and Dakota and of their influence
on the migration and succession of forms of life.
Third, the whole method of attack on problems of
vertebrate paleontology has been developed; we seek
to know the entire living animal, its musculature, its
mode of locomotion, and its feeding habits, in order
to insure the complete restoration of the body. Fourth,
the study of the many branches of this group has given
the most convincing demonstration that evolution,
even in any one geographic region, seldom moves along
a single line of descent ; more frequently it moves along
many lines — it is polyphyletic; in other words, it
radiates, following the principles of local adaptive
radiation. Finally, the history of the titanothere
family in its evolution from very small and relatively
weak forms into titanic quadrupeds, second in size
only to the elephants, has afforded us a unique oppor-
tunity to enlarge our previous knowledge of the actual
modes of evolution as well as to revise our theories as
to the causes of evolution and of extinction.
I desire to express my appreciation of the support
given by the Geological Survey under your direction
in the completion and publication of this work.
With the aid of many coworkers I have endeavored
to set a new standard of broad, thorough, and ex-
haustive research in vertebrate paleontology which
shall be worthy of the great geologic traditions of our
national Geological Survey. I trust that this mono-
graph, like Leidy's classic memoir of 1869, may ex-
ercise a permanent influence upon future studies of
the geologic history of the great West.
Henry Fairfield Osborn,
Vertebrate Paleontologist.
American Museum of Natural History,
December 19, 1919.
PREFACE
VERTEBRATE PALEONTOLOGY IN THE NATIONAL
SURVEYS
Joseph Leidy, Edward Drinker Cope, and Othniel
Charles Marsh, who successively served as members of
United States Government surveys of the West, were
the founders of American vertebrate paleontology.
Leidy's memoir of 1869, entitled "The extinct mam-
malian fauna of Dakota and Nebraska, including an
account of some allied forms from other localities,
together with a synopsis of the mammalian remains of
North America," marked the end of the first period of
exploration. Cope's great memoir of 1885, entitled
"The Vertebrata of the Tertiary formations of the
West," marked the end of the second period of explor-
ation.
Meanwhile the subject had become too broad to be
comprehended in a single work. Accordingly Marsh,
as vertebrate paleontologist, planned a series of ex-
haustive monographs on special groups of extinct
birds, reptiles, and mammals, which should treat in
great detail the anatomical structure and form the
basis of a systematic classification. For these mono-
graphs he carried out the most intensive field explora-
tions known to science and published a large number of
preliminary papers, which fairly revolutionized our
knowledge of these and many other groups. In 1880
the Fortieth Parallel Survey published his monograph
on the Odontornithes, an extinct group of birds of North
America. In 1883 the United States Geological Sur-
vey published his paper entitled "Birds with teeth,"
and in 1886 his monograph on the Dinocerata, an
extinct order of gigantic mammals. This was the
first of the series of five monographs projected for pub-
lication by the United States Geological Survey on
the Dinocerata, the Stegosauria, the Sauropoda, the
Ceratopsia, the Brontotheridae. The monograph last
indicated has developed into the present monograph
on the titanotheres, which covers a much broader field
than that contemplated by Marsh for the monograph
on the Bronototheridae.
For the monographs on the Ceratopsia and on the
Brontotheridae exploration on an unprecedented scale
was begun by the United States Geological Survey
under the direction of Marsh. For the four mono-
graphs on the Stegosauria, Sauropoda, Ceratopsia, and
Brontotheridae 204 superb lithographic plates were
completed under Marsh's direction. Altogether he
had been engaged on this work nearly 17 years when
death interrupted his monumental labors on March
18, 1899.
PREPARATION OF THE PRESENT MONOGRAPH
The first important step taken by Marsh in his series
of contributions to our knowledge of this extinct fam-
ily was the publication of his paper on "The structure
and affinities of the Brontotheridae," published in
1874, based on the collections at Yale University.
The second was his paper entitled "Principal charac-
ters of the Brontotheridae," published in 1876. In
the meantime he had made a geologic excursion to
White River in South Dakota, in the vicinity of the
Red Cloud Agency. This visit marks an interesting
epoch in the history of paleontologic exploration for
the titanotheres.
Late in the autumn of 1875 Marsh, accompanied by
an escort from Fort Laramie to the Red Cloud Agency,
went to the Badlands of Nebraska and Dakota. The
consent of the Indians was deemed necessary to permit
safe search for fossil bones in their country. This con-
sent was obtained with difficulty, and after it had been
obtained the Indians withheld their assistance. An
account of Marsh's visit is given in a manuscript en-
titled "Sketches of the life of Red Cloud," by Capt.
James H. Cook, of Agate, Nebr., at that time serving as
a scout for the United States Army. Captain Cook
writes:
It was in the autumn of 1875 that I visited the Red Cloud
Agency, which was at that time located on the White River, in
the northwestern part of Nebraska, the agency buildings stand-
ing about 2 miles up the river from the place where the city of
Crawford is now situated. The chief of the Sioux, Red Cloud,
made me welcome to his lodge.
It was on this visit that I first learned of the petrified bones
of strange creatures that had once occupied the lands to the
eastward of the agency. Two of Red Cloud's subchiefs,
American Horse and Little Wound, took me to the lodge of
Afraid of Horses, where I was shown a piece of bone, perfectly
petrified, containing a molar tooth .3 inches or more in diameter.
American Horse explained that the tooth had belonged to a
"Thunder Horse" that had lived "away back" and that then
this creature would sometimes come down to earth in thunder-
storms and chase and kill buffalo.
His old people told stories of how on one occasion man_v,
many j^ears back, this big Thunder Horse had driven a herd
of buffalo right into a camp of Lacota people during a bad
thunderstorm, when these people were about to starve, and
that they had killed many of these buffalo with their lances
and arrows. The "Great Spirit" had sent the Thunder Horse
to help them get food when it was needed most badly. This
story was handed down from the time when the Indians had
no horses;
While I was the guest of Red Cloud on this occasion, Prof.
O. C. Marsh, of the Smithsonian Institution and Yale Uni-
versity, came over from Fort Laramie to Camp Robinson and
the Red Cloud Agency to get permission to collect fossils in
XXI
XXII
the Sioux country. The Sioux, however, did not take kindly
to this proposition, thinking it was yellow lead (gold) that the
white chief wanted, not stone bones.
I met Professor Marsh at that time and talked with him.
I showed him the tooth the Indians had shown me. When I
returned to Red Cloud's lodge I told Red Cloud that Professor
Marsh was a friend of the "Great Father" (the President) at
Washington, and that I thought if he were allowed to hunt for
stone bones in the Sioux country he would be a good friend
to the Sioux people. Red Cloud said that if Professor Marsh
was a good man he would help the Sioux people to get rid of
the agent that was then in charge of the agency, whom the.y
disliked very much. This being brought to the attention of
Professor Marsh, he took the matter in hand, and an investi-
gation of affairs took place at the Red Cloud Agency, the re-
sult of which was at least pleasing to the Indians concerned,
as the agent was removed.
Professor Marsh was allowed to collect with his field parties
unmolested from that time on. He was named by Red Cloud
"Wicasa Pahi Hohu" (pronounced we-ch5-shJl pa-he ho-hii),
Man-that-Pioks-Up-Bones. The professor and Red Cloud
became friends to the extent that Red Cloud was entertained
at the home of the professor in New Haven, Conn., and the
two were photographed there with clasped hands and the
"peace pipe" between them.
The first collections made for this monograph were
those brought together from Colorado and South
Dakota, part of them under the direction of Marsh,
for the Peabody Museum of Yale University. By far
the greatest collection was that brought together by
John Bell Hatcher for the Geological Survey, now
preserved in the United States National Museum.
Between 1870 and 1891 Marsh published 14 papers
on these collections. These papers relate more or
less directly to the Brontotheridae; the last appeared
in 1891 and contained descriptions of three new types
from South Dakota — AUops crassicornis, Brontops
dispar, and Brontotherium medium.
WORK BY THE AUTHOR, 1878-1919
In the meantime the present author made his first
contribution to the history of this family in 1878 in
a paper on the results of the Princeton collections of
1877 and 1878 in the Bridger Basin. His second
contribution was made in 1887 in a paper entitled
"Preliminary report on the vertebrate fossils of the
Uinta formation collected by the Princeton expedition
of 1886." His third and fourth contributions were
made in 1890, in the two papers entitled, respectively,
"Preliminary account of the fossil mammals from the
White River and Loup Fork formations," which
related to a collection made in South Dakota by Dr.
S. Garman for the Harvard University Museum, and
"The MammaUa of the Uinta formation," Parts III
and IV, on the Perissodactyla. These have been
followed by 38 papers by the author, based chiefly
on his paleontologic and geologic expeditions in the
field for the American Museum of Natural History,
planned by the author and ably directed by Dr. J. L.
Wortman, Mr. O. A. Peterson, and Mr. Walter
Granger. To these indefatigable field explorers science
is indebted for the wonderful series of Eocene titano-
theres which have enabled us to trace the ancestry
of the Oligocene titanotheres and to establish all the
early phases in the history of this family. To Peter-
son, Earl Douglass, and Elmer S. Riggs in the Uinta,
and especially to Granger in the entire series from the
basal Eocene to the base of the Uinta, is due the
remarkable precision of the geologic records by which
the faunal life zones of the Eocene have been deter-
mined.
The stratigraphic succession of the Eocene and of
the lower Oligocene mammal life has been determined
chiefly by the field observations and collections of
Granger in the Eocene and of John Bell Hatcher in
the lower Oligocene and by the systematic examina-
tions of species by Dr. William Diller Matthew and
by the author.
RESEARCH AND COLLABORATION
Prof. William K. Gregory has been in the closest
cooperation with the author in all the details of the
preparation of the monograph since the beginning of
the work in the year 1900. Words are inadequate
to express the author's sense of indebtedness to his
former student and present colleague in the American
Museum and in Columbia University.
The author desires also to acknowledge his special
indebtedness to Mr. Granger for his valuable notes
and his cooperation in the preparation of the text
and the geologic sections, as presented in Chapter II,
on the Eocene and Oligocene formations of the Rocky
Mountains, as well as to Prof. William J. Sinclair
for his work on the volcanic nature of the middle
Eocene deposits and to Mr. Albert Johannsen of the
Geological Survey for his analyses of the material
of these deposits. It is hoped that that chapter will
furnish a key to future exploration of this mountain-
basin region as well as to the Oligocene sections of the
Great Plains. Matthew, by means of the rich col-
lections in the American Museum, has furnished
critical determinations for the discrimination of mam-
malian species in the sixteen life zones and has cooper-
ated with the author in the preparation of "Cenozoic
mammal horizons of western North America," pub-
lished by the Geological Survey in 1909 as its Bulletin
361, which forms the foundation of the more de-
tailed life-zone work whose results are presented in
Chapter II.
Details of the history of the collections at home and
abroad are presented in Chapter III under the head-
ing "History of explorations and discoveries and
original descriptions of the Eocene and Oligocene
titanotheres." Every known significant specimen
is referred to, its species and its sex are determined,
and its principal characters are described. This
monograph will furnish a much desired key to the
present and future collections and surveys in Wyo-
ming, Nebraska, Colorado, the Dakotas, and Assin-
iboia.
XXIII
COOPERATION OF MUSEUMS
To the museums of the United States, Great
Britain, and Bavaria, where titanothere remains are
preserved, the author is indebted for cordial coopera-
tion in furnishing materials for study and in affording
every possible facility for measurements and illustra-
tions. The author would mention especially Prof.
Charles Schuchert and Prof. Richard S. Lull, of the Yale
University Museum, present custodians of the great
Marsh collections, as well as their assistant, Mr.
Thomas A. Bostwick, who is in charge of all the field
records of Marsh. In connection with the superb
Hatcher collection in the United States National
Museum, which far surpasses any other in existence.
Dr. Charles W. Gilmore and Dr. James W. Gidley
have rendered every possible assistance. The author
is especially indebted to the director of the Carnegie
Museum at Pittsburgh, Dr. W. J. Holland, and to
Mr. O. A. Peterson of that museum for the liberal
use of collections of the Uinta titanotheres; also to
Mr. Earl Douglass of the same institution for his
invaluable field notes and observations on the Uinta
section. The systematic location of species in the
great Uinta section is due to the precise field work of
Mr. Elmer S. Riggs of the Field Museum of Natural
History, Chicago, an institution that is especially
rich in remains from the horizon known as Uinta B.
To his former colleague Prof. William B. Scott of
Princeton University, as well as to his colleague
Prof. William J. Sinclair, the author is indebted for
the liberal use of valuable collections, including many
types from several levels of the Bridger and from the
uppermost Eocene horizon, known as Uinta C.
From 1846, when the earliest remains of titanotheres
were found, until 1918 almost every year has added
one or more stages or types to the history of this
great family. The stages still to be discovered are
in the unknown interval between the uppermost
Eocene horizon, known as Uinta C, and the lowermost
Oligocene horizon, known as Chadron A.
WORK ON TEXT AND ILLUSTRATIONS
The great task of preparing the bulk of the manu-
script— a task performed between 1902 and 1918 —
fell upon Miss H. Ernestine Ripley, the work being
done chiefly from the dictation and notes of thp
author. The preparation of the bibliography and
the first revision of the entire manuscript were also
undertaken by Miss Ripley with interest and per-
formed with precision. The author warmly appreci-
ates this invaluable service to paleontology. The
final arrangement of the illustrations together with
the preparation of the accompanying legends, was
undertaken, under the author's general direction, by
Doctor Gregory with the cooperation of Miss Chris-
tine D. Matthew, Mr. Erwin S. Christman, and Mrs.
Lindsey Morris Sterling. The preliminary editorial
work has been performed with celerity and skill by
Miss Mabel Rice Percy, of the American Museum.
The final arrangement and verification of illustra-
tions and captions were the work of Miss Christine
D. Matthew.
The final editorial work and preparation of the text
for the printers were accomplished by Messrs. George
M. Wood and Bernard H. Lane, Mr. Wood continuing
the work as a member of the American Museum staff
after his retirement from the Geological Survey.
The illustrations, which are taken from many
sources, date back to the early lithographic figures of
Leidy. They include the unpublished lithographic
plates prepared under the direction of Cope, and
especially the superb lithographic drawings made
for the United States Geological Survey by Mr. F.
Berger under Marsh's direction. These lithographic
plates are supplemented by numerous plates based
upon photographs taken chiefly by Mr. A. E. Ander-
son of the American Museum staff.
The text and plates are adorned with reproductions
of the fine series of drawings from the pen and brush
of Mr. Christman and from the numerous pen draw-
ings of Mrs. Sterling. The geologic sections in
Chapter II are the work of Mr. William E. Belanske.
To Mrs. Sterling, Mr. Christman, and Mr. C. A. Weck-
erly of the Geological Survey were assigned the final
preparation for the photoengraver of all the illustra-
tions for the monograph, which, including those in the
Appendix, consist of 797 figures and 236 plates.
This review affords a partial explanation of the
great length of the period of time occupied by the
author in the preparation of this monograph. The
work has involved repeated explorations in the West
in search of the remains of all the ancestors of the
family and in establishing the full chronology. It has
necessitated repeated journeys to all the museums of
the country and long and painstaking research. The
greatest effort, however, has been expended on the
solution of the series of related problems in stratig-
raphy, in adaptation, in the origin of new characters,
in the mechanics of locomotion, in the modes of
evolution, and in the causes of evolution and of
extinction that presented themselves as essential to
the exposition of the life history of a long extinct
family. To restore the living and the lifeless environ-
ment of the Rocky Mountain region and to present
the titanotheres as living forms has been the persistent
purpose of this monograph.
SUMMARY OF GEOLOGIC AND ANATOMIC
PRINCIPLES
The following is a brief statement of the principles
developed and discriminated in this monograph :
1. The principle of the division and correlation of geologic
formations in Eocene and lower Oligocene time by mamma-
lian life zones and bv the subdivision of these zones.
XXIV
2. The principle of the correlation of local physiographic
diversity with the adaptive radiation, local and continental,
of titanotheres and other ungulates.
3. The principle of adaptive radiation as expressed in adap-
tations to aquatic, forest, savanna, and plains life at different
altitudes.
4. The principle of multiple lines of descent in the same
regions, of polyphyly and of polyphyletic evolution as more
common among ungulates than monophyletic evolution.
5. The principle of distinguishing phyla by contrasting pro-
portions of the head (dolichocephaly and brachycephaly) , of
the Umbs (dohchomely and brachj'mely) , of the feet (dolicho-
pody and brachypody), and of the teeth (hypsodonty and
brachyodonty) .
6. The principles of the lengthening and shortening of the
limb segments in harmony, respectively, with adaptation to
speed and to weight.
7. The principles of evolution by rectigradation (origination
of new characters) and by allometry (changes of proportion)
as effecting the chief changes in the hard parts.
8. The principles of continuity and of orthogenesis — the
direct continuation of animal form estabUshed in adaptation
to environment and of the evolution of new types irrespective
of external influences. '
The theoretic causes underlying these principles of
evolution are briefly stated in Chapter I, and the con-
clusions reached are summarized in Chapter XI.
THE TITANOTHERES OF ANCIENT WYOMING, DAKOTA,
AND NEBRASKA
By Henry Fairfield Osborn
CHAPTER I
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
SECTION 1. EXPLORATION AND RESEARCH MADE IN
THE PREPARATION OF THIS MONOGRAPH
The preparation of this monograph was actually
begun in 1846, when a part of a jawbone of a titano-
there was found in the region now known as South
Dakota and sent first to Dr. Hiram A. Prout of St.
Louis and then to Dr.
Joseph Leidy of Phila-
delphia for description.
This bit of bone gave the
first hint to science of
the wonderful deposits
of vertebrate fossils in
the Rocky Mountain
region that have revo-
lutionized vertebrate
paleontology. The de-
tails of this epoch-mak-
ing discovery are given
in Chapter III. The
original fragment bears
the generic name Meno-
dus, which was assigned
to it by the keen system-
atic paleontologist of
France, Nicolas Auguste Pomel,who gave it the specific
name giganteus. Menodus giganteus is thus the first
titanothere known to science, and it is a representative
of the most imposing family of mammals that was
evolved in ancient North America.
Figure 1. — "Fragment of the inferior maxillary of the left side'
Front's "gigantic Palaeotherium," the first titanothere discovered
After Prout (1847). One-fourth natural size.
of America — Joseph Leidy, Edward Drinker Cope,
Othniel Charles Marsh, John Bell Hatcher — up to
the time when the whole long and difficult study of
family history, of geologic succession, and of environ-
ment was intrusted to the present author.
From the first it seemed desirable that this study
should encompass more
than a dry, systematic
description — that these
animals and their envi-
ronment should, so far
as possible through pale-
ontology, be made to
live again as the domi-
nant animals of a long
and very interesting
epoch in the history of
North America — the
first third of the Terti-
ary period. The field
explorations made in
the prosecution of this
research should, more-
over, sustain the guiding
principles of the union
of paleontology and geology established by the pioneers
of our national surveys, as seen especially in the com-
bined work of the geologist, Frederick V. Hayden, and
the paleontologists, Charles A.White and Joseph Leidy,^
whose reports are still fundamental standards of Terti-
FiGURE 2. — Tj'pe of Palaeotherium ? proutii
Owen's specimen, Nat. Mus. 113. After Leidy (1852). One-third natural size. This was one of the specimens referred to by
Leidy (1852.1)1 in proposing the name Titanolherium.
This family, from its earliest known beginnings in the
Wind River Mountains of the present State of Wyoming
to the height of its development on the plains of the
ancient Dakota- Nebraska -Colorado region, attracted
the attention of the leading vertebrate paleontologists
1 The figures in parentheses refer to entries i
chapter.
, the bibliography at the end of this
ary geologic and paleontologic history. Subsequent
works have surpassed these in specialization and in
number and variety of animal forms, and the geologic
areas and life zones have been greatly increased by
subsequent discovery, but none have surpassed them
' See reports of Hayden and White (1867-73.1, 1868.1), based on surveys begin-
ning in 1802, and Leidy's great memoh (1869.1).
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
in scientific method — in the constant union of paleon-
tologic with geologic evidence in the reconstruction of
the slow succession of events in the wonderful history
of this western resion.
ROCKY
MOUNTAIN
LARAMIDE
SIERRA NEVADA
APPALACHIAN
7/ , , , , ■ ,,
/ r^y /// /// ,,.
/PENNSYLVANjAN
/ ' MISSfeSIPPlKN^ /
.''//•> / ^, / /'///
? PRE-CAMBRIAN
? ARCHEAN
FiGtTRE 3. — Geologic ages and orogenic
periods in North America
Age of mammals, stipple; age of reptiles, vertical lines;
age of amphibians and fishes, oblique lines. The peri-
ods of the birth and elevation of the chief American
mountain systems, notably the Rocky Mountains
(including the Laramide revolution), are indicated
approximately by incisions on the right. Modified
from diagram by Henry Shaler Williams.
The present monograph is made up of this introduc-
tory chapter and of ten other chapters, covering the
following six main lines of exploration and research
that have been followed in order to restore, at least
in part, the life and times of the titanotheres :
1. Geologic, physiographic, climatic, and faunal
environmental conditions of the titanothere epoch — •
the Eocene and lower Oligocene divisions of the
Tertiary. Principles of adaptive radiation in animals
as explaining the variation of the titanotheres.
(Chap. II.)
2. History of the discoveries of the remains of
titanotheres, the original published descriptions, and
the previous and present classification of genera and
species. (Chaps. Ill and IV.)
3. Systematic study of the titanotheres: Eocene
and lower Oligocene subfamilies, genera, and species.
Characters of the skull, dentition, and postcranial
skeleton. (Chaps. V, VI, and VII.)
4. Muscular anatomy of the titanotheres: Princi-
ples of locomotion and evolution of limb structure in
the hoofed mammals (Ungulata) in relation to habits.
(Chaps. VIII and IX.)
5. Origin, ancestry, and adaptive radiations of the
titanotheres and other odd-toed ungulates. (Chap. X.)
6. Evolution and extinction of the titanotheres:
Evidence regarding modes and causes of evolutionary
development and decline in mammals. (Chap. XI.)
SECTION 2. PRELIMINARY SURVEY OF THE MONO-
GRAPH AND OF THE CONCLUSIONS PRESENTED
RANGE OF THE TITANOTHERES IN GEOLOGIC TIME
Geographic distribution. — The earliest known titano-
theres lived near the end of early Eocene time, after
the appearance in the Rocky Mountain region of
three kinds of quadrupeds — the horses, the related
forest-living tapirs, and the more remotely related
rhinoceroses, which still exist elsewhere.
The successive immigrations of related odd-toed
ungulates are recorded in the Eocene deposits of the
region now included in the State of Wyoming, which
during Eocene time was a fertile land inhabited by
an abundant fauna. The Eocene titanothere epoch
in northern Utah, south of the great Uinta Mountain
range, which, according to Powell, rose to majestic
heights, ended in late Eocene time.
In lower Oligocene time the titanotheres had
seemingly become the largest mammals in North
America. They were second in size to the existing
elephants only, but recent paleontologic evidence
indicates the existence in Oligocene time in India of
mammals that exceeded in size both the titanotheres
and the elephants. In 1913 Mr. C. Forster-Cooper
(1913. 1) described a new genus of perissodactyls from
the upper Oligocene deposits of the Bugti Hills of
Baluchistan, BalucTiiiherium (Thaumastotherium) os-
borni, an animal of proportions so gigantic that it
dwarfs the largest known titanothere.
Sedimentary divisions and faunal life zones. — The
lower Eocene to lower Oligocene sediments in which
titanothere remains have been found occur here and
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
there in several of the ancient river drainage basins
of Wyoming. While the remains of the animals and
plants of the period were accumulating in these sedi-
ments the titanotheres and other herbivorous quadru-
peds and the carnivores that preyed upon them, as
well as the other mammals and invertebrates of the
land, of the water, and of the air, were constantly
evolving, appearing and disappearing through mi-
gration and extinction. Thus where the sediments
of Front's "gigantic PalaeotJierium" (Menodus gigan-
teus) in 1846 to the present time, it has been found
that the lower division of this zone is distinguished by
the presence of 85 species of vertebrates. The names
of the dominant form or forms of each zone are used
to designate the several life zones. For the designa-
tion of the Titanotherium zone the name of this single
genus Titanotherium (Menodus) is used, for it is the
most distinctive form in that zone.
swEETGRAss co.jNTRODUCTION OF ANCESTORS
(MONT.)
2 ^I'aunal Period
g::=a ARCHAIC MAMMALS
ONLY
1-1/ ^^Faunal Period
CorypTiodon - Ayriblypods
and.
Eohippjis -^vrstSorses
W$mwm MAMMALS
Pan.toZcmibda -^-mhlypods
Polynhastodan.
Figure 4. — Successive and overlapping Oligocene and early Eocene formations of the Rocky Mountains
The duration of the titanothere epoch is indicated by the arrow.
are very rich in fossils of all kinds — mammals, reptiles,
iishes, and rarely birds — we are able to restore the
life that was distinctive of certain more or less con-
tinuous phases of geologic sedimentation. These
time divisions are designated life zones, as distin-
guished from the sedimentary divisions of groups and
formations.
After an exploration of the Titanotherium zone that
covered a period of over 70 years, from the discovery
Many genera persist through several successive life
zones. Two genera, the large-hoofed Coryphodon and
the small primitive horse EoMppus, persist through
four lower Eocene geologic phases or life zones, during
which a succession of other species, as well as migra-
tions, extinctions, etc., may be clearly observed. It
may therefore be necessary to select more than one
genus, perhaps as many as three genera, in order to
define clearly a certain life zone. For example, the
TITANOTHEHES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
amblypod Coryphodon, the horse EoMppus, the tapir
Systemodon unite to define the Systemodon-Coryphodon-
EoMppus life zone of the lower Eocene.
It is through these zonal resemblances in the
mammalian life, and more rarely in the plant life,
that relatively sure estimates are made of the time
during which the sediments containing certain fossils
were deposited, irrespective of such geologic data as
whether the sediments are thick or thin, whether they
are products of erosion or of volcanic eruption, whether
they were deposited in still water or in rapidly moving
water, or whether they are composed of clay, sand,
gravel, conglomerate, or tuff. The life zone, when
adequately defined, is an absolutely reliable means of
time correlation as distinguished from other means —
physiographic, geologic, or lithologic.
Similar sediments. — It is true that in the Rocky
Mountain region there prevailed at times over wide
mentation in one region (for example, the Cypress
Hills, Saskatchewan) and with excessively slow sedi-
mentation on river flood plains in another region (Chey-
enne and White Rivers, S. Dak.), or with a fall of
volcanic ash in still another region (Beaver Divide,
Wyo.).
Evolution of mammals a stable process. — ^By com-
paring all the events in the history of the American
continent for which the records afforded by geology
and paleontology harmonize with others afforded by
paleontology alone we reach the conclusion that one of
the most uniform, the most stable geographically, and
the most widespread is the evolution of mammalian life.
This evolution proceeds more or less uniformly in
Europe, in Asia, and in North and South America. The
apparently sensitive protoplasm (body substance) and
germ plasm (hereditary substance) are far more stable
and far more uniform in their progressive evolution
Former land orea^ Former migration areas Known fossil areas
Figure 5. — Map showing areas throughout the world in which remains of titanotheres have been
found (solid black) and areas in which titanotheres were probably in migration during Eocene
and Oligocene time (oblique lines)
Titanotheres have been found in the northwestern United States, the Gobi Desert (Mongolia), Burma, and southeastern
Europe.
areas similar physiographic, climatic, and eruptive
volcanic conditions, as, for example, during what we
designate Fort Union time, Wasatch time, upper
Bridger time. During such periods of uniform con-
ditions the geologic evidence is concordant or harmoni-
ous with the paleontologic evidence afforded by life
zones, and doubtless any paleobotanic evidence that
may be found must also be concordant. In basal
Eocene (Fort Union) time, for instance, the forests,
the mammals, the reptiles, the climate, the physiogra-
phy of the chief areas of sedimentation of the whole
Rocky Mountain region were all more or less similar,
and in this particular epoch these several means of time
correlation afford more or less harmonious evidence.
Unlilce sediments. — Such similar sediments, however,
become increasingly rare in the continental deposits
of Eocene and Oligocene time. A single life zone,
such as the Titanotherium zone, may be contempo-
raneous with violent fluviatile action and heavy sedi-
than the surface of the earth. For this reason they
form superior data for time correlation. This is one
of the chief generalizations that have grown out of the
long series of observations and studies of the correla-
tion of Tertiary geologic events in America and Europe
that were specially made in the preparation of this
monograph.
Life zones of the titanothere epoch. — By the method
of determining geologic time by discriminating life
zones the whole epoch of the evolution of the titano-
theres has been subdivided into titanothere zones,
distinguished not only by successive genera and species
of titanotheres but by corresponding changes in all the
environmental forms of life. Each of these life zones
probably represents a very long period of time, for in
each there was a very considerable evolution of the
titanotheres as well as of other forms. These zones
(17-9; see table, p. 9), named in descending order,
are as follows:
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
17. Titanotherium-Mesohippus zone (Brontops robustus zone,
fauna; Chadron C fauna; Brontops dispar zone, Chadron B
and Brontops brachycephalus zone, Chadron A fauna) .
16. Theoretic zone (Uinta C 2).
15. Diplacodon-Protitanotherium-Epihippus zone (Uinta C 1
fauna) .
14. Eobasileus-DoKchorhinus zone (Uinta B 2 and Washakie B 2
faunas) .
1.3. Metarhinus zone (Uinta B 1 and Washakie B 1 faunas).
12. Uintatherium-Manteoceras-Mesatirhinus zone (Bridger C
and D and Washakie A faunas).
11. Paleosyops paludosus-Orohippus zone (Bridger B
fauna) .
10. Eornetarhinus-Trogosus-Palaeosyops fontinalis zone
(Bridger A and Huerfano B faunas).
9. Lambdotherium-E otitanops-C oryphodon zone
(Wind River B and Huerfano A faunas) .
Estimated duration of tJie titanothere epoch. —
The duration of the titanothere epoch, from
the time of the earhest known member of
the family {Larnbdotherium) to that of the
last product of titanothere evolution {Bron-
totherium) is estimated as 600,000 years.
This estimate is based on the calculation of
Walcott, made from measurements of the
rate of geologic sedimentation, that the
entire Tertiary period covered not more than
3,000,000 years. If estimates made by Bar-
rell (1917.1, p. 892), based on radioactivity,
can be verified the duration of Tertiary time
should be extended to 54,000,000 years. If
this estimate is accepted the duration of
the titanothere epoch alone would extend to
11,000,000 years. Though the geologic esti-
mate of 600,000 years for titanothere evolu-
tion seems to be too small, the physical esti-
mate seems to be too great, and for the
present we may regard the estimate based on
geologic data as ranging between 600,000 and
1,000,000 years.
HAYDEN'S SUBDIVISIONS OF THE EOCENE AND THE
OLIGOCENE
The geologic formations in which titano-
there remains occur and the life zones into
which these formations are subdivided have
been discovered and described during the last
56 years, the first report on them being that
of Meek and Hayden (1862.1), in which the
entire Tertiary geologic column is represented
in a "General section of the Tertiary rocks
of Nebraska," reproduced here in facsimile.
There is little doubt that when Hayden described
the White River group as "1,000 feet or more" in
thickness, as including the "Bad Lands of White
River; under the Loup River beds, on the Niobrara,
and across the country to the Platte," and as com-
posed of "white and light-drab clays, with some beds
sandstone, and local layers limestone," he had in mind
the area extending from Cheyenne River of South
Dakota to the region south of North Platte River,
displayed in the accompanying map and panoramic
section. This section includes at its base the Titano-
therium and Oreodon zones (Chadron and Brule for-
mations), from which Hayden listed certain char-
acteristic forms of animal life, such as TitanotJierium
{=Menodus), Choeropotamus {=Ancodus, Hyopota-
mus), "Rhinoceros" {=Caenopus), AncJdtherium
{= Mesohippus)' , Hyaenonodon {= Hyaenodon) , Ma-
chair odus { = Dinictis).
Gemral Section of the Tertiary rocks of Nebraska.
Names.
SUBDIVISIONS.
Thick-
LOCALITIES.
Foreign
Equiva-
lents.
.a
>
2
3
Fine loose sand, with some
layers of limestope, — contains
bones of Canis, Felis, Castor,
Eguiis, Mastodon, Tesiudo, &c.,
some of which are scarcely dis-
tinguishable from living spe-
cies. Also Helix, Pliysa succinea,
probably of recent species. All
fresh water and land types.
o
o
o
On Loup fork of
Platte River ; extend-
ing north to Niobrara
River, and south to
an unknown distance
beyond the Platte.
a
u
>
5
White and light drab clays,
with some beds sandstone, and
local laj-ers limestone. Fossils,
Oreodon, Titanotherium, Ch(?ro-
potamus, Rhinoceras, Anchitke-
rium,Bycenonodon,A/achairodus,
Trionyx, Testudo, Helix, Plan-
orbis, Limnma, Petrified wood,
&c. &c. All extinct. No
brackish water or marine re-
mains.
i
a
o
o
o
r-t
Bad Lands of White
River ; under the
Loup River beds, on
Niobrara, and across
the country to the
Platte.
o
13 P.
Light gray and ash colored
sandstones, with moro or less
argillaceous layers. Fossils, —
fragments of Trionyx, Testudo,
with large Helix, Vivipara,
Petrified wood, &c. No marine
or brackish water types.
O
O
O
IM .
Wind River valley.
Also west of Wind
River Mountains.
-
'3
'a
1
Beds of clay and sand, with
round ferruginous concretions,
and numerous beds, seams and
local deposits of Lignite ; great
numbers of dicotyledonous
leaves, stems, &c.of the genera
Platanus, Acer, Ulmus, Populus,
&e., with very large leaves of
true fan Palms. Also, Helix,
Melania, Vivipara, Corbicula,
Unio, Ostrea, Potamomya, and
scales Lepidotus, with bones of
Trionyx, Emys, Compsemys,
Orocodilus, he.
U
o
a
u
o
Occupies the whole
country around Fort
Union, — extending
north into the British
possessions, to un-
known distances ;
also southward to
Fort Clark. Seen un-
der the White River
Group on North Plat-
te River above Fort
Laramie. Also on
west side Wind River
Mountains.
1
Figure 6. — The Meek and Hayden Tertiary section of 1862
The deposits named are now known to include the following:
"Loup River beds." The lower Pleistocene fauna listed is found in an area that includes
deposits of the Pliocene and upper Miocene (Ogalalla formation of Darton) .
"White River group," including lower Miocene (Arikaree formation of Darton) and Oligo-
cene (Brule and Chadron formations of Darton). The " Choeropotamus" is Ancodus
ameTicanus, the ancodont of the Chadron formation (Titanotlierium zone).
"Wind River deposits" (summit of the lower Eocene).
"Fort Union or Great Lignite group" (basal Eocene).
These Titanotherium and Oreodon zones are now
regarded as lower and middle Oligocene, respectively,
and above them have been discovered the Protoceras
and Leptauchenia zones, which embrace the highest
sediments assigned to the Oligocene. The combined
thickness of the Oligocene at this point is 600 to 650
feet. Above it, to the east, are "light-drab clays,"
having a total thickness of 500 feet, and these, when
combined (1,150 feet), correspond to the "1,000 feet
6
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
or more" of Hayden's section. It therefore appears
that Hayden's description of the White River group
conforms with the accompanying panoramic section
of the Oligocene and lower Miocene exposed on the
south side of White River, South Dakota, shown in
of his White River group apparently came from
beds now classified as Oligocene. The name White
River group has therefore for years been restricted
to the beds of Oligocene age (Brule and Chadron
formations).
DISCOVERY OF THE TITANOTHEEES OF
THE PLAINS
At the base of this great section lies
the Titanotherium zone, or " Titano-
therium beds" of the Hayden-Leidy
memoirs, fully described in Chapter II,
composed in part of clays, in part of
river-channel sandstones, in which
titanothere remains are extraordi-
narily abundant.
The northern borders of this wonder-
ful region appear to have been first
explored around Bear Creek, a dry
tributary on the south side of Cheyenne
River, from which Thaddeus A. Cul-
bertson brought back the first collection
of fossils in 1850. From these expo-
sures of the Titanotherium. and Oreodon
life zones were obtained the greater
part of Leidy's types, which are de-c
scribed in Chapter III. The Brule and
Arikaree formations, which overlie the
Chadron, belong to a period succeed-
ing the titanothere epoch, with which
this monograph closes.
The physiography of this ancient
flood-plain region — its broad level
stretches, its meandering rivers, its
fringing river-border forests, its distant
mountains and active volcanic peaks —
as restored from our present knowledge
of its fauna and flora, is described in
Chapter II. It forms a wide contrast
to the mountain-basin region, in the
heart of which lie the Wind River de-
posits, described by Hayden in 1862.
DISCOVERY OF THE MOUNTAIN - BASIN
ENVIRONMENT OF THE TITANOTHEEES
As the entire lower Oligocene history
Figure 7 —Panoramic section of the Big Badlands of South Dakota, looking of the titanotheres is recorded chiefly
southeastward across Cheyenne and White Rivers to Porcupine Butte
This section of the ancient flood-plain sediments now expoi^ed cuts through five great life zones — the
Titanotherium, Oreodon, Leptauchenia, Promenjcochoerus, and Merycochoerus zones. It includes also four
ancient river-channel sandstones and conglomerates— the •'Titanotherium sandstones," " Metamynodon
sandstones," "Protoceras sandstones," and " Fromcrycochoerus sandstones" — each of which includes a
more or less distinct river-border and forest fauna. (See map, fig. 69, vicinity of section B.)
Figure 7, as sketched under the direction of Osborn
for the United States Geological Survey in 1909.
(Osborn and Matthew, 1909.321.) Hayden did not,
however, specifically define the upper limit of his
group, and all the fossils listed by him as characteristic
in the Great Plains region east of the
Front Range of the Rocky Mountains,
so their entne Eocene history is
recorded almost exclusively in the
mountain -basin region west of the
Front Range, in western Wyoming, northwestern
Colorado, and northeastern Utah. The interpreta-
tion of these remnants of the great Eocene sediments
(given in Chapter II) involves far more difficult
problems and has required more prolonged and in-
INTRODUCTION TO MAMM.tLIAN PALEONTOLOGY
D A K O T A
Chadron Formation
(Tit<xjvo i^herTZi^n/ZoThe)
Figure 8. — Map showing the type geologic locaUty (X) of the Titanotherium zone on Bear Creek, branch of
Cheyenne River, S. Dak.
The map shows the present exposures of the Cbadron formation in South Dakota, Nebraska, Montana, eastern Wyoming, and Colorado. These exposures
of the Titanotherium zone form the northern and western fringes of the overlying sediments, composing the Brule and Arikaree formations {the great
"White River group" of Hayden). Map after Darton, U. S. Geological Survey, 1905, modified from observations of Matthew and Thomson, 1906, 1907.
TITANOTHEEES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
tensive geologic researcli than tlie interpretation of
the Oligocene sediments. The program for this
exploration was proposed by the author to the Director
of the United States Geological Survey in 1900.
Each of the typical lower, middle, and upper Eocene
basins shown in the accompanying map has had its
Wind River Valley. Also west of Wind River Moun-
tains." It is possible that Hayden here refers to the
Wasatch or the Bridger formation, which lie southwest
of the Wind River Range.
Subsequent exploration by Hayden revealed the
typical Bridger, Wasatch, and "Washakie"^ forma-
tions, each affording portions of separate
chapters in the history of the ancient
mammalian life of the mountain-basin
region, which has proved to be no less
remarkable than that of the Great Plains.
Hayden was aided by the early paleon-
tologic observations of Leidy on the
Bridger fauna.
The survey along the fortieth parallel
by Clarence King was supplemented by
the paleontologic observations of Marsh,
who described the life areas south of
the Uinta Mountains and defined the
Diplacodon zone of the Uinta. Cope
hastened to describe the life of the Wind
River, of the Wasatch, and of the
"Washakie" formations and made known
a very rich fauna contemporaneous with
the Wasatch of the Big Horn Basin, to
the north, and of the San Juan Basin of
northern New Mexico, to the south, where
he also discovered the basal Eocene fauna
(Puerco). Five of these six geologic for-
mations were long regarded also as fau-
nistic units and were described as single
life zones, namely, the "Diplacodon beds"
(Uinta formation), the "Dinoceras beds"
(Bridger and "Washakie" formations),
the " LambdotJierium beds" (Wind River
formation), the " Coryphodon beds"
(Wasatch formation), and the Puerco
formation.
The intensive observation of these
six formations and the analysis of their
fauna has enabled us to divide them
into sLxteen known life zones, which in
turn afford the key to the time of origin
and of cessation of sedimentation in each
basin.
Figure 9. — Map showing cluster of typical lower, middle, and upper Eocene
sedimentary basins in the heart of the Rocky Mountain region
DISCOVERY AND DELIMITATION OF PERIODS
OF SEDIMENTATION AND OF LIFE ZONES
Mapped chiefly aftst the explorations of Hayden, King, and Powell of successive Government
surveys. The arrows indicate the lines along which were taken the chief geologic sections de-
scribed and illtistrated in Chapter II. Modified from Osbom and Matthew, 1909.321.
antecedent historic and its recent analytic treatment,
beginning with the Wind River deposits of Hayden
(Meek and Hayden, 1862.1, p. 433), who described
these deposits as "light-gray and ash-colored sand-
stones, with more or less argillaceous layers. Fossils —
fragments of Trionyx, Testudo, with large Helix,
Vivipara, petrified wood, etc. No marine or braclcish-
water types * * * 1,500 to 2,000 feet * * *
The fact that these sediments ac-
cumulated very slowly, during long
periods of geologic time and in the course of profound
changes in climatic and physiographic environment,
with consequent variations in the fauna and flora,
has gradually become recognized, and the explorations
and researches that have led to this recognition have
s The Washakie was contemporaneous with the upper two-thirds of the Bridger
formation, and the name Washakie is now abandoned by the United States Geolog-
ical Survey for the name Bridger.
INTEODUCTION TO MAMMALIAN PALEONTOLOGY
9
formed a considerable part of the work done for this
monograph. At first the periods of sedimentation
were regarded as broadly equivalent to a similar
number of life zones. For example, up to the year
1900 the two chief formations, the Wasatch and the
Bridger, were treated as containing one fauna each.
It was not known that the Wasatch represents five
distinct life zones, that the Bridger represents four
and perhaps five life zone^, and that the partly con-
temporaneous Washakie represents three distinct life
zones. The correlation of different areas of sedimen-
tation by means of fossils was similarly loose and in-
exact. The evidence discovered since 1900 by parties
sent out from the American Museum of Natural
History proves that there was considerable change of
environment as well as a great faunal change during
Bridger time. The careful recording of the precise
geologic level at which every specimen, especially
every type specimen, was collected, together with
close analysis of lithologic evidence that the rocks
afford as to modes of deposition, has worked a com-
plete revolution in our knowledge of the history of
these mountain basins in Eocene time and of the
flood plains in early Oligocene time and has afforded
the relatively precise and far more interesting sequence
of events that is described in Chapter II.
Our geologic studies show that from basal Eocene to
early Oligocene time there were six great physiographic
and climatic epochs of sedimentation, shown in the
accompanying table.
Epochs of sedimentation and life zones from iasal Eocene to early Oligocene time in hasins in the Rocky Mountain
region
Physiographic epochs
6. Lower Oligocene, represented by Chadron formation. Flood plains east of
the Rocky Mountains. Sedimentation extremely slow. Moderate rain-
fall. Warm temperate climate.
5. Latest upper Eocene, represented by Uinta formation (Uinta C). Flood-
plain basins south of the Uinta Mountains. Sedimentation relatively
rapid; fine material. Heavy rainfall, diminishing.
4. Upper Eocene, represented by contemporaneous deposits in Washakie and
Uinta Basins (horizons Washakie B and Uinta B) and probably by upper-
most part of Bridger formation, or Bridger E. Violent river and stream
action from the north and south sides of the Uinta Mountains. Erup-
tions of volcanic dust; coarse material. Heavy rainfall.
3. Middle Eocene, represented by Bridger formation (horizons Bridger A, B,
C, andD). More quiescent flood-plain conditions in the Bridger Basin;
eruptions of volcanic dust; intervals of evaporation. Sediments com-
posed in part of eroded material, generally laid down on lacustrine
deposits.
2. Lower Eocene, represented by Wasatch, Wind River, and Green River
formations. Warm temperate climate of the Green River lake period,
and evidently arid conditions in the contemporaneous Wind River sedi-
ments. Alternation of arid and fluviatile conditions characteristic of
Wind River and Wasatch time. Evidence of open country, favorable to
cursorial mammals.
1. Basal Eocene, represented by the Puerco, Torrejon, and Fort Union forma-
tions. Forests, base-leveled areas, flood plains, and swamps widespread.
Evidence of somewhat cooler climate.
17. Titanotherium-Mesohippus.
16. Theoretic zone. No fauna discovered.
1 5 . Diplacodon-Protitanotherium-Epihippus.
14. Eobasileus-Doliohorhinus.
13. Metarhinus.
12. Uintatherium-Manteoceras-Mesatirhinus.
11. Palaeosyops paludosus-Orohippus.
10. Eornetarhinus-Trogosus-Palaeosyops fonti-
nalis.
9. Lambdotherium-Eotitanops-Coryphodon.
8. Heptodon-Coryphodon-Eohippus.
7. Systemodon-Coryphodon-Eohippus.
6. Eohippus-Coryphodon.
5. Phenacodus-Nothodectes-Coryphodon.
4. Pantolambda.
3. Deltatherium.
2. Polymastodon.
1. Ectoconus.
The evidence of the existence of these successive
climatic, physiographic, and biologic conditions is
derived from studies by Berry of the flora; by Hay of
the reptiles; by Osborn, Scott, Wortman, Granger,
Matthew, Peterson, Douglass, and Riggs of the mam-
mals; and by Sinclair and Johannsen of the lithology.
These studies, the results of which are in part set forth
in Chapter II, show a great advance upon the pioneer
studies by Leidy, Marsh, and Cope, which were based
chiefly on characters of the skeleton and teeth.
101959^29— VOL 1 3
Our paleontologic division of the strata of the Eocene
and lower Oligocene epochs into sixteen known life zones
and one theoretic life zone enables us to fix the date of
the immigrations of animals into this region, as well as
the emigrations and extinctions, with much greater
precision than formerly. Remains of titanotheres
have been found in the upper eight of the known life
zones.
Extremely important is the realization that the zonal
fossil fauna reflects local conditions of sedimentation.
10
TITANOTHEKES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
which have a significant bearing on the kinds of animals
preserved. For example, violent fluviatile action
may preserve for us chiefly the river-border and aquatic
fauna; but remains of the animals of the surrounding
plains and of the distant forests may not have entered
the river-channel sandstones. Forest-living animals,
like the chalicotheres [Moroyus], are relatively rare;
and arboreal animals, like the lemurs (NotJiardus) , are
seldom preserved in channel sandstones. Certain
mammals apparently arriving as new immigrants, like
the giant uintatheres, which suddenly appear in
Bridger C, doubtless came from the surrounding plains
or mountain regions, where the conditions were un-
favorable for their entombment and fossilization.
The threefold division of the Wasatch and Bridger
mammals by Matthew (1909.1) and Loomis (1907.1)
according to their habitats, into meadow, forest, and
Figure 10.-
- Restorations of Eotilanops borealis (A) and Brontotherium plaly-
ceras (B) , drawn to the same scale
titanotheres has extended we invariably find more than
one of the branches of the titanotheres, as in Wind
River and early Bridger time, and in some areas as many
as five or six contemporaneous branches. Altogether
twenty branches of the great titanothere family tree
have thus far been discovered in Eocene and lower Oligo-
cene strata. This multiple branching, known as poly-
phyletic evolution, has made the study of the titano-
theres more difficult and at the same time more
fascinating than if these mammals presented only
a single line of descent, as in monophyletic evolution.
Some of the phyla of the titanotheres can be traced
through a long series of successive evolutionary stages,
such as Palaeosyops, Manteoceras, and Dolichorhinus
in the Eocene, Brontops, Menodus, and Brontotherium
in the Oligocene. Other phyla, such as the supposed
river-dwelling Eometarhinus and Metarhinus, appear
in two life zones only, in the middle Eocene,
Huerfano B, and the upper Eocene, Uinta
B 1, under fluviatile conditions of sedi-
mentation favorable to fossilization.
Extremes of evolution. — Members of these
twenty branches wandered in and out of
the regions favorable to fossilization, and
consequently no single branch (phylum) can
be traced over the whole period of time.
Even if this period covered 600,000 years
(minimum estimate), or 11,000,000 years
(maxunum estimate), the descent of a gi-
gantic horned quadruped, such as Bronto-
therium platyceras, from a small and de-
fenseless animal akin to Eotitanops borealis
would appear almost incredible were it not
that unremitting exploration during the last
half century has unearthed many phyla of
One of the earliest members of the titanothere family (£. borealis of the Wind River formation, lower
Eocene) and one of the latest and most formidable ( B. platyceras of the White River group, lower
Oligocene). Frommodelsinthe American Museum of Natural History made by Erwin S. Christman spCcicS that are mOre Or IcSS intermediate
imder the direction of the author and of William K. Gregory.
river living groups, not only has important bearing on
the gaps in the fossil record and on the interpretation
of the evidence relating to immigration and emigration
but is in accord with the principle of local adaptive
radiation developed by Osborn, as fully set forth at the
end of Chapter II.
PRINCIPIE OF LOCAL AND CONTINENTAL ADAPTIVE
RADIATION
The changes in the climatic and physiographic
conditions during the Eocene epoch, which favored not
only the evolution but the fossilization of this or that
type of animal, supply the key to the divergence in
anatomical structure and to the presence in the
diversified Eocky Mountain region and adjacent
plains of a great variety of titanotheres, in a measure
comparable to the great variety of ruminants found
to-day in the plain and plateau regions of the continent
of Africa.
Twenty branches of titanotheres. — In the eight life
zones through which the observed evolution of the
between these two extremes. Although the
whole period of life of the titanotheres was relatively brief
as compared with that of the surviving horses, tapirs,
and rhinoceroses, yet within this period the titano-
theres became much more specialized than the modern
tapirs; in fact, although in lower Eocene time they
resembled superficially the existing tapir {Tapirus
terrestris), by middle Eocene time they had reached and
passed the tapir-like stage of evolution. As compared
also with the contemporary horses they were more
rapidly progressive; the difference between the lower
Oligocene Brontotherium and the lower Eocene Eoti-
tanops is vastly greater than that between the lower
Oligocene horse Mesohippus and the lower Eocene
Eohippus. The titanotheres evolved rapidly, partly
because the environment was peculiarly favorable
to their rapid evolution; partly because their internal
germinal hereditary conditions favored their rapid
evolution and differentiation.
Competition oj the titanotheres with other ungulates. —
In the course of their evolution the titanotheres came
into competition as herbivorous quadrupeds with
members of four orders of hoofed mammals. They
INTEODUCTION TO MAMMALIAN PALEONTOLOGY
11
competed with members of two archaic orders, the
Amblypoda, typified by Coryphodon, and the Condy-
larthra, typified by Phenacodus. The titanotheres
survived both these archaic orders. They came
into competition with members of several other
families of the Perissodactyla and rapidly outstripped
them in evolution. The period of the extinction of
the titanotheres, at the end of lower Oligocene time,
marked also the decline of several other of the great
rhinoceroses are the only odd-toed ungulates that
outlived the titanotheres and survived to the present
time. The fourth order of quadrupeds that competed
with the titanotheres were the Artiodactyla, the dimin-
utive ancestors of the even-toed ungulates, including
the ruminants, which entered a great era of expansion
soon after the titanotheres became extinct.
The earliest known types of titanothere evolution,
Lambdotherium and Eotitanops, which were contem-
FiGUEE 11. — Amblypoda: Skeletons and restorations of an ancestral form (A) and a specialized form (B)
A, Pantolamida of the basal Eocene Torrejon formation; B, Coryphodon of the Wasatch formation, persisting throughout five life zones of
lower Eocene time, contemporaneous in its later stages of development with Eotitanops and Lambdothtriiim, ancestral titanotheres.
families of perissodactyls, especially the aquatic
rhinoceroses (amynodonts), the cursorial rhinoceroses
(hyracodonts), and the fleet lophiodonts (Colodon),
all of which became extinct soon after the titanotheres
disappeared. The aberrant perissodactyl chalico-
theres, which are in many respects similar to the titan-
otheres, survived, perhaps because they retreated,
like the okapi of the Congo region of Africa, into the
recesses of the forests. The tapirs, horses, and true
poraneous, appear in the fourth Coryphodon life zone.
Coryphodon is a clumsy but powerful mammal of very
archaic type, heavily armed with great canine tusks.
It is descended from Pantolamhda of the basal Eocene.
As Coryphodon appears in the far distant region of
the Sparnacian of France as the companion of a giant
bird (Gastornis) and of a primitive horse {Hyracothe-
rium) similar to the American Eohippus, France and
western America are brought close together in their
12
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBKASKA
mammalian life during lower Eocene time, so that we
shall probably discover a similar Coryphodon fauna
in the intermediate regions of eastern Europe, northern
Asia, and British Columbia.
COMPARISON OF THE FOUE LIFE PHASES IN EUROPE
AND IN NORTH AMERICA DURING EOCENE AND EARIY
OLIGOCENE TIME
Length of Eocene time. — It is the comparison of the
ancient life of the Old and the New World, especially
by means of the results of the successive studies of
Cope, Filhol, Deperet, Osborn, and Matthew, that
has led to the demonstration by Osborn of four great
continental faunal phases in Eocene and lower Oligo-
cene time — phases that probably extended over the
entire Northern Hemisphere and that were separated
by the rise and fall of the archaic forms of life, by the
union or separation of western Europe and western
America into one single or two distinct centers of mam-
malian life, and by the severance of all connection be-
tween North and South America. Together these
three series of events form a sequence that affords evi-
dence of the great length of Eocene time. In other
words, the biologic evidences of very marked evolution
in single families like the titanotheres, of the zoogeo-
graphic events of migration, and of the succession and
extinction of faunas together indicate that the Eocene
epoch alone may have been longer than the 600,000 to
1,000,000 years allotted to the titanothere epoch in
accordance with Walcott's estimates of Tertiary time
based upon purely geologic data.
The archaic succeeded hy the modernized mammals. —
The long duration of Eocene time is further indicated
by the subdivision of the Wasatch {Coryphodon) epoch
(the "Coryphodon beds" of Marsh and Cope) into
five lesser time divisions. Thus the term Coryphodon
alone no longer serves as the designation of a life zone,
because Coryphodon is now known to have survived
through at least five life zones, Nos. 5-9 in the
zonal series (p. 57), as follows:
"Coryphodon beds" oj Marsh and Cope
9. Lambdotherium-Eotitanops-Coryphodon zone o( Osborn
8. Heptodon-Coryphodon-Eohippus zone
7. Systemodon-Coryphodon-Eohippus zone__
6. Eohippus-Coryphodon zone
5. Phenacodus-Nothodedes-Coryphodon zone.
The modernized mammals in the series tabulated
above are the titanotheres, lophiodonts, tapirs, horses;
the archaic mammals are the condylarths (Phenacodus)
and amblypods (Coryphodon).
As remarked above, no single biologic phenomenon
affords stronger evidence of the long duration of
Eocene time than the complete replacement of the
archaic fauna of North America, which exclusively
held the stage during basal Eocene time, in itself a
very long epoch, by the ancestors of modern mam-
mals, as shown in the accompanying diagram (fig. 12)
and indicated precisely in the transition between the
Phenacodus and Eohippus zones. The modernized
mammals came in not suddenly or en masse, as we
formerly supposed, but gradually, family by family,
the first apparently being the swiftest and most vita-
tive family — the horses (Eohippus).
We infer that western Eiu-ope witnessed a similar
replacement, for, although sparsely loiown, the basal
Eocene life of western Europe was broadly similar to
that of western North America.
The archaic life of American basal Eocene time,
first made known by Cope, then studied by Osborn
and Earle, and finally given very full and precise
geologic and zoologic determinations by Matthew,
_ First appearance of the titanotheres in America. " Wind
River" fauna of Cope.
.First appearance of lophiodonts in America. "Lysite"
fauna of Granger.
.First appearance of tapirs in America. "Gray Bull" fauna
of Granger.
.First appearance of horses in America. "Sand Coulee"
fauna of Granger.
_ Phenacodus extremely abundant. "Clark Forli" and "Tif-
fany" fauna of Granger. The closing phase of the reign of
the archaic mammals of North America, Pantolambda, Cory-
phodon, Phenacodus.
Granger, and Gidley, affords the basis of our present
knowledge of the wonderfully rich and varied fauna
embraced within the four basal Eocene life zones.
The precision with which we are now able to note
the extinction or disappearance of the archaic mam-
mals and their replacement, one by one, by members
of modernized families is due especially to the ex-
plorations of the American Museum of Natural His-
tory, led by Granger with the assistance of Sinclair,
and to the analyses of the fauna by Matthew and
Granger in a series of researches which are classic not
only for their precision but for the revelation of new
and hitherto unsuspected affinities of the mammals
of North America with those of South America and
with the existing mammals of the oriental region of
the Old World.
Relation of the titanotheres to other quadrupeds. — In
their broadest relations the titanotheres were mam-
mals of the cohort Ungulata, which possess hoofs as
distinguished from claws. We know that eleven great
orders of ungulates (see accompanying table) were
distributed through different parts of the earth during
ancient and modern time. Of these eleven orders,
which were the sources of the herbivorous quadrupeds
of the world, only five have survived to the present
time.
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
13
TTie eleven orders of Tertiary ungulates
I. Archaic ungulates:
I America and Eurasia. Originating in Cretaceous time and contemporaneous
in Eocene time (Coryphodon) witli the titanotheres, becoming extinct in late
Eocene time {Uintatherium and Eobasileus).
I America, Eurasia, and possibly South America. For a short period contempo-
raneous with the titanotheres, becoming extinct in the lower Eocene (Phena-
codontidae) .
II. Modernized ungulates:
A. Primarily North American and north Eurasian:
..,,,, . , , , f America, Eurasia, and subsequently South America. First appearing in early
,' „ . , ^ , „ i.j_^_ _I,'""7 ~ I Eocene time. The Perissodactyla gradually gave way to the Artiodactj^la.
The chalicotheres were in part contemporaneous with the titanotheres near
the end of their Ufe period.
4. Perissodactyla (horses, titanotheres,
tapirs, rhinoceroses).
B. Originally African-Asiatic ungulates:
„ -H rv, ^ [First appearing on the African continent; subsequently, in part, entering
„■ „ , ., , , , _ ~j ----------- 1 southern Eurasia and North America. None of these orders is known to
Proboscidea (elephants and masto-
dons).
7. Sirenia (sirenians') .
have been contemporaneous (in Europe) with the titanotheres or to have
[ entered into competition with them.
[Aquatic mammals, first known in Africa, possibly related to the same ancestors
' I as the Proboscidea; believed to have sprung from ungulate ancestors.
8. Embrithopoda (arsinoitheres) Known solely on the African continent; Oligocene.
C. Distinctively South American ungulates:
Q p iv^ . / .i.- j_N [Exclusively South American in history and evolution.'' None of these orders
in T rl +• f f l'\ J entered into competition with the titanotheres. Part of them (Litopterna)
,,'-i X , \- J.N I imitated the other orders of ungulates, and part (Toxodontia) evolved into
11. Litopterna (extmct) ■ , & > f ^
[ unique forms.
* A single jaw attributed to one of the aberrant Southi American ungulates has been found in the Eohippus- Corypho
Basin, Wyo.
I lite zone, "Sand Coulee beds" of Clark Fork
Only three of the eleven ungulate orders shown
in the table were living in the Rocky Mountain
region when the titanotheres arrived — (1) the archaic
Amblypoda, represented, as we have seen, by Cory-
phodon, extremely smaU-brained, of very clumsy
build, heavy-footed, in general proportions somewhat
like the African rhinoceroses, RTiinoceros (Cerato-
therium) simus and R. (Opsiceros) iicornis; (2) the
Condylarthra, represented by a diminutive Phena-
codus, also extremely small-brained, contrasting with
Coryphodon in its small size and cursorial build,
formerly but no longer believed to be ancestral to the
higher ungulates; (3) the modernized Perissodactyla,
including the ancestors of the horses {Eohippus),
tapirs (Systemodon) , and lophiodonts (Heptodon).
The newly arriving perissodactyl titanotheres
equaled in size and resembled in their general cursorial
limb structure the condylarths as well as the horses,
tapirs, and lophiodonts. They were greatly surpassed
in size by members of the Coryphodon family, some
species of which were quadruple the size of the earliest
known titanotheres. However, certain of the titano-
theres of this stage (Eofitanops) exceeded the condy-
larths (Phenacodus) in size.
It is noteworthy that the archaic Condylarthra
(Phenacodus) were numerically preponderant in the
Phenacodus zone, just prior to the arrival of the
earliest perissodactyl horses. There was doubtless an
incessant competition between all these modernized,
alert, large-brained perissodactyl ungulates and the
archaic, small-brained ungulates {Coryphodon and
Phenacodus), which were especially inferior in the
mechanics of their foot structure. When, in the
upper Eocene, the clumsily built Amblypoda reached
the final phase of their evolution in the gigantic
Uintatherium and Eohasileus, they apparently became
suddenly extinct, and at the same time the titanotheres
suddenly began to develop into more formidable
animals. At no time in the Tertiary period was the
earth populated in the same region with more than
one type of very large quadruped. In the Northern
Hemisphere the dominance of the amblypods (in the
Eohasileus- Coryphodon epoch) was succeeded by the
dominance of the titanotheres (in the closing titano-
there epoch), and the titanotheres in turn, when
they had reached their largest development, suddenly
became extinct with no trace of a preliminary stage of
decline.
OID AND NEW SYSTEMS OF CLASSIFICATION
OLD TERMINOLOGY RETAINED
The studies for this monograph were begun by
Professor Marsh under the old ideas of classification
in mammalogy, derived from Linnaeus and his suc-
cessors. These studies were continued by Osborn on
the same old lines, as shown in his first paper on the
titanotheres. (Osborn, 1896.107.) The discovery
of adaptive radiation and of polyphyletic evolution,
which was one result of the researches made for this
monograph, has developed a new phyletic system of
classification. Yet even in this new system it is
necessary to adhere to the old Linnaean terminology,
for the reason that Linnaean methods have been used
during the long period of systematic description in
14
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
which the greater number of genera and species of
titanotheres have been described; and the Linnaean
generic and specific names can not be replaced unless
two systematic names have been given to the same
animal. Rather than introduce a new terminology
we attempt to place each Linnaean species in its proper
phyletic position — that is, in its true phylum — and to
connect it with other species by intermediate or transi-
tion stages, which are termed mutations, the "ascend-
ing mutations" of Waagen as distinguished from the
contemporaneous "mutations" of De Vries.
LINNAEAN METHODS OF DEFINING SPECIES, GENERA,
AND PHYLA OF TITANOTHERES
Between 1847 and 1902 as many as 29 genera and
67 species of Eocene and Oligocene titanotheres were
defined, but of all the definitions given hardly a single
Proceeding along these lines Marsh and Cope defined
a number of genera of titanotheres, certain of which
have since proved to be closely successive members of
the same phylum and consequently members of the same
genus. Osborn went to the opposite extreme in attempt-
ing to reduce all the titanotheres to a single genus. In
his paper of 1896, entitled "The cranial evolution of Ti-
tanotJierium" (Osborn, 1896.110),hereached the wholly
erroneous conclusion that there had been only a single
distinct and definable genus of titanotheres — the origiaal
Titanotherium of Leidy — and that all the variations
among the titanotheres were of the rank of species, rep-
resenting different stages of development. This has
proved to be a greater error than that of Marsh, because
it was based on the hypothesis that the titanotheres
belonged to a single — monophyletic — line of descent.
APPEARANCE AND EXTINCTION OF MAMMAL ORDERS IN NORTH AMERICA
Archaic J\fa7n7na2s - soled, hlacTo. Mode?-nLze<t .MajnTnaZs - outiine
Figure 12. — Diagram showing the gradual extinction of orders of archaic mammals (solid black) of earliest
Eocene time and their gradual replacement during later Eocene time by the ancestors of modernized orders
of mammals (outline), including related forms that are now extinct
one has proved to be distinctive and valid. The
main characters utilized in the old classifications by
the chief contributors to the history of the Oligocene
titanotheres — that is, by Leidy, Marsh, Cope, Scott,
and Osborn — were the following:
1. The presence or the absence and the number of incisor
teeth (Cope and Marsh, in generic definition).
2. The number of premolar teeth (Marsh, in generic definition).
3. The development of the cingulum on the premolar teeth
(Cope andMarsh, in generic definition).
4. The presence of a second cone on the last superior molar
(Marsh, in generic definition).
5. The length and shape of the nasal bones (Cope, Marsh,
Scott, and Osborn, in generic definition).
6. The length and shape of the fronto-nasal horns (Cope
and Marsh, in generic definition).
7. The presence or absence of a trapezium in the carpus
(Hatcher, in phyletic definition).
RECOGNITION OF MANY LINES OF DESCENT) POLYPHYLY
THE KEY TO INTERPRETATION OF THE FAMILY
In January, 1901, a few months after the studies
for this monograph were begun, all the data, observa-
tions, skull sections, and measurements were assembled,
and by July of the same year it was demonstrated by
Osborn that at least four lines of separate descent
are to be found among the lower Oligocene titanotheres,
and this number has since been increased to five or
eight.
In 1902 Osborn established the fact that throughout
lower Oligocene time, when the Titanotherium-he&vmg
beds were being deposited, as many as eight more or
less different phyla, or series, were independently
evolving in the same region. Certain of these phyla
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
15
embrace one or more of the genera originally proposed
by Pomel, Leidy, Cope, and Marsh. Other phyla
correspond with certain genera — for example, Menodus
Pomel (syn. Titanotherium Leidy), Brontops Marsh,
Allops Marsh, Megacerops Leidy, BrontotheriumMarsh.
These five generic names correspond to members of
five phyla that persisted throughout a very long period
of geologic time. The remaining phyla of titanotheres
are branches that persisted only for a short time, so
far as we know at present — for example, Diploclonus.
As shown in the accompanying diagram (fig. 15)
these generic phyla are branches of the family tree of
has adopted in expressing the relationships and
descent of the rhinoceroses, animals whose evolution
presents in many respects analogies to the evolution
of the titanotheres, especially in the modes of the
evolution of horns, in the loss or retention of cutting
teeth (incisors), and in the adaptations of limb struc-
ture to swift and slow movement.
RELATION OF THE PHYLOGENETIC CLASSIFICATION TO
THE LINNAEAN CLASSIFICATION
Linnaeus described one or more species of mammals
geographically distributed in space (see table on p. 16),
FiGUEE 13. — Phenacodus (A) and Coryphodon (B) drawn to the same scale
Restorations made by Charles R. Knight under the author's direction.
the titanotheres. When two of these branches run
close together they may for convenience be united into
a single subfamily. Thus, for purposes of description
the graphic presentation of the titanothere family tree
in the accompanying diagrams may be supplemented
by the systematic subdivision of these animals into
12 subfamilies and 24 genera, as shown on a subse-
quent page.
The free use of subfamily divisions to distinguish
the branches of Eocene and Oligocene generic phyla
from one another is similar to that which the author
whereas the phylogenetic classification of the titano-
theres covers species extending over both space and
time. The geographic range of the existing red deer
(Cervus) and of the extinct titanotheres lies within
the same zoogeographic region — Holarctica, which
includes Europe, Asia, and North America. The
comparison is therefore significant. (See table on
p. 18.)
The classification presented in this monograph is
more than phylogenetic: it is polyphyletic. Lin-
naeus (1758.1), when he wrote the several editions
16
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
of his "Systema naturae" (1735-1768), did not
dream of the succession of species of mammals in
time; he did not know of a single phylum, much less
of polyphyla. Darwin's theory of descent and
divergence implied the existence of phyla, but when
he published "The origin of species" (1859.2) he also
did not ls;now of a single phylum or a single direct line
of descent. Waagen (1869.1) was the first to dis-
tively rapid gain or loss of certain characters. This
definition relates to the hard parts, which are pre-
served in fossUization; the principle applies equally
to characters of all kinds.
In contemporaneous Linnaean genera and species
we observe differences of many kinds, such as differ-
ences in color and proportion, and, more rarely, we
may note the presence or absence of simple characters
A Contemporojieous LiruzcbeoJi Syste/n Zoology
X
c
' Su2).\Fam..l
JFAM I LY
_A ^^
Sub. Fam,. 2
)T
"^
. Spec.
Spec./
\ I -^^f^^^ /'^Si^.Fa.nv.4^
.' , I ' Gerv. Gerv)
;■"■•• J \ ^ ^ ^
Spec./ \ y V -^
\
RECENT EXISTING PERIOD ^ --^ '
B Geologic /. Phylogenetlc SystejTh Ihleontolo^y
/ ^^ ^ /T?
/ — ~~-^'^^, \ \\
/
LOWER/
0LI60CENE ,^ /
/A
\
/! \
1 ^
\
UPPER^ ,
EOCENE \
\ >^j
\
MIDDLE EOCEP _
\
\
\
^ jA V-.^\\ V:. V" ^ \\ \\ i i ,, I \\ I y'l
EARLY EOCENE
\
/
Figure 14. — Contrast between the Linnaean and phylogenetic systems of classification of sub-
families, genera, and species
cover a continuous phylum (namely, of ammonites) —
that is, successive hereditary stages, which he named
"mutations." Many direct phyla of invertebrate
animals have since been made known.
In this monograph we first learn the full meaning of
a mammalian phylum — namely, a phylum is a con-
tinuous geologic line of descent diverging from other
phyla (1) in the gradual transformation of every
character in size and proportion and (2) in the rela-
of teeth, vertebrae, or claws. The "species" of Lin-
naeus are now known to be actually superspecies and
to include one or more modern species, subspecies, and
geographic races and varieties, distinguished by differ-
ences in coloring, habit, proportion, or otherwise.
These differences are due in part to environment and
in part to habit. They represent the different bodily
effects produced on animals of similar ancestral stock
under different environments, in which somatic changes
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
17
are rapid and conspicuous. They are in part hereditary
(germinal) diflferences, which pass down for generations
unmodified by habit or environment.
For example, the American genus Peromyscus (the
white-footed mouse), as studied by Osgood (1909.1),
(dolichocephalic). Peromyscus may have been widely
distributed from some common center during the last
40,000 years, and during this long period there may
have been both geographic or space evolution and
geologic or time evolution, the evolution in time being
Figure 15.-
-The family tree of the titanotheres, showing the relation between the branches (phyla), sub-
families, and genera, as known to science in 1919
The shaded areas show connections that f
3 well established; the dotted lines show gaps that remain to be filled by future discovery, especially
in the Uinta formation of Utah.
presents a continuous series of transition changes in
color and form in species having a geographic range
from Tehuantepec to Alaska. In the northern re-
gions Peromyscus is larger and has relatively longer
teeth and a skull that may be somewhat elongate
comparable to that which we observe in the geologic
phyla of the titanotheres.
The existing genus Cervus affords another example,
significant because its geographic range is similar to
that of certain Oligocene titanotheres.
18
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Species and subspecies of the genus Cervus
[Table prepared by Qeirit S. Miller, 1918]
Name
Habitat
Nature of habitat
Climate
Cervus canadensis (American elk =
New York and New Jersey southward to
Open plains, badlands,
Humid to extremely
wapiti) .
the Carolinas; central western States;
sand hills; forests and
arid.
Nebraska, the Dakotas, and the coun-
meadows.
try farther west, across the Rockies.
C. c. merriami (Merriam's elk) .-
New Mexico and Arizona _. _ .
Mountains and plateaus;
Generally arid; for-
forests and meadows.
ests wet.
C. nannodes (dwarf elk) _ .
San Joaquin Valley, Calif., and adjoining
Plains and tule swamps
Generally arid.
foothills.
C. occidentalis (Olympic elk)
Washington, Oregon, California; formerly
Chiefly forested regions;
Humid.
south to San Francisco Bay. Van-
some meadows.
couver Island?
C. xanthopygus (Bedford's deer;
Manchuria and adjoining parts of Siberia
Forests .. ...
Do.
Manchurian stag).
C. sibiricus (Altai maral)
Baikal, Saiansk, and Altai Mountains;
Formerly forests and open
Extremely humid to
southern Siberia and northern Mon-
timberless country;
extremely arid.
golia.
even open high desert
mountains. Now re-
stricted to forests and
meadows.
C. songaricus (Tien-Shan stag)
Tien-Shan Mountains ^ ._
Mostly arid(?)
C. yarkandensis (Yarkand stag)
Eastern Turkestan _ _ _
Do.
C. macneilli (Kansu stag)
Kansu and Szechwan border of Tibet
Tibet.. - _ ... .-...-
do
C. wardi (Ward's stag)
do
C. hanglu (Kashmir deer; hangul,
Vale of Kashmir and adjacent mountains.
Chiefly forest; some open
Humid.
hanglu) .
parks.
C. bactrianus
Russian Turkestan . .
Chiefly arid.
C. maral (maral) .
Persia, Crimea, Caucasus
C. e. atlanticus (Norwegian red
West coast of Norwav _
Do.
deer) .
C. e. germanica (red deer).
Middle Europe
Do.
C. e. bolivari (red deer of central
Central Spain..
Spain).
C. e. hispanicus (red deer of south-
Southern Spain.
Humid and semiarid.
ern Spain) . ,
C. corsicanus (Corsican stag)
Corsica and Sardinia. _ .
Semiarid.
C. barbarus (Barbary deer) _ _ _ _ _
Morocco, Algiers, palearctic north Africa
Chieflv arid.
COMPARISON BETWEEN ZOOLOGIC AND PALEONTOLOGIC
SPECIES
The difference between zoologic and paleontologic
species is represented in tlie accompanying diagram
(fig. 16), showing the descent and relationship of cer-
tain members of the dog family (Canidae). A theo-
retic stem or central form is shown from which geo-
graphic races have been given off horizontally, as it
were, and the ascending mutations and species of the
evolutionary line of development from the ancestral
form have arisen geologically.
It follows that in making an anatomic comparison
between the existing geographic species and sub-
species of such genera as Peromyscus or Cervus and a
geologic phylum of species such as that of Menodus or
Brontotherium the same comparative anatomical
methods of measurement and observation should be
employed. Direct measurements of the length and
breadth of the skull should be recorded, by which
indices (proportions of single structures like the skull)
and ratios (proportions between different parts like
the upper and lower segments of the limbs) should be
established.
The proportional changes technically known as
dolichocephaly (elongation of the head), brachy-
cephaly (broadening of the head), dolichopy (elonga-
tion of the face), brachyopy (abbreviation of the
face), dolichopody (elongation of the feet), brachy-
pody (abbreviation of the feet), dolichomely (elonga-
tion of the limbs), brachymely (abbreviation of the
limbs) occur in geographic species and subspecies in
their corresponding stages exactly as they occur in
geologic phyletic time series. The chief difference is
that in the geologic time phyla these differences of
INTEODXJCTION TO MAMMALIAN PALEONTOLOGY
19
proportion may be followed through long periods of
time from their incipient to their final stages, in
which various climaxes of change of proportion are
reached, such as extreme length or breadth of head or
extreme length or shortening of the feet.
PROPORTIONS OF THE SKULL IN BEARS AND IN
TITANOTHERES
In comparing the Eocene and Oligocene titanotheres
with the modern bears {Ursus), for example, as
studied by C. Hart Merriam (1918.1), we may note
certain parallelisms. The members of each of the
eleven subfamilies of titanotheres are distinguished by
certain proportions of the skull — that is, they are
broad-headed, round-headed, or long-headed — by
the shape of the horns and the acceleration or retarda-
tion in their development, by the presence or
absence of cutting (incisor) teeth, by certain
proportions of limb, according as they are
swift-footed (cursorial), slow-footed (medi-
portal), or heavy-footed (graviportal), and
by other minor features. The methods ap-
plied to the study of the existing bears may
be applied to the study of the skull or other
hard parts of the titanotheres. In the titano-
theres, however, we may observe all these
changes of proportion actually in progress
from stage to stage as revealed by paleontol-
ogy, whereas in the bears we can observe only
certain structural forms, which, so far as our
observation goes, appear to be fixed or com-
pleted, although they undoubtedly represent
stages in a state of actual progression.
B. Bridger and succeeding titanotheres — Continued.
6. Manteoceratinae; mesatioephalic to brachycephalic;
accelerated development of the horns; mediportal
{Manteoceras, Prolitanotherium) .
5. Diplacodontinae; dolichocephalic; accelerated molar-
ization of the premolars; imperfectly known
(Diplacodon) .
4. Telmatheriinae; mesaticephalic to dolichocephalic
{Telmatherium, Sthenodecles) .
3. Palaeosyopinae; brachycephalic; short-limbed {Palae-
osyops, Limnohyops) .
A. Wind River titanotheres ; face longer than cranium :
2. Eotitanopinae; medium-limbed, mediportal (Eoti-
1. Lambdotheriinae; light-limbed, cursorial {Lambdo-
thenum) .
The above scheme presents the eleven subfamilies
of titanotheres as they were distinguished in 1914.
GEOGRAPHIC DISTRIBUTION
AT PRESENT TIME
True R
ZcoUfiical WlPES
-GEOGRAPHIC DISTRIBUTION
IN PAST TIME
FEATURES
DISTINGUISHING
TITANOTHERES
PHYLA OF
The first application of changes of propor-
tion to the arrangement of the subfamilies of
titanotheres is the following synopsis, pre-
pared in 1914:
Proportions of skull and limbs; presence and absence of characters
distinguishing the subfamilies {main phyla) of titanotheres
lOsborn, 1914.409]
B. Bridger and succeeding titanotheres; cranium longer than
face:
11. Brontotheriinae; mesaticephalic to brachycephalic;
horns long, transversely flattened, and divergent
(Brontolherium) .
10. Megaceropinae; mesaticephalic to extreme brachy-
cephalic; horns long, vertically placed; no incisor
teeth (Megacerops (—Symborodon)).
9. Brontopinae; brachycephalic; horns short, rounded,
or oval; incisors persistent (Brontops {=Mega-
ceratops), Diploclonus) .
8. Menodontinae; mesaticephalic to dolichocephaUc;
short triangular horns; incisor teeth reduced or
wanting; feet and limbs long {Menodus {=Titano-
therium), Allops).
7. Dolichorhininae; mesaticephaUc to doUchocephalic;
limbs, so far as known, short (Dolichorhinus,
Mesatirhinus, Sphenocoelus, Metarhinus, Rhadi-
norhinus) .
Figure 16. — Theoretic descent of existing members of the dog family
(Canidae) from a common ancestor
A represents the ancestral type. Dots represent intergradations indicated by paleontologic
observations (vertical lines) covering five periods of geologic time. A', B, B', C, and C rep-
resent existing forms, and dots represent a few existing intergradations demonstrated by zoo-
logic observations (horizontal lines). Heavy lines and the adjacent dots represent the phyla;
also the past and present distribution of geographic (ontogenetic and environmental) sub-
species, races, and intergrades.
Since that time certain phyla have been condensed by
the discovery of titanotheres that link together some
of these subfamilies, and others have been expanded
by the discovery of new subfamilies, such as the
Rhadinorhininae.
MUTATIONS OF WAAGEN
Where the fossil material is abundant the genera
and species are found to be connected by a series of
intergradations. These intergradations, though con-
tinuous, are measurable, and therefore a species is
subdivisible into a series of intergrading forms. The
monophyletic, systematic, or taxonomic unit division
of these species is the mutation of Waagen, which is a
subspecific stage in the development of one or more
characters. Such an actual sequence of mutations of
Waagen may be illustrated in the genus BrontotJierium,
as indicated on the following page.
20
TITANOTHEEES OP ANCIENT WYOMING, DAKOTA, AND NEBEASKA
Oligocene stages of titanotheres of the Brontotherium 'phylum in the Titanotherium zone
Division of zone
Stage
Species
Tiieoretic ascending
mutations
Brontotherium platyceras
Brontotherium ramosum
Brontotherium ourtum
Species.
Subspecies.
Mutation.
Do.
Species.
Subspecies.
Mutation.
Do.
Species.
Subspecies.
Mutation.
Do.
Species.
Subspecies.
Mutation.
Do.
Upper.
Do
Subspecies.
Mutation.
Do.
Species.
Subspecies.
Mutation.
Do.
Middle.
Subspecies.
Mutation.
Do.
Brontotherium hypoceras
Lower.
Subspecies.
Mutation.
Do.
Subspecies.
Mutation.
Do.
" Genus Titanops IVIarsli.
ZOOLOGIC AND PALEONTOLOGIC NOMENCLATURE
Significance of the table. — The sequence shown in the
accompanying table, which presents what is believed
to be a generic, monophyletic, or nearly single phyletic
series of changes of form, evolving in a single geographic
region of South Dakota, illustrates the manner in
which the Linnaean binomial system and the muta-
tion substages of Waagen may be adapted to express a
phyletic sequence. The newer trinomial names of
modern mammalogy and the subspecific names may
be employed to connect the intergrading mutations.
The most primitive species, Brontotherium leidyi, is
so notably distinct in size and skull structure from the
most advanced species, Brontotherium platyceras, that,
if named by zoological standards, it might well be
' Type of genus Brontotherium (Marsli).
placed in a separate genus — in fact, several generic
names have been suggested for members of this
phylum, namely, Brontotherium, Titanops, Bronto-
theridion (MS.) — but the subdivision of such a phylum
into a number of genera would obscure the all-im-
portant monophyletic unity, for such a phyletic genus
is defined by its peculiar and distinct evolutionary
tendencies. For example, the genus Brontotherium
tends toward the evolution of flattened horns, a charac-
teristic which begins in a very slight flattening of the
posterior side of the horn, as observed in B. leidyi, and
develops into the extraordinarily broad, flattened
horns of B. platyceras.
New phyletic meaning of species. — The species repre-
sented by large collections of mammals like those of
some of the phyla of the titanotheres, especially the
INTBODUCTION TO MAMMALIAN PALEONTOLOGY
21
Brontops phylum, are so closely intergraded and con-
nected by "ascending mutations" that the dividing
lines between them can be drawn only arbitrarily,
according to individual judgment. In the Brontops
phylum, for example, the species Brontops hrachyce-
pJialus grades imperceptibly into the species Brontops
dispar through gradual transitions in a great number
of characters, as may be seen in the Hatcher collection
in the United States National Museum. There is no
evidence of brusque transitions, saltations, or jumps
in any structure, such as are presupposed in the
mutation theory of De Vries. By contrast, the
mutations of Waagen are intergradations between
arbitrarily defined species, and through these muta-
tions species and genera pass imperceptibly one into
another.
Evolutionary characters of each phylum. — Thus we
reach a clear conception of a phylum of the titanotheres
in its osteologic and dental characters. A phylum may
be further defined as a succession of interbreeding
(syngamic, Poulton) individuals of similar (synepi-
gamic, Poulton) ancestry, which may or may not
occupy a similar range of country (synpatric, Poul-
ton), which follow in every structural character a sim-
ilar line of evolution (synphyletic, Osborn) and adap-
tation (syntelic, Osborn).
In each horn, in each tooth, in every bone of the
skull and skeleton, and by inference in all the hard
parts as well as in all the soft parts, each phylum has
its distinctive mode and rate of transformation in each
character, as follows: (1) Distinctive hereditary pro-
portion; (2) distinctive tendencies to change of propor-
tion; (3) distinctive progressive changes of proportion;
(4) distinctive retrogressive changes of proportion; (5)
distinctive accelerations and retardations in ontogeny
(individual development) ; (6) distinctive rates (veloc-
ities) of progression and retrogression in phylogeny in
each character. In each phylum are consequently
developed distinctive and ever changing proportions
and ratios between different single characters and
groups of characters, measurable by indices and ratios.
Such indices express the degrees of broad-headed, long-
headed, broad-footed, short-footed structure and pro-
portion, and so on. Each phylum has also its distinc-
tive but constantly changing indices and ratios of
teeth to skull, of skull to body, of body to limbs, etc.,
which also are constantly changing as we pass from
the lower to the higher geologic levels.
Old and new meanings of taxonomic terms. — In the
following table a comparison is made between the old
and the new meanings of the taxonomic terms used
by mammalogists. The definitions given in the sec-
ond column are those of the old "special creation"
system — followed by Linnaeus — which is based on
geographic distribution alone; the definitions given in
the third column are those of the new phyletic sys-
tem— that of Osborn — which is based on both geologic
and geographic distribution. The new system was
first used for the rhinoceroses (Osborn, 1900.192) and
for the titanotheres (Osborn, 1902.208).
Comparison oj ike Linnaean and the phyletic systems of taxonomic terms
Term
Definitions
Old system
New system
Family
A contemporaneous group of similar subfamilies
A smaller contemporaneous group of similar genera__
A still smaller contemporaneous group of similar or
related species.
A group of related subspecies and geographic va-
rieties.
Nothing corresponding to the geologic mutation of
Waagen.
Contemporaneous and ancestral phyla that exhibit
similar family tendencies of evolution.
A branch composed of one or more phyla which exhibit
similar generic tendencies of evolution.
Part of a single phylum of successive species and muta-
tions exhibiting similar tendencies.
A series of ascending mutations.
Geologic mutation (of Waagen); ascending substages
within a specific phylum.
Subfamily
Genus __
Species
Mutation
Desired harmony of mammalian paleontology and
zoology. — The methods employed by all zoologists,
paleontologists, and anthropologists in their observa-
tion and measurement of the hard parts of mammals
should be the same. The methods pointed out above,
first presented by Osborn (1914.412), are founded
on the comparison in time of geologic ascending
evolutionary phyla of mammals — such as the rhi-
noceroses and the titanotheres — with contemporaneous
geographic series of species, subspecies, and varieties
that may be grouped within a single genus. What
applies to the systematic terms used in the classifica-
tion and description of animals applies with equal
force to those used for single characters, for it is
the cumulative sum of evolutionary change in a very
large number of single characters which makes up
the mutation of Waagen, the species, or the genus, as
the case may be.
22
TITANOTHEEKS OF ANCIENT AVYOMING, DAKOTA, AND NEBEASKA
SUMMARY OF DIFFERENCES BETWEEN OLD AND NEW
SYSTEMS
To sum up: (1) The Linnaean genus or species is
defined (statically) by the presence of certain propor-
tions and by the presence or the absence of certain
characters, whereas the phyletic genus or species is
defined (dynamically) by the progressive evolution of
certain proportions and by the gradual gain or loss of
certain characters; (2) the Linnaean genus or species
was clearly distinguished from a related genus or
species, whereas the phyletic genus or species may
gradually fade into its ancestor or successor, and the
point where we make the dividing line is largely arbi-
trary; (3) consequently the phyletic genus actually
has a new meaning, but to avoid innovation in nomen-
clature we apply the phyletic term genus to a number
of species having a wide range in time and space, in
the same manner that Linnaeus applied the term
genus to a number of species having a wide range in
space only.
STUDY OF THE EVOLUTION OF SINGLE CHARACTERS
In the hard parts of living as of extinct animals
only three kinds of changes are observable — (1)
changes of proportion, which the author terms "allo-
metrons"; (2) the appearance of absolutely new char-
acters, which the author terms "rectigradations"; (3)
the disappearance or retrogression of characters.
Changes of proportion. — Changes of proportion
(allometrons) make up by far the larger part of the
evolution of the titanotheres, as of that of all other
mammals. At least 95 per cent of the differences
between the skeletons of Eotitanops horealis and
Brontotherium plafyceras are due to changes of pro-
portion, and not more than 5 per cent to additions of
absolutely new characters, such as horns. Conse-
quently a very careful study has been made of allo-
metry — that is, of the methods of calculating, measur-
ing, recording, and describing changes of proportion —
and the result has been the discovery of a number of
general principles that apply to all mammals, extinct
and living, including man. Probably also the un-
discovered causes of changes in proportions are the
same in all mammals, but their discovery constitutes
a very difficult problem. (See Chap. XL) In this
difficult work the paleontologists may be greatly aided
by the zoologists, especially by very precise field
observers, such as Allen, Merriam, Miller, Osgood, and
Sumner.
Although the mammalogists have demonstrated that
there is an apparently causal relation, direct or in-
direct, between certain types of coloration and of size
(harmonic increase or decrease) and the geographic
environment, the relation between change of environ-
ment and changes in proportion (disharmonic) is very
obscure. It is known that a harmonic increase or
decrease in size of the entire mammal is correlated
with certain differences in habitat, often for the
obvious reason that a favorable environment favors
development of larger races, whereas an unfavorable
environment dwarfs growth. It remains to be
determined, however, whether certain environments
induce uniformly similar disharmonic changes of pro-
portion. Anthropologists, for example, have failed
to establish a definite causal relation between environ-
ment and the broad-headed (brachycephalic) or the
long-headed (dolichocephalic) form of the human head.
The chief contribution that the paleontologist has
made to this obscure matter is to show that when a
proportionate change of head form is once established
in a certain direction there is a tendency to go to
extremes, so that, for example, extremely long heads
or extremely broad heads tend to evolve longer or
broader heads. These evolutionary tendencies are
illustrated in the titanotheres.
Adaptive new characters. — The second mode of mam-
malian evolution — by the appearance of absolutely
new characters — lies in a field where the paleontologist
has a great advantage over the zoologist, because in
a series of fossils a new character (rectigradation) can
be traced back to its incipient, rudimentary stage, in
which it is so inconspicuous that it would not attract
the attention of the zoologist. Many characters that
eventually may exert a most profound influence on
the evolution of a race — that may, in fact, dominate
aU other characters — arise, so far as observed, from
excessively minute beginnings. These origins of new
characters are pointed out with great precision in
Chapters V and VI, in which the evolution of the
skuU and teeth is described in detail as observed in the
Eocene and lower Oligocene titanotheres. This very
precise study of the origin and evolution of similar
characters in many different lines of descent has led
to the important discovery that phyla differ less
through the possession of this or that new character
than through the different rates of evolution at which
the same character arises and evolves. In one
phylum a new character like the horns will arise in
an early geologic stage and evolve very rapidly,
whereas in a related phylum it will arise relatively late
in geologic time and will evolve very slowly. Thus a
phyletic genus is defined not only by the characters
which it exhibits but by the rate of the evolution of
these characters. This principle, again, is observable
only thi-ough paleontology.
The origin of new characters, as manifested in dif-
ferent ways in the members of twelve subfamilies of the
titanotheres and as indicated by comparison with the
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
23
origin of similar characters in other families of Peris-
sodactyla, has accordingly been studied with great
care.
Retrogressive characters. — The retrogression or disap-
pearance of characters is illustrated in the history
of the titanotheres by the features enumerated below.
1. Reduction of the canine teeth in many later titanotheres.
2. Reduction and occasional loss of incisors.
3. Reduction and frequent loss of first lower premolar.
appearance in North America and western Europe of
members of nine different families of Perissodactyla,
the odd-toed ungulates, which were probably all de-
scended from a common ancestral or stem form which
lived in Upper Cretaceous time. The probable charac-
ters of this stem form are fully described in Chapter X,
where it is shown that the ancestral perissodactyl was
a comparatively small and simple quadruped not ex-
Perissodxictyls
CRETACEOUS
Peris sodaxtyL Stem,
Figure 17.-
-Successive invasion of nine perissodactyl families in Nortli America and western Europe between
latitudes 40° and 50°
Th2 chalicotheres (aberrant clawed perissodactyls with affinities to the titanotheres) are regarded as members of a separate superfamily, the
Chalicotheroidea. Diagonal shading indicates the extent to which each phylum is represented by fossil remains.
4. Reduction and loss of protoconule and metaconule in
upper molars.
5. Reduction of nasals and their coalescence with frontals.
6. Reduction of the trapezium in later titanotheres.
PHYLOGENY OF THE NINE TYPICAL FAMILIES OF THE
PERISSODACTYLA
The competition of the titanotheres through natural
selection was naturally closest with other members of
the order Perissodactyla. As shown in the ordinal
phylogenetic tree (fig. 17), we observe the successive
ceeding half a meter in height, and that it was origi-
nally confined to a definite geographic area, feeding
ground, and range, very possibly in northern Asia.
The eight families that appear in North America and
the paleotheres, which appear only in western Europe,
were by no means equally distinct from one another.
They were originally separated from the stem form
not into nine branches but into five great main
branches, termed superfamilies, as shown in Figure 17
and in the accompanying table.
24
TITANOTHEEES OF ANCIENT WTfOMING, DAKOTA, AND NEBRASKA
Phyla of the odd-toed ungulates
Superfamilies
Families
1. Titanotheroidea
1. Brontotheriidae : The titanotheres, known chiefly in North America and in
eastern Europe.
2. Chalicotheroidea
2. Chalicotheriidae: The chalicotheres, first known in Europe and North America;
then in Asia.
3. Hippoidea: Horselike forms _ ._
3. Palaeotheriidae : The paleotheres, known in western Europe only.
4. Equidae: The horses, first known in Europe; then simultaneously in North
America and Europe; subsequently in Asia, Africa, and South America.
4. Tapiroidea: Tapir-like forms
5. Tapiridae: The tapirs, first known in North America; then in Europe and Asia.
5. Rhinocerotoidea: Rhinoceros-like forms _-
6. Lophiodontidae: The lophiodonts, known in North America and Europe.
7. Amynodontidae: The amynodonts, aquatic rhinoceroses; first known in North
America; then in Europe.
8. Hyracodontidae : The hyracodonts, cursorial rhinoceroses; upper Eocene and
OUgocene of North America only, so far as known.
9. Rhinocerotidae : The rhinoceroses, the typical rhinoceroses; first known in
North America and Europe; then in Asia and Africa.
In North America the horses (Eohippus) were the
first perissodactyls to arrive. They were followed by
the tapirs (Systemodon) , which in turn were succeeded
by the lophiodonts {Heptodon). It is possible that
ancestral titanotheres were living iu northern parts
of the American continent, but apparently thej^ did
not reach the region near the fortieth parallel until
it had become well populated with horses, tapirs, and
lophiodonts. By middle Eocene time three more
families had appeared — the paleotheres, in Europe
only; the rhinoceros-like amynodonts (semiaquatic
forms), which first appear in North America and
subsequently in Europe; and the cursorial rhinocer-
oses known as hyracodonts (Hyrachyus), which appear
in North America only and preceded the amynodonts.
Toward the beginning of upper Eocene time there
first appear in North America, as well as in Europe,
ancestors (Eomoropus) of the chalicotheres, animals
closely related in tooth structure to the titanotheres,
which were separated into a distinct order (Ancylo-
poda) by Cope and are here regarded as forms some-
what parallel to the Titanotheroidea.
WIDE GEOGRAPHIC DISTRIBUTION OF THE
PERISSODACTYLA
We are first struck with the remarkably wide
holarctic distribution of the perissodactyls in Eocene
and lower Oligocene time, a fact which points to
facihty of migration over the whole Northern Hemis-
phere. Only one family, the paleotheres, is exclu-
sively European, and one other, the hyracodonts, is,
so far as known, exclusively North American. The
titanotheres were formerly beheved to be exclusively
North American, but two forms have been found in
eastern Europe, which correspond very closely with
the titanotheres of upper Eocene age from the Uinta
Basin in northern Utah.
Members of all the other perissodactyl families —
the chalicotheres, tapirs, lophiodonts, amynodonts,
and rhinoceroses — probably ranged freely to and fro
over the great northern continent of Em-asia and
North America combined, the geographic region
known as Holarctica.
The second important fact regarding the Peris-
sodactyla is that, although the environment dining
middle and upper Eocene time, after the extinction
of the archaic imgulates — the Condylarthra and
Amblypoda — was especially favorable to the existence
of the Perissodactyla, this order reached its maxi-
mum expansion in the lower Ohgocene epoch, when
all the nine families were existing and apparently
flourishing at the same time. It would appear that
in upper Eocene and lower Oligocene time Holarctica
was dominated by perissodactyls. This period was
immediately followed by a period when either the
environment was adverse to the existence of the peris-
sodactyls or competition with other types of imgu-
lates was disastrous to them, because at or before the
end of the lower Oligocene epoch five perissodactyl
families suddenly disappeared — the titanotheres, paleo-
theres, lophiodonts, amynodonts, and hyracodonts.
The aberrant chalicotheres, apparently through retreat
to forested regions, survived in Europe and probably
also in North America until the Pliocene epoch.
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
25
.^"^-
Figure 18. — Outlines of the body form of the perissodactyls, drawn to the same scale
The largest known member o( each family is selected for comparison. The smallest known stem forms of each family are illustrated
in Chapter X. The animals are grouped according to their natural relationships, as indicated especially by the pattern of the
molar teeth, as follows:
Bhinocerotoid group: A, Mclamynodon; family Amynodontidae; graviportal; aquatic; lower Oligocene. B, Hyracodon, family
Hyracodontidae; cursorial; middle Oligocene. C, Ceratotherium simum; living white rhinoceros; family Rhinocerotidae;
graviportal.
Tapiroid group: D, Tapirus terresiris; existing tapir; family Tapiridae; mediportal.
Hippoid group: E, Palaeotherium; family Palaeotheriidae; lower Eocene; mediportal. F, Equus pTzewalskii; existing horse; family
Equidae; cursorial.
Chalicotheroid group: (?, MoTopus; family Chalicotheriidae; clawed perissodactyl; lower Miocene.
Titanotheroid group: B, Brontotherium platyceras; family Brontotheriidae; graviportal; lower Oligocene.
101959— 29— VOL 1 4
TITAXOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
27
Thus out of the nine original famiUes of the great
order of Perissodactyla only three — the horses, tapirs,
and rhinoceroses — have survived to the present time,
and these during the glacial epoch were greatly
reduced both in numbers and in geographic dis-
tribution.
The consideration of these facts raises the whole
problem of the origin and adaptive radiation of the
perissodactyls (see Chap. X) and the general problem
of the causes of the extinction of the perissodactyls
and of other quadrupeds (see Chap. XI).
adaptive origin of new characters. The moment of
origin of each new character is a very important
moment in the history of that character. Does each
new character arise fortuitously at this point or that, in
an adaptive or inadaptive condition, or does each new
character arise in a mechanically adaptive condition,
although this condition may be merely incipient?
The biologic purpose of the long and dry descrip-
tions and tables of measurements given in Chapters
V, VI, and VII of this monograph is to direct obser-
vation continuously to this problem of the origin of
Figure 20. — Periods of expansion and extinction of the perissodactyls and contemporary forms
Showing that the expansion of the perissodactyls was coincident with the extinction of the archaic Condylarthra and Amblypoda and that the
extinction of many perissodactyls was coincident with the expansion and adaptive radiation of the artiodactyls.
CAUSES OF EVOLUTION
There can be no doubt as to the survival value of
certain finished types of tooth structure and Hmb
structure (see pp. 880-881), a principle first formulated
by the distinguished Russian paleontologist Kova-
levsky (1873.1). Two important questions that the
reader must keep in mind in considering the origin
of innumerable new characters are (1) whether
there is evidence of chance origins and chance rudi-
ments of certain types of structure possessing suffi-
cient survival value to establish themselves through
the principle of the survival of the fittest, or natural
selection; or (2) whether there is some other ortho-
genetic principle at work causing the definite and
new characters. Our general conclusions concerning
these two questions are presented in Chapter XI.
ADAPTIVE EVOLUTION AND OVEEEVOLUTION OF THE
FORM OF SKULL, TOOTH, AND FOOT
Whatever may be the causes of evolution its re-
sults are definite. The visible evolution of all the
hard parts of the body in herbivorous animals is
originally mechanical and manifests general adapta-
tion to two broad groups of purposes:
1. Prehension of food (lips, teeth, and jaws); com-
minution of food (teeth and jaws); conservation and
transportation of stored food energy (body and limbs).
These purposes involve all the mechanical changes of
structure of skull and tooth.
28
TITANOTHEEES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
2. Motion and locomotion; migration in search of
food and to escape enemies; adaptation to perform
the act of reproduction and to protect the young.
These purposes involve all the mechanical changes of
the structure of limb and body.
The operation of the principle that, under the domi-
nance of these modes of mechanical adaptation each
organ, structure, and character is adaptively evolved
for some special service to the organism is not invariably
evident in respect to all changes in the proportion of
characters. Certain characters of proportion, such as
extreme broad-headedness or extreme long-headedness,
seem to interfere with adaptation; they appear to be
carried so far in one direction as to render the animal
less adapted to survive than its less specialized ances-
tral forms. In other words, certain tendencies of
evolution may carry a phylum beyond its require-
ments in adaptation.
Aside from this question of the different degrees of
survival or actual elimination value of certain tend-
encies of evolution, there can be little doubt that in
its origin and development each character, sooner or
later, responds and reacts independently to the con-
ditions of the environment, quite apart from the
question as to the causes of such response. The teeth
react to the kinds of food; the feet and limbs to the
kinds of soil.
The principles of the divergence of quadrupeds
from each other in their independent adaptations in
the skull, teeth, limbs, and feet are fully discussed
elsewhere (see p. 123) in the treatment of the principle
of adaptive radiation. Though they may have lived
apparently in the same region and have been fossilized
side by side in the same sediments, all distinct species
of quadrupeds have locally different habits and habi-
tats. The structure of the skull, jaws, and teeth re-
sponds to their habits and tastes ; the structure of the
feet and limbs responds to their habitats — the nature
of the ground, etc.
PHYLETIC DIVERGENCE IN THE EVOIUTION OF NEW
PROPORTIONS IN HORSES AND IN TITANOTHERES
All the families of an order of Perissodactyla start
their career from a similarly proportioned ancestral
stem form such as that described in Chapter X (p. 760)
as the stem perissodactyl. Starting with the same
complement of characters, divergence in proportions
separates the families of perissodactyls more and more
widely from one another. In the Equidae (horses),
for example, the head form of the earliest known
ancestor (EoMppus) is very similar to that of the
earliest known ancestor (Eotitanops) of the family
Brontotheridae. In both these primitive skulls the
orbit is near the center of the head, and in the later
forms it apparently moves backward or forward, but
what really happens is that the skull is elongated in
front of the orbit in the horse and is elongated behind
the orbit in the titanothere. (See fig. 21.)
A comparison of the forms shown in Figure 21
with those shown in the following figures will demon-
strate the marked similarity of the lower Eocene
forms and the very wide divergence of the modern
forms. The skulls of the ancestral tapir, horse, and
titanothere {Systemodon, Eohippus, and Eotitanops)
are in many ways much alike, the chief differences
consisting in (1) the details of the characteristics of
the dentition, (2) the relative position of the orbits,
(3) the depth of the head through the back part
of the lower jaw, and (4) the size of the muzzle.
The primitive titanothere prophetically suggests the
titanothere characters in the relatively heavy muzzle
and stout lower jaw. The primitive horse Eohippus
prophetically suggests the modern horse in the taper-
ing form of the slender lower jaw and in the general
contour of the skull, except that the eye is placed near
the middle of the head, as in other primitive perisso-
dactyls. The primitive perissodactyl Systemodon,
regarded by Osborn as an ancestral tapiroid, had a
somewhat longer, more pointed muzzle but was
otherwise very similar to the contemporary horse
Eohippus.
These differences of proportion between the facial
region in front of the orbit and the cranial region
behind the orbit are partly correlated in adaptation
to the elongation (hypsodonty) of the crowns of the
grinding teeth. In the horse and in most of the rumi-
nant artiodactyls the face is elongated to accommodate
the vertically elongated (hypsodont) grinding teeth.
In the titanotheres, which are browsing animals,
and in the browsing rhinoceroses of India and of
Africa the orbit is directly above the grinding teeth
and the cranium is slightly elongated, as shown in
Figure 22. Thus it may be stated as a general prin-
ciple of skull evolution that in browsing ungulates
the cranium tends to be elongated and the face tends
to be abbreviated, whereas in grazing ungulates,
like the white rhinoceros of Africa, in which the grind-
ing teeth are elongated, the face is elongated, and
the cranium is abbreviated.
It follows that these respective proportions of the
region in front and back of the eyes are adaptive;
they are part of the general correlation of skull
proportions with the functions of the grinding teeth
employed in the prehension of food, as provided for
chiefly in the shape of the upper and lower lips,
which are obtrusible and flexible both in the browsing
rhinoceroses and in the grazing horse, which occasion-
ally browses. When the horse is browsing it extends
its lips very much in the manner of the browsing
rhinoceros, except that in the rhinoceros the independ-
ent motion and the pointing of the upper lip are more
extreme. In the grazing white rhinoceros the upper
lip is extremely broad and square. The animal
subsists largely on grasses, which it crops with its
square lips, exactly in the manner that the horse
INTEODUCTION^ TO MAMMALIAN PALEONTOLOGY
29
crops grass with its lips and front teeth. In all
the rhinoceroses cropping front teeth are atrophied,
the four pairs of incisors and the canines being
reduced to a single large pair on either side and being
thus analogous to those of certain titanotheres.
From these comparisons we deduce the structure
of the mouth parts in the titanotheres as restored by
Gregory. (See p. 704.) We also deduce the various
adaptations to the browsing and grazing habit re-
spectively in the different genera of titanotheres, for
undoubtedly some were purely browsers and others
of the face, with a relatively short skull, and with a
very powerful neck, a feature that is also especially
characteristic of the titanotheres.
Thus there is a general resemblance between the
side profile of Brontotherium platyceras and that of the
Indian rhinoceros, which is due to analogous mechan-
ical evolution, through the principles known as homo-
plasy, parallelism, or convergence. The titanotheres
pass through a long lower and middle Eocene phase of
tapir-like analogies, but when, in middle Eocene time,
horns begin to appear the head region develops
Figure 21. — Phyletio divergence in the evolution of new proportions in horses and in titanotheres
Lower Eocene ancestral horse Eohippus (A) and lower Eocene ancestral titanothere Eotilaijops (C) (both with the orbit in the same relative
position on the skull) compared with a modern horse (B) with face extended in front of the orbit and a titanothere of the latest stage (D) with
slvuU extended behind the orbit. Thus two very similar heads (A, C) become increasingly dissimilar (B, D). Scales various.
tended toward grazing. Thus the orbits, the face,
the grinding teeth, the front teeth, the lips, and the
bones supporting these structures are respectively
transformed in adaptation to the function of prehen-
sion and to browsing or grazing habits. The front
part of the skull of the rhinoceros, with its terminal
dermal horn, is comparable to that of the large-horned
titanotheres, with their terminal bony horns. It will
be observed that the entire front part of the head of
the rhinoceros, in adaptation to the great strain of
the horn used as a weapon of offense and defense, is
correlated with a flat or a concave line along the top
rhinoceros-like analogies. Similar analogous phases
also occur to a greater or less extent in the feet of the
rhinoceros and the titanothere.
On comparing the heads of the types of perisso-
dactyls, ancient and modern, we observe that different
modes of feeding and of offense and defense guide the
dominant adaptations in evolution. The evolution
operates under the principles of anatomical correla-
tion and compensation, gain or loss in one part being
mechanically balanced by gains and losses in every
other part. This process includes the principle of
physiologic compensation, whereby loss of function in
30
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 22. — Contours of the head and of parts of the mouth in browsing and grazing perissodactyls
A, Asiatic rhinoceros {Rhinoceros ' inikusi, chiefly a browser; B, black rhinoceros ot Africa (R. (.Opsiceros) bicornis), chiefly a browser; C, white
rhinoceros of Africa (R. (Ccraioiherium) simum), chiefly a grazer; D, domestic horse {Equns caballus), chiefly a grazer; E, American
tapir ( Tapirus tenestTis), a browser.
' The generic terminology of the rhinoceroses is not yet fully agreed upon by zoologists. The family tree, like that of the titanotheres,
is polyphyletic.
INTBODUCTION TO MAMMALIAN" PALEONTOLOGY
31
Figure 23. — Heads of lower Eocene and modern jjerissodactyls, showing changes of proportion and of the lip
structure
Based on materials in the American Museum of Natural History. Scales various. A, Head of the lower Eocene tapiroid Sijstemodon, very sim-
ilar to that of Eohippus and of Lambdotherium; B, head of middle Eocene tapir Hdaleies, in which a prehensile upper lip first appears;
C, head of the modern tapir Tapirus, whose prehensile upper lip forms a short proboscis; D, head of middle Eocene cursorial rhinoceros
Hyrachyus, still of primitive proportions; E, head of existing white rhinoceros {EhiTicccrcs ( Caalothcriuw) simum) with extremely
broad, grazing type of lip structure.
32
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
one part is taken up by some other part. For exam-
ple, the loss of the function of the incisors in the pre-
hension of food is compensated for by changes in the
form and function of the lips.
upper lip like that of the tapir necessitates space for
the superior retractor muscles, which curl the lip
upward and backward. An example of the results of
the evoUition of the lower jaw may be seen by compar-
FiGtTRE 24. — Restorations to the same scale of the heads of some of the principal t3'pes of titanotheres
Drawn by Charles E. Knight, after models made by him under the author's direction. About one-seyenteenth natural size. A,
Brontops roiustus Marsh, oblique yiew, middle Titanotherium zone; B, Menodus giganleus, upper Titanotherium zone; C,Megacerops
copei Cope, partly oblique side yiew, summit of the Titanotherium zone of Colorado; D, Broniotherium platyceras Scott and Osborn,
the final stage in the eyolution of the horns of the titanotheres, summit of the Titanotherium zone of South Dakota; E, Protitano-
iherium sp., summit of the Eocene.
With the evolution of the lips the structure of the
anterior parts of both the upper and lower jaws, of
the anterior teeth, and the anterior nasal openings is
closely correlated. The development of a prehensile
ing Eotitanops gregoryi and Brontotherium {medium)
gigas, the whole jaw of the former hardly exceeding in
length a single posterior grinding tooth of the latter.
(See fig. 25.)
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
33
EVOIUTION OF THE LIMBS AND FEET OF THE
TITANOTHERES
The feet of the titanotheres, like their skulls, pass
through a lower Eocene tapir-like phase, which is
followed by a middle and upper Eocene rhinoceros-
like phase and finally they attain a structure similar
to that of the rhinoceroses, as shown in Figure 26,
except that all the titanotheres, like the existing
tapirs, retained four distinct and functional digits
in the fore foot.
The fore foot of the tapir resembles the fore foot of
the lower Eocene titanothere except that in the latter
D. II, III, IV, V were all of nearly equal size, as
shown in the diagram (B). This is known as the
mediportal stage, for it is adapted to carrying a
moderate amount of weight. The
foot of the rhinoceros (C, C, C)
is like that of the upper Eocene
and lower Oligocene titanotheres
except that in these there were
four weight-bearing digits instead
of three. This is known as the
graviportal type of foot, in which
a large cushion pad is developed
at the back to relieve the shock
of impact, and the end phalanges
of the digits are incased in the
horny sheath in front. In the
tapir and rhinoceros the main
weight passes directly through the
center of the median phalanx
(D. Ill), but in the tetradactyl
titanotheres the main weight
passes between D. Ill and D. IV.
The concentration of the weight
on the central digit of the horse
and its resultant monodactylism,
correlated with the expansion of the horny hoof and
the contraction of the pad, is part of the evolution
of a cursorial type of foot, which presents the widest
contrast to the graviportal type.
In addition to comparing the head structure it was
found necessary to compare the foot and limb struc-
ture of the titanotheres with that of all the other
perissodactyls — not only the bony parts but the
musculature. The work done on the musculature led
to an exhaustive study of all that is known of the
muscular anatomy of the members of the three exist-
ing families of perissodactyls. This study, which was
directed by William K. Gregory, formed the basis of
the restoration of the muscular anatomy of the giant
Brontops rolustus presented in Chapter VIII (pp. 722,
723). This restoration of an extinct animal is the first
that has been based upon exact comparative study. It
presents the titanothere as a superb example of the
graviportal type of musculature and skeleton, sur-
passed only by the existing elephants.
The study of the structure of the foot led to a special
investigation of the proportions of the limb bones in
the ungulates. This investigation, directed by Osborn
and cooperated in by Gregory, resulted in the striking
discovery that the proportions of the upper and lower
segments of the limbs and of the feet are invariably
adjusted, first, to the weight that the limb must carry,
and second, to speed of locomotion. These propor-
tions are evolved, quite irrespective of ancestry, in
adaptation to different modes of progression. Thus
similar proportions of limb segments are observed not
only in all mammals but in reptiles as well. A study,
'^^-^Ga^;^^' I
Figure 25. — Lower jaws of the first and the last of the titanotheres
One-sixth natural size. A, EotUanops gregoryi, a small-jawed species from the Wind River formation (lower
Eocene); B, BrontotTierium medium, from Chadron C level of Chadron formation (lower Oligocene).
therefore, which was designed to disclose the habits of
the titanotheres led to a thorough investigation of the
principles of limb evolution in all the hoofed mam-
mals in adaptation to various modes of locomotion
and to various loads. This special study forms the
subject of Chapter IX, in which acknowledgment is
made to previous investigators.
Not only the proportions of the upper and lower
segments of the limbs but all the bones of the shoulder
and pelvic girdles are gradually transformed from the
subcursorial stages of Lambdoiherium and Eotitanops
through the mediportal tapir-like stages to the gravi-
portal stages of the ponderous Oligocene titanotheres.
This transformation is continuous, not sudden; it is
brought about gradually by the simultaneous and
correlated modification of all the bones and muscles
involved in locomotion. Function (habit) is evi-
34
TITANOTHEKES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
dently far more potent than ancestry (heredity) in
the determination of general form, yet in comparing
the limbs of all the members of the different perisso-
dactyl families with one another we can generally, by
some family characteristic inherited from the ancestral
stem form, distinguish the tapir type, the rhinoceros
type, the titanothere type, etc. In the limbs, as in
the skull and teeth, the titanothere, rhinoceros, or
tapu' ancestry respectively seems to keep the evolution
of proportion and form within certain limits, so that,
for example, the resemblance between the graviportal
scapula of the titanothere and that of the rhinoceros,
though it may be very close and deceptive, is never
quite complete. The stages of muscular and skeletal
the origin of new characters (rectigradations). In
this problem of the origin of new characters in the
titanotheres we have two principal subjects of study,
namely, the origin of horns on the skull and the
origin of cusps on the grinding teeth.
In the evolution of the grinding teeth the titano-
theres are very conservative; in them few new cusp
elements originate, though several of the old cusp
elements disappear. These animals thus present a
striking contrast to the horses in the evolution of the
grinding teeth, for in the horses a large number of
new cusp elements are successively added. Yet the
grinding tooth of the earliest titanotheres {Lambdo-
therium and Eotitanops) is in general similar to that
Figure 26. — Structure of the feet in extinct and living odd-toed ungulates (perissodactj-ls)
A, Sole of the left fore foot of a tapir (Taphus ierresiris), showing the tripod-like arrangement of digits II, III, and IV, and
the reduced condition of V; B, sole of the left fore foot of an Eocene titanothere (Mesatirkinus petersoni), restoration based
on Princeton Museum specimen No. 10013; C, sole of the fore foot of a rhinoceros, showing the enlarged hoofs of the
three digits (II, III, IV) ; C^, side view of same; C^ longitudinal section of same; D', sole of the fore foot of a horse, show-
ing the expanded nail; D^, longitudinal section of same. The central pad (/} in A, B, and C is homologous with the
relatively reduced pad or frog (/) in the foot of the horse (DO- All but B after Eber.
evolution, arranged from latest to earliest, are as
follows :
4. Graviportal; ponderous, relatively slow-moving types,
such as Brontolherium, Rhinoceros {C eratotherium) simum.
.3. Mediportal; of moderate weight and speed, such as
Limnohyops, Tapirus.
2. Subcursorial; of light weight and relatively swift move-
ments, such as Eolilanops of the lower Eocene.
1. Cursorial; swift moving, Ught frame, such as Lanibdo-
therium of the lower Eocene.
ORIGIN OF NEW CHARACTERS AS DISTINGUISHED FROM
CHANGES IN PROPORTION
The continuous gradual changes of proportion in
the head, trunk, and limbs (allometrons), as already
outlined, present a problem distinct from that of
of the earliest horses {Eohippus). In these lower
Eocene contemporary mammals the grinding teeth
are the same, cusp for cusp. In the horse all these
cusp elements are preserved and utilized, and the
highest degree of mechanical adaptation to the graz-
ing habit is gradually evolved; in the titanotheres
the browsing habit is generally conserved, and there
is little marked increase of mechanical adaptation;
in fact, mechanical inadaptation or imperfection of
the grinders may have been one of the probable
causes of the extinction of the titanotheres at a time
when the conditions favorable to grazing gradually
replaced those favorable to browsing.
The adaptive radiation of the grinding teeth in the
several families of the Perissodactyla from somewhat
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
35
similar ancestral forms is shown in Figure 29. The
earliest members of every family had low-crowned
(brachyodont) molar teeth, of relatively simple
pattern, composed of six principal cusps ranged in
three pairs — an external pair, the paracone and meta-
cone; an intermediate pair, the protoconule and
metaconule; and an internal pair, the protocone
and hypocone.
In the titanotheres, chalicotheres, paleotheres,
and horses the internal pair of cusps assume the
conical, rounded shape (bunoid), whereas the
two external cusps assume the double crescentic
shape (selenoid), together forming a W, hence
this type of tooth is termed bunoselenodont.
These bunoselenodonts apparently formed origi-
nally a natural group from which the horses
(Eohippus), the titanotheres (Eotitanops), and
the chalicotheres (Eomoropus) gradually diverged
very early in Eocene time. This is shown in
Figure 30.
Another group of perissodactyls is the bunolo-
phodonts, which includes the tapirs and lophio-
donts, in which the internal and external pairs
of cusps alike assume an elongate, crested, or
lophoid pattern. This group has two main
branches, the tapirs and the lophiodonts. The
tapirs as forest-seeking animals escaped fossiliza-
tion and are rarely found; only isolated remains
of them have been found in Europe and America;
yet they constituted one of the most persistent
of all the perissodactyl phyla. The lophiodonts
were tapir-like animals, in which the posterior
outer molar cusps were flattened and thus are
intermediate in shape between the tapir tooth
and the rhinoceros tooth. These animals doubt-
less had a wide expansion in the luxuriant
forests of Eocene France, and they attained
very great size just before their extinction,
which occurred contemporaneously with the
extinction of the titanotheres in America — that
is, in lower Oligocene time. Only one branch of
the lophiodonts, the swift-footed Helaletinae,
reached North America in lower Eocene time,
soon after the arrival of the tapirs (Systemodon)
and the horses (EoTiippus).
The grinding tooth of the rhinoceroses is lopho-
dont — that is, all the cusps are turned into elon-
gate crests, of lophoid type, and the posterior
outer cusps of the upper grinding teeth are
elongated as well as flattened, producing an asym-
metry of the cusps of the outer wall (ectoloph) of the
crown. A grinding tooth of this kind is far more
effective than that of the bunoselenodont titano-
theres or of the bunolophodont tapirs. Such a tooth
is a very efficient cutting instrument for an animal
of either the browsing or the grazing habit. It is
also capable of elongation (hypsodonty), and in
two subfamilies of the rhinoceroses, the white rhinoc-
eroses and the elasmotheres, the grinding teeth
become hypsodont, greatly increasing the longevity
and consequent reproductive power of each indi-
vidual.
Figure 27. — Restorations of nine species of titanotheres from the
lower, middle, and upper Eocene and the lower Oligocene
Drawn by Mrs. E. M. Fulda. About one-fiftietb natural size.
The rhinoceroses gave off at least twelve distinct
branches (phyla) and were thus more plastic in adapta-
tion than the titanotheres. These branches became
adapted to every habitat, aquatic as well as terrestrial,
to every mode of locomotion — cursorial, mediportal,
and graviportal — and to every kind of feeding — brows-
ing and grazing. Like the titanotheres some of the
rhinoceroses passed from the mediportal to the gravi-
36
TITANOTHBEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
portal stage of locomotion. In doing so they acquired
an entirely new set of proportions, which are shown
in detail in Chapter IX.
The teeth form the readiest means of distinguishing
different branches and subbranches of the Perisso-
dactyla from one another. The ancestral pattern,
whether bunoselenodont or lophodont, is so marked
and persistent that it is only partly modified through
their evolution, and these give off one mediportal,
forest-living branch, HypoMppus. The horses are
paralleled by cursorial or subcursorial titanotheres,
such as LamhdotJierium, by cursorial paleotheres
{Palaeotherium and Paloplotherium) , mistakenly sup-
posed by Huxley to be the ancestors of the horses, by
two cursorial branches of the lophiodonts (the helale-
tids and the chasmotheres), and by two cursorial
Figure 28. — Evolution of the skeleton of the titanotheres
A, First stage (subcursorial), lower Eocene, Lambdotlierium popoagicum; B, second stage (subcursorial), lower Eocene, Eotitanops
horealis; O, intermediate stage (mediportal), middle Eocene, Palaeosyops leidyi; D, final stage (graviportal), lower Oligocene, Brontops
Tobustus. From one twenty-eighth to one-thirtieth natural size.
branches of the rhinoceroses (the triplopodines and the
hyracodonts) . It is shown elsewhere (see Chap. IX)
how the cursorial habit, independently assumed in
each of these subfamilies, modified not only the limbs
but the skull and the entire skeleton into analogous
forms that simulate real affinity. In Figure 32 all
these cursorial branches, independently evolving in
analogous adaptation. The manner in which the
skeleton and limbs similarly became adapted inde-
pendently to various modes of locomotion and thus
assumed analogous forms and proportions is no less
remarkable than the independent adaptation of the
teeth to similar kinds of food.
Of the nine typical perissodactyl families the horses
alone are cursorial through the entire period of
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
37
ProtitartotfieriuTrh emargrinaticm.
Upper Eocene ( upper Uinta)
Manteocems manteoceras
Middle Eocene (upperBridger)
Zimnohyops priscus
M/ddle Eocene (lowerBridger)
Eotitanops borecdzs
Lower Eocene (Wind River)
Figure 29. — Evolution of the skull and molar teeth in the titanotheres
In EoUtanops the facial part of the skull is longer than the brain case (cranium). In Brontotherium the face is very short and the brain
ease is very long. The horn swellings (H) first appear in Manteoceras and become very prominent in the succeeding stages. The top of
the skull becomes deeply concave. The outer wall and the V-shaped cusps of the upper molar teeth (paracone, metacone) become
very deep, while the inner cusps (protocone, hypocone) retain their low, conical form. The lower molars retain the W-shaped crown
throughout, which increases considerably in depth.
38
TITANOTHEEES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
different perissodactyl families, are indicated by dif-
ferent kinds of shading.
Forest-living habits among perissodactyls are some-
what more rare, especially the extreme adaptation to
forest living, consisting of relatively slow locomotion
and marked special adaptation to browsing on the
leaves of trees. Types that are more or less fully
Aquatic branches of the perissodactyls are also more
or less readily distinguishable. Among the titano-
theres we have a group of swamp or river living forms,
with short limbs and spreading feet, whose remains
are preserved in many river-channel sandstones,
namely, the genera DolichorMnus and Metarhinus,
which are clearly distinguished from all other titano-
FiGURE 30. — Adaptive radiation in tlie evolution of the upper molar teeth in the odd-toed hoofed mammals
(perissodactyls)
After W. D. Matthew. The earliest members of each family had low-crowned (brachyodont) teeth, of relatively simple pattern. In the
titanotheres and paleotheres the internal cusps remain low and the two outer main cusps form a W. In the horses (hypsodont) the whole sur-
face of the crown is thrown into complex crests and ridges and the crown becomes very long. In the tapirs (brachyodont) the molar crown
takes the form of two sharp cross crests. A somewhat similar pattern is seen in the lophiodonts, e.Kcept that in this family (brachyodont)
the outer cusps form an irregular outer wall. In the rhinoceroses (brachyodont to hypsodont) the outer wall (ectoloph) becomes very much
flattened, elongate, and oblique, and the cross crests also become oblique.
adapted to forest living are represented, we believe,
among the chalicotheres, among certain forest-living
horses {HypoJiippus), and among certain forest-living
tapirs (Tapirus terrestris), all relatively slow in move-
ment and all without conspicuous weapons of offense
or defense, except that the chalicotheres, such as
Moropus, are provided with heavy claws.
theres by their apparent adaptations to river-border
or aquatic life. Certain tapirs frequent river borders
and swim freely for long distances, but they do not
acquire distinctive aquatic adaptations. Among the
rhinoceroses the pronounced aquatic division is the
amynodonts, which have marked aquatic features
about the head, simulating those of the hippopotami.
INTRODUCTION TO MAMMALIAN PALEONTOLOGY
The great family tree of the perissodactyls may be interpreted as shown below.
Family tree of the perissodactyls
39
Primitive ancestors
Ancient branches
Families, extinct and living
A. Bunoselenodont branch of basal Eocene
time: inner cusps bunoid, conical; outer
cusps selenoid, crescentic.
1. Titanotheres.
2. Chalicotheres.
3. Paleotheres.
4. Horses.
Perissodactyls of Upper Cretaceous
and basal Eocene time: four digits on
the fore foot, three on the hind foot;
six rounded cusps on the upper grind-
B. Bunolophodont branch of basal Eocene
time; inner cusps crested, outer cusps
symmetrically crested and more or less
flattened.
5. Tapirs.
6. Lophiodonts, mediportal and graviportal;
confined to Europe. Helaletids, cursorial
lophodonts; reaching America.
ing teeth.
C. Lophodont branch of upper Eocene
time; inner cusps crested, outer cusps
asymmetrical, greatly flattened.
7. Amynodonts (aquatic).
8. Hyraoodontidae (cursorial and medi-
portal) .
9. True rhinoceroses (mediportal and gravi-
portal), variously adapted to browsing
and grazing; distinguished by variations
in the evolution of the horns.
The mediportal structure, in which the skeleton
and limbs are adapted to moderate speed and weight,
embraces those intermediate stages in several different
families in which there was moderate body weight
and moderate speed, as in the tapirs. In the tapirs
this is the last term of evolution, but in the titanotheres
and in many rhinoceroses the mediportal stage is
simply a gateway to the graviportal stage, in which
the proportions of the limbs and trunk are adapted to
weight bearing, more or less rapid progression, and
active offense and defense.
The interpretation of these phenomena of analogous,
parallel, and convergent evolution under the princi-
ple of adaptive radiation, presented on pages 121-127,
simplifies the problem of the anatomy of the group
as a whole as well as of the several adaptations
seen in the skull, skeleton, limbs, and teeth. Each
perissodactyl family appears to exhibit an innate
potentiality to evolve in many different directions
and thus to meet new conditions of life. In this
sense each family is plastic. Here we are not wit-
nessing the direct action of the environment: we
are witnessing the direct response of the organism,
through largely unknown causes, to develop its poten-
tial heredity characters along certain new lines. If
the supply of new potential characters is exhausted,
if a mechanical stage is reached out of which no addi-
tional stages can be developed, the animal will tend
to become extinct unless it can retire to the recesses
of the forests, as did the chalicotheres, and thus escape
a struggle for existence in competition with more
plastic forms, better adapted to the grazing life. The
interpretation of these processes, however, has been
the most difficult and baffling of all the problems that
have arisen in the research made for this monograph.
The interpretation of the modes and causes of the
origin and evolution of new characters and of new
proportions in response to new conditions of life
(see pp. 834-849) is extremely difficult. Explanations
that at first seem obvious appear on close analysis
not to be explanations at all. As this monograph is
the most exhaustive and most detailed study thus far
made of any group of mammals it seems important
to show the bearing of all the observations on each of
Figure 31. — Three types of teeth of members
of nine typical families of perissodactyls
Bunoselenodont (A), bunolophodont (B), and lophodont (C)
types of teeth displayed in the short-crowned Cbrachyodont)
stage.
the current theories of evolution. It appears that,
as is fully set forth in Chapter XI, we are still very
far from even a preliminary understanding of the
causes of many of the processes of mammalian
evolution.
VELOCITY IN THE DEVELOPMENT OF CHARACTERS AND IN
PHYLOGENY
The earliest explanations of evolution were purely
mechanical; we are now passing through a phase of
40
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
chemical explanations; but it appears that we may
be led to the adoption of certain physical conceptions
and the use of certain physical terms (Osborn, 1917.
462) for what has been described above as the rate of
evolution of certain characters as distinguishing
genera. For the term "rate" we will substitute the
term "velocity."
Ontogenetic velocity. — The velocity of the evolution
of certain characters in embryonic development — in
fact, throughout the whole course of individual
development — has long been a very familiar feature
of adaptation. From the embryo onward a char-
isms, and we shall see that the most plausible explana-
tion of it thus far offered is the theory of natural selec-
tion proposed by Darwin.
Phylogenetic velocity. — Another kind of velocity,
however, may be noted in the perissodactyls and may
be measured and calculated with great precision in
the numerous phyla of titanotheres here considered.
This velocity may be called phylogenetic velocity. Its
postulation rests upon the fact that a given character
may evolve much more rapidly in the members of one
phylum than in the members of a related phylum, al-
though the environment of both phyla may be the
FiGiTRB 32. — The family tree of the perissodactyls, showing adaptive radiation of the nine families and thirty-
five subfamilies
Exhibiting their divergence In limb and foot structure into cursorial, forest-hving, mediportal, and graviportal types and in tooth structure into
browsing and grazing types.
acter may be either hurried along or slowed down in
its rate of development, and in consequence it will
appear in earlier or later stages of individual life. For
example, certain adult proportions of the limbs are
needed at birth in all cursorial animals; these adult
proportions are consequently hurried forward during
the foetal life, so that the animal is at birth able to
run immediately with almost the same speed as the
parent. This kind of velocity of development is
called ontogenetic; it is appurtenant to every char-
acter in every stage of its development, it is closely
connected with the survival of certain young organ-
same. For example, in twelve subfamilies of titano-
theres we observe homogeneous characters evolving
independently — the same cusps on the teeth, the same
horns on the skull. How, then, do the subfamilies
differ from one another? They differ because the
evolution of each character in each phylum proceeds
with its distinctive velocity. In a phylum that is
evolving rapidly a certain character appears early in
geologic time; in a phylum that is evolving slowly the
same character appears late in geologic time. The
titanotheres of one phylum may at a particular geologic
period be completely hornless, whereas those of a con-
INTBODUGTION TO MAMMALIAN PALEONTOLOGY
41
temporaneous phylum may have well-developed horns.
In the former the horns may appear much later and
may never acquire very great momentum in develop-
ment. We can thus note the incipiency of the differ-
ences between the short-horned titanotheres and the
long-horned titanotheres.
This principle of unequal phylogenetic velocity in
the development of the same characters enables us to
distinguish different genera and species. In one genus
the development of the internal cusps of the premolar
grinding teeth shows high velocity; in another genus
it shows low velocity. Apparently these internal
cusps are useful adjuncts of the tooth; they make the
tooth more effective for grinding up food. Similarly
the horns are useful adjuncts of the head in warding
off enemies. Yet these characters evolve so slowly
in certain phyla that it is unreasonable to believe that
utility and natural selection are the prime causes of
their evolution. There would seem to be physiological
and physical (or chemico-physical) causes of these
different velocities. It is the data on the different
velocities of the developmen-t of the same characters in
related phyla which give the principal biologic value to
the long series of detailed measurements and justify
the large number of figures that are presented in Chap-
ters V and VI. This suggests a summary of the bio-
logic aspects of the problems of this monograph and
of the features that distinguish this particular field of
biologic research.
SUMMARY OF THE EVOLUTION OF THE TITANOTHEREG
The known titanotheres were confined to a relatively
small area near the fortieth parallel in western North
America and to Europe and Asia. The direct lines of
descent and the continuous changes in many branches
in different or successive life zones were complicated
by the occasional incursion of new families from out-
side larger regions, probably from northern America
and perhaps from northern Asia. (See appendix.)
Nevertheless the localities in western North America
where the remains of titanotheres have been found
were apparently near the main geographic center of
the evolution of the family, for the series of known
fossils enables us to follow almost every step in the
slow transformation of forms that were small and
defenseless to forms that were huge and well armed.
The remains of the titanotheres now collected repre-
sent the most complete evolutionary series of mammals
thus far discovered except those of the horses. The
horses, however, are much less highly differentiated.
In the titanotheres we see the growth of a great and
vigorous family tree, giving off numerous branches
(phyla), which diverge in characters and habits while
retaining hereditary resemblances and certain heredi-
tary trends and tendencies of transformation. Each of
these branches is made up of slowly transforming
successive stages (mutations of Waagen), which appear
101959^29— VOL 1 5
to be the more continuous and unbroken by sudden
change the more thoroughly we explore the geologic
levels where they successively occur. The evolution
of the soft parts can only be inferred. The hard parts
evolve in a variety of ways, chiefly through increase
of size, through changes in proportion, through addi-
tion of new parts, and in less measure through loss of
parts. Actual addition or loss of parts in the titano-
theres is rare; general increase in size is almost uni-
versal, though in a few branches the size is diminished
or arrested.
Changes in the proportions (allometrons) of struc-
ture were brought about by different velocities of
phylogenetic evolution (acceleration and retardation)
in the skeletal framework as a whole and in each of
its parts. No less important is the definite and
successive addition of new characters (rectigra-
dations), each developing from infinitesimal begin-
nings until it reaches a stage of usefulness and
each apparently having its individuality (biocharacter)
and its separate history.
Throughout this wonderful transformation, which
is in general adaptive, there were certain manifest
germinal (hereditaiy) tendencies and certain unkno\vn
interactions between these germinal changes and' the
external, habitudrnal, and environmental influences.
The more carefully we study the detailed characters in
each branch the more evident it becomes that the
causes of evolutionary development are neither exclu-
sively external nor exclusively internal but are to
be sought hypothetically in the interactions between
germinal, habitudinal, and environmental forces. The
changes in the proportions of the skeletal characters
and the new elements added to the teeth and skull,
which are the outward expressions of these hypothetic
germinal and environmental reactions, become visible
more or less contemporaneously but not simultane-
ously in all members of the branches and sub-
branches of the great family tree — that is, the
same characters appear, but at different periods
and with different velocities of development. The
whole process is an orderly one, which is, however,
not predetermined in the germinal constitution of the
titanotheres but results from certain innate or germinal
potentialities of evolution, which are evoked in response
to certain environmental and habitudinal conditions.
The struggle for existence, or natural selection, is
operating continuously and more or less strongly on
every single character according as its survival value
is greater or less. In each successive geologic level
we witness alterations of the family tree — its impover-
ishment through the extinction of certain branches or
its augmentation through the survival of other
branches and the immigration of branches which
evolved in other regions. The individual members of
all the branches (with two exceptions) become more
imposing and more diverse as time goes on. Finally,
42
TITANOTHEEES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
at the climax of the general trend of transformation
and at the very height of the grandeur of development,
we observe the apparently simultaneous extinction of
the whole titano there family, seemingly through failure
to cope with changed environmental conditions or to
compete successfully with other herbivorous types.
This contribution to biology is therefore important
chiefly as a study of the actual modes of evolution as
observed in the skeleton and teeth of many different
members of a great family of extinct animals which
existed throughout a long period of geologic time —
from the early Eocene through the early Oligocene —
a time reckoned as hundreds of thousands of years.
It is merely suggestive as to the causes of evolution.
SECTION 3. BIBLIOGRAPHY OF LITERATURE CITED OR
CONSULTED IN THE PREPARATION OF CHAPTER I
Barrell, Joseph.
1917.1. Rhythms and the measurement of geologic time:
Geol. Soc. America Bull., vol. 28, pp. 745-904,
Dec. 4, 1917.
Darwin, Charles.
1859-2. The origin of species by means of natural selection,
or the preservation of favored races in the
struggle for life, 502 pp. London, John
Murray, 1859.
FOBSTER-COOPER, C.
1913.1. Thaumastotherium osborni, a new genus of perisso-
dactyls from the upper Oligocene deposits of
the Bugti Hills of Baluchistan (preUminary
notice): Annals and Mag. Nat. Hist., 8th ser.,
vol. 12, pp. 376-381, October, 1913.
1913.2. Correction of generic name [to Baluchitherium]:
Annals and Mag. Nat. Hist., 8th ser., vol. 12,
p. 504, November, 1913.
Hayden, Ferdinand Vandiveer.
1873.1. United States Geological Survey of the Territories;
First, Second, and Third Annual Reports,
reprinted in one 8vo volume, 1873. (First
and second first issued in Rept. Commissioner
of General Land Office for year 1867, Wash-
ington, 1867; Third issued independently as
Preliminary Field Report of U. S. Geol. Survey
of Colorado and New Mexico, 1869.)
Kovalevsky, Dr. Woldemar.
1873.1. Monographic der Gattung Anthracotherium Cuv.
und Versuch einer nattlrlichen Classification
der fossilen Hufthiere: Palaeontographica,
Band 22, Heft 3, pp. 131-210, Taf. 7-9, 1873;
pp. 211-346, Taf. 10-17, 1874.
Leidy, Joseph.
1869.1. The extinct mammalian fauna of Dakota and
Nebraska, including an account of some allied
forms from other localities, together with a
synopsis of the mammalian remains of North
America: Acad. Nat. Sci. Philadelphia Jour.,
2d ser., vol. 7, pp. 1-472, 30 pis.
Linnaeus, Caroltjs.
1758.1. Systema naturae . . . Editio decima, reformata,
Holmiae, 1758.
Matthew, William Diller.
1901.1. The Carnivora and Insectivora of the Bridger
Basin, middle Eocene: Am. Mus. Nat. Hist.
Mem., vol. 9, pt. 6, pp. 291-567, pis. 43-52,
1901.
Meek, F. B.,
1862.1 (and Hayden, F. V.). Descriptions of new lower
Silurian (Primordial), Jurassic, Cretaceous, and
Tertiary fossils collected in Nebraska Territory
by the exploring expedition under the command
of Capt. Wm. F. Raynolds, U. S. Top. Engrs.,
with some remarks on the rocks from which
they were obtained: Acad. Nat. Sci. Philadel-
phia Proc, vol. 13, pp. 415-447, 1862.
Merriam, C. Hart.
1918.1. Review of the grizzly and big brown bears of North
America (genus Ursus), with description of a
new genus, Vetularctos: U. S. Dept. Agr. Bur.
Biol. Survey North Am. Fauna, No. 41, 136 pp.,
16 pis., Feb. 9, 1918.
OsBORN, Henry Fairfield.
1896.107. Titanotheres of the American Museum of
Natural History: Am. Naturalist, vol. 30,
No. 350, pp. 162-163, February, 1896.
1896.110. The cranial evolution of Titanoiherium: Am.
Mus. Nat. Hist. Bull., vol. 8, pp. 157-197,
July 31, 1896.
1900.192. Phylogeny of the rhinoceroses of Europe,
Rhinoceros Contributions No. 5; Am. Mus,
Nat. Hist. BuU., vol. 13, pp. 229-267, Dec. 11,
1900.
1909.321. Cenozoic mammal horizons of western North
America, with appendix, Faunal lists of the
Tertiary Mammalia of the West by William
Diller Matthew: U. S. Geol. Survey BuU. 361,
138 pp., 1909.
1914.409. Recent results in the phylogeny of the titano-
theres: Geol. Soc. America BuU., vol. 25, No.
3, pp. 403-405, Sept. 15, 1914.
1914.412. Rectigradations and allometrons in relation to
the conception of the "mutations of Waagen"
of species, genera, and phyla: Geol. Soc.
America BuU., vol. 25, No. 3, pp. 411-416,
Sept. 15, 1914.
1917.462. The origin and evolution of life on the theory of
action, reaction, and interaction. New York,
Charles Scribner's Sons, 1917.
Osgood, Wilfred H.
1909.1. Revision of the mice of the American genus
Peromyscus: U. S. Dept. Agr. Bur. Biol.
Survey North Am. Fauna, No. 28, 285 pp.,
7 pis., map, Apr. 17, 1909.
Waagen, W.
1869.1. Die Formenreihe des Ammonites subradiatus,
Versuch einer palaontologischen Monographic:
Geognostisch-palaontologische Beitrage heraus-
gegeben * * * von Dr. E. W. Benecke,
Band 2, pp. 179-257, 1869.
White, C. A.
1868.1. First and Second annual reports of progress by
the State geologist and the assistant and
chemist on the Geological Survey of the State
of Iowa, etc., 284 pp., Des Moines, 1868.
CHAPTER II
ENVIRONMENT OF THE TITANOTHERES AND EFFECT OF ADAPTIVE RADIATION ON THEIR
VARIATION
SECTION 1. GEOLOGY AND GEOGRAPHY
CORRELATION OF EARLY TERTIARY EVENTS IN THE ROCKY
MOUNTAIN REGION WITH THOSE IN WESTERN EUROPE
The recorded history of the titanotheres extends
from the upper horizons of the lower Eocene series
(upper Ypresian or upper Wind River horizon)
through the middle and upper Eocene to the top of
the lower Oligocene (Sannoisian or Chadron horizon),
covering a period estimated at 450,000 to 600,000
years. This estimate is based on the assumption
that 9,000 to 12,000 feet of sediment was deposited
during the period from basal Eocene to lower Oligocene
time and that the average rate of deposition was 1
foot in every 100 years.
The Eocene type formations (Wasatch, Bridger,
etc.) of the Rocky Mountain region in North America
have gradually acquired a time significance, similar
to the stages (etages) into which the Eocene and
lower Oligocene of Europe are divided, as shown in
the following table. The correlation in time between
France and America is close for some periods, as, for
example, between the Sparnacian and lower Wasatch
and between the Sannoisian and Chadron. For
other periods the correlation is provisional, because
the faunal relations are interrupted.
Provisional correlation of European and American geologic stages and life zones of the tifanothere epoch
Epochs
Stages (Stages) ot Europe
Type formations ot America
Major type life zones
Maximum
thiclvuess of
sediments in
feet, deducting
overlaps
Lower Oligocene.
Sannoisian.
Chadron (Nebraska and South
Dakota).
(Extinction of titanotheres.)
Tilanotherium-Mesohippiis.
500
Upper Eocene.
Ludian.
Uinta (northeastern Utah).
Diplacodon-Protitanotherium-Epi-
hippus.
600
Bartonian.
Lutetian.
Ypresian (upper).
Bridger (southwestern Wyoming).
Uintatherium-Manteoceras-Mesa-
tirhinus.
Palaeosyops paludosus-Orohippus.
Eometarhinus - Trogosus - Palaeo-
syops fontinalis.
Middle Eocene.
1,875
Lower Eocene.
Ypresian (lower).
Sparnacian.
Wasatch (western W3'oming).
Coryphodon.
(First titanotheres.)
2,025
Transition.
Cernaysian.
Thanetian.
O 03
Torrejon (northwestern
New Mexico).
Puerco (northwestern New
Mexico) .
Pantolambda.
Polymastodon.
6,000
Basal Eocene.
Total 11,000
Cretaceous. °
Montian.
Danian.
Lance ( = in part Laramie and
Denver)."
Triceratops.
" The United States Geological Survey classifles the Lance formation as Tertiary (?), the Laramie formation as Upper Cretaceous, and the Denver as Eo
author of this monograph believes that the Lance formation is equivalent in part to the Laramie and Denver formations and that it is of Cretaceous age.
43
44
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
All estimates orgeologic time are highly provisional,
because they involve two unknown quantities — the
amount of overlap and the relative rate of deposition.
The rate of the deposition of sediments varies enor-
mously. For example, certain Fort Union sediments
of Montana, aggregating 6,000 feet in thickness, are
at present considered contemporaneous with Torre-
Lambdotheruum popoaoicum £otUanoDS orinceps
Figure 33.-
-Outlincs of the bodies of titanotheres at different stages
of evolution
jon sediments of New Mexico, which aggregate only
385 feet. It would therefore appear that sedimenta-
tion in Montana was more than thirteen times as
rapid as in New Mexico. The only sedimentary
stage which appears fairly uniform in several geo-
graphic localities is the Wasatch, which exhibits beds
of approximately the same thickness in many different
regions.
If an average rate of deposition of a foot in a century
is assumed, the period from basal Eocene to lower
Oligocene time, inclusive, is estimated as not exceeding
1,100,000 years, a moderate estimate considering the
great biologic changes that took place in the titano-
theres and other groups during this period. The
epoch of the titanotheres is roughly estimated at
500,000 years or more, during which they steadily
increased in size, from the geologically earliest
animals, which are no larger than a sheep, to some
of the latest members of the race, which exceeded
in size the largest rhinoceroses, standing over SJ^
feet at the shoulders.
The recorded history of the titanotheres is
nearly unbroken, but there have been two evolu-
tionary gaps, one between the lower and the mid-
dle Eocene, which was filled in 1918 by explora-
tions of the Huerfano (Osborn, 1919.494), and one
between the upper Eocene and the lower Oligo-
cene, which will be filled by the exploration of
the upper part of the Uinta formation (theoretic
faunal zone 16, still unknown). The record also
shows sudden transitions caused by invasions of
animals from other regions.
The geographic range of the titanotheres was
probably continent wide in America and also ex-
tended across Asia into the Balkan region of south-
eastern Europe. In the relatively small Rocky
Mountain and western plains region, where most
of the fossil remains have been discovered, we
observe the successive invasion of new kinds of
titanotheres, which had apparently evolved pre-
viously in other regions, probably in areas to the
north and east.
The geologic age of the little-known European
titanotheres is somewhat uncertain. The type
and only known specimen of Brachydiastemaihe-
rium, an animal about the size of Diplacodon, is
recorded fi'om a formation in eastern Hungary
that was originally assigned to the lower Eocene,
but this animal is in a stage of evolution corre-
sponding to that of the uppermost Eocene titano-
theres of America, and the same European forma-
tion has yielded remains of a primitive rhinoce-
ros (ProTiyracodon) of upper Eocene or even lower
Oligocene type. Brachydiastematlierium is there-
fore probably not of lower Eocene age. The
animals described as Menodus rumelicus and
Titanotherium hohemicum are in all respects sim-
ilar to American titanotheres of lower Oligocene
age, but as the localities and horizons from which
these fragmentary specimens were obtained are in
doubt they may be imported American fossils to which
a European origin has been erroneously imputed.
The correlation of the chief geographic, geologic,
climatic, and faunistic events during the Tertiary
period in the Eocky Mountain region with those in
western Europe has been studied by the author con-
tinuously during the last 20 years, with the coopera-
ENVIRONMENT OF THE TITANOTHERES
45
tion of Depgret in France and of Matthew, Merriam,
Granger, Brown, Peterson, Douglass, Riggs, Darton,
Stanton, Berry, Knowlton, and others in this country.
The theoretic correlations reached are shown in the
accompanying tables (pp. 43, 48). The comparison of
similar stages in the evolution and migration of floras
and faunas is partly independent of changes in the
surface of the earth and in climate and is partly
related to them. The general succession (Osborn and
Matthew, 1909.321; Osborn, 1910.346) of the four
Eocene and Oligocene life phases of North America is
as follows:
Phase IV (lower Oligocene) , approximation. — A similar mam-
mal fauna in western America and western Europe. Extinction
of archaic fauna and invasion of modern fauna.
Phase III {upper and middle Eocene), estrangement. — Inde-
pendent mammal fauna of western America and western Europe;
gradual diminution of archaic fauna.
Phase II (lower Eocene), approximation. — Closely allied and
similar fauna of western America and western Europe; first
invasion of modernized fauna.
place this after the first Rocky Mountain (Laramide)
revolution in post-Laramie time — that is, after the
end of typical Laramie deposition in Colorado.
Others, among them the author of this monograph,
place it at the time of the extinction of the great land
and marine reptiles of Europe and America — that is,
after Lance time.'' The Fox Hills formation, which
underlies the Lance, represents the end of uniform
widespread marine sedimentation. At some places
the Fox Hills is continuous with overlying fresh-
water deposits laiown as Laramie; at others it is con-
tinuous with overlying deposits known as the Lance.
Thus Laramie time and Lance time, in our opinion,
are in part the same — that is, they overlap at some
places.
Lance and Fort Union flora. — New physiographic
and climatic conditions arose during the initial period
of the Rocky Mountain uplift, when uplands and
plateaus were formed. Knowlton and Berry have
shown that the Fort Union flora extends back into
land areas
Forme
Known fossil areas
iigration areas
Figure 34. — Map showing the known areas (black) and the hypothetical areas (oblique lines)
of titanothere migration and habitat
Phase I (basal Eocene) , approximation. — Partly similar archaic
mammal fauna of western America and western Europe.
Final Mesozoic phase. — Gradual extinction of the upper
Cretaceous dinosaur fauna and appearance of ancestors of the
archaic Eocene fauna.
This alternate approximation and estrangement of
the mammal life of western America and western
Europe points to periods during which conditions
favored intermigration and intervening periods when
geographic, climatic, or forest barriers may have stood
between these widely separated regions. The basal
Eocene American forests — those of the Fort Union
epoch, for example — were very luxuriant and were
unfavorable to migration.
lATE CRETACEOUS AND EARLY TERTIARY CLIMATES
End of the Cretaceous period. — The initial point in
the correlation of geologic time in both the Eastern
and the Western Hemisphere is-the end of Cretaceous
deposition. (See table on p. 48.) Some geologists
Lance dinosaur time, regarded by the author as late
Cretaceous. The Lance flora is prevailingly a rela-
tively warm temperate flora as compared with the
antecedent Laramie and other Upper Cretaceous
floras in the same region, and the climate in Lance
time was about like that of the present Atlantic Coast
States from North Carolina southward. In the Rocky
Mountain province (Berry, 1914.1, pp. 153-154), in
the zone of transition from the Cretaceous to the
Eocene, a large number of local floras appear, such
as those in the Arapahoe and Denver formations of
Colorado, the Livingston formation and the Lance
formation ("Hell Creek beds") of Montana, and the
typical Lance formation of Wyoming. The forma-
tions in which they occur consist of lacustrine,
fluviatfle, and terrestrial deposits eroded from the
rising land area of the Rocky Mountain province.
These early so-called post-Laramie floras are said to
8 The United States Geological Survey classifies the Lance formation as Terti-
I ary (?). The author of this monograph regards it as Cretaceous.
46
TITANOTHEEES OF ANCIENT "VA^yOMING, DAKOTA, AND NEBRASKA
1. Sweet Grass County, Mont. Fort
Union formation.
2. P. T., San Juan Basin, N. Mex. and
Colo. Puerco and Torrejon forma-
tions and "Tiffany beds."
3. W., near Evanston, Wyo. Typical
Wasatch group.
4 Big Horn Basin, Wyo. Wasatch for-
mation.
5. W. R.. Wind Eirer Basin, Wyo.
Typical Wind River formation.
6. Beaver Divide, Wyo. Eocene and
Oligocene section.
7. H., Huerfano Basin, Colo. Typical
Huerfano formation.
8. B., Bridger Basin, Wyo. Typical
Bridger formation.
9. W. K., Washakie Basin, Wyo. Typi-
cal "Washakie formation" of Hay-
den.
10. U., Uinta Basin, Utah. Typical
Uinta and older Eocene deposits.
11. Wh. R., White River, S. Dak. Typi-
cal White River group.
12. Powder River and Pumpkin Buttes,
Wyo. Fort Union and Wasatch
formations.
13. F. U., Fort Union, N. Dak. Typi-
cal Fort Union formation.
14. P., Red Deer River, Alberta. Paska-
poo formation.
Oligo
Oligocene flood plain
Figure 35.— General geologic sketch map of the Rocky Mountain region, showing existing topography and drainage areas and
their relation to areas of Eocene and lower Oligocene sedimentation
Each of the numbered areas e.'cccpt 13 and 14 is also represented in geologic section in this chapter. Topography after the United States Geological Survey, 1911
(See tables on pp. 48, 57, .=.8.)
ENVrROlSTMENT OF THE TITANOTHERES
47
be distinct from those of the true Laramie and to be
more closely allied to those of the true Fort Union
above.
The true Fort Union floras of basal Eocene (Thane-
tian) age include between 500 and 600 species of trees,
which were apparently derived from areas farther
north, certainly not from areas farther south.
These forests, which were contemporaneous with the
Puerco and Torrejon mammals, indicate a climate
in the Rocky Mountain region between the fortieth
and fiftieth parallels that was far from tropical, yet
moderately warm and humid, with mild winters, favor-
able to the growth of palm, fig, and camphor trees,
as well as other warm-temperate trees, including gink-
gos and sequoias. This flora, which is characteristic
of the early uplift period of the Rocky and Uinta
Mountains in Colorado and Wyoming, indicates a
somewhat cooler climate than that of the subsequent
lower Eocene (Green River) epoch in the same region
and a much cooler climate than the subtropical climate
of the South Atlantic States in early Eocene time. In
fact, both in the Rocky Mountain region and farther
south the American climate became milder and more
tropical as the Eocene epoch advanced.
EOCENE GEOGRAPHY OF WESTERN NORTH AMERICA AND
ITS RELATION TO FAUNAI MIGRATIONS
GEOGRAPHIC DIVISIONS AND THEIR BEARING ON
MIGRATION
The main topographic features of western North
America were established between late Cretaceous and
middle Eocene time. In late Cretaceous and early
Eocene time certain routes of migration connected
the animal life of the central Rocky Mountain
region with that of Eurasia and probably with
that of South America. The key to these routes
of migration and to the geographic distribution of
these animals is afforded by the results of researches
made since 1853 by the geological surveys of the United
States and Canada. The foundation of the descriptive
geologic history of the Rocky Mountain region is laid
in the report of F. B. Meek and F. V. Hayden (Meek
and Hayden, 1862.1).
The entire Cordillera region extends from Bering
Strait to the Isthmus of Tehuantepec, a distance of
4,500 miles, and has an average width of 500 to 600
miles. The main geographic divisions of the Cor-
dilleran region, named in order from east to west, are
the following :
Rocky Mountain Range, Bering Sea to Colorado,
including —
Front or eastern range, facing the Great Plains.
Rocky Mountain basins between the eastern and
western ranges, forming the central north and
south migration routes of mammals.
Westerly ranges, facing the interior plateaus.
Central interior plateaus, intermontane belt region
(main migration routes of herbivorous mammals) :
Northern interior plateaus, Alaska to Washington.
Columbia Plateau.
Nevada-Sonora Plateau (Great Basin).
Colorado Plateaus.
Mexican Plateau.
Pacific mountain system, British Columbia:
Sierra Nevada.
Pacific mountain basins between the Sierra
Nevada and the Coast Ranges. Coastal
migration routes of mammals.
Pacific Coast Range.
A transverse section of the Cordillera on the 41st
parallel exhibits clearly the main confines of these
mountain ranges, basins, and plateaus. The great
plateaus and the mountain basins may have pre-
sented bordering forests and central grassy plains
and jungles, interspersed with swamps, marsh lands,
rivers, and lakes similar to those in the plateau
and mountain (Kenya, Kilimanjaro) region of equa-
torial Africa to-day. Migration from north to south
or from south to north was possible along three
routes.
Our only knowledge of the late Cretaceous and
Eocene mammal life of North America is afforded
by the remains of mammals of the Rocky Mountain
basins and foothills from Alberta to northern New
Mexico. During the Oligocene epoch the life of the
Columbia Plateau is revealed in the John Day forma-
tion of Oregon. The life of the Great Plains first
appears in the lower Oligocene formations in South
Dakota, Wyoming, Nebraska, and Colorado, which
border the Rocky Mountains on the east. The Eocene
mammalian life of the country that stretches east-
ward from the Rocky Mountain Front Range to the
Mississippi and the Atlantic coast is entirely unknown.
48 TITAKOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBEASKA
Correlation of late Cretaceous and early Tertiary stages in Europe and in North America
Epochs
European stages
Rocky Mountain and Plains formations
Changes in flora and climate
Chief forms of reptile and mammal
Upper Eocene.
Ludian.
Bartonian.
Uinta formation {Diplacodon
zone), upper part of "Wa-
shakie" formation (Washakie
B), and (?) upper part of
Bridger formation (Bridger
E). .
Ancestors of horned titan-
otheres.
Middle Eocene.
Lutetian.
Upper Ypresian.
Lower Eocene.
Lower Ypresian.
Sparnacian.
Transition.
Cernaysian.
Basal Eocene.
Uppermost Creta-
ceous."
Upper Cretaceous.
Danian.
Maestrichtian.
Lower part of Bridger forma-
tion (Bridger A, B, C, and D),
lower part of "Washakie"
formation (Washakie A), and
upper part of Huerfano for-
mation (Huerfano B).
Rapid evolution of titano-
theres (upper Bridger).
Post- Wasatch and post-Green River uplift, Uinta Mountains, Utah.-
Wind River, Green River, and
Wasatch formations and low-
er part of Huerfano forma-
tion.
Green River flora, show-
ing affinity to tropical
flora of the south;
climate warmer than
Fort Union.
Post-Fort Union mountain uplift, Montana and Colorado.
Fort Union, Torrejon, and Pu-
erco formations. Swamp, la-
goon, forested flood-plain
sediments; lignitic and coal
' deposits.
Lance (upper part) , Denver and
Arapahoe formations. Ris-
ing land area of Rocky Moun-
tain region; brackish-water
estuarine, fluviatile, and chan-
nel sediments.
Fort Union flora of mod-
ernized types.
Appearance of titanothe-
res (Wind River time).
Appearance of modernized
families (lower Wasatch
time) .
Archaic mammals of Pu-
erco, Torrejon, and Fort
Union time.
Extinction of the dino-
saurs and large marine
reptiles.
Fort Union flora. Warm
and humid climate
similar to that of south-
eastern coastal States;
mild winters, flora not
tropical. Low-lying
forested swamps in the
plateau region. Open
flood plains surround-
ing the mountain
slopes.
Triceratops-Tyrannosaurus
fauna.
Mammals of Lance time.
Ancestors of Puerco and
Torrejon placentals, mar-
supials, multitubercu-
lates. Paskapoo mam-
mal fauna of Alberta
(more recent).
Beginnings of Laramide revolution; Rocky Mountains (Colo.), Uinta Mountains (Utah),
Wasatch Mountains (Utah).
Uppermost of the conformable
series sediments of Rocky
Mountain and Plains region:
Laramie formation ( = low-
er part of Lance).
Fox Hills sandstone.
Pierre shale.
Edmonton flora of Al-
berta (similar to Fort
Union) .
Laramie flora transitional
to modern.
Upper Cretaceous flora.
Climate warmer than
Fort Union.
Edmonton dinosaur {Lep-
toceratopa) fauna (suc-
ceeding Belly River), of
Fox Hills (?) age; Ojo
Alamo (N. Hex.) dino-
saur fauna similar to
Judith River fauna;
Judith River (Mont.)
and Belly River (Al-
berta) dinosaur fauna;
Monoclonius of Pierre
age.
■■ The Lance formation is classified by the United States Geological Survey as Tertiary (?) and the Denver and Arapahoe formations as Eocene.
Note.— Near the end of Cretaceous time the chief uplift of the Laramide revolution in the Rocky Mountains began in the Front Range of the Colorado Rockies
after the Laramie and before the Arapahoe. In the northern (the Montana) Rooky Mountains the chief uplift occurred at the end of the Fort Union. In southern Colo-
rado and in northern New Mexico uplifts occurred both before the Puerco and after the Torrejon. (Ransome, 1915.1, pp. 360-3C2.)
ENVIKONMENT OF THE TITANOTHERES
49
Figure 36. — Map of western North America showing supposed routes of migration of animals
This map shows the general early Tertiary topography of the Great Plains, mountain ranges, northern and southern plateaus, and coast basins
and illustrates the supposed lines of Asiatic migration from the north and South American migration from the south. Modified after
F. L. Hansome (1915.1).
50
TITANOTHEKES OF ANCIENT WYOMING, DAKOTA, AND NEBEASKA
Figure 37. — Map showing the orogeny of the western mountain and plateau region
After F. L. Ransome (1915.1). Key to the numerals is given in Figure 36.
ENVIRONMENT OF THE TITANOTHEKES
51
CHARACTER OF THE MOUNTAIN-BASIN, PLATEAU, AND
PLAINS REGIONS
The geographic history of the mountain-basin
region and of the Plains region presents some resem-
blances and some contrasts. Both regions were
subject to slowly progressive elevation during this
period. Nearly all the Eocene deposits of the moun-
tain basins were laid down in broad, fiat valleys and
on mountain plateaus, which were drained largely
by the same great river systems that drain them
to-day, whereas those of the Plains region were
widely scattered over broad flood-plain areas in
which the rivers frequently changed their courses, the
present river courses being cornparatively modern.
In the mountain basins, from the basal Eocene of
the Fort Union, Puerco, and Torrejon formations to
the summit of the upper Oligocene as represented
in the John Day formation of the Columbia Plateau,
the older Tertiary rocks were at very few places
worked over into newer deposits, but at many places
deposition was continuous. Despite continuous ero-
sion since Oligocene time large areas of the historic
Eocene sediments of the mountain-basin region have
been preserved in their original purity and con-
tinuity for the geologist and paleontologist. By
contrast, in the Plains region large areas of the
original Oligocene strata were in part worked over
to form Miocene strata, and part of these in turn
were eroded to form Pliocene strata; again all three
contributed to the Pleistocene strata; and finally all
four are now contributing to the alluvium of the
Great Plains.
EOCENE TOPOGRAPHY IN THE ROCKY MOUNTAIN REGION,
MONTANA TO NORTHERN NEW MEXICO
By middle Eocene time the topography of the Rocky
Mountain region from Montana to northern New
Mexico had become broadly similar to that of to-day.
The existing sharply sculptured ranges of the Big
Horn, Wasatch, Uinta, and San Juan Mountains are
remnants of much loftier ranges, which had their
birth in late Cretaceous and early Eocene time. The
two great drainage systems of the region — (1) Big
Horn, Yellowstone, and Missouri Rivers on the north
and (2) Green, White, San Juan, and Colorado Rivers
on the south — were probably well established at the
end of Eocene time.
According to Ransome (1915.1) and Lindgren
(1915.1) the general uplift of the land in the Rocky
Mountain region near the end of Cretaceous time
was not uniform at different points either in its incep-
tion or in its intensity. Apparently the earliest move-
ment occurred after the deposition in the Denver
Basin of the conformable series of Cretaceous beds
that is now called the Laramie formation, which over-
hes the Fox Hills sandstone. The Front Range of
central Colorado arose at this time, before the deposi-
tion of the Arapahoe formation of Colorado (Ran-
some, 1915.1, p. 361). Andesitic tuffs and flows occur
in the Denver formation, which immediately overlies
the Arapahoe. At the south end of the Rocky Moun-
tains, in northern New Mexico, great uphfts occurred
both before and after the deposition of the basal
Eocene Puerco and Torrejon formations. In con-
trast, in the typical Rocky Mountains of Montana the
principal uplift appears to have taken place at the
end of Fort Union time — that is, subsequent to basal
Eocene time. In the Park Range province of Colorado
there was uplift and vigorous erosion at the end of
the Cretaceous period and renewed uplift after the
deposition of the lower Eocene Wasatch and Green
River sediments.
The separate history of the great mountain ranges
in the basin region also shows that the upward move-
ments began early in Eocene time. The Big Horn
Range of northern Wyoming (Darton, 1906.1) arose
as an anticline from the nearly horizontal strata of
the Plains to a height of 9,000 feet in early Eocene
time. Its uplifted peaks were truncated, and the larger
features of the present topography were outlined.
The major uplift of the Wind River Mountains, which
produced a broad, low, somewhat broken anticline,
also took place in early Eocene time (Fisher, 1906.1).
In the Wasatch Range of western Wyoming, an east-
ward-dipping monocline cut off along its western side
by a great fault, there was only a slight uplift at the
end of the Jurassic, the main uplift taking place at the
end of the Cretaceous (Boutwell, 1907.1). Subse-
quent movement took place in post-Eocene time.
East of the Wasatch Range is the exceptional east and
west anticline of the Uinta Mountains, which extends
eastward and westward as a broad central plateau,
150 miles long and 30 miles wide, forming a dividing
line between the Bridger and Uinta Basins. The for-
mation of the Uinta arch began at the end of the
Cretaceous period (Emmons, 1907.1, p. 302), as is
shown by the fact that the flanking Tertiary beds lie
unconformably over the upturned edges of the older
strata, which stand at angles of 30° or more. The Eo-
cene formations — the Wasatch, Green River, Bridger,
and Uinta — are upturned against the flanks of the
Uinta Mountains, in a position which means that the
continued rise of the mountain mass has dragged up
the edges of the adjoining beds.
Powell estimated that the summit of the Uinta
anticline rose 25,000 feet above the level of the ad-
jacent country — the Bridger and Uinta Basins. This
altitude is equivalent to that of the Himalaya Moun-
tains. Certainly in Eocene time the Uinta was a
lofty, majestic range. The Colorado Front Range arose
between the time of the deposition of the Laramie and
Arapahoe formations, to the south, and the San Juan
Mountains arose at the end of Cretaceous time and
again after the deposition of the basal Eocene Puerco
and Torrejon formations.
52
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
ENVIRONMENT OF THE TITANOTHEKES
53
The entire topography of the mountain-basin region
was thus broadly defined at the end of the Cretaceous
period and was accented by uplifts during and after
Fort Union (Puerco and Torrejon) time; also after
Wasatch and Green River time, following which, from
the present Canadian border to northern New Mexico,
there was a continuous very gradual uplift. In gen-
eral this uplift was earlier and more rapid in Colorado
and New Mexico — that is, it occurred before the Fort
Union epoch — and more retarded in Montana, where
it occurred after the Fort Union epoch. In the Huer-
fano Basin the upturn of the western edge of the
Huerfano beds amounts to 84°, and although this
uplift is local it indicates a considerable movement
in the Sangre de Cristo Range after Wind River
time (W. Granger, letter, 1919). Ransome (1915.1,
p. 362) believes that a large part of the Rocky
Mountain uplift followed the deposition of the Fort
Union formation.
CONTRAST IN PHYSIOGRAPHIC CONDITIONS EAST AND
WEST OF THE ROCKY MOUNTAIN FRONT RANGE
During and after the deposition of the conformable
Cretaceous formations (such as the Fox Hills and the
Laramie) the country bordering the Rocky Mountain
range on the east presented a marked physiographic
contrast to that lying within the Rocky Mountain
basins. Sedimentation east and west of the Rockies
was not contemporaneous.
East of the Rockies. — On the east flanks of the Front
Range great river flood-plain systems began in the
north in Pierre time and extended toward the south
after Fox Hills time. Thus on the western borders
of the present Great Plains region rivers had long been
spreading out sand over their flood plains in Alberta,
forming such deposits as the Belly River sandstone in
Pierre time and the Edmonton sandstone in Fox Hills
time, and extending southward through Montana to
deposit the Judith River sandstone in Pierre time, the
Laramie formation of Colorado, the "Hell Creek beds"
of Montana, the great Lance sandstones of Converse
County, Wyoming, and the Denver and Arapahoe
formations of Colorado after Fox Hills time.
The fact that the Lance sandstones were laid down
at the end of Cretaceous time ^ is shown by the
remains of the horned and carnivorous dinosaurs found
in them, especially Triceratops and Tyrannosaurus.
At about the same time Triceratops alticornis flour-
ished east of the Front Range of Colorado, during the
deposition of the Denver formation, wliich overlies
unconformably (by erosion and uplift) the Laramie,
the topmost formation of the "conformable Cretaceous
series." These great flood-plain deposits, correlated
both by their dinosaurs and by flora of the older Fort
' The United States Geological Survey classifies the Lance formation as Ter-
tiaryC?), but the author regards it as of Upper Cretaceous age.
Union type, mark the beginning of the Rocky Moun-
tain revolution as it affected the country to the east.
At certain localities, notably along Hell Creek, Mont.,
south of the Missouri, these fans of much disturbed
channel sand and gravel are contemporaneous with
undisturbed beds that appear to be lithologically
exactly like those of the Fort Union; consequently
Fort Union sedimentation began in some regions early
in post-Laramie time.
This long period of mountain erosion and sedi-
mentation east of the Rockies came to an end either
through heavy forestation or high-gradient river ero-
sion, which deposited materials farther east. It is a
very significant fact that in the region east of the Rocky
Mountains, between South Dakota and northern New
Mexico, only sparse lower Eocene sediments (Huer-
fano A and Cuchara) are known between Fort Union
(basal Eocene) and Chadron (lower Oligocene) time,
whereas in the region west of the Front Range sedi-
mentation continued through the entire Eocene epoch.
West of the RocTcies. — In the mountain-basin region
from southern Montana to New Mexico the condi-
tions during Lance time were very different from those
that prevailed east of the Rockies. There was ap-
parently erosion and rapid transportation rather than
deposition. Within the mountain basins — except
around Medicine Bow, near Laramie, and around the
Agathaumas sylvestris locality, near Black Buttes,
Wyo. — relatively few deposits of Lance age {Tricera-
tops zone) have thus far been identified by means of
fossils. The Evanston formation, above the Adaville
formation, in the typical Wasatch section of south-
western Wyoming, according to Berry, contains plants
of Fort Union and of Wasatch rather than of Denver
age. Similarly the oldest Eocene deposits of the San
Juan Basin (the Puerco and Torrejon) are comparable
with the Fort Union and not with the older Lance
formation; they overlie unconformably beds of prob-
able Montana age. In brief, few deposits of Lance
time (Triceratops zone) have thus far been identified
within the mountain-basin region, although they may
be found hereafter. At many places the oldest sedi-
ments of the mountain basins lie upon the eroded sur-
faces of unquestioned Cretaceous and older formations
with pronounced unconformity.
Physiographic conditions again changed, apparently,
for after Lance time sedimentation began vigorously
in the mountain-basin region and continued through-
out the Eocene until it formed deposits having a com-
bined thickness of 9,000 to 11,000 feet. (See table on
p. 43.) Not until Oligocene time, when the deposi-
tion of these mountain-basin beds probably ceased,
was great fluviatile and flood-plain sedimentation re-
sumed east of the Front Range, forming the lower
Oligocene Chadron beds.
54
TITANOTHEEES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
LATERAL AND MAIN RIVER SYSTEMS IN THE MOUNTAIN-
BASIN REGION
The great mounfain-hasin valley. — The contour lines
of the basal Eocene and lower Eocene sediments of
the mountain-basin region in northern New Mexico
and Montana are very illuminating. They show the
presence of a series of broad, relatively level basins — ■
a chain of flat uplands or valleys — in which the prod-
UINTA BAS
Period of volcanic
^ dust eruption in
^ soutlnern Wyoming,
Q> Utah. and Colorado,
sandstones washakie basinQ, mingled with erosion
'^ products.
Chiefly dacite
> tuffs and
sandstones
Figure 39. — Chronologic relations of formations in the mountain-basin region
This diagram exhibits the overlapping of sediments and the falls of volcanic ash in eight widely separated areas (Nos,
2-5 and 7-10, flg. 35), which, when combined, cover the entire Eocene epoch.
ucts of erosion and the volcanic dust that were gath-
ered by streams from the surrounding mountains were
spread wide, indicating that although the mountain
streams had high gradients and great erosive power
the larger rivers had low gradients and little trans-
porting power. The uniform elevation of the moun-
tain-basin region at the north and the south and the
low river gradients were favorable to sedimentation.
We observe, moreover, that in basal Eocene time the
conditions of climate and of sedimentation were some-
what uniform in the Puerco and Torrejon deposits
of the San Juan Basin in New Mexico, laid down by
tributaries of Colorado River, and in the typical Fort
Union deposits of Montana, laid down by tributaries
of Missouri River. The rates of sedimentation were
different. Similar basal Eocene sediments probably
underlie some of the Wasatch (lower Eocene) deposits
in the intermediate basins of
Wyoming and Utah, for they
have been exposed in the San
Juan Basin only by the removal
of the overlying Wasatch. A
new sedimentary phase was
begun in Wasatch time, and a
third phase in Bridger time.
The contrast in the physio-
graphic conditions east and
west of the Front Range has
a very important bearing upon
the paleontologic records. The
mountain-basin sediments af-
ford a marvelous and almost
unbroken record of mammalian
evolution in the Eocene, but
little or nothing in the Oligo-
cene, doubtless because large
areas of Oligocene sediments
have been eroded away. Only
two spots remain — Bates Hole
and Beaver Divide, in Wyo-
ming.
Piedmont, flood-plain, and la-
custrine deposits. — King led the
earlier geologists in presenting
the theory that the mountain
basins were once filled with a
chain of lakes. This theory was
adopted by Marsh, Cope, Scott,
and Osborn. Leidy, as early as
1869, cast doubt upon the lake
theory as applied to the White
River group east of the moun-
tains. The lake theory has grad-
ually been replaced by the flood-
plain theory through the studies
of Haworth (1897. 1), Gilbert
(1896.1), Matthew (1899.2),
Davis (1900.1), Johnson (1901.1), and Hatcher (1902.3).
For the highly diversified mountain-basin region
throughout the very long period of the Eocene, with
its considerable climatic vicissitudes, no single theory of
deposition is adequate. We have seen that in the
basal Eocene, during Fort Union, Puerco, and Torrejon
(Thanetian) time, there were doubtless great level
areas, heavily forested, with dense undergrowth, favor-
able to the formation of peat and lignitic deposits
Period of mountain
erosion of granitic,
calcareous, rhyolitic
early volcanic and
sedimentary areas
ENVIRONMENT OP THE TITANOTHEEES
55
and subject to heavy silting of fine sediments from
annual floods. These were like the flooded areas of
the forest belt in the Amazon delta. Such still-water
areas were contemporaneous with areas in the pied-
mont regions close to the mountains, where stream
erosion was active. The conditions that prevailed in
general during Wasatch (Sparnacian) time are nearly-
paralleled by those now found in the flood plains of
Parana, Paraguay, and Uruguay Eivers, which are
carrying down vast masses of gravel, sand, and clay from
the mountain chains of Brazil, as reported by John Ball
in his " Notes of a naturalist in South
America "(1887.1). The annual rain-
fall in these mountains ranges from 100
to 136 inches, and it rapidly disinte-
grates the yieldingrocks and discharges
a vast quantity of detrital matter over
the broad plains of Argentina and
Uruguay. The mountain streams
have thus built up wide, level areas
in these countries, and the lower
rivers, ploughing their channels
through the vast deposits over which
they must make their way, extend
their banks with every increment
and thus continually make additions
to the outskirts of the formation
they are depositing. In this way
deposits covering an area of 200,000
square miles have been formed from
the mountains of Brazil.
The period of flood-plain and
piedmont deposition in the Rockies
was followed by the great lacustrine
period of Green River time and of
Wind River (Ypresian) time, in which
the climate was much warmer. In
the same region there ensued the
flood-plain period of the Bridger.
Eocene basin deposition of another
kind and climatic change are indi-
cated in the widespread horizontal
' ' white layers "that divide the Bridger
into several geologic and faunistic
levels. These white layers indicate
periods of lagoon leveling by annual
uniform flooding and evaporation, similar to that of the
existing playa lakes of the Great Basin in Nevada.
In middle Eocene time new conditions of foresta-
tion and erosion and the presence of volcanic atmo-
spheric dust in the Bridger and Washakie Basins are
indicated. Sinclair showed (1906-1909) that the
Bridger formation was composed chiefly of volcanic
material that has been more or less rearranged by
stream action, and that clouds of volcanic dust
doubtless filled the atmosphere during the Bridger
epoch (middle and upper (?) Eocene). This interest-
ing discovery was confirmed by thorough analyses
made by Johannsen in 1914. The rocks of the upper
and middle Eocene formations consist chiefly of
volcanic tuff. Although the minerals of this tufl' are
those of a dacite (quartz andesite), the quartz grains
may be of sedimentary origin and the volcanic rock
may be andesite (Johannsen, 1914.1, p. 210). The
presence of dacite tuffs in the lower Bridger levels (B
and C) indicates that the atmosphere was charged
with volcanic dust, which also settled upon the con-
temporaneous deposits of the Washakie Basin, 100
miles to the east, as well as on the Uinta Range, 60
.25, 26a, 26 b
STACK MT^
CO >
< / U
LOWER BROV
FiGURK 40. — Section of deposits near Barrel Springs, Washakie Basin, southern
Wyoming (No. 9, fig. 35)
Showing alternating beds of tuff, siliceous and calcareous deposits, and sandstone. Johannsen (1914.1), after
Granger, with modifications. The numbers refer to lithologic specimens examined by Johannsen.
miles to the south. Thus during middle and upper
Eocene time the atmosphere over the present Bridger,
Washakie, and Uinta region was at times charged with
volcanic dust. Specimens of lower and basal Eocene
rocks indicate sediments of more normal type, and
whatever volcanic material they contain is so much
altered by re-sorting and mixing with normal sedi-
ments that it is not clearly recognizable.
The manner in which the layers of dacite and glass
tuffs alternate -with the heavy river-channel sand-
stones is clearly displayed in the analysis of sediments
from the Washakie Basin by Johannsen. Tuffs are
56
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
also scattered, but more sparingly, through the earlier
Torrejon, Wasatch, and Wind River formations, along
with river-borne material derived from the decay and
erosion of older rocks.
SECTION 2. EOCENE AND LOWER OLIGOCENE
FORMATIONS AND FAUNAL ZONES
FIRST FAUUAI PHASE (BASAL EOCENE)
SEVENTEEN LIFE ZONES
Largely as the result of explorations and researches
made for this monograph, the major Eocene and Oligo-
cene type life zones that were recognized by Leidy,
Cope, and Marsh up to the year 1900, such as the
" Coryphodon beds," " JJintaiherium beds, "Diplacodon
beds," and " TitanotJierium beds," have gradually been
differentiated, through the work of Osborn, Granger, and
Matthew, into 16 known life zones, each distinguished
by the presence of a highly varied mammal fauna and
by the appearance or disappearance of certain groups
of mammals and reptiles. There is also one theoretic
life zone, between known upper Eocene and known
lower Oligocene time, making 17 in all. Each of these
life zones corresponds with a series of sediments rang-
ing in thickness from 300 to 600 feet. Many of them
correspond with changes in climate, temperature, and
forestation, and some of them are clearly defined and
sharply demarcated from others. A single generic
name, such as Coryphodon, rarely suffices to distin-
guish them, because many genera and even certain
species may survive for long periods of time.
Each of these faunal zones is defined paleontologic-
ally by one or more of the life forms it contains, geo-
logically by the locality v/here it is best preserved, to
some extent botanically by the flora it contains, and
lithologically by the character of its rocks as shown
by microscopic analyses. Thus, for example, we have
the typical upper Wind River zone — the "Lost Cabin "
or LamhdotJierium-Eotitanops-CorypJiodon zone, a rather
cumbersome designation, which indicates that only
during this period did these three types of mammals
exist together. In this zone Lambdotherium is the
most distinctive genus. Sediments in different geo-
graphic basins are correlated in such a manner as to
present the whole life story of the Eocene epoch, as
shown in the accompanying diagram. Of the two
chief faunistic gaps that have been recognized, that
between the Wind River and the Bridger has now been
filled by explorations of the Huerfano, so that there
remains only that between the Uinta and the White
River. These 16 known life zones will doubtless be
multiplied to 20 or more by future discovery. They
are shown in the following table:
ENVIfiONMENT OF THE TITANOTHEEES
Synopsis of life zones
57
Epoch
Life zones
Horizon
Characteristic species and genera
17. Titanotherium-Mesohippus.
Brontops robustus, Menodus gigan-
Lower Oligocene.
Chadron B
teus, Brontotherium platyceras,
Allops crassicornis.
Brontops dispar, Menodus trigono-
ceras, Allops marshi, Brontothe-
rium hatcheri.
Brontops braohycephalus, Menodus
heloceras, Brontotherium leidyi.
16. Theoretic zone (no fauna)
15. Diplacodon-Protitanotherium-
Epihippus.
14. Eobasileus-Dolichorhinus
13. Metarhinus
Uinta C 2.
Uinta CI
Diplacodon, Protitanotherium, Epi-
Upper Eocene.
Uinta B 2 and Washakie
B 2.
Uinta B 1 and Washakie
B 1.
hippus, Protoreodon.
Eobasileus, Dolichorhinus.
Metarhinus, Amynodon.
12. Uintatherium-M a n t e 0 c eras -
Mesatirhinus.
11. Palaeosyops paludosus-Orohip-
pus.
10. Eometarhinus-Trogosus-Palaeo-
syops fontinalis.
Washakie A and Bridger
C and D.
Uintatherium, Manteoceras, Mesati-
rhinus.
Palaeosyops paludosus, Orohippus.
Bridger A and Huerfano
B.
Palaeosyops fontinalis, Eometarhinus.
Lower Eocene.
9. Lambdotheri um-Eotitanops-
Coryphodon.
8. Heptodon-Coryphodon-Eohippus.
7. Systemodon-Coryphodon-Eohip-
pus.
6. Eohippus-Coryphodon.
Huerfano A, Wind River
B, and Big Horn E.
Big Horn (Wasatch) D
and Wind River A.
Big Horn (Wasatch) C__.
Big Horn (Wasatch) B-._
Lambdotheri um, Eotitanops, Cory-
phodon, Meniscotherium.
Heptodon, Eohippus, Coryphodon.
Systemodon, Eohippus, Coryphodon.
Eohippus, Pelycodus, Coryphodon.
Transition basal Eo-
cene to lower Eocene.
5. Phenacodus-Nothodectes-Cory-
phodon.
Big Horn (Wasatch) A___
Phenacodus, Nothodectes, Corypho-
don, Champsosaurus.
Pantolambda, Tetraclaenodon, Claen-
odon.
Torrejon A._ . . .- --
Deltatherium, Mioclaenus, Haplo-
Basal Eocene.
conus.
Polymastodon, Oxyclaenus.
Puerco A._ _
Ectoconus, Champsosaurus.
Cretaceous."
Triceratops-Tyrannosaurus
Lance and Denver forma-
tions.
Triceratops, Tyrannosaurus, Champ-
sosaurus, Meniscoessus.
Judith River and Belly
River formations.
Monoclonius, Deinodon, Eodelphis.
» The United States Geological Survey classifies the Denver formation as Eocene and the Lance formation as Tertiary (?) .
101959— 29— VOL 1 6
58
TITA^TOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Lower Tertiary geologic horizons and life zones and their hoofed mammals
Epoch
Geologic horizon
Chiet life zones. (Titanotheres, horses, and
other mammals.)
Chief titanotheres and other perissodactyls
Oligocene.
Leptauchenia, Miohippus, and
Oreodon.
Extinction of titanotheres
Chadron A, B, and C.
17. Titanotherium-Mesohippus.
Giant titanotheres — Menodus,
Brontops, Brontotherium, etc.
Upper Eocene.
Uinta C.
16. Theoretic zone (fauna un-
. known).
15. Diplacodon-Protitan othe-
rium-Epihippus.
Protitanotherium, early horned ti-
tanotheres.
Washakie B and Uinta B.
14. Eobasileus-Dolichorhinus.
13. Metarhinus.
Dolichorhinus cornutus, Mesati-
rhinus.
Bridger C and D and Washakie A.
12. Uintatherium-Manteoceras-
Mesatirhinus.
Manteoceras, ancestors of Ohgo-
cene titanotheres.
Middle Eocene.
Bridger B.
11. Palaeosyops paludosus-Oro-
hippus.
Palaeosyops and numerous other
titanotheres.
Bridger A and Huerfano B.
10. Eometarhinus-Trog osus-
Palaeosyops fontinalis.
Palaeosyops fontinalis (primitive).
Big Horn E ("Lost Cabin"), Wind
River B (typical Wind River of
Cope), and Huerfano A.
9. Lambdotherium-Eotitanops-
Coryphodon.
Appearance of titanotheres
Lower Eocene.
Big Horn D ("Lysite") and Wind
River A.
8. Heptodon-Coryphodon-Eo-
hippus.
Big Horn C ("Gray BuU").
7. S3'stemodon-Coryphodon-
Eohippus.
Earliest tapiroids, Tapiridae.
Big Horn B ("Sand Coulee").
6. Eohippus-Coryphodon.
Earliest Equidae (horses).
Transition.
Big Horn A ("Clark Fork") of
Wyoming and " Tiffany beds " of
southwestern Colorado.
5. Phenacodus-Nothodectes-
Coryphodon.
Earliest Phenacodus (condylarths) .
Earliest Coryphodon. Notho-
dectes, similar to Pleisiadapis of
Cernay.
Upper horizon of Torrejon forma-
tion.
4. Pantolambda.
Ancestors of the Amblypoda.
Basal Eocene.
Lower horizon of Torrejon forma-
tion.
3. Deltatlierium.
^
Upper horizon of Puerco forma-
tion.
2. Polymastodon.
M ultitub erculata.
Lower horizon of Puerco forma-
tion.
1. Ectoconus.
Earliest known Taligrada.
ENVIEONMENT OP THE TITANOTHEEES
59
Figure 41. — Eocene and lower Oligocene mammalian life zones in eleven typical correlated areas in New
Mexico, Colorado, Utah, Wyoming, South Dakota, and Montana, located as shown on the general geologic
map (fig. 35)
Arranged by Osborn (1919) after original studies made in the field, chiefly by Granger, but also by Hatcher (Oligocene), Hills, Peterson, and
Gidley (Eocene) . The 16 Icnown life zones numbered 1 to 15 and 17 are indicated in the diagram by darlc horizontal lines. The nonfossiliferous
areas are indicated by light oblique lines. These life zones and sections also correspond with the detailed geologic sections in this chapter.
The United States Geolojical Survey classifies the Lance formation as Tertiary (?), Eocene (?). The author of this monograph regards it
as Cretaceous.
60
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
BASAL EOCENE TIME IN MONTANA AND NEW MEXICO
Fori Union formation of Montana. — The typical Fort
Union formation of Hay den (Meek and Hayden, 1862.1,
p. 433), at the junction of Yellowstone and Missouri
Rivers, lies east of the center of an ancient forested
swamp in which was laid down the upper part of
Hayden's "Great Lignite Group." One of the most
interesting results of discoveries made in 1901 (Doug-
lass, 1902.1) is revealed in an exposure of the Fort
Union in Sweet Grass County, Mont., near the head-
waters of the Musselshell, containing a rich fauna of
the archaic species of basal Eocene animals, some of
which are identical with those found on the head-
waters of San Juan River, in northern New Mexico,
a thousand miles to the south. Both lie near the
one hundred and seventh meridian. The presence in
large numbers of animals belonging to similar species
shows that uniform climatic and physiographic con-
ditions existed in this great mountain-basin region
for a very long time, because similar generic fonns of
life (Olaenodon, Pantolambda) persist through 3,000
feet of Fort Union sediments. The remains of the
oldest of these mammals are found immediately above
the dinosaur-bearing beds at a level which is here
identical with that of the Lance formation; and the
present opinion is that sedimentation may have been
continuous throughout Upper Cretaceous and basal
Eocene time in this region in Montana.
The mode in which these Fort Union beds were
formed has not yet been positively determined, but
the masses of fresh-water shells which they contain in
certain localities indicate that they were in part laid
down in shallow lagoons and swamps, which were in
some places crossed by river channels. At some
places the beds contain multitudes of leaves, which
give us a complete record of the forest life of the time.
Vast areas of warm-temperate and more hardy trees
were interspersed with areas where swamp vegetation
accumulated rapidly enough to form great beds of
lignite. Amid the glades of these forests there wan-
dered swamp turtles, alligators, and especially the
choristoderan reptiles of the characteristic genus
Champsosaurus.
Puerco and Torrejon formations of New Mexico. —
A southern center of this archaic mammal life is the
type locality of the Puerco formation, on the divide
between the Rio Grande and the San Juan, in north-
western New Mexico, a formation described by Cope
(1875.1) as the "Puerco marls." Cope listed the
first mammalian fauna from those beds in 1881,
opening a new epoch in mammalian paleontology. In
1885 he assigned to the formation a thickness of 850
feet and distinguished it from the underlying beds,
which he supposed to be Laramie but which have
since been divided into the Qjo Alamo sandstone, the
Kirtland shale, and the Fruitland formation, all
probably of Montana age, older than Laramie. The
Puerco of Cope appears to be a single formation geo-
logically, deposited with apparent conformity between
the upper and lower divisions, but it is sharply divided
faunistically into two main life zones, a lower, which
retains the name Puerco, and an upper, to which the
name Torrejon was given by Wortman in 1895
(Osborn and Earle, 1895.95, pp. 1-3A). In 1910
Gardner (1910.1) applied the name Nacimiento
group to both divisions. In 1897 Matthew (1897.2)
separated the mammal fauna of the two levels, and
in 1912 and 1913 Sinclair and Granger (1914.1) estab-
lished in this group no less than four faunistic levels,
which are shown in the accompanying section (fig.
43). Two faunistic levels were observed by Wortman
in the Puerco, and two distinct faunistic levels are dis-
tinguished by Granger, Sinclair, and Matthew in the
Torrejon.
These four successive changes in the archaic fauna
occurred during a period of continuous sedimentation,
for no unconformity has been observed between the
Puerco and Torrejon. The rate of deposition of the
800 feet of Puerco and Torrejon sediments was rela-
tively slow as compared with that of the deposition
of the 6,000 feet of the corresponding Fort Union sedi-
ments to the north. As the mammals distributed
through 4,000 feet of the northern part of the Fort
Union deposits correspond chiefly with those of the
Torrejon, it appears possible that the underlying
Puerco fauna may belong in part in upper Lance time.
We observe that the Fort Union was deposited upon
the Lance continuously, without recognized notable
unconformity, whereas the Puerco lies upon the eroded
surface of the Ojo Alamo, which, because of its
dinosaur fauna, is considered of probable Judith River
and Belly River age.
The close resemblance of the crestless trachodont
dinosaur, Kritosaurus navajovius, from the Ojo Alamo,
to a corresponding form from the Belly River forma-
tion of Alberta also suggests a close correlation in
time.*
In 1912 and 1913 Sinclair and Granger thoroughly
explored the basal Eocene deposits of the San Juan
Basin, with the results enumerated above.
SUMMARY OF FAUNAE EVENTS OF BASAL EOCENE TIME
In addition to the four fossiliferous zones observed
tn the Puerco and Torrejon formations, all distinc-
tively basal Eocene, there is an overlying zone in the
"Tiffany beds," beyond the border of Colorado, deter-
mined by Gidley (1909) and Granger (1916). These
beds contain a fifth fauna, which is strictly interme-
diate between basal Eocene and lower Eocene. This
transitional basal-lower Eocene zone is described on
pages 64-65. The basal Eocene mammalian life
> See Parks, W. A., The osteology of the trachodont dinosaur Kritosaurus incur-
limanus: Univ. Toronto Studies, Qeol. series, 1920.
ENVIRONMENT OF THE TITANOTHERES
61
Figure 42. — Section of Upper Cretaceous and basal Eocene (Fort Union) deposits in
Sweet Grass County, Mont.
After Stanton (1909.1), Stone and Calvert (1910.1), and Gidley (1919). This very significant exposure (No. I, flg.35) is in
an outlying area of the Fort Union formation and its mammal fauna corresponds broadly with that of the Torrejon
formation of northwestern New Me-xico, although the section has not yet been divided into separate life zones.
It affords the most satisfactory means of correlating the Fort Union and Puerco and Torrejon formations. The
United States Qeological Survey classifies the Lance formation as Tertiary(?), but the author of this monograph re-
gards It as of UpperTCretaceous age.
62
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 43. — Section of Eocene deposits in the San Juan Basin, northwestern New Mexico (No. 2,
fig. 35), showing the base of the Puerco formation resting upon the eroded surface of the Ojo
Alamo sandstone, as observed by Sinclair and Granger (1914.1)
This section displays the close geologic continuity of the Puerco and Torrejon beds, which are subdivided faunistically into four
distinct life zones, named, in ascending order, (1) Ecloconus and (2) Poly mastodon zones, Puerco formation; (3) Deltatherium and
(4) Fantolambda zones, Torrejon formation. The Ojo Alamo sandstone is perhaps of Judith River age.
ENVIRONMENT OF THE TITANOTHEEES
63
of the Puerco and Torrejon formations in northwestern
New Mexico and southwestern Colorado is accord-
ingly related as follows:
Transition epoch:
5. Phenacodus-Nothodectes-Conjphodon zone. Represented
in the "Tiffany beds" of southwestern Colorado and
in the Wasatch formation (horizon Big Horn A,
"Clark Fork") of Big Horn Basin, Wyo.
Basal Eocene epoch:
4. Pantolamhda zone. Represented in the upper levels of
the Torrejon formation of northwestern New Mexico
and in the upper part of the Fort Union formation of
Montana.
3. Deliatherium zone. Represented in the basal part of
the Torrejon formation and in part of the Fort
Union formation of Montana.
2. Polymastodon zone. Represented in the upper part of
the Puerco formation of northwestern New Mexico.
Not yet recorded in the Fort Union formation.
1. Ectoconus zone. Represented in the lower part of the
Puerco formation. Not yet recorded in the Fort
Union formation.
BASAL EOCENE FAUNAL ZONES
ZONES 1 AND 2: ECTOCONUS AND POLYMASTODON ZONES
[Puerco fauna; part of Thanetian of Europe]
No equivalent of the most ancient Puerco fauna has
thus far been discovered in the Fort Union beds of
the North or in Europe; it is at present unique.
Puerco mammals and reptiles. — The Puerco mammals
are extremely archaic, mostly Meseutheria (Osborn)
or paleoplacentals (Matthew), representing groups of
placentals that became extinct during the Eocene.
The Puerco contains no remains of modern orders or
families of mammals except three, one (Miacidae)
which is related to the doglike Carnivora, a second
which is related to the primitive Insectivora, and a
third which is related to the primitive Edentata. No
rodents or lemuroid primates have been discovered,
and certainly no perissodactyl or artiodactyl ungu-
lates were in this region at this time. Matthew
(1914.1, p. 383) is of the opinion that most of these
archaic placentals have "no known predecessors in the
Lance formation.
About 10 per cent of the fauna consists of rodent-
like multituberculates, an extremely ancient order re-
lated to the existing monotremes or to the marsupials.
These animals are nearly related to ancestral forms in'
the Lance. Didelphiid marsupials are also present.
Similarly the reptiles all belong to families that
originated in Belly River or Pierre time (Upper
Cretaceous) or earlier. The Choristodera (Champso-
saurus) became extinct in basal Eocene time. Note-
worthy is the absence of the prevailing Tertiary
families of chelonians (Emydidae, Testudinidae),
which, with the modernized mammals, first appear in
the lower Eocene.
On comparing the life of the Puerco with that of
the Lance we find a mammalian fauna that indicates
no very wide gap in time — a fauna that is somewhat
more ancient than the Torrejon and known Fort
Union, also more ancient than the Cernaysian and
upper Thanetian of France. It is therefore probable
that the Puerco corresponds with the lower Thanetian
of France, but its life has no known equivalent either
in Europe or in this country.
The opinion of Cope that the ancestry of the
modernized mammals should be sought among these
Puerco forms lacks adequate confirmation. The op-
posite opinion — that the Puerco mammals are not
ancestral to the modern mammals — was developed by
Osborn (1893.82, 1894.89) when he applied to them
the name Mesoplacentalia (Meseutheria), indicative of
their archaic or Mesozoic characteristics. They repre-
sent the first known adaptive radiation of the placen-
tals into archaic flesh eaters and herbivores. We note
the presence of three families of archaic Carnivora
(Creodonta) and remote relatives {Psittacotherium) of
the Edentata. Among the archaic ungulates we find
one varied family (Periptychidae) of the Amblypoda
(Taligrada) and two families (Phenacodontidae, Mio-
claenidae) of the Condylarthra.
Puerco sedimentation and physiography . — The Puerco
formation is not separated from the overlying Torrejon
formation by any lithologic or stratigraphic break.
(Sinclair and Granger, 1914.1, p. 308.) The absence
of erosional unconformity between the Puerco and
Torrejon was also observed by Gardner (1910.1, pp.
722-723) and by Bauer (1916.1, p. 277). That the
Puerco and Torrejon formations represent a very
long period of geologic time is demonstrated by the
recorded 6,000 feet of Fort Union sediments, which
have yielded the Torrejon fauna alone; and, like the
Fort Union, they represent a very long period of uniform
conditions of climate and sedimentation. The pres-
ence of fish, crocodiles, turtles (Trionyx), and other
genera in the same strata with the bones of mammals
and at the same level shows conclusively that these
deposits were formed by water. That the streams
were of low gradient is shown by the complete absence
of pebbles in the Puerco and by the wide horizontal
extent of some of the clay bands. Bogs, apparently
formed in back waters in the channels, were filled with
accumulations that preserved impressions of the leaves
of figs (Kc-Ms), plane trees (Platanus), poplars {Populus),
relatives of the bread fruit (Artocarpus) , and numerous
shrubs (Paliurus, Viburnum). The quantity and vari-
ety of these plant remains, together with the abundant
large drift logs in the clays, indicate a heavy growth
of vegetation along the streams. The species of Ficus.
Paliurus, Viburnum, and Artocarpus are also found in
the Denver and Raton formations of eastern Colorado ;
and other species indicate Fort Union age (Knowlton,
cited by Sinclair and Granger, 1914.1, p. 306). The
mode of occurrence of the fossils in the still-water
clays and occasionally in the river-channel sandstones
shows that some of the skeletons may have been
64
TITANOTHEEES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
washed into the streams during heavy rains and
scattered by the action of crocodiles, carnivores, tur-
tles, and fish. Other skeletons show traces of gnaw-
ing, probably by small Ptilodontidae, which proves
that many of the bones lay for some time on the surface
of the ground before reaching the streams or being
covered in flood time by water-borne sediments.
ZONES 3 AND 4: DEITATHERIUM AND PANTOLAMBDA ZONES
[Torrejon and Fort Union faunas; in part Thanetian of Europe]
The mammals of the Torrejon formation of north-
western New Mexico, whose remains are found in a
stratum about 250 feet above the base of the Puerco
mammal-bearing level, are somewhat larger, con-
siderably more diversified (perhaps because more fully
known), and of slightly more progressive type. They
show very close affinity to the Fort Union mammals of
Montana and some affinity to the Cernaysian forms
discovered in the conglomerat de Cernay, near Rheims,
France.
The multituberculates, which occur in the Holarctic
region in upper Triassic (Rhaetic) time, now make
their last appearance abundantly; of the Ptilodonti-
dae, Ptilodus (or Neoplagiaulax) is found in New
Mexico, Montana, and Cernay; the large Polymasto-
don that distinguishes the upper Puerco zone does
not recur.
Here also are five families of archaic carnivores
(Creodonta), among which, in the Miacidae, there is a
genus (Didymidis) which appears to lead through the
civet and doglike forms of the lower and middle
Eocene into forms related to the modern Carnivora.
Among the three Torrejon families of Insectivora the
existing Centetidae (tenrecs) are possibly related to
the genus Palaeorydes , a very primitive form resem-
bling the modern Cape golden moles {(JhrysocMoris of
South Africa, Necrolestes of South America). The
ancestors of the modern edentates are highly diversi-
fied (Edentata, Ganodonta) and include slothlike
animals, indicative of present or former migrations
into South America. Of the families of archaic ungu-
lates two (Phenacodontidae and Mioclaenidae) repre-
sent the Condylarthra, and two (Periptychidae and
Pantolambdidae) represent the Amblypoda. Of the
Amblypoda Pantolambda cavirictus, which is also
found in the Fort Union, is very characteristic. Of
the bearlike Creodonta (Arctocyonidae) Claenodon
ferox, which is closely related to the Arctocyon of the
Thanetian of France, occurs also in the Fort Union
of Montana.
Most of these mammals of the Torrejon, like those
of the Puerco, were ancient adaptive radiations of
the Mammalia. They were small-brained, had de-
fective foot structure, and were unfitted to compete
with the ancestors of the modernized mammals,
which begin to appear immediately above the Noiho-
dedes zone. Six families approached extinction at the
end of the Torrejon — the Plagiaulacidae of the Multi-
tuberculata; the Oxyclaenidae of the Carnivora; the
Conoryctidae of the Edentata; the Periptychidae and
Pantolambdidae of the Ambyploda (which, however,
are related to the succeeding coryphodons) ; and the
Mioclaenidae of the Condylarthra. The Plagiaula-
cidae and Oxyclaenidae, however, survive into the
early Wasatch, the Periptychidae into the "Tiffany
beds." Torrejon time thus ends with the extinction
of a large number of families of archaic mammals,
though several families survived, passing into the
succeeding lower Eocene.
Unconformities of the Torrejon with the underlying
Puerco have not been found. (Sinclair and Granger,
1914.1, p. 312; also Gardner, 1910.1, p. 722, and Bauer,
1916.1, pp. 273-277.) There is no doubt about the
aqueous origin of either the Puerco or the Torrejon
deposits. The Torrejon carries less petrified wood
than the Puerco, but it contains Z7mo-bearing beds,
which occur repeatedly in the gray clays, and abundant
shells of land moUusks {Pupa), which are found in the
clays that contain bones of mammals. Lithologically,
the Torrejon closely resembles the Puerco, except that
gravels of quartzite, jasper, red shale, etc., occur in
some of the channel sandstones. Mammals appear
principally in the zones filled with small rusty calca-
reous concretions, which occur in clays that range in
color from red mottled with green to gray. The upper
boundary of the Torrejon is everywhere marked by
the presence of Tetradaenodon (ancestor of PJiena-
codus) and of the two amblypods PeriptycTius rhabdo-
don and Pantolambda. The total thickness of the
Torrejon differs at different places, ranging from 240
to 660 feet, whereas the approximately contempora-
neous Fort Union of Montana, which possibly also
represents the Puerco, attains a thickness recorded as
nearly 6,000 feet.
The top of the Tori'ejon is in unconformable contact
with sandstone that indicates a cycle of deposition of
coarse sediments and alluvial fans, attributed to
Wasatch time.
SECOND FAUNAI PHASE (LOWER EOCENE)
TRANSITIONAL BASAL EOCENE FAUNAS
ZONE 5; PHENACODUS-NOTHODECTES-COETPHODON ZONE
[Base of Wasatch formation of Big Horn Basin, first Wasatch life zone, Big Horn
A; Cernaysian of Europe]
The first Wasatch life zone is represented in the
"Tiffany beds" of southwestern Colorado, in the basal
part of the Wasatch formation (horizon Big Horn
A= "Clark Fork") of the Big Horn Basin, Wyo., and
probably in the summit of the Fort Union formation
of Montana. In southwestern Colorado, near the
headwaters of the San Juan, are the "Tiffany beds"
of Granger, which contain a fauna characterized by
the last appearance of PeriptycTius and by the first
appearance of Phenacodus and of Coryphodon, a genus
ENVIKONMENT OF THE TITANOTHEEES
65
characteristic of the Sparnacian of France. Notho-
dectes of the "Tiffany beds" is particularly interest-
ing because of its structural affinity to Plesiadapis
of the Cernaysian of France. The multituberculates
are represented in Wyoming by Ptilodus (" Sand
Coulee" and "Clark Fork"?). Of the four specimens
of ptilodontids from Wyoming, one found by Granger
was from the Big Horn B horizon ("Sand Coulee
beds"). Three found by Stein were probably from
the same horizon but may have been from the under-
lying Big Horn A horizon (the "Clark Fork beds").
Undoubtedly ptilodontids occur in the "Clark Fork,"
but we can not furnish any positive evidence (W.
Granger, 1919).
This mammal fauna as a whole actually resembles
that of the Torrejon more closely than that of the
lowest overljang horizon (Big Horn B, "Sand Coulee")
of the Wasatch. A significant discovery in the No-
thodectes zone is Zanyderis, a bat showing affinities
with the vampires (Phyllostomatidae) of South
America.
The NotTiodedes zone ("Tiffany" and "Clark Fork")
is basal Eocene, as indicated by the absence of the
four orders Primates, Perissodactyla, Artiodactyla,
Rodentia; it is lower Eocene, as indicated by the
presence of Phenacodus and Coryphodon.
The mammalian life of the "Clark Fork" beds in
the Big Horn Basin of Wyoming, to the north, is very
similar (Granger, 1914.1, p. 204) to that of the "Tif-
fany beds" in Colorado. These "Clark Fork beds,"
500 feet in thickness, are characterized by the pre-
dominance of the Condylarthra (PJienacodus and Ec-
tocion), remains of which constitute three-fourths of
the fossils collected from them. The amblypod un-
gulates are represented by CorypJiodon and by the
first appearance of an animal (Eohathyopsis) ancestral
to Baihyopsis, of the Wind River formation, which in
turn is ancestral to the horned UintatJierium of the
Bridger formation. Among the Reptilia is the last
surviving Champsosaurus from the Fort Union and
the Cretaceous, a distinctively basal Eocene type.
EARLY EOCENE TIME
General correlation. — Lower Eocene (Wasatch) time
began, it may be said, with the first appearance of
Coryphodon and Phenacodus in the "Clark Fork" and
"Tiffany beds" described above as the Phenacodus-
Nothodectes- Coryphodon zone, in which is found the first
phase of the Coryphodon fauna. The modernization
occurred later, in the "Sand Coulee beds" (Eohippus
zone), which overlie the "Clark Fork."
The Sparnacian of Europe is broadly parallel with
part of the Wasatch formation {Coryphodon zone) of
America. It is typified in France by the deposits of
Soissons, Meudon, and Vaugirard; in England by the
Woolwich beds, which contain a rich flora. In these
fluviomarine, lagoon, and lacustrine deposits of Europe
mammals are rare, and homotaxis with America is
afforded through the large coryphodons, the perisso-
dactyl Lophiodon, and the creodonts Palaeonictis and
Pachyaena. This sparse European fauna, which in its
early stages lacks Equidae (Hyracotherium) , has almost
a counterpart in that of the Nothodectes zone of the
Rocky Mountain region.
The two upper zones of the lower Eocene (Wasatch)
are correlated with the Ypresian of Europe.
Wasatch and Sparnacian floras. — According to Berry
(1914.1, p. 148) the earliest Eocene beds of Europe
(Sparnacian and Ypresian stages) contain the flora
found in the Oldhaven, Woolwich, and Red beds of
England, largely unstudied, and the small flora from
the Paris Basin recently described. The Woolwich
beds have yielded the fig (Ficus), the locust (JRohinia),
the tulip tree (Liriodendron) , and Grevillea, a pro-
teaceous plant now confined to Australia. Berry
believes (letter to the author, April 1, 1918) that in
lower Wasatch time the Fort Union flora persisted over
the Rocky Mountain basin region. This belief implies
that the climate was then prevailingly warm-temperate
but that there were occasional incursions of trees of
subtropical type.
Sedimentation during Wasatch time. — As the Sparna-
cian stage of Europe, which is equivalent to part of the
Wasatch, derives its name from Epernay (Latin Spar-
nacum), so the Wasatch stage of mammalian life
derives its name from the typical Wasatch group of
Hayden in western Wyoming, a single mammal-bearing
member of which is the Knight formation (Veatch,
1907.1), 1,750 feet in thickness, containing Cope's
types of Eohippus index, E. vasacciensis, Phenacodus
primaevus, Coryphodon radians, C. semicinctus, C.
latipes. These species of mammals do not represent
the oldest Wasatch fauna; they are of the same age as
the species found at the "Lysite" horizon (life zone
No. 8) of the Big Horn Basin.
Among the chief sources of Wasatch mammals are
the following:
Feet
1. Knight formation (Veatch), top of typical Wasatch
group (Hayden), southwestern Wyoming; red and
yellow sandy clays 1,750
2. Wasatch formation, Big Horn Basin, Wye; red,
brown, and gray sandstones and clays 2, 025
3. Wasatch formation ("Bitter Creek" of Powell and
"Vermilion Creek" of King), Washakie Basin,
Wyo.; red and gray clays and sandstones 4, 000-5, 500
4. Wasatch formation of the San Juan Basin, N.
Mex 1,500
5. Wasatch formation of the Uinta Basin, Utah (White,
1878) 2,000
6. Wasatch formation of the Powder River Basin, Pump-
kin Buttes, Wyo 2,400
The estimate made by King (1878.1) of the thick-
ness of the sediments in the Washakie Basin (4,000-
5,500 feet) is considered high (Granger). It is inter-
esting to note that the mean thickness (about 2,300
feet) of the Wasatch sediments in the six areas listed
above exceeds somewhat that of the Bridger formation
(1,875 feet).
66
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The earliest Wasatch sediments are those in the
Big Horn and Clark Fork Basins of northern Wyoming,
from which we obtain the whole range of lower Eocene
fossil mammals, beginning with the end of basal
Eocene time.
fossils. Douglass found a considerable faima in the
Wasatch of the Uinta Basin. Systemodon occurs
there. Wortman has reported (letters) a Coryphodon
from the Wasatch of the Washakie Basin. To the
south, in the San Juan Basin, there were laid
down, over the Torre j on, thick beds of sand
and fluviatile sediments, which form the New
Mexico Wasatch. These beds, which are
divided by Granger (1914.1) into an upper
("Largo') and a lower ("Almagre") divi-
sion, have a combined thickness of 1,500 feet,
throughout the greater part of which mam-
malian fossils are found. These Wasatch beds
in New Mexico have much the same general
appearance as the Wasatch in other localities,
consisting of red, gray, and ocherous bands
of shale and sandstone, without evidence of
unconformity throughout the series. The most
recently identified Wasatch sediments are
those of Pumpkin Buttes, in the Powder
River Basin, Wyo.
The correlation of the faunal horizons in
these sedimentary areas by the species of
mammals which they contain was determined
with remarkable precision by the American
Museum expedition under Granger, as shown
in the accompanying table (p. 67).
Wasatch pJiysiograpliic and climatic condi-
tions favorable to modernized fauna. — All the
Wasatch sediments indicate a profound and
somewhat abrupt change in the physiographic
and climatic conditions of the mountain-basin
region from those that prevailed during Fort
Union, Puerco, and Torrejon time. In general,
still-water sedimentation in level forests and
lagoons ceased. Fluviatile, flood-plain, fluvial-
fan, and channel deposits containing a larger
percentage of coarser materials were wide-
spread. There is evidence of open stretches
of country exposing sand, gravel, and clay,
subject to occasional desiccation and aridity.
The Wasatch of the Big Horn Basin repre-
sents the filling of an intermontane trough of
downwarp. (Sinclair and Granger, 1912.1,
p. 66.) Materials were transported by streams
Figure 44.— Columnar section of Cretaceous and Eocene sediments ex- f ^.^^ ^^le surrounding mountains, as shown by
posed -along Bear River near Evanston, in extreme southwestern ^^e lithology of the gravel, sand, and clay. The
Wyoming (No. 3, fig. 35) , sliowing tlie typical Wasatcli group of Hayden
(1869). Chiefly after A. C. Veatch (1907.1)
Mammals similar to those in zone 8 (Eohippus, Phe-nncodus, lieplodon, and Coryphodon) occur
in a narrow fossiliferous stratum which may be referred to the Heptodon- Coryphodon- Eohippus
zone. Above are Oreen River and Bridget beds; below are 4,600 feet of beds (without mammals)
belonging to the Wasatch group (Fowkes and Almy formations), which are imderlain by the
Evanston formation, containing Fort Union plants, and the AdaviUe formation, containing
Montana plants and invertebrates. The author of this monograph regards the Evanston
formation as uppermost Cretaceous.
Similar heavy and continuous sedimentation also
occurred during Wasatch time, in both the northern
and the southern Uinta region, in the Bridger and
Washakie Basins on the north, and in the great Uinta
Basin south of the mountains. Few of these vast
masses of sediment have thus far yielded mammalian
underlying Fort Union was uplifted before
sedimentation began, and the synclinal basin
was inclosed more or less completely to the east,
south, and west by anticlinal mountains.
Erosion from the mountain rocks represents
all the members of the typical section from
the Archean to the Fort Union, usually
by stream transportation and deposition in river
channels and over broad flood plains. No beds of
volcanic ash have been found, nor is there evidence
of transportation by wind. The deposits of clay show
a more or less regular alternation of red and bluish-
gray layers, which may be due to climatic changes.
ENVIRONMENT OF THE TITANOTHEKES
67
The excess of iron salts in the red clays may have
accumulated and oxidized to hematite during dry
climatic cycles; the blue clays were probably deposited
in a moister climate, which is less favorable to the
concentration and oxidation of the iron. Similar
alternations of red and blue clays in the desert basins
of Lop and of Sewistan have been described by Hunt-
ington, who also associates the colors with the recur-
rence of moist and arid climatic cycles. Sinclair and
Granger (1914.1) ascribe the color banding of the
Wasatch and Wind River clays to a similar cause — the
alternation of moist and dry climatic conditions — but
they have not found any other evidence of excessive
aridity, the fauna of the red and blue bands being
the same. The fact that the blue clays of the Wasatch
are here and there lignitic and are at some places
associated with skeletal remains suggests that they
may have been formed during cycles of rather abun-
dant rainfall, when the surface of the intermontane
basin was prevented from drying out rapidly. That
these climatic and physiographic conditions were not
local is shown by similar color banding in the Wasatch
of all the mountain-basin regions. The name "Ver-
milion Creek" was applied by King to the Wasatch
because of the red color of the rocks through which
that creek flows in southern Wyoming and north-
western Colorado.
Microscopic examination of the feldspars in the
Wasatch deposits of the Big Horn Basin does not favor
the idea of luxuriant subtropical forests and a warm,
humid climate, with the formation of a deeply decayed
humus, but rather suggests a dry, not necessarily arid
climate, with rapid changes of temperature, favorable
to splintering of the ledges of hard rock; rapid trans-
portation of the fragments for short distances; and
burial of these beyond reach of carbonated waters.
A cursorial ungulate fauna. — This conception of a
drier lower Eocene climate in the basins during
Wasatch time accords with the successive appearance
in this region of four families of the modernized types
of perissodactyl mammals — horses, tapirs, lophio-
donts, and titanotheres — with light, cursorial limb
and foot structure adapted to rapid locomotion and
wide seasonal migration.
Correlation of lower Eocene life zones of Wyoming and New Mexico {after Granger, with modifications)
New Mexico (Wasatch —
"Largo" and "Almagre")
'Largo" (typical).
Eohippus, Menis-
cotherium, Am-
Vjloctonus.
'Almagro" (typical).
Eohippus, Anaco-
don.
Unconformit3' be-
tween Wasatch and
Torrejon. In south-
ern Colorado "Tif-
fany" (typical) .
No perissodactyls.
Torrejon. No peris-
sodactyls. Fauna
more primitive
than in "Clark
Fork."
Evanston (typical Wasatch)
Green River.
Knight (typical) .
Heptodon, Eohip-
Wind River Basin (typical
Wind River)
Lambdotherium zone
("Lost Cabin";
typical) . H y-
rachyus, Eotita-
nops, Lambdothe-
rium, Heptodon,
Eohippus, Menis-
cotherium.
Heptodon zone ("Lj'-
site"; typical) .
Heptodon, Eohip-
pus.
Big Horn and Clark Fork Basins
Lambdotherium zone ("Lost
Cabin"). Lambdothe-
rium, Heptodon, Eo-
hippus, Ambloctonus.
Heptodon zone ("Lj'site").
Heptodon, Eohippus,
Anacodon.
Systemodon zone ("Gray
Bull," typical). Syste-
modon, Eohippus.
Eohippus zone ("Sand Cou-
lee," typical). Eohippus
(abundant), etc., first ar-
tiodactyls, rodents, and
primates.
Phenacodus zone ("Clark
Fork"; typical). No
perissodactyls, artiodac-
tyls, rodents, or primates.
Fauna more advanced
than in Torrejon.
End of lower Eocene.
First titanotheres ap-
pear.
First lophiodonts ap-
pear.
First tapirs appear.
First horses appear.
Arrival of modern-
ized mammals.
End of basal Eocene.
Archaic mammals
onlv.
68
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
A very significant fact, clearly presented in tlie table
on page 67, is that these small, light-limbed, cursorial
ungulates appear not simultaneously but at successive
horizons. At the lowest level are the horses (Eohip-
pus) ; at a higher level the pseudo tapirs (Systemodon) ;
at a still higher level the lophiodonts (Heptodon) ; and
then, toward the end of the lower Eocene, the titano-
theres {Lambdoiherium) .
Figure 45. — Generalized section through Upper Cretaceous and basal and
lower Eocene deposits near Pumpkin Buttes, Powder River Valley, AVyo.
(No. 12, fig. 3.5)
Adapted from C. H. Wegemann (1917.1).
Though the results of our observations may be modi-
fied by further discoveries the successive rather than
simultaneous appearance of these advancing waves of
perissodactyl migration is what a study of modern
migrations should lead us to expect. All these animals,
as shown elsewhere in this monograph, have similar
cursorial foot structure, which indicates extensive areas
of dry land and open meadow, in which the small,
defenseless Herbivora could easily escape the attacks of
the Carnivora.
Habitat of Wasatch mammals. — The conditions that
prevailed in Wasatch time have been determined very
interestingly by Loomis in his "Origin of the Wasatch
deposits" (1907. 1, pp. 356-364). In adaptation to
various habitats the known species of vertebrates are
divided as follows: Aerial, 3 per cent; cursorial,
terrestrial, and arboreal, 75 per cent; amphibious, 12
per cent; aquatic, 10 per cent. The light-limbed
horse, Eohippus, typical of a plains
or partly open country, alone makes
up 32 per cent of the total collections
from the Systemodon zone ("Gray
Bull" horizon). All the other odd-
toed ungulates are light-limbed, in-
cluding the tapiroids {Systemodon),
lophiodonts (Heptodon), and primitive
titanotheres {LamhdotJierium) , as well
as the surviving archaic condylarths
(PJienacodus and Edocion). The feet
of all these animals indicate dry rather
than swampy or forested land, because
they are more slender than those of
the modern tapir. On the other hand,
the coryphodons were certainly marsh
dwellers and perhaps in part stream
dwellers. The small percentage of
species of truly aquatic animals, such
as crocodiles, fishes, and turtles, whose
remains are mingled with those of the
prevailing land animals, probably be-
came stranded in lagoons far from the
rivers. The presence in the rivers of
rather large fishes is shown by the re-
mains of the large Clastes. Remains
of river-living turtles (Trionyx) have
also been found in the Wasatch.
LOWER EOCENE FAUNAL ZONES
ZONE 6: EOHIPPUS-CORYPHODON ZONE
[Second Wasatch life zone. Big Horn B; lower Sparnacian
of Europe]
Below the EoMppus-CorypTiodon
zone in the Clark Fork Basin of Wyo-
ming lies the first Wasatch life zone
{PJienacodus - Nothodectes - CorypJiodon
zone) described on pages 64-66.
Near the head of the Big Sand
Coulee, on the Clark Fork of the Yellowstone, which
adjoins the Big Horn River basin on the west,
is a series of about 200 feet of red-banded shales,
which overlie the Phenacodus zone ("Clark Fork
beds," transition basal Eocene) and contain a mam-
malian fauna that is radically different from that of
the underlying "Clark Fork." These beds (the
"Sand Coulee beds " of Granger) mark the first appear-
ance in the Rocky Mountain basin region of four
modernized orders of mammals — the lemuroids.
ENVIEONMENT OF THE TITANOTHEEES
69
rodents, artiodactyls, and perissodactyls. Of the
Perissodactyla only one family occurs, the Equidae,
represented by a prhnitive specific form of EoJiippus
(E. horealis). There are two or possibly three species
of Eohippus in these "Sand Coulee beds," which are
not yet separable from the species found in the "Gray
Bull" horizon above. Here also occurs Palaeanodon,
an ancestral armadillo, which left descendants in lower
and middle Eocene time.
This lower Eocene horizon, described by Granger
(1914.1, p. 205), appears to constitute the beginning
of Sparnacian time in the Rocky Mountain region. It
contains the oldest known modernized fauna (perisso-
dactyls, artiodactyls, rodents, etc.) found in America.
The antiquity of these beds is indicated by the last
recorded appearance of the primitive order Multi-
tuberculata, as represented by remains of Ptilodusf sp.
The horizon is also distinguished by the absence of
tapirs (Systemodon) . Here occur the first known
species of the primitive lemuroid Notharctidae (Pely-
codus) and the peculiar ungulate Hyopsodus, now re-
garded as a condylarth. No other exposures con-
taining this very primitive Wasatch fauna have thus
far been discovered.
ZONE 7: SYSTEMODON-COEYPHODON-EOHIPPUS ZONE
[Third Wasatch life zone, Big Horn C; upper Sparnacian of Europe]
The "Gray Bull beds" of Granger (1914.1, pp. 203,
204), in the Big Horn Wasatch, lie at a horizon that
is distinguished by the presence of the earliest tapirs —
the pseudotapirs (Systemodon). These beds were at
first called the Ralston,^ a name that had been pre-
occupied. They are exposed principally in the Clark
Fork and Big Horn Basins south of the Yellowstone
(PI. V, B) and are at least 600 feet thick. They may be
correlated with part of the "Almagre" of the Wasatch
of New Mexico. As this is the first appearance of the
tapirs, and as their remains are mingled with those
of horses, this horizon may be known as the Syste-
modon-Corypliodon-EoMppus zone. These beds are
exposed chiefly along the south side of GreybuU
River, where they extend over many miles. From
this horizon was made the larger part of Cope's col-
lection from the lower Eocene of the Big Horn
Basin, including the classic skeleton of Phenacodus
primaevus, as well as the skeleton of P. copei '° and
that of Eohippus, besides many species of CorypTiodon.
One of the most common forms is the pseudotapir
>"Ealston" was the name given by Sinclair and Granger (1912) to the Clark
Fork beds. "Clark Fork" was substituted by Granger (1914) because "Ralston"
had been previously used otherwise. Sinclair and Granger (1912) referred the beds
between the "Lysite" and the "Ralston" to the "Knight" formation. Granger
(1914) separated the " Knight beds " into two horizons, which he called " Gray Bull "
and "Sand Coulee." The "Gray Bull" and the overlying "Lysite" of Buffalo
Basin constitute the "Big Horn Wasatch" of Cope and Wortman. The "Gray
Bull" is exposed almost entirely in the Big Horn Basin, although a small area of it
overlies the " Sand Coulee" beds at the head of Big Sand Coulee in the Clark Fork
Basin (Granger, 1919).
1 'The type of Phenacodus wortmani is from Wind River. Cope's reference of the
small Big Horn skeleton to this species is not correct. Granger (1915) renamed the
skeleton P. copei.
Systemodon, which includes the species S. tapirinum,
and it is noteworthy that this genus, which is in-
directly related to true tapirs, does not appear in
the overlying beds.
ZONE 8: HEPTODON-COEYPHODON-EOHIPPUS ZONE
[?ourth Wasatch life zone. Big Horn D and Wind River A ; lower Ypresian of Europe]
To zone 8 belong the "Lysite beds" (PI. V, A) of
the Big Horn Basin Wasatch, Wyoming (Big Horn D);
the lower level of the Wind River formation (Wind
River A) ; a part of the Knight formation of the typical
Wasatch group; and parts of the "Almagre" and
"Largo" of the New Mexico Wasatch. In this life
zone Heptodon takes the place of Systemodon, which
disappears or is not thus far recorded. The grace-
ful lophiodont Heptodon appears at the very summit
of the underlying "Gray Bull beds," is abundant in
the "Lysite," and continues into the "Lost Cabin,"
its presence being one of the means of correlating the
fauna of these beds with that of the typical Wasatch
group in the Knight formation. This Knight fauna
occurs in the CorypJiodon-henrmg layer, which Cope
describes as 500 feet above the base of this division
of the typical Wasatch of the Evanston region, or
about the middle third of the formation according to
Granger.
The typical Heptodon zone (= "Lysite") of the
Wind River beds, 350 feet in thickness, is distinguished
by the absence of titanotheres (LamidotJierium,
Eotitanops), which are very abundant in the super-
imposed "Lost Cabin beds." The "Lysite" or
Heptodon zone in the Big Horn Basin is 400 feet thick.
Anacodon, one of the arctocyonid creodonts, which
has flattened or pavement-like teeth, is characteristic
of the Heptodon zone. This zone is faunistically but
not lithologically separated from the overlying Lamh-
dotJierium zone.
ZONE 9: 1AMBD0THEEHTM-E0TITAN0PS-C0RYPH0D0N ZONE
[Fifth Wasatch life zone. Big Horn E, Wind River B, and Huerfano A; upper
Ypresian of Europe]
Geology and fauna. — To zone 9 belong the typical
Wind River of Hayden and of Cope in the Wind
River Basin, Wyo. ( = the "Lost Cabin" of Granger
and Sinclair); the "Lost Cabin" (Granger) of the
Big Horn Basin Wasatch; part of the "Largo beds"
(Granger) of the San Juan Wasatch of New Mexico;
part of the Green River lacustrine formation of
Wyoming; and the lower level of the Huerfano for-
mation (Hills) of Colorado or Huerfano A. This is
the typical Wind River life of all the literature of
Cope. (See PI. VI.)
The Lamldotherium life zone is distinguished by the
arrival in the Rocky Mountain basin region of the
first titanotheres, which are abundantly represented
in remains of the smaller, cursorial Lambdotherium
and the larger, mediportal Eotitanops. It includes the
70
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
FiGUHE 46. — Composite section of the Eocene deposits of tlie Big Horn and Clark Fork Basins, Wyo.
This section contains ttie entire Big Horn Basin Wasatch of Cope's descriptions, which is now divided into very clearly defined
ascending life zones, as follows: 5, Phenacodus-Nothoiectes-Coryphodon zone; 6, EoMppus- Corijphodm zone; 7. Systemodon-Corypho-
don- Eohippus zone; 8, Heplodon- Coryphadon-Eohippus zone; 9, Lambdotliermm-Eotitanops-Coryphodon zone. A few characteristic
species of mammals from each horizon are indicated in the right-hand column. Chiefly after Granger (1918).
ENVIEONMENT OF THE TITANOTHERES
71
last surviving species of Coryphodon and of the con-
dylarth Phenacodus among the archaic ungulates.
The presence of the condylarth Meniscotherium
serves to correlate the Wind River with the upper
levels ("Largo beds") of the Wasatch of New Mexico.
While the Wind River life on the whole represents
a continuation of that of the preceding stages of the
Wasatch, with which it possesses several genera and
eleven species in common, it also includes nine new
genera that survive in the Bridger formation of
middle Eocene time. The Wind River marks the
end of the lower Eocene, the last period of certain
highly distinctive lower Eocene forms like Cory-
phodon, but it is also prophetic of the middle Eocene
in the presence of lemuroids like Notharctus, Anapto-
somewhat like a slender, diminutive tapir in body
proportions. In skull structure and dentition Eoti-
tanops foreshadows the true titanotheres of the
middle Eocene; its feet are more slender than those of
its successors, and it was doubtless a more agile animal.
The special life conditions surrounding these early
titanotheres are more fully set forth in the descriptions
of the Wind River titanotheres in Chapter V, section 3.
Olimate and physiography during the deposition of
the Wind River and Green River sediments. — For Wind
River life in general the reader is referred to section 3
of this chapter. Here we may speak of the whole
basin region.
While fluviatile and flood-plain sediments were
being deposited in the Wind River Basin of northern
f!t» - ""."-i**^*,
Figure 47. — A typical "Lost Cabin" locality, on the north side of AlkaU Creek about 8 miles east of Lost
Cabin, Wind River Basin, Wyo.
Lambdotherium-Eotitanops- Coryphodon zone (Wind River B). A characteristic view of tlie red-banded beds that have yielded the greater part of
the fauna of the Lambdotherittm zone. (Compare PI. VI, B.) After Granger (1910.1), Am. Mus. negative 17792.
morphus, and Shoshonius; of true doglike or civet-like
carnivores like Viverravus and Vulpavus; or of rodents
like Sciuravus and Par amy s. Remains of Equidae
are rather rare and are represented by several species
of Eohippus, of which E. venticolus is the most pro-
gressive, and those of titanotheres, especially Lamb-
dotherium, are very abundant.
Lambdotherium, one of the earliest titanotheres,
was a small, light-limbed form, about the size of a
coyote {Canis latrans). It represents a distinct
cursorial side branch of the titanothere family, re-
sembling the contemporary horses and lophiodonts
in its light limb and foot structure. Eotitanops
("the dawn titanothere") was a true and very primi-
tive titanothere about the size of a sheep {Ovis aries),
Wyoming there lay to the south a large, shallow lake,
covering about 5,000 square miles, in which were
deposited 800 feet of impure limestone at the base,
followed by about 1,200 feet of thin, fissile calcareous
shale. (King, 1878.1, p. 381.) The deposition of
these lake sediments (Green River) began near the
end of Wasatch time. They contain abundant and
well-preserved remains of insects and fishes. The
presence of sting-rays and other fishes of marine or
coastal type indicates that these originally marine
forms had become landlocked, as did the existing
marine survivors in the Caspian Sea and Lake Titicaca.
Many of the fishes of the Green River shales are related
to forms now found chiefly in the southern continents,
especially South America.
72
TITA.NOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Green River forests. — Our present knowledge of the
Green River flora, which, according to Berry (1914.1,
p. 164) was mid-Eocene, indicates a considerably
warmer climate than that of the basal Eocene Fort
forests differed from the tropical forests of the Georgia
coast in the presence of genera Hke Bex, Juglans,
Myrica, Planera, Quercus, Rhus, Salix, and Zizyphus,
most of which are temperate types. Thus the Green
MESOZOIC and PALEOZOIC
Pe/ycoc/us
Sh osh on fas
/^bsaroh/L/S
' Didyrnic^iS
\//verr3\/us
Miacis
Vu/pavus
Pafripfe/is
Pro/i'mnocyon
S/nopa
Trif-emnoc/on
^Hapa/odecfes
CynodonfomyS
Diacodon
Pariofops
D/de/pnodaS
,Pa/seosinopa
{Esfhonyx
' Paramys
Mysops
? Palaeanodon
5/y/inodon
Phenacodus
Pcfoconus
Menisco^her/um
Pyopsodus
Coryphodon
Baihyops/s
Eoffranops
Lambdoiher/um
Hyrachyus
Hep^odon
fiobippus
[O'Scodex/s
Pe/ycodus
Omomys
Te fortius
Abs3ro/<ius
Didymicii's
Vulpa vus
Oxyasna
S/nopa
Hapa/odecfes
Cynodon fomys
En'fomo/esfes
■acodon
Esfhonyx
Paramys
PPa/asanodon
Phenacodus
Pyopsodus
Coryphodon
Pephodon
Eobippus
Wasdi-chia
Diacodexis
Figure 48. — Section through the Wind River formation (lower Eocene) near Lost Cabin, Wind River Basin
Wyo. (No. 5, fig. 35)
A complete list or genera from each horizon is given in the right-hand column. First appearance of the primitive titanotheres Lamidotherium
and Eotitanops. Chiefly after Granger (1910.1) and Sinclair and Granger (1911.1).
Union, for it includes such types as Acrostichum and
Arunio, which are also represented in the contempo-
raneous Eocene flora of Georgia, as well as the genera
Ficus and Sapindus. The Green River lake-border
River exhibits a commingling of warm-temperate and
tropical trees such as are now found in subtropical
forests in regions where there is a mean annual temper-
ature of about 14° C, uniform humidity, and an
ENVIEONMENT OF THE TITANOTHEHES
73
annual rainfall exceeding 200 centimeters. This flora
is not very difl'erent from that found in the upper
Ypresian of France.
These forests are so interesting in respect to the
environment of the first titanotheres which appeared
in North America that the principal genera cited by
Berry may be quoted in full. The figures
appended to the names of the genera
show the number of species in each genus.
(Sinclair and Granger, op. cit., p. 105.) There is
evidence also of an uplift of the Big Horn Range
subsequent to the deposition of the Wasatch. In
the Wind River Basin material washed down from
the mountains continued to be spread over the basin
floor by streams until the end of upper Eocene time.
Acrostichum, 1.
Alnus, 1.
Ampelopsis, 1.
Aralia, 1.
Arundo, 2.
Brasenia?, 1.
Cheilanthes, 1.
Eucalyptus?, 1. Myrica, 1.
Ficus, 4. Phragmites, 1.
Ilex, 2. Planera, 2.
Juglans, 3. Quercus, 2.
Leguminosites, Rhus, 1.
1. Sabal, 1.
Lygodium, 1. Salix, 2.
Cissus, 1. Manicaria, 1. Sapindus, 1.
Cyperus, 1. Musophyllum, Sphaeria, 1.
Equisetum, 1. 1. Zizyphus, 2.
The Green River flora is the only de-
scribed middle Eocene flora known from
latitude 40°. The nearly contempora-
neous Claiborne flora of Georgia shows
(Berry, op. cit., p. 161) that the main
elements of the modern flora of tropical
America reached at least as far north as
latitude 33° and, in the middle Eocene,
probably several degrees farther north.
Wasatch and later events. — In areas that
lay north of the great lake in this region
in Wasatch time were laid down the sedi-
ments of the Wind River and Big Horn
Basins, the deposition of which began in
the first phase of Wasatch time and prob-
ably continued into middle Oligocene
time. (Sinclair and Granger, 1911.1, p.
85.) The Wind River sandstones in the
vicinity of the Beaver Divide are stream-
channel deposits, probably laid down in
broad, shifting streams of low gradient
which flowed across clay-covered flats,
into which they sunli their channels or
over which in seasons of flooding they
spread coarse detritus. The materials
composing these sandstones were derived
from the granites and other pre-Ter-
tiary rocks of the surrounding moun-
tains. Below the Lambdoiherium zone
Figure 49. — Map showing cluster of lower, middle, and upper Eocene sedi-
mentary basins in southwestern Wyoming and northern Utah, exhibiting
parts of areas of the Wasatch, Wind River, Bridger, and Uinta formations
at other localities, interstratified with Extensive areas of the Wasatch are purposely omitted. After Osborn and Matthew (1909.321), U. S.
Wi T~» • 1 1 ^ A ,1 Geol. Survey Bull. 361. Arrows show lines along which sections were taken
md River clays and sandstones, there
are layers of white volcanic tuff, 13 feet thick, in-
dicating the presence of active volcanoes. The floor
of the Big Horn Basin, to the north, was modified
by erosion that took place subsequent to the main
uplift of the Big Horn Mountains, which occurred
after the deposition of the Fort Union formation.
101959— 29— VOL 1 7
Fluviatile and flood-plain deposition is indicated
throughout Eocene time. The lignitic shales that
cap the Lamhdof-herium zone of the Big Horn Basin,
containing fresh-water mollusks (Planorhis) and crus-
taceans (Entomostraca), are certainly both fluviatile
and palustrine.
74
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
TRANSITIONAL LOWER TO MIDDLE EOCENE DEPOSITS
HUERFANO FORMATION OF COLORADO (LOWER AND MIDDLE EOCENE)
While the lacustrine and flood-plain Green River
and Wind River formations were being deposited in
Wyoming there were accumulating in southeastern
Colorado the lower fossiliferous beds of the Huerfano
formation, described by Hills (1888.1), explored by
Osborn and Wortman in 1896 and by Granger and
Olsen in 1918, and now known as Huerfano A. The
deposition of this formation apparently began near
the end of Wasatch time and extended into early
although part of its fauna is doubtless transitional
from the summit of the underlying lower Eocene.
In the upper half of the Huerfano formation (Huer-
fano B) are found mammals that are also characteristic
of the lower Bridger (A) . The imperfectly known life
of the upper level includes the tillodont Trogosus and
two kinds of small titanotheres, one (Eometarhinus)
resembling Metarhinus and the other Palaeosyops fon-
tinalis of Bridger A; also a horse {OroMppus?) and ani-
mals resembling the Bridger genera Hyrachyus, Hyop-
sodus, Microsyops, as well as more ancient genera —
the creodonts Amhloctonus and Didymidis — which
Figure 50. — Sketch map of the region of the Huerfano and Cuchara formations in southern Colorado
After Hayden (1880), Hills (1888.1), and Granger (1918).
Bridger time. Among the mammals of the lower
Huerfano, which corresponds with the upper Wind
River (="Lost Cabin"), are the rare Coryphodon, the
small-limbed titanothere Lambdotherium, EoTiippus,
Oxyaena, Didymidis, and Heptodon, a purely upper
Wind River (="Lost Cabin") fauna.
The whole Huerfano formation is 3,500 feet thick,
and a large part of it (see fig. 51) lies below horizon
A. (Granger, 1918.) Huerfano B, although it lies
immediately above Huerfano A, contains the genus
Palaeosyops, a distinctive middle Eocene form. Con-
sequently Huerfano B is placed at the base of the
middle Eocene and is correlated with Bridger A,
suggest a fauna more ancient than that of Bridger B,
corresponding perhaps with the still unknown fauna
of Bridger A. It appears probable that the Huerfano
will give us a complete faimistic transition between the
end of Wasatch and the beginnuig of Bridger B (middle
Eocene) time.
WIND RIVER BEDS AND THEIR FAUNA
The discovery of the geologic section at Beaver
Divide, between Wind River and Sweetwater River, is
one of the most significant recently made in the study
of Rocky Mountain basin geology. Here deposition
without angular unconformity extends from the third
ENVIRONMENT OF THE TITANOTHEEES
75
Figure 61. — Section of the Huerfano formation in southeastern Colorado as exposed west of
Gardner, Huerfano Basin
Thickness 3,500 feet. Near the summit is a Wind River B fauna (.Lambdotherium zone), and above that a Bridger A fauna
(.Palaeosyops fontinalis zone). After Granger (1918).
76
TITANOTHEKBS OF ANCIENT AVYOMING, DAKOTA, AND NEBEASKA
Oj^eodoTL zoTie
Oreodon
Cylindrodon
> Caenopus
Ischyromys
Poebro therium
Menodus he/oceres
?I)ipioicodo7v zoTze^
Amynodon ? anficjuus
Protoreodon
Camelodon
^ Pro tifan other/ um
Zamhdotherhwv zone
Lambdofherium
Coryphodon, Phenacoo/us_,
Hepfodon, Eohippus
Figure 52.— Section of exposures from lower Eocene to lower Oligocene at Green Cove,
on Beaver Divide, at the southwestern border of the Wind River Basin, Fremont
County, Wyo.
Includes deposits in Wind River, Bridger (?), Uinta (?) and White River time. Chiefly after Granger (1910.1). This
is a most significant section, for the base ot the TiimolUnum zone (Chadron A) uncontormably overlies beds
originally referred to Uinta C 1 (Diplacodon zone).
ENVIEONMENT OF THE TITANOTHEEES
77
Wasatch Heptodon-Coryphodon-EoMppus zone through
the Wind River Lamhdotherium-Eotitanops-CorypTiodon
zone upward into the Oreodon zone of Ohgocene time.
This is the only undoubted Eocene-Oligocene sedi-
ment thus far determined in the Rocky Mountain
basin region. Its total thiclmess is 1,080 feet, and it
represents relatively slow sedimentation. There is a
single period of erosional unconformity at the end of
the upper Eocene.
The life of the Wind River beds of this section is
distinctly of upper Wind River ("Lost Cabin") time,
corresponding with Wind River B and Huerfano A,
for it includes the titanothere Lambdotherium popo-
agicum, a CorypTiodon, two species of Equidae {Eohippus
craspedotus and E. venticolus), and two species of
Heptodon (H. calciculus and H. ventorum), which are
characteristic of closing Wasatch time. The presence
of remains of garpikes and crocodiles in this fauna
shows that these deposits were fiuviatile and indicates
that Wind River shales were of flood-plain origin,
though they include many channel fillings of coarse
arkose.
We thus have glimpses of a faunistic period broadly
corresponding with the lower Ypresian of France, cer-
tainly extending from Wyoming to Colorado, and
probably spreading much more widely in the Rocky
Mountain and the adjacent Plains region. Though it
includes surviving members of the older Wasatch life
and incoming members of the succeeding Bridger life,
the Wind River and Huerfano life stands directly
intermediate between these; in fact, the representa-
tives of archaic families destined to become extinct
and those of modernized families destined to populate
the earth are very nearly balanced, including 21 genera
(30 species) of archaic mammals and 22 genera (36
species) of modernized mammals.
Simultaneously with the decline of the coryphodons
the uintatheres reappeared in the genus Bathyopsis,
ancestral to the giant Uintatherium, which character-
izes Bridger C and D.
THIRD FAUNAI PHASE (MIDDLE AND UPPER EOCENE)
CORRELATION OF AMERICAN ZONES WITH EUROPEAN
STAGES
There is strong evidence of uniform and favorable en-
vironment and persistent evolution throughout middle
and upper Eocene time in the Rocky Mountain basin
region. The changes show progressive modification
and adaptation rather than breaks by migration or
extinction. Both the archaic and the modernized
families increased in size and variety. The surviving
archaic mammals appear to have flourished and in-
creased, especially in size and muscular power. Near
the very end of Eocene time only two new famihes of
quadrupeds appear, the ancestral camels (Camelidae)
and the oreodonts (Oreodontidae), whereas in western
Europe new families repeatedly appear from the south.
east, and north. The general correlation of the Euro-
pean stages and the American zones is given on page 78.
78 TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBEASKA.
Correlation of middle Eocene and upper Eocene American life zones and European stages
Epoch
American life zone
Approximate European
stage
15. Diplacodon-Protitanotherium-Epihippus zone (Uinta C).
Ludian.
Upper Eocene.
14. Eobasileus-Dolichorhinus zone (Washakie B 2, Uinta B 2).
13. Metarhinus zone (Washakie B 1, Uinta B 1).
Bartonian.
12. Uintatherium-Manteoceras-Mesatirhinus zone (Washakie A, Bridger C and
D).
Lutetian.
Middle Eocene.
I 11. Palaeosyops paludosus-Orohippus zone (Bridger B).
i 10. Eometarhinus-Trogosus-Palaeosyops fontinalis zone (Bridger A, Huerfano B) .
TYPICAL BBIDGER FORMATION; MIDDLE EOCENE
(LUTETIAN AND BARTONIAN OF EUROPE)
Character of sediments. — The Bridger formation, the
most important and the most thoroughly explored of
the Eocene Tertiary, occupies a great area north of
ditions. The Bridger formation attains its maximum
thickness of 1,875 feet near the Uinta Mountains and
thins out northward. Beyond the margins of the
Green River lacustrine deposits the Bridger overlies
upper members of the Wasatch group.
Figure 54. — Map showing the Eocene sediments encircling the Uinta Mountains of southwestern Wyoming
and northern Utah
Modified after King (1876.1). U, Uinta Basin, typical Uinta formation of King and Marsh. (The area mapped includes older and possibly
younger rocks than the true Uinta formation— Z)/piacoion zone.) B, Bridger Basin, typical Bridger formation of Hayden. WK, Washakie
Basin, typical " Washakie" formation of Hayden. G, Green River formation. W, Typical Wasatch group of Hayden. X, Type locality
of Coryphodon and associated Wasatch fossils. C, Cretaceous.
the Uinta Mountains and east of the Wasatch. Unlike
the Wasatch, the lower Bridger (horizon B) is unique;
no contemporaneous fossiliferous deposition is known.
At the base its sediments pass gently into the Green
River shales, and the lower levels of Bridger A show
gradual transition from lacustrine to flood-plain con-
Unlike the lower Eocene Wasatch and Wind River
sediments the Bridger is not composed chiefly of
material derived by erosion from the adjacent moim-
tains (Sinclair, 1906.1, p. 278) but consists of great
series of deposits of volcanic ash and dust, solidified
into tuffs, which weather into picturesque and in
ENVIRONMENT OF THE TITANOTHEHES
79
places highly colored "badlands." Apparently the
greater part if not all of these tuffs were distributed
from unlinown eruptive volcanic centers by wind; but
at four periods they were deposited in great shallow
playa lakes and partly worked over by stream, delta,
and flood-plain deposition. From the general absence
of coarse materials such as would be transported by
streams of high gradient, it is inferred that the Bridger
formation accumulated in a relatively level area.
(Sinclair, 1906.1, p. 279.)
Exploration of the Bridger formation. — The Bridger
formation has been explored almost continuously by
geologists and paleontologists, first by Hayden (1869-
1871), next by King (1878), who regarded the Bridger
as an ancient lake basin deposit, then by Osborn and
Scott (1877-1878), and again by Endlich (1879).
In 1902 the American Museum parties, guided by
Matthew and Granger, under the direction of Osborn,
undertook to determine whether the Bridger can be
divided into a series of life zones. After four years of
careful geologic field work by Granger and Matthew
(1902-1905), who had at hand the level record of every
specimen, the Bridger was subdivided lithologically
and faunistically into five levels, A to E. Bridger A
is relatively barren. Of these levels A and B were
grouped into the lower Bridger (Palaeosyops paludosus-
OroMppus zone), characterized by the absence of
Uintatherimn, and C and D, the upper Bridger
{Uintatherium- Manteoceras- Mesatirhinus zone), distin-
guished by the appearance and great abundance of
TJintaiherium. Similar faunistic surveys in the
Washakie Basin, east of the Bridger Basin, and
in the Uinta Basin, south of the Uinta Mountains,
have given very complete correlation of the local
subdivisions of the section as follows:
Correlation of middle and upper (?) Eocene sections of the Uinta, WashaTcie, and Bridger Basins
Uinta Basin
Washakie Basin
Bridger Basin
Life zones
Uinta C.
Absent.
Absent.
Theoretic zone (No. 16); fauna unknown.
15. Diplacodon-Protitanotherium-Epihippus.
Uinta B.
Washakie B.
Bridger E (barren beds).
14. Eobasileus-Dolichorhinus.
13. Mctarhinus.
Uinta A (barren).
Washakie A.
Bridger D.
Bridger C.
12. Uintatherium-Manteoceras-Mesatirhinus.
Barren beds.
Barren beds.
Bridger B.
11. Palaeosyops paludosus-Orohippus.
Bridger A.
10. Eometarhinus-Trogosus-Palaeosyops fontinalis.
Volcanic ash deposits. — The petrographic analysis of
the rocks of the Bridger formation serves to support
their correlation with the deposits of the Washakie
Basin, to the east, and of the Uinta Basin, to the south.
The recognition by Sinclair (1906.1, pp. 273-280) of
the fact that the entire Bridger series was in large
part originally volcanic dust and the later careful petro-
graphic analysis by Johannsen (1914.1) led to the
conclusion that the Bridger rocks are largely tuffs
perhaps modified in part by sufficient transportation to
add the numerous grains of quartz they contain, and
that these grains may be of sedimentary origin
although the material of the tuffs is mostly andesitic.
Johannsen's analysis of the Bridger rocks is essentially
as follows:
Bridger D. Irregular grains of quartzite, feldspar, hornblende,
etc. : dacite tuff.
Bridger C. Fragments of quartz and hornblende; groundmass of
glass tuff.
Bridger B. Smith's Fork; fragments of quartz, feldspar, horn-
blende: ?dacite tuff.
Bridger B. Church Buttes; fragments of quartz, feldspar, etc.:
altered tuff, probably dacite tuff.
Bridger A. North of Church Buttes, fragments of quartz, feld-
spar, hornblende. No glass tuff seen.
Thus the Bridger is composed chiefly of dacite tuff,
of altered dacite, and of glass tuff containing irregular
grains of quartz, feldspar, and hornblende, which
are at some places contained in a groundmass made up
of entirely coarse angular particles of stringy glass
full of bubbles. The Huerfano formation of Colorado,
which is in large part older than the Bridger, is com-
posed largely of glass tuff. The deposits in the
Washakie Basin, east of the Bridger Basin, are com-
posed chiefly of dacite and glass tuffs.
Playa lalce deposits. — Conspicuous features of the
Bridger formation are four hard "white layers,''
which were laid down at intervals in the series of beds.
80
TITANOTHEBES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Some of these "white layers" have been traced over
many square miles. They are composed of tuffaceous
shale and marl or of calcareous shale and are in places
filled with fresh-water shells. They mark periods
during which the deposition of volcanic dust was less
rapid, when the Bridger Basin was temporarily base-
leveled and the waters rose into wide, shallow playa
lakes, in which sedimentation was slow. That these
four relatively thin "white layers," which are vari-
FlGURE
-Geologic section of the entire Bridger formation in the
Bridger Basin, Wj'o.
Shows the division by the four chief "white layers" and the main divisions by three principal
zones — A, Palaeosyops fontinalis zone; B, Palaeosyops paludosus-Orohippus zone; C and
Utniatherium zone.
ously known geographically as the "Cottonwood
white layer," the "Burnt Fork white layer," the
"Lone Tree white layer," and the "upper white layer,"
correspond with long periods of geologic time is shown
by the marked faunal differences that separate them,
which indicate that extensive migration occurred
before and after the deposition of each of these layers,
but especially the first, which separates the lower
from the upper Bridger life zone.
Life environment in Bridger time. — From observa-
tions made by Hay (1905.1, pp. 327-329) while he was
collecting fossil turtles in the Bridger in 1903, he con-
cluded that the Bridger deposits were almost solely
the result of fluviatile and flood-plain action, that this
basin was a nearly level country, which was probably
covered with vegetation and well forested. The dis-
tribution of fossil remains in all parts of the Bridger
area indicates that the animals lived near the places
where they became buried and that they were chiefly
such as may inhabit well-wooded regions. The river-
channel beds, which are composed of coarse ma-
terials, show that streams with rapid currents
traversed the basin. These streams were bor-
dered by swamps in which were formed beds of
impure lignite, or by fresh-water bays in which
the shells of fresh-water mussels accumulated.
The finer deposits indicate shallow, muddy
bays, in which the remains of the larger quad-
rupeds are occasionally found in positions
indicating that they had been mired in a
standing posture. The old stream channels
have yielded remains of several species of
bowfins (Amiidae), garpikes (Lepidosteus) ,
and siluroids. Crocodiles were numerous and
diversified. The reptiles suggest that the
climate was Floridian, or south temperate, and
we may picture a partly open, partly forested
country, somewhat similar to the existing
bayou region of the Mississippi Delta of
Louisiana. Analysis of the Testudinata by
Hay (1908.1) has also afforded a clear idea of
the physiographic conditions in Bridger time.
The soft-shelled river turtles (Trionychoidea)
were represented by at least 25 species, and
there are now in the world only 26; the
Bridger rivers and brooks fairly swarmed
with these creatures, some of them equal in size
to the largest existing Asiatic species. There
are indications of 4 species of the family
Emydidae (order Cryptodira), as compared
with the 12 species now living in the Missis-
sippi Valley. The genus Baptemys, of the
same order, has its nearest relatives at present
in Central America, and a third genus
(Anosteira) is reported by Lydekker in the
upper Eocene of England. The presence of
: and D, extcusive stretches of land is indicated by the
true land tortoises (Testudinidae) of the genus
Hadrianus, including giant tortoises nearly 3 feet
long, which probably lived on dry lands bordering
the sluggish Bridger streams. The ancient Lower
Cretaceous order Amphichelydia is also represented
here by four species belonging to two genera.
The environmental adaptations of the animals of
the Bridger Basin were classified by Matthew (1901.1,
pp. 309, 310) as follow?:
Land animals:
1. Aerial: Remains of birds rare and fragmentary, as in
nearly all geologic formations.
ENVIHONMENT OF THE TITANOTHEEES
81
2. Arboreal: Primates, many Carnivora, and some Insec-
tivora and Rodentia. Out of 1,007 specimens, belong-
ing to 46 genera, 13 genera (184 specimens) were
certainly arboreal and 11 genera (485 specimens)
were probably arboreal.
3. Terrestrial (cursorial and ambulatory) : Some of the
carnivores and all the ungulates (17 genera, 314
specimens). Also some lizards and chelonians.
4. Fossorial: Certainly fossorial, 3 genera (S specimens).
Some of the insectivores may also have been fossorial.
5. Amphibious: One insectivore (Pantolestes) (1 genus).
Probably certain carnivores.
Water animals:
6. Fresh- water: Numerous crocodiles, aquatic turtles,
fish, and fresh-water mollusks.
7. Marine: No marine animals. Contrast this lack of
types with the types of fish in the preceding Green
River formation.
The Bridger life thus included many arboreal,
terrestrial, and aquatic forms, the last mostly reptiles,
fishes, and invertebrates. The slow-moving, ambula-
tory quadrupeds form a relatively large proportion of
the mammals, but the cursorial types, such as the
Equidae (Orohippus), are relatively rare; also the fos-
sorial types. The Bridger life seems to be that of a
partly forested flood plain. The remains of large
mammals are so numerous as to indicate abundant
open, gladed areas, comparable to the partly forested
and partly open delta regions along certain rivers of
modern time.
The foot structure of the Bridger quadrupeds gives
less certain evidence of an open plains country,
favorable to cursorial types, than that of the Wasatch
(lower Eocene) quadrupeds of the same region.
No impressions of leaves from the Bridger forests
have been discovered. It is probable that the forests
of Green River type, described on pages 72-73, per-
sisted into Bridger time and that the climate then
was warm-temperate, almost subtropical.
The faunal history of the Bridger as a whole shows
a gradual reduction in the number of archaic mammals
of Mesozoic stock and a steady increase in the number
of their competitors among the modernized mammals,
the numerical relations between these two groups in
upper Bridger time being as follows:
Genera Species
Archaic mammals 15 35
Modernized mammals 57 146
Duration oj the Bridger epoch. — Matthew (1909.1),
following the earlier geologists, believes that the
lacustrine conditions in Green River time arose from
the uplift of the Uinta Mountain range, which blocked
the basin and caused the formation of the great lake
in which the material that formed Green River shale
was laid down. As the river gradually cut its way
through the east end of the Uinta Range the lake
gave way to the broad Bridger flood plain, in which
was deposited the volcanic ash washed down from
the slopes of the Uinta Mountains to the south, and
the deposit was worked over and sorted by the streams
that flowed across the plain. The Bridger Basin was
subject to intermittent overflow, which gave rise to
large but shallow lakes of clear water. If we should
assume that the Bridger formation occupied one-
tenth of total estimated Eocene time — 90,000 to
100,000 years — the fossiliferous beds, which are 1,100
feet thick, have accumulated at an average rate of
12 inches per century. This estimate would allow
110,000 years for the deposition of the Bridger forma-
tion exclusive of the "white layers " formed at intervals
when deposition was arrested. The titanothere re-
mains of the Bridger indicate a long period of evolu-
tion, but not so long as that of the Chadron (lower
Oligocene).
Chief localities and exposures of the Bridger formation in ike
Bridger Basin
Bridger E:
Uppermost exposures:
Sage Creek Mountain.
Henrys Fork Table.
Twin Buttes.
Bridger D:
Upper exposures: Level
Twin Buttes D 1-5
Spanish John's Meadow D 1-5
Cat Tail Spring D 1-5
Henrys Fork, Burnt Fork post office D 1-5
Henrys Fork, Lone Tree post office D 1-5
Summers Dry Creek D 1-5
Henrys Fork Hill D 1-5
Sage Creek Spring D 1-5
Lane Meadow D 1-5
Bridger C:
Lower exposures:
Henrys Fork, Lone Tree post office C 4-5
Lane Meadow C 3-5
Spanish John's Meadow C 3-5
Henrys Fork Hill C 3-5
Twin Buttes C 1-5
Dry Creek C 1-5
Henrys Fork, Burnt Fork post office C 1-5
Church Buttes, third bench C 1-3
Bridger B:
Upper exposures :
Cottonwood Creek Typical B 4^5
Millers ville, 6 miles southeast of B 4-5
Cottonwood Creek, middle of B3
Grizzly Buttes B 3
Church Buttes B 2-3
Lower exposures:
Cottonwood Creek B2
Grizzly Buttes Typical B 2
Exposure B, 5 miles south of Granger B 2
Millersville B 1-2
Cottonwood Corral, Blacks Fork ,. B 1-3
Exposure A, 5 miles south of Granger B 1
Church Buttes B 1
Bridger A:
Hams Fork Bluff; Granger to Opal, 25 miles.
Mouth of Big Sandy Creek {IPalaeosyops -fontinalis
zone).
Big Muddy exposures between Carter and Granger.
Blacks Fork Bluffs, east of Granger.
82
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
ZONE 10: EOMETAEHINUS-TEOGOS0S-PALAEOSYOPS FONTINAIIS ZONE
[Bridger A and Huerfano B ; lower Lutetian of Europe]
The lower Bridger (Bridger A and B) corresponds
with the "calcaire grossier superieur" of the Paris
Basin. The correlation of Bridger A with Huerfano
of some 200 feet of calcareous shale alternating with
tuff (Matthew, 1909.1), which are exposed principally
around the eastern, northern, and western margins of
the Bridger Basin. It is therefore supposed that
Bridger A, which passes down into the lacustrine
Green River shales, is partly of lacustrine, partly of
Figure 56. — Map of the Bridger Basin, Wyo. (No. 8, fig. 35)
Showing the principal topographic features, Twin Buttes and Henrys Fork Table, and a diagrammatic section of the Bridger formation (A, B, C, D, and E) capped by
the Bishop {"Wyoming") conglomerate (W). After Matthew and Granger, 1902, 1909.
B has recently been established through the discovery
in each of the mammalian species Palaeosyops (Lim-
nohyops) fontinalis Cope. (Osborn, 1919.494.) In
these beds vertebrate fossils are rare and include,
besides the titanothere above mentioned, remains of
crocodiles, turtles, and fishes only. Bridger A consists
fluviatile origin and is transitional both geologically
and in its fauna between Green River (upper Wind
River) and Bridger B time. Sinclair describes this
horizon as consisting of "buff and pale-green tuffaceous
shales and sandstones, often containing in enormous
numbers shells of Paludina and Unio."
ENVIRONMENT OF THE TITANOTHEEES
83
z
LJ
O
O
liJ
U
_J
Q
Q
no
q:
DQ
B3
Pa/aeosyops
majo/; neofype
{?)Limnohyops
laevidens, type
Palaeosypps palicdosits-
Pa/aeosyops major, ref.
Limnohyops monoconus, type
L/mnohyops matfhewi, type
L/mnohyops priscus, type
Limnohyops
monoconus
OroTiippus typicus
GRIZZLY BUTTES FAUNA
Notharctus
Harpagolestes
Hyradiyus agrarizis
MetadieiroTnys dasypus
Orohippus atavus
TUlothjeriLawfodiens
A5
,'' Pa/aeosyops
fonf/na/is
cr
Uo
q:
CD
NO MAMMALS
oo
cc
^ LL t_>
u bj F
UJ > <
a ^ 5
o '^ e
Figure 57. — Section of the lower part of the Bridger formation in the Bridger Basin, Wyo. (No. 8, fig. 35),
showing the succession of the species of titanotheres and other mammals
The section is 650 feet thick. The principal geologic features are represented in the center. After the studies of Osborn, Granger, and Matthew.
84
TITANOTHEBES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
ZONE 11: PAIEOSYOPS PALUDOSUS-OEOHIPPUS ZONE
[Bridger B; upper Intetian of Europe]
The richly fossiliferous deposits belonging to the
Paleosyops paludosus-OroMppus zone (Bridger B)
are exposed chiefly in the northern area of the Bridger
formation, near Fort Bridger, along Blacks Fork and
its tributaries. They represent a very long period and
consist of 450 feet of tuffs and sandstones (fig. 4)
divided into two principal escarpments, which are
separated by Cottonwood Creek valley.
In this zone the titanotheres and other mammals
undergo notable progressive evolution, and there is a
marked succession of species. (See fig. 57.)
The succession of the species of titanotheres in
Bridger B, in descending geologic order, is as follows:
Limnohyops monoconus Os-
born, type.
Limnohyops matthewi Osborn,
type.
Palaeosyops paludosus Leidy.
Palaeosyops paludosus, re-
ferred.
Palaeosyops paludosus, type.
?Mesatirhmus Junius Leidy.
Palaeosyops major Leidy, hy-
potype.
Limnohyops laevidens Cope,
type.
Palaeosyops major Leidy, re-
ferred.
Limnohyops monoconus Os-
born, referred.
The species of titanotheres found in Bridger B
belong exclusively to the subfamily Palaeosyopinae
and represent the two generic branches Palaeosyops
and LimnoTiyops, closely related animals with broad
spreading feet and heavy limbs, slow in gait. The
reference to Mesatirhinus of the species P. Junius
Leidy is somewhat doubtful. The lower half of
Bridger B at Grizzly Buttes (PI. VII, B) , an escarp-
ment along Smiths Fork, is by far the richest collect-
ing ground in the Bridger Basin; thousands of speci-
mens have been taken from it, including many more
or less complete skulls and skeletons, all recorded
from Bridger B 2. Beds at a slightly higher level, in
Bridger B 2 and in Bridger B 3, on the escarpment
along Cottonwood Creek, have yielded a number of
complete skeletons, including those of several species
of Equidae (OroTiippus), a variety of catlike and dog-
like creodonts {Limnocyon, Harpagolestes), abundant
small civet-like creodonts (Viverravus, Sinopa), an-
cestral canids (Miacis, Uintacyon), a surviving (?)
condylarth (Hyopsodus) ; also ancestral Edentata
{Metacheiromys, armadillo-like) and the rodent-like
tillodonts {Tilloiherium fodiens, Trogosus). They
have also yielded many rodents (Paramys, Sciuravus),
as well as a rich primate fauna of lemuroids (Noth-
arctus). The entire fauna has been very carefully
reviewed and analyzed by Matthew (1909.1, pp.
298-302).
Rich as is the fossil life of the lower Bridger, many
mammalian subfamilies and many genera and species
are lacking which occur abundantly in the upper
Bridger. Noticeable is the absence of uintatheres
(Uintatherium) and of three important genera of ti-
tanotheres {Manteoceras, Telmatherium, Mesatirhinus),
which appear abundantly in the upper Bridger.
The "Cottonwood Creek white layer," marking
the summit of Bridger B, indicates a long period of
shallow lake flooding of the Bridger Basin during
which the large amblypod uintatheres and the more
advanced titanotheres entered the basin. Vintaihe-
rium is not found in Bridger B, but it occurs at the
very base of Bridger C, the lowest level of the upper
Bridger.
ZONE 12: UINTATHERIUM-MANTEOCERAS-MESATniHINUS ZONE
[Bridger C and D, Wasliakie A, and Uinta A; part of Bartonian of Europe]
The fauna of zone 12 in the Bridger Basin, which
includes deposits 725 feet thick (Bridger D, 375 feet;
Bridger C, 350 feet; see fig. 58), may be clearly dis-
tinguished from that of zone 1 1 (lower Bridger = Bridger
B and A) by its content of the remains of the animals
listed below:
Titanotheres :
Palaeosyops robustus Leidy.
Palaeosyops copei, type.
?Telmatherium validum, type.
Manteoceras manteoceras.
Mesatirhinus petersoni, type.
Palaeosyops leidyi, type.
Limnohyops laticeps, type.
Mesatirhinus megarhinus, type.
?Telmatherium cultridens.
Other mammals :
Hyrachyus princeps (cursorial rhinoceros).
Patriofelis ferox (catlike creodont).
Isectolophus latidens (tapir).
Uintatherium robustum (four-horned amblj-pod) .
Notharctus crassus (large lemuroid).
Pantolestes natans (aquatic insectivore) .
Homacodon vagans (primitive artiodactyl) .
LTintatherium mirabile (amblj'pod uintathere).
Orohippus sylvaticus (primitive equine).
Bridger C. — The lowest beds of the horizon Icnown
as Bridger C are exposed at the foot of Sage Creek
Mountain, along the southern slope of Henrys Fork
Table; also at the foot of Twin Buttes and along the
slopes north of Twin Buttes. They consist of 350 feet
of gray and greenish-gray tuffs, divided into a lower
and an upper half by the "Burnt Fork white layer"
and bounded above by the "Lone Tree white layer."
After careful analysis of the fauna of Bridger C, Mat-
thew concluded (1909.1, p. 304) that its marked dis-
tinction from the fauna of Bridger B was due to the
immigration of several new genera into the Bridger
Basin. Among these especially are the titanothere
genera Manteoceras, TelmatJieriurn, and MesatirJiinus,
which appear to be really newcomers and not in any
sense descendants of the lower Bridger genera Palaeo-
syops and Limnohyops. The two genera last named,
however, are represented in Bridger C by distinctly
new specific forms, much more progressive than those
in Bridger B. Thus Bridger C is characterized both
by marked evolutionary changes in mammals that pass
over from the lower levels and by the introduction of
a fauna that is more or less new. Of this new fauna
BNVIKONMENT OP THE TITANOTHERES
85
TJintaiherium is closely related to the ancestral BatJiyop-
sis, which is found in the Wind River Lambdotherium
zone and in the long antecedent first Wasatch zone.
The pseudotapir Isedoloiyhus is related in tooth struc-
ture to Systemodon, which is characteristic of the third
Wasatch zone. We are therefore disposed to regard
the life of the upper Bridger ( Uintatherium) zone as
the result of a local immigration from the adjacent
Rocky Mountain or Plains region into the Bridger
Basin, and not as the result of a continental immigra-
tion such as is made manifest in the lower Eocene.
Bridger D. — Upon the "Lone Tree white layer" lie
the 375 feet of strata that form Bridger D, in which
are found five faunistic levels, D 1 to D 5. The fos-
siliferous sediments of this closing period of the
Bridger consist of 350 feet of "gray and greenish-gray
sandy and clayey tuffs, with one or more ash beds,"
including the upper "white layer," which lies about 75
feet below the top of the formation. Among the tita-
notheres of this zone are descendants of species of
Palaeosyops, Limnohyops, Manteoceras, and Telma-
therium, which continue to increase in size and which
represent advancing mutations that are exhibited in
the comparative measurements shown in the tables on
pages 304, 313, 341, 364. It is noteworthy that there
is no very marked faunistic change in the species of
titanotheres that persisted from Bridger C to Bridger D.
For example, Manteoceras manteoceras persists from the
lower to the higher levels, and Mesatirhinus peter soni is
recorded in both C 2 and D 3. Exceptions to this
slow evolution are seen in two species — Palaeosyops
copei, which represents in certain characters an ad-
vanced stage of evolution allied to a stage found in the
lower sediments of the Washakie Basin, and Telma-
therium validum, assigned to Bridger D, which shows a
distinct advance upon Telmaiherium cultridens, as-
signed to Bridger C 5.
Bridger E. — Bridger E is theoretically correlated
with Washakie B and Uinta B (upper Eocene). The
topmost beds of the Bridger formation, 500 feet thick,
include sediments that are almost barren of fossils,
but the few fragments of mammals they have yielded
are of undoubted Bridger age. The 500 feet of soft
banded tuff containing at intervals thick layers of
volcanic ash indicate increasingly active volcanism.
The layers of gypsum found at this horizon were
probably deposited in playa lakes (Sinclair, 1906.1),
like those in the Humboldt Basin of the present time.
The dark-red bands in Bridger E may indicate an arid
climate. The correlation of Bridger E with Washakie
B, to the east, is purely conjectural, for neither con-
tains determinable remains of mammals. Matthew
(1909.1, p. 306) attributes the paucity of life in this
zone to violent volcanic eruptions, observing that the
thick and generally unsorted beds of ash indicate great
volcanic activity and that the large amount of gyp-
sum and the absence of fossils might be due to the
consequent destruction of vegetal and animal life,
which converted the region into a barren plain that
was alternately submerged and desiccated.
The UintatJierium zone in the Washakie Basin
(Washakie A) is described on pages 85, 87, in the
description of the deposits of that basin. The barren
deposits in the Uinta Basin (Uinta A) that correspond
to the Uintatherium zone are described on pages
91-92, in the description of the Uinta Basin.
WASHAKIE BASIN, WYO.
STRATIGRAPHY OF THE BASIW
Deposits and faunal zones. — The Washakie Basin
lies about 50 miles east of the Bridger Basin, and the
two contain similar volcanic sediments. The basin
was described by Hayden in 1869-70 (1871.2, p. 73),
and more fully by Cope in 1873 (1873.4). Its fau-
nistic levels were studied by the Princeton expedition
(Osborn and McMaster, 1881.8) and by expeditions
of the American Museum of Natural History, under
Wortman (1893, 1895) and Granger (1906). Granger
(1909.1, pp. 13-32) gave the first complete and accu-
rate description of the geology of the Washakie Basin
in his "Faujial horizons of the Washakie formation
of southern Wyoming" (1909.1, pp. 13-32). King
treated the deposits of the Washakie Basin as of
Bridger age and of lacustrine origin. Osborn (Osborn
and McMaster, 1881.8) favored the theory of separate
deposition, and Scott (1899.1) showed that where the
fauna of the Washakie Basin departs from that of the
Bridger it approaches that of the Uinta. The dis-
covery of the true upper Bridger fauna in horizon A
of the Washakie Basin was due to the American
Museum expeditions of 1893, 1895, under Wortman.
The Washakie Basin, with its vivid coloring and its
alternation of hard and soft layers of tttft' and sand-
stone, affords the most picturesque geologic views to
be found in the Rocky Mountain Eocene basins.
Haystack Mountain ("Mammoth Buttes" of Cope),
a long ridge of badlands near the north end of the basin,
which in places rises 400 feet above the plain, forms
the northern border of an extensive semicircular
"central basin" that has the appearance of a gigantic
crater. The floor of this basin is rather level and
regular, being broken only by a few low tables and
buttes, which have long been preserved from erosion
by their capping of hard sandstone, though their
sides are trenched by innumerable deep, vertical-
walled canyons, which present a great variety of
architectural forms that are illuminated by brilliant
coloring.
Washalcie A {TJintaiherium zone, middle Eocene). — '
The "lower brown sandstone" of the Washakie Basin,
known as Washakie A (fig. 60), contains the fauna of the
Uintatherium- Manteoceras- Mesatirhinus zone. It was
deposited contemporaneously with the upper Bridger
(Bridger D), to the west, and probably with the non-
86
TITANOTHEKES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
3 lIj
u
z
UJ
u
O
Ld
U
-J
Q
Q
D3
TITANOTHERES GEOLOGIC SECTION OTHER MAMMALS
Pa/aeosyops robustus
Mesai/rh/nusjan/us
Pa/aeosyops robusius
Pa/aeosyops cope/, type
" robustus
Mesat/r/j/nuspetersor?/, type
Pa/aeosyops /e/dy/
Manteoceras manteoceras
Limnohyops /at/ceps
Pa/aeosyops /e/'c/yJ
No t/i a/ 'c/./JS CI -assi fs
Pairiofclis fcro.x-
Z//'/ilcU/ior/u/n
leic(ya/iurrh
(?JJsec/o/op/ias latldens
~~ Uintat/ieriL/yn.
HENRYS FORK
H/LL
L ONE TREE WH/TE Z A YER
UirifafJteri'iim att/rrps
'^'Elac/i oce? -as "par vu/ri
C5
fPj Pa/aeosyops /e/oy/
'Je/mather/umcu/ir/dens,type
/^esat/rh/nus megarb 's
xoTie --
Pa/aeosyops /e/dyi, type
CD
C3
C2
BURNT FORK WH/TE LAYER
? Limnohyops /at/ceps,type ^L,^,j^j_ i— j_j.^'
Te/matber/um cu/tr/c/ens
/^esatirb/nus peterson/
Pa/aeosyops granger/.tjpe
?/i^anteoceras manteoceras
Homacodon vaf/ans
Jfyraehyus inipe/ia/L-i
Oro/i/ppiis sy/vat/C7/s
COTTONWOOD W/i/TE LAYER BENCH
UintaiJi eriurn
^^ B5
_SANO^.O.fii£
Palaeosyops paZudosits-
Ir^roTiippiis :zoTie
\ ^ ' / COTTONWOOD
CREEK BENCH
Figure 58.— Section of the upper part of the Bridger formation in the Bridger Basin, Wyo.
Shows the vertical distribution of the titanothere species on the left, the principal geologic features in the center, and the distribution of the
other species of mammals on the right. Principally after Osborn, Granger, and Matthew.
ENVIRONMENT OP THE TITANOTHERES
87
fossiliferous Uinta A, to the south. Its contempo-
raneity with Bridger D is established through the
common presence of the following species:
Uintatherium robustum.
Uintatherium mirabile.
Manteoceras manteoceras.
Notharctus tenebrosus Leidy.
Hyrachyus princeps Leidy.
Sinopa.
Palaeosyops copei Osborn.
Mesatirhinus megarhinus.
Mesatirhinus petersoni.
Hyopsodus.
Paramys.
I bench, which constitutes the lower rim of the basin
I on its northern border. This "lower brown sand-
I stone" passes at a low angle southward beneath the
floor of the basin. Below it, and apparently conform-
able with it, lie gray sandy shales, which are pro-
visionally referred to the Green River but which
were probably laid down in lower Bridger time
(Bridger A and perhaps Bridger B). As these deposits
I show no marked evidence of erosion it seems probable
fShale?
Nos. 35,36,37
[Sandstone
Nos. 25, 26a, 26 b
Nos. 20,21, 22
STACK MT^
LOWER BROV
Figure 59. — Diagrammatic vertical section of deposits near Barrel Springs, Washakie
Basin, southern Wyoming
Shows the alternation of tuffs, siliceous, calcareous, and sandstone materials. Johannsen (1914.1), after
Granger, with modifications. The numbers refer to lithologic specimens examined by Johannsen.
This fauna of the Uintatherium zone occurs in 260
feet of Washalde A, which is composed largely of
altered eruptive rocks, probably dacite tuffs, of cal-
careous and siliceous shales, and of glass tuffs mingled
with grains of quartz, hornblende, feldspar, according
to the analysis of Johannsen (1914.1, p. 214).
The "lower brown sandstone" layer yields a rich
fauna of uintatheres. This layer forms a persistent
that the Washakie Basin was filled wi^h a lake in
Green River time whUe Bridger A was being deposited
to the west.
Washakie B {Metarhinus and Eohasileus-DolicJio-
rhinus zones, upper Eocene). — The upper Eocene
Washakie B horizon is described on pages 89-90, in
the description of upper Eocene faunal zones 13 and
14, to which it belongs.
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 60. — Diagrammatic horizontal section of the Washaliie Basin, southern Wyoming, from north to
After Granger (1909.1). This section stiows tlie Uintatheriiim- Manteoceras zone (Washakie A, lower brown sandstones), discovered by the American Museum in
1893; Metarhinus zone (Washalcie B 1), base of the original "Washakie" formation of Hayden and Cope; DolichoThinus-Eobasileus zone (Washakie B 2), sumrnit
of the original "Washakie" formation of Hayden and Cope; "Adobe Town," rougtJy eroded area in which Amymion antiquus, Achaenodon, etc., were dis-
covered by the Princeton espedition of 1878. The numbers show locations of lithologic specimens studied by Johannsen.
FiGTJEE 61. — Sketch map of the Washakie Basin region, in southern Wyoming
After Granger (1909.1) from Clarence King (1876.1). The shaded area is the "Washakie" formation of Hayden, mapped by King and
the United States Geological Survey as the Bridger formation.
ENVIRONMENT OF THE TITANOTHERES
Mammalian life of the WashaMe Basin
89
Washakie A (Uintatherium-Manteooeras-Mesatirhinus zone).
Washakie B 1 and B 2 (Eobasileus-Dolichorhinus zone and Meta-
These forms are found also in Bridger C and D, to the west
rhinus zone). These forms are found also in Uinta A and B,
south of the Uinta Range.
Primates :
Lemnroids
Notharctus sp.
Hemiacodon sp.
Rodents
Paramys cf. P. delicatus.
Paramys leptodus, type.
Paramys grangeri.
Carnivores :
Creodonts
Thinocyon cledentis, type.
Harpagolestes immanis. (Giant creodont of the
Patriofelis ferox.
family Mesonychidae.)
Sinopa rapax var. lania, type.
Synoplotherium lanius, type.
Miacids (doglike
Miacis washakius, type.
Limnocyon potens. (An oxyaenid creodont.)
carnivores) .
Miacis medius.
Oodectes? pugnax, type.
Condy larths
Hyopsodus cf. H. despiciens.
Uintatherium grande, type.
Ambly pods
Eobasileus cornutus, type. (Giant amblypod
Uintatherium speirianum, type.
with the front horn directly above the eyes;
first appearance.)
Eobasileus galeatus, type.
Eobasileus furcatus, type.
Eobasileus pressicornis, type.
Artiodactyls
Homacodon sp.
Achaenodon insolens, type. (First of the elo-
theres.)
Achaenodon robustus, type.
?Protylopus sp. (A cameloid form.)
Perissodactyls :
Titanotheres
Palaeosyops copei? (Last of Palaeosyops.
Palaeosyops sp.
Manteoceras manteoceras, type. ("Prophet-
horn" titanotheres.)
Manteoceras washakiensis, type.
Mesatirhinus megarhinus, type. (Ancestor of
Metarhinus earlei, type. (Fluviatile type.)
DoHchorhinus.)
DoUchorhinus hyognathus, type. (Dohchorhi-
Mesatirhinus petersoni.
nus cornutus stage.)
Dolichorhinus vallidens, type.
Rhinoceroses and rhi-
Hyrachyus sp. (Cursorial rhinoceros.)
Hyrachyus sp. (Cursorial rhinoceros of Bridge
noceratoids.
Triplopus cubitalis.
type.)
Triplopus sp.
Amynodon antiquus, type. (First of the amyno-
donts (aquatic rhinoceroses).)
Chalicotheroids
Eomoropus amarorum, type. (Forest-living
ancestral chalicothere; ancestor of Moropus.)
Lophiodonts
Helaletes sp.
Desmatotherium guyoti, type.
Dilophodon minusculus, type.
Dilophodon minusculus?
ZONES 13 AND 14: METARHINUS ZONE AND EOBASHEUS-DOIICHORHmUS
ZONE
[Uinta B 1 and Washakie B 1; Uinta B 8]
The great life division known as Washalde B, 380
feet in thickness, contains a new and dominant fauna,
which is not represented at all in Bridger D or Wash-
akie A. It is significant that this unit is divided into
two zones by its fauna, exactly as Uinta B is divided
into two zones, the Eoiasileus-DoIichorMnus zone
(Washakie B 2 = Uinta B 2), and the Metarhinus
zone (Washakie B 1 = Uinta B 1 = (in part) Bartonian
of Europe). Certain of the older mammalian families
and genera (as Uintatherium) begin to disappear and
101959— 29— VOL 1 S
new generic and specific forms replace them. Con-
spicuous among these is the amblypod Eobasileus,
first described from this region by Cope, which re-
places Uintatherium. A full list of this fauna is given
above. Among the distinctive forms are the fol-
lowing :
Eobasileus cornutus Cope,
type;
DoHchorhinus hyognathus
Scott and Osborn, type.
Uintatherium speirianum
Osborn, type.
Triplopus cubitalis Cope, type.
Eomoropus amarorum Cope,
type.
Amynodon antiquus Scott and
Osborn.
Achaenodon insolens Cope.
Metarhinus earlei Osborn,
type.
90
TITANOTHEKES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Washakie A is characterized by "rusty brown nodu-
lar sandstones," and Washakie B by "coarse white,
pink, and sabnon-colored sandstones" and by "the
extremely coarse green sandstones or feldspar con-
glomerates. The rocks first recognized as sandstones
GEOLOGIC LEVELS OFSPECIES
SUMMIT OF HAYSTACK M
Zp^y\Eobas/7eus comufus, type
"°?S-<5-v>.\ EobcLsileiLS-
ict)_i-s^E=^ Dolichorhimis
zone
ZONAL LEVEL APPROXinATE
Eomoropus amarorum, type
Lepforeodon marsh/, type
\Do/ichorhinus hyoqnathus, type level
(A.M. No. 13164, Co/U906)
'■ a.^i^?l%'^'^^^i^hMelarl>/nus earlei, type
L£V£L OF
S^r^ A my no don
_=ti -"^ —^''^pnt/quas type
ADOBE TOW/V
Achaenodon insolens, type
-■-'■'■ -~ '^-^Manleoceras washak/ensis, type level
Winlalherium speirlanum, type level
Mesatirhinus megarhinus
type
Palaeosyops cope/
Manteoceras manteoceras
type
Uintafherium sp. div.
Figure 62. — Columnar section of the Washakie Basin, Wyo.
life zones
Shows the principal genera of the lower and upper life zones and the actual level of certain characteristic species.
Chiefly after Granger (1909.1). This section includes the Uintaiherium zone (Washakie A), lower brown sand-
stones; Meiarhinus zone (Washakie B 1); and Eobasileus-DolichorMnus zone (Washakie B 2) , upper gray-green
beds. ' Numbers in column show position of lithologic specimens examined by Johannsen.
prove to be interspersed with dacite and glass tuffs."
(Johannsen, 1914.1, p. 215.) The sandstones, which
were derived from granite by erosion, consist of grains
of quartz, hornblende, and feldspar embedded in a
shghtly devitrified groundmass. (See PI. IX.)
The composition of these sediments indicates the
presence in this region of active volcanoes, which
were discharging great clouds of dust. Unlike the
sediments of the Bridger Basin the sediments of the
Washakie Basin were deposited in rather turbulent
water and contain none of the
"white layers" that indicate
the still water that prevailed
in the upper Bridger. Turbu-
lent water is not favorable
to the preservation of the
remains of small mammals.
Only one of the smaller
perissodactyl ungulates
(Triplopus) has been found,
and no remains of Equidae.
The first aquatic rhinoceroses
{Amynodon) belong to a river-
frequenting type; the first of
the entelodonts {Achaenodon)
is also a river-frequenting
form; the first of the forest-
dwelling chalicotheres {Eomo-
ropus) also occurs. Thus the
Washakie Basin has preserved
for us mainly the larger
swamp and river-border fauna
but has yielded little record
of either the arboreal or
plains -living cursorial fauna
of the time.
In the fauna of the Wa-
shakie Basin (a list of which
is given in the table on p. 89)
the large hoofed animals pre-
dominate, especially those
adapted to stream borders,
swampy land, rivers, and
streams. A small fauna of in-
sectivores, lemuroids, carni-
vores, and ancestral artiodac-
tyls doubtless abounded, but
the environment was unfavor-
able to the preservation of
such remains, and the micro-
fauna has been found only
rarely. The small titanothere
MetarUnus is highly distinc-
tive of this Washakie B 1 life
zone. (Kiggs, 1912.1.)
Uinta .B.— Exactly the
same physiographic condi-
tions prevailed at the same time in the great basin
south of the Uinta Mountams while the sediments
known as Uinta B were being deposited. These sedi-
ments, which have a combined thickness of 800 feet,
contain exactly the same riparian fauna, including a
35), showing
ENVIRONMENT OF THE TITANOTHERES
91
AMVNODON SANDSrONE
large number of identical species, and therefore con-
stitute an extension of the Eobasileus-DolichorMnus
and Metarhinus life zones to the south. The fauna
and deposits of Uinta B are more fully described on
pages 91-99, in the description of the Uinta Basin.
UINTA BASIN, UTAH
PHYSIOGRAPHIC, CUMATIC, AND VOLCANIC CONDITIONS IN THE UINTA
BASIN DURING MIDDLE (?) AND LATER EOCENE TIME
It is a striking fact that the later Eocene sediments
in the Uinta Basin are composed mainly of altered
eruptives, probably dacite
tuffs, as indicated by analyses
of nine samples by Johannsen
(1914.1, pp. 212-214). The
rocks of the lower levels ^
described as "brown sand-
stones" comparable in litho-
logic appearance to Washakie
A, contain a large element of
tuff and consist microscop-
ically of irregularly broken
and rounded fragments of
quartz, lime-soda feldspar,
hornblende, biotite, and frag-
ments of andesite or basalt in
a brown groundmass, which
is chiefly chlorite but contains
some calcite. On the lower
levels (in Uinta A) brown is
the prevailing color, as in
Washakie A. In Uinta B
sediments of this color pass
into pinkish-brown and red-
dish-brown sediments, and in
Uinta C into pale-green and
gray fine-grained rocks con-
taining considerable glass.
Many rocks that look like
sandstones prove under the
microscope to resemble flow
breccias.
Uinta A as now defined is entirely unfossiliferous
but is here correlated with the middle Eocene fossil-
iferous horizon A of the Washakie Basin {Uintathe-
rium zone).
Uinta B 1 (in some previous reports included in
Uinta A) contains a rich river-border fauna, like that
of Washakie B 1.
Uinta B 2 (formerly constituting all of Uinta B)
contains a larger land and river-border fauna, like
that of Washakie B 2.
BARREN
{DiplacodoTV-
EpUtippus
zone
EobcLsiZeus -
DolichorTuruLS
zoTte
TYPICAL U/NTA MEADOW FAUNA
TRANSITION FAUNA
^:metarhinus sandstone"
'XfluviatiTe' ^=r
£^?^ ^_^1 MetarhiniMS
FLU VI ATI LE FAUNA
Figure 63. — Diagrammatic section of the Uinta formation exposed in tiie nortii wall of
White River Canyon 3 miles below mouth of Evacuation Creek, Utah
GEOLOGIC HORIZONS IN THE UINTA
BASIN
Uinta fauna of Marsh 100 feet above "Amynodon sandstone."
The deposits of horizons
A and B of the Uinta Basin
are not those of the typical Uinta formation of Marsh
(1871.3), of King (1878), or of Scott and Osborn
(1891.1), all of which belong to Uinta C, the Diplaco-
don zone; they form the lower part of the section
(Uinta A and Uinta B), determined by the American
Museum expedition of 1894 under Peterson (Osborn,
1895.98) and successively explored with remarkable
results by Peterson, Douglass, and Riggs, whose obser-
vations and exact records of the vertical distribution
of genera and species have firmly established the
stratigraphy of the Uinta Basin section as presented
in Figure 65. (See PL IX.)
After observations of Peterson, Douglass, and Riggs. Uinta A, columnar sandstones, unfossiliferous; Uinta B 1, MetarUnus
zone capped by ** Metarhinus sandstone," containing a fiuviatile fauna; Uinta B 2, Eobasileus-DaUchorhinus zone, capped
by "Amynodon sandstone," containing a transition fauna; Uinta C, Diplacodon-EpiMppus zone, containing the typical
Uinta C contains the typical Uinta (Diplacodon)
fauna.
The sediments in the Uinta Basin between the
Diplacodon zone and the Green River formation were
classified by White (1878.1, p. 37) as Bridger, although
no fossils were found in it, and wore treated as con-
temporaneous with the Bridger deposits. We now
know that the sediments that form Uinta B were cer-
tainly laid down after Bridger C and D had been
deposited, but they may have been contemporaneous
with the unfossiliferous Bridger E. During the
American Museum explorations of 1893-94 Peterson
92
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
discovered 150 fossil mammals in the lower series, I by the subsequent explorations and publications of
which were first correlated by Osborn (1895.98, p. 72) 1 Douglass (1909.1) and Eiggs (1912.1). The strati-
N.
Mesatirhirvus superior, type
Metarhinus ripartus , •■
r Sthenodectes
Anvynodon.CF>inte^-medzzts A
Figure 64. — Section of the Uinta formation (No. 10, fig. 35) from Kennedy's Basin to White River
Canyon, Utah
This section includes Uinta A, the barren sandstones; Uinta B 1, the Metarhinus zone capped by prominent bluffs of "Metarhinus sand-
stone"; above this Uinta B 2, the Eobasileus-Dolichorhinus zone, capped by the "Amynodon sandstone." After E. S. Eiggs (1912.1); see
also F- B. Weeks (1907.1).
with the typical "upper Washakie," now known as
Washakie B. The determination of the stratigraphy
as well as the faunistic succession has been modified
graphic order of the later Eocene deposits of the Uinta
Basin and the correlated fauna may be presented as
follows :
Later Eocene deposits and fauna in the Uinta Basin, Utah
Formation and nature of deposits
Geographic conditions and mammalian fauna
Uinta of King, Marsh, and White: Diplacodon elatus beds of
Marsh; horizon C of Peterson, Douglass, and Riggs. Dacite
tuffs and sandstones, grayish and greenish. Ferruginous.
Thickness, about 600 feet.
Uinta B 2 of Peterson and Osborn: Doliehorhinus cornutus
zone of Osborn (1895.98). Amynodon beds of Riggs (1912. 1,
p. 22). Coarse brownish dacite tuffs and sandstones,
capped at the summit by the "Amynodon sandstone,"
immediately underlying Uinta C. Thickness, 285 feet.
Uinta A of Peterson and Osborn, in part [Telmatotherium
megarhinum beds of Osborn = Metarhinus fluviatilis zone,
Osborn, upper Metarhinus zone of Riggs]: Capped by the
"Metarhinus sandstones" of Riggs, with underlying coarse-
grained brownish dacite tuffs and sandstone ledges; channel
beds, varying in thickness from 5 to 30 feet, containing
abundant remains of Metarhinus. Thickness, 266 feet.
Uinta A of Peterson and Osborn, lower levels (lower Metarhi-
nus zone of Riggs) : Capping of columnar sandstones, under-
lain by friable sandy shales, interspersed with ledges.
Thickness, 585 feet (Douglass, 1913). Unfossiliferous.
Underlain by Green Ri\'or formation.
Meadow, forest, and river fauna. Large titanotheres: Diplacodon
elatus, Protitanotherium emarginatum, etc. Artiodactyla:
Protoreodon, Leptotragulus, primitive camels. Small equines
(Epihippus uintensis). Other large and small members of the
true Uinta fauna. No traces of Amblypoda.
Fauna chiefly stream border and fluviatile and some small
forms. Last uintathere (Eobasileus). Aquatic rhinoceros
(Amynodon intermedius) abundant. Entelodonts (Protelothe-
rium uintense). Rare cameloids (Protylopus) . Rare eden-
tates (Stylinodon). Numerous long-headed titanotheres (Doh-
chorhinus cornutus, D. fluminalis, Sthenodectes). In the
upper levels, first long-horned titanothere (Eotitanotherium of
Peterson) ; ancestral Symborodon-like titanotheres (Rhadinorhi-
nus). Titanotheres e.xtinct at this level or not recorded from
it are Mesatirhinus and Metarhinus.
Abundant fluviatile and forest fauna, of small variety. Tita-
notheres: SmaU lowland varieties of Metarhinus very abund-
ant, including several distinct specific forms; also the long-
headed Doliehorhinus superior, the short-headed Sphenocoelus,
Metarhinus earlei, M. riparius, M. fluviatihs, Rhadinorhinus,
Doliehorhinus longiceps, an ancestral form of Dohchorhinus
cornutus. The amblypods Eobasileus or Uintatherium. The
large creodont Mesonyx obtusidens.
No fossil mammals certainly recorded by Peterson, Douglass, or
Riggs from this level.
ENVIEONMENT OF THE TITANOTHEEES
93
U
DQ
CD
<
h
Z
100
300'
400'
Dolichorhinus
fluminalis
Dolichorhinus y-
cornufus, type '
Sfhienodectes
incisivus, iype
Dolichorhinus
heterodon
Dolichorhinus
hyognafhus
(cornulus)
Sfhenodectes
incisivus
R had in orhinus
diploconus
Dolichorhinus
longiceps, type
Dolichorhinus I =-,^^-==2:=".
T Diplcuxtdort
Amynodon skel-.ATn.Mus. N9J933
26"ATrvynocLoTh sandstoTve^j^ArrvyTzodon, irvtermediics
ProtelotheriuTTh idntense
; Eobasileus-
400
super/or, type
Melarhinus
n'parius, lype
Melarhinus
earlei
Telmalolherium
D olich orhinus
longiceps
Rhadinorhinus
abbotti
Melarhinus
fluvialilis, lype
Sphenocoelus
uinlensis, lype
Melarhinus
cristalus, lype
(?Dolichorhinus
longiceps)
Melarhinus
ripar/us
1:
I EobasileiLS
< StyLLnodorv
I ProtylopzLS
sWidstone^^ Harpccg'olestes
Eobasileits iiznte.nsis, type
FieldMus. 12170
iEobcLsiLeiis
Triplopics
MesoTvyx obtusidens
Crocodilus
? TriplopiLS
i<
500
NO MAMMALS RECORDED
Figure 65. — Section of the Eobasileus-Dolichorhinus and Melarhinus zones in tlie Uinta Basin, Utah, show-
ing stratigraphic distribution of species of titanotheres
The species of titanotheres are shown in the left-hand column, the geologic strata in the middle column, other characteristic mammals in the
right-hand column. After observations made by Peterson, Douglass, Eiggs, and Osborn.
94
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBEASKA
The researches of Peterson, Douglass, Riggs, and
Osborn prove that Washakie B and Uinta B comprise
two distinct faunal divisions — a lower, Uinta B 1
{MetarTiinus fluviatilis, M. earlei zone), probably cor-
responding with the lower levels (B 1) of Washakie B,
and an upper, Uinta B 2 {Eohasileus-DolichorJiinus
(cornutus) Tiyognathus zone), probably corresponding
with the upper levels (B 2) of Washakie B.
UINTA B 1 (METARHINUS ZONE = ZONE 13)
Riparian fauna. — The fauna of the MetarTiinus zone
was evidently that of a riparian lowland and was in
part fluviatile or aquatic, as is indicated by its adapta-
tions to aquatic and lowland life, which are inde-
pendently developed in members of several different
families. These adaptations are indicated by some of
the specific names, such as MetarTiinus riparius, M.
fluviatilis, two diminutive titanotheres, and DolicTio-
rTiinus fluminalis. The animal last named is a short-
limbed swamp-dwelling form, a fit companion of the
river-seeking rhinoceros Amynodon intermedius , which
begins to show aquatic adaptations in the structure of
the orbit. The generic aspect of this fauna is almost
identical with that of Washakie B, with the single
exception that near the summit of Uinta B the ente-
lodont ProtelotTierium replaces AcTiaenodon of Washa-
kie B. The fauna contains a single new titanothere,
RTiadinorTiinus , which is closely related to MetarTiinus.
Some of the specific forms are identical with those of
Washakie B and some exhibit more recent phases of
evolution, which may be represented in the unfossilif-
erous upper levels of Washakie B. We consequently
reach the broad generalization that Washakie B 1 and
Uinta B 1 were not only contemporaneous sediments
but that they indicate the prevalence of similar
physiographic and climatic conditions at this time on
the north and the south sides of the Uinta Range.
River-cTiannel fauna. — Remains of the small titano-
there MetarTiinus have been found in ancient river
channels, as determined by Riggs. This genus is by
far the most distinctive fossil of this life zone and is
apparently confined to it, although at certain levels
primitive species of DolicTiorTiinus {D. longiceps) are
found in equal abundance (Riggs, 1912.1, p. 20).
This life zone, which is 400 feet thick, is composed
chiefly of massive ledges of sandstone alternating
with layers of sandy shales or indurated clays. In all
the ledges there are traces of cross-bedding, and at many
places there are beds of coarser river sand containing
pebbles of quartzose material, sandstone, and clay
shale. In these beds are found disarticulated bones
of mammals, as well as the branches and at some
places the trunks of trees, all pointing to the action of
swiftly flowing streams that swept through a flood
plain. Many skulls are found embedded in gravel,
with their narial or orbital cavities filled with pebbles
such as could be carried only by rapidly flowing water.
Another evidence of stream action lies in the dissocia-
tion of the parts of single skeletons. Whole skeletons
have exceptionally been found but little disturbed,
lying in a fine-grained homogeneous sandstone, ap-
parently deposited in quieter water, such as deep
pools or eddies. Remains of the long-headed titano-
there DolicTiorTiinus are found only in the heavy sand-
stones, so that this animal was apparently confined to
the vicinity of streams. Supposed river-frequenting
species of MetarTiinus, always found in sandstone, in-
clude M. fluviatilis, M. riparius, and M. earlei. The
species last named is found also in the lower levels of
Washakie B, north of the Uinta Mountains. As we
ascend in this MetarTiinus zone we find, according to
Riggs (1912, p. 24), increasing numbers of upland
forms. The "MetarTiinus sandstone" ledge that caps
this zone has yielded the type specimens of Doli-
cTiorTiinus superior, MetarTiinus riparius, and M. earlei,
the last-recorded appearance of these animals in the
Uinta Basin.
UINTA B 2 (EOBASILEUS-DOLICHORHINUS ZONE = ZONE U)
In the beds of the Eohasileus-DolicTiorTiinus zone the
river sandstones and channel deposits gradually give
place to shales and clays, indicating physiographic
changes in this part of the Uinta Basin. In the lower
100 feet of bluish or grayish shales, which are overlain
by 40 feet or more of fine red clays, little evidence of
mammal life is found, but certain thin beds contain frag-
ments of Eohasileus. As we rise in the formation the
gray "clays" begin to yield a mixed fauna of lowland
and plains forms, including Protylopus and Stylinodon,
together with remains of DolicTiorTiinus (cornutus)
TiyognatTius and Amynodon intermedius. The massive
"Amynodon sandstone," which forms the summit of
this life zone, yields the type specimen of the long-
headed titanotheres DolicTiorTiinus (cornutus) Tiyog-
natTius, D. fluminalis; also of StJienodectes incisivus.
This is the last appearance of the genus DolicTiorTiinus
in the Uinta Basin. Doubtless the massive "Amyno-
don sandstone" terminated the active period of
fluviatile and flood-plain deposition in this locality.
The D. (cornutus) TiyognatTius zone yields the large
enteledont RrotelotTierium uintense, which is inter-
mediate between AcTiaenodon insolens of Washakie B
and ElotTierium of the White River group. This
sandstone contains also the lophiodont DesmatotTie-
rium guyoti, which is a forerunner of Colodon of the
White River (Oligocene) group.
ZONE 15: DIPLACODON-PEOTITANOTHERIUM-EPIHIPPnS ZONE
[Uinta C I; Xudian of Europe]
To zone 15 belong the Uinta of King and Marsh, the
"Brown's Park beds" of Powell, and the Uinta(?) of
the Beaver Divide, Wind River Basin. This zone
(Uinta C 1) is correlated with the European stage
that was named Ludian, after the "marnes de Ludes"
in the Paris Basin, a stage typified by the "gypse de
Montmartre," made famous by the classic researches
ENVIRONMENT OF THE TITANOTHEEES
95
of Cuvier. The lower Ludian yielded the type speci-
men of the equine LophiotJierium, a horse in the same
stage of evolution as the diminutive American Epihip-
pus of the Uinta. The American beds contain a rich
titanothere fauna. They include the "Diplacodon
beds" of Marsh (1877.1, p. 354) and contain the
robust titanothere Protitanotherium, which is inter-
mediate between the "prophet-horn" Manteoceras and
the horned titanotheres of the lower Oligocene; also the
type of Protitanotherium superbum, an animal greatly
exceeding in size the earlier Oligocene titanotheres.
Of great interest is the survival of the ancestral genus
Manteoceras in the species M. uintensis, a genus first
occurring in the upper Bridger, and the Bridger genus
TelmatJierium in the species T. ultimum.
plains fauna (Hypertragulidae and Camelidae) rep-
resented respectively by genera believed to be ancestral
to the tragulids (Leptotragulus) and to the camels
(Protylopus, Camelomeryx) ; also members of the
oreodonts (Protoreodon), and the agriochoerids {Agrio-
choerus). The diminutive tylopod Protylopus has
been selected as the possible ancestor of the great
family of American camels.
The fauna found near the base of the true Uinta
thus includes a considerable light-limbed meadow
and plains element, transitional to the plains fauna
of the lowest Oligocene of the White River group.
The occurrence of this fauna near the base of Uinta
C indicates that the Uinta formation probably passes
up into lower Oligocene time. The beds represent
Figure 66. — Badlands near the mouth of White River, Uinta Basin, Utah (No. 10, fig. 35)
Wortman and Peterson collecting. This view shows the typical Uinta formation (Uinta C 1) and the Diplacodon zone in the fore-
ground, with Uinta C 2 (unfossiliferous) in the distance. After Osborn (1910.346). Am. Mus. negative 17663. Compare Plate
XII, B.
The amphibious rhinoceros Amynodon occurs in the
species A. antiquus. It should be noted that the
remains of all these large mammals were found not
far above the base of Uinta C, and that all the speci-
mens in the chief collections of small Artiodactyla
(Protoreodon, Leptotragulus) and of Perissodactyla
(Triplopus, LopModon, Isectolophus , a tapiroid, Epi-
Jiippus) (Peterson) were obtained from the lower
levels of Uinta C. With Epihippus were found the
only primate that has been discovered in the Uinta
Basin, NotJiarctus? uintensis, a lemuroid, and the
supposed condylarth or insectivore Hyopsodus. The
few surviving ancient creodonts are represented by
Oxyaenodon and by the giant Harpagolestes uintensis.
Especially important is our first knowledge of the
a considerable change in local physiographic condi-
tions from those of Uinta B. The fine-grained soft
material, composed of altered eruptives, probably
dacite tuffs, is of much the same texture as the char-
acteristic " Titanotherium beds" (Chadron formation)
of South Dakota, except as to its color, which is brick-
red ; in fact, a reddish tinge prevails throughout the
sediments of Uinta C
During this latest part of the Eocene epoch the
titanotheres of the Rocky Mountain basin south of the
Uinta Mountains distinctly approach in character the
titanotheres of the Great Plains. The appearance
in this layer and near the summit of Uinta B of two
or three entirely new forms of titanotheres (Eotitano-
tJierium, Diplacodon, Protitanotherium) is less indica-
96
TITANOTHBEES OF ANCIENT AVYOMING, DAKOTA, AND NEBEASK.V
tive of new migrations into the Rocky Mountain
region than of new physiographic conditions favor-
able to the fossilization of some of the upland and
meadow Herbivora that had been evolving in the
adjacent Plains region but had not mingled with the
fluviatile, swamp, and forest-border fauna that
inhabited the Uinta Basin in Uinta B time.
The following summary of the later Eocene faunas
of the Uinta Basin should be examined in connection
with Figures 63-66.
Composite section of mammalian faunas of tlie late?' Eocene sediments of the Uinta Basin
[After Peterson, Osborn, Riggs, and Douglass]
Uinta C (true Uinta formation = Diplacodon
zone); 600 feet. Badlands like those of
South Dakota, but of brick-red color.
Brownish and reddish ferruginous sand-
stones and clays (Peterson).
Uinta B 2 (Eobasileus-Dolichorhiniis zone) ;
300-400 feet. Section along gilsonite vein
No. 2 (Riggs). Includes "Amynodon sand-
stone," gray and greenish clays, ferruginous
sandstones, bluish and greenish shales.
Two red layers with fossiliferous sandstone
between (Douglass). Supposed base of
horizon B 2.
Uinta B 1 (Metarhinus zone = upper Meta-
rhinus zone of Riggs) ; 400 feet. Section on
divide between White River Canyon and
Coyote Basin (Riggs) . Also section 3 miles
below mouth of Evacuation Creek (Riggs) :
"Metarhinus sandstone."
"Eobasileus sandstone" = massive ledges
of reddish sandstone, alternating with
layers of sandy shales.
Indurated clays.
Uinta A (lower A of Peterson, lower Meta-
rhinus zone of Riggs) ; 500 feet (Riggs) ; 585
feet (Douglass). Section in north wall of
White River Canyon (Riggs) :
"Columnar sandstones, about 300 feet
thick, weathering as bold cliffs, or but-
tresses along the canyon of White
River. Color slightly more grayish
than the underlying shales, but brown
predominates (Riggs).
"Two hundred feet friable sandy shales,
weathering in steep slopes, with hori-
zontal outcroppings of nodular or
sandy layers, or by massive ledges of
limited extent" (Riggs).
Green River (?) formation. Shaly gray sand-
stone of lacustrine origin.
Titanotheres
Diplacodon elatus, type.
Protitanotherium emarginatum.
Protitanotherium superbum, type.
Telmatherium ultimum, type.
Manteoceras uintensis, type.
Eotitanotherium osljorni, type.
Dolichorhinus cornutus ( = hyogna-
thus), type.
Dolichorhinus fluminalis, type.
Dolichorhinus heterodon, type.
Sthenodectes incisivus, type.
?Rhadinorhinus diploconus, type.
Dolichorhinus longiceps, type (near
base) .
Dolichorhinus superior, type.
Metarhinus riparius, type.
Metarhinus earlei.
"Telmatherium," large jaw.
Dolichorhinus longiceps.
Rhadinorhinus abbotti, tj-pe.
Metarhinus fluviatilis, tyjje.
Dolichorhinus longiceps, skeleton.
Metarhinus riparius (from base).
Sphenocoelus.
Heterotitanops parvus. (=?Meta-
rhinus), from base of B 1.
None.
Other forms of life
First oreodonts.
Cameloids.
Dichobunids.
Aquatic rhinoceros (Amynodon).
Small perissodactyls (tapiroids, lophio-
donts, Epihippus).
Last creodonts, Meson\-chidae and
Oxyaenidae.
No amblypods found.
Last ambly pod (Eobasileus).
First cameloid (Protylopus).
Last taeniodonts (Stylinodon).
Aquatic rhinoceros (Amynodon).
Primitive entelodont (Protelotherium
uintense) .
Remains of plants and fishes; oc-
casionally plentiful in sandstone
ledges (Peterson) .
Last primates ("Notharctus").
Giant creodonts, Mesonychidae (Har-
pagolestes) .
Giant creodont (Harpagolestes).
Crocodiles.
Turtles.
Giant amblypod (Eobasileus).
Light-limbed perissodactyl (Triplopus).
Creodonts, Mesonychidae (Mesonyx).
None.
No mammals (Peterson, Riggs).
Fragments of turtles.
Unios.
Remains of plants and occasionally
large tree trunks in sandstone ledges
(Peterson).
Remains of plants, fishes, and insects
in the shales (Peterson).
ENVIEONMENT OF THE TITANOTHERES
97
SUMMARY OF FAUNAS OF UINTA B AND C
Though the whole later Eocene section of the
Uinta is 1,900 feet thick it includes 500 feet of un-
fossiliferous beds both at its base and at its summit,
so that the fossiliferous beds cover only about 900 feet.
The stages of evolution are best measured in the suc-
cessive species of DolichorJiinus, which are found both
at low and at high levels in the fossiliferous part of
the section.
The archaic mammals that play so large a part
through lower and middle Eocene time diminish in
number and approach extinction at the end of Eocene
time. The numerical inferiority of the waning archaic
mammals and the rapid increase in the numbers of
modernized mammals are indicated in the following
table, prepared in 1910:
Transition in mammalian life at end of Eocene time
Genera
Species
Archaic mammals:
.2
1
5
6
Condylarthra ( H vopsodontidae)
2
5
8
13
Modernized mammals:
Primates _
2
1
3
3
9
?3
Rodentia __
3
Carnivora (Miacidae)
4
4
Perissodactyla
16
18
30
The Amblypoda culminate in the gigantic Eobasileus,
which disappears at the end of Uinta B, when the
gigantic creodont Mesonychidae and the catlike
Oxyaenidae appear for the last time. It is note-
worthy that these animals attain their largest size in
this, their waning period. The lemuroid primates
are found in greatly diminished numbers as compared
with those in the Bridger, possibly because the con-
ditions were unfavorable to the fossilization of re-
mains of arboreal animals; in fact, we know nothing
of the forest or the arboreal fauna during the entire
period of Washakie B and Uinta B because of pre-
vailing fluviatile conditions of deposition.
ADAPXrVE RADIATION OF THE TITANOTHERES IN THE UINTA BASIN
GENERA AND SPECIES HEPRESENTED
Through these 650 feet of fossiliferous sediments
the titanothere fauna of the Uinta Basin is revealed
as extraordinarily large and varied, no less than 11
genera and 22 species having been described. The
animals range in size from the small Metarhinus flu-
viatilis, some of which were not so large as a tapir,
to the huge ProtitanotJierium superbum.
The titanothere Metarhinus is abundant and char-
acteristic in Uinta B 1, ranging from the base to the
summit but not extending into Uinta B 2 as here de-
fined. (In previous reports horizon B 2 has been
included in Uinta A.) According to Riggs (1912.1, p.
27) the genus includes two phyla — the first comprising
the small MetarJiinus fluviatilis Osborn and M. riparius
Riggs, with long, narrow skull; the second including
the broad-skulled forms M. earlei Osborn (which is
also found in Washakie B) and M. cristatus Riggs.
Metarhinus was a companion of its long-skulled rela-
tive Dolichorhinus in and near the rapidly flowing
streams, its remains being usually found in coarse
and semigravelly sandstones. (Riggs, op. cit., p. 24.)
In Uinta B 2 rapid streams, apparently the favorite
haunt of Metarhinus, were less abundant than in
Uinta B 1 (Riggs, op. cit., p. 25), which partly ac-
counts for the apparently sudden disappearance of
these animals from the sediments.
Sphenocoelus uintensis, which is also probably from
the Metarhinus zone (Uinta B 1), is known only from
the hinder half of a skull. This strange animal is
clearly a member of the Metarhinus-Dolichorhinus
group and may be closely related to the long-skulled
Metarhinus riparius. The Metarhinus series as a
whole is clearly related to the older and more primitive
Mesatirhinus megarhinus of Washakie A and Bridger
C and D, which is also structurally ancestral to
Dolichorhinus .
The name Heterotitanops parvus Peterson has been
applied to the skeleton of a very young animal from
Uinta B 1. It was found, articulated, in a hard sand-
stone concretion and lower down in Uinta B 1 than
any mammalian remains heretofore described from
that horizon. (Peterson, 1914.2.) In the opinion of
Gregory the characters of the deciduous dentition
and of the facial region of the skull of this animal
indicate that it probably represents the newly born
young of some of the Metarhinus-Rhadinorhinus group.
Rhadinorhinus is distinguished from Metarhinus by
its tapering nasals and by the reduced infraorbital
process of the malar bones. One species, R. abhotti
Riggs, is found in Uinta B 1 , and another, R. diploconus
Osborn, is recorded from Uinta B 2. Riggs suggests
that Rhadinorhinus was an upland rather than semi-
aquatic form. Gregory noted in 1902 that it fore-
shadows the long-horned titanothere Megacerops
(Symiorodon) of the lower Oligocene in the abbrevia-
tion of the face and in the characters of the dentition.
The long-skulled Dolichorhinus is represented by
two species in Uinta B 1 (one of which, D. longiceps
Douglass, extends into the base of Uinta B 2) and by
four species in Uinta B 2. The most primitive species,
D. superior, is in general intermediate in structure
between the ancestral Mesatirhinus and the later
species of Dolichorhinus. The most advanced species,
D.fluminalis Riggs, is from the upper levels of Uinta
98
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
B 2. The allied D. cornutus is believed to be specifi-
cally identical with D. hyognathus of Washakie B.
In Uinta B 1 remains of Dolicliorliinus, as noted by
Riggs, are frequently found associated with those of
Metarhinus in coarse pebbly sandstone laid down in
rapid streams; but in Uinta B 2 they are frequently
found in lenticular sandstones, which were apparently
deposited in quiet water, for they show little evidence
of water currents, and which yield skulls associated
with mandibles and parts of skeletons. Riggs accord-
ingly infers that Dolichorliinus inhabited the low
grounds near quiet waters rather than the swift
currents preferred by Metarhinus.
Contrasting with the elongate and straight-sided
skull of Bolichorhinus is the broad, robust skull of
StJienodedes incisivus (Douglass), a titano there with
exceptionally massive incisor and canine teeth and
broad, low-crowned upper molars. The type skull
was found by Douglass in a thick deposit of sandstone
and small gravel, evidently of stream origin, near the
middle of Uinta B 2, whereas Riggs's specimen was
found in lenticular sandstones at about the same level.
Sthenodectes shares many characters in common with
the Bridger genera Manteoceras and Telmaiherium and
appears to be an advanced member of that macrodont
group.
With the possible exception of Rhadinorhinus all
the titanotheres so far noted from Uinta B 1 and B 2
belonged to aberrantly specialized side lines, which
are not found in later formations and apparently
became extinct.
One titanothere recorded from near the summit of
Uinta B 2, namely, Eotitanoiherium osborni Peterson
(1914.1 ; 1914.4), is highly progressive toward the giant
plains-living types of the uppermost Eocene (Uinta C)
and of the lower Oligocene. This animal, represented
by parts of two skeletons in the Carnegie Museum,
surpasses even ProtitanotTierium of Uinta C in the
development of a pair of large oval bony protuber-
ances above the eyes. One of the most remarkable
features of Eotitanotherium is the progressive sub-
molariform character of its third and fourth upper
molars, which are, indeed, slightly more advanced in
type than those of certain lower Oligocene titano-
theres. The animal was about as large as a rhinoceros,
and throughout the skeleton are mingled the earlier
characters of its Eocene predecessors with progressive,
plains-living adaptations prophetic of some of the
Oligocene titanotheres, especially those of the meno-
dontine group.
The titanothere fauna of the true Uinta (Uinta C),
though less extensive in genera and species than that
of Uinta B 1 and B 2, is none the less of prime impor-
tance to the historian of the family. Thus the titano-
theres of the true Uinta include, first, certain conserva-
tive phyla {Manteoceras, Telmatherium) , which repre-
sent the little-altered descendants of genera either of
Uinta B 1 and B 2 or of Washakie and Bridger types;
second, two very progressive and different phyla,
Diplacodon and ProtitanotTierium, of uncertain relation-
ships, which appear to be immigrants from other
localities.
ADAPTIVE RADIATION OF PHYLA
There seem to be at least four contemporary phyla,
representing wide local adaptive radiation :
First, the robust, short-limbed forms, one of which,
Manteoceras uintensis Douglass, found in gray sand-
stone in the red beds of the lower portion of Uinta C,
is considerably larger than the more primitive species
of Manteoceras in the upper Bridger and Washakie A
but is little modified otherwise. Its horn swellings,
if developed at all, were not large, and it is strongly
macrodont in type, like Telmatherium and Sthenodectes.
Second, the long-limbed, long-headed, relatively
hornless Telmatherium, which is distinguished espe-
cially by its deep malar bones and the high sagittal
crest and is represented in Uinta C by the great
Telmatherium ultimum and the gigantic T. altidens.
T. ultimum is practically hornless, having only the
slightest rugosity at the naso-frontal junction in the
type skull. Accordingly the species Manteoceras
uintensis and Telmatherium ultimum and the genus
Sthenodectes, while advancing in the direction of the
Oligocene type in various characters, are apparently
excluded from direct ancestry to the later types by
certain specializations, such as marked enlargement
of the incisors and canines, and by the lack of develop-
ment of effective horn swellings.
Third, Diplacodon elatus Marsh of Uinta C, a
progressive titanothere, which is known chiefly from
the upper dentition and takes its generic name from
its submolariform third and fourth premolars. The
precise relations of this animal are still in doubt.
The premolars and molars may have been derived
from the type represented by Rhadinorhinus diploconus
of Uinta B 2, which is the only one of the older titano-
theres that has the dentition and skull at all like
those of Diplacodon. In other respects the Diplacodon
dentition suggests that of the lower Oligocene titano-
there Menodus trigonoceras , and in still another respect
it resembles that of Eotitanotherium of Uinta B 2, though
it differs from that genus in the more molariform con-
dition of the third upper molar.
Fourth, Protitanotherium emarginatum Hatcher,
which is known from the facial part of the skull and
the lower jaw of the type specimen. It is a large
animal, which, so far as Imown, approaches the
Oligocene type of Brontops. It has oval horn swellings
which are less protruding than those of the type of
Eotitanotherium; its nasals are wide distally, in con-
trast to the tapering nasals of Eotitanotherium; its
canines are very stout and acutely conical; its stout
upper incisors form a flattened arch. Altogether it
seems to represent a phylum distinct from Eotitano-
ENVIRONMENT OF THE TITANOTHEEES
99
therium and of doubtful relationship both to earlier
or to later titanotheres, although it was possibly
derived from Manteoceras. Of the same phylum is
Protitanotherium swperhum, a gigantic animal, with a
jaw 24 inches long and premolars and molars of very
progressive type. It is much larger than the smaller
titanotheres of the lower Oligocene.
Another titanothere of uncertain relationship is
BrachydiasUmatherium from Transylvania, eastern
Hungary (now Rumania). The geologic age of this
animal is not certain, but it is in an upper Eocene
stage of evolution as compared with the titanotheres
of America.
FAUNA UNREPRESENTED
The sequence of titanothere species in the Uinta
Basin illustrates the vagaries of the fossil records of
the Rocky Mountain basin region caused by local
physiographic changes; each kind of sedimentation
exhibits only a part of the fauna. For the entire
period covered by the lower sediments of the Uinta
Basin little or no knowledge of the small terrestrial
fauna has come to light, none of the arboreal fauna,
and none of the plains and upland fauna, in contrast
with the surprisingly extensive knowledge of the
fluviatile and the swamp-dwelling fauna. Gradually
conditions changed, and Uinta B 2, as we ascend,
affords an increasing knowledge of the cursorial
meadow fauna; but it is not until Uinta C (true Uinta)
that local conditions became favorable to the pres-
ervation and fossilization of the small cursorial mam-
mals of the artiodactyl and perissodactyl divisions
of the ungulates. The sudden appearance of these
animals might be attributed to immigration, but it
is equally probable that they were all evolving in the
same region or in the adjacent Plains region. Thus
the data do not necessarily suggest immigration or
migration; these animals may have been brought into
the field of observation by changing conditions of
fossilization. The manner in which these numerous
phyla of titanotheres enter this field is shown in the
following table:
Geologic and geographic range of phyla {here "subfamilies" and certain genera) of titanotheres
[Showing their successive immigration from the north and their evolution in their i
column; the later immigrants are named in order from bottom to top.
ew habitat. The earliest immigrants are those named at the bottom ot the first
The difference in the length of the blaclc bars has no significance]
Phylum
Wind River B «
(" Lost Cabin ")
(middle Eocene)
Bridger (upper middle Eo-
cene)
Washakie
(upper
Eocene)
Uinta (upper-
most Eocene)
Chadron (Oligo-
cene)
A'
B
C =
D »
E
A '
B
A
B"
C
A
B
C
1
1
^^^
___
1
—
—
■^
----
—
^
!
» Wind River B = Huerfano A.
I Bridger A = Huerfano B.
■ Bridger 0 and D = Washakie A.
' Washakie A = Bridger C and D.
• Uinta B = Washakie B.
'Diplacodon, Eotitanotherium, Protitanotherium.
ZONE 16: THEORETIC UINTA C 2
Titanotheres have thus far been determined from
only the lower 100 feet of Uinta C. They are at
present only partly known. When fully known we
shall probably find close generic if not specific corre-
lation between the upper fauna (now unknown) of
Uinta C and the fauna of the lower levels (Chadron
A) of the White River group. The passage from
Eocene to Oligocene time probably occurs within the
period of Uinta C (true Uinta) deposition. Scott is
disposed to put all of Uinta C in the Oligocene.
COMPOSITE EOCENE AND LOWER OIIGOCENE SECTION AT BEAVER
DIVIDE, WIND RIVER BASIN, WYO.
Most of the Oligocene sediments in the Rocky
Mountain basin region have been eroded away. The
only locality where fossil-bearing lower Oligocene sedi-
ments still overlie those of the upper Eocene is on the
southern border of the Wind River Basin, Wyo.,
where the true Titanotherium zone overlies sediments
containing a fauna similar to that of the Diplacodon
zone (Uinta C). The geologic section observed at this
point by the American Museum expedition of 1909
100
TITANOTHEHBS OF ANCIENT WYOMING, DAKOTA, AND NEBEASKA
under Granger and N. H. Brown, who discovered this
fauna in 1908, is as follows:
Oreodon zone = Brule for
Summit of lower Oligocene
mation
Base of lower Oligocene, Titanotherium zone=Chadron
formation
Upper Eocene, Diplacodon zone = Uinta(?) formation —
Middle Eocene (?), unfossiliferous = Bridger (?) forma-
tion
Lower Eocene, Lamhdotherium zone = upper part of
Wind River formation
Feet
540
O/'eodoTL zone
A single tooth of either Diplacodon or Protitano-
therium has been found at Beaver Divide, Wyo.
The correlation with Uinta C rests upon Camelodon
arapahovius Granger, a species somewhat more pro-
gressive than Protylopus of Uinta C and somewhat
more simple than Leptotragulus , characteristics that
combine to place it among ancestral camels, in the
Camelidae. In certain characters it agrees with Lep-
totragulus profedus of the Titanotlierium zone of Pipe-
stone Springs, Mont. The Amy-
nodon found here agrees with the
species A. antiquus, originally
determined in Washakie B
( = Uinta B). Two specimens
of Protoreodon are referable to
P. parvus, from the base of Uin-
ta C or the summit of Uinta B.
Above this Diplacodon (?)
level is a very marked erosional
unconformity between the up-
per Eocene and the lower Oli-
gocene; broad, shallow valleys
(Sinclair and Granger, 1911.1,
p. 99), indicating fairly mature
topography, were excavated in
the sediments of the Diplaco-
don(1) zone. After these val-
leys were cut the first deposits
laid down were fine-grained
buff-colored tuffaceous shales.
In this tuff the American Mu-
seum exploring party of 1909
found a skull of Menodus heJoce-
ras, which belongs to the lower
level of the Titanotherium zone,
corresponding with Chadron A.
The volcanic ash comprising
the sediments of the Oreodon
titanothere zone, a few feet
thick, is covered with a mud
flow of volcanic material 46 feet
thick, above which lies 540 feet
of fine, wind-blown buff ash and
dust. No clays have been found
at this middle Oligocene horizon ,
which corresponds in age with
the Brule formation of the
White River group — only wind-
laid ash and coarse gravel,
perhaps deposited by torrents
during occasional heavy rains.
None of these sediments ap-
pear to have been much dis-
turbed by water, and Sinclair
Diagrammatic section of deposits at Green Cove, Beaver Divide, Wyo. ^^^ Grander (1911.1 p. 114)
Oreodon
Cyllndrodon
Caenopus
Ischyromys
Poebro therlum
Menodus he/oceras
? DiplcLcodoTz zone
Amynodon ? anfiquus
Protoreodon
Camelodon
Pro titan o ttierium
Lamhdotherium zone
Lambdotherium
Coryphodon, PlienacoduSj
hieptodon , Eohippus
Figure 67 ^^^ ^^„„ „ ^„ ,
(No. 6, fig. 35), from the Lambdotherntm zone (Wind River) to the Oreodon zone """ .""'"""" t"*h lie e that
they accumulated under a drier
(White River) at the summit
Chiefly after Granger (1910.1).
ENVIEONMENT OF THE TITANOTHEEES
101
climate than that which prevailed in Eocene time.
These upper sediments contain a true Oreodon zone
fauna.
FOURTH FAUNAI PHASE (LOWER OLIGOCENE)
LOWER OLIGOCENE MAMMALS
COERELATION OF EUROPEAN AND AMEEICAN FORMS
The lower Oligocene mammals represented by the
fossils thus far discovered are listed below.
Peculiar to Europe:
Paleotheres.
Anoplotheres.
Oenotheras.
Gelooids.
Amphicyonids.
Viverrids.
Cricetines (hamsters).
Theridomyids.
Sirenians (Hahtherium).
(Horses not recorded.)
Common to Europe and North America:
Titanotheres (central Europe).
Chalicotheres.
Rhinoceroses (aceratheres and diceratheres) .
Amynodonts.
Anthracotheres.
Suillines.
Entelodonts.
Opossums.
Hyaenodonts.
Canids (dogs).
Mustelids (martens).
Machaerodonts (saber-tooth cats).
Peculiar to North America:
Horses.
Hyracodonts (rhinoceroses) .
Oreodonts.
Camelids.
Hypertragulids.
Leptiotids.
Chrysochlorids? (inseotivores) .
Ischyromyids (rodents).
Leporids (hares).
ZONE 17: TITANOTHERIUM-MESOHIPPUS ZONE
[Chadron A, B, and C; Sannoisian of Europe]
The forms that constituted this rich world of lower
Oligocene mammalian life were distributed through
the Rocky Mountain basin region, but the sediments
that contained the fossils have been eroded away
except in a few isolated areas, such as those along
Pipestone Creek, Mont.; at Beaver Divide, Wyo.,
south of the Wind River Basin; and at Bates Hole,
Wyo. The areas in which these sediments were
deposited lie east of the Rocky Mountains, in Sas-
katchewan, North Dakota, South Dakota, and Colo-
rado. The chief fossil-bearing sediments exposed are
in the localities shown below.
Recorded thickness of the Titanotherium zone in thirteen exposures
of lower Oligocene deposits
Feet
1. Cypress Hills, Saskatchewan (Lambe, 1908) 50-500
2. Pipestone Creek, Jefferson County, Mont.
(Douglass, 1903) 300 +
3. White Butte, N. Dak. (Douglass, 1903) 120
4. Big Badlands, S. Dak. (Hatcher, Darton) (typical
area of Titanotherium zone) 180
5. Goshen Hole (Scotts Bluff), southeastern Wyoming
(Darton), maximum thickness 200
6. Hat Creek, South Fork, Cheyenne River, Dawes
County, Nebr 100 ±
102
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
ENVIRONMENT OF THE TITANOTHERES
103
7. Near Dickinson, S. Dak. (Douglass) 40-50
8. Pine Ridge, S. Dak. (Darton) 30-60
9. Beaver Divide, Wyo. (Granger)
10. Bates Hole, Natrona County, Wyo
11. Adelia, Sioux County, Nebr. (Darton), about
Between Platte River and Arkansas River drainage
(Darton). ("Monument Creek group"
of Hayden; Castle Rock conglomerate
of Richardson, 1912.1) 300
Horsetail Creek, northeastern Colorado
(Matthew,1901.1), not over
46
(?)
12.
13
between the upper fauna (now unknown) of Uinta C
and the fauna of the lower levels (Chadron A) of the
White River group. The passage from Eocene to
Oligocene time probably occurs within the period of
deposition of Uinta C. Scott is disposed to put all of
Uinta C in the Oligocene.
General Section of the Tertiary rocks of Nebraska.
100
The deposits at these localities, some of
them indicated on the accompanying map,
represent only the exposed parts of the lower
Oligocene deposits of the great flood-plain sys-
tem now known as the Chadron and corre-
lated formations, the larger part of which is
covered by the Brule and Arikaree formations.
This flood plain extends 325 miles north and
south and 300 miles east and west. We do
not know whether it was wholly continuous.
Such an area would embrace 97,500 square
miles, which would not exceed the present
Andean flood plains.
At the base of these sediments in South Da-
kota and northern Colorado there are abundant
remains of titanotheres, certain of which are
in stages of evolution no more advanced than
those found at the base of Uinta C, Diplacodon
zone. Consequently the faunistic relation be-
tween the titanotheres living in the mountain
basins and those living on the Plains remains
to be solved by future discovery. This rela-
tion may be revealed in the "missing" faunal
zone. At present we may divide the life zones,
in descending order, as follows:
17. Titanotherium zone:
Chadron C, levels 3, 2, 1:
Brontops robustus.
Menodus giganteus.
Brontotherium platyceras.
Chadron B:
Brontops dispar.
Menodus trigonoceras.
Brontotherium hatched.
Chadron A, levels 1, 2, 3:
Brontops brachycephalus.
Menodus heloceras.
Brontotherium leidyi.
16. Theoretic zone of Uinta C (upper levels, or Uinta
C 2): Unknown or "missing."
15. Diplacodon zone of Uinta C (lower levels, or Uinta C 1) :
Protitanotherium emarginatum.
P. superbum.
Diplacodon elatus.
It is very important to recall the fact that
titanotheres have thus far been determined from only
the lower 100 feet of Uinta C, that they are only
partly known, and that when fully known we shall
probably find a close generic if not specific correlation
NameB.
SUBDIVISIONS.
Thick-
LOCALITIES.
Foreig:n
Equiva-
lentB.
S
3
Fine loose sand, with some
layers of limestope, — contains
bones of Canis, Felis, Caxtor,
Equus, Mastodon, Testudo, &c..,
some of which are scarcely dis-
tinguishable from living spe-
cies. Also Helix, Physasucclnea,
probably of recent species. All
fresh water and land types.
o
On Loup fork of
Platte River ; extend-
ing north lo Niobrara
River, and south to
an unknown distance
beyowd the Platte.
a
§
>
s
s
White and light drab clays,
with some beds sandstone, and
local layers limestone. Fossils,
Oreodon, Titanotherium, Cliaro-
potamus, Rhinoceros, Anchithe-
rium, Hycenonodon, Afachairodus,
Trionyx, Testudo, Helix, Plan-
orbis, Limncea, Petrified wood,
&c. &c. All extinct. No
brackish water or marine re-
mains.
o
a
o
8
O
Bad Lands of White
River ; under the
Loup River beds, on
Niobrara, and across
the country to the
Platte.
a
o
o
.£■2
Pig
Ti p.
Light gray and ash colored
sandstones, with more or less
argillaceous layers. Fossils, —
fragments of Trionyx, Testudo,
with large Helix, Vivipara,
Petrified wood, &c. No marine
or brackish water types.
o
o
§ .
-21
r-t
Wind River valley.
Also west of Wind
River Mountains.
»-
5
'S
13
"§
3
D
Beds of clay and sand, with
round ferruginous concretions,
and numerous beds, seams and
local deposits of Lignite ; great
numbers of dicotyledonous
leaves, stems, &c. of the genera
Platanus, Acer, Ulmus, Populus,
&o., with very large leaves of
true fan Palms. Also, Helix,
Mclania, Vivipara, Corbicula,
Unio, Ostrea, Potamomya, and
scales Lepidotus, with bones of
Trionyx, Emys, Compsemys,
Crocodilus, &c.
a
u
o
o
o
o
Occupies the whole
country around Fort
Union, — extending
north into the Britisli
possessions, to un-
known distances ;
also southward to
Fort Clark. Seen un-
der the White River
Group on North Plat-
te River above Fort
Laramie. Also on
west side Wind River
Mountains.
§
Figure 70. — Facsimile of the Meek and Hayden Tertiary section of 1862,
showing original definitions of White River group and Wind River
formation
The deposits are now known to include the following:
"Loup River beds" (lower Pleistocene fauna listed). The area includes deposits of the Plio-
cene and Miocene (Ogalalla formation of Darton).
"White River group," including lower Miocene (Arilcaree formation of Darton) and Oli-
gocene (Brule and Chadron formations of Darton). The " Clioeropotamus" is Ancodus
amcricoKKs.the ancodont of the Chadron formation (Titanotherium zone).
"Wind River deposits" (summit of the lower Eocene).
"Fort Union or Great Lignite group" (basal Eocene).
OLIGOCENE FLOOD-PLAIN SEDIMENTATION IN THE
WESTERN GREAT PLAINS REGION
CONDITIONS OF DEPOSITION
A very long period of extremely slow sedimentation,
east of the Rocky Mountains of Wyoming and Colo-
rado, began in lower Oligocene time and extended
without interruption to lower Miocene time, laying
down the great deposits originally described as the
White River group by Meek and Hayden (1862.1,
p. 433) in the following language:
104
TITANOTHEBES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
White River group * * * Wliite and light drab clays,
with some beds sandstone, and local layers limestone. Fossils:
Oreodon, Titanotherium, Choeropotamus, Rhinoceros, Anchithe-
rium, Hyaenodon, Machairodus, Trionyx, Testudo, Helix,
Planorbis, Limnaea, petrified wood, &c., &o. All extinct. No
brackish- water or marine remains * * * I^OOO feet or
more * * * Badlands of White River; under the Loup
River beds, on Niobrara, and across the country to the Platte.
* * * Miocene.
This original definition appears to include all that
has been determined subsequently and mapped by
the United States Geological Survey (Darton, 1905)
under three formations, namely, Chadron, Brule, and
Arikaree, as shown in the accompanying illustration
east. This fact is significant. It would appear, as
stated at the beginning of this chapter, that sedimen-
tation in this region was suspended after Denver,
Lance, and Fort Union time. The Eocene gradients
were so high that there were long periods of erosion,
during which large areas of Upper Cretaceous beds
were laid bare in the region that now includes North
and South Dakota, western Nebraska, and Colorado,
so that the lowest Oligocene sediments of the White
River group, composing the Titanotherium zone
(Chadron A), lie in gentle valleys of ancient formation
that range in age from the Algonkian to the Denver
formation and Dawson arkose. In Hayden's typical
Figure 71. — Map showing tributaries of Cheyenne River, S. Dak., from the southeast and tlie type locality
(X) of the " Titanotherium beds" of Hayden (Chadron formation), on Bear Creek; also principal collecting
ground of Hatcher (dotted area) , the chief fossiliferous area in the Big Badlands
(fig. 69). Meek and Hayden did not, however, specif-
ically define the upper limit of theii' White River group,
and all the fossils listed by them as characteristic
of the White River group apparently came from beds
now classified as Oligocene. The name White River
group has therefore for years been restricted to the
beds of Oligocene age (Brule and Chadron formations).
This great flood-plain deposition was preceded by
a long period of erosion in Eocene time. No sedi-
ments of Wasatch, Bridger, or Uinta age have been
found on the Plains east of the Front Range of the
Rocky Mountains, except in a small area of Huerfano
sediment which lies within a mountain basin farther
locality of the White River group — the Mauvaises
Terres of early explorers — the Big Badlands between
the Cheyenne and the White River of South Dakota —
the underlying beds are composed entirely of the
Pierre (Upper Cretaceous). At some places (Loomis,
1904.1, p. 432) the rivers depositing the Titanotherium-
bearing beds washed out along theu" banks masses of
the Pierre shale that contained characteristic Pierre
fossils — Baculites and the bones of Cretaceous rep-
tiles— and redeposited them in Oligocene sediments.
On this level, the gently undulating surface of the
Pierre, east of the Rocky Mountains and the Black
Hills, meandered broad, sluggish streams, whose chan-
ENVIEONMENT OF THE TITANOTHEEES
105
nels ranged in width from a few hundred feet to half
a mile. Beside these stream channels there were
lagoons and areas of back water, some of them spread-
ing into shallow lakes but none into vast sheets of
fresh water. Savannas were interspersed with grass-
covered pampas traversed by wide, meandering rivers
that frequently changed their course. In these chan-
nels were deposited conglomerates and river sandstones,
marked by cross-bedding, as well as calcareous grits
In the shallow lagoons and back waters were deposited
the fine clays and layers of fuller's Qarth. The de-
posits of gypsum represent periods of evaporation.
In the lower part of the Titanotherium zone the de-
it spread over the great area on which it has left its
traces by the deposition of its peculiar sediments.
* * * The basin-like character of this formation
is most admirably shown." In the same memoir,
Leidy (1869.1, p. 25) expressed some doubt as to the
lacustrine theory, observing: "It is a remarkable cir-
cumstance that among the large quantity of fossil
bones brought from the Mauvaises Terres and sub-
mitted to the examination of the author, certainly
amounting to several tons in weight, there was de-
tected no trace of remains of birds or fishes; and the
same may be said of reptiles, except one species of
turtle."
Juan Ot/l^^ Ul 'lo he.iLi' ol UjJ.yiU-J
Figure 72. — Type locality of the " Tiianoiheriuvi beds ol llaj deu, oii Bear Creek, S. Dak.
Panoramic view, connecting at X. Upper section, looking southeastward, up Bear Creek; lower section, looking northwestward, down Bear Creek. Am. Mas.
negatives 104722-104726.
posits consist chiefly of fine flood-plain or overflow
sediments interspersed with river sandstones and
conglomerates, perhaps locally lacustrine, and occa-
sional layers of volcanic ash.
This theory that the deposits of the western Great
Plains region are of flood-plain and fluviatile origin
has gradually replaced the older lacustrine theory that
they were laid down in great fresh-water lakes. The
lacustrine theory originated with Hayden, who, in his
geologic introduction to Leidy's memoir of 1869
(1869.1, p. 18), observes: "One of the most interesting
features in regard to this great fresh-water lake is the
evidence of its growth from a germ, as it were, until
101959^29— VOL 1 9
The lake-basin theory was generally adopted by
geologists and paleontologists, reaching its apex in
King's development of the lake theory both for the
Plains and the mountain region. Johnson (1901.1),
Gilbert (1896.1), Haworth (1897.1), and especially
Davis (1900.1) reviewed the whole subject broadly in
a critical way, developing the theory of fluviatile and
flood-plain origin. Fraas (1901.1), Hatcher (1902.3),
and more recently Darton (1905.2) set forth strong
evidence for the theory of deposition in river channels,
flood plains, back waters, lagoons, and shallow lakes.
Among paleontologists Matthew (1899.2; 1901.1)
was the first to attack the lacustrine theory of the
106
TITANOTHEEES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
origin of the Brule clay of the White River group and
to advance reasons for believing that the sandstones
were formed by river and flood-plain sedimentation
and the clays in part by back water and lagoon and
chiefly by eolian sedimentation. His paleontologic
Porcupine Butte
CANIC ASH LAYER
t-:<Wjg ,,
"■""'saite-,,,.
jifery'cocha
Steneofiber
Prom erycoch ce
Figure 73. — Panoramic section of the Big Badlands of South Dakota
Modified from United States Geological Survey Bulletin 361, PI. in. (Osborn and Matthew, 1909.321.) View
southeastward from Cheyenne River, along line indicated on Figure 09, toward Porcupine Butte, across
the Chadron, Brule, and Arikaree formations. This section illustrates the intrusion of river-channel
deposits (the " T'/VaTioMfriiim sandstones," "JV/c/amyTiodon sandstones," and "Pro^ncfras sandstones") and plete skeletons at present knOWn maV
river-channel conglomerates in "clays" of the ritanof/iOTitm and Orfodoji zones. It shows also the charae- , j-j+ufi f 1^"
teristic erosion forms of these different layers. (See map forming fig. 69, vicinity of section B.) be COUntcd On the nngerS 01 One nana.
The Testudinata as analyzed by Hay (1908.1)
furnish evidence that during lower Oligocene time the
Great Plains region was prevailingly dry land. In
the sediments of the White River group there occur
eight species of the Testudinidae, including one of
the land tortoises, Stylemys, and one
species of Testudo. Testudo hrontops
Marsh occurs in the Titanotherium
zone (Chadron formation) and is
generally found in the White River
deposits of Colorado. Of water-living
forms the White River group of South
Dakota has furnished one species of
small turtles related to the Chelydridae
and now confined to Central America.
In 1904 Loomis (1904.1) advocated the
flood-plain origin of the "Titanotherium
beds" and described two new river-hv-
ing reptiles — Chrysemys, similar to the
Emys lativertehralis Cope of the
Wasatch; and Alligator prenasalis
(Loomis), recently found by Loomis
in the beds of Indian Draw, the first
appearance of this genus in the
Tertiary.
The nature of the sandstone or the
clay in which their remains are found
makes it impossible to separate the
mammals of the Chadron formation
(Titanotherium zone) into plains-
dwelling and river-dwelling forms,
because during floods both were swept
into the streams, the skeletons being
dissociated and the skulls and jaws
separated. Doubtless also the remains
of decaying carcasses were pulled apart
by crocodiles and garpikes. Only three
complete skeletons with skulls have
been found intact, namely, the famous
Brontops rohustus of the Yale Museum,
the Brontops rohustus of the American
Museum, and the Menodus trigonoceras
of the Munich Museum. "For every
even approximately complete skele-
ton," observes Hatcher (1902.3, p.
124), "there are scores of isolated
skulls and other bones. Taking Titano-
therium as an example, I have myself
collected nearly 200 skulls of this
animal, while the number of fairly com-
analysis showed that the fine Brule clays contain
chiefly terrestrial and plains animals, whereas the
river-channel sandstones that traverse these clays,
although contemporaneous, contain chiefly forest and
fluviatile animals.
SOUTH DAKOTA IN TITANGTHEEIUM TIME
The best description of the conditions in the South
Dakota region while it was inhabited by titanotheres
is that given by Hatcher (1902.3, pp. 125-127),
based on his own keen observations, which extended
ENVIRONMENT OP THE TITANOTHEBES
107
75
100
over many years of arduous exploration for remains
of titanotheres:
The distribution, state of preservation, nature, and character
of the animal and plant remains found in the clays and sand-
stones, as well as the distribution of the latter, absolutely pre-
clude the possibility of their having been deposited in a vast
lake and favor the presence of streams meandering through
low, broad, level, open or wooded valleys subjected in part at
least to frequent inundations, con-
ditions very similar to those at
present prevailing in the interior of
South America, about the head-
waters of the Orinoco, the Amazon,
and the Paraguay and Parana
Rivers.
Now it is evident that if such
conditions prevailed in this region
during the deposition of the White
River beds there should remain cer-
tain evidences concerning it, such
as flUed-in river channels and small
lagoons with their characteristic
deposits and remains of the animal
and vegetable life peculiar to each.
Moreover, some indication at least
of the forests should remain and be
found somewhere in this vast region.
With these and many other points
constantly in mind the writer passed
a considerable portion of the seasons
of 1900 and 1901 in exploring these
deposits. Particular attention was
given to ascertaining whether or not
they contained an aquatic fauna
and flora. The sandstone lenses
were especially examined with ref-
erence to this, for whether the de-
posits as a whole were of lacustrine
origin or not, there could be little
doubt as to the aqueous origin of the
sandstones. Though for the most
part remarkably barren of acjuatic
life, remains of Trionyx, fishes, and
crocodiles were found, and in one
locality the casts of unios were ob-
served in great numbers. A search
in the clays of the Titanotherium
and overlying Oreodon beds was re-
warded with greater success, for
numerous thin layers of limestone,
varying in thickness from a fraction
of an inch to a foot or more and
always of limited areal extent, were
discovered at many horizons rich
in the remains of fresh-water plants
and MoUusca, such characteristi-
cally shallow-water forms as Chara,
Ldmnaea, Physa, and Planorbis
occurring in the greatest abundance.
plants and MoUusca as are Chara and P/ji/sa at various horizons
throughout the White River series, and in the very midst of the
region which was supposed to have been occupied by a great
lake, and intercalated with the clays which advocates of the
the lake theory maintain were deposited in the deep and quiet
waters, would appear to preclude the possibility of the existence
of such a lake in White River times. Moreover, remains of
forests were found at several places and at different horizons
O
''Leptcaicheniev
y^-^^^^--^-^ zone
.^rr-f^ Promerycochoerus
zone
~^^s^^^^^ zo7ie\
pN CHANNEL 5A NDSTONESh
^:=--^^^=^^ OreodoTh
^^^^^^f^^zorieX
j^~^-^2r-^^-^^ = (upper) _\
r^AL TERN A tTnG RED A NiTgRA Y LA yEPS
^^^^^^^^^^Tttcuxotherizurv
~Z One^r^'DDL E BEDS
Miohippus
Mefamynodon
Meso/iippus bairdi
Brontops robusfus
Brontops d/'spar
Brontops brachycephalus
Figure 74. — Section of the Big Badlands of South Dakota showing the chief faunal zones
of the Oligocene (White River group. No. 11, fig. 35) and the Miocene
The Chadron formation ( Tilanotheriiim zone) is shown as determined by the surveys of Hatcher; the divisions of the
Brule formation (.Oreodon and Leptauchenia zones) were first established by Wortraan's observations; above is the
Arikaree formation of Darton (PromeTycochoerus zone).
I have submitted these
MoUusca to Drs. Dall, Pilsbry, and Stanton, and all have
assured me that they belong to species inhabiting swamps
and small ponds and could not have lived in the midst of a
great lake; while Dr. Knowlton, who has examined the
plants, finds in great abundance the stems and seeds of
Chara, which, as aU know, is distinctly an inhabitant of small
springs, shallow ponds, and brooks. The presence of these
thin limestone layers with such characteristically swamp
throughout these beds. At various localities in the Hat Creek
basin in Sioux County, Nebr., I discovered remains of the
silicified trunks of trees and seeds belonging especially to
Hicoria and Celtis. These were found at various horizons from
the middle Titanotherium beds to the very top of the Loup
Fork. And in South Dakota, some 12 mUes north of White
River, opposite the mouth of Corn Creek, I discovered the
remains of a not inconsiderable forest. Here in the upper
Titanotherium beds and lower Oreodon beds there occur, actu-
108
TITANOTHEKES OF ANCIENT WYOMING, DAKOTA, AND NEBRASEA.
Chadron. Fonnatioii
(TitanotheT-ibUTL 7,OTie}
C OLORAD O
Castle Rock
K A N" S A S
100 120 140 160 MILES
Figure 75. — Map showing principal exposures of tlie Chadron formation {Titanotherium zone) of Montana, Soutli
Dakota, Wyoming, Nebraska, and Colorado
Chiefly after Darton, 1905, United States Geological Survey. Includes the exposures at Castle Rock (Castle Rock conglomerate), south of Denver, in
Weld and Logan Counties, Colo., where the early collections of Marsh and Cope were made: iu the outlying localities of Bates Hole and Hat Creek,
Wyo., from which the Reed and Hatcher collections were made; and in the Big Badlands of Pennington, Custer, and Washington Counties, S. Dak.
The Titanotherium zone was first observed by Hayden at point X on map and was first subdivided faunistically by Hatcher.
ENVIRONMENT OF THE TITANOTHEEES
109
ally by hundreds, the silicified stumps and partially decayed
trunks of trees, weathering out of the fine clays of these deposits.
It was noticeable that only the knots and lower stumps had
been preserved. Nothing like complete trunks were to be
observed, and the entire aspect was that of the remains of a
dead and decayed forest on the margin of some streams, where
only the less destructible knots and stumps would endure
sufficiently long to be finally covered up and preserved. In this
same region there were discernible certain strata which seemed
to indicate that during the deposition of these beds there has
been at several horizons an accumulation of vegetable mold
or humus, and on Dry Creek, some 5 miles northeast of Chadron,
in Dawes County, Nebr., I observed near the base of the
Oreodon beds a stratum of some 2 feet of dark-colored humus,
clearly indicating that this region had not been occupied by a
great lake while this stratum was being deposited.
Hatcher concludes that the sandstone, the con-
glomerate, and a part of the clay were deposited in
river channels and that the lenses of limestone, which
are rich in remains of aquatic plants and moUusks, were
formed in shallow ponds and lakes that were scattered
over the higher tablelands and the broad flood plains,
where most of the finer clays were deposited by
occasional inundations in the rivers and by wind.
These conditions are similar to those now prevailing
about the sources of Parana and Paraguay Rivers in
central South America. ' There the rainy season
extends from October to April, and the heaviest rains
fall near its end, when the small rivers from the
highlands are flooded and pour their waters over the
flood-plain. The water, however, takes a long time to
spread over the plain, and it is there highest in July
and August and lowest in February. The flood plain
of the Paraguay is 1.50 miles wide and broadens up-
stream. The flood plains of the upper Paraguay, the
Amazon, and the Orinoco are confluent. Here we
have a group of regions that are together probably
larger than that occupied by the great White River
group during Oligocene time — namely, 97,500 square
miles.
RAPID FIUVIATIIE SEDIMENTATION IN THE CYPRESS HILLS,
SASKATCHEWAN
While the conditions thus described existed in the
Big Badlands of South Dakota, the streams were
much more active at places in areas to the south and
north. "That the Cypress Hills Oligocene deposits
were the result of rapidly flowing water from the west
is evident," observes Lambe (1908.1, p. 7). He con-
tinues :
The thick basal beds of rounded pebbles represent the work
of a strong transporting force, such as would be supplied by
a turbulent stream of considerable size carrying eastward
material from the Rocky Mountains. The sands show false
bedding as a result of varying currents. With the accumulation
of material eastward, and consequent reduction of the trans-
porting force, beds of finer material were deposited at a higher
level and probably on extensive areas of overflow.
The beds that yield the most fossils are composed of
a fine conglomerate, which on disintegration has freed
the fossils. Associated beds of a rich brown coarse
sand have also yielded some interesting remains. The
vertebrate fauna of this region in Saskatchewan
includes about 37 genera and 58 species, comprising
among the fishes Amia, Lepidosteus, and catfishes;
among the reptiles land tortoises, chelydrids, leather-
backs (Trionyx), anguid lizards, palaeophid snakes,
and true crocodiles; among the mammals opossums,
anthracotheres, elotheres, agriochoerids, camels, tragu-
lids (Leptomeryx) , horses (Mesohippus), hyracodonts,
true rhinoceroses (aceratheres), titanotheres (several
lower Oligocene types), sciurid and ischyromyid
rodents, beavers, hares, hyaenodonts, dogs {Oynodictis,
DapTiaenus), and cursorial saber-tooths (Dinictis).
These Saskatchewan beds are not only more than
twice as thick as those in South Dakota but they have
afforded a truer picture of the highly diversified rep-
tilian and mammalian life during the time represented
by the Titanotherium zone. The species of titano-
theres that they contain belong chiefly to the lower
(Chadron A) and middle (Chadron B) levels of the
Titanotherium zone of South Dakota.
SLOW SEDIMENTATION IN SOUTH DAKOTA
As compared with the 500 feet of fluviatile deposits
of the Titanotherium zone in Saskatchewan the bare
180 feet of sediments that represent the Titanotherium
zone in South Dakota are very misleading as to the
length of geologic time they represent. Deposition in
South Dakota must have been extremely slow. The
finer materials that border the river channels and
compose the clays must have accumulated very
gradually. That a very long period of geologic time
elapsed while these sediments were being laid down is
evident also from the great span of evolutionary
change indicated by members of each phylum of the
titanotheres found in this region. On the lowest
levels are found primitive small-horned titanotheres,
inferior in size to the smaller existing rhinoceroses; on
the highest levels are found gigantic animals, of almost
elephantine proportions, armed with great, powerful
horns. As a basis for estimating the time required for
the deposition of the South Dakota sediments, com-
parison may be made with existing conditions along the
River Nile, which between Aswan and Cairo is build-
ing up its bed at the slow average rate of 10 centimeters,
or 0.32 foot, per century. (Lyons, 1906.1, p. 334.)
At this rate the deposition of the 180 feet of "Titano-
therium beds," if composed entirely of fine clays, would
have required some 55,000 years. On the other hand,
if we apply Humphreys and Abbot's estimates for the
Mississippi River system, namely 0.5 foot in 100 years,
about 36,000 years would have been necessary for the
deposition of the fine clay materials of the Titano-
therium zone. The present author inclines to the
opinion that the lower Oligocene evolution of the
titanotheres demands a period of not less than 55,000
years, which would correspond with the present rate
of sedimentation in the flood plain of the Nile.
no
TITANOTHERES OF ANCIENT WTOMING, DAKOTA, AND NEBRASKA
GEOGRAPHIC DISTRIBUTION OF THE CHADRON FORMATION
The following list of localities at which remains of
titanotheres have been collected includes some
isolated spots where the Chadron formation is rec-
ognized by a few bones or a single skull, as well as
points in the classic areas of the Great Plains where
the history of the titanotheres is chiefly recorded.
Localities at wMcli tlie principal types and collections of Oligocene titanotheres Jiave heen obtained
South Dakota
Eegion
' Mauvaises Terres of Nebraska,"
Big Badlands of Chej'enne and
White Rivers of South Dakota.
This region, the one most exten-
sively explored, commonly known
as the Big Badlands, lies between
White and Cheyenne Rivers,
southwestern South Dakota, ex-
tending over the border into Ne-
braska and Wyoming, including
the basin of Hat Creek. The
lower Oligocene has been well dif-
ferentiated in this region, and the
records are generally definite.
Explorations
Successively explored by mem-
bers of the American Fur Co.
(1845), John Evans (1853),
Meek and Hayden (1853), and
by more recent explorers:
Hatcher (for the U. S. Geologi-
cal Survey, U. S. National
Museum, and Yale University
collections, 1886, 1887, 1888,
1902), Garman (for the Muse-
um of Comparative Zoology,
1885), Wortman (for the
American Museum of Natural
History, 1892, 1894), Gidley
and Thomson (for the Amer-
ican Museum of Natural
History, 1902), Thomson (for
the American Museum of
Natural History, 1904), Dar-
ton (for the U. S. Geological
Survey, 1905).
Locality
Bear Creek _
Indian Draw.
Quinn Draw_
Corral Draw
Type
Menodus (Titanotherium) proutii
(Owen, Norwood, and Evans),
Diploclonus (Megacerops) tyleri
(Lull), Brontotherium (Tita-
nops elatus) gigas Marsh, Men-
odus (Menops) varians (Marsh)
Brontotherium tichoceras
(Scott and Osborn), B. doli-
choceras (Scott and Osborn),
B. platyceras, B. leidyi Osborn,
Brontops validus Marsh, Al-
lops crassioornis Marsh, A.
serotinus Marsh.
Brontotherium medium Osborn,
B. hatcheri Osborn, Mega-
cerops (Symborodon) copei
Osborn, Allops (Megacerops)
marshi Osborn.
Diploclonus bicornutus (Osborn) ,
Brontops brachycephalus Os-
born, Brontotherium (Titano-
therium) ramosum Osborn.
Allops walcotti Osborn.
Nebraska
White River, northern Nebraska
Hat Creek, Sioux County, Nebr.
Hat Creek, a branch of the South
Fork of Cheyenne River, rises in
the canyon in the north front of
Pine Ridge, Sioux County, and re-
ceives numerous branches, also
heading in this front.
Big Cottonwood Creek, Sioux Coun-
ty, Nebr. The exposures of the
Titanotherium zone at the head of
Big Cottonwood Creek are coex-
tensive with those of the Hat
Creek basin, which lies north of
this locality. Much of Hatcher's
collecting was done in exposures
on the low divide connecting Big
Cottonwood Creek with the ex-
posures in the Hat Creek basin.
Adelia is a station on the Chicago,
Burlington & Quincy R. R., on
the outskirts of this particular
region.
Marsh and Clifford (for Yale
University, 1874).
Hatcher (for the U. S. Geologi-
cal Survey, 1886; for the Car-
negie Museum, 1900), Peter-
son(forthe Carnegie Museum,
1901, 1902).
Hatcher (for the U. S. Geo-
logical Survey, 1886; for the
Carnegie Museum, 1900).
Dry Creek_
Hat Creek.
Brontops rdbustus Marsh.
Brontops dispar Marsh.
ENVIEONMENT OF THE TITANOTHEKES
111
Localities at which the principal types and collections of Oligocene titanotheres have been obtained — Continued
Colorado
Region
Explorations
Locality
Type
Northeastern Colorado, Logan and
Weld Counties, exposures south
of the Pawnee Buttes escarpment
and some distance north of the
South Platte River. The lower
Oligocene was differentiated and
explored by Cope in 1873, but it
has hardly been touched by any
subsequent work (Horsetail Creek
beds of Matthew).
Lower Oligocene has been recognized
at other points in Colorado, as in
the vicinity of Akron, but no
adequate collections have been
made for the identification of
species.
Castle Rock conglomerate (upper
part of " Monument Creek beds"),
Colorado.
Marsh (for the Yale Museum,
1870), also field collectors.
Cope 1873, Matthew, Brown,
Martin (for the American
Museum of Natural History,
1898), Matthew, Brown (for
the American Museum of
Natural History, 1901).
Darton (for the U. S. Geological
Survey, 1905, 1906). Richard-
son (for the U. S. Geological
Survey, 1912).
Probably in Weld
County, Colo.;
near Gerry's ranch,
Colo.; also 4
miles south of
Pond Springs,
Colo. Collector,
Devendorf.
Horsetail Creek ,
Colo.
Brontotherium gigas Marsh,
M e n o d u s (Brontotherium)
ingens Marsh, Brontotherium
(Titanops) curtum Marsh.
Megacerops (Symborodon) acer
Cope, M. (S.) altirostris Cope,
M. (S.) bucco Cope, M. (S.) tor-
vus Cope, Menodus (Symboro-
don) trigonoceras Cope, M. (S.)
heloceras Cope, M. (S.) hypo-
ceras Cope, Megacerops riggsi
Osborn.
Wyoming
Beaver Divide, Fremont County,
Wyo.
Bates Hole, Natrona County, Wyo.
Exposures lying between Bates
Hole, at the north end of the
Laramie Plains, and Beaver
Divide, at some distance to the
west, have been casually examined
by W. H. Reed and W. D. Mat-
thew, who have recognized a
lower Oligocene fauna, but no
S3'stematic exploration has been
made. A number of well pre-
served specimens of titanotheres
(Univ. Wyoming Mus.) were
found in this area.
Granger (for the American Mu-
seum of Natural History,
1910).
Reed (for the University of
Wyoming, 1907, 1908).
Pipestone Creek and Thompson
Creek, Jefferson County, Mont.
In southwestern Montana, at
Pipestone Springs, McCarty
Mountain, north of Dillon, and
elsewhere, small scattered expo-
sures. A considerable fauna, of
small species, has been described
from these beds.
Douglass (for the Carnegie Mu-
seum, 1899, 1901, 1903),
Matthew (for the American
Museum of Natural History,
1902).
Saskatchewan
Swift Current Creek, Cypress Hills,
McConnell and Weston (1883),
Menodus angustigenis, M. sel
Saskatchewan.
Weston (1888, 1889), Lambe
wynianus, Megacerops syceras
(1904).
M. assiniboiensis, M. primi-
tivus.
112
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The Chadron formation was especially explored by
Darton, under whose direction a map showing its gen-
eral distribution Csee fig. 8) was prepared. His de-
scription of this map may be paraphrased as follows:
The most extensive outcrops are in the soutliwestern portion
of South Dalcota, in a district known as the Big Badlands.
These extend along the valley of White River and in the ad-
jacent divide between White River and Chej'enne River
West of the latter the formation caps many of the divides
extending to and up the eastern slopes of the Black Hills. The
formation extends eastward to the vicinity of longitude 100°,
and it extends westward along the foot of Pine Ridge through
Dawes and Sioux Counties in Nebraska and Converse County,
Wyo., to the north end of the Laramie Range. The formation
probably underlies a large portion of western Nebraska, but it
only reaches the surface at the foot of Pine Ridge and along the
north Platte Valley west of Scotts Bluff. Isolated outcrops
are also reported at Valentine and Lone Pine. The formation
appears extensively in northeastern Colorado, on both sides of
the valley of South Platte River. There are prominent ex-
posures west of Akron, south of which the formation extends
across the greater part of Washington County. South of
Denver an extensive area caps the divide between the South
Platte and the Arkansas Rivers, at the foot of the Rocky
Mountains. The deposits in this area have been designated
the Monument Creek formation. This consists of two members
of which the upper [now called Castle Rock conglomerate;
(Richardson, 1912.1)] has recently yielded Titanotherium
remains. Other outlying areas of the formation occur in Bates
Hole west of the Laramie Range [Wyoming], in Butte County,
S. Dak., and in the southeastern corner of Montana.
An important additional exposure of the Titanothe-
rium zone discovered by the American ]\luseum expe-
dition of 1909 is at Beaver Divide (Wagonbed Spring),
in the southern rim of the Wind River Basin, near
Hailey, Wyo. Here a deposit containing a skull of
a primitive Oligocene titanothere was found overlying
a deposit of upper Eocene age containing Amynodon.
The thickness of the Chadron formation varies, but
in some places it reaches 180 feet. It consists of
clays, sands, gravels, and sandstones, clay predominating greatly
over the other materials. * * * This clay is of pale-greenish
color, weathering in typical badland form and often having the
peculiar character of fuller's earth. At the base of the formation
there is usually a bed of gravel and sand merging upward into
sands and sandy clays, which in the Big Badlands and western
Nebraska are often of a reddish color. At various horizons
through the formation there are beds of sandstone from a few
inches to 4 feet in thickness and of local extent. Ordinarily
these coarse materials exhibit current bedding and from their
character and relations are clearly the products of running
water. Beds of volcanic ash occur extensively in the Chadron
formation in the Big Badlands and at intervals along the
northern front of Pine Ridge (South Dakota).
Hatcher observes (1893.1, pp. 206-207):
The clays- greatly predominate, consist of very fine particles,
and are quite compact. In places they are composed almost
entirely of pure kaolin, but they often contain a considerable
portion of sand. Near the bottom of the beds the color is often
red or variegated, due to the presence in them of small quanti-
ties of red oxide of iron; but the prevailing color is a very char-
acteristic and delicate greenish white. * * * Owing to
the extreme minuteness of the particles forming the clays and
the absence of sufficient cementing material in them, in most
places they readily yield to the action of water and are quite
rapidly eroded. The clays of the Titanotherium beds were
probably derived from two sources, viz, from the Cretaceous
clays and shales and from the kaolinization of granite feldspars.
The sandstones are never entirely continuous and never more
than a few feet thick. They present everj- degree of compact-
ness, from loose beds of sand to the most solid sandstones.
They are composed of quartz, feldspar, and mica and are evi-
dently of granite origin. When solidified the cementing sub-
stance is carbonate of lime.
The conglomerates, like the sandstones, are not constant, are
of very limited extent, never more than a few feet thick. They
are usually quite hard, being firmly held together by carbonate
of lime. A section of the beds taken at any point and showing
the relative position and thickness of the sandstones, clays, and
conglomerates is of little [stratigraphic] value, since these vary
much at different and quite adjacent localities.
These descriptions by Darton and Hatcher reveal
a wide contrast between the composition of the
Titanotherium-heen-ing beds and that of the upper
Eocene deposits of the mountain-basin region.
COMPARISONS OF BASINS IN WESTERN UNITED STATES WITH THE FIOOD
PLAIN OF THE NUE
The flood-plain deposition of the Nile, which has
been very carefully studied, also throws light on the
mode of formation of parts of the Chadron formation.
The following passages are taken from "The physi-
ography of the River Nile and its basin," by Capt.
H. G. Lyons (1906.1, pp. 241, 311, 334):
When rivers already loaded with sediment emerge from their
mountain valleys of high slope on to a level plain under arid
climatic conditions where the water table is at somie distance
from the surface their water sinks in almost at once instead of
flowing on the surface and therefore deposits its load of sediment
as an alluvial fan. This fan is built up most rapidly at its
head, and as the floods of successive years come down new tem-
porary channels are formed which divide and reunite, forming
a network of channels, each b}' deposit building up banks for
itself, which are probably cut through in the next season of the
summer rainfall.
While it is the finer silt which is deposited in the irrigation
basins, on the shelving banks of the river, and on such parts of
the flood plain as are annually flooded, it is the bottom load
which is deposited in the bed of the river itself, and this con-
sists of the coarser sand which the current can not carry so
readily as the finer material. If the Nile mud is treated by
levigation so as to remove the finest particles of clay and sand
the residue is a fine whitish-gray sand, such as is seen forming
sand banks in the Nile wherever the conformation of the river
is such that the velocity of the fiood current is reduced at that
point.
In this Aswan-Cairo reach of the Nile, then, we have to do
with a river which is fiowing with a low slope through an alluvial
plain which it has formed and which, if uncontroUed, it annually
floods, depositing on the flood plains part of its load of silt as
the velocity of flood water is diminished.
The Nile between Aswan and Cairo follows a depression in
which it has gradually deposited a considerable thickness of
alluvial mud, and now it meanders on the flood plain which it
has formed. In earlier times side channels followed the lower
margins of the valleys, and lagoons and swamps existed in the
same part of the valley.
ENVIRONMENT OF THE TITANOTHERES
113
FAUNAL DIVISIONS IN THE CHADRON FORMATION
THEEE FAUNISTIC LEVELS DETERMINED
In the series of sediments that were deposited on
the uneven surfaces of the Pierre shale and that con-
sist of fine clays, which were traversed and at many
places secondarily eroded by river channels and which
were very gradually accumulated during an extremely
long period, we should not expect to find anything re-
sembling clearly defined stratification or horizontal
and vertical succession of species and genera. Never-
theless, we owe to the genius and the untiring explora-
tion of Hatcher a division of the Chadron formation
into lower, middle, and upper levels, which we shall
designate Chadron A, Chadron B, and Chadron C,
and which correspond to similar divisions of the deposits
of the Rocky Mountain basins.
In his paper of 1893 (1893.1), "The Titanoiherium
beds," Hatcher remarked that these beds were so
named by Meek and Hayden in 1857 from the genus
TitanotJierium, established by Leidy in 1852. Al-
though we are obliged to replace the generic name
TitanotJierium by Menodus it seems best to retain
Titanotherium as the historic zonal name for these sig-
nificant beds.
The thickness of the "Titanotherium beds" at dif-
ferent localities in Wyoming, Colorado, the Dakotas,
and Saskatchewan, as recorded above, varies, having
a maximum of 500 feet and a minimum of 30 feet.
Hatcher, accepting a total of 180 feet as the maxi-
mum thickness of these beds in the Big Badlands of
South Dakota, assigned 50 feet to the lower level, 100
feet to the middle, and 30 feet to the upper (1893.1,
p. 210). During the field seasons of 1886, 1887, and
1888 Hatcher collected for the present monograph
material including 105 nearly complete Titanotherium
{Menodus) skulls and parts of numerous skeletons, as
well as disarticulated bones, besides remains of many
other associated animals. Early in the season of
1886 it became apparent that certain forms of skulls
were characteristic of certain horizons in the "Titano-
therium beds." This fact indicated the desirability of
keeping, so far as possible, an exact record of the
horizon from which each skull or skeleton was taken.
From actual measurement the vertical range of the
titanotheres in the Big Badlands was found to be
about 180 feet. For convenience in keeping a record
of horizons the beds were divided into three divisions
of 60 feet each, and each of these three divisions was
subdivided into three divisions of 20 feet each. As
each skull or skeleton was dug out a separate letter or
number was given to it and it was assigned to that
subdivision of the beds from which it was taken.
STRATIGEAPHIC DISTRIBUTION OF SPECIES OF OLIGOCENE
TITANOTHERES
In 1888 Hatcher drew up a manuscript table for
Professor Marsh in which the lower, middle, and
upper divisions of the "Titanotherium beds" were
each subdivided into three levels, and in which he
placed the letters assigned to many of the skulls
found by him. In 1901 Hatcher revised this table for
Osborn for use in the present monograph. In the
summer of 1902 the United States Geological Survey
sent Messrs. N. H. Darton and J. B. Hatcher to the
Big Badlands of South Dakota for a resurvey of the
localities where some of the skulls were found by
Hatcher in order to determine precisely the elevation
of these localities above the Pierre shale, at the
base of the beds. Prof. Eberhard Fraas, of Stuttgart,
accompanied the party and made some interesting
observations on the mode of deposition of these beds.
(Fraas, 1901.1.) This experienced geologist con-
cluded that the "Titanotherium beds" consisted of
river and flood-plain deposits whose surfaces were
exposed during the dry seasons of the year; that
parts of the overlying Brule clay — the beds in the mid-
dle Oreodon zone — were deposited in shallow lakes, the
dissolved materials, of varying concentration, giving
rise to the banded layers; and that the reddish strata
of the Oreodon zone (Brule) were formed of eolian
loess.
In the following table the results of records and ob-
servations made by Hatcher, indicated by the abbre-
viation J. B. H., are supplemented by the results of a
few observations made by N. H. Darton of the United
States Geological Survey, E. S. Riggs of the Field
Museum, W. H. Reed of the University of Wyoming,
and Walter Granger of the American Museum. The
species are arranged in the five generic phyla deter-
mined by Osborn, namely, Brontops, Allops, Menodus
{= Titanotherium) , Megacerops { = Symborodon) , Bron-
totherium.
114
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Geologic succession of Oligocene titanotheres in the Ohadron formation
[Levels (above Pierre shale except as otherwise indicated) mostly taken from J. B. Hatcher's field records of 1886, 1887, 1888. Genera and species of fossils determined by
H. r. Osborn and W. K. Gregory]
Level
Brontops
Allops
Menodus
Megacerops
Brontotherium
Classification
uncertain
Brontotherium platyceras.
Skull 12161, Field Mus.
Upper levels (E. S.
Menodus giganteus.
Riggs).
Brontotherium medium.
Field Mus. skull
P5927. Near top of
Skull w, Nat. Mus. 4256
upper Titanothe-
(type); "from the ex-
rium beds (E. S.
treme top of the Bronto-
3
Brontops dispar?. Skull G, Nat. Mas.
4248. Record and level uncertain
(J. B. H.).
Allops serotinus.
Skull I, Nat. Mus.
2151. -l-80feet.
Same locality as H.
Riggs).
Menodus giganteus.
Univ. Wyoming
skulls. Upper beds
(W.H.Reed). Bates
Hole, Wyo.
therium beds. Oreodon
teeth were found in dig-
ging it up" (J. B. H.).
4-81 feet, "well up in red
clays" (J. B. H.).
Brontotherium curtum.
Skulls.
Menodus giganteus.
Skull Y', Nat. Mus.
o
Skull r, Nat. Mus.
1211. 4-93.3 feet. Skull
e
1212.
q, Nat. Mus. 4946. 4-89
3
Megacerops copei.
feet. Skull g, Nat. Mus.
1
Allops serotinus.
Skull V, Nat. Mus.
4244.
O
Skull H, Nat. Mus.
4251. 4-77 feet; 34
4711. -1-65.4 feet,
possibly In C2 (J. B.
S
P
Brontops dispar. Skull p, Nat. Mus.
feet below top.
H.). ?SkullL',Nat.
s
1217.
Mus. 4700.
d
Brontops dispar?. Skull d, Nat. Mus.
4696,
Brontops sp. Mounted skeleton, Am,
Allops serotinus
?Menodus giganteus.
Skull G', Nat. Mus.
4291. "From mid-
?Brontotherium hatcheri?.
Univ. Wyoming skull 1.
Skulls R', W'.
2
Mus. 618. "Very high, 8-10 feet
from top." (J. B. H.) "32 feet be-
low the 3-foot siliceous limestone
layer at top of Titanotherium series."
(N. H. Darton, 1901.)
Skull i, Nat. Mus.
4938.
dle beds or toward
the top" (J. B. H.).
"Upper beds." Bates
Hole, Wyo.
Megacerops acer.
Brontotherium medium?.
Allops crassicornis.
Menodus proutii.
Univ. Wyoming
Skull N', Nat. Mus.
1
?Brontops dispar. Skull h, Nat. Mus.
Skull Z', Nat. Mus.
4289. "J. B. H. in-
Skull e, Nat. Mus.
skull 2. "Upper
beds about two-
4699. Level?
Skulls M', U'.
4944.
4701. Level doubt-
Brontotherium gigas.
clined to place this
ful (J. B. H.).
thirdsupfrombase"
Skull H', Nat. Mus.
skull higher up."
(W. H. Reed).
Bates Hole, Wyo.
4262.
Brontops dispar. Skull D', Nat. Mus.
Menodus giganteus.
Megacerops bucco.
Skull 0', Nat. Mus.
Brontotherium medium?.
4706. Level essentially correct (J. B.
Skull r, Nat. Mus.
Skullu, Nat. Mus. 4716.
3
H.).
1220 (very large).
4705. Level rather
Level?
Skulls S', C, F',
Brontops robustus. Type skeleton.
Menodus trigonoceras.
doubtful (J. B.H.).
4-46,7 feet.
Brontotherium hatcheri.
E'.
Yale Mus. 12048. 60 feet below
Skull 0, Nat. Mus.
Type skull a, Nat. Mus.
summit (J. B. H.).
4257.
1216.
Megacerops sp. (or
g
Brontops dispar. Skull f, Nat. Mus.
Allops marshi?.
Brontotherium
g
4703. Level certainly B2 (J. B. H.).
Skull t, Nat. Mus.
hatcheri). Skull Q',
Brontotherium hypoceras.
3
Skull D (type), Nat. Mus. 4941.
4942.
Nat. Mus. 4255.
Skull 1, Nat. Mus.
Skulls T', P', C
^
2
Hat Creek. Levelcorrect (J. B. H.).
Allops marshi?.
"Certainly in mid-
4273(?). Level proba-
B'.
Skull K, Nat. Mus. 4290 (type of
Skull A', Nat. Mus.
dle beds, perhaps in
bly correct (J. B. H.).
1
B. validus).
1215.
middlelevel"(J. B.
'f>
H.).
Id
s
Brontops dispar. Skull (erroneously
m
1
lettered P). Nat. Mus. 4245. Skull
J, Nat; Mus. 4738. Hat Creek,
Lower B, probably correct (J. B. H.).
Brontops brachycephalus?. Skull M,
Nat. Mus. 4259. -(-.55.6feet(J.B.H.).
Menodus trigonoceras.
Skull G', Nat. Mus.
4291.
Skulls N, A, B
No.? (a large
skull).
Level B, probably correct (J. B. H.).
Brontops brachycephalus. Skull i",
Nat. Mus. 4258. -f71.4 feet (J. B.
H.); -f48.5feet(N.H. Darton).
ENVIKONMENT OF THE TITANOTHEKES
Geologic succession of Oligocene titanoiheres in the Chadron formation — Continued
115
Level
Brontops
A Hops
Menodus
Megacerops
Brontotherimn
Classification
uncertain
3
Brontops brachycephalus. Skull X',
Nat. Mus. 1214. Level probably
correct (J. B. H.).
Brontops brachycephalus?. Skull m,
Nat. Mus. 4940. Level probably
correct (J. B. H.).
Brontotherium ?hypoceras.
Skull K', Nat. Mus.
4702. Level very doubt-
ful (J. B. H.).
Skulls V, I'.
B
s
2
Diploclonus tyleri. Type skull. 35
feet above Pierre shale and 165 feet
below top of formation (R. S. Lull).
Allops marshi?.
Skull E, Nat. Mus.
1213. Level proba-
blycorrect (J.B.H.).
Brontotherium leidyi
(type). Skull R, Nat.
Mus. 4249. Level cor-
rect (J. B. H.).
1
1
Brontops dispar??. Skull P, Nat.
Mus.? (not 4245). "This skull in
pieces found July 4, 1887, right at
base of beds" (J. B. H.).
Brontops brachycephalus. Skull c,
Nat. Mus. 4261 (type) . Lower levels
(J. B. H.). Skull b, Nat. Mus. 4947
(marked 1991). +14.4 feet; 130.6 feet
below top (J. B. H.). Very young
skull (new born?), Univ. Wyoming.
"From extremely low level" (W.
H. Heed). Bates Hole, Wyo.
Allops walcotti (type) .
Skull Q, Nat. Mus.
4260. "Probably
lower beds, level A,
fine-grained sand-
stones" (J. B. 11.).
Menodus heloceras.
Skull, Am. Mus.
14576. At base of
beds near Hailey,
Wyo. (W. Granger).
Brontotherium leidyi.
Skull, Carnegie Mus. 93.
Hat Creek, Nebr., 15 or
20 feet from bottom of
lower beds (J. B.H.).
HATCHER'S COHECTIONS, 1886-1888
According to a report delivered orally by Hatcher to
the author in 1901, the collections made by him in 1886
included 24 skulls, some from Hat Creek, Nebr., and
some from the Big Badlands of South Dakota, which
were designated in his records and field notebooks by
the letters A, B, C, etc., but which now bear United
States National Museum numbers. In 1887 Hatcher
collected from Sioux County, Nebr., mostly from Big
Cottonwood Creek (adjoining Hat Creek), a second
series of skulls, which he similarly designated by the
letters A to K. Later in the same season he moved
camp to the South Dakota Badlands and collected the
skulls L to Z and a to w. Thus during the season of
1887 he collected 45 skulls. During the season of
1888 he collected another series of 24 skulls in the
South Dakota Badlands, which he designated by the
letters a', b', c', to z' but which Professor Marsh
later relettered A' to Z' . Subsequently the catalogu ers
of the United States National Museum assigned
numbers to all these skulls. These revisions of the
records have caused confusion, so that it is now doubt-
ful whether certain skulls that bear capital letters and
United States National Museum numbers belong to
the series of 1886 from Hat Creek, Nebr., or to the
series of 1887 from Big Cottonwood Creek and the
South Dakota Badlands. Such uncertainty, of course,
involves equal uncertainty as to the localities and
geologic levels at which the specimens were obtained,
but nearly all uncertainties have been settled by Dr.
W. K. Gregory through careful examination of all
the available evidence, with the assistance of Mr.
J. W. Gidley, of the United States National Museum.
The above table is based on these original and revised
records.
This remarkable collection, now preserved in the
United States National Museum, constitutes the
reference standard as specifically determined by the
author with the assistance of Messrs. Gidley and
Gilmore and includes the skulls and jaws indicated
below, which are enumerated in detail under the
respective genera in Chapter VI :
Allops phylum: 24 skulls and lower jaws in four specific stages.
Diploclonus phylum: 1 skull in one specific stage.
Brontops phylum: 58 skulls and jaws in three specific stages.
Brontotherium phylum: 42 skulls and jaws in nine specific
Megacerops phylum: 7 skulls and jaws in three specific stages.
Menodus phylum: 26 skulls and jaws in four specific stages.
Figure 76. — Section showing the results of stratigraphic
leveling in the Chadron formation (Titanotherium zone) in
the badlands of White River, S. Dak., in June, 1901, by
N. H. Darton
The results are affected by dip, by unconformity, and by variation in the thickness
of the beds. In determining the dip the beds showing the nearest reliable con-
tacts of the Chadron with the Pierre formation were selected for all the levelings,
and as most of the distances determined were short and were measured along the
strike of the low-dipping beds the angle of the dip is unimportant . The Chadron
formation lies on a smooth plane of unconformity, and its basal member is gen-
erally continuous but was doubtless laid down against a sloping shore, and the
layers are not synchronous throughout its extent. Nearly all the bones listed in
the text, however, were found in an area so small that this unconformity is unim-
portant. The variation in the thickness of beds is the most important factor
aflecting the determination of the stratigraphic levels and one that could not be
accurately determined, for the beds present so much variation in character that
they can not be followed for a distance long enough to afford a basis for strati-
graphic subdivision of the formation. A horizon 30 feet above the base of the
formation at one point may represent a horizon 45 feet above it at another point;
thus a bone found at A may have been deposited at the same time as a bone
found at B.
116
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
SOURCES OF ERROR IM DETERMINING STRATIGEAPHIC LEVELS
It should be borne in mind that owing to the great
difference in the thickness of the " Titanotherium
beds" in different locaUties and to the irregular topog-
raphy of the Pierre shale upon which the beds rest
it often happens that the base of these beds at one
point may correspond to the middle Titanotherium
zone at others, so that an exact stratigraphic subdi-
vision of the Chadron formation over wide areas is
some 165 feet below their summit. Hence this skull
is assigned to Hatcher's level A 3, although its large
size and progressive structure would lead one to infer
that it came from the upper Titanotherium zone
(Chadron C).
Notwithstanding these discrepancies we are able
to follow the evolution of five separate phyla of
titanotheres, from the small animals of the lower
Titanotherium zone (Chadron A), which have small
Adaptive
radiatiorL of the
subfaTnilies of
PerLssodcLctyls
Haiiits OTui habitats
AQUATIC Kv-3 MEDIPORTAL WpM GRAVI PORTAL ESS
Figure 77. — The family tree of the Perissodactyla
Adaptive radiation of the 9 families and 35 subfamilies. Their divergence in limb and foot structure into cursorial, forest-living, mediportal,
and graviportal types, and in tooth structure into browsing and grazing types, is indicated by respective symbols.
not at all possible. In spite of such opportunity
for error only a few well-authenticated records (such
as that of the type of Brontops dispar) appear to con-
tradict Hatcher's statement that the titanotheres of
advanced structural development are confined to
the upper levels of the beds. One such striking
exception is recorded by Prof. R. S. Lull (1905.1),
who states that he found the type of Diploclonus
tyleri at a point only 35 feet above the Pierre shalei
at the base of the Titanotherimn beds, which was there
horns, through intermediate types to the latest forms,
which have highly specialized skulls, from the top
of the " Titanotherium beds." This evolution was
rather even and regular in the phyla Brontotherium
and Menodus {= Titanotherium) , but in the Brontops
phylum it appears that some of the primitive types
of the lower zone survived with little change into the
middle zone (Chadron B), and that other primitive
types evolved gradually into the more specialized
species of the middle and upper zones.
ENVIEONMBNT OP THE TITAN0THEEE5
117
Height in feet above Pierre shale at which remains of titanotheres were found as determined in 1901 hy J. B.
Hatcher and N. H. Barton'^
Skull Y' (?). "Big flat-horned skull in National Museum"; Nat. Mus. 1211 (?)
Skull Q. Indian Draw. Probably skull "small q" was meant (Nat. Mus. 4946, Brontotherium curtum),
' ' wrongly lettered Q"
Large-horned red skull. Nat. Mus. 4256, Brontotherium medium (type)
Skeleton. Am. Mus. 518, Brontops robustus?
Skull v. Indian Draw. Nat. Mus. 4711, Megacerops copei (type)
Long-horned skull. West branch of Indian Draw. Brontotherium ramosum
Skull M. Near Middle Corral Draw. Nat. Mus. 4259, Brontops brachycephalus
Skull "F." Quinn Draw, South Dakota. ?Nat. Mus. 4258, Brontops brachycephalus
Skull O'. South side of west fork of Corral Draw. Nat. Mus. 4705, Megacerops "bucco," female
Skull "I." Quinn Draw, South Dakota. Nat. Mus. 2151, "AUops serotinus," female
Skull "H." Quinn Draw, South Dakota. Nat. Mus. 4251, Allops serotinus (type)
Skull "little F." Quinn Draw, South Dakota. Nat. Mus. 4703, Brontops dispar
Little skull "B." On fork of west fork of Corral (?Quinn, J. B. H.) Draw. Probably skull b, Nat. Mus.
4947, Brontops brachycephalus, female, aged
81
C)
65.4
"93. 3
81.
n
55. 6
71.4
46. 7
80
77-34
62
65. 3
62. 3
55. 6
48. 5
46. 7
43.5
40.7
39.0
14.4
• In a letter to the author, dated July 31, 1901, Hatcher expressed grave doubts as to the accuracy of these levels on account of practical difficulties encountered in the
field.
» From horizon ot skull O', Nat. Mus. 4705, to horizon of this skull there is a vertical upgrade of 46.6 feet.
' 27 leet above skull V.
•I Very high, 8 to 10 leet from top of titanothere zone (Hatcherl.
• 32 leet below the 3-foot sihceous limestone layer at top of Titamtherium zone. The Pierre shale contact was far away, and although it was on a line of levels the dip
in interval could not be ascertained precisely (Darton).
MAMMALIAN LIFE OF THE LOWER OLIGOCENE TITANOTHEEIUM ZONE
The most highly characteristic feature of the Ohgo-
cene mammals as a whole, compared with the Eocene
mammals, is their decided modernization, which is
shown in the following table giving the percentages
of the modern and the archaic families of the Oligo-
cene Plains fauna as compared with those of the Eocene
mountain-basin fauna.
Percentages of modern and archaic families in Eocene, Oligocene,
and Miocene time
Basal Eocene
Lower and middle Eocene
Upper Eocene
Lower Oligocene
Miocene
Modern fami-
lies or those
closely related
or ancestral
to modern
families
Archaic fami-
lies supposed
to be wholly
extinct and
not closely
related to
modern types
This modernization of mammalian life is in part
real and in part apparent, because the Plains fauna
presents for the first time the full aspect of the upland,
plains, and meadow life, especially the smaller and
larger herbivorous ungulates. This life is, however,
only partly revealed in the Titanotherium zone, in
which conditions for the fossilization and preserva-
tion of the land fauna were less favorable than in the
overlying Oreodon zone (Brule clay). In fact, re-
mains of the small ungulates, such as the horses of the
period (Mesohippus) , are very rarely preserved in
either the coarser or the finer sediments of the Chad-
ron of South Dakota but are found more abundantly
in the sediments of Pipestone Creek, Thompson
Creek, and other areas in Montana and in the Swift
Current Creek area of Saskatchewan. The entire
Titanotherium zone fauna as listed by Osborn and
Matthew (1909.321, pp. 103, 104) contains representa-
tives of 6 orders and 24 families of mammals, which
are of interest and value as showing the principal
types of mamnxals that were in competition with the
titanotheres in the struggle for existence.
118
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Fish, reptile, and mammal Jauna contemporary with the titanotheres
Common name or comparable form, habits or habitat,
nature of deposits, etc.
Classific name
Region inhabited
PISCES
Actinopterygii:
Amiidae —
Do.
Do.
Gar pikes (Lepidosteus) ; rivers and streams. -
Lepidosteidae —
Do.
Catfislies (siluroids) ; rivers and streams
Siluridae —
Do.
Do.
Do.
Crocodiles and alligators; rivers and streams __
REPTILIA
Crocodilia:
Crocodilidae —
Do.
South Dakota.
Squamata:
Anguidae —
Helodermoides tuberculatus Douglass
Montana.
Do.
Palaeophidae —
Subterrestrial ; wet and forested places
Chelonia:
Dermatemydidae —
Xenochelys formosa Hay
South Dakota.
Emydidae —
Land tortoises; characteristic of uplands, open
country.
Testudinidae —
Do.
Do.
Soft-shelled turtles; aquatic; remains found in
fluviatile sandstones.
Trionychidae—
PIat3'peltis leucopotamica Cope
Saskatchewan, South Da-
MAMMALIA
Marsupiaha:
Didelphidae —
kota.
Cursorial, predacious, like the Thylacinus of
Tasmania; resembling modern wolves in
Ferae :
Hyaenodontidae —
"Pseudopterodon" minutus" (Douglass)
Montana.
Comparable with hyena of Africa; powerful
Saskatchewan, South
jaws.
Hyaenodon cf. H. crucians Leidy
Dakota.
Do.
Canidae —
Daphaenus cf. D. hartshornianus Cope
Daphaenus cf . D. felinus Scott **
Saskatchewan.
Do.
South Dakota.
Analogous to the marten and mongoose
Nothocyon "^ paterculus (Matthew)
Montana.
South Dakota.
Mustelidae —
Bunaelurus inf elix
Analogous to the marten and polecat
Montana.
" This is, in fact, an undescribed genus, more primitive than Hyaenodon and Pterodon, allied apparently to Sinopa and Tritemnodon.
>> If Mr. Lambe's figure is correct this can hardly be Z>. felinus; it agrees much better with D. dodgei Scott.
' CynodicHs is not applicable to the American OlJgocene species that have been called by that name. Nothocyon is next in priority among available names. The type,
however (N. geismarianus) , is a large and rather specialized species from the John Day formation. It maybe necessary to separate the small species from the middle and
lower Oligocene under the name Pseudocynodictis (Schlosser).
ENVIKONMENT OP THE TITAN OTHERES
Fish, reptile, and mammal fauna contemporary with the titanotheres — Continued
119
Common name or comparable form, habits or habitat,
nature of deposits, etc.
Classific name
Region inhabited
Analogous to the leopard and cheetah
MAMMALIA — Continued
Ferae — Continued.
Felidae—
South Dakota.
Possibly like the modern Gymnura of the East
Insectivora:
Leptictidae —
Montana.
Do.
Do.
Do.
Do.
Do.
Allied to the Solenodon of the West Indies(?) _
?Solenodontidae —
Montana.
Fossorial, like the Cape golden moles
Burrowing; analogous to marmots, prairie
dogs, and the like.
Squirrels and ground squirrels (spermophiles).
? Chry sochloridae —
Montana, Wyoming.
Montana.
Saskatchewan.
Montana, Wyoming.
Assiniboia.
Montana.
"Xenotherium" "^ unicum Douglass
Nothocyon "lippincottianus" (Cope)
Glires (Rodentia) :
Ischyromyidae —
Titanotheriomys veterior (Matthew)
Titanotheriomys ' ' Ischyromys typus Leidy ' ' _
Do.
Prosciurus ?saskatohewensis (Lambe)
Saskatchewan.
Castoridae —
katchewan.
Do.
Found in Chadron clays; like pocket mice;
Perognathus.
Rabbits; remains found in the Chadron clays-.
Heteromyidae —
Adjidaumo(Gymnoptychus) minor Douglass-
Adjidaumo (Gymnoptychus) minimus Mat-
thew.
Leporidae —
Montana.
Do.
Do.
Do.
Saskatchewan.
Grazing, upland rhinoceroses; cursorial; found
in the Chadron clays.
Perissodactyla:
Hyracodontidae —
Amphibious rhinoceroses; found in the channel
Amynodontidae —
South Dakota.
Do.
Small rhinoceroses of rather slender propor-
tions, probably of browsing habit; remains
found chiefly in clays.
Rhinocerotidae —
Do.
Do.
Leptaceratherium trigonodum Osborn
Do.
Assiniboia, South Dakota.
Do.
Caenopus cf. C. platycephalus Osborn
South Dakota.
Very small, slender-limbed horses, cursorial;
grazers and browsers. Abundant in clays;
Lophiodontidae —
Colodon ( = Mesotapirus) occidentalis Leidy.
Equidae —
Dakota.
South Dakota.
Colorado, South Dakota.
Mesohippus westoni Cope
Saskatchewan.
Do.
Do.
Mesohippus precocidens Lambe
Saskatchewan.
I Name preoccupied by Xenotherium Ameghino, 1904, a genus of edentates.
120
TITANOTHEEES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Fish, reptile, and mammal Jauna contemporary with the titanotheres — Continued
Common name or comparable form, habits or habitat,
nature of deposits, etc.
Classiflc name
Region inhabited
Suillines abundant and characteristic. Diffi-
cult to place. Small, compact, dldactyl
feet and fairly long limbs, cursorial. Ribs
and abdomen small. Common in clays and
sandstones.
MAMMALIA — Continued
Artiodactyla:
Entelodontidae —
South Dakota.
Do.
South Dakota.
Dicotylidae ( = Tagassuidae) —
South Dakota.-
Leptochoeridae —
Stibarus montanus Matthew --
Montana.
Analogous to pigs. Occur chiefly in the clays-_
Anthracotheriidae —
South Dakota.
Do.
Peccary-like, but of grazing habits. Rather
Oreodontidae ( = Agriochoeridae) —
Montana.
Do.
Browsing. Agriochoerus partly arboreal
proportions like the larger cats.
Do.
Oreodon (= Merycoidodon) hybridus Leidy.
Oreodon (= Merycoidodon) buUatus Leidy_-
Oreodon (= Merycoidodon) affinis Leidy
Oreodon ( = Merycoidodon) " culbertsonil
Leidy."
Agriochoerus maximus Douglass
Agriochoerus minimus Douglass
South Dakota.
Do.
Do.
Saskatchewan.
Montana.
Do.
South Dakota.
Saskatchewan.
Analogous to existing chevrotains of Africa —
Hypertragulidae —
Trigenicus socialis Douglass
?Trigenicus mammifer Cope
Montana.
Saskatchewan.
Do.
Do.
South Dakota.
Do.
Grazing, upland, cursorial, like the smaller
antelopes of Africa and the guanacos of
South America.
?Heteromeryx transversus Cope
" Anthracotherium pygmaeum" Lambe "
Camelidae —
Saskatchewan.
Do.
?Leptotragulus profectus Matthew
Montana.
« Based upon a part of a "right upper molar," which from Mr. Lambe's figure appears to be a left lower molar, probably of a hypertragulid comparable
to Heieromeryx.
NOTES ON THE HABITAT OF THE FAUNA OF THE CLAY AND
SANDSTONE AS A WHOLE
Matthew was the first to distinguish between the
upland forms, found chiefly in the clays (flooded
plains), and the lowland and aquatic forms, found in
the sandstones (river channels). The following dis-
criminations have been made:
1 . Typical grazing group oj open plains. — Hyracodon,
Oreodon, Mesohippus, Eotylopus, Poehrotherium. Note
the cropping front teeth, associated with delicately
cut and progressively long-crowned grinders, small,
compact feet, and, except in Oreodon, long, slender
limbs. Colodon may belong here.
2. Browsing group of hush country and forest. —
Titanotheres, Metamynodon, Caenopus, lEntelodon,
1 Anthracotherium, lAncodon, ''.Agriochoerus. All large-
sized fighting beasts, with coarse, heavy enamel on
cheek teeth; front teeth adapted to lip browsing.
Metamynodon may very likely have been amphibious ;
the others probably were not. Entelodon is somewhat
of an enigma; Sus is the nearest analogue but not a
close one.
3. Small hush or forest-dwelling browsers. — Hetero-
meryx, Leptomeryx, Trigenicus. Analogous to the
modern tragulines and probably of similar habits.
4. Carnivora. — The hyaenodonts are analogues of
the wolves. The ancestral canids are analogues of the
mustelines and viverrines. True mustelines are
scarce. Dinictis is the only cat.
5. Rodentia. — Rabbits much like modern "cotton-
tails" of the Great Plains. Heteromyids have ap-
peared, but no true mice (Muridae) imtil the middle
ENVIHONMENT OF THE TITANOTHEEES
121
Oligocene. Ischyromyids are abundant and include
terrestrial (?) and arboreal (?) forms; whether fos-
sorial forms existed or not is not proved. Eutypomys,
though referred to the Castoridae, is not at all analo-
gous to the modern beaver but rather to a large
squirrel or spermophile.
6. Insedivora. — The leptictids have rather sharp-
cusped teeth and are intermediate in type between
opossums and tree-living erinaceids. The moderate
wear of the teeth is evidence against the theory that
their food was worms or other terrigenous forms.
There are no obvious arboreal adaptations in the
limbs and feet; perhaps they may have been semi-
arboreal. Their survival, unaltered as to cheek
teeth, from the basal Eocene is suggestive of some
special protection, such as spines. As for the zalamb-
dodonts, they may have been fossorial, Xenotherium
being molelike, but the evidence is insufficient.
7. Marsupialia. — Rare. Precisely like small opos-
sums in the structure of the teeth.
8. Aves. — No birds have been recorded in this
fauna, although they probably existed and may have
been even numerous and varied.
9. Reptilia. — Crocodiles and trionychids occur in
the sandstone lenses; probably they were aquatic
forms analogous to modern crocodiles and soft-shell
turtles. In the clays Testudo occurs; also Xenochelys,
probably similar in habits to modern land tortoises
and marsh turtles. The lizards are apparently analo-
gous to the Gila monster and to some of the swift-
footed anguid lizards. Burrowing amphisbaenids
occur in the Oreodon zone but have not yet been dis-
covered in the lower Oligocene; no doubt they formed
part of the fauna; also other lizards and many snakes.
10. Batrachia. — No batrachians have been recorded,
but there is no reason to suppose that they were
absent or rare.
11. Pisces. — A few fragments of fresh-water fishes,
similar to those characteristic of muddy rivers of
to-day, are recorded from the Swift Current beds in
Canada. They will doubtless be found in the sand-
stones and other stream deposits of the Titanotherium-
bearing beds of the United States.
SECTION 3. ADAPTIVE RADIATION, PRIMARY AND
SECONDARY, THROUGH CHANGE OF ENVIRONMENT
A CAUSE OF THE DIVERSIFICATION OF THE TITANO-
THERES
HABITAT OF THE UNGULATES
The present geographic features of modern equato-
rial Africa, consisting of a high central plateau, river
borders, savannas, and forests, exhibit a close parallel
to what we believe were those of the known titanothere
region of North America in Eocene and lower Oligocene
time. These conditions may also be compared with
those found in the existing flood plains at the head-
waters of the great rivers of South America east of
the Andes in the warm temperate and subtropical
but not in the tropical belt.
101959— 29— VOL 1 10
Adaptive radiation: Favorite habitats of existing
perissodactyls and elephants
[See fig. 78]
RHINOCEROSES
Rhinoceros sondaicus. Java. Typically a forest dweller,
occasionally found in alluvial swamps. A browser.
Rhinoceros (Dicerorhinus) sumatrensis. Hilly forest districts
of Sumatra. A browser.
Rhinoceros {Opsiceros) bicornis. Bush-covered country and
open plains; forested foothills in the dry season. Fairly abun-
dant on the top of the Aberdare, British East Africa (elevation
9,000 feet). A browser, feeding on shrubs, roots, leaves, etc.
Rhinoceros unicornis. Grassy jungles of India. A grazer.
Rhinoceros (Ceratotherium) simus. Savannas and grassy
plains, with swamps or water holes for wallowing. A grazer.
TAPIRS AND ELEPHANTS
Tapirus roulini. Pinchaque tapir of the high region of the
Andes and Cordilleras. A browser.
Elasmognalhus bairdi. A hill dweller, seeking lowlands
during rainy seasons. A browser.
Tapirus terrestris. A forest dweller. Lowlands of Brazil
and Paraguay. A browser, feeding on palm leaves, fruits,
water plants.
Tapirus indicus. Lowlands and forests of India. A browser.
Elephas (Loxodonta) africanus. Less typically a forest
animal than E. indicus. Savannas, dry country, and forests.
Ranges from the seacoast to points beyond the alpine heath
zone of Mount Kenia and the bamboo belt of other African
mountains. Ascends and descends steep places with wonderful
facility. A browser and grazer.
HORSES AND ZEBRAS
Equus burchelli. Essentially a plains dweller; often found
in sparse savannas.
E. grevyi. Grevy's zebra. Low plateaus, thorn bush and
feather grass country that has gravelly soil. Essentially a
dweller in open plains and savannas.
E. quagga (extinct). The quagga. A karoo dweller. Fre-
quents open, arid plains.
E. zebra. Mountain zebra. Hilly and mountainous country.
E. przewalskii. Przewalski horse. Gobi Desert. A steppe
dweller.
ASSES
Equus asinus. Abyssinian ass. Wiry hedge and upland
country.
E. hemionus kiang. The kiang. Desolate plains in the
vicinity of lakes and rivers. High table-lands of Tibet (15,000
feet) . Coarse wiry pasture and rough hard yellow grass.
E. asinus somalicus. Striped African ass. Borders of the
Nubian Desert.
E. hemionus onager. Persian wild ass. Migrates from the
plains to the hills in summer. The onager of Persia.
POLYPHYLY AMONG HOOFED MAMMALS
THE TITANOTHEEES AND OTHER EXTINCT FORMS
It is astonishing to find within relatively small
geographic areas both Eocene and Oligocene remains
of many kinds of titanotheres, which lived close
together under very similar climatic conditions, the
more so because the known geographic distribution of
the titanotheres in Eocene time is confined to the cen-
tral Rocky Mountain region and extends only from the
Wind River Basin of Wyoming on the north to the
White River Basin of Utah on the south, a distance of
122
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
kl
^ ^
■5
1^
■<5 ^
~ci ^
\
ro
.fo
1
~^
S
^
f:;
^
/2.O00 to /O.OOO
6.000 fo 2,000
/oo'
^ ^ ^
« ^
I S
>0
Vi
I
I
/-/yr3<zoc/o/?/s
D. s'ama-
frensis
(Suma/ra)
/?. i/n/cor/7/s
O. 6 /cor/7 is
C/Hfr^'ca)
/Icerafheres
(£i/rope, /is/ 3. M /l/rjer/c^)
/?. SO/7c/a/CUS
/l/77y/?oc/o///s
Me/a/77y/?oc/o/7/s
Z~. a/^r/C(9r7us-
£'. 3/'/-/C3/?tJS
fSat^ y/7n3/7S o/~ /^/r/ c^)
^/ep//as //7£7/
Ps/ae o/r?3sfo -
A^as? oa^o/7/s
fFar ?s/s o/'/ne//'^)
^//>p opo fa/77/
Z?C/^0/7^
N
£ as/'nus
{/Africa)
£.przew3/sA/
(y^sia s/eppesj
£. ze.&r^
fj^fr/ca p/3:^e.aus)
NypoAippic/S—
ttypc h/pp/c/s
£. zebrs
£. iurc/ie///
£.^revy/
£. ^cjsgga
■5
Tap/rus
C/t/g/7
(Soc/M
/Indi'^s)
A/7?e r/ca)
Tap/rus
ba/rc//
^oL//h /imer/'ca)
Tap/ras
/nd/cus
77 smer/canus
(/.Oiv/anc/.r S. Amer/ca)
Te/ma'fheres
(H/^h p/3ins)
Menodonfs—
Symboroc/onfs
(Hi'//y regions)
Man/eoceraf/nes
Bronfofheres
^f/oocf p/aJ/7s)
-Menodonts
Bron fop/n es
DoZ/chorh/nes
Mefarhines
Pa/aeosyop/nes
Figure 78. — Geographic cross section showing the nature of the habitats of the larger existing ungulates
and of the titanotheres as illustrating adaptive radiation
The upper row shows the present geographic distribution ot the ungulates in continental Africa and the theoretic geographic features of the
Rocky Mountain region in Eocene and Oligocene time— namely, high valleys, plateaus, foothills, plains, river valleys, flood plains, bot-
tom lands, and river and lake borders. The second row shows the corresponding present distribution of the plant foods of different
types of browsing and grazing, cursorial, graviportal, and semiaquatic quadrupeds. The four next lower rows show, in descending order,
the corresponding adaptive radiation of the rhinoceroses, extinct and living; of the elephants and mastodons and the typical aquatic hippo-
potami and sirenians; of the plateau, plains, and forest types of horses; of the mountain, foothill, and lowland types of tapirs. The bot-
tom row shows the theoretic adaptive radiation of the principal types of titanotheres— telmathcres and menodonts of the higher levels;
symborodonts in the foothills; manteoceratines, brontothercs, brontopines, and menodonts on the flood plains; dolichorhines, metarhines,
and palaeosyopines on the lowlands and river borders.
ENVIRONMENT OF THE TITAN OTHEEES
123
480 kilometers (298 miles). The continental extent
of the distribution of the titanotheres, which is still
unknown, was undoubtedly far greater, including,
perhaps, the larger part of the North American
continent and certainly extending into Asia. In
Oligocene time the known geographic distribution
was somewhat larger, including an area extending from
Colorado to southern Alberta and measui'ing from
north to south about 1,200 kilometers (746 miles).
Titanotheres lived also in eastern Europe, both in
Transylvania' and Rumelia, also in Mongolia.
Our present Icnowledge of the geologic horizons of
the titanotheres is still extremely meager regarding
certain strata. The extent of our knowledge is sum-
marized below.
Geologic horizons of the known genera and subgenera of the
titanotheres
Lower Oligocene; upper, middle, and lower levels: Brontops,
Diplocloniis, Alloys, Menodus, Brontotherium, Megacerops.
Upper Eocene; Uinta C (true Uinta) : Telmatherium, Man-
teoceras, Diplacodon, Prolitanoiherium, Eotilanolherium.
Upper Eocene; Washakie B and Uinta B: Metarhinus, Rhadi-
norhinus, Mesatirhinus, DoUchorhinus, Manteoceras, Tel-
matherium, Diploceras.
Middle Eocene; upper part of Bridger formation : Mesatirhinus,
Manteoceras, Palaeosyops, Telmatherium.
Middle Eocene; lower part of Bridger formation: Limno-
hyops, Palaeosyops, Eometarhinus.
Lower Eocene; Wind River formation: Lamhdotherium, Eoti-
tanops.
As compared with what we observe among the hoofed
animals living to-day these titanotheres certainly
dwelt near one another under very similar conditions
of climate but in different feeding ranges and local
habitats; they sought the same watercourses, and
their remains were entombed in similar deposits. As
the whole tendency of discovery up to the present
time has been to multiply the phyla, to separate and
diversify the titanotheres, the probability is that many
other kinds of titanotheres lived in other parts of
North America and Asia.
The evolutionary principle underlying these diversi-
ties Osborn (1902. 214, p. 353) has called adaptive
radiation, which is the application to paleontology of
the idea of divergence as conceived and developed
successively by the studies of Lamarck, Darwin,
Huxley, and Cope." Radiation is a broader principle
than divergence, because it implies evolution in every
direction possible to the organism. The idea of
radiating branches from central forms assists the
imagination, because the known radiations of extinct
animals must be supplemented by the unknown radia-
tions, and it is most remarkable how these missing
radii have been discovered in group after group of
animals. Such adaptive "radiation" is either "con-
tinental"— that is, it occurs where diversities in food,
" See also Osborn, H. F., 1902.214; 1905.267; 1910. 345; 1910. 346; Stevenson-
Hamilton, J., 1912. 1; Sclater, P. L., 1894. 1; Lydekker, K., 1893. 1; Gregory, J. W.^
1896.1; Blanford, W. T., 1888.1; Kobelt, W., 1902.1; Schimper, A. F. W., 1903.1;
Lonnberg, E,, 1912. 1; Roosevelt and Heller, 1914. 1.
soil, or climate prevail over large areas — or "local" —
that is, it occurs where marked diversities prevail in
relatively small areas. The radiation among the
titanotheres in southern Wyoming and northern Utah
seems to have been largely "local," indicating that the
physiography of the mountain basin was highly
diversified.
One of the results of adaptive radiation is poly-
phyletic evolution, the existence within families of a
large number of independent minor branches that
may pursue more or less divergent evolution in local
or continental regions but that may come together in
river and flood-plain basins, so that their fossil re-
LIMBS AND FEET
Short-limbed, plantigrade,] AMBULATORY
pentadactyl, ung:uicu-> or
late stem ^ ^^JTORRESTRIAL
CURSORIAL
Digitigrade
OMNIVOROUS
(Grass
Herb
Shrub
Fruit
^. „.„„o|..„.. ™<"
"~-^^[Carr'
MYRMECOPHAGOUS
Dentition reduced
Stem INSECTIVOROUS
Figure 79. — Original radiation of the unguligrade
Herbivora, Carnivora, and Inseotivora, showing
the adaptations of teeth, limbs, and feet to various
habits and environments
mains are found in the same localities and deposits.
Polyphyletic evolution has been discovered so fre-
quently, among both the mammals and the lower forms
of life, that it may be considered the rule and mono-
phyletic evolution along single lines the exception.
Some of the examples of polyphyletic evolution among
extinct mammals that ha^'e been determined in com-
paratively recent years are the following:
Contemporaneous
branches, or phyla
Oreodonts (Cope, Wortman, Peterson, Matthew, Doug-
lass) 7-9
Lophiodonts (Osborn, Deperet) 5-7
Anthracotheres (Stehlin, Deperet, Andrews) 6-8
Rhinoceroses (Osborn) 8-9
Horses (Osborn, Gidley, Matthew) 8-9
Titanotheres (Osborn) 10-12
Elephants and mastodons (Osborn) 7-10
124
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
THE EXISTING AFRICAN ANTELOPES
The polyphyly among the titanotheres and other
extinct Perissodactyla presents a marked contrast to
the impoverished conditions among tlie existing mem-
bers of the same order wlien we consider that in all
parts of Asia and Africa only five kinds of existing
rhinoceroses can be distinguished by the characters of
the skeleton and teeth alone, that only six or eight
kinds of horses, asses, and zebras in the same great
region can be distinguished by their hard parts, and
that, similarly, among the tapirs of Asia and South
America only three kinds can be distinguished. This
contrast between present monophyly and former poly-
phyly is due to the fact that the order Perissodactyla,
though formerly a dominant group, is now a declining
group.
In the existing Bovidae, especially those in the groat
continent of Africa, we have a parallel to the ancient
polyphyly of the titanotheres and other Perissodactyla.
The Bovidae is a family that includes the cattle and
antelopes and that is now in the highest stage of ra-
diation and adapted to a great variety of physiographic
and biotic conditions, as shown in the primary and
secondary adaptations in the seven subfamilies of the
African antelopes.
The African antelopes: Subfamilies, habits, and environment
Subfamilies and habits
Antilopinae (browsers and grazers)
Gazelles
Pallahs (impalas)
Springbucks
Gerenuks
Saigas
Bubalidinae (mostly grazers) :
Gnus
Hartebeests
Blesboks
Sassabies
Tragelaphinae (browsers and graz-
ers):
Elands
Koodoos
Bush bucks
Bongos
Situtungas
Hippotraginae (grazers) :
Roan antelopes
Sable antelopes
Gemsboks
Addaxes
Neotraginae (browsers and graz-
ers):
KUpspringers
Oribis
Dik-diks
Cephalophinae (mostly browsers) ;
Duikers
Environment
Plains and deserts.
Thorny bush and glades.
High veldts.
Deserts and bush.
Steppes.
Open plains.
Open forests or plains.
Open rolling country.
Open forests and flats.
Stony hills.
Forests.
Dense forests.
Swamps and lagoons.
Thin forests.
Rolling uplands.
Open deserts.
Waterless deserts.
Hills, mountains.
Thin forests.
Dense forests and bush.
The African anteloyes: Subfamilies, habits, and environment —
Continued
Subfamilies and habits
Environment
Cervicaprinae (grazers on suc-
culent plants near water) :
Open forests and stony
hills.
Reed swamps, river bor-
ders.
Open swampy plains.
Slopes of hills.
An incipient or attempted adaptation to a grazing
life is seen in the teeth of certain titanotheres. Most
titanotheres are browsers. Broadly speaking, her-
bivorous animals that live on open plains are grass
eaters and tend to become gregarious in habit and
cursorial in locomotion, whereas those that prefer the
shady depths of the forests are browsers, are of soli-
tary habit, and are mediportal in locomotion. There
are exceptions, such as the black rhinoceros (Rhi-
noceros {Opsiceros) Mcornis), which now frequents the
treeless plains of East Africa but which is habitually
a browser, although it is at times a grazer. The long-
necked giraffes are fond of rather dry and fairly open
country and are not found in strictly forested regions,
yet they are wholly browsers, being especially fond of
the leaves of certain thorny acacias, notably Acacia
girajfa, and the related short-necked okapi, which is
found only in the dense forests of the Congo, is a
browser.
The principles of adaptation shown in the skull and
teeth of Perissodactyla to browsing and grazing habits
are described in Chapters V and VI of this monograph.
The adaptation of the limbs of the Perissodactyla to
speed and weight are described in Chapter VII.
In general, the competition and range for food
among hoofed animals is accompanied by lengthen-
ing of the limbs from medium-paced (mediportal)
types to either swift-moving (cursorial) types or heavy-
bodied (graviportal) types. Similarly, adaptation of
the grinding teeth to browsing habits is seen in the
short-crowned (brachyodont) types, and transition to
the grazing habit is accompanied by lengthening
(hypsodontism) of the crowns of the grinding teeth.
Such changes are accompanied by changes in the pro-
portions of the head to adapt the action of the teeth
to browsing or to grazing. We observe a passage
from short-headed (brachycephalic) to long-headed
(dolichocephalic) forms of skull. In adaptive radia-
tion every possible combination of lengthening and of
shortening of skull, tooth, limb, and foot may arise,
as well as notable coincidences of structure in different
forms, for similar kinds of food may be found and
ENVIRONMENT OP THE TITANOTHEEES
125
similar feeding habits may be acquired in widely
separated habitats or greatly different environments.
Contrasts in structure, such as those shown below,
are equally notably.
Contrast in structure between browsing and grazing types
Browsing types (brachyodont)
Grazing types (hypsodont)
Short-headed (brachycephalic)
Long-headed (dolichocepha-
lic).
Straight-headed (orthooephalic) _ _
Bent-headed (cyptocepha-
hc).
Short-limbed (braehymelic, bra-
Long-limbed (dohchomelic,
ch ypodal) .
dolichopodal) .
Grazing on the harder
siliceous grasses of dry
plains and uplands
Gazelle
Addax
Grazing and browzing on
the tender leaves and
twigs of plains, thin for-
est and brush country
Sable and roan
Grazing and browzing
on the tender grasses
of moister land and
swampy plains
"Puku
Cob
Reedbuck
Leaf, bark, and twig
eaters in forests
Duiker
Browsing on tender
leaves and shrubs
of partly forested
countries
Bushbuck
Waterbuck
Figure
Browsing on succulent
aquatic plants
of swampy lands
Sitatunga
Lechwe
). — Adaptions in the structure of the skull and teeth
of Herbivora to diverse habits of feeding
Double or even multiple adaptive radiation is
continually in operation, first, in the structure of
skull and tooth, which is dependent on the nature of
the food, and, second, in the structure of foot and
body, which is dependent on the nature of the soil.
Thus may arise cursorial (long-limbed) grazers (long-
toothed), graviportal (heavy-limbed) grazers (long-
toothed), or cursorial (long-limbed) browsers. There
is no fixed law of correlation of structure of skull and
tooth such as was supposed by Cuvier. The law of
correlation as restated by Osborn (1902.214) is as
follows :
Structure of feet (correlated chiefly with structure
of limb and body) and structure of teeth (correlated
chiefly with structure of skull and neck) diverge inde-
pendently in adaptation respectively to obtaining
food (by feet) and eating food (by teeth) in different
environments. Each structural feature is evolved
directly to perform its own mechanical functions or
purposes, yet in such a manner that each is consonant
with the other.
CONTINENTAL ADAPTIVE RADIATION OF THE AFRICAN
ANTELOPES
The African antelopes are divided into seven sub-
families, all mediportal to cursorial in limb structure
but widely different in tooth and skull structure, as
shown in the table on page 124.
The 133 or more species (Sclater, 1894.1) embraced
in these seven subfamilies seek food and protection
from enemies on the varied surface of the African
continent in habitats including no less than 17 differ-
ent kinds of country.
Each type of habitat has food peculiarly favorable
to certain feeding habits to which the structure of the
teeth and skull is speciflcally adapted. Each type of
CURSORIAL
Distance carrying
Gazelle
Gemsbuck (Oryx)
Addax
SALTATORIAL
Leaping and springing
Springbuck
MEDIPORTAL ^ Klipspringer
Medium-weight bearing
Hartebeest
Gnu
Sable and
antelopes
SEMI-ARBOREAL
Progressively on
branches of trees
Impala
GRAVIPORTAL
Heavy-weight bearing
Greater kudu
E'and \ \ FOSSORIAL
Digging and uprooting
AMPHIBIOUS
Swamp and river-living
Waterbuck
Puku
Cob
AQUATIC Reedbuck
Partly fluviatile, largely
feeding and seeking
safety in the water
Lechwe
Sitatunga
Figure 81. — Convergent adaptations in the structure of the
limbs and feet of ungulates
Certain gazelles are independent of watercourses. The adaptive radiations
indicated above occur independently within different subfamilies.
habitat demands modiflcations of limb, foot, and hoof
structure for movement in the search for food and
escape from enemies.
The theory of the evolution of the antelope is that
in mid-Tertiary time a divergent primary radiation
126
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
divided them into seven subfamilies, each with its
distinctive mode of life. During a long period of
geologic time the Bovidae have undergone secondary
radiations (A-Q, fig. 82), by which certain branches of
these subfamilies have become adaptively convergent
toward certain branches of other subfamilies through
the adoption of similar habits and habitats. Thus,
analogous genera and species arise independently in
each subfamily. For example, waterless deserts were
sought both by the addax, among the Hippotraginae,
and by the gazelle, among the Antilopinae; reeds,
river borders, and lagoons were sought both by certain
Figure 82.-
c
-Adaptive radiation in the feeding habits of antelopes, as observed by
Stevenson-Hamilton in 1912
1-7, Primary radiations; A-Q, secondary radiations
Cervicaprinae, such as the lechwes and kobs, and by
certain of the Tragelaphinae, such as the situtungas.
ADAPTIVE RADIATION IN THE FEEDING HABITS
ANTELOPES
OF
The habits and habitats of the antelopes, as noted
by Stevenson-Hamilton (1912), are as follows:
1. Antilopinae. — The impalas {Apyceros melampus) eliiig to
neighborhoods of dense thorny bush, to which they fly for
refuge. More partial to brovv^sing than to grazing. Food con-
sists largely of leaves and shoots, but they eat young and tender
grass freely after early rains. Staple diet leaves and fruit of
certain acacias, also twisted bean pods of the same. In March
fruit of marula is eaten. Toward the end of the dry season
they completely strip the bush of everything edible up to the
extreme height which they are able to reach. * * * The
springbucks (Antidorcas euchore) are typical of the high veldt
fauna of South Africa. The only member of the gazelle group
in this region. Love high, open tablelands. * * * Xhe
typical African races of gazelles include 14 species. Grant's,
Thomson's, Speke's, etc. Inhabitants of wide, open plains or
sandy deserts. Largely independent of water.
2. Bubalidinae. — Antelopes of large size, large, moi.st rhinari-
um; including Buhalis (= hartebeest), Damaliscus (= bastard
hartebeest), Connochaetes (= gnu, or wildebeest). Buhalis
(= hartebeest), eight species, with everywhere same charac-
teristics; frequent open or forest countrj' or treeless plains;
essentially grass eaters; like to drink
regularly. Young carried about eight
months. * * * Damaliscus, bonte-
buck {D. pycargus), blesbuck, tsessebe
(sassaby) (both D. albifrons), grass-eat-
ing antelopes, favoring rather open and
fairly flat country, never hills or thick
jungle, partiality for shady patches of
bush or forest for shelter during the hot
hours. * * * Connochaetes (gnu, or
wildebeest), white-tailed or black (C
()7iu) and brindle or blue (C. tautrinus).
Prefer open, rolling country interspersed
with thick thorn or other bush. Some-
times remain in the open, bare spaces
or plains where they can see for long dis-
tances. Essentially grass-eating ani-
mals. Pasture cropped closely. Socia-
l5le, gregarious.
3. Tragelaphinae. — Elands and bush-
bucks, inyalas, kudus, situtungas. Elands
{Taurotragus oryx), plains type, graze
with horses, donkeys, and cattle but
browse by preference, favoring the grass
only when fresh and green, sometimes
cropping the tops of young river reeds.
Gestation period eight and a half months.
* * * Bongo {Boocercus eurycoros),
fond of the most dense forest, leaves and
twigs of a certain kind of undergrowth,
which grows from 6 to 8 feet in height.
Young shoots all nipped off if bongo
have been feeding. Also (?) bark eaters.
* * * Bushbucks {Tragelaphus scrip-
txis), forest dwellers, solitarj', nocturnal,
prefer densely wooded gullies, or kloofs,
of South Africa. Browsers on the leaves
of various small shrubs and trees; eat
grass sparingly when the latter is fresh
and green; roots and tubers form fur-
* * * Inyalas {Tragelaphus angasi),
Exhibit intense localization, probably due
to the presence of some peculiar foodstuff, limited in quantity
but necessary to the health of the individual animal. Probably
browse on various leaves, shrubs, and fruits, bean pods and
acacias, fruit of the marula; grass eaten when it is young and
of good quality. * * * Sitiitnnga.s (Tragelaphus spekei) , semi-
aquatic animals, almost amphibious by nature. Great elonga-
tion of hoofs. Strong swimmers. Rapid locomotion upon dry
land very difficult. Frequent extensive reed and papyrus swamps
bordering lakes and large rivers. * * * Kudus {Strepsiceros
strepsiceros) , love stony or rather broken ground, covered with
thorn scrub. Gregarious, more than most antelopes, a browser,
subsisting chiefly on the leaves of thorn acacias and bush shrubs,
together with the fruits of the marula and other trees.
ther articles of diet,
very local and rare.
ENVIRONMENT OF THE TITANOTHERES
127
4. Hippotraginae. — Sable and roan antelopes Hippotragus,
oryx and addax, distinguished by the presence of horns in both
sexes and small rhinarium or bare space on the muzzle. Sable
(H. niger) to a great extent, though not entirely, a grass eater.
Prefers thin forest country, interspersed with alternate thickets
for shade, and open vleis for grazing. Regular drinker, seldom
found more than a few hours from water. Gestation period
about 270 days. * * * Roan antelope {H. equinus) favors
rather upland, rolling country, not too thickly wooded, such
as the middle veldt, but when persecuted takes readily to
forest or the same environment as the sable antelope. A
grass eater, and drinks regularly * * * [Genus Oryx.] The
gemisbuck (0. gazella) of South Africa separated from its nearest
generic relative (0. beisa) of German East Africa by an interval
of 1,500 miles. Fairly numerous in Kalahari Desert, main-
taining its security owing to its independence of water, able
to quench thirst from moist tubers and roots. Generally found
in small troops. The beisa (0. beisa) inhabits the Kilimanjaro
district, British East Africa, Somaliland and the Sudan, east of
the Nile. Sometimes found in herds of 50. Period of gestation
eight and a half to ten months. White oryx (0. leucoryx) is
found west of the Nile. Essentially a desert animal and like
the gemsbuck apparenth' associates in small parties. [Genus
Addax.] The addax [A. nasomacidatus\ distantly related to
both oryxes and roan and sable antelopes, pale sandy color.
An inhabitant of waterless sandy deserts of northern Africa.
5. Neotraginae. — Klipspringers {Oreotragus orcotra.gus), like
the chamois, prefer small shrubs and grasses growing among the
stones. Live on natural moisture of the grass and nightly
dews. * * * Oribi (Ourebia) frequents open grass country
or plains not too thickly forested. Grass feeders, seldom found
any distance from water. Eight species.
6. Cephalophinae. — Lovers of dense bush and forest of central
and southern Africa. Thirty-eight species. Duiker (Cephalo-
phus grimmi), solitary animal, fond of bush country. Never
far from covert. Mainly browsers. Nibbles leaves and young
shoots of various acacias, small shrubs. Grass consumed
when. young and fresh. Red duiker (C natalensis), dense
forests and bush. Blue duiker (C monticola), essentially a
browser, favors shelter or dense covert.
7. Ceruicaprinae. — Animals of large or medium size. Water
buck (Cobus ellipsiprymiius), open forest country, eastern
Africa, favor banks of large rivers, prefer succulent herbage,
but are partial to rough and broken country, stony hillsides,
and vicinity of fairly thick bush; grass feeders. During dry
season frequent banks of streams for succulent herbage.
* * * Sing-sing water buck {Cobus defassa), habits similar
to above. * * * The lech we (C. lechwe) is smaller than
the water buck. Hoofs elongated and pointed. Frequent
great reed swamps and river borders, northern Rhodesia.
Next to the situtunga, the most aquatic of all antelopes, stand-
ing knee or even belly deep in large shallow lagoons. Come
ashore to graze, food consisting of grass and young reeds.
* * * Gray's water buck (C maria), frequent river bottoms
and reedy grass. Stand in shallow water. * * * Puku
(C. vardoni), less aquatic than the lechwe, approaching in this
respect the water buck — that is, found close to but not in the
water. Frequent swampy plains. * * * Uganda cob (C.
thomasi), fond of open, rather swampy plains, near rivers or
permanent water. Grazes on young shoots of grass. * * *
Common reedbuck {Cervicapra arundinum) , lowlands of Natal
and Zululand, Transvaal bush country, etc. Favors grassy or
reedy valleys near streams or permanent water of some kind.
Occasionally met with in thin bush. Food consists entirely of
grass. Do not take to water when alarmed. * * * Moun-
tain reedbuck {Ceruicapra fulvorufula), lower slopes of hills
covered with rocks and loose stones, mingled with scattered
bush and long grass. Grass eaters, at night descending from
hills to nearest w-ater. Affecting sides rather than tops of
hills. * * * Bohor reedbuck {Cervicapra redunca),, favor-
ing open vleis and bush or swamp land. Like the neighbor-
hood of water. * * * Gray rhebuck {Pelea capreolus) ,
unlike mountain reedbuck, frequent flat tops of the table
mountains; common in South Africa as well as higher levels of
the ranges. Grass feeders, and descending at night to drink
after the manner of the mountain reedbuck.
[Note vertical physiographic distribution of the genus
Cervicapra.]
CAUSES OF VARIATION AND POLYPHYLY AMONG
QUADRUPEDS
Change of physical environment. — A series of meteoric
and biotic changes — that is, changes of season, of
climate, or of rainfall, the appearance of new enemies,
the introduction of new plants or the crowding out of
old ones — will cause a change of food supply, which
will cause a change of habitat, which in turn will
cause a change of browsing or grazing habits that
will affect locomotion — the use of the limbs in the
search for food — and modify the form of the hoofs,
because of the change of soil. The browsing mountain
moose (Alces) of eastern Idaho, for example, has a
hoof of very different form from that of the water-
living forest moose of Maine. Among the new
enemies that may appear are certain insect pests,
such as flies or ticks, which may drive quadrupeds
away from feeding ranges that are otherwise favorable
into regions, perhaps not far distant, where food is
scarcer and the general conditions are more adverse,
and where, perhaps, the young are exposed to new
dangers.
Such changes may bring about (1) a change of
habit or (2) a change in habitat or environment,
either of which, as a general law, culminates in (3)
change of function, followed by (4) change of struc-
ture. (5) A change of function or habit certainly
brings about a new "incidence" of selection or new
set of causes tending to survival or extinction.
Change of appetite. — Variations in appetite are un-
doubtedly among the chief causes of local divergence.
Stevenson-Hamilton (1912.1, pp. 97-158) noted the
fastidious choice of food by each of the principal
species of African antelopes, and other wild animals are
very fastidious and seek an astonishing variety of food
in the course of a single season. The predilection for
certain kinds of food is very strong, and departures
from it lead to adaptive radiation. Similarly Sampson
(1905.1) records that the white-tailed deer (Odocoileus
virginianus) browse on many kinds of plants in the
course of a year.
Local polyphyly through reunion of phyla. — Animals
that have diverged through migration or through
geographic segregation or separation may later be
brought together in one region. For example, the
mule deer {Odocoileus hemionus) and the white-tailed
deer (0. virginianus), which have probably evolved in
different regions of the United States, are now found
128
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
together in the same region in the West. In Miocene
time the American rhinoceroses were joined in the
western plains by certain European rhinoceroses.
Thus continental radiations from great countries like
Africa, Asia, or America may pour some of their
branches into a single small region, mingling many
distinct phyla.
Hypsodont or grazing types may mingle with brach-
yodont or browsing types in the same locality through
their choice of grasses or of shrubs as their principal
article of diet. Independently in the same region in
southern Wyoming two of the branches of the titano-
theres {Telmatherium and DolichorTiinus) began to
acquire long-crowned teeth, while two others {Palaeo-
syops and Limnohyops) retained persistently short-
crowned teeth.
HABITS OF THE EHINOCEROSES PARALLEL TO THOSE OF
THE OLIGOCENE TITANOTHERES
Mingling of hrowsing and grazing rJiinoceroses in
Africa. — In equatorial Africa the Nile is an insuperable
barrier between two species of rhinoceros, the "white
rhinoceros," which is confined to the west bank, and
the "black rhinoceros," which ranges along the east
bank; yet these two species were formerly found
together in the same regions of South Africa. The
large grazing "white rhinoceros," R. (Ceraiotheriwn)
simus, has hypsodont teeth and grazes in the open
country, particularly in the wide, grassy valleys,
though it was frequently met on the high veldt of
Matabele and Mashonaland, feeding at night or in the
cooler parts of the morning and evening. Its food
consists entirely of grasses. Its sight is bad, but its
scent and hearing are acute. On the other hand, the
smaller browsing "black rhinoceros," R. (Opsiceros)
hicornis, which has brachyodont teeth, was formerly
common on the slopes of Table Mountain and on the
Cape Flats and closely overlapped R. (Ceratotherium)
simus in certain parts of its range; it frequented bush-
covered country more than open grass lands and was
often found in rocky, stony districts. It is partly
nocturnal in its habits. Its food consists entirely of
leaves, twigs, and sometimes of the roots of certain
bushes and shrubs, but seldom of grass (Roosevelt and
Heller, 1914.1). Its adaptations are essentially those of
a browser, for it prefers the twigs and small roots of
certain shrubs which it finds on the treeless plains of
East Africa (Stevenson-Hamilton, 1912.1). It has a
considerable vertical geographic range,'- being found
also on the high plateau near the glaciers of Mount
Kenya. (J. W. Gregory, 1896.1, p. 267.)
Habits of Asiatic rhinoceroses. — The existing species
of Asiatic rhinoceroses differ in habitat; they do not
mingle. Rhinoceros unicornis or indicus, which has
relatively hypsodont grinders, frequents the swampy,
grassy jungles of the plains of India. The R. sondaicus
" Gregory attributes this range to the white rhinoceros, but his observation
actually refers to the black rhinoceros, as Heller has pointed out.
of Burma and Java has shorter grinders. As observed
by Blanford (Lydekker, 1893.1, vol. 2, sec. 4, p.
470), it "is more an inhabitant of the forest than of
the grass, and although it is found in the alluvial
swamps of the sudarbans, its usual habitat appears
to be in hilly countries. It has been observed at
considerable elevations both in Burma and Java."
Indeed there is much evidence that it probably ascends
occasionally to as much as 7,000 feet above sea level.
Its food consists largely of twigs and smaller branches.
The third species of Asiatic rhinoceros, the Sumatran
rhinoceros {R. (Dicerorhinus) sumatrensis) , which has
relatively short-crowned teeth, inhabits hilly forest
districts and has been observed in Tenasserim at an
altitude of 4,000 feet above the sea. According to
Lydekker, it is a good swimmer and is said to have
been seen swimming in the sea in the Mergui Archi-
pelago, possibly traveling in search of new feeding
grounds or to avoid certain unfavorable conditions.
Thus we find among the rhinoceroses three lines of
adaptation to habitat and to food radiation — first, both
hypsodont (grass-loving) and brachyodont (browsing)
forms; second, a considerable geographic vertical range
both in R. (Ceratotherium) simus and R. sondaicus;
third, the occasional assumption of semiaquatic habits.
All these conditions were partly paralleled among
the Oligocene titanotheres, which, however, attained
no extreme hypsodontism.
HABITS OF THE EXISTING TAPIRS PARALLEL TO THOSE OF
THE EOCENE TITANOTHERES
The Eocene titanotheres, although inferior in the
structure of their grinding teeth, were nearest in form
and in body adaptations to the existing tapirs. In
the Tapiridae we find these principles of adaptive
radiation — great vertical geographic range, including
choice between upland and lowland habitat, and
assumption of more or less aquatic life. The teeth
are short-crowned (brachyodont), are crested (lopho-
dont), and are superior in mechanism to the cone and
crescent (bunoselenodont) grinders of the titano-
theres. These principles are observed as follows:
1. According to J. E. Gray (1872.1, p. 486) Tapirus
pincJiaque ascends to very great heights in the Andes.
M. Goudot "obtained a young female tapir at an
elevation of about 1,400 meters — nearly up to the
snow level on the Peak of Tolima in New Granada —
about 1843." According to Gray (1872.1, pp. 487,
488) Tschudi, in the "Fauna peruana" (p. 213), says,
" This species of tapir [T. roulini] is found in Peru on
the eastern slope of the Cordilleras at an elevation of
7,000 or 8,000 feet, which is above the snow line. "
2. On the other hand, the tapirs (T. hairdi) from
Mexico and the Isthmus of Panama, which have
been referred to the genus Elasmognathus by Gill, are
more generally confined to the lower hills or occupy
an intermediate habitat. Captain Dow observes
(1867.1, p. 214):
ENVIRONMENT OF THE TITANOTHEKES
129
Thus far all examples of T. bairdi [Elasmognathus] have been
found exclusively on the Atlantic side of the Isthmus [of
Panama], and north of the Chagres River. Their favorite
haunts appear to be in the hills lying at the back of Sion Hill
and the adjoining stations of the Panama Railway. It is only
during the rainj' season that they seem to seek the lowlands,
for it is only in that season they are captured.
Similarly Tapirus {Elasmognathus) dowi was found
in the highlands of Guatemala, Nicaragua, and Costa
Rica.
3. The opposite extreme from mountain-living
habits is furnished by the typical South American
tapir (T. terrestris), which inhabits the forest districts
of Brazil, Paraguay, and the northern part of Argen-
tina. This species is fond of gamboling in the water
and rolling in soft mud and swims and dives like a
capybara; it is not improbable that it may also walk
along the beds of shallow rivers and lakes, as was
observed to be the habit of a specimen of the Malayan
tapir (Tapirus indicus). In Brazil, in districts remote
from cultivation, the food of the tapir is composed
largely of palm leaves, but at certain seasons of the
year these animals subsist almost exclusively on
fallen fruits, and in some districts swampy grasses
and water plants form their chief food.
VERTICAL GEOGRAPHIC RANGE OF QUADRUPEDS
The rhinoceroses as a group have a wide vertical
geographic distribution, ranging from sea level to the
snow belt. The black rhinoceros, although it prefers
the lower grassy plains, is found also on the high
plateaus near the glaciers of Mount Kenya. (Gregory,
1898, op. cit., p. 263. '') As above noted, the tapirs
as a group range from sea level to the snow belt,
8,000 feet above sea level. Some species are exclu-
sively low-level forms {T. terrestris); others range
from sea level well up into the mountains {T. iairdi);
still others inhabit the higher Andes {T. pinchaque).
The elephants also enjoy a wide vertical range;
Elephas (Loxodonta) africanus is said to ascend and
descend steep places with wonderful facility, and
t footprints of the modern Asiatic elephants have been
seen among the eternal snows of the highest mountains
(Pohlig, 1891.1, p. 328).
VERTICAL GEOGRAPHIC RANGE OF THE TITANOTHERES
Thus, judging by analogy with the other Perisso-
dactyla and from what we know to be true also of the
horses, it is probable that the titanotheres enjoyed a
considerable vertical geographic range in the Rocky
Mountain region in Eocene time and that this may
have entered into the causation of their local adaptive
radiation.
TEN CHIEF HABITAT ZONES OF MAMMALS
Wide climatic and physiographic differences, if
concentrated in a geographically restricted area,
facilitate local adaptive radiation. For example,
" Gregory inadvertently attributes this range to the white rhinoceros.
grassy meadows favorable to shrubs bring grazers and
browsers together. That much more extreme con-
trasts are by no means unnatural is shown along the
coasts of Mexico, where there is an abrupt transition
from an extremely moist, warm lowland to a dry, .
cool upland. Similarly abrupt transitions are ob-
served in parts of the Andes and the Himalayas.
It is consequently not difficult to account for the
fact that seven or eight different phyla of titanotheres
lived together in southern Wyoming and northern
Utah in middle and upper Eocene time, for the entire
region was varied and rnountainous.
The life zones of mammals have been set forth
admirably by Kobelt (1902.1) and should be studied in
connection with the vegetation zones of Schimper
(1903.1). Some mammals are strictly confined to
their typical habitat zones — that is, they are intensely
localized. Certain antelopes, such as Tragelaphus
angasi, the inyala (Stevenson-Hamilton, 1912.1, p.
135), probably feed upon only a single plant and are
limited in range to its distribution. Many Herbivora,
such as elephants, rhinoceroses, and horses, are very
plastic and have great diversity of habitat in the
course of the change in seasons and under varying
conditions of competition.
Life zones are defined by land and water, by mois-
ture and aridity, by depression and elevation, by low
and high temperature, by the distribution of insects,
and especially by the presence of vegetation adapted
to grazing or browsing. Life zones are therefore de-
fined sharply in some places and feebly in others.
The ten zones discriminated are described below.
1 . Mountain or alpine liaiitat. — High mountains and
mountain ranges with the snow and timber lines at
altitudes of 6,000 to 12,000 feet or more. Thinly
forested or tundra-like lands, adapted both to grazing
and browsing ungulates having relatively short limbs
and feet adapted to climbing. The Artiodactyla are
represented by many forms, some of which range far
above timber line, including goats (Capra), rupi-
caprines {Rupicapra, NemorJiaedus, Oreamnos), moun-
tain sheep (Ovis), vicunas {Lama vicunna) at certain
seasons, Pudu deer {Pudua). The Perissodactyla that
invade these high forest zones are only certain tapirs
of the Andes {Tapirus pinchaque and T. roulini).
2. Mountain forest habitat. — Lower mountain ranges
and foothills, dry or well watered, well wooded, with
river valleys. This zone includes the dry tropical
woodlands (such as those of India), which are favorable
to the larger ungulates; also the tropical rain forests
(Asia, Africa, North America), generally unfavorable
to large ungulates. In Asia the especial habitat of
many deer, bovines, antelope, browsing perissodactyls,
such as Rhinoceros sondaicus of Java, typically a forest
dweller, R. {Dicerorhinus) sumatrensis of Sumatra.
In the northern latitudes of North America, the typical
home of the deer {Odocoileus), moose {Alces), wapiti
130
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
(Cervus), mountain caribou (Rangij'er) , at elevations
of 2,000 to 8,000 feet. On these levels in South
America are found among the Tapiridae T. {Elasmo-
gnathus) bairdi, a hill dweller seeking the lowlands
during the rainy season ; also T. (ElasmognatTius) dowi.
In the equatorial belt of Africa both the high forests
and lower forested foothills favorable to the growth
of shrubs and trees attract also the elephants.
Vertical distribution qf lije zones oj ungulates
Alluvial bottom lands.
Family or phylum
Peaks and high-
est mountain
ranges; 6,000
to 12,000 feet.
Browsers
High desert and drier
uplands and plains,
table-lands, plateaus,
mesas; 5,000 to 15,000
feet. Grazers
Lower mountain ranges,
foothills, well wooded
and watered; forest
lands; tributary river
valleys; 2,000 to 8,000
feet. Browsers
Great plains and larger
river valleys, broad
grassy meadows, rolling
country; sea level to
6,000 feet. Grazers
delta and flood-plain
deposits, swamps and
jungles, forests and
partly forested low-
lands; river or sea
level. Browsers and
grazers
Rivers and lakes,
river or lake level.
Euminants
Mountain sheep,
goat, deer, and
elk (summer).
Pronghorn antelope
Deer, moose, elk, cari-
bou (winter).
Buffalo and wapiti, or elk.
Rbinocerotidae (re-
Hyracodon nebrascensis.
Rhinoceros sondaicus.
Rhinoceros unicornis- in-
Rhinoceros sondaicus.
Metamynodon
cent and extinct
Three long toes.
Java; typically a forest
habits grass jungles.
Occasionally seen in
A mph ibious
types).
Rhinoceros (Geratothe-
dweller.
R. (Ccratotherlum) simus.
alluvial swamps.
rhinoceros.
rium) simus. Meadows
R. (Dicerorhinus) su-
Large two-horned rhi-
Aceratherium. Four-
and sparse forests.
matrensis. Inhabit-
ant of hilly forest dis-
tricts.
R. (Opsiceros) bicornls.
Two-horned " black "
rhinoceros of Africa.
Often seen on slopes
of table mountains;
feeds on roots, leaves,
etc.
noceros of Africa: inhab-
itant of grassy valleys
on high veldt.
R. (Opsiceros) bicornls.
Found on Cape flats, in
bush-covered country.
Coenopus.and Dlcerathe-
rium. Three-toed ani-
mals.
toed rhinoceros.
Equidae (horses,
zebras, asses).
Equus Demionus, E. ki-
Equus zebra, mountain
Equus hemionus onager.
ang. Kianganddzigge-
zebra.
Migrates to the hills in
tai. Inhabit table-
summer.
lands of Tibet, 16,000 feet
E. asinus somallcus. In-
high. Prefer desolate
habits Nubian desert.
places near lakes and
E. quagga. The quagga
rivers, and coarse wiry
of South Africa; extinct.
pasture of rough, hard
A karroo dweller.
yellow grass.
E. burchelll. Burchell's
E. onager. The onager of
zebra; found north of
Persia.
Orange River; often seen
E. hemippus. Syrian
in sparse forests, but
wild ass.
predominantly a plains
E. asinus. Feeds on wiry
dweller.
desert grasses.
E. grevyi. Low plateaus
E. zebra. Feeds on plains
with gravelly soil. Seen
grasses.
in thick thorn bush and
E. przewalskil. Inhabits
tall feathery grass. Es-
northern deserts.
sentially an inhabitant
of the open plains.
Tapiridae (tapirs)..
Tapirus roulini
Tapirus roulini..
Tapirus bairdi. Hill
Tapirus americanus.
T. pinchaque.
T. pinchaque. Tapir of
dweller, seeking low-
Common tapir of for-
Inhabits -slopes
the high regions of the
lands at rainy season.
ests and lowlands of
of Cordilleras.
Andes.
T. dowi.
Brazil and Paraguay.
A forest dweller, feed-
ing on palm leaves,
fruits, and water
plants.
T. indicus.
Proboscidea (ele-
Elephas africanus. As-
Elephas africanus. Less
phants).
cends and descends
steep places with won-
derful facility.
typically a forest animal
than E. indicus; found
in comparatively open
country; also in forests.
E. indicus. Typically a
forest animal.
Sirenians, chalico-
Macrotherium.
Moropus.
Manatee.
theres, hippopo-
Dugong.
tami.
Chalicotherium.
Hippopotamus.
ENVIRONMENT OF THE TITANOTHEEES
131
3. Boreal forest Tiaiitat. — Characteristic of north
temperate zones with cold winters. The "temperate
rain forests" of Schimper, partly interspersed with
meadowlands. This zone includes the whole of primi-
tive northern Europe and North America south of the
tundra zone. In Asia it includes the whole of Siberia,
grading on the south into the high "steppe" and high
"plateau" regions and on the north into the Arctic
tundras or barren grounds. It is the great boreal
zone of North America, favored both by woodlands
and meadows and by sufficient rainfall. The ungulates
are very numerous, especially genera of Bovidae,
Cervidae, and Suidae.
4. Tundras and barren ground habitat. — In this low-
lying, north circumpolar region trees are scarce or
absent, except the willows and birches of the river
bottoms, and the subsoil is frozen throughout the year.
The ungulates are now represented only by the musk
ox {Ovibos moschatus) and several species of reindeer
(Rangifer); formerly by the mammoth and the horse
in Alaska and Siberia during the period of greater
forestation.
5. Higher plains and plateaus. — Mesas, table-lands
(as in Tibet and the Himalayas), and the desert
plateaus of the Rocky Mountains and Andes, altitude
3,000 to 6,000 feet or more; vegetation scattered,
sparsely forested, both grasses and shrubs abundant;
or rocky and open country with occasional forests.
Climate generally severe in winter. This zone grades
into the "high steppes" of Asia, the veldt of South
Africa, the high plains of North America. It is mostly
open country adapted to grazers with hypsodont teeth,
long limbs, and slender feet, or to the cursorial and
gregarious Herbivora.
6. High steppe and desert habitat. — Treeless and arid
wastes, steppes, and deserts of central Asia (such as
the Desert of Gobi) or of Persia and Asia Minor,
reaching an altitude of 6,000 feet, usually not so rich
in flora and fauna as the high plateau. Climate
extremely severe in winter. Inhabited chiefly by
grazers. In Asia, among the Equidae we find the
kiang {Eguus Jciang) of Tibet, the dziggetai {E.
hemionus) of Mongolia, the wild horse {E. przewalslcii)
of the Desert of Gobi or the Kobdo district of western
Mongolia. The kiang of Tibet and Turkestan prefers
desert places near lakes and rivers, seeking coarse, wiry
pasture and rough, hard grasses. The dziggetai ranges
from the lowland steppes of Turkestan to the high
plateaus (1,680 meters) of the deserts of Mongolia.
In this zone among the Artiodactyla we find the wild
Bactrian camel {Camelus bactrianus), the saiga ante-
lope {Saiga tartarica), and the Persian gazelle {Gazella
gutturosa).
7. Low desert habitats. — Steppes and sandy deserts
of northern Africa, Syria, Arabia, Mesopotamia, and
the northern borders of the Arabian Sea; rocky
countries covered with sparsely vegetated areas and
thin forests, scattered shrubs, and thorny bushes.
Except in temperature and altitude this zone is like
that of the high steppes; its vegetation is sought
mostly by cursorial browsers and grazers with colora-
tion of the desert; in Africa Gazella dorcas, Addax,
Oryx leucoryx, and among the Equidae the north
African wild ass {Equus asin'us), the Somaliland ass
{Equus somaliensis) , the Assyrian E. hemippus, and
the onager {E. onager), which grazes in the low deserts
of Kutch and Rajputana. Neither the rhinoceroses
nor the tapirs have ever had representatives in these
low-lying desert belts.
8. Plains habitat. — Great plains and larger river
valleys; broad, grassy meadows bordering glades
partly forested or not forested at all, extending from
sea level to an altitude of 6,000 feet in northern
latitudes. The tropical grasslands or savannas of
Africa, the llanos of the Orinoco, the campos of
Brazil, the semiarid karoos and veldts of South Africa
are partly included in this zone, although they also
approach the high steppe habitat. This zone is
generally adapted to grazing, hypsodont types, mostly
long-headed and cursorial. It is the natural habitat
on the Great Plains of North America of the buffalo
(Bison bison), of the pronghorn antelope (Antilocapra
americana), and formerly of the wapiti (Cervus
canadensis). Similarly on the plains of equatorial
Africa are found numerous species of antelope (mostly
grazers), oxen (grazers), giraffes (tfue browsers), the
black rhinoceros, R. (Opsiceros) bicornis (browsers and
grazers), and all species of zebra. The ungulates in
this open country are either cursorial or graviportal
and are well defended by horns. The Tapiridae
have never been adapted to a country of this kind.
The giraffes frequent the savanna and the thorn-
forested country (xerophilous woodland of Schimper).
9. Lower river valleys habitat. — Alluvial bottom
lands, delta and flood-plain deposits, swamps and
jungles, forested or grassy lowlands near rivers or sea
level, typically the home of browsers rather than
grazers, with feet and limbs adapted to soft soil,
j limbs both of mediportal and graviportal type, with
some cursorial types (such as situtungas) having
spreading feet. The Artiodactyla include many
bovines, some antelopes (such as situtungas), chevro-
tains, suillines, the Liberian hippopotamus {Choer-
opsis liberiensis) , and the primitive traguline (Dorca-
therium) of West Africa. Among the Asiatic rhino-
ceroses R. sondaicus, a browsing, brachyodont type,
132
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
is occasionally seen in these alluvial bottoms. Simi-
larly, the Sumatran rhinoceros, R. (DicerorJiinus)
sumatrensis, also brachyodont, occasionally frequents
such a region. Tapirus indicus inhabits this low
forest belt in India, and T. terrestris is the common
tapir of the forests and lowlands of Brazil and Para-
guay. Among extinct forms the chalicotheres
{Moropus, etc.) are found here.
10. Aquatic, river and lalce iorder habitat. — Rivers,
bayous, and lakes, frequented especially by aquatic
browsing types with limbs adapted to swimming or
to aquatic life and the teeth adapted to the softer
kinds of food. Here we find the ungulates represented
by their partly degenerate and specialized offshoots the
sirenians, the Artiodactyla by the common hippopot-
amus or the water buffalo of the Philippines. Either
the lower river valleys or the rivers themselves were
undoubtedly the habitat of the extinct rhinoceroses
known as amynodonts; also, possibly, of the Miocene
Teleoceras, a short-limbed river-frequenting animal.
Among the titanotheres there are evidences of aquatic
adaptation in species of the genus Mesatirhinus.
CONCLUSIONS AS TO HABITATS OF THE TITANOTHERES
We have no evidence that titanotheres formerly
inhabited alpine, high steppe, or low desert regions.
Neither the teeth nor the feet predispose us to specu-
late upon such a habitat, nor have we any geologic evi-
dence of it. There remain to be considered the "moun-
tain," the "forest," the "boreal," or north temperate,
the "plains," the "river valleys," the "rivers and la-
goons."
The earliest known types of titanotheres, which are
subcursorial in limb structure, developed in a partly
open and partly forested country, frequenting mead-
ows, lower river valleys, and plains that were flooded
during certain seasons of the year. There is reason
to believe that one of their upper Eocene radiations
(MetarJiinus) became amphibious or even aquatic.
Some authors (Riggs, 1912.1, p. 36) believe that
DolichorMnus , as well as the short-footed Palaeosyops,
was semiaquatic. The habits of these animals are
more 'fully considered in Chapter V.
In Oligocene time the titanotheres entered the
savanna-like Great Plains region of western North
America, which was in part open country, in part
country traversed by undulating rivers and by river
bottoms bordered with forests.
In dentition the titanotheres, both in Eocene and
Oligocene phyla, are chiefly a browsing family, though
they show incipient indications of adaptation to the
grazing habit.
SECTION 4. BIBLIOGRAPHY FOR CHAPTER II
Ball, John.
1887.1. Notes of a naturalist in South America, .xiii, 416
pp., 1 map, London, 1887.
Bauer, Clyde Max.
1916.1. Contributions to the geology and paleontology
of San Juan County, N. Mex. — 1, Stratig-
raphy of a part of the Chaco River valley:
U. S. Gaol. Survey Prof. Paper 98, pp. 271-
278, pis. 6^71, Nov. 24, 1916.
Maps western part of Puerco-Torrejon area and gives short
suraraary of the two formations.
Berby, Edward Wilber.
1914.1. The Upper Cretaceous and Eocene floras of South
Carolina and Georgia: U. S. Geol. Survey
Prof. Paper 84, 200 pp., 29 pis., 12 figs., 1914.
Blanford, W. T.
1888.1. The fauna of British India, including Ceylon and
Burma — Mammalia, 1888-1891.
BouTWELL, John Mason.
1907.1. Stratigraphy and structure of the Park City min-
ing district, Utah: Jour. Geology, vol. 15, pp.
434-458, 1907.
Brown, Barnum.
1914.1. Cretaceous-Eocene correlation in New Mexico,
Wyoming, Montana, Alberta: Geol. Soc.
America Bull., vol. 25, pp. 355-380, Sept. 15,
1914.
"Through this eastern exposure I have often found it im-
possible to establish any definite line of demarcation between
the two beds (Fox Hills and Lance)" (p. 3.5S).
"The insensible gradation from marine through brackish-
water into fresh-water sandstones is not confined to the
eastern exposures of the 'Lance' on Hell Creek. The same
transition is found on the border of the Lance formation on
Alkali Creek, Sevenmile Creek, and Robber's Koost, all
tributaries of the Cheyenne River in Weston County,
Wyo."
Calvert, William R.
1910.1. See Stone, Ralph Walter, 1910.1.
Clark, William Bullock.
1891.1. Correlation papers — Eocene: U. S. Geol. Survey
Bull. 83, 173 pp., 1891.
Resume of work of various writers. Deposits of Bridger
and Washakie Basins considered as one formation. Regards
the Puerco as probably of Eocene age (p. 138). Eocene of
the Atlantic coast. Gulf States, Pacific coast; historical sketch
of the Eocene of the interior. Table showing relative posi-
tion of interior Eocene deposits. Extensive bibliography.
1896.1. The Eocene deposits of the middle Atlantic slope
in Delaware, Maryland, and Virginia: U. S.
Geol. Survey Bull.' 141, 167, pp., 40 pis., 1896.
COCKERELL, THEODORE DrU AlISON.
1906.1. The fossil fauna and flora of the Florissant
(Colorado) shales: Colorado Univ. Studies,
vol. 3, pp. 157-176, 5 figs., June, 1906.
Birds, fishes, insects, mollusks, plants.
CoMSTOCK, Theodore Bryant.
1873.1. On the geology of western Wyoming: Am. Jour.
Sci., 3d ser., vol. 6, pp. 426-432, 1873.
Bridger classed as upper Miocene.
Cope, Edward Drinker.
1872.3. On Bathmodon, an extinct genus of ungulates:
Am. Philos. Soc. Proc, vol. 12, pp. 417-420,
1872.
Describes the first mammal from this horizon (.Bathmoion) .
1872.4. On a new genus of Pleurodira from the Eocene of
Wyoming: Am. Philos. Soc. Proc, vol. 12,
pp. 472-477, 1872.
Gives a detailed account of the fossil-bearing beds along
Bear River, near Evanston, Wyo.
ENVIRONMENT OF THE TITANOTHEEES
133
Cope, Edward Drinker — Continued.
1873.4. The monster of Mammoth Buttes: Penn Monthly,
vol. 4, pp. 521-534, 1 pi., August, 1873.
A popular account of the finding of tlie sltull of Eobasileus
cornutus.
1875.1. Report on the geology of that part of north-
western New Mexico examined during the
field season of 1874: U. S. Geog. Surveys
W. 100th Mer. Ann. Rept. for 1875, pp. 61-97,
pis. 2-6, 18 figs., 1875.
Tlie original description of " Puerco marls. " Type locality,
head of Puerco River. Gives section (p. 96) showing relation
of Puerco and other beds in that vicinity. No mammalian
fossils, but the marls are referred to the Eocene for stratigraphic
reasons.
1877.1. Report upon the extinct Vertebrata obtained in
New Mexico by parties of the expedition of
1874: U. S. Geog. Surveys W. 100th Mer.
Rept., vol. 4, pt. 2, 370 pp., pis. 22-73, 1877.
Extensive account of geology of the Wasatch beds and their
fauna. Quotes former article (pp. 17, 18), but says the beds
may represent Fort Union or the lignites of upper Missouri.
The thickness of the Puerco is given as 500 feet.
1879.1. The relations of the horizons of extinct Vertebrata
of Europe and North America: U. S. Geol.
and Geog. Survey Terr. Bull., vol. 5, pp. 33-54,
1879.
Correlation of Mesozoic and Cenozoic horizons of Europe
and North America.
1879.2. Second contribution to a knowledge of the Miocene
fauna of Oregon: Am. Philos. Soc. Proc, vol.
18, pp. 370-376, Dec. 30, 1879.
John Day formation, Oligoeene.
1880.1. The badlands of Wind River and their fauna:
Am. Naturahst, vol. 14, pp. 745-748, October,
1880.
Eocene.
1880.2. Observations on the faunae of the Miocene Ter-
tiaries of Oregon: U. S. Geol. and Geog.
Survey Terr. Bull., vol. 5, pp. 55-69, 1880.
See also Paleont. Bull. No. 30, Dec. 3, 1878,
and Am. Philos. Soc. Proc, vol. 18, pp. 63-78,
Dec. 30, 1878.
John Day formation, Oligoeene.
1881.1. Mammalia of the lower Eocene beds : Am. Natural-
ist, vol. 15, pp. 337-338, April, 1881.
The first mammals are described, but they were not l^nown
definitely at that time to be from the Puerco formation.
1885.1. The Vertebrata of the Tertiary formations of
the West: U. S. Geol. Survey Terr. Rept.,
vol. 3, XXXV, 1009 pp., 134 pis. (pis. l-75a),
38 figs., 1885.
Contains a general r63Um6 of the Wasatch. The deposits of
the Bridger and W'ashaliie Basins and small area on White
River in the Uinta Basin considered contemporary. Table
of formations in this worli places Puerco as "post-Cretaceous,"
but in the text the author places it definitely in the Eocene.
1885.2. The relations of the Puerco and Laramie deposits:
Am. Naturalist, vol. 19, pp. 985-986, October,
1885.
states that the thickness of the beds near the type locality is
850 feet. The author points out the distinctions from Laramie
but considers the possibility of "post-Cretaceous" age.
1885.3. The White River beds of Swift Current River,
Northwest Territory: Am. Naturalist, vol. 19,
p. 163, February, 1885.
Oligoeene, White River.
Cope, Edward Drinker — Continued.
1886.1. The Vertebrata of the Swift Current Creek region
of the Cypress Hills: Canada Geol. and Nat.
Hist. Survey Ann. Rept., new ser., vol. 1, for
1885, appendix I to article C, pp. 79-85, 1886.
Oligoeene.
Cdlbertson, Thaddeus a.
1851.1. Journal of an expedition to the Mauvaises Terres
and the upper Missouri in 1850: Smithsonian
Inst. Fifth Ann. Rept., appendix 4, pp. 84-145,
1851.
Bear River [=Bear Creek] (p. 9.3), a southern tributary of
the Cheyenne. First collection [in the Oreodon zone, Brule
clays] (p. 94), rhinoceros skull (,A. oiddenlah) , several good
heads, excellent teeth and jawbones, etc. Report to Baird
(p. 105).
Dall, William Healey.
1892.2 (and Harris, G. D.). Correlation papers — The
Neocene of North America: U. S. Geol. Survey
Bull. 84, 349 pp., 3 pis., 43 figs., 1892.
See especially chapter 6, on the supposed Neocene of the in-
terior region, considered by States (pp. 280-317); table showing
the vertical range of the Neocene of the interior (p. 279);
map (p. 178); list of names applied to the Cenozoic beds and
formations of the United States (p. 320).
1898.1. A table of the North American Tertiary horizons
correlated with one another and with those of
western Europe, with annotations: U. S.
Geol. Survey Eighteenth Ann. Rept., pt. 2,
- pp. 327-348, 1898.
Marine Tertiary horizons of the Atlantic coast'and the Gulf
States correlated with one another, with those of the western
United States, and with those of western Europe.
Darton, Nelson Hokatio.
1896.1. Catalogue and index of contributions to North
American geology, 1732-1891: U. S. Geol. Sur-
vey Bull. 127, 1045 pp., 1896.
1903.1. Preliminary report on the geology and water
resources of Nebraska west of the one hundred
and third meridian: U. S. Geol. Survey Prof.
Paper 17, 69 pp., 43 pis. (inch 9 maps), 23 figs.,
1903.
Titanotherium zone (Chadron formation) of western
Nebraska, along the North Platte, Scott Bluffs, Sioux County,
etc.
1905.1. Age of the Monument Creek formation: Am. Jour.
Sci., 4th ser., vol. 20, pp. 178-180, 1905.
Menodus {TiianoiheriuTTi) remains. Oligoeene.
1905.2. Preliminary report on the geology and underground
water resources of the central Great Plains:
U. S. Geol. Survey Prof. Paper 32, 433 pp., 72
pis., 18 figs., 1905.
1906.1. Geology and underground waters of the Arkansas
Valley in eastern Colorado: U. S. Geol. Survey
Prof. Paper 52, 90 pp., 27 pis., 2 figs., 1906.
" Monument Creek formation," containing Menodus (^Titano-
iherium) of White River age (p. 34) . Nussbaum formation,
of late Tertiary age (p. 34) .
1906.2. Geology of the Big Horn Mountains: U. S. Geol.
Survey Prof. Paper 51, 129 pp., 47 pis., 14
figs., 1906.
Brief reference to the Eocene rocks (p. C7). See especially
Bridger [Wind River] formation (p. 70).
Davis, William Morris.
1900.1. The fresh-water Tertiary formations of the Rocky
Mountain region: Am. Acad. Arts and Sci. Proc,
vol. 35, pp. 346-373, 1900.
History of opinion on mode of formation; evidence against
lake-bed hypothesis and in favor of fluviatile origin.
134
TITANOTHEBES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Dawkins, W. Boyd.
1880.1. The classification of the Tertiary period by means
of the Ivlammalia: Geol. Soc. London Quart.
Jour., 1880, pp. 379-405.
Tertiary and Quaternary horizons and faunas of Great
Britain, France, and Italy correlatRd.
Deperet, Charles.
1893.1. Note sur la succession stratigraphique des faunes
de mammifcres pliocenes d' Europe et du Pla-
teau central en particulier: Soc. g<Sol. France
Bull., 3d ser., vol. 21, pp. 524-540, 1893.
1906.1. L'evolution des mammiferes tertiaires, importance
des migrations, epoque miocene: Compt. Rend.,
vol. 143, No. 26, pp. 1120-1123, 1906. The
evolution of Tertiary mammals and the impor-
tance of their migrations (translation) : Am.
Naturalist, vol. 42, pp. 109-114, 166-170,
303-307.
DOLLO, Loui.?.
1909.1. The fossil vertebrates of Belgium [Correlation
Bull. No. 2] (translation by W. D. Matthew) :
New York Acad. Sci. Annals, vol. 19, No. 4,
pt. 1, pp. 99-119, pis. 4-10, July 31, 1909.
DouGL.ASs, Earl.
1S99.1. The Neocene lake beds of western Montana and
descriptions of some new vertebrates from the
Loup Fork: Montana LTniv. thesis, 27 pp.,
4 pis., June, 1899.
Geology, faunas, and correlation of White River, "Deep
River," and "Madison Valley." "Loup Fork" horizons in
Montana. Systematic descriptions of certain fossil camels, etc-
1902.1. Fossil Mammalia of the White River beds of
Montana: Am. Philos. Soc. Trans., new ser.,
vol. 20, pp. 227-278, pi. 9, 1902.
"Pipestone beds," "Toston beds," "Blaclitail Deer Creelc
beds." Geology and faunas; new genera and species of mam-
mals.
1902.2. A Cretaceous and lower Tertiary section in south-
central Montana: Am. Philos. Soc. Proc, vol.
41, No. 170, pp. 207-224, pi. 29, April, 1902.
Sketch of the Jurassic and Cretaceous deposits. Probable
relations of the "Laramie" and overlying beds. Fossil mam.
mals of the Fort Union beds. Describes the Fort Union beds
of Montana; considers them as of practically the same age as
the Torrejon Tertiary. Places Puerco as Upper Cretaceous-
1902.3. The discovery of Torrejon mammals in Montana:
Science, new ser., vol. 15, No. 372, pp. 272-273,
Feb. 14, 1902.
First record of mammals from Fort Union beds of Crazy
Mountains region.
1903.1. New vertebrates from the Montana Tertiary:
Carnegie Mus. Annals, vol. 2, No. 2, pp.
145-199, pi. 2, 37 figs., November, 1903.
"Sage Greek" (Eocene?), White River deposits, "Fort
'Logan beds" (upper Oligocene), "Deep" and "Flint Creek"
beds. New mammals described.
1909.1. Preliminary descriptions of some new titanotheres
from the Uinta deposits: Carnegie Mus. Annals,
vol. 6, No. 2, pp. 304-313, pis. 13-15, 8 figs.,
August, 1909.
Describes new faunas from liorizon B.
Dow, John M.
1867.1. Extracts from letters relating to Tapirus bairdi
(read by P. L. Sclater) : Zool. Soc. London Proc,
1867, p. 241, 1867.
Earle, Charles.
1895.1. See Osborn, Henry Fairfield, 1895.95.
Emmons, Samuel Franklin.
1907.1. Uinta Mountains: Geol. Soc. America Bull., vol.
17, pp. 287-302, pi. 24, 2 figs., July 13, 1907.
Endlich, Frederic Miller.
1877.1. Report on the San Juan region: U. S. Geol. and
Geog. Survey Terr. Ninth Ann. Rept., pp. 176-
191, 1877.
Tertiary (p. 189). Puerco beds of Animas Valley, southern
Colorado (1,000 to 1,200 feet), are considered the basal member
of the Wasatch.
1879.1. Report on the geology of the Sweetwater district:
U. S. Geol. and Geog. Survey Terr. Eleventh
Ann. Rept., pp. 5-158, 1879. '
Refers to the lower Bridger exposed in the northern part o!
the basin, Big Sandy^Creek, etc. (p. 132). Considers a portion
of the "Wasatch" of Beaver Creek, Wyo., as parallelwith the
Puerco marls.
Filhol, Henri.
1885.1. Observations sur le memoire de M. Cope intitul(5
"Relations des horizons * * * d'animaux
vertebres fossiles en Europe et en Amerique":
Annales sci. g6ol., vol. 17, art. 2, pp. 1-18, pi.
6, 1885.
FiNLAY, George Irving.
1916.1. U. S. Geol. Survey Geol. Atlas, Colorado Springs
folio (No. 203), 17 pp., 3 maps.
Laramie, Dawson, and Denver of Colorado; flora, fauna.
Fisher, Cassius Asa.
1906.1. Geology and water resources of the Big Horn
Basin, Wyo.: U. S. Geol. Survey Prof. Paper 53,
72 pp., 16 pis., 1906.
Discusses briefly the character, thickness, and distribution
of the Wasatch formation (p. 33) .
Fraas, Eberhard.
1901.1. On the aqueous vs. eolian deposition of the White
River Oligocene of South Dakota (translation
by H. F. Osborn): Science, new ser., vol. 14,
No. 345, pp. 210-212, Aug. 9, 1901.
" Titanotherlum beds" formed by river and flood-pli.in
deposits exposed during dry season. Middle " Oreodon beds "
deposited by a shallow lake with dissolved materials of varying
concentration (cf. banded layers). Upper "Oreodon beds"
formed by eolian loess.
Gardner, James Heney".
1910.1. The Puerco and Torrejon formations of the
Nacimiento group: Jour. Geology, vol. 18,
pp. 702-741, 1 pi., 9 figs., 1910.
Gives historical review. Topography, structure, and physi-
ographic record of the Puerco-Torrejon district described.
Considers that an unconformity exists between the two for-
mations, to which the group name Nacimiento is given.
Gidley, James Williams.
1904.1. See Matthew, William Diller, 1904.1.
1917.1. [The 1910 collection near the Davis ranch, Powder
River valley, Wyo.] In Wegemann, C. H.,
Wasatch fossils in so-called Fort Union beds of
the Powder River basin, Wyo., and their bear-
ing on the stratigraphy of the region: U. S.
Geol. Survey Prof. Paper 108, p. 59, 1917.
Gilbert, Grove Karl.
1898.1. The underground waters of the Arkansas Valley
in eastern Colorado: U. S. Geol. Survey
Seventeenth Ann. Rept., pt. 2, pp. 553-601,
pis. 56-68, figs. 45-49, 1896.
Rocky Mountain deposits may be of fluviatile and not of
lacustrine origin.
ENVIKONMENT OF THE TITANOTHERES
135
Granger, Walter.
1909.1. Faunal horizons of the Washakie formation of
southern Wyoming: Am. Mus. Nat. Hist.
Bull., vol. 26, pp. 13-23, pis. 2-6, 1 map, Jan.
19, 1909.
Divides "Washakie beds" into two horizons, characterized
by fauna and lithology. Lower horizon=upper Bridger;
upper horizon=lower and middle "Uinta" (Uinta A and B).
1910.1. Tertiary faunal horizons in the Wind River Basin,
Wyo., with descriptions of new Eocene mam-
mals: Am. Mus. Nat. Hist. Bull., vol. 28, pp.
235-251, pis. 20-23, 6 figs., July 16, 1910.
Determines two distinct faunal horizons in Wind Hiver
beds — the Lambdotherittm zone and an earlier horizon.
1911.1. See Sinclair, William John, 1911.1.
1912.1. See Sinclair, William John, 1912.1.
1914.1. On the names of lower Eocene faunal horizons of
Wyoming and New Mexico: Am. Mus. Nat.
Hist. Bull., vol. 33, pp. 201-207, Mar. 31, 1914.
Names "Clark Fork," "Sand Coulee," and "Gray Bull
beds" of Big Horn, Wyo., and "Almagre" and "Largo beds"
of New Mexico. Correlates the lower Eocene of New Mexico
with that of the various Wyoming basins.
1914.2. See Sinclair, William John, 1914.1.
1917.2. Notes on Paleocene and lower Eocene mammal
horizons of northern New Mexico and southern
Colorado: Am. Mus. Nat. Hist. Bull., vol. 37,
pp. 821-830, Dec. 5, 1917.
1918.1 (and Matthew, W. D.). A revision of the lower
Eocene Wasatch and Wind River faunas:
Am. Mus. Nat. Hist. Bull., vol. 38, pp. 565-
657, 1918.
Gray, Dr. J. E.
1872.1. Notes on a new species of tapir {Tapirus leucogenys)
from the snowy regions of the Cordilleras of
Ecuador and on the young spotted tapirs of
tropical America: Zool. Soc. London Proc,
1872, pp. 483-492, pis. 21-22.
Gregory, John Walter.
1896.1. The Great Rift Valley, 422 pp., London, John
Murray, 1896.
Harris, Gilbert Dennison.
1892.2. See Dall, William Healey, 1892.2.
Hatcher, John Bell.
1893.1. The Titanotherium beds: Am. Naturahst, vol. 27,
pp. 204-221, 3 figs.. Mar., 1893.
General description. Accepts lacustrine theory of deposi-
tion.
1894.1. Discovery of Diceratherium, the two-horned
rhinoceros, in the White River beds of South
Dakota: Am. Geologist, vol. 13, pp. 360-361,
May, 1894.
Top of White River correlated with John Day formation.
1895.1. On a new species of Diplacodon, with a discussion
of the relations of that genus to Telmatherium:
Am. Naturahst, vol. 29, pp. 1084-1093, pis.
38-40, fig. 1, Dec, 1895.
1902.3. Origin of the Oligocene and Miocene deposits of
the Great Plains: Am. Philos. Soc. Proc, vol.
41, pp. 113-131, 1902.
Summarizes facts and accepts theory of small lakes, flood
plains, river channels, and pampas as prevailing conditions
during Oligocene and Miocene time. Gering, Arikaree,
Ogalalla, Monroe Creek, Harrison, and "Nebraska" of Scott.
Classification of the Oligocene and Miocene. "Lake-bed"
hypothesis of origin disproved in favor of fluviatile, flood-plain,
and eolian hypothesis .
Haworth, Erasmus.
1897.1. Physical properties of the Tertiary [of Kansas]:
Kansas Univ. Geol. Survey, vol. 2, pp. 247-284,
pis. 36-44, 1897.
Rejects "lake-basin" hypothesis in favor of hypothesis of
fluviatile origin of Tertiary of Kansas.
Hay, (Jliver Perry.
1905.1 The fossil turtles of the Bridger Basin: Am.
Geologist, vol. 35, pp. 327-342, June, 1905.
Evidence for flood-plain rather than lacustrine origin of the
Bridger. Discussion of life and climatic conditions.
1908.1 The fossil turtles of North America: Carnegie Inst.
Wash. Pub. 75, 568 pp., 113 pis., 704 figs., 1908.
H.\y, Robert.
1889.1. Northwest Kansas, its topography, geology, cli-
mate, and resources: Kansas State Board Agr.
Sixth Bienn. Rept., pp. 92-116, 2 maps, 4 figs.,
1889.
See especially discussion of the Tertiary geology of Kansas.
Hayden, Ferdinand Vandiveer.
1858.1. Notes on the geology of the Mauvaises Terres of
White River, Nebr. : Acad. Nat. Soi. Phila-
delphia Proc, vol. 9, pp. 151-165, 1858.
Refers to Bear Creek, Pennington County, S. Dak. Type
locality of Mesohippus bairdii.
1862.1. See Meek, Fielding Bradford, 1862.1.
1869.1. Geological report of the exploration of the Yellow-
stone and Missouri Rivers, by F. V. Hayden,
under the direction of William F. Raynolds,
174 pp., 1 map, Washington, 1869.
1871.2. Report of F. V. Hayden. In [Fourth Annual]
Preliminary report of the United States geolog-
ical survey of Wyoming and portions of con-
tiguous territories, pp. 9-81, 1871.
A general account of the topography and geology (type
description) of the Bridger Basin (pp. 54-58) . Considers upper
portion of Washakie Basin sediments, as either an extension
eastward of the Bridger beds or as a separate deposit of the
same age. Notes occurrence of vertebrate fossils.
1873.1. PreUminary field report of the United States
geological survey of Colorado and New Mexico:
U. S. Geol. Survey Terr. Third Ann. Rept.,
pp. 105-251, 1869, reprinted 1873. [Reprinted
in 1873 in First, Second, and Third Annual
Reports of the Geological Survey of the Terri-
tories. In the text of this monograph refer-
ence is made to the reprinted edition.]
Names and briefly describes "Green River shales," Bridger
"group," Wasatch "group," and Bear River "group." Desig-
nates Tertiary deposits between Creston and Bitter Creek
along Union Pacific RaUroad as "Washakie group" (p. 190).
1881.1. Geological and geographical atlas of Colorado and
portions of adjacent territory, U. S. Geol. and
Geog. Survey Terr., 1877, corrected to 1881.
Heller, Edmund.
1914.1. See Roosevelt, Theodore, 1914.1.
Hills, Richard Charles.
1888.1. The recently discovered Tertiary beds of the
Huerfano River basin, Colo. : Colorado Sei.
Soc Proc, vol. 3, pp. 148-164, 1 map, 1888.
Beds first described. Upper half suspected to be of Wasatch
age.
1889.2. Additional notes on the Huerfano .beds: Colorado
Sci. Soc. Proc, vol. 3, pp. 217-223, 1889.
Mammals reported from upper division. Bridger age
indicated.
136
TITANOTHEBES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Hills, Richard Charles — Continued.
1891.1. Remarks on the classification of the Huerfano
Eocene: Colorado Sci. Soc. Proc, vol. 4, pp.
7-9, 1891.
Series divided into Huerfano, Cuchara, and Poison Canyon
beds. Huerfano=Bridger; other two=lower Eocene.
HovEY, Edmund Otis.
1908.1. See Willis, Bailey, 1908.1.
Irving, John Duer.
1896.1 The stratigraphical relations of the Browns
Park beds of Utah: New York Acad. Sci.
Trans., vol. 15, p. 252, pi. 18, Sept., 1896.
The beds in Browns Park valley assigned to the Pliocene.
JOHANNSEN, ALBERT.
1914.1. Petrographio analysis of the Bridger, Washakie,
and other Eocene formations of the Rocky
Mountains: Am. Mus. Nat. Hist. Bull., vol.
33, pp. 209-222, 2 figs.. Mar. 31, 1914.
Considers Bridger and "Washakie" rocks largely tufls
modified by slight transportation. The older Eocene rocks
are considered more strictly sedimentary.
Johnson, Willard Drake.
1901.1. The High Plains and their utilization: U. S.
Geol. Survey Twenty-first Ann. Rept., pt. 4,
pp. 601-741, pis. 113-116, figs. 300-329, 1901;
Twenty-second Ann. Rept., pt. 4, pp. 631-669,
pis. 55-65, figs. 236-244, 1902.
Tertiary deposits of the Plains, of fluviatile and flood-plain
origin.
King, Cl.arence.
1876.1. Geological and topographical atlas accompanying
the report of the Geological Exploration of the
40th Parallel, 1876.
1878.1. Systematic geology: U. S. Geol. Expl. 40th Par.
Rept., vol. 1, 803 pp., 21 pis., 12 maps, 1878.
Gives the name "Vermilion Creek" to the Wasatch beds of
southern Wyoming; considers them as lowest Eocene and
unconformable with the overlying Green Eiver beds. The
name "Uinta group" is given to the uppermost 400 feet of the
sediments in the vallej' of White River; considered to lie un-
cnnformably on lower beds and to represent uppermost Eocene
Mammals collected by Marsh are listed. Area is mapped,
and relationships of Bridger with other Eocene deposits of the
basin are set forth.
Knowlton, Frank Hall.
1902.1. Fossil flora of the John Day Basin, Oreg.: U. S.
Geol. Survey Bull. 204, i53 pp., 17 pis., 1902.
Geology (pp. U-20, 102-108). Mascall formation referred
to upper Miocene.
1909.1. The stratigraphic relations and paleontology of
the "Hell Creek beds," " Ceratops beds," and
equivalents, and their reference to the Fort
Union formation: Washington Acad. Sci. Proc,
vol. 11, No. 3, pp. 179-238, Aug. 14, 1909.
Kobelt, W.
1902.1. Die Verbreitung der Tierwelt, 576 pp., Leipzig,
1902.
Lambb, Lawrence Morris.
1908.1. The Vertebrata of the Oligocene of the Cypress
Hills, Saskatchewan: Canada Geol. Survey
Contr. Canadian Paleontology, vol. 3, pt. 4,
65 pp., 8 pis., 1908.
Leidy, Joseph.
1869.1. The extinct mammalian fauna of Dakota and
Nebraska, including an account of some allied
forms from other localities, together with a
synopsis of the mammalian remains of North
America: Acad. Nat. Sci. Philadelphia Jour.
2d ser., vol. 7, 472 pp., 30 pis., 1869.
Lindgren, Waldemar.
1915.1. The igneous geology of the Cordilleras and its
problems. In Problems of American geology
(Silliman Memorial Lectures, 1913), pp. 234-
286, 1 map, Yale Univ. Press, 1915.
Lonnbekg, Einar.
1912.1. Mammals collected by the Swedish zoological
expedition to British East Africa, 1911: K.
svenska Vet.-Akad. Handlingar, Bd. 48, No. 5,
1912.
LooMis, Frederic Brewbteh.
1904.1. Two new river reptiles from the titanothere beds:
Am. Jour. Sci., 4th ser., vol. 18, pp. 427-432,
4 figs., Dec, 1904.
Flood-plain origin of the " Titanotherium beds."
1906.1. The Tertiary of Montana: Carnegie Mus. Mem.,
vol. 2, pp. 203-224, pi. 22, 1905.
Chiefly a description of Iciops, Xenotherium, and other lower
White River mammals.
1907.1. Origin of the Wasatch deposits: Am. Jour. Sci.,
4th ser., vol. 23, pp. 356-364, 3 figs.. May, 1907.
Treats of the Big Uorn Basin Wasatch; divides the beds into
three faunal levels, lists fossils from each level, and gives sec-
tions. The Wasatch is considered a flood-plain deposit, the
upper 1,000 feet of which appear to overlap in time the base
of the Wind River.
Lx'LL, Richard Swann.
1905.1 Megacerops tyleri, a new species of titanothere
from the Bad Lands of South Dakota: Jour.
Geology, vol. 13, No. 5, pp. 443-456, pis. 3-4,
1905.
Lydekkeb, Richard (editor).
?1893.1. The new natural history, vols. 1-4 (American
reprint of "The Royal natural histor}'," pub-
lished 1893-1896).
Lyons, H. G.
1906.1 The physiography of the River Nile and its basin,
441 pp., 48 pis., 1 map, Egypt Survey Dept.,
1906.
Rate of deposition (p. 334).
McMaster, John Bach.
1881.1. See Osborn, Henry Fairfield, 1881.8.
Marsh, Othniel Charles.
1871. 3. On the geology of the eastern Uintah Mountains:
Am. Jour. Sci., 3d ser., vol. 1, pp. 191-198, 1871.
Short account of the expedition to Uinta Basin in 1870. Con-
siders Uinta Basin deposits synchronous with those of Bridger
Basin on paleontologic evidence. Considers the fossils as
indicating much greater age than Miocene of eastern Rocky
Mountain basins.
1875. 2. Ancient lake basins of the Rocky Mountain region:
Am. Jour. Sci., 3d ser., vol. 9, pp. 49-52, Janu-
ary, 1875.
1877. 1. Introduction and succession of vertebrate life in
America: Am. Jour. Sci., 3d ser., vol. 14, pp.
337-378, 1877.
Plate showing successive horizons named from characteris-
tic genera. Names Diplacodon zone (p. 354).
1891. 2. Geologic horizons as determined by vertebrate
fossils: Am. Jour. Sci., 3d ser., vol. 42, pp. 336-
338, October, 1891.
1898. 1. The comparative value of different kinds of fossils
in determining geological age: Am. Jour. Sci.,
4th ser., vol. 6, pp. 483-486, December, 1898.
Value of a form depends upon its modiflability in accordance
with changing conditions.
ENVIBONMENT OF THE TITANOTHEKES
137
Matthew,
1897. 2
1899. 2.
1901. 1
1903. 1
1906. 1
1908. 1
1909. 1.
1909. 2.
1914. 1
1918. 1
William Dillbe.
. A revision of the Puerco fauna: Am. Mus. Nat.
Hist. Bull., vol. 9, pp. 259-323, Nov. 16, 1897.
Points out the distinct separation of species of upper and
lower beds and adopts Wortman's proposed name, Torrejon
for the upper beds.
, A provisional classification of tlie fresh-water
Tertiary of the West: Am. Mus. Nat. Hist.
Bull., vol. 12, pp. 19-75, Mar. 31, 1899.
Is the White River Tertiary an eolian formation?
Am. Naturalist, vol. 33, pp. 403-408, May, 1899.
Summary of the paleontologic evidence against the "lake-
basin" hypothesis.
Fossil mammals of the Tertiary of northeastern
Colorado: Am. Mus. Nat. Hist. Mem., vol. 1,
pt. 7, pp. 353-447, 1901.
Stratigraphy of White River deposits ("Horsetail Creek,"
"Cedar Creek," and "Martin Canyon beds") and of "Loup
Fork" formation ("Pawnee Creek beds"). Evidence as to
mode of deposition (chiefly eolian); analysis of faunas: correla-
tion of horizons: systematic descriptions.
List of the Pleistocene fauna from Hay Springs,
Nebr.: Am. Mus. Nat. Hist. Bull., vol. 16, pp.
317-322, Sept. 25, 1902.
Lists for comparison the faunas of Hay Springs (Nebr.),
Silver Lake (Oreg.), and Washtucna Lake (Wash,).
The fauna of the Titanotherium beds at Pipestone
Springs, Mont.: Am. Mus. Nat. Hist. Bull.,
vol. 19, pp. 197-226, 19 figs.. May 9, 1903.
(and Gidley, J. W.). New or little-known mam-
mals from the Miocene of South Dakota: Am.
Mus. Nat. Hist. Bull., vol. 20, pp. 241-268, 15
figs., July 20, 1904.
Upper Miocene "Loup Fork beds," geology and faunal list.
Lower Miocene "Rosebud beds" (new name). New Carni-
vora and Rodentia.
Hypothetical outlines of the continents in Tertiary
times: Am. Mus. Nat. Hist. Bull., vol. 22, pp.
353-384, 7 figs., Oct. 25, 1906.
A lower Miocene fauna from South Dakota: Am.
Mus. Nat. Hist. Bull., vol. 23, pp. 169-219,
26 figs., 1907.
"Lower Rosebud" and "Upper Rosebud" deposits and
faunas: comparison with American Oligocene and Miocene
faunas. New Carnivora, Rodentia, Artiodactyla.
Mammalian migrations between Europe and
North America: Am. Jour. Sci., 4th ser., vol.
25, pp. 68-70, January, 1908.
The Carnivora and Insectivora of the Bridger
Basin, middle Eocene: Am. Mus. Nat. Hist.
Mem., vol. 9, pt. 6, pp. 289-559, pis. 44-52,
118 figs., 1909.
History of exploration. Stratigraphy and faunal divisions.
Condition of deposition.
See Osborn, Henry Fairfield, 1909. 321.
Evidence of the Paleocene vertebrate fauna on the
Cretaceous - Tertiary problem: Geol. Soc.
America Bull., vol. 25, pp. 381-402, Sept. 15,
1914.
See Granger, Walter, 1918. 1.
Meek, Fielding Bradford.
1862. 1 (and Hayden, F. V.). Descriptions of new Lower
Silurian (Primordial), Jurassic, Cretaceous, and
Tertiary fossils, collected in Nebraska by the
exploring expedition under command of Wm. F.
Raynolds, with some remarks on the rocks from
which they were obtained: Acad. Nat. .Sci.
Philadelphia Proc, vol. 13, pp. 415-447, 1862.
Wind River deposits considered intermediate in age between
Fort Union and White River.
101959— 2S— VOL 1 11
Mercer, Henry Chapman.
1899. 1. The bone cave at Port Kennedy, Pa., and its
partial excavation in 1894, 1895, 1896: Acad.
Nat. Sci. Philadelphia Jour., 2d ser., vol. 11,
pt. 2, pp. 269-286, Feb. 4, 1899.
Referred to the Pleistocene, but without comparison with
other cave formations and faunas.
Merriam, John Campbell.
1901.1. A contribution to the geology of the John Day
Basin [Oreg.]: California Univ. Dept. Geology
Bull., vol. 2, pp. 269-314, pis. 6-8, fig. 1, 1901.
Geology, faunas, and floras of the Cretaceous (Chieo and
Kno.\ville), Eocene (Clarno), Oligocene (John Day), Columbia
River lava, Miocene (Mascall), Pliocene (Rattlesnake),
Quaternary (p. 2C9),
NicKLBs, John M.
1924.1. Geologic literature on North America, 1785-1918:
U. S. Geol. Survey Bull. 746 (Bibliography),
1167 pp.; Bull. 747 (Index), 658 pp., 1924.
Osborn, Henry Fairfield.
1878.3 (and Scott, W. B.). Palaeontologieal report of
the Princeton Scientific Expedition of 1877:
Princeton Coll. E. M. Mus. Geol. Archaeol.
Contr., No. 1, 107 pp., Sept. 1, 1878.
A general account of the Bridger badlands, with notes on
analysis of the rocks.
1881.8 (and McMaster, J. B.). A memoir upon Loxolo-
phodon and Uintaiherium, two genera of the
I suborder Dinocerata, accompanied by a strati-
graphical report of the Bridger beds in the
Washakie Basin by J. B. McMaster: Princeton
Coll. E. M. Mus. Geol. Archaeol. Contr., vol.
1, No. 1, pp. 5-54, 1881.
Topography and geology described. Section given and im-
portant fossil localities indicated. Osborn notes for first time
difference in fauna between beds of the two basins and con-
siders "Washakie" as somewhat later than Bridger. First
stratigraphic section with geologic location of species. Error
in stratigraphy.
1887.30. See Scott, William Berryman, 1887.1.
1887.37 (and Scott, W. B.). Preliminary report on the
vertebrate fossils of the Uinta formation col-
lected by the Princeton expedition of 1886:
Am. Philos. Soc. Proc, vol. 24, No. 126, pp.
255-264, 1887.
1890.51. See Scott, William Berryman, 1890.1.
1892.67 (and Wortman, J. L.). Fossil mammals of the
Wasatch and Wind River beds, collections of
1891: Am. Mus. Nat. Hist. Bull., vol. 4, pp.
81-147, Oct. 20, 1892.
Geology of the Big Horn Basin (Wortman), p. 135, Analysis
and description of the fauna (Osborn) . Considers Wind River
beds distinct from and successive to the Wasatch of Big Horn
Basin.
1893-82. Rise of the Mammalia [vice-presidential address
before American Association for the Advance-
ment of Science]: Am. Jour. Sci., 3d ser., vol.
46, pp. 379-392, 448-466, November, Decem-
ber, 1893; Am. Assoc. Adv. Sci. Proc, vol.
42, pp 189-227, 1894.
1894.89. A division of the eutherian mammals into the
Mesoplacentaha and Cenoplacentalia [terms
subsequently altered to Meseutheria and Ce-
neutheria]: New York Acad. Sci. Trans., vol.
13, pp. 234-237, June 4, 1894.
138
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
OsBORN, Henrt Fairfield — Continued.
1894.90 (and Wortman, J. L.). Fossil mammals of the
lower Miocene White River beds, collection
of 1892: Am. Mus. Nat. Hist. Bull., vol. 6,
pp. 199-228, pis. 2-3, July 28, 1894.
Succession of species in the White River "Miocene" [Oligo-
cene],
1895.95 (and Earle, Charles). Fossil mammals of the
Puerco beds, collection of 1892: Am. Mus.
Nat. Hist. BuU., vol. 7, pp. 1-70, Mar. 8, 1895.
Wortman (field notes, pp. 1, 2) divides the Puerco into upper
and lower beds, with two thin yet distinct fossfl-bearing
strata in the lower bed and one (?) thicker stratum in the
upper. Gives the localities of both levels and estimates the
thickness of the combined beds (upper and lower) at 800 to
1,000 feet.
1895.98 (and Peterson, O. A.). Fossil mammals of the
Uinta Basin, expedition of 1894 (geologic
levels by O. A. Peterson) : Am. Mus. Nat.
Hist. Bull., vol. 7, pp. 71-105, 17 figs.. May 18,
1895.
Divides Uinta Basin deposits into three horizons; the three
faunal levels (horizons A, B, C), with faunal lists. New
genera and species, especially of Mesonyx, TelmatheTium,
SpheTWCoelits, and Elotherium.
1895.105 (and Wortman, J. L.). Perissodactyls of the
lower Miocene White River beds: Am. Mus.
Nat. Hist. Bull., vol. 7, pp. 343-375, pis. 8-11,
Dec. 23, 1895.
Oligocene.
1897.126. The Huerfano lake basin, southern Colorado,
and its Wind River and Bridger fauna: Am.
Mus. Nat. Hist. BuU., vol. 9, pp. 247-258,
Oct. 20, 1897.
1900.182. The geological and faunal relations of Europe
and America during the Tertiary period and
the theory of the successive invasions of an
African fauna: Science, new ser., vol. 11, No.
276, pp. 561-574, Apr. 13, 1900.
1900.187. Correlation between Tertiary mammal horizons
of Europe and America, an introduction to the
more exact investigation of Tertiary zoogeog-
raphy, prehminary study with third trial sheet:
New York Acad. Sci. Annals, vol. 13, No. 1, pp.
1-72, July 21, 1900.
1901.200. Correlation des horizons mammiferes tertiaires
en Europe et en Am^rique: Cong. g^ol. internat.,
8« sess., Compt. rend., pp. 357-363, 1901.
1901.201. See Fraas, Eberhard, 1901.1.
1902.214. The law of adaptive radiation: Am. Naturalist,
vol. 36, pp. 353-363, May, 1902.
1905.267. Ten years progress in the mammalian palaeon-
tology of North America: Cong, internat.
zool., 6« sess. (Bern, 1904), Compt. rend.,
pp. 86-113m, pis. 1-15; Am. Geologist, vol.
36, pp. 199-229, October, 1905.
1907.294. Tertiary mammal horizons of North America:
Am. Mus. Nat. Hist. Bull., vol. 23, pp. 237-
253, 3 figs.. Mar. 30, 1907.
1909.321 (and Matthew, W. D.). Cenozoic mammal
horizons of western North America, with ap-
pendix, Faunal lists of the Tertiary Mammalia
of the West, by William DiUer Matthew: U. S.
Geol. Survey BuU. 361, 138 pp., 2 pis., 14 figs.,
January, 1909.
1910.341. The paleontologic correlation through the Bache
fund: Science, new ser., vol. 31, No. 794, pp.
407-408, Mar. 18, 1910.
OsBORN, Hbnet Fairfield — Continued.
1910.342. Correlation of the Cenozoic through its mam-
malian fife: Jour. Geology, vol. 18, No. 3, pp.,
201-215, April-May, 1910; OutUnes of geologic
history, pp. 251-264, Chicago Univ. Press
July, 1910.
1910.345. Paleontologic evidences of adaptive radiation:
Pop. Sci. Monthly, vol. 77, pp. 77-81, July,
1910.
1910.346. The age of mammals in Europe, Asia, and North
America, 635 pp., 220 figs.. New York, Mac-
millan Co., 1910.
1912.376. Correlation and paleogeography : Geol. Soc.
America BuU., vol. 23, pp. 232-256, 1912.
1919.494. New titanotheres of the Huerfano: Am. Mus.
Nat. Hist. BuU., vol. 41, pp. 557-569, 7 figs.,
1919.
Peale, Albert Charle.s.
1876.1. Report on stratigraphy — Cenozoic formations:
U. S. Geol. and Geog. Survey Terr. Eighth
Ann. Rept., pp. 75-180, 13 pis., 5 maps, 1876.
"Washakie" treated as distinct group. Contains table of
localities, authorities, references, etc. (p. 140).
1879.1. Report on the geology of the Green River district:
U. S. Geol. and Geog. Survey Terr. Eleventh
Ann. Rept., pp. 511-542, pis. 47-54, 1879.
Peterson, Olof August.
1895.1. See Osborn, Henry Fairfield, 1895.98.
1914.1. A new titanothere from the Uinta Eocene: Car-
negie Mus. Annals, vol. 9, pp. 29-52, pis. 6-10,
figs. 1-14, 1914.
Diploceras osborni, from horizon B.
1914.2. A smaU titanothere from the lower Uinta beds:
Carnegie Mus. Annals, vol. 9, pp. 53-57, pi. 11,
figs. 1-2, 1914.
Beterotitanops parvus, from horizon A.
1914.3. Some undescribed remains of the Uinta titanothere
Dolichorhinus: Carnegie Mus. Annals, vol. 9,
pp. 129-138, figs. 1-7, 1914.
From Uinta A.
1914.4. A correction of generic name: Carnegie Mus.
Annals, vol. 9, p. 220, 1914.
PoHLiG, Hans.
1891.1. Dentition und Kraniologie des Elephas aniiquus
Falc. * * * Nachtrage: K. Leop. -Carol,
deutsche Akad. Naturforscher Nova Acta,
Band 57, pp. 285-459, 1891.
Powell, John Wesley.
1876.1. (and White, C. A.). Invertebrate paleontology of
the Plateau province. In Powell, J. W., Report
on the geology of the eastern portion of the
Uinta Mountains and a region of country adja-
cent thereto, pp. 74-135, U. S. Geol. and
Geog. Survey Terr., 1876.
Ransome, Frederick Leslie.
1915.1. The Tertiary orogeny of the North American Cor-
diUera and its problems. In Problems of Amer-
ican geology (SilUman Memorial Lectures,
1913), pp. 287-376, 1 map, Yale Univ. Press,
1915.
Richardson, George Burr.
1912.1. The Monument Creek group: Geol. Soc. America
BuU., vol. 23, pp. 267-276, 1 fig., 1912.
Describes the Dawson arkose and Castle Rock conglomerate
forming the "Monument Creek group" of Colorado and dis-
cusses their relation to the Denver and Arapahoe formations
ENVIRONMENT OF THE TITANOTHEEES
139
RiGGS, Elmer Samuel.
1912.1. New or little known titanotheres from the lower
Uintah formations: Field Mus. Nat. Hist. Pub.
159 (Geol. ser., vol. 4, No. 2), pp. 17-41, pis.
4-12, figs. 1-2, June, 1912.
Discusses stratigraphy of lower part of lower horizon of Uinta
Basin and gives general and detailed sections, with exact strati-
graphic position of various species of titanotheres.
Roosevelt, Theodore.
1914.1. (and HeUer, Edmund). Life histories of African
game animals, vols. 1-2, New York, Charles
Scribner's Sons, 1914.
Sampson, J. A.
1905.1. A deer's bill of fare: Sierra Club Bull., vol. 5,
pp. 194-210, 1905.
SCHIMPER, A. F. W.
1903.1. Plant geography upon a physiological basis (trans-
lation by W. R. Fisher, revised and edited by
Percy Groom and I. B. Balfour), 839 pp.,
Oxford, 1903.
ScLATER, Philip Lutley.
1894.1 (and Thomas, Oldfield). The book of antelopes,
vols. 1-4, London, 1894. [Issued in parts dated
consecutively 1894-1900.)
Scott, William Bbrrtman.
1878.1. See Osborn, Henry Fairfield, 1878.3.
1887.1 (and Osborn, H. F.). Preliminary account of the
fossil mammals from the White River forma-
tion contained in the Museum of Comparative
Zoology: Harvard Coll. Mus. Comp. Zoology
Bull., vol. 13, pp. 151-171, pis. 1-2, September,
1887.
1887.2. See Osborn, Henry Fairfield, 1887.37.
1888.1. The upper Eocene lacustrine formations of the
United States (abstract) : Am Assoc. Adv.
Sci. Proc, 1887, p. 217, March, 1888.
1890.1 (and Osborn, H. F.). The Mammalia of the
Uinta formation: Part I, The geological and
faunal relations of the Uinta formation, by
W. B. Scott; Part II, The Creodonta, Rodentia,
and Artiodactyla, by W. B. Scott; Part III,
The Perissodactyla, by H. F. Osborn; Part IV,
The evolution of the ungulate foot, by H. F.
Osborn: Am. Philos. Soc. Trans., new ser.,
vol. 16, pt. 3, pp. 461-572, pis. 7-9, 1890.
Considers the " Washaliie" a later substage of the Bridger
formation, and notes that several forms of animals found in the
beds are more similar to the Uinta Basin stages than to the
Bridger stages. Uinta considered top of Eocene, but strong
affinities with the White River Oligooene shown in the fauna.
1893.1. The mammals of the Deep River beds: Am.
Naturalist, vol. 27, pp. 659-662, July, 1893.
Preliminary description.
1894.1. The later lacustrine formations of the West: Geol.
Soc. America BuU., vol. 5, pp. 594, 595, 1894.
"Nebraska formation," " Corsoryx beds." Type reference.
1895.1. The Mammalia of the Deep River beds: Am.
Philos. Soc. Trans., new ser., vol. 8, pp. 55-185,
6 pis., 1895.
Geology (pp. 55-63) . European homotaxis with Sanson and
Simorre (middle Miocene).
1895.2. The Tertiary lacustrine formations of America:
Science, new ser., vol. 2, No. 42, p. 499, Oct.
18, 1895.
Tabular correlation of Tertiary horizons of Europe and
America.
Scott, William Berryman — Continued.
1899.1. The selenodont artiodaotyls of the Uinta Eocene:
Wagner Free Inst. Sci. Trans., vol. 6, pp. i-xiii,
15-122, pis. 1-4, May, 1899.
Angular unconformity between horizons B and C. White
River beds homotaxial with Ronzon of France. Uinta
compared with Paris gypsum (Lutftien).
Scudder, Samuel Hubbard.
1890.1. The Tertiary insects of North America: U. S.
Geol. Survey Terr. Rept., vol. 13, 734 pp., 28
pis., 1 map, 1890.
Map of the Tertiary lake basin at Florissant, Colo. Geology
of the deposits yielding Tertiary insects in America. Regards
Florissant ( Amyzon) beds as Oligocene? Volcanic origin of the
deposits. Now regarded as Miocene.
1894.1. The effect of glaciation and of the glacial period
on the present fauna of North America:
Am. Jour. Sci., 3d ser., vol. 48, pp. 179-187,
September, 1894.
Sinclair, William John.
1906.1. Volcanic ash in the Bridger beds of Wyoming:
Am. Mus. Nat. Hist. Bull., vol. 22, pp. 273-
280, pis. 35-38, July 31, 1906.
General features of the geology. Lithologic and stratigraphic
classification of the Bridger "group." The entire series of
Bridger rocks is determined as of volcanic origin.
1909.1. The Washakie, a volcanic ash formation: Am.
Mus. Nat. Hist. Bull., vol. 26, pp. 25-27,
Jan. 19, 1909.
Determines the nature of the volcanic material of the " Wash-
akie" to be different from that of the Bridger, which argues
against contemporaneity of deposition in the two basins.
1911.1 (and Granger, Walter). Eocene and Oliogocene of
the Wind River and Big Horn Basins: Am.
Mus. Nat. Hist. Bull., vol. 30, pp. 83-117,
July 11, 1911.
General account of Wasatch and later beds of the Big Horn
Basin. Discusses origin and mode of deposition of the sedi-
ments.
1912.1 (and Granger, Walter). Notes on the Tertiary
deposits of the Big Horn Basin: Am. Mus.
Nat. Hist. Bull., vol. 31, pp. 57-67, Mar. 30,
1912.
Additional observations as to deposition, extent, and chron-
ological subdivision of Big Horn sediments. Describes "Ly-
site" and "Lost Cabin formations" in the Big Horn Basin.
1912.2. Contributions to geologic theory and method by
American workers in vertebrate paleontology:
Geol. Soc. America BuU., vol. 23, pp. 262-266,
June, 1912.
1914.1 (and Granger, Walter). Paleocene deposits of the
San Juan Basin, N. Mex.: Am. Mus. Nat.
Hist. Bull., vol. 33, pp. 297-316, pis. 20-27, 2
figs., June 3, 1914.
Gives descriptions, measurements, and sections of Puerco
and Torrejon formations at various points. Determines that
mammal remains come from two layers in the Puerco and two
in the Torrejon. Considers both formations of fluviatile origin .
Lists important fossil localities.
Smith, James Henry.
1900.1. The Eocene of North America west of the 100th
meridian (Greenwich) : Jour. Geology, vol.
8, pp. 444^471, 1 map, 1900.
R§sum6 of literature on these deposits (pp. 452-454).
Stanton, Timothy William.
1909.1. The age and stratigraphic relations of the "Cera-
tops beds" of Wyoming and Montana: Wash-
ington Acad. Sci. Proc, vol. 9, No. 3, pp. 239-
293, 1909.
140
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Stanton, Timothy William — Continued.
1914.1. Boundary between Cretaceous and Tertiary in
North America as indicated by stratigraphy
and invertebrate faunas: Geol. Soc. America
Bull., vol. 25, pp. 349-351, Sept. 15, 1914.
Records Lance resting conformably on Fox Hills along
Missouri River in tlie Dakotas.
1916.1. Contributions to the geology and paleontology of
San Juan County, N. Mex. — 3, Nonmarine
Cretaceous invertebrates of the San Juan Basin;
U. S. Geol. Survey Prof. Paper 98, pp. 309-326,
pis. 79-83, 1916.
Stevenson-Hamilton, Major J.
1912.1. Animal hfe in Africa, 539 pp., iUus., London,
Wm. Heinemann, 1912.
Stone, Ralph Walter.
1910.1 (and Calvert, W. R.). Stratigraphic relations of the
Livingston formation of Montana: Econ. Geol-
ogy, vol. 5, pp. 551-557, 1 pi.; pp. 652-669;
pp. 741-764, 1 fig., 1910.
Thomas, Oldpield.
1894.1. See Sclater, Philip Lutley, 1894.1.
Veatch, Arthur Clifford.
1907.1. Geography and geology of a portion of south-
western Wyoming, with special reference to
coal and oil: U. S. Geol. Survey Prof. Paper 56,
178 pp., 26 pis., 9 figs., 1907.
Deposits of the period between the known Cretaceous and
the known Eocene: Evanston formation (Eocene?) (p. 86);
"Wasatch group (pp. 87-96); Green River formation (p. 97);
Bridger formation (p. 99). Divides Wasatch group into three
formations, only the uppermost of which (Knight formation)
contains vertebrate remains.
Weed, Walter Harvey.
1896.1. The Fort Union formation: Am. Geologist, vol. 18,
pp. 201-211, 1896.
Reviews early descriptions of the Fort Union "group,"
gives two sections of Fort Union strata in Montana, and dis-
cusses physical and faunal characters.
Weeks, Fred Boughton.
1902.2. North American geologic formation names, bibli-
ography, synonymy, and distribution: U. S.
Geol. Survey Bull. 191, 448 pp., 1902.
Weeks, Fred Boughton — Continued.
1907.1. Stratigraphy and structure of the Uinta Range:
Geol. Soc. America BuU., vol. 18, pp. 427-448,
6 pis., 1907.
Describes the occurrence and relations of pre-Cambrian,
Paleozoic, Mesozoic, and Tertiary formations and the geologic
structure of the region.
Wegemann, Carroll Harvey.
1917.1. Wasatch fossils in so-called Fort Union beds of the
Powder River Basin, Wyo., and their bearing
on the stratigraphy of the region: U. S. Geol.
Survey Prof. Paper 108, pp. 57-60, pis. 22-23,
fig. 16, 1917.
White, Charles Abiathar.
1876.1. See Powell, John Wesley, 1876.1.
1878.1. Report on the geology of a portion of northwestern
Colorado: U. S. Geol. and Geog. Survey Terr.
Tenth Ann. Rept., pt. 1, pp. 3-60, 1 map, 1878.
General account of Uinta formation; thickness 1,200 feet,
resting unconformably upon other Tertiary. Refers to ex-
posures of Bridger "group" in the Uinta Basin (p. 37).
Willis, Bailey.
1908.1 (and Hovey, E. O.). Symposium on correlation.
Section E, American Association for the Ad-
vancement of Science and Geological Society
of America: Science, new ser., vol. 28, No. 729,
pp. 878-879, 1908.
Wortman, Jacob Lawson.
1882.1. The geology of the Big Horn Basin. In Cope,
. E. D., Contributions to the history of the Verte-
brata * * * Qf Wyoming: Am. Philos. Soc.
Proc, vol. 20, pp. 139-142, 1882.
1892.1. See Osborn, Henry Fairfield, 1892.67.
1893.1. On the divisions of the White River or Lower
Miocene of Dakota: Am. Mus. Nat. Hist.
Bull., vol. 5, pp. 95-105, June 27, 1893.
Division of White River into three zones; Titanotherium,
Oreodon, and Protoceras.
1894.1. See Osborn, Henry Fairfield, 1894.90.
1895.1. See Osborn, Henry Fairfield, 1895.105.
1903.1. Studies of Eocene Mammalia in the Marsh collec-
tion, Peabody Museum, Part II, Primates: Am.
Jour. Sci., 4th ser., vol. 15, pp. 163-176, 399-
414, 419-436; vol. 16, pp. 345-368, pis. 11-12;
vol. 17, pp. 23-33, 133-140, 203-214, figs. 100-
146, 1903-4.
Origin of tlie primates (vol. 15, pp. 419-436).
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE I
A. ERUPTION OF THE CRATER OF TAAL, PHILIP-
PINE ISLANDS, JANUARY, 1911
Submerged layers
of volcanic cinders, ashes, a
C. Worce^er.) Comparable
B. FLOODED AREA
I mud, in vi^hich are entombed the bodies of men and th(
> volcanic ash deposits of Bridger age in southern Wyoming
of animals. (After De
■D. S. GEOLOGICAL SURVEY
MONOGRAPH 65 PLATE II
A. OJO ALAMO, SAN JUAN COUNTY, N. MEX., LOOKING NORTH
Contadl (indicated by arrows) betw^een Ojo Alamo sanditone and clay of Puerco formation is observed dire(5tly back of the
trading store
EjJqc
'"*^*f*^/$0'^
;i. EXPOSURES OF PUERCO FORMATION EAST OF OJO ALAMO, N. MEX.
sfts on eroded surface of Ojo Alamo sandftones; contaiS: indicated by dotted line. The dark Stratum at top to the right is the lower
level i^£tocoyiu.s 2:one) of the Puerco formation
CRETACEOUS AND BASAL EOCENE CONTACTS IN NEW MEXICO
Photographs by W. J. Sinclair,'1913. (After Sinclair>nd Granger, 1914.1)
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE III
/ rlij-(jf }!tt^4/> X -y
A. UPPER BEDS OF TORREJON FORMATION, WEST FORK OF TORREJON ARROYO, SANDOVAL COUNTY, N. MEX.
The Torrejon is overlain unconformably^y the basal sand^ones of the Wasatch(?) formation
JPotyrn-fy-stodnTi zo/ie
•■' iliiiii if^,. .
^^T u ' . ■ < ^}*^mmi
B. EXPOSURES OF PUERCO FORMATION 3 MILES EAST OF OJO ALAMO, N. MEX.
BASAL EOCENE AND LOWER EOCENE CONTACTS IN NEW MEXICO
Photographs by W. J. Sinclair, 1913. (After Sinclair and Granger, 1914.1)
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE IV
jM-'Of ** * J ^
A. EOHIPPUS-CORYPHODON ZONE (LOWER PART OF WASATCH FORMATION, LEVEL BIG HORN B), LITTLE SAND
COULEE
Fir^ appearance of Eohipptis. Am. Mus. negative 18565
B. PHENACODUS-NOTHODECTES-CORYPHODON ZONE (BASAL PART OF WASATCH FORMATION, LEVEL BIG HORN A),
ABOUT 4 MILES NORTH OF RALSTON
Am. Mus. negative 18563
LOWER WASATCH STRATA RESTING ON BASAL WASATCH STRATA, CLARK FORK BASIN,
PARK COUNTY, WYO.
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE V
A. TYPICAL "LYSITE" LOCALITY, AT COTTONWOOD DRAW, NORTH OF LOST CABIN, WIND RIVER BASIN, WYO.
Shows the Heptodon-Coryphodon-Eohippus zone (level Wind River A), with Paleojoic hills in the background. (After Granger,
1910.1.) Am. Mus. negative 18393
B. TYPICAL "GRAY BULL" LOCALITY, 4 MILES SOUTH OF OTTO, BIG HORN BASIN, WYO.
Shows the Syilemodon-Coryphodon-Eohippus sone flevcl Big Horn C), with the excavation of the skeleton of Eohippus osboi
in the foreground. Am. Mus. negative 18450
EXPOSURES OF WASATCH FORMATION AND TYPICAL WIND RIVER DEPOSITS IN WYOMING
TJ. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE VI
( SiA.'Cr'f'cu-vo JJ }
I*,
iMtoik^''
■Msiti--
»m
^^f
^^avjiKKsaitM
A. A T^'I'ICAI. HL'ERFANO LOCALITY, 2 MILES WEST OF GARDNER, HUERFANO BASIN, COLO.
Palaeosyops fontinalis zone (level Huerfano B). A cedar-covered ridge in midale diftance, and eruptive peaks in the background. The
type of Eometarhinus and referred specimens of Palaeosyops fontinalis were found at this site. Am. Mus. negative 104715
J^ cirrti do ih4^riz<^n^ zorve
imm
B. A TYPICAI- "LOST CABIN" LOCALITY, ON ALKALI CREEK, EAST OF LOST CABIN, WIND RIVER BASIN, WYO.
Lamhdotheriunt sone (level Wind River B). The types of Lambdotherium po^oagicum, Eotitanops horealis, and E. gregoryi -were
found at this site. Am. Mus. negative 18392
TYPICAL HUERFANO FORMATION OF COLORADO AND WIND RIVER FORMATION OF
WYOMING
V. S. GEOLOGICAL STJBVET
MONOGRAPH 55 PLATE VII
A. HENRYS FORK TABLE, LOOKING NORTHWARD ACROSS HENRYS FORK, BRIDGER BASIN, WYO.
[UintatheriUTti zone (levels Bridger C and D) and ^etarhinus zone (level Bridger E) -with Bishop ("Wyoming") conglomerate at the
top. The Burnt Fork "white layer" (w and arrow) separates level Bridger C 2 from Bridger C 3. Am. Mus. negative 18152
B. GRIZZLY BUTTES, SOUTH OF MOUNTAIN VIEW, UINTA COUNTY, WYO.
Palaeosyops paludosuS'Orohippiis zione (level Bridger B). Excavation by Granger of the type skull of Limnohyops priscus (Am. Mus.
11687). Am. Mus. negative 18089
TYPICAL BRIDGER FORMATION (LEVELS UPPER C, D, E, AND LOWER B) OF WYOMING,
MIDDLE AND UPPER (?) EOCENE
n. s. cKoi.onirAL survey
MONOGRAPH 55 PLATE VIII
A. NORTHWEST POINT OF HAYSTACK MOUNTAIN, HEAD OF BITTER CREEK, SWEETWATER COLINTY, WYO.
Eohasileus-Dolichorhinus and Metarhinus aones (levels Washakie B 2 and B 1). (After Granger.) Am. Mus. negative 18213. (Se
figs. 60 and 61)
Wash-akyoe £
WasTtaJtie- A
B. VIEW SOUTHEASTWARD FROM LACLEDE STATION ON OVERLAND STAGE TRAIL, SWEETWATER COUNTY, WYO.
Lower brown sandstones of \Jiy\tat\\,^rium 2;one (level Washakie A) in middle di^ance. Hayrack Mountain and the Eobasileus'
1 the background. (After Granger.) Am. Mus. negative 18223
:i sandilonei
of \Jiy\taihe.rium 5c
lis and Metarhinus :
TYPICAL "WASHAKIE" FORMATION (LEVELS B 2 AND B 1 OVERLYING LEVEL A 1), WYOMING;
MIDDLE AND UPPER EOCENE
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE IX
A. COLUMNAR SANDSTONES, TOP OF LEVEL XJINTA A, WHITE RIVER CANYON, UINTA BASIN, UTAH
(After Riggs, 1912.1.) Field Mus. negative
B. PANORAMIC VIEW, WHITE RIVER CANYON, 4 MILES BELOW WAGONHOUND BEND, UINTA BASIN, UTAH
Bluffs on right bank of river belong to the unfossiliferous level Uinta A. Photograph by Riggs. Field Mus. negative
MIDDLE EOCENE OF NORTHERN UTAH (LEVEL UINTA A)
tr. S. GEOLOGICAL SURVET
MONOGRAPH 55 PLATE X
A 7 11
J\.myru)do7T^ sandstoTve
JSobcLsile-us -
DolicJioj'Tzisn
A. NORTHERN BOUNDARY OF COYOTE BASIN, UINTA BASIN. UTAH
Showing greenish clays of the 'Exibas^Xe^KS'■T>6iich<yr^^iy^,us zone (level Uinta B 2) capped by '^* Amyyiodotx, sandstone." These
clays have yielded mo^ of the smaller mammaliin fauna of this middle horison of Uinta Basin. (After Riggs, 1912.1.)
Field Mus. negative
B. DIVIDE BETWEEN WHITE RIVER CANYON AND COYOTE BASIN, UINTA BASIN, UTAH
Showing fossil-bearing sand^one of the Metarhinus ?one (level Uinta B 1). (After Riggs, 1912.1.) Field Mus. negative
UPPER EOCENE OF NORTHERN UTAH (LEVEL UINTA B)
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XI
I (fanolht^/ iinii
^\rn\ liOiioTL (?)aj'tiLou-u^s
A. NORTH FACE OF BEAVER DIVIDE, WIND RIVER BASIN, WYO.
View we^^vard from point near Wagonbed Spring, showing lo\ver Oligocene beds (.Titanotherium zone), with Menodus heloceras
(level Chadron A), reding on upper Eocene {'Diplacodon sone?). Skull provisionally referred to Amynodon antiqutis was
taken from left foreground. (After Granger, 1910.1.) Am. Mus. negative 18388
-^ft'v"*.^-
B. EXPOSURES AT WAGONBED SPRING, BEAVER DIVIDE, FREMONT COUNTY, WYO.
Sho\ving contact between upper Eocene and lower Oligocene. The skull of Menodus heloceras came from the dra-w ju^ to the right
of this view. Am. Mus. negative 18391
LOWER OLIGOCENE OVERLYING UPPER EOCENE OF CENTRAL WYOMING
V. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XII
A. CONTACT BETWEEN TITANOTHERIUM ZONE (LOWER OLIGOCENE) AND PIERRE SHALE (CRETACEOUS), NEAR
MOUTH OF CEDAR CREEK, BIG BADLANDS, S. DAK.
Orcodo-n zone in the di^ance to the right. Am. Mus. negative 35997
v>
Uirvtco C J
Dif^^a?^
B. BADLANDS SOUTH OF WHITE RIVER, UTAH
Showing Diplacodon 5one (level Uinta C 1, upper Eocene) in foreground and level Uinta C 2 in diilance- (Compare fig. 66.) Am.
Mus. negative 17665
101959— 29— VOL 1
D. S. GEOLOGICAL SURVEY
MONOGRAPH 56 PLATE XIV
EXPOSURES AT QUINN DRAW, BIG BADLANDS, S. DAK.
Showing summit of lower Oligocene Chadron formation (Titanotherium zone) and, at the top, base of younger Brule formation (Oreodon :;one). The
sandstone columns in the center indicate a river channel betvveen underlying and overlying claya. Am. Mus. negative 36012
V. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XV
A. SOUTH END OF SHEEP MOUNTAIN, NEAR HEAD OF CORRAL DRAW, BIG BADLANDS, S. DAK.
Showing Oreodon zone (Brule formation). Am. Mus- negative 36006
B. CEDAR CREEK, BIG BADLANDS, S. DAK.
Showing Oreodon zone (Brule formation) overlying Titanotherium zone (Chadron formation). Am. Mus. negative 36013
BRULE AND CHADRON FORMATIONS OF SOUTH DAKOTA
CHAPTER III
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS OF THE TYPES
SECTION 1. HISTORY OF DISCOVERY
Full descriptions of the geologic and geographic
positions of the several types and kinds of titano-
theres are given in Chapter II. The present chapter
relates the history of the explorations and of the
gradual discovery of the character and relations of the
titanotheres.
THE OLIGOCENE TITANOTHERES
THE PIONEER PERIOD: PROUT, OWEN, EVANS, LEIDY
(1846-1873)
The Big Badlands of South Dakota and north-
western Nebraska are even now practically unknown
to most Americans. As these lands lie in an arid
region far from navigable rivers — a region that was
formerly occupied by hostile Indians and that offers
little attraction to either the prospector or the set-
tler— it is not surprising that their fossil wonders long
lay hidden from the world. The fossil remains of the
great animals described in this monograph were known
to the Indians and referred to in their mythology as
"thunder horses." (See Preface, p. xxi.)
In 1846 Dr. Hiram A. Prout, of St. Louis, sent to
Professors Dana and Silliman of Yale College a cast
of a remarkable fossil that he had received from "a
friend residing at one of the trading posts of the St.
Louis Fur Co. on the Missouri River." Front's brief
notes, together with a crude sketch of one of the lower
molars, were accordingly published in the American
Journal of Science and Arts. (Prout, 1846.1, pp.
288, 289.) In a later communication Prout (1847.1)
stated that this fossil (fig. 85) was discovered in the
"Mauvais Terre, on the White River, one of the west-
ern confluents of the Missouri." This was the famous
specimen described by Prout as a "gigantic Palaeo-
therium," which Leidy tells us (1852.1, p. 551) was
"the first of the many mammalian remains which have
been brought to the notice of the scientific world from
the vast Eocene cemetery of Nebraska." It thus gave
the first hint to scientists that "the region of Nebraska
Territory of the United States appears to be as rich in
the remains of Mammalia and Chelonia of the Eocene
period as the deposits of the same age of the Paris
Basin." (Leidy, 1852.1, p. 539.)
The fossil jaw described by Prout represented an
animal of great size. "The entire jawbone," he says,
"must have been at least 30 inches long, which far
exceeds in size the PalaeotJierium magnum." The
reference to Cuvier's PalaeotJierium was, under the
circumstances, very natural, because the lower molars
of Front's specimen were surmounted by crescentic
cutting surfaces somewhat like those of Palaeotherium.
This discovery evidently attracted attention abroad,
for in 1849 the French paleontologist Pomel (1849.1,
pp. 73-75), after carefully considering Prout's descrip-
tion and figures, stated that the fossil represented a
new subgenus of paleotheres, for which he proposed
the name Menodus giganteus, the generic name re-
ferring to the crescents of the lower molars, the specific
name to the great size of the animal.
Meanwhile (in 1839, 1840-1849) the United States
Government geologist. Dr. David Dale Owen, was
making his extensive geologic reconnaissance of Wis-
consin, Iowa, and adjacent States. In his final report
(Owen, 1852.1, p. 194) he tells us that he was "de-
sirous, if possible, to connect the geology of the Missis-
sippi Valley, through Iowa, with the Cretaceous and
Tertiary formations of the upper Missouri, a matter
very important to the proper understanding of the
formations of the intervening country, which it had
been made my particular duty to explore." Finding
it impracticable to explore the Missouri region himself
he detailed to this work one of his assistants, Mr. John
Evans. Late in the field season of 1849 Evans "finally
reached that most curious unexplored region, the corner
of the 'Badlands' (Mauvaises Terres), lying high up
on White River, a locality which seemed likely, above
all others, to furnish satisfactory information regard-
ing the precise character and age of the Tertiary de-
posits of the upper Missouri country." (Owen, 1852.1,
p. 195.)
From Evans's report (p. 197) Owen gives the fol-
lowing description of the Mauvaises Terres of White
River:
To the surrounding country, however, the Mauvaises Terres
present the most striking contrast. From the uniform, monoto-
nous open prairie, the traveler suddenly descends, one or two
hundred feet, into a valley that looks as if it liad sunk away
from the surrounding world, leaving standing, all over it,
thousands of abrupt, irregular, prismatic, and columnar masses,
frequently capped with irregular pyramids and stretching up
to a height of from one to two hundred feet or more.
So thickly are these natural towers studded over the surface
of this extraordinary region tliat the traveler threads his way
through deep, confined, labyrinthine passages, not unlike the
narrow, irregular streets and lanes of some quaint old town of
the European continent. Viewed in the distance, indeed, these
rocky piles, in their endless succession, assume the appearance
of massive artificial structures, decked out with all the acces-
sories of buttress and turret, arched doorway and clustered
shaft, pinnacle and finial, and tapering spire.
One might almost imagine oneself approaching some magnifi-
cent city of the dead, where the labor and the genius of for-
gotten nations had left behind them a multitude of monuments
of art and skill.
On descending from the heights, however, and proceeding
to thread this vast labyrinth and inspect, in detail, its deep,
intricate recesses, the realities of the scene soon dissipate the
141
142
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
delusions of the distance. Tlie castellated forms which fancy
had conjured up have vanished, and around one, on every
side, is bleak and barren desolation.
Then, too, if the exploration be made in midsummer, the
scorching rays of the sun, pouring down in the hundred defiles
that conduct the wayfarer through this pathless waste, are
reflected back from ' the white or ash-colored walls that- rise
around, unmitigated by a breath of air or the shelter of a soli-
tary shrub.
The drooping spirits of the scorched geologist are not per-
mitted, however, to flag. The fossil treasures of the way well
repay its sultriness and fatigue. At every step objects of the
highest interest present themselves. Embedded in the debris
lie strewn, in the greatest profusion, organic relics of extinct
animals. All speak of a vast fresh-water deposit of the early
Tertiary period and disclose the former existence of most re-
markable races that roamed about in bygone ages high up in
characters belonging now to the above three orders; for the
molar teeth are constructed after the model of those of the
hog, peccary, and babyroussa; the canines as in the bear;
while the upper part of the skull, the cheek bones, and the
temporal fossa assume the form and dimensions which belong
to the cat tribe. Another, the Oreodon of Leidy, has grinding
teeth like the elk and deer, with canines resembling the omnivo-
rous thick-skinned animals, being, in fact, a race which lived
both on flesh and vegetables and yet chewed the cud like our
cloven-footed grazers.
Associated with these extinct races we behold also, in the
Mauvaises Terres, abundant remains of fossil Pachydermata of
gigantic dimensions and allied in their anatomy to that sin-
gular family of proboscidate animals of which the tapir may be
taken as a living type. These form a connecting link between
the tapir and the rhinoceros; while, in the structure of their
grinders, they are intermediate between the daman and rhinoc-
FiGUKE 83 — Mauvaises Terres, Nebraska. After David Dale On en, 1851
the valley of the Missouri, toward the sources of its western
tributaries, where now pastures the big-horned Ovis montana,
the shaggy buffalo or American bison, and the elegant and
slenderly constructed antelope.
Owen continues (p. 198) with a popular description
of the extinct animals found:
Every specimen as yet brought from the Badlands proves to
be of species that became exterminated before the mammoth
and mastodon lived and differ in their specific character, not
alone from all living animals, but also from all fossils obtained
even from cotemporaneous geological formations elsewhere.
Along with a single existing genus, the Rhinoceros, many new
genera never before known to science have been discovered,
and some, to us at this day, anomalous families, which com-
bine in their anatomy structures now found only in different
orders. They form, indeed, connecting links between the
pachyderms, plantigrades, and digitigrades. For example, in
one of the specimens from this strange locality, described by
Dr. Leidy under the name Archiotherium, we find united
eros; by their canines and incisors, they connect the tapir with
the horse, on the one hand, and with the peccary and hog on
the other. They belong to the same genus of which the labors
of the great Cuvier first disclosed the history, under the name
of Palaeotherium, in publishing his description of the fossil bones
exhumed from the gypsum quarries of Montmartre, near Paris,
but are of distinct species; and one at least, of this genus, dis-
covered in the Badlands (Palaeotherium proutii), must have
attained a much larger size than any which the Paris Basin
afforded. In a green, argillo-calcareous, indurated stratum,
situated within 10 feet of the base of the section, a jaw of this
species was found, measuring, as it lay in its matrix, 5 feet
along the range of the teeth, but in such a friable condition,
that only a portion of it could be dislodged; and this, notwith-
standing all the precautions used in packing and transportation,
fell to pieces before reaching Indiana.
A nearly entire skeleton of the same animal was discovered,
in a similar position, which measured, as it lay embedded, 18
feet in length, and 9 feet in height. But here, as in the former
case, the crumbling condition of the bones rendered it impos-
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
sible to disinter them whole; and the means of transportation to
the Missouri were insufficient, even if these interesting remains
could have been extracted in good condition.
Figure 84. — "Vertical view of the posterior tooth belonging to the lower jaw of
Mr. Prout's Palaeoiherium; natural size." After Prout, 1846
Owen also gives (1852.1, p. 200) a tabular "Section
of beds constituting the early Tertiary (Eocene) of
the Badlands." This section, reproduced below, was
doubtless taken by Evans.
143
The detailed description of the mammalian and
chelonian fossils collected by Owen, Evans, and others
was intrusted to Dr. Joseph Leidy, of Philadelphia,
and was published in Owen's report
of 1852 (1852.1, pp. 551, 552),
already cited. In this publication
Leidy describes Prout's original
specimen and gives a poor figure
of Evans's principal specimen, an
imperfect lower jaw of a titanothere.
He adopts provisionally the
name Palaeotherium? proutii Owen,
Norwood, and Evans but con-
cludes his description of these fos-
sils with the following significant
remark :
AU the preceding specimens, except, probablj', the latter two
(fragments of upper molars), I suspect belong to a different
genus from either Palaeotherium or A nchitherium, and should the
suspicion prove correct, Titanotherium would be a good name
for the animal, as expressive of its very great size.
Section of beds constituting the early Tertiary {Eocene)
of the Badlands (Mauvaises Terres)
30
[Numbered in descending order]
1. Ash-colored clay, cracking in the sun; con-
taining siliceous concretions
2. Compact white limestone
3. Light-gray marly limestone
4. Light-gray indurated siliceous clay (not effer-
vescent)
5. Aggregate of small angular grains of quartz,
or conglomerate, cemented by calcareous
earth; slightly effervescent 8
6. Layer of quartz and chalcedony (probably
only partial) 1
7. Light-gray indurated siliceous clay, similar
to No. 4 but more calcareous, passing down-
ward into pale flesh-colored indurated
siliceous marly limestone (effervescent) ;
turtle and bone bed 25
8. White and light-gray calcareous grit; slightly
effervescent 15
9. Similar aggregate to No. 5 but coarser 8
10. Light-green indurated argillaceous stratum
(slightly effervescent) ; ' ' palaeotherian bed" - 20
Some of the specimens brought back by
Evans were referred to in a brief notice pub-
lished by Owen, Norwood, and Evans (1850.1),
in which the name "Palaeotherium proutii"
was proposed "in compliment to Dr. Prout,
of St. Louis."
The next year (1850) after Evans's journey
Mr. Thaddeus A. Culbertson visited, under
the auspices of the Smithsonian Institution,
the same region (Leidy, 1854.1, p. 12)
and "made a good collection of its animal
remains," including skulls of Oreodon culbertsoni
and the titanothere upper premolars which Leidy
afterward described (1852.2, p. 2) under the names
Rhinoceros americanus and Eotherium americanus.
The locality was Bear Creek, a dry tributary of
Cheyenne Kiver. (See Chap. II, p. 104.)
Figure 85. — Original figures of Prout's "gigantic Palaeotherium," the
first titanothere discovered. After Prout, 1847
A, "Fragment of the inferior maxillary of the left side," one-fourth natural size; B, last lower
molar, left side, four-fifths natural size.
Thus was proposed the name Titanotherium, which
has figured so largely in the literature of American
paleontology and was consequently chosen as the basis
for the title of this monograph.
Two years later Leidy (1854.1) gave a revised and
extended description of the available remains of titano-
144
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
theres, which was accompanied by excellent litho-
graphs of Prout's and other fragmentary specimens.
At this stage of knowledge the only parts figured
under the name Titanotherium proutii included the
lower molars, a lower premolar, a lower canine, a frag-
mentary upper molar, and two upper premolars
(Leidy's types of Rhinoceros americanus). Fragments
of large upper molars were named Palaeotherium
giganteum.
The "palaeotherian bed" of Owen and Evans is
referred to by Leidy (op. cit., p. 13) as the "Titano-
therium bed." This appears to be the first use of this
term, which was afterward, in the form " Titano-
therium beds" (now Titanotherium zone), so widely
used by geologists and paleontologists.
An interval of 15 years in the literature of the sub-
ject, from 1854 to 1869, is broken only by Prout's
brief notice of an indeterminate molar (now lost) of a
titanothere, which he named Leidy otherium. But
during this seemingly barren interval Meek and
Hayden were making their historical explorations
(Merrill, 1906.1, pp. 585-592), which resulted in
notable advances in our knowledge of the relations
of the geologic deposits of the Rocky Mountains and
Great Plains. They also brought back many verte-
brate fossils, including specimens of Titanotherium.
One of the specimens of titanotheres collected by
Meek and Hayden included a nearly complete series
of upper teeth. This specimen, which belonged to
Prof. James Hall and is now No. 433 of the Hall
collection of the American Museum of Natural
History, was described and figured by Leidy in his
memoir of 1869 (1869.1, pp. 206, 207, pi. 24) and was
by far the best spec'men that had been described up
to that time. Leidy referred it to his species Titano-
therium proutii, but it probably belongs in the genus
that Marsh afterward named Brontotherium. This
specimen misled Leidy into assigning Titanotherium
to the Artiodactyla. "From the form of its lower
true molars, which were first discovered," he says,
"it was supposed to be more nearly alHed with the
Palaeotherium and was hence placed among the uneven-
toed pachyderms, or Perissodactyla, but the nearly
complete dentition of both jaws, since discovered,
appears to indicate its position to be as above stated " —
that is, it appeared to be "nearly allied with Chali-
cotherium, and, like it, approximates the even-toed
pachyderms, or Artiodactyla * * * with the Ru-
minantia."
In 1870 Leidy (1870.1, pp. 1, 2) described a frag-
mentary fossil from Colorado that had been submitted
to him by Doctor Hayden. We now know that this
specimen consists of the horn cores and attached
coossified nasal bones of a titanothere of some sort,
but to Leidy, who knew practically nothing of the
skull of the titanothere, it proved "singularly puzzling
in character." He at first thought it might pertain
to Titanotherium, "but in the state of extreme uncer-
tainty as to its collocation, it may with equal proba-
bility be referred to other genera, perhaps to Megalo-
meryx, or it may have been an American species of
Sivatherium. Under the circumstances it may be
referred to a new genus, with the name of Megacerops
color adensis ."
This problematical fossil was redescribed and figured
by Leidy in his memoir of 1873 (1873.1, p. 239). He
states that the specimen "appears to correspond
with that portion of the face * * * [of Siva-
therium] which comprises the upper part of the nose,
together with the forehead and the anterior horn
cores." He compares the specimen with the corres-
ponding parts of the Sivatherium, the rhinoceros, the
tapir, and the mastodon. He decides that the frag-
mentary horn core formerly attributed to Titano-
therium may perhaps belong to another species of
Megacerops.
This erroneous determination, together with the
previous assignment of Titanotherium to the Artio-
dactyla, shows how greatly Leidy, even with all his
skill and caution, was deceived by the lack of well-
preserved and definitely associated feet and skulls,
a lack which is felt to some extent even at the present
time.
Leidy's description of Megacerops may be regarded
as marking the close of the first or pioneer period in
the study of the titanotheres, a period characterized
by (1) the chance discovery of "Prout's specimen,"
(2) the exploration of the White River badlands
by Evans, Hayden, and others and the resulting
knowledge of the general geologic age of the beds, (3)
the description of fragmentary remains of titanotheres,
chiefly teeth, by Prout, by Pomel, and by Leidy in
successive publications, together with the beginnings
of the systematic nomenclature, (4) the erroneous
reference of Titanotherium to the Anoplotheriidae
among the Artiodactyla.
TAXONOMIC ARRANGEMENT AND COMPARISON
WORK OF MARSH AND COPE (1870-1887)
The second period in the study of titanotheres,
which may be called the period of systematic descrip-
tion, really began before the first period had closed
(1873).
From 1873 to 1891, inclusive, the literature of the
Oligocene titanotheres is dominated almost exclu-
sively by the explorations and systematic contribu-
tions of Marsh and Cope. During this time Marsh
described eight genera and fourteen species as new,
and Cope described three genera and twelve species
as new. The solution of the exact systematic and
phylogenetic interrelations of these genera and species
is one of the principal themes of Chapters IV to VII of
the present monograph.
In 1870 Prof. Othniel C. Marsh (1870.1) headed an
expedition sent from Yale College to northern Colo-
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
145
rado, where he not only discovered and explored "an
extensive outcrop of the true Mauvaises Terres, or
White River formation," but also procured some mag-
nificent specimens of titanotheres (including the types
of Brontotherium gigas and B. ingens), which he de-
scribed and figured three years later. Marsh was also
able to solve the problem of the ordinal relationships
of the titanotheres (1873.1, p. 486), showing that his
Brontotherium gigas was a "true perissodactyl with
limb bones resembling those of Rhinoceros. The genus
is related to Titanotherium, and the two appear to form
a distinct family, which may be called Brontotheridae."
He was able in a very few words to throw a flood of
light upon the characters of the skeleton, hitherto
known chiefly from fragments :
It closely resembles that in recent perissodactj^ls but shows
some approach to the Proboscidea. The femur has a third
trochanter, and its head a pit for the round ligament. The
fibula is entire and slender. The astragalus is remarkably
short. It has a deep groove on its upper surface, and the
articular facets for the navicular and cuboid are nearly equal.
In the manus there are four toes of nearly equal size, the first
digits being rudimentary or wanting. There were three digits
only in the pes, the first and fifth being entirely wanting. The
toes were short and thick, as in proboscidians. The meta-
carpals and metatarsals are longer than in the elephant, and
the phalanges shorter. The foot was also more inclined. The
carpal and tarsal bones are very short and form interlocking
series. The tail was long and slender.
An important point not touched upon in this com-
munication was the presence or absence of horns.
Prof. Edward D. Cope was not far behind Marsh in
contributions to the literature of the titanotheres.
Two years after Marsh had made his explorations in
Colorado, Cope, in 1872, discovered a number of re-
markable skulls (now in the Cope collection of the
American Museum of Natural History) which, in bul-
letins pubhshed in 1873 and 1874, he made the types
of Symhorodon torvus, Megaceratops acer, M. heloceras,
Symborodon hucco, S. dltirosfris, S. trigonoceras. He
states (1873.2, pp. 2, 3) that "Leidy and Marsh have
described two genera of this group, viz, Titanotherium
and Brontotherium, but without certain indications of
their possession of horns." He regards them as "all
true perissodactyls and allied to the Rhinoceros and
Palaeotherium." His genus Symhorodon, like Menodus,
Titanotherium, and Brontotherium, was "established
on mandibular rami only, which can not be certainly
associated with crania," the last phrase suggesting one
of the most troublesome and obdurate of titanothere
problems, which from the first has caused confusion in
the systematic nomenclature. Cope regarded the
absence of incisors as one of the generic characters
that separated Symhorodon from Titanotherium and
Brontotherium, thus first raising the problem how far
differences in the number of incisors may correspond
to true generic differences. The discovery of so many
more or less complete skulls enabled Cope to infer
specific and generic characters from the variations in
form of the horn cores, skull top, nasals, and zygo-
matic arches. Thus the discoveries of Cope and
Marsh, although they settled the ordinal relationships
of the titanotheres, began to complicate the problem
of their interrelationships.
SUMMARY OF MAESH'S CONTRIBTTTIOKS
In Marsh's paper "On the structure and affinities
of the Brontotheridae" (1874.1) he developed further
the family characters of the group, separating them
from the Rhinocerotidae, "apparently their near
allies," establishing the number of digits in the fore
and hind feet and the general characters of the skull,
lower jaw, vertebrae, and limbs. This paper is ac-
companied by the first of a series of excellent litho-
graphic plates, illustrating some of Professor Marsh's
superb specimens of titanotheres from Colorado.
Marsh contributed another short but pregnant article
on the "Principal characters of the Brontotheridae"
in 1876 (1876.1), and after that he published at in-
tervals brief descriptions of supposedly new genera
and species, not all of them accompanied by illustra-
tions, until September, 1891, the date of his last
published contribution to the subject.
Marsh's most valuable contributions to our knowl-
edge of the titanotheres may be summarized as follows:
(1) He and his party explored the White River forma-
tion in Colorado and collected from it many remark-
ably fine specimens; (2) he demonstrated the ordinal
position of the group, classifying its members as
perissodactyls; (3) he recognized the fact that the
titanotheres constitute a distinct family, which he
named the Brontotheridae; (4) he made the illuminat-
ing observation that his upper Eocene genus Diplaco-
don served to connect the Oligocene Brontotheridae
with the Eocene "Limnohyidae"; (5) he published
many excellent lithographs and woodcuts, showing
chiefly the skulls and dentition of titanotheres, but
including also (1889) an excellent restoration of
Brontops rolustus; (6) he supervised the preparation
of a fine series of lithographic plates for the present
work; (7) under the auspices of the United States
Geological Survey he founded the present series of
monographs on fossil vertebrates; (8) he began the
preparation of the present monograph, although he
left no manuscript for it; (9) he obtained for the
National and Yale Museums their superb specimens of
titanotheres, most of which were collected by his field
assistant J. B. Hatcher, who in turn also made valuable
scientific contributions to our knowledge of these
animals.
Marsh's detailed systematic work on the titanotheres
was less fortunate than his broader contributions,
owing chiefly to confusion in regard to features of the
skull and jaw. After founding the genus and species
Brontotherium gigas upon a lower jaw, he referred to the
146
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
same genus as the type of B. ingens, a skull that
certainly belongs to another genus (Menodus). In
consequence of this initial confusion he erected a new
genus (Titanops) for skulls that should have been
referred to Brontotherium. Many of his conceptions
of the interrelations of the genera and species proposed
by him were erroneous. Although recognizing the
fact that the genera Brontops, Allops, and Teleodus
were all allied to " BrontotTierium" — that is, as repre-
sented by the skull of "BrontotTierium [Menodus]
ingens" — he nevertheless thought thsLt Diplodonus was
related to Titanops (the true BrontotTierium), and he
referred to Menops (a near ally of his "BrontotTierium''
ingens) a well-preserved skull that is now known to
belong to BrontotTierium proper. In fact, in common
with Cope and others. Marsh apparently faUed to
recognize the comparatively wide phyletic gap between
the true BrontotTierium (his Titanops) and Cope's
Symborodon on the one hand and the supposed genera
Brontops, Allops, Menops, and Menodus (his "Bronto-
tTierium ingens") on the other. Consequently his
generic definitions are unsatisfactory, and he was
certainly not overconservative in proposing new
generic and specific terms.
SUMMARY OF COPE'S CONTRIBUTIONS
The next year (1874) after publishing his prelimi-
nary descriptions of the several species of Symborodon
and allied genera Cope (1874.2) gave full descriptions
of these forms in his "Report on the vertebrate pale-
ontology of Colorado," which was accompanied by
eight lithographic plates. He presented a careful
review of the general morphology of the skull, includ-
ing the brain case and cranial antra, which was fol-
lowed by a review of the work of preceding authors and
by a tabular analysis and detailed description of the
species of Symborodon. He recorded many interesting
facts, such as the similarity of the dentition of Sym-
borodon to that of Palaeosyops and of CTialicotTierium
and the mingling of proboscidian and rhinoceros
analogies in the limbs. He considered the indications
that Symborodon possessed a short proboscis. In his
tabular analysis of species he indicated the differences
in the shape of the horns and noted that in S. trigono-
ceras and S. Tiypoceras the upper premolars have a
strong internal basal cingulum, whereas in S. bucco
and S. altirostris the premolars are "without inner
basal cingulum."
Cope, like Marsh, failed to distinguish the sexes as
well as the separate groups or phyla of titanotheres.
His "S." trigonoceras , for example, is a Menodus, a
member of an altogether different group from his "S."
Tiypoceras, which is a BrontotTierium.
After an interval of 12 years, in 1886, Cope (1886.1)
described the first Canadian species, Menodus angusti-
genis, basing it upon fragments discovered by Mc-
Connell and Weston for the Geological and Natural
History Survey of Canada. Three years later (1889.1,
p. 153) he referred this form to a new genus, Hapla-
codon, and in the same year (1889.2, pp. 628, 629) he
described two other Canadian species. His review
(1891.2, p. 17) of these forms and attempted revision
of the nomenclature were involved and unsatisfactory.
He recognized only two genera, Menodus and Symbo-
rodon. The last species of titanothere described by
him was his Menodus peltoceras (1891.1), which is
probably a female of Marsh's BrontotTierium curtum.
EEINTERPRETATION AND PHYLOGENETIC STUDY
(OSBOEN, 1887-1919)
STUDY OF CERTAIN FEATURES
Before Marsh and Cope had ceased naming new or
supposedly new genera of titanotheres a turn was
given to the trend of study by a paper by Scott and
Osborn (1887.1, pp. 157, 158), entitled "Preliminary
account of the fossil mammals from the White River
formation contained in the Museum of Comparative
Zoology." This paper, which was a description of the
interesting collection made by Mr. Samuel Garman
under the auspices of Prof. Alexander Agassiz, reacted
from the polynomial systems of Marsh and Cope and
tended toward a mononomial system. In this paper
the Perissodactyla were described by Osborn, the
Artiodactyla and Carnivora by Scott. Before de-
scribing the new titanothere material the authors
noted the difficulty in deciding where to draw generic
lines, a difficulty that is increased by the fact that the
mandibles are seldom found with the skulls.
As in Uintatherium, the variability in the various portions of
the skull, especially in the region of the horns, is so extreme that
no two skulls are found which are exactly alike. But the denti-
tion, which is constant among the Dinocerata, here greatly com-
plicates the problems of classification. The premolars vary in
number, and the incisors, always of relatively small size and
fairly constant in number in the upper jaw, vary from three to
none in the lower jaw." In all the lower jaws found in Professor
Cope's collection of Menodontidae from northern Colorado there
are no incisors, and the mandibular symphj'sis is extremely
narrow. In the lower jaws of the Cambridge and Princeton
collections, which are all from the Nebraska and Dakota
exposures, the symphysis is broad, and the incisors, where pre-
served, are two in number, while in one of the Cambridge
specimens no less than three incisor alveoli may be counted
upon one side of the symphysis.
We might infer from this that Symborodon can be clearly
separated from Menodus by the absence of the lower incisors,
accompanied by a narrowing of the symphysis; but Professor
Cope has recently described a new species, M. angustigenis,
from the Swift Current Creek region (Cope, 1886.1, p. 81c),
which combines the narrow type of symphysis with the presence
of two incisors. The separation of these genera is rendered
still more improbable by the parallelism which exists between
the skulls from the Nebraska and Colorado localities, especially
in respect to the conformation of the nasal bones and the
horns. The genus Symborodon is, however, provisionally
adopted at present to include the species with a narrow man-
dibular symphysis and no lower incisors.
The genus Brontotherium Marsh (that is. Marsh's "Bronto-
therium" ingens, not the true Brontotherium) can not be dis-
tinguished from Menodus. It rests in part upon the premolar
I' One of the Cambridge skulls, M. coloradensis, has but a single upper incisor.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
147
formula |^, in the synopsis given by Professor Marsh (1876.1,
p. 339), as distinguished from Menodus, with ?pm J^. One
of the lower jaws of the Princeton collection, however, has the
premolar formula 57^, demonstrating that the first lower pre-
molar is a variable tooth and can not in this case be used in
classification. The same rule applies to the second cone upon
the last upper molar, the supposed generic character of Dicono-
don Marsh. This is found in different species in all degrees
of development, from a small prominence upon the basal cin-
gulum to a well-developed cone {M. proutii) .
From this evidence Osborn draws the following
partly erroneous conclusion:
Such characters as the invariable absence of lower incisors
may subsequently be found to separate one genus of the Meno-
dontidae from another; but our present evidence goes to show
that they simply characterize the extremes of a closely related
series of animals, from the same horizon, of which the inter-
mediate forms are represented by numerous species. The
safest basis of specific determination seems to be the correla-
tion between the developnaent and proportion of the horns and
of the nasals, the rule being that where the horns are long the
nasals are short, and conversely. The number of the teeth
does not at present seem to be absolutely constant, even within
the limits of the species.
The following determination of the species in the Cambridge
collection is, for the above and other obvious reasons, provi-
sional. The classification can be finally settled only when the
lower jaws and skulls are found in association.
Thus the validity of the several genera recognized
by Marsh and Cope and of the chief criteria used by
them as generic characters was called in question.
The species are treated as belonging mainly to the
single genus Menodus. Taking up the description of
the new material, the authors mistakenly refer to
Leidy's Megacerops coloradensis, a well-preserved
skull, which at present is referred to Alhps marshi.
They then describe two new species — "Menodus"
tichoceras, based on a skull, and "Menodus" platyceras,
based on a pair of bony horns. Both these species
are at present referred to the true Brontotherium or
flat-horned genus. The authors conclude their dis-
cussion of the "Menodontidae" by presenting the
first published restoration of the skeleton, made up
of material in several museums, forming a composite
animal representing Menodus proutii. In connection
with a table of measurements arranged to show pro-
gressive and correlated changes in the horns and nasals,
they make the following remarks (op. cit., p. 16):
The above measurements bring out very clearly the decrease
in the proportions of the nasals pari passu with the gradual
elongation of the horns. Another very interesting fact is
brought out by the comparison of the transverse and longitudi-
nal diameters of the horns at the base. As we pass from the
short to the long horned types, through M. coloradensis, ticho-
ceras, doUchoceras, and platyceras, there is a gradual rotation of
the longer axis of the horn section from a fore and aft to a trans-
verse plane, the species last named representing the extreme of
the transverse type.
The fuller development and more or less radical
modification of the hypotheses put forward in this
paper have been the subject of successive contribu-
tions by Osborn, culminating in the present work.
GEOLOGIC lEVEIS AND SUCCESSION OF TYPES (HATCHEE, 1886-1893)
The work of Marsh and Cope had been exclusively
descriptive and systematic. Osborn's observation of
the correlated progressive reduction of the nasals
and the enlargement and flattening of the horns
seems to have been the first important application of
evolutionary principles to tlie study of the Oligocene
titanotheres. But materials for an exact knowledge
of the phyletic succession, resting securely upon a
knowledge of the precise geologic levels of a large
series of specimens, had hitherto been entirely lacking.
This all-important element of the time relations of
the different species was largely supplied by the
labors and study of J. B. Hatcher. In 1886, 1887, and
1888 Hatcher spent 15 months in the White River
beds of South Dakota and Nebraska, collecting
material for Professor Marsh's monograph on the
Titanotheridae. In an interesting article in the
American Naturalist for March, 1893, Hatcher (1893.1,
pp. 214, 215) tells us that he collected or purchased
"nearly 200 complete skulls and many more or less
complete skeletons," a part of which are now on
exhibition in the National and Yale Museums. The
superb Hatcher collection in the United States
National Museum contains skulls and jaws of 157
individuals; as completely listed in the generic
sections of this monograph, it furnishes the classic
standard of reference. Hatcher writes:
Early in the season of 1886 it became apparent that certain
forms of skulls were characteristic of certain horizons in the
beds. This fact showed the importance of keeping, so far as
possible, an exact record of the horizon from which each skull
or skeleton was taken. From actual measurement the vertical
range of the Titanotheridae was found to be about 180 feet.
For convenience in keeping a record of horizons the beds were
divided into three divisions of 60 feet each, and each of these
three divisions was subdivided into three divisions of 20 feet
each. The difi'erent skulls and skeletons, when dug out, were
each given a separate letter or number, and this letter or num-
ber was placed in that subdivision of the beds from which the
skull or skeleton was taken.
At present about 60 of these skulls and several more or less
complete skeletons have been freed from their matrix. When
studied in connection with the horizons from which they were
taken, these remains show that a regular and systematic
development took place in these animals from the base to the
top of the beds. The most noticeable change which took place
in the Titanotheridae was a gradual and decided increase in
their size from the lowest to the uppermost beds, as is shown
by the increase in the size of the skulls, fore and hind limbs, and
other portions of the skeleton. Individuals found near the
bottom of the beds are little, if any, larger than the living
rhinoceros. From this they gradually increase in size as we go
up until at the top we find a type described by Professor Marsh
as Titanops, rivaling the modern elephant in size.
This increase in size from the base to the summit of the beds
was attended by a very marked development in certain portions
of the skeleton, noticeable among which are the following:
A variation in shape and an increase in the size and length of
the horn cores as compared with the size of the skulls, attended,
near the summit of the beds at least, by a decided shortening
of the nasals.
148
TITAJSfOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Hatcher was less fortunate in his observations on the
evolutionary changes in the dentition, stating that
"the number of incisors, though probably never
constant, even in the same species, shows a tendency
to decrease in skulls found near the summit of the
beds," and concluding that "the number of incisors
can hardly be considered as of either generic or specific
importance in the Titanotheridae, where they are no
longer functional and vary with individuals in the
same species and with age in the same individual.
The same may be said of the presence or absence of
the first premolar."
After noting certain other changes rightly believed
by him to be progressive, such as (1) the loss of the
trapezium, (2) the development of a postero-internal
cone on the third upper molar, and (3) the flattening of
the horns, Hatcher concludes his paper by giving a
tabular paleontologic section of the " Titanotlierium
beds," with a general description of the forms char-
acterizing the three ascending divisions. As to the
number of genera, he gives the impression that he
•regards all the various species ascribed by Marsh and
Cope to different genera (except Teleodus avus Marsh)
as referable to the single highly variable genus
Titanotherium Leidy.
FIRST EUEOPEAn NOTICE (TOTJIA, 1892)
The next important event is the discovery of a
titanothere of Oligocene type in Europe, described as
Menodus rumelicus by Toula (1892.1). This dis-
covery, in connection with that of the Transylvanian
BracJiydiastematJierium, described by Bockh and
Maty in 1876 (1876.1), extended the known range of
the titanotheres to the Old World.
DISTINCTIONS OF SEX (OSBOEN AND WOETMAN, 1895)
In 1895 Osborn and J. L. Wortman (1895.105)
published a corrected restoration of Titanotlierium
based upon the fine skeletal material secured by the
American Museum field parties in 1892 and 1894. They
ventured the conclusion that "it is probable that
certain wide differences in the development of the
horns, which have been assigned a generic value,
are merely sexual characters. "
MONOPHYIETIC INTEEPEETATION (OSBOEN, 1898)
The extreme development of the erroneous theor}'
that all the various species of Oligocene titanotheres
belonged to the single genus Titanotherium and were
practically monophyletic is worked out in a very
elaborate way in Osborn's paper "The cranial evolu-
tion of Titanotlierium," published in 1896. This
'was the most comprehensive review of the subject
that had hitherto appeared and was illustrated by
numerous text figures and several folded plates.
The direct observations were based chiefly on the
large collection of titanotheres in the American
Museum and to a less extent upon figures and descrip-
tions previously published. Part I, the systematic
introduction, includes a chronologic list of generic
and specific terms, with references and a brief history
of the progressive complication of the nomenclature,
after which the author says (Osborn, 1896.110, p.
162):
It is obvious that the only method of clearing up this hetero-
geneous list [of nominal genera and species] is first to establish
certain laws of cranial development, and second to apply
these laws to the distinction of genera and species in chrono-
logical order. Examined in this way, the vast array of genera
and species is resolved into one or possibly two genera and
about fourteen definable species.
Accordingly in Part II, "Principles of cranial and
dental evolution," we find a study of the differences
in size of skull, shape of horns, nasals, zygomatic
arches, auditory meatus, cingula on grinding teeth,
incisors, canines, second internal cone of last molar,
etc., aU considered as indicating either specific or
sexual or individual differences within the limits
of a single genus, Titanotlierium. This is followed by
the "Revision and definition of species," in which
some 27 species, including the new T. ramosum, are
discussed. The known species from the lower, middle,
and upper beds are arranged in a single or monophy-
letic series, beginning with the T. heloceras-trigonoceras
ingens series, continuing with torvum, rohusfum,
doliclioceras, elatum, amplum, acer, and culminating
with ramosum and platyceras.
This analysis, although wholly wrong in treating
all the species as members of a monophyletic series,
not only laid the foundation for the present evolution-
ary and phylogenetic treatment of the group but
established, as it were, the technique of investigation
POIYPHYIETIC INTEEPEETATION (OSBOEN, 1902-1919)
The reaction against the monophyletic theory was
felt by the same author as a result of more extended
research. In his paper of 1902 on "The four phyla
of Oligocene titanotheres, " after acknowledging the
services of the late Professor Marsh and admitting the
incorrectness of the monophyletic theory, Osborn
says (1902.208, p. 91):
This second review is an abstract of a portion of the results
obtained for the United States Geological Survey monograph
"The titanotheres," now in preparation. It covers practically
aU the type material in the Yale, National, American, and
Harvard Museums, and advantage has been taken of the
invaluable field observations by Hatcher of the levels on which
the different skulls in the National Museum collection were
discovered. The section method also has been very greatly
extended and, taken in connection with the teeth and the
detailed structure of the skull, has proved to be a sure criterion
of specific and phjdetic character.
Four important considerations had led Osborn to
give up the monophyletic theory: First, from his
phylogenetic studies on the rhinoceroses of Europe
and America (Osborn, 1898.143; 1900.192) he had
concluded that, contrary to earlier opinions, this
DISCOVERY OF THE TtTANOTHERES AND ORIGINAL DESCRIPTIONS
149
group was in a high degree polyphletic, embracing
many parallel phyla and having a wide adaptive
radiation; second, the principle of dolichocephaly and
brachycephaly (Osborn, 1902.207), as interpreted by
him in the rhinoceroses and other groups, raised the
presumption that similar differences would be found
to distinguish genera and phyla among the titano-
theres; third, he had learned to realize that the extent
to which parallel and convergent evolution had oper-
ated in many allied phyla had been but little appre-
ciated by earlier writers, who had largely failed also
to distinguish between persistent, progressive, and
retrogressive characters; fourth, an examination of
the titanothere skulls collected by Hatcher and now
in the National Museum, which Hatcher had recorded
exactly as to level, enabled him, with the aid of prin-
ciples just stated, to distinguish several distinct phyla
and to foUow them from the lower part through the
middle and into the very top of the " Titanotherium
beds." The characters of these phyla were summar-
ized by Osborn as follows (1902.208, pp. 92-94):
THE FOUR GENERA
Titanotherium Leidy applies to .long-limbed animals with long
skulls, persistently long and broad nasals, short triangular
horns placed slightly in front of the e3'es, vestigial incisors ^o'
large canine teeth. Known from the base to the summit of the
[lower] Oligocene.
Megacerops Leidy applies to titanotheres with broad skulls,
nasals progressively shortening, short horns rounded or oval
in section, shifting anteriorly, one or two pairs of incisor teeth,
.|r}, medium-sized canine teeth. Known from the base to the
summit of the [lower] Oligocene.
Probably related to this are the subgenera of the t3'pes named
Allops and Diploclonus by Marsh, differing from the above in
horn characters. Known chiefly from the upper beds.
Symborodon Cope includes titanotheres with skulls of varying
proportion, nasals slender and progressively shortening, horns
elongate and peculiar in being placed above the eyes instead of
shifting forward, incisors vestigial l^, canines small, approx-
imated. Known only from the middle and upper beds.
Brontotherium iVlarsh embraces the largest titanotheres, with
very broad zygomatic arches, nasals shortening while horns
elongate and shift forward; incisors persistent, f in the males,
canines stout and obtuse.
Representatives of Titanotherium and Megacerops can be now
continuously traced from the base to the summit of the [lower]
Oligocene. Primitive species of Brontotherium also appear at
the base, although the phyletio sequence through the middle
to the upper beds is not so clear. Symborodon suddenly appears
in the middle beds.
Viewed in the light of the foregoing principles, the
variations in the horns, nasals, incisors, cingula, etc.,
took on new meanings — biologic, phylogenetic, and
systematic; so that, after more than half a centm-y of
research (1846-1902) the systematic problem presented
by the Oligocene titanotheres appeared in its main
features to be solved. Subsequent research, however,
has led to certain regrettable but apparently necessary
changes in nomenclature: (a) The name " Megacerops"
Leidy, as defined above, has been set aside for Brontops
IVIarsh, for the reasons given below; (b) the name
Titanotherium Leidy has been abandoned for the prior
name Menodus Pomel; (c) the name Symhorodon Cope
has been replaced by the prior name Megacerops
Leidy.
RECENT DISCOVERIES BY LULI, lAMBE, AND OTHERS
There remain to be recorded the following contri-
butions: (1) The description of Megacerops tyleri by
Lull (1905.1), based upon a fine skull and lower jaws
with associated limbs, discovered by the Amherst
CoUege paleontologic expedition of 1903; (2) the
description of Brontotherium hatcheri and Symhorodon
copei by Osborn in 1908 (1908.318), based on skulls
in the National ]Museum; (3) the description of Mega-
cerops primitivus and M. assinihoiensis, based on frag-
ments obtained from Saskatchewan, Canada, by
Lambe in 1908 (1908.1); (4) W. K. Gregory observed
(a) that there is an alliance between Brontops, Allops,
and Menodus as these terms are now used by Osborn,
indicated by certain intermediate forms between the
extremely brachycephalic Brontops roiustus and the
dolichocephalic Menodus giganteus, (b) that there is
also an alliance between Brontotherium and Megacerops
(Symhorodon) in spite of the differences in the incisors.
Hence the former group — Brontops, Allops, Menodus — ■
has been called the menodontine group, and the latter
group — Brontotherium, Megacerops — has been called
the brontotheriine group.
Possibly the most valuable general result of the
study of the titanotheres has been the fact that it
has made possible the close examination of an extensive
evolutionary history, stretching from the lower Eocene
to the summit of the lower Oligocene. IVTany observa-
tions have been made on the precise modes of evolu-
tion, especially with regard to the way in which char-
acters first appear and subsequently develop. The
results of this evolutionary study are set forth in
Chapters V, VI, VH, and XI of the present work.
THE EOCENE TITANOTHERES
PIONEER DISCOVERIES
WORK IN THE BRIDGER, WASHAKIE, AND UINTA BASINS BY lEIDY,
MARSH, COPE, SCOTT, OSBORN, AND OTHERS (1870-1889)
Prof. F. V. Hayden, in the course of his historic
explorations in the fossiliferous beds of the Kocky
IMountains and Great Plains, obtained at Church
Buttes, near Fort Bridger, Wyo., a number of isolated
teeth, which were described by Leidy (1870.2) under
the name Palaeosyops paludosus. This was the first
Eocene titanothere made laiown to science, 24 years
after the discovery of Prout's "gigantic Palaeotherium"
(Titanotherium) in South Dakota. Although Leidy
noted that the lower molar of Palaeosyops "resembles
in its constitution those of Palaeotherium, (Jlialicothe-
rium, and Titanotherium," he did not classify the new
genus with the titanotheres, for the reason that at
that time he thought Titanotherium and Chalicotherium
were allied to the Artiodactyla. (See p. 247.) Soon
150
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
afterward Leidy (1873.1, p. 27) described a skull, some
teeth, and parts of the limb bones of Palaeosyops and,
noting the similarities of this species to its supposed
allies Tapirus and Palaeoiherium, correctly referred it
to the perissodactyls — the odd-toed pachyderms.
Three other species (P. major, P. Tiumilis, P. Junius)
were also described by him from the Bridger beds
upon very fragmentary material.
In developing our knowledge of the Eocene titano-
theres of the Bridger Basin, as in developing that of
the Oligocene titanotheres. Marsh and Cope were not
far behind Leidy. The first specimen of an Eocene
titanothere described by Marsh, however (1871.2),
was not recognized as such by him, as he mistook the
isolated second lower premolar of a Bridger Palaeosyops
for the fourth upper premolar of a dog and named it
"Canis montanus." The next year (1872.1) he de-
scribed some well-preserved remains under the name
Palaeosyops laticeps and also founded the genus
Telmatherium. Marsh's subsequent contributions to
our knowledge of middle Eocene titanotheres were not
especially significant, but in 1875 he described the
very important genus Diplacodon from the upper
Eocene Uinta beds of Utah and recognized its inter-
mediate position both in time and in structural
characters between his "Limnohyidae" (Palaeosyo-
pinae) and Brontotherium.
Cope's explorations of the Bitter Creek or Wash-
akie Basin (middle Eocene) of Wyoming m 1872 led
to his describing the species "Palaeosyops" vallidens
and "Limnoliyus" laevidens, both represented by
imperfect remains. The former is now known to
belong to the long-headed genus Dolichorfiinus.
The next year, 1873, Cope (1873.5) described the
species Limnoliyus ( = Palaeosyops) fonfinalis from the
lower levels of the Bridger formation (supposed
Bridger A), which is the oldest middle Eocene titano-
there yet discovered. His Palaeosyops diaconus,
from the upper levels of the Bridger Basin, is prob-
ably a synonym of Palaeosyops rohustus (Marsh).
DISCOVERY IN HUNGAEY
Shortly after these pioneer discoveries in America
Bockh and Maty (1876.1) described a large lower
jaw from Eocene deposits in Transylvania, in Hun-
gary. The animal was supposed to be allied to
Palaeoiherium and was named Brachydiastematherium
transilvanicum. Its affinities with the Palaeosyops
group long remained unnoted, and even to this day
it is the only known specimen of its kind in Europe.
PEINCETOM AND COPE-WOETMAN EXPEDITIONS
The Princeton expeditions sent to the Bridger and
Washakie Basins in 1877 and 1878 under Scott,
Osborn, and Speir brought to light much valuable
material of Palaeosyops and allied genera, especially
the types of "Leurocephalus" cultridens and the pecu-
liar form which was later described by Earle as
Palaeosyops megarMnus. Thus by the end of 1878 re-
mains of the genus Palaeosyops and its allies had
been discovered in the middle Eocene Bridger and
Washakie Basins and in the upper Eocene Uinta
Basin.
The next year (1879) Dr. J. L. Wortman, who was
collecting for Cope, extended the kiiown range of the
group into the lower Eocene Wind River formation of
Wyoming, where he discovered the very primitive
form which Cope in 1880 named Palaeosyops horealis
and which is now recognized as approximately ances-
tral to the middle Eocene titanotheres. Wortman
also discovered a very small form, which was described
by Cope in 1880 (1880.1) as Lamhdotherium popo-
agicum and recognized as more or less closely allied to
the Palaeosyops group.
The next important expedition was that made by a
Princeton party under Scott and Speir in 1886 into
the Uinta Basin (upper Eocene). They collected
skeletal material, referred at that time to Diplacodon,
which was described by Osborn in 1890 (1890.51) and
which demonstrated the intermediate characters of
"Diplacodon" {Protitanoiherium) between the Oligo-
cene and middle Eocene titanotheres. In the same
publication Osborn also described "Palaeosyops"
hyognathus, a species based on a jaw that is now known
to represent the long-skulled genus DolichorMnus.
FIRST SYSTEMATIC AND EVOLUTIONARY REVISION
(EARLE, 1889-1891)
Although Cope in 1884 (1885.1) had republished
and partly extended the original descriptions of his
own species, with lithographic figures, no satisfactory
revision of the Palaeosyops group was possible at that
time or for many years later.
In 1889 Charles Earle, at the invitation of Prof.
H. F. Osborn, began a careful study of the material
in the Princeton Museum and other collections, and
in 1892 he published a memoir "On the genus Palaeo-
syops Leidy and its allies" (1892.1). Earle gave a
very detailed description of the osteology of Palaeo-
syops and of the first attempted reconstruction of the
skeleton of an Eocene titanothere by Osborn. (See
fig. 86.) Owing in part to the lack of sufficient well-
associated material, in part to the confusing practice
of the earlier writers in designating and founding
species upon several specimens of doubtful specific
association, Earle's revision of the species and genera
was, as he himself recognized, by no means final. He
rightly regarded as distinct the genera Lamhdotherium,
Limnohyops, Palaeosyops, and "Telmatotherium," but
as he showed in his tentative phylogenetic scheme,
he, like other paleontologists at that time, did not
appreciate the polyphyletic character of groups and
consequently referred to a single main line of descent
a number of forms that belong to widely separated
phyla.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
151
AMERICAN MUSEUM AND OTHER EXPLORATIONS OF THE
EOCENE BASINS (1891-1895)
The problems relating to Palaeosyops and its
allies, which had been barely made evident by the
pioneer discoveries and had now been partly formu-
lated by Earle, were of course only particular results
of the general explorations of the fossil-bearing forma-
tions of the West. The early explorations had been in
part reconnaissances, and their results were accordingly
incomplete as regards both the nature of the material
and the records of the stratigraphic levels at which the
specimens were found, both absolutely prerequisite to
a detailed knowledge of the phylogeny.
exhibit a mounted composite skeleton of this animal.
Much other material was also collected by the same
party. All this material has been used profitably in
the present monograph, especially the specimens
representing the "prophet-horn stage," to which
Doctor Wortman in a letter from the field applied the
name Manteoceras.
Another American Museum expedition, under Mr.
O. A. Peterson, went into the Uinta Basin in 1894
and examined two hitherto unexplored horizons (Uinta
B 2 and Uinta B 1 of this monograph), which underlie
the true Uinta (Uinta C). This expedition collected
many new forms and worked out the faunal sequence
of the three horizons indicated. Among the results
Figure 86. — Osborn's first restoration of Palaeosyops paludosus Leidy
This restoration is a composite one— the skull from the fine specimens in the Academy of Natural Sciences of Philadelphia, and the axial
skeleton from the material in the Princeton Museum,
twelfth natural size.
The fore feet were afterward referred to Mesaiirhinus peterst
The founding (in 1890) of the department of verte-
brate paleontology in the American Museum of
Natural History by Prof. Henry Fairfield Osborn and
the consequent establishment of continuous and syste-
matic exploration began a new era of exact investiga-
tion not only of the titanotheres but of the whole
series of vertebrate remains to be found in the Rocky
Mountains and Great Plains regions, as well as the
stratigraphic horizons at which they occur.
The first of these expeditions, led by Dr. J. L.
Wortman, procured some important skeletal material
of "Palaeosyops" horealis from the Wind River forma-
tion. Another expedition, sent out under Doctor
Wortman in 1893, procured from the Bridger and
Washakie Basins extensive material of the true
Palaeosyops, enabling the American Museum to
of this expedition, as reported in 1895, were the discus-
sion by Osborn and Peterson (Osborn, 1895.98) of the
three faunal levels (Uinta B 1, B 2, and C) and the
description by Osborn of the specialized and interest-
ing titanotheres named " TelmatotJierium" diploconum
and T. cornutum. Wortman's "prophet-horn" skulls
were referred to " Telmatotherium vallidens," so that
animals showing a wide range of form were here
erroneously included under a single genus. The
very aberrant form SpJienocoelus was also described,
but its ordinal and family positions were left "Incertae
sedis," on account of the lack of the teeth in the type
and the peculiar characters of the base of the skull.
In the same year (1894) Mr. J. B. Hatcher, of the
Princeton Museum, also went into the true Uinta
area and discovered specimens representing the very
152
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
advanced stage which in 1895 (1895.1) he described
as Diplacodon ejnarginatus. In a brief postscript
to this description he noted the wide range of forms
that had been erroneously grouped by Osborn under
the genus " TelmatotTierium," and he formally proposed
the generic names Manteoceras for the "prophet-
horn" and Dolichorhinus for the long-skulled form.
Figure 87. — Four stages in the origin and evolution of tlie
horns in titanotheres
After Hatcher's original plate (1895). A, Palaeosyops laiiceps (= Limnohyops lali-
ceps), after Earle; B, Telmatotherium vallidens ( = Manteoceras manteoceras), after Os-
born; C, Diplacodon emarginatus (,= PTOtitanotherium emarginatum), after Hatcher;
D, Titanotherium varians (= Bronioihermin leidyi), after Marsh. One-eighth
natural size.
Both these terms, as well as the name Protitanotherium
of Hatcher, have proved to be valid. Hatcher's separa-
tion of these genera was a very important move toward
a correct understanding of their phylogenetic rela-
tions. He also figured a series of four stages ranging
from the middle Eocene to the lower Oligocene, inclu-
sive, showing the origin of the "horns." (See fig. 87.)
INVESTIGATIONS AND EXPLORATIONS MADE IN PREPARA-
TION FOR THE PRESENT MONOGRAPH (1900-1919)
Between 1895 and 1900 no very important work on
the Eocene titanotheres was done. By the end of the
nineteenth century some 12 nominal genera and 25
nominal species of the titanotheres had been proposed,
but many of the real generic limits and phyloge-
netic relations were still obscure except for the preg-
nant suggestions of Hatcher. In 1900 Professor
Osborn, at the invitation of Director Charles D.
Walcott, undertook to revise and monograph the
Eocene titanotheres in connection with the United
States Geological Survey monograph on the Oligocene
titanotheres that had been begun by Professor Marsh.
The work on the Eocene titanotheres has proved to be
by far the most difficult and most extensive part of
this task. During the last 28 years Professor Osborn,
with the assistance of Dr. W. K. Gregory, has studied
the great and growing collection in the American
Museum of Natural History and in other institutions
and has set forth the results in several prelimmary
articles and more fully in the present work.
A long series of parties of exploration, beginnLng in
1903, sent out from the American Museum by Osborn
(1909.321) and conducted chiefly by Mr. Walter
Granger, have carefully examined the various lower,
middle, and upper Eocene basins of the West with
special reference to the exact succession of species.
This very precise work has shown that the Bridger
and other formations are divided into a succession of
zonal levels characterized by the remains of titano-
theres and other mammals in different generic and
specific stages of evolution. The stratigraphic rela-
tions of the Eocene to the Oligocene deposits have also
been in part explored. The results are fully set forth
in this monograph. Although this work in the
Eocene basins has been carried on chiefly by the Ameri-
can Museum of Natural History, the Carnegie and
Field Museums have sent expeditions into the Uinta
Basin under Douglass (1909.1) and under Riggs
(1912.1), which have yielded similar results as to
specific and generic succession.
The distinction of numerous independent Eocene
phyla by Osborn has followed the discovery of the
Oligocene phyla, some of which arise from those of the
Eocene.
Thus have been established secure bases of fact,
first, for a general history of the early Tertiary faunas
of the West; second, for a demonstration of the evolu-
tion of certain phyla of titanotheres through long
periods of time; and, third, for a consideration of the
modes and factors of evolution of titanotheres in par-
ticular and of mammals in general.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
153
SECTION 2. ORIGINAL DESCRIPTIONS OF EOCENE
TITANOTHERES
FIVE RULES FOE DETERMINING THE NAMES OF
TITANOTHERES
The systematic revision of the Eocene and Oligo-
cene titanotheres was begun in 1900 by the author
with the cooperation of W. K. Gregory and has been
continued up to the day of the deUvery of the manu-
script of this volume to the Geological Survey. The
great difficulties and the labor involved in determin-
ing the correct prior names for the genera and species
have been due to the imperfection of the fossil types,
to loose methods of description and comparison, and
to the mingling as cotypes of animals belonging to dif-
ferent species or even to different genera.
Experience has shown that the following five rules
are absolutely necessary for future vertebrate paleon-
tologic work.
Rule 1. Accept the "law of priority," as defined by the
International Committee on Nomenclature.
In this revision the author has accepted as authori-
tative the rules of nomenclature based upon the "law
of priority," as defined by the "Code" of the Ameri-
can Ornithologists' Union and by the recommenda-
tions of the committee on nomenclature of the Twelfth
International Congress of Zoology. Special acknowl-
egments a,re due to the eminent authority Dr. J. A.
Allen for frequent aid in deciding troublesome prob-
lems of nomenclature.
Rule 2. Determine the geologic level and characters of the
type, as the starting point of monographic inquiry.
Experience teaches that the characters of the holo-
type specimen and the geologic level on which it was
found afford the permanent facts to which all questions
of nomenclature must be referred as the basis of mono-
graphic investigation.
Rule 3. Avoid confusion of characters of th& type and cotype
or paratype.
All the early systematic work on the titanotheres
was done without regard to precise discrimination
between the certain or permanent nomenclatural
value of the holotype specimen and the uncertain
value of "specific" characters based on cotype, para-
type, and neotype specimens.
For example, take the case of the classic species
Palaeosyops paludosus Leidy. Leidy used as types
the very fragmentary teeth from the lower levels of
Bridger B, which first came into his hands; he later
erroneously associated with these fragments, practi-
cally as cotypes, other more complete specimens, which
are now known to belong to two or three different
species from higher geologic levels. Subsequently
Leidy himself. Cope, Marsh, Scott, Osborn, and Earle
all accepted Leidy's erroneous associations, and P.
paludosus came to be known by certain of its falsely
associated cotype and paratype characters instead of
by its true type characters.
101959— 29— VOL 1 13
Thus the entire nomenclature of the subject became
a mass of confusion, and the difficulties encountered in
clearing it up have been almost insuperable.
The rule is that specific definitions must be based on
holotypes only, unless there is absolutely no possibility
of doubt that the associated types are from the same
geologic level and belong to the same species.
Rule 4. Distinguish the different values and kinds of types.
The use of the terms type (or holotype), cotype,
paratype, lectotype, hypotype, neotype has been dis-
cussed critically by Oldfield Thomas (1893.1, p. 241),
by Schuchert (1905.1, pp. 9-14), and by Osborn
(1918.473). The distinctions indicated below should
be noted.
Type, individual, or holotype. — -A holotype is a
particular individual specimen "deliberately selected
by the author of a species; or it may be the only
example of a species known at the time of original
publication. A holotype, therefore, is always a single
individual but may embrace one or more parts, as the
skin, skeleton, or other portions." (Schuchert, op.
cit.) The holotype must usually be determined from
the original description.
Cotype, coordinate or equivalent type. — The term co-
type is applied to specimens when an author's type
description refers to remains of two or more individuals
without selecting or distinguishing one as the holotype,
so that all appear to be equally identified with the
specific name given.
Lectotype. — "Where the origina' diagnosis is with-
out illustrations or is accompanied by figures based on
two or more specimens, the first subsequent author is
at liberty to select from these cotypes a type for the
old species, adhering, so far as can be ascertained, to
the intention of the original author. Such a type
specimen is to be designated a lectotype ( = a chosen
type)." (Schuchert, idem.) The practice of Osborn
as to lectotypes in paleontology is either (a) to select
the first individual specimen named by the original
author, because the second individual specimen may
belong to a distinct species, or (&) to select the speci-
men to which the specific name obviously refers —
for example, Cope's Menodus angustigenis.
Hypotype and plesiotype. — As shown by Schuchert
(idem), the terms hypotype and plesiotype have been
used in two different senses to cover "supplementary
types." They may well be dropped.
Neotype. — A neotype is defined by Schuchert (idem)
as a [new] "supplementary type selected by an [a sub-
sequent] author, on which a species is to rest because
of the loss of the original type, or where the original
material still extant is so poor or fragmentary that from
it the characters of the species can not be determined
with certainty." Great care must be taken that the
neotype comes from the same geologic level as the
type.
154
TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
Summary. — The usage adopted in this monograph
is as follows:
Holotype (of original autlior) : The original individual type
specimen selected by the author.
Cotypes: Different individual specimens rightly or wrongly put
together by the author as "types."
Paratype (of original author) : Additional individual specimen
or specimens noted by the author in the original description
and used by him in defining the species.
Lectotype (of subsequent author) : The specimen selected by
a subsequent author, from among the "cotypes," for pur-
poses of subsequent description or redefinition. This may
be (a) the specimen first mentioned by the author, or (6) the
specimen to which the specific name obviously applies.
Neotype (of second or subsequent author) : A new specimen
selected in a subsequent description because of the loss or im-
perfection of the holotype or type.
These five terms are all that are necessary in verte-
brate paleontology. The terms plesiotype and hypo-
type are discarded in this monograph because they are
too indefinite.
Monographic revision in the use of above terms. —
Leidy founded the species Palaeosyops paludosus
upon some isolated teeth from the low levels of
Church Buttes. In the original description these
teeth, which probably represent more than one in-
dividual, were treated as coordinate or equivalent
types or "cotypes." Out of this lot the second
lower molar (m2), which was the first specimen men-
tioned and described by Leidy, has been selected by
Osborn in the present volume as the final standard,
or "lectotype," of the species.
In the same original description by Leidy of P.
paludosus a second lot of teeth, from the high levels
of Henrys Fork, were mentioned, and the characters,
of these teeth entered into Leidy's original conception
of the species. These teeth are now called "paratypes."
In the present revision, since there is little doubt
that Leidy's paratypes are not really conspecific with
the specimen first mentioned (lectotype), Osborn
has selected from the same geologic level. Church
Buttes, a lower jaw in which m2 agrees most clearly
with the lectotype m2 and which is to serve as a
secondary type, or "neotype."
It wUl be seen that cotypes, paratypes, or neotypes
may sometimes be wrongly associated specifically
with the holotype, in which case the specific name
must cling to the holotype and lectotype as the
ultimate standard means of identification.
The first step toward permanence, therefore, is the
settlement of the holotype characters, which is some-
times an almost impossible task, owing to the poor
quality of the holotype selected — for example, the
holotype of Palaeosyops major Leidy, a jaw fragment
without teeth; the holotype of P. humilis Leidy, a
single deciduous premolar.
Rule 5. Avoid mingling as types and cotypes specimens from
different geologic levels.
The mingling of types and cotypes from different
geologic levels has been the second chief source of
confusion. To cite a prominent instance. Cope's
cotypes of Palaeosyops laevidens were two skulls col-
lected at widely separated localities, and in his original
description no regard was shown for their possible
difl'erence of geologic age. It appears almost certain
that the lectotype belongs to a lower level and is
perhaps some thousands of years more ancient than
the paratype. Similarly we have shown that the
lectotype of Leidy's P. paludosus is from Bridger
level B 1 or B 2; the paratypes are from level C 2 or
C 3, a difference of geologic level representing a very
long period of time, in which it is now certain that a
very marked progressive evolution took place in teeth,
skull, and skeleton.
Our geologic leveling of the Bridger formation,
described in Chapter II, has therefore not only
afforded us the means of determining the evolutionary
succession of the species of titanotheres but, if the
localities of the types were properly recorded by the
authors, it has enabled us to separate many er-
roneously associated type specimens. The geologic
levels of the materials recently acquired by the
American Museum have been ascertained precisely;
on the whole, the successive species correspond very
closely with the successive levels — that is, in no case
have different species in the same line of descent been
found at the same level, although species in different
lines of descent (that is, in different genera) are found
in analogous stages of evolution.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
THE GENERA AND SPECIES OF EOCENE TITANOTHERES
155
The accompanying list shows, in chronologic order,
the names assigned to Eocene titanotheres. The
numbers in the first column indicate the chronologic
order or rank of the systematic names, the roman
numerals indicating generic names, the arable numerals
specific names.
Chronologic list of original descriptions of Eocene titanotheres
[Generic names accepted in tliis work as valid are printed In small capitals; abandoned names are inclosed in brackets]
I
1
2
3
4
6
6
II
7
Ilia
8
9
Illb
10
11
12
IV
13
V
14
VI
15
16
VII
17
18
19
VIII
20
21
22
23
24
IX
25
X,
26
XI
XII
27
28
29
30
31
XIII
32
33
34
35
36
1870
1870
1871
1871
1872
1872
1872
1872
1872
1872
1872
1872
1872
1873
1873
1873
1875
1875
1876
1876
1878
1878
1880
1880
1880
1881
1889
1890
1891
1891
1892
1895
1895
1895
1895
1895
1895
1895
1895
1897
1899
1899
1899
1902
1907
1907
1908
1908
1908
1908
Palaeosyops
Palaeosy ops
Palaeosyops
[Canis]
Palaeosyops
[Palaeosyops]
[Palaeosyops]
Telmatherium
Telmatheriu m
[Limnohyus]
[Limnohyus]
[Palaeosyops]
[Limnohyus]
[Limnohyus]
[Limnohyus]
Palaeosyops
DiPLACODON
Diplaoodon
Bkachydiastbmathebium
Brachydiastematherium.
[Leurocephalus]_
[Leurocephalu s] .
paludosus..
major
[montanus]_
[hu mills]
Junius
laticeps
validus.
robustus.
vaUidens _
laevidens-.
fontinalis. .
[diaconus].
elatus_
transilvanioum.
cultridens.
[Palaeosyops] borealis
Lambdotherium
Lambdotherium
[Lambdotherium]
[Palaeosyops]
LiMNOHYOPS
[Palaeosyops]
[Palaeosyops]
Palaeosyops longirostris
[Telmatotherium] diplooonum_
[Telmatotherium] [cornutum]_
Sphenocoelus
Sphenoooelus uintensis
PrOTITANOTHERIUM ':
[Diplaoodon] emarginatus
megarhinus.
[minor]
Manteoceras
DoLICHORHINtrS- -
[Palaeosyops]
[Palaeosyops]
[Telmatotherium],
[Canis?]
Manteoceras
EOTITANOPS
Lambdotheriu m_
Limnohyops
Limnohy ops
Limnohyops
Palaeosyops
Present determination
Leidy.-
do.
do.
Marsh.
Leidy. .
do.
Marsh.
do.
do.
do
do
Cope .
Leidy (not Marsh).
Cope
do
do
Marsh
do
Bockh and Maty.
do
Osborn, Scott, and
Spelr.
do
Cope.
.do.
popoagicum do__
brownianum do_.
hyognathus Osborn..
Marsh. -
Earle
do._
do_.
Osborn..
do_-
do..
do.-
Hatcher.
do..
ultimus
manteoceras
[diploconum var.
minus.
[marshii]
manteoceras
pnmaevum.
priscus
matthewi
monoconus.
.! leidyi.
do...
do...
Matthew.
do...
de-
Hay. __
do.
Osborn.
Loomis.
Osborn.
do.
.do.
.do.
Palaeosyops Leidy.
Palaeosyops paludosus Leidy.
Palaeosyops major Leidy.
Palaeosyops major? Leidy.
Palaeosyops sp.
Mesatirhinus Junius (Leidy).
Limnohyops laticeps (Marsh).
Telmatherium Marsh.
Telmatherium validum Marsh.
Palaeosyops Leidy.
Palaeosyops robustus (Marsh).
Dolichorhinus vallidens (Cope).
(Preoccupied.)
Limnohyops laevidens (Cope).
?Palaeosyops fontinalis (Cope).
Palaeosyops robustus (Marsh).
Diplaoodon Marsh.
Diplaoodon elatus Marsh.
Brachydiastematherium Bockh and
Maty.
Brachydiastematherium transilvani-
cum Bockh and Maty.
Telmatherium Marsh.
Telmatherium cultridens (Osborn,
Scott, and Speir).
Eotitanops borealis (Cope).
Lambdotherium Cope.
Lambdotherium popoagicum Cope.
Eotitanops brownianus (Cope).
Dolichorhinus hyognathus (Osborn) .
Limnohyops Marsh.
Mesatirhinus megarhinus (Earle).
Palaeosyops paludosus Leidy.
Palaeosyops longirostris Earle.
Rhadinorhinus diploconus (Osborn).
Dolichorhinus hyognathus (Osborn) .
Sphenocoelus Osborn.
Sphenocoelus uintensis Osborn.
Protitanotherium Hatcher.
Protitanotherium emarginatum
Hatcher.
Manteoceras Hatcher.
Dolichorhinus Hatcher.
Telmatherium ultimum Osborn.
Manteoceras manteoceras Hay.
Metarhinus fluviatilis Osborn.
Palaeosyops major? Leidy.
Manteoceras manteoceras Hay.
Eotitanops Osborn.
Lambdotherium primaevum Loomis.
Limnohyops priscus Osborn.
Limnohyops matthewi Osborn.
Limnohyops monoconus Osborn.
Palaeosyops leidyi Osborn.
156
TITAJSrOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
Chronologic list oj original descriptions oj Eocene titanotheres — Continued
[Generic names accepted in this work as valid are printed in small capitals; abandoned names are inclosed in brackets]
Bank Date
37
38
39
XIV
40
• XV
41
42
43
44
45
46
47
48
49
50
XVI
51
52
53
54
XVII
55
56
57
58
59
60
XVIII
61
XIX
62
XX
63
64
65
XXI
66
1908
1908
1908
1908
1908
1908
1908
1908
1908
1908
1908
1908
1909
1909
1909
1909
1912
1912
1912
1912
1912
1912
1912
1913
1913
1913
1913
1913
1914
1914
1914
1914
1914
1916
1919
1919
1919
1919
grangeri
copei
washakiensis.
petersoni.
fluviatilis
earlei
intermedius.
ultimum
altidens
superbum_ _
incisivum
uintensis
heterodon__
longiceps
Palaeosyops
Palaeosy ops
Manteoceras
Mesatirhinus
Mesatirhinus
Metarhinus
Metarhinus
Metarhinus
Dolichorhinu s
Telmatheriu m
Telmatherium?
Protitanotheriu m
[Telmatherium?]
Manteoceras
Dolichorhinus
DoUohorhinus
Sthenodectes
[Mesatirhinus]
Metarhinus
Metarhinus
Dolichorhinus
Rh ADINORHINUS
Rhadinorhinus
Eotitanops
Eotitanops
Eotitanops
Lambdotherium
Lambdotherium ' progressum.
[ Diploceras]
[Diplooeras] orborni
[Heterotitanops]
[Heterotitanops] I parvus
EOTITANOTHERIUM I
[Telmatherium?] j birmanicum
Lambdotherium
Eotitanops
EOMETARHINUS
Eometarhinus
superior..
riparius
cristatus__
fluminalis-
abbotti.-
gregoryi-
princeps.
major
prisoum.
magnum,
minimus.
huerfanensis _
Osborn.
do-
.do.
.do.
.do.
.do_
.do_
.do.
.do_
.do.
.do.
do...
Douglass.
do.._
do._
do..
Gregory.
Riggs...
do._
.do_
.do.
.do.
do.
Osborn.
do.
-do.
.do-
.do.
Peterson.
do._
do__
.do_
.do.
Pilgrim and Cotter .
Osborn
do
.do_
.do_
Present determination
Palaeosyops grangeri Osborn.
Palaeosyops copei Osborn.
Manteoceras washakiensis Osborn.
Mesatirhinus Osborn.
Mesatirhinus petersoni Osborn.
Metarhinus Osborn.
Metarhinus fluviatilis Osborn.
Metarhinus earlei Osborn.
Dolichorhinus intermedius Osborn.
Telmatherium ultimum Osborn.
Telmatherium altidens Osborn.
Protitanotherium superbum Osborn.
Sthenodectes incisivus (Douglass).
Manteoceras uintensis Douglass.
Dolichorhinus heterodon Douglass.
Dolichorhinus longiceps Douglass.
Sthenodectes Gregory.
DoUohorhinus superior (Riggs).
Metarhinus riparius Riggs.
Metarhinus cristatus Riggs.
Dolichorhinus fluminalis Riggs.
Rhadinorhinus Riggs.
Rhadinorhinus abbotti Riggs.
Eotitanops gregoryi Osborn.
Eotitanops princeps Osborn.
Eotitanops major Osborn.
Lambdotherium priscum Osborn.
Lambdotherium progressum Osborn.
Eotitanotherium Peterson.
Eotitanotherium osborni Peterson.
? Metarhinus.
?Metarhinus sp.
Eotitanotherium Peterson.
Uncertain.
Lambdotherium magnum Osborn.
Eotitanops minimus Osborn.
Eometarhinus Osborn.
Eometarhinus huerfanensis Osborn.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
157
OEIGINAI DESCRIPTIONS OF THE SPECIES
Palaeosyops Leidy, 1870
Cf. Palaeosyops, this monograph, page 312
Original reference. — Acad. Nat. Sci. Philadelphia
Proc, 1870, p. 113 (Leidy, 1870.2).
Type species. — Palaeosyops paludosus Leidy. (See
p. 319.)
Generic characters. — Leidy, in his description of the
fragmentary type, very properly refrained from at-
tempting to distinguish generic from specific charac-
ters. Generic characters are given below.
Etymology. — TraXaio?, ancient; am, boar; ih\p, face
(appearance). The name was probably suggested by
the fact that the "upper true molars exhibit the outer
part of a crown composed of a pair of lobes, exactly
as in Hyopotamus." (Leidy.)
Present determination. — The generic name is a valid
one.
Palaeosyops paludosus Leidy, 1870
Cf. Palaeosyops paludosus, this monograph, page 319
Original reference. — Acad. Nat. Sci. Philadelphia
Proc, 1870, p. 113 (Leidy, 1870.2).
Subsequent reference. — Leidy, Extinct vertebrate
fauna of the Western Territories, p. 28, pi. 23, figs. 3-6
(fig. 5 lectotype), 1873 (Leidy, 1873.1).
Type locality and geologic horizon. — Church Buttes,
near Fort Bridger, Bridger Basin, Wyo.; Palaeosyops
paludosus-Orohippus zone (Bridger B 1 or Bridger
B2).
Leidy's cotypes. — M2, p*, m^, m^ (Nat. Mus. 759,
758, 762). (Extinct vertebrate fauna, p. 28, 1873.)
"The species Palaeosyops paludosus * * * -nras
founded on a number of isolated teeth * * * qI^.
tained by Professor Hayden at Church Buttes, Wyo."
(Leidy.) (See fig. 88.)
Characters. — Leidy (1870.1, p. 113) writes:
The crown of a lower true molar [mj of the right side, the
Osborn lectotype (fig. 88)] resembles in its constitution those of
Palaeotherium, Chalicotherium, and Titanotherium, being com-
posed of a pair of fore and aft conjoined pyramidal lobes with
crescentic summits. It measures 16 lines anteroposteriorly and
10 lines transversely. Fragments of upper true molars [m^ left,
m' right] exhibit the outer part of the crown composed of a pair
of lobes exactly as in Hyopotamus. The inner portion of the
crown is composed of a pair of simple cones, broad and low,
the front one considerably larger than the back one. One of
the specimens in the entire condition of the crown measured
about 22 lines fore and aft and 18 lines transversely. The crown
of an upper premolar [p*] has its outer part composed of a pair
of conjoined cones with acute summits and sides. The inner
portion of the crown [p* of the opposite side] consists of a single
broad, simple cone embraced in front and behind by a basal
ridge. The anteroposterior diameter of the crown externally
measures 9}/2 lines; the transverse diameter is an inch.
Leidy's cotypes. — The first lot of specimens from
Church Buttes (Bridger B 1), upon which the species
was originally established, consist of a second lower
molar (ma, Nat. Mus. 759; see Leidy, 1870.1, p. 113;
1873.1, pi. 23, fig. 5); "of an upper fourth premolar
nearly unworn" (p*, Nat. IVIus. 762; see Leidy, 1873.1,
pi. 5, fig. 5); of the anterior half of a second upper
molar (m^ Nat. IVIus. 758; see Leidy; 1873.1, pi, 23,
fig. 6); and of the inner side of a premolar (p*) of the
opposite side. This lot constitutes the cotypes, which
are here refigured. Of these, the second lower molar
agrees with the specimens described in this monograph
as P. paludosus. The upper teeth do not certainly
belong to the same animal; it appears necessary,
therefore, to base the genus and species on the first
specimen described in the original description, namely,
the second lower molar, which may be taken as the
lectotype.
Leidy's paratypes. — Specimens of a second lot,
from Henrys Fork, belonging to a much older individ-
ual, were treated practically as paratypes of this species
in the original notice; they were described in Leidy's
memoir of 1873 (1873.1, pp. 29, last line, and 30),
were figured in Plate 5, Figures 4, 6, 7, 8, 9, and are
Figure 88. — Leidy's cotypes of Palaeosyops paludosus
Specimens upon which the species was originally established. Hayden's collection
of 1870. After Leidy, 1873; Nat. Mus. 758, 759, 762. Natural size. The second
lower molar (Nat. Mus. 759) is the lectotype.
preserved in the United States National IVTuseum.
These are the specimens that Cope, IVEarsh, Osborn,
Earle, and others may have taken for the types, but
they are from a higher geologic level and may pertain
to P. major or P. leidyi. A third lot of specimens,
from Grizzly Buttes, included the "facial portion of a
skull containing nearly all the molars and the canines
of both sides." This specimen was treated virtually
as a paratype by Leidy (1873.1, pp. 30-34, pi. 18,
fig. 51, and pi. 4, fig. 3) and was described at length
by him. It is probably but not certainly conspecific
with the lectotype m2.
Osiorn's neotype. — The determination of P. palu-
dosus therefore rests positively on the second lower
molar alone. To supplement this lectotype the
present author has selected as a neotype a lower jaw
(PI. LVI, B; fig. 268, C) with dentition, Am. IVIus. 1 1680,
in which m2 agrees closely with the lectotype and with
the measurements given by Leidy for P. paludosus
(1873.1, p. 57 and pi. 5, figs. 10, 11). The locality
158
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
(Millersville) is about 10 miles distant from that of
the holotype (Church Buttes), but the geologic level
is believed to be identical, namely, Bridger B 1 .
Etymology. — paludosus, marshy, dwelling in the
marshes, probably because the remains were found in
one of the supposed "ancient lake basins."
Present determination. — Palaeosyops paludosus Leidy
is a valid species, but the "P. paludosus" of other
authors refers to related species of more recent geologic
age (P. major, P. leidyi, P. rolustus).
Palaeosyops major Leidy, 1871
Cf. Palaeosyops major, this monograph, page 321
Original reference. — Acad. Nat. Sci. Philadelphia
Proc, 1871, p. 229 (Leidy, 1871.1).
Subsequent reference. — Leidy, Extract vertebrate
faima of the Western Territories, p. 45, pi. 20, fig.
8, 1873 (Leidy, 1873.1).
Figure 89. — Leidy's type (.holotypej of Palaeosyops major in
the collection of the Academy of Natural Sciences of Phila-
delphia
Part of the right ramus of a lower jaw. After Leidy, 1873. One-half natural size.
"The specimen is somewhat swollen and altered from disease and is one of those
upon which the species was first indicated. Discovered by Dr. Carter at Grizzly
Buttes." (Leidy.)
Type locality and geologic liorizon. — Grizzly Buttes,
Bridger Basin, Wyo.; Palaeosyops paludosus-OroTiip-
pus zone (Bridger B 2 or B 3). Dr. J. Van A. Carter,
collector.
Holotype. — "A jaw fragment with the retained
fragments of the true molars." This type is now in
the collection of the Philadelphia Academy of Natural
Sciences. (See fig. 89.)
Characters (Leidy). — Size apparently "much larger
than P. paludosus."
The true molars occupied a space of 4J^ lines. The last
molar measured IJ^ inches fore and aft and an inch trans-
versely in front. "
In his "Extinct vertebrate fauna" of 1873 (1873.1,
pp. 45, 46) Leidy gives a fuller description of the very
imperfect holotype and figures it on Plate 20, Figure
8. He believed the jaw specimen to be
in some degree abnormal in form, due to inflammation or
some other affection connected with the second molar tooth.
* * * In its proportions the jaw, in a normal condition,
would appear to be of more robust character than in Palaeo-
syops paludosus. * * * In its present state the base is
more convex fore and aft than in the latter, and the alveolar
border more ascending posteriorly.
The remains of the molar fangs at the entrance of the alveoli
appear to indicate teeth of the same form and construction as in
Palaeosyops paludosus, for which reason the fragment was
referred to the same genus. The true molars appear to have
occupied a space of 4^ inches, though this is probably some-
what exaggerated, as the interval occupied by the last inter-
mediate molar appears proportionately somewhat too large.
The crown of the last molar, which was clearly trilobate as in
Palaeosyops paludosus, had an antero-posterior diameter of 2
inches.
Leidy's paratype, "consisting of the left ramus of the
lower jaw, containing six molar teeth," was obtained
by Doctor Carter "in Dry Creek Canyon, 40 mUes
from Fort Bridger" (Bridger Basin, Bridger C) and
together with a second similar specimen from the
same locality is described by Leidy (1873.1, p. 46, pi.
23, fig. 1; second specimen, fig. 2).
The holotype, it is important to note, is from the
low level (probably Bridger B 2) of Grizzly Buttes,
but Leidy's paratype, which has the characters of the
more progressive Palaeosyops leidyi Osborn, is from
the higher level (Bridger C) of Dry Creek. The
paratype is thus certainly not conspecific with the
holotype.
Osborn's neotype. — In order to supplement the
characters of Leidy's imperfect holotype, the present
writer has selected as a neotype a lower jaw (fig.
268, C) with dentition (Am. Mus. 12181) from Cotton-
wood Creek and from about the same level (B 3)
as the holotype, with which it agrees closely. (See
Chap. V.)
Etymology. — major, in allusion to the larger size as
compared with P. paludosus.
Present determination. — The species P. major is
believed to be a valid one.
Canis montanus Marsh, 1871
Cf. Canisf marshii Hay, below {Palaeosyops major?), page 178
Original reference. — Am. Jour. Sci., 3d ser., vol. 2,
p. 123, August, 1871 (Marsh, 1871.2).
Type locality and geologic liorizon. — Grizzly Buttes,
Bridger Basin, Wyo.; Palaeosyops paludosus-Oro-
hippus zone (Bridger B, probably B 2).
Marsh's cotypes. — "A last upper premolar tooth in
good preservation, a canine, wanting most of the
crown, and a number of the larger bones of a skeleton,
all apparently of the same species, but pertaining to
three individuals, differing somewhat in size " (Marsh).
Of these materials the "last upper premolar" (first
lower premolar) alone is described and measiu'ed, and
it is also the first specimen mentioned. It should
therefore be taken as the lectotype (Yale Mus. 11770).
Characters. — "The last upper premolar * * *
is robust, has a short compressed crown. The princi-
pal cusp is conical, with subacute edges, the anterior
being about twice the length of the posterior. Behind
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
159
Figure 90. — Leidy's
type of Palaeosyops
humilis
the main cusp there is a large triangular tubercle,
with its apex exterior to the fore and aft axis of the
crown." (Marsh.)
Anteroposterior diameter of last upper pre-
molar 9 lines [19 mm.]
Greatest transverse diameter of same 4.25 lines [8 mm.l
Height of main cusp 6 lines [12.7 mm.]
Height of posterior tubercle 3.75 lines [7 mm.]
Synonym. — Canis? marsMi Hay was proposed in
place of 0. montanus IVIarsh, name preoccupied by
0. montanus Pearson (Hay,
1899.1).
Etymology. — montanus, dwell-
ing in the mountains — that is,
from the Eocky Mountain
region.
Present determination. — The
type specimen of "Canis mon-
specimen in the collection of tdnus" is a first lowcr premolar
the Academy of Natural of somc Undetermined member
S c i e n c e s of Philadelphia, <■ ,i n v
After Leidy, 1S73. Natural 01 t^ie gCnUS FttlaeOSyOpS, pOSSl-
size. Regarded by Leidy as jjly PaltteOSyOpS paludoSUS Or P.
"A last upper molar of the . rm p , /-y •
leftside. • • • Found by mc-jor. ihe reference to Cams
Doctor Corson on the buttes ^^Q,s, doubtless made by rcason
of Dry Creek Canyon." r ,i i . • ii n
of the deceptive resemblance oi
one of the lower premolars to the upper carnassial
tooth of a dog.
Palaeosyops humilis Leidy, 1872
Cf. Palaeosyops major, this monograph, page 321
Original reference. — Acad. Nat. Sci. Philadelphia
Proc, 1872, p. 168 (Leidy, 1872).
Subsequent reference. — Leidy, Extinct vertebrate
fauna of the Western Territories, p. 58, pi. 24, fig. 8,
1873 (Leidy, 1873.1).
Type locality and geologic Tiorizon. — "Valley of Dry
Creek 40 miles from Fort Bridger (Wyo.)." Doctor
Corson, discoverer. "Buttes of Dry Creek Canyon,"
Bridger Basin; horizon probably Bridger C (Uinta-
therium- Manteoceras- MesatirMnus zone) .
Holotype. — "An upper molar." (See fig. 90.)
Characters (Leidy). — "An upper molar tooth of this
animal measures three-foiu-ths of an inch in diame-
ter." In his later description Leidy recognized that
the specimen belonged to the milk series.
Etymology. — Tiumilis, lowly, small; in allusion to
the small size in comparison with P. paludosus.
Present determination. — This milk tooth probably
pertains to the genus Palaeosyops, but comparison
with P. major and P. leidyi leaves the species unde-
termined.
Palaeosyops Junius Leidy, 1872
Cf . Mesatirhinus Junius (Leidy) , this monograph, page 388
Original reference. — Acad. Nat. Sci. Philadelphia
Proc, 1872, p. 277 (Leidy, 1872.3).
Subsequent reference. — Leidy, Extinct vertebrate
fauna of the Western Territories, p. 57, no figure,
1873 (Leidy, 1873.1).
Type locality and geologic horizon. — Fort Bridger,
Bridger Basin, Wyo.; level not recorded. Dr. J.
Van A. Carter, collector.
Holotype. — -"Doctor Carter recently sent the writer
several small fragments of the right side of a lower
jaw, together with a sketch of a larger fragment of
the left side, containing the last premolar and the
succeeding molars." Of this type material only p4
(right) and the posterior half of ma (right) were
located (1906) in the collection of the Academy of
Natural Sciences of Philadelphia.
Characters. — Leidy writes:
Intermediate in size to P. paludosus and P. humilis. Founded
on portions of a lower jaw agreeing in character with the cor-
responding parts of P. paludosus but smaller. Space occupied
by the last premolar and the true molars, 4 inches. Antero-
posterior diameter of last premolar, 8 lines; of last molar,
173^ hnes.
In the fuller description in his memoir of 1873,
Leidy says:
The specimens * * * appear to indicate a small species
of Palaeosyops, though it is not improbable that they pertain
to a small variety of P. paludosus.
The parts agree closely with the corresponding parts of the
lower jaw and teeth of the latter, except in size. They have
been viewed as representatives of a species with the name of
Junius.
Figure 91. — Leidy's cotypes of Palaeosyops
Junius
Specimens in museum of Academy of Natural Sciences of
Philadelphia; Bridger B(?), level doubtful. A, Eight fourth
lower premolar (pO; B, posterior part of third lower molar
Cms) . Natural size.
The measurements of the teeth (fig. 91) in comparison with
those of P. paludosus are as follows:
Space occupied by the last pre-
molar and molars
Space occupied by the molars
Breadth [anteroposterior] of last
premolar
Thickness [transverse] of last
premolar
Breadth [anteroposterior] of first
molar
Breadth [anteroposterior] of sec-
ond molar
Breadth [anteroposterior] of
third molar
Thickness [transverse] of third
molar at middle
39J^
8
10
12
17
7
[Milli-
meters]
[102]
[94]
[17]
[12]
[21]
[25]
[10]
[14]
55
46
6M
12J^
15
19
93^
[MUli-
meters]
[116]
[96]
[19]
[12]
[38]
[32]
[39]
[19]
160
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Etymology. — Junius, younger, in allusion to its small
size.
Present determination. — From the two teeth (p4 and
part of ma) preserved it appears that this species
probably pertains to the genus Mesatirhinus. It is
smaller than Mesatirhinus megarhinus. No other
material has been certainly identified with it. (See
p. 388.)
Palaeosyops laticeps Marsh, 1872
Cf. Limnohyops laticeps Marsh, this monograph, page 311
Original reference. — Am. Jour. Sci., 3d ser., vol. 4,
p. 122, August, 1872, dated "July 18, 1872" (Marsh,
1872.1).
Type locality and geologic horizon.— NesiT Marsh's
Fork, about 15 miles from Fort Bridger, Wyo. A. H.
Ewing, discoverer. Level not recorded.
Holotype. — "A__nearly complete skeleton" (Yale
Mus. lioOO).
Etymology. — latus, broad; caput (in compounds ceps),
head; in allusion to the width across the zygomata.
Present determination. — Marsh's accurate diagnosis
of this excellent type was made before the generic
characters of Palaeosyops were fully known. The
species was subsequently chosen by Marsh as the
type of the genus Limnohyops Marsh, and both the
genus and the species are valid.
Telmatherlum Marsh, 1872
Cf. Telmatherium, this monograph, page 340
Original reference. — Am. Jour. Sci., 3d ser., vol. 4,
p. 123, August, 1872 (Marsh, 1872.1).
Type species. — Telmatherium validum Marsh. (See
pp. 160, 344 of this monograph.)
Generic characters. — See T. validum (p. 340).
Etymology. — reXfia, a pool, marsh (cf. "paludosus");
dripiov, beast.
Figure 92. — Marsh's type of Palaeosyops laticeps
Natural size.
Characters. — Marsh writes :
The teeth in this specimen have apparently the same general
structure as those in the type of P. paludosus but differ in
being nearly smooth, and this is not the result of age, as this
individual was younger than the original of the larger species.
The proportions, moreover, given for the molar described
(Leidy, 1870.2, p. 113), "22 lines fore and aft and 18 trans-
versely," would not apply to any of the series in the present
specimen. The last upper molar of the latter has two well-
developed internal cones. * * * The upper teeth form a
complete series. The canine is large and broadly oval at its
base. The outer incisor is the largest, and at its posterior
edge the premaxillary is subtriangular in transverse section.
The sagittal and occipital crests are strongly developed, and
the coronoid process of the lower jaw is short and recurved.
Measurements [Marsh] ^
Millimeters
Length of entire upper molar series 155
Anteroposterior extent of three true upper molars 94 [90]
Anteroposterior diameter of last upper molar 36 [33]
Transverse diameter [protocone to mesostyle] 40
Anteroposterior diameter of upper canine at base
[alveolar portion 28] 29
Transverse diameter 22
Present determination. — The generic term as re-
defined in the present monograph is a valid one.
Telmatherium validus Marsh, 1872
Cf. Telmatherium validum, this monograph, page 344
Original reference. — Am. Jour. Sci., 3d ser., vol. 4,
p. 123, August, 1872; dated "July 18, 1872" (Marsh,
1872.1).
Type locality and geologic horizon. — "Near Henrys
Fork of the Green River in Wyoming." (Bridger
Basin, level C or D.) J. F. Quigley, discoverer.
Holotype. — "The greater portion of a skull, with
teeth" (Yale Mus. 11120). (See fig. 93.)
Characters. — Marsh writes:
The dentition of this genus, so far as known, appears to be
similar to that of Palaeosyops; but the two may readily be dis-
tinguished by the anterior portion of the skull, which in the
present genus has the premaxillaries compressed, with an
elongated median suture. The zygomatic arch is also much
less strongly developed, and the squamosal portion of it is com-
paratively slender.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
Limnohyus Marsh, 1872
The upper molar teeth have the inner cones more elevated
and more pointed than in Palaeosyops, and the basal ridge is
well developed. The last upper molar has but a single internal
cone. The upper canines are large, pointed, and have strong
cutting edges. The outer incisors are the largest, and all these
teeth have a strong inner basal ridge. The roof of the mouth is
deeply excavated between the premolars. The nasals are de-
curved laterally and much compressed.
Cf. Palaeosyops, this monograph, page 331
Original reference. — Am. Jour. Sci., 3d ser.
p. 124, August, 1872 (Marsh, 1872.1).
Type species. — Limnohyus rohustus Marsh.
161
vol. 4,
Figure 93. — Marsh's type of Telmatherium validus
Natural size.
Measurements [Marsh]
Millimeters
Extent of upper molar series : 224
Extent of upper true molars 130
Anteroposterior diameter of last upper molar 54
Anteroposterior diameter of last upper premolar 28
Transverse diameter 33
Anteroposterior diameter of upper canine at base 27
Transverse diameter 22
Etymology. — validus, strong, stout; perhaps in allu-
sion to the large size of the upper canines.
Generic characters (Marsh). — The term Palaeosyops
is restricted to those specimens which, like P. paludo-
sus, possess two inner cones on m^
The other specimens have but a single internal cone on the
last upper molar, and for the genus thus represented the name
Limnohyus is proposed. These genera may be distinguished
from Telmatherium by the premaxillaries, which are short,
stout, and depressed, with a small median suture.
Etymology. — Xi^ufi?, a marshy lake; Cs, boar.
Present determination. — Since the type species Lim-
nohyus rohustus is now believed to be congeneric with
Figure 94. — Marsh's type of Limnohyus rohustus
Natural size.
Present determination. — This is a valid genus and
species. The name Telmatherium was amended to
Telmatotherium by Marsh in 1880 (1880.1) in his
"List of genera established by Prof. O. C. Marsh,
1862-1879," and the amended form was accepted by
Earle, Osborn, Hatcher, and later writers, but accord-
ing to the rules of nomenclature now generally ac-
cepted the amended form has no standing and the
original form Telmatherium should be used.
Leidy's Palaeosyops paludosus, Marsh's genus Limno-
hyus becomes a synonym of Palaeosyops.
Limnohyus rohustus Marsh, 1872
Cf. Palaeosyops rohustus (Marsh), this monograph, page 331
Original reference. — Preliminary description of new
Tertiary mammals: Am. Jour. Sci., 3d ser., vol. 4,
p. 124, August, 1872; dated "July 18, 1872" (Marsh,
1872.1).
162
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Type locality and geologic horizon. — Near Henrys
Fork, Wyo.; Uintatherium- Manteoceras- Mesatirhinus
zone (Bridger C or D). F. Mead, jr., collector.
Holotype. — A fragmentary skull including nasals and
palate with teeth (Yale Mus. 11122).
Characters. — Marsh writes:
The present species may be distinguished from those above
described [Palaeosyops laticeps Marsh, Tdmatherium validum
Marsh], especially by the strong basal ridge of the molars. On
the last lower molar it extends entirely around the posterior
lobe. The first of the upper true molars has the two inner
cones nearly of the same size. The small intermediate median
FiGTJRE 95. — Cope's cotypes of Palaeosyop.
After Cope, 1884. Ai, Lett mandibular ramus, superior view Oeetotype, Am. Mus. 5098); Aj, internal view otsame;
B, left upper premolars and molars (Am. Mus. 5097); Ci, third left superior molar, crown view; Ct, the same, buccal
view (Am. Mus. 5099); Di, fourth superior premolar, crown view; Dj, the same, buccal view (Am. Mus.).
tubercles are well developed on the upper molars, and all the
teeth are strongly rugose, even in fully adult animals. The
nasal bones contract anteriorly and are rounded in front.
The outer margin is decurved and thickened. The premaxil-
laries unite by a very short median suture, similar to that in
Palaeosyops laticeps. The zygomatic process of the squamosal
is stout but much compressed, thus differing widely from both
the species already described.
Measurements [Marsh]
Millimeters
Anteroposterior extent of last three upper molars 110
Anteroposterior diameter of last upper molar 41
Transverse diameter 43. 5
Anteroposterior diameter of last lower molar 51
Etymology. — robustus, robust; in allusion to the stout
skull and dentition.
Present determination. — The species is probably a
valid one, referable to the genus Palaeosyops.
Limnohyus Leidy (not Marsh), 1872
Cf. Palaeosyops, this monograph, page 155
Original reference. — Acad. Nat. Sci. Philadelphia
Proc, 1872, pp. 240-242; published December 17,
1872 (Leidy, 1872.1).
^ As we have seen above, Marsh's genus Limnohyus is
simply a synonym of Palaeosyops, which had been
defined by Leidy as having
"but a single lobe to the inner
part of the crown " of the "last
upper molar." In 1872 Leidy,
after pointing out this fact,
says that the name Limnohyus
"might with propriety be
applied to the animal with
molars like those of Palaeo-
syops except that the last upper
one has two inner cones to the
crown." This doubtless sug-
gested Marsh's subsequent
term Limnohyops. Lim-
nohyus Leidy is thus preoccu-
pied by Limnohyus Marsh,
which is a sj'^nonym of
Palaeosyops.
Etymology. — Xifivrj, a marshy
lake; Cs, boar.
Palaeosyops vallidens Cope, 1872
Cf. DoKchorhinus vallidens (Cope),
this monograph, page 401
Original rejerence. — -Pal.
Bull. No. 7, dated "Aug. 22,
1872"; Am. Philos. Soc. Proc,
vol. 12, p. 487, 1873 (Cope,
1872.1).
Subsequent reference. — Ter-
tiary Vertebrata, p. 699, pis.
51, fig. 1; 52, fig. 3; 53, fig. 1;
36, figs. 10, 10a, 11, 11a, 1884
[1885] (Cope, 1885.1).
Type locality and geologic
horizon. — -"Mammoth Buttes, southwestern Wyoming,
near the headwaters of Bitter Creek," Washakie Basin;
Eobasileus-Dolichorhinus zone (Washakie B 2).
Characters. — Cope writes:
Represented by the dentition of one maxillary bone with
other bones of one individual [Cope, Am. Mus. 5097]; a portion of
the same dentition of a second [No. 5099]; with both rami of
the mandible with complete dentition of a third [No. 5098].
The species is distinguished by the details of the dental struc-
ture and by the superior size. It exceeds, in this respect, the
Palaeosyops major Leidy; while the three posterior lower molars
measure 4.5 inches in length, the same teeth of the present
animal measure 6.25 inches. The last superior molar of an-
other specimen measures 2 inches in length; in the third the
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
163
first true molar is 1.5 inches in length, while the last inferior
molar is 2.25 inches long. The pecuharity in the structure of
the superior molars consists in the existence of two strong
transverse ridges, which connect the inner tubercle with the
outer crescents, inclosing a pit between them. These are most
marked on the premolars, where also is found the peculiarity
of the almost entire fusion of the outer crescents into a single
ridge. These united crescents are narrower than in P. major,
and the summits of all the crescents are relatively more ele-
vated. The number of inner tubercles is the same as in that
species; all the teeth have very strong basal cingula, which
rise up on the inner tubercle. The last inferior molar is rela-
tively narrower than in P. major, and the posterior tubercle is
larger and longer and is an elevated cone.
From the foregoing description it will be seen that
Cope based his specific description upon three speci-
mens (cotypes) without designating any one of the
three as being more typical than the others. (See
fig. 95.) The name vallidens, however, from vallum,
a wall, seems to refer to the peculiarity in the struc-
ture of the superior molars, which
consists in the existence of two strong transverse ridges, which
connect the inner tubercle with the outer crescents. * * *
These ridges are most marked on the premolars, where also is
found the pecuharity of the almost entire fusion of the outer
crescents into a single ridge.
If we had nothing further to guide us we would thus
be led to infer that the upper dentition (Am. Mus. 5097),
which best shows these peculiarities alluded to in the
name vallidens, should be regarded as the most typical
of the three specimens and should be chosen as the
lectotype. But in his "Tertiary Vertebrata" Cope
(1885.1, p. 700) says:
The bones containing the maxillary and mandibular teeth
were not found together in any instance, so that it is possible
that the different series may represent different species. No
other species of the genus was, however, found in the locaUties
to which the respective parts could be referred. Should these
prove not to pertain together, the lower jaws may be regarded
as typical of the species.
As Cope was the "first reviser" of the species there
seems to be no escape from the conclusion, if modern
rules of nomenclature are to be followed, that the lower
jaw (Am. Mus. 5098) must be treated as Cope's
lectotype.
Etymology. — vallum, wall or redoubt; dens, tooth;
allusion as explained above.
Present determination. — This little-known species is
allied to but probably specifically distinct from Doli-
cJiorTiinus Tiyognathus of Washakie B and Uinta B.
It is also more primitive than that species (see below).
Limnohyus laevidens Cope, 1873
Cf. Limnohyops laevidens (Cope), this monograph, page 305
Original reference. — Pal. Bull. No. 11 ("issued Jan.
31, 1873"); Am. Philos. Soc. Proc, vol. 13, pp. 35,
36, 1873 (Cope, 1873.5).
Subsequent references. — Cope, On the extinct Verte-
brata of the Eocene of Wyoming, observed by the
expedition of 1872: U. S. Geol. and Geog. Survey
Terr. (Hayden) Sixth Ann. Kept., p. 591, 1873 (Cope,
1873.6); Tertiary Vertebrata, p. 701, cotype skull, pi.
50, figs. 1, 2 (holo type) , fig. 3 (paratype), 1884 [1885]
(Cope, 1885.1).
Type locality and geologic horizon. — Type ("No. 1 "),
Cottonwood Creek, Bridger Basin, Wyo.; Palaeosyops
paludosus-Orohippus zone (Bridger B). Cotype ("No.
2"), Bitter Creek, Washakie Basin, Wyo.; horizon
uncertain.
Cope's cotypes: "A cranium lacking the posterior
part of one side and the lower jaw," from Cottonwood
Creek ("No. 1," now Cope collection. Am. Mus. 5104).
Also "a nearly complete cranium with dentition from
Figure 96. — Cope's cotypes of Limnohyops laevidens
After Cope, 1885. One-fourth natural size. A, Am. Mus. 5104, lectotype: Ai,
" Cranium lacking posterior part of one side and lower jaw, from Cottonwood
Creek" (Cope), "No. 1"; As, upper teeth of the same. B, Am. Mus. 6105,
now retered to Palaeosyopst copei, right maxilla, p^m'.
Bitter Creek" ("No. 2," now Cope collection. Am.
Mus. 5105). (See fig. 96.)
Cope's lectotype: Cope's first-mentioned specimen
is the one from Bitter Creek (Washakie B?) (Am. Mus.
5105), now referred to Palaeosyops? copei. But the
"No. 1" of Cope's description and measurements and
the specimen to which the name "laevidens" refers is
unquestionably the skull Am. Mus. No. 5104, from
Cottonwood Creek (level Bridger B), Bridger Basin,
now referable to Limnohyops. Furthermore, in the
"Tertiary Vertebrata" (Cope, 1885.1, pp. 701-703,
pi. 50, figs. 1, 2) Cope definitely selects, describes, and
figures this specimen as the type, again referring to
the Washakie specimen as "No. 2" and admitting
that its specific association with the other specimen
was doubtful. We therefore follow Cope in regarding
164
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
"No. 1," the Bridger specimen (Am. Mus. 5104), as
the lectotype.
Characters.- — Cope writes:
This species is one of the larger forms of the group originally
represented by Palaeosyops, and which has turned out to be so
numerous in species. [This statement refers apparently to
"No. 2."]
The anterior median small tubercle of the first true molar is
wanting. The last true molar has but one interior cone. [All
these statements apply evidently to " No. 1," the Bridger or Cot-
tonwood Creek specimen.]
The canine tooth is powerful and bearhke; the outer incisor
is the largest. The premaxillary bones are short, and the side
of the face elevated and plane to the convex nasal bones. Zygo-
matic arch massive.
Am. Mus. 5107.
Figure 97.—
After Cope, 1885. Ai,
natural
A3
-Cope's type (holotype) of Limnohyus fontinalis
Young skull seen from above, one-half natural size: As, the same, right side, one-half
size; A3, right maxilla with dp<, m', m', natural size.
The molars have the general form of those of L. robustus,
but the second superior premolar has but one outer tubercle.
The cingula are much less developed than in that species,
those between the inner cones of the molars being entirely
absent. These cones are low and, with the rest of the crowns
of all the teeth, covered with smooth and shining enamel.
Measurements [Cope, condensed and corrected]
IVtillimeters
Length of molar series (No. 1) 141
Length of true molars 84
Length of crown canine (anteroposterior) 20
Length of crown last molar (anteroposterior) 30
Width of crown last molar (transverse) 34
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
165
Etymology. — laevis, levis, smooth, shining; dens,
tooth; apparently in allusion to the "smooth and
shining enamel."
Present determination. — This is a valid species refer-
able to the genus Limnohyops.
Limnohyus fontinalis Cope, 1873
Cf. fPalaeosyops fontinalis Cope, this monograph, page 317.
Original reference. — Pal. Bull. No. 11, "issued Jan-
uary 31, 1873"; Am. Philos. Soc. Proc, vol. 13, pp. 35,
36, 1873 (Cope, 1873.5).
Subsequent references. — Cope, On the extinct Verte-
brata of the Eocene of Wyoming, observed by the
expedition of 1872: U. S. Geol. and Geog. Survey
Terr. (Hayden) Sixth Ann. Kept., p. 594, 1873 (Cope,
1873.6); Tertiary Vertebrata, p. 707, pi. 49, fig. 9;
pi. 50, fig. 4; pi. 58a, figs. 4, 5, 1884 [1885], (Cope,
1885.1).
Type locality and geologic Tiorizon. — "Found by the
writer on a bluff on Green River, near the mouth of
the Big Sandy, Wyoming." ("Isolated patch lying
northeast of the badlands." Probably Eometarliinus-
Trogosus-Palaeosyops fontinalis zone (Bridger A).)
Holotype. — A young, fragmentary skull (Cope col-
lection. Am. Mus. 5107, retaining dp*, m', - -
m^ of the right side. (See fig. 97.)
Characters. — Cope writes:
A small species agreeing with the P. paludosus
in the two interior cones of the last superior
molar. It is represented especially by a consider-
able part of the cranium of an individual in which
the last superior molar is not quite protruded,
but with the other molars and last premolar of
the permanent dentition in place. The enamel
of these teeth is in accordance with the age, delicately rugose,
and while the cingulum is present fore and aft, it is wanting
internally and externally. The anterior median tubercle is
present on all the true molars, and the bases of the acute inner
cones are in contact. The sagittal crest is truncate, and the
squamosal portion of the zygoma very stout. The nasal bones
are together very convex in transverse section.
Palaeosyops diaconus Cope, 1873
Cf. Palaeosyops robustus (Marsh), this monograph, page 331.
Original reference. — Pal. Bull. No. 12, p. 4, "pub-
lished March 8, 1873" (Cope, 1873.1).
Subsequent references. — Cope, On the extinct Verte-
brata of the Eocene of Wyoming observed by the
expedition of 1872: U. S. Geol. and Geog. Survey
Terr. (Hayden) Sixth Ann. Kept., p. 593, 1873 (Cope,
1873.6); Tertiary Vertebrata, p. 706, pi. 51, fig. 3,
1884 [1885] (Cope, 1885.1).
Type locality and geologic horizon. — Henrys Fork of
Green River, Wyo.; Uintatherium- Manteoceras- Mesa-
tirhinus zone (Bridger C or D).
Holotype. — "Represented by parts of the two
maxUlary bones, which present the crowns of the third
and fourth premolars, and of the second and third
true molars, with the bases of the other molars and
premolars." (Cope collection. Am. Mus. 5106.)
(See fig. 98.)
Characters. — Cope writes:
Belonging to the genus Palaeosyops as understood by Marsh —
that is, with two cones on the inner side of the last superior
molar. The species is as large as the Limnohyus major of
Leidy but differs in the relative proportions of the teeth.
Measurements [Cope]
Millimeters
Length of true molar series (2.75 inches) 67
Length of last molar 25
Width of last molar 26
Etymology. — fontinalis, of or from a spring, hence
original; in allusion to the primitive characters.
Present determination. — Cope was in error in inter-
preting the teeth of this skull, which belong to a very
juvenUe animal, the teeth exposed being the last
upper mUk tooth, dp*, the first and second molars,
m*, m^. The cranial characters, too, are very juve-
nUe. So far as they serve to guide us, the animal
probably belongs to the genus Palaeosyops, and also
probably to a distinct species, from a low geologic
level, possibly Bridger A.
Figure 98. — Cope's type (holotype) of Palaeosyops diaconus
Left upper teeth. Am. Mus. 5106. After Cope, 1885. One-half natural size.
Thus the last three molars have the same anteroposterior
length, while the space occupied by four premolars is shorter.
The anterior and posterior cingula of the true molars are very
strong, but it is not weU marked on the inner side between the
cones. The latter are acutely conic, and the median anterior
tubercle is strongly developed. Although the wearing of the
teeth indicates maturity, the enamel is coarsely and obtusely
rugose. The fourth premolar differs from that of L. major
in its smaller size relatively and absolutely and in the presence
of a prominent vertical tubercle on the outer face, rising to the
angle of the deep notch between the lobes. The third premolar
is as wide as the fourth and about as large as the corresponding
tooth in L. major, but different from it in the absence of tubercle
and ridge that mark its external face. The first premolar has
two roots, and the canine is large and short.
Measurements [Cope]
Millimeters
Length of entire molar series 171
Length of true molars 106
Length of last molar (crown) 42
Width of last molar (crown) 43. 7
In comparison with Marsh's description of his P. laticeps,
the measurements are all larger, and the enamel is as rugose as
in L. major, instead of smooth. The shortening of the pre-
molar series is greater in P. diaconus; thus in P. laticeps the
two sets of molars are related as 94 to 61 millimeters; in the
present one, as 106:65; were the proportions similar, the length
of the premolar series should be 69 millimeters.
166
TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
Etymology. — 5t?, double; kcows, cone; because the
third upper molar had two mner cones.
Present determination. — The name P. diaconus Cope
is probably a synonym of Palaeosyops rohustus
(Marsh), as explaiaed in Chapter V of this mono-
graph.
Diplacodon Marsh, 1875
Cf. Diplacodon Marsh, this monograph, pages 155, 439
Original reference. — Am. Jour. Sci., 3d ser., vol. 9, p.
246, March, 1875, "dated February 20, 1875" (Marsh,
1875.1).
Type species. — Diplacodon elatus Marsh. (See
p. 439.)
Generic characters. — Marsh writes:
The genus here established presents characters in some
respects intermediate between Limnohyus and Brontotherium.
It agrees with the former in its complete dentition (44 teeth)
and in the general form of the incisors, canines, and true molars.
It resembles the latter still more closely in the premolar and
molar teeth, and parts of the skeleton, especially in the verte-
brae, and bones of the extremities. From the Eocene Limno-
hyidae, already described, this genus is sharply distinguished
by the last upper premolar, which has two distinct inner cones,
thus agreeing essentially with the first true molar. This char-
acter, which has suggested the name of the genus, is one step
^9
Characters. — The specific characters were not for-
mally separated from the generic characters above
given under Diplacodon.
Measurements [Marsh]
Millimeters
Extent of upper molar series 242
Extent of upper true molars 152
Anteroposterior diameter of last upper molar 60
Transverse diameter 59
Anteroposterior diameter of [upper] canine, at base 32
Height of crown 27
Etymology. — elatus, lofty; apparently in allusion
either to the large size or to the advanced stage of
evolution.
Present determination. — This important genus and
species was based upon an excellent type. The genus
and species are vahd. (See p. 439.)
Brachydiastematherium Bockh and Maty, 1876
Cf. Brachydiastematherium Bockh and Maty, this monograph,
page 382
Original reference — Mittheilungen aus Jahrb. K. k.
geol. Anstalt, Band 4, pp. 125-150, 1876 (1876.1).
Type species. — Brachydiastematherium transilvani-
cum Bockh and Maty. (See p. 382.)
Figure
-Marsh's type of Diplacodon elatus
One-third natuial size.
toward the modern type of perissodactyl dentition. The dental
formula of the genus is the same as Limnohyus, viz, incisors f ,
canines \, premolars |, molars f. In other respects the teeth
most resemble those of the Brontotheridae. From this family
Diplacodon differs widely in its dentition and the absence of
horns.
Etymology. — SlwXoos, double; aK-q, a point; 65ovs, tooth;
because the upper premolars had two inner cones.
Present determination. — This genus is certainly
valid so far as it applies to the type species. (See
p. 439.)
Diplacodon elatus Marsh, 1875
Cf. Diplacodon elatus Marsh, this monograph, page 439
Original reference. — Notice of new Tertiary mam-
mals: Am. Jour. Sci., 3d ser., vol. 9, p. 246, March,
1875; dated "February 20, 1875" (Marsh, 1875.1).
Type locality and geologic horizon. — "Upper Eocene
beds of Utah"; horizon probably Diplacodon-Proti-
tanotherium-Epihippus zone (Uinta C, true Uinta for-
mation).
Holotype. — A palate with dentition nearly complete,
parts of the skull and skeleton (Yale Mus. No. 11180).
Generic characters. — The generic characters mingled
with the specific characters are given below. (See
also p. 382.)
Etymology. — jSpaxvs, short; didaTrnxa, an interval;
dr]piov, beast; in allusion to the short diastema
between the lower canines and first premolars.
Present determination. — The type of this genus is an
animal closely similar in size and in stage of evolution
to the Protitanotherium superhum of the upper Eocene
of Utah but differs in certain characters, which are
probably of generic value. (See p. 382.)
Brachydiastematherium transilvanicum Bockh and Maty, 1876
Cf. Brachydiastematherium transilvanicum Bockh and Maty,
this monograph, pages 382, 941
Original reference. — -Mittheilungen aus Jahrb. K. k.
geol. Anstalt, Band 4, pp. 125-150, pis. 17, 18, 1876
(1876.1). Cf. Toula, Akad. Wiss. Wien Sitzungsber. ,
Band 101, pp. 612 et seq., 1892 (1892.1).
Type locality. — Andrashaza (Siebenbtirgen), Hun-
gary (Transylvania, eastern Hungary, about 150 miles
northeast of Belgrade). Collected in 1871 by Dr.
Alex. Pavay.
DISCOVERY OP THE TITANOTHEEES AND ORIGINAL DESCRIPTIONS
167
Geologic level. — The specimen, according to Pro-
fessor Koch (Bockh, 1876.1, p. 149), was found in
"buntes Thongebilde" of "lower" Eocene age, but
the assignment of a form of this advanced stage to a
level so low appears unwarrantable, and later evidence
indicates that the age of this specimen is more prob-
ably upper Eocene. (See p. 382.)
Holotype. — Anterior part of lower jaw, containing
incisors, canines, four premolars, and one molar.
Originally described and defined by Bockh (1876.1)
as a palaeotherioid. (See fig. 100.)
Generic and specific characters. — Bockh and Maty
(p. 148) write:
einer dreieckigen Emailzunge. An der Krone sammtlicher
Zahne sind die Reste einer diinnen cementartigen Kruste zu
sehen.
The following measurements are taken from the
original figures:
Millimeters
I3, anteroposterior 22
I3, transverse 20
C, maximum anteroposterior diameter (horizontal measure-
ment near base) 38
C, maximum transverse 31
C, lieight of crown (estimated) 40
Postcanine diastema (at top) 12
Pi-p4, anteroposterior 107
Pi, anteroposterior 18
Figure 100. — Type (holotype) lower jaw of Brachydiastematherium transilvanicum
After Bockh and Maty, 1876. A', Side view; K', inner side; A', rear view of mi; A<, front view of mi; A^, top view of jaw; A', outer view of right lower canine; A',
section of root of right lower canine; A^, fragment of right lower incisor. Two-fifths natural size.
I3, mit flachliegelformiger Krone, welche mit warziger
Emailwulst versehen ist; Ci, mit kegelformiger Krone, welche
gleichfalls eine warzige, starke Emailwulst besitzt; seine
Wurzel ist iiberaus stark, lang und gerade. Die Zahnliicke ist
sehr kurz; pi, deren erster am kleinsten, und seine nur eine
Wurzel besitzende Krone stellt nur einen einfachen Kegel dar;
die iibrigen drei wachsen gradatim und die warzige Wulst der
Basis fehlt an der inneren Seite dieser letzteren. Die drei
letzten Praemolare ahmen wohl die Form der entsprechenden
Zahne der echten Palaeotherien nach, wirkliche Halbmonde
an der Oberflache seiner abgewetzten Krone zeigt indessen nur
der vierte Praemolar; an den demselben vorangehenden zwei
Zahnen kann die Verzierung noch niclit als Halbmond bezeich-
net warden. Die hintere Bucht des vierten Praemolares 1st
durch eine Scheide in zwei Theile abgetheilt, und heizu ist
der Keim auch schon beim dritten Praemolar zu bemerken;
m (?)3, die innere Seite des ersten echten Molares zeigt gleich-
falls keine Emailwulst, an der Mitte der hinteren Seite des
hinteren Halbmondes vereinigt sich indessen die Wulst mit
P2, anteroposterior 26
P2, transverse (through anterior lobe) (estimated) 17
P3, anteroposterior 31
P3, transverse (estimated) 22
P4, anteroposterior 38
P4, transverse (estimated) 28
Ml, anteroposterior 50
Ml, transverse (estimated) 30
Etymology. — transilvanicum, Transylvanian.
Present determination. — The species is probably
valid.
Leurocephaius Osborn, Scott, and Speir, 1878
Cf. Telmatherium Marsh, this monograph, page 341
Original reference. — ^E. M. Mus. Geol. and Arch.
Princeton Coll. Contr. No. 1, p. 42, pi. 4, 1878 (Osborn,
Scott, and Speir, 1878.3).
168
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Type species. — LeurocepTialus cultridens Osborn,
Scott, and Speir. (See p. 341.)
Generic characters. — Scott and Osborn write:
Upper incisors acute, with strong posterior ridges, lower in-
cisors compressed and laniariform, canines compressed, with
serrated cutting edges; first upper premolar with rudimentary
anterior lobe, last upper molar with rudimentary postero-
internal cusp. Molars constructed as in Palaeosyops but higher,
with sharper cones and more erect external lobes. Internal
median valley very much deeper. Little or no depression at the
forehead; zygomatic arch round, comparatively straight and
does not project outward, and with obscure postorbital
process. Premaxillaries short and straight. Mandible with
nearly straight lower margin and shallow masseteric fossa;
mental foramen single.
Etymology. — Xeupos, smooth; Ke(j>a\r], head; in
allusion to the smooth texture of the bone.
Present determination. — LeurocepTialus is a synonym
of Telmatherium Marsh.
Figure 101. — Type (holotype) of Leurocephalus cultridens
Right prematiUa, maxilla, and mandibular ramus. Princeton Mus. IQ027. Atter Osborn, 187S. One-third
natural size.
Leurocephalus cultridens Osborn, Scott, and Speir, 1878
Cf. Telmatherium cultridens (Osborn, Scott, and Speir), this
monograph, page 341
Original reference. — E. M. Mus. Geol. and Arch.
Princeton Coll. Contr. No. 1, p. 42, pi. 4, 1878 (Osborn,
Scott, and Speir, 1878.3).
Subsequent reference. — Earle, A memoir on the genus
Palaeosyops Leidy and its allies: Acad. Nat. Sci. Phil-
adelphia Jour., 2d ser., vol. 9, pp. 343-348, pi. 10,
tig. 3, 1892; type (Earle, 1892.1).
Type locality and geologic horizon. — Henrys Fork
divide, near Fort Bridger, Wyo.; Vintatherium-
Manteoceras- Mesatirhinus zone (Bridger C or D).
Holotype. — "Established on specimen having a
nearly complete dentition and portions of the cran-
ium" (Princeton Mus. 10027). (See fig. 101.)
Specific characters. — The specific and generic char-
acters were not distinguished in the original descrip-
tion.
Etymology. — culter, a knife; dens, tooth; in allusion
to the sharp-edged reciurved canines.
Present determination. — This genus is a synonym of
Telmatherium Marsh. The species is valid. (See
p. 341.)
Palaeosyops borealls Cope, 1880 '
Cf. Eotitanops horealis (Cope), this monograph, pages 156, 292
Original reference. — Am. Naturalist, vol. 14, p. 746,
1880 (Cope, 1880.1).
Subsequent reference. — Cope, Tertiary Vertebrata,
p. 703, pi. 58a, fig. 3, 1884 [1885], (Cope, 1885.1).
Type locality and geologic horizon. — "Badlands in the
upper drainage basin of the Big Horn River in western-
central Wyoming"; Wind River formation, horizon
not determined, probably Lambdotherium-Eotitanops-
Coryphodon zone (Wind River B). J. L. Wortman,
collector.
Holotype. — "Founded on a portion of the right
maxillary bone, which supports the three true molars
and one premolar" (Cope collection. Am. Mus. 4892).
(See fig. 102.)
Characters. — Cope writes :
Size of Limnohyus fontinalis, or much
smaller than P. laevidens. Anterior median
tubercle well developed; anterior and pos-
terior cingula strong, not rising to inner
cones. A low ridge extending outward and
forward from posterior cone. Enamel
smooth. Differs from P. junior Leidy in
the presence of the intermediate tubercle
and crest and in the weak ex-ternal cin-
gulum. Length of true molar series 63 [mil-
hmeters]; diameters of first true molar,
anteroposterior, 19; transverse, 20.
Etymology. — borealis, relating to
Boreas; in allusion to the Wind River
formation.
Present determination. — The species
is valid but generically distinct from
Palaeosyops. It is the type of the
genus Eotitanops Osborn. (See p. 289 .)
Lambdotherium Cope, 1880
Cf . Lambdotherium Cope, this monograph, page 279
Original reference. — Am. Naturalist, vol. 14, p. 746,
1880 (Cope, 1880.1).
Subsequent reference. — Cope, Tertiary Vertebrata,
p. 710, 1884 [1885] (Cope, 1S85.1).
Figure 102. — Type (holotype) of Palaeosyops borealis
Right upper part of right maxilla with p<-m3. Am. Mus. 4892. After Cope,
18S5. Natural size.
Type species. — Lambdotherium popoagicum Cope.
(Seep. 281.)
Generic characters. — Cope writes :
Dentition much as in Limnohyus, excepting that there is a
diastema in front of the second inferior premolar. Presence of
first inferior premolar not ascertained. Fourth inferior pre-
molar without posterior cusps. Superior molars with an
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
169
angular ridge extending inward from each inner cusp. Last
inferior molar with heel. * * * The V-shaped crests of the
inferior molars separate it from Hyracotherium.
Etymology. — Xa/i/35a, lambda; dTjplov, beast; in allusion
to the A-shaped crests of the lower molars.
Present determination. — The genus is valid and is
now referred to the Eocene titanotheres.
Lambdotherium popoagicum Cope, 1880
Cf. Lambdotherium popoagicum Cope, this monograph, page 281
Original reference. — Am. Naturalist, vol. 14, pp.
746, 747, 1880 (Cope, 1880.1).
Subsequent reference. — Tertiary Vertebrata, p. 710,
pi. 58b, figs. 7 and 7a, 1884 [1885] (Cope, 1885. 1).
Type locality and geologic Tiorizon. — Badlands of
Wind Elver, western-central Wyoming; Lambdo-
iherium-Eotitanojps-Coryphodon zone (Wind River B).
Type. — A lower jaw with dentition (Am. Mus. 4863).
(See fig. 103.)
Specific cliaracters. — Cope writes :
The heels of the second and third premolars have
a median keel; the third only has an anterior
tubercle. The crest of the heel of the fourth forms
an imperfect V. Heel of the last true molar small.
No cingula; enamel smooth. Length of molar series
80 [millimeters]; of true molars 44; of last molar
19; depth of ramus at first premolar 21; at last
molar 31. * * * About the size of Hyrachyus
agresiis.
Etymology. — popoagicum, in allusion to
Popo Agie River, a tributary of Wind River.
Present determination. — The species is
valid. (See p. 283.)
Lambdotherium brownianum Cope, 1881
Cf. EoHtanops brownianus (Cope), this monograph,
page 292
Original reference. — U. S. Geol. and Geog.
Survey Terr. Bull., vol. 6, p. 196, 1881
(Cope, 1881.2).
Subsequent reference. — Cope, Tertiary
Vertebrata, p. 709, pi. 56a, fig. 10 (not the
type), 1884 [1885] (Cope, 1885.1).
Type locality and geologic horizon. — Badlands of
Wind River, western-central Wyoming; Lambdo-
therium-Eotitanops-CorypJiodon zone (Wind River B).
Holotype. — "The greater part of a lower jaw,"
with p^, m'-m' (Cope collection. Am. Mus. 4885).
(See fig. 104.)
Characters. — Cope writes:
Considerably larger than the L. popoagicum and about
equal to the Tapirus ierreslris. The greater part of a lower
jaw represents the species, and on this, unfortunately, only
one of the premolar teeth remains. The three premolars are
all two-rooted, and the posterior lobe of the last true molar is
well developed. The inferior part of the e.xternal side of the
ramus contracts or retreats rather abruptly posteriorly, below
the last molar. It presents a slight external convexity below
the second and third premolars. The alveolar line rises rapidly
101959— 29— VOL 1 14
posteriorly, so that the last true molar is quite oblique. The
second (first) premolar has a considerable heel, which is narrow
and elevated on the middle line. The principal cusp is large
and compressed but obtuse and has no anterior basal tubercle.
Measurements [Cope]
Millimeters
Length of six molars 90
Length of true molars 55
Diameters of second (first) premolar:
Vertical 9
Anteroposterior 12
Transverse 6
Length of base of first true molar 15
Width of base of first true molar 9
Length of base of third true molar 23
Width of base of third true molar 11
Depth of ramus at second premolar 30
Depth of ramus at ma:
At front of tooth 39
At end of tooth 47
Etymology. — "Dedicated to my friend Arthur E.
Brown, superintendent of the Philadelphia Zoological
Garden" (Cope).
FiGUEE 103. — T3'pe (holotype) of Lambdotherium popoagicum
Left mandibular ramus, with pj-ms. Am. Mus. 4863. After Cope, 1885. Natural size.
Present determination. — The species is valid,
generic reference is to Eotitanops. (See p. 292.)
The
Palaeosyops hyognathus Osborn, 1889
Cf. Dolichorhinus hyognathus (Osborn), this monograph,
page 409
Original reference. — Am. Philos. Soc. Trans., new
ser., vol. 16, p. 513, 1890 [author's reprint issued
Aug. 20, 1889; O. P. Hay] (Scott and Osborn, 1890.51).
Subsequent reference. — Earle, A memoir upon the
genus Palaeosyops Leidy and its allies: Acad. Nat. Sci.
Philadelphia Jour., 2d ser., vol. 9, pi. 11, figs. 10,
11 [type], 1892 (Earle, 1892.1).
Type locality and geologic horizon. — Washaliie, White
River, northeastern Utah; Washakie B.
170
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Eolotype and specific characters. — Scott and Osborn
write :
In the Washakie beds is found a large species, about the
same size as P. vallidens Cope, which is provisionally referred
to Palaeosyops (P. hyognathus, sp. nov., Princeton collection.
No. 10273). This is represented by a lower jaw seven-eighths
as large as the type mandible of Diplacodon. [See fig. 105.]
;-->;x..i4::
Figure 104. — Cope's type of Lainhdotherium brownianum
One-half natural size.
As in the latter, the incisors form a close procumbent series;
the tips forming a gently arched line when seen from above.
The symphj'sis is extremely long (11 centimeters) and shallow;
the canines are rather small and semiprooumbent. The molar-
premolar series measures 24.5 centimeters, the last molar
measures 6.5 centimeters, the transverse measurement outside
of the canines is 9.6 centimeters; in Diplacodon elalus
the same measurement is 10 centimeters. Unfortu-
nately, the premolar crowns are broken; it is probable
that one or two of the premolars will be found to be
like the molars. The characters of the chin and sym-
physis are significant of close relationship to Dipla-
codon elatus.
Etymology. — vs, boar; yvados, jaw; in allusion
to the forward-pointing lower incisors and
shallow mandibular symphasis.
Present determination. — The species is valid.
The generic reference is to DolicliorTiinus.
(See p. 409.)
Liitinohyops Marsh, 1890
Cf. Limnohyops Marsh, this monograph, page 303
Original reference. — Am. Jour. Sci., 3d ser.,
vol. 39, p. 525, 1890 (Marsh, 1890.1).
Type species. — Palaeosyops laticeps Marsh
(Am. Jour. Sci., 3d ser., vol. 4, p. 122, 1872).
(Seep. 311.)
Generic characters. — Marsh says:
In 1872 the writer described a large mammal from the Eocene
of Wyoming under the name of Palaeosyops laticeps. As the
name Palaeosyops has since been restricted, this species must
be regarded as representing a distinct genus, which may be
called Limnohyops. In this form the last upper molar has two
inner cones, and in Palaeosyops, as now defined, there is only one.
Etymology. — Xiixfrj, shore; vs, boar; &}//, face.
Present determination. — This is a valid genus and
species. For fuller descriptions, see page 303.
Palaeosyops megarhinus Earle, 1891
Cf. Mesatirhinus megarhinus (Earle), this monograph,
page 388
.--''" "~\ Original reference. — Am. Naturalist,
vol. 25, No. 289, pp. 45-47, 1 fig., Jan-
;-''' uary, 1891 (Earle, 1891.1).
\__,,— ,^ Subsequent reference. — Earle, A mem-
oir on the genus Palaeosyops Leidy and
its allies: Acad. Nat. Sci. Philadelphia
Jour., 2d ser., vol. 9, pp. 320-329, pi.
10, fig. 2; pi. 11, figs. 4, 5, 1892
(1892.1).
Type locality and geologic horizon. —
Washakie Basin of Wyoming; level unde-
termined, probably Uintatherium- Man-
teoceras- Mesatirhinus zone (Washakie A).
Type.— "A fine skull (No. 10008) in
the Princeton collection" (Earle). (See
fig. 106.)
Paratype. — Earle writes:
There is also another portion of a skull (No. 10041), probably
belonging to this species, with the occiput well preserved, from
the Bridger proper [Earle, 1891.1, p. 45]. This paratype probably
belongs to a more advanced species of this genus. (See p. 388.)
Figure 105. — Type (holotype) of Palaeosyops hyognathus
Incomplete lower jaw. Princeton Mus. 10273. After Earle, 1892. One-fifth natural size.
Specific characters. — Earle writes:
Cranium: The characters of this skull are quite unique
and depart widely from any of the species of the family that
I have examined. The general form of the skull is broad and
depressed. Its dorsal contour is very like that of Palaeotherium
crassum — namely, there is no frontal depression, which is so
characteristic of Palaeosyops paludosus, and the occipital
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
171
region is only slightly higher than the frontal. The temporal
fossae are not deeply excavated, and the occipital crests are
weakly developed when compared to this region of the slcull in
Limnohyops. The occiput itself is high and rather narrow.
The foramen magnum is wide, bordered by very large condyles.
The auditory processes are widely separated. The post-
tympanics are broad and heavy. The postglenoid is peculiar
in form; it is very short and thick; its form is very different
from other species in the collection. An internal glenoid
process is present in this species. The mastoid was probably
exposed. The form of the zygomatic arch is striking; it is
very light, nearly straight, with the temporal portion strongly
compressed. The malar portion is also peculiar; the malar
insertion is very abrupt and strongly depressed, With the
external part very broad, thin, and shelf-like. The infra-
orbital foramen is not exposed. The form of the malar in this
species is totally different from all other allied forms
that I have examined. The orbit is very small, termi-
nates anteriorly above the anterior border of the second
superior molar; the postorbital processes are well
marked. The facial region of the skull is very short,
compared to the total length of the cranium. The
nasals are very long and heavy; their distal portion is
expanded and broader than the middle part. The nasal
notches are very deep and high. The premaxillaries are
triangular in outline; their symphysis is short and nar-
row, with a prominent anterior keel. The canine al-
veolus is very prominent. The palate is long and nar-
row, the roof of the same being strongly arched. The
posterior termination of the palate is at the second
superior molar. The incisive foramina are not divided.
Teeth: The crowns of the teeth in this skull are
badly damaged, but enough remains to give the total
measurements and the characters of the last molar.
The superior molars in this species form a continuous
series, being not interrupted by a diastema. The
sections of the incisors are very small. The canines
are also very small and diverge widely. Only the
second and third molar of each side are partially pre-
served. They have a square form with low crowns;
externally they are totally without a cingulum. The
external V's are rather wide and angular, in this respect
approaching that of Telmalotherium. The last molar
is without any intermediate conules.
Measurements
Millimeters
Length of skull, from premaxillary symphysis to end
postglenoid ' 285
Length from orbit to premaxillary symphysis 125
Length from orbit to postglenoid 160
Depth of nasal notch 84
Length of nasals 100
Entire molar series 148
Last superior molar:
Anteroposterior 37
Transverse 39
Etymology. — fie-yas, great, pis, nose; in allusion to the
length of the nasal bones.
Present determination. — This is a valid species which
has been made the type of the genus MesatirJiinus
by Osborn. (See p. 388.)
Palaeosyops minor Earle, 1891
Cf. Palaeosyops paludosus, this monograph, page 319
Original reference. — Acad. Nat. Sci. Philadelphia
Proc. for 1891, p. 112, issued March 31, 1891 (Earle,
1891.2).
Subsequent reference. — Earle, A memoir upon the
genus Palaeosyops Leidy and its allies: Acad. Nat.
Sci. Philadelphia Jour., 2d ser., vol. 9, pp. 269, 331,
332, 1892 (1892.1).
Earle's cotypes. — In his original description Earle
says (1891.2, p. 112), "P. minor embraces specimens
which Leidy erroneously described as P. paludosus, pi.
4, figs. 3-6, of Leidy's report for 1873." In his memoir
(1892.1, pp. 269, 330) Earle refers again to Leidy's
Plate 4, Figures 3-6, as the types of P. minor, but on
page 332 he says, "We may consider as the type
specimen" the "beautifully preserved mandible fig-
FiGURB 106. — Type (holotype) of skull of Palaeosyops megarhinus
Princeton Mus. 10008. After Earle, 1892. No scale given.
ured by Leidy" (Leidy, 1873.1, pi. 5, figs. 10, 11);
and again on page 387 he states that the specimen
figured in his (Earle's) Plate 12, Figure 14, is "the
type of this species and is in the collection of the
Academy of Natural Sciences of Philadelphia." But
this specimen is apparently the same one figured in
Leidy's Plate 4, Figure 5. (See fig. 107.)
Specific characters. — Earle writes:
Second superior premolar with two external lobes, external
lobes of last superior premolar equal. Intermediate conules of
true molars reduced, a strong external cingulum present.
Etymology. — minor, in allusion to the relatively
small size.
Present determination. — Of the first-mentioned speci-
mens (Leidy, 1873.1, pi. 4, figs. 3-6) Figures 3 and 4
represent an upper dentition, which is probably con-
specific with P. paludosus as determined in this mono-
graph; hence if this is taken as Earle's type P. minor
becomes a synonym of P. paludosus.
172
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
On the other hand, Leidy's Plate 4, Figures 5 and 6,
represent an upper dentition of uncertain specific ref-
erence. The "beautifully preserved mandible"
(Leidy's pi. 5, figs. 10, 11) is probably referable to P.
paludosus. Hence we may regard P. minor as a syno-
nym of P. paludosus.
Type locality and geologic horizon. — Cottonwood
Creek, Bridger Basin, Wyo.; Palaeosyops paludosus-
Orohippus zone (Bridger B).
Holotype. — "A jaw, No. 10275 [Princeton Mus.],
associated with a well-preserved radius, ulna, and two
metacarpals."
Ci
FiGUKJs 107. — Earle's ootypes of Palaeosyops minor in the collection of the Academy of Natural
Sciences of Philadelphia
Ai, Lett maxilla with root of canine and premolar-molar series. After Leidy, 1873. Two-thirds natural size. Aj, The same;
outer view of premolar-molar series. B, Another specimen; left upper premolar-molar series. After Leidy, 1873. Two-
thirds natural size. (A reversed view of this specimen, which is of uncertain specific reference, was figured by Earle
as the type (Earle's pi. 12, fig. 14).) Ci, Left mandibular ramus with p3-m3. After Leidy, 1873 (pi. 5, fig. II). One-halt
natural size. C2, The same, pj-ma; crown view. After Leidy, 1873 (pi. 5, fig. 10). One-half natural size. The last two
specimens are referable to Palaeosyops paludosus.
Palaeosyops longirostris Earle, 1892
Cf. Palaeosyops longirostris Earle, this monograph, page 319
Original reference. — Acad. Nat. Sci. Philadelphia
Jour., 2d ser., vol. 9, p. 338, 1892 (Earle, 1892.1).
Characters. — Earle writes:
The type jaw of this species, with the parts of the skeleton
associated with it, was referred by Scott and Osborn [Osborn,
1878.3, pp. 37, 38] to our P. minor (equal, in part, to P. palu-
dosus Leidy). After comparing Leidy's type specimen [prob-
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
173
ably the specimen figured in Leidy's memoir of 1873, pi. 5,
fig. 11] with this jaw, I find that there is such a marked difference
in some of its characters that I have to give it a specific rank.
The following characters distinguish it from Leidy's type: (1)
The great posterior extension of the jaw behind the last molar
(this is a unique character of this jaw — I have not observed it
in any other species of this subfamil}') ; (2) the symphysis is
much more elongated than in P. minor; (3) the lower border is
straighter and less inflected than in P. minor; (4) the posterior
en'f
Figure 108. — Earle's type of Palaeosyops longirostris
Princeton Mus. 1027-5. One-foarth natural size.
tubercle of the last inferior molar is much larger than in the
last named species; (5) the V's of premolar 4 are not so well
developed as in P. ininor, and there is also a well marked dif-
ference in the size of the first molars of the two species.
In this jaw the first true molar is con-
siderably smaller than in P. minor. The
canine is very large and semiprooumbent,
its position in the jaw resembling that of
T. hyognathus.
Etymology. — longus, loia^g; rostrum,
bill, snout (hence, in this instance,
jaw); in allusion to the great
posterior extension of the j aw behind
the last molar. (Earle.)
Present determination. — This prob-
ably valid species is certainly refer-
able to the Palaeosyopinae and
probably to Palaeosyops. (See
p. 319.)
Telmatotherium diploconum Osborn, 1895
Cf . Rhadinorhinus diploconus (Osborn) , this
monograph, page 431
Original reference. — Am. Mus.
Nat. Hist. Bull., vol. 7, p. 85, fig. 6,
1895 (Osborn, 1895.98).
Type locality and geologic horizon. —
NoTthe&sternlJtah;" Telmatotherium
cornutum" beds, Eohasileus-Doli-
chorhinus zone (Uinta B).
Holotype. — "The type is a skull
(No. 1863) [Am. Mus.] in which the
nasals are wanting and the mid-
region of the cranium was ci'ushed."
(See fig. 109.)
Characters. — Osborn writes:
Superior premolar-molar series, 174 millimeters. A large
hypocone upon last upper molar. Nasofrontal without horn.
Long sagittal crest. Canines small, rounded.
This species differs from T. megarhinum in the absence of the
infraorbital shelf and in the presence of a large hypocone upon
the last upper molar. The premolar-molar dentition is similar in
size and form to that of T. cultridens, but there are the following
important general differences: (1) Canines small and circular in
section; (2) a very short diastema, if any, behind the canine;
(3) a large hypocone upon m'; (4) the infraorbital foramen
close beneath the anterior border of the molar [malar]. [Com-
parisons with T. cultridens follow.]
Etymology. — StxXoos, double; Kcofos, cone; in allu-
sion to the presence of two internal cones on the third
upper molar.
Present determination. — The species is valid; it is
now referred to the genus Rhadinorhinus. (See p. 431.)
Telmatotherium cornutum Osborn, 1895
Cf. Dolichorhinus hyogiiathus (Osborn), this monograph, page 409
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
7, p. 90, figs. 10, 11, 1895 (Osborn, 1895.98).
Type locality and geologic horizon. — Northeastern
Utah; "Telmatotherium cornutum beds," Eohasileus-
Dolichorhinus zone (Uinta B 2).
Holotype and paratypes. — Osborn writes:
The type of this species is a fine skull (No. 1851) [Am. Mus.],
while several other well-preserved skulls from the same levels
give us all the cranial characters and the superior dentition
(Nos. 1850, 1847, 1848, 1852, 1837). [See fig. 110.]
Figure 109. — Type (holotype) of Telmatotherium diploconum
Superior and lateral views of skull. The nasals are broken off. Am, Mus. 1863. After Osborn, 1895.
One-fourth natural size.
Characters. — Osborn writes:
Incisors f . Premolar-molar series, 208 millimeters. A nar-
row diastema. Upper canines lanceolate. Long premaxillary
symphysis. A well-developed nasofrontal protuberance. Top
of cranium completely flattened. No sagittal crest. An
infraorbital process upon malar.
This species is remarkable for its very long flat-topped cranium
and its incipient knoblike osseous horns borne chiefly upon the
nasals but partly upon the frontals. These horns project
laterally and rise slightly above the general surface, and are best
174
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
seen in the anterior view, Figure 110. Ttiese characters and
the absence of the frontoparietal and interparietal sutures all
point well toward Titanotherium, but the premolars are still
absolutely simple, showing no trace of the postero-internal cusps
which characterize Diplacodon elatus.
Other striking peculiarities are the upward-arching mid-
cranial region, the extremely long, narrow, and laterally de-
curved nasals; the strong infraorbital shelf upon the molars
[malars] (seen also in T. megarhinum) , the slender zygomatic
arch, the low occiput, the backward extension of the posterior
nares by the palatines, and the partial inclosing of the roof of
the pharynx by the pterygoids.
Sphenocoelus Osborn, 1895
Cf. Sphenocoelus Osborn, this monograph, page 417
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
7, pp. 98-102, 1895 (Osborn, 1895.98).
Generic characters. — Osborn writes:
The distinctive features of the skull may therefore be summed
up as follows: Deep paired pits in the aUsphenoids, and orbito-
sphenoids upon either side of the thin presphenoid [basisphenoid] ;
a long alisphenoid canal; foramen ovale widely separated from
Figure 110. — Type (holoty
Side, front, and top views of skull. Am. Mus.
This general description of character was followed
by a more detailed description.
Etymology. — cornutus, horned; in allusion to the
osseous "horns."
Present determination. — Comparison of the lower
jaw with the type of Palaeosyops hyognathus Osborn
indicates that the species T. cornutum is a synonym of
P. hyognathus, a species which is now referred to the
genus Dolichorhinus.
pe) of Tebnatotherium cornutum
1851. After Osborn, 1S95. One-fourth natural size.
1 for. lac. medium; condjdes very broad; foramen magnum
large; occipital crest extending anteriorlj' into a short sagittal
crest with convex sagittal ridges; skull apparently long and
narrow.
Etymology. — <T4>riv, a wedge; koIXos, hollow; in allusion
to the paired cavities in the basisphenoid bone.
Present determination. — This is a valid genus of
Eocene titanotheres related to the long-skulled
Dolichorhinus. (See p. 417.)
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
175
Sphenocoelus uintensis Osborn, 1895
Cf. Sphenocoelus uintensis Osborn, this monograph, page 419
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
7, pp. 98-102, figs. 12-15, 1895 (Osborn, 1895.98).
Type locality and geologic horizon. — Northeastern
Utah; Metarhinus zone (Uinta B 1).
Holotype. — "Represented by the posterior portion
of a slvull" (Am. Mus. 1501). (See fig. 111.)
convex sagittal ridges. The occiput is rather broad, and below
it are two widely set occipital condyles which are directed
obliquely downward and backward. On either side of these
the exoccipitals extend down into obtuse paroccipital processes,
which are closely joined to the post-tympanics. The external
auditory meatus is open inferiorly. In front of this the post-
glenoid process faces somewhat inward; the glenoid facet is
L-shaped, two narrow arms extending out upon the squamosal,
and a broad arm descending upon the postglenoid. The dis-
tinctive feature of the zygoma is the presence of a deep depres-
sion just behind the lateral arm of the glenoid facet.
FiGUKE 111. — Type (holotype) of Sphenocoelus uintensis
Posterior half of cranium. Am. Mus. 1601. After Osborn, 1895.
, Basal view; Aa, top view; As, occipital view;
natural size.
fiew of left side. One-third
Specific characters. — Osborn writes:
This new genus is represented by the posterior portion of a
skull, which is distinct from any cranium known to the writer.
Its most distinctive feature is the presence of a pair of pits in
the floor of the skull upon either side of the narrow presphenoid
[basisphenoid]. These pits were at first mistaken for the for.
lac. media, but more careful investigation shows that they are
roofed over by bone and apparently do not communicate at
all with the cranial cavity. The pit on the right side is per-
fectly preserved and clearly exhibits these characters. The
pits are 42 millimeters long, 14 millimeters wide, and' 2 milli-
meters deep.
The skull has a long, narrow cranium surmounted posteriorly
by a sagittal crest, which diverges anteriorly into two decidedly
Skull measurements
Millimeters
Width across zygomatic arches 230
Height of occiput 142
Breadth 117
Breadth of occipital condyles 130
Basioccipital to top of sagittal crest 114
The foramina of the skull are related to those of the Peris-
sodactyla, for there is a long alisphenoid canal, upon the outer
side of the anterior opening of which is the foramen. Just
behind the posterior opening of the canal is the foramen ovale,
and between these foramina are the two pits above mentioned.
This foramen is separated by a very wide plate of bone from the
for. lac. medium, which is partly filled by the periotic mass.
176
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Etymology. — uintensis, from Uinta, in allusion to the
Uinta Basin.
Present determination. — This is a valid genus and
species. (See p. 419.)
Protitanotherium Hatcher, 1895
Cf. Protitanoiheriiim Hatcher, this monograj)h, page 374
Original reference. — Am. Naturalist, vol. 29, p. 1084,
December, 1895 (Hatcher, 1895.1).
would seem that Professor Marsh's conclusion is entirely con-
jectural, since his material does not show whether there were
horns or not. The present skull has a well-developed pair of
frontonasal horns, and, since it agrees in all the characters
known to that genus, I have preferred to refer it to that genus
rather than to propose for it a new one on the strength of this
purely conjectural character ascribed to Diplacodon by Pro-
fessor Marsh. Should future discoveries show that there are
hornless forms with the same dental characters as Diplacodon,
it will then be necessary to establish for the present specimen a
new genus, which may be called Protitanotherium.
FiGUKE 112. — Type (holotype) of Diplacodon emarginatus
Facial part of skull and anterior part of mandible. Princeton Mus. 11242. After Hatcher, 1895. Ai, Az, Aj, Side, top, and front views of
skull; Bi, B2, B3, side, top, and front views of mandible. One-fourth natural si:e.
Type species. — Diplacodon emarginatus Hatcher.
Generic characters. — Hatcher writes:
In referring this skull to Diplacodon, I have been compelled
to ignore certain characters ascribed to that genus by Professor
Marsh. That author, in speaking of the relations of this genus
to the Titanotheriidae (Brontotheridae) , in his original descrip-
tion of the type specimen, says (Marsh, 1875.1, p. 24) : "From
this family, Diplacodon differs widely in its dentition and the
absence of horns." In describing Diplacodon as hornless, it
Etymology. — pro, before; Titanotherium — that is,
forerunner of Titanotherium.
Present determination. — It is not yet settled whether
Diplacodon elatus Marsh had horns or not, but it is
now believed that even if this character is set aside
D. elatus is generically distinct from D. emarginatus,
and we may therefore regard Hatcher's Protitanothe-
rium as a valid genus.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
177
Diplacodon emarginatus Hatcher, 1895
Cf. Protitanotherium emarginatum Hatcher, this monograph,
page 377
Original reference. — Am. Naturalist, vol. 29, pp.
1084-1087, pi. 38, figs. 1-4, December, 1895 (Hatcher,
1895.1).
Type locality and geologic horizon. — Found by J. B.
Hatcher "near the base of the Diplacodon elatus beds
[Uinta C of Osborn], in the upper Eocene or Uinta of
Marsh. The locality is about 8 miles north of White
River and 25 miles east of Ouray Agency, Utah, and is
locally known as Kennedy's Hole."
Holotype. — A skull with lower jaw (Princeton Mus.
11242). The anterior part only of the skull is well
preserved. (See fig. 112.)
Characters. — Hatcher writes:
The present species is at once clistinguished from D. elatus by
its greater size, as is shown by a comparison of the length of the
premolar and molar series, which is* 310 millimeters in the
former and 242 in the latter.
In general appearance the cranium of D. emarginatus is re-
markably like some of the smaller forms of Titanotherium.
Etymology. — emarginatus, referring to the emargi-
nate form of the distal end of the nasals.
Present determination. — The species is valid but
generically distinct from Diplacodon Marsh and is now
referred to ProtitanotTierium Hatcher. (See p. 377.)
Manteoceras Hatcher, 1895
Cf. Manteoceras Hatcher, this monograph, page 362
Original reference. — Am. Naturalist, vol. 29, p. 1090,
1895 (Hatcher, 1895.1).
Type species. — By designation Telmatotherium valli-
dens (of Osborn, not Palaeosyops vallidens Cope) =
Manteoceras manteoceras Hay ex Osborn, MS., 1902.
Hotelier's description — Hatcher writes:
The genus Telmatotherium as it now stands should be divided,
since it embraces at least three distinct forms. The type of T.
vallidens should be removed from that genus and made the
type of a new genus. This new genus may be called Man-
teoceras, as suggested by Wortman from the field; it would be
distinguished from Telmatotherium by the absence of the infra-
orbital shelf, the stronger and more expanded zygomata, and
the concave superior aspect of the skull and incipient fronto-
nasal horns.
In the above passage the reference to "the type of
T. vallidens" if taken by itself would lead one to regard
"Palaeosyops" vallidens Cope as the type of the genus
Manteoceras Hatcher. But a careful study of
Hatcher's full text and a knowledge of the history
of the subject proves that Hatcher had in mind the
"Telmatotherium vallidens" of Osborn, not of Cope:
because (a) Hatcher refers to his Plate 29, Figure 2,
as "Telmatotherium vallidens," and this figure is
copied from Osborn's " Telmatotherium vallidens,"
Figure 7; (6) these figures represent Wortman's
original "prophet horn" skull, to which he had
applied the name Manteoceras "in a letter from the
field" (Osborn); (c) the generic characters assigned by
Hatcher refer most clearly to this skull and are utterly
inapplicable to Telmatotherium {"Palaeosyops") valli-
dens Cope, in which only the dentition and not the
skull is known.
Thus the type of the genus Manteoceras Hatcher is
Telmatotherium vallidens of Osborn not Cope, which
is equivalent to Manteoceras manteoceras Hay ex
Osborn MS. The generic name can not be credited
to Wortman, because he never published it, although
Osborn (1895.98), mentions it as a manuscript name.
Etymology. — juavrtj, prophet; Ktpas, horn; in allu-
sion to the incipient "horns" above the orbits.
Present determination. — This valid genus is fully de-
scribed on page 362.
Dolichorhlnus Hatcher, 1895
Cf. DoKchorhinus Hatcher, this monograph, page 396
Original reference. — Am. Naturalist, vol. 29, p. 1090,
1895 (Hatcher, 1895.1).
Type species. — Telmatotherium cornutum Osborn.
Characters. — Hatcher writes:
The genus Telmatotherium as it now stands should be divided,
since it embraces at least three quite distinct forms * * *
The type of T. cornutum should also be made the type of a new
genus which may be called Dolichorhinus; it would be dis-
tinguished from Manteoceras and Telmatotherium by the
reduced number of inferior incisors, presence of incipient
horns, presence of infraorbital shelf, and position of posterior
nares.
Etymology . — SoXixos, long; pis, nose.
Present determination. — This is a valid genus.
p. 396.)
(See
Palaeosyops ultimus Matthew, 1897 (ex Osborn MS.)
Cf. Telmatherium ultimum Osborn, 1908, this monograph,
page 345
Original reference. — Am. Naturalist, vol. 31, pp.
57-58, 1897 (Matthew, 1897.1).
Subsequent reference. — Bibliography and catalogue
of the fossil Vertebrata of North America: U. S.
Geol. Survey Bull. 179, p. 631, 1902 (Hay, 1902.1).
Doctor Matthew had no intention of describing a
new species. He merely stated incidentally that
P. ultimus, as established in manuscript by Osborn,
and P. paludosus both have a short-necked astragalus.
No type was mentioned, and the single character
given does not separate the species from P. paludosus.
Hence "Palaeosyops ultimus Matthew" (cited by
Hay, 1902, p. 631) remained a nomen nudum until
the type was fixed by Osborn in 1908. (See p. 345.)
Etymology. — ultimus, last, latest; in allusion to the
relatively late geologic horizon and to the apparent
extinction of the race.
Palaeosyops manteoceras Matthew, 1899 (ex Osborn MS.)
Cf. Manteoceras manteoceras Hay, this monograph, page 395
Original reference. — Am. Mus. Nat. Hist. Bull.,
vol. 12, p. 47, 1899 (Matthew, 1899.1).
178
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
In this faunal list the present specific name is
merely mentioned. No type is specified, and no
characters are given, so that Palaeosyops manteoceras
Matthew was a nomen nudum until the type was
designated by Hay in 1902. (See p. 365.)
Etymology. — /xavris, prophet; Ktpas, horn; in allusion
to the incipient "horns" above the orbits.
Telmatotherium diploconum var. minus Matthew, 1899
(Nomen nudum)
Original reference. — Am. Mus. Nat. Hist. Bull.,
vol. 12, p. 50, 1899 (Matthew, 1899.1).
In Matthew's faunal list " Telmatotherium diploconum
var. minus" is recorded but not defined, and no type
is specified. Hence Telmatotherium diploconum minus
is a nomen nudum.
Etymology. — minus, less; because smaller than the
tyjDical T. diploconum.
Canis? marshii Hay, 1899
Of. Palaeosyops major? Leidy, this monograph, page 321
Original reference. — Science, new ser., vol. 10, p. 253,
1899 (Hay, 1899.1). Founded on "Canis montanus"
Marsh (see p. 158), which was preoccupied by Canis
montanus Pearson.
Etymology. ^-'Named in honor of Prof. O. C. Marsh.
Present determination. — As explained above, the
type of Canis montanus Marsh (not Pearson) and
Canis? marshii Hay is a second lower premolar of an
Eocene titanothere, probably Palaeosyops paludosus
or P. major. Canis? marshii Hay is therefore either
indeterminate or a synonym of P. paludosus or P.
major.
Manteoceras manteoceras Hay, 1902 (ex Osborn MS.)
Cf . Manteoceras manteoceras Hay, this monograph, pages 365-370
Original reference. — U. S. Geol. Survey Bull. 179,
p. 632, 1902 (Hay, 1902.1).
Lectotype. — A skull (Am. Mus. 1569) lacking the
dentition, described and figured by Osborn as "Telma-
totherium vallidens" (Osborn, 1895.98, pp. 87-90, figs.
7-8). (See fig. 113.)
Paratype. — An incomplete skull (Am. Mus. 1570)
with dentition (op. cit., fig. 9) from the same locality
and level.
Type locality and geologic horizon. — Washakie Basin,
Wyo.; discovered by J. L. Wortman, of the American
Museum Bridger expedition of 1893, "in a brown
layer of sandstone 3 miles north of the base of Haj''-
stack Mountain, upon Bitter Creek" (op. cit., p. 87).
Uintatherium- Manteoceras- Mesatirhinus zone (Wash-
akie A).
. Hay's type. — We have seen above that the name
Palaeosyops manteoceras Matthew (ex Osborn MS.)
was a nomen nudum, because no type had been desig-
nated. The type was for the first time clearly indi-
cated by Hay (1902.1, p. 632), who refers to this
species Hatcher's (1895.1) Plate 39, Figure 2 (p. 368,
this monograph) and Osborn's (1895.98) Figures 7-9
(pp. 366, 368). These are clearly the same two "prophet
horn" skulls (Wortman's first "Manteoceras" speci-
mens) that had been at first erroneously referred by
Osborn to "Telmatotherium vallidens Cope." Of these
two skulls. Am. Mus. 1569 — that is, Osborn's Figures
7, 8 and Hatcher's Figure 2 (copied from Osborn's
fig. 7) — may be taken as the lectotype.
The generic name Manteoceras and the specific
name manteoceras were first brought together by Hay
in the reference now under consideration.
Specific characters. — In Osborn's original descrip-
tion (Osborn, 1895.98, p. 87) these skulls were errone-
ously identified as conspecific with the type of Palaeo-
syops vallidens Cope, under the name "Telmatotherium
vallidens Cope." The specific characters given by
Osborn were as follows:
Superior premolar-molar series, 184-220 millimeters. A
narrow diastema. Molar cusps less elevated. A rudimentary
nasofrontal tuberosity. Premaxillary symphysis short. Top
of cranium flattened; very short bifid sagittal crest.
Etymology. — fiavrLs, prophet; Kepas, horn; in allu-
sion to the incipient "horns" above the orbits.
Present determination. — The species is a valid one
and is fully described on pages 365-370.
Lambdotherlum primaevum Loomis, 1907
Cf . Lambdotherium -primaevum Loomis, this monograph, page 283
Original reference. — Am. Jour. Sci., 4th ser., vol. 23,
p. 363, fig. 2, May, 1907 (Loomis, 1907.1).
Type locality and geologic horizon. — Buffalo Basin,
near Meeteetse, Wyo. "Wasatch beds of the Big
Basin." Horizon regarded by Loomis as equivalent
to the base of the Wind Eiver formation — that is,
the Heptodon-Coryphodon-Eohippus zone (Wind River
A).
Holotype. — Amherst Mus. 254, "consisting of upper
molars 1 and 2 of the right side and lower molars 1, 2,
and 3 from the same side, the specimen being from the
Buffalo Basin, near Meeteetse, Wyo. This species is
fau'ly abundant at this horizon and is intermediate in
size between L. hrownianum and L. popoagicum."
(See fig. 114.)
Characters. — Loomis writes:
On the upper molars the parastyle, though strong, is not so
well developed as in the foregoing forms; the paraoonule is well
developed, but the metaconule is so annexed to the metaoone as
to appear like a buttress of this cusp. The second molar
measures 12 milUmeters transversely [anteroposteriorly] by 17
millimeters lengthwise [transversely]. The robust lower molars
have the protoconid markedly bifid, while the paraconid and
hypoconid are each high crescents. The heel of the last molar
is a high shallow basin completely surrounded by an outer rim.
The three molars occupy 41 milhmeters.
The brackets above indicate that in the foregoing
description the measurements of the molar teeth
have been inadvertently transposed. The description
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
179
should read: "Second superior molar, antei'oposterior,
12 millimeters; transverse, 17 millimeters."
Etymology. — primaevus, earliest in age; in allusion to
the supposedly low geologic horizon.
Present determination. — Provisionally recognized as
a valid species.
Mus. Nat. Hist. Bull., vol. 24, pp. 600, 601, 1908
(Osborn, 1908.318).
Type species. — Palaeosyops horealis Cope.
Generic characters (Osborn, 1908.318, p. 601). —
Superior molars subquadrate and rounded in form;
conules reduced, sublophoid; m'-m^ 63 millimeters
Figure 113. — Cotypes of Manteoceras manteoceras {Telmalotheriam vallidens)
After Osborn. Ai, Composite Am. Mus. 1569, 1570; side view otslsull; As, Am. Mus. 1669 (lectotype), superior view ot slcull; B, Am.
Mus. 1570, superior view of slcull. All one-fourth natural size.
Eotitanops Ogborn, 1907
Cf. Eotitanops Osborn, this monograph, page 289
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
23, p. 242, 1907; type species designated (Osborn
1907.294).
Subsequent reference. — Osborn, New or little-known
titanotheres from the Eocene and Oligocene: Am.
(estimated). Inferior molars without metastylids.
Hypoconulid of m' subconic. Fii'st inferior premolar
present. Manus tetradactyl, functionally tridactyl
with a tendency to mesaxonic structure. From Wind
River formation.
Etymology. — ^cbs, dawn; Tltclv, a titan; w^, face — ■
that is, first of the titanotheres.
Present determination. — This genus is valid. (See
p. 289.)
180
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Limnohyops priscus Osborn, 1908
Cf. Limnohyops priscus Osborn, this monograph, page 306
Original reference. — Am. Miis. Nat. Hist. Bull.,
vol. 24, pp. 601-602, fig. 5, 1908 (Osborn, 1908.318).
Type locality and geologic horizon. — Grizzly Buttes,
Bridger Basin, Wyo.; Palaeosijops paludosus-Oroliippus
zone (Bridger B 2).
A
Figure 114. — Type (holotype) of Lambdotherium
primaevum
Amherst Mus. 254. After Loomis, 1907. A, Right upper molars 1 and
2; B, right lower molars (mi-ms). Natural size.
Holotype. — A crushed skull with excellent dentition
(Am. Mus. 11687), discovered by the American
Museum expedition of 1903. (See fig. 115.)
Figure 115. — Type (holotype) of Limnohyops priscus
Am. Mus. 11687. Pi-m' left. After Osborn, 1908. One-half natural size.
Characters. — Osborn writes:
P'-m', 148 (type) to 161 millimeters. Distinguished from
the contemporary Limnohyops laevidens Cope by its larger
size and by the more progressive character of pm^-pm'.
Second superior premolar obliquely elongate with a very rudi-
mentary tritocone. Large hypocone on m'.
Etymology. — priscus, ancient, in allusion "to the
low geological level and primitive characters of this
species." (Osborn.)
Present determination. — The species and generic
reference are probably valid. (See p. 306.)
Figure 116.-
-Tj'pe (holotype) skull of Limnohyops
matthewi
Am. Mus. 11684. After Osborn, 1908. One-fourth natural size.
Limnohyops matthewi Osborn, 1908
Cf. Limnohyops matthewi Osborn, this monograph, page 308
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, p. 602, fig. 6, 1908 (Osborn, 1908.318).
Type locality and geologic horizon. — Grizzly Buttes,
Bridger Basin, Wyo. ; Palaeosyops paludosus-Orohippus
zone (Bridger B 2).
Holotype. — A skull (Am. Mus. 11684) lacking the
anterior portion and dentition. Discovered by the
American Museum expedition of 1903. (See fig. 116.)
Specific characters. — Osborn writes:
Intermediate in size between L. laevidens and L. monoconus.
M' of small size with large hypocone and quadrate inner half
Occiput very high and narrow. Cranial portion of skull greatly
abbreviated, bringing post-tympanic and postglenoid processes
into broad union. Temporal openings subcircular as defined
by zygomatic arches.
Etymology. — Named "in honor of Dr. W. D.
Matthew, of the American Museum staff." (Osborn.)
Present determination. — The species is probably
valid. (See p. 308.)
Figure 117. — Type (holotype) skull of Limnohyops monoconus
Am. Mus. 11679. After Osborn, 1908. One-fourth natural size.
Limnohyops monoconus Osborn, 1908
Cf. Limnohyops monoconus Osborn, this monograph, page 309
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, p. 603, fig. 7, 1908 (Osborn, 1908.318).
Type locality and geologic horizon. — Grizzly Buttes,
Bridger Basin, Wyo. ; Palaeosyops paludosus-Orohippus
zone (Bridger B 2).
Holotype. — A crushed skull with dentition (Am.
Mus. 11679). Discovered by Mr. Quackenbush, of
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
181
the American Museum expedition of 1903. (See fig.
117.)
Specific characters. — Osborn writes:
M^ without hypocone, roundly triangular in form, with
broadly extended ectoloph and parastyLe. P^-m' 150, p'-m^
163 milUmeters. Condyle to incisive border 510. Occiput
very high, cranium relatively elongated, with space (4 miUi-
meters) between post-tympanic and postglenoid processes.
Temporal openings as defined by zygomatic arches elongate.
Figure 118. — Type (holotype) skull of Palaeosyops leidyi
Inferior view. Am. Mus. 1544. After OsbDrn, 1908. One-fourth natural size.
Etymology. — yibvo's, single; kccws, cone; named "in
reference to the presence of but a single cone on the
inner side of the third superior molar, an exceptional
condition in the genus Limnohyops." (Osborn.)
Present determination. — The specific and generic
references are probably valid. (See p. 309.)
Palaeosyops leidyi Osborn, 1908
Cf. Palaeosyops leidyi Osborn, this monograph, page 323
Original reference. — Am. Mus. Nat. Hist. Bull.,
vol. 24, p. 604, fig. 8, 1908 (Osborn, 1908.318).
Type locality and geologic horizon. — Henrys Fork,
Bridger Basin, Wyo.; Uintatherium- Manteoceras- Mesa-
tirliinus zone (Bridger C 2 to C 4). Discovered by the
American Museum expedition of 1893, under Dr.
J. L. Wortman.
Holotype. — A well-preserved skull (Am. Mus. 1544)
associated with considerable portions of the skeleton.
This specimen, which is associated with a considerable
portion of the postcranial skeleton, is now mounted
in the American Museum, the missing parts having
been supplied from other individuals. (See p. 323;
Pis. XXVII, L, LXI; and fig. 118.)
Specific characters. — Osborn writes:
Of larger size; total length of skull 415 millimeters; p'-m^, 158;
P2-m3, 168; diastema behind canines; p-', p^ superior, with
mesostyles. Barely defined sweUings representing the rudi-
ments of osseous frontonasal horns.
Etymology. — Named "in honor of Joseph Leidy,
the discoverer of the family and [founder] of the
genera Palaeosyops, Titanotherium, and Megacerops."
(Osborn.)
Present determination. — The species is probably
valid.
Palaeosyops granger! Osborn, 1908
Cf. Palaeosyops grangeri Osborn, this monograph, page 335
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, p. 604, fig. 9, 1908 (Osborn, 1908.318).
Type locality and geologic horizon. — Twin Buttes,
Bridger Basin, Wyo. ; Uintatherium- Manteoceras- Mesa-
tirhinus zone (Bridger C 1).
Holotype. — A palate and grinding teeth with por-
tions of the lower jaw and skull (Am. Mus. 12189),
American Museum expedition of 1904. (See fig. 119.)
Specific characters. — Osborn writes:
Exceeding P. robustus in certain dental proportions; p^-m^,
165 millimeters. Fourth superior premolar enlarged (trans-
verse, 31 mm.). Molars with extremely prominent parastyles
and oblique ectolophs.
Etymology. — Named "in honor of Mr. Walter Gran-
ger, of the American Museum staff, whose explora-
tions have transformed our knowledge of the Bridger
animals. " (Osborn.)
Present determination. — The species is probably
^alid. (See p. 335.)
Figure 119. — Tj'pe (holotype) of Palaeosyops grangeri
; maxillary with p'-mi. Am. Mus. 12189. After Osborn, 1908. One-half natural size.
Palaeosyops copei Osborn, 1908
Cf. Palaeosyops copei Osborn, this monograph, page 336
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, p. 606, fig. 10, 1908 (Osborn, 1908.318).
Type locality and geologic horizon. — Lone Tree Hen-
rys Fork, Bridger Basin, Wyo.; Uintatherium- Man-
teoceras-Mesatirhinus zone (Bridger D 3).
Holotype. — A series of superior grinding teeth (Am.
Mus. 11708). (See fig. 120.)
182
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Specific characters. — Osborn writes:
Of more diminutive size (p'-m', 153 mm.), but the miost
progressive species of Palaeosyops known in the evolution of its
superior premolars and molars. Heavy oingula embracing the
inner sides of the crowns. A rudimentary tetartooone on p^.
Etymology. — Named "in honor of the late Prof.
E. D. Cope, the describer of Lambdotherium, 'Palaeo-
FiGTjRE 120. — Type (holotype) of Palaeosyops copei
P'-ms, right. Am. Mus. 11708. After Osborn, 1908. One-half natural size.
syops' horealis, and other species of Eocene titano-
theres." (Osborn.)
Present determination. — The species is probably
valid.
Manteoceras washakiensis Osborn, 1908
Cf . Manteoceras washakiensis Osborn, this monograph, page 371
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, p. 607, fig. 11, 1908 (Osborn, 1908.318).
Type locality and geologic Tiorizon. — Base of Hay-
stack Mountain, Washakie Basin, Wyo.; summit of
UintatJierium- Manteoceras- Mesatirliinus zone (Washa-
kie A).
Holotype. — A well-preserved skull with dentition
(Am. Mus. 13165). Discovered by Mr. Paul Miller,
of the American Museum expedition of 1906. (See
fig- 121.)
Specific characters. — Osborn writes:
Distinguished from M. manteoceras of a somewhat lower
geological level by its more progressive characters, as follows:
Canines short, obtuse, recurved; internal lobes of pm^, pm^
broadening, with shelf for development of deuterocone; p^ (ap.
19 mm., tr. 17) with marked external convexities and a re-
duced external cingulum; p^ (ap. 19, tr. 25) exhibits
the tetartocone fold somewhat more conspicuously
than in the most progressive Bridger level D speci-
mens. ?■* (ap. 24, tr. 30) is progressive in transverse
measurement and in the development of the tetar-
tocone shelf. The molars are progressive in their
large size (m^ ap. 42, tr. 48), in the strong develop-
ment of the internal cingulum, and in the elongate
ectolophs.
Etymology. — washaJciensis; "so named be-
cause it is a more recent phase, probably
characteristic of the Washakie rather than of
the Bridger." (Osborn.)
Present determination. — The species and the generic
reference are valid. (See p. 371.)
Mesatirhinus Osborn, 1908
Cf. Mesatirhinus Osborn, this monograph, page 387
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, p. 608, 1908 (Osborn, 1908.318).
Type species and geologic horizon. — Palaeosyops
megarhinus Earle. Bridger Basin, Wyo., levels Bridger
C and D; Washalcie Basin, Wyo., levels Washakie A
and base of Washakie B.
Specific characters. — Osborn writes:
Titanotheres of small size (skull length 354^425 mm.)
typically mesaticephalic, persistent or progressing to dolicho-
cephalic. The horns when present incipient or rudimentary,
chiefly borne on the nasals. An infraorbital shelf. Cranium
with a sagittal crest. Humerus relatively abbreviated — that
is, with refeYence to Palaeosyops — carpus and tarsus narrow,
astragalus with elongate neck, the sustentacular distal and
ouboidal facets continuous and forming a reversed L (j) ; meta-
podials slender.
Etymology. — fiiaaros, middle; pis, nose; because the
length of the snout is moderate compared with that
in the allied genus Dolichorhinus.
Present determination. — The genus is valid. (See
p. 387.)
Mesatirhinus petersoni Osborn, 1908
Cf. Mesatirhinus petersoni Osborn, this monograph, page 389
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, p. 608, fig. 12, 1908 (Osborn, 1908.318).
Holotype. — A skull with dentition (Am. Mus. 12184)
from Cattail Spring, Bridger Basin, Wyo., levels
Bridger D 3 and C 3. (See fig. 122.) The species is
also recorded from Washalde Basin, Wyo., level
Washakie A.
Specific characters. — Osborn writes:
Pm'-m^, 156 millimeters; m'-m', 90. Skull length, pre-
maxillaries to condyles 412 (estimated) ; preorbital facial region
more elongate (217). Other characters as in Mesatirhinus
megarhinus — that is, broad occipital condyles, broad infra-
orbital shelf on malar, etc.
Comparison of this animal with the type of M. megarhinus
can leave no doubt that we have to do here with a much more
advanced stage of evolution. The skull is longer, the pre-
orbital region especially. The grinding teeth occupy more
space, and there is an average advance in all the rectigradations
which proves that these differences in form and size are not
merely due to fluctuations of size or differences of sex.
ffK?-
Figure 121. — Type (holotype) skull of Manteoceras washakiensis
Left side. Am. Mus. 13165. After Osborn, 1908. One-flfth natural size.
Etymology. — "The species is named in honor of Mr.
O. A. Peterson, now of the Carnegie Museum, whose
titanothere collections in the Uinta formation greatly
extended our knowledge." (Osborn.)
Present determination. — The species and generic ref-
erence are valid. (See p. 389.)
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
183
Metarhinus Osborn, 1908
Cf. Metarhinus Osborn, this monograph, page 420
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, p. 609, 1908 (Osborn, 1908.318).
Type species and geologic liorizon. — Metarhinus flu-
viatilis Osborn. Washakie Basin, Wyo., level Washa-
kie B; Uinta Basin, Utah, levels Uinta B 1 and B 2.
Specific charac-
t er s . — Osborn
writes :
Small tltanotheres
(skull length 355 to 440
mm.), persistently mes-
aticephalic. Narrow,
abbreviated preorbital
region, premaxillary
symphysis greatly
elongated, and anterior
narial openings deeply
recessed in side view.
Infraorbital shelf pres-
ent, or wanting (M.
diploconus) ; occipital
condyles narrow.
Grinding teeth sub-
hypsodont; premolars
progressive; hypoco-
nulid of ms small,
conic.
Etymology. — iiera,
after; rJiinus (that
i s , MesatirMnus) .
"The name alludes
to the somewhat
later geological ap-
pearance of this
genus as compared with MesatirMnus." (Osborn.)
Present determination. — The genus is valid. (See
p. 420.)
Metarhinus fluviatilis Osborn, 1908
Cf. Metarhinus fluviatilis Osborn, this monograph, page 421
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, p. 609, fig. 13, 1908 (Osborn, 1908.318).
Figure 122. — Type (holotype) skull
of MesatirMnus petersoni
Top view. Am. Mus. 12184. After Osborn, 1908.
One-fourth natural size.
Figure 123. — Type (holotype) skull of Metarhinus fluviatilis
Lett side. Nasals broken off.
Am. Mus. 1500.
natural size.
After Osborn, 1908. One-fourth
Type locality and geologic liorizon. — Uinta Basin,
Utah; Metarhinus zone (Uinta B 1).
Holotype. — A skull (Am. Mus. 1500) discovered by
the American Museum expedition of 1894 in horizon
B 1 of the Uinta Basin. (See fig. 123.)
Specific characters. — Osborn writes:
Pm'-m^ — 144 millimeters. A relatively short (355 mm.,
estimated), broad (200 mm., estimated) skull. Eye sockets
small and very prominent. Premaxillary symphysis elongate,
grinding teeth subhypsodont, m' with a cingulum-hypocone in
the type.
Etymology. — fluviatilis, fluviatile. "The name is
given in allusion to the possibly river-living or am-
phibious habits of the animal." (Osborn.)
Present determination. — The species and the generic
reference are valid. For fuller specific distinctions
see page 421.
Metarhinus earlei Osborn, 1908
Cf. Metarhinus earlei Osborn, this monograph, page 426
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, p. 610, fig. 14, 1908 (Osborn, 1908.318.)
Figure 124. — Type (holotype) .skull of
Metarhinus earlei
Top view. Am. Mus. 13166. After Osborn, 1908.
One-fourth natural size.
Type locality and geologic horizon. — North side of
Haystack Mountain, Washakie Basin, Wyo.; Meta-
rhinus zone (Washakie B 1 ) .
Type. — A skull (Am. Mus. 13166) lacking the nasals,
American Museum expedition of 1906. (See fig. 124.)
Specific characters. — Osborn writes:
Pmi-ni' = 167 millimeters. Skull proportions, length 380,
breadth 230. Narrow occipital condyles. Extremely elongate
premaxillar}' symphysis. A short sagittal crest. No hypocone
on m^.
This animal is readily distinguished from M. diploconus by
(1) the infraorbital shelf of the malars; (2) the elongate premaxil-
184
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
lary; (3) the absence of a double cone on m^ In many other
respects it resembles M. diploconus, especially in its proportions.
It is distinguished from M. megarhimis by (1) the elongate
premaxillary symphy-
sis, correlated with the
long, narrow facial re-
gion; (2) the narrowness
of its occipital condyles.
It is distinguished from
M. fluviatilis by (1) its
greatly superior size
and (2) the lesser prom-
inence of the orbits.
Etymolo g y. —
Named "in honor of
Charles Earle, the
first monographer of
the genus Palaeo-
syops and its allies."
(Osborn.)
Present determina-
tion.— The species
is probably valid.
(See p. 426.)
Dolichorhinus interme-
dius Osborn, 1908
Cf. Dolichorhinus inter-
niedius Osborn, this
monograph, page 405
Original refer-
ence.— Am. Mus.
Nat. Hist. Bull., vol.
24, p. 611, fig. 15,
1908 (Osborn,
1908.318).
Type locality and
Top view. Am. Mus. 1837. After Osborn, 1908. neoloqic llOrizOn. —
One-fourth natural size. .^t. , -r» • i.^
Umta Basin, north-
eastern Utah; Eobasileus - Dolicliorhinus zone
(Uinta B 2).
Hohtype. — A skull with dentition (Am. Mus. 1837),
discovered by the American Museum expedition of
1894. (See fig. 125.)
Specific characters. — Osborn writes :
Distinguished from D. hyognathus Scott and Osborn by (1)
its inferior size (pm'-m^ 179, m'-m' 109 mm.);
(2) premolars less progressive, with subconic
deuterocones; (3) all oingula less robust; (4)
nasals more pointed and less expanded distally;
(5) infraorbital shelf of malar relatively narrow.
Etymology. — "The name 'intermedins'
is given because in some characters this
species is intermediate between Mesati-
rhinus petersoni and DolichorMnus Tiyo-
gnathus, although on the whole it is Figure 126. — Type (holotype) skull of Telmatheriiim ultimum
much more nearly allied to the latter." side ™w. Am. Mus. 2O6O. After Osbom, 19O8. One-flfth natural size. The skull has been some-
//-j V s what deformed by pressure.
Present determination. — The generic reference ap- I pointed. P,, pz laterally compressed, nonmolariform; ps, p4
pears certain; the species is probably valid. (See submolariform; dolichocephalic, anterior portion of face
p. 405.) I elongate.
Figure 125. — Type (holotype) skull of
Dolichorhinus inter medius
Telmatherium ultimum Osborn, 1908
Cf . Telmatherium ultimum Osborn, this monograph, page 345
Original reference. — Am. Mus. Nat. Hist. Bull.,
vol. 24, p. 613, fig. 17, 1908 (Osborn, 1908.318).
Type locality and geologic horizon. — Uinta Basin,
northeastern Utah; Diplacodon-Protitanotherium-Epi-
hippus zone (Uinta C, lower levels).
Holotype. — A well-preserved skull with dentition
(Am. Mus. 2060). Discovered by Mr. Peterson, of
the American Museum expedition of 1895. (See fig.
126.)
Synonymy. — This species was mentioned by Mat-
thew as Palaeosyops ultimus Osborn MS. (see p. 177),
but as no type was indicated or specific diagnosis
given the name remained a nomen nudum until a type
was designated and a diagnosis given by Osborn in
1908.
Specific characters. — Osborn writes:
P'-m*, 226 mm. Lateral superior incisors greatly en-
larged, caniniform. Pm^- ^. * with Internal subcrescentic
deuterocone ridges, with faint rudiments of tetartocones
posteriorly. Ectolophs of premolars elevated and biconvex.
Etymology. — ultimus, latest. "The specific name is
given because this appears to be the last representative
of the Palaeosyops-Limnohyops-Telmatherium group."
(Osborn.)
Present determination. — This species is certainly a
valid one. The grounds for regarding it as allied to
the genus Telmatherium are given on page 345.
Telmatherium? altidens Osborn, 1908
Cf . Telmatherium altidens Osborn, this monograph, page 351
Original reference. — Am. Mus. Nat. Hist. Bull.,
vol. 24, p. 614, fig. 18, 1908 (Osborn, 1908.318).
Type locality and geologic horizon. — Uinta Basin,
northeastern Utah; Diplacodon-Protitanotherium-Epi-
hippus zone (Uinta C).
Holotype. — A lower jaw with dentition (Am. Mus.
2025) discovered by the American Museum expedition
of 1895. (See fig. 127.)
Specific characters. — Osborn writes:
Pmj-mj, 330 milUmeters; a wide diastema (70 mm.) behind
the canines. Canines in male exceptionally elevated (76) and
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
185
The specific characters are more fully given on
page 351 of this monograph.
Etymology. — "The specific name refers to the high-
crowned piercing canine."
Figure 127. — Type (holotype) of Telmaiherium? altidens
Lower jaw. Am. Mus. 2025. After Osborn, 1908. One-sixth natural size.
Present determination. — The species is probably
valid. The generic reference is somewhat less certain.
(Seep. 351.)
Protltanotherium superbum Osborn, 1908
Cf. Protitanotherium superbum Osborn, this monograph, page 379
Original reference. — Am. Mus. Nat. Hist. Bull.,
vol. 24, p. 615, fig. 19, 1908 (Osborn, 1908.318).
Type locality and geologic Jiorizon. — Uinta Basin,
northeastern Utah; Diplacodon-Pi'otitanotherium-E'pi-
Jiippus zone (Uinta C) ; probably higher levels.
Holotype. — A well-preserved lower jaw with denti-
tion (Am. Mus. 2501). (See fig. 128.)
Specific characters. — Osborn writes:
Pi-m3, 318 millimeters. Canines in males very robust; pi
double fanged; postoanine diastema abbreviated; premolar
series relatively abbreviated; p2 with very large talonid and
crescentic protoconid; ps, p4 with talonid heavy and promi-
nent— that is, submolariform — but no entoconid. Ma with
hypooonulid sharply constricted off at base.
Etymology. — superbum, haughty, arrogant. "The
name is given in reference to the great size and pre-
sumed power of this Uinta titanothere, which con-
siderably exceeds that of the smaller [lower]
Oligocene titanotheres." (Osborn.)
Present determination. — The species is
probably valid. The generic reference is
somewhat less certain. (See p. 379.)
Telmatherium? incisivum Douglass, 1909
Cf. Sthenodecies incisivus (Douglass), this mono-
graph, page 354
Original reference. — Carnegie Mus. Annals,
vol. 6, No. 2, pp. 305-307, text figs. 1, 2, 3, pi.
13, fig. 1, 1909; "issued November 6, 1909"
(Douglass, 1909.1).
Type locality and geologic horizon. — Uinta
Basin, Utah, about 3 miles northeast of well
2, from "a thick deposit of sandstone and
small gravel evidently of stream origin, near the mid-
dle of horizon B." Near the summit of Eohasileus-
Dolichorhinus zone (Uinta B 2). Discovered by Mr.
J. F. Goetschius.
101959— 29— VOL 1 15
Type. — A skull, lacking the ends of the nasals
(Carnegie Mus. 2398). (See figs. 129, 130.)
Specific characters. — Douglass writes:
I think that this skull represents a different genus
from Telmatherium, but I prefer to place it provi-
sionally here rather than establish another genus. The
skull is broad and short, but not high. The forehead
is broad and flat. ' The premaxillaries are oblique,
not transverse. The face is short and concave.
Apparently there are vacuities anterior to the orbits.
Beneath these there is a rounded angle on the malar,
but there is no flattened shelf beneath the orbit. The
zygomatic arch is spreading and moderately heavy.
The sagittal crest is quite high and thin. The
superior wings of the occiput are also thin. The brain
case^is small; the outward-projecting zygomatic proc-
esses of the squamosals shelf -like and broad anteropos-
teriorly. The paroccipital processes extend laterally
and are continuous with the paramastoid processes
posterior to the external auditory meatus and the postglenoid
process. The anterior portion of the opening of the posterior
nares is between the anterior portions of the last molars.
The teeth increase quite regularly in size from p2 to m'. The
premolars have heavy cingula. The deuterocones on p" and
p^ are oblong anteroposteriorly, while that on p* is high and
conical.
Measurements [Douglass]
Millimeters
Length of skull, basal 490
Width of skull 330
Length of dental series 295
Length of molar-premolar series 212
Transverse diameter of i' 21
Anteroposterior diameter of i' 22
Transverse diameter of i^ 27
Anteroposterior diameter of i^ 25
Transverse diameter of i' 22
Anteroposterior diameter ofi^ 25
Transverse diameter of canine 24
Anteroposterior diameter of canine 27
Transverse diameter of p2 22
Anteroposterior diameter of p^ 20
Transverse diameter of p^ 30
Anteroposterior diameter of p^ 24
Transverse diameter ofp* 37
'T\ Anteroposterior diameter of p"" 27
Figure 128. — Type (holotype) of Protitanotherium superbum
Lower jaw. Am. Mus. 2501. After Osborn, 1908. One-sixth natural size.
Anteroposterior diameter ofm' 44
Transverse diameter of m^ 53
Anteroposterior diameter of m' 46
Transverse diameter of m^ 53
Anteroposterior diameter of m' 46
186
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Etymology. — incisivum, provided with incisors; in
allusion to the great size of the superior incisors.
Present determination. — The species probably repre-
sents a genus distinct from Telmatherium (see p. 353)
named Sthenodectes by Gregory.
Type. — A skull lacking the posterior portion
(Carnegie Mus. 2888). (See figs. 131 and 132.)
Specific characters. — Douglass writes:
The skull is high, the forehead broad, and the zygomatic
arches spreading. The premaxillary region as seen from the
front is broad, though the incisors are only moder-
ately large. The canines are directed outward. The
free nasals are short and moderately broad. Appar-
ently the infraorbital foramen is not excessively large.
The malar is rounded beneath the orbit and has no
protuberance or shelf. The zygomatic arch is not
very heavy and is only moderately deep anterior to
the glenoid articular surface. It is not nearly so
heavy as in Telmatherium uliimum. The opening
of the posterior nares extends forward to the middle
of the second molars. Their border is rounded and
thickened.
The incisors are moderately large but not cupped.
They are arranged in an oblique line about halfway
between a transverse and anteroposterior direction.
The crowns of i' and i^ are low. The anterior faces
are very convex. There are two posterior flattened
surfaces separated by a rounded ridge. There are
no cups, but the posterior portion forms a kind of
ledge or keel. P is higher and is directed more down-
ward. The posterior portion is flattened, and there is
a low flat ledge behind the conical cusp. The canine
has a moderately high curved crown, on which there
are antero-internal and postero-external ridges, pass-
ing downward from the base to the apex. There is
also a narrow postero-internal ledge.
Unless the skull is more crushed laterally than it
appears to be, there is a sudden contraction posterior
to the canine, so that the first two premolars are
much nearer to the median line of the palate than are
the canines. The diastema between the canine and
p' is about 3 centimeters in length.
P' is a simple oblong conical tooth, which has a
small antero-internal depression, and a small ridge
passes backward from the apex to the posterior por-
tion of the rudimentary keel. P 2, 3, and 4 have low
cusps. The teeth increase nearly uniformly in width
and size from p^ to the last molar. The two outer
elements in each are well defined and are subequal
in size, although the anterior cusp is slightly the
larger. The internal cusp on p^ is small, oblong
anteroposteriorly, and is placed far back. The inter-
nal cusp on p' is much larger and is crescent-shaped.
On p* it is more nearly conical. There are rudi-
mentary cingula on the inner faces of the last three
premolars. The postero-internal cusp on m' is repre-
sented by a low crescent-shaped ridge.
FiGUBE 129. — Type (holutype) skull of Telmatherium? incisivum
palatal view; A3,
Carnegie Mus. 2398. After Douglass, 1906. Ai, Superior view;
view. One-fifth natural size.
Manteoceras uintensis Douglass, 1909
Cf. Manteoceras uintensis Douglass, this monograph, page 372
Type reference. — Carnegie Mus. Annals, vol. 6, No.
2, pp. 307-310, text figs. 4, 5, pi. 13, fig. 4, 1909;
"issued November 6, 1909" (Douglass, 1909.1).
Type locality and geologic horizon. — Uinta Basin,
Utah, about 5 miles northeast of well 2, from "gray
sandstone in red Uinta beds. Lower portion of
horizon C." Diplacodon-Protitanotherium-Epihippus
zone (Uinta C) .
Measurements [Douglass]
Millimeters
Is, lateral Lgjjgth of skull, anterior portion to glenoid 430
Length of dental series 356
Length of molar-premolar series ^-_-- 247
Length of premolar series 106
Length of molar series 141
Transverse diameter of i' 16
Anteroposterior diameter ofii 18
Transverse diameter ofi^ 16
Anteroposterior diameter of i^ 18
Transverse diameter of i' 20
Anteroposterior diameter ofi^ 22
Transverse diameter of canine 22
Anteroposterior diameter of canine 26
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
187
Figure 130. — Type (holotype) of Telmatherium? incisivum
Lett upper teeth, crown view. After Douglass, 1909. One-half natural size.
Millimeters
Transverse diameter of p' 12
Anteroposterior diameter of p' 22
Transverse diameter of p^ 21
Anteroposterior diameter ofp^ 28
Transverse diameter of p^ 28
Anteroposterior diameter ofp' 27
Transverse diameter of p* 33
Anteroposterior diameter ofp' 30
Transverse diameter of m' 44
Anteroposterior diameter of m' 40
Transverse diameter ofm.^ 63
Anteroposterior diameter ofm^ 55
Transverse diameter ofm^ 56
Anteroposterior diameter ofm^ 51
Width of palate between canines? 68
Width of palate between first premolars 54
Width of palate between last molars 83
Etymology. — uintensis, in reference to the
Uinta Basin.
Present determination. — The generic refer-
ence to Manteqceras appears to be correct.
The species is a valid one.
Dolichorhinus heferodon Douglass, 1909
Cf. Dolichorhinus heterodon Douglass, this monograph,
page 416
Original reference. — Carnegie Mus. Annals,
vol. 6, No. 2, pp. 310-311, text figs. 6, 7,
pi. 13, fig. 3, 1909; "issued November 6, 1909"
(Douglass, 1909.1).
Type locality and geologic horizon. — Uinta Basin,
Utah, 6 or 7 miles northeast of well 2; from "upper
part of horizon B or lower part of horizon C";
Eohasileus-DolicJiorJiinus zone (Uinta B 2).
Type. — A skull lacking the front teeth and both
zygomatic arches (Carnegie Mus. 2340). (See figs. 133
and 134.) Discovered by Mr. J. F. Goetschius.
The infraorbital foramen is large. The infraorbital shelf is
represented by a protuberance, which is thickened on the free
Figure 132. — Type (holotype) of Manteoceras uintensis
Upper teeth. Carnegie Mus. 2388. After Douglass, 1909. One-third natural size.
Specific characters. — Douglas writes:
The skull is long, narrow, and moderately high. The face is
short and the brain case long. The free nasals are long, the
posterior opening of the anterior nares extending well backward
toward the orbit. The lower borders of the nasals approach
each other, but this is probably in part due to lateral crushing.
Figure 131. — Type (holotype) skull of Manteoceras uintensis
Carnegie Mus. 2388. After Douglass, 1909. Ai, Palatal view; Aj, view of right side.
One-fifth natural size.
outer surface. If there were horn cores
above the orbit they ■v\'ere very small. The
long brain case was apparently arched
from before backward, the posterior de-
scent to the crest of the occiput being very
steep, though this may be somewhat ex-
aggerated by crushing. The occipital con-
dyles are very large. The median portion
of the occiput above them is convex, while
above this there is a large concavity. The
postglenoid processes are not excessively large.
The premolars are small, the last being very decidedly
smaller than the first molar. The first premolar is not pre-
served, but it was evidently a simple tooth. In the last three
premolars there is a lobe or buttress on the antero-external
portion of the tooth, which makes the anterior margin oblique.
188
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The inner cusps (deuterocones) are low with rounded summits.
They are more nearly opposite the postero-external than the
antero-external cusi5. There are inner cingula on p' and p*.
The antero-internal cusp in m^ is quite high and m' conical.
The postero-internal cusp is due simply to an increase in height
of the cingulum.
Measurements
Millimeters
Total length of top of skull 500
From anterior orbit to front of nasals 160
Width of occiput 128
Height of occiput 140
Length of molar-premolar series 190
Length of premolar series 75
Length of molar series 115
Length of p2 20
Width of p2 16
Dolichorhinus longiceps Douglass, 1909
Cf. Dolichorhinus longiceps Douglass, this monograph, page 406
Original reference. — Carnegie Mus. Annals, vol. 6,
No. 2, pp. 312-313, text fig. 8; pi. 13, fig. 2; pis. 14,
15, 1909; "issued November 6, 1909" (Douglass,
1909.1).
Type locality and geologic horizon. — Uinta Basin,
Utah, "about 1^ miles east of well No. 2," from
Figure 133. — Type (holotype) skull of Dolichorhinus heterodon
Carnegie Mus. 2340. After Douglass, 1909. Ai, Palatal view; As, right lateral
view. One-fifth natural size.
Millimeters
Length of p3 21
Width of p3 20
Length of p< 24
Width of p^ 27
Length ofm' 34
Width of ml 35
Length ofm^ 46
Width of m2 42
Length ofm^ 48
Width of m3 42
Figure 134. — Type (holotype) of Dolichorhinus heterodon
Upper premolar series. Carnegie Mus. 2340. After Douglass, 1909. Slightly less
than one-half natural size.
"the lowest level at which fossils were
found in horizon 'B' of the Uinta, about
700 feet below the bottom of the Uinta
red beds (horizon 'C')." Eohasileus-
DolicJiorJiinus zone (Uinta B 2).
Type. — A skull lacking the incisors, part
of the dentition, and the basioccipital
region (Carnegie Mus. 2347). (See figs.
135 and 136.)
Specific characters. — Douglass writes:
Phis skull in general outline is very much like
that of Dolichorhinus hyognathus, though broader.
In describing it I prefer to point out the char-
acters which distinguish it from that species.
Apparently it is somewhat broader proportionally
than that of D. hyognathus. The skull is some-
what crushed, but it evidently was not flattened
on top. The present specimen had no heavy
protuberances or horn cores, though there may
Figure 135. — Type (holotype) skull of Dolichorhinus longiceps
Top view. Carnegie Mus. 2347. After Douglass, 1909. One-sixth natural size.
Etymology. — crepos, difl^erent, or various; bbobs, tooth.
Allusion not clear; name possibly given because no
two teeth in the superior premolar-molar series are
alike.
Present determination. — The form is closely allied
to D. intermedins, of which it may be the successor.
Its specific separateness is somewhat doubtful.
have been the slightest beginning of such. There is a
rather narrow shelf, or lateral expansion of the malars, with
rounded outer borders, beneath the anterior portion of the
orbit, but it is not like the infraorbital process of D. hyognathus.
The postorbital hook does not appear to have been long or
prominent. Evidently the zygomatic arches extend laterally
outward more than in the last-named species; the postglenoid
processes are not nearly so heavy; the palate is broader; the
top of the cranium, though there is no zygomatic arch, becomes
narrower anterior to the crest of the occiput.
DISCO'V'EEY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
189
The teeth are very similar to those of Dolichorhinus heterodon,
so much so that, if only the teeth were known, they might be
referred to that species. They, as well as the skul), are larger.
Measurements [Douglass]
Millimeters
Length of top of skull 590
Length of free nasals 150
Length of skull posterior to anterior portion of orbit 393
Width of skull at glenoid articular surface 267
Width at infraorbital shelves 247
Millimeters
Length of p' 24
Width of p3 ■. 25
Length of p^ 27
Width of p* 31
Length ofm' 30
Width of m', about 37
Length of m^ 37
Width of m2 44
Length of m^, about • 41
Width of m3, about 43
A2
FiGUKE 136. — Type (holotype) of Dolichorhinus longiceps
Carnegie Mus, 2347. After Douglass, 1909. Ai, Palatal view of skull, somewhat less than one-third natural size; Aj, left lateral view of skull, somewhat less than
one-third natural size; A3, crown view of right upper premolar series, one-half natural size.
Length of molar-premolar series 192
Length of premolar series 88
Length-of molar series 112
Length of p' 15
Width of pi 11
Length of p^ 20
Width ofp2 20
Etymology. — longiceps, in allusion to the long skull.
Present determination. — For the reasons stated above
it appears that this form is connected with the typi-
cal D. Jiyognathus by a skull of intermediate char-
acters. Its status as a distinct species is therefore
somewhat doubtful.
190
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Sthenodectes Gregory, 1912
Cf. Sthenodectes, this monograph, page 353
Original reference. — Science, new ser., vol. 35, No.
901, p. 545, April, 1912 (Gregory, 1912.1).
Subsequent reference. — Riggs, New or little known
titanotheres from the lower Uinta formations: Field
Mus. Nat. Hist. Pub. 159, Geol. ser., vol. 4, No. 2,
p. ^8, June, 1912 (Riggs, 1912.1).
Figure 137. — Type (holotype) skull of Mesatirhinus superior
Field Mils. 12188. After Riggs, 1912. Side, top, and palatal views. Less than one-fourth natural size
Type species. — Telmatherium? incisivum Douglass.
Generic characters. — Gregory writes:
This, genus is distinguished from Telmatherium ultimum Os-
born by the following assemblage of characters: (1) The in-
cisors are far larger and more advanced in evolution, i' being
closely appressed to its fellowr in the median line, with anterior
face elongate, antero-internal tip blunt, median basin large,
posterior wall or cingulum very massive, i^ i^ extremely large
with low recurved tips and very heavy posterior cingula. (2)
The postcanine diastema is reduced or absent. (3) Superior
premolars 2, 3, 4 are much more advanced than in T. ultimum,
having very heavy internal cingula, pronounced external cin-
gula, high slender internal cusps (deuterocones) ; p^ especially
is in a relatively advanced stage as compared with T. ultimum.
(4) The least tranverse diameters of p* and of the anterior lobe
of m' are greater, that of m^ much less, than in T. ultimum.
(5) The basicranial region differs in many details, such as the
apparent junction of the postglenoid and post-tympanic proc-
esses below the auditory meatus. (6) The occiput is low,
with a sharp, long sagittal crest. (7) The forehead is
relatively wide. (8) The nasals taper dis-
tally.
From Manteoceras (especially M. uinten-
sis) the genus under consideration is dis-
tinguished by (1) the form and size of the
incisors and canines, (2) the much more
advanced stage of evolution of the premo-
lars, (3) the shorter anteroposterior diam-
eter of m^, (4) the reduction of the post-
canine diastema, (6) the arched and
spreading zygomata, etc.
From Dolichorhinus and Mesatirhinus it
is separated by the shortness and relative
breadth of the skull, the great size of the
incisors, the relatively heavy zygomata,
and many other details.
Etymology. — adho^, strength, driKT-qs,
a biter; in allusion to the great
power and development of the in-
cisors and canines.
Present determination. — A valid
genus, offshoot of the typical Telma-
therium phylum.
Mesatirhinus superior Riggs, 1912
Cf. Dolichorhinus superior (Riggs), this
monograph, page 405
Original reference. — Field Mus.
Nat. Hist. Pub. 159, Geol. ser., vol.
4, No. 2, p. 26, pi. 6, June, 1912
(Riggs, 1912.1).
Type locality and geologic hori-
zon.— White River divide, north-
eastern Utah; upper " Metarhinus
sandstones," summit of Metarhinus
zone (Uinta B 1). (See fig. 137.)
Holotype.— A skull (Field Mus.
12188).
Specific characters. — Riggs writes:
Skull 485 by 255 millimeters, molar series 182 millimeters,
nasals free to a point over last premolar, infra-orbital process
present, arches slender anteriorly, nasals infolded at margins,
sagittal area expanded, canines small, p^ and p' oblique to axis
of series. Molars relatively small, strong hypocone on m^, pos-
terior nares opening opposite the anterior margin of last molar.
Etymology. — superior, in allusion to its large size
and high stage of evolution.
Present determination. — This is a valid stage im-
mediately ancestral to the Dolichorhinus stage.
DISCOVERY OF THE TITANOTHEEES AND ORIGINAL DESCRIPTIONS
191
Metarhinus riparius Riggs, 1912
Cf. Metarhinus riparius, this monograph, page 429
Original reference. — -Field Miis. Nat. Hist. Pub. 159,
Geol. ser., vol. 4, No. 2, p. 28, pi. 7, fig. 1, June, 1912
(Eiggs, 1912.1).
Type locality and geologic horizon. — White River
canyon and divide, northeastern Utah; "entire upper
Metarhinus beds," base of Metarhinus zone (Uinta B 1).
Figure 138. — Type (holotype) skull of Metarhinus riparius
Field Mus. 12186. After Eiggs, 1912. About one-fourth natural size.
Holotype.— Skull (Field Mus. 12186). (See fig. 138.)
Paratype {"cotype"). — "Lower jaws" (Riggs, pi. 7,
figs. 2, 3).
Specific characters. — Riggs writes:
Skull long and narrow (405 by 210 mm.).
Anterior cranial region expanded, sagittal crest
short. Interorbital region relatively narrow and
rounded, rudimentary horn cores above orbits,
canines large, molar series short (88-93 mm.),
hypocone usually present on m', mandible
straight in the ramus, lower canine long and
recurved.
Etymology. — cristatus, crested; in allusion to the
high sagittal crest.
Present determination. — A valid stage in the Meta-
rhinus fluviatilis phylum.
Dolichorhinus fluminalis Riggs, 1912
Cf. Dolichorhinus fluminalis, this monograph, page 417
Original reference. — Field Mus. Nat. Hist. Pub. 159,
Geol. ser., vol. 4, No. 2, p. 33, pi. 10,
figs. 1-3, June, 1912 (Riggs, 1912.1).
Type locality and geologic horizon. —
Uinta Basin, northeastern Utah; "Amy-
nodon sandstone," summit of Eohasileus-
Dolichorhinus zone (Uinta B 2).
Holotype. — A fine skull. Field Mus.
12205; collector M. G. Mehl. (See fig.
140.)
Specific characters. — Riggs writes:
Skull small and narrow (520 by 230 mm.),
facial region much shorter than cranial, nasals
narrow and slightly tapering, posterior nares
opening between hamular processes, postorbital
process of jugal back of the last molar, molar-
premolar series 171 millimeters; canines short and recurved, in-
cipient horn cores in the form of high, narrow ridges. * * *
The skull is slender, light and complex in structure as com-
pared with the massive and rounded D. cornutus. The molar
teeth are no longer in the crown than those of Metarhinus
Etymology. — riparius, riparian, in allu-
sion to the nature of the habitat.
Present determination. — A valid species
in the Metarhinus phylum.
Metarhinus cristatus Riggs, 1912
Cf . Metarhinus cristatus, this monograph, page 429
Original reference. — Field Mus. Nat.
Hist. Pub. 159, Geol. ser., vol. 4, No. 2,
p. 28, pi. 9, fig. 3, June, 1912 (Riggs,
1912.1).
Type locality and geologic horizon. —
White River canyon, northeastern Utah;
"upper Metarhinus beds," lower section
of Metarhinus zone (Uinta B 1).
Holotype. — A skull, lacking the muzzle (Field Mus.
12194). (See fig. 139.)
Specific characters. — Riggs writes:
Skull length approximately 380 millimeters, molar series 94
millimeters. Frontal region broad, sagittal crest long and
high, molars short-crowned, no hypocone on m', arches rela-
tively heavy. Represented by a single skull lacking the
nasals and the premaxillaries.
Figure 139. — -Type (holotype) skull of Metarhinus cristatus
Field Mus. 12194. After Eiggs, 1912. One-third natural size.
earlei. The jugal process of the maxillaries arises at a point
back of the last molar rather than beside it as in Z). longiceps.
There is no offset in the palate between the last molars, though
the primary position of the posterior narial opening is marked
by a slight rugosity.
D. fluminalis is most nearly related to D. intermedins. The
skull exceeds in length the type of that species in the ratio of
520:465 millimeters. The molar teeth are proportionately
much smaller; the series measures relatively 99:109 millimeters.
192
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The position of the posterior narial opening is the most distinc-
tive character, appearing much farther back in D. fluminalis
than in any other described species. The two forms agree more
closely in the tapering form of the nasals and in the narrow
recess separating them, from the maxillaries.
Etymology. — fluminalis, pertaining to rivers; in
allusion to the habitat.
Present determination. — A stage in the DolichorTiinus
phylum, not very clearly distinguished specifically
from other progressive stacc?.
,to»
Figure 140. — Type (holotype) skull of Dolichorhinus fluminalis
Field Mus. 12205. After Eiggs, ] 912. Side, top, and palatal views. About one-fifth natural size
Rhadlnorhinus Riggs, 1912
Cf. Rhadinorhinus, this monograph, page 430
Original reference. — Field Mus. Nat. Hist. Pub. 159,
Geol. ser., vol. 4, No. 2, p. 36, June, 1912 (Riggs,
1912.1).
Type species. — Rhadinorhinus ahbotti Riggs.
Generic characters. — Riggs writes :
Titanotheres with slender skulls, nasals deeply recessed later-
ally and tapering, molars long-crowned, p-- ^- * subrectangular,
a wide median area between the incisors, no infra-orbital
process. The name Rhadinorhinus alludes to the tapering
nasals which characterize this genus.
Etymology. — pa8iv6s, slender; pis, nose.
Present determination. — Probably a valid stage, an
extreme offshoot of the Metarhinus phylum. (See
p. 17, fig. 15.)
Rhadinorinus abbotti Riggs, 1912
Cf. Rhadinorhinus ahbotti, tliis monograph, page 430
Original reference. — Field Mus. Nat. Hist. Pub. 159,
Geol. ser., vol. 4, No. 2, p. 36, pi. 11, figs. 2, 3, June,
1912 (Riggs, 1912.1).
Type locality and geologic horizon. —
Northeastern Utah; "upper Meta-
rhinus beds," center of Metarhinus
zone (Uinta B 1).
Holotype. — A fine skull (Field Mus.
12179). (See fig. 141.)
Specific characters. — Riggs says:
Length of skull 435 millimeters, molar-
premolar series 168 millimeters, nasals shorter
than premaxillaries, thickened at suture, and
tapering toward a terminal rugosity. Arches
slender, posterior nares open opposite middle
of m2. Sagittal crest long and narrow.
Hypocone of m' vestigial, diastema short.
Etymology. — Named in honor of
Mr. J. B. Abbott, of the Field
Museum of Natural History.
Present determination. — A valid spe-
cific stage.
Eotitanops gregoryi Osborn, 1913
Cf. Eotitanops gregoryi, this monograph,
page 291
Original reference. — Am. Mus. Nat.
Hist. Bull., vol. 32, p. 407, fig. 1;
p. 411, fig. 4B, September 2, 1913
(Osborn, 1913.400).
Type locality and geologic horizon. —
Type from Wind River Basin, Wyo.,
100 feet above Alkah Creek "red
stratum . ' ' Lamhdotherium-Eotitanops-
Coryphodon zone (Wind River B,
"Lost Cabin").
Type. — An incomplete lower jaw,
containing the right lower premolar-
molar series (pa-ms), also fragments
of left maxilla containing m", m^ (Am. Mus. 14889).
(See fig. 142.)
Specific characters. — Osborn writes:
Of inferior size. P2-m3, 78.4 millimeters; mi_3, 49; P2-3
with the internal cusps, paraconid and metaoonid, consisting
of rectigradations of most rudimentary stage; hypoconulid of
ms very small; m^ with a single internal cone, no hypocone.
This very sharply defined species may represent a
persistent primitive stage, because its recorded
(Granger) geologic level, 100 feet above the Alkali
Creek "red stratum," is higher than that of the
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
193
typical and relatively progressive E. horealis. Its
primitive condition is shown in the comparison of the
premolars with the same teeth in E. horealis (Cope)
and E. princeps Osborn.
The third inferior premolar is seen to be much less
progressive than in E. princeps or even in Lamhdo-
FiGURE 141. — Type (holotype) skull of Rhadinorhinus abboiti
Field Mus. 12179. After Riggs, 1912. About one-fourth natural size.
tJierium; the other premolars are also very primitive.
P2 short, compressed, with a very rudimentary hypo-
conid; ps laterally compressed, hypoconid distinct,
paraconid, metaconid, and entoconid extremely rudi-
mentary rectigradations. In the molar teeth, mi_3,
the metastylid and entostylid are also in an extremely
rudimentary or rectigradational stage. In ms the
hypoconulid is small, subconic, external in position.
Etymology. — Named in honor of Dr. W. K. Gregory,
of the American Museum of Natural History, the
colleague of the author in the preparation of this
monograph.
Present determination. — A valid specific stage.
Eotitanops princeps Osborn, 1913
('f. Eotitanops princeps, this monograph, page 295
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
32, pp. 410-411, fig. 4E, September 2, 1913 (Osborn,
1913.400).
Type locality and geologic horizon. — Wind River
Basin, Wyo.; Lamidotherium-Eotitanops-CorpJiyodon
zone (Wind River B, "Lost Cabin," exact level not
recorded). J. L. Wortman, collector.
Type. — Am. Mus. 296, including lower jaw, femur,
humerus, right manus, one cervical, three dorsal, and
one caudal vertebrae. (See figs. 143,
144.)
Specific cJiaracters. — Osborn writes:
Of still larger size, pa-ms 105 millimeters
(estimated). Inferior premolar teeth some-
what more complicated, as shown in the type
specimen. P2 with elevated, distinct, but very
rudimentary paraconid and metaconid; ento-
conid very rudimentary; talonid narrow. P3,
paraconid quite distinct, elevated; metaconid
small, distinct; entoconid rudimentary ; talonid
broad. P4, talonid broad; entoconid distinct.
Hypoconulid of ma rounded, more robust.
Ramus, larger and more robust.
The more advanced development of the
premolar rectigradations, the increased size of
the teeth and of the jaw, the larger size of the
hind feet in the referred specimen (Am. Mus.
4902) combine to distinguish this specimen as
a mutation or subspecific stage between E.
horealis and E. maJQr.
Etymology. — princeps, chief; in allu-
sion to its comparatively large size.
Present determination. — A valid spe-
cific stage.
Eotitanops major Osborn, 1913
Cf . Eotitanops major, this monograph, page 296
Original reference. — Am. Mus. Nat.
Hist. Bull., vol. 32, pp. 412-413, figs.
5D, 6, September 2, 1913 (Osborn,
1913.400).
Type locality and geologic horizon. — •
From Alkali Creek, Wind River Basin,
Lambdoiherium-Eotitanops-CorypTiodon zone
Wyo.;
Figure 142. — Type (holotype) teeth of Eotitanops gregoryi
Am. Mus. 14889. After Osborn, 1913. A, Left m'-m'; B, right lower premolar
series (P2-ms). Natural size.
(Wind River B, "Lost Cabin"; exact level unre-
corded).
Type. — Am. Mus. 14894, a left median metatarsal;
also the distal end of the tibia. (See fig. 145.)
194
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Specific cJiaracters. — Osborn writes:
Of superior size, Mts III 104 millimeters longitudinal, 16
transverse, index 15.
This ill-defined species indicates the existence in Wind River
times of a relatively large, short-footed titanothere, v^hich is
Figure 143. — Lower jaws of Lambdotherium and
EoHlanops
A, Lambdothenum popoagicum; B, Eotitanops gregoryi (holotype);
C, Eoiiianops browniamis; D, Eotitanops boreatis; E, Eotitanops
princcps (type). .One-fourth natural size. After Osborn, 1913.
possibly ancestral to some of the short-footed middle Eocene
types. The comparative measurements with the median
metatarsal of E. borealis are as follows:
E. borealis
E. major
Mm.
86
13
15
21
Mm.
104
Width of shaft -
16
15
25
Etymology. — major, larger; in allusion to the supe-
rior size of this animal compared with others of the
same genus.
Present determination. — A valid specific stage.
Lambdotherium priscum Osborn, 1913
Cf. Lambdotherium priscum, this monograph, page 286
Original rejerence. — Am. Mus. Nat. Hist. Bull., vol.
32, pp. 413-414, figs. 7A, 9A, September 2, 1913
(Osborn, 1913.400).
Type locality and geologic horizon. — -Wind River
Basin, 3 miles east of Lost Cabin, Wyo.; Lambdo-
therium-Eotitanops-Ooryphodon zone (Wind River B).
Granger, American Museum expedition, 1905.
Type. — Am. Mus. 12822, anterior portion of jaw
with P2-P4, nil of right side, also ps, mi, m2 of left
side. Rami fragmentary. (See fig. 146.)
Specific cJiaracters. — Osborn gives the following
description:
P2-P4, 25 millimeters. Second and third lower premolars
extremely simple, with rudimentary paraconid. Metaconid
of p3 rudimentary, placed very low upon slope of protoconid;
talonid narrow, depressed, with cingular rudiment of entoconid.
The extremely simple or primitive structure of the second
lower premolar clearly distinguishes this stage.
A referred specimen (Am. Mus. 14908) is slightly more
advanced in the structure of the second lower premolar, but is
still much more primitive than the type of L. popoagicum.
This specimen was found in the Wind River Basin, Dry
Muddy Creek, 18 miles up (Granger, Am. Mus. expedition,
1909).
The measurements of these two specimens are:
Type
(No. 12822)
Mm.
Second to fourth premolar, inclusive 25
Third premolar, anteroposterior 8
Third premolar, transverse 5
Fourth premolar, anteroposterior 9
Fourth premolar, transverse 6. 5
First molar, anteroposterior 11.5
First molar, transverse 7. 5
First to third molar, inclusive
Beferred
specimen
(No. 14908)
10
7
37
Etymology. — priscus, ancient; in allusion to the
primitive character of the species.
Present determination. — A valid specific stage.
Lambdotherium progressum Osborn, 1913
Cf. Lambdotherium progressum, this mongraph, page 286
Original rejerence. — Am. Mus. Nat. Hist. Bull.,
vol. 32, p. 415, fig. 8, September 2, 1913 (Osborn,
1913.400).-
Type locality and geologic horizon. — Wind River
Basin, Wyo. (Alkali Creek, Buck Spring); Lambdo-
therium- Eotitanops- Cor yphodon zone (Wind River B).
Granger, American Museum expedition, 1909.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
195
Type. — Am. Mus. 14917. Right ramus and sym-
physis of jaw containing ps-mz of right side, also left
canine. (See fig. 147.)
Specific characters. — Osborn writes:
P2-P4 16.5 millimeters. Second, third, and fourth lower
premolars progressive. Rudiment of metaconid on p2. Pswith
Figure 144. — Type (holotype) of Eotitanops
Left lower grinding teetii. Am. Mus. 296. After Osborn, 1913.
elevated metaconid subequal with protoconid, broad talonid
with rudimentary entoconid. P4 with bifid metaconid and
distinct entoconid.
This species is readily distinguished from both
L. priscum and L. popoagicum by the advanced con-
dition of p3, which may be described as submolariform.
pnnceps
Natural size.
AM 14894^
Figure 145. — Type (holotype) of Eotitanops
major
Metatarsal (A) and fragment of tibia (B). Am. Mus. 14894. After Osborn, 1913.
A, Median metatarsal: A', posterior view; A', anterior; A', distal; A*, projdmal.
B', Distal end of left tibia, anterior view; B^ the same, distal view. All one-
half natural size.
Measurements of type
Millimeters
Second to fourth lower premolar, inclusive 26
Second premolar, anteroposterior 8
Second premolar, transverse (trigonid) 4. 8
Third premolar, anteroposterior 9
Third premolar, transverse 6
Fourth premolar, anteroposterior 9. 3
Fourth premolar, transverse 7. 3
First molar, anteroposterior 11.5
First molar, transverse 8. 5
Second molar, anteroposterior 12. 5
Second molar, transverse 9. 5
Etymology . — progressum , progressive .
Present determination. — A valid specific stage.
Diploceras Peterson, 1914
Cf. Eolitanolherium, this monograph, page 435
Original reference. — Carnegie Mus. Annals, vol. 9,
Nos. 1-2, pp. 29-52, text figs. 1-15, pis. 6-10, 1914;
"issued August 17, 1914" (Peterson, 1914.1).
Type species. — Diploceras oshorni.
Generic characters. — Peterson writes:
Dentition: I|, C^, P-J, M|; premolar series proportionally
long; p5 with two distinct internal tubercles; horn cores well
developed; limbs relatively long and slender; tibial trochlea not
extended back on the calcaneum.
Astragalus high, with long neck, cal-
caneal and cuboidal facets laterally
located.
Etymology. — SittAoj, double ;
Kepai, horn.
Present determination. — The
name Diploceras being preoccu-
pied, Eotitanotherium was later
substituted. (See below.) The
genus itself is probably related
to the typical Diplacodon Marsh.
Diploceras osborn! Peterson, 1914
Cf. Eotitanotherium oshorni, this monograph, page 435
Original reference. — Carnegie Mus. Annals, vol. 9,
Nos. 1-2, pp. 29-52, text figs. 1-15, pis. 6, 7, 1914;
"issued August 17, 1914" (Peterson, 1914.1).
Type locality and geologic horizon. — On Duchesne
River near Myton, Uinta County, Utah; Eohasileus-
Dolichorhinus zone (upper levels of Uinta B 2).
Type. — Front of skull, lower jaws, portion of pelvis,
atlas, portion of axis, fragments of scapula and foot
bones, No. 2859 (Peterson, figs. 2, 3, 4, 7, 12; pis. 6,
7, 10). (See figs. 148, 149.)
Paratypes. — Front of skull, No. 2858; vertebral
column, fragments of ribs, bones of limb and foot, No.
2860; crowns of two upper molars, No. 2860a; hu-
merus, No. 2861; tibiae. No. 2862 (Peterson, figs. 1, 5,
6, 8, 9, 10, 11, 13, 14, 15; pi. 8).
Figure 146.-
-Type (holotype) of Lamhdotherium
priscum
Am. Mus. 12822. Ai, Anterior part of lower jaw; As, inner view
of right pj-p (reversed). After Osborn, 1913. Natural size.
Specific characters. — Peterson writes:
Alveolar borders of the premaxillaries extending well in front
of the canines; nasals long and relatively thin, their anterior
portion abruptly turned downward and convex on the anterior
border; incisors well in front of the canines and relatively sub-
equal in size; canines proportionally small.
196
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Etymology. — Named in honor of Prof. H. F. Osborn.
Present determination. — The genus is doubtfully sep-
arable from Diplacodon Marsh, but the species differs
in the more advanced development of the third upper
premolar.
FiGUKE 147. — Type (holotype) of Lambdotherium progressum
Lower jaw. Am. Mus. 14917. After Osborn, 1913. Natural size.
Heterotitanops Peterson, 1914
Cf. Metarhinus, this monograph, page 420
Original reference. — Carnegie Mus. Annals, vol. 9,
Nos. 1-2, pp. 53-57, text figs. 1, 2; pi. 11, "issued
August 17, 1914" (Peterson, 1914.2).
Type species. — Heterotitanops parvus Peterson.
Generic characters. — Peterson writes:
Dentition: If?, C-}-?, P|?, M|. Deciduous dentition: If?,
C-r, Mf?. Rapid increase in size of the deciduous upper
cheek teeth from first to last tooth. D * with perfectly formed
internal tubercles (proto- and hypocones) and the antero-
external angle very greatly developed. Molars hypsodont.
Ml with large conical proto- and hypocones, the external faces
of the ectoloph less emarginated anteroposteriorly
than in the titanotheres generally and the median
vertical ridge of the ectoloph projecting forward to a
greater degree.
Etymology. — erepoj, other, different; Ttrdi',
Titan; ajf, face; in allusion to its supposed
possible relationship to such forms as Eoti-
tanops.
Present determination. — According to Dr.
W. K. Gregory, who has studied the type
specimen of Heterotitanops parvus, the animal
probably represents a very young individual of
Metarhinus or Rhadinorhinus.
Heterotitanops parvus Peterson, 1914
Cf. Metarhinus sp. or Rhadinorhinus sp., this mono-
graph, page 198
Original reference. — Carnegie Mus. Annals,
vol. 9, Nos. 1-2, pp. 53-57, text figs. 1, 2,
pi. 11, 1914; "issued August 17, 1914"
(Peterson, 1914.2). ^"'""'°
Type locality and geologic horizon. — White River,
Uinta County, Utah; base of Metarhinus zone (Uinta
B 1). The type specimen "was found articulated
in a hard sandstone concretion, and lower down in
horizon A ['^J of the Uinta sediment than any mam-
" The upper or fosslliferous part of Uinta A of previous reports is Uinta B 1 of
this monograph.
malian remains hitherto described from that forma-
tion." (Peterson.)
Type. — Skull, lower javrs, vertebral column, ribs,
limb bones, calcaneum, and astragalus of young indivi-
dual (Carnegie Mus. 2909). (See figs. 150, 151, 152, 360.)
Specific characters. — Not determined.
Etymology. — parvus, poor, small.
Present determination. — According to Dr. W. K.
Gregory the type specimen probably represents a
very young individual of an undetermined species of
one of the previously described genera of Uinta
Basin titanotheres, probably of Metarhinus.
Eotitanotherium Peterson, 1914
(To replace Diploceras Peterson, 1913, preoccupied)
Cf. Eotitanotherium, this monograph, page 435
Original reference. — Carnegie Mus. Annals, vol. 9,
p. 220, September 12, 1914 (Peterson, 19U.4); Eotitano-
therium, a new generic name to replace Diploceras
Peterson. (See Peterson, 1914.1.)
In my article entitled "A new titanothere from the Uinta
Eocene" I employed the generic name Diploceras, having
overlooked the fact that this name is already preoccupied,
having been employed by Conrad as early as 1844 to designate
a genus belonging to the Mollusca. For this name I now sub-
stitute the name Eotitanotherium, which, after a diligent search
of the literature, I believe is not preoccupied. (Peterson.)
Etymology. — ^cos, dawn; Ttrav, a Titan; drjpiov, a
beast.
FiGUEE 148. — Type of Diploceras osborni
I lower jaw. Carnegie Mus. 2859. After Peterson, 1914. One-fourth natural size.
Present determination. — The genus is doubtfully
separable from Diplacodon Marsh.
Telmatherium? birmanicum Pilgrim and Cotter, 1916
Cf. Telmatherium f birmanicum, this monograph, pages 196-199
Original reference. — India Geol. Survey Records, vol.
47, pt. 1, pp. 72-74, pi. 5, figs. 9-11, 1916 (Pilgrim and
Cotter, 1916.1).
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
197
Type locality and geologic Jiorizon. — Myaing Town-
ship of the Pakokku district, Burma; Pondaung sand-
stone (upper to middle Eocene).
Cotypes. — Pilgrim and Cotter write:
This species is represented by five fragments of upper
molars, two of which are ahnost identical in shape and com-
prise the antero-internal quarter of two of the upper molars
probably occupying successive positions in the maxilla and
ably more behind the level of the paracone than is the case in
the Chalicotheriidae; thirdly, because in pm^ there is a single
large rounded and isolated inner cusp — the protocone, which is
totally unconnected with the two main outer cusps — a condi-
tion which never occurs in any chalicotheroid. In that family
the protocone in the premolars is connected to the outer cusps
either by a single or by a double crest. In addition to these
specific differences, the general structure of the tooth is unlike
that of any chalicotheroid that is known to us.
Figure 149. — Type of Diploceras {Eolitaiiotherimn) osborni
Palatal view. Carnegie Mus. 2859. After Peterson, 1914. One-half natural size.
being either m^ and m' or m' and m', two other portions of
the wall of the external crescents, and another an isolated proto-
cone. A sixth fragment consists only of the internal half of
what we take to be the last upper premolar. Three of these
pieces are figured in Plate 5, Figure 11 [9-11]. (See fig. 153.]
Systematic characters. — Pilgrim and Cotter write:
It is obvious that these are not chalicotheroid; first because
there is no trace of a protoconule, which in the Chalicotheriidae
is always present between the protocone and the paracone,
being invariably united to the latter by a transverse crest;
secondly, because the protocone in our specimens lies consider-
On the other hand, it approximates so nearly to that of many
of the Titanotheriidae that we have no hesitation in assigning
these fragments to that family. A careful comparison with the
various known species of the Titanotheriidae convinces us
that the Burmese fragments belong to a new species, but whether
this is to be referred to one of the known genera of that family
or whether it belongs to a new genus is a point which the
material at our disposal is insufficient to enable us to deter-
mine. We shall therefore do no more than indicate its prob-
able affinities, leaving a definite conclusion to the future,
when we may hope that more abundant material may come
to fight.
198
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
One of the most crucial points which has presented itself to
us for decision in connection with the material belonging to this
species is the position in the jaw of the tooth (G. S. I. No. C.
widening which we must assume to have taken place in m' of
this species. Again the faint V-ing of the line which connects
the two external crescents points to these being more closely
FiGUEE 150. — Type (holotype) skeleton of Heterotitanops parvus
Carnegie Mus. 2909. After Peterson, 1914. One-fourth natural size.
315) figured in Plate 5, Figure II. Although in some respects
this specimen reminds us of the last upper molar in some of the
FiGtTBE 151. — Type (holotype) skull of Heterotitanops parvus
Carnegie Miis. 2909. After Peterson, 1914. One-half natural size.
upper Eocene members of the Palaeosyopinae, yet its small
size as compared with the two other specimens of the upper
connected than is the case in the last upper molar of a titano-
there. On the other hand these features are such as the last
upper premolar of that family
would present, the only peculiar-
ities being the rounded nature of
the inner cone and the highly de-
veloped cingula on the anterior
and posterior margins of the frag-
ment, dying away internally and
apparently also on either side of
the two main external cusps.
It is evident that this simple struc-
ture of pm* prohibits the possibility
of this species being one of theTitan-
otheriinae of the Oligocene, while
on the other hand the increased
development of the cingulum and
the absence of an intermediate
tubercle point to its representing one of the latest develop-
mental stages of the Eocene subfamily of the Palaeosyopinae.
A similar indication is afforded by the fragmentary upper
Figure 152. — Type (holo-
type) of Heterotitanops
parvus
Upper and lower teeth. Carnegie
Mus. 2909. After Peterson,
1914. 1, Deciduous upper pre-
molars, first permanent molar;
2, permanent mi. One-half
natural size.
Figure 153. — Cotypes of Telmatherium? birmanicum
In the collection of the Geological Survey of India. After Pilgrim and Cotter, 1916. Naturalsize. A, "The antero-internal
portion of a right upper molar, surface view"; B, "e-xternal portion of an upper molar, showing the gently rounded
median fold, external view"; C, "internal portion of last upper premolar, surface view."
molars militates against this view. Further, the almost
rectangular shape of the inner portion of the tooth, which
alone is preserved to us, is inconsistent with the external
molars, in which the protocone is rather lofty and the only
vestige of a protoconule is the presence of a minute row of
fringing the protocone between it and the paracone.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
199
These start from the prominent cingular protostyle and cul-
minate in a more elevated portion some 13 millimeters to the
rear, diminishing again behind this point.
Attention may also be called to the presence in one of the
specimens of a broad, gently rounded median fold iu the
center of the external paraconal wall of the tooth, although in
the other specimen no such fold is visible. According to
Earle such a median rib is characteristic of all the early titano-
theres, tending to vanish in the upper Eocene and being
entirely absent in the Oligocene subfamily of the Titano-
Measurements of inferior teeth Pi-m^ and superior teeth m'-m^
Millimeters
P2-m3: Huerfano A. L. priscum (ref.), Am. Mus. 17526. 67
Wind River B. L. popoagicum (type), Am. Mus.
4863 69
Wind River B. L. progressum (type), Am. Mus.
14917 (estimated) 71
Huerfano A. L. magnum (type), Am. Mus. 17527. 74
Mi-m^: Huerfano A. L. priscum (ref.), Am. Mus. 17529. 21. 5
Huerfano A. L. priscum (ref.). Am. Mus. 2688.. 22. 5
Wind River B. L. popoagicum (ref.), Am.
Mus. 14902 25
Huerfano A. L. progressum (ref.), Am.
Mus. 17530 23.5
Wind River B. L. magnum (ref.). Am.
Mus. 15600 27.5
These measurements show that there is not a
great range in size between the smaller and the
larger animals referred to this genus.
Etymology. — magnum, large.
Present determination. — A valid specific
Figure 154. — Type (holotype) of Lambdotherium magni
Lower jaw. Am. Mus. 17527. After Osborn, 1919. Natural size.
theriinae. In any case the external lobes are broad and flat
and considerably elevated, hli;e those of the latest members of
the Palaeosyops-Diplacodon phyla.
Perhaps taking everything into consideration the present
species shows greater affinities with Telmatherium than any
other known titanotherid genus.
Etymology.- — Mrmanicum, relating to Burma.
Present determination. — Position uncertain. The
very close beading and massive cones of the single
grinding tooth figured suggest comparison with
Palaeosyops, a progressive species like P. copei.
These teeth might belong to a chalicothere, such as
Macrotherium or Moropus, but the resemblance is
not close.
Lambdotherium magnum Osborn, 1919
Cf. Lambdotherium magnum, this monograph, page 288
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
41, p. 562, fig. 3, 1919 (Osborn, 1919.494).
Type locality and geologic Jiorizon. — Lower horizon of
the Huerfano formation (Huerfano A) of Colorado.
Specific characters. — Osborn writes:
Exceeding in size any other known lambdothere is the type
jaw (Am. Mus. 17527) from the Garcia Canon, lower Huerfano,
containing a complete inferior series, p2-m3 of both sides,
represented in Figure 3. (1) These teeth exceed in length over
all (74 mm.) those of the type of L. popoagicum, in which the
same teeth measure 69 millimeters. (2) P3 has a rudimentary
metaconid and paraconid, in the same stage of evolution as in L.
popoagicum. (3) Of similar large size is a referred specimen.
Am. Mus. 15600, from the Big Horn, west end of Tatman Moun-
tain. "These_ referred grinders, m', m^, coincide closely in size
with the type of L. magnum and may be regarded as a paratype.
[See fig. 154.]
stage.
Eotitanops minimus Osborn, 1919
Cf. Eotitanops minimus, this monograph, page 296
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
41, p. 564, fig. 4, A, A\ 1919 (Osborn 1919.494).
Type locality and geologic Jiorizon. — Two miles north
of Gardner, Huerfano Basin, Colorado; from the lower
level of the upper horizon of the Huerfano formation
(Huerfano B).
Specific characters. — Osborn writes:
In reference to the fact that it is the smallest true titanothere
known, these type lower molar teeth, pi-ms. Am. Mus. 17439
(fig. 4, A, A'), * * * are assigned a new specific name
on the following grounds: (1) The measurement of p4-m3 (53
mm.) is much less than that (58) of the corresponding teeth
E.minlmus, Type
Figure 155. — Type (holotype) of Eotitanops minimus
Lower teeth. Am. Mus. 17439. After Osborn, 1919. A, Lingual or internal view;
A^, crown view. Natural size.
in E. gregoryi; (2) the other characters are so similar to those
of E. gregoryi as to suggest that this is a related form. [See
fig. 155.]
The accompanying figures (fig. 4, A, B, C) exhibit the
dimensional proportions of the above species of Eoiilaiiops.
It has been found from the large number of measurements of
Eocene titanotheres that no single species exhibits so great
a range of size.
Etymology. — minimus, small.
Present determination. — ^A valid specific stage.
200
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Eometarhinus Osborn, 1919 Eometarhinus huerfanensis Osborn, 1919
Cf. Eometarhinus, this monograph, page 419
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
41, pp. 568, 569, 1919 (Osborn, 1919.494).
Generic characters. — Osborn writes:
Small; ancestral to Metarhinus; with rudimentary frontonasal
horn; nasals elongate; overhanging premaxillaries, decurved as
Cf. Eometarhinus huerfanensis, this monograph, page 420
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
41, pp. 567-569, fig. 6, 1919 (Osborn, 1919.494).
Type locality and geologic horizon. — Huerf ano-Muddy
divide, 3 miles west of Gardner, Huerfano Basin, Colo.;
Huerfano formation, 205 feet below top (Huerfano B).
Figure 156. — Type (holotype) skull of Eometarhinus huerfanensis
11. Mus. 17412. After Osborn, 1919. A, Nasals, superior view; Ai, Aj, sections; B, skull, view of
left side; C, right upper jaw and teeth. One-half natural size.
in Metarhinus; no infraorbital shelf; characters apparently in-
termediate between those of the Metarhinus and Mesatirhinus
phyla.
Etymology. — rjcos, dawn; Metarhinus, a genus of the
middle Bridger beds; indicating an ancestral form of
Metarhinus.
Present determination. — This genus appears to be
ancestral to the Dolichorhinus phylum.
Type. — Anterior portion of skull (Am. Mus. 17412).
(See fig. 156.)
Specific characters. — Inferior in all measurements to
Mesatirhinus megarhinus. Premolars with small deu-
terocone. p'-m', 124 millimeters ; p'-p*, 53; m'-m', 72.
Etymology. — huerfanensis, in allusion to type locality.
Present determination. — A valid specific stage.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
201
SECTION 3. ORIGINAL DESCRIPTIONS OF TYPES OF
OLIGOCENE TITANOTHERES
IIST OF GENERA AND SPECIES
The following list shows that 49 species of Oligocene
titanotheres in North America and Europe have been
described and made the types of 17 different genera,
of which seven are regarded as valid. The types come
from many geologic levels. In finally determining the
genera wo are reluctantly compelled to adopt Menodus
Pomel in preference to Titanotherium Leidy, to adopt
Megacerops Leidy although it is based on a poor type,
and to reject Symhorodon Cope, because the genotype
species belongs to Menodus. The genera that rest on
the genotypic specimens are Brontotherium Marsh and
Brontops Marsh. Diplacodon Marsh is the least
soundly determined. It is close to Brontops and may
represent a sport. Teleodus Marsh represents an in-
ferior stage of the Brontops phylum, transitional to
Protitanotherium.
Chronologic list of the genera and species of Oligocene titanotheres
[Generic names accepted in this work as valid are printed in small capitals; abandoned names are inclosed in brackets.]
Date
1846
I
1849
1
1849
2
1850
II
1852
3
1852
4
1852
III
1853
6
1854
IV
1860
V
1870
6
1870
VI
1873
7
1873
VII
1873
8
1873
VIII
1873
9
1873
10
1873
11
1873
12
1873
13
1873
14
1873
1.5
1873
16
1874
IX
1875
17
1875
X
1876
18
1886
19
1887
20
1887
21
1887
XI
1887
22
1887
23
1887
XII
1887
24
1887
XIII
1887
25
1887
26
1887
XIV
1887
Present determination
["Gigantic Palaeotheri-
um."]
Menodus
Menodus
[Palaeotherium(?)]
giganteus.
proutii
[Titanotheriu m]
[ Palaeothieriu m]
[Rhinoceros]
[Eotherium.] (Type Rhi-
noceros americanus
Leidy.)
[Palaeotherium]
[Leidyotherium]
Megacerops
Megacerops
Brontotherium
Brontotherium
[Symborodon]
[Symborodon]
[Miobasileus]
[ Miobasileus]
Megaceratops
[ Megaceratops]
[Symborodon]
[Symborodon]
[Symborodon]
[Brontotherium]
[Symborodon]
[ Anisacodon]
[Anisaoodon]
[Diconodon (not Anisa-
codon).]
[Menodus]
[ Menodus]
[maximum],
[americanus]
[giganteum].
coloradensis.
gigas
torvus.
[ophryas]
acer
heloceras
bucco
[altirostris] . _
trigonoceras-
[ingens]
hypoceras
[montanus].
angustigenis.
tichoceras
[Menodus]-
[Menodus]_
dolichoceras _
platyceras
101959
Brontops
Brontops
Brontops
[M'enops]
[Menops]
[Titanops]
[Titanops]
[Titanops]
Allops
—29— VOL 1 16
robustus-
dispar
curtus..
[elatus]_
Prout-
Pomel _
do_
Owen, Norwood, and
Evans.
Leidy
do
_do_
-do.
do_
Prout_.
Leidy- .
do_
Marsh.
do_
Cope.-
do.
.do_
.do.
_do.
.do.
_do.
.do_
do.
Marsh.
Cope.-
Marsh.
do.
do.
Cope
Scott and Osborn.
.do.
Marsh.
do.
do.
.do.
.do.
.do.
.do.
-do.
.do.
Menodus Pomel.
Do.
Menodus giganteus Pomel.
Menodus proutii (Owen, Norwood,
and Evans).
Do.
(Indeterminate.)
Do.
Subfamily Menodontinae, genus in-
determinate.
(Indeterminate.)
Do.
Megacerops Leidy.
Megacerops coloradensis Leidy.
Brontotherium Marsh.
Brontotherium gigas Marsh.
Menodus Pomel.
Menodus torvus (Cope).
(Indeterminate.)
Do.
Megacerops acer Cope.
Menodus heloceras (Cope).
Megacerops bucco (Cope).
Megacerops acer Cope.
Menodus trigonoceras (Cope).
Menodus giganteus Pomel.
Brontotherium hypoceras (Cope).
(Indeterminate.)
Menodus giganteus Pomel.
Menodus giganteus? Pomel.
?Brontops angustigenis (Cope).
Brontotherium tichoceras (Scott and
Osborn) .
Brontotherium dolichoceras (Scott
and Osborn).
Brontotherium platyceras (Scott and
Osborn) .
Brontops Marsh.
Brontops robustus Marsh.
Brontops dispar Marsh.
Menodus Pomel.
Menodus varians (Marsh).
Brontotherium Marsh.
Brontotherium curtum (Marsh).
Brontotherium gigas Marsh.
Allops Marsh.
202
TITANOTHERES OP ANCIENT "WTOMING, DAKOTA, AND NEBRASKA
Chronologic list of the genera and species of Oligocene titanotheres — Continued
[Generic names accepted in ttiis work as valid are printed in small capitals; abandoned names are inclosed in brackets.]
Present determination
27
XV
28
29
XVI
30
XVII
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
1887
1S89
1889
1889
1890
1890
1890
1890
1891
1891
1891
1891
1892
1896
1902
1902
1902
1902
1905
1908
1908
1908
1908
1913
1916
1916
serotinus.
selwynianus .
syceras
amplus.
AUops
[Haplacodon] "
[Menodus]
[Menodus]
DiPLOCLONUS
Diploclonus
Teleodus
Teleodus
Allops
Brontops
[Titanops]
[Menodus]
[Menodus(?)]
[Titanotherium]
[ Megacerops]
[ Megacerops]
[Megacerops]
Brontotherium
[Megacerops]
Brontotherium
[Symborodon]
[ Megacerops] j primitivus —
Megacerops assiniboiensis
[Titanotheriu m] | [bohemicum] _
Allops ' walcotti
Megacerops j riggsi
Marsh.
Gope_-
do_
avus
crassicornis
[validus]
medius
[peltoceras]
rumelicus
ramosum
braehycephalus -
bicornutus
marshi
leidyi
tyleri
hatcheri
copei
do_
Marsh.
do-
.do.
.do.
-do.
do_
do_
Cope..
Toula..
Osborn.
do.
.do.
.do.
.do.
LuU...
Osborn.
do-
Lambe.-
do..
Kiernik.
Osborn..
do_.
Allops serotinus Marsh.
Allops sp.
Diploclonus selwynianus (Cope).
? Megacerops syceras (Cope).
Diploclonus Marsh.
Diploclonus amplus Marsh.
Teleodus Marsh.
Teleodus avus Marsh.
Allops crassicornis Marsh.
Brontops dispar Marsh.
Brontotherium medium (Marsh).
? Brontotherium curtum (Marsh).
? Brontotherium rumelicum (Toula).
Brontotherium ramosum (Osborn).
Brontops braehycephalus (Osborn) .
?Diploclonus bicornutus (Osborn).
Allops marshi (Osborn).
Brontotherium leidyi Osborn.
? Diploclonus tyleri (Lull).
Brontotherium hatcheri Osborn.
Megacerops copei (Osborn).
Teleodus primitivus (Lambe).
Megacerops assiniboiensis Lambe.
Menodus giganteus Pomel.
AUops walcotti Osborn.
Megacerops riggsi Osborn.
« Genotype Menodus angastigenis, upper teeth only. See No. 18, above.
PROUT'S DESCRIPTIONS OF A FRAGMENTARY lOWER JAW,
THE FIRST TITANOTHERE MADE KNOWN TO SCIENCE
"Gigantic Palaeotherium," Prout, 1846
Original reference. — Am. Jour. Sci., 2d ser., vol. 2,
pp. 288-289, 1 fig., 1846 (Prout, 1846.1).
Subsequent references. — Leidy, Description of the
remains of extinct IVIammalia and Clielonia from
Nebraska Territory, in Owen, Report of a geological
survey of Wisconsin, Iowa, and JMinnesota, p. 551,
1852 [Tab. 9, figs. 3, 3a, is not Prout's specimen]
(Leidy, 1852.1); The ancient fauna of Nebraska, pp.
72, 114, pi. 16, fig. 1, 1853 (Leidy, 1854.1).
Original description. — Dana and Silliman write:
Gigantic Palaeotherium. — We have recently received infor-
mation from Mr. H. A. Prout, of his discovery of the remains
of a Palaeotherium in the Tertiary near St. Louis, and we are
also indebted to him for a cast of the jaw, a view of the pos-
terior tooth of which is represented below. Mr. Prout is pre-
paring a memoir on the subject; and in the meantime we
state the following facts from his letter.
This fossil was found in the great northwestern Tertiary
belt, which is deflected from the north by the Black Hills and
which crosses the Missouri River at about latitude 43°. It
was accompanied by several Baculites compressus, an Inocera-
mus concentricus, a vertebra of a large fish, and some crystallized
gypsum. [As noted later by Prout these were from the Creta-
ceous and from another locality.] The entire jawbone, judg-
ing from the decrease in size of the teeth, must have been at
least 30 inches long, which far exceeds in size the Palaeotherium
magnum. The face of the posterior tooth is 4^ inches in
length; and from the posterior side of the last tooth to the
anterior side of the antepenultimate molar of the same side
the distance in the specimen is 11 inches. [See fig. 157.] This
is the aggregate length, in the line of the jaw, of but three out
of seven teeth; and with the most liberal allowance for decrease
of size in the other four the whole of the seven could not have
measured less than 16 or 18 inches, which is about one-half
larger than in the P. magnum.
Remarlcs. — This specimen was "the first of the
many mammalian remains which have been brought
to the notice of the scientific world from the vast
Eocene cemetery of Nebraska" (Leidy, 1852.1, p. 551).
It was the subject of Prout's second article cited below
and was the type of Menodus giganteus Pomel and one
of the cotypes of Palaeotherium? proutii Owen, Nor-
wood, and Evans (1850.1) and of Titanotherium,
proutii.
" Fossil maxillary bone of a Palaeotherium," Prout, 1847
Original reference. — Am. Jour. Sci., 2d ser., vol. 3,
pp. 249, 250, 1 fig., 1847 (Prout, 1847.1).
Subsequent references. — (See p. 204.)
Prout's description. — The following notice, written
by Dr. Prout himself, is a full description of the same
lower jawbone mentioned in his letter of the preceding
year:
The palaeotherial bone here described was sent to me some
time ago by a friend residing at one of the trading posts of the
St. Louis Fur Co., on the Missouri River. From information
since obtained from him, I learn that it was discovered in the
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
203
Mauvais Terre, on the White River, one of the western confluents
of the Missouri, about 150 miles south of St. Pierre, and 60
east of the Black Hills, at a point which would very nearly
The fifth and sixth molars (first and second true molars) re-
semble the one described, except that they want the third lobe,
and the dentine area on the crown of each lobe is much larger.
Figure 157. — "Vertical view of the posterior tooth belonging to the lower jaw of
Mr. Prout's Palaeotherium"
After Prout, 1846. Natural size.
The sixth is 33^ inches from front to posterior side. The
posterior lobe is 2 inches from the outer to the inner surface
and 1% inches long in the line of the jaw. The whole distance
on the jaw occupied by the three teeth is 11 inches. In the
correspond with the intersection of latitude 43" with longitude
26° west of Washington.
The Baculites and the Inoceramus which accompanied it and
which I at first supposed belonged to the same locality were
found in another formation — probably the Cretaceous-
distant about 100 miles, and included in the Grande
Detour or Great Bend of the Missouri River.
This fossil bone is a fragment of the inferior maxillary
of the left side, consisting of the posterior part of the
bone, together with the last three molar teeth. The
ramus is much fractured and presents an irregular sur-
face; yet the general direction of its outline may be
made out. The length of this fragment is 15 inches,
its depth from the liighest point of the ramus (a) to
the lowest (h) is 9K inches: it narrows regularly forward
so as to measure only 3}4 inches from the lower sur-
face of the bone at (d) to the alveolar process of the
antepenultimate tooth at (c). The inner surface of the
bone is more uniform, being marked merely by depres-
sions for the attachment of muscles. The alveolar por-
tion is here very prominent and well rounded, the teeth
being planted more than an inch from a vertical line which
is tangential to the inner surface of the bone. It is
covered in places with a concretionary matter which
could not be removed without injury to the specimen; on
analysis, this was found to consist chiefly of carbonate of
lime, with some alumina, and a small proportion of silex.
The last molar tooth has the three lobes of the Pa-
laeotheria, as shown in Figure 2. The inner surface is
nearly smooth and flat and shows no trace of lobes.
The size of the tooth from posterior to anterior sides is 4}/^
inches, of which 1^ inches belong to the anterior lobe, the
same to the middle, and 134 inches to the posterior. In
an upper view the two larger lobes have a deltoid form,
with the sides somewhat convex, and a rounded outer
angle. The thickness through from the outer to the op-
posite side is 15^ inches. The enamel of the inner side
folds over the surface, covering nearly a semicircular space
and leaving between it and the edge of the posterior en-
amel a subcrescent-shaped space (deltoido-lunate) of den-
tine, somewhat concave, which is nearly seven-eighths of
an inch broad at its widest part. These crescent-shaped Figure 158. — Original figures of Prout's "gigantic Palaeotherium," the
areas of the two lobes are connected by a continuous tract first titanothere discovered
of dentine, nearly IJ^ lines wide at the narrowest part;
and the same tract continues from the middle lobe to the
posterior; upon the latter it does not widen over the in-
terior, as the reflexed inner enamel covers the whole of the crown,
excepting a narrow space adjoining the posterior enamel. The
prominent points of the crown between the lobes project about
half an inch; and probably much more in the perfect tooth.
After Prout, 1847. A, "Fragment of the inferior maiillary of tlie left side," one-fourth natural
size; B, last lower molar on the left side, four-fLIths natural size.
largest Palaeotherium hitherto described, the P. magnum, the
same teeth occupy a space scarcely one-third that of the Mis-
souri animal.
St. Louis, December 10, 1846.
204
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
POMEL'S GENUS MENODUS, BASED ON PROUT'S
DESCRIPTION AND FIGURE OF THE FRAG-
MENTARY LOWER JAW
Menodus Pomel, 1849
Cf. Menodus, this monograph, page 522
Original reference. — Bibliotheque universelle de
Geneve (Supp.) Arch. sci. phys. nat., vol. 10, pp.
73-75, January, 1849 (Pomel, 1849.1).
Type species. — Menodus giganteus Pomel.
Original description. — Pomel writes :
Ce fossile a 6t6 dScouvert k Mauvais-Terre sur la Riviere
Blanche a 43° latitude nord et 26° longitude ouest de Wash-
ington, sur le versant occidental du bassin du Missouri. C'est
un fragment de mandibule portant les deux dernieres molaires
et I'alveole de I'antepenultieme, qui montrent tous les caracteres
du genre palaeotherium. La derniere molaire, la mieux oon-
servee, indique une espece plus voisine des vrais palaeotherium
(dont les P. magnum, medium, etc., sont les types) ou du sous-
Malheureusement on ignore I'age du terrain ou ce fossile
remarquable a 6t6 decouvert, quoiqu'il soit probable que c'est
dans la serie des formations de I'epoque alluviale qu'il faudra le
ranger. Cette difference d'age entre ce palaeothere et ceux de
I'Europe ocoidentale, ne doit pas etonner, puisque Ton trouve
dans I'Amerique du sud, dans des formations de meme age, un
animal de la meme tribu (on pourrait dire du meme grand genre),
le macrauchenia qui, lui aussi, est d'une taille superieure aux
esptices d' Europe. On salt, du reste, que sa derniere molaire
inf^rieure n'a que deux collines, comme dans le paloplotherium,
et que ses membres sont assez greles, tandis qu'il est probable
qu'un animal aussi gigantesque que ce nouveau palaeotherium
a ete assez trapu. Nous proposons de designer cette forme
animale fossile sous le nom de Menodus giganteus, en la consi-
derant comme un sous-genre des palaeotherium.
Etymology. — fxrivrj, the moon; 65ovs, tooth; in allu-
sion to the crescents of the lower molars.
Present determination. — Pomel proposed Menodus
as a subgenus of PalaeotJierium, using the latter term
in a very comprehensive sense, as later authors would
genre plagiolophus, que des anchitherium et des paloplotherium,
en ce que la troisieme coUine est bien d^veloppee, et forme un
troisieme croissant k la couronne; les autres croissants sont un
peu anguleux (croissants deltoides, dit I'auteur). La base de
la couronne est entouree d'un petit bourrelet comme dans les
palaeotherium d'Europe; mais si le dessin est exact, la maniere
dont les croissants principaux se reunissent indiquerait quelque
rapport avec ce qui existe chez les anchitheriums et les paloplo-
theriums, cette partie dtant plus 6paissie. II serait n^cessaire
d'en connaitre une molaire superieure pour fixer sa veritable
place; nous serions porte a presumer toutefois, que ce palaeothe-
rium est le type d'un sous-genre particulier; car independam-
ment de la brievetiS du fAt de la couronne des molaires, sa taille
est trop au-dessus de ceUe de nos plus grandes espfeoes euro-
p^ennes, pour qu'on puisse admettre sans hesitation son identity
subg^nerique avec ceUes-ci. En effet, I'arriere-molaire du
palaeotherium magnum est k peine le tiers de celle de I'espece
americaine: aussi cette derniere est-elle r^ellement colossale,
mesurant 0m,116, dont 0m,032 appartiennent k la troisieme
coUine; son ^paisseur est 0m,045. L'os mandibulaire est,
comme on devait s'y attendre, tres-robuste; il a 0m,112 de
diametre vertical entre les deux arriere-molaires; il s'elargit
consid^rablement k la partie du bord inf^rieur situ^e sous la
branche montante.
Figure 159. — Type of Menodus giganteus
Prout's original specimen. After Leidy, 1854. One-tlurd natural size.
speak of a family. In 1873 Marsh (1873.1, p. 486)
rejected the name Menodus on the ground that it was
essentially the same word as Menodon Meyer, 1838, a
genus of reptiles (Palmer, 1904.1, p. 410); but, as the
two names are spelled differently, according to the
modern rules of nomenclature' Menodus Pomel can
not be rejected on that ground. As shown below, the
type species Menodus giganteus rests upon Prout's
specimen, of which an excellent figure was given later
by Leidy (1854.1, pi. 16, fig. 1).
Menodus giganteus Pomel, 1849
Cf. Menodus giganteus, this monograph, pages 530, 535
Original reference. — See genus Menodus, above.
Type specimen. — As noted above, the species rests
upon Prout's original specimen, which was figured by
Prout in 1847 (1847.1, p. 249, and 1 fig.) and by
Leidy under the name Titanotherium proutii in 1854
(1854.1, pi. 16, fig. 1 only). The type may have been
destroyed in the "great fire" of St. Louis.
DISCOVEKY OF THE TITANOTHEEES AND ORIGINAL DESCRIPTIONS
205
Neotype (Osborn). — A carefully made model, based
on Leidy's figures and measurements of the lower jaw,
was compared with various specimens of Menodus
untn an upper dentition was found (in a skull, Am.
Mus. 505) which appears to fit very well the lower
teeth of the type. Hence the skull (Am. Mus. 505)
has been selected as a neotype of Menodus giganteus.
Specific characters. — Not separated from the generic
characters in Pomel's description. (See p. 530.)
Etymology. — giganteus, gigantic ; because larger than
the Palaeotherium magnum.
Present determination. — Although Prout's original
specimen, the type of Menodus giganteus Pomel, has
been lost, Leidy's carefully executed figure of this
specimen, together with his measurements and descrip-
tions, reveals generic and specific identity with the
dolichocephalic titanothere which Osborn in 1902
designated (1902.208, p. 96) Titanotherium ingens
Marsh. Titanotherium ingens is therefore to be
regarded as a synonym of Menodus giganteus Pomel.
Type. — From a study of the foregoing references it
is evident that Owen, Norwood, and Evans intended
the name Palaeotherium? proutii to cover both Prout's
original specimen and "Owen's specimen," discovered
by Evans, the lower jaw which was figured by Leidy
in 1852 (1852.1, pi. 9, figs. 3, 3a) and is still preserved
in the United States National Museum (No. 113;
our fig. 160). Prout's specimen is the type of Meno-
dus giganteus Pomel; hence, by the method of elimina-
tion, Owen's specimen becomes the type of Palaeothe-
rium? proutii Owen, Norwood, and Evans.
Etymology. — Named in honor of Dr. Hiram Prout.
Present determination. — "Owen's specimen" (Nat.
Mus. 113) appears to represent a Menodus, of a stage
slightly smaller than M. trigonoceras. (See p. 528.)
Titanotherium Leidy, 1852
Cf. Menodus, this monograph, page 522
Original reference. — "Palaeotherium? proutii Owen,
Norwood, and Evans," Owen, Eeport of a geological
Y
'<j.
\y\
Figure 160. — Owen's specimens of Palaeotherium? -proutii
After Leidy, 1862. A, Type of Palaeolheriumf proutii (Owen's specimen), Nat. Mus. 113. One-third natural size. Ttiis was the
principal specimen referred to by Leidy in proposing the name Titanotherium (1852.1). B, Third left lower molar, another of
Owen's specimens used by Leidy in describing Titanotherium. Two-thirds natural size.
EARLY NOTICES BY lEIDY AND OTHERS, 1850-1870
Palaeotherium? proutii Owen, Norwood, and Evans, 1850
Cf. Titanotherium proutii Leidy
Original reference. — Acad. Nat. Sci. Philadelphia
Proc, vol. 5, p. 66, August, 1850 (Owen, Norwood,
and Evans, 1850.1).
Subsequent reference . — " Palaeotherium? proutii Owen,
Norwood, and Evans," Leidy, Description of the
remains of extinct Mammalia and Chelonia from
Nebraska Territory, in Owen, Report of a geological
survey of Wisconsin, Iowa, and Minnesota, pp. 551-
552, tab. 9, figs. 3a, 3, 1852 [Owen's specimens, not
Prout's] (Leidy, 1852.1); "Titanotherium proutii
Leidy," The ancient fauna of Nebraska, pp. 72-73,
pi. 16, figs. 1-3, 1853 (Leidy, 1854.1).
Original description. — Owen, Norwood, and Evans
state that
These remarkable remains are thus named in compliment to
Dr. Prout of St. Louis who first noticed them in the American
Journal of Science and Arts. The generic characters, however,
are not yet satisfactorily decided.
survey of Wisconsin, Iowa, and Minnesota, p. 552
1852 (Titanotherium) (Leidy, 1852.1V
Subsequent reference. — " Titanotherium proutii
Leidy," Leidy, The ancient fauna of Nebraska, pp.
72, 114, 1853 (Leidy, 1854.1).
Type species. — Palaeotherium? proutii Owen, Nor-
wood, and Evans."* (See p. 526.)
Generic characters. — Not separated by Leidy from
specific characters.
Etymology. — 'Yirkv, a Titan; driplov, beast.
Present determination. — Leidy based the genus Ti-
tanotherium collectively upon a number of specimens,
including, first, Prout's original specimen; second,
"Owen's specimen" (Nat. Mus. 113); and third, cer-
tain other fragmentary specimens. Prout's specimen
was already the type of Menodus giganteus Pomel,
hence by elimination the genus Titanotherium rests
upon the species Palaeotherium? proutii Owen, Nor-
i« In his work of 1853 Leidy placed his own name after the speoihc name proutii,
evidently following the practice of those who placed the name of the author of the
genus after the specific name.
206
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
wood, and Evans, the type of which is the second
specimen described by Leidy, namely, Evans's speci-
men (Nat. Mus. 113). This specimen is believed by
FiGUEE 161. — Type (holotype) of Palaeoiherium maximum
Parts of the outer wall (the ectoloph) ot two upper molars. After Leidy, 1862. Natural size.
Osborn to be congeneric with the type of Menodus
giganteus Pomel.
Proposal of the generic name TitanotJierium. — After
describing under the name Palaeotherium?
proutii the specimens made known by Prout
and by Owen, Norwood, and Evans, Leidy /'
(1852.1, p. 552) says:
All the preceding specimens, except probably the
latter two, I suspect belong to a different genus from
either Palaeotherium or Anchitherium, and should the
suspicion prove correct, Tilanotherium would be a
good name for the animal, as expressive of its very
great size.
Palaeotherium maximum Leidy, 1852
Original reference. — Leidy, in Owen, Report
of a geological survey of Wisconsin, Iowa,
and Minnesota, description of tab. 12 B, figs.
3, 4, 1852 (Leidy, 1852.1).
Type locality and geologic horizon. — White
River, "Nebraska" [South Dakota]; Chadron
formation {Titanotherium zone).
Type. — Parts of the outer wall or ectoloph
of two superior molars. Types now lost-
(See fig. 161.)
Characters. — Leidy writes: "I am at pres-
ent very much inclined to consider these as
belonging to a true species of Palaeotherium,
which from its very great size might be
appropriately named Palaeotherium maxi-
mum."
Etymology. — maximum, greatest — that is,
greater than P. magnum.
Present determination.— These fragments belong to
a large Oligocene titanothere of wholly uncertain
reference.
Rhinoceros americanus Leidy, 1852
Original reference. — Acad. Nat. Sci. Philadelphia
Proc, vol. 6, p. 2, 1852 (Leidy, 1852.2).
Subsequent reference. — Leidy,
The ancient fauna of Nebraska,
p. 76, pi. 17, figs. 1-4, 1853 CLeidy,
1854.1).
Type locality. — White River,
"Nebraska" [South Dakota].
Type. — Two superior premolars
belonging upon opposite sides of
the jaw. Part of a collection pro-
cured by Mr. Thaddeus A. Cul-
bertson for the Smithsonian Insti-
tution. Types not located. (See
fig. 162.)
Characters. — The proceedings
of the Philadelphia Academy con-
tain the following note:
Dr. Leidy called the attention of the members to a fossil
tooth and a fragment of a second, from the collection made by
Mr. Culbertson in Nebraska Territory, which, he observed,
belonged to a new species of Rhinoceros, or probably Acero-
FiGURE 162. — Cotypes of Rhinoceros americanus
Two upper fourth premolars. After Leidy, 1853. Natural size.
therium. The former specimen is probably a third premolar,
the latter a portion of the fourth. A great peculiarity in the
teeth is the confluence of the inner lobes with each other and
their separation to the base from the outer lobes. They
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
207
possess a remarkably strong basal ridge and indicate an animal
larger than any species of existing Rhinoceros; the greatest
transverse diameter of the third premolar being 2J^ inches;
its anteroposterior diameter 1% inches. For the species the
name Rhinoceros americanus is proposed.
Etymology. — americanus, in allusion to the then
novel fact that a supposed rhinoceros had once in-
habited America.
Present determination. — It does not seem possible
to determine positively whether these isolated pre-
molar teeth belong to Allops or to Menodus; the
affinity to one or the other of these genera is indicated
by the pronounced internal and external cingula and
by the large tetartocone on p*. In view of the doubt
and the disappearance of the type, it seems best to re-
gard "Rhinoceros" americanus as indeterminate.
Eotherium Leidy, 1853
Cf. Menodus Pome!, this monograph, page 522
Original reference. — Acad. Nat. Sci. Philadelphia
Proc, vol. 6, p. 392, 1853 (Leidy, 1853.1).
Present determination. — The specimens indicated
were first chosen the types of Rhinoceros americanus.
(See above.) The very pronounced internal and
external cingula of the type (fig. 162), however,
appear to indicate that they belong generically to
Menodus. The genus Eotherium was subsequently
treated by Leidy as a synonym of Titanotherium. The
name Eotherium was subsequently (1875) applied by
Owen to a genus of sirenians.
Palaeotherium giganteum Leidy, 1854
(Indeterminate)
Original reference. — The ancient fauna of Nebraska:
Smithsonian Contr. Knowledge, vol. 6, p. 78, pi. 17,
figs. 11-13, 1853 (Leidy, 1854.1).
Type locality. — White River, "Nebraska" [South
Dakota].
Types. — Portions of the ectoloph of five molars "in
the collections of Mr. Culbertson and Dr. Owen."
Lectotypes (Osborn). — The fragmentary ectoloph
figured in Plate 17, Figure 11, of Leidy's work. (See
fig. 163.)
Figure 163. — Cotypes of Palaeotherium giganteum
Parts of the ectoloph of upper molars. After Leidy, 1853. Natural size.
Subsequent reference. — Leidy, The ancient fauna of
Nebraska, pi. 17, figs. 1-7, 1853 (Leidy, 1854.1).
Type species (monotypic). — Rhinoceros americanus
Leidy. (See above.)
Generic description. — Leidy says:
Of the huge Titanotherium proutii there are numerous small
fragments of bones and teeth and also several entire superior
molars, which have served to remove some of the obscurity
in regard to the characters of the animal. From the last-
mentioned specimens it appears that those which have been
described as probably indicating a new species of Palaeotherium^
under the name P. giganteum (Ancient fauna of Nebraska, pi.
17, figs. 11-13), belong to Titanotherium ■proutii, while several
superior molars (ib., figs. 1-7), attributed to the latter, belong
to a new genus associating characters of Rhinoceros and Palaeo-
therium. For this genus and species, represented by Figures
1-7, Plate 17, in the Ancient fauna of Nebraska, I propose the
name of Eotherium americanum.
Etymology. — rjois, dawn, dripiov, beast; possibly in
allusion to the relatively early geologic age of the
animal.
Characters. — Leidy writes:
The fragments, of which there are five, are only single ex-
ternal lobes of the upper molars. These, externally, correspond
to the description of Cuvier of the teeth of Palaeotherium. A
conjoined pair of the lobes, forming the outer part of a tooth,
"present the external face strongly inclined inward in descend-
ing and divided by three salient ridges into two concavities,
which are rounded toward the fangs and terminate in a tri-
angular cusp at the masticating surface, the basal angles of
which rest upon the termination of the salient ridges." The
median ridge is a thick obtuse fold outward of the tooth, and
the anterior and posterior ridges are acute, roughened offsets
from the basal ridge, descending to the masticating surface.
The measurements of the more perfect specimens are as
follows :
In. hnes
Length of the longest lobe 2 4
Length of a second specimen 2
Breadth of the second specimen at the basal angles of the
cusp 1 8
Length of the shortest lobe 1 7
Breadth of the shortest lobe at the basal angles of the
cusp 1 3
208
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Etymology. — giganteum, gigantic; in obvious allu-
sion to the great size, which seems to have impressed
all the early observers of Titanotherium.
Present determination. — Leidy himself subsequently
(1854.1, p. 157) transferred this species to T. proutii.
It is indeterminate.
Figure 164. — Type (holotype)
coloradensis. Nasals and
One-third natural size.
After Leidy, 1873. A, Top view; B, front view; C, view of left side,
Leidyotherium Prout, 1860
(Indeterminate)
Original rejerence. — Acad. Sci. St. Louis Trans., vol.
1, pp. 699-700, 1860 (Prout, 1860.1).
Subsequent rejerence. — Leidy, Extinct Mammalia of
Dakota and Nebraska, p. 390, 1869 (Leidy, 1869.1).
Type species. — None designated.
Type locality. — The specimen was reported to have
been obtained near Abingdon, in Virginia, but was
later stated by Leidy (op. cit., p. 390) to be "a fossil
from the Mauvaises Terres of White Kiver, Dakota."
Type. — "The fragment of a large molar tooth."
Generic characters. — Prout writes:
The lobed or indented border of the enamel would seem to
show that this animal was nearly allied to Titanotherium, while
the great width and depth of the groove between the outer and
what may have been the inner border of the tooth would sepa-
rate it from this genus. * * * It is distinguished, more-
over, from these [Lophiodon] by the greater length of the fangs
and the comparative shortness of the enamel on
the outer surface of the tooth. * * * j^ must
have been a phytivorous pachyderm, as large if
not larger than the Titanotherium.
Etymology. — Named in honor of Joseph
Leidy.
Present determination. — No specific name
■;. is given. Leidy treated the genus as syn-
onymous with Titanotherium. It is an
indeterminate member of the family.
Megacerops Leidy, 1870
Cf. Megacerops, this monograph, page 541
Original reference. — Acad. Nat. Sci. Phila-
delphia Proc, vol. 22, p. 2, 1870 (Leidy,
1870.1).
Subsequent reference. — Leidy, Extinct ver-
tebrate fauna of the Western Territories,
p. 239, pi. 1, figs. 2, 3; pi. 2, fig. 2, 1873
(Leidy, 1873.1).
Type species. — Megacerops coloradensis
Leidy.
Generic characters. — In the original refer-
ence a detailed description of the type
specimen of Megacerops coloradensis is given,
comparisons being made with the anterior
horn cores and nasals. of the Siwalik Sivathe-
rium, with which it was thought possibly to
be allied. Leidy concludes as follows:
It is probable that the fossil may pertain to the
same animal as the remains from the Mauvaises
Terres of Nebraska, described under the name of
Titanotherium, but in the state of extreme uncer-
tainty as to its collocation, it may with equal
probability be referred to other genera, perhaps
to Megalomeryx, or it may have been an American
species of the Sivatherium. Under the circum-
stances it may be referred to a new genus, with the
name of Megacerops coloradensis.
Etymology. — jxkya^, great; Kfpas, horn; &4',
face.
Present determination. — Leidy's carefully
executed figures of the type, in the opinion
of the present writer (Osborn), reveal the
generic relationship of this animal with that later de-
scribed by Cope (1873.2, p. 4) as Megaceratops acer.
Megacerops coloradensis Leidy, 1870
Cf. Megacerops coloradensis, this monograph, page 544
Original reference. — Acad. Nat. Sci. Philadelphia
Proc, vol. 22, p. 2, 1870 (Leidy, 1870.1).
Subsequent reference. — Leidy, Extinct vertebrate
fauna of the Western Territories, pp. 239-242, pi. 1,
figs. 2, 3; pi. 2, fig. 2, 1873 (Leidy, 1873.1).
Type locality and geologic horizon. — Colorado ; Chad-
ron formation {Titanotherium zone), level not ascer-
tained.
DISCOVERY OP THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
209
Type. — Fractured horns and nasals. The present
location of this type has not been determined. It is
not in the collection of the Philadelphia Academy,
nor is any record of its loan to be found. (See fig.
164.)
Characters of type. — Leidy's description is too long
to quote here. The specimen may be described briefly
as follows: Seen from above the nasals are of moderate
length and taper toward the extremities; from the
side and front they appear decidedly long and thin
and are strongly decurved at the tip, at which point
there is a median notch. The horns project forward
and outward and pass from an elongate oval section
at the base to rounded, transversely oval tips. The
greatest diameter of the horns at the base is antero-
posterior, with flattened outer and convex inner
faces. The following approximate measurements are
taken from Leidy's descriptions and figures:
Millimeters
Free width of nasals 108
Free length 104
Outside measurement of horns 140
Etymology. — coloradensis , in allusion to the type
locality.
Present determination. — The type of Megacerops
coloradensis, consisting of the osseous horns and nasals,
is apparently distinct specifically from Cope's M. acer,
M. hucco.
SPECIES DESCRIBED BY MARSH AND COPE IN 1873-1876
Brontotherium Marsh, 1873
Cf. Brontotherium, this monograph, pages 555-557
Original reference. — Am. Jour. Sci., 3d ser., vol. 5,
p. 486, 1873 (Marsh, 1873.1).
Type species. — Brontotherium gigas Marsh. (See
below.)
Generic characters. — Marsh writes:
An examination of the remains, in the Yale Museum, of the
huge mammals allied to Titanotherium has led to the discovery
that two different animals have hitherto been referred to the
species known as T. prouti. These animals are generically dis-
tinct and probably are from separate geological horizons.
The one here described differs from Titanotherium in its denti-
tion, having but three lower premolars, the series being as fol-
lows: Incisors 2, canine 1, premolars 3, molars 3. The animal
was, moreover, a true perissodactyl, with limb bones resembling
those of Rhinoceros. The genus is related to Titanotherium, and
the two appear to form a distinct family, which may be called
Brontotheridae. The present species is based on portions of
three individuals, one of which has the lower jaws and en-
tire molar series complete. They indicate an animal fully
equal to T. prouti in size, and but little inferior in bulk to the
elephant. The lower molars resemble those in the type speci-
men of T. prouti, but the jaw below them is not so deep, and its
lower margin is more nearly straight, descending but very
slightly toward the angle. The front part of the lower jaws is
somewhat suilline in form. The incisors are quite small, and
the two next to the symphysis are separated from each other.
There is a short diastema between the canine and the first
premolar. [This is followed by remarks on the skeleton based
on the "other specimens."]
Etymology: fipovTi], thunder; drjpiov, beast.
Present determination. — This was the most impor-
tant contribution to the knowledge of the titanotheres
made up to that time. The characters of the lower
jaw and of the skeleton are correctly described, and the
family is referred to the Perissodactyla. Subsequent
research has shown that the genus Brontotherium is
distinct from Menodus and Megacerops; "Brontothe-
rium ingens," as used in later publications by Marsh,
referred to the skull, the type of "B. ingens," and not
to the jaw, the type of Brontotherium gigas.
Brontotherium gigas Marsh, 1873
Cf. Brontotherium gigas, this monograph, page 567
Original reference. — Am. Jour. Sci., 3d ser., vol. 5,
p. 486, 1873 (Marsh, 1873.1).
Subsequent reference. — Principal characters of the
Brontotheriidae : Am. Jour. Sci., 3d ser., vol. 11, pi.
12, figs. 1-3, 1876 (Marsh, 1876.1).
Type locality and geologic horizon. — Colorado; exact
locality and level not published. Sargent, Griswold,
and Marsh, collectors.
Type. — "The present species is based on portions
of three individuals, one of which has the lower jaws
and entire molar series complete [lectotype]." Yale
Mus. 12009. (See fig. 165.)
Characters of type. — The specific characters were
not separated by Marsh from the generic charcters.
Measurements of the lower jaw were given, some of
which (now verified) are as follows:
Millimeters
Length of lower jaw, from condyle to front of symphysis 634
Depthof lower jaw, from top of coronoid process to angle 367
Length of last lower molar 117
Length of last lower premolar (Marsh gives this as 51) [49]
Etymology. — yiyas, giant.
Present determination. — This valid species is fully
discussed in Chapter VI of this monograph (p. 567).
Symborodon Cope, 1873
Cf. Menodus, this monograph, page 525
Original reference. — Pal. Bull. No. 15, p. 2, "issued
August 20, 1873" (Cope, 1873.2).
Subsequent reference. — Cope, Eeport on the verte-
brate paleontology of Colorado, pi. 2, fig. 1; pis. 3, 4,
1874 (Cope, 1874.2).
Type species. — Symborodon tonus Cope. (See
below.)
Generic characters. — Cope writes:
Dentition: I.? 0; C. 1; Pm. 3; M. 3; the canines slightly
separated from each other, but not from the first premolar.
Crowns of the premolars with L-shaped crescents as in Rhi-
noceros; of the molars with completed crescents; the last molar
with third posterior crescent. Symphysis mandibuli coossi-
fied, crowns of canines not projecting, conic. * * * Xhe
genus differs from Titanotherium and Brontotherium in the
absence of incisors and from the former in the presence of but
three premolars. If there had been a deciduous incisor on each
side I was unable to detect any trace of it.
Etymology. — (tvv, together; |Sop6s, devouring; oSous,
tooth; in reference to the approximation of the op-
posite canines toward the middle line.
210
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Present determination. — Subsequent research has
proved that this genus is a synonym of Menodus. It
is fully described on page 522.
Symborodon torvus Cope, 1875
Cf. Menodus torvus, this monograph, page 525, Figure 166
Original reference. — Pal. Bull. No. 15, p. 2, "issued
August 20, 1873" (Cope, 1873.2).
FiGUEB 165. — Type (lectotype) of Bronluihenum gigas
Lower jaw, with nearly complete dentition. Yale Mus. 12009. After Marsh, 1876. One-sixth natural size
Subsequent reference. — Report on the vertebrate
paleontology of Colorado, p. 486, 1874. The jaw
figured in Plate 2, Figure 1, is not the type of torvus
(Cope, 1874.2).
Type locality and geologic Tiorizon. — Horsetail Creek,
Logan County, northeastern Colorado; Chadron for-
mation {Titanotlierium zone), level not ascertained.
Ootypes. — Cope writes: "The present genus is
established on mandibular rami only, which can not be
certainly associated with crania." These rami (Cope
collection. Am. Mus. 6365, 6345) are accordingly
CO types. In his "Report on the vertebrate paleon-
tology of Colorado" Cope says, "I append a de-
scription of the mandible, on which the species
Symborodon torvus was established." Careful com-
parison of Cope's original and subsequent descriptions
and measurements shows that the species
was established largely upon the lower
jaw (Am. Mus. 6365, fig. 166) which
is accordingly regarded as the lecto-
type.
Etymology. — torvus, wild, grim.
Present determination. — The species is
now regarded by Osborn as belonging in
the genus Menodus. In size the type
is intermediate between M. Jieloceras and
M. trigonoceras.
Miobasileus Cope, 1873
(Indeterminate)
Original reference. — Pal. Bull. No. 15,
p. 3, "issued August 20, 1873" (Cope,
1873.2).
Subsequent references. — On some ex-
tinct types of horned perissodactyls,
p. 108, 1874 (Cope, 1874.1); Synopsis
of new Vertebrata from the Tertiary of
Colorado, p. 14, 1873 (Cope, 1873.3);
Report on the vertebrate paleontology of
Colorado, p. 490, 1874 (Cope, 1874.2);
U. S. Geol. Survey Terr. Rept. for 1873,
p. 490, 1874.
Type species. — Miobasileus ophryas
Cope. (See below.)
Generic characters. — Not separated
by Cope from specific characters.
(Seep. 201.)
Established on a cranium with nearly com-
plete dentition but without mandibular ramus.
Head elongate, concave in profile from the
interorbital region to the supraoccipital crest.
This is transverse and concave, the posterior
borders of the temporal fossae extending behind
it. These fossae leave a narrow flat vertex
between them. Zygomatic arch stout and rather
deep; a strong postglenoid process. Nasal bones
very massive, their free portion elongate,
hornless. A massive horn core rising from
above each orbit, no superciliary angle or ridge. Orbit not
inclosed behind. Of molar teeth only Pm. 2, M. 3, pre-
served, the M. with two, the Pm. with one inner cone, and
two outer continuous crescents. The latter send inward to
one side of the cones a transverse ridge. Incisors and canines
unknown.
Char, specif. — Front concave transverse just behind between
the horns. Latter massive and little compressed. Nasal
bones convex longitudinally and transversely, slightly rugose.
Transverse ridges of teeth with transverse expansions at their
inner extremity, being thus T-shaped.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
211
Millimeters
Length from apex of nasals to occipital condyles (axial) 684
Length from occipital cond3'les to femoris of palate 376
Length from occipital condyles to end of palatine lamina
pteryzoidea 270
Length of four last molars 242
Length of three last molars 195
Length of last molar 68
Width of palate at nareal notch 116
Etymology. — Mio, Miocene; /SacriXeiis, king — that is,
monarch of the Miocene.
Present determination. — -The genus is indeterminate.
(See M. ophryas.)
Miobasileus ophryas Cope, 1873
(Indeterminate)
Original reference. — Pal. Bull. No. 15, p. 3, "issued
August 20, 1873" (Cope, 1873.2).
Subsequent references. — Cope, On some extinct types
of horned perissodactyls, p. 108, 1874 (Cope,
1874.1); Synopsis of new Vertebrata from the
Tertiary of Colorado, p. 14, 1873 (Cope,
1873.3) ; Report on the vertebrate paleontology
of Colorado, p. 490, 1874 (Cope, 1874.2).
Type locality and geologic Jiorizon. — Cedar
Creek, Logan County, Colo.; Chadron forma-
tion {Titanotherium zone), level not ascertained.
Type. — A cranium with incomplete dentition,
without mandibular ramus. (In a later descrip-
tion Cope (1874.2, p. 490) remarks, "of molar
teeth only pm 3-4, m 1, 2, 3, preserved. ") This
type was left in the field and is now lost.
Generic and specific cJiaracters (summarized
from Cope). — Supraoccipital crest concave.
Zygomatic arch stout and relatively deep.
Nasal bones very massive, elongate, convex
longitudinally and transversely; a massive horn core,
little compressed, rising above each orbit. In a later
communication Cope (1874.2, p. 491) gives the length
from apex of nasals to occipital condyles as 664
millimeters and observes:
The dental characters of this species ally it to the S. trigo-
noceras, but the form as well as the position of the horns is
quite different. Instead of being triangular, a section of the
base of these is elliptic. Extremity conical.
Millimeters
Length from apex of nasals to occipital condyles 684
Length of three last molars 195
Length of last molar _. 68
Etymology. — 64>pvs, eyebrow*; possibly in allusion
to the form of the orbit.
Present determination. — Owing to the loss of the
type and the uncertain character of the description,
this genus and species is indeterminate.
Megaceratops Cope, 1873
Original reference. — Pal. Bull. No. 15, p. 4, "issued
August 20, 1873" (Cope, 1873.2).
Present determination. — This name Megaceratops was
not proposed in order to denominate a new genus but
was merely an emendation on etymologic grounds of
Leidy's term Megacerops, of which it must be regarded
as a synonym.
Megaceratops acer Cope, 1873
Cf. Megacerops acer, this monograph, page 545
Original reference. — Pal. Bull. No. 15, p. 4, "issued
August 20, 1873" (Cope, 1873.2).
Subsequent reference. — Cope, Report on the verte-
brate paleontology of Colorado, p. 488, pi. 7; pi. 8,
fig. 3, 1874 (Cope, 1874.2).
Type locality and geologic Jiorizon. — Horsetail Creek,
northeastern Colorado; Chadron formation {Titano-
therium zone), level not ascertained.
Type. — "A single cranium without under jaw."
Am. Mus. 6348. (See figs. 167, 170.)
Figure 166. — Type (lectotype) jaw of Symhorodon torvus
One-sixth natural size.
Specific cJiaracters. — Cope writes:
Top of head flat, forming a narrow plane between the temporal
fossae; latter produced backward. Orbit not inclosed behind,
an overhanging superciliary ridge. Nasal exceedingly short
and massive, each supporting a large acute horn core, which is
connected with its fellow by a ridge at the base and diverges
widely from it with an outward and forward curve to the
acutely compressed apex. Each horn core about 1 foot long.
The top of the head is plane between the orbits, and little
concave fore and aft. The zygoma is very deep, and the post-
glenoid process well developed. End of nasal bones short and
thick but flat.
Measurements
Millimeters
Length of cranium (35 inches) 895
Length from posterior rim temporal fossa to middle of super-
ciliary ridge 345
Width front between eyebrows 210
Length horn core on inner side (10 inches) 254
The elemental origin of the horn cores is probably different
in this genus from that which exists in Miobasileus.
Etymology. — acer, fierce, in allusion to the somewhat
ferocious appearance.
Present determination. — This valid species, which
pertains to the genus Megacerops, is fully described on
page 545.
212
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Megaceratops heloceras Cope, 1873
Cf. Menodus heloceras, this monograph, pages 524, 681
Original reference. — Pal. Bull. No. 15, p. 4, "issued
August 20, 1873" (Cope, 1873.2).
Subsequent reference. — Cope, Report on the verte-
brate paleontology of Colorado, pp. 487-488, 1874
(Cope, 1874.2).
Type locality and geologic horizon. — Horsetail Creek,
northeastern Colorado; Chadron formation {Titano-
therium zone), level not ascertained.
Type. — "A cranium * * * with nearly com-
plete maxillary dentition," anterior teeth and part of
frontals wanting. Am. Mus. 6360. (See fig. 168.)
Specific characters. — Cope writes:
There is a prominent horizontal superciliary ridge without
horns, and two short obtuse horn cores on the muzzle. These
Figure 167. — Type (holotype) skuU of Megaceratops acer
Am. Mus. 6348. After Cope, 1874. One-sixth natural size.
diverge outward, the outer sides being flattened and the sum-
mits contracted and truncate. They are mere rudiments of
the horns seen in M. aceronsor [sic], M. coloradoensis. The
molar teeth do not exhibit the T-shaped cross ridges seen in
Miobasileus, and the two outer crescents are continuous with
each other.
Measurements
Millimeters
Length from posterior rim of temporal fossa to middle of
osseous eyebrow 472
Least width of parietal plane 104
Superciliary width 260
Elevation of horn core 50
Etymology. — ^Xos, wart; xepas, horn; in allusion to
the wartlike appearance of the horn.
Present determination. — The species
referable to the genus Menodus. (See p.
Symborodon bucco Cope, 1873
Cf. Megacerops bucco, this monograph, page 544
Original reference. — Synopsis of new Vertebrata
from the Tertiary of Colorado, p. 11, 1873 (Cope,
1873.3).
is valid
524.)
but
Subsequent reference. — Cope, Report on the verte-
brate paleontology of Colorado, pp. 484, 485, pis.
2-4. 1874 (Cope, 1874.2).
Type locality and geologic horizon. — Horsetail Creek,
northeastern Colorado; Chadron formation {Titano-
therium zone), level not ascertained.
Cotypes. — In the original description Cope says the
species is represented "by an imperfect cranium; by
a cranium almost perfect, including very probably
both mandibular rami, with entire dentition; a frag-
mentary skeleton, including parts of cranium, teeth,
and vertebrae; and by a series of cervical and dorsal
vertebrae." Which of these cotype individuals thus
mentioned shall we select as the lectotype? If we
should take the first specimen mentioned, namely, the
imperfect cranium (known to be Am. Mus. 6346), we
find that since it consists of only the posterior portion
it lacks most of the characters given in the specific
description, except the single one of possessing ex-
panded zygomata (hence the name hucco). On the
other hand the "cranium almost perfect" (Am. Mus.
6345) also has expanded zygomata and was evidently
the chief specimen, since it furnished most of the
specific characters and measurements given in the origi-
nal description; moreover, in Cope's
fuller report (1874.2) it was figured
in Plates 2, 3, 4, imder the name
Symhorodon hucco, and in the key to
the species (p. 484), in which S. hucco
is contrasted with S. altirostris, the
diagnostic characters (referring to
the horns, premolars, nasals, de-
pressed cranium) are evidently from
the "cranium almost perfect" (No.
6345) rather than from the "im-
perfect cranium."
Lectotype. — From these clear indi-
cations of the author's intention the
skull (Am. Mus. 6345) may therefore be regarded
as the lectotype. (See figs. 169, 170.)
Specific characters. — Cope mentions especially the
enormous buccal expansion of the zygomata, the char-
acters of the horns, nasals, skull top, orbits, etc.
Specific characters are fully given on page 544.
Etymology. — hucco, one having extended cheeks.
Present determination. — This species is provisionally
regarded as a valid one.
Symborodon altirostris Cope, 1873
Cf. Megacerops acer, this monograph, page 545
Original reference. — Synopsis of new Vertebrata
from the Tertiary of Colorado, p. 12, 1873 (Cope,
1873.3).
Subsequent references. — Cope, Report on the verte-
brate paleontology of Colorado, p. 486, pis. 5, 6, 8,
fig. 1, 1874 (Cope, 1874.2); The Perissodactyla, pi. 33,
fig. a, opposite p. 1062, 1887 (Cope, 1887.1).
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
213
Type locality. — Cedar Creek, Logan County, Colo.
Type. — A cranium with premolar-molar teeth, zygo-
matic arches fractured (Am. Mus. 6350). (See figs.
170, 171.)
Characters of type (summarized from Cope). — Nasal
bones very short, broad, obtuse, massive, and "stand-
ing on a plane above that of the front." Orbit very
far forward. Horns straight, with approximated bases
Present determination. — As shown (p. 545), there are
reasons for regarding the type of S. altirostris as repre-
senting a female skull of Megacerops acer.
Symborodon trigonoceras Cope, 1873
Cf. Menodus trigonoceras, this monograph, page 528
Original reference. — Synopsis of new Vertebrata from
the Tertiary of Colorado, p. 13, 1873 (Cope, 1873.3).
1^
^^5
^^^^^N^^^H
H
Hk ';^^v^H
1
and moderately divergent, subcylindrical at base and
compressed inward and forward at the narrow apex.
"The first premolar and two incisors are very insig-
nificant; canines with short stout crowns." The pre-
molars with two smooth cones. Many other charac-
ters are given.
Etymology. — altus, high; rostris, beak, snout; in
allusion to the high position of the nasals.
FiGUEE 168. — Type (holotype) skull of Megaceratops heloceras
Am. Mus. 6360. After Cope. One-flfth natural size.
Subsequent references. — Cope, Eeport on vertebrate
paleontology of Colorado, 1874, p. 488, 1874 (Cope,
1874.2); The Perissodactyla, p. 1065, figs. 29, 30, 1887
(Cope, 1887.1).
Type locality and geologic Tiorizon. — ^Horsetail Creek,
northeastern Colorado; Chadron formation {Titano-
therium zone), level not ascertained.
214
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Cotypes. — Skull (Am. Mus. 6355) lacking all the
teeth except m^; Am. Mus. 6356, anterior-inferior por-
tion of skull, including horns, nasals, right zygoma,
and teeth. Of these two cotypes we may regard No.
6355 as the lectotype. (See fig. 172.)
Specific characters (summarized from Cope). — A
strong basal cingulum, on the inner side of the pre-
molars, which is continued in a less prominent form
Present determination. — This is a valid
described on page 528, referable to Menodus.
species,
Brontotherium ingens Marsh, 1873
Cf. Menodus giganteus, this monograph, page 530
Original reference. — Am. Jour. Sci., 3d ser., vol. 7,
pp. 85, 86, pis. 1, 2, January, 1874; "published Dec.
30, 1873" (Marsh, 1874.1).
■"-^fS^ ^
Figure 169. — Type (lectotype) skull of Symborodon bucco
Am. Mus. 6345. After Cope, 1874. One-ninth natural size. The mandible in the upper figure probably does not belong with the skull.
between the bases of the cones of the molars. Bases
of cones of premolars strongly plicate. Horns tri-
quetrous, dii-ected outward and upward. Squamosals
not expanded, nasals elongate transversely plane.
Etymology. — rpis, three; yuvla, angle; Kepas, horn;
in allusion to the three-sided section of the horn.
Subsequent reference. — Marsh, The principal char-
acters of the Brontotheridae, p. 335, text figs. 1, 2,
pis. 10, 11, 1876 (Marsh, 1876.1).
Type locality and geologic horizon. — Colorado; Chad-
ron formation {Titanotherium zone); exact locality
and level not recorded.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
215
Type. — A complete skull; premaxillaries with in-
cisors and canines wanting; nasals and horns partly
restored. Yale Mus. 2010. (See fig. 173.)
Characters. — Marsh says:
to the median line. The upper part of the horn cores is rugose,
and the base contains large air cavities. The free extremities
of the nasals are coossified and much elongated. They are
rounded in front, slightly decurved, and the surface at the end
is rugose. [Many other characters are listed.]
Figure 170. — Type skulls of Symhorodon altirostris (1), S. bucco (2), and
Megacerops acer (3)
Front views. After Cope, 1874. One-sixth natural size.
The most striking peculiarity of this cranium is the pair of
huge horn cores on the nasals. They are about 8 inches in
length and extend upward and outward. They are triangular
at the base, with the broadest face external. The two inner
faces of each core are separated by a ridge, which is continued |
Millimeters
Length of skull from occipital condyles to end of nasals (36
inches) 915
Distance on median line from occipital crest to end of
nasals 762
216
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Millimeters
Expanse of zygomatic arches 558
Least distance across vertex 157
Space occupied by four upper premolars 162
Space occupied by three upper true molars 266
Space occupied by molar-premolar series 428
Etymology. — ingens, vast.
Present determination. — The species is a synonym of
Menodus giganteus Pomel.
Symborodon hypoceras Cope, 1874
Cf. Brontotherium hypoceras, this monograph, page 562
Original reference. — U. S. Geol. and Geog. Survey
Terr. Ann. Rept. for 1873 (Hayden), p. 491 [no
figure], 1874 (Cope, 1874. 2).
cores of very different shape described below. (See
fig. 174.)
Specific characters. — Infraorbital foramen "fiat with
a wide external face, instead of being a cylindric col-
umn as in S. acer, altirostris, bucco, and ophryas."
One of the horn cores "consists of the extremital
part. * * * j^g section is a compressed oval
narrowed in front; its profile with parallel outlines and
a little recurved and not very rugose. Its size as
compared with the rest of the skull is the smallest in
the genus, and not more than half the proportions of
the S. altirostris." Another fragment Cope deter-
mined as a portion of the frontal bearing a "large
osseous tuberosity, which consists of a mass of bone
FiGTJRE 171. — Type (holotype) skull of Symborodon altirostris
Am. Mus. 6350. After Cope, 1874. One-sixth natural size.
Type locality. — ?Cedar Creek, Logan County, Colo.
Type. — Cope writes:
This species reposes on a fragmentary cranium only, which
embraces nasal, maxillary, frontal, malar bones, etc., both
zygomata, premolar, and parts of molar teeth. These frag-
ments were taken out of the matrix by the writer and were
found in juxtaposition. They represent parts of the same
skuU and, as no other was found in the same bank, are prob-
ably without admixture.
The only remains representing this type which are
now preserved in the American Museum of Natural
History (Am. Mus. 6361) include two portions of the
malar bones, a fragment of the orbit and infraorbital
canal, a fragment of the alveolar region, and two horn
coossified with the upper surface as in the horn of
the girafi'e." Cope concluded that "it is probable
that this species possessed two pairs of osseous proc-
esses or cores on each side, the one on the nasal, the
other on the frontal bone." The name "hypoceras"
doubtless referred to the supposed presence of the
second horn core (the rounded tuberosity) behind and
below the oval-sectioned horn on the nasals. Cope
gives 14 measurements, including the following:
Millimeters
Length from front of orbit to glenoid fossa (axial) 365
Depth of malar below orbit 20
Length of molars and last three premolars 293
Length of last three premolars 110
Diameter of horn core, transverse 38
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
217
Fixation of ledotype. — Cope's conclusion that the
above-mentioned fragments "are probably without
admixture" appears open to doubt. The "frontal
tuberosity" referred to is shown by comparison with
well-preserved material to be the horn core of the left
side of an immature individual resembling Allops
marshi, a reference favored by Cope's observation of
the wide bridge over the infraorbital foramen, which
contrasts with the narrow columnar bridge in Bron-
totheriwn and Symhorodon. The oval-sectioned horn
core which Cope supposed to be borne on the nasals is
a right horn core of very different shape, agreeing closely
with that in skull No. 4702, U. S. National Museum,
which Osborn selected (1902.208, p. 106) as the neo-
type of this species. The oval-sectioned horn core
may, therefore, be regarded as the ledotype.
Etymology. — invb, below; /cepas, horn; in allusion to
the supposed presence of a low horn swelling on the
frontal, behind the one on the nasal.
Present determination. — As thus interpreted, hypo-
ceras is a valid species of the genus BrontotJierium.
Anisacodon Marsh, 1875
Cf. Menodus, this monograph, page 522
Original reference. — Am. Jour. Sci., 3d ser., vol. 9,
p. 246, March, 1875 (Marsh, 1875.1).
Type species. — Anisacodon montanus Marsh (see
below).
Generic characters (Marsh). — "Dentition: Incisors-?;
canines y; premolars f ; molars f. No superior dias-
tema. Strong inner basal ridge on upper pre-
molars. Last upper molar with two inner cones. No
postorbital process."
Etymology. — iivtcT-os, unequal; (xktj, point; 66ous,
tooth. Possibly in allusion to the unequal develop-
ment of the two inner cones on the third upper molar.
Present determination. — In view of the strong
similarities to Menodus in the vestigial condition of
the incisors, in the strong internal cingulum in the
premolars, in the shape of the nasals, and in the second
internal cone of the third molar, this genus is now
regarded as a synonym of Menodus.
Anisacodon montanus Marsh, 1875
Cf. Menodus giganteus?, this monograph, page 537
Original reference. — ^Am. Jour. Sci., 3d ser., vol. 9,
p. 246, March, 1875 (Marsh, 1875.1).
Type locality and geologic horizon. — "Northern
Nebraska" (Big Badlands, White River, S. Dak.);
Chadron formation {Titanotherium zone); exact local-
ity and level not recorded.
Type. — A fragmentary skull including the maxil-
laries and fragmentary molar teeth. Yale Mus.
10022. (See fig. 175.)
Specific characters. ^Mursh. writes:
This species is especially distinguished by the emargination
of the extremity of the nasals, the short premaxillaries, and
101959— 29— VOL 1 17
the rectangular form of the last upper molar. The inner pos-
terior cone of this molar is smaller than the one in front, and
quite distinct from the posterior basal ridge.
Measurements [selected from Marsh]
Millimeters
Width of nasals above end of premaxillaries 95
Anteroposterior diameter of last upper premolar 43
Anteroposterior diameter of penultimate upper molar 77
Anteroposterior diameter of last upper molar 84
Etymology. — montanus, dwelling in the mountains.
Exact allusion uncertain, unless the badland topogra-
phy of South Dakota is thought of as mountainous.
\
Figure 172. — Type (holotype) skviH of Sy7nborodontrigonoceras
Am. Mus. 6355. One-ninth natural size.
Present determination. — In the form of its premolars
and third molar as well as in its vestigial incisors this
animal resembles Menodus giganteus; the emarginate
nasals with processes on either side of the median
notch also recall female Menodus skulls. Anisacodon
(Diconodon) montanus is probably referable to Meno-
dus cf. M. giganteus.
"Diconodon non Anisacodon" Marsh, 1876
Cf. Menodus giganteus, this monograph, page 530
Original reference. — Am. Jour. Sci., 3d ser., vol. 11,
p. 339, April, 1876 (Marsh, 1876.1). In this paper
Marsh gives diagnosis of four genera of Brontothe-
ridae. No. 4 is called "Diconodon Marsh (Anisaco-
218
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 173. — Type (holotype) skull of Brontotherium {=Menodus) ingens
Yale Mus. 2010. After Marsh, 1S74. About one-sixth natural size.
Figure 174. — Type (lectotype) of Symborodon {= Brontotherium) hypoceras
Am. Mus. 6361. One-half natural size. Fragment of right horn core: A', front view; A^, rear view; A', top view.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
219
Ann.
don). * * * Type D. montanus Marsh." The
term Anisacodon had been preoccupied by Anisacodon
Marsh, 1872, a genus of insectivores.
Etymology. — Sis, double; kuvos, cone; 65ovs, tooth.
Present determination. — See remarks under Anisaco-
don, above.
FIRST NOTICE OF CANADIAN TITANOTHERES BY COPE, 1886
Menodus angustigenis Cope, 1886
Cf. Megacerops angustigenis, this monograph, page 482,
fig. 176, Ci
Original reference. — Canada Geol. Survey
Kept., new ser., vol. 1, p. 81, 1886 (Cope,
1886.1).
Subseguent references. — " Haplacodon
angustigenis," The Vertebrata of the
Swift Current River, II, p. 153, 1889
(Cope, 1889.1); On Vertebrata from the
Tertiary and Cretaceous rocks of the
Northwest Territory, I, p. 13, pi. 5, figs.
1, 2; pi. 6, figs. 2, 2a; pi. 7, figs. 1, la, la
[bis], 1891 (Cop, 1891.2).
Type locality and geologic liorizon. —
Swift Current River, Assiniboia, Canada;
Cypress Hills beds, level not determined.
McConnell and Weston, collectors.
Cope's cotypes. — Cope writes:
This large mammal is represented by numer-
ous specimens. I select for present description
two maxillary bones from the same skull [fig.
176, A] [Cope, 1891.2, pi. 5, figs. 1, 2], each of
which contains the first [fourth] premolar and
the true molars; and two lower jaws from second
and third individuals [fig. 176, B]. One of these
[op. cit., pi. 7, figs. 1, la, la [bis], our fig. 176 C,
now regarded as the lectotype] consists of little
more than the symphysis. The other [op. cit.,
pi. 5, fig. 2; pi. 6, figs. 2, 2a] includes part of
the symphysis and the left ramus, which con-
tains all the molar teeth except the first and last.
[See fig. 176.]
Lectotype. — Of these semingly coequal
types or cotypes, which is to be regarded
as the lectotype? The one mentioned
first is "the two maxillary bones from
the same skull," but the mandibular
symphysis (op. cit., pi. 7, figs. 1, la, la
[bis]), from which the species evidently
takes its name (meaning narrow chin) , is
certainly to be selected as the lectotype.
Specific cTiaracters. — Cope's description
is too long to quote here. He compared
Menodus angustigenis with " Symhorodon trigonoceras"
and other species and gave numerous measurements.
Among the chief characters noted are the contracted
shape of the mandibular symphysis and the square
outline of the molars.
Etymology. — angustus, narrow; gena, chin.
Present determination. — As defined from the lecto-
type the species is provisionally referred to Megacerops,
although its generic reference is uncertain.
The maxilla with the dentition belongs to a very
different animal. It is apparently referable to Allops
sp. (See below.) The lower jaw appears to be
referable to Menodus cf . M. proutii.
SPECIES DESCRIBED BY SCOTT AUD OSBORN IN 1887
Menodus tichoceras Scott and Osborn, 1887
Cf. Brontotherium tichoceras, this monograph, page 565
Original reference. — Mus. Comp. Zoology BuL., vol.
13, No. 5, p. 159, text figs. 3, 2; 5, 2; 6, 2, 1887 (Scott
and Osborn, 1887.1).
Figure 175. — Type (holotype) of Anisacodon montanus
Yale Mus. 10022. A, Third right upper molar; B, fourth upper premolar and first and second molars; C,
alveoli of upper canines and incisors; Di, nasals, top view; Da, nasals, front view. All one-half
natural size.
Type locality and geologic Tvorizon. — Big Badlands,
South Dakota; exact locality and horizon not recorded.
S. Garman, collector.
Type.^&coit and Osborn describe the type as "a
large skull with the dentition complete, lacking the
upper part of the horns and the crest of the occiput."
Now in the Museum of Comparative Zoology at Cam-
bridge, Mass. (See fig. 177.)
220
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Characters (abbreviated from Scott and Osborn).
Dentition: I 2, C 1, P 4, M 3. The skull is described
as 29 inches [736 mm.] in length; with a narrow and
elevated anterior portion; nasals of medium length,
with short, obliquely placed horns, zygomatic arch
very massive, presenting a bulge in the posterior half
which is much less prominent than in S. iucco.
Etymology. — relxos, wall; Kipas, horn; possibly in allu-
sion to the high connecting crest.
Type locality and geologic horizon. — South Dakota;
Chadron formation {TitanotJierium zone); exact local-
ity and level not recorded.
Type. — A skull incomplete in the supraoccipital
region; zygomatic arch fragmentary; maxillary, pala-
tine, and basioccipital regions much distorted. Now
in the Museum of Comparative Zoology, Harvard
University. (See fig. 177.)
Characters. — Scott and Osborn write:
Figure 176. — Cope's cotypes of Menodus angustigenis
A, Right maxilla (subsequently made the type of Haplacodon angustigenis), three-sixteenths natural size; B, left halt of a lower jaw (now referred to Menodus
sp.) , three-sixteenths natural size; C, symphysis mandibulae (leetotype), one-third natural size (Ci, front; Ci, right side; C3, under side).
Present determination. — This species is provisionally
referred to Brontotherium, but its exact position in
that phylum is uncertain. (See p. 565.)
Menodus dolichoceras Scott and Osborn, 1887
Cf. Brontotherium dolichoceras, this monograph, page 572
Original reference. — Mus. Comp. Zoology Bull., vol.
13, No. 5, p. 160, figs. 3, 3; 5, 3; 6, 3, 1887 (Scott and
Osborn, 1887.1).
Dentition: I ?, C^-, P^, M^. Upper premolars with a faint
internal cingulum. Nasal bones extremely short and obtuse.
Horns extremely long and powerful, directed obliquely forward
and outward, projecting beyond the nasals in side view. The
section is suboval at the base, with the long axis obliquely
transverse. Cranium very broad and saddle-shaped above
the orbits, narrowing somewhat posteriorly. A prominent and
overhanging superciliary ridge. Postglenoid and post-tym-
panic processes united for a short distance. The skull which
we have made the type of this species is much larger and more
powerful than Professor Cope's type of jS. acer. The horns are
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
221
longer and more widely divergent at the base. The angle of
inclination of the horns and the diminutive proportions of the
nasals, as well as the form of the top of the cranium, all bring
this specimen near S. acer and separate it from other known
species. Unlike S. acer, the horns are not united by a ridge.
[This is an error.] The specimen is incomplete in the supra-
occipital region, the zygomatic arch is fragmentary, and the
maxillary, palatine, and basioccipital regions are much dis-
torted.
Menodus platyceras Scott and Osborn, 1887
Cf. Brontotherium platyceras, this monograph, page 578
Original reference. — Mus. Comp. Zoology Bull., vol.
13, No. 5, pp. 160, 161, fig. 4, 1887 (Scott and Osborn,
1887.1).
Subsequent reference. — The cranial evolution of Tita-
notherium, p. 186, fig. 7A, 1896 (Osborn, 1896.110).
The specimen figured is not the type.
Figure 177. — Anterior part of skulls of (1) " Megacerops colorodensis" (not the type), now referred to Allops
marshi; (2) Menodus tichoceras (type) (present determination, Brontotherium tichoceras) ; and (3) Menodus
dolichoceras (type) (present determination, Brontotherium dolichoceras)
Specimens in the Museum of Comparative Zoology, Harvard University. After Scott and Osborn, 1887. Greatly reduced in size. A, Side
views: B, front views, showing the variations in the horns, nasals, and anterior nares; O, top views, showing the nasals and horns, and sections
of the bases of the horns.
Revised measurements
IVIilUmeters
Occipital condyles to nasal tips 690
Free length of nasals 46
Free breadth of nasals 90
Outside measurement of horns 310
Anteroposterior diameter of horns 85
Transverse diameter of horns 125
Etymology. — 56\ixoi, long; Kepas, horn.
Present determination. — As shown in Chapter VI
the present species probably pertains to Brontotherium
rather than to Symhorodon.
Type locality and geologic horizon. — Big Badlands,
South Dakota; Chadron forma'tion (Titanotherium
zone, Chadron C); exact locality and level not re-
corded. S. Garman, collector.
Type. — A pair of horns with the nasal bones at-
tached. Now in the Museum of Comparative Zoology
at Cambridge, Mass. (See fig. 178.)
Neotype.SknW (Am. Mus. 1448).
Characters. — Scott and Osborn write:
Nasal bones e.xtremely short and obtuse, as in M. dolicho-
ceras and M. acer. The inner [posterior] contour of the horns
222
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
is concave; they are greatly flattened anteroposteriorly, with
a ridgelilie outer margin, and connected by a well-raised median
Figure 178. — Type (holotype) horns of
Menodus platyceras
In the collection of the ]Museum of Comparative
Zoology, Harvard University. After Scott and
Osborn, 1887. Greatly reduced. A, Front view; B,
cross section: C, side view.
ridge. The posterior face is nearly plane, the anterior is con-
vex, so that the section of the horn is plano-convex from base
SPECIES DESCRIBED BY MARSH IN 1887
Brontops Marsh, 1887
Cf. Brontops, this monograph, page 482
Original reference. — Am. Jour. Sci., 3d ser., vol. 34,
p. 326, October, 1887 (Marsh, 1887.1).
Type species. — Brontops rohustus Marsh. (See
below.)
Generic characters. — Marsh writes:
The present genus is quite distinct from any of the forms
previously described. * * * xhe skull is large and massive,
with widely expanded zygomatic arches, and short and robust
horn cores, projecting well forward. In general form it re-
sembles the skull of Brontotherium but may be readily dis-
tinguished from it by the dental formula, which is as follows:
Incisors xi canines-}-; premolars |; molars |.
The presence of four premolars in each ramus of the lower
jaw is a distinctive feature in this genus. This character, with
the single, well-developed incisor, marks both the known species
[B. robusius, B. dispar].
FitiUHE 179. — Tj'pe (holotype) skeleton of Brontops, robusius
Yale Mus. 12048. After Marsh, 1889. One twenty-fourth natural size.
to tip. In side view the horns completely overhang the nasals
and are slightly recurved. The long axis of the horn section is
[almost or quite] directly transverse.
Measure7nents
Millimeter.'!
Outside length of horns 315
Transverse diameter of horns 125
Anteroposterior diameter of horns . 67
The type probably belongs to a young male in which the
horns are not fully developed, because the horns increase in
width and flatness and the basal section becomes more truly
transverse, with age.
Etymology. — irXarvs, flat; /cepas, horn.
Present determination. — This valid species, described
on page 578, belongs in the genus Brontotherium.
Etymology. — Brontotherium; &^, face, "having the
face or appearance of"; resembling Brontotherium.
Present determination.— In 1902 Osborn (1902.208)
treated Brontops as a synonym of Megacerops Leidy,
but renewed examination of Leidy's figure of M.
coloradensis indicates that it is not congeneric with
Brontops, which is here regarded as a valid genus.
Brontops robustus Marsh, 1887
Cf. Brontops robustus, this monograph, page 492
Original reference. — Am. Jour. Sci., 3d ser., vol. 34,
p. 326, October, 1887 (Marsh, 1887.1).
Subsequent references. — Restoration of Brontops ro-
iustus: Am. Jour. Sci., 3d ser., vol. 37, pp. 163-165, pi.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
223
6, 1889 (Marsh, 1889.1); skeleton and restoration, this
monograph. Plates XCVI-CIII, CXCV-CCXXIX.
Type locality and geologic horizon. — "Near the
White River in northern Nebraska. " "The geological
horizon is in the upper part of the Brontotheriwn
beds [Chadron formation, Titanotherium zone] "
(Marsh). "Upper levels of middle beds at least 60
feet below the top of the upper beds" (Hatcher, 1901).
Figure 180. — Type (holotype) lower jaw of Brontops dispar
Nat. Mas. 4941. After Marsh, 1887. One-eighth natural size.
Type. — A skull and skeleton, Yale Mus. 12048.
(See fig. 179.)
Specific characters. — Marsh did not formally sepa-
rate the specific from the generic characters. He
records the fact that the skull is large and massive,
with widely expanding zygomatic arches and stout,
robust horn cores, projecting well forward.
Etymology. — roiustus, robust (that is, strong as an
oak, rohur).
Present determination. — The genus and species are
valid. The species is described also on pages 492-499.
Brontops dispar Marsh, 1887
Cf. Brontops dispar, this monograph, page 488
Original reference. — Am. Jour. Sci., 3d ser., vol. 34,
pp. 327, 329, figs. 7, 8 (jaw); not figs. 5, 6 (skull),
October, 1887 (Marsh, 1887.1).
Type locality and geologic horizon. — Found on Hat
Creek, Sioux County, Nebr., by J. B. Hatcher, May
14, 1886; Chadron formation {Titanotherium zone),
middle level.
Type. — "A nearly complete skull with lower jaws
and entire dentition." (Marsh.) Nat. Mus. 4941
(skull D). (See fig. 180.)
Characters. — Marsh writes: "The skull is less mas-
sive and proportionately more elongate than in the
type species, and the lower jaw more slender." In
the same brief passage Marsh described a young skull
(Nat. Mus. 4258) as belonging to the same species;
this is a somewhat more primitive type (Brontops
brachycephalus) belonging to a younger individual
(p. 483).
Etymology. — dispar, uneven, probably in allusion to
the asymmetrical distortion of the type skull.
Present determination. — The species is valid and is
now referred to Brontops.
Menops Marsh, 1887
Cf. Menodus, this monograph, page 522
Original reference. — Am. Jour. Sci., 3d ser., vol. 34,
p. 328, October, 1887 (Marsh, 1887.1).
Type species. — Menops varians. (See below.)
Generic characters. — Marsh writes:
The present genus is most nearly related to Diconodon and
in its molar teeth agrees with that form. It differs in the
presence of two upper incisors on each side. The superior
dentition is as follows: Incisors, 2; canine, 1; premolars, 4;
molars, 3.
Etymology. — Menodus; &\p,i ace; resembling Menodus
(cf. Brontops, above).
Present determination. — The incisors are vestigial,
the alveoli being very small. The skull presents re-
semblance to both Menodus and Allops. The generic
reference is to Menodus.
Menops varians Marsh, 1887
Cf. Menodus varians, this monograph, page 535
Original reference. — Am. Jour. Sci., 3d ser., vol. 34,
p. 328, fig. 9, October, 1887 (Marsh, 1887.1).
Type locality and geologic horizon. — " Brontotherium
beds of Dakota" (Chadron formation, Titanotherium
zone); exact locality and level not stated. George
A. Clarke, collector.
FiGUKE 181. — Type (holotype) skull of Menops varians
Yale Mus. 120G0. Front view. One-eighth natural size.
Type. — A well-preserved skull (Yale Mus. 12060).
(See fig. 181.)
Specific characters.— Not separated by Marsh from
generic characters. (See above.)
Etymology. — varians, variant; allusion doubtful, but
possibly to the somewhat aberrant character of the
type skull.
Present determination. — The species is valid and is
referred to Menodus.
224
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Titanops Marsh, 1887
Cf. Brontolherium, this monograph, page 555
Original reference. — Am. Jour. Sci., 3d ser., vol. 34,
p. 330, October, 1887 (Marsh, 1887.1).
Type species. — Titanops curtus. (See below.)
Figure 182. — Type (holotype) skull of Titanops curtus
Front view. Yale Mus. 12013. After Marsh, 1887. One-eighth natural size.
Generic characters. — Marsh writes:
This genus contains the largest members of the Brouto-
theridae and some of the last survivors of the group. They are
distinguished from all the other known types by the long,
narrow skulls, lofty, flat horn cores, and short nasals. The
upper dentition corresponds nearly to that of Brontotherium,
but the upper molars have all two inner cones. * * * The
nasals are the shortest known in the group.
Etymology. — Titanotlierium; u^/, face — that is, like
Titanotlierium.
Present determination. — The genus is a synonym of
Brontotherium Marsh.
Titanops curtus Marsh, 1887
Cf. Brontotherium curtum, this monograph, page 574
Original reference. — Am. Jour. Sci. 3d ser., vol. 34,
p. 330, fig. 11, October, 1887 (Marsh, 1887.1).
Type locality and geologic horizon. — Colorado; e.xact
locality not stated but recorded by Hatcher (1901) as
from the upper Titanotlierium zone [of Chadron
formation].
Type. — A complete skull with teeth (Yale Mus.
12013). (See fig. 182.)
Specific characters. — Not separated from generic
characters by Marsh.
Etymology. — curtus, short; in allusion to the short
nasals.
Present determination. — The species is valid and is
referred to Brontotherium.
Titanops elatus Marsh, 1887
Cf. Brontotherium gigas, this monograph, page 567
Original reference. — -Am. Jour. Sci., 3d ser., vol. 34,
p. 330, fig. 12, October, 1887 (Marsh, 1887.1).
Type locality and geologic horizon. — "Upper Titano.
therium zone, South Dakota" (Chadron formation).
Type.— A skull and jaw (Yale Mus. 12061). (See
fig. 183.)
Specific characters. — Marsh writes:
The nasals are much longer, and the occipital crest much
higher, than in the type species [T. curtus]. The zygomatic
arches are unfortunately wanting, but the lower jaw is present,
nearly in place. It shows no marked characters different
from that of Brontops.
Etymology. — elatus, lofty; possibly in allusion to the
high stage of specialization.
Present determination. — The species is synonymous
with Brontotherium gigas Marsh.
Allops Marsh, 1887
Cf. Allops, this monograph, page 506
Original reference. — Am. Jour. Sci., 3d ser., vol. 34,
p. 331, October, 1887 (Marsh, 1887.1).
Type species. — Allops serotinus. (See below.)
Generic and specific characters. — Marsh writes:
This skull in its general form resembles that of Brontotherium,
but differs in having only a single upper incisor, and the last
molar has the posterior inner cone more strongly developed.
The superior dentition is as follows: Incisor, 1; canine, 1;
premolars, 4; molars, 3.
In the tj'pe specimen the canine is small, extending but
httle below the premolars. There is no diastema. The upper
premolars have a very strong inner basal ridge. The nasals
are wide, expand forward in the free portion, and are notched
in front. The entire length of the skull is 31 inches (79 centi-
meters), the distance across the zygomatic arches 21 inches
(53 centimeters), and the length of the horn cores about 10
inches (25 centimeters).
Figure 183. — Type (holotype) skull of Titanops elatus
Front view. Yale Mus. 12061. After Marsh, 1887. One-eighth natural size.
Etymology. — aXXos, strange; ciiA, face.
Present determination. — Allops is intermediate be-
tween Menodus and Brontops and is here regarded as
a valid genus. (See p. 506.)
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
225
Allops serotinus Marsh, 1887
Cf. Allops serotinus, this monograph, page 515
Original reference. — Am. Jour. Sci., 3d ser., vol. 34,
p. 331, October, 1887 (Marsh, 1887.1).
Type locality and geologic horizon. — Quinn Draw,
South Dakota, "near the top of the Brontotherium
beds," Chadron formation {TitanotTierium zone).
Figure 184. — Type (liolotype) skull of Allops serotinus
After Marsh. Nat. Mus. 4251. One-seventh natural size.
Type. — "A well-preserved skull and various other
remains." U. S. Nat. Mus. 4251. J. B. Hatcher,
collector. (See fig. 184.)
Specific characters. — Not separated from generic
characters in original description.
Etymology. — serotinus, from sero{1), to bind, connect;
possibly because the characters appeared to be more
or less annectant with those of other species.
Present determination. — The species is valid. It is
described on page 515.
CANADIAN SPECIES DESCRIBED BY COPE IN 1889
Haplacodon Cope, 1889
Cf. Allops, this monograph, page 506
Original reference. — Am. Naturalist, vol. 23, p. 153,
March, 1889 (Cope, 1889.1). (See p. 202.)
Type species. — Menodus angustigenis Cope. The
genus was founded on the characters of one of the
several "types" of Menodus angustigenis, namely, a
maxilla containing the fourth upper premolar and the
three molars.
Generic characters. — Cope writes:
It differs from all the genera of the Menodontidae in the
presence of but a single internal cusp of the first (posterior)
superior premolar, a fact which renders it highly probable
that the premolars which precede it in the maxillary bone
were similarly constituted. It differs from all other genera of
Lambdotheriidae and also from Diplacodon, to which it is
allied, in the presence of but two inferior incisors on each side.
It is not certain whether it possesses horns or not.
Comparative measurements of the type of "Haplacodon" Cope, in
millimeters
M'-m'
P', ap. by tr__
M', ap. by tr_
M\ ap. by tr_
M^, ap. by tr_
187
38X62
50X52
66X62
65X62
169
35X51
45X51
61X61
60X61
Etymology. — awXoos, simple; aKri, cone; 65ous, tooth; in
allusion to the "single internal cusp" of the fourth
upper premolar.
Present determination. — The upper teeth agree
closely in general characters with those of Allops and
are intermediate in size between Allops walcotti and
Allops marshi.
A2
A3
Figure 185. — Type of Menodus selwynianus
Coossifled nasal. Ai, Left side; Aj, upper side; A3, under side. Three-eighths
natural size.
Menodus selwynianus Cope, 1889
Cf. Diploclonus selwynianus, this monograph, page 502
Original reference. — Am. Naturalist, vol. 23, p. 628,
July, 1889 (Cope, 1889.2).
Subsequent reference. — On Vertebrata from the
Tertiary and Cretaceous rocks of the Northwest
Territory, I, p. 17, pi. 5, figs. 3, 3a, 3b, 1891 (Cope,
1891.2).
226
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Type locality and geologic horizon. — Swift Current
River, Assiniboia, Canada; Cypress Hills beds, level
not recorded.
Type. — Coossified nasal bones detached from skull.
Ottawa Mus. (See fig. 185.)
Figure 186. — Type of Menodus syceras
Coossified nasal and left horn eoie. After Cope. Ai, Leftside; A2, front; A3, section of left horn. One-half
Characters of type. — Cope writes:
Represented by a nasal process, which consists of the coossi-
fied nasal bones, of peculiar form. They are elongate as com-
pared with their width and are vaulted. The lateral borders
are nearly parallel, and the extremity viewed from above is
rounded. Owing to the thickness of the body, the profile
descends abruptly at the extremity, and the convex surface is
roughened as though for the attachment of some fixed body,
tegumentary or muscular. From this tuberosity the surface
descends steeply to a thin border. A short distance posterior
to the extremity the lateral margins are
decurved, forming the lateral walls of a
deep longitudinal median gutter-like nasal
meatus, which is deeper than in any other
species. The horns are broken off, but the
median inferior surface is so little recurved
laterally that it is evident that the former
were not only small but laterally placed.
Length of fragment above, millimeters, 130;
length of nasal border, 70; width at nasal
notch, 80; width near extremity, 65; depth
at apical tuberosity, 26.
Additional ohservations. — The
lower surface of the horns in the
type exhibits a portion of the frontal
sinus. The nasals are shorter than
in the type of M. coloradensis. The
measurements are as follows:
Millimeteis
Free length of nasals 80
Free width of nasals 101
Outside measurement of horns 67
Anteroposterior measurement of horns 79
Etymology. — "This species is dedi-
cated to Dr. A. R. C. Selwyn, the
accomplished Director of the Survey
of Canada." (Cope.)
Present determination. — The species
is probably allied to Diploclonus
hicornutus (Osborn).
Menodus syceras Cope, 1889
Cf. Megacerops syceras, this monograph,
page 549
Original reference. — Am. Natu-
ralist, vol. 23, pp. 628-629, July,
1889 (Cope, 1889.2).
Subsequent reference. — Cope, On
Vertebrata from the Tertiary and
Cretaceous rocks of the Northwest
Territory, I, p. 18, pis. 7, fig. 2; 8,
figs. 4, 5, 1891 (Cope, 1891.2).
Type locality and geologic hori-
zon.— Swift Current River, Assiniboia,
Canada; Cypress Hills beds, level
not recorded.
The nasal bones of three individuals
present the characters above given." Of these we may
select as the lectotype the specimen figured by Cope
(1891.2, pi. 8) that shows the character from which
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
227
the name syceras is derived, in reference to the approxi-
mation of the horns at their bases. Portion of right
frontal, coossified nasals, and right horn. (See fig.
186.)
Characters of type. — Cope writes:
It differs from the two species of that group now known,
the M. proutii Leidy and the M. Hchoceras S. and 0., in the
very close approximation of the basis of the horns and the pres-
ence of a strong angle or ridge connecting them, so that the nasal
bones are in a different plane from that of the front. The
entire width of the skull at the basis of the horns is not greater
than the length of each horn above the nasal notch. The
horns are not long, and the section of their base is a longitudinal
oval, flattened on the external side. Summit subround. The
nasal bones are fiat, with broadly rounded extremitj', and are
much wider than long.
The width of the nasals at the base of the horns is 116 milli-
meters; length of do. from do., 70; diameters of bases of horns;
anteroposterior, 94; transverse, 67; length of horn from nasal
notch, 160; width of muzzle at bases of horns inclusive, 160.
Etymology. — aiiv, together; /cepas, horn; because the
horns were set very near to each other at the base.
Present determination. — M. syceras is at present
indeterminate or possibly a synonym of M. angustigenis,
both are provisionally referred to the genus Megacerops.
LAST FIVE SPECIES DESCRIBED BY MARSH, 1890-91
Diploclonus Marsh, 1890
Cf. Diploclonus, this monograph, page 499
Original reference. — Am. Jour. Sci., 3d ser., vol. 39,
p. 523, June, 1890 (Marsh, 1890.1).
Type species. — Diploclonus amplus. (See below.)
Characters. — Marsh writes:
One of the most marked features is seen in the horn cores,
which are high, compressed transversely, and have a prominent
knob on the inner superior margin about one-third of the dis-
tance to the summit. Seen from the front the horn cores thus
appear to be branched. It is probable that in life this feature
was still more evident, and the covering of the horn core may
have shown an actual division, but this can not be determined
from the present specimen. There is a sharp ridge at the base
of the horn cores on the outside. The nasals project but very
little in front of the horn cores. The zygomatic arches are
especially strong and widely expanded. The posterior nares
have their front margin opposite the back of the last upper
molars.
There were apparently but two upper incisors — that is, one on
each side — and no diastema exists behind the canines. The
premolars have a strong inner basal ridge, and the last upper
molar has two inner cones. This genus appears to be most
nearly related to T'iianops, but the horn cores will distinguish
it readily from all known forms of the Brontotheridae.
Etymology. — 5t7rX6os, double; kXwv, a twig; in allusion
to the branched appearance of the "horn core."
Present determination. — The genus is now regarded
as valid by Osborn.
Diploclonus amplus Marsh, 1890
Cf. Brontops amplus, this monograph, page 504
Original reference. — Am. Jour. Sci., 3d ser., vol. 39,
p. 523, June, 1890 (Marsh, 1890.1).
Type locality and geologic horizon. — South Dakota;
" Brontotherium beds" (= Chadron formation, or Titan-
otherium zone).
Type. — "Nearly complete skull, in good preserva-
tion, but without the lower jaws." Yale Mus. 12015a.
(See fig. 187.)
Specific characters. — Marsh writes:
The skull measures 28 inches from the front of the nasals to
the back of the occipital condyles and 24 inches in greatest
width across the zygomatic arches. The space occupied by the
upper dental series is 13J^ inches, and by the true molars 8
inches.
Etymology. — amplus, broad, in allusion to the great
breadth of the skull.
FiGUEE 187. — Type skull of Diploclonus amplus
After Marsh. One-eighth natural size. A', Side view; A', front view
Present determination. — This species is cither an
aberrant stage in the evolution of Brontops — that is,
a sport — or a lateral phylum of Brontops.
Teleodus Marsh, 1890
Cf. Teleodus, this monograph, page 481
Original reference. — Am. Jour. Sci., 3d ser., vol.
39, p 524, June, 1890 (Marsh, 1890.1).
Type species. — Teleodus avus Marsh. (See below.)
Generic characters. — Marsh writes:
The present genus is allied to Brontotherium but differs from
it in having six lower incisors instead of four. It has the same
number of inferior premolars and molars, and these teeth are
similar in the two genera. From Diplacodon of the upper
228
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Eocene the present genus may be distinguished by having
only three lower premolars on a side instead of four. * * *
Of the three lower incisors in place on each side, the middle
one is the largest. There is a short diastema behind the
lower canine, but no first premolar. The dental formula of
the lower jaws is as follows: Incisors, 3; canine, 1; premolars,
3; molars, 3.
The space occupied by the lower dental series is 143^ inches,
and by the last three molars 8H inches.
Etymology. — reXeos, distant; dSovs, tooth; in allusion
to the peculiar character of the incisors.
Present determination. — The genus may either be
valid or synonymous with an early stage in the evolu-
tion of Brontops.
Teleodus avus Marsh, 1890
Cf. Teleodus avus, this monograph, page 481
Original reference. — Am. Jour. Sci., 3d ser., vol. 39,
pp. 523, 524, June, 1890 (Marsh, 1890.1).
Type locality and geologic horizon. — " BrontotJierium
beds of Dakota" ( = Chadron formation, or Titano-
therium zone) ; exact
geologic level not re-
corded but probably
lower beds (Chad-
ron A).
Type . — A lo wer j aw.
Yale Mus. 10321.
(See fig. 188.)
Specific cliaracters. —
Not separated by
Marsh from the gen-
eric characters. (See
Type ol Teleodus avus p. 481.)
Etymology. — avus,
grandfather; in allu-
sion to the primitive character of the animal.
Present determination. — The species is probably
valid.
Allops crassicornis Marsh, 1891
Cf. Allops crassicornis, this monograph, page 517
Original reference. — Am. Jour. Sci., 3d ser., vol. 42,
p. 268, September, 1891 (Marsh, 1891.1).
Type locality and geologic horizon. — "BrontotJierium
beds of South Dakota" ( = Chadron formation, or
Titanotherium zone). Geologic level as recorded by
J. B. Hatcher, collector, is the lower portion of the
upper Titanotherium zone (Chadron C).
Type. — A "nearly perfect skull of an adult but not
old animal." Nat. Mus. 4289. (See fig. 189.)
Specific characters. — Marsh writes:
The skull is of medium size, with the zygomatic arches moder-
ately expanded. The nasal bones do not project beyond the
premaxillaries. The horn cores are very short and massive,
with rounded summits, and thus form one of the striking fea-
tures of the skull. The dentition is complete and in fine pres-
ervation. The single incisor is quite small and situated close
Figure 188.-
Lower incisors and canines,
natural size.
to the canine. The latter is of moderate size and projects
but little above the rest of the dental series. There is no
diastema between the canine and the first premolar, which is
small and has its inner face on a line between the canine and
the second premolar. The second, third, and fourth premolars
are large and have a strong inner basal ridge. The last molar
has its anterior margin somewhat in advance of the front
border of the posterior nares.
The length of this skull on the median line is about 30
inches, and the width across the zygomatic arches 23 inches.
The width across the horn cores is 14 inches. The extent of
the superior dental series is 16 inches.
Etymology. — crassus, thick; cornus, horn.
Present determination. — The species is valid. It is
fully described on page 517.
Brontops validus Marsh, 1891
Cf. Brontops dispar, this monograph, pages 230, 488
Original reference. — Am. Jour. Sci., 3d ser., vol. 42,
p. 269, September, 1891 (Marsh, 1891.1).
Type locality and geologic horizon. — The geologic
level as recorded by J. B. Hatcher, collector, is the
"middle portion of the middle Titanotherium beds,
White River, S. Dak." (Chadron formation, horizon
Chadron B).
Type. — A "skull in fine preservation." Nat. Mus.
4290 (skull K). (See fig. 190.)
Specific characters. — Marsh writes:
[The skull] agrees in its main characters with the other
species of the genus but is particularly short and robust. The
zygomatic arches are widely expanded, almost as much as in
any skull of this group. The nasal bones have only a moderate
extension in front and do not reach the end of the premaxil-
laries. The free portion is broad and massive. The horn cores
are of moderate size, nearly round in section, and have their
obtuse summits directed somewhat backward. The occipital
crest slopes forward and is expanded transversely. The length
of this skull on the median line is about 26 inches. The great-
est transverse diameter across the zygomatic arches is 22 inches,
and across the summits of the horn cores 14 inches.
Etymology. — validus, stout, brave.
Present determination. — As shown on page 202, this
species is probably synonymous with Brontops dispar.
Titanops medius Marsh, 1891
Cf. Broniotherium medium, this monograph, page 573
Original reference. — Am. Jour. Sci., 3d ser., vol. 42,
p. 269, September, 1891 (Marsh, 1891.1).
Type locality and geologic horizon. — "Near the top of
the Brontotherium beds of South Dakota" (Chadron
formation, Titanotherium zone). J. B. Hatcher, col-
lector.
Type. — "One skull in fair preservation with the
horn cores and dentition complete." Nat. Mus. 4256.
(See fig. 191.)
Specific characters. — Marsh writes:
The free portion of the nasals is very small and projects but
slightly beyond the anterior line of the horn cores. The latter
are compressed anteroposteriorly and project laterally nearly
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
229
at right angles to the median line of the skull. The two in-
cisors on each side are quite small and separated from each
other and from the canine. There is a slight diastema behind
the canine. The first premolar is small and triangular in out-
line. The second premolar is of moderate size, and the third
and fourth premolars have only an incomplete inner basal ridge.
The width of this skull across the horn cores is 23 inches,
and the distance from the end of the nasals to the front of the
XAST SPECIES DESCRIBED BY COPE, 1891
Menodus peltoceras Cope, 1891
Cf. Brontotherium curtum, this monograph, page 574
Original reference. — Am. Naturalist, vol. 25, p. 48,
January, 1891 (Cope, 1891.1).
FiGUKE 189. — Type skull of Allops crassicornis
Palatal view. Nat. Mus. 4289. After Marsh. One-fifth natural size.
posterior nares is 16 inches. The extent of the upper dental
series is 17 inches.
Etymology. — medius, middle; in allusion to the in-
termediate character (between the species elatus and
curtus) of this form.
Present determination. — The species is probably
valid and is referable to Brontotherium.
Type locality and geologic horizon. — " Titanotherium
beds of northern Nebraska" (Chadron formation).
Type. — "Represented by the nasal region and the
horn cores; the apex of one of the latter being broken
away." Am. Mus. 10719. Dr. Hobart Hare, col-
lector, Nebraska. Presented by the Museum of the
University of Pennsylvania. (See fig. 192.)
230
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Specific characters. — Cope writes: •
The peculiarity of the species consists in the immense trans-
verse extent of the horn cores and their complete fusion into an
osseous wall which extends across the muzzle, forming a huge
plate or shield. The superior border of this shield is moderately
concave, a protuberant angle on each side representing the apex
of each horn core. The nasal bones form a flattened protuber-
ance much wider than long, which overhangs the nares. * * *
Measurements: Elevation of horn-core plate at middle line
behind, 180 millimeters; do. at lateral apex, 190 millimeters;
total width of do. at middle, 300 millimeters. Projection of
nasal bones beyond lateral base of horn-core plate, 20 milli-
meters; width of nasal meatus at base of nasal bones, 65 milli-
mieters; width of base of horn-core plate outside of nasal
meatus, 90 millimeters. Anteroposterior diameter of base of
, Figure 190. — Type (holot3'pe) skull of Brontops validus
After Marsh. Nat. Mus. 4290. One-eighth natural size,
horn core above side of and parallel to nasal meatus, 85 milli-
meters. This species is nearest the M. platyceras S. and O.,
which has transverse compressed horn cores. They are, how-
ever, distinct from each other, and not nearly so expanded
transversely as in the present form. The M. pelioceras, in fact,
carried a transverse shield on the end of its nose, which must
have given it an extraordinary appearance.
Etymology. — we\Tri, small shield; Kepa^, horn; be-
cause the bases of the horns formed together a "huge
plate or shield" extending across the muzzle.
Present determination. — The type specimen (fig. 192)
possibly represents a female of one of the long-horned
species of Brontotherium, perhaps B. curtum, B. platy-
ceras, or B. ramosum. The species is therefore practi-
cally indeterminate at present.
FIRST EUROPEAN OLIGOCENE SPECIES, DESCRIBED BY
TOULA, 1892
Menodus? rumelicus Toula, 1892
Cf. Brontothermm rumelicum, this monograph, pages 660, 941
Original reference. — Akad. Wiss. Wien, Math.-nat.
Classe, Sitzungsber., Band 101, Abt. 1, pp. 608-615,
1 pi., May, 1892 (Toula, 1892.1).
Subsequent reference. — Ueber einen neuen Rest von
Leptodonf (Titanotherium?) rumelicus Toula spec, pp.
922-924, 1896 (Toula, 1896.1).
Type locality and geologic horizon. — Near the railroad
on the Jambol line near Kajali, northwest of Burgas,
eastern Rumelia. Level, lower Oligocene (?).
The formation from which the type was re-
corded was correlated by Toula with the
" Belvedereschotter."
On account of the extreme rarity of titano-
theres in Europe it seems important to
note the published evidence concerning the
provenience of the type and referred speci-
mens of this species. According to Toula the
specimens were received from his friend G. N.
Zlatarski in Sofia. Toula does not state that
Zlatarski himself collected the specimens.
He states only that they must have come
from near the railroad at Kajali, from the
great heaps of material which had been dug
up in the search for usable rubble ("taugli-
chem Schotter"), and that these "Schotter-
massen" should correspond at best with that
isolated remnant of a formation at Lidscha,
northwest of Burgas, of which he had already
spoken in his first report on the geology of the
eastern Balkans. He writes : "I have referred
to these 'Schotter' as Belvedereschotter, and
I believe, from the condition of preservation
of the specimens from Kajali, and especially
from the rusty sand grains still adhering to
them, that they must be referred to the
same kind of rock." Besides the specimens
of titanotheres Toula records a lower molar
and a canine of a "middle-sized rhinoceros"
from the same locality. Later he received
from the same locality, this also from Zla-
tarski, a fragment of the lower jaw of a titanothere
that included the symphyseal region (Toula, 1896.1,
pp. 922-924). But Toula has not disproved the
possibility that these specimens may have been im-
ported from America, perhaps by laborers returning
home from the western United States. (See p. 560.)
Lectotype. — Third right lower molar and part of the
right ramus of the lower jaw. (See fig. 193.)
Paratypes. — A second right lower molar and a
canine.
Referred specimen. — The symphyseal region of the
jaw with the roots of pi, p2, Ps, and the worn p4,
in place.
DISCOVERY OF THE TITANOTHEEES AND ORIGINAL DESCRIPTIONS
231
Specific characters. — Toula's description is too long
to quote here. The principal characters revealed by
his figures are, symphysis massive, canines (?) large,
four lower premolars, lower molars with faint external
cingula, hypoconulid of ms without strong internal
crest.
Etymology. — rumelicus, in allusion to Rumelia, the
region in Hungary where the type was discovered.
Present determination. — The species is probably
valid, and its generic reference is probably to
Brontotherium.
SPECIES DESCRIBED BY OSBOEN IN 1896 AND 1902
Titanotherium ramosum Osborn, 1896
Cf. Brontotherium ramosum, this monograph, page 577
Original reference. — Am. Mus. Nat. Hist. Bull.,
vol. 8, p. 1941, pi. 4, text fig. 13, 1896 (Osborn,
1896.110).
Type locality and geologic horizon. — "Upper
Titanotherium beds, South Dakota." Chadron for-
mation, Quinn Draw, Big Badlands, S. Dak.
Type. — A complete male skull lacking incisive
border. Am. Mus. 1447. (See fig. 194.)
Characters of type. — Osborn writes:
The distal spreading or branching of the horns is the
character by which this species is designated. It differs
from T. elatum in this character, but more especially in
the great depth of the "connecting crest" and the ex-
treme flattening of the horns, the section, as shown in
diagram 1, being intermediate between that of the T.
elatum and of T. plaiyceras. It is remarkable that the
teetli in this large skull are relatively of 3mall size; the
last upper molar has no second cone.
Etymology. — ramosum, branched, in allusion to
the "distal spreading or branching of the horns."
Present determination. — The species is probably
valid.
Megacerops brachycephalus Osborn, 1902
Cf. Brontops brachycephalus, this monograph, page 483
Original reference. — Am. Mus. Nat. Hist. Bull.,
vol. 16, pp. 97-98, fig. 3 (not the type), 1902
(Osborn, 1902.208).
Type locality and geologic horizon. — Big Bad-
lands, S. Dak.; Chadron formation, Chadron A, base
or level A of lower Titanotherium zone.
Type. — A complete skull (Nat. Mus. 4261, skull a),
collected by J. B. Hatcher in 1887. (See fig. 195.)
Specific characters. — Osborn writes:
The type of this species is No. 4261, U. S. Nat. Mus. It
includes very small, broad-skulled titanotheres with very rudi-
mentary second internal cones upon the upper premolars;
nasals elongate, narrowing anteriorly, as in Palaeosyops. Horns
of anteroposterior oval section placed above orbits. It is
represented in the National Museum by numerous skulls
besides the type, all collected and recorded by Hatcher. One
of these skulls was provisionally referred by him to Teleodus avus,
from which this species is quite distinct.
Etymology. — /3paxi's, short; Ki4>a\i), head, in allusion
to the brachycephalic form of the skull.
Present determination. — The species is probably
valid.
Megacerops bicornutus Osborn, 1902
Cf. Diploclonus bicornutus, this monograph, pages 234, 501
Original reference. — Am. Mus. Nat. Hist. Bull.,
vol. 16, p. 99, fig. 5, 1902 (Osborn, 1902.208).
Figure 191. — Type (holotype) skull of Titanops medius
After Marsh. Nat. Mus. 4256. One-eightli natural size.
Type locality and geologic horizon. — Quinn Draw,
White River, S. Dak.; exact level not recorded. Col-
lected by J. W. Gidley, 1896.
Type. — Skull and lower jaws (Am. Mus. 1476).
(See fig. 196.) Paratype, skull (Am. Mus. 1081).
One of these skulls (No. 1081) was first described by
Osborn (1896.110, p. 176) as Titanotherium torvum ox
rohustum.
Specific characters. — Osborn writes:
Horns directed anteriorly. Hornlets upon the inner and
anterior midportion of the horn. Basal section of the horn
slightly oval, subtransverse. Nasals narrow and relatively
elongate. Sharp malar bridge in front of orbit. Orbit large.
232
TITANOTHERES OF AKCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 192. — Type (holotype) nasofrontal shield
of Menodus peltoceras
Am. Mus. 10719. One-fourth natural size.
FiGORB 193. — Cotypes of Menodus? rumelicus
After Toula, 1892. Two-thirds natural size. The right lower molar
(two upper figures) is the leetotype.
Figure 194. — Type (holotype) skuU of Tiianoiherium ramosum
After Osborn, 1896. Am. Mus. 1447. Ai, Side view, one-twelfth natural size; A2, top view, one-
thirteenth natural size; A3, front view, one-thirteenth natural size.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
233
This animal stands nearest M. selwynianus, though dis-
tinguished by the greater size and slightly greater width
of the nasals. The sharp malar bridge is the most abso-
lute character. The two hornlets are possibly variations.
Type locality and geologic horizon. — -Big Badlands
(probably Cheyenne River badlands), S. Dak.;
Chadron formation (TitanotJierium zone), exact level
not determined.
FiGTjHE 195. — Type skull of Megacerops brachycephalus
Nat. Mus. 4261. One-fourth natural size.
Etymology. — lis, twice; cornutus, horned; in allusion
to the presence of small accessory horn swellings.
Present determination. — The species is probably
valid.
Megacerops marshi Osborn, 1902
Of. Allops marshi, this monograph, pages 511-515
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
16, pp. 100-101, fig. 6, 1902 (Osborn, 1902.208).
101959— 29— VOL 1-^18
Type. — A complete skull (Am. Mus. 501). (See
fig. 197.)
Paratype. — Skull (Am. Mus. 1445). Collected by
American Museum expedition, 1892.
Specific characters. — Osborn writes:
Type, skull No. 501; cotype, skull No. 1445, Am. Mus.
Nasals elongate and square distaUy, horns short, of oblique
oval basal section, overhanging the maxillae, or projecting
forward or outward. Incisors, ?f . Canines short, tetartocones
234
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
of premolars moderately developed. These skulls were pre-
viously confused by the writer with T. trigonoceras, from which
they are readily separated by the horn section, which relates
them to some of the primitive types of M. hrachycephalus and
equally to M. robustus. The canines are more obtuse than in
Figure 196. — Type (holotype) skull and lower jaw of Megacerops
bicornutus
Am. Mus. 14/6. After Osborn, 1902. One-eighth natural .size.
M. dispar, and the superior incisors resemble those in Bronto-
iherium rather than in M. robustus.
Etymology. — Named in honor of the late Prof. O. C.
Marsh, who estabUshed the remarkable collections of
titanotheres in the Yale and National Museums, pro-
posed the family name Brontotheridae, gave names to
many of the genera and species, and projected the
present monograph.
Present determination. — The species is probably
valid.
Brontotherium leidyi Osborn, 1902
Cf. Bronlhotherium leidyi, this monograph, page 558
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
16, pp. 105-106, figs. 9, 10, 1902 (Osborn, 1902.208).
Type locality and geologic horizon. — Big Badlands,
S. Dak.; Chadron formation, lower levels of lower
Titanotherium zone (Chadron A).
Type.— A complete skull (Nat. Mus. 4249, skull R)
collected by J. B. Hatcher in 1887. (See figs. 198,
199.)
Specific characters. — Osborn writes:
Nasals elongate, narrowing anteriorly. Horns very short,
slightly recurved, of transverse oval section. Canines stout
and blunt. Premolars noncingulate, with rounded contours
and weU-developed tetartocones. Incisors?^.
Etymology. — Named in honor of Joseph Leidy, the
first of the three great founders of American vertebrate
paleontology, describer of Titanotherium, Megacerops,
Palaeosyops, author of "The ancient fauna of Ne-
braska" and of "The extinct mammalian fauna of
Dakota and Nebraska."
Present determination. — The species is probably
valid.
SPECIES DESCRIBED BY LULL IN 1905
Megacerops tyleri Lull, 1905
Cf. Diploclonus tyleri, this monograph, page 502.
Original reference. — Jour. Geology, vol. 13, No. 5,
pp. 443-456, pis. 3, 4, August, 1905 (Lull, 1905.1).
Type locality and geologic horizon. — North side
of Spring Draw Basin, about 10 miles from the
mouth of Bear Creek, a tributary of Cheyenne
River, S. Dak. Type specimen found 35 feet
above the base of 200 feet of the Chadron
formation {Titanotherium zone) lying upon Pierre
deposits, "hence in the upper part of the lower
division," as defined by Hatcher in 1893 (1893.1,
p. 218).
Type. — Skull, limbs, and many vertebrae of a
single individual (Amherst Mus. 327). (See
figs. 200 and 201.) Found by T. C. Brown, of
the Amherst College paleontologic expedition of
1903.
Specific characters. — Lull writes:
Horns well in front of orbits, directed somewhat
forward and outward, an elongate oval in basal section
with the long axes in line, rounded oval at the summit.
Hornlets quite conspicuous, on the inner face of the
horns midway between the base and summit. Con-
necting crest low and inconspicuous. Nasals broad,
well rounded in front, and but sHghtly arched beneath.
Zygomata expanded and deep, with a well-rounded outer
face. Dentition: Superior incisors represented by the deep
and well-defined median alveoli and by the lateral teeth,
which remain in place and which have hemispherical crowns
which show little sign of wear. The canines are lanceolate,
with a well-developed postero-internal cingulum. There is a
short diastema in front of, and a longer one behind, the canine.
Premolars with a smooth internal cingulum, less pronounced
in the middle of the tooth, and with no external cingulum. The
deuterocone is well developed, while the tetartocone, especially
of premolar 4, is inconspicuous.
The jaw is deep and robust, with the alveoli of two incisors,
probably of the second and third, deep and distinct. There is
Figure 197. — Type skull of Megacerops marshi
After Osborn, 1902. Am. Mus. 501. One-tenth natural size. The lower
jaw (Am. Mus. 516) figured with this skull does not belong with it.
It is probably referable to Brontotherium leidyi.
'no space between the lateral incisors and the canine, though
between the two median alveoli a considerable gap occurs.
There seems to have been a small diastema behind the lower
canines, which are lanceolate, though with a less prominent
cingulum, and not so strongly recurved as the upper ones.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
235
Etymology. — Named in honor of Prof. John M.
Tyler, of Amherst College, "a teacher of men, who,
by his earnest efforts, as well as by his own generosity,
was mainly instrumental in maldng possible the ex-
pedition which secured the specimen" (Lull).
Present determination. — This species is probably
valid. It is discussed on page 502 of this monograph.
Figure 198. — Type (holotype) skull of Brontotherium leidyi
After Osborn, 1902. Nat. Mus. 4249. One-eighth natural size. The side view of this
skull was figured by Marsh (Am. Jour. Soi., October, 1887) as Menops varians.
SPECIES DESCRIBED BY OSBORN IN 1908
Brontotherium hatched Osborn, 1908
Cf. Brontotherium hatcheri, this monograph, page 563
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, pp. 615-616, fig. 20, 1908 (Osborn, 1908.318).
Type locality and geologic horizon. — South Dakota;
Chadron formation, middle Titanotherium zone (Chad-
ron B), lower levels. J. B. Hatcher, collector.
Type. — A nearly complete skull (Nat. Mus. 1216,
skull a) lacking the premaxillaries and anterior por-
tion of the maxillaries. (See fig. 202.)
Specific characters. — Osborn writes:
If, Pf. Nasals moderately long (97 mm.), thin at the
edges. Horns 250 -f millimeters, two-thirds the length oj
B. gigas horns. Skull length (pm.x-condyles) , 710 (estimated),
width across zygomata, 530 (estimated) . This species appears
to represent an early phase of evolution of B. gigas. The horns
are very round or convex in section and have a well-defined
malar ridge on the lower outer portion. The connecting crest
is relatively shallow, and the nasals are thin. The premolars
are well advanced, the tetartocone of p' being well rounded and
quite distinct.
Etymology. — Named "in honor of the late J. B.
Hatcher, who discovered many of Professor Marsh's
titanothere types, brought together the great collec-
tion of titanotheres ia the National and Yale Mu-
seums, and placed the stratigraphic succession of the
species upon a secure basis." (Osborn.)
Present determination. — The species is probably
valid.
Symborodon copei Osborn, 1908
Cf. Megacerops copei Osborn, this monograph, page 548
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
24, pp. 616, 617, fig. 21, 1908 (Osborn, 1908.318).
Type locality and geologic horizon. — South Dakota,
Big Badlands, Indian Draw; Chadron formation,
level probably middle Titanotherium zone (Chadron
B). J. B. Hatcher, collector.
Type. — A complete skull (Nat. Mus. 4711, skull
V), collected by J. B. Hatcher, 1888. (See fig. 203.)
Specific and generic characters. — Osborn writes:
Incisors (type) persistent but greatly reduced; canines
very small, reduced (28 mm.) ; premolars with cingula reduced
or absent; tetartocones connected with deuterocones by a
longitudinal ridge. Skull: nasals thin, short and broad in pro-
portion, 80 by 125 millimeters; horns, S , 300, no connecting
crest, transverse oval near summit; buccal processes of zygomata
t? stout and conve.x; malar in front of buccal process very deep,
beneath postorbital process stout, convex; occipital pillars
not greatly expanded at the summits.
Etymology. — Named in honor of the late Prof.
E. D. Cope, prolific author of "The Vertebrata of the
Tertiary formations of the West," original describer of
Symborodon, founder of the "Cope collection," now
in the American Museum of Natural History.
Present determination. — The species is probably
valid.
CANADIAN SPECIES DESCRIBED BY LAMBE IN 1908
Megacerops primitivus Lambe, 1908
Cf. Teleodus primitivus, this monograph, page 482
Original reference.— Contr. Canadian Paleontology,
vol. 3, pt. 4, pp. 49-51, pi. 6, figs. 4, 5, 1908 (Lambe,
1908.1).
FiauRE 199.^Upper premolars of type
skull of Brontotherium leidyi
After Osborn, 1902. Nat. Mus. 4249. One-halt natural size.
Type locality and geologic horizon. — "Oligocene
deposits of the Cypress Hills," Saskatchewan. Col-
lector, L. M. Lambe, 1904.
Type. — Both halves of the lower jaw, with the denti-
tion of the left side complete. Ottawa Museum. (See
fig. 204.)
Specific characters. — Lambe writes:
Incisors, in three pairs, with a space between the inner pair;
canines, of small diameter, apparently short; a diastema
236
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 200. — Type (holotype) skull of Megacerops iyleri
After Lull. Amherst Mus. 327. A, Dorsal aspect of skull; B, lateral aspect of skull and jaw; C,
anterior aspect of horns and nasals; all about one-eighth natural size. D, Upper dentition (incisor,
canine, and premolar series), one-fourth natural size.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
237
Figure 201. — Right manus and right hind limb of the type of Megacerops tyleri
Alter Lull, 1905. Amherst Mus. 327. A, Pro.dmal row of carpals, proximal aspect; B, distal row of carpals, proximal
aspect (sc. /., scaphoid facet, lu.f., lunar facet, en. /., cuneiform facet); C, right manus; all one-fourth natural
size. D, Right hind limb, one-eighth natural size.
238
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
between the canine and the first premolar; first premolar
small; third premolar becoming molariform; fourth premolar
molariform; symphysis long; symphyseal surface between
canines narrow; jaw contracted at the diastema; external
cingula moderately developed; internal cingula wanting;
mental foramen beneath the second premolar; coronoid process
short.
Megacerops avus (Marsh), from the Oligocene of South
Dakota, has three pairs of inferior incisors but only three pre-
molars below on each side, and there is a short diastema behind
P
^
^B
r
T'tif' ^^
^^H
L_
_ r^^
^^
Jr- , "^^B
i^Tr^^H
n
W'^^^P"
^H
F 1
B^
H^^'
PI
f 1
^m
kw ~
1
i^m
1 :
-rf»r-"^
l^^l
iH
mn
H
Figure 202. — Type (holotype) skull of Bronlo-
therium hatcheri
Top view. Nat. Mus. 1216. After Osborn. 1908. One-tenth
natural size.
the lower canine. Its dimensions are greater than those of
M. primitivus. These two species are apparently the only ones
of the Oligocene titanotheres in which there are three pairs
of incisors in the lower jaw.
In the Cypress Hills specimen the crowns of the incisors are
of a depressed spherical shape, with a tendency to come to a
rounded central point above. The second incisor is the largest,
and the first is slightly smaller than the third, which is the
most upright. The first is more procumbent than the second.
Between the inner pair is a very decided interval, leaving a
space of 6.5 millimeters between the crowns of the two teeth.
The crowns of the canines are broken ofi' (that of the right tooth
being restored in fig. 5 of pi. 6) and the right first premolar is
lost from its alveolus. * * *
Keeping in mind the differences due to sex in titanotheres
generally and the apparent variability, both specific and indi-
vidual, of certain dental characters, such as the degree of devel-
opment of the cingula, the presence or absence of the first
premolar, the size of the canines, and the number of the incisors,
M. primitivus is apparently a well-marked species, characterized
principally, so far as known at present, by the breadth of the
mandible anteriorly (as compared with M. angustigenis) and
the presence of the fuU number of teeth, with a comparatively
long diastema behind the canines.
This species, for which the name primitivus is used, is regarded
as representing a rather early stage in the development of the
titanotheres. The general character of the dentition suggests
the appropriateness of referring the species to the genus
Megacerops.
Measurements
Millimeters
Length of ramus 475
Depth of same at posterior end of fourth premolar 74
Depth of same at posterior end of second molar 81
Depth of same from tip of coronoid process to lower
border 247
Maximum thickness of same beneath third molar 46
Length of symphysis 144
Distance apart of inside surface of base of canines" 31
Length of premolar series 103
Length of molar series 183
Diameter of canines at base:
Anteroposterior 18
Transverse 16
Diameter of second premolar:
Anteroposterior 26
Transverse 18
Diameter of third premolar:
Anteroposterior 32
Transverse 23
Diameter of fourth is premolar:
Anteroposterior 35
Transverse 27
Etymology. — primitivus, primitive; in reference to
the presence of three lower incisors.
Present determination. — The species is probably
valid. It is probably referable to Teleodus.
Megacerops assiniboiensis nom. prov., Lambe, 1908
Cf. Brontotherium curium, this monograph, page 574
Original rejerence. — Contr. Canadian Paleontology,
vol. 3, pt. 4, pp. 51-53, pi. 5, fig. 6, 1908 CLambe,
1908.1).
Figure 203. — Type (holotype) skull of Symborodon copei
After Osborn, 1908. Nat. Mus. 4711. One-tenth-naturai size.
Type locality and geologic horizon. — Oligocene de-
posits of the Cypress Hills, Saskatchewan. Collection
of 1904.
" In the mandible of M. angustigenis (No. II) figured by Cope, op. cit. [1891.2]
this measurement is about 18 millimeters, and in the symphysis of the jaw (No. I,
also figured) a like measurement given, by the same authority, as 27 millimeters ,
should be 22 millimetei s.
>' First premolar in Cope's description of M. angustigenis.
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
239
Type. — "A robust, short left mandibular ramus,"
lacking the posterior end. The three molars and the
the fourth premolar and the first molar. The bone is massive
and heavy throughout. The mental foramen is placed beneath
the posterior root of the third premolar, farther back than in
M. ■primiiivus.
The cingula are very slightly developed. The external cingu-
lum is present for a short distance only, on the anterior face
of each of the four teeth, and in the third molar in advance of
the heel. The only trace of an internal cingulum is to be seen
in the third molar on the posterior slope of the heel.
Figure 204. — Type (holotype) jaw of Megacerops primiiivus
In the collection or the Ottawa Museum. After Lambe, 1908. A, Superior aspect, one-half natural size; B, lateral aspect, one-third natural size.
fourth premolar are preserved, as well as part of the
symphyseal region. Ottawa Museum. (See fig. 205.)
The fourth premolar is fully molariform. The teeth are
stout and of about the size of the corresponding ones in M.
Figure 205. — Type (holotype) jaw of Megacerops assiniboiensis
In the collection of the Ottawa Museum. After Lambe, 1908. One-third natural size.
Characters. — Lambe writes:
The jaw is much deeper, thicker, and relatively shorter than
in angustigenis and primiiivus, and the teeth are much larger
than in these species. It is narrow anteriorly, and the sym-
physis extends back to a point in line with the division between
marshi Osborn, but the jaw is relatively shorter than in this
species.
From the material available, the species, for which the
provisional name assiniboiensis is proposed, can not be defi-
nitely characterized.
240
TITAJSrOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Measurements of ramus {type)
Millimeters
Depth of ramus at posterior end of fourth premolar 80
Depth of ramus at posterior end of third molar 156
Thiclcness of ramus above lower border beneath posterior
end of first molar 55
Vertical thickness of symphysis a little in advance of its
posterior termination 53
Vertical thickness of symphysis in line with front root of
third premolar 31
Space occupied by fourth premolar and the molars 260
Diameter of fourth premolar:
Anteroposterior 41
Transverse 31
Diameter of first molar:
Anteroposterior 55
Transverse 36
SECOND EUROPEAN OIIGOCENE SPECIES, DESCRIBED BY
KIERNIK, 1913
Titanotherium bohemicum Kiernik, 1913
Cf. Menodus giganteus, this monograph, page 530
Original reference. — Acad. sci. Cracovie Bull., ser. B,
vol. lOB, pp. 1211-1225, pi. 63, 1913 (Kiernik, 1913.1).
Type locality. — Uncertain. The specimen, a frag-
ment of the lower jaw containing the third right lower
molar, was received with a lot of fossils from the dilu-
vium near Prague. It was supposed to have come
from the lime pits of Podbaba, near Prague, and to
have been sold by one of the workers in the lime pits
C D
Figure 206. — Type of Titanotherium bohemicutn Kiernik
Fragment of a lower jaw with third right lower molar. After Kiernik. A, Outer side view; B, inner side view; C, top view, showing
the grinding surface of ms; D, front view, showing the exposed posterior roots of mj. Ahout one-fourth natural size.
Diameter of second molar:
Anteroposterior 71
Transverse 41
Diameter of third molar:
Anteroposterior 99
Transverse 43
Space occupied by roots of third premolar (anteroposterior) 34
Space between fourth premolars (twice the distance of
fourth premolar from vertical plane through symphysis) - 60
Etymology. — assiniboiensis , in allusion to the geo-
graphic occurrence of the type.
Present determination. — This species apparently be-
longs in the Brontotheriinae. It is smaller than
Brontotherium Tiatcheri. The nasals doubtfuUy referred
by Lambe to this species suggest those of Bronto-
therium curtum.
to Herr Baumeister Kuchta (died 1910). He gave it,
along with other prehistoric specimens, to Herr
EoJanek, who in turn gave it to Herr Jira, who pre-
sented it to the Institute for Comparative Anatomy
at Prague. After carefully considering the possi-
bility that the specimen might have been of American
provenience the author, Herr Kiernik, inclines rather
to the view that it really came from Bohemia, al-
though not from Pfodbaba, but from the fresh-water
Tertiary deposits of Tuchofitz (northwestern Bo-
hemia). The well-known fauna of Tuchofitz is,
however, of lower Miocene facies.
Type. — A lower jaw fragment containing the third
right lower molar. (See fig. 206.)
DISCOVERY OF THE TITANOTHERES AND ORIGINAL DESCRIPTIONS
241
Characters. — Kiernik carefully compares the frag-
ment with the types of Brachydiastemafherium tran-
sihanicum Bockh and Maty, Menodus rumelicus Toula,
and Titanotherium proutii Leidy. He shows that the
third lower molar is much larger than that of either
Brachydiastematherium or Menodus rumelicus, but
that it is nearer in its measurements to the type of
Titanotherium proutii, as indicated in the following
table :
Measurements of Titanotherium hohemicum, T. -proutii, and
Menodus rumelicus, in millimeters
T. bohemicum M. rumelicus
Total length of the wear-
ing surface
Breadth of the first section
(lobe) of the tooth
Breadth of the second sec-
tion (lobe) of the tooth...
27
T. proutii
The author concludes that this species is widely
distinct from the Ivnown European forms but that
possibly it may eventually prove to be identical with
either Titanotherium proutii or another species of the
same genus. This, however, he considers unlikely, in
view of its [supposed] European origin, so that he
thinks he is quite justified in retaining the name
Titanotherium hohemicum.
Etymology. — hohemicum, in allusion to the country
where the specimen was supposedly found.
Present determination. — According to Dr. W. K.
Gregory, who has compared a cast of the type of
Titanotherium hohemicum with various American ti-
tanotheres, the type specimen is closely similar to one
in the American Museum of Natural History referred
to Menodus giganteus (Am. Mus. 1007). It differs
chiefly in the greater width of the anterior lobe of
m3. It appears indeed to be specifically referable to
Menodus giganteus, and it seems possible that it is
in reality an American specimen which became mixed
with the collection of fossils from Podbaba, near
Prague. (Cf. pp. 230, 560, 941.)
Measurements of Menodus bohemicus and M. giganteus, in
millimeters
M3, total length (estimated)
M3, breadth of first lobe at base
M3, breadth of second lobe
M3, breadth of third lobe
Center of protooonid to center of
hypoconid
Center of metaconid to center of
entoconid
Depth of jaw below front edge of ms-.
Depth of jaw just behind ms
108
52
47
33
39
39 +
111
152
M. giganteus
(trigonoceras).
Am. Mus,
1007
109
47
47
111
143
FINAI OLIGOCENE SPECIES DESCRIBED BY OSBOEN IN
1916-1919
Allops walcotti Osborn, 1916
See page 509
Original reference. — Am. Mus. Nat. Hist. Bull.,
vol. 35, pp. 721, 722, fig. 1, 1916 (Osborn, 1916.433).
Type locality and geologic horizon. — "Big Badlands,"
S. Dak., probably Corral Draw; Chadron formation
{Titanotherium zone), lower levels (Chadron A).
Type. — A nearly complete skull in the National
Museum (No. 4260, skull Q). (See fig. 207.)
Figure 207. — Type (holotype) skull of Allops walcotti
Nat. Mus. 4260. After Osborn, 1916. One-eighth natural size.
Specific characters. — Osborn writes:
Premolars with small tetartocones; p'-m^ 285 millimeters.
Incisors f . Horns elongate oval, no connecting crest. Mesa-
ticephaUo. Nasals elongate, broad. Face relatively elongate.
The type skull of this species (U. S. Nat. Mus. 4260) from
level A is narrow and elongate, partly owing to lateral crushing.
This feature conceals its resemblance to Allops marshi, which
is apparent in other features — namely, (1) primitive, long
nasals, (2) horns primitively short and obhquely oval, (3) large
lateral incisor (ij) and small first (ij) or median incisor, (4)
premolars accelerated, tetartocones more advanced than in
Brontops robustus of level C.
242
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Observations on the measurements oj AUops wal-
cotti. — The type and only known specimen of this
species exhibits the following comparison in measure-
ments with skulls of B. hrachycephalus and Menodus
Jieloceras, which show that the type of AUops walcotti
has relatively large premolars and small molars.
Measurements of AUops walcotti, Menodus heloceras, and Brontops
brachycephalus, in millimeters
A.walcotti,
Nat. Mus.
4260 (type)
M. helo-
ceras. Am.
Mus.
14576
B. brachycephalus
Nat. Mus.
4940
Nat. Mus.
42S1
Pi-m3
Pi-p*
Mi-m3.
285
112
169
640
105
100
35X51
60X61
265
170
603
132
70
265
101
160
280
" 104
178
Pmx-condyles
680
102
32X51
62X70
85
33X53
68X73
Etymology. — -"The species is named in honor of
the Secretary of the Smithsonian Institution, Charles
D. Walcott." (Osborn.)
Present determination. — The skull is 'crushed later-
ally but probably had a low zygomatic index — that is,
it was mesaticephalic. While its reference to AUops
requires confirmation, its nearer affinities appear to be
with this genus rather than with Brontops or Menodus.
The external cingula of the premolars are not as
sharply defined as in other primitive members of the
menodontine group.
Megacerops riggsi Osborn, 1916
See page 550
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
35, p. 723, fig. 2, 1916 (Osborn, 1916.433).
Type locality and geologic horizon. — Northeastern
Colorado, Horsetail Creek; Chadron formation (Titano-
therium zone), upper (?) levels.
Type. — A nearly complete lower jaw in the American
Museum (No. 6364). E. D. Cope, collector. (See
fig. 208.)
Specific characters (Osborn). — Of small size, smaller
than any known individual of Megacerops or Bronto-
therium. Very massive jaw with a small coronoid
FiGUEE 208. — Type (holotype) jaw of Megacerops riggsi
Am. Mus. 6364. After Osborn, 1916. One-sixth natural size.
process and a very short symphysis. Premolar series
greatly abbreviated (85 mm.). Premolars and molars
with reduced external cingula.
Measurements of type
Millimeters
Symphysis to condyle (estimated) 465
Premolar-molar series (pi-ms) 282
Premolar series (pi-pi) 85
Molar series (mi-ma) 194
Etymology. — Named "in honor of Mr. E. S. Riggs,
of the Field Museum of Natural History, in recogni-
tion of his discoveries of Eocene titanotheres." (Os-
born.)
Present determination. — The type of this species
is a jaw in the Cope collection (Am. Mus. 4636),
which was wrongly referred by Cope to his species
" Symhorodon" acer. It represents a highly specialized
and small form of Megacerops.
Note. — For descriptions of upper Eocene and lower Oligo-
cene titanotheres from MongoUa described by Osborn in 1923
see appendix; also the final opinion regarding the titanotheres
of eastern Europe, page 941.
CHAPTER IV
SYSTEMATIC CLASSIFICATION OF THE TITANOTHERES
SECTION 1.
PHYLETIC VERSUS LINNAEAN SYSTEM
OF CLASSIFICATION
NEO-LINNAEAN SYSTEMATIC DIVISIONS (ZOOLOGIC) AND
EVOIUTIONARY PHYIA (PALEONTOIOGIC)
As explained in the introduction, the Linnaean
system was based on the theory of the special creation
of all systematic divisions coinciding in geographic
space, so that its application to our modern paleonto-
logic phyla, which succeed one another over long
periods of geologic time, is beset with great difficulties
and has led to different uses of systematic terms by
different authors. The present monograph employs
a phyletic system which has been used by the author
since 1892 in the classification of the Perissodactyla
(Osborn, 1892.67, pp. 90-94).
The taxonomic principle is that ancestral affinity is
stronger than contemporary resemblance. Thus an
animal that is directly ancestral to the titanotheres is
placed in the family Brontotheriidae; an animal that
is directly ancestral to BrontotJierium is placed in the
subfamily Brontotheriinae; a series of ascending
species in the same line are placed in the genus
BrontotJierium; a series of "ascending mutations" may
be placed within the single species BrontotJierium
gigas.
Such a vertical or phyletic application of the Lin-
naean system involves, it is true, a departure from the
traditional Linnaean methods, but in the author's
opinion it is far preferable to the introduction of a
new systematic terminology. If necessary the author's
system may be distinguished as neo-Linnaean. It is
an adaptation of the Linnaean system to phylogeny
as revealed by paleontology.
The degrees or steps in the evolution of neomorphic
and heteromorphic characters, or rectigradations and
allometrons, afford the real basis of our division of the
great family tree of the titanotheres into branches
and subbranches as follows:
Family, a branch of the Perissodactyla having a large num-
ber of similar characters and similar tendencies of evolution.
Subfamily, a branch of the main family embracing one or
more genera retaining certain similar characters and developing
certain peculiar evolutionary tendencies.
Genus, a branch of a subfamily or a stage of a subfamily
distinguished by the prominent position of certain distinctive
characters, which may be in widely different stages of develop-
ment— for example, Brontotherium leidyi, B. platyceras.
Species and subspecies, divisions distinguished by certain
gradations in the development of characters common to the
genus, also by certain rectigradations and allometrons.
Ascending mutations, divisions distinguished by various
intermediate stages of development of rectigradations and
allometrons.
These principles of phyletic classification as devel-
oped and adopted in this monograph are also fully
explained in Chapter I.
Classification is simply a convenient and condensed
expression of our knowledge of hereditary lines of
descent. It is constantly shifting and changing with
discovery. The final classification can be attained
only after we have worked out all the lines of descent
of this great family. In the meantime we may review
the history of the successive attempts at classification
made up to the present time.
SUPEEFAMILY NAMES PROPOSED BY OSBORN (1898) AND
HAY (1902)
Superfamily Titanotherioidea Osborn, 1898
Original reference. — Am. Mus. Nat. Hist. Mem., vol.
1, pt. 3, p. 79, 1898 (Osborn, 1898.143).
Osborn divided the Perissodactyla into five super-
families :
I. Titanotherioidea ("including the single family Titano-
theriidae"), understood by Osborn to include both Eocene and
Oligocene titanotheres.
II. Hippoidea, including Equidae and Palaeotheriidae.
III. Tapiroidea, including Tapiridae and Lophiodontidae.
IV. Rhinocerotoidea, Including Hyracodontidae, Amyno-
dontidae, Rhinocerotidae.
V. Chalicotherioidea, Chalicotheriidae.
Present determination. — Superfamily names are
formed by adding oidea to the stem of the family
name, and as Brontotheriidae is now regarded as valid,
it appeared necessary to Doctor Hay to substitute for
Osborn's term Titanotherioidea the term Bronto-
therioidea, first used by Hay in 1902.
Superfamily Brontotherioidea Hay, 1902
Original reference. — U. S. Geol. Survey Bull. 179,
p. 629, 1902 (Hay, 1902.1).
The content of this term is as follows:
Brontotherioidea :
Brontotheriidae:
Lambdotheriinae (Eocene titanotheres) .
Brontotheriinae (Oligocene titanotheres).
The content of the term Brontotherioidea Hay, 1902,
is thus the same as that of Titanotherioidea Osborn,
1898.
FAMILY NAMES PROPOSED OR ADOPTED BY MARSH (1873),
FLOWER (1875), COPE (1879-1889), AND OSBORN (1889)
Family Brontotheridae Marsh, 1873
Original reference. — Am. Jour. Sci., 3d ser., vol. 5,
p. 486, 1873 (Marsh, 1873.1).
Included genera. — Titanotherium Leidy and Bronto-
therium Marsh.
243
244
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Family characters. — Not distinguished, but state-
ment is made that Brontotherium was a " true perisso-
dactyl with hmb bones resembling those of RM-
noceros." Marsh gave the famUy characters fully in
a paper entitled "On the structure and affinities of the
Brontotheridae." He writes (Marsh, 1874.1, p. 82):
Among the more marked characters of the Brontotheridae,
which readily distinguished them from the Rhinocerotidae,
apparently their near allies, may be mentioned the following:
There are four short and thick toes in the manus, and three in i
the pes. The skull supports a pair of large horn cores, placed
transversely, as in modern artiodactyls.^' There are well- i
developed canine teeth in both jaws. The molar teeth, above
and below, are not of the Rhinoceros type but resemble those of
Chalicotherium.
Present determination. — As long as Brontotherium
was regarded as a synonym of Titanotherium the term
Brontotheriidae had no standing, but since Bronto-
therium has been shown to be a good genus the term
Brontotheriidae must be held valid.
Family Limnohyidae Marsh, 1875
Cf. Palaeosyopinae, this monograph, page 298
Origin^ reference. — Am. Jour. Sci., 3d ser., vol. 9,
p. 246, 1875 (Marsh, 1875.1).
Present determination. — In defining the genus Dipla-
codon, Marsh says: "From the Eocene Limnohyidae,
already described, this genus is sharply distiuguished."
The name Limnohyidae does not occur in Marsh's
previous descriptions, and so far as one can judge the
famUy had not been defined. As Limnohyus is a
synonym of Palaeosyops the family name is invalid.
Family Titanotherlidae Flower, 1876
Cf. Brontotheriidae Marsh, this monograph, page 279
Original reference. — Nature, vol. 13, p. 328, 1876
(Flower, 1876.1).
Present determination. — Flower regarded Bronto-
therium as synonymous with Titanotherium and so
naturally called the family Titanotherlidae; but siace
Brontotherium is now regarded as valid, Flower's term
becomes a synonym of Brontotheriidae Marsh.
Family Chalicotherlidae Cope, 1879
Original reference.- — U. S. Geol. and Geog. Survey
Terr. Bull., vol. 5, p. 228, 1879 (Cope, 1879.1).
Included genera. — "Limnohyus Leidy [ = Limnohyops
Marsh], Palaeosyops Leidy, ' Leurocephalus S., O. &
S.' [= Telmatherium cultridens], Menodus Pomel, Sym-
lorodon Cope, Daeodon Cope, Chalicotherium Kaup,
Nestor itherium Kaup."
Present determination. — The titanotheres should
never have been included in the same family with
Chalicotherium.
19 Ehinoceros pleuToceTOS Duv., from the Miocene of France, has a transverse pair
of small horn cores on the nasals, not unlike those in DiTtoceras. R. mxnutua Cuv.
has somewhat similar processes.
Menodontidae Cope, 1881
Cf. Brontotheriidae Marsh
Original reference. — Am. Philos. Soc. Proc, vol. 19,
pp. 378, 379, 397, 1881 (Cope, 1881.1).
Present determination. — The name Menodontidae as
applied to the Ohgocene titanotheres is invalid because
antedated by Brontotheriidae Marsh.
Family Lambdotheriidae Cope, 1889
Cf. Lambdotheriinae, this monograph, page 279
Original reference. — Am. Naturalist, March, 188 9
p. 153 (Cope, 1889.1).
Included genera. — From Cope's description it is
plain that he intended to refer to the Lambdotheriidae
not only the type genus Lambdotherium but all
titanotheres with "but a single internal cusp on the
first (posterior) superior premolar." He thus con-
trasts the Lambdotheriidae with the Menodontidae
( = Brontotheriidae). Cope then also referred to the
family Lambdotheriidae an Oligocene genus "Hapla-
codon" (= Megacerops angustigenis) .
Synonymy. — The term Lambdotheriidae as used by
Nicholson and Lydekker (1889.1, vol. 2, p. 1371) had
the same connotation. It was apparently first limited
to the genera Lambdotherium, Palaeosyops, and "Lim-
nosyops" { = Limnohyops) by Flower and Lydekker
(1891.1, p. 413) in 1891. Later authors, as Earle in
1892 (1892.1) and Zittel in 1893 (1893.1, p. 300), used
the term Palaeosyopidae or Palaeosyopinae to include
the same genera.
Present determination. — In this monograph the group
under consideration is treated as a subfamily Lambdo-
theriinae of the Brontotheriidae.
Family Titanotherlidae Osborn, 1889 (1890?)
Cf. Brontotheriidae Marsh, this monograph, page 279
Original reference. — Am. Philos. Soc. Trans., new
ser., vol. 16, p. 514, 1889 (1890) (Scott and Osborn,
1890.1).
Included genera. — Osborn writes:
Palaeosyops has hitherto been referred to the ChaUcothe-
riidae, but the discovery of the footbones of Chalicotherium
by Filhol shows that the genera are widely separated. The
discovery of the skeleton of Diplacodon, however, links Palaeo-
syops very closely to Titanotherium. * * * It seems best
to group the three genera [Palaeosyops, Diplacodon, and
Titanotherium] in the single family Titanotherlidae.
Present determination. — This was the first descrip-
tion which included the true titanotheres of the
Eocene and Oligocene without extraneous elements
(Chalicotherium). The term is nevertheless pre-
occupied by Titanotherlidae Flower, 1876, which is
in turn a synonym of Brontotheriidae Marsh, 1873.
SYSTEMATIC CLASSIFICATION OF THE TITANOTHERES
245
SUBFAMILY NAMES AND PHYLA PROPOSED BY STEIN-
MANN AND DODERIEIN (1890), EARLE (1892), AND RIGGS
(1912)
Subfamily Falaeosyopinae Steinmann and Doderlein, 1890
Original reference. — Elemente der Palaontologie,
p. 777, 1890 (Steinmann and Doderlein, 1890.1).
Included genera. — The authors divide the Chali-
cotheriidae into three subfamihes — Falaeosyopinae,
Brontotheriinae, Chalicotheriinae. The Falaeosyo-
pinae include the genera Palaeosyops, "Limnohyus"
{Limnohyops) , Diplacodon.
Present determination. — Falaeosyopinae Steinmann
and Doderlein, 1890, thus has priority over Falaeo-
syopinae Earle, 1892.
Subfamily Falaeosyopinae Earle, 1892
Cf. Falaeosyopinae Steinmann and Doderlein
Original type reference. — Acad. Nat. Sci. Fhila-
delphia Jour., 2d ser., vol. 9, pp. 272 et seq., 1892
(Earle, 1892.1).
Included genera. — Lamidotherium, Limnohyops,
Palaeosyops, Telmatherium, Haplacodon [Megacerops
angustigenis].
Present determination. — Earle gives a detailed and
accurate description of the subfamily characters
(pp. 274-276). The term is preoccupied by Falaeo-
syopinae Steinmann and Doderlein, 1890, and in its
content is preoccupied by Lambdotheriidae Cope,
1889.
Subfamily Dolichorhinae Riggs, 1912
Cf. Dolichorhininae
Original reference. — Field Mus. Nat. Hist. Fub. 159,
Geol. ser., vol. 4, No. 2, p. 25, June, 1912 (Riggs,
1912.1).
Included genera. — Middle Eocene titanotheres
having nasals elongate and deeply recessed laterally,
face shorter than cranium, an infraorbital process
more or less developed, and molars only moderately
expanded.
This group is proposed in order to designate those
long-nosed titanotheres which evidently sprang from
a common stock and form a natural and homogeneous
group. It includes the genera MesatirJiinus, Meta-
rJiinus, Dolichorhinus, and RhadinorJiinus.
DIVISION OF THE OLIGOCENE TITANOTHERES INTO FOUR
CONTEMPORARY PHYLA, OSBORN (1902)
Original reference. — Am. Mus. Nat. Hist. Bull., vol.
16, pp. 91-109, February 18, 1902 (Osborn, 1902.208).
Included genera. — Osborn writes:
The Oligocene titanotheres consisted of at least four contem-
porary phyla, to which the prior generic names Titanotherium,
Megacerops, Symborodon, and Brontotherium may be applied.
They represent an adaptive radiation for different local hab-
itat, different modes of feeding, fighting, locomotion, etc., which
took origin, in part at least, in the middle or upper Eocene.
Europe and Asia also may have shared in this radiation, since
titanotheres are now definitely known in the Balkan region.
The main phyletic characters are analogous to those recently
(Osborn, 1900, p. 231) determined among rhinoceroses; the
great antiquity of the lines leading to the existing species of
rhinoceroses necessitated the revival of a number of discarded
generic names to distinguish them. Similarly the separateness
of four of the titanothere phyla, throughout the Ohgocene and
possibly from the Eocene, renders it desirable to revive certain
generic names which in my first review I considered undefinable.
Radiation involved three main sets of characters, two of
which were correlated:
First, doliohocephaly and brachycephaly, associated with
numerous changes in the skull and teeth and, in at least two
phyla, with longer and shorter limbs.
Second, four distinct types in the shape and position of the
horns, correlated with the structures of the nasals and frontals
and indicative of different modes of combat among the males.
(See fig. 209.)
Third, canines of different form; and, finally, the presence
of one or two pairs of functional incisor teeth, or the total
degeneration of these teeth.
Titanotherium Leidy applies to long-limbed animals with
long skulls, persistently long and broad nasals, short triangu-
MegaceropSy Upper Beds.
Diplodonus, Upper Beds.
Syiiiborodoti. Upper Beds,
.Brontotkcriuv!, Upper Beds,
Figure 209. — Characteristic basal sections of horns of
Oligocene titanotheres
l.ar horns placed slightly in front of the eyes, vestigial incisors,
^:?, large canine teeth. Known from the base to the summit of
the Oligocene.
Megacerops Leidy applies to titanotheres with broad skulls,
nasals progressively shortening, short horns rounded or oval in
section, shifting anteriorly, one or two pairs of incisor teeth,
^, medium-sized canine teeth. Known from the base to the
summit of the Oligocene.
Probably related to this are the subgenera of the types named
Allops and Diploclonus by Marsh, differing from the above in
horn characters. Known chiefly from the upper beds.
Symborodon Cope includes titanotheres with skulls of vary-
ing proportion, nasals slender and progressively shortening,
horns elongate and peculiar in being placed above the eye
instead of shifting forward, incisors vestigial, |^, canines small,
appro.ximated. Known only from the middle and upper beds.
Brontotherium Marsh embraces the largest titanotheres, with
very broad zygomatic arches, nasals shortening while horns
elongate and shift forward; incisors persistent, f in the males,
canines stout and obtuse.
Representatives of Titanotherium and Megacerops can now
be continuously traced from the base to the summit of the
Oligocene. Primitive species of Brontotherium also appear at
246
TITANOTHEKES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
the base, although the phyletic sequence through the middle
to the upper beds is not so clear. Symborodon suddenly appears
in the middle beds.
The names of three of the genera thus recognized
were subsequently changed, for the reasons given, as
follows: For " Titanotherium Leidy" was substituted
Menodus Pomel; for " Ilegacerops Leidy" was substi-
tuted Brontops Marsh; for "Symborodon Cope" was
substituted Megacerops Leidy; " Brontotherium Marsh"
was permanently accepted. The phyla subsequently
were called subfamilies. (See below.)
RECLASSIFICATION OF THE EOCENE AND OLIGOCENE
SUBFAMILIES BY OSBORN (1914)
Original reference. — Geol. Soc. America Bull., vol. 25,
pp. 403-405, Sept. 15, 1914 (Osborn, 1914.409).
Reasons for reclassification. — Osborn makes the
following statement :
Recent discoveries have modified the author 's earlier opinions
as to the lines of descent of the titanotheres, and still further
changes are anticipated with increase of knowledge of the
connections between upper Eocene, or Uinta, titanotheres and
those of the lower Oligocene, or White River.
The main lines of division are indicated in the proportions of
the limbs, whether cursorial, mediportal, or graviportal; the
proportions of the skull, whether mesaticephalic, brachy-
cephalic, or dolichocephalic; the development of frontonasal
horns, whether accelerated or retarded; the molarization of the
premolar teeth, whether accelerated or retarded; the presence
or absence of incisor teeth; the abbreviate or elongate, the tri-
angular or oval form of the frontonasal horns as developed in
Oligocene times.
The new arrangement. — With these criteria the vari-
ous phyla were distinguished in 1914 as follows:
A. Wind River titanotheres, face longer than cranium:
I. Lambdotheriinae, light-limbed, cursorial:
Lambdotherium.
II. Eotitanopinae, medium-limbed, mediportal:
Eotitanops.
B. Bridger and succeeding titanotheres, cranium longer than
face:
III. Palaeosyopinae, short-limbed, brachycephalic :
Palaeosyops, Limnohyops.
IV. Telmatheriinae, mesaticephalic to dolichocephalic:
Telmatherium, Sthenodecies.
V. Diplacodontinae, dolichocepahlic, with accelerated
molarization of the premolars, imperfectly known:
Diplacodon.
VI. Manteoceratinae, mesaticephalic to brachycephalic,
accelerated development of the horns, mediportal:
Manteoceras, Protitanotherium.
VII. Dolichorhinae, mesaticephalic to dolichocephalic,
limbs, so far as known, abbreviate:
Dolichorhinus, Mesatirhinus, Sphenocoelus, Meta-
rhinus, Rhadinorhinus.
VIII. Menodontinae, mesaticephalic to dolichocephalic,
with abbreviate, triangular horns, with incisor teeth
reduced or wanting, feet and limbs elongate:
Menodus { = Titanotherium), Allops.
IX. Brontopinae, brachycephalic, horns abbreviated,
rounded, or oval, incisors persistent:
Brontops {= Megaceratops^") , Diploclonus.
" Error; should have been Megacerops.
B. Bridger and succeeding titanotheres — Continued.
X. Megaceropinae, mesaticephalic to extreme brachy-
cephalic, horns elongate, vertically placed, no in-
cisor teeth:
Megacerops {^Symborodon).
XI. Brontotheriinae, mesaticephalic to brachycephalic,
horns elongate, transversely flattened and diver-
gent:
Brontotherium.
The free use of subfamily divisions to express the
distinct phyletic series is similar to that which the
author adopted in the phylogeny of the rhinoceroses.
More conservative usage would have divided the titano-
theres into four subfamilies only. Of these names
of phyla those assigned to Nos. II, IV, V, VI, VIII,
IX, X, and XI had apparently not hitherto been pub-
lished, and those assigned to Nos. I, III, and VII,
although they had been used in previous publications,
mostly by other authors, were now used in a more
restricted sense.
Other subfamilies awaited further study and the
discovery of connecting forms, namely :
Diplacodontinae = ancestors of Menodontinae or Bronto-
theriinae.
Eotitanopinae = ancestors of Palaeosyopinae.
Rhadinorhininae = ancestors of Megaceropinae.
Each subfamily name is carried back as far as possi-
ble— that is, to the point, even very remote, where the
subfamily characters and tendencies of evolution are
first clearly and unmistakably manifested.
SPECIES WRONGLY REFERRED TO THE TITANOTHERES
Palaeosyops minor Marsh, 1871 (=Anchippodus minor)
Original reference. — Am. Jour. Sci., 3d ser., vol. 2,
p. 36, 1871 (Marsh, 1871.1).
Type. — "A molar tooth, from the right lower jaw,
and probably by some other less characteristic re-
mains" from Grizzly Buttes, Bridger Basin, Wyo.
Present determination. — This specimen was wrongly
referred to Palaeosyops, as was recognized by Marsh,
Cope, and others. The specimen pertains to the order
Tillodontia.
Helotherium procyoninum Cope, 1872
Original reference.— Fed. Bull. No. 2, p. 466, 1872
(Cope, 1872.2).
Synonymy. — LambdotTierium procyoninum Cope,
Tertiary Vertebrata, pp. 631, 711, pi. 24, fig. 22, 1884
[1885] (Cope, 1885.1).
"Syn.? of OroUppus pumilis," Hay (1902.1, p. 612).
Hyracotherium procyoninum Matthew, Am. Mus.
Nat. Hist. Bull., vol. 12, p. 45, 1899 (Matthew,
1899.1).
OroMppus sp. Granger, Am. Mus. Nat. Hist. Bull.,
vol. 24, p. 227, 1908 (Granger, 1908.1).
Daeodon shoshonensis Cope, 1878
Original reference. — Pal. Bull. No. 30, "December 3,
1878" (Cope, 1878.1).
SYSTEMATIC CLASSIFICATION OF THE TITANOTHERES
247
Type and geologic horizon. — "The terminal portion
of the lower jaw of a huge mammal " (Am. Mus. 7387),
from the Miocene of Oregon.
Present determination. — The genus and species be-
long in the family Entelodontidae (Peterson, 1909.1,
p. 63).
SECTION 2. CLASSIFICATION OF THE TITANOTHERES
ADOPTED IN THIS MONOGRAPH
SYNOPSIS OF THE CLASSIFICATION
The natural classification or ancestral tree of the
titanotheres is based on the characters of the skull
and teeth, as set forth in Chapters V and VI, com-
bined with those of the limbs and feet, as set forth in '
Chapter VII. The full definitions of the family and
of the 12 subfamilies into which the titanotheres are
now divided are presented in Chapters V and VI, of
which the following classification is a synopsis. It
should be compared with the phylogenetic tree given
m Chapter X (p. 769). Each of the chief phyla has
H subfamily name.
A. Wind River titanotheres, face longer than cranium:
I. Lambdotheriinae, light-limbed, cursorial:
Larnhdoiherium.
II. Eotitanopinae ( = ?Palaeosyopinae), medium-
limbed, mediportal:
Eoiitano-ps.
B. Bridger and succeeding titanotheres, cranium longer than
face:
III. Palaeosyopinae ( = ?Eotitanopinae), short-limbed,
brachycephalic:
Palaeosyops, Limnohyops.
IV. Telmatheriinae, mesatioephalic to dolichocephalic:
Telmatherium, Sthenodectes.
V. Manteoceratinae ( = Brontopinae), mesaticephalic to
brachycephalic, accelerated development of the
horns, mediportal:
Manteoceras, Protitanoiherium, Brachydiasie-
ntatherium.
VI. Dolichorhininae, mesaticephalic to dolichocephalic;
limbs, so far as known, abbreviate; facial region
downturned:
Eomeiarhinus, Dolichorhinus, Mesatirhinus,
Sphenocoelus, Metarhinus.
VII. Rhadinorhininae ( = ?Megaceropinae), mesatice-
phalic, facial region cyptocephalic, upturned:
Rhadinorhinus.
VIII. Diplacodontinae ( = ?Menodontinae, =7Bronto-
theriinae), dolichocephalic, with accelerated
molarization of the premolars, imperfectly
known:
Diplacodon, Eotitanoiherium.
IX. Brontopinae (= Manteoceratinae), brachyce-
phalic, horns abbreviated, rounded or oval,
incisors persistent; premolars retarded:
Teleodus, Brontops { = Megacerops), Diplo-
clonus.
X. Menodontinae ( = ?Diplacodontinae), mesatice-
phalic to dolichocephalic, with abbreviate, tri-
angular horns, with incisor teeth reduced or
wanting, feet and limbs elongate, premolars
accelerated:
Menodus {= Titanotherium) , Allops.
B. Bridger and succeeding titanotheres — Continued.
XI. Megaceropinae ( = ?Rhadinorhininae), mesatice-
phalic to extreme brachycephalic, horns elon-
gate, vertically placed, no incisor teeth:
Megacerops {=^ Symhorodon) .
XII. Brontotheriinae ( = ?Diplacodontinae), mesatice-
phalic to brachycephalic, horns elongate, trans-
versely flattened and divergent, premolars
accelerated:
Bronlotherium.
Suggestions as to resemblance or the affinity between
subfamilies are given above in parentheses, and the
families are arranged according to the general geologic
sequence. One of these suggestions of ancestral
affinity is now apparently well established, namely,
that the Manteoceratinae are ancestors of the Bron-
topinae.
I. TITANOTHERES OF LOWER EOCENE TIME
(Face elongate)
Group I. Hornless:
1. Subfamily Lambdotheriinae Osborn. "Lamb-
dotheres." (Lower Eocene titanotheres.
Long-headed, very small; body and limbs
slender and cursorial; face longer than cra-
nium, slender.) Pages
Genus Lambdolherium Cope 168,279,690
Species priscum Osborn 194, 286, 590
primaevum Loomis 178, 283, 590
popoagicum Cope 168, 281, 590
progressum Osborn 194, 286, 590
magnum Osborn 199,288,590
2. Subfamily Eotitanopinae (— ?Palaeosyopinae)
Osborn. " Eotitanopines. " (Lower Eocene
titanotheres of intermediate size. Head of
medium length; body and limbs less slender
and cursorial than in the lambdotheres; gait
submediportal; face longer than cranium.)
Genus Eoiitanops Osborn 179, 289, 591
Species gregoryi Osborn 192, 291, 593
brownianus (Cope) 169,292
borealis (Cope) 168,292,593
princeps Osborn 193,295,593
major Osborn 193, 296, 597
minimus Osborn 199,296
II. TITANOTHERES OF MIDDLE AND UPPER EOCENE TIME
(Face abbreviate)
Group II. Retarded horn rudiments:
3. Subfamily Palaeosyopinae (=? Eotitanopinae)
Steinmann and Doderlein. "Palaeosyo-
pines." (Titanotheres larger than tapirs.
Broad-headed, skull and limb proportions be-
coming stout; skull broad; zygomata progres-
sively brachycephalic; grinders small; nasals
tapering distaUy; face shorter than cranium;
feet abbreviate, brachypodal; gait gravi-
portal.) Pages
Genus Limnohyops Marsh (mesaticephalic
to brachycephalic) 170,303,612
Species prisons Osborn 180, 306
laevidens (Cope) 163,305
matthewi Osborn 180, 308
monoconus Osborn 180, 309, 614
laticeps Marsh 160, 311, 618
248
TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
Group II. Retarded horn rudiments — Continued.
3. Subfamily Palaeosyopinae — Continued. Pages
Genus Palaeosyops Leidy (bracliycephalic,
hyperbrachy cephalic) 157, 312, 619
Species ffoniinalis Cope 165,317
longirostris Earle 172,319
paludosus Leidy 157, 319
major Leidy 158, 321, 620
grangeri Osborn 181, 335
leidyi Osborn 18 1, 323, 620
robustus (Marsh) 161, 331, 626
copei Osborn 181, 336, 629
4. Subfamily Telmatheriinae Osborn. "Telma-
theres." (Middle and upper Eocene titano-
theres of larger size. Heads of medium length,
with large cingulate incisors and heavy, sub-
lanceolate canines; grinders large; mesatice-
phaUc {Telmatherium) or subbrachycephalic
(Sthenodectes) ; of mediportal gait.)
Genus Telmatherium Marsh (mesaticephalic,
narrow sagittal crest) 160, 340
Species cuUridens (Osborn, Scott, and
Speir) 168, 341
validum Marsh 160, 344
altidens Osborn 184, 351
ultimum Osborn 184, 345
Genus Sthenodectes Gregory (mesaticephalic
to subbrachycephalic) 190, 353
Species incisivus (Douglass) 185, 354
Group III. Accelerated horn rudiments :
5. Subfamily Manteoceratinae (Brontopinae) Osborn.
" Manteoceratines " (prophet-horn), "bronto-
pines." (Precociously horned titanotheres, of
the same stock as the Dolichorhininae. Skull
mesaticephalic, face abbreviate; feet abbre-
viate, brachypodal, gait graviportal; premolars
retarded, incisors rounded; ancestral or related
to the Brontops phylum of the Oligocene.)
Genus Manteoceras Hatcher (horns rudi-
mentary) 177, 362, 631
Species manteoceras Hay 177, 365, 631
washakiensis Osborn 182,371
uintensis Douglass 186, 372
Genus Protitanotherium Hatcher (horns elon-
gate, oval, more prominent than in Man-
teoceras) 176, 375
Species emarginatum Hatcher 177, 377
superhum Osborn 185, 379
Genus Brachydiastematherium Bockh and
Maty (large size; upper Eocene of Tran-
sylvania) 166, 382
Species transilvanicum Bockh and
Maty 166, 382
6. Subfamily Dolichorhininae Riggs. "DoUcho-
rhines" (long-snouted). (Middle and upper
Eocene titanotheres. Typically dolichocepha-
lic and dolichopic; nasals typically long and
expanding distally; precocious horn rudiments;
infraorbital shelf usually conspicuous.)
Genus Eometarhinus Osborn (ancestral to
Metarhinus, primitive, mesaticephalic). 200, 419
Species huerfanensis Osborn 200, 420
Group III. Accelerated horn rudiments — Continued.
6. Subfamily Dolichorhininae — Continued. Pages
Genus Mesatirhinus Osborn (ancestral to
Dolichorhinus; subdolichocephalic) . 182,387,636
Species Junius (Leidy) 159, 388
megarhinus (Earle) 170, 388
peiersoni Osborn 182, 389, 641
Genus Dolichorhinus Hatcher (extremely
dolichocephalic, cyptocephalic; becoming
extinct) 177, 396, 645
Species superior (Riggs) 190, 395, 405
longiceps Douglass 188,406,651
vallidens (Cope) 162, 401
heterodon Douglass 187,416
intermedins Osborn 1 84, 405
hyognathus (Osborn) _ 169, 173, 409, 646
Jluminalis Riggs 191, 417
Genus Metarhinus Osborn (dwarfed, aber-
rant, mesaticephalic) 183, 420, 648
Species earlei Osborn 183, 420
fluviatilis Osborn 183, 421
cristatus Riggs 191, 429
riparius Riggs 191, 429
Genus Sphenocoelus Osborn (little known; (?)
branch of Mesatirhinus) 174, 417
Species uintensis Osborn 175,419
7. Subfamily Rhadinorhininae ( = ?Megaceropinae)
Osborn. " Rhadinorhines " (slender - nosed) .
(Middle Eocene titanotheres. Mesaticephalic,
cyptocephalic; infraorbital shelf reduced.)
Genus Rhadinorhinus Riggs (nasals short,
pointed; possibly ancestral to the Oligo-
cene Megacerops; cyptocephalic) 192, 430
Species abbotti Riggs 192, 430
diploconus (Osborn) 173, 431
Group IV. Short-horned:
8. Subfamily Diplacodontinae (=?Menodontinae,
Brontotheriinae) Osborn. " Diplacodonts. "
(Upper Eocene ancestors of the Oligocene sub-
family Menodontinae. Heads probably mesa-
ticephalic; grinding teeth foreshadowing the
menodont type.)
Genus Diplacodon Marsh (horns well devel-
oped) 166,439
Species eZa^Ms Marsh 166,439
Genus Eoiitanotherium Peterson (horns well
developed) 196, 435, 656
Species osborni Peterson 195, 435, 656
in. TITANOTHERES OF LOWER OIIGOCENE TIME
(Face extremely abbreviate)
Group I. Short-horned:
9. Subfamily Brontopinae (Manteoceratinae) Osborn.
" Brontopines. " (Lower Oligocene and possibly
middle to upper Eocene titanotheres. Progres-
sively brachycephalio, with short-crowned teeth
and moderately short feet; horns short, sub-
oval; incisor teeth persistent, rounded crowns,
one or two pair; premolars with retarded
tetartocones.) Pages
Genus Teleodus Marsh (with three lower
incisors; basal Oligocene) 227,481
Species avus Marsh 228, 481
primitivus (Lambe) 235, 482
SYSTEMATIC CLASSIFICATION OF THE TITANOTHERES
249
Group I. Short-horned — Continued. Pages
9. Subfamily Brontopinae — Continued.
Genus Brontops Marsh (with two or one
lower incisors; lower Oligooene) . 222, 482, 664-676
Species brachycephalus (Osborn) 231,
483, 675, 676
dispar Marsh 223, 488, 664
robustus Marsh i___ 222, 492, 666
fangustigenis (Cope) 219,482
Genus Diploclonus Marsh (with internal
branching horns; lower Oligocene) 227,
499, 675-678
Species ftyleri (Lull) 234, 502, 675
fbicornutus (Osborn) 231,501
amplus Marsh 227, 504
selwynianus (Cope) 225,502
10. Subfamily Menodontinae ( = ?Diplacodontinae)
Osborn. "Menodonts." (Lower Oligocene
and possibly upper Eocene titanotheres. Heads
of medium width, progressively elongating
(Menodus) or broadening (Allops); horns
short, trihedral in section; incisor teeth vesti-
gial; grinding teeth long-crowned with promi-
nent cingula; premolars with accelerated
tetartocones.)
Genus Menodus Pomel {— Titanolherium
Leidy) 204, 522, 681
Species heloceras (Cope) 212, 524, 681
iorvus (Cope) 210,525
proutii (Owen, Norwood, and
Evans) 205,526
trigonoceras {Cope) 213,528,683
varians (Marsh) 223,535
giganieus Pomel 204, 530, 687
Genus Allops Marsh 224, 506, 678
Species walcotli Osborn 241, 509
marshi (Osborn) 233,511,678
serotinus Marsh 225, 515
crassicornis Marsh 228,517,679
Group II. Long-horned: Pages
11. Subfamily Megaceropinae ( = ?Rhadinorhininae)
Osborn. "Megaceropines," "symborodonts."
(Relatively small, long-horned titanotheres,
possibly descended from Rhadinorhinus. Of
lower Oligocene age. Horns precociously
evolved, with little or no connecting crest;
head mesaticephalic to brachycephalic, oypto-
oephalic; narrow-lipped; premolars with pre-
cocious tetartocones; grinding teeth without
cingulum; vestigial incisor teeth.)
Genus Megacerops Leidy { = Symborodon
Cope) (horns rounded, erect) 208,541,691
Species riggsi Osborn 242, 550
assiniboiensis Lambe 239, 549
copei (Osborn) 235, 548
acer Cope 211, 545
bucco (Cope) 212, 544
coloradensis Leidy 208, 544
?syceras (Cope) 226, 549
12. Subfamily Brontotheriinae ( = ?Diplacodontinae)
Osborn. "Brontotheres." (Lower Oligocene
titanotheres. Primitively dolichocephalic, pro-
gressively mesaticephalic and brachycephalic,
slightly cyptocephalic; broad-lipped; very pre-
cocious development of the horns; accelerated
development of internal cones of superior pre-
molars; prominent cingulate incisor teeth in
males.)
Genus Broniotherium Marsh (horns progres-
sively elongate, nasals abbreviate; lower
Oligocene) 209, 555, 690
Species leidy i Osborn 234, 558, 690
hypoceras (Cope) 216,562
hatcheri Osborn 235, 563, 695
tichoceras (Scott and Osborn). 219, 565
gigas Marsh 209, 567
dolichoceras (Scott and Osborn). 220, 572
medium (Marsh) 228,576
curtum (Marsh) 224, 574
ramosum (Osborn) 231,577
platyceras (Scott and Osborn). 221, 578
frumelicum (Toula) 230, 560
Note. — Additional species are described and classified in the appendix,
including Mongolian, east European, and Burmese titanotheres.
-29— VOL 1 19
CHAPTER V
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
SECTION 1. GENERAL PRINCIPLES OF THE STUDY OF
THE CHARACTERS OF THE SKULL AND TEETH
PROPORTION CHARACTERS AND TENDENCIES OF EVOLU-
TION DISTINGUISHED BY ANALYSIS AND SYNTHESIS
The key to the evolution of the titanotheres is
afforded by the analysis and synthesis of the separate
characters of which the individuals in each line of
ascent are composed — characters large or small, single
or multiple — in correlation with one another and in
correlation with the individual as a whole; characters
progressing or retrogressing in successive generations;
characters evolving rapidly or evolving slowly: such
is the composition of each individual titanothere, as
well as of each phylum.
Out of an almost infinite number of characters that
are independently evolving we select a few that are
visible and measurable. In a few individuals we
observe the origin of new characters, but generally
we observe the changes of form and of proportion in
existing characters, which make up the greater part
of the transformation of the individuals composing
the family. In heredity each character is a separate
unit, completely separable from all others; in adap-
tation it is correlated with other characters of the
individual, as is fully explained in Chapter IX.
DISTINCTIONS BETWEEN PROPORTION CHARACTERS AND
NEW RECTIGRADATION CHARACTERS
Methods employed. — The present chapter explains
how the characters of titanotheres have been observed,
examined, and measured, partly by new methods,
largely devised especially to solve the problems that
have arisen in the task of working out the genealogy
of this family, and partly by old methods, which have
been in use by paleontologists and systematists. To
distinguish the characters of the teeth, skull, and
limbs, which are generally but fragments, necessitates
very refined and precise systems of measurement and
comparison, because the individual members of differ-
ent lines of descent may be very close to each other
in certain characters yet readily separable in others.
Animals that the zoologist would readily distinguish
as species and subspecies by their external coloring,
bodily form, or habits of life may be extremely simi-
lar in skeletal characters, yet the close methods of
measurement and analysis that we have been com-
pelled to adopt prove that every character has dis-
tinctions that may be revealed by minute and precise
observation.
Researches on proportion characters. — The chief
papers on the principles of evolution of the mammaJian
skull and teeth which the author has published in the
investigation of the titanotheres are the following:
1896.110. Th • cranial evolution of Titanotherium: Am. Mus.
ra . Hist. Bull., vol. 8, art. 9, pp. 157-197, July
31, 1896.
1902.207. Dolichocephaly and brachycephaly in the lower mam-
mals: Am. Mus. Nat. Hist. Bull., vol. 16, art. 7,
pp. 77-89, Feb. 3, 1902.
1902.208. The four phyla of Oligocene titanotheres: Am. Mus.
Nat. Hist. Bull., vol. 16, art. 8, pp. 91-109, Feb.
18, 1902.
1907.301. Evolution of mammalian molar teeth to and from
the triangular type, 250 pp., New York and Lon-
don, Macmillan Co., September, 1907.
1912.368. Skuh measurements in man and the hoofed mammals:
Science, new ser., vol. 35, No. 902, p. 596, Apr. 12,
1912.
1912.372. The continuous origin of certain unit characters as
observed by a paleontologist (Harvey lecture) :
Am. Naturalist, vol. 46, No. 544, pp. 185-206,
April, 1912; No. 545, pp. 249-278, May, 1912;
Harvey Soc. Volume, 7th ser., pp. 153-204, No-
vember, 1912.
1912.382. Craniometry of the Equidae: Am. Mus. Nat. Hist.
Mem., new ser., vol. 1, pt. 3, pp. 57-100, figs. 1-17,
June, 1912.
1915.416. Origin of single characters as observed in fossil and
living animals and plants (Presidential address
before the Paleontological Society of America,
Dec. 31, 1914) : Am. Naturalist, vol. 49, No. 580,
pp. 193-239, AprO, 1915.
Proportion characters defined hy indices and ratios. —
Many specific, generic, and subfamily characters of
animals can be best expressed in mathematical ratios
and indices, for these figures record most precisely the
movements or tendencies of development that lead
from species to species. In all ascending series of
titanotheres every measurable character is in a state
of movement either progressively or retrogressively.
Significance of tendencies in proportion. — A tendency
or trend to evolve in a certain proportional direction
is found to be a phyletic distinction of prime im-
portance, which leads us through all the stages of
mutative, specific, generic, and subfamily characteris-
tics. For example, certain titanotheres become more
and more broad-headed from lower to higher geologic
levels; progressive brachycephaly thus becomes a
phyletic character of taxonomic value. One genus
may be defined as "progressively brachycephalic,"
whereas a related genus, in which the tendency to
become long-headed prevails, may be defined as
" progressively dolichocephalic."
Taxonomic value oj proportion tendencies.- — Thus the
true relations of most of the lines of descent among the
Eocene and Oligocene titanotheres have been grad-
ually discovered, partly by the old methods of descrip-
tive anatomy, used by Leidy, Marsh, and Cope, and
partly by the new methods which have been developed
since 1900 in the preparation of this monograph. It
251
252
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
has been found that the changing proportions of the
various parts of the slvull, of the individual grinding
teeth, of every part of the skeleton, especially the
limbs, are highly distinctive systematic and phyletic
characters.
Five distinctions of 'phyla. — Each line of descent is
distinguished by five different methods: First, by the
presence or absence of certain characters; second, by
the new proportions of certain characters; third, by
the tendencies or directions in which proportions are
being changed; fourth, by the rates of change in
proportion characters, whether retarded or acceler-
ated; fifth, by the appearance of new rectigradation
characters.
Numerous extinct iranclies or phyla. — The distinc-
tions in characters multiply with the multiplication of
the phyla. In 1914 no less than 20 branches of the
titanothere famdy were known, and probably many
more existed that had not yet been discovered.
Throughout Eocene time titanotheres continued to
migrate into the mountain region of the Bridger and
Washakie Basins of Wyoming. Allowing for certain
branches that drop out, we find that the number of
their known branches constantly increases from lower
to higher levels, as shown below.
Oligocene : White River group 7-8
Summit of upper Eocene: Lower part of Uinta C (true
Uinta formation), Uinta Basin, Utah 4
Upper Eocene: Uinta B 2 of Uinta Basin, Utah; Washakie
B 2 of Washakie Basin, Wyo 6
Upper Eocene: Washakie B 1 of Washakie Basin, Wyo.;
and Uinta B 1 of Uinta Basin, Utah 8
Middle Eocene: Bridger C and D of Bridger Basin, Wyo.;
Wasliakie A of Washakie Basin, Wyo 5
Middle Eocene: Bridger A and B of Bridger Basin, Wyo.;
Huerfano B, Huerfano Park, Colo 2
Lower Eocene: Wind River formation, Wind River Basin,
Wyo.; Huerfano A, Huerfano Park, Colo 2
Universal change oj form. — No characters in any
genus or phylum are stationary. During the long
intervals of geologic time the members of each of these
branches were constantly diverging in some characters
and converging in others and becoming more and
more unlike one another both as a whole and, so far as
we can observe, in each one of their single characters.
ATLometrons and rectigradations. — The term allome-
trons (Osborn, 1912.372, pp. 249-278) designates
characters that arise through continuous changes of
size or proportion in old features — that is, purely
quantitative changes — such as may be expressed in
differences of measurement as well as in indices and
ratios. Rectigradations are new characters that tend
to evolve in a definite direction — the earliest "rudi-
ments" or discernible stages of absolutely new forms.
In 1889 Osborn called such characters "definite vari-
ations" (Osborn, 1907.301, p. 239).
Six points in the distinction between allometrons
and rectigradations may be readily grasped: (1)
When the shadowy beginning of a new cusp on the
grinding teeth or the rudiment of a horn is first dis-
cernible as a new character it appears as a "rectigrada-
tion"; (2) when this same rudiment of a cusp or horn
takes on a new shape the change of form appears as
an " allometron " ; (3) in the hard parts of a titano-
there, as of any other mammal, the rectigradations —
the numerically new characters of any kind — are com-
paratively few and uncommon, but the allometrons —
the transformations of existing characters — comprise
the larger number of changes; (4) both allometrons and
rectigradations are distinctly heritable characters; (5)
in the genesis of rudiments (rectigradations) of new
cusps or of horns all the branches or phyla of titano-
theres sooner or later tend to resemble one another —
that is, to develop the same cusps and the same horn
swellings — and thus to become convergent; (6) on the
other hand, in changes in the proportions (allometrons)
of the skull, the different phyla may differ widely from
one another and through dissimilar allometrons may
become divergent. (See fig. 210.)
STEPS IN TRANSFORMATION OF CHARACTERS
So far as we have observed, all absolutely new char-
acters that we have traced to their very beginnings in
titanotheres arise gradually and continuously; there is
no evidence of sudden leaps from mutation to muta-
tion or from species to species. This continuous mode
of evolution is more fully considered in Chapter IX.
The addition (rectigradation) or the modification
(allometron) of a single character is theoretically the
first step in transformation, but as a matter of fact
all characters are being simultaneously more or less
modified, and in the individual as a whole new char-
acters are constantly being added. Only when fully
developed after the lapse of many generations does a
rectigradation or an allometron become of sufficient
systematic value to define the mutation or the species.
None the less each of these changes forms one in a
series of steps in the transformation of species.
One by one the characters, either rectigradations or
allometrons, in many parts of the titanothere are inde-
pendently changed until the changes build up what
paleontologists call an "ascending mutation" in the
sense in which the German invertebrate paleontologist
Waagen defined this term in 1869. An ascending mu-
tation is a stage in a continuous evolutional ascent in
one or more characters from one species to another;
there is no evidence that it is a saltation or "mutation"
in the sense of that word as used by De Vries.
Finally these rectigradations and allometrons attain
by accumulation sufficient importance to enable us to
call a stage a "species" in the Linnaean sense or a
"subspecies" in the modern sense.
The divergence between the several branches of the
titanothere family therefore actually consists of the
sum total of changes in an almost infinite number of
single characters, only a few of which are measurable.
These changes are of the following principal lands:
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITAN0THERE3
253
V
w
m
I
Recit^ra^daiionS
NetM cliarcioiers, similar, hu-i
ept
Dnylu.)
jillomeiroyvs
New proporiions disii-noiiz>e
oj ea-ch phylum. Dijjeveni
-ow alt oiher phyla.
t,n eveyxj phyL
FiGUBE 210. — Skulls showing different numerical and proportional characters in five separate phyla of
titanotheres
Similar numerical characters (rectigradations, A, B) and dissimilar proportional characters (allometrons, C, D), all arising independently in
descendants of the same ancestors. Each of the five phyla (I-V) exhibits similar rectigradations of the premolar teeth and in the osseous
horn rudiments (H) but dissimilar allometrons of the sliull ( C) and of the foot bones {D). pa'', mes^, New cuspules on the teeth! H, rudi-
diments of the newly arising horns. I, Eotitanops, a dolichocephalic ancestral form; II, Palaeosyops, brachycephalic; III, Telmatherium,
mesaticephalic; IV, Manteoceras, mesaticephalic; V, Dolichorhmus, dolichocephalic.
254
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
1. Loss of old characters (paleomorphs) : Absolute
loss of character is rare in the titanotheres. The tra-
pezium is the only bone known to be lost duriag the
recorded history of the family, whereas in the horses
many bones disappear. The incisor teeth disappear in
several phyla. The grinding teeth lose certain cusps.
2. Appearance of new characters (neomorphs), in-
cluding rectigradations : A large number of new cusps
appear on the premolar grinding teeth, and a few new
features appear on the molar grinding teeth. A pair
of new hornlets appear on the frontonasal region of
the skull.
3. Increase or diminution of size as a whole (hetero-
morphs) : Increase of size is the prevailing tendency
among the titanotheres, but in some phyla size is per-
sistent or is even arrested, as in the dwarf Metarhinus;
or it is reduced, as in the species Palaeosyops copei.
4. Change in proportions of different regions of the
skeleton (heteromorphs, including allometrons) : Such
changes are expressed in ratios — for example,
length of tibia
length of femur
These ratios are extremely significant. Thus one phy-
lum may become short limbed or brachymelic, another
long limbed or dolichomelic ; one short footed or
brachypodal, another long footed or dolichopodal; one
phylum may become large headed or macrocephalic,
another small headed or microcephalic.
5. Change in proportions of single parts of the
cranial skeleton (allometrons) : Such changes are best
expressed in indices, such as
width of skull
length of skull
Some series become long headed or dolichocephalic,
others broad headed or brachy cephalic; some become
long faced or dolichopic, others become short faced or
brachyopic.
The manner in which these changes of proportion
(allometrons) and the successive addition of rectigra-
dations serve to distinguish the genera of Eocene
titanotheres from one another is clearly shown in the
following descriptive characterizations of ten Eocene
genera :
Lambdotherium: Small, long headed, long limbed, without
horns.
Eotitanops: Larger, long faced, limbs somewhat heavier, horn-
less.
Limnohyops: Still larger, broad headed, short faced, light
limbed, broad footed, hornless.
Palaeosyops: Massive, broad headed, short faced, heavy limbed,
short footed, rudiments of horns (rectigradations).
Telmatherium: Large, long headed, short faced, light limbed.
rudiments of horns (rectigradations).
Manteoceras: Large, medium headed, short faced, medium
limbed, short footed, small, distinct horns (rectigradations).
Mesatirhinus: Of medium size, narrow headed, short faced, light
limbed, long footed, small horns quite distinct.
Metarhinus: Very small, medium headed, short faced, light
limbed, horns not very distinct.
Dolichorhinus: Large, extremely long headed, short limbed, short
footed, horn rudiments very prominent.
Rhadinorhinus: Medium size, medium headed, light limbed,
horn rudiments indistinct.
The degrees of change among the "species" con-
stituting each of these "genera" are exemplified in
the "standard measurement tables" that accompany
the description of every genus in Chapters V and VI.
In these tables it is demonstrated, first, that the
Linnaean lines of division between species do not exist ;
second, that occasionally the type and paratype spec-
imens of a single species selected by the pioneer pale-
ontologists belong to separate stages because they
were found at different geologic levels. Some of these
ancient specific names have historic value and are
retained for convenience, although some that were
applied to forms on the border line between two specific
stages are very inconvenient.
PROPORTIONS AND FLEXURES OF THE SKULL
The skull is the chief center of evolution movement
in the titanotheres, both in the transformation of its pro-
portions and in the development of horns, and with the
teeth it furnishes a complete key to the evolution,
relationship, and ascent of these mammals, although
the proportions of the skeleton and the feet also
furnish valuable indications. The forms of the feet,
which are evolving so rapidly and are so significant in
the horses, are relatively stationary in the titanotheres.
The chief principles in the transformation of the
skull through changes of proportion are the following:
1. Elongation and narrowing of the skull as a whole — that
is, dolichocephaly.
2. Abbreviation of the cranium and elongation of the face —
that is, proopic dolichocephaly or dohchopy, as in Equus.
3. Abbreviation of the face and elongation of the cranium. —
that is, postopic dolichocephaly or brachyopy, as in all Oligocene
titanotheres.
4. Abbreviation and broadening of the skull as a whole —
that is, brachycephaly.
5. Flexure of the facial upon the cranial region — that is,
cyptocephaly.
The principal measurements of the skull in the
titanotheres differ somewhat from those employed in
the craniometry of the Equidae. (Osborn, 1912.382.)
They are listed below and are illustrated in Figure 211.
Direct measurements of skull
1. Cephalic or basilar length from incisive border to occipital
condyles, inclusive.
2. Facial length from postorbital process to incisive border,
inclusive (projected on the basilar Une).
3. Cranial length from postorbital processes to occipital
condyles, inclusive (projected on the basilar line).
4. Facial breadth, or frontal width, as measured in the
horses across the postorbital processes to establish the cephalic
index. This measurement is not practicable in titanotheres.
5. Zygomatic breadth across the widest part of the zygomatic
arches.
6. Premolar-molar length, or superior grinding series (p' to
m^ if the first premolar is retained, otherwise p^ to m^).
7. Molar length, anteroposterior measurement along middle
of crowns (m' to m') .
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
255
8. Molar length and width, anteroposterior measurement of
first superior molar (m') and transverse measurement of the
same (least width, near center of crown).
Indices of skull (expressed as per cent)
- „ , . , ,. ■ J breadth across zygomata
9. Zygomatic-cephalio mdex=r — n , .. ^ — —nr
■^ ° basilar or cephalic length
length of six superior
r, 1 1 1, 1- • J grinding teeth, p^-m'
Premolar-molar cephahc index=^ iT't — j — <£
iiiujc^i ^^ij^ia.j^ iiiiic^v cephalic length
I, ,, , u 1- ■ J length of upper true grinders, m'-3
II. Molar-cephalic index =■ — ^, -,.-, -,-,
'^ cephalic length
, „ - , , . , breadth of a molar, mi or m^
12. Molar index= -. rr — c i
length of a molar
lensth of face ^'
13. Faciocephalic index= ^r^. — , — -r;—
'^ cephalic length
length of cranium -'
cephalic length
14. Craniocephalic index =-
Flexures of skull
15. Palatooranial flexure = angle that the palate, from the
posterior to the incisive border, makes with the basal line of the
cranium. (Not used in the study of titanotheres.)
16. Faciocranial flexure = angle that the preorbital part of the
skuU, determined from the optic foramen (see figs. 213, 214) to
the incisive alveolus, makes with the line from the optic foramen
to the middle of the occipital condyle.
SUMMARY AS TO CRANIOMETRY
1. Direct measurement. — Since the fossil skulls and
dental series are rarely complete or perfect, the paleon-
tologist requires an additional series of direct detailed
measurements of parts of the skull and teeth that are
not needed by the zoologist.
2. Significance. — Every one of these direct measure-
ments, indices, and angles is significant, because all
skulls are in a continuous process of movement, or
evolution. The indices are even more significant than
the direct measurements, because every genus and
probably every species has its distinctive indices in
adult specimens, and the direct measurements vary
greatly with the age, sex, and individual variation of
the specimen.
3. Imperfection. — In fossil skulls the indices are
often difficult to determine; a slight crushing or dis-
tortion seriously disturbs the index, for a skull that is
crushed on its side is narrowed and lengthened at the
same time. Nevertheless, the indices and ratios
should be used wherever obtainable.
4. Age. — The proportions between the several parts
are largely altered with the age of the animal; this
statement is especially true of progressive allometrons,
such as the proportion between the face and the
cranium. Thus the faciocephalic or craniocephaHc
index may alter rapidly as the titanothere advances
from youth to maturity; similarly the flexure (cyp-
tocephaly) becomes extreme only in mature skulls.
The age of the animal measured is thus to be con-
sidered in all the indices and ratios of the skull, teeth,
and skeleton.
'* As projected on basilar line.
5. Sexual correlations. — Certain proportions and indi-
ces are correlated sexual characters — that is, in brachy-
cephalic phyla the males have relatively broader heads
than the females. For example, we observe in the
genus Brontops the following proportions :
Males of Brontops validus (braohycephalio), indices 73-87.
Females of Brontops validus (mesaticephalic), indices
60-70.
FiGUBE 211. — Standard measurements of Eocene
titanothere skulls
Palatal view of a skull of a middle Eocene titanothere, Mesatirhinus
petersoni, showing (by arrows) how the basilar or cephalic length,
the zygomatic width, and the length and width of the upper pre-
molars and molars are measured.
6. Effects of crushing. — The indices of fossil skulls
are profoundly modified by vertical or lateral crush-
ing: vertical crushing tends to give brachycephalic
indices; lateral crushing tends to give dolichocephalic
indices. To these facts are due in part the wide
variations in the tables of indices, especially in the
chapter on the Oligocene titanotheres.
256
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
CHANGING PROPORTIONS OF THE CRANIUM AND FACE
The back of the eye socket, or orbit, is the dividing
line between the face, which lies in front of it, and the
cranium, which lies behind it, as indicated by the
shaded and unshaded parts of the skulls in Figure 212,
which shows that in the oldest true titanothere
(Eotitanops) of lower Eocene time, as in the oldest true
horse (Eohippus) of the same period, the face (shaded)
and the cranium (outline) are equally long — in fact, in
Eotitanops the face is a little longer than the cranium.
The imiversal allometric character of titanotheres is
abbreviation of the face (proopic region) and elonga-
tion of the cranium (postopic region) — that is, brachy-
opy and dolichocrany. Thus in all middle and upper
Eocene titanotheres the cranium is longer than the
face. This disparity keeps increasing until in the
Oligocene titanotheres, such as Brontotherium (fig.
212), the face is greatly abbreviated and the cranium
greatly elongated. In the horses (Equus) this allome-
try is just reversed: the face becomes very long (fig.
E7otita7iop5
Figure 212. — Unequal elongation of face and cranium in
titanotheres and horses
Eoliianops and Eohippus, primitive perissodactyl type; face and cranium sub-
equal in length. BTontoiheTium, titanothere type; face abbreviated, cranium
elongated. B^uus, Equidae type; face elongated, cranium abbreviated.
212), but the cranium remains very short
(dolichopy and brachycrany). These differ-
ences are expressed in the so-called facio-
cephalic index, which is obtained as follows:
length of face including orbits X 100
basilar length of skull
The relative faciocephalic indices in titano-
theres and horses are as follows:
author (Osborn, 1912.382); the term "cyptocephaly"
is correct.
In primitive ungulate skulls and in the fetal skull
the anteroposterior planes of the face and palate and
of the basicranial axis are more nearly in parallel
lines — that is, the skulls are "orthocephalic."
In certain specialized ungulates there is either an
upward or a downward deflection of the face on the
Figure 213. — Faciocranial flexure, or cypto-
cephaly
In the reindeer {RoTigifeT) the face is much less bent upon the
cranium than in the hartebeest (Bubalis). A similar but less
pronounced contrast is seen in the Eocene titanotheres Palaco-
syops and Solichorhinus.
cranium, which appear respectively to be adapted to
different forms of feeding, as follows: (1) Horizontal
and upward flexure of the face is characteristic of
certain browsing types, such as Alces and Rangifer;
(2) downward flexure of the face and palate on the
basicranial axis is characteristic of certain grazing
types, such as the hartebeest (Bubalis) and other
grazing antelopes (see fig. 213); (3) in the young of
certain species of Equidae, Bovidae, and Cervidae
the palatal line makes an angle of 19° to 25° with the
Titanotheres :
Brontotherium, 33.
Eotitanops, 56.
Horses:
Equus, 70.
Eohippus, 53.
In the titanotheres the imiversal tendency
of facial abbreviation and cranial elongation
distinguishes all the branches alike, but since
the allometric movement takes place at un-
equal rates each genus or phylum has its YiGvnn 214.— Faciocranial flexure in Palaeosyops (A), orthocephalic, and
Dolichorhinus (B), cyptocephalic
distinctive faciocephalic index
CYPTOCEPHALY, OR FACIOCRANIAL FLEXURE
The upward or downward flexure of the facial and
palatal parts of the skull upon the basicranial axis
was first erroneously termed "cytocephaly" by the
basicranial line, which may increase to 53° in the
adults of extremely deflected types.
Among the many authors who have more or less
directly contributed to this subject are Riitimeyer
EVOLUTION OF THE SKULL AND TEETH OP EOCENE TITANOTHERES
257
(Cervidae, 1882.1), Flower (1885.1, pp. 185-201), Lan-
kester (Giraffidae, 1902.1), Ewart (Equidae, 1907.1).
Riitimeyer pointed out that the Cervidae, as brows-
ing animals, have an extremely horizontal axis of the
skull, in contrast with that of the Bovinae, grazing
animals, in which the face is strongly deflected.
Lankester pointed out that in the hornless Olcapia,
a forest animal that browses on the leaves of trees, the
facial and cranial regions are orthocephalic, or nearly
in the same horizontal plane; in the related Giraffa,
however, which also browses on high trees but possesses
horns, the face is deflected on the cranium almost as
much as in the grazing sheep (Ovis), which feeds upon
the ground. Lankester consequently attributed the
Osborn's examination of the horses (1912.382, p. 96)
shows that in the adult domesticated horse the
palatocranial angle ranges from 10° to 17°; in the
Burchell zebra the palatocranial angle increases with
age from 19° at three years to 25° at six years. In
the domesticated horses it varies from 20° to 23°. In
the Burchell zebras it increases from 15° at birth to
25° at the fourth year. Since the asses are more
given to browsing than the horses or zebras the
slight difference in flexure may be attributed to the
prevailing browsing habit. (See p. 259.)
In titanotheres the faciocranial angle is measured as
shown in Figure 214. In a comparison of the brachy-
cephalic Palaeosyops, presumably a browsing type with
Figure 215. — Cranial proportions of Eocene titanotheres — Palaeosyops, Manteoceras, and Dolichorhinus
A, Brachycephalic (Palaeosyops major), zygomatic-cephalio index 77. B, Mesaticephalic (J/orjieocfrasTnaHfeoccras), zygomatio-cephalio index
63-68. C, Dolichocephalic (DoUchorMnus Jiyognathus) , zygomatic-cephalic index 43^6. h, Eudiments (rectigradations) of the horns.
deflection of the face to the possession of horns.
Ewart applied cyptocephaly as a means of distinguish-
ing the various phyla of horses and, like Riitimeyer,
explained the flexures as adaptations to a prevailing
browsing or grazing habit, respectively. He pointed
out that Alces and Ovis illustrate the two extreme
types of skull: (1) the elk (Alces) is a short-necked,
forest form adapted to feeding on shrubs and trees —
that is, to holding the head in a nearly horizontal
position, (2) whereas the sheep {Ovis) grazes or
browses on the ground and is adapted to holding the
head when feeding in a nearly vertical position.
There are excellent reasons for believing that a bent
skull facilitates grazing on short herbage.
short-crowned teeth, the lines of the face and the
cranium are more nearly parallel, the angle being
154°. In the extremely dolichocephalic DolicJiorTiinus ,
which has more hypsodont teeth and presumably
subgrazing habits, the face is sharply bent down on
the cranium, forming an angle of 135°. (See fig. 214.)
Cyptocephaly, whatever its adaptive significance,
is certainly one of the important progressive characters
in the transformation of the ungulate skull and is
decidedly marked in certain titanotheres.
DOIICHGCEPHAIY, BRACHYCEPHAIY, AND CORRELATION
In 1902, when the measurements of Oligocene
titanotheres were brought together, the conclusion
was again reached that dolichocephaly and brachy-
258
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
cephaly are among the dominating tendencies affecting
the skull and grinding teeth of titanotheres, but
that they are not invariably correlated with similar
abbreviation or elongation iu the trunk and limbs.
The principle of correlation, however, was found
to explain a vast number of dry detailed facts concern-
ing all parts of the skulls of titanotheres, including the
teeth, which had been recorded by Cope, Marsh, Earle,
Osborn, and others without any appreciation of their
morphologic significance.
In brief, the progressive doUchocephaly or brachy-
cephaly of the skull is found to dominate the shape
of every bone in the skull but more particularly that
of the nasals, horns, zygomatic arches, and palate,
as well as the confluence or separation of the foramina
Manteoceras
J)oiic?>orfiinus
Figure 216. — Cranial proportions in man (A) and
in the titanotheres (B)
Viewed from above. Used in reference to man the words brachyce-
phalic, mesaticephalic, and dolichocephalic denote, respectively,
brachycranial, mesaticranial, and dolichocranial. In other words,
they describe the proportions of the cranial cavity. Used in refer-
ence to the titanotheres the same words describe the relative length
and breadth of the entire skull.
in the base and sides of the skull, the form of the
occiput and of the mastoid, and the relations of other
bones around the auditory meatus, the shape of the
premaxillary and mandibular symphyses, the dias-
temata between and behind the teeth, the number
and shape of the teeth, the shape, number, and rela-
tions of the cusps of the teeth, and even, it would
appear, the cingulum around the grinding teeth.
The three skulls of middle and upper Eocene
titanotheres shown in Figure 216 also illustrate
admirably the extremes of brachycephaly and doUcho-
cephaly and the intermediate condition of mesa-
ticephaly.
The skulls are those of the species Palaeosyops
major, with its extremely broad head, of the mod-
erately broad-headed Manteoceras manteoceras, and of
the extremely long and narrow-headed DoKcJiorMnus
Jiyognathus.
The cranium of these titanotheres varies in width
like the skulls of man (fig. 216) and the rhinoceros,
and the excessive width is contributed chiefly by the
great expansion of the zygomata. The skull of Doli-
chorhinus as a whole, however, is far more elongate than
that of Palaeosyops. We find also very pronounced
differences of proportion in every bone and every
Aa
Figure 217. — -Natural and artificial brachycephaly and doUcho-
cephaly. After Osborn, Science, 1908, pp. 750, 751
A, Palaeosyops, extreme brachycephalic type, superior view; Aa, Palaeosyops, arti-
ficial dolichocephaly produced by stretching A to length of C; B, Mesatirhinus,
primitive dolichocephaly, palatal view; Ba, Mesatirhinus, artificial dolicho-
cephaly, palatal view, produced by stretching B to length of C ; C, Dolkhorhinus,
progressive dolichocephaly, superior view; C, BolichoThinus, progressive dolicho-
cephaly, palatal view.
tooth when we compare Palaeosyops and DolichorMnus
minutely. The table on page 259 presents some of
the extremes of structure observed especially in the
titanotheres, in which the most careful comparison of
dolichocephalic and brachycephalic skulls has been
made.
When we compare a long-skulled with a short-
skulled titanothere the skull at first appears to be
EVOLUTION OF THE SKULL AND TEETH OP EOCENE TITANOTHERES
259
compressed laterally or stretched out longitudinally as
if composed of india rubber, all parts being elongated
alike. (See fig. 217.) This appearance is entirely
deceptive, because every bony element of the skull
has a different rate of elongation. As above noted,
although the face of the titanothere is abbreviated
(brachyopic), the cranium is elongated (dolichocranic),
and it is chiefly the midregion of the cranium that is
notably elongated — that is, the region between the
orbits and the postglenoid processes. There is no
predetermined innate or invariable correlation of
brachycephaly or dolichocephaly in all parts of the
skull, nor is there any fixed correlation between elonga-
tion or abbreviation of the skull and of the limbs
respectively. (See law of correlation, Chap. XI.)
ZYGOMATIC-CEPHAHC INDICES IN THE TITANOTHERES
AND OTHER PERISSODACTYLS
RELATIVE VALUE OF INDICES
The proportions and indices applied to different
groups of mammals are largely relative. The terms
applied to the human cranium and the indices are as
follows :
Brachycephalic = brachycranial, 80.1-100
Mesaticephalic == mesaticranial, 75. 1-80
Dolichocephalic = dolichocranial 60-75
Among the perissodactyl ungulates that have ordinal
relationships to the titanotheres some of the indices of
total length and width of skull are shown in the
accompanying table.
Indices oj length of sTcuU of perissodactyl ungulates related to the titanotheres
Sumatran rhinoceros (Rhinoceros (Dicerorhinus) sumatrensis) .
Black African rhinoceros (R. (Opsiceros) bicornis)
South American tapir (Tapirus terreStris)
White African rhinoceros (R. (Ceratotherium) simum)
Indian tapir (Tapirus indicus)
Domestic horse (Equus caballus)
Domestic horse (Equus caballus)
Domestic ass (Equus asinus)
Basilar length,
premaxillaries
to condyles
(millimeters)
580
568
355
387
513
Transverse
width across
zygomatic arches
(millimeters)
345
320
178
183
202
59
56
50
43-50
47
39
40. 4-44 1
46. 9-49. 9
Most of the skulls of the above-named species are dolichocephalic in comparison with the skulls of titanotheres.
A standard of skuU proportions among the perissodactyl ungulates, including the rhinoceroses, tapirs, horses,
and titanotheres, may be established as shown in the table below.
Summary of the zygomacephalic indices of the perissodactyls
Form of skull
Index
Perissodactyl ungulates
Dolichocephalic :
H yperdolichocephalic
Dolichocephalic
Subdolichocephalic
Mesaticephalic - -
39. 0-50
50. 1-55
55. 1-60
60. 1-70
70. 1-75
75. 1-85
85. 1-91 +
Horse, tapir (Indian), white rhinoceros, Dolichorhinus, tapir (South American),
Eotitanops.
Mesatirhinus petersoni, Rhadinorhinus.
Black rhinoceros, Sumatran rhinoceros, Mesatirhinus megarhinus.
Telmatherium ultimum, Metarhinus earlei, Manteoceras, Menodus sp., Allops
marshi, Menodus giganteus.
Limnohyops laticeps, AUops serotinus, Brontotherium curtum.
Brachycephalic :
Subbrachy cephalic . _
Palaeosyops leidyi, Brontotherium platyceras.
Brontops robustus, Diploclonus amplus.
Hyperbrachycephalic
INDICES OF SKULLS OF EOCENE AND
TITANOTHERES
OLIGOCENE
The study of these proportions and indices of the
titanotheres demonstrates that the skull in this
family presents an ascending scale from primitive
dolichocephalic ancestors like Lambdotherium and
Eotitanops, which, on the one hand, evolved into
extremely broad-headed forms like Palaeosyops and
Megacerops, and, on the other, into extremely long-
headed forms like Mesatirhinus and Dolichorhinus,
as shown in the following table:
260
TIT.-USrOTHERES OP ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
Zygomatic-cephalic indices in the tiianotheres
[Arranged in chronologic and taxonomic order]
Lower Eocene:
Eotitanops borealis (dolichocephalic)
Lambdotherium popoagicum (dolichocephalic)
Middle Eocene:
Limnohy ops monoconus (brachy cephalic)
Limnohyops laticeps (brachycephalic)
Palaeosyops leidyi (brachycephalic)
Upper Eocene:
Telmatherium ultimum (crushed) (mesaticephalic)-
Sthenodectes incisivus (mesaticephalic)
Manteoceras manteoceras (mesaticephalic)
Mesatirhinus megarhinus (dolichocephalic)
Mesatirhinus petersoni (dolichocephalic)
Doliohorhinus superior (dolichocephalic)
Dolichorhinus hyognathus (dolichocephalic)
Metarhinus fluviatilis (dolichocephalic)
Metarhinus earlei (dolichocephalic)
Rhadinorhinus abbotti (dolichocephalic)
Lower Oligocene:
Brontops brachycephalus, 9 (mesaticephalic)
Brontops brachycephalus, & (brachycephalic)
Brontops validus (brachycephalic)
Brontops robustus (brachycephalic)
Allops marshi (mesaticephalic)
Allops serotinus (brachycephalic)
Diploclonus amplus (brachycephalic)
Diploclonus tyleri (brachycephalic)
Menodus giganteus (mesaticephalic)
Menodus varians (brachycephalic)
Menodus heloceras (brachycephalic)
Megacerops coloradensis (brachycephalic)
Megacerops acer (brachycephalic)
Brontotherium leidyi (brachycephalic)
Brontotherium gigas (brachycephalic)
Brontotherium sp. div. (brachycephalic)
Index
50
(?)
72
75
74-75
61
63-65
63-68
56-59
49-52
49
46
58
59-63
51-52
64
72
85
76-83
64^69
72-78
91
85
62-70
73
79(?)
76
84
66
84
74-80
Generic tendencies. — Certain generic ascending series
are progressively brachycephalic — for example, Bron-
tops and Megacerops; others are progressively dolichoce-
phalic— for example, Menodus.
Zygomatic-cephalic indices of the titanotheres and other
perissodactyls
[Arranged in ascending numerical order]
Dolichocephalic :
Hyperdolichocephalic — Index
Domestic horse (Equus caballus) 39, 40. 4^44. 1
Dolichorhinus hyognathus (Eocene) 43-49
White African rhinoceros (Ceratotherium
simum) 43-50
Domestic ass (Equus asinus) 46. 9-49. 9
Indian tapir (Tapirus indicus) 47
Mesatirhinus petersoni (Eocene) 49
Dolichorhinus superior (Eocene) 49
Eotitanops princeps (Eocene) 50
Dolichocephalic —
Mesatirhinus petersoni (Eocene) 51, 52
Rhadinorhinus abbotti (Eocene).. 51-52, 53, 54, 54. 9
Subdolichocephalic —
Black African rhinoceros (Opsiceros bicornis) _ _ 56
Mesatirhinus megacrhinus (Eocene) 56-57
Metarhinus fluviatilis (Eocene) 58
Metarhinus earlei (Eocene) 59
Sumatran rhinoceros (Rhinoceros sumatrensis).. 59
Mesaticephalic : index
Telmatherium ultimum (Eocene) 61
Metarhinus earlei (Eocene) 60. 1-61, 62-63
Menodus giganteus (Oligocene) 62-70
Sthenodectes incisivus (Eocene) 63-65
Manteoceras manteoceras (Eocene) 63-68
Allops marshi (Oligocene) 64-69
Brontops brachycephalus, $ (Oligocene) 64
Brontotherium leidyi (Oligocene) 66
Menodus giganteus (Oligocene) 69. 9, 70
Brachycephalic :
Subbrachycephalic —
Allops serotinus (Oligocene) 72
Brontops brachycephalus, c? (Oligocene) 72
Menops varians (Oligocene) 73
Brontotherium curtum (Oligocene) 74
Palaeosyops leidyi (Eocene) 74-77
Limnohyops laticeps (Eocene) 75
Brachycephalic —
Megacerops? coloradensis (Oligocene) 76
Brontotherium curtum (Oligocene) 78
Allops crassicornis (Oligocene) 76
Palaeosyops major (Eocene) 77
Allops serotinus (Oligocene) 78
Brontops dispar (Oligocene) 78-79
Brontotherium platyceras (Oligocene) 80
Brontotherium peltoceras (Oligocene) 80
Brontotherium curtum, 9 (Oligocene) - 80
Brontotherium gigas (Oligocene) 82, 84
Brontops validus (Oligocene) 83
Megacerops acer (Oligocene) 84
Diploclonus tyleri (Oligocene) 85
Hyperbrachycephalic —
Brontotherium gigas (Oligocene) 87?
Brontops dispar (Oligocene) 87
Brontops robustus (Oligocene) 87?
Diploclonus amplus (Oligocene) 91?
DIFFERENCES IN TERMINOLOGY OF SKULL PROPORTIONS
IN TITANOTHERES AND MAN
In the Oligocene genus Menodus the zygomatic-
cephalic index rises above 60, and the skuU is there-
fore "mesaticephalic," as defined above, rather than
"dolichocephalic," as described in the author's earlier
papers. The term "stenocephalic," meaning narrow
headed, may therefore be used to describe the narrow
cranium and dentition of Menodus, in contrast with
the wide and truly brachycephalic cranium and denti-
tion of Brontops, Megacerops, and Brontotherium.
According to the indices adopted for all perisso-
dactyls, no Oligocene titanotheres are truly dolicho-
cephalic; all are mesaticephalic or brachycephalic as
compared with many Eocene titanotheres. The term
dolichocephalic, as used in the section on the Oligocene,
may be considered equivalent to "stenocephalic" (see
Chap. V) as applied to Oligocene titanotheres having
relatively narrow face and teeth but a zygomatic
index of 64-70.
The terms given above are not used in the same sense
as in anthropology. The anthropologists for cranial
form should have introduced the terms "dolicho-
cranial" and "brachycranial," but as a matter of fact
they used "brachycephalic" and "dolichocephalic."
There is no other word left for craniometry, because
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHEKES
261
there is no other Greek word descriptive of the head
as a whole. We can not use "dolichocranial," because
our indices measure the whole skull, not the calrarium
alone.
The standard zygomatic-cephalic indices adopted
in this monograph are as follows:
Dolichocephalic 39-60
Mesaticephalic 60. 1-70
Brachycephalic 70. 1-91 +
It is true that in the top view of Brontoiherium
platyceras the skull top itself is long, but the indices
show that the head as a whole is extremely brachy-
cephalic. In uncrushed skulls of Bro7itotherium gigas
the index is 82-S7, which is doubtfully exceeded only
by B. rohustus and Diploclonus amplus.
CONTRAST IN FEATURES OF BRACHYCEPHALIC AND
DOLICHOCEPHALIC SKULLS AND TEETH
The later brontotheres are excessively brachyopic,
not only in measurements but in all the characters of
the teeth, including molars and premolars. This
brachyopy, no doubt, supervened upon an earlier
stage in which the middle portion of the cranium was
elongate, and it is the elongation of the middle part
of the cranium that gives a dolichocephalic tendency.
Far from being elongate as viewed from below, the
palatal and basicranial regions of brontotheres are
excessively wide and short, as well as the zygomata
and the face, and thus contrast very strongly with the
narrow face and unexpanded zygomata of Menodus.
Hence there is no available substitute for the term
brachycephaly for the brontotheres.
Contrasts in features oj hracJiycepJialic and dolichocepTialic teeth and slculls
Brachycephalic type
DolichocephaUc type
Teeth:
Grinding series
Diastema between cutting and grinding
Anterior premolars, p \
Intermediate tubercles or conules of
molars.
Opposite cutting and grinding series
Incisor series
Canine teeth
Grinding teeth, or molars
Cingulum between molar teeth
Skull:
Entire skull
Most of the constituent bones
Palate
Nasals
Abbreviated.
Closed
Suppressed, or one fang suppressed.
Persistent
Bridge over infraorbital foramen.
Infraorbital foramen
Lacrimal bone
Lacrimal foramen
Zygomatic arches
Areas of insertion of masseteric and
temporal muscles.
Mastoid portion of periotic
Exoccipital, postglenoid and post-tympanic
processes.
Postglenoid and post-tympanic processes. .
Tympanic bulla
Foramen ovale and foramen lacerum
medius.
Alisphenoid canal
Presphenoid
Vomer
Premaxillary symphysis
Frontonasal horns
Converging or arched. .
Placed transversely
Rounded or broadened.
Shortened and widened.
Suppressed
Shortened and broadened .
do
Broadened and flattened..
Shortened and spreading.
Narrowed
Not seen on side of face
Crowded toward orbit
Crowded into orbit
Broadened, especially in the buccal
plates; in section broad rather than
deep.
Increased in thickness
Exposure abbreviated or covered.
Broadened
Jaw:
Ramus of jaw.
Area of insertion of temporal muscle.
Coronoid process
Mandibular symphysis
Approximated, especially below, in-
closing the external auditory meatus
inferiorly.
Thrust inward
Approximated
Abbreviated
do
Thrust backward
Abbreviated and massive
Transversely expanded
Shortened, thickened, deepened
Reduced
do
Abbreviated and massive
Elongated.
Open.
Persistent and spaced.
Reduced or aborted.
Parallel or elongated.
Convergent anteriorly.
Elongated or compressed.
Lengthened and narrowed.
Persistent.
Lengthened and narrowed.
Do.
Narrowed and transversely arched.
Elongated with curved and straight
borders.
Broadened.
Conspicuous on side of face.
Exposed on side of face.
Seen on edge of orbit.
Elongated and vertically deepened; in
section deep rather than broad.
Elongated horizontally.
Expanded and exposed.-
Deepened and narrowed.
External auditory meatus not closed
inferiorly.
Exposed laterally.
Separated by a bridge o i bone.
Elongated.
Do.
Not thrust backward.
Elongated.
Less expanded transversely.
Elongated with straight lower borders
and backward produced angle.
Balance maintained.
Lengthened anteroposteriorly.
Elongated.
262
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Comparison of the auditory region in the skull of
these seven divergent Eocene species shows that the
auditory meatus tends to close in the brachycephalic
types, is moderately open in the mesaticephahc types,
and is widely open in the dolichocephalic types of
titanotheres, and thus parallels the auditory region of
some modern perissodactyls, as shown in Figure 379.
The chief aUometric characters that distinguish the
Eocene titanotheres are the following:
1. Abbreviation of the face (brachyopy) and elongation of the
cranium (dolichocrany), combined with general brachycephaly
or with general doliohooephaly.
2. Flexure of the face upon the cranium ( = cyptocephaly,
upward or downward bending).
3. Broadening and elongating of the nasals or narrowing and
recession of the anterior narial openings.
4. Pneumaticity, development of great pneumatic cavities in
the upper part of the face and cranium.
We observe a long series of modifications of all the
great adaptive functions in the evolution of the
mammalian skull — namely, (1) prehension of food,
(2) mastication of food, (3) passage of food to the
fauces, (4) channels of respiration, (5) lodgment of
sense organs, (6) lodgment of the brain, (7) offensive
use of the canine teeth, (8) offensive use of the horns.
It would appear that the two functions last indicated
(7, 8) exert little influence in the middle Eocene
titanotheres. In wide contrast are the latest Eocene
and the Oligocene titanotheres, in which the horns
predominate. The canines vary greatly in size and in
the telmatheres become effective offensive weapons.
The prehensile functions of the lips and anterior teeth
vary with the development of a broad muzzle, the
firmly united premaxUlae, the greater or less recession
of the nasal bones ; but there is little evidence of strong
development of prehensile powers in the upper lip such
as is seen in the skull of the tapirs.
nST OF ABBREVIATIONS USED IN ILLUSTRATIONS
OF SKULLS
The following abbreviations are used in this mono-
graph in the illustrations of skulls. Names of com-
plete bones are begun with capital letters; names of
parts of bones with small letters.
As. Alisphenoideum.
Bo. Basioccipitale.
Bs. Basisphenoideum.
c. As (car. ex. mx.). canalis alisphenoideus, carotis externa,
ramus maxiUaris.
c. i. o. (V2 car. ex.). canalis infraorbitalis, nervus maxillaris,
carotis externa,
cond. ac. condylus occipitalis accessorius.
cond. condylus occipitalis,
cr. lamb, crista lambdoidea.
cr. sag. crista sagittalis.
or. tem. crista temporalis,
dct. la. ductus nasolacrymalis.
em. ar. eminentia articularis.
em. Fr. eminentia frontalis.
Ex. o. Os exoccipitale.
f. 0. (XII). foramen condylare, nervus XII.
f. la. foramen lacrimale.
f. 1. a. (Ill, IV, V, VI). foramen lacerum anterius, nervi III,
IV, V, VI.
f. 1. m. (car. in.), foramen lacerum medium, carotis interna,
f. 1. p. (IX, X, XI). foramen lacerum posterius, nervi IX,
X, XI.
f. mg. foramen magnum,
f. mn. (V3). foramen mentale, nervus V3.
f. ms. (jug.), foramen mastoideum.
f. op. (II). foramen opticum, nervus II.
f. ov. (V3). foramen ovale, nervus V3.
f. pi. a. (N. nas. pi.), foramen palatinum anterius, nervus
nasopalatinus.
f. p. gl. foramen postglenoideum.
f. pi. p. (N. pi. post.), foramen palatinum posterius, nervus
palatinus posterior,
f. r. (V2). foramen rotundum, nervus Vj.
f. sph. pi. (V2). foramen sphenopalatinum.
f. stm. (VII). foramen stylomastoideum, nervus VII.
f. su. or. foramen supraorbitale.
f. ven. foramen venosum.
fis. nar. ant. fissura narialis anterior,
fos. gl. fossa glenoidea.
fos. interpa. fossa interparietale.
fos. men. in. fossa menisci interna,
fos. nar. ant. fossa narialis anterior,
fos. nar. post, fossa narialis posterior,
fos. st. hy. fossa ossis stylohyoidei.
Fr. Os frontale.
HI Umbo cornu ("horn").
I. P. Os interparietale.
La. Os lacrimale.
lig. nu. ligamentum nuchae.
m. a. e. meatus acusticus externus.
m. obi. cap. sup. musculus obliquus capitis'
superior.
m. rect. cap. lat. musculus rectus capitis later-
alis.
m. rect. cap. post, musculus rectus capitis
posticus.
Ms. mastoideum.
ms. Per. pars mastoidea, ossis periotici.
Mx. maxiUa.
Mx. (alv.) maxilla (processus alveolaris).
Mx. tb. maxilloturbinale.
Na. Os nasalis.
nar. post, naris posterior.
obi. cap. sup. musculus obliquus capitis superior.
Orb. Orbis.
Pa. Os parietale.
petr. pars petrosa ossis periotici.
p. gl. Sq. Processus postglenoideus ossis squamosi.
PL Os palatinum.
Pmx. Premaxilla.
p. o. Ex. Processus paroecipitalis ossis exoccipitalis.
p. o. Fr. Processus postorbitalis ossis frontalis.
p. o. Mai. Processus postorbitalis ossis malaris.
pr. cor. processus coronoideus.
pr. i. o. Mai. processus infraorbitalis ossis malaris.
pr. p. o. processus postorbitalis.
Psph. Os presphenoideum.
Pt. Os pterygoideum.
pt. As. Processus pterygoideum ossis alisphenoidei.
p. ty. Sq. processus post-tympanicus ossis squamosi.
rect. cap. musculus rectus capitis (antice) .
rect. cap. post, musculus rectus capitis posticus.
sin. lat. eth. sinus lateralis ethmoidalis.
So. Os supraoccipitale.
Sq. Os squamosum.
(attachments).
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
263
t. la. tuberculum ossis lacrimalis.
XII. foramen condylare (nervus XII).
z. Sq. processus zygomaticus ossis squamosi.
7i. Mx. processus zygomaticus maxillae.
TERMINOLOGY OF THE UPPER MOIAR TEETH
The accompanying table shows the terms used to
designate the upper molar teeth of ungulates:
Comparative terminology of the superior molar teeth
All ungulates: Terminology
based upon evolution from
a tritubercular, bunodont
ancestral molar type. Os-
born (1888, 1892)
Primary molar cones:
Protocone
Paracone-.
Metacone.
Hypocone.
Intermediate molar
cones:
Protoconule
Metaconule
Premolar cones " :
Protocone
Deuterocone
Tritocone
Tetartocone
Secondary pillars or
styles:
Parastyle
Mesostyle.
Metastyle..
Hypostyle.
Secondary crests:
Ectoloph
Protoloph-
Metaloph_
Valleys :
Medisinus_.
Postsinus
Secondary folds:
Crochet
Antecrochet-
Crista
Secondary pits:
Pre-, medi-, and
postfossettes.
Gingulum
Titanotberes: Leidy
(1873), Cope (1883),
Marsb (1877)
Antero-external
lobe.
Postero-external
lobe.
Postero-internal
lobe.
■' Tubercles "-
Antero-internal
be.
Buttress
Median fold
Basal ridge.
Horses: Huxley (1876)
and Lydekker (1886,
p. 67)
Anterior pillar
Anterior crescent.
Posterior crescent
Posterior pillar
Anterior ridge.
Middle ridge
Posterior ridge..
Posterior promi-
nence.
Rhinoceroses: Cuvier (1836), De
Blainville (1846), Gaudry (1878),
Pavlow (1892)
Dentioule interne du pre-
mier lobe.
Dentticule externe du pre-
mier lobe.
Denticule externe du second
lobe.
Denticule interne du second
lobe.
Crete externe.
Colline seconde = Crete ou
lobe antSrieur.
La troisieme colline = crete
ou lobe post^rieur.
Vallon oblique
Fossette post^rieure.
Crochet.
Crochet ant^rieur.
Antecrochet
Fossette post&ieure=fos-
settes.
Bourrelet
Rhinoceroses, English au-
thors: Boyd-Dawkins
(1867), Busk (1877) and
Lydekker (1882), Joote
(1874)
Second costa.
Costae (in part)
First costa (buttress) .
Posterior collis (in
part) .
External lamina (dor-
sum).
Anterior collis
Median coUis.
Anterior valley..
Posterior valley.
Posterior combing
plate (uncus, cro-
chet) .
Antecrochet
Anterior combing
plate = crista.
Posterior collis (in
part) = cingulum,
guard.
Rhinoceroses and
ungulates, Ger-
man and Russian
authors: Riiti-
meyer (1863) and
Kowalevsky (1873)
Innenpfeiler
des Vorjochs.
Innenpfeiler
des Nach-
jochs.
Pericones,
Randgipfeln.
Aussenwand.
Vorjoch.
Nachjoch.
Wulst.
■ Premolar cusp; term proposed by W. B. Scott (1891).
264
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
SECTION 2. INTRODUCTION TO THE ANATOMY OF
THE SKULL AND TEETH OF THE EOCENE TITANO-
THERES
TYPES OF SKULL OF EOCENE TITANOTHERES
For reasons that are fully set forth in Chapter VIII,
on the origm and descent of the titanotheres, we regard
the skull of Eotitanops horealis (figs. 250, 251), from
the lower Eocene, as the ancestral type from which
all the highly modified Eocene skulls were derived.
The structure of the middle Eocene skulls is corre-
lated with certain feeding habits and exhibits a marked
contrast to that of the Ohgocene skulls. In middle
Eocene time the horns had not yet become weapons of.
offense and defense.
The forms of the skulls of the following Eocene
titanotheres are noteworthy:
1. LimnoJiyops priscus, a primitive hornless titano-
there, had a moderately brachycephalic skull and
primitive low-crowned grinding teeth.
Figure 218. — Contrasting forms of upper teeth in Eocene titanotheres
Brachycephalic (A) and dolichocephalic (B) types of upper premolar-molar series. One-half natural
size. A, Pa!(icos!;op5Zei(iv!,Am.Mus.l544(t5'pe); B, DoUcliorhinusliyogmthus, Am. Mas. ISbl.
2. Palaeosyops leidyi (figs. 275-278) was entirely
hornless and represents the extreme brachycephaUc
and brachyodont type.
3. The skull of TelmatJierium ultimum (figs. 294-296)
may be regarded as an elongated or mesaticephalic
modification of the primitive Limnohyops type. The
horn rudiments are retarded, and the crowns of the
teeth are more elongated than in Palaeosyops. In
many respects this skull resembles that of the suc-
ceeding type, Manteoceras.
4. The skull of Manteoceras manteoceras (figs. SOS-
SOS) differs from that of TelmatJierium ultimum in the
vigorous development of the very precocious horn rudi-
ments, which are seen in profile above and in front of
the eyes. It is also mesaticephaUc, and the molars
are more brachyodont than those of TelmatJierium.
5. The skull of MesatirJiinus petersoni has passed
from mesaticephaly into dolichocephaly. It resembles
an elongated skull of Manteoceras in having similar
rudiments of horns above the eyes, and the conforma-
tion of the zygomatic arch is similar to that of Manteo-
ceras and very distinct from that of Palaeosyops and
TelmatJierium.
6. The skuU of DolicJiorJiinus JiyognatJius (figs. 347-
349) is a decidedly long-headed derivative of Mesati-
rJiinus petersoni (figs. 327-329). The horn rudiments
are much more prominent and show some progressive
characters, such as the flattening of the top of the
cranium, which is analogous even to the cranium of
the Oligocene titanotheres. It is also decidedly
cyptocephalic, the face being strongly bent down on
the cranium. This is perhaps a river-loving type.
7. The skull of MetarJiinus earlei (fig. 361) presents
a striking contrast to that of MesatirJiinus petersoni.
It is less dolichocephalic and shows a marked recession
of the narial openings and very prominent orbits,
indicative, perhaps, of semiaquatic habits.
FEEDING HABITS OF BROAD-HEADED AND LONG-HEADED
TITANOTHERES
Peculiar forms of tJie teetJi. — Tlue mode of
feeding and the food of the titanotheres can
not be inferred with certainty, because their
dentition differs considerably from that of
any modern mammal. In middle Eocene
titanotheres the grinding teeth were per-
fectly adapted to a combination of cutting
and crushing the food, as noted below.
This adaptation implies a choice of succu-
lent food consisting of relatively coarse
leaves, grasses, buds, twigs, roots, and
tubers such as would be found in forest
and stream habitats. In later Eocene and
Ohgocene titanotheres the shearing action
of the teeth was more perfect and the food
may have included smaller objects of tougher
fiber.
Although the structure of the grinding teeth of the
titanotheres is very different from that of the grinding
teeth of members of related families — the tapirs,
rhinoceroses, and horses — the titanotheres neverthe-
less present certain analogies in the form of the head,
from which we may infer that analogies existed also
in the feeding habits.
Again, a survey of the feeding habits of the existing
Perissodactyla reveals a certain family likeness running
throughout the families of this order, which was prob-
ably manifest also among the extinct Perissodactyla.
Primitive types. — The primitive form of head and
tooth of Eotitanops is analogous to that of the primitive
paleotheres and horses, in which the proportions of
the cranium and face and the structure of the grinding
teeth are again similar. We may infer that all these
animals had a marked similarity of diet, from which
the Eocene titanotheres diverged in two directions,
developing into the extremely brachycephalic Palaeo-
syops and into the extremely dolichocephalic Doli-
cJiorJiinus.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
265
Brachycephalic types. — In the short-skuUed Palaeo-
syops we observe heavy canine tusks, large canini-
form outer incisor teeth, deep and heavy zygomatic
arches and lower jaw, high and relatively thin sagittal
for plucking and tearing up succulent bulbs, tubers,
and roots from the ground as well as for browsing on
twigs and leaves, a diet much in favor with the Amer-
ican tapir. But it differed from the tapir in that the
£o tita n ops
' A^anteocercLS
Figure 219. — Skulls of Eocene titanotheres of the principal genera
Side views. One-eighth natural size. A, Eotitanops borealis, lower Eocene, Wind River formation; B, Limnoliyops prisms, middle Eocene,
Bridger formation, horizon Bridger B; C, Palaeosyops leidyi, middle Eocene, Bridger formation, horizon Bridger D; D, Telmatherium
uHimum, upper Eocene, Uinta formation (Uinta C); E, Manteoceras manieoceras, middle Eocene, Bridger formation, horizon Bridger
D; F, Mesatirhinus petersoni, middle Eocene, Bridger formation, horizon Bridger D; G, Meiarhinus earlei, upper Eocene of Uinta
Basin, level Uinta B 1; H, DolichorUnus hyognathus, upper Eocene of Uinta Basin, level Uinta B 2. H, Horn.
crest, and large areas of attachment for the temporal
and masseter muscles. This indicates a notably ver-
tical movement of the jaw and great power in crush-
ing the food. Such an animal would seem well fitted
101959— 29— VOL 1 20
nasals are not retracted, and there is no evidence that
the upper lip had exceptional prehensile power. (See
fig. 220.) This titanothere presented the extreme of
the browsing type. It had a lumbering gait and
266
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
spreading feet, and to judge from the associated fauna
in comparison with that of the Uving tapirs it inhab-
ited semitropical forests, especially those near streams.
Dolichocephalic types. — The other extreme of struc-
ture among Eocene titanotheres is the long-skulled
Dolichorhinus, which succeeds Palaeosyops in geologic
time, belonging more to the upper Eocene. The
muzzle of this animal (fig. 219) was rather expanded
The cheek teeth were relatively long-crowned with
pointed cusps and constituted a relatively elaborate
cutting and triturating apparatus, as compared with
the very short-crowned grinders of Palaeosyops. The
excursion of the more slender mandible was partly
vertical, partly oblique. The oblique position of
the grinding teeth produced an oblique shearing
action. Conditioning these changes the length and
proportions of the masticating muscles
and their angles of action were also
changed. (See Chaps. V, VIII for de-
tails.
These features of the head of Dolicho-
rhinus indicate that the food of this
animal required finer cutting and better
trituration than that of Palaeosyops.
Although in no sense a grazing animal
as compared with the grazing Equidae
and Bovinae, Dolichorhinus was better
adapted to grazing than Palaeosyops.
Its remains are very frequently found in
coarse sandstones laid down by rapid
streams, and it may well have hved partly
in the rivers and along their banks.
Intermediate types. — The other Eocene
titanotheres, such as Manteoceras (fig.
220, C) and Telmatherium (fig. 220, B)
are more or less intermediate between
these extremes in the form of the head.
Thus Manteoceras has very heavy,
almost boarlike tusks and large, blunt
incisor teeth, together with cheek teeth
that are more elongate than those of
Palaeosyops. Telmatherium had much
more trenchant canine tusks, pointed
incisor teeth, and somewhat elongated
grinding teeth.
ORIGIN AND STRUCTURE OF THE
IN TITANOTHERES
'HORNS"
-Heads of Eocene titanotheres of four phyla
Modeled by Charles R. Knight. A, Palaeosyops, brachycephalic; B, Telmatherium, mesaticephalic; C'
Manteoceras, mesaticephalic; D, Dolichorhinus, dolichocephalic. The nostrils were actually more
nearly terminal than those shown in the models, and the upper lip may have been more markedly
pointed or prehensile. H, Horn rudiments.
or truncate. The incisors were arranged in a semi-
circle and made some approach to the cropping in-
cisors of a ruminant, being also partly cupped as in
Oligocene species of the horse. The space behind
the canine tusk was longer, as in typical herbivorous
forms. The canines were recurved, compressed, or
sharp-edged and may have been used in fighting,
as in the camels. The offensive power of the front
teeth was, however, much less than in Palaeosyops.
The so-called horns of titanotheres
arise as rectigradations; they consist of
osseous protuberances of the skull above
the eyes, where the frontals overlap
the nasal bones. In life they were prob-
ably covered with tough sldn, rather than with horn.
In the earliest titanotheres, of lower Eocene age {Lamb-
dotherium, Eotitanops), the frontonasal junction shows
no beginning of the horns. In the genera Palaeosyops
and Limnohyops (middle Eocene) most of the skulls
were equally hornless, but some very old males of
Palaeosyops show an incipient nasofrontal protuber-
ance and roughening of the outer tabula of the bone.
(See PI. XVI.) In the middle Eocene contemporary
D. S. GEOLOGICAL SURVEY
MONOGUAPH 55 PLATE XVI
r
^
■^
B
\
THE REGION OF THE HORN SWELLING IN PALAEOSYOPS, MANTEOCERAS, AND TELMATHERIUM
A', Palaeos^iops robuiiui (Am. Mus. 1554), superior view; A', the same, vertical longitudinal secftion. B, Manteoceroj manteoceras (Am. Mus. 1569).
C, Telmatherium ultimum (Am. Mus. 2004). fr, Frontals; h, rudimentary horn; mx, maxillary; na, nasal. All natural sise
U. S. GEOLOGICAL StrKVET
MONOGRAPH 55 PLATE XVII
THE REGION OF THE HORN SWELLING IN MANTEOCERAS, MESATIRHINUS, AND DOLICHORHINUS
Manteoceras manteoceras (Am. Mus. 23S3), lateral view, left side (right side reversed). B, DoUchorhinus hyognathus (Am.
Mus. 1851), lateral view, left side (right side reversed). C, Mesatirhinus petersoni (Am. Mus. 1556), lateral view, left side.
D, DoUchorhinus hyognathus (Am. Mus. 1851), superior view. c. i. o. /., c. i. o., Infraorbital foramen; /r, frontal; h, rudi-
mentary horn; la, lacrimal; mo, molar; mx, maxillary; na, nasal; or, orbit; s. nar., external narial aperture
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
267
Telmatherium cultridens there appears to have been
a distinct nasofrontal protuberance, but in the geologi-
cally later Telmatherium ultimum only the faintest
indication of its presence is found; it is possible that
the horn retrogressed in this phylum. In Manteo-
ceras manfeoceras of the upper part of the Bridger
formation (middle Eocene) the protuberance, although
small, is perfectly distinct and fully characteristic in
form. In ProtitanotJierium of the upper Eocene the
horns (figs. 317-319) consist of oval protuberances
about 20 millimeters high and 90 millimeters long.
In nearly all the lower Oligocene titanotheres the
horns are of large size and finally become the domi-
nant feature of the whole skull, affording generic and
specific characters.
The horns are believed to have evolved concomi-
tantly with the fighting habits of these animals and
with the general increase in size and body. The
conditions of the horns in the titanotheres may be
summarized as follows:
Summary of character or condition of the horns in Eocene and Oligocene titanotheres
Subfamily or genus
Character or condition of horns
Lambdotheriinae _
Eotitanopinae
Palaeosyopinae, . .
Lambdotherium zone (Wind River B) .
do
Telmatheriinae _
Sthenodectes
Manteoceratinae
Dolichorhininae :
Mesatirhinus-Doliohorhinus_
Metarhinus
Rhadinorhininae :
Rhadinorhinus
Manteoceratinae :
Protitanotherium _
Eotitanotherium. _
Brontopinae
Lower horizons of Bridger Basin to lower
horizons of Washakie Basin, inclusive.
Upper horizon of Bridger Basin to Uinta
C, inclusive.
Uinta B 1 only
Upper horizons of Bridger Basin to Uinta C.
Upper horizons of Bridger Basin to Uinta
B 2.
Lower horizons of Washakie Basin to
Uinta B 1, inclusive.
Uinta B 1 only_
Uinta C onlv_-
Uinta B 2
Chadron A to C, inclusive.
Menodontinae. .
Megaceropinae.
.do.
Brontotheriinae.
.do_
_do_
Hornless.
Do.
Hornless, or nasofrontal horn swelling barely
perceptible.
Do.
Horn swelling small.
Horn swelling small but distinct.
Horn swelling more pronounced; on nasals
only.
Horn swelling small; chiefly on nasals.
Horn swelling small.
Nasofrontal horn swelling pronounced and
progressive.
Do.
Nasofrontal horn swelling at first small,
slowly Isecoming progressively larger.
Do.
Nasofrontal horn swelling of medium to
large size.
Nasofrontal horn swelling at first small,
rapidly becoming progressively larger.
PROPORTION AND RECTIGRADATION IN THE GRINDING
TEETH OF EOCENE TITANOTHERES
The chief characters of the grinding teeth in the
Eocene titanotheres were evolved from the bunose-
lenodont pattern (see fig. 221) and were modified by
changes of proportion and rectigradation, under eight
principles, as follows :
1. The primitive grinders seen in Lambdotherium
and Eotitanops are extremely low crowned, or brachyo-
dont. The numerous phyla can be distinguished
chiefly by the different degrees and rates of elonga-
tion of the crown, which shows progressive hypso-
dontism.
2. The sLx main grinding teeth in the upper and
lower jaws, p^-m^, p^-nis, are closely crowded together,
and this crowding causes the crowns of the grinders to
be closely proportioned to the brachycephaly or
dolichocephaly of the skull. In brachycephalic titano-
theres the transverse diameters of the grinding
teeth generally exceed the anteroposterior diameters,
whereas in dolichocephalic skulls the reverse is true.
Thus we shall speak of the grinders as of the "brachy-
cephalic" or of the "dolichocephalic" type.
3. The general tendency of the grinders in titano-
theres is to become macrodont, because the pattern
of the grinding teeth is mechanically imperfect, and
the grinders compensate in size, in some degree, for
what they lack in mechanical perfection,
4. The transformation of the "cone and crescent"
or bunoselenodont pattern of the upper and lower
grinding teeth in the titanotheres can be best under-
stood by comparing that pattern with that seen in the
other bunoselenodonts — the primitive chalicotheres,
horses, and paleotheres — a pattern similarly derived
from the same primitive type of upper and lower
grinding teeth (fig. 222), which presents four main
cones above and four main cones below, known as the
"primary molar cones."
268
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
5. The secondary accessory folds, known as "styles,"
or pillars, and "lophs," or crests, as well as the
"fossettes," or pits, may also be homologized by
comparing the superior and inferior molars of the
ECTOLOPH
MESOSTYLE''
PARASTYLE
PROTOCONULE-
-METACONULE
-HYPOCONE
PROTOCONW
PROTOLOPHID
PARACONID-
-HYPOCONID
METALOPHW
ENTOCONID
METACONID \ METACRJSTID
METASTYUD
FiGUEE 221. — Upper (A) and lower (B) molars of bunosele-
nodont pattern
Molars of Lambdotherium, a lower Eocene titanothere. Enlarged. The worn
enamel surface is cross hatched; the exposed dentine is shown in dense black.
titanothere (TelmatJierium) and of equine forms
{AncMtherium) with those of the rhinoceros.
6. Arrested or retro-
gressive development is
the chief characteristic
of the titanothere molar
evolution- — that is,
parts are arrested or
vestigial in titanotheres
that evolve rapidly and
strongly in paleotheres,
chalicotheres, and
equine s. Thus the
titanothere molar be-
gins its evolution in the
form of the Lamhdothe-
rium molar (figs. 221,
235) or of the Eofitanops
molar (figs. 229, 249), in which 32 primary and
secondary elements may be more or less clearly dis-
tinguished in the typical upper and lower grinders.
In the course of evolution these grinders, through
arrested development, lose six or more of these ele-
ments. Thus the grinding teeth are impoverished as
compared with those of the other bimoselenodonts
(fig. 223).
The parts that gradually become vestigial or dis-
appear in titanotheres are the following:
Protoconule, anterior intermediate cusp of superior molars;
degenerates.
Metaconule, posterior intermediate cusp of superior molars;
degenerates.
Protoloph, anterior transverse crest of superior molars, formed
of protocone, protoconule, paracone (inner base) ; disappears.
Metaloph, posterior transverse crest of superior molars,
formed of hypocone, metaconule, metaoone (inner base) ; dis-
appears.
Paraconid, antero-internal cusp of infe-
rior molars, reduced or vestigial in all Peris-
sodactyla; disappears.
Hypoconulid, posteromedian cusp of in-
ferior molars; abortive except in third
inferior molar.
7. All stages in the reduction and
disappearance of these six or more
elements in the upper and lower
grinding teeth are observed among the
Eocene titanotheres (Pis. LIV-LXV),
whereas the lower Oligocene titano-
theres exhibit grinding teeth (fig. 381)
in which all these parts have totally
disappeared and certain new secondary
rectigradations, such as the "fossette"
and crochet, have appeared.
8. The appearance or disappearance
of these single elements is generally
gradual or continuous; yet it is much
more rapid in certain phyla than in others. The
variation in the rate of degeneration distinguishes
the phyla from one another and thus becomes a char-
B
Figure 2 2 2. —
Upper (A) and
lower (B) molar
patterns of
Hyracotherium,
a primitive
Eocene equine
perissodactyl
(ancestor o f
the horse)
Enlarged.
a
Figure 223. — Bunoselenodont patterns of upper and lower molars in Tertiary perissodactyls
Telmatherium vltimum, an upper Eocene titanothere, upper molar; a, Telmatherium cuUridens, a middle Eocene titanothere, lower
molar; B, b, Moropus sp.
Miocene hippoid.
a Miocene chalicothere; C, c, Palaeotheriuvi sp., an Eocene paleothere; D, d, Atichitherium sp., a
acter of generic value. The numerical gain or loss of
one of these elements is of specific value and marks off
the subspecific stages or mutations.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
269
MECHANISM OF THE TITANOTHERE GRINDING TEETH
The pattern of the upper and lower grinding teeth
of the titanotheres is one that has entirely disappeared
among the existing mammals. It has no counterpart
among any living ungulates, but it is closely analogous
PARAGON e:
FOSSA METACONE
FOSSA
Jvyd
772-2
METACONID ENTOCONID
FOSSA FOSSA
Figure 224. — Relations of upper and lower molars in an Eocene
titanothere, Telmatherium cultridens
Princeton Mus. 10027 (type). Two-thirds natural size. A, Second upper molar
showing the crushing parts and the cutting parts, also the pits or fossae for the
reception of the projections on the lower teeth; B, second lower molar, showing
the fossae for the parts of the upper teeth; C, diagram showing how each lower
molar articulates with two upper molars, the trigonid of ma wedging between mi
and m^, and the talonid of ma receiving the protocone of m^.
to that of many Eocene and Oligocene mam-
mals, both artiodactyls and perissodactyls.
There is no perfected grinding function, such
as that between the upper and lower molars
of the horse, nor could such a grinding function
evolve out of the titanothere molar tooth.
The dental mechanism is a combination
of two functions which may be described as
crushing and cutting, the crushing being
effected by a double pestle and mortar or peg
in socket mechanism, and the cutting by a
mechanism of double shears or reversed double-
bladed crescents.
In the accompanying diagram (fig. 224) the
manner in which this double function is sub-
served in the upper and lower grinders is
clearly shown.
(1) The superior molars consist of a double-
cutting shear W, composed of the paracone and
metacone crescents, which oppose the reversed
double shear M, composed of the crescents developed
from the protoconid and hypoconid. (2) The chief
crushing action is performed by the superior cones, the
protocone and hypocone, which fit into the protocone
and hypocone fossae of the inferior molars. (3) The
apex (O) of each of the four primary superior cones (pro-
tocone, paracone, metacone, hypocone) has its contact
or abrasion point (oblique shading in figure), or fossa,
in the inferior molars. (4) Similarly the apex of each
of the four inferior .primary cones (protoconid, meta-
conid, entoconid, hypoconid) has its contact or ab-
rasion point (oblique shading), or fossa, in the superior
molars.
This double cutting and crushing function is me-
chanically imperfect in the short-crowned molars of
Lambdotherium and Eotitanops (figs. 229, 235, 242,
253). It becomes more efficient as the crowns become
higher and the cones and crescents are vertically
elongated in Telmatherium (fig. 225). The evolution
of the titanothere grinders is directed to overcome
the deficiency of this cone and crescent mechanism,
which proves to be inherently defective in design.
The crushing function of the grinders is best ob-
served in the internal view (fig. 226) of the upper
and lower grinders of a telmathere, in which the
pestles (p-h) are sinldng into the mortars (m-e) in
exactly the same manner as in the primitive insecti-
vores. This closely correlated mechanism of the
upper and lower grinding teeth, which was first
studied by Cope (1889.3) and more fully by Gregory
(1916.1), indicates that every new character (recti-
gradation, allometron) added to the upper grinders
must be correlated with a new and mechanically
adaptive character (rectigradation, allometron) in
the lower grinders. The cutting function performed
by the W of the upper ectoloph and the M of the
lower ectoloph is illustrated (figs. 224, 225) and dis-
FlGTTRB 225
-Dental mechanism of titanotheres
Interlocking relations of upper and lower premolar-molar series. One-half natural size. A, Telma-
therium cultridens, Princeton Mus. 10027 (type); upper teeth (light line), with pattern of lower
teeth (heavy line) projected upon them. Crown view. B, Internal view of the same teeth,
showing the crushing action of the cones and conids.
plays the close mechanical relation of the alternating
crests as well as the simultaneous development of
the new cusps (rectigradations) of the premolar
teeth.
270
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The partial transformation in the titanotheres of
a more vertical chopping, crushing, and cutting
motion of the jaw into a more oblique sweeping
or true grinding action of the molars finally results
in the entire loss of the conules and transverse crests
(protoloph, metaloph), in the partial molarization of
the premolars, and in the development of two gigantic
internal crushing cones (protocone, hypocone) and
of a very powerful external cutting crest (ectoloph).
2. The excursion of the mandible was made from the
outer side upward and inward, as in rhinoceroses and
horses, in contrast with the opposite motion in ruminants.
3. In Eocene titanotheres the lack of a tetartocone
in the upper premolars and of an entoconid in the
lower premolars leaves an open space when the jaws
are shut. This open space is filled in the Oligocene
titanotheres by the opposing tetartocone and entoconid.
MOLARIZATION OF THE PREMOLARS
The titanotheres resemble all the other fami-
lies of Perissodactyla in the gradual molariza-
tion of the premolar teeth — that is, in the
Figure 226. — Dental mechanism: Grinding teeth of a titanothere (A)
and an insectivore (B)
After W. K. Gregory. Internal view of the opposed upper and lower grinding teeth of
Telmatherium cuUridens (A), natural size, and Erinaccus (B), much enlarged. The pro-
tocones (p) fit into the talonid basins (fossae) between the metaconids (m) and entoconids
(e). The hypocones (ft) fit into the trigonid basins (fossae) between the entoconids (e) and
metaconids (m) . Similar relations are found in all primitive mammals.
Three other interesting features in the evolution of the
dental mechanism of the titanotheres are the following:
1. The marked protrusion of the roots on the outer sides
of the upper molars in old Oligocene titanotheres is a result
HYPOCONID
M ESOSTYLE
PROTOCONE
Figure 227. — Contrast of molars of a brachyodont Eocene titanothere
(A) and a semihypsodont Oligocene titanothere (B, B')
A, Third left upper molar of Palaeosj/ops leiiyi, seen from the rear. The internal and external
cones are subequal in height. B, Third left upper molar (unworn) of Menodus giganteus,
seen from the rear. The internal cones are low; the external cones have greatly increased
in height and have grown inward at the tip. B', The same seen from the outer side,
showing the much deepened ectoloph.
of the bunoselenodont pattern of the molars and
of the vertical-oblique pressure of the lower teeth.
As the outer side of the molar crowns becomes more
hypsodont, in passing from lower Eocene to Oligocene
titanotheres, so the external roots protrude more
prominently.
Figure 228. — Cross sections through second
upper and lower molars of Lambdotherium and
Menodus
A, A brachyodont lower Eocene titanothere, Lambdotherium
popoagkum, three-halves natural size; B, a semihypsodont
lower Oligocene titanothere, Menodus giganteus, one-half nat-
ural size. In A the excursion of the mandible was more trans-
verse in direction than it was in B, where, in correlation with
the deepening of the ectoloph, the movement of the mandible
was more vertical.
transformation of the premolars into the molar
pattern. The mechanical inferiority of the teeth
of the titanotheres lies in the fact that this trans-
formation is never perfected ; it is very slow or
retarded, and the premolars never completely
acquire the molar pattern, as they do in the
Equidae, for example, in which the premolars
become actually superior to the molars both in pattern
and in mechanical perfection. The arrested transfor-
mation of the premolars in the titanotheres is
undoubtedly a defect that is correlated with the
abbreviation of the facial region and with the great
increase in the relative size of the molars.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERE3
271
The primitive superior premolars of Eotitanops
(fig. 229, A) when contrasted with those of Bron-
totherium (fig. 229, B) are seen to be triangular in
outline and surmounted with three more or less
perfectly developed cusps. The transition between
these primitive and specialized extremes, by the addi-
tion of neomorphs or rectigradations, furnishes a
whole series of specific characters and mutations in
the ascending phyla of titanotheres.
All the phyla exhibit a similar orthogenic tendency to
molarization, but in some phyla this tendency is rapid
and in others it is slow. Thus the different rates of
molarization are of taxonomic value: they furnish
distinctions between different generic or phyletic series.
(See Velocity of movement. Chap. XI, pp. 810-812.)
A careful review of the premolars in lower and
middle Eocene titanotheres shows that the order of
appearance of the premolar cusps is as follows:
Lower premolars Upper premolars
L Protoconid. 1. Protocone.
2. Hypoconid. 2. Deuterocone.
3. Paraoonid. 3. Tritocone.
4. Metacristid. 4. Parastyle.
5. Metaoonid. j 5. Tetartocone.
6. Entooonid. I 6. Mesostyle.
In the above terminology of the upper premolar
cusps the names are those proposed by Scott in 1892
in his paper on the evolution of the premolar teeth
in the mammals (1892.1), but they are used through-
out this monograph simply as names, without re-
ference to the phylogenetic order of evolution, which
Figure 229. — Upper premolar-molar teeth of the earliest (A)
and latest (B) known titanotheres
Both tooth rows drawn to the same length. A, Eotitanops lorealis, Am. Mus.
14887; lower Eocene. Premolars very primitive, a wide diastema behind pm'.
B, Brontotlicrium gigas. Am. Mus. 492; lower Oligocene. The later type, showing
the molarized pattern of the premolars and the great size of the molars. Pre-
molars very advanced in pattern, diastema closed.
is given above. The names of the lower premolar
cusps are adapted from the terminology used by'
Osborn for the lower true molars, but again without
reference to the phylogenetic order of appearance.
Summary of premolar evolution of the titanothe
res
Wind River Basin,
Wyo. (Wind River B,
"Lost Cabin")
Bridger Basin, Wyo.
Washakie Basin,
Wyo.
(Washakie A, B;
B = Uinta A, B)
Uinta Basin, Utah
Titanotherium
Brldger
A, B
Bridger C, D
( = Washakie A)
Uinta A
Uinta B
Uinta C
zone (Chadron
A, B, C)
pi with one cusp, except in later
forms; p^ p', p' with three
cusps:
Primitive: evolving
slowly.
larly.
More advanced -
A little more
advanced.
Well advanced.
But little
changed.
Of high type
do
Evolving slowly.. .
Of high type
Of high t
do...
Rhadinorhininae —
Intermediate conditions:
Diplacodontinae—
Advanced
pi-p' with four cusps; p' with
three or four cusps (fourth
cusp (te) absent or connected
with third cusp (de)):
Advanced, but
changing very
little.
Do.
Menodontinae—
More advanced.
but changing
very little.
Well advanced and
P'-p* with four cusps (p^ with
fourth cusp very large and not
connected with third cusp) :
progressing mod-
erately.
Very advanced, but
changing little.
Do.
'
272
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
CORRELATION OF DIMENSIONS OF UPPER AND LOWER
TEETH
In considering the generic and specific relations of
isolated upper or lower jaws, it is frequently desirable
to estimate some of the dimensions of unknown upper
teeth from corresponding dimensions of the lower
teeth and vice versa. The following table of equa-
tions of measurements, prepared by W. K. Gregory,
was based primarily upon the type of Telmatherium
cultridens but has been verified as to other titano there s.
Dimensions in the upper row of teeth (anteroposterior measure-
ment) approximately equal to dimensions in the lower row
M '-111^ = tip paraconid mi to tip entoconid 1113.
P'-p* = hinder border pi to tip protoconid mi.
P2-p'' = tip protoconid p2 to tip protoconid mi.
P* = tip protoconid p4 to tip protoconid mi.
M' = tip protoconid mi to tip protoconid m2.
M2 = tip protoconid mj to tip protoconid ms.
M3 = tip protoconid ms to tip hypoconulid mi.
Tip metacone (tritooone) p'' to tip metastyle m^ = anterior border
mi to tip hypoconulid ms.
Tip paracone m' to tip paracone m^ = tip metaconid mi to tip
metaconid ms.
Tip parastyle m' to tip parastyle m3 = tip protoconid mi to tip
protoconid mj.
Some of these correlations have been frequently
used in the identification of Eocene and Oligocene
titanotheres, but they are equally applicable to many
other groups of mammals, because they are based
upon constant interlocking relations of the cusps in
the upper and lower molars.
GEOLOGIC SUCCESSION AND GEOGRAPHIC DISTRIBUTION
OF THE EOCENE TITANOTHERES
The following table shows the geologic succession
and geographic distribution of the Eocene titanotheres
as known in 1914:
Geologic succession and distribution of the species of Eocene titanotheres
[See flg. 41, p. 59)
s
.9
ffl
0
a
0
1
m
is
1
1
a
i3
S
B
0
.1
0
^\
X
•so
X
4(>
X
48
X
47
X
4fi
X
45
X
X
X
X
X
X
X
X
44
43
.. __ -
4?
X
41
40
3P
X
SS
S7
X
Sfi
X
X
X
X
X
X
X
X
X
X
X
S^i
S4
•^s
3?
M. earlei
X
31
30
W
98
X
?7
X
X
X
X
?6
25a
X
''^
X
X
94
X
X
X
X
X
X
X
X
X
9S
?,•>,
"^1
?0
X
X
IP
IS
P. leidyi
17
16
15
L. monoconus
X
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHBRES
Geologic succession and distrihution of the species of Eocene titanotheres — Continued
273
<
s
0
o
a
a
3
1
3
1
P
14
X
X
X
X
X
n
T>
11
in
X
X
Pi
q
X
X
X
X
X
X
X
X
X
X
8
7
R
5
4
S
?,
In,
1
1
_ .
SECTION 3. THE LOWER EOCENE TITANOTHERES
ANCESTRAL TITANOTHERES OF THE lAMBDOTHERIUM ZONE
OF WYOMING AT THE END OF LOWER EOCENE TIME
The sudden appearance of two forms of titanotheres
in the region now linown as Wyoming and Colorado
at the end of lower Eocene time in the Lambdothe-
rium zone is a very striking fact. The two forms are
LamidotJierium, relatively small, swift, with slender
limbs, very abundant (upland type); and Eotitanops,
larger, with medium-sized limbs, less abundant (low-
land type).
No trace of the ancestors of either of these animals
has thus far been found in the immediately underlying
Heptodon zone and earlier beds, although the remains
of horses (Eohippus) occur there in abundance. The
evidence favors the theory that the titanotheres mi-
grated into the ancient mountain region of North
America near the end of early Eocene time. The
beds in which they first appear belong to what is called
the Lamhdotherium zone, because of the great abun-
dance of the remains of this delicately formed titano-
there, which is found there in numbers exceeded only
by the remains of horses. In the typical Wind River
formation Granger, in his collections made from the
Lamhdotherium zone in 1905 and 1909, assembled out
of a total of 727 specimens remains of 191 Eohippus,
111 Lamhdotherium, and 14 Eotitanops.
The localities of the Lamhdotherium zone known up
to the year 1912 were as follows:
Thickness
in feet
Wyoming, Wind River Basin, Lost Cabin section 400
Wyoming, Big Horn Basin, Tatman Mountain section 325
Wyoming, Beaver Divide, Green Cove section 265
Colorado, Huerfano Basin, Garcia Canyon 400
The three sections in Wyoming are of nearly uni-
form thickness throughout. The geology of the Lamh-
dotherium zone in Colorado is described in Chapter II.
There is also considerable uniformity in the size and
character of the remains of Lamhdotherium. Most of
the remains are referred to a single species, L. popoagi-
cum, which, however, appears to split up into several
subspecies. We Icnow only one phase in the evolu-
tion of this animal. Other phases await discovery.
PHYSIOGRAPHIC ENVIRONMENT AT THE END OF LOWER
EOCENE TIME
It is especially interesting to picture the geographic,
climatic, and biotic conditions surrounding these early
titanotheres. The picture may be drawn partly from
the study of the rocks in which their remains occur
and partly from the remains of the numerous mammals
that are found with them in these Wind River deposits.
The complete geologic relations of the deposits of the
Lamhdotherium zone have been presented in Chapter
II. We comment here chiefly on the geographic and
climatic features of the period.
WIND RIVER BASIN, WYOMING
In lower Eocene time {Lamhdotherium zone) the
Wind River Basin was a broad flood-plain valley with
mountain barriers to the north, west, and southwest
and an easterly drainage. (Sinclair and Granger,
1911.1, pp. 87-103, 105.) The materials of which
the beds are composed came from these surrounding
mountains. The fine material consists of highly
colored clay, in places banded alternately red and
blue, interstratified with pale greenish-buff and yel-
low-brown sandstone in more or less continuous
274
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
lenses. The sandstone, which is composed largely of
quartz sand in which fossils are rare, appears to have
been laid down by swift-flowing streams. The de-
posits of blue clay contain layers of lignite, ranging
from mere dark bands to rather thick beds, indicating
still water and a humid climate when vegetation was
accumulating rapidly. Skeletons of mammals found
in this blue clay were evidently swept into still-water
areas and covered with river sediment; but fossils are
rare in this stratum also. In many of the bands of
red clay, on the contrary, or at the contact of the red
and blue strata, great numbers of fragments of jaws
and scattered teeth are found. Such levels probably
represent parts of the basin floor as it was when these
creatures died. The beds of red clay, according to
Loomis, Granger, and Sinclair, were formed during the
drier cycles, when the carbonaceous matter of decaying
plants was completely oxidized, when iron compounds
were concentrated and oxidized, and when the bones of
animals exposed at the surface were weathered and
broken before they were entombed. These signs of the
alternation of moist and dry climate, indicated respec-
tively by blue and red clays, are not accompanied by
signs of excessive aridity, the mammals in the red and
blue clay bands being the same. Similar alternations
of red and blue clays are now found in the desert
basins of Asia.
BIG HORN BASIN, WYOMING
The discovery of Lamhdotlierium by the Amherst
College expedition of 1904 under Loomis and its
localization by the American Museum expedition of
1911 (Sinclair and Granger, 1911.1) in the uppermost
levels of the red-banded clay beneath the lignitic
beds of Tatman Mountain demonstrated the deposi-
tion of sediments of Wind River age in the Big Horn
Basin. The true Lambdoiherium zone is exposed on
all sides of Tatman Mountain and consists chiefly
of red-banded beds. Granger and Sinclair observe
(1912.1, p. 66) that the lower Eocene sediments of the
Big Horn Basin, like those of the Wind River Basin,
represent the filling in of a great trough surrounded
by mountains. No volcanic ash occurs. The moun-
tain streams have borne down gravel, sand, and clay
and deposited them in stream channels or spread
them over flood plains. No evidence of wind trans-
portation has been observed. The red and blue
banding of the clays occurs in more or less regular
alternation.
BEAVER DIVIDE, WYOMING
The discovery of a typical Wind River fauna on
Beaver Divide by Olsen, of the American Museum
party of 1910, was a most important one, because it
extends the range of this fauna many miles to the
southwest. The entire fauna was obtained at or near
a certain stratum of bluish-green shale resting on a
band of red shale, the fossiliferous zone not exceeding
10 feet in thickness. (See Chap. II.) Remains of
the animals listed below were obtained:
Equidae:
Eohippus craspedotus.
Eohippus? venticolus.
Lophiodontidae :
Heptodon calciculus.
Heptodon ventorum.
Heptodon n. sp.
Titanotheriidae :
Lambdotherium popoagi-
cum.
Amblypoda:
Coryphodon sp.
Reptilia:
Glyptosaurus (scutes).
Crocodile (scutes, verte-
brae, and teeth).
Turtles (numerous frag-
ments) .
Insectivora:
Hyopsodus n. sp.
Hyopsodus sp.
Creodonta:
Didymictis? altidens.
Primates:
Microsyops sp.
The fish and aquatic reptiles in this fauna indicate
plainly that the deposit on Beaver Divide was fluvia-
tile, and, as Granger and Sinclair observe, go far toward
establishing the theory that the Wind River shales
were flood-plain deposits, a theory that is further
supported by the presence of numerous channel
fillings of coarse sandstone. All the fossils from the
shales are fragmentary and consist mostly of teeth
whose roots are worn off, indicative of water trans-
portation and abrasion.
HUERFANO BASIN, COLORADO
The Lambdotherium zone was discovered in Colo-
rado by Dr. J. L. Wortman while he accompanied
the writer in 1897 (Osborn, 1897.126) on a survey of
the Huerfano Eocene deposits, which were first
announced by Hills in 1888 (Hills, 1888.1). The
zoogeographic significance of this discovery is evident
from the fact that it carries the Lambdoiherium
fauna eastward to the foothills of the Rocky Moun-
tains, between the famous extinct volcanoes known as
the Spanish Peaks.
Wortman described these beds as follows:
These beds of the lower division [Lambdoiherium zone] are
indistinguishable, so far as their general appearance and litho-
logical characters are concerned, from those of the upper level
[Palaeosyops fonlinalis zone]. The fossils occur apparently
in a single stratum not exceeding 10 or 15 feet in thickness and
not more than 30 or 40 feet from the base of the formation.
They underlie the beds of the upper division with perfect
conformity, and there is at present no means of determining
exactly where the one ends and the other begins. That
sedimentation was continuous and iminterrupted from the
beginning to the close of the whole [Huerfano] deposit, I do
not think there can be the slightest question. The exact
locality from which the greater number of the fossils of the
lower beds were obtained is Garcias Caiion, about IJ-^ miles
south of Talpa or the mouth of Turkey Creek. [Osborn,
1897.126, pp. 253-254.]
The animals associated with Lambdotherium in this
zone are provisionally identified by Osborn and
Matthew as follows:
Titanotheriidae Lambdotherium popoagi-
cum.
Creodonta Didymictis altidens.
Didymictis leptomylus.
Oxyaena lupina.
Insectivora Hyopsodus sp.
Amblypoda Coryphodon ventanus.
Artiod actyla Trigonolestes secans.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
275
SUMMARY
A summary of the life conditions in the Wind River
and Big Horn Basins during lower Eocene time shows
that there is no evidence of climatic change throughout
lower Eocene time; that the conditions through the
Systemodon zone and the succeeding Heptodon zone,
into the overlying Lambdoiherium zone, remain sub-
stantially similar. This fact accords with the sub-
stantial similarity in the general character of the mam-
malian fauna throughout the lower Eocene. The
fauna evolves during this very long period; old forms
give way to new; but it does not change in its
general adaptation to conditions.
Thus in the great mountain valleys sediments were
being continually derived by erosion from the older
rocks of the mountains and deposited in these great
basins. Throughout Wind River time fluviatile dep-
osition is indicated by numerous channels filled with
coarse sandstones which irregularly traverse the finer
clays or interstratify with them in the form of lenses.
In the clays are found fish, crocodiles, and turtles, and
occasionally beds of JJnio. Local swamps are indi-
cated by the presence of lignitic areas in the blue clays
and in the sandstones, but never among the red clays.
The feldspars that wash down from the surrounding
granitic mountains are fresh and angular, a fact that
suggests rapid transportation of the fragments for
short distances and burial beyond the reach of car-
bonated waters. These conditions do not favor the
idea of luxuriant Eocene tropical forests or of a warm,
humid climate with the formation of a deeply decay-
ing humus, but suggest rather a dry although not nec-
essarily arid climate, with rapid changes of tempera-
ture favorable to splintering the ledges of the granite
cliffs. There were no frosts, but the climate may have
been stimulating to a vigorous and actively competing
fauna. In the Big Horn Basin fluviatile deposition
is indicated throughout the entire Eocene epoch.
The lignitic shales that cap the Lambdoiherium zone,
as indicated by the fresh-water moUusks and plant
contents, are both fluviatile and palustrine.
The above picture of the physiography and the cli-
mate of these Rocky Mountain basins of Wyoming in
early Eocene time accords thoroughly with the analysis
of the chief adaptive types of mammals whose re-
naains are found in the lower Eocene rocks. These
mammals are broadly divided into three types — flu-
viatile, or river-living; river border, or palustrine;
upland, or plains-living.
The upland type sought and found hard ground, to
which their narrow feet and compressed hoofs were
adapted. Conspicuous among these dry-ground forms
is Lambdoiherium itself, a swift-moving, or cursorial
animal broadly analogous in structure to the horses
of that time (Eohippus) and lophiodonts (Heptodon),
as weU as to the archaic condylarths (Phenacodus) .
The Wind River fauna of the Lambdotherium zone
represents the closing chapter of lower Eocene mam-
malian life. It is closely affiliated with the fauna of
the typical lower Eocene or Systemodon zone, because
the two contain 25 genera and 1 1 species of mammals
in common. It is doubtful whether a single family
of mammals of the Systemodon zone becomes extinct
in the Lambdotherium zone, yet some of the archaic
mammals begin to show a numerical reduction.
On the other hand, the Wind River fauna is pro-
gressive; the first appearance of these two genera of
titanotheres, Lambdotherium and Eotitanops, and of
seven other new genera of mammals is prophetic of
the oncoming middle Eocene or Bridger life.
During this transition of the Wind River mammals
from their Wasatch forerunners to their Bridger suc-
cessors the physiography and the climate apparently
remained the same as in earlier Wasatch time, a fact
attested not only by the geologic and physiographic
evidence just considered but by the similar relative
abundance of the adaptive types of mammals found in
these two formations.
The extent of the collections in the American
j Museum, the total number of specimens collected, and
the field records of Granger show that fossil mammals
are about four times as abundant in the Systemodon
zone as in the Lambdotherium zone.
The relative numbers of the mammals in the
Systemodon zone of the Big Horn Basin of Wyoming,
as indicated by the number of specimens collected,
are shown below:
Mammals in the Systemodon zone
Specimens
Perissodactyla (horses, 1,202; tapirs, 370) 1, 572
Insectivora (Hyopsodus, 254) 306
Condylarthra (mostly Phenacodus) 264
Amblypoda (corj'phodons only) 209
Creodonta (various carnivores) 203
Primates (Pel3'codus, Anaptomorphus, etc.) 151
Artiodactyla (mostly Trigonolestes) 120
TiUodontia (Esthonyx) 73
Rodentia (Paramys) 16
Taeniodonta (?edentates, Calamodon) 2
Edentata (new type) 1
2,917
This table naturally is only approximately repre-
sentative. The rodents, for example, were probably
far more abundant numerically than the horses.
The great number of horses, tapirs, and hyopsodonts
in the assemblage listed above is due in part to the
abundance of these animals in the "red beds." The
table is valuable chiefly in expressing the relative
abundance of the adaptive types of ungulates.
In the Lambdotherium zone the relative abundance
of the remains of ungulates undergoes a marked
change: both of the archaic types of mammals, the
condylarths and the amblypods, are relatively less
abundant than in the Systemodon zone.
276
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Relative frequency of mammals in the Lambdotherium zone of
Wyoming as indicated by the number of specimens collected
Specimens
Perissodactyla (horses, 191; titanotheres, 124; heptodonts,
56; hyrachyids, 2) 373
Primates (pelycodonts, 42; anaptomorphs, 9; Microsyops,
30) 81
Insectivora (Hyopsodus, 71) 75
Condylarthra (Phenacodus, 47; Meniscotherium, 1) 48
Rodentia (Paramys, 39) 42
Creodonta (various carnivores) 35
Amblypoda (Coryphodon, 29; Bathyopsis, 2) 31
Artiodactyla (Trigonolestes) 20
Tillodontia (Esthonyx) 12
Edentata (Taeniodonta) (Stylinodon, Calamodon) 5
722
It will be observed that the Condylarthra {Phena-
codus) here drop to the fourth place in relative fre-
quency, and the Amblypoda {Coryphodon) drop from
the fourth to the seventh place. This reduction is
partly in accord with the reduction of the archaic types
of mammals generally, as shown also in the following
faunistic comparison; in the Systemodon zone both the
genera and species of archaic mammals numerically
exceed those of modern type, whereas in the Lambdo-
therium zone the genera and species of archaic and of
modern types are evenly balanced.
The relative frequency of the various adaptations
to cursorial, ambulatory, and arboreal life is still more
significant. The comparison of the adaptive types of
the Systemodon and Lambdotherium zones is as follows :
System- Lambdo-
odon therimn
zone zone
Ungulates; cursorial, small, light-limbed 1,692 373
Ungulates; mediportal, medium in size 264 63
Ungulates; graviportal, large, heavy-limbed 209 31
Primates; arboreal, climbing -types 194 81
Insectivores; rodents, etc., ambulatory, small
terrestrial and fossorial mammals 395 139
Carnivores; larger and smaller creodonts 203 35
2,957
722
It will be observed that although the fossils collected
from the Lambdotherium zone are only one-third as
numerous as those of the Systemodon zone the relative
abundance of the adaptive types is approximately the
same, a fact that sustains the inferences as to geologic
and physiographic continuity or the absence of any
marked changes of environment during lower Eocene
time. Also, in the Lambdotherium zone, as in the
Systemodon zone, there is still a numerical predomi-
nance among the ungulates of cursorial types, the
horses {Eohippus), the lophiodonts {Eeptodon), the
titanotheres {Lambdotherium). Singularly, no primi-
tive tapirs {Systemodon) have been found. Among
the mediportal types may be noted the remains of
Phenacodus and the titanothere Eotitanops. The
graviportal coryphodonts, which may have been am-
phibious or partly aquatic in habit, are comparatively
The first author to analyze the Wasatch fauna with
reference to adaptive types in their bearing on physi-
ography and climate was Loomis (1907.1), who divided
the Wasatch fauna into percentages, substantially as
follows :
Terrestrial and arboreal types 75
Aerial 3
Amphibious 12
Aquatic, including crocodiles, turtles, and fishes 10
The present analysis of the adaptations of foot
structure in Wasatch and Wind River time combined
gives the following relative degrees of abundance
among the hoofed mammals:
Ungulates; small, cursorial, light-Umbed types 2, 065
Ungulates; medium, mediportal types, proportioned like
the tapir 327
Ungulates; graviportal, heavy-hmbed types {Cory-
phodon) , proportioned hke the hippopotamus 240
The analysis both of the geologic and paleontologic
evidence appears to show that in Wind River time
there was a warm but relatively dry and invigorating
climate in the Rocky Mountain region; that there
were streams, swamps, and river borders for the coryph-
odonts, forests and meadow borders for the true
titanotheres {Eotitanops), and open spaces with harder
ground for the diminutive horses, lambdotheres, and
heptodonts. In the forests there were numerous lemu-
roid or monkey types, as well as arboreal rodents, and
on the borders of the savannas there were terrestrial
and partly fossorial edentate-like mammals. Periods
of aridity and areas of drier ground favored the de-
velopment of the light-limbed ungulates.
CONTRASTS AND RESEMBLANCES BETWEEN LAMBDO-
THERIUM AND EOTITANOPS
Geologic and biologic evidence of the existence of
areas of dry, hard ground in Wind River time is thus
adduced to explain the surprising fact that the feet
and limbs of the little Lambdotherium are more highly
specialized for cursorial locomotion than the feet of
any of the known middle Eocene titanotheres. An
alternative interpretation is that the ancestral peris-
sodactyls wei'e small cursorial forms with narrow feet
like Heptodon and Systemodon and that the widening
of the feet is a secondary adaptation to mediportal
habits. (See p. 586.) The skull of Lambdotherium
is elongate and relatively Eohippus-like. Lambdo-
therium was probably an early specialized cursorial
member of the great titanothere family, a member
that died out without leaving descendants. Under
the law of local adaptive radiation it may have lived
in the drier uplands; at all events its remains are
especially abundant in the "red beds," in which all
together no less than 111 specimens have been found
in comparison with 14 of the bulkier Eotitanops.
(See fig. 230.)
Judging by the 14 specimens of Eotitanops that
have been found in the typical Wind River Lambdo~
EVOLUTION OP THE SKULL AND TEETH OF EOCENE TITANOTHBRES
277
therium zone, they had already shown considerable
specialization both in structure and in size. The
smaller and more primitive forms, such as E. gregoryi,
only slightly exceed Larabdotherium popoagicum in
central digit — it still resembles Lambdotherium, as
shown in Figure 220. Its feet are adapted to softer
ground, and we may conjecture that it resembled the
tapir in its habits.
Figure 230. — Reconstructed skeletons and restorations of Lambdotherium 'popoagicum (Ai, A2) and Eotitanops
borealis (Bi, B2)
Drawn by E. S. Christman under the direction of W. K. Gregory. About one-tenth natural size. These provisional reconstructions of skeleton
and body are based on material in the American Museum.
size, but species like Eotitanops princeps and E. major
are little inferior to the existing American tapir
{Tapirus terrestris). The gradations in size between
these five or six species and mutations of Eotitanops
may be judged from the accompanying outlines (fig.
231) of the lower jaws.
Eotitanops is a more typical titanothere than Lamb-
dotherium. In its limb structure it approaches espe-
cially Mesatirhinus, of the middle Eocene, although
in its mesaxonic foot structure — that is, its enlarged
The principal contrasts
theres are the following:
Lambdotheri um
Cursorial, light limbed, small,
like Eohippus.
Face decidedly elongate.
Muzzle attenuate, pointed.
Cranium abbreviate.
Three premolars.
Cropping teeth procumbent.
between these two titano-
Eotitanops
Cursorial to mediportal, small
to large, almost equaling
Tapirus.
Face moderately elongate.
Muzzle relatively abbreviate,
broad.
Cranium intermediate.
Four premolars.
Cropping teeth more erect.
278
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Both animals possess a number of features in com-
mon: (1) Tliey show a similar cone and crescent (or
Figure 231. — Lower jaws of Lamhdotherium., Eotitanops, and
Tapirus
One-fourth natural size. A, Lambdoiherium popoagicum. Am. Mus. 14906; Alkali
Creek, Buck Spring, Wind River Basin. B, Eotitanops grcgoryi, Am. Mus.
14889 (type); Alkali Creek, Buck .Spring, Wind River Basin, C, E. brownianus,
Am. Mus. 4885 (type); Wind River Basin. D, E. borealis, Am. Mus. 14891;
west bank of Wind River, 3 miles above canyon (top of banded beds). E, E.
princeps. Am. Mus. 296 (type); Wind River Basin. F, Tapirus ierrestris, Am.
Mus. 1135; immature rpecimen.
bunoselenodont) pattern of the superior grinding
teeth, with either reduced or vestigial intermediate
cusps or conules; (2) in both the preorbital (or facial)
part of the skull is longer than the cranial (see fig. 232),
a primitive characteristic of perissodactyls which at
once allies these animals to other primitive perisso-
dactyls and distinguishes them from the middle
Eocene titanotheres, in which the cranium is longer
than the face; (3) they show enlargement of the median
digit (D. Ill), or mesaxonic structure of the fore and
hind feet.
It is quite possible (see Chap. VIII) that this
mesaxonic specialization is itself common among
perissodactyls — that is, that all perissodactyls are
descended from quadrupeds with narrow feet, as sug-
gested by Gregory. We should interpret this swift-
footed structure as an adaptation that enabled the
small, defenseless perissodactyls, without horns or
tusks, to escape their pursuers.
CBp7ialic70O%
FiGTJKE 232. — Restored contours of skulls of La?7ibdothe-
rium and Eotitanops
Showing estimated proportions of face (shaded) to cranium. One-fourth
natural size. A, Lambiotheriam, face 65, cranium 35. B, Eotitanops, lace 56,
cranium 44.
The special similarities of foot structure between
Lamhdotherium, and Eotitanops embrace the follow-
ing characters: (1) Enlargement of the median digit
(D. Ill), which is provided with a broad proximal
phalanx; (2) reduction of digits II and IV, with re-
lative narrowing of the proximal phalanges; (3) reduc-
tion of digit V (correlated with this metapodial
structure we find that the lunar immediately rests on
the unciform, with a narrow {Lambdotherium) or
oblique (Eotitanops) facet on the magnum) ; (4) small,
relatively high and narrow magnum. These points
are well illustrated in the accompanying figures of the
manus of Lamhdotherium and Eotitanops.
The interpretation of this somewhat specialized
form of foot as one derived from a cursorial type is
considered in the introduction to the study of the
skeleton of the Eocene titanotheres (p. 586).
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
279
EXPLORATIONS AND DISCOVERIES
The explorations in the Wind River Basin for re-
mains of titanotheres and the types collected were as
follows :
1880. J. L. Wortman, for E. D. Cope; EoHtanops borealis
(Cope), E. brownianus (Cope), and Lambdotherium
popoagicum (Cope).
1891. J. L. Wortman, for American Museum of Natural
History; EoHtanops princeps Osborn.
1894. F. B. Loomis, for Amherst College Museum; Lamb-
dolheriavi primaevum Loomis.
1909, 1910, 1911. Walter Granger, for American Museum of
Natural History; L. priscum Osborn, L. progressum
Osborn, and E. gregoryi Osborn.
The American Museum expeditions of 1909, 1910,
and 1911, under Granger and Sinclair, made a series of
important discoveries — first, in locating all these
titanotheres in the Lamhdotherium zone; second, in
proving that EoHtanops and Lambdotherium were
contemporaneous; third, in collecting the remains of
more than eight specimens of EoHtanops (Am. Mus.
14887-14894); fourth, in collecting a complete skull
and jaws of Eotitanops iorealis, affording proof that
EoHtanops is more primitive than the middle Eocene
or Bridger forms and belongs to a distinct generic
stage.
The Wind River Lamhdotherium zone is 400 feet
thick and may be divided into four levels, each in-
cluding 100 feet. It is a remarkable fact that nearly
all the remains of the 124 specimens of titanotheres
found by the American Museum parties were collected
between the 250 and 400 foot levels, as shown in the
section taken from Granger's field records (fig. 48).
Thus Eotitanops and Lamhdotherium occur contem-
poraneously. It appears that the mutations or spe-
cies of Eotitanops do not exhibit a continuously graded
evolution or succession in ascending levels, for it hap-
pens that the smallest and most primitive form known,
E. gregoryi, occurs on a high geologic level, showing
that Eotitanops was already polyphyletic in early
Eocene time. This is an example of the extreme
importance of an exact record of levels.
SYSTEMATIC DESCRIPTIONS OF THE LOWER EOCENE
TITANOTHERES
Superfamily Titanotheroidea Osborn
Perissodactyls with bunoselenodont superior molars
and selenodont inferior molars. Distinguished from
the Chalicotherioidea by normal limbs and hoofs.
Distinguished from the Hippoidea by a persistently
tetradactyl manus and tridactyl pes.
Family Brontotheriidae Marsh
Related to the type of Brontotherium. Evolving
between early Eocene and early Oligocene time.
Primitively cursorial in gait but early evolving into
mediportal and graviportal forms. Terminal pha-
langes and hoofs progressively reduced. Forward
101959— 29— VOL 1 21
portion of skull originally elongate but early becoming
abbreviate; cranial portion progressively elongate.
Earlier genera hornless; paired nasofrontal horns
developing in middle Eocene time and becoming the
dominant character of the skull.
Subfamily Lambdotlieriinae Osborn
Lower Eocene titanotheres, small, light limbed, of
slender, cursorial proportions. Skull mesaticephalic;
facial region elongate; cranial region abbreviate.
Superior molars brachyodont, incipiently bunoseleno-
dont, with lophoid paraconules and metaconules;
lower molars with pronounced metastylids; premolars \
absent; lower premolars p2_4, progressively molariform.
Lambdotherium Cope
Plate LIV; text figures 27, 33, 103, 114, 143, 146, 154, 221,
228, 230-242, 244, 483, 484, 486-492, 503, 504, 510, 512, 521,
522, 661, 685, 688, 692, 694, 700, 701, 723
[For original description and type reference see p. 168. For skeletal characters see
p. 590]
Generic characters. — Skull of decided proopic doli-
chocephaly. Anterior premolars wanting. Superior
molars broadened transversely, with prominent para-
styles and mesostyles, oblique ectolophs, large, free
protoconules, and low metalophs; m' with prominent
hypocone; inferior molars with metastylids; first up
per and lower premolars wanting; posterior lower pre-
molars progressive; ra^ with large crescentic hypoco-
nulid. Manus numerically tetradactyl, functionally
aniso tridactyl; lunar resting chiefly on unciform an-
teriorly; magnum small, high, and narrow.
We know neither the ancestors nor the descendants
of this animal. It is already more highly specialized
in many respects than certain of the middle Eocene
titanotheres. Lamhdotherium is readily distinguished
by its fine, delicate construction for swift movement,
in which it has analogies to the lower Eocene horses.
It has a long, slender snout and delicately prehensile
jaw. We note especially that (1) the face is much
elongated, the faciocephalic index being 65 as com-
pared with 56 in Eotitanops and 60 in Eohippus; (2)
this elongation is correlated with a very slender snout-,
but the first upper and lower premolars are wanting;
(3) the third and fourth lower premolars are more
complicated than in Eotitanops, p4 in some forms
closely resembling a molar tooth; (4) its sharply
piercing canine teeth and chisel-shaped incisors are
other features of specialization.
History of discovery of Lamhdotherium. — Wortman' s
discovery in the Wind River valley (1880) and Cope's
original description have already been cited. Cope
at once recognized the ancestral relationship of this
form to the titanotheres. In 1889 he made Lamhdo-
therium the type of a distinct family — "Lambdo-
theriidae"- — to embrace all the Eocene titanotheres, a
family name that was adopted by Flower and Lydek-
ker. In his "Tertiary Vertebrata" (1885.1, pp. 709,
280
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
711) Cope mistakenly referred to this genus the two
species L. ( = Eotitanops) hrownianum. and L. procyo-
niiMm. In the same memoir he placed all the Eocene
titanotheres in the family "Chalicotheriidae." In his
memoir of 1892 Earle correctly considered this animal
a probable member of a side line of titanotheres. In
1893 Osborn recognized the division as a subfamily —
"Lambdotheriinae." In 1897 Osborn mistakenly pro-
posed to remove Lamhdotherium from the titanotheres
and related it to the Equidae on the ground of its slen-
der foot structure. The renewal of the demonstration
of its relation to the titanotheres is due to W. K.
Gregory. In 1907 Loomis discovered five specimens
of this genus in the deposits of Buffalo Basin, one of
which he selected as the type of the new species L.
primaevum, believing it to represent a stage somewhat
more primitive and perhaps geologically older than L.
popoagicum. In 1905 and 1909 the American Museum
party under Granger finally determined that these
animals were geologically contemporaneous with Eoti-
tanops but confined to a comparatively narrow geologic
zone.
The principal collectors and the areas in which they
worked were as follows:
1880. E. D. Cope, J. L. Wortman, for American Museum of
Natural History; Wind River Basin, Wyo.
1896. J. L. Wortman, for American Museum of Natural His-
tory; Huerfano Park, Colo.
1905. Walter Granger, for American Museum of Natural His-
tory; Wind River Basin, Wyo.
F. B. Loomis, for Amherst Museum; Big Horn Basin,
Wyo.
Walter Granger, for American Museum of Natural His-
tory; Wind River Basin, Wyo.
1907.
1909.
1916. Walter Granger, for American Museum of Natural His-
tory; Huerfano Park, Colo.
1918. Walter Granger, for American Museum of Natural His-
tory; Huerfano Park, Colo.
Geologic horizons. — The animals found in four ex-
posures of the Lamidotherium zone in the Wind River
Basin, the Big Horn Basin, the Beaver Divide, and
Huerfano Park, each 300 to 400 feet thick, are not
separated by marked differentiation or evolution; in
all these basins and on all the levels of each formation
the lambdotheres, so far as known, are substantially
similar in size but differ markedly in the degree of
evolution of the third and fourth lower premolar
teeth. The range in size is indicated on page 282 . A
very striking fact is that the extremes of premolar
structure (-fig. 234) were found in animals collected
around the great Alkali Creek "red stratum," which is
such a conspicuous level mark (fig. 47), many of the
specimens being just below the "red stratum," or 50
to 100 feet above the base of the Wind River forma-
tion. The greater part of the collections have been
made within a vertical distance of 200 feet, which
would represent time for considerable evolution; but
as the stages of evolution do not occur successively in
the ascending levels, it does not appear practicable to
separate any but the extreme forms as species or
mutations, and the systematic order therefore appears
as follows:
Lamhdotherium progressum Osborn (most progressive) .
L. popoagicum Cope, L. primaevum Loomis (inter-
mediate) .
L. priscum Osborn (most primitive).
Measurements of teeth of Lamhdotherium, in millimeters
[The numbers are those of specimens in the American Museum of Natural History]
L
. priscum
L. popoagicum
L.
progressum
L. sp.,
14922
L. mag-
U916
14912
14914
12822
(type)
14908
4863
(type)
14899
14904 14907
14902
14917
(type)
14918
14919
17527
(type)
68.7
60.9
50.9
26.7
42
7. 5
9.3
9.4
11.4
12
17.5
64
57
47
25
39
7.6
8.2
9
10.3
11. 1
16.2
69
61
51
26.4
42.4
9
9.5
1L4
12.6
17.4
70.6
62.3
53.3
27.4
43. 1
8.2
9
9.3
12.8
13.2
17.3
9.3
10.8
12.5
74
58
59
64
51.8
26.4
54
P2-P4
25
36
28. 5
39
41
41
44
P2 (ap.)
7.6
9
9.5
11.4
12.5
8.5
P3 (ap.)
8.5
9
10.9
11.8
16
9
9
10. 8
12
9.5
9.3
11
12. 5
16. 5
8
9
12
8.2
10
10. 1
15
9.5
P4 (ap.)
10
Ml (ap.)
11
11. 4
17.7
12
13
18.5
P2 m'
58.6
51
24
34
7
7.8
8.5
10.4
11. 1
12
pS jqS
50.6
V- D* 1
M' m^ 1
36
34
P2 (an )
P2 (an )
7
8
10.3
10.9
11.4
P* (au )
1
M' Can 1 1
11.5
12
12
M2 faD ■)
M« (an )
EVOLUTION HE SKULL AND TEETH OF EOCENE TITANOTHERES
281
Lambdotherium popoagicum Cope
Plate LIV; text figures 27, 33, 103, 143, 228, 230, 231, 233-237,
244, 483, 484, 486-492, 503, 504, 512, 521, 522, 661, 694, 700
[For original description and type references see p. 168. For slieletal characters
see p. 590]
Type locality and geologic horizon. — Wind River
Basin, Wyo.; Lambdotherium zone (Wind River B).
Specific characters. — P2-m3 69-56 millimeters; ps with
paraconid, metaconid, and hypoconid intermediate
in development. In superior molars the protoconules
more or less free and distinct, metaloph low but dis-
tinct, cingula not surrounding the crown internally.
Materials. — The type species (figs. 234, H; 236, C)
of a series of mutations of specific character is repre-
sented by over 70 specimens in the collections of the
American Museum, chiefly from the typical Wind
River formation, but also from contemporaneous
deposits in the Big Horn Basin, Beaver Divide, and
Huerfano Basin. These specimens consist mostly of
scattered upper and lower teeth and fragments of
jaws but include several nearly complete jaws. One
specimen (Am. Mus. 4880) affords a limited but signi-
ficant knowledge of the skeleton.
SJciill.— The imperfectly known skull is analo-
gous to that of the primitive horses rather than
that of Eotitanops or any of the typical middle
Eocene titanotheres. There are only two speci-
mens (Am. Mus. 14903, 14907) in which fragments
of the skull are associated with the teeth, from
which the conjectural restoration (fig. 233) is as-
sembled, the outlines of the anterior part, or pre-
maxillaries and nasals, being inferred from the
attenuate structure of the lower jaw. The prin-
General features of the teeth. — The dental formula, so
far as known, is If, C\, Pf , M|. The inferior incisors,
as observed in Am. Mus. 14899, 14906, 14920, repre-
sented in Figures 233, 236, 237, are semicircular in
arrangement, semiprocumbent, with spatulate or
chisel-shaped crowns; the median incisors especially,
which are distinctly chisel-shaped, are quite different
from those of Palaeosyops, which are bluntly pointed.
The inferior and superior canines are rounded, slightly
compressed laterally, and sharply pointed.
Ty2)e preinolars. — No trace of p^ or pi is to be found
in any of the specimens; this tooth is ordinarily very
persistent in the Perissodactyla. In the L. popoagicum
type premolar series p2 is an elevated, laterally com-
pressed cone, with a rudimentary paraconid and low,
narrow heel bearing a hypoconid; ps presents an
anterior lobe composed of a low paraconid, an elevated
protoconid, a postero-internal metaconid elevated but
slightly developed, a somewhat broader posterior
Figure 233. — Skull of Lambdotherium "popoagicum, reconstructed
cipal characters are the following: (1) Dolicho- MadebyL. M. sterling under the direction otW. K.Gregory. About two-fifths natural size.
,. . iUi"! IJ '^'''^ reconstruction is made from three specimens in the American Ivluseum, collected in
Cephaly OI prOOpiC type that is, long, slender ^^^ ^^^^ jji^^j. Basin-No. 14899,Alkali creek, Buck Spring, lower jaw; No. 14907, Alkali
skuU, in which the facial greatly exceeds the cranial creek, Wolton, maxilla, malar, and skull top; No. 14903, Alkali Creek, Buck Spring,
, ,1,1 <•• 11--1 1- nr- squamosal and condyle. Missing parts conjecturally restored by comparison with Systemo-
length, the faciocephalic mdex being 65, as com- ionB.niEoMppm.
pared with 56 in Eotitanops; (2) sagittal crest rather
low and slender; (3) external auditory meatus open
inferiorly; (4) infraorbital foramen placed above the
second premolar — that is, decidedly anterior in posi-
tion as compared with that in the typical titano-
theres; (5) an attenuated rostrum associated with
the elongated symphysis of the jaw, suggesting the
conformation of the skull of a ruminant rather than
of a perissodactyl.
Comparison. — This skull is very close in its propor-
tions to that of Eohippus, and if it were not for the
differences in the teeth might be mistaken for it.
The faciocranial indices are similar, namely:
Eohippus
Lamb-
dotherium
Eotitanops
Faciocephalic _ __ -
68-60
42-40
41
65
35
(?)
56
44
(?)
heel or hypoconid, with a rudiment ry internal crest
representing the entoconid; p4 is a more progressive
or submolariform tooth with an anterior transverse
crest composed of protoconid and metaconid behind
which is a low, incomplete posterior crest supporting
an elevated hypoconid and a depressed internal ridge.
Lower premolars, primitive and progressive muta-
tions or specific forms. — The structure of the cusps in
P2, Ps, P4 is very important. The accompanying
diagram (fig. 234) shows the wide range of progressive
evolution in the lower premolar teeth which are
exhibited in the large number of specimens in the
American Museum collection. They embrace stages
ranging from far less primitive to stages far more
primitive than the lower premolar teeth of the type
of L. popoagicum. These stages are especially impor-
tant and interesting because they are recorded as
coming from similar geologic levels. These records of
geologic levels may be confused, but accepting them
282
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBjRASKA
as correct, it seems impracticable to divide Lamb-
dotherium into a large series of species, although
the development of the
premolar cusps certainly
warrants specific separa-
tion.
These extreme stages are
therefore grouped together
for the present as muta-
tions between Lambdothe-
rium priscum, represented
by the simplest type (No.
12822), and Lambdotherium
progressum, represented by
the most advanced type
(No. 14917). Halfway
between the extremes is
the type species L. popo-
agicum.
The cusp evolution is very
interesting, including the
following elements: Ante-
rior lobe — protoconid (pr*),
paraconid (pa*^), metaconid
(me"*), metastylid (ms"^);
posterior lobe — hypoconid
(h**), entoconid (en'').
The series represented in
Figure 234, including L.
priscum at the bottom and
L. progressum at the top,
presents a complete mor-
phologic transition or epit-
ome of premolar evolution,
it being essential to note
that we do not know
Figure 234.— Lower premo- whether this corresponds
lars of three "species" or .,•■ i
f T<7,fh o T-ooi succession m
mutations of Lambdothe-
rium, illustrating progres-
with
time.
real
At the base p2 and
sive evolution of the pre- p^ are excessively simple,
NatoalsL^'selnd, third, and fourth but at the Summit Pa haS
lower premolars of the left side, inner a large paraCOnid and mcta-
sideview. A-E, i. priscum; F-H, i. . , , ,
popoagimm; I-L, i. progressum. AH COnid, and P4 haS a para-
American Museum specimens from „ -J rnptflponid metfl-
the Wind River Basin. A, No. 12822 COUIQ, metaCOniQ, meia-
stylid, and rudiment of an
entoconid, which is devel-
oped as a distinct cusp in
certain specimens (such as
(type), 3 miles east of Lost Cabin; B,
No. 14916, Alkali Creek, Buck Spring;
C, No. 14900, Alkali Creek, Wolton;
D, No. 14912, Dry Muddy Creek, 18
miles above mouth; E, No. 14914,
Alkali Creek, Wolton; F, No. 14915,
Alkali Creek, Wolton; G, No. 14924,
Dry Muddy Creek, 18 mUes above Am. MuS. 14924).
mouth; H, No. 4863 (type). Wind 7-7
River valley; I, No. 14913, Muddy PremoiaTS relatively prO-
Creek, south side, 18 miles above • tj_ •
mouth; J, No. 14919, Alkali Creek, gressive.—lt IS vcry impor-
Wolton; K, No. 14918, lower Alkali font in nntA tViot n nnrl
Creek; L, No. 14917 (type). Alkali ^^^^ ^^ ^^^^ ^'^^^ Ps ^^^
Creek, Buck Spring. p^ in the progressive forms
are much more advanced in evolution than the
corresponding teeth in the contemporary Eotitanops;
in fact, p4 is submolariform and lacks only the prom-
inence of the metaconid to be like a molar. P2 in
Lambdotherium is almost as progressive as ps in
Eotitanops. Even in the middle Eocene species
Palaeosyops leidyi, ps, p4 are not so far advanced as
in Lambdotherium. We observe also another dis-
tinctive character: Whereas in Palaeosyops the meta-
conid arises as a bud or reduplication of the proto-
conid, in Lambdotherium it springs from the posterior
side of the protoconid. These details are of impor-
tance as demonstrating the accelerated rate of evolu-
tion of the premolar cusps as a character of lamb-
do theres.
Superior premolars. — Comparatively few well-pre-
served superior premolar series are Icnown, so it can
not be determined whether there is a corresponding
series of mutations in the evolution of the upper teeth.
In the specimens Am. Mus. 14902, 14900, 14911,
14907 the following characters are observed: (1) p^
very simple, with single external protocone and rudi-
ments of the deuterocone and sometimes of the
tritocone; (2) p', p* with rudimentary parastyle,
deuterocone, tritocone, rudimentary crests connecting
deuterocone with protocone and tritocone, respec-
tively, faint conules sometimes observed on these
crests.
A series of deciduous premolars (Am. Mus. 14934)
exhibits dp^ somewhat more complex than p^, dp^
elongate, quadricuspidate, with prominent parastyle
and mesostyle.
Molars. — The inferior molars are highly character-
istic teeth, distinguished especially by the elevation of
their crescents; the protolophid consists of an elevated
protoconid, metaconid, and metastylid, or double
internal cusp, which is very distinct in unworn teeth.
This reduplicate cusp, which develops in the upper
Eocene species of horses and also in the true chali-
cotheres, is not present in the titanotheres of the
middle Eocene, such as Palaeosyops. A rudimentary
hypoconuHd is usually observed in nii and m2 and
develops into a strong crescentic third or posterior
lobe in 1113.
The superior molars, as observed in five specimens
in the American Museum (Nos. 14900, 14902, 14904,
14907, 14911), are fairly uniform in character, with
very prominent parastyles, mesostyles, variable pro-
toconules, rudimentary or lophoid metaconules. (PI.
LIV, A, B; fig. 235, A.) A very distinctive feature is
the large hypocone on m^ The following characters
should also be noted: (1) The transverse diameter
always exceeds the anteroposterior; (2) m' is a rela-
tively small tooth; m^ is usually the largest tooth of
the series; m' is usually intermediate in size but some-
times is the largest tooth of the series.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
283
Measurements of superior molars of Lambdotherium popoagicum,
in millimeters
Ml
M!
M>
Ap.
Tr.
Ap.
Tr.
Ap.
Tr.
Am. Mus. 4664 .._
10.5
1L5
16
Am. Mus. 4880 .
11.5
11.5
15
14.2
Am. Mus. 14902, J.
10.5
13
12
14.5
Other distinctive characters of the superior teeth
are the very obUque ectolophs, the prominent para-
styles and mesostyles, the sculptured form of the
cusps, especially apparent in the unworn specimens.
The protoconule is distinct and usually of subtriangular
form; it is connected with the protocone by a low
crest. The metaconule proper is rather sessile, indis-
tinct, or wanting; there is, in fact, a depressed metaloph
or rudimentary posterior crest. The third superior
molar (Am. Mus. 4664) exhibits an especially oblique
ectoloph, also a prominent hypocone and low but
distinct metaloph.
The jaw.— The type jaw (Am. Mus. 4863, fig. 236)
consists of two separate and incomplete rami figured
by Cope (Tertiary Vertebrata, PL LVIII, B). The
distinctive feature of the type species is the inter-
mediate condition of the third premolar, which dis-
tinguishes this animal from primitive and more pro-
gressive specimens.
There are five more complete jaws, namely, Am.
Mus. 14899 (figs. 231, 236), 14905, 14906 (figs. 231,
236), 14909, which together afford a full knowledge
of the characters of the jaw except the angular and
condylar region ; the elongate and laterally compressed
chin (fig. 236), even more extreme than that of the
Eocene horses; the wide diastema between the canine
and the second premolar; the incisive border extending
somewhat to support the slender, recurved, prehensile
canines and the row of chisel-shaped, semiprocumbent
teeth; the coronoid process high, vertically placed,
sharply defined, with flat anterior face.
The extremes of measurement are shown below.
Measurements of jaw of Lambdotherium, in millimeters
L. popoagicum,
Am. Mus.
4863 (type
jaw)
L.progressum,
Am. Mus.
14919 Oargest
L. priscum,
Am. Mus.
14908 (smaJlest
jaw)
P2-m3, anteroposterior
Mi-m3, anteroposterior
M], anteroposterior
M2, anteroposterior
M3 , anteroposterior
Depth of jaw below ms
68
41
11
12
16
32
70.6
43. 1
66
37
The premolars are 63 per cent of the length of the
molars.
Lambdotherium primaevum Loomis
Plate LIV, C, D; text figure 114
[For original description and type reference see p. 178]
Type locality and geologic horizon. — ^Big Horn Basin,
Wyo. ; Lamhdotherium-Eotitanops-Coryphodon zone (Big
Horn D).
Specific characters. — Superior molars with crescentic
protoconules; cingula completely surrounding the
crowns. Measurements as in L. popoagicum.
This type is significant as coming from the Big
Horn Basin. The type superior first and second
molar teeth (fig. 114, p. 178) may be readily dis-
tinguished by the greater development of the internal
cingulum, which completely surrounds the crown.
Another feature is that the protoconules are large
and subcrescentic, and the metaconules are lost in the
metaloph. These measurements^^ are:
D
Figure 235. — Upper and lower grinding teeth of
Lambdotherium
Natural size. A, L. popoagicum, Am. Mus. 14902, Alliali Creek,
Wolton; left upper premolar-molar series. B, L. priscum.
Am. Mus. 14908, Dry Muddy Creek, 18 miles above mouth;
light piemolar-molar series. C, L. progressum, Am. Mus.
14917 (type). Alkali Creek, Buck Spring; rightlower premolar-
molar series. D, L. progTcssum, Am. JNIus. 14918, lower Alkali
Creek; first lower molar of the left side, crown view. All
from Wind River Basin.
Measurements of molar teeth of Lainhdolhcrium primaevum
Millimeters
M' and m^, combined, anteroposterior 23. 5
M', anteroposterior 11
M>, transverse 13
M^, anteroposterior : 12
M^, transverse 15
M^, transverse, maximum along anterior border 18
Ml to M3, combined, anteroposterior 41
Ml, anteroposterior 11
M2, anteroposterior 12. 5
M3, anteroposterior _17
22 The measurements of mi and m* were accidentally transposed in Loomis'
original description.
284
TITANOTHBRES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
^4x^ kid^
Figure 236. — Lower jaws and teeth of Lambdotherium popoagicum
Natural size. A, Am. Mus. 14899, Alkali Creek, Buck Spring; lower jaw, inferioi surface. Bi, Am. IMus. 14906, Alkali Creek, Buck Spring; lower
jaw, infeiior surface; an older individual. Bi, The same, showing upper surface of symphyseal region. Ci, Am. Mus. 4863, Wind River valley;
type jaw, outer side view. Cs, The same, lower piemolar-molar series, crown view.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
285
Figure 237. — Lower jaws and teeth of Lambdotherium popoagicum
Side view. Natural size. A, Am. Mus. 4863 (type); Wind River valley; front part restored from Am. Mus. 14899. B, Am. Mus. 14906; Alkali Creek, Buck
Spiing; an old individual. C, Am. Mus. 14899; Alkali Creek, Buck Spring. D, Am. Mus. 2989; Wind River valley.
286
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The three inferior molar teeth, mi-ma, measure
longitudinally 41 millimeters, as compared with 42 in
the type of L. popoagicum. The external crescents
(protoconid, hypoconid), the internal cones (meta-
conid, distinct metastylid, entoconid), and the cres-
centic third lobe of ms (hypoconulid) are characteristic.
Lambdotherium priscum Osborn
Text figures 146, 234, 235, 238-240
[ For original description and type references see p. 194]
Type locality and geologic horizon. — Wind River
Basin, 3 miles east of Lost Cabin, Wyo.; Lambdoihe-
FiGURE 238. — Front part of type lower jaw of
Lambdoiherium priscum
Natural size. Am. Mus. 12822, reversed, showing the long postcanine
diastema, the three lower premolars, and the first lower molar.
Locality, 3 miles east of Lost Cabin; Wind River formation.
rium-Eotitanops-CorypJiodon zone (Wind River B);
Granger, collector, American Museum expedition,
1905.
Specific characters. — P2-P4, 25 millimeters; nii-ms
(referred specimen), 37 millimeters; second and third
lower premolars extremely simple, with rudimentary
paraconid; ps, metaconid rudimentary, placed very
low upon slope of protoconid, talonid narrow, de-
pressed, with cingular rudiment of entoconid. (See
fig. 238.)
The extremely simple or primitive structure of the
second lower premolar clearly distinguishes this
stage.
A referred specimen (Am. Mus. 14908) collected
by Granger (American Museum expedition, 1909),
is slightly more advanced in the structure of the second
lower premolar (fig. 239) but is still much more
primitive than the type of L. popoagicum.
The measurements of these two specimens are
shown below.
Measurements of Lambdoiherium -priscum, in millimeters
12822 (type)
14908 (referred)
P2-P4— . -. _.
25
7
8
5
9
6.5
12
7.5
P2, anteroposterior.
8
P3, transverse
5
P4, anteroposterior .
8. 5
P4, transverse
Ml, anteroposterior
10
Mi, transverse. _. .
7
Mi-ms-
37
This Wind River species is identified in Huerfano
A by a fine pair of jaws from Garcia Canyon (Am.
Mus. 17526). The specific character of pa, without
trace of metaconid, is clearly shown in Figure 240.
This species is represented by another jaw, with teeth
of the same size (Am. Mus. 17528) in which ps,
also without metaconid, is in a slightly more advanced
stage of evolution, the talonid being broader.
This species is also doubtfully represented by the
imperfect specimen of upper teeth referred to L.
popoagicum by Wortman (Am. Mus. 2688), as well as
by a newly found specimen (Am. Mus. 17529) of
approximately the same size. In this new specimen,
found 3 miles east of Gardner Butte, the isolated
upper teeth of two sides, including p'-m^, show the
following characters: (1) Molars sUghtly smaller than
in the referred specimen of L. progressum, (2) conules
and cingulum not so well developed, (3) measurements
slightly inferior to those of the type of L. popoagicum.
(See p. 283.)
The types of L. popoagicum and L. priscum are both
lower jaws from the Wind River, and as there are
two lower jaws from the Huerfano positively referable
to L. priscum and none referable to L. popoagicum
it seems best to assign these two sets of upper teeth to
L. priscum also.
Lambdotherium progressum Osborn
Text figures 147, 234, 235, 241, 242
[For original description and type references see p. 194]
Type locality and geologic Tiorizon. — Wind River
Basin, Alkali Creek, Buck Spring; Lamhdotherium-
Trh2_
P2
Figure 239. — Incomplete lower jaw of Lambdoiherium priscum
Natural size. Am. Mus. 14908; Dry Muddy Creek, 18 miles above mouth. A
referred specimen. Oblique view of dentition.
Eotitanops-Coryphodon zone (Wind River B); Granger,
collector, American Museum expedition, 1909.
Specific characters. — P2-P4, 16.5 millimeters. Sec-
ond, third, and fourth lower premolars progressive:
rudiment of metaconid on P2; Ps with elevated meta-
conid subequal with protoconid, broad talonid with
rudimentary entoconid; p4 with bifid metaconid and
distinct entoconid.
This is readily distinguished from both L. priscum
and L. popoagicum by the advanced condition of ps,
which may be described as submolariform.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
287
Measurements of type of Lambdoiherium progressum {Am.
Mus. 14917)
Millimeters
P2-P1 26
P2, anteroposterior 8
P2, transverse (trigonid) 4. 8
P3, anteroposterior 9
This Wind River type is distinguished by ps, which
has a strong metaconid — that is, it is submolariform.
A series of molar teeth, p^-m^ (fig. 242), from the
highest level of the lower Huerfano, is referred to L.
progressum on the following grounds: (1) The upper
teeth fit pretty well those of the type of L. progressum
L. macjnum type
Figure 240. — Jaws and teeth of Lambdoiherium
Am. Mus. 17526, lower jaw of L. priscum, referred specimen from Huerfano A, outside and crown views. Am. Mus.
17527, outer view of type jaw of L. magnum, Huerfano A. Am. Mus. 15000, first and second upper molars of L.
magnum, referred specimen from the Wind River horizon of the Big Horn Basin, Wyo. Natural size. After
Osborn, 1919.
P3, transverse 6
P4, anteroposterior 9. 3
P4, transverse 7. 3
Ml, anteroposterior 11. 5
Ml, transverse . 8. 5
M2, anteroposterior 12. 5
Mj, transverse 9. 5
from the Wind River; (2) the parastyle is especially
prominent at the antero-external angle of m^, m'; (3)
m^ has prominent hypocone and cingular hypostyle;
(4) the outer cusps of the premolars are approximated,
conules prominent; (5) the cingulum is strong on p^
and m'.
288
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
With these specimens (association doubtful) were
found the calcaneum, portion of a tibia, and a proximal
phalanx.
Lambdotherium magnum Osborn
Text figures 154, 240
[For original description and type references see p. 199]
Type locality and geologic horizon. — Huerfano Park,
Colo.; lower horizon of the Huerfano formation,
Lamidotherium-Eotitanops-Coryphodon zone (Huerfano
A).
Specific characters. — Osborn writes:
Exceeding in size any other known lambdothere is the type
jaw (Am. Mus. 17527) from the Garcia Canyon, lower Huer-
FiGUKE 241. — Lower jaw and teeth of Lambdotherium progressum
Natural size. Am. Mus. 14917, type, reversed; Alkali Creek, Buck Spring; Wind
River formation.
fano, containing a complete inferior series, P2-m3 of both sides,
represented in Figure 240. These teeth exceed in length over
all (74 mm.) those of the type of L. popoagicum, in which the
same teeth measure 69. P3 has a rudimentary metaconid and
paraconid, in the same stage of evolution as in L. popoagicum.
Of similar large size is a referred specimen, Am. Mus. 15600
(fig. 240), from the Big Horn, west end of Tatman Mountain.
These referred grinders, m', m^, coincide closely in size with
the type of L. magnum and may be regarded as a paratype.
Below are given the measurements of teeth of
species of lambdotheres. The numbers following the
specific names are those assigned to the specimens in
the American Museum of Natural History.
Measurements of teeth of lambdotheres
INIillimeters
P2-m3; Huerfano A, L. priscum 17526 (referred) 67
Wind River B, L. popoagicum 4863 (type) 69
Wind River B, L. progressum 14917 (type) (esti-
mated) 71
Huerfano A, L. magnum 17527 (type) 74
M'-m^ :Huerfano A, L. priscum 17529 (referred) 21.5
Huerfano A, L. priscMTO 2688 (referred) 22.5
Huerfano A, L. progressum 17530 (referred) 23.5
Wind River B, L. popoagicum 14902 (referred) _. 25
Wind River B, L. magnum 15600 (referred) 27.5
These measurements show that there is not a great
range in size between the smaller and the larger animals
referred to this genus (Osborn, 1919.494).
Subfamily Eotitanopinae Osborn
Lower Eocene titanotheres of intermediate size.
Body proportions slender, submediportal rather than
cursorial. Skull dolichocephalic; facial region longer
than cranial region. Superior molars brachyodont;
molar tooth proportions much as in the Palaeosyopinae,
with reduced paraconules and metaconules. Inferior
molars without metastylids. Premolars j- present;
molarization of premolars retarded.
Discovery. — The details of the discovery of Eoti-
tanops {Palaeosyops horealis) in 1850 and the early
history of opinion are in part related above. In
Cope's "Tertiary Vertebrata" (1885.1, p. 703, pi.
58a, fig. 3) a full description is given of the type molar
teeth and the imperfect radii of the animal that Cope
called Palaeosyops horealis (Am. Mus. 4892).
Materials. — In 1891 Dr. J. L. Wortman, who had
discovered the type, enlarged our knowledge of this
genus by the discovery of another specimen (Am. Mus.
296), including a complete lower jaw, two cervical,
three dorsal, and one caudal vertebra, the femur,
humerus, and the greater part of the fore foot. These
bones were described by Osborn and Wortman in
1892 (1892.67) and were referred to the type species,
Palaeosyops horealis; they are now known as E.
princeps.
Soon afterward Earle's memoir (1892.1) appeared,
in which he treated Palaeosyops horealis as probably
ancestral to the Bridger Telmatherium cultridens. In
1908 Osborn (1908. 318) revised" the Eocene titano-
theres and placed P. horealis in the new genus
Eotitanops.
Granger's explorations in 1909 to 1911 resulted in
the discovery of the type of E. gregoryi and have
enabled us to make a systematic revision of these
animals based upon materials in the American
Museum collections, which are arranged below accord-
ing to size and morphologic succession; their geologic
succession is sho\vn in Figure 48. The numbers are
AM.I7S30
Figure 242. — Upper teeth of Lambdotherium
progressum
Am. Mus. 17530, refened specimen from Huerfano A. Natural
size. After Osborn, 1919.
those assigned to the specimens in the American
Museum of Natural History.
E. major Osborn, 14894 (type), a third metatarsal of the
left side (figs. 145, 506).
E. princeps Osborn, 296 (type), jaw, manus, humerus, femur,
etc. (figs. 144, 231, 246, 252, 484, 490, 494, 496, 498-600, 512,
686, 692, 700, 724 (Pis. XXVI, LIV).
E. princeps Osborn, 4902 (referred?), fragments of pes.
E. borealis (Cope), 4892 (type), superior molars p<-m^ radius,
etc. (figs. 102, 497, 498; PI. LIV).
E. borealis (Cope), 14887 (neotype), skull, jaw, atlas, pelvis,
etc. (figs. 229, 232, 244, 250, 251, 494, 495, 501, 515, 721; PI.
LIV).
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
289
E. borealis (Cope), 14890, portions of right and left jaws.
E. borealis (Cope), 14891, complete jaws (figs. 231, 248, 249).
E. borealis (Cope), 4886, anterior portion of jaw (figs. 246,
249).
E. borealis (Cope), 14895, caleaneum, astragalus (reference
doubtful) (figs. 503, 505, 522).
E. borealis (Cope), 14888, jaws and fragments of skeleton
and feet, right pes (figs. 246, 249, 494, 501-503, 521, 701).
E. brownianus (Cope), 4885 (type), jaw fragment (figs. 104,
231, 246, 247).
E. gregoryi Osborn, 14889 (type), jaws, also m'-s (figs. 142,
231, 245-247; PI. LIV).
E. gregoryi Osborn, 14933 (referred) , portions of pes and tibia
(fig. 503).
The specimens listed above are arranged not in the
ascending geologic order but according to size, E.
gregoryi being the smallest and E. major the largest
Manus numerically tetradactyl but functionally tri-
dactyl, with a tendency to mesaxonic structure.
This animal is separated generically from Larnbdo-
therium by the possession of full eutherian dentition,
including p|. The fact that the face is longer than
the cranium constitutes its principal generic distinction
from the middle Eocene titanotheres (fig. 256). As
has been shown above, the genus is represented by
five specific stages or mutations, which are distin-
guished partly by size but more clearly, at least in
four species, by the development of cuspules on the
inferior premolar teeth, as follows :
E. major Osborn, distinguished only by its large size.
. E. princeps Osborn, distinguished by size and premolar com-
plication.
Figure 243. — Restoration of Eotitanops borealis, of the Wind River formation (Eocene)
About one-twelfth natural size. Made by E. S. Christman in 1917 under the direction of W. K. Gregory.
of the animals represented by these types and other
specimens. The specific reference of the separate
foot bones (Nos. 14893, 4902, 14895, 14933) is doubtful.
Eotitanops Osborn
Plates XXVI, LIV; text figures 10, 21, 25, 27, 28, 29, 33 142-
145, 155, 210, 212, 219, 229-232,243-253,405-408,482-485,
490, 492-503, 505-507, 512, 515, 521-523, 646, 648, 649,
661, 686, 688, 690, 692, 694, 695, 700, 701, 704, 709, 711, 717,
721-727, 733, 740, 742, 745
[For original description and type references see p. 179; for skeletal characters see
p. 591]
Generic characters. — Skull of proopic dolichoceph-
aly. Incisor series obliquely anteroposterior. Pj
with small, compressed single fang; p^-p* with single
internal cusps; p'-p* with rudimentary lophoid proto-
conules; p4 rather progressive. Superior molars sub-
quadrate and rounded in form; protoconules small;
metaconules wanting or rudimentary; inferior molars
without metastylids; hypoconulid of ms subconic.
E. borealis (Cope), distinguished by intermediate size and
premolar simplicity.
E. brownianus (Cope), distinguished by smaller size and
premolar simplicity.
E. gregoryi Osborn, distinguished by the smallest size and
extreme premolar simplicity.
The range of measurement in the species and muta-
tions is shown in the following tables :
Comparative measurements of Eotitanops, in millimeters
[All specimens in American Museum]
Pj-ma
Mtc III
Mts III,
E. major, 14894 (type) .
103
E. princeps, 296 (type)
E. borealis, 14891 (referred),
E. borealis, 14890 (referred ).
E. borealis, 14888 (referred) _
E. brownianus, 4885 (type).
E. gregoryi, 14889 (type) — -
E. minimus, 17439 (type) . .
- 105
98
96
94
90
78.4
» 72
87
290
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Standard measurements of teeth of the species of Eotiianops, in
millimeters
[The numbers are those of specimens in the American Museum of Natural History]
'S.D,
II
f4
E. borealis
1
i
1
i
i
1
a-s-
is
Pi-m'
108
92
P2-m3
1
P2-p^
36
M'-m'-
35.5
54
17
16
18
20
19
P\ ap
11
14.6
18
18.3
P«, tr .-
M', ap
MS tr .-.
M2, ap
14
15 5
W, tr _
MS, ap- .
13 3
18
M3, tr
17
22
Pz-ms '78
90
35
55
— -
94
36
58
98
38
59
96
36
60
11. 5
6
12
6.5
13
7.5
-105
P2-P4
29.4
49
8.8
6.2
9.5
5.2
39
66
P2, ap. -
12
6
12
7
12.5
8
- —
13
P2, tr .
63
P3, ap
125
P3, tr_
63
P4, ap- -. _
12
8
15.5
10
18
11.7
22
11.5
16.3
11
19
13
23.2
P,, tr
Ml, ap_ _
14.5
8.5
15.6
10.5
19.5
10. 7
183
Mi, tr .
12
Mj, ap
21
M2, tr
14
23
— -
25
M3, tr
14
Figure 244. — Skulls of the oldest known titanotheres
Reconstructions by L. M. Sterling under the direction of W. K. Gregory. One-
fourth natural size. A, Lamhdotherium popoagicum, Am. Mus. 14907, Alkali
Creek, Wolton; 14899 and 14903, Alkali Creek, Buck Spring. B, Eotiianops bo-
realis, Am. Mus. 14887, Dry Muddy Creek 12 miles above mouth. All specimens
from the Wind River Basin, Wind River formation.
Range of evolution. — As Eotitanops gregoryi, the
smallest and simplest form, occurs on a high level,
having been found 100 feet above the "red stra-
tum" on Alkali Creek (see figs. 47, 48), and as
specimens referred to E. horealis and E. princeps
range from the "red stratum" on Alkali Creek, on
the 250-foot level, to the 400-foot level, these species
and mutations can not be arranged in monophyletic
succession, but they afford evidence that even at this
time the titanotheres were polyphyletic.
Range in size. — The smallest of the Wind River
titanotheres, E. gregoryi, measures about 18J^ inches,
or 45.6 centimeters, at the shoulders. A larger form,
E. princeps, measures about 26 inches, or 66 centi-
meters, at the shoulders. The intermediate form,
E. borealis, is more slenderly proportioned than the
American tapir (T. terrestris); it is between 75 and
78 per cent of the height of the tapir, and thus
about 75 per cent of the height of Mesatirhinus of
the upper levels of the Bridger Eocene. E. major,
judged only by the size of the pes, more nearly
approaches T. terrestris in size, the median metatarsal
of E. major measuring 103 millimeters and that of T.
terrestris 108.
It should be noted that Eotitanops includes the only
known large lower Eocene perissodactyl. Even
Eotitanops major, the largest Wind River species,
appears to be considerably smaller than Palaeosyops
jontinalis, the smallest Bridger species.
Measurements of upper teeth of Eotitanops horealis and Palaeo-
syops fontinalis, in millimeters
M', anteroposterior
M', transverse
M', ectoloph, maxiUa
M', transverse maxilla
M^, transverse maxilla (pr.-pas.)
Length of left zygoma (anterior bor-
der malar to posterior border post-
glenoid process)
E. borealis
from Wind
River B, Am.
Mus. 148S7
17.5
"18.5
19.5
23
25.5
» 129
P. fontinalis
from Bridger
A (type). Am.
Mus. 5107
137
The measurements given show that in its dentition
P. Jontinalis of Bridger A was much larger than
E. horealis of the Wind River formation. The rela-
tively small size of the zygoma in the type of P. fonti-
nalis is consistent with the fact that the animal was
very young, its milk dentition being still functional.
A comparison of E. horealis (summit of known lower
Eocene) with P. fontinalis (lower middle Eocene or
lower Bridger) indicates a long period of titanothere
evolution between these two species. P. fontinalis,
although the oldest known Bridger titanothere, differs
in two points — the superior dental series is 25 per
cent larger than that of E. horealis; the cranium is
elongate and the face abbreviate.
EVOLUTION OP THE SKULL AND TEETH OF EOCENE TITANOTHERES
291
Eotitanops gregoryi Osborn
Plate LIV; text figures 25, 27, 33, 142, 143, 231, 245-247,
253, 483, 492, 493, 503, 661, 726, 727, 742
[For original description and type references see p. 192]
Type locality and geologic Jiorizon.- — Alkali Creek,
Buck Spring, Wind River Basin, Wyo.; Wind River
This very sharply defined species is named in honor
of Dr. William K. Gregory. Its especial interest lies
in the fact that it is the most primitive titanothere
known. It represents, however, a persistent primitive
stage, because its geologic level, 100 feet above the
alkali "red stratum," is higher than that of the
Figure 245. — Model of skuU of Eotitanops gregoryi
Based on type specimen (Am. Mus. 14889) and on Eotiiamps iorealis. One-half natural size.
formation, Lambdotherium zone, horizon Wind River B
("Lost Cabin"), 100 feet above heavy "red stratum."
Specific characters. — -Very primitive and of inferior
size, p2-m3, 78.4 millimeters; mi_3, 49; P2-3 with the I (See fig. 246.)
typical and relatively progressive E. borealis. Its
primitive condition is apparent in the comparison of
P3 with the same tooth in E. borealis and E. princeps.
FiGTJBE 246. — Lower premolars and molars of Eotitanops
Natural size. American iVEuseum specimens from the Wind River formation. Wind River Basin. A, B, C, Inner side view of the
third left lower premolar: A, E. gregoryi, No. 14889 (type), Alkali Creek, Buck Spring, upper level of "Big Red Pocket," 100
feet above heavy red stratum; B, E. iorealis, No. 14888, Alkali Creek Davis's ranch; C, E. princeps, No. 296 (type). D, E. gregoryi,
No. 14889 (type), left lower premolars (p2, ps), inner side view. E, E. brownianus. No. 4885 (type), second left lower premolar,
inner side view. F, E. borealis, No. 14891, west bank of Wind River, 3 miles above canyon (top of banded beds); left lower
molars (mi, ma) , inner side view.
internal cusps, paraconid and metaconid, consisting of
rectigradations in a most rudimentary stage; hypo-
conulid of m3 very small; m^ with a single internal
cone, no hypocone.
This third inferior premolar, ps, is much less pro-
gressive than in E. princeps or even in Lambdotherium;
the other premolars are correspondingly primitive, p2
short, compressed, with a very rudimentary hypo-
292
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
conid, Pa laterally compressed, hypoconid distinct,
paraconid, metaconid, and entoconid extremely rudi-
mentary rectigradations. In the molar teeth, mi_3,
the metastylid and entostylid are also in an extremely
rudimentary or rectigradational stage. In ma the
hypoconulid is small, subconic, external in position.
(See fig. 235.)
Eotitanops brownianus (Cope)
Text figures 104, 143, 231, 246, 247, 253
[For original description and type references see p. 169]
Type locality and geologic horizon. — Wind River
Basin, Wyo.; Wind Eiver formation, Lambdotherium-
c-fSS^'
Figure 247. — Lower jaws of Eotitanops gregoryi and E. brownianus
One-half natural size. A, E. gregoryi, Am. Mus. 14889 (type), reversed; Alkali Creek, Buck Spring,
upper level ol "Big Red Pocket"; Wind River formation, upper part (Wind River B, "Lost
Cabin"). Contours partly restored from E. borealis. B, E. brownianus, Am. Mus. 4885 (type),
reversed; Wind River Basin. Contours partly restored from E. borealis and E. princeps.
Eotitanops- Cory phodon zone (Wind River B), exact
level not recorded.
Specific characters. — Size greater than E. gregoryi;
P2-m3, 90 millimeters; mi_3, 55; fang of pi placed in
close proximity to the canine; p2 compressed, hypo-
conid distinct, elevated, entoconid invisible, paraconid
and rudimentary rectigradations placed very low on
the crown, metaconid extremely rudimentary if pres-
ent; metastylid rudimentary.
P2 (see fig. 246) is in a less advanced stage of
evolution than ps in E. gregoryi.
As shown in the comparative series of the jaws
(fig. 231), in the table of measurements (p. 290; see
also fig. 483), and in the accompanying figures, the
type of this species belonged to an animal in size
midway between E. gregoryi and E. borealis. The
ramus of the jaw rather resembles that of E. borealis
but with a pronounced swelling below ms; its vertical
depth below the anterior face of ms is 40 millimeters;
the symphysis is decidedly broad and massive.
Eotitanops borealis (Cope)
Cf. Palaeosyops borealis Cope
Plate LIV; text figures 10, 28, 29, 102, 143, 219, 229-231, 243,
244, 246, 248-251, 405, 406, 482, 493-495, 497, 498, 501-503,
507, 515, 521-523, 646, 648, 649, 690, 694, 700,
/ 701, 717, 721, 724, 725, 745
(For original description and typa references see p. 168]
Type locality and geologic Twrizon. —
Wind River Basin, Wyo.; Wind River
formation, Lambdotherium- Eotitanops-
Coryphodon zone (Wind River B, "Lost
Cabin").
Specific characters. — Of larger size; p2-m3,
94-98 millimeters; premolar teeth more
complicated, as shown in neotype and
associated specimens; p2 with very, rudi-
mentary paraconid and metastylid; p^~*
with progressively developing tritocones
and single internal deuterocones back-
wardly inclined, crowns sub triangular;
m'~^ with distinct protoconules.
Materials. — The fragmentary type speci-
men is the historical Palaeosyops borealis
(Am. Mus. 4892) of Cope, figured in the
"Tertiary Vertebrata," Plate LVIII, A,
Figures. It is marked No. 16 in the Wind
River valley collection of J. L. Wortman,
July, 1880. The very fine specimen se-
lected as a neotype (Am. Mus. 14887, figs.
250, 251) consists of the skull and jaws
found by Granger in 1909 on Dry Muddy
Creek, 100 feet above the alkali "red
stratum," and represents a slightly larger
and somewhat more progressive mutation.
Incisors of neotype. — The incisors show the char-
acteristic titanothere feature of increase in size from
i' to i^, the transverse measurement of the crowns
being respectively i' 6 millimeters, i^ 6, i^ 8 (estimated).
The crowns of i' and i^ are bluntly spatulate or chisel-
shaped. P has a faint antero-internal cingulum; i^
is rounded and subcaniniform. The general arrange-
ment of the series is obliquely anteroposterior rather
than transverse. The canine is prominent, laterally
compressed, the alveolus measuring, transverse, 13
millimeters (estimated); anteroposterior, 17.
EVOLUTION OF THE SKULL AND TEETH OP EOCENE TITANOTHERES
293
Premolars of neotype. — P' is placed midway between
the canine and p^ consistently with the relatively
elongate preorbital region. The chief features of p^^*
are the simple, backwardly directed deuterocones
with low crests connecting them with the protocones
and tritocones; the tritocones (see figs. 229, 250) in-
crease progressively in p'"^; external cingula faintly
indicated in p^ p*. The three premolars taken to-
gether are subordinate in measurement (36 mm.) to
the molars (54 mm.).
Superior molars oj neotype. — The superior molars
exhibit the characteristic
bicrescentic ectoloph with
prominent parastyles and
mesostyles, and median ridges
opposite the paracones and
metacones (PL LIV, fig. 229);
the protoconules are fairly
prominent and faintly cres-
centic in m'~', forming a ves-
tigial protoloph; the internal
cingula festoon but do not sur-
round the inner sides of the
crown; m^ entirely lacks the
hypocone; protolophs distinct
but sessile are observed on
m'"^, also faint rudiments of
metalophs on m\ m^
Lower molars of referred speci-
mens.— The valuable series of
jaws (Am. Mus. 14887, 14890,
14891, 14888, and 4886) com-
plete our knowledge of the
inferior dentition except the
incisors, which are unknown
(figs. 248-250). The premolar-
molar series, p2-m3, exhibit
progressive gradations of
length from 94 to 98 milli-
meters. (See table on p. 290.)
They are thus superior to E.
irownianus and inferior to E.
princeps in measurement. The
premolars afford the distinctive
specific characters or muta-
tions in the progressive stages
of the internal cuspules or rectigradations.
Besides the somewhat arbitrary association of the
type and neotype, we also refer to this species the
materials listed above, including a number of jaws and
portions of the skeleton. (See figs. 231, 246, 248, 249,
494, 501-503, 505, 521, 522, 701.)
Characters of the teeth. — The fourth superior pre-
molar (Am. Mus. 4892; PI. LIV, H) measures antero-
posteriorly 12 millimeters, transversely 14; it exhibits
a faint external, distinct anterior and posterior, but no
internal cingula, conical deuterocone, small proto-
conule and larger convex protocone, a somewhat
smaller and more plane tritocone and small meta-
conule ridge. The superior molars exhibit faint
external, more prominent anterior and posterior, and
incomplete internal cingula; the ectoloph consists
of sharply defined parastyle, paracones and meta-
cones with median external ridges, and a prominent
mesostyle; the most distinctive feature of the inner
half of the crown in m' is the sublophoid character of
the protolophs and hypolophs, which unite respec-
FiGUEE 248. — Lower jaw of Eoiitanops borealis
One-half natural size. Am. Mus. 14891; west bank of Wind River, 3 miles above canyon (top of banded beds); Wind
River formation. Ai, Outer side view; A2, inferior view.
tively with the distinct protoconule and a much less
distinct metaconule to form a low or sessile crest.
This rudimentary or vestigial lophoid character is
even less evident in the middle Eocene species of
titanotheres. M' measures 18 by 17 millimeters
(ap. by tr.); it is a nearly quadrate tooth, in wide
contrast to the transversely expanded tooth of L.
popoagicum. In m^ the less worn paracone measures
7 millimeters in height ; the ectoloph is thus somewhat
elevated in these molars, but its crescents are not
294
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
strongly concave and incurved as in the Bridger
titanotheres; the protoconule is distinct, the meta-
FiGURE 249. — Lower teeth of Eotitanops borealis
Natural size. A, Crown view of left lower premolars and molars (P2--m3); Am. Mus. 1488S; Alkali
Creek, Davis's ranch, Wind Hiver. B, Crown view of molars (mj-ms); Am. Mus. 14891; west
bank of Wind River, 3 miles above canyon (top of banded beds). C, Crown view of premolars
Cp!-P4); Am. Mus. 4886; Wind River valley, Wind River formation.
Figure 250. — Skull of Eotitanops borealis
One-fourth natural size. Am. Mus. 14887, Dry Muddy Creek 12 miles above mouth;
Wind River formation. Ai, Palatal view of crushed specimen; Az, As, recon-
struction of the palatal and side views of the same skull made by L. M. Sterling
under the direction of W. K. Gregory.
conule is faint. In m' similar characters are observed
on the anterior half of the crown; the posterior half
is broken away.
The lower molars exhibit low brachyodont crowns,
the crescentic external cusps alternating with the
subconic internal cusps; extremely rudi-
mentary hypoconulids, metastylids, and
vestigial paraconids are observed; external
cingula rudimentary, internal cingula en-
tirely wanting, as in all titanotheres; hypo-
conulid of m^ central, small, sublophoid.
Slcull.—The discovery of the skull of E.
borealis (Am. Mus. 14887, neotype) was an
important event in the work of determining
the morphology of the titanotheres because
it connected these mammals closely with
other early Eocene perissodactyls and sepa-
rated them from the middle Eocene forms.
The chief feature of the
skull is that the proopic or
facial region is longer than
the opisthopic or cranial
region, whereas in all the
middle Eocene titanothere
skulls yet known the face is
shorter than the cranium
and becoming progres-
sively shorter throughout
Eocene and lower Oligo-
cene time. The skull is
also relatively long and
narrow, and the true molar
series is relatively short as
compared with the total
length of the skull. These
characters are well shown
in the reconstruction of the
skull (figs. 250, 251) and
in the model of the head
(figs. 646, 648, 649); they are
expressed in the following
indices, which are estimates
only, because the skull is
considerably crushed :
Cephalic index 50 (vs^idth
across zygomata -=- basal
length = 160 millimeters h- Figure 251. — Skull of Eotitan-
313 [estimated]).
Faciocephalio index 56 (length
of face -;- basal length =
185-^313).
Molar index 17 (length m"-
m^ -^ basal length of skull =
54 -H 313).
These fundamental proportions give the skull of
Eotitanops (fig. 250) a striking resemblance to that
of other lower Eocene perissodactyls. The type is
technically known as proopic dolichocephaly.
Attention may be called to the following details:
(1) Premaxillaries slender, symphyseal union very
slight, indicating feeble use of superior incisors, pre-
maxillaries joining nasals superiorly, a primitive fea-
ture; (2) infraorbital foramen placed above p'-p*,
ops borealis
A I, Top view; A2, occipital view. One-
fourth natural size. Am. Mus. 14887,
Dry Muddy Creek 12 miles above
mouth; Wind River formation. Re-
construction made by L. M. Sterling
under the direction of W. K. Gregory.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
295
widely separate from orbit (closer to orbits in Eocene
forms); (3) malars gently rounded, and zygomata
moderately projecting; (4) superior profile believed
to be plane or gently convex, slightly convex above
the orbits; (5) greatest width of skull opposite glenoid
fossae; (6) temporal fossae deep, brain case small, sur-
mounted by high, thin parietal crest; (7) occipital
crest overhanging condyle superiorly, deeply indented
in median line; (8) postglenoid and post-tympanic open
below auditory meatus; (9) small exposure of the
mastoid between the post-tympanic and paroccipital
process; (10) in occipital view (fig. 251) the parietal
crest is narrow and flaring superiorly; (11) in palatal
view basioccipital and basisphenoid keeled or com-
pressed; (12) typical perissodactyl foramina separate^
namely, condylar, lacerum medium and posterius,
ovale, and alisphenoid; (13) posterior nares deeply
inclosed by pterygoids and
pterygoid wings of alisphe-
noids; (14) posterior borders
of palatines not preserved;
(15) palate relatively elon-
gate, narrow and arched
from side to side; (16) post-
glenoid process narrow,
internal in position ; (17)
occipital condyles sharply
convex, prominent, sepa-
rated in median line.
ing behind ms; symphysis moderately elongate, gently
convex, incisive alveoli, indicating progressive increase
of size from ii to is and semiprocumbent position of the
incisors.
Eotitanops prlnceps Osborn
Plates XXVI, LIV; text figures 27, 33, 143, 144, 231, 246, 252,
407, 408, 483, 484, 490, 492-494, 496, 498-500, 512, 661,
686, 692, 700, 704, 709, 724
[For original description and type references see p. 193. For sljeletal characters
see p. 690]
Type locality and geologic
horizon. — Wind River Basin,
Wyo. ; Wind River formation,
Lambdotherium-E otitanops-
CorypJiodon zone (Wind River
B, "Lost Cabin").
Measurements of Eotitanops borealis and E. princeps, in milli-
meters
Basilar length, premaxillaries to con-
dyles (estimated)
Zygomatic or transverse width (esti-
mated)
Width across occipital condyles
Cranial length, postorbital process to
occipital condyles
Facial length, postorbital process to
maxillary symphysis
Length of lower jaw, symphysis to
condyles (estimated)
Height of jaw, condyle to bottom of
angle
Lower jaw, depth behind ma
E. borealis.
Am. Mus.
14887 (neo-
type)
E. princeps,
Am. Mus.
296 (type)
The jaws are well displayed in the neotype (Am.
Mus. 14887) and in the referred specimens, especiaUy
in the well-preserved jaw shown in Figure 248 (Am.
Mus. 14891).
The chief characters are the following: Ramus
elongate, gently convex in vertical section, expanding
toward symphysis; lower border suddenly compressed
and descending below angle, thin posterior border;
delicately retroverted coronoid, ramus slowly ascend-
101959— 29— VOL 1 22
riGTTHB 252. — Lower jaw of Eotitanops princeps
One-halt natural size. Am. Mus. 296 (type), reversed. Wind River Basin; Wind River formation.
Specific cJiaracters. — Of still larger size; Pa-ms, 105
millimeters (estimated). Inferior premolar teeth some-
what more complicated, as shown in the type specimen;
P2 with elevated, distinct, but very rudimentary para-
conid and metastylid, entoconid very rudimentary if
present, talonid narrow; ps, paraconid quite distinct,
elevated, metastylid small, distinct, entoconid rudi-
mentary, talonid broad; p4 submolariform, talonid
broad, entoconid shelf distinct. Hypoconulid of ma
rounded, more robust. Ramus larger and more robust.
The more advanced development of the premolar
rectigradations, the increased size of the teeth and of
the jaw, the larger size of the hind feet in the referred
specimen (Am. Mus. 4902) combine to distinguish
this specimen as a mutation or subspecific stage
between E. borealis and E. major.
Lower jaw of type. — The weU-preserved jaw (fig.
252) of the type specimen (Am. Mus. 296) measures
253 millimeters from the back of the condyle to the
symphysis, 99 from the condyle to the bottom of the
angle, and 53 vertical depth of the ramus just behind
ms. Its distinguishing features are (1) the elevation
of the condyle above the grinders; (2) the rather
slender, recurved coronoid with sharply angulated
and flattened anterior border, which reminds us of the
coronoid of the middle Eocene Mesatirhinus and
DolichorJiinus rather than of that of Palaeosyops;
(3) the well-defined superior fossa between the angle
and the coronoid; (4) the depressed or delicate incurved
513
162
52
128
245
97
296
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
posterior border of the angle; (5) the elongate (70
mm., estimated) symphysis laterally compressed
l)ehind the canines; (6) the slope of the anterior border
of the coronoid directly into the fang of ms; (7) the
moderately thick rami (18 mm.). The lower profile
or contour of the jaw is convex below the molars,
concave below the coronoid, extending backward into
the angle.
Inferior teeth oj E. princeps {type; PI. LIV). — There
were apparently three inferior incisors, the crowns of
which are not preserved. The alveoli of the inferior
canines are slightly compressed laterally; the esti-
mated measurements are 15 millimeters (antero-
posterior) by 13 (transverse). The crown of pi is
not preserved; its fang is single; the fang is separated
from that of the canine by a very narrow diastema
Figure 253. — Lower grinding teeth of three species of Eotitanops from
the upper Huerfano formation (Huerfano B)
Natural size. After Osborn, 1919. A, A', E. minimus (type), lower level of the upper
horizon of the Huerfano formation; B, Bi, E. gregoriji (referred specimen), from the
upper Huerfano; C, E. hrownianus (referred specimen), from near the base of the
lower Huerfano.
(3 mm.); behind it is a continuation of the diastema,
16 millimeters in width. This diastema points to a
somewhat elongate character of skuU and jaw, since
the total length between the canine and p2 is 31
millimeters. P2, measuring 13 by 6 millimeters, is
an elongate, laterally compressed, bifanged tooth with
an elevated paraconid and depressed metastylid or
posterior cusp, noncingulate and with the faintest
indication of valleys on the inner surface. P3, meas-
uring 12 by 7 millimeters, is slightly more progressive,
with its metaconid externally placed and a more clearly
indicated posterior valley. P4, measuring 13 by 8
millimeters, exhibits a broader talonid and is thus
submolariform. The molars are perfectly pi'eserved.
aggregating 65 millimeters in length, 14 in maximum
width of crown. The individual total measurements
(ap. by tr.) are as follows: Mj, 17 by 12 millimeters;
m2, 21 by 13; ma, 26 by 14.
This progressive increase posteriorly accords with
a similar increase of the upper molars posteriorly, as
observed also in Larnhdotherium. The inferior molars
exhibit faint external and no internal cingula; rather
low but well-defined crescents; a progressive increase
in size; paraconids partly defined on ms. The most
distinctive primitive feature in ma is the small size,
subconic form, and mesial position of the hypoconulid
as compared with its backward extension and cres-
centic form in some of the middle Eocene types.
Eotitanops major Osborn
Text figures 145, 483, 506
[For original description and type references see p. 193. For skeletal
characters see p. 697.]
Type locality and geologic Tiorizon. — Alkali
Creek, Wind River Basin, Wyo.; Wind Eiver
formation, Lamidotherium- Eotitanops - Cory-
;., phodon zone (Wind River B, "Lost Cabin").
7 -•iMNoi744i Specific characters. — The type and only known
specimen (Am. Mus. 14894) consists of a left me-
dian metatarsal associated with the distal end of a
tibia. It is distinguished from E. princeps by its
notably larger size (length of Mts III, 104 mm.,
' greatest width, 16). The skull and dentition are
not known.
Eotitanops minimus Osborn
Text figures 155, 253
[For original description and type references see p. 199.]
Type locality and geologic horizon. — Huerfano
Park, Colo.; Huerfano formation, Eometarhinus-
j_ Palaeosyops fontinalis zone (Huerfano B; lower
1 level).
Specific characters. — As this is the smallest true
titanothere known, Osborn (1919.494, p. 564) as-
signed to the type lower molar teeth P4-m3 (Am.
Mus. 17439) the specific name minimus. The
measurement of p4-m3 (53 mm.) is much less than
that (58 mm.) of the corresponding teeth in E.
gregoryi, yet the other characters are so similar to
those of E. gregoryi as to suggest that this is a related
form. Figure 253 exhibits the form and size of three
species, minimus, gregoryi, and hrownianus. A large
number of measurements of Eocene titanotheres show
that no single species exhibits so great a range in size.
The discovery of this dwarf titanothere, together
with the presence of titanotheres of the same size as
E. gregoryi and E. hrownianus in Huerfano B and
Wind River B, reveals the existence of what is probably
a distinct phylum of diminutive titanotheres separable
from the Eotitanopinae. We must, however, await
the discovery of the skeletons before this supposition
can be confirmed.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
297
SECTION 4. THE MIDDLE AND UPPER EOCENE
TITANOTHERES
PHYIA DISTINGUISHED
Some of the middle Eocene titanotheres represented
in the lower Bridger beds may have been evolved from
forms related to the Eotitanopinae of the Wind River
formation. A geologic interval covering a long period
(including Huerfano B= Bridger A) separates the
titanotheres of the Wind River B from those of
Bridger B, and during this period there was a marked
transformation in the proportions of the head, for in
the titanotheres of Wind River B the face is longer
than the cranium (dolichopic), whereas in those of
Bridger B and succeeding subdivisions the cranium is
longer than the face (brachyopic).
In the titanotheres now to be described this change
in faciocranial proportions probably occurred during
the deposition of Bridger A and Huerfano B. In the
10 or 12 genera of titanotheres of the middle and upper
Eocene the cranium is longer than the face. These
animals fall broadly into two large groups, which are
more or less theoretically subdivided (1917) into two
groups and into six chief phyla or lines of descent as
shown in the accompanying table.
Characteristic features of groups of titanotheres
[Compare flg. 219, p. 265]
Palaeosyopine group: Palaeosyops, Limnohyops, Telmatherium, Sthenodectes
Manteoceras-Dolichorhinus group: Manteoceras, MesatirUnus, Dolichorhinus,
Metarhinus, Ehadinorlimus, Diplacodon
Skull brachycephalic to mesaticephalic.
Horn rudiments retarded in evolution.
Occiput rounded or high.
Zygomata deepened vertically.
Canines more pointed, erect.
Third superior incisor oaniniform.
1. Subfamily Palaeosyopinae "(Limnohyops, Palaeosyops), ex-
tremely brachycephalic.
2. Subfamily Telmatheriinae (Telmatherium), mesaticephalic
to brachycephalic.
Skull mesaticephalic to dolichocephalic.
Horn rudiments precocious in evolution.
Occiput primitively low and broad.
Zygomata shallow vertically.
Canines more obtuse, recurved.
Third superior incisor incisiform.
3. SubfamilyManteoceratinae = Brontopinae (Manteoceras, Pro-
titanotherium, Brontops), progressively brachycephalic.
4. Subfamily Dolichorhininae (Mesatirhinus, Metarhinus, Doli-
chorhinus) , mesaticephalic to dolichocephalic, facial region
downturned.
?5. Subfamily Megaceropinae=?Rhadinorhininae (?Rhadino-
rhinus, Megacerops), mesaticephalic, facial region up-
turned.
6. Subfamily Brontotheriinae=?Diplacodontinae (Diplaco-
don, Brontotherium) , horns precociously developed.
The extreme forms of the two groups — namely,
Palaeosyops and Dolichorhinus — also contrast widely
in the detailed characters of the skull, as shown in
longitudinal and cross sections in Figure 254.
The subfamilies 1-6, according to the Osborn system
(see Chap. I), correspond with the phyla, or vertical
lines of descent, which have been established among
the Eocene titanotheres, also between the Eocene
and Oligocene titanotheres. Thus it is now Icnown
that Manteoceras and Protitanotherium are related to
Brontops of the Oligocene. Diplacodon of the upper
Eocene is of uncertain affinities with the lower Oligo-
cene genera. It is possible but by no means demon-
strated that Rhadinorhinus is related to the Oligocene
brontotheres and Megacerops, as suggested by Gregory.
SPECIES OF PALAEOSyOPINAE AND DOLICHORHININAE
FROM THE UPPER HUERFANO (TROGOSUS ZONE)
The discovery of two very distinct phyla of true
titanotheres in the lower Eocene confirms the theo-
retic separation of the titanotheres into subfamilies
as occurring in lower Eocene time. In the Huerfano
formation we have evidence of two subfamilies, as
follows :
Palaeosyopinae (perhaps derived from
the Eotitanopinae)
Hornless.
Slender nasals.
Subbrachycephalic.
More robust proportions.
Dolichorhininae (Manteoceras-Me tarhi
nuS'MesaUrkinus-VolichOThinus grouTp)
Osseous horn rudiments at
nasofrontal junction.
Nasals very stout, laterally
decurved.
Mesaticephalic.
Smaller proportions.
The first subfamily is represented by numerous
specimens of Palaeosyops fontinalis Cope; the second
group is represented by the single type specimen of
the new genus Eometarhinus {E. huerfanensis) .
SYSTEMATIC DESCRIPTIONS OF THE MIDDLE AND UPPER
EOCENE TITANOTHERES
THE PALAEOSYOPINE GROUP (PALAEOSYOPS, LIMNO-
HYOPS, TELMATHERIUM, STHENODECTES)
DISTINCTIVE FEATURES AND GEOLOGIC HORIZONS
Osborn finally included the genus Telmatherium
within the palaeosyopine group, although there are
some grounds for placing it closer to the Manteo-
ceratinae. The telmatheres appear to have had a
long and independent evolution of their own (see fig.
697) and thus constitute the distinct subfamily Tel-
matheriinae.
298
TITANOTHEEES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The resemblances and contrasts between the three I proportional and numerical evolution explained on
chief genera included in these two subfamilies are in- j pages 251-262.
dicated in the following manner, on the principles of i
Proportional and numerical characters oj titanotheres of the palaeosyopine group
Telmatherium
Skull and skeleton very robust; skuU de-
cidedly broad and massive; feet short.
Fore feet short, more paraxonic, the fifth
digit larger.
Skull rounded, occiput stout, sagittal
crest of medium length, forehead pro-
tuberant or convex.
Jaws robust, chin prominent, angulate;
mandibular rami massive below grind-
ing teeth; coronoid at base very broad
and concave anteriorly.
Premaxillary symphysis short and
rounded.
Maxillary splint on side of malars; malar
section rounded.
Incisors more transverse; canines sub-
lanceolate to round; premolars com-
pressed anteroposteriorly; molars broad
or quadrate, with strong, rounded para-
styles.
Conules on superior molars more or less
persistent, rounded.
Grinders persistently brachyodont.
Last superior molar usually without hypo-
cone or second postero-internal cusp;
crown subtriangular, rounded.
Ectolophs of superior premolars in some
specimens resembling those of molars
(that is, with mesostyles).
Skull and skeleton more slender; skull
broad, brachycephalic, less massive;
feet narrow.
Fore feet short, more mesaxonic, with the
fifth digit reduced.
A more elevated occiput, higher and
thinner sagittal crest; forehead con-
cave, without protuberance.
Jaws somewhat more slender, chin slop-
ing, rami less massive below grinders,
anterior face of coronoid less broadened.
Premaxillary symphysis rounded.
Maxillary splint extending from side to
beneath malars; malar section de-
pressed.
Incisor series obliquely placed; canines
slightly more compressed and ridged;
molars with ridged parastyles and
ridged conules.
Conules on superior molars persistent,
ridged or lophoid.
Grinders persistently brachyodont.
Last superior molar usually with a dis-
tinct hypocone; crown more quadrate.
Skull and skeleton rather slender; skull
decidedly elongate, dolichocephalic.
Fore feet long, more mesaxonic, the fifth
digit elongate.
Occiput very high; sagittal crest elongate;
forehead plane, no protuberance.
Jaws more slender; chin deep; symph3'sis
elongate.
Premaxillary symphysis elongate.
Maxillary splint elongate, extending be-
neath malars; malar section rectangular.
Incisor series more parallel; canines high,
sublanceelate; premolars elongate; mo-
lars narrow, more sharply crested or hyp-
sodont, with feeble parastyles or none.
Conules on superior molars disappearing
early.
Grinders progressively hypsodont.
Last superior molar without hypocone;
crown quadrate.
SUBFAMILY PAIAEOSYOPINAE (STEINMANN AND DODEELEIN)
The Palaeosyopinae consist of the Limnohyops and
Palaeosyops generic phyla. They were abundant
chiefly in lower and middle Bridger time, beginning
to decline in upper Bridger time. Limnohyops is sub-
brachycephalic to brachycephalic, mediportal; Palaeo-
syops is brachycephalic to hyperbrachycephalic,
graviportal. They were larger than tapirs, propor-
tions stout, becoming graviportal; feet of brachypodal
type; skuU broad, progressively brachycephalic, facial
region abbreviate, nasals tapering distally, nasofrontal
horns retarded in development; cranial region and
zygomata broadening; grinding teeth persistently
brachyodont; canines stout, subconical.
They make their appearance at the base of the
Bridger or in Bridger A, in the species Palaeosyops
fontinalis, and the last member known is the species
Palaeosyops copei, of Bridger D or Washakie A. The
Palaeosyopinae thus formed the first titano there sub-
family to appear in the middle Eocene and also the
first, so far as known, to disappear geologically.
The subfamily name Palaeosyopinae is taken from
the name of the classic genus Palaeosyops leidyi, the
first Eocene titanothere discovered. These titano-
theres are broad-headed, chiefly of lower and middle
Bridger age, reaching a climax and beginning to decline
in upper Bridger time. The cranial region of the skull
is longer than the facial region; the head is short and
broad (brachycephalic); the horns are relatively late
or retarded in development; the feet are short and
broad (Palaeosyops), or less broad (Limnohyops).
The two phyla, Palaeosyops and Limnohyops, were
contemporaneous, their remains being found in the
same deposits.
SEPARATION OF PAIAEOSYOPS AND IIMNOHYOPS GENEKIC PHYIA
In the middle Eocene of the Bridger region in west-
ern Wyoming the animals known as Palaeosyops and
Limnohyops are the earliest to occur geologically —
namely, in Bridger A, B, and C. They were browsing
animals, with short-crowned teeth and broad heads,
which increase in breadth in the successive descendants
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHEEES
299
of the original forms. They exhibit many features in
common, yet they seem to represent two distinct
contemporaneous phyla. Of these two phyla Lim-
nohyops is the more primitive; it is in many features
more central or intermediate than the second phylum,
ferent lines of descent, the ancestral members of
each line (Bridger A and B) are not easily distin-
guished.
The abundance of remains of Palaeosyops is wel-
comed by the student of evolution because so many
Figure 254. — Sections of skull of brachycephalic (A, B) and dolichocephalic (C, D) Eocene titanotheres
One-fourth natural size. Ai, Palaeosyops Icidyi, Am. Mus. 1544 (type). Median section of cranial region. Note the back part of the
olfactory chamber (with remnants of the ethmoturbinals), the cribriform plate, the frontal sinus, and the cellular character of the expanded
cranial vertex above the brain chamber. (Section line shown in A2.) As, The same, rear view. The left side shows the widened vertex oi the
occiput; the right side shows in section the frontal and ethmoidal sinuses. B, Limnohyops prisms, Princeton Mus. 10044. Cross section
through the basioccipital, periotic, squamosal, and paiietal, seen from front. In this primitive form the sagittal crest has not expanded
into a flattened cranial vertex, and hence this region is without any large cavities. C, Dolichorhinus Jiyognathus, Am. Mus. 1851. Median
section of the whole skull slightly to the right of the median plane, showing the enormous olfactory chamber, the elongate fronto-occipital
sinus, and the small brain chamber. The much enlarged maxilloturbinal (mx. tb.) is produced backward and downward, appearing as a
prominent swelling in the roof of the narial channel; the primary border of the posterior nares is at pn', the secondary at pn'. D, Dolichorhinus
longicepsf (Jiyognathus f) , Am. Mus. 1852. Cross section through the middle part of the brain chamber (near line A-A of figure C) looking
forward. Note the fossae for the anterior lobes of the brain, the lateral ethmoid sinus, the mesethmoid septum, the remains of the
ethmoturbinal scrolls, and the large fronto-occipital sinus.
consisting of the very massive, broad-headeu Palaeo-
syops.
The genus Palaeosyops of Leidy was the first known,
and the LimnoJiyops of Marsh may be regarded as a
subgenus. Although the animals belong to two dif-
mtergradations or mutations are found. But this
very abundance renders more difficult the definition of
species because the sharp lines of specific separation
and distinction breah down ; the forms merge into one
another.
300
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 255. — Cross sections of the skull in middle Eocene titanotheres
One-fourth natural size. Ai-Ei, Sections across the face just behind the lacrimal and through the malar and m'; A2-E!, sections
across parietals, alisphenoids, and zygomatic process ol squamosal. Ai, Aj, Palaeosyops leidyi, Am. Mus. 1516; Bi, Ba,
Telmathermm ultimum, Am. Mus. 2060 (type; crushed laterally); Ci, Cj, Manteoceras manteoceras, Am. Mus. 12678; Di, Ds,
Mesatirhinus pclersoni, formerly Am. Mus. 1566, now in British Museum; Ei, Ej, Bolkhorhinus hyognathus, Am. Mus. 1851.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
301
Probably the pbysiographic conditions in this region
during the early stages of the Bridger deposition were
peculiarly favorable to these animals. Whatever the
cause in Bridger B and C their remains are as plentiful
as those of other titanotheres are rare. In Bridger
D, however, remains of Palaeosyops become mingled
with those of titanotheres of other kinds, which are
Z. i'Ti n ohyops
Figure 256. — Three skulls typical of the palaeo-
syopine group
One-eighth natural size. A, Zimnohyops prisms. Am. Mus.
11687 (type), middle Eocene, lower Bridger; B, Palaeosyops
leidyi, Am. Mus. 1544 (type), middle Eocene, upper Bridger;
C, Telmatherium ultimum, Am. Mus. 2060 (type), upper
Eocene, TJinta C (true Uinta formation).
equally or even more abundant and include forms that
apparently had undergone their antecedent evolution
in another part of the mountain region of the continent.
(See fig. 257.)
COMMON CHARACTERS OF THE PALAEOSYOPS AND IIMNOHYOPS
GENERIC PHYLA
The three most distinctive features of Palaeosyops
and Liranohyops, as stated above, are brachyodonty,
or persistently short-crowned grinding teeth; brachy-
cephaly, or progressively increasing head width;
brachypody, or broad and abbreviated foot structure
(less defined in Limnohyops).
The members of all the species known in both phyla
are thus short-toothed, short-skulled, and more or less
short-footed. In the accompanying outline of the
Figure 257. — Distribution of the species of Palaeosyops and
associated fauna in the Bridger formation, Bridger Basin,
Wyo.
parallel geologic distribution of the species belonging
to these two phyla in the Bridger formation we observe
that in about every 200 feet of sediment there is accu-
302
TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
mulated a change in proportions and in a number of
details of cranial and dental structure which we may
collectively dignify by the term species, in the neo-
Linnaean sense. The transitional or intermediate
stages, distraguished by the infinitesimal advance in
certain new characters, are mutations in the sense of
that term as used by Waagen. The orthogenetic or
direct and adaptive origins of single new characters are
rectigradations in the sense of that term as used by
Osborn. (See p. 812.) The progressive changes in
certain characters — for example, in the rectigradations
of the premolar teeth and in the rise of the horn
rudiments on the frontals — occur nearly contemporane-
ously in members of the two phyla. In some other
characters the progression is dissimilar, or at different
rates.
Geologic horizons of Limnohyops and Palaeosyops
Part of lormation
Horizon
Limnohyops
Palaeosyops
D3 .
P. copei.
Upper Bridger _.
D2
D 1
D
L. laticeps..
L. laticeps..
C4
C3
L. laticeps. -
P. leidyi.
P. leidyi.
P. leidyi.
P. grangeri.
C 2. -
C 1
B 4
P. major.
P. major.
P. major.
P. paliidosus.
P. longirostris.
P. paludosus.
B 3
Lower Bridger
B 2
B l?-5?
[L. matthewi
L. monoco-
nus.
L. priscus- .
[L. laevidens.
B 1
A?
PROGRESSIVE DISTINCTIONS BETWEEN PALAEOSYOPS AND IIMNOHYOPS
It is extremely difficult — indeed, it may be impossi-
ble— to distinguish parts of individuals belonging to
Palaeosyops from parts of those belonging to Limno-
hyops. Means of recognizing the differences and
resemblances have been afforded by the cumulative
work of Marsh, Earle, Osborn, and Gregory.
The supposed distinctive generic character (Marsh)
of LimnoTiyops, namely, the presence of a hypocone on
m^, is possibly a primitive character, because of its
presence in Lamidotherium. It prevails but does not
appear to be constant in aU species of Limnohyops.
It is typically absent but exceptionally present, by
reversion perhaps, in Palaeosyops. In the proportions
of the skull Palaeosyops is more robust and Limno-
hyops is more slender, and this quantitative or pro-
portional contrast prevails throughout all the cranial,
dental, and skeletal parts, although it is often difficult
to measure or define the finer shades of difference.
When we compare the ancestral members of the
two phyla in Bridger B, some of them are difficult to
separate. As the successive specific stages of Limno-
hyops are contemporaneous geologically with those of
Palaeosyops it is well to enumerate the chief known
distinctions which gradually develop and become fully
apparent only after the two lines of descent have
diverged from each other, as observed in the higher
geologic levels — for example, in comparing P. rolustus
and L. laticeps of Bridger D. These distinctions are
as follows:
1. The upper and lower molar teeth of Palaeosyops
are relatively larger, more rounded, and more robust
than those of Limnohyops.
2. The vertical striations on the cones of the upper
and especially of the lower molars of Palaeosyops are
more distinctly marked.
3. On the upper molars (m'~^) of PaJaeosyops the
conules are more variable, more rounded, and sepa-
rate; in Limnohyops they are more constant, lophoid,
ridged, or conjoined with the protocone and hypo-
cone; this distinction, however, is not invariably
reliable.
4. In m' of Limnohyops the hypocone is typically
though not invariably present (L. laticeps), and the
metaconule is extremely reduced, owing to the large
size of the adjacent hypocone. In m^ of Palaeosyops
the hypocone is typicaUy absent but sometimes pre-
sent, as in the type of P. diaconus. In m^ of Palaeosyops
the metaconule is generally present and in some
specimens is so close to the raised posterior cingulum
as to appear like a hypocone; thus the m' of Palaeo-
syops is generally more triangular, whereas that of
Limnohyops is more quadrate and sometimes actually
bUobed internally.
5. The parastyle in Palaeosyops is rounded and
obliquely placed across the outer angle of the crown,
whereas in Limnohyops it is sharp and extends out-
ward as a ridge, analogous to the parastyle of the
Telmatherium type (Pis. LX, LXIII).
6. The nasals taper toward the extremities and are
more pointed in Palaeosyops, whereas in Limnohyops
the sides of the nasals are more parallel and they are
more truncate at the extremities.
7. The suborbital bar in the two genera becomes
quite different; in Palaeosyops the bar is roimded and
the overlying maxillary process extends back on its
outer side as a broad splint, whereas in Limnohyops
the bar becomes more depressed and slightly rectan-
gular in section and the maxillary process extends
back as a long, slender splint on the lower side; in
Telmatherium the suborbital bar is distinctly rectan-
giilar and the maxillary process extends back as a
long, narrow splint beneath the malar projection.
8. In LimMohyops the top of the cranium is slightly
concave; in Palaeosyops there is a strong median con-
vexity near the frontoparietal junction some distance
behind the orbits.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
303
9. The sagittal crest of Palaeosyops is lower, broader,
and passes more rapidly into the temporal ridges,
whereas in Limnohyops as in Telmatherium the crest
is higher and thinner and extends well forward before
spreading into the temporal ridges.
10. The male jaws of Palaeosyops are at once recog-
nized by (a) the more prominent, massive chin and,
as seen from below, the short, depressed area for
the digastric muscle on the posterior symphyseal
line, features that contrast with the longer, more slop-
ing chin of Limnohyops and its elongate median fossa
for the digastric; (6) the massive breadth of the
Palaeosyops rami, as seen from below, in contrast with
the somewhat more slender inferior borders of the jaw
in Limnohyops; (c) the extremely distinctive base of
the anterior border of the coronoid process which in
Palaeosyops is very broad and in progressive stages
deeply hollowed out in front, whereas in progressive
stages of Limnohyops it is somewhat narrower, less
deeply excavated, and lies more to the outer side of
the line of the molar teeth.
Additional means of distinction are set forth in the
descriptions of the genera and species.
W. K. Gregory has observed that the above and
other differences are in part quantitative; they are
differences in the proportion of one and the same
character, as in the form of the nasals, of the sub-
orbital bar, and of the sagittal crest. The divergence
is far less than that seen in the modern genus Cervus,
for example. It may be noted also that certain of
the numerical differences are variable; for example,
the hypocone on m^. The known forms of Limnohyops
are rather slender; thus a male jaw of this animal
would resemble a female jaw of Palaeosyops.
Limnohyops Marsh
Plates LVI, LVII, LX, LXII; text figures 29, 87, 96, 115-117,
219, 254, 256, 258-266, 274, 484, 485, 510-514, 516, 518-523,
525, 527-532, 538, 672, 685, 686, 690, 701, 714, 717, 722,
723, 745, 760
[For original description and type reference see p. 170; for skeletal characters see p. 605]
Generic characters. — Brachy cephalic; grinding teeth
persistently brachyodont; conules on the molars per-
sistent, usually lophoid; third superior molar sub-
quadrate and usually with distinct hypocone. Pro-
portions of skull and skeleton moderately robust.
Manus slender. Five sacral vertebrae (type).
Geographic and geologic distribution. — Limnohyops
has thus far been found only in the geologic levels
B, C, and D of the western or Bridger Basin (see
geographic map on p. 8). As compared with Palaeo-
syops the materials representing this genus or sub-
genus are limited; we can not therefore trace at present
the successive stages of its evolution. It is subdi-
vided into five species — L. Inevidens, L. priscus, Z.
laticeps, i. matthewi, L. monoconus.
Resemblances to Palaeosyops. — From our present
knowledge the geologically early species L. laevidens
and L. priscus are so close to the type of Palaeosyops
(P. paludosus) that they might readily be embraced
within one and the same genus. In fact, material at
first referred by the present author to L. priscus now
appears to belong to P. paludosus, which is itself so
primitive that it may almost be regarded as the an-
cestor of Limnohyops. However, the sum total of
the distinctions between these animals — in external
form, color, and habits — was probably very consid-
erable, and as we progress into geologically higher
stages the cranial and dental differences become more
apparent, as summarized below.
Materials. — Besides the admirable type specimen
from Bridger C 4 in the Yale Museum, on which
Marsh founded the genus and species, American
Museum parties have found five specimens of L.
laticeps in levels Bridger C 4 and D 2. In the lower
level of Bridger B 2 occurs the type of L. laevidens
Cope, and here we have also found two specimens of
the somewhat more progressive stage L. priscus.
At present the species L. matthewi and L. monoconus
are represented only by a single specimen each, and
it is noteworthy that these also are of lower Bridger
age. Thus our knowledge of Limnohyops at present
rests on portions of about 16 individuals.
Chief progressive distinctions from Palaeosyops. —
(Compare pp. 302, 618, vertebrae; p. 612, limbs and feet.)
So far as we know at present Limnohyops is distin-
guished by somewhat more slender proportions. The
skull in the larger species is equally broad but less
massive ; the j aws are decidedly less massive. The long
bones of the limbs referred to L. laticeps are practically
of the same length as those of the contemporary
Palaeosyops leidyi, but the foot bones of the manus
of Limnohyops appear to be shorter (figs. 512, 520).
Limnohyops may be described briefly as a relatively
light-limbed, broad-skulled, short-footed type.
As we have already given many of the details by
which Limnohyops in its advanced stages may be
distinguished from Palaeosyops, it is only necessary
to summarize its chief diagnostic characters.
Cranium: (1) The skull of Limnohyops has a concave
instead of a convex forehead (fig. 256); (2) it has a
high, thin sagittal crest; (3) the nasals are relatively
broad anteriorly; (4) the suborbital bridge of the
malars is shallow, narrow, and more or less quad-
rangular in section or broadly depressed, with a
rounder outer border, and the fiange for masseteric
insertion is not extended so far forward as in Palaeo-
syops; (5) the splint of the maxilla extends backward
under the side of the malar; (6) only slight promi-
nences indicate the osseous horn areas, and no horn
rugosities have been observed; (7) the jaws have more
slender rami, the chin is sloping, the digastric fossa
is elongate, deep, and sharply defined, the lower border
is less thickened below the grinders, the coronoid base
is less broad and flaring anteriorly, when seen from
304
TITANOTHERES OF ANCIENT "WYOMING, DAEIOTA, AND NEBRASKA
the front, and is set on the outer side of the line of the
grinders.
Dentition: In the superior teeth we note especially
that (1) m^ is small; (2) the crested metaconules are
confluent internally with the hypocone; (3) the
parastyle is elongated on the outer side of the ectoloph
and somewhat more sharply ridged; (4) the ectolophs
of premolars have more sharply defined convex ridges
opposite the paraconules and metaconules; (5) the
hypoconulid of ms is rather sharp and prominent,
laterally depressed, and very slightly crescentic.
Measurements of species of Limnohyops, in millimeters
Level in
Bridger
formation
Species
Basilar
length of
slsull
pi-m3
P2-m»
C-D
B 2
B 1-2
B 1-2
L. laticeps Marsh
L. monoeonus Osborn
L. priscus Osborn
L. laevidens (Cope) _. -
"410
"439
375
153
156
149
141
139
142
133
129
• Estimated.
The distinctions in measurement and proportion
noted above are based upon our present loiowledge
and are by no means so full and definite as we should
wish.
Comparison of the species. — The known species of
Limnohyops do not form a progressive phyletic series
in the ascending geologic scale, whereas the known
species of Palaeosyops form a finely progressive
phyletic series.
The type of L. (Palaeosyops) laevidens Cope, from
Bridger B 1-2, is a relatively small and primitive
form, an undoubted Limnohyops in skull structure,
distinguished by a small p^ of rounded form. All
the teeth in the type are greatly worn.
Considered in ascending geologic order from Bridger
B 1 to D the species may be distinguished as follows:
The type of L. priscus Osborn is recorded as from the
same geologic horizon — namely, Bridger B 2 — but it
is an animal of greater size and has a p^ of elongate,
triangular form.
The large and heavy type of L. monoeonus Osborn
also occurs in Bridger B at Grizzly Buttes. It is
named the "single-coned species" in reference to the
absence of the hypocone on m'. Except in this
character it is a typical Limnohyops.
The type of L. matthewi also occurs in Bridger B 2,
a surprising fact because of its extremely short and
broad skull proportions. It exhibits the extreme of
brachycrany.
L. laticeps, the type species of the genus, occurs
much higher up — namely, in Bridger C and D. It is
less specialized in skull structure than L. matthewi
and has the generic character of the presence of the
hypocone on m^ very marked.
QXTANTITATIVE EVOIUTION OF IIMNOHYOPS
The accompanying table gives the measurements of
29 characters of proportion in 11 specimens, repre-
senting 5 species, collected at ascending geologic levels
in the Bridger formation.
Evolution of proportions of Limnohyops
[Measurements in millimeters]
L.
laevidens,
Am. Mus.
5104 (type);
Bridger B
L. priscus
L.
matthewi,
Am. Mus.
11684
(type);
Bridger B 2
L. monoeonus
L. laticeps
Am. Mus.
11688
(cotype);
Bridger B 2
Am. Mus.
11687
Bridger B 2
Am. Mus.
11679
(type);
Bridger B 2
Am. Mus.
6102;
Bridger B
(3-5?)
Yale Mus.
11000
(type)
Am. Mus.
11710;
Bridger D 2
Am. Mus.
12201;
Bridger C 4
Am. Mus.
12198;
Bridger D 2
Skull:
"375
270
-439
"■320
°410
»310
452
118
-395
Height of occiput above for.
118
160
169
122
Width across postglenoid proc-
pi-m3
141
129
57
85
15
21
23
27
31
34
149
137
64
87
18
23
24
27
32
35
"156
142
172
159
153
139
64
90
18
24
24
27
33
38
"75
p2-m'
P'-p«
M'-m^
» 30
'■93
19
23
26
29
33
36
103
20
26
27
33
36
40
32
38
» Estimated.
EVOLUTION OF THE SKULL AND TEETH OP EOCENE TITANOTHERES
Evolution of proportions of LimnoTiyops — Continued
305
L.
laevidens,
Am. Mus.
6104 (type);
Bridget B
L. priscus
L.
matthewi,
Am. Mus.
11684
(type);
Bridger B 2
L. monoconus
L. laticeps
Am. Mus.
11688
(cotype) ;
Bridger B 2
Am. Mus.
11687
(type);
Bridger B 2
Am. Mus.
11679
(type);
Bridger B 2
Am. Mus.
5102;
Bridger B
(3-6?)
Yale Mus.
11000
(type)
Am. Mus.
11710;
Bridger B 2
Am. IVjus.
12201;
Bridger C 4
Am. Mus
12198:
Bridger D 2
Lower jaw:
»278
-75
71
123
160
146
91
19
13
24
16
41
20
368
93
93
-175
191
172
111
22
14
27
17
48
24
357
"95
94
» 153
lower teeth:
Pi-ms
196
173
116
20
14
28
19
52
25
" Estimated.
The foregoing table brings out the following facts:
1. The cranial increases in length and width are not
accompanied by proportional dental increases.
2. Relatively large-skulled and extremely brachy-
cephalic animals (i. monoconus) occur in the lower
Bridger levels.
3. There is no evidence of progressive monophyletic
change such as we see in Palaeosyops. (See p. 313.)
Limnohyops laevidens (Cope)
Plate LVII; text figures 96, 258, 259
[For original description and type references see p. 163]
Type locality and geologic horizon. — The type speci-
men represents the smallest, most primitive, and
geologically earliest LimnoTiyops at present known. It
is somewhat doubtfully recorded from Bridger B 2(?),
as represented in the deposits of Cottonwood Creek,
Bridger Basin, Wyo. If from this level, it is slightly
more recent than the geologically early and most primi-
tive discovered stage of Palaesoyops, known as P.
paludosus, referred specimens of which have been
found in Bridger B 1 .
Specific characters. — Inferior in all dimensions to
type of L. laticeps; p'-m', 141 millimeters; p^-m^
129; p'-p*, 57; p^ rounded or transversely oval, with
tritocone rudimentary or absent. A large hypocone
on m'.
Materials. — The type specimen of L. laevidens (Am.
Mus. 5104; see revision of the nomenclature, Chap.
Ill) is a part of an aged skull containing a well-worn
dental series, from which it is difficult to determine
positively the characters of this animal. The specific
name laevidens (from laevis, (Zens = imperfect tooth)
was assigned to this specimen by Cope in recognition
of the supposed absence of the tritocone on p^. It is
probable that in the unworn condition this cusp was
present but very rudimentary.
Cope's type of Palaeosyops laevidens
Teeth. — The teeth of the type of P. laevidens (Am.
Mus. 5104), belonging to an aged animal, are especially
Figure 258. — Anterior part of skull of Lim7iohyops laevidens
One-fourtli natural size. Am. Mus. 5104 (type), reversed; Cottonwood Creeli,
Bridger Basin; level Bridger B 3. Ai, Side view; A2, front view.
interesting because they exhibit the influences of age
in rounding off the angles, wearing away the cusps,
and smoothing down the cingula. Professor Cope
was probably misled when he described this type as
lacking a tritocone on p'; we now recognize that the
306
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
apparent absence of this tritocone may be due to age;
it probably has been almost completely worn off.
The superior incisors arch gently forward, the total
transverse series when in place measuring 56 milli-
meters; there is a regular increase from i^ to i^; the
anterior faces of the crowns are slightly crenulate,
and a convex swelling or low ridge surrounds the base;
the posterior faces are marked off by lateral ridges
and by a posterior cingulum, which is irregularly
pitted above; the transverse measurements of the
anterior portions of the crown are: i' 10 millimeters,
i^ 11 (ap. 12), i' 12. A narrow diastema (7 mm.),
separates i^ from the canine.
The canine has a stout fang and a crown laterally
compressed at the base, the measurements being,
anteroposterior, 21 millimeters; transverse, 18; height,
36 (estimated). The crown is defined by faint
anterior and posterior ridges; it is slightly retro verted
and inverted.
In the premolar-molar series a very narrow diastema
(5 mm.) separates the canine from p', a tooth which
is continuous with the remaining grinders, the total
length of the whole grinding series being exactly 141
millimeters, less than in the type of L. prisons (149
mm.). In MesatirMnus megarhinus the premolar-molar
series ranges from 140 to 147 millimeters. P' is a simple,
bifanged cone (ap. 11 mm., tr. 8), with faint anterior
and posterior concavities on the inner side. P^ is
an obliquely placed oval, measuring (ap. by tr.) 12
(ectoloph 15) by 15 millimeters, whereas in Mesati-
rMnus megarhinus the anteroposterior diameter greatly
exceeds the transverse. As Cope pointed out, this
tooth is distinguished specifically by the simple
rounded protocone, with a more sessile and internally
placed tritocone, and a relatively small deuterocone
on its lingual side. P^, measuring 15 (ectoloph 18)
by 19 millimeters, is broader than long and exhibits
relatively more prominent tritocones and deutero-
cones. P*, 15 (ectoloph 18) by 21 millimeters, is
also broader than long, the tritocones and deutero-
cones are still larger, and the external cingulum begins
to be defined, as well as the rudimentary anterior
and posterior cingula; very rudimentary cingula are
also observed in p^, p^, except on the lingual side of
the deuterocones, which cusps are absolutely smooth
and rounded in all these teeth, presenting in this
respect a sharp contrast to the condition in M. mega-
rhinus, or even (in less degree) to that in the type of L.
priscus. The entire length of the premolar series
is 58 millimeters, as compared with 64 in a small
individual of M. megarhinus. The molar series
measures 84 millimeters. The imperfectly preserved
m' (ap. 2.3 mm., tr. 27) exhibits rudimentary external
and antero-internal cingula and a subquadrate crown;
in m^, also badly damaged, we observe evidence of
sessile conules and a low anterior cingulum; in the
better preserved m' (ap. 30 mm., tr. 35), also a sub-
quadrate tooth, there are rudimentary external.
anterior, and postero-internal cingula, the last giving
rise to a low cingulate hypocone; there is some evi-
dence of small, well-worn protoconules and metaco-
nules; the crown, as in the other molars, is singularly
smooth.
SlcuU. — Although only the anterior portion of this
cranium is present (see fig. 258), it affords conclusive
evidence of ancestral relationship to L. laticeps in
the rounded shape of the nasals. Its general or
palaeosyopine characters are especially seen in (1) the
downward V-shaped extension of the nasals on the
sides of the face; (2) the prominent antorbital Icnob of
the lacrimals; (3) the backward extension of the
infraorbital portion of the maxillaries beneath the
malar arch; (4) the narrow median symphysis between
the premaxiUaries; (5) the comparatively slight lateral
decurvature of the nasals; (6) the extreme upward
arching of the zygoma as a whole, the mid-depth
being 51 millimeters, and the "depression and angula-
tion" of the malar 19 millimeters behind the orbits.
This cranium not only differs in its smaller size but
in a number of other proportional characters from that
of L. laticeps. The nasals are relatively more elongate,
narrower posteriorly, and relatively broader an-
teriorly— that is, the sides of the nasal in front of the
narial notch are more nearly parallel, the narrowest
midportion measuring 44 millimeters, the broadest
terminal portion measuring also 44. From the an-
terior border of the orbit to the narial notch the
measurement is 61 millimeters. The zygomatic bar
immediately below the orbit is more angulate and less
rounded than in P. leidyi, the inferior face of the
malars being more flattened and the sharp masseteric
ridge defining the malars inferiorly being less extended
fore and aft. In palatal view we observe the trans-
verse extension of the glenoid facets for the condyles
of the jaw, the opening of the nares behind m^, the
abbreviation of the palatines, and the relative flatness
of the palate.
Limnohyops prlscus Osborn
Plates LVI, LX, LXII; text figures 29, 115, 219, 254, 256,
259, 260, 266, 274, 690, 717, 745
[For original description and type references see p. 180]
Type locality and geologic horizon. — Grizzly Buttes,
west Bridger Basin, Wyo. ; Bridger formation, Palaeo-
syops paludosus-Orohippus zone, level Bridger B 2.
Specific characters. — P^-m^ 148 (type) to ?161 milli-
meters. Second superior premolar obliquely elongate
with a very rudimentary tritocone. Large hypocone
on m^.
This is apparently a larger and relatively more
advanced animal than L. laevidens, but, as the specific
designation priscus indicates, it is still very primitive.
Materials. — L. priscus is represented by the type
skull (Am. Mus. 11687; see fig. 259), from the
Bridger formation at Grizzly Buttes, level B 2, and
by the type or cotype jaw (Am. Mus. 11688), found
EVOLUTION OP THE SKULL AND TEETH OF EOCENE TITANOTHERES
307
near the same place. An occiput in the Princeton
Museum (No. 10044), also from Cottonwood Creek,
which was described and figured by Earle (1892.1,
p. 353) as L. laticeps, may also be provisionally referred
to L. priscus.
The generic or LimnoJiyops character is shown in the
presence of a hypocone on m^. Divergence from the
type of L. laevidens is indicated by the elongate form
of p^ as distinguished from the rounded or transversely
oval form which this tooth presents in L. laevidens.
The progressive stage is indicated by the rudimentary
condition of the tritocone on p' in the type specimen
of L. priscus.
The type sTcull. — The skull of the type specimen
(fig. 259) is somewhat larger than that of the American
tapir. It is extremely interesting to note that it more
closely resembles that of the common ancestral form of
the titanotheres than does the skull of the contem-
porary species Palaeosyops paludosus. The
skull and m^ are therefore more primitive
than those of Palaeosyops. The specimen in
hand is much crushed laterally, and the
restoration represented in Figure 259 is an
approximation to the complete form. The
estimated total length from condyles to
symphysis is 375 millimeters; from the crest
of the occiput to the tip of the nasals 395.
The cranium shows the typical LimnoJiyops
characters enumerated above — namely,
short supraoccipital exposure on top of the
skull; moderately high, thin sagittal crest,
which diverges into the supratemporal crests
about 95 millimeters in front of the occiput;
occiput moderately high (proportions not
to be exactly ascertained owing to its crushed
condition). The nasals measure 143 milli-
meters (ap.) and taper gradually toward the
extremities. In the base of the skull the
paroccipital processes are separated by shal-
low grooves from the post-tympanic processes, and the
external auditory meatus is apparently open below.
Dentition. — The superior teeth are well represented
in the type cranium. The lateral incisor is enlarged
as in Palaeosyops; the canine is slender, slightly
recurved, and followed by a very narrow diastema;
the grinding series (Pis. LVI, LX, LXII) is con-
tinuous, p'-m^ measuring 149 millimeters; pm^ pre-
sents a very large convex protocone and rudimentary
tritocone. The inferior teeth are well shown in the
cotype jaw (Am. Mus. 11688), found not far from the
skull and probably belonging to the same individual.
The two incisors preserved are cingulate posteriorly;
the canine is rather slender, laterally compressed, and
slightly recurved; behind this is a small, simple pi,
followed by a narrow diastema; p2 and ps are com-
paratively narrow and simple teeth, exhibiting ex-
tremely rudimentary paraconids and slightly cres-
centic hypoconids, and a faint rudiment of the meta-
conid in ps; p4, on the contrary, is submolariform,
exhibiting a well-developed metalophid and a rudi-
mentary hypolophid or posterior crescent which lacks
only the entoconid. The three true grinders are
simple, with rudiments of vertical striations, with
distinct paraconids, and with a sharply pointed sub-
crescentic hypoconulid on la^. The characters of the
jaw are well shown in Figure 259.
Back of the cranium. — The back of a cranium in the
Princeton collection (Princeton Mus. 10044) may be
provisionally referred to this species. It was found
on Cottonwood Creek, Bridger Basin, Wyo., in the
Bridger formation, level B 3(?) by Francis Speir, of the
Princeton expedition of 1877, and was rightly referred
to Limnohyops by Earle. The deep (45 mm.) and nar-
row (9 mm.) sagittal crest is not quite so elongate as
that of the Yale Museum type of L. laticeps, the supra-
FiGUKB 259. — Skull of Limnohyops priscus
One-fourth natural size. Skull Am. Mus. 11687 (type); Grizzly Buttes (west), Bridger Basin, Wyo.;
Bridger formation, level B 2. Partial reconstruction of crushed skull made by L. M. Sterling
under the direction of W. K. Gregory. Lower jaw. Am. Mus. 11688, possibly belonging to same
individual. Details of zygoma restored from Am. Mus. 5104 (type of L. laevidens); details of
occiput from Princeton Mus. 10044, L. priscus.
temporal ridges beginning to diverge rapidly 77 milli-
meters in front of the crest of the occiput. In
addition to the points noted in the type skull above
described, we observe in this occiput (1) two large
mastoid foramina, (2) two prominences just above the
foramen magnum, (3) the broadly extended (48 mm.)
post-tympanic processes closely conjoined at the base
with the relatively narrow (12 mm.) paroccipital proc-
esses; (4) the relatively broad (37 mm.) postglenoid
processes. In palatal view the zygomata are seen to
have a transverse measurement of 268 millimeters, the
post-tympanic and postglenoid processes are slightly (3
mm.) separated; the articular facets for the condyles
of the jaw are nearly transverse; a bridge of bone, 19
milluneters in width, separates the foramen ovale and
foramen lacerum medium. These features are well
illustrated in Figure 260 as compared with similar views
of L. laticeps and L. matthewi.
308
TIT.\KOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Limnohyops matthewi Osborn
Text figures 116, 261, 262
[For original description and type references see p. 180]
Type locality and geologic horizon. — -Grizzly Buttes
(west), Bridger Basin, Wyo.; Bridger formation,
Palaeosyops palvdosus-OroMppus zone, level B 2.
Specific characters. — Type of intermediate size as
compared with L. laevidens and L. monoconus. M}
small, with large hypocone and quadrate inner half.
Occiput very high and narrow. Cranial portion of skull
greatly abbreviated, bringing post-tympanic and post-
glenoid processes into broad union. Temporal open-
ings subcircular as defined by zygomatic arches. (See
figs. 261, 262.)
Figure 260. — Back part of skull of Limnohyops priscus
One-fourth natural size. Princeton Mus. 10044; Cottonwood Creels:, Bridger Basin, Wyo.; lower part of Bridger formation.
Ai, Side view; Az, basal view; A3, top view; A4, occipital view.
This species is named in honor of Dr. W. D.
Matthew, of the American Museum, who has done
so much to advance our knowledge of the Bridger
fauna and geology.
As compared with the primitive type skulls of L.
laevidens and L. priscus, above described, the cranial
region of this animal is surprisingly specialized in its
abbreviation, or extreme brachycephaly. This pro-
gressive character is difficult to reconcile with the fact
that it is geologically recorded in the same low level as
that of L. laevidens.
Materials. — The type (Am. Mus. 11684), consisting
of the posterior portion of a skull which includes the
orbits, is the only specimen at present referable to this
species. It was found in 1903 by the American
Museum expedition.
Distinctive characters. — The brachycranial type of
this species is readily distinguished from that of
L. laevidens by the very specialized condition of the
posterior portion of the skull, including the elevated
occiput and the firm inclosure of the auditory meatus
below. It differs from L. laticeps when seen from
above in the greater height of the occiput and in the
transversely oval form of the temporal openings as
defined by the zygomatic arches. It differs from
L. monoconus in the presence of a large hypocone on
m' and in the quadrate form of the inner side of this
tooth, also in the rounded temporal openings.
In the lateral view of the skull we observe that the
zygoma descends rapidly anteriorly and thins out as
it passes into the anterior
portion of the malar, which
constitutes the suborbital
bridge; this bridge is de-
pressed in section but ex-
hibits a rounded rather
than angulate outer bor-
der. This species differs
from Palaeosyops in that
the masseteric insertion
ridge is not carried for-
ward any great distance,
and that there is a slender
splint of the maxillary ex-
tending back below the
malar bridge. The sagit-
tal crest extends 50 milli-
meters above the brain
case proper; it is even
higher and thinner than in
L. laticeps. As seen from
below the temporal open-
ings are transversely oval,
the anteroposterior meas-
urement being 82 milli-
meters, and the transverse
measurement from the
malars to the alisphenoids
95. The sagittal crest divides these great insertion
cavities of the temporal muscles and is thin at the
summit, measuring 8 millimeters at the thinnest point;
it is also elongate, extending 94 millimeters from the
tip of the occiput to the point where it bifurcates into
the supratemporal ridges.
The chief feature of the occiput is the foreshortening
and compression of its lower portion against the back
portion of the squamosals and zygomatic arch,
causing a broad junction of the postglenoid and
post-tympanic processes and a very characteristic
flattening of this region. Seen from behind (fig. 262)
the occiput measures 116 millimeters above the
foramen magnum, and 157 above the bottom of the
occipital condyles. Conforming to the smaller size
of the skull as a whole, the condyles measure 86 milli-
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
309
meters transversely as compared with 99 in L. mono-
conus. The width of the occiput across the top is
108 millimeters. At the sides of the exoccipital por-
tion are indistinctly seen the mastoid foramina. Just
Figure 261. — Skulls of three species of Limnohyops
Occipital view. One-fourth natural size. A, L. monocomis, Am. Mus. 11679
(type); Grizzly Buttcs, west Bridger Basin, Wyo.; Bridger formation, level B 2.
B, L. laticeps, Yale jMus. 11000 (type, vertically crushed); Bridger Basin; upper?
Bridger. C, L. matlhewi, Am. Mus. 11684 (type); Grizzly Buttes, west Bridger
Basin; Bridger formation, level B 2.
above the foramen magnum are very faintly indi-
cated the pair of facets characteristic of this genus,
more distinctly marked in Telmatherium.
Limnohyops monoconus Osborn
Plate LXII; te.xt figures 117, 261-263, 484, 485, 510-514, 516,
519-523, 525, 527-530, 685, 686, 701, 723
[For original description and type reference see p. 130. For skeletal characters see
pp. 604, 612]
Type locality and geologic Tiorizon. — Grizzly Buttes
(west), Bridger Basin, Wyo.; Bridger formation, Pal-
aeosyops paludosus-Orohippus zone, level B 2. This
specimen was discovered by the American Museum
expedition of 1903. It was on the level of L. laevidens
and L. matthewi and thus belongs to a much lower
horizon than L. laticeps. This specimen may rep-
resent, however, the geologic intrusion of a fauna from
a higher level, and it is therefore quite possible that it
was not contemporaneous with L. laevidens. The
presence of these specialized forms, L. mattheivi and
Figure 262. — Skull of Limnohyops matthewi
One-fourth natural size. Am. Mus. 11684 (type) ; Grizzly Buttes (west) , Bridger
Basin, Wyo.; Bridger formation, level B 2. Ai, Side view; As, basal view.
L. monoconus, on the same geologic level as the
primitive forms L. laevidens and L. priscus is contrary
to the general law of succession observed among
other forms in the Bridger Basin. It may indicate
either some source of error in the geologic records
or some deviation from the generally horizontal dis-
tribution of the Bridger titanotheres. Another ex-
planation is possible: that Limnohyops was evolving
more rapidly in other geographic centers, from which
these progressive forms may have migrated.
Specific characters. — M^ without hypocone, roundly
triangular in form, with broadly extended ectoloph
and parastyle. P*-m', 142 millimeters; p'-m^, 156.
Condyle to incisive border 439 millimeters (estimated);
310
TITAKOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
occiput very high; cranium relatively elongated,
with space (4 mm.) between post-tympanic and post-
glenoid processes; temporal openings as defined by
zygomatic arches elongated.
This type (Am. Mus. 11679) is an exception to all
the other species referred to Limnohyops in the ab-
sence of the hypocone on m^. The specific name,
monoconus, refers to the existence of but a single
internal cusp (protocone) on this tooth. The animal
is provisionally referred to the genus Limnohyops on
strong evidence in five other points of cranial structure
which are cited below.
As compared with the type of L. matthewi, the
animal on which this species is founded is of very
FiGUBE 263. — Skull of Limnohyops monoconus
Palatal view. One-fourth natural size. Am. Mus. 11679 (type); Grizzly Buttes,
west Bridger Basin, Wyo.; Bridger formation, level B 2.
robust size; its size is especially robust for the low
geologic level from which it is recorded — namely, B 2.
Its proportions agree fairly well with those of L. lati-
ceps, which is referred to level D 2.
Materials. — The species is certainly Icnown only
from the type (Am. Mus. 11679), which consists of
the crushed cranium of a robust male specimen, as
indicated by the large, recurved canines. Other
doubtfully referred specimens (Am. Mus.' 5102, 12680)
are recorded from B 5.
As the hypocone on m^ is the alleged "generic"
character of LimnoTiyops and is present in the other
species L. laevidens, L. matthewi, and L. laticeps, its
absence in this form is very exceptional and causes
some doubt as to the propriety of the generic refer-
ence of this species to Limnohyops. The other grounds
for referring this animal to Limnohyops rather than to
Palaeosyops are strong, however — namely, (1) the
doubly ridged ectolophs of p"-p*; (2) the very elevated
occiput; (3) the more or less angulate form of the
suborbital bridge in the malars; (4) the splint of the
maxillary extending on the lower side of the malar
bridge; (5) nasals not perceptibly narrowing anteriorly.
Specifically this animal is readily distinguished from
all other species of Limnohyops by the absence of the
hypocone on m^. As compared with L. laevidens it is
an exceptionally large form, the measurements (esti-
mated) being, from the condyle to the incisive border
439 millimeters, width across the zygomata 320. It
is also distinguished by the very robust and recurved
canines. From L. laticeps it is distinguished by the
higher occiput, by the absence of a hypocone on m^,
and by its supposed lower geologic level. It is dis-
tinguished from L. matthewi by the more elongate
cranium, correlated with which are the oval openings
circumscribed by the zygomatic arches, and by the
separation between the postglenoid and post-tympanic
processes.
The skull is of massive proportions, with widely
arched zygomata (320 mm., estimated) as compared
with the total length (439 mm., estimated). (See fig.
263.) The anterior portion is too much crushed for
recognition, except that the nasals have the form char-
acteristic of Limnohyops, with more parallel sides than
in Palaeosyops. Seen from behind the occiput is more
elevated than in L. laticeps and resembles that of
L. matthewi. (See fig. 262.) The extreme height of
the sagittal crest above the foramen magnum is 133
millimeters, and above the bottom of the condyles
179; the condyles measure 99 millimeters in width.
Seen from above the nasals are narrow and long,
measuring 175 to 186 millimeters as compared with
168 in L. laticeps. In the region of the frontonasal
horn swelling there is a rugose area which may have
exhibited a rudimentary frontonasal horn.
Dentition. — The canine is exceptional in its robust
size and recurved form, the height being estimated at
41 millimeters and the diameters at the base of the
enamel being, anteroposterior, 25; transverse, 25.
The crown approaches that of Manteoceras in the
swelling of the base. The premolars are primitive
in lack of complication: (1) there is a small trito-
cone on p^, (2) p^ and p^ lack the internal cingula
entirely, a primitive condition, (3) the double ridging
of the ectolophs is a Limnohyops character. The mo-
lars, m'-m^, measuring 93 millimeters, are very distinc-
tive in form, with oblique ectolophs and prominent
parastyles. The triangidar m^ especially is of quite
different form from that of L. matthewi, the tooth
narrowing toward the inner side and having a long
oblique outer border.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
311
Limnohyops laticeps Marsh
Plates LVII, LXII; text figures 87, 92, 261, 264-266, 511, 531,
532, 714, 760
[For original description and type references see p. 160. For skeletal character
see p. 6181
Type locality and geologic horizon. — Bridger Basin,
Wyo. Marsh's Fork, the level of which is not cer-
tainly known, is the type locality. The American
Museum specimens closely resembling the type are
from Bridger C 4, D 1, and D 2.
Specific characters. — Of intermediate size; p'-m^,
153 millimeters; p^-m', 139; second and third in-
ferior premolars of more advanced type than in L.
FiGiTRB 264. — Skull of Limnohyops laticeps
Top view. One-fotirth natural size. Yale Mus. 11000 (type). Bridger
Basin, Wye; upper (?) part of Bridger formation.
priscus; p^ with well-developed tritocone; m' with a
large hypocone (fig. 265). Condyles to incisive border
410 millimeters (estimated); breadth across zygomata
320; smooth and extremely rudimentary horn swell-
ings on nasofrontal sutures. Cephalic index 75 (esti-
mated).
This was one of the earliest of the Palaeosyopinae
to be described, and for a long time it was not clearly
separated from the genus Palaeosyops. The most dis-
tinctive character assigned by Marsh was the hypocone
oa the last superior molar (fig. 265). Many of the
distinctive cranial characters were clearly pointed out
101959— 29— VOL 1 23
by Earle. Additional materials in the American Mu-
seum collections enable us to fully define this species,
especially from the full characters of the skull, the
carpus, and the manus.
Materials. — L. laticeps is represented by two speci-
mens— by the type cranium and parts of the skeleton
(Yale Mus. 11000), belonging to an individual not
fully grown, and by Am. Mus. 11710, a fragment of
the maxilla containing two molars, from level D 2,
Bridger.
Specific characters of the type. — In addition to the
specific characters enumerated above the occiput is
moderately high (144 mm. above the condyles, 118
above the foramen magnum) ; the condyles are moder-
ately broad (95 mm.); the mesostyle on p*, which is
seen as a shadow rudiment or rectigradation in L.
priscus, is here quite distinct; in p^ the tritocone is
much larger and more distinct than in L. priscus; m'
as in L. priscus is small (ap. 24 mm., tr. 27); m^ is
much larger than in that species; the metaconules
are larger than in L. priscus and confluent internally
with the hyipocones; there
is little or no diastema be-
hind the canine.
The type slcull. — Our
knowledge of the skull is
based mainly on the crushed
but very complete type
cranium of L. laticeps (Yale
Mus. 11000), which gives
us the principal characters.
(See fig. 264.) (1) The pro-
portions of the skull are
approximately the same as
in P. leidyi — namely, 410
millimeters in length and
310 across the zygomata.
(2) The skull of L. laticeps
is distinguished from that of L. priscus and approaches
that of P. leidyi in the slight narrowing of the nasals
anteriorly: posteriorly they measure 53 millimeters in
width ; anteriorly they diniinish to 43 millimeters, being
still much broader anteriorly than in P. leidyi. Other
distinctions from P. major and P. leidyi are found in the
following principal characters : (3) The sagittal crest is
very high, extending 65 millimeters above the brain
case, and thin at the summit (9 mm.), extending for-
ward a considerable distance (103 mm.) before the
crest begins to spread into the supratemporal ridges,
whereas in the least progressive specimen of P. leidyi
described below the thinnest portion of the crest
measures 13 millimeters and begins to expand rapidly
into the plane of the vertex; (4) the occiput as seen
from behind is well defined by a sharp crest and is
rounded superiorly, extending 118 millimeters above
the foramen magnum and 125 millimeters trans-
versely; (5) the postglenoid and post- tympanic proc-
esses are slightly separated; (6) the zygomata arch
widely, the malars being compressed inferiorly and
Figure 265. — Third right up-
per molar of Limnohyops
laticeps
Natural size. Am. Mus. 11710, re-
versed. Henrys Fork, Lone Tree,
Bridger Basin, Wyo.; Bridger for-
mation, level D 2.
312
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
forming a sharply convex ridge below the orbits, with
a narrow V-shaped union with the maxillaries anteri-
orly ; (7) in lateral view the nasals are slightly deciu-ved
anteroposteriorly; (8) the infraorbital foramen is above
the third and fourth premolars, or about 34 milli-
meters below and anterior to the orbit.
Dentition of the type. — The premolars of the type of
L. laticeps are much more progressive than those of
the types of L. laevidens and L. priscus. The range
of progression is parallel with that of P. leidyi in the
following respects: (1) A larger tritocone is developed
on p^; (2) a distinct mesostyle is observed in the
ectoloph of p*; (3) the protoconule of p'* is larger.
Comparison of these teeth with Leidy's superior pre-
molar and molar types of P. paludosus shows that the
ectoloph of p^ is identical in measurement in the two
specimens and that the measurements and characters
of m^ are very similar except that in the type of L.
laticeps the anterior cingulum is somewhat more
accented.
In the type of L. laticeps p^-m^ measures a few
millimeters less than in Am. Mus. 2361. The inner
side of the crown of p^ (ap. 1 1 mm.^^) is preserved, show-
ing a postero-internal cingulum; p^ is a transversely
oval tooth like that of P. laevidens; p^ (ap. 14 mm.,
tr. 19) is more progressive than in L. priscus in its
slightly better developed tritocone, more sharply
ribbed protocone, stronger internal cingula; in p*
(ap. 17 mm., tr. 24) we observe the slightly more
distinct development of the conule above mentioned
and the presence of a mesostyle on the tooth of the
right side, which, however, is wanting in that of the
left.
The molar series measures 90 millimeters in length,
and the anteroposterior by transverse dimensions of
the teeth are, m' 24 by 26 millimeters, m^ 31 by 35,
m^ 33 by 38, the teeth thus being broader than long;
the slopes of the cusps are vertically striated but less
strongly so than in Palaeosyops; both protoconules
and metaconules are present, small, of transversely
lophoid shape; in m^, however, the protoconule is
more distinct, the metaconule is vestigial, and the
cingulum rises into a distinct cingular hypocone (see
fig. 265); the vertical ridges of the ectoloph are
slightly fainter than in Leidy's cotype of P. jmludosus,
and the external crescents are slightly more open.
Jaw oj the type. — The posterior portion of the jaw
as preserved shows that the angle descends almost
vertically below the condyle and does not extend
backward so far as in Palaeosyops. The lower border
of the angle is 157 millimeters below the condyle.
The coronoid is stout, relatively low and broad.
23 The anteroposterior measurements given here and elsewhere are talien across
the middle of the crown.
Palaeosyops Leidy
Plates XVI, XXVII, XXVIII, XLIV, L, LIII, LV, LVI
LVIII-LXII; text figures 27, 28, 33, 88, 88, 89, 97, 108, 118-
120, 210, 214-220, 227, 254r-257, 266-288, 305, 405, 407, 482-
485, 508-516, 519-523, 528, 533-537, 539-550, 552, 559, 571,
645, 661, 685, 686, 703, 711, 713, 714, 716, 718, 721, 724, 727,
732, 733, 737-742, 745, 760
[For original description and type reference see p. 157. For skeletal character
see p. 619]
Type locality and geologic horizon. — Bridger Basin,
Wyo., Bridger formation, levels B, C, and D; Wash-
akie Basin, Wyo., Uintatherium-Manteoceras-Mesati-
rhinus zone (Washakie A).
Specific characters. — Brachycephalic. Grinding
teeth persistently brachyodont; metaconules on the
molars persistent or absent; third superior molar with-
out hypocone. Skull and skeleton robust; feet broad;
manus with well-developed fifth digit; lunar resting
subequally on magnum and unciform. Four sacral
vertebrae.
Geographic and geologic distribution. — This type
genus of the family Palaeosyopinae embraces a re-
markable series of stages of evolution of animals, which
are subdivided into eight species, found in ascending
geologic succession — namely, P. fontinalis, P. longiros-
trisf, P. paludosus, P. major, P. grangeri, P. leidyi,
P. roiustus, and P. copei. The members of the phylum
belong chiefly to the Bridger formation and increase
steadily in size from its base to its summit, or from
level B to D. P. grangeri and P. copei are not in-
cluded in the direct line.
The specimens listed below were collected from the
Bridger formation at the levels indicated.
13032, B 1.
10276, B 1 or B 2.
12182, B 3.
13116, B 2.
12185, C 3.
10009, C or D.
12196, C 2.
1544, C 4.
11678, D 4.
5106, D (?).
10282, D (?).
11683, D 3.
12189, C 1.
11708, D 3.
11692, B 2.
11680, B 1.
12181, B 3.
12165, B 2.
12183, B 3.
Type of P. r,
2(?).
12200, C 4.
12188, C 5.
12205 a, D 1.
Comparatively few specimens of Palaeosyops have
been found in the more easterly Washakie Basin, only
80 kilometers (50 miles) distant. This fact implies a
difference of living conditions, because the deposits of
the Washakie Basin are much less rich in fossils and
because the greater part of the fossiliferous "Washa-
kie" exposures are of more recent age than the Bridger.
Only the lower "Washakie" (horizon A) was syn-
chronous with the summit of the fossiliferous Bridger
(horizon D).
The evidence afforded by our present knowledge in-
dicates that the quadrupeds now known as Palaeosyops
disappeared from this region or died out during or
soon after the deposition of Bridger D. Palaeosyops
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
313
was by far the most abundant of the Bridger Basin
quadrupeds. The identified materials in the American
Museum from the Bridger Basin embrace parts of more
than 60 individuals, including 12 more or less com-
plete skulls.
Materials. — Altogether, more than 70 specimens
have been examined for this monograph, including
those in the collections in the American, Yale, Na-
tional, and Philadelphia Museums. Most parts of the
skeletons are dissociated from skulls; even the jaws
are rarely found with skulls; in only three of the speci-
mens of Palaeosyops studied are the jaws associated
with the skull.
The following table shows the length of the lower
grinding series, the length of the upper grinding
series, the length along the bottom line of the skull,
and the breadth across the zygomatic arches :
Comparative measurements oj species of Palaeosyops, in millimeters
Bridger D 3 P. copei Osborn
Bridger D ' P. robustus (Marsh) .
Bridger C 2-4 _
Bridger C 1 _ _ .
Bridger B 2-4.
Bridger B l-2_
Bridger B
Bridger A
Huerfano B
P. leidyi Osborn
P. grangeri Osborn..
P. major Leidy
P. paludosus Leidy _.
P. longirostris Earle.
P. fontinalis Cope
P. fontinalis Cope__.
(?)
164
151
143
170
180
174
180
155
137?
153
170
158
165
147
124?
(?)
(?)
(?)
(?)
(?)
(?)
'•440
415
(?)
389-435
(?)
(?)
(?)
146
(?)
'•340
310
(?)
' 290-335
(?)
(?)
(?)
(?)
There are steady increases in every dimension
measured until we reach the single specimen known
of the ultimate species, P. copei, from Bridger D 3, in
which a sudden falling off in size is observed. (See
also detailed table on p. 316.)
Palaeosyops in general steadily evolves from a
smaller although very robust animal into a larger, ex-
tremely broad-skulled, or brachycephalic animal.
Certain characters are persistent in all the species.
There are added through orthogenesis other new char-
acters, or "rectigradations," which distinguish the
more advanced stages from those found in the lower
levels. The chief rectigradations and allometrons are
the following: (1) Increase in size and brachycephaly
of the skull; (2) steady increase in all dimensions of
the grinding teeth; (3) certain cusps in the lower pre-
molars (p2, Ps) analogous to the paraconid, metaconid,
and hypoconid of the true molars ; (4) the second cusps
(or tritocones) on the outer wall of the second upper
premolars (p^), cusps foreshadowed in the ancestral
species; (5) the median ridges (or mesostyles) added
on the outer walls of the two posterior upper premolars
(p^, p*), which happen to be quite distinctive because
these ridges do not arise in all titanotheres as they
do in all horses; (6) the second postero-internal cusp,
or tetartocone, on the superior premolar teeth is late in
evolution in all titanotheres and only appears in very
rudimentary form in P. copei; (7) the horns, or naso-
frontal bosses, are rectigradations, which appear to
be more retarded in evolution in Palaeosyops than in
several other phyla.
Summarized, these seven principal cranial and den-
tal changes are shown below.
I. Allometrons (proportional, or metatrophic) :
1. Increase of the skull in size and in brachycephaly.
2. Steady increase in all dimensions of the grinding
teeth (see table on p. 316).
3. Cusps on the lower premolars P2, Pa analogous to the
paraconid, metaconid, and hypoconid of the true
molars; foreshadowed in P. paludosus and in-
creasing progressively in the higher stages.
II. Rectigradations (numerical, or the addition of new elements):
4. Tritocone on p^ first occurring in P. major and in-
creasing in all higher stages.
5. Mesostyles added to p', p*, first occurring in P.
leidyi and characterizing all higher stages.
6. Tetartooones first appearing on superior premolars
in P. copei (no higher stages at present known).
7. Osseous frontonasal horns first appearing in P.
leidyi and increasing in all higher stages.
Incisors. — The superior incisors exhibit low crowns
with irregularly folded posterior cingula; the opposite
series are usually separated by a median diastema and
from the canines by lateral diastemata. The incisors
increase regularly in size from i' to i'; the crowns are
convex anteriorly, with a basal subcingulate expan-
sion; the posterior faces slope gradually into an
irregularly folded basal cingulum; the lateral incisors
i' are much the largest and decidedly subcaniniform,
with obliquely sloping internal cingulum; i' is more
symmetrical, with short, compressed root; i^ is less
symmetrical, with longer root; i' is very asymmetrical,
with a very long, large root. The inferior incisors
are readily distinguished by their narrow crown,
sharply convex in front, and sloping posterior face
divided by a median ridge. As shown in P. leidyi
the crowns increase slightly but regularly in height
from ii to is. In the lower incisors the opposite ij
have very long roots nearly meeting in the midline
314
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
below; is have short roots and are being crowded out
byia.
Canines. — The superior canines are rounded at the
base and when unworn are very sHghtly recurved and
sharply pointed, differing from those of Telmatherium
Figure 266. — Lower jaws of Limnohyops and Palaeosyops
One-fourth natural size. A, L. prisms. Am. Mus. 11688 (cotype), reversed; Grizzly
Buttcs (east), Bridger Basin, Wyo.; Bridger formation, level B 2; coronoid and
condylar region reconstructed. Bi, P. copeiT, Am. Mus. 12205a; Lone Tree,
Henrys Fork, Bridger Basin; level D 1, 15 feet above white stratum. Bj, The same,
section through mj. C, L. laticeps. Am. Mus. 12201,$, reversed; Henrys Fork
Hill, Bridger Basin, level C 4; angle restored from P. paludosus, Am. Mus. 11690
in their circular section and strongly convex inner
sides. In females the canines are long, pointed, and
less robust, whUe in males they are much more rounded
and robust. The inferior canines are of very similar
form, nearly erect, with crowns rounded at the base,
differing from those of Manteoceras in the absence ol
the marked posterior expansion of the base, and from
those of Telmatherium in the rounded, nonlanceolate
form. A peculiar feature of the lower canines is the
directly internal or lingual position of the posterior
ridge, the two ridges being thus brought very close
together.
Premolars. — The law of cusp addition in the pre-
molars, which constitutes a most important means of
distinguishing the specific stages, is clearly set forth
above. In the general brachycephaly the premolars
acquire a rounded form rather than the elongate or
angulate form seen in Telmatherium. P' is either
slightly separated from the canine or placed im-
mediately behind it, according to the degree of
brachycephaly; in form it is either an extremely small,
rounded, bifanged tooth or considerably larger, with a
posterior internal cingulum in the higher stages. The
Figure 267. — Lower jaws of
, male and female
Inferior view. One-fourth natural size. A, P. paludosus, Am. Mus. 11680, S
(neotype); Little Dry Creek, Bridger Basin, Wyo.; Bridger formation, middle
beds, level B 1. B, P. copeif, Am. Mus. 12206a, V; Henrys Fork, Lone Tree,
Bridger Basin; Bridger D 1 , 15 feet above white stratum.
chief point to note in p^ is the strengthening and in-
creasing convexity of the tritocone or posterior outer
cusp (P. major). A very sharp distinction is seen in
the progressive transformation of p' and p* in Palaeo-
syops— namely, in the gradual development of a
mesostyle or median ridge separating the protocone
from the tritocone (P. leidyi), which appears to in-
dicate that in Palaeosyops the ectolophs of the premo-
lars are tending to become somewhat like those of the
molars, whereas in Manteoceras the ectolophs never
assume the molar form. The outer cusps of the
premolars of titanotheres never become crescentic
externally, however, and are thus essentially different
from those of the molars.
In general, the grinding teeth are distinguished by
persistent brachyodont or short crowns, and the
persistence of the conules is undoubtedly correlated
with the persistent breadth of the grinders.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
315
Molars. — The superior molars are progressively
distinguished (1) by the increased development of
the cingula; (2) by the intensified striations of the sides
of the cones; (3) by the robust and rounded form
of the parastyle and increasing prominence of this
style; (4) by the relative persistence of the conic or
sublophoid protoconules; (5) by the more variable,
rounded, or lophoid metaconules^*; (6) by the subtrian-
gular or subselenoid form of the hypocone in m' and m^,
the protocone remaining rounded or bunoid; (7) by
the fact that in no specimen of Palaeosyops has a
distinct hypocone on m^ been observed, although the
metaconule is often enlarged and might easily be
mistaken for a hypocone, and the posterior cingulum
is often elevated. The inferior molars are readily
distinguished progressively (1) by the rounded, bulb-
ous character of the sides of the main cusps; (2) by
the vertical grooving or striation of the sides of the
cusps; (3) by the festooning of the external cingula;
(4) by the prominence of the paraconids, the increased
development of the metastylid, entostylid, and
parastylid; (5) by the increasingly central position
and rounded form of the hypoconulid in m.3, a cusp
which is subcrescentic when entirely unworn but, as
pointed out by Earle, wears into a circle in old age;
(6) by the median ridges strengthening the internal
slopes of the protoconid and hypoconid crescents,
which are evidently dynamically correlated with the
presence of the analogous median external ridges on
the outer slopes of the paracone and metacone of the
upper molars. The internal ridges also serve to
comminute the food by pressing against the pro-
tocones and hypocones.
Slcull and jaws. — The detailed primitive and pro-
gressive characters of the jaw and of the male skull
may best be indicated under the descriptions of the
various species. In general, the skull (fig. 276) is
distinguished by (1) nasals tapering anteriorly; (2)
a prominent convexity above the orbits covering a
large frontal sinus; (3) rudimentary osseous horns
which appear on the sides of the face; (4) a relatively
short, free sagittal crest; (5) a moderately elevated
occiput; (6) deep and outwardly arched zygomatic
arches; (7) in earlier stages post-tympanic and post-
glenoid processes slightly separate; (8) a very large
mastoid foramen; (9) occipital condyles moderately
expanded; (10) constantly increasing expansion of
the masseteric insertion beneath the orbits; (11)
premaxUlary symphysis short and rounded. In the
jaw marked differences are seen between the sexes
(figs. 266, 267). In the males the chin is very deep
and prominent, correlated with the insertion of the
long fangs of the extremely robust canines. In the
females the chin is more shallow and more sloping
and rather resembles that of Limnohyops. The dis-
tinctive characters of the male jaw (fig. 268) are (1)
the marked curvature of the lower border; (2) the
great thickening of the lower border beneath the
2* These cusps appear to be entirely wanting in the primitive species Palaeosyops
fontinalis, from Bridger A.
I alveoli of the grinding teeth, which first appears
beneath the premolars and then extends progressively
backward beneath the molars; (3) the comparatively
short digastric fossa for the insertion of the digastric
muscle below the posterior portion of the symphysis;
(4) the greater distance between the posterior molar
and the back of the angle as compared with Limno-
hyops; (5) the breadth of the anterior border of the
Figure 268. — Lower jaws of three species of Palaeosyops
One-fourth natural size. A, P. longirosMs, Princeton Mus. 10275 (type), Bridger
Basin, Wyo.; B, P. paludosus. Am. Mus. 11680 (neotype), Millersville, Little
Dry Creek, Bridger Basin, level B 1; C, P. major, Am. Mus. 12181 (neotype jaw),
Cottonwood Creek, Bridger Basin, level B 3.
coronoid, which becomes progressively hollowed out
in front; (6) the reduction of the upper portion of
the coronoid process, which in some higher stages
assumes a more simple form with an abbreviated
posterior hook.
The characters of the skeleton are fully described
on page 619.
QUAHTITATITE EVOLUTION OF PALAEOSYOPS
The following table gives the measurements of 28
characters of proportion in 16 specimens, belonging to
6 species, collected at ascending geologic levels in the
Bridger formation:
316
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Evolution of proportions of Palaeosyops
[Measurements in millimeters. Level, if known, given after specimen number]
3
g
fi
a"
1
P. major
P. leidyi
P. robustus
3
1
on
.p
si
<
a
a
D
1
a
<
a
<
1
3 ■-
a""
<
3q
a
<
if
||
3
g
■3
><
0
S
a
-<
a"-
if
so
It
F
Sliull and upper teeth
Condyles to incisive border
-389
'■290
435
335
414
•276
405
416
310
397
112
196
174
158
120
73
100
19
25
"341
I
Widtli across postglenoid proc-
188
169
156
119
Pi-ms
161
144
79
71
91
18
22
25
26
32
34
1
162
150
115
- «170
174
163
127
180
165
125
71
107
21
31
30
35
38
43
P2-m'...
! 147
160
'im .
•170
169
131
P!-m'..- _ -
116
P>-p< .
70
102
20
27
■ 29
35
37
42
M>-m'
17
25
32
37
94
19
24
27
31
33
40
74
98
18
24
97
19
25
29
»31
35
39
96
18
26
29
32
32
41
102
18
25
29
32
34
39
94
17
24
-102
19
26
100
20
27
•103
20
26
30
36
37
42
96
31
34
39
36
38
74
33
37
36
40
42
Cephalic index...
1
P. paludosus
P. major
P. leidyi
■o
a.-
li
M-'
pi
1
4
1^
0
a
o
P
a
1"
a
a
i
a
2
s
a
3
SO
3 ."
a*^
1
s
a
<
1
3
a
a
1
3
a
a
1
a
r
p^
Lower Jaw and lower teeth
-340
-85
85
165
-169
152
98
17
12
26
18
43
23
370
100
97
170
182
168
112
20
14
26
19
49
358
93
98
174
193
166
107
19
14
29
19
46
Length of symphysis
Depth of ramus behind ms.
86
167
164
106
19
13
26
17
45
24
95
87
182
91
192
171
113
20
15
-107
-170
190
173
109
21
15
•90
184
91
Pi-ms. . -.
180
169
109
19
14
27
19
47
27
189
192
178
115
21
15
•126
170
113
20
14
98
19
12
24
15
41
22
118
22
15
29
19
60
26
«115
105
112
108
20
14
27
18
46
115
21
14
29
21
52
112
20
13
29
19
62
114
26
19
45
25
29
19
48
25
-50
48
26
47
25
60
60
31
52
28
'Specific reference doubtful.
The accompanying table brings out the following
facts:
1. The law of regular progressive increase in size as we pass
from P. paludosus (of Bridger B 1) through P. major (of levels
B 2-3), P. leidyi (C 2 to C 4?), to P. robustus.
2. The exceptionally large P. major skull, Am. Mus. 13116,
from B 2 interrupts this regular increase, but its tooth row is
not larger than in other P. major skulls (that is, the skull and
the teeth are differential) .
3. So far as the measurements are evidence, the larger speci-
mens of P. major overlap the smaller specimens of P. leidyi,
and the larger P. leidyi overlap the smaller P. robustus.
4. P. grangeri, although it is not so advanced in its stage of
premolar evolution as P. leidyi and although it comes from a
lower level, yet exceeds P. leidyi in size.
It
5. P. leidyi seems to be very closely allied to P. robustus.
exhibits considerable fluctuation in size.
6. P. copei, although more advanced in premolar evolution
than P. robustus, has a somewhat smaller dentition (differential
evolution) .
7. While the premolar series remains relatively stationary in
length the molar series increases from 91 millimeters in P.
paludosus to about 102 in P. robustus (differential evolution).
8. The fourth premolar and first molar exhibit a marked
increase in size, especially in width (differential evolution).
Other conclusions from the measurements are con-
sidered under the heading "Differential allometrons,"
on page 825.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
317
Palaeosyops fontinalis (Cope)
Test figures 97, 269-271
[For original description ^nd type references see p. 165]
Type locality and geologic horizon. — Bridger Basin,
Wyo.; Bridger formation, level not clearly recorded
but probably EometarTiinus-Palaeosyops fontinalis zone
(Bridger A), as the type skull is recorded from Green
River. Also recorded from Huerfano Park, Colo., in
Bridger formation, horizon Huerfano B ( = Bridger A).
The six specimens from Huerfano B (see below) range
from 250 feet to 500 feet below the top of the Huerfano
formation.
Specific characters .—01 relatively
small size. Superior molars with
sharply defined crescents and cusps;
m' extremely small (ap. 22 mm., tr. 25),
with protoconule and no metaconule.
This little-known animal resembles
Palaeosyops in its cranial structure but
differs quite widely from any known
species in the form of the cusps of its
grinding teeth. It is the smallest,
probably the most primitive, and
certainly the oldest Bridger titanothere
known, and these facts in connection
with the very low geologic level give it
great importance.
Materials. — This species is repre-
sented in Bridger (A?) merely by the
very immature skull (Am. Mus. 5107)
collected on Green River near Big
Sandy Creek. In Huerfano B it is
represented by six specimens in the
American Museum of Natural History,
as follows :
17411. Superior dentition and portion of
palate (fig. 271, A), Huerfano-Muddy divide,
2 miles west of Gardner, Colo., 414 feet
below the top of tlie Huerfano formation.
17413. Two upper molars and incisor (fig.
271, C), 3 miles north of Gardner, 400 to 500
feet below the top of the Huerfano formation.
17414. Three superior molars, fragmentary,
3 miles north of Gardner, 400 to 500 feet
below the top of the Huerfano formation.
17417. Ml, p', and milk teeth (fig. 271, B), 2 miles north
of Gardner, 400 to 500 feet below the top of the Huerfano
formation.
17425. Series of right upper grinders, p'-m^ (fig. 271, D), 2
miles north of Gardner, 400 to 500 feet below the top of the
Huerfano formation.
17450. Lower canine and fragment of ma, Huerfano-Muddy
divide, 2 miles west of Gardner, about 250 feet below the top
of the Huerfano formation.
The immaturity of the type specimen is determined
by the fact that only one true upper molar (m') has
come into use, the second molar (m^) being still
deeply embedded in the jaw. In superior view
(fig. 269, A3) the cranium is valuable as exhibiting
the suture between the supraoccipital and the parietals,
a suture which closes very early in Palaeosyops.
Similarly in the lateral view (fig. 269, Ai) the maxilla
sends out a broad, spurlike process directly on the
outer side of the malar as in Palaeosyops, instead of
on the under side as in Limnohyops. The infraorbital
bridge is rounded as in Palaeosyops, rather than
FiGiTEE 269. — Young skull of Palaeosyops fontinalis
One-third natural size. Am. Mus. 5107 (type).
Basin, Wyo.; Bridger formation, level A?.
Green River near mouth of Big Sandy Creek, Bridger
Li, Side view reversed: A2, palatal view; A3, top view.
The geologic horizons of these specimens range
rom 250 to 500 feet below the top of the Huerfano
formation.
angulate as in Limnohyops. The depressed or sessile
character of the supratemporal crest is probably due
to immaturity. Behind the post-tympanic process
the mastoid is exposed as a narrow strip.
The specimen thus probably pertains to the genus
Palaeosyops, though its grinding teeth are not entirely
of the typical Palaeosyops form ; its specific distinction
from the larger and more robust forms is readily
determined from the teeth.
Dentition. — The part of the crown of dp* that is
preserved exhibits the protocone more internally
318
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Comparative measurements of teeth of P. fontinalis, in millimeters
placed than the hypocone; this tooth is therefore fully
quadrate though not precisely molariform. The well
preserved first superior molar (m') exhibits trenchant
or pointed main crescents and cones; the ectoloph
is divided by small, sharply ridged parastyles and
Figure 270. — Upper molars of Palaeosyops
fontinalis
Natural size. Am. Mus. 5107 (type) . Green River near mouth
of Big Sandy Creek, Bridger Basin, Wyo.; level Bridger A?
mesostyles; a protoconule but no metaconule is pres-
ent; in size this tooth is diminutive, measuring (ap.
by tr.) 22 by 25 milHmeters. The a.m. /7413 i
anterior half of the second molar (m^)
is preserved; it measures 35 millimeters
from the outer side of the parastyle to
the inner side of the protocone; the
protocone and paracone are prominent;
the protoconule is reduced; the meta-
conule is not preserved.
As compared with the molars of
P. leidyi, we note the following differ-
ences: (1) Ectolophs and cusps more
trenchant and flatter; (2) outer border
of the metacone ectoloph more in-
clined; (3) styles more sharply ridged;
(4) hypocone of dp* projecting more
internally than in corresponding tooth
of P. leidyi. This species is certainly
not a typical Palaeosyops in its denti-
tion, as stated above. Measurements
may be taken from the natural-size
figures of the teeth in Figure 270.
Of the six specimens from Huerfano
B (see above) a finely preserved palate
(Am. Mus. 17411; fig. 271, A) of an
aged individual and the unworn upper
teeth (Am. Mus. 17425, fig. 271, D)
of a young individual afford a close
comparison with the two permanent teeth of the
type of Palaeosyops jontinalis from Bridger A
and are very similar both in characters and in
measurement.
Am.
Mus.
6107
(type);
Bridger
Am.
Mus.
1742S;
Huer-
fano B
Am.
Mus.
17411;
Huer-
fano B
Am.
Mus.
17414;
Huer-
fano B
Am.
Mus.
17413;
Huer-
fano B
P. palu-
dosus
Am.
Mus.
13032;
Bridger
B 1
M'-m3. _
83
16. 5
21. 5
23
26
29
34
"146
■^63
77
16.3
21. 5
22. 5
26
27
31.5
141
63
91
18
22
M', anteroposterior
M', transverse
M^, anteroposterior
M^, transverse
22
26
30
34
23.5
°26
28
32. 5
25
26
32
34
P'-m' . . _
160
P'-p^
71
General specific characters of P. fontinalis. — The dis-
tinctive specific characters of P. fontinalis are (1)
parastyle very prominent; (2) protoconules reduced;
(3) meta- and hypocones closely compressed; (4) no
Figure 271. — Teeth of Palaeosyops fontinalis
One-half natural size. Referred specimens from Huerfano B. (See p. 317.)
trace of metaconules; (5) no hypocone on m^; (6) pre-
molars very simple; (7) restored palate of considerable
breadth; and (8) the maxillaries send back a spHnt on
the outer side of the malars, as in typical Palaeosyops.
EVOLUTION OP THE SKULL AND TEETH OF EOCENE TITANOTHERES
319
Since all these six individuals from Huerfano B
agree closely in size with each other and with the type
of P.fontinalis, they establish Huerfano B and Bridger
A as the Palaeosyops fontinalis zone, distinguished by
true ancestors of Palaeosyops inferior in size to any
known members of this genus in Bridger B.
Palaeosyops longirostris Earle
Plate LVI; text figures 108, 268
[For original description and type references see p. 172]
Type locality and geologic horizon. —
Cottonwood Creek, Bridger Basin, Wyo;
Bridger formation, Palaeosyops paludosus-
Orohippus zone (Bridger B).
Specific characters. — Pi-ma, 156 milli-
meters (estimated); p2-m3, 143; mi_3, 90;
P4 rather simple. Wide posterior extension
of the ramus of the jaw behind ms. Canine
large, semiprocumbent. Mandibular sym-
physis elongate.
This little-known form appears to be
closely related to but somewhat smaller
than the typical P. paludosus. The type
specimen, the left ramus of a lower jaw
(fig. 268, A), was well described by Earle in
his memoir of 1892. No other known
material has been referred with certainty
to this species.
Palaeosyops paludosus Leidy
Plates LII, LVI, LVIII, LIX, LXII; text figures
86, 88, 267 A, 268 B, 405, 407, 732
[For original description and type references see p. 157]
Type locality and geologic horizon. —
Bridger Basin, Wyo.; Palaeosyops paludo-
sus-Orohippus zone (Bridger B). Bridger
B 1, as represented by the base of the sec-
tion of Church Buttes, is apparently the
type geologic level. The Bridger B 2
specimen is from the lower portion of the
Grizzly Buttes deposition.
Specific characters of type and of referred
specimens. — Of small size; pa-nis, 152 milli-
meters; p^-m^, 144; m2 (ap. by tr.), 33 by
20.5; second and third superior and inferior
premolars more primitive than in succeed-
ing stages ; p^ with a single external cone —
that is, no tritocone ; p^ with a rudimentary
tritocone; superior molars subquadrate and
premolars without mesostyles. No rudi-
ments of osseous horns.
Joseph Leidy, over 50 years ago, aptly characterized
this animal specifically as "swamp or marsh loving"
ijpaludosus) because its teeth are manifestly adapted to
the softer kinds of herbage.
By Leidy himself and by subsequent authors the
term "paludosus" was erroneously applied to the
more progressive species which are classified in this
monograph under P. major, P. leidyi, and P. roiustus.
Materials. — Many of the specimens other than the
type which were referred to this species in the volumi-
nous literature belong on higher geologic levels, such
as those from upper Cottonwood Creek (Bridger B
4 and 5) and from Henrys Fork (Bridger C and D),
and consequently to higher stages of evolution. P.
paludosus in the restricted sense is represented by
Leidy's isolated type teeth in the National Museum
FiGUEE 272. — Skull of Palaeosyops major
One-fourth natural size. Am. Mus. 12182 (neotype skull). Cottonwood Creek, Bridger Basin,
Wyo.; Bridger formation, level B 3. Crushed downward but slightly reconstructed from Am.
Mus. 1516 (P. leidyi), especially in the infraorbital region. Ai, Side view; As, top view.
(No. 762 in part), which are carefully distinguished
in Chapter III of this monograph. An upper dental
series probably preserved in the Philadelphia Acad-
emy collection was also referred by Leidy to P. palu-
dosus (Leidy, 1873.1, pi. 4, fig. 3) and may possibly
be conspecific with the type. In the same stage of
evolution apparently is the fine lower jaw (Am. Mus.
11680) from MillersvUle, Bridger Basin, level B, which
320
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
may be regarded as a neotype. Agreeing well with
this neotype lower jaw is the finely preserved upper
dentition Am. Mus. 13032, also from Bridger B 1.
Three other jaws (Am. Mus. 11692, 11711, from
Grizzly Buttes, Bridger B 2, and 13118, from Moun-
tain View, Bridger B 2, Grizzly Buttes) may also be
referred to this very primitive species.
The inferior molar, m^, of Leidy's original type,
now taken as the lectotype, measuring (ap. by tr.)
33 by 20.5 millimeters, exhibits crescents with striated
sides, very low antero-internal cusp, very rudimentary
metastylid, prominent entoconid, with a distinctly de-
veloped entostylid, also anterior and posterior cingula
wanting, even between the valleys.
The superior premolar type (Nat. Mus. 762, in part)
consists of the ectoloph of p^ of the left side and of
the inner half of the crown of p* of the right side,
two teeth conjoined by wax. The accurate transverse
measurement of the crown, therefore, can not be given;
the ectoloph measures 19 millimeters anteroposteri-
orly and 13 millimeters (estimated) from the base of
the crown to the tip of the partially worn protocone;
it consists of a prominent anterior style, a protocone,
sharply convex externally, spreading into an external
cingulum which surrounds the less convex tritocone, a
cone which is slightly smaller than its fellow the pro-
tocone. The inner half of the crown of p^ of the oppo-
site side consists of a deuterocone, a faint ridge ex-
tending anteriorly toward the protocone, measuring
16 millimeters anteroposteriorly; on either side are
anterior and posterior cingula which rise gently to-
ward the apex but do not tend to surround the smooth
inner side of the protocone; this condition is exactly
intermediate between that observed in L. laevidens
and P. major.
As remarked above, the association of these upper
teeth with the lower is doubtful; the upper teeth may
belong to LimnoTiyops. The second superior molar
(Nat. Mus. 758) consists only of the anterior half of
the tooth, estimated at 35 millimeters transversely;
it exhibits a stout parastyle, a median rib opposite
the paracone, a distinct and somewhat transversely
elongated protoconule, a low, conical protocone, a low
and slightly worn cingulum which rises at the antero-
internal edge of the protocone. The anterior cingu-
lum is incomplete at a point anterior to the tip of the
paracone (cf. P. major, Am. Mus. 12182). The slopes
of the cusps, like those of the molars of P. major, are
vertically striated. The anterior crescent is consid-
erably smaller, whereas in higher stages the two are
subequal.
Specific characters oj the types. — It is difficult to
define this species clearly from the lectotype specimens
which are described in detail above. Reference should
be made to the very carefully prepared natural-size
drawings of these teeth on Plates LII and LIX. The
following is a provisional definition:
Second inferior molar (lectotype) with distinct but
depressed entoconid, and metastylid and entostylid
folds; median ridges within the crescents; cingula not
prominent; dimensions, anteroposterior, 33 millimeters;
transverse, 20.5; superior premolars without trace of
internal cingulum at base of deuterocone; subequal
protocone and tritocone on p*, no mesostyle; superior
molars with moderately developed cingula and proto-
conule, moderately open external crescents.
The superior teeth referred by Leidy to this species
are the principal ones among those described by Leidy
as P. paludosus which possibly belong to this primitive
species. The figure copied herewith is taken from
Leidy's memoir of 1873 (pi. 4, fig. 3). The specimen
is recorded from Grizzly Buttes, Bridger Basin, Wyo.,
probably Bridger B 2. It is seen at once to belong to a
small animal in a very simple stage of evolution. The
premolar teeth are readily distinguished specifically,
as shown in the figure, by the extremely simple charac-
ter of the ectoloph of p^; it exhibits not even a rudiment
of the tritocone, a cusp which is strongly developed in
the geologically successive P. major. The measure-
ments of the teeth as figured on Leidy's plate, said to
be of natural size, are suspiciously small — namely,
p'-m' 137 millimeters, p^-m^ 133 — and may indicate,
if the drawing is correct, that the specimen belongs to a
smaller form, such as P. longirostris.
Oshorn's neotype and other referred material. — The
neotype jaw (Am. Mus. 11680, fig. 268, B) is referred
to the same species as the lectotype (see above) on
account of the absolute similarity in form and size of
the second inferior molar (m2) in the two specimens.
Although small it belongs to a male animal, as indi-
cated by the very deep and prominent chin and robust
canines, which measure (ap. by tr.) 24 by 22 milli-
meters at the base of the crown. The lower borders
of the rami are especially thickened beneath the pre-
molar series, the vertical depth of ramus in front of p2
being 60 millimeters and behind ms 81 or 84. While
broad, the anterior face of the lower portion of the
coronoid process lacks the concavity which distin-
guishes P. major. The grinding series, P2-ni3,
measures 152 millimeters as compared with 163 to 169
in P. major and 174 in P. rohustus. P2 is in a simple
stage, not showing the distinct rudiments either of the
paraconid or of the crescentic conformation of the
hypoconid. In ps faint rudiments of the paraconid, of
the metastylid, fold, and well-developed hypoconid
crescents are seen. In p4 all these characters are
strongly accented, but this tooth can not be described
as molariform, as it still lacks the entoconid, the eleva-
tion of the hypoconid, and the equalization of the two
lobes. The outline form of this jaw is represented in
Figures 267, 268, B, and the detailed characters of the
teeth are shown in Plates LVI, LXII, Figure 268.
Another jaw (Am. Mus. 11711) from the bluff above
Mountain View, Bridger Basin, Wyo., Bridger forma-
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
321
tion, level B, has the same general characters but is of
smaller size and unfortunately lacks the teeth.
In a third, rather young jaw (Am. Mus. 11692) from
Grizzly Buttes, level B 2, the molars are only a shade
smaller than those of the type, with which they agree
in most details; but this younger jaw is slightly more
progressive in the structure of p4, of the paraconids,
and of the metastylid folds in mi-ma.
In a fourth jaw (Am. Mus. 13118), from B 2, the
characters and measurements agree perfectly with
those of the type and neotype.
In a fifth jaw (Am. Mus. 12679), from B 3, the dental
measurements are slightly smaller than in the neotype.
The finely preserved upper teeth (Am. Mus. 13032)
fit exactly with the neotype jaw. These upper teeth
represent one of the oldest (level B 1) and certainly
most primitive Bridger titanotheres known. They
differ from the type of L. priscus in the following re-
spects:
1. They are from a lower level, very low in B 1
(Granger).
2. They are larger.
3. The premolars are far more primitive — the most
primitive known, in fact, among Bridger titanotheres.
4. The hypocone-cingule on m^ is imperfect.
They agree well with the neotype lower jaw of P.
paludosus in the following respects :
1 . They are from the same general level (B 1 ) .
2. They correspond in general size.
3. They show exact fitting of upper teeth (Am.
Mus. 13032) with lower teeth (neotype of P. paludosus,
Am. Mus. 11680) — that is, certain measurements be-
tween cusps in the upper jaw agree with corresponding
measurements between interspaces and valleys of the
lower teeth. The accuracy of this correspondence is
highly significant.
4. They show correspondingly backward develop-
ment in the upper premolars of No. 13032 and of the
lower premolars in the neotype of P. paludosus.
5. The canines are large and rounded and are charac-
teristic of the genus Palaeosyops.
In short, this specimen. No. 13032, appears to fill the
great want of an upper dentition of P. paludosus. The
presence of a cingule-hypocone on m^ does not neces-
sarily excluded it from Palaeosyops, because this cusp
is more or less variable, a fact shown by its absence
in L. monoconus and its presence in P. diaconus ( =
rolustus).
Specific cTiaracters of the neotype and other referred
specimens. — Although the first of the Eocene titano-
theres to be discovered and constituting the classic
type of the genus Palaeosyops, and also the oldest in
point of evolution, this primitive species is still im-
perfectly known because of the rarity of the speci-
mens on this low geologic level. A vast amount of
confusion has attended the previous description of P.
paludosus. We are now for the first time enabled to
characterize it sharply as a stage in which the second
and third superior and inferior premolars are ex-
tremely simple in point of cusp evolution. Referring
the reader to the previous systematic discussion in
Chapter III, we may here summarize our knowledge
of the neotype and the referred specimens. The
lower jaw (Am. Mus. 11680) is taken as the neotype, in
which m2 agrees exactly with that of the type. This
jaw exhibits the following specific characters: P2-m3,
151 millimeters; p2 extremely simple, with faint trace
of paraconid fold and noncrescentic hypoconid; ps
with rudimentary paraconid, metastylid fold, and
crescentic hypoconid ; p4 with very decided paraconid,
elevated metaconid, distinct metastylid fold, broadly
crescentic hypoconid, extremely rudimentary ento-
conid; m2 closely agreeing in form and measurement
with that of the type; ms with a narrow, subcrescentic
hypoconulid, median in position.
The following measurements of two specimens re-
ferred to Palaeosyops paludosus should be compared
with the table of measurements on page 316:
Measurements of Palaeosyops paludosus
Upper teeth, Am. Mus. 13032, Bridger B 1: Millimeters
Pi-m3 102
P2-m3 144
P>-p< 71
M>-m3 : 91
PS ap. by tr 18X22
Ml, ap. by tr 25X26
MS, ap. by tr 32X34
Neotype lower jaw, Am. Mus. 11680, Bridger B 1:
Condyle to symphysis (estimated) 340
Length of symphysis (estimated) 85
Depth of ramus behind ms 85
Condyle to angle 165
Pi-m3 (estimated) 169
Mi-ms 98
PS ap. by tr 17X12
Ml, ap. by tr 26X18
MS ap. by tr 43X23
Palaeosyops major Leidy
Plates LVIII, LXII; text figures 89, 215, 268 C, 272, 279, 515,
516, 533-535, 546, 550, 686, 721, 741
[For original description and type reference see p. 158. For skeletal characters see
p. 620]
Type locality and geologic horizon. — Bridger Basin,
Wyo. ; Bridger formation, levels B 2 to 4, Palaeosyops
paludosus-Orohippus zone Leidy's type jaw is simply
recorded from Grizzly Buttes, equivalent to Bridger
B 2. The geologic range of this species, as exposed in
the upper portions of Grizzly Buttes and the lower
portions of the Cottonwood Creek section, covers
Bridger B 2, B 3, and probably B 4.
Specific characters. — Of intermediate to large size;
total length of skull (estimated), 389 to 436 millime-
ters; p2-m3, 164; p^-m', 147; lower premolars somewhat
more progressive; superior premolars without meso-
styles; p^ with two external cones — -that is, both pro-
tocone and tritocone. No rudiments of osseous horns.
Cephalic index 74 to 77.
322
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
In 1873 Leidy named this quadruped P. major, in
reference to its larger size as compared with P. palu-
dosus, but he was unable to characterize it fuUy.
At least six rudimentary new characters, or recti-
gradations, may now be observed in the grinding teeth
alone of this mutation, or subspecific stage — so much
for numerical change. The quantitative or propor-
tional changes (allometrons) are equally significant of
progressive evolution.
Type and neotype. — Finding it impossible to define
this species from Leidy's type specimen, which con-
sists of an incomplete and abnormal fragment of a
ramus only, a neotype jaw and skull were selected by
Osborn from the same geologic level — namely, Am.
Mus. 12181 and 12182, from middle Cottonwood
Creek, level B 3 in the Bridger Basin. From these
specimens discovered by the American Museum expe-
ditions the species is defined as above. In continua-
tion of the systematic description above, P. major may
be clearly distinguished as an important early stage of
evolution, successive to P. paludosus.
Materials. — This species is represented by Leidy's
pathologic and fragmentary type ramus (Philadelphia
Acad. Nat. Sci. collection, fig. 89) and by six referred
specimens in the American Museum from levels B 2
and B 3 of the Bridger Basin, as follows: A crushed
but finely preserved skull (Am. Mus. 12182), probably
belonging with the mandibular ramus (Am. Mus.
12181), both from Cottonwood Creek, Bridger Basin,
level B 3; a stiU finer specimen (Am. Mus. 13116),
from middle Cottonwood Creek, level Bridger B 3; a
skuU (toothless) and excellent limb bones (Princeton
Mus. 10276); an imperfect palate from Smiths Fork,
probably from level B 3 of the Bridger Basin; lower
jaws (Am. Mus. 12183) from middle Cottonwood
Creek, level Bridger B 3; a fine set of lower teeth
(Am. Mus. 12165) from the same locality, level
Bridger B 2; fine jaws (Am. Mus. 5101), locality and
level unknown. The measurements of the jaws
indicated above agree approximately with the few
measurements that may be taken from Leidy's very
imperfect type, and there is consequently little doubt
about the final identification of this species, which, like
P. paludosus, has been confused with species belonging
to higher stages and higher geologic levels.
Specific cJiaracfers of type. — The only specific char-
acters that can be drawn from the fragmentary type
specimen (fig. 89) are the measurements of the lower
jaw (97 mm. below ms) and the length of the inferior
molar series (mi_3, 115 mm.).
P. major can not, however, be distinguished specifi-
cally by characters taken from the type, such as the
length of the true molar series or the depth of the
ramus behind ms, because specimens of Palaeosyops
leidyi having the same measurements are found at
higher levels. It must therefore be distinguished by
the characters of the neotype specimens found at the
same geologic level as the type.
Characters of neotype jaw. — The jaw (Am. Mus.
12181, fig. 268, C), from Bridger level B 3, middle of
Cottonwood Creek, is taken as a neotype, although it
belongs to a young adult or a female and is somewhat
less robust than the type. This jaw was found near
the skull (Am. Mus. 12182) and may belong to it.
It affords, however, the following distinctive characters
as compared with the referred jaws of P. paludosus:
Ps-ms 164 millimeters; p2 with rudimentary para-
conid, metastylid fold, and subcrescentic hypoconid
(characters all of which are wanting in P. paludosus) ;
Pa with distinct paraconid, very rudimentary meta-
conid, with metastylid fold subcrescentic and elevated
hypoconid; p4 with broadly prominent meta^'.onid,
metastylid fold, hypoconid low, broad, and relatively
elevated. More in detail, in p2 we see a tooth slightly
more progressive than that of P. paludosus in the
following respects, as shown in Plate LXII: It exhibits
a very rudimentary paraconid and rudimentary
metastylid fold and the beginning of a hypoconid
crescent, yet these rectigradations are in their very
inception. In pa the paraconid, metastylid fold, and
hypoconid crescents are accompanied by the ento-
conid in its most rudimentary form. In p^ in the
unworn condition (Am. Mus. 12165) we see a distinct
entoconid and a relatively more elevated hypoconid
than in P. paludosus. P. major is therefore dis-
tinguished as a mutation or higher stage than P.
paludosus by a number of rudimentary cusplets on
P2_4 and by the general progress of these teeth toward
the molar form. Similarly, in the molar teeth,
paraconids, striations on the sides, ridges within the
crescents, and festoonings of the external cingulum
seem more clearly defined, as well as the entostylids.
A very distinctive character also is the hollowing out
or concavity of the base of the coronoid process
behind m.3, not observed in P. paludosus; the free
portion of the coronoid process is still quite high and
recurved, not having assumed the triangular form
seen in P. leidyi; the thickening of the lower borders
of the rami now extends back below the first molar.
The characters of the jaw of P. major are also
exhibited in Am. Mus. 12183 and 5101. They are
distinguished by the following principal features:
(1) The decided curvature of the lower border; (2) the
posterior thickening of the symphysis (ap. 89 to 103
mm.); (3) the depth of the ramus behuid ma (86 mm.,
cotype, female; 96, type, male; 97, Am. Mus. 12183);
(4) the slight reduction of the free portion of the
coronoid process; (5) the deep excavation of the ante-
rior border at the base of the coronoid process.
Characters of the neotype sTcuU. — The fine skull.
Am. Mus. 12182 (figs. 272, 279), from level B 3 of the
Bridger Basin, top of Grizzly Buttes, may also be
taken as a neotype and possibly belongs with the neo-
type jaw above described. It exhibits the following
characters, which are well displayed in Plates LVIII,
LXII: Superior teeth, p^-m', 147 millimeters; no meso-
styles thus far observed on p^ very faint shadow
rudiment of a mesostyle on p*, cingulum not completely
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
323
embracing inner sides of superior premolars, tritocone
on p^ small, no external ciagulum ; ectolophs of molars
with "wide-angle" or open crescents; parastyle but-
tresses not very prominent; rudimentary external
cingula opposite valleys; the protoconules very large,
angulate; the metaconules small.
The skull is that of a young adult; in an old adult
the masseteric ridge below the zygoma would be more
strongly developed. It exhibits clearly the convexity
of the forehead, the abbreviated sagittal crest, the
broad, low occiput, the separation of postglenoid and
post-tympanic processes, the absence of even a rudi-
ment of the froiitonasal horn; and these characters in
connection with its inferior size indicate a decidedly
lower stage of development than that of the skull of
P. leidyi or P. rohustus. (See table on p. 316.)
The superior cutting teeth are unknown.
Superior premolar-molar series. — This series is well
represented in the neotype skull (Am. Mus. 12182),
possibly also in Am. Mus. 5105, Cope's paratype
of P. laevidens, and in Princeton Mus. 10276. The
molars are readily distinguished from those of P.
leidyi and P. rohustus by their smaller dimensions.
(See table, p. 316.) As pointed out above, the pre-
molars are distinguished by the absence of well
developed mesostyles in p^~^ and by the smaller size
of the tritocone of p".
Measurements of another specimen. — A bea-utifully
preserved superior premolar-molar series (Am. Mus.
2361, Pis. LVIII, LXII), probably from level A of the
Washakie Basin, is in a more advanced stage of pre-
molar evolution than P. major and differs from the
neotype of that species in having a larger tritocone
on p^, a larger deuterocone and stronger "protoconule
ridges" onp^~*, more prominent protoconule ridges on
m'~^ and a well-marked metaconule ridge on m'. The
specimen also differs in details from those referred to
P. leidyi, P. rohustus, and P. copei, and it may repre-
sent a new species or subspecies characteristic of
Washakie A. Comparative measurements of this
specimen are as follows:
Comparative measurements of species of Palaeosyops, in milli-
meters
Palaeo-
syops
sp., Am,
Mus.
2361
p. major
P.
copei?,
Am.
Mus.
13177
P. copei,
Am.
Mus.
11708
(type)
Am.
Mus.
6105
Am.
Mus.
12182
pi-m^
155
145
112
94
17
24
26
30
33
35
" 170
P2-m3 - --- -- --
''US
112
95
17
24
26
"31
38
147
116
94
19
24
27
31
33
40
17
26
26
29
" 153
P2-m2 .- . -
118
M'-mS
P^, anteroposterior
P^, transverse
M', anteroposterior
96
19
26
24
27
M^, anteroposterior
M', transverse
34
40
Palaeosyops leidyi Osborn
Plates XVI, XXVII, XLIV, L, LIII, LVI, LIX-LXII;
te.xt figures 27, 28, 33, 118, 217-220, 227, 254-256, 273-283,
305, 482, 483, 485, 511-513, 520, 522, 523, 536, 537, 539,
540, 643, 546, 552, 559, 645, 661, 703, 713, 716, 727, 737,
741, 742, 745
(For original description and type references see p. 181. For skeletal characters
see p. 620]
Type locality and geologic horizon. — Henrys Fork,
Bridger Basin, Wyo.; Bridger formation, Uinta-
therium- Manteoceras- MesatirTiinus zone. Bridger levels
C 2, C 3, C 4, and C 5?, as exposed on Henrys Fork in
the Bridger Basin, are the geologic levels of this species,
which is well above that of Palaeosyops major — ■
approximately 200 feet.
Specific characters. — Of larger size; total length of
skull 415 millimeters; p^-m', 158; P2-m3, 168. Diaste-
mata behind canines. Posterior superior premolars
with mesostyles. Barely defined swellings repre-
senting the rudiments of osseous frontonasal horns.
At least four new numerical characters, or rectigrada-
tions, in the grinding teeth. Cephalic index, 74.
This species is named in honor of Joseph Leidy, the
founder of American vertebrate paleontology and
first contributor to our knowledge of the titanotheres.
P. leidyi is noteworthy as the earliest form to have the
visible beginnings of horns (PL XVI, fig. 281). It is
the only species of which the bony structure is known
in nearly every part; this knowledge is obtained
chiefly from materials collected by the American
Museum expeditions under Dr. J. L. Wortman.
The form of the occiput in the type skull (figs. 277,
279, Bi) is very exceptional and may be an individual
variation. The premolars of P. leidyi are more
advanced than those of P. major, but this and the
incipient horn bosses are about the only decisive
characters separating the two. The average size of the
skull in P. leidyi was larger than in P. major, but
exceptionally large individuals of P. major are actually
larger than small individuals of P. leidyi (see table on
p. 316); in fact, P. major and P. leidyi might by some
systematists erroneously be regarded as successive
mutations (in the sense used by Waagen) from
P. paludosus rather than as species in the Linnaean
sense, and this conception might be conveniently
expressed by trinominal names, such as P. paludosus
paludosus, P. paludosus major, and P. paludosus
leidyi.
In the opinion of the present author we should hold
a heredity conception — that of germ evolution through
the independent advance of a very considerable num-
ber of single characters, including new rectigradations,
such as cusplets on the teeth and horn bosses on the
skull; and new proportions or quantitative characters
(allometrons). Expressed in another way, P. leidyi
succeeded P. paludosus after a vast interval of time,
as indicated by the intervening 400 to 600 feet of
sediment. In the long .series of generations that
separated these stages new tendencies of character
324
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
and proportion, such as brachycephaly, were one by
one added in heredity, so that the offspring of P. leidyi
were born on a more specialized heredity plane than
those of P. paludosus. P. leidyi was by no means the
same animal; it was an enriched stock; it possessed
in its germ substance a number of characters not found
in its ancestor, and probably it lost some other germ
characters.
Materials. — This species is by far the most richly
and abundantly represented in remains of skull,
type of P. leidyi, a broad-topped skull with lower jaw
from Henrys Fork, with a large part of the skeleton
associated, level probably upper C. Also the following
superior teeth: No. 12208, m'-m\ level C 4; No. 1552,
c-m^ left, p^-m^ right, a small female, very progressive,
level probably upper C; No. 12196, p^-m' of right side,
level C 2; also No. 1565, milk premolars and m'-m^,
from Henrys Fork, levelprobably C; in the Princeton
Museum, skull and jaws. No. 10009, level probably
upper C
Figure 273. — Skull and head of Palaeosyops leidyi
Restoration by Erwin S. Christman made under the direction of W. K. Gregory.
tooth, and skeleton. The American Museum collec-
tions contain the following principal specimens:
No. 1516, a perfectly preserved female skull, level not
definitely ascertained, probably C 4; No. 12185, a male
skull transitional between P. leidyi and P. major,
level C 3; No. 1581, a laterally crushed skull with a
convex forehead and faint rudiments of the horns,
associated with portions of the skeleton (possibly
P. rohustus), level probably Bridger C; No. 1544, the
The best specimen of a lower jaw is that associated
with the type skull (No. 1544), certainly a male.
There are also Am. Mus. 1585, 1522, 1564, 12200, all
probably males; 12197, which is in the milk stage,
corresponding closely in size with the upper milk teeth
(1565); 5103, possibly a female; and 1549, a female;
also Leidy's cotype of P. major (Acad. Nat. Sci.
Philadelphia). None of the foregoing lower jaws are
positively recorded as to level.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
325
Specific and age characters. — The materials enum-
erated above exhibit a considerable range of measure-
ment (see table on p. 316), as well as progressive
development of all the mutational characters. The
skull form in Am. Mus. 12185 is the most primi-
tive of the series and in many respects takes a
position intermediate between that of P. major and
that of P. leidyi; on the other hand, the type skull
(Am. Mus. 1544) has an extremely broad cranial roof
and differs from all known specimens of Palaeosyops in
this respect. In addition to these differences, which
are due to actual progressive stages of development,
there are apparent differences due to age. For ex-
ample, as explained above, the rudimentary cusplets
(such as the paraconid and metastylid), seen especially
on the unworn premolar teeth of young individuals,
disappear on the worn premolar teeth of old indi-
viduals. Another very important age character is the
faint frontal rugosity prophetic of a horn seen in old
male skulls such as Am. Mus. 1581 but absent in all
the younger male and female skulls. With these ex-
ceptions P. leidyi may be provisionally characterized
as follows:
Specific cliaracters of P. leidyi hased on the type. —
Premolars, especially in specimens from the upper
levels, slightly more complex than in P. major; para-
conid distinct and metaconid rudimentary on p2;
metaconid distinct on ps; entoconid distinct on p4;
tritocone larger on p^ becoming convex or ridged; p^
with mesostyle rudimentary or variable ; p^ with meso-
style distinct, sometimes large. Superior molars with
parastyle somewhat more prominent, ectoloph conse-
quently more oblique, external cingula more distinct,
all cingula heavier in specimens from the upper levels,
valleys of the external crescents somewhat narrower;
conules variable, often reduced, sometimes very large;
metaconules lophoid on m' and m^, often much re-
duced, sometimes very large on m'. In the skull,
sagittal crest variable, sometimes narrow, in the type
broadened into a fiat crest (figs. 276-279), occiput
confluent with cranial roof superiorly (in type), post-
tympanic and postglenoid approximated, almost
touching, coronoid process of jaw broadly concave
inferiorly.
Incisors. — The superior series measures 70 milli-
meters from side to side (Am. Mus. 1544); the incisor
teeth increase in size from i' to i^, transversely measur-
ing i' 11 millimeters, i^ 12, i^ 16. Similarly the inferior
incisors increase slightly but regularly in height and
breadth, the lateral teeth being either continuous with
or slightly separated from the canine. There are su-
perior diastemata between the grinding series and the
canines, also between the lateral incisors and the
canines.
Canines. — The canines are more slender in females,
measuring vertically 32 millimeters (Princeton Mus.
10009); the superior canines in this specimen measure
34 millimeters. In the males the canines are more
robust, the fangs at the base measuring (ap. by tr.) 21
by 20 millimeters, and when unworn are very slightly
recurved, sharply pointed, differing from those of
Telmatherium in their circular section and strongly
convex inner sides. In one specimen (Am. Mus. 1549)
the posterior base of the crown is horizontally grooved,
apparently as a re-
sult of use of this
tooth in uprooting
plants or pidling
down twigs.
Superior grind-
ing teeth in the type
and other speci-
mens.— P' exhibits
diastemata both tn
front and behind
(seePls.LX,LXII),
whereas in the more
progressive speci-
mens of P. rolusfus
the diastema be-
hind p^ is closed; p'
is occasionally
large (Am. Mus.
1552, 5102). The
succeeding premo-
lars are distin-
guished by sharply
convex protocones,
flattened or very
slightly convex
tritocones, internal
cingula variable,
mesostyle wanting
on p', variable,
often very distinct
on p^ ; external cin-
gula of the pre-
molar and molar
ectolophs are vari-
able but especially
strong in progres-
sive specimens; in figure 274.— Incisors and canines of
general, more Limnohyops and Palaeosyops
sharply marked One-half natural size. A, L. priscus, Am. Mus.
than in P maior ^'"^^ (type); Orizzly Buttes, west Brldger Basin,
. '. ^ ' Wyo.; Bridger formation, lower beds. B, P.
It IS a Strikmg fact Iddyi, Am. Mus. 1616; Sage Creek, Bridger Basin,
that in all the tvpi- '^' ^' '"'''''' ^™- ^^''^- IS*^ (type); Henrys Fork,
, . Bridger Bafin; probably Bridger upper C or D.
cai specimens
(Am. Mus. 1544, 1516; Princeton Mus. 10009) refer-
red to this species the metaconules are wanting, while
the protoconules are quite distinct. M^ is a large tooth
but still inferior in measurement and especially in de-
velopment of the parastyle to that of P. rohustus. The
series p^-m^ measures from 150 to 159 millimeters, as
compared with 145 to 147 in P. major and 163 to 170
in P. rohustus.
326
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Injerior grinding teeth. — The most distinctive char-
acters of the lower grinding teeth (PI. LVI) are the
prominence of the paraconid, the distinctness of the
metaconid, and the crescentic form of the hypoconid
on P2 as compared with that of P. major; in more
progressive specimens of P. leidyi p2 is almost as pro-
gressive as p3 in P. major. Similarly, ps and p^ are
gressive than those of P. major, and several stages are
represented in the five skulls described below.
Some of these stages belong to animals related to
the type of P. leidyi; others are intermediate between
P. leidyi, P. major, and P. rolustus. Of para-
mount interest is the origin of the osseous horns.
First stage: Transitional skull (Am. Mus. 12185)
from level Bridger C 3, found at the mouth
of Summers Dry Creek, appears to be the
most primitive in its dentition, ps lacldng
the mesostyle, p^ comparatively primitive,
and the premolar-molar series measuring
162 millimeters, yet the sagittal crest is
much broader (17 mm. at narrowest part)
than in the P. major hypo type skull; the
zygomata are more massive and widely
expanded, and the measurements through-
out are more robust. In this specimen
the protoconules are greatly reduced and
the metaconules are small and lophoid.
It is apparently a male, the canines meas-
uring 36 millimeters vertically and 20
across the base of the crown. The molar
crescents are of the "wide-angle" type
seen in the hypotype of P. major.
Second stage: A higher stage is repre-
sented by the female skull Princeton Mus.
10009, in which as a progressive feature a
very distinct mesostyle is observed on p*
and a rudimentary mesostyle on p^; p^-m'
estimated at 150 millimeters. The supe-
rior molars exhibit very distinct protoco-
nules but no metaconules. The lower jaw
exhibits the more oblique chin of the female
type, and the mandibular ramus meas-
ures 87 millimeters behind nis, in which
the hypoconulid is progressively conic in
form. In this specimen, however, m2 is
very simple.
Third stage : The third stage exhibits the
horn swellings without rugosity. It is rep-
resented by the finely preserved, apparently
female skull Am. Mus. 1516 (fig. 275), in
which the zygomata are moderately ex-
panded and the sagittal crest has a mini-
mum breadth of 13 millimeters. More
in detail, the superior aspect represents
FiGUEE 275. — Skull of Palaeosyops leidyi an adult but not aged animal. Although
One-fourth natural size. Am. Mus. 1516; Sage Creek, Bridger Basin, Wyo.; level probably Bridger practicaUy of the Same size in itS length (415
mm.) it is somewhat narrower (275 mm. as
against 310) than the type (Am. Mus. 1544) and ap-
pears to be in a stage only slightly advanced beyond
that of skull Princeton Mus. 10009, because the
sagittal crest is just beginning to broaden out into the
plane of the vertex, the vertex of the crest measuring
13 millimeters transversely. The supratemporal crests
are very prominent, sharply overhanging the tempo-
ral fossae and terminating anteriorly in prominent
C or D. Ai, Side view, reversed (canines from Am. Mus. 12185J; Aj, top view.
much more advanced than the corresponding teeth of
P. major. In the true molars a marked feature is the
more conic form of the hypoconulid of nis. All these
specific characters, however, exhibit fluctuations
either toward a more primitive or a more progressive
type.
Stages of evolution represented hy the sJculls. — All the
characters of the cranium of P. leidyi are more pro-
EVOLUTION OP THE SKULL AND TEETH OF EOCENE TITANOTHERES
327
\
postorbital processes. Between the orbits the skull is
broadly convex. The nasals extend 175 millimeters
anteroposteriorly; posterolaterally they send down a
broad flange beneath the adjoining parts of the max-
illaries, lacrimals, and frontals. This flange is not
(as in Manteoceras, etc.) largely covered by the forward
extension of the frontals but appears in side view as a
V-shaped area lying between the frontals and the
maxillaries, a feature which is very characteristic
of Palaeosyops and Limnohyops. The maxillaries
rise in front of this point «<^'5*?>.
and arch over the maxillary
notch. Another character-
istic feature is that the
nasals diminish toward the
tips both in width and in
the depth of the lateral
decurvature. Still another
feature is the broad en-
trance of the maxillaries
below the anterior portion
of the zygomatic arch.
Horn swellings: By far
the most important feature
of this skull is the lateral
horn swelling (fig. 275) on
each of the frontals behind
the nasofrontal sutures,
which are entirely smooth.
They would certainly never
have been observed if atten-
tion had not been directed
to this particular region of
the skull by the distinct
and rugose horn rudiments
seen in a subsequent stage
(P. rohustus).
The type stage (muta-
tion): A more progressive
stage is the type cranium.
Am. Mus. 1544. In this
male skull the frontonasal
horn swellings are so slight
that they can barely be
distinguished. This is a
very important point because in the female skull just
described the horn swellings are quite apparent. This
fact, in connection with corroborative evidence in
other phyla, tends to prove that in their inception the
horns are not sexual characters. The supratemporal
crests at the narrowest point are separated 36 milli-
meters by the broadly plane vertex of the skull, which
passes uninterruptedly into the occiput by a gentle
curve, there being no definite supraoccipital border.
Immediately behind the orbits the vertex measures
136 millimeters transversely; between the orbits, 119
101959— 29— VOL 1 24
transversely. The frontoparietal, intei'frontal, and
frontonasal sutures are all closed by age. The nasals
narrow from 76 millimeters at the broadest point
posteriorly to 47 at the tips. In the palatal aspect
we observe that the palate is relatively short and broad
and not decidedly arched, the horizontal plates of the
palatines being abbreviated. The postnarial space is
relatively short and deep but less excavated than in
Telmatherium ultimum. A relatively narrow bridge
of bone (14 mm.) separates the foramen ovale from
.pr-%
r\
yf
Figure 276. — Type skull of Palaeosyops leidyi
Am. Mus. 1544; upper part of Bridger formation, level C or D. Drawings by R. Weber. One-sixth natural size.
the foramen lacerum medium. The postglenoid proc-
esses are transversely extended but less decidedly so
than in T. ultimum. A very characteristic feature is
the broad triangular plate formed by the basioccipital
and basisphenoid, slightly keeled in the median line,
with a prominent rugosity (insertion of rectus capitus
and constrictor muscles) at the junction of the basi-
occipital and basisphenoid and a very narrow bridge
between the condylar foramen and foramen lacerum
posterius. This is very different from the more
elongate and laterally compressed region in Tel-
328
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
matJierium uitimum. The post-tympanic processes
are broadly oval, and the paroccipital processes are
more slender and acute than in T. uitimum. A dis-
tinctive feature is the relatively sharp ridge con-
stituting the inferior border of the malars from the
point where the arch leaves the skull to their junction
with the squamosals. The occipital view of the skull
fu o (J^ ^car.ex.)
Figure 277. — Type skull of Palaeosyo'ps leidyi
Top view. One-fourth natural size. Am. Mus. 1544 (type) . Henrys Fork
Bridger Basin, Wyo.; Bridger formation, level probably upper C or D.
reveals a large mastoid foramen. This is also observed
in LimnoJiyops laticeps and appears to be a distinctive
feature of this series of crania. Unlike that of L.
laticeps (figs. 261, 264) the summit of the occiput is
not defined by a crest but passes directly into the
vertex.
In the lateral view, in addition to the features
already mentioned, attention should be called (1) to
the prominent antorbital knob or process on the
lacrimals, to the abbreviation of the face, the space
between the orbit and narial notch measuring but
70 millimeters; (2) to the absence of any horn rudi-
ment; (3) to the convexity of the maxillaries as seen
in side view; (4) to the abbreviation of the premaxil-
laries; (5) to the semicircular section of the malars
below the orbits, which passes into the deep and later-
ally compressed section, with the inferior masseteric
ridge beneath and behind the postorbital processes.
The anterior and superior views of the cranium (figs.
277, 279) illustrate the characteristic rounded pre-
maxillary symphysis and the abbreviation of these
elements and the correlated rounding and depression
of the maxillaries below the narial notch.
Type lower j aw : The perfectly preserved j aw of this
specimen (fig. 280, C) exhibits the massive chin of the
male form. On the lower surface we see the short
(24 mm.) digastric fossa, the massive thickening of the
mandibular rami (35 mm.), now extending back
below m2, the curvature of the lower border of the
ramus, the increased depth (98 mm.) behind m^, the
triangular form of the free portion of the coronoid
process, the broadly transverse expansion of the base
of the coronoid process behind nis, the transverse
extension (74 mm.) of the condyles, the great elevation
(175 mm.) of the condyles above the bottom of the
angle, the extension of the angle backward to a less
degree than in Manteoceras, also on the outer surface
the subdivision of the masseteric fossa by a horizontal
xictr
\ -ms.per
Figure 278. — Type skull of Palaeosyops leidyi
Palatal view. One-fourth natural size. Am. Mus. 1544 (type). Henrys
Fork, Bridger Basin, Wyo.; Bridger formation, level probably upper C or D.
ridge extending forward and downward from the
condyle.
Dentition of the type: Diastemata appear between
the opposite incisor series and behind both the incisors
and the canines. P* is too much worn to show
the mesostyle. On the molars the conules are almost
completely worn ofi'. External cingula are well
marked, especially on m'. In the lower jaw the
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
329
incisors exhibit a compact transverse series measuring
53 millimeters. In p, the paraconid is well marked.
The worn grinding teeth, although perfectly preserved,
in such a specimen as the type do not present the
distinctive characters of the series; but they demon-
strate most conclusively that every element in the
crown finally comes into some degree of use in the
comminution of food and therefore has an adaptive
significance; the styles, the cingula, the cones, the
crescents, the conules, all are distinctly worn and
blunted either by opposing elements in the lower
teeth or by attrition of the food. The crown in this
stage of wear presents an ineffective grinding and
cutting apparatus and serves little more than a
crushing function, because the low crenulated cres-
cents and cones exhibit none of the hypsodont tendency
so characteristic of the Telmatherium series.
Fifth stage: Mutation transitional to P. robustus:
The male skull. Am. Mus. 1581 (represented in fig.
281), exhibits as its most interesting and important
feature a rugosity and a very slight elevation of the
frontals just behind their junction with the nasals,
which represents the horn rudiment in this species in
the incipient rugose stage. Viewed from above (fig. 28 1 ,
A2) this rugosity is seen to be very slightly convex —
that is, it rises above the surrounding surface as an
elevation involving the posterior border of the nasals
and a portion of the nasofrontal suture. This horn
is thus slightly posterior to the position which it occu-
pies in the skull of the allied genus Telmatherium. In
this skull the facial convexity is very prominent.
Although m3 is well worn, the conules are persistent
and very slightly affected by the wear. The meso-
style on p*, originally present, has been worn away.
Comparison with dentition of other forms: Tho
measurements of several other maxillary series are
given in the table on page 316. Among the large num-
ber of specimens examined some (such as Princeton
Mus. 10009) are smaller and are more primitive in
structure, approaching P. major, whereas others, such
as Am. Mus. 5102 (P. leidyi advanced, but no meso-
style on p*) are larger and more progressive. In the
former the ectoloph and styles of the premolars are
less strongly developed; in the latter the ectoloph,
styles, and cingula are very strongly developed. In
Am. Mus. 12208, from C 4, the protoconules are pres-
ent in the second and third molar teeth. In Am. Mus.
5102, an aberrant form, the third and fourth superior
premolars are without mesostyles but at the same
time exhibit a very marked tendency toward molari-
zation and have strongly pinched ridges opposite
the protocones and tritocones. This specimen is pro-
gressive in the great prominence of the parastyle.
Fluctuations and progressive characters in the in-
ferior teeth: In the several fine jaws referred to this
species (Am. Mus. 1585, 1564, 1546, 1549) we observe
fluctuations or individual variations of mutative and
specific characters. In P2 the paraconid is always dis-
tinctly marked; in ps the protolophid and hypolophid
crests are very well defined; in p4 the metaconid ap-
pears in a rudimentary cuspule; in ms the hypocon-
ulid varies from a rounded (progressive) to a more
crescentric (primitive) form; No. 1549 exhibits the
triangular form similar to that seen in the type jaw;
in No. 1585 we observe the more recurved form seen in
P. major.
Figure 279. — Skulls of Palaeosyops major and P. leidyi
One-fourth natural siie. A, P. major, Am. Mus. 12182 (neotype skull);
middle Cottonwood Creek, Bridger Basin, Wyo.; Bridger formation, level
B 3; occipital view, crushed downward. Bi, P. leidyi, Am. Mus. 1544
(type); Henrys Fork, Bridger Basin; Bridger, level probably upper C or
D; occipital view. B2, The same, front view.
Juvenile dentition and crania: In many ungulates
the deciduous premolars are more molariform than
their permanent successors, and this law is well
illustrated in Palaeosyops, as in the milk dentition
referred to P. leidyi (see below) dp^, dp', and especially
dp"* are more molariform than their successors p^, p', p*.
It seems not impossible that p^ in the titanotheres
represents a persistent milk tooth, or dp^ In the
milk dentition under discussion (Am. Mus. 1565) the
330
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
first premolar (dp^) is about as large as in the adult
Princeton Mus. 10009. The large alveolus of the
canine appears to leave no room for the formation of
Stage with milk teeth and one permanent molar in
use: This stage is represented by a maxilla (Am. Mus.
1565, PL LIX, figs. 282, 283), probably referable to
Palaeosyops leidyi, and by a lower jaw (No.
12197, level C 2) belonging to another indi-
vidual of the same size. In these specimens
the length of the ramus is estimated at 300
millimeters. In the upper jaw two upper
milk incisors and the milk canines are in
place; dp^ is a simple, conical tooth, smaller
than its successor p^; dp^ exhibits two ex-
ternal crescents, a parastyle and mesostyle,
and two internal cones, but the crown is
oblong and not molariform; dp' exhibits the
same elements more fullj'' developed, the
crown small and more quadrate in form,
small conules, submolarif orm ; dp* is entirely
molariform, with protoconules and meta-
conules and fully quadrate contours; m' is
also in use.
The juvenile jaw, Am. Mus. 12197, belong-
ing to an animal of exactly the same size as
that to which the above-described maxilla
pertained, exhibits the following characters:
Dp2 is of small size but similar in form and
not more progressive than its permanent
successor p^; dps and dp4 are in every respect
fully molariform.
Leidy's type of Palaeosyops Tiumilis: The
single tooth from the "Dry Creek beds,"
Bridger formation, figured in Leidy's memoir
of 1873 (1873.1, pi. 24, fig. 8) is the third
superior milk molar (dp') of the left side of
the jaw, belonging to an animal slightly
inferior in size to P. leidyi. Its agreement
in all essentials with dp' of the milk dentition,
Am. Mus. 1565, above described, makes it
clearly referable to the genus Palaeosyops,
but geologic evidence for this reference is
lacking, as its exact level is unrecorded; it
might therefore belong to P. major, P. leidyi,
or P. rohustus. Hence it seems best to regard
P. Tiumilis as an indeterminate species.
Conclusions. — (1) There is evidence in
Palaeosyops of the presence of at least two
upper milk incisors, one milk canine, and
Figure 280. — Lower jaws of Palaeosyops leidyi three milk premolars on each side above, or
One-fourth natural size. Ai, Am. Mus. 1546 (reversed); Bridger Basin, Wyo.; side view; condyle 10 rlppirliinnci tpptli nhnvo nil +ncrot>iAr' (9^
restored from Am. Mus. 6103, angle from Am. Mus. 1622. A,, The same, anterior view of section -^^ aeCmUOUS tCetU aDOVe ail tOgCtUer, {Z)
through line b. Ai, The same, anterior view of section through line c, with coronoid and condyle, the first Upper and loWCr premolars (Pt) are
B, Am. Mus. 16C4; Henrys Fork, Bridger Basin; upper Bridger; coronoid and chin supplied from . . i i • i • i
Am. Mus. 1585, incisors and canine from Am. Mus. 1544. C, Am. Mus. 1644 (type); Henrys Simple, COniCal tCCth, whlch are retained
Fork, Bridger Basin; Bridger, level probably upper C or D. i. Abnormal, extra premolar. • .1 rl 1 +
a permanent successor, or p'. The evidence is still
insufficient, however. In the Oligocene titanotheres
p' may also be a persistent milk tooth (dp'), as it is in
recent horses.
dentition — that is, they
are probably without predecessors, or milk teeth;
(3) the second lower deciduous premolar (dp2)
resembles in form its successor p2 except that the
posterior lobe is better developed; (4) in correlation
EVOLUTION OF THE SKULL AND TEETH OP EOCENE TITANOTHERES
331
with the submolariform shape of the corresponding-
upper teeth, the third and fourth lower deciduous pre-
molars (dps and dp4) are more molariform than their
successors ps, p4, especially in having large posterior
lobes with high entoconids, which latter are lacking in
the permanent teeth; (5) the second and third upper
deciduous premolars are quadricuspidate, not fully
quadrate, and may be described as sub-
molariform; (6) the fourth upper decid-
uous premolar (dp^) is fully molariform.
Palaeosyops robustus (Marsh)
Plates LV, LVI, LVIII, LXII; text figures 94,
284-288, 508-511, 521-523, 542, 544-546, 571,
685, 714, 718, 724, 737, 741, 760
[For original description and type references see p. 161. For
skeletal characters see p. 626]
Type locality and geologic Jiorizon. —
Bridger Basin, Wyo., especially the upper
exposures of Henrys Fork; Bridger forma-
tion, Vintaiherium - Manteoceras -Mesati-
rJiinus zone (Bridger D).
Specific cliaracters. — Of massive breadth
and proportions; slightly more brachy-
cephalic than P. major and P. leidyi; meas-
urements (all estimated), total length of
skull, 440 millimeters; basilar length, 440;
zygomatic breadth, 340; p^-m^, 163-170;
mesostyles variable on p^, more constant
on p*; molars with oblique ectolophs;
m^ enlarged, with prominent parastyle,
ectoloph oblique, molar conules strong;
distinctly rugose frontonasal horn
swellings.
This stage is less perfectly known than
P. leidyi, our knowledge being confined
to the structure of the cranium, of the
superior dentition, of a few of the ver-
tebrae and of numerous fragmentary limb
bones.
Materials. — The type maxillary teeth
are in the Yale University collection (No.
11122). In the American Museum collec-
tion the following are the principal speci-
mens: No. 11683, Bridger level D 3, ca-
nines to m^, progressive, close to P. robustus,
type; No. 1580, crushed skull with rudi-
mentary horns and associated skeletal
fragments, from Henrys Fork; No. 1554,
skull with rudimentary horns, same stage as type; No.
11678, Bridger level D 4, a broad skull with smaller
teeth; No. 5106, Cope's type of Palaeosyops diaconus,
from Henrys Fork, progressive; Princeton Mus.
10282b, maxillary from Henrys Fork. Also the fol-
lowing less progressive forms: Am. Mus. 1584, palate
and teeth; Am. Mus. 1552, palate and teeth, from
Twin Buttes, level Bridger C or D; Am. Mus. 1558,
also Twin Buttes, level Bridger C or D, palate with
p^-m^; Am. Mus. 1586, i'-m^ unprogressive; and Am.
Mus. 1590, fragments of lower jaw, mi_3. We still
lack the complete lower jaw.
General specific cliaracters of P. robustus. — The horn
swellings, as compared with those of three of the skulls
Figure 281. — Skulls of Palaeosyops leidyi and P. copei
robustus)
One-fourth natural size. Ai, P. leidyi, Am. IVlus. 1581; Henrys Fork Hill, Bridger Basin, Wyo.; Bridger
formation, level probably upper C or D; old male skull, side view, showing rugose horn swelling
(purposely emphasized in the drawing); skull straightened from Am. Mus. 1544. As, The same; top
view of the region of the horn swelling. B, P. copeil, Am Mus. 12205a; Lone Tree, Henrys Fork,
Bridger Basin; level Bridger D 1; top view of skull.
described under P. leidyi, are here more strongly
developed. The upper grinding series averages 6
millimeters longer than that of P. leidyi. The depth
of the ramus behind ma is estimated at 98 millimeters.
A metatrophic character is the relatively large size of
m^ measuring 42 millimeters transversely in the type,
exhibiting large conules and more distinct external
332
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
cingula. There seems to be considerable range of
variation in the size of m^, as indicated below, so that
its large size in the type may not be truly specific.
P. major, 32 by 38 to 33 by 40 millimeters.
P. leidyi, 34 by 39 to 36 by 38 millimeters.
P. robustus, 33 by 37 to 37 by 42 millimeters (type).
P. granger!, 38 by 43 millimeters.
P. copei, 34 by 40 millimeters.
The size of the conules is also variable, as would be
expected in a character which was losing its hold in
Figure 282. — Jaws and deciduous teeth of Palaeosyops leidyi?
One-fourth natural size. Am. Mus. 1565; Bridger Basin, Wyo.; upper jaw. Am.
Mus. 12197; Henrys Fork Hill, Bridger Basin; Bridger formation, level C 2; lower
jaw.
most titanotheres. The forward extension of the an-
terior part of the masseter and deepening of the malar
is a progressive metatrophy, as are also the closing
of the diastema (postcanine), the closure of the ex-
ternal auditory meatus, and the large size of the
skull. A very constant brachycephalic character is
the closing up of the postcanine diastemata, which
are either reduced or wanting The sagittal crest is
powerful but is differently
formed from that of the type
of P. leidyi, although not dis-
similar to that of other P
leidyi skulls.
In addition to the specimens
which exactly or very closely
resemble Marsh's type, there
are others which appear to oc-
cupy an intermediate position
between P. robustus and the
older form P. leidyi of level C.
Fluctuations. — The mesostyle and other premolar
characters (Pis. LVIII, LXII) in these intermediate
forms also show considerable fluctuation, but on the
whole there is a recognizable metatrophic advance
over P. leidyi. It appears that at each actual period
of geologic time Palaeosyops would show a consider-
able range of variation, partly individual, partly vari-
etal. For example, large size appears as an excep-
tional variation in a B level P. major (Am. Mus.
1.3116) and in the P. grangeri of Bridger C 1, while
small size of grinding series appears as an exceptional
character in the high level (Bridger D 3) P. copei and
in the species of Palaeosyops from the Washakie Basin.
Mutations. — Tliere seems to be considerable evi-
dence for the view that these "transitional" dentitions
and skulls bridge over the structural gap between
P. leidyi and P. robustus; indeed, it would appear
that this is clear. This view contradicts the idea
expressed elsewhere that P. leidyi "stands apart and
does not appear to form a connecting link between
P. major and P. robustus"; but that statement applies
only to the broadened occiput of the type of P. leidyi,
and since other skulls with narrow crests make up
the bulk of the species P. leidyi, too much should not
be made of the exceptional condition in the type.
It may well be that in one or two trifling characters
P. robustus may be shown ultimately to be descended
not from the true race of P. leidyi which lived at
Henrys Fork Hill during Bridger C 3 time but from
some other race of P. leidyi living to the north and
perhaps during Bridger C 1 to 3 time. However, by
such hairsplitting we obscure the grand evolution
lesson that P. major, leidyi, and robustus form suc-
cessive mutations which are very nearly if not quite
in a direct line, which might perhaps have been desig-
nated by trinomial names such as P. paludosus
paludosus, P. paludosus major, P. paludosus leidyi,
and P. paludosus robustus.
Less progressive mutations, transitional from the P.
leidyi stage. — Some of the less progressive forms
are so much more primitive than the typical P.
Figure 283. — Deciduous cheek teeth of Palaeosyops leidyi?
Am. Mus. 1565. Natural size. The identity of mi, m' is positively established by comparison with the adult dentition.
The deciduous molars dp^, dp', dp*, are more molariform than the teeth which succeed them, p2, ps, p<. In this speci-
men the alveolus for the permanent canine lies closely appressed to p^ while p> has been forced out into association
with the milk molars dp^, dp', dp*. Thus in this middle Eocene titanothere the relations of the milk and permanent
teeth are the same as in the Oligocene titanotheres.
robustus that they might be placed with equal exactness
in P. leidyi. They are especially interesting biologi-
cally in demonstrating the gradual inception of such
specific characters as are seen in Palaeosyops robustus
rather than the sudden saltation of this species out
of its predecessors. In comparing the following four
mutations we note especially the very gradual evolu-
tion of the rectigradations — namely, of the premolar
mesostyles — also the gradual atrophy of the conules.
First mutation: Of these the least progressive is
Am. Mus. 1586, consisting of a palate with full denti-
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
333
tion. There is a short diastema behind the canine;
p' is of small size. The following measurements are
much inferior to those of the type of P. roiustus:
P* 23 millimeters, transverse; width of last molar 39,
-Fragments of jaws of
Palaeosyops
One-fourth natural size. A, P. granger!, Am. Mus'
12189 (type); Twin Buttes, Bridger Basin
Bridger formation, level CI. B, P. roSus/MS, Am-
Mus. 1590; Bridger Basin, Wyo.; inner view.
as in P. leidyi; small conules on the molars; a faint
"shadow" rudiment of the mesostyle on p^.
Second mutation: Am. Mus. 1558 also exhibits a
faint "shadow" mesostyle rudiment on p^; a very
strong mesostyle on p*. The following measurements
are similar to those of P. leidyi: Width of p* 26 mil-
limeters; width of m' 39. Small conules on the molars.
P' is unusually large.
Third mutation: Am. Mus. 1552 is the palate of a
small female individual. Premolars and molars ex-
hibit strong external cingula; mesostyle is rudimentary
but distinct on p^, very strong on, p*. The measure-
ments are, however, the same as in P. leidyi, namely,
p*, transverse, 23 millimeters; m^ transverse, 39.
Fourth mutation: Am. Mus. 1584 exhibits an
advanced mesostyle on p^ and p''. The diastemata
are closed up, as in the type of P. roiustus, and p*
shows a slightly increased width (27 mm.) transversely.
The molars exhibit very small conules and faint
cingula.
Detailed characters of the type and other progressive
forms. — The dentition of the type specimen is fully
described above and figured on Plates LVI, LVIII.
In this specimen p* attains a width of 26 millimeters,
and m^ a width of 41; the measurement of m^ as
seen in its oblique diameter, measured from the para-
style to the hypocone, is 52 as compared with 48
in P. leidyi and 56 in the type of P. grangeri. This
diagonal expansion of m^ is characteristic of the spe-
cies. The mesostyles of p^ and p* are worn off or
possibly were not present in the type. The cingulum
nearly closes in around the inner sides of the premolars.
Both pro to- and metaconules on the molars are large.
Stages similar to the type : Very close to the stage
represented by the type is the dental series Am. Mus.
11683 (level Bridger D 3), measuring, p^-m^ 167
millimeters; width of m', 42; width of p"*, 27.
The cingula are progressive on the inner side of p*;
as in the type the postcanine diastema is very narrow,
and, as observed above, the parastyle expansion is very
marked. The masseteric ridge of the malar is very
deep below the orbit.
In a similarly advanced stage is Princeton Mus.
10282b, with heavy cingula and a large mesostyle
on p*.
Cope's type of Palaeosyops diaconus. Am. Mus.
5106 (Henrys Fork, Bridger Basin, Wyo., level
Bridger D?), is also very progressive, with "shadow"
mesostyle on p' (nearly worn off) and p^. The
internal premolar cingula are nearly in contact on the
inner sides of the deuterocones; similarly the cingula
nearly embrace the protocones of the molars internally.
The transverse measurement of p* is 26 millimeters.
FiGtiRE 285. — Skull of Palaeosyops rohustas
Top view. One-fourthnatural size. Am. Mus. 1S64; Henrys Fork,
Bridger Basin, Wyo.; Bridger formation, probably level D. A-A,
Section tine across born swelling (cf . fig. 210) .
A pecuhar feature is the expanded metaconule of m^
which is unfortunately broken away in the posterior
half. Cope believed that this metaconule represented
a second internal cusp or hypocone, as in Limnohyops;
334
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
but by comparison with other specimens this cuspule
is seen to be certainly a metaconule and not a true
hypocone.
A similar " pseudo-hypocone " condition of the meta-
conule on m^ is also observed in a skull from Henrys
Figure 286. — Hyperbrachycephalic old male skull
robustus
One-fourth natural size. Am. Mus. 1580; Henrys Fork, Bridger Basin, Wyo.; Bridger formation.
conules, prominent parastyles, little or no postcanine
diastema. (See fig. 286.)
The close concurrence of measurements and pro-
gressive and retrogressive characters in the above-
mentioned specimens with those of the type of P.
rohustus fully establishes this species as a dis-
tinct stage of evolution.
Jaws. — The jaws of this species are not yet
fully known. There are portions of the rami
of the young adult (Am. Mus. 1590) contain-
ing mi_3, which measure 119 millimeters (esti-
mated), as compared with 107 in P. leidyi.
In these molars the cingulum is progressive and
the paraconids are large. These teeth appear
to correspond with the P. rohustus stage.
Sicull of P. rohustus. — Three skulls are at-
tributed to this species in the American Mu-
seum collection, namely, Nos. 11678, 1554, and
1580.
Horns. — The matter of chief interest is the
structure of the osseous horn knobs. Kudi-
mentary frontal horns are seen on Am. Mus.
1554 (fig. 285), which are even more prominent
than in the transitional skull Am. Mus. 1581
attributed to P. leidyi. The position of the
protuberances is on the frontals behind the
nasal sutures; they are a little more posterior
in position than the rugosities observed in P.
leidyi, the center of the protuberances being 18
millimeters behind the suture and 59 milli-
meters from the median or internasal suture.
The convexity of the horn is a complete oval,
approximately 29 millimeters in diameter in
transverse and longitudinal sections (PI. XVI).
The frontal bones are thickened and more can-
cellous beneath the horn.
In a very aged and robust skull. Am. Mus. 1580
Palaeosyops
probably level D. Skull crushed downward. Shows horn swelling (?i) and extreme rugosity, fr- 286) which is COVCrcd with CXOStOSCS the horn
Fork (Am. Mus. 11678), level Bridger D 4; this tooth,
like that in Cope's type of P. diaconus, entirely lacks
the true hypocone. Although this skull is a male its
molar-premolar series is relatively short, the three true
molars measuring only 94 millimeters, as compared
with 101 in the type of P. rohustus. Mesostyles on
p'"^ if originally present are worn away. Progressive
features are the marked external cingula of the molars.
This important skull is illustrated in Figure 287.
A skull that exhibits rudimentary horns (Am. Mus.
1554) shows in its detailed measurements (p^, trans-
verse, 28 mm. (estimated); m^, transverse, 40; m',
oblique, 52) close approximation to the type measure-
ments. The mesostyle is absent on p", wanting or
worn off on p^ (See fig. 285.)
A more robust skull with horns (Am. Mus. 1580),
Henrys Fork, level Bridger D?, exhibits measurements
(p^ 28 mm.; m^ transverse, 41; m^ oblique, 52)
which are very close to those of the type; also large
knobs are more prominent and rugose and still more
Figure 287. — Basicranial region of Palaeosyops rohustus
One-fourth natural size. Am. Mus. 11678; Henrys Fork, Lone Tree, Bridger Basin,
Wyo.; Bridger formation, level D 4.
posterior in position than in the specimen above
described, the center being 23 millimeters back of the
frontonasal suture.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
335
Cranial evolution. — The breadth of the skull, even
in the earlier stages (P. major), is the most conspicuous
feature of the quantitative and differential evolution.
The comparative measurements in the three succes-
sive stages are as follows :
Measurements of skull in species of Palaeosyops, in millimeters
P. major (Am. Mus. 12182)
P. major (Am. Mus. 13116)
P. leidyi (Am. Mus. 1544)
P. robustus (Am. Mus. 11678)__
Longitu-
dinal
Trans-
verse
390
''290
435
335
415
310
"440
°340
Assuming that these measurements are fairly repre-
sentative, between P. major and P. rohustus we note
little if any rise in the cranial index — that is, relative
increase in breadth over length of skull.
In the same period the grinding series (p--m') has
increased in length from 10 to 15 per cent over that
of P. major, or about as rapidly as the cranial length.
Prominent features of the aged skull (Am. Mus.
1580) are (1) the width and power of the zygomatic
arch, including the deepening of the forward extension
of the masseteric insertion, which now has a depth of
62 millimeters below the orbits; (2) pterygoid wings of
the alisphenoid are very heavy, for the insertion of the
external pterygoid muscles opposing the temporals and
masseters; (3) sagittal crest, while largely broken away,
apparently broadened, as seen in the aged skull; in the
younger skull (Am. Mus. 1554) still narrow (11 mm.);
(4) occiput apparently broad and low, not confluent
superiorly with the vertex of the cranium, as in the
type of P. leidyi, resembling rather that of P. major,
with the broadly flaring pillars above the condyles;
(5) similarly paroccipital and post-tympanic processes
suturally separate, as in P. major, and not closely
conjoined, as in P. leidyi. Viewed from below the
basioccipitals are sharply keeled, the keel bifurcating
posteriorly into the occipital condyles, as in P. major,
and dissimilar from the same region in the type of
P. leidyi. While these advances upon P. leidyi are
bridged over by several other skulls, yet they are all
metatrophic and thus significant.
These characters (assuming them to be specific and
not merely individual) would seem at first to indicate
that P. rohustus is to be regarded as a successor of
P. major rather than of P. leidyi, and that in the
intermediate levels (Bridger C) we should look for the
species contemporaneous with P. leidyi but directly
intermediate between P. major and P. robustus. In
this connection, however, we should bear in mind the
apparent variability in metatrophic characters which is
displayed in the skulls referred to P. leidyi (see also
remarks under "Mutations, " above).
Nasals. — Skull Am. Mus. 1510 exhibits the nasals
(fig. 288) robust and tapering anteriorly, strongly
arching from side to side and anteriorly, extremely
solid in section.
Palaeosyops granger! Osborn
Plates LIX, LXII; text figures 119, 284 B
[For original description and type references see p. 181]
Type locality and geologic horizon. — Bridger forma-
tion, UinfatJierium- Manteoceras- MesatirJiinus zone,
level Bridger C 1, is recorded as the geologic horizon
of this species. The type is from Twin Buttes,
Bridger Basin, Wyo., 200 feet below the "red
stratum. "
Specific characters. — Exceeding P. rohustus in certain
dental proportions; p^-m^, 165 millimeters; fourth
Figure 288. — Nasals of Palaeosyops
rohustus
One-fourth natural size. A, Yale Mus. 11122 (type).
Bridger Basin, Wyo.; upper (?) part of Bridger
formation. Bi, Am. Mus. 1510; Bridger Basin,
top view. B2, The same, side view.
superior premolar enlarged; molars with extremely
prominent parastyles and oblique ectolophs.
This species was named in honor of Walter Granger,
associate curator of fossil mammals in the American
Museum of Natural History, whose explorations have
done so much to advance our knowledge of the Bridger
titanotheres and of Bridger stratigraphy. The ani-
mal seems to be a collateral rather than a main-line
form, distinguished by several peculiarities of its
grinding teeth.
Materials. — The only specimen known is the type,
consisting of a palate and grinding teeth, with por-
tions of the jaw and skull (Am. Mus. 12189).
General specific characters. — The species appears to
be collateral to the stage represented by P. rohustus.
It comes from Bridger C and was found 200 feet
below the "red stratum"; there is no exposure of
Bridger D at Twin Buttes. It is more progressive
than P. rohustus, chiefly in its enormous size, for the
premolars (p^~^) are quite backward in development,
in both the ectoloph and internal border.
336
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
P. grangeri may be a descendant of some large
strain of P. major, such as Am. Mus. 13116; in fact,
very projecting parastyles are seen in Am. Mus.
12185, transitional between the P. leidyi and the
P. major stage, and also in Am. Mus. 12680, from B 5,
referred provisionally to P. paludosus.
The species is especially distinguished by the
extremely prominent parastyles of the niolar teeth,
which result in the very oblique direction of the
ectoloph. The principal measurements are, p'-m^
180 millimeters, breadth of p* 31, of m^ 43, oblique or
diagonal measurement of m' 57. The grinding teeth
form a continuous series behind the large and laterally
compressed canines. The single incisor preserved
exhibits a subcaniform crown, 20 millimeters in
height; the mesostyles are not observable on p^ or p''
but were possibly present (though small) in the
unworn condition. The molars exhibit sharply defined
median ridges in the valleys of the ectolophs. Strong
development of the cingulum, which surrounds the
entire crown excepting only the inner sides of the
protocones, is a very characteristic feature. The
conules are also well developed but relatively less
than in typical specimens of P. roiustus. In p^ the
internal cingulum is complete though faint.
The fragment of the lower jaw which has been
preserved (fig. 284) indicates that the thickening of
the lower border which we have traced in successive
stages from P. major is now carried back below m.3.
The ramus measures 86 millimeters below m^; mi_3
estimated at 126 millimeters. The anterior lobe of
nis measures 31 millimeters transversely. This tooth
has strongly striated sides, festooned external cingula,
and strong entoconules.
Palaeosyops copei Osborn
Plate LX; text figures 120, 266, 267, 281, 484, 511-514, 519,
543, 547-550, 724
For original description and type references see p. 181. For skeletal characters
see p. 629]
Type locality and geologic Tiorizon. — Henrys Fork,
Lone Tree, Bridger Basin, Wyo.; Bridger formation,
Vintatherium- Manteoceras- MesatirM.nus zone, level
Bridger D 3. Probably also from level A of Washakie
Basin, Wyo.
Specific characters. — Tooth row of somewhat smaller
size. The most progressive species of Palaeosyops
known in superior premolar and molar evolution.
Heavy cingula embracing the inner sides of the
crowns. P*, p', p^ very advanced, with subquadrate
contours and subequal protocones and tritocones, in-
cipient tetartocones on p*, and distinct tetartocone
constriction on p^ (rectigradations).
This little-known animal represents a most ad-
vanced stage. It is, so far as known, the terminal
stage of Palaeosyops evolution. In view of its pro-
gressive character this species is appropriately named
in honor of Edward D. Cope, one of the founders of
American vertebrate paleontology and the describer
of Lambdotheriura, "Palaeosyops" horealis, and other
species of Eocene titanotheres.
Materials. — This species is positively known only
from the American Museum series of superior grinding
teeth (No. 11708) from Lone Tree, Henrys Fork,
Bridger level D 3, including the premolars and molars
of opposite sides (PL LX; fig. 120). Detailed meas-
urements are given above. As shown in Plate LX
this is by far the most specialized or advanced of the
species of Palaeosyops in respect to the molarization
of the premolars. It shows the following features:
(1) The cingida are carried broadly around the inner
sides of p'""*, a character approached but not so fully
attained in any of the previous stages of the evolution
of the premolar teeth of this genus; (2) a rudimentary
tetartocone is present on p^, as indicated by a con-
striction of the deuterocone to form this cusp, very
apparent on the outer side of the deuterocone and less
strongly marked on the inner side; (3) the decided
convexity of the protocone and tritocone ridges of
the ectoloph approaches that of some of the uppermost
Eocene titanotheres and is quite different from that in
P. rohustus or P. leidyi; (4) p^ is a very progressive
elongate tooth (17 mm. as compared with 12 in P.
leidyi), with rudimentary deuterocone; (5) the molar
cingula are very broad and heavy, continuous around
the inner side of the protocone in m^; (6) the inner
side of all the premolars is more filled out, more sub-
quadrate.
P. copei is also very probably represented in Washa-
kie A by Am. Mus. 13177, a very aged skull, in which
the teeth, so far as preserved, closely resemble those
of the type but are a little larger. Portions of the
skull indicate an animal about the size of the P.
leidyi type, resembling the leidyi-rolustus group in
its very convex forehead, nasals, and basicranial re-
gion. The nasofrontal horn swelling was if anything
more pronounced than in P. rohustus. The nasal
sinus beneath the horn, so prominently developed in
Oligocene titanotheres, was present.
A specimen doubtfully referred to P. copei? is the
young jaw (Am. Mus. 12205a, level Bridger D 1)
that belongs with the cranium and skeleton described
on page 629. The associated top of the cranium
(fig. 281) is almost certainly that of a Palaeosyops,
but the specific reference is uncertain. In the jaw of
this specimen (fig. 266, B) the measurement from the
angle to the incisive border is 340 millimeters; p2-m3,
172. The second and third incisors are approximately
equal in size. The canine is comparatively small and
probably indicates that this animal is a female. Close
behind it is pi, followed by a narrow diastema (5 mm.).
P2 and Pa are very narrow, simple teeth, but slightly
more progressive than in Limnohyops prisons, the
metaconid being quite distinctly formed on the inner
side of ps. P4 is decidedly more progressive than that
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
337
of L. prisons, the posterior crest being well defined
and the entoconid ridge being somewhat more decided.
A larger and more progressive jaw, also of doubtful
specific reference, is Am. Mus. 12201 (fig. 266, C),
from Bridger level C 4. In this the measurement
from the condyle to the incisive border is 365 milli-
meters; p2-m3, 171. The paraconids are distinctly de-
fined on p2 and pa. The teeth are otherwise very
simple, although the posterior crescent (hypolophid)
is deepened and slightly broadened. In p4 a distinct
entoconid is seen. The true molars measure 111 milli-
meters; the ramus behind ma, 94 millimeters.
A jaw (Am. Mus. 12198) from Henrys Fork (level
D 2) is in a slightly more advanced stage of evolution
on the evidence presented in its inferior premolar
teeth.
Palaeosyops copei is the last known species of this
palaeosyopine race. Its specialized condition and its
reduction in size may indicate that it was declining
and on the point of extinction. On the other hand, it
may have migrated from this region.
SUBFAMILY TEIMATHEEHNAE OSBORN
Middle and upper Eocene titanotheres of larger size.
Skull with elongated cranial and abbreviated facial
region, mesaticephalic {Telmatherium) or sub-brachy-
cephalic {Sthenodectes) . Basicranium abbreviated.
Nasofrontal horns retarded in development. Denti-
tion of macrodont type; large cingula; incisors heavy,
sublanceolate; premolars progressively molariform;
molars large, progressively subhypsodont.
Geologic liorizon and geographic distriiution. — The
genera are Telmatherium, mesaticephalic, of levels
Bridger C 3 to Uinta C 1, and Sthenodectes, mesati-
cephalic to brachycephalic, of level Uinta B 2.
These are the least known of the middle and upper
Eocene titanotheres. Remains are infrequently found,
and the parts preserved are incomplete. The Bridger
region was probably beyond the center of their favorite
habitat. The telmatheres appear to have been the
most elegant and graceful as well as the most progres-
sive and active of the middle Eocene titanotheres; in
this respect they correspond with the Menodontinae
of the lower Oligocene. As shown in Figure 257 they
appear suddenly in the upper Bridger (levels C and
D) and extend up into Uinta C 1, which probably rep-
resents the end of upper Eocene time. No known
telmathere is directly related to the Oligocene Menodus,
yet certain telmatherines may have given rise to the
Oligocene offshoots known as the Menodontinae. The
resemblances and differences between Telmatherium
and Menodus are set forth below.
Resemllances to contemporary titanotheres. — The tel-
matheres may be regarded as intermediate in anatomy
between the Limnohyops-Palaeosyops type and the
Manteoceras type. The earliest species known, T.
cultridens, exhibits certain resemblances to Manteo-
ceras, others to Limnohyops. The skull as a whole is
long; the basicranial region is short; the zygomata
are moderately arched. A distinctive feature of the
face is that the frontonasal horn swelling is feeble or
wanting, perhaps because the horns were compensated
for by the large, powerful canine tusks.
These relatively dolichocephalic, subhypsodont,
supposedly subdolichopodal, and subcursorial charac-
ters doubtless indicate that the telmatheres frequented
firmer ground and made longer excursions for harder
kinds of food than did members of the Palaeosyops
phylum. They were also probably more intelligent
and alert. Since the rise of Telmatherium, Manteo-
ceras, and other phyla possessing relatively long-
crowned molars occurred simultaneously with the
decline of the earlier group of Palaeosyops and
Limnohyops, the general replacement of the latter by
the former may be attributed to the mechanical
superiority of their grinding teeth as well as to physio-
graphic changes from forests and lakes to more open
flood-plain country.
Compensation for small horns hy large tusTcs. — The
pronounced development of the canine tusks in the
telmatheres indicates that they were probably com-
bative and vigorous fighters; another respect in which
they resemble the Oligocene menodonts. This
development of the tusks may have compensated for
the absence or retarded development of the osseous
nasofrontal horns. In the earliest known species,
T. cultridens, this horn rudiment was evidently
represented by a slight nasofrontal convexity. The
condition of the horns in T. validum is unlcnown, but
even in a male of T. ultvmum the horn rudiment is only
slightly developed (PL XVI), while the canine tusks
and lateral incisors are both enlarged and tusldike
(PI. LV).
Two suhphyla. — There is evidence of an early
division of the Telmatherium phylum into Telmathe-
rium (mesaticephalic to dolichocephalic, index 60,
incisors moderately large) and Sthenodectes (mesati-
cephalic, index 63-65, incisors greatly enlarged).
Geologic succession. — The telmatheres appear in
Bridger C 3 and extend into Uinta C 1, through a
vertical thickness of over 1,500 feet. Their span of
life thus covers a very long period of geologic time.
History of discovery. — As stated above, the Tel-
matherium phylum is comparatively little known,
probably because the known areas of deposition did not
present a habitat favorable to these animals; they are
very rare in the upper Bridger deposits; they are as
yet unknown in the deposits of Washakie Basin; and
only a few specimens have been found in the deposits
of the Uinta Basin.
The first remains of a member of the group to be dis-
covered were the maxilla and superior teeth that
Marsh described as Telmatherium validum in 1872.
His type description was brief and was published
without illustration; the geologic entry is simply the
338
TITANOTHBRES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Bridger formation, but the type specimen may have
been found in Bridger D. A maxilla discovered by
the Princeton expedition of 1877 in Bridger C or D
was considered by Scott and Osborn to represent a
new genus of animals, to which they gave the name
Leurocephalus, the type being the species L. cultridens ;
but in 1891 Earle pointed out that Leurocephalus is
generically identical with Telmatherium. All the
exploration of many subsequent years in the Bridger
Basin has not revealed anything certainly similar to
these types. In the upper deposits of the Uinta
Basin (Uinta C or true Uinta formation) Peterson, of
the American Museum expedition of 1894, secured a
fine skull of a female specimen to which the name
Telmatherium ultimum has been given; and in the same
deposits was found the anterior portion of a male skull
of the same species (fig. 297). These skulls fortunately
throw a flood of light on the cranial structure of these
animals, which were previously Ivnown only by upper
and lower jaws.
These animals reappear (Riggs, 1912.1) in the up-
permost levels of horizon B 1 of the Uinta Basin (in
the " Metarhinus sandstones"), and in the middle of
horizon Uinta B 2 Douglass discovered in 1908 the
type of T. incisivum to which Gregory (1912.1) gave
the name Sthenodectes, a telmathere with very large
incisor teeth.
Finally a large jaw was found by Peterson in
Uinta C which apparently represents the latest known
member of this series, to which the name T. altidens
has been given by Osborn.
Irregular geologic distribution. — It is important to
note that these animals appear simultaneously with
the first species of Mesatirhinus and Manteoceras in
the upper Bridger levels; that they have not been
recorded thus far in the Washakie Basin levels nor
in the lower levels of the Uinta Basin; and, finally,
that they are sparsely found in the upper or true
Uinta levels. As above intimated, this rarity of geo-
logic distribution appears to prove that they dwelt
apart or in another food region and rarely invaded
the region inhabited by Manteoceras and Mesatirhinus.
General structure and habits. — Little can be said re-
garding the general proportions of these animals until
the skeleton has been discovered. The known indi-
viduals of T. cultridens somewhat exceed in size the
largest tapirs, T. validum is somewhat larger, and
T. ultimum of Uinta C is still larger, having a skull
about 20 inches long. In divergent adaptation the
telmatheres were probably swifter and of more grace-
ful build than Palaeosyops and Limnohyops. The
incisor, canine, and grinding teeth are much more
elevated, sharp, and trenchant (hence the specific
name T. cultridens) than in Palaeosyops and were
reinforced with distinctly defined cingula.
Phyletic affinities of the telmatheres. — The first ques-
tion that arises is, Are these animals more closely
related to Limnohyops, to Palaeosyops, or to Manteo-
ceras? The answer in brief is that although they are
somewhat intermediate in position they are related
by most of their ancestral or hereditary characters to
Limnohyops and Palaeosyops. This real ancestral
affinity was long obscured by the general mesatice-
phalic character and correlations of the different parts
of the skull, jaws, and teeth in Telmatherium, which
are the dominant distinguishing features of this
animal.
Affinities to the Palaeosyopinae. — The ancestral affin-
ities of the telmatheres to the Palaeosyopinae are indi-
cated (1) in the transversely subconvex contour above
and in front of the orbit, correlated with the very
retarded development of the frontonasal horns {T. ulti-
mum); (2) in the subrectangular, rounded rather than
shelf -like section of the malars below the orbits; (3)
in the deep, laterally compressed form of the zygo-
matic arches and the progressive development of a
vertical flange {T. ultimum); (4) in the tusklike en-
largement of the third or outer superior incisors; (5)
in the rounded rather than angular posterior borders
of the temporal fossae {T. ultimum); (6) in the ele-
vated and rounded superior contours of the occiput
(T. ultimum); (7) in the absence of distinct lateral
occipital pillars above and on either side of the fora-
men magnum {T. ultimum); (8) in the presence of two
facets for the atlas just above the foramen magnum
(T. ultimum); (9) in the shape of the base of the
skull, which is sub-brachycephalic or mesaticephalic,
perhaps the strongest indication of affinity with the
Palaeosyopinae. The concurrence of these resem-
blances in so many different parts of the skull is strong
evidence of a community of descent; moreover, the
species Limnohyops monoconus exhibits several char-
acters that are seen also in Telmatherium — (1) the
very high, rounded occiput, with thin sagittal crest;
(2) a deep superior flange on the squamosal portion of
the zygoma; (3) a similar, though less quadrate infra-
orbital portion of the malar.
Comparative indices in telmatheres and related species
Cephalic
Faciocephalic
Molar
Telmatherium ultimum
Sthenodectes incisivus
Manteoceras manteoceras.-
Palaeosyops leidyi
Palaeos3'ops major
60
63-65
60-66
70
74-75
" 49
" 48
49
46
26
"27
20-23
24-28
Influence of dolichocephaly. — It appears that the
Telmatherium craniurri is a partly elongate or drawn
out Limnohyops type of cranium, and that correlated
with this incipient dolichocephaly are the beginnings
of numerous familiar dolichocephalic characters; but
this incipient dolichocephaly affects chiefly the facial
and midcranial regions, while the base of the cranium
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
339
proper remains relatively short. In the teeth the
incipient dolichocephaly appears in the following
characters: (1) Incisors somewhat compressed, oppo-
site pairs ranged in convergent series; (2) canines
laterally compressed, or lanceolate, rather than
rounded; (3) premolars and molars generally with
elevated crowns somewhat compressed transversely,
and with decidedly compressed crescents and sharply
pointed cones; (4) conules reduced or vestigial; (5)
first inferior premolars laterally compressed, with dia-
stemata on either side; (6) molars laterally compressed.
In the skull we first observe the elongate, deep, and
narrow premaxillary symphysis and the corresponding
form of the "median suture." This is the generic
character originally pointed out by Marsh and em-
phasized by Earle, in contrast with the shallow,
rounded symphysis and median suture of Palaeosyops.
The anterior aspect (fig. 295) of the symphysis is very
characteristic of the species of this genus as compared
with Palaeosyops and Mesatirhinus but is not greatly
different from the Manteoceras type. The zygomata
bend outward widely but not so much as in Palaeosyops;
they are deeply extended vertically into flanges. The
external auditory meatus remains widely open below.
Distinctions from Manteoceras. — The distinctions
from- Manteoceras are seen in a number of prominent
characters in the Telmatherium series: (1) The horn
rudiments are less prominent and the facial concavities
less pronounced (T. ultimum); (2) the malar section
below the orbits in T. cultridens is roundly angulate
and in T. ultimum it is more rounded, approaching
that in Palaeosyops, whereas in Manteoceras it is
sharply angulate externally, foreshadowing the shelf-
like flattening and rudiment of the infraorbital shelf
which is so prominent a feature in MesatirJiinus and
Dolicliorliinus; (3) the canines are elongate, laterally
compressed, and lanceolate, while in Manteoceras they
are suboval and incurved rather than vertical; (4) the
lateral superior incisors of TelmatJierium (Pis. LV^
LXIV) rapidly increase in size, progressively becoming
caniniform (T. validum, T. ultimum), but in Manteo-
ceras the lateral incisors are moderately large and
increase in size progressively, though the disparity
between i^ and i^ is less marked than in T. cultridens:
in DolicTiorTiinus they progressively diminish in size,
but the lateral incisor, while the largest of the three,
is both relatively and absolutely smaller than in
Telmatherium and Manteoceras; (5) in Telmatherium
ultimum the ectolophs of the superior premolars (PI.
LXV) exhibit a very pronounced development of the
cingulum, which rises in a festoon upon the protocone,
producing an asymmetry of the outer face (a highly
progressive character), whereas in Manteoceras the
cingula are less pronounced and the protocones and
tritocones are less subequal on the ectoloph; (6) in
Telmatherium the deuterocones of the premolars are
longitudinally compressed (Pis. LXIII, LXIV, LXV,
fig. 291), with a tendency to a ridged apex, which
becomes more marked in T. validum and very decided
in T. ultimum, whereas in Manteoceras the deuterocones
of the premolars are more oval or conical; (7) in Tel-
matherium the mesostyles of the superior molars are
sharply compressed (Pis. LXIV, LXV, fig. 292), the
buttress rising into a horizontal ridge, which becomes
a very decided character in T. validum and T. ultimum,
while in Manteoceras the mesostyles are more robust
and rounded; (8) in the members of both genera the
conules tend rapidly to disappear owing to the lateral
compression of the crown and the elongation of the
ectoloph, but the ectolophs in Telmatherium seem to
be even more elongate, progressive, and trenchant
than in Manteoceras.
There are, however, some peculiar features which
distinguish the incipient dolichocephaly of this phyhun
from the more pronounced dolichocephaly seen in the
genera Mesatirhinus and Dolichorhinus — namely, the
free nasals are relatively short; the sagittal crest is
elongate and relatively persistent; the basicranial
region is relatively abbreviate. These differences are
consistent with the general law that dolichocephaly is
a process of differential growth of different parts of the
cranium, not all parts being elongated equally.
Affinity to Manteoceras. — There are important fea-
tures in which T. cultridens, from the upper Bridger,
the earliest known member of this series, resembles the
contemporary representatives of Manteoceras, as shown
in a comparison of Figures 290 and 308. There appears
to be a similar development of the nasofrontal horn
rudiment and a somewhat similar concavity in front of
the orbits, though unfortunately this region of the
type of T. cultridens is fragmentary (fig. 290). A de-
cided resemblance to Manteoceras and Limnohyops and
distinction from Palaeosyops are seen in the form of the
nasal bones, which in T. ultimum are elongate but
with a short free portion which is laterally decurved
and truncate instead of pointed distally (contrast
Palaeosyops).
With these exceptions the progressive affinities of
Telmatherium to Manteoceras and Mesatirhinus appear
to be adaptive and convergent characters rather than
ancestral or genetic characters.
It thus appears that the distinctions from Manteo-
ceras outweigh the resemblances and that the resem-
blances to Manteoceras and Mesatirhinus are in part
attributable to parallel or convergent adaptation, in
part to similarity of origin. '°
Progressive and conservative or stationary characters. —
It appears at present that the horn rudiments are not
progressive in the telmatheres; they are found to be
even less prominent in the Uinta T. ultimum than
in the Bridger T. cultridens — a feature possibly com-
" W. K. Gregory regards the species T. cultridens as linked by intermediate stages
(Am. Mus. 12193, 12194) with M. manteoceras and as very closely related in all char-
acters, a resemblance not due to convergence. T. cultridens, according to this view,
is intermediate between the manteoceratine and the palaeosyopine divisions.
340
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
pensated for, as above noted, by the development of
the tusks. As a second conservative character it is
important to note that the first and the second lower
premolar teeth exhibit in T. culfridens a distinctively
high, laterally compressed, and secant character, and
that vestiges of this character are conserved in the
species T. altidens of Uinta C. The free or projecting
portion of the nasals remains relatively short (T.
ultimum) .
All the other distinctive incipient dolichocephalic
characters of Telmatherium appear to be progressive:
(1) the lateral superior incisors are decidedly progres-
sive, becoming elongate and tusklike; (2) the hypso-
donty in the superior grinding teeth becomes more
marked; (3) the posterior superior premolars (p^"*)
acquire similar internal ridges longitudinally placed,
which in T. ultimum tend to develop tetartocones;
(4) there is a decided elongation of the postcanine
diastema, culminating in the very long diastema
of T. altidens; (5) there is a marked elongation of the
third inferior molar (ms); (6) while the canines of
T. ultimum are not relatively larger than those of
T. cultridens, the canines of T. altidens of Uinta C are
exceptionally large and show progressive development
of this character; (7) the ectolophs of the superior
premolars as seen in T. ultimum of Uinta C tend to
develop symmetrical convexities of protocones and
tritocones such as are characteristic of all Oligocene
titanotheres.
Sex cliaraders. — Differences in sex are indicated
very markedly in the male and female specimens of
T. ultimum in the inferior size of the canines in the
female, and apparently also in the absence or faint
development of the horn rudiments. Contrary to an
earlier opinion of the author it now appears that even
in their first development the horn swellings are less
prominent and rugose in the female than in the male
Eocene titanotheres.
Resemhlaiices to the Oligocene Menodus. — There are
many resemblances in Telmatherium ultimum to the
characters of the Oligocene genus Menodus, as follows:
(1) Middle region of the skull between the orbits and
the postglenoid processes lengthened; (2) molar series
enlarged absolutely and proportionally both in length
and in breadth; (3) canines of sublauceolate form;
(4) grinding teeth sharp and hypsodont; (5) post-
temporal and occipital regions similar in their rounded
and elevated form, also in the absence of the separate
condylar pillars at the back of the occiput and in the
presence of accessory articular facets above the fora-
men magnum; (6) zygoma deep in section in both
Telmatherium and Menodus, with a vertical flange;
(7) coronoid process high and slender.
Contrast with Menodus. — On the other hand, Tel-
matherium appears to be excluded from the ancestry
of Menodus by the wholly different trend of develop-
ment of certain parts: (1) the elongation of the post-
canine diastema seems to be a progressive feature
culminating in T. ultimum, whereas in Menodus this
diastema is much reduced; (2) the progressive increase
in size of the incisors contrasts with the extremely
vestigial condition of the incisors in Menodus; (3) all
the Uinta Basin species of Telmatherium and Dolicho-
rhinus are characterized by the great prominence of
the incisor series and by sharp constriction of the face
back of the enlarged canines, probably indicating
grazing habits, and no species seems to fulfill all the
ancestral conditions of any of the Oligocene titanotheres.
It therefore can not be said that we now know any
species of Telmatherium that would fill the ancestral
characters of the Oligocene Menodus.
Telmatherium Marsh
(Leurocephalus Osborn, Scott, and Speir)
Plates XVI, XLVI, LI, LV, LXIII-LXV; text figures 126,
127, 210, 219, 220, 223-226, 255, 256, 289-300, 508, 516,
588, 592, 593, 647, 717, 733, 735, 745
Generic characters. — Skull as a whole long, basi-
cranial region short, zygomata spreading, with deep
malar flanges; frontonasal horn swellings rudimentary
or wanting. Incisors large, with heavy posterior
cingulum; i^ very large and pointed; canines large,
compressed, pointed, with sharp anterior and posterior
borders; upper premolars relatively advanced, with
complete internal cingulum; upper molars sub-
hypsodont, relatively large, with rudimentary conules
and slender parastyles and mesostyles, m^ without
hypocone. These animals are without difficulty dis-
tinguished from the species of Mesatirhinus and
Dolichorhinus, but the earlier stages show certain
resemblances to Manteoceras manteoceras.
The known specific stages are as follows :
T. cultridens, from Bridger ?C or D. The premolar-
molar series measures 180 millimeters. The premolars
are somewhat simpler in structure than in T. validum.
The animal is inferior in size to T. validum and is of
the same size as the smaller members of M. manteoceras.
T. validum, Bridger D. The premolar-molar series
measures 195 millimeters. This animal is represented
by part of a male skull, the only specimen Icnown.
The lateral superior incisor is more distinctly canini-
form than in T. cultridens, and the premolars and
molars are somewhat more progressive.
T. ultimum, from Uinta C. The premolar-molar
series measures 217 millimeters. The lateral superior
incisors are greatly enlarged and caniniform; the
superior premolars are progressive and have more
symmetrically convex protocones and tritocones and
well-defined external cingula. The second superior
premolar is simpler than the third and fourth.
T. altidens, from Uinta C (? lower levels). The
inferior premolar-molar series measures 330 milli-
meters. The anterior premolars are primitive and
laterally compressed. The canines are exceptionally
prominent, hence the name T. altidens.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
341
Measurements of skull and teeth of species of telmatheres, in
■millimeters
Basal length of skull
Zygomatic breadth of
skull
Length of dental series
Pi-m3
P'-p<
M'-m3
I', ap. by tr__
P, ap. by tr_.
C, ap. by tr__
P*, ap. by tr_.
M', ap. by tr_
IVP, ap. by tr_
255
180
80
103
9X 8
MX 13
20X ?
21X26
29X29
36X39
270
79
113
UX 9
22X30
30X30
39X38
330
295
207
82
128
22X21
25X22
27X24
22X36
38X38
38X43
S5.
as
300+
305
218
89
130
14X14
19X18
25X35
38X40'
46X52'
229
95
137
14X14
22X20
2.5X 22
27X35
40X37
49X48
The history of this species has already been given.
(See pp. 167-168.)
Materials. — As noted above, the type (Princeton
Mus. 10027) is from the Bridger Basin, Henrys
Fork Hill, level C or D. It represents a smaller and
considerably more primitive animal than T. validum,
especially in the more incisiform character of the
superior lateral incisor. Another specimen from
Bridger C 2 referable to this species (Am. Mus. 12209)
consists of p"*, m', and m^ Another specimen that is
certainly referable to this species is a young lower jaw
(Am. Mus. 1560), with ms not yet entirely exposed,
recorded from Twin Buttes, level Bridger C or D.
Another well-preserved lower jaw (Am. Mus. 12193),
from Henrys Fork, level Bridger C 3, agrees closely
with the type in the dentition. Am. Mus. 12685,
which includes m\ with a deciduous molar, and an
unerupted p-, from Sage Creek Spring, level Bridger
C 3, may represent a primitive phase of this species.
A lower jaw (Am. Mus. 12687) from Henrys Fork Hill,
level Bridger D 3, is somewhat more progressive than
the type in p2 and ps. Another specimen from the
Bridger Basin (Am. Mus. 1546a), consists of pMn^,
B "=^ C
Figure 289. — Progressive hypsodonty of the molars in Telmaiherium
Natural size. Posterior view of third left upper molar. A, T. cuUridens, upper Bridger (C or D) ; B, T. ralidum, Bridger D; C, T. uUimum, Uinta C (true Uinta) .
The table illustrates (1) the marked increase in the
size of the skull and dentition as we pass from T.
cultridens of the upper Bridger to T. ultimum of
Uinta C; (2) the relatively larger size, in the later
stages, of the true molars as compared with the pre-
molars; (3) the increase in both the length and the
breadth of the molars.
Telmatherium cultridens (Osborn, Scott, and Speir)
(? Leurbcephalus cultridens)
Plates LV, LXIII, LXV; text figures 101, 223-226, 289-293,
735
[For original description and type references see p. 168]
Type locality and geologicliorizon. — Henrys Fork Hill,
Bridger Basin, Wyo.; Bridger formation, level C or D.
Also recorded from Bridger C 2 and probably other
horizons, as described below.
Specific characters. — P'-m', 180 millimeters. In
males superior canines elongate (46 rnm.), laterally
compressed; premolars less progressive than in T.
validum.
right and left, and the lower border of the orbit.
It differs from the type in the more complete exter-
nal cingulum on p*, but the malar closely resembles
that of the type. This specimen also approaches
M. manteoceras in some respects. A young lower jaw
from the Washakie Basin (Am. Mus. 2356), with the
milk molai's in place, is more advanced than the type
in the characters of the permanent p2.
Type of T. cultridens. — So little is preserved of the
cranium of this type (Princeton Mus. 10027) that it
can only be partially characterized. As shown in
Figure 290 there is a slight concavity at the side of
the face and distinct evidence of the existence of a
nasofrontal horn rudiment. The premaxUlary in side
view approaches the Manteoceras type but is dis-
tinguished by the greater depth and by the emphasis of
the dorsal symphyseal keel. It is more elongate and
more angulate superiorly than the short, rounded
premaxillary of Palaeosyops and is vertically deeper
than in Mesatirhinus, Metarliinus, and Dolichorhinus.
Its dimensions are, depth from symphyseal crest to
342
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
internal alveolar border 38 millimeters, length from
anterior border to the anterior edge of the canine 45,
extreme length 96. Behind the canine convexity
the sides of the maxillaries are somewhat flat. The
infraorbital foramen is well exposed on the side of the
face, the distance from the antorbital border of the
malar being 36 millimeters. Only the anterior portion
of the malar is preserved, but the relations of the malar
and maxillary are shown by close examination and
comparison to have been about as they were in
Manteoceras and Limnohyops — that is, the maxillary
contributed an antorbital process and a long internal
inferior sliver; there is no infraorbital shelf; imme-
diately below the orbits the malars are gently convex
on the outer surface and broadly flattened on the
inferior surface, the vertical extent of the outer face
pas. pci- nts.
FiGUEB 290. — Upper jaw of Telmatherium cultridens
One-half natural size. Princeton Mus. 10027 (type), reversed, showing the region of the horn swelling (A) and the overlap
of the maxilla on the nasal. The fragment of the nasofrontal region here figured is said to be associated with this type-
Upper part of Bridger formation, Bridger Basin, Wyo.
being 27 millimeters, and the transverse extent of
the slightly concave inferior face opposite m^ 23.
The teeth in general are distinguished by the
sharply defined, finely sculptured character of all their
elements.
Incisors. — In the type (Princeton Mus. 10027) the
superior incisors have the typical generic character
of the opposite sets, forming acutely convergent or
V-shaped rather than gently convergent series; the
incisors increase rapidly in size from i' to i^, the fangs
measuring S, 10, and 14 millimeters, respectively. The
antero-external faces of the crowns are readily distin-
guished from those of Mesatirhinus megarlnnus by a
sharp anterior ridge, wliich divides the crown into a
flattened external portion, feebly convex and cingulate,
and a narrow anterior portion, feebly concave. Simi-
larly the postero-internal face is flattened, with a median
basal ridge and a very prominent postero-internal
cingulum, especially upon i^ 1^, which is less per-
fectly preserved, is a large subcaniniform tooth, with
a sloping and less prominent internal cingulum. The
single infei'ior incisor preserved (is. Am. Mus. 1560) is,
in contrast to its mate above, typically incisiform,
with a more uniformly convex antero-external face,
feebly cingulate, and a nearly plane postero-internal
face, ^vith a median ridge and sessile postero-internal
cingulum; the long axis of this tooth is oblique, the
diameters being 14 by 11 millimeters.
Canines. — In the type the superior canines have not
fully emerged, the crown measuring 49 millimeters verti-
cal, 24 anteroposterior, 22 transverse (estimated); the
tusk has the true generic lanceolate or lateraUy com-
pressed character, the anterior and posterior ridges being
sharply defined and terminating in the piercing apex;
the outer face is broadly convex; the inner is more
nearly plane, with a convex median swelling. The
inferior canines belonging
to an animal of the same
size (fig. 293; Am. Mus.
1560) are somewhat smaller
(vertical 30 millimeters,
anteroposterior 17, trans-
verse 15), distinctly lanceo-
late, with a sharply defined
anterior ridge, which be-
comes especially prominent
and inflected near the base
of the crown; the posterior
ridge is much less sharply
defined.
Premolars. — The superior
molar-premolar teeth in
the type specimen (see PI.
LXIII, fig. 291) have the
laterally compressed, sharp-
cusped, secant, and pierc-
ing form so characteristic
of T. validuin, the type of
this genus; they are distinctly smaller — ISO milli-
meters, as against 195 millimeters in T. validuin.
The superior premolars (81 mm.) exhibit nearly
complete internal cingula, excepting p^ also an exter-
nal cingula except directly opposite the protocone
sweUing. P' is a narrow tooth (ap. 15 mm., tr. 9),
with sharply compressed ridges extending forward and
backward from the protocone, a rudimentarj^ trito-
cone, and well-defined but low anterior style. P^ is a
subtriangular tooth; the deuterocone in this tooth is
double, extending backward to the postero-internal
portion of the crown. As seen externally (fig. 292)
the protocones and tritocones of p^~* are of equal
height, but the convexity of the protocone broadens
characteristically at the base into an anterior and
posterior cingulum; the tritocones present narrow
vertical external ridges, which enable us to distinguish
these teeth from the typically more flattened tritocones
of Mesatirhinus mega.rhinus and Manteoceras manteo-
ceras. In p' the ectoloph exhibits the same characters
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
343
and the same length as in p^ but p^ is relatively much
broader, the length of the ectoloph being 21 millimeters
and the breadth across the crown 22; the deuterocone
is more median in position, elongate, and flattened
internally, as in T. ultimum; this tooth also exhibits
a low antero-external style. In p'' this style
is still more prominent, the protocone and
tritocone convexities of the ectoloph are
more symmetrical, and the crown seen
from above is more quadrangular, the
deuterocone being sharp and slightly flat-
tened internally.
Of the inferior premolars pi (preserved
in Am. Mus. 1560) is spaced, lying 13 milli-
meters behind the canine and 3 millimeters
in front of P2; it is laterally compressed
(12 by 7 mm.), sharply pointed, and simply
plano-convex in section, with a notice-
able paraconid, or anterior style. P2 is
elongate, laterally compressed, 22 by 10
millimeters (in Am. Mus. 1560), with an
elevated protocone, rudimentary antero-
internal cusp (=metaconid) and well-
developed postero-external cusp (=hypo-
conid), slightly concave internally. P3 ex-
hibits more symmetry; the protoconid still being more
elevated than the postero-external cusp (=hypoconid),
the proportions of the crown being a shade larger in the
type than in Am. Mus. 1560 (ap. 22 mm., tr. 14); the
analogous to the metaconid; the anterior lobe (=tri-
gonid) is much higher, however (15 mm.), than the
posterior (12 mm.).
Molars. — The superior molars (103 mm. in type)
exhibit prominent external cingula feebly continuous
P3 ^ ^p^
FiGUHE 291. — Upper and lower teeth of Telmatherium cuUridens, showing
their mechanical relations
One-half natural size. Princeton Mus. 10027 (type). A, Crown view; upper teeth (light line),
with pattern of lower teeth (heavy line) projected upon them. B, Internal view of the same,
showing the crushing action of the cones and conids.
around the styles in m\ m^, and internal cingula
nearly continuous on the inner sides of m^, m'.
The conules are represented merely by a vestige in
m'. The internal cones (protocones and hypocones)
Figure 292. — Upper and lower teeth of Telmatherium cuUridens, interlocked
Outer side view. One-half natural size. Princeton Mus. 10027 (type), reversed; upper part of Bridger formation (level C or D), Bridger
Basin, Wyo. This view shows the shearing or cutting action of the grinders by the interaction of the superior and inferior crescents
form is thus feebly molariform. In Am. Mus. 1560 the
posterior half of ps, p4 is widely expanded transversely.
P4 is submolariform (ap. 23 mm., tr. 15 in the type),
with cusps analogous to the paraconid, entoconid, and
metastylid of the molars, and a very prominent cusp
101959— 29— VOL 1 25
are sharply pointed and slightly more flattened in-
ternally than externally. In m^ the unworn proto-
cone and ectoloph measure in height 18 and 31 milli-
meters, respectively, exhibiting incipient hypsodonty.
They are sharply pointed and closely approximated.
344
TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
the apices being only 12 millimeters apart. M' ex-
hibits a rudimentary hypocone, also a rudimentary
swelling of the anterior cingulum corresponding with
the protostyle of T. ultimum and Menodus. The in-
ferior molars constitute a long, narrow series (122
mm.) in the type specimen and exhibit distinctly
defined paraconids as well as rudimentary metastj^-
lids and entostylids. In the elongate nis (54 by 22
mm.) the apex of the hypoconulid is placed nearly in
line with the outer wall, and this cone is deeply cres-
centic within.
Mechanical correlation oj upper and lower teeth. — The
significance of these accessory cuspules as well as the
mechanical relations of the upper and lower cusps is
well demonstrated in the accompanying drawings
(figs. 291, 292), which show that the metastylid be-
low serves to press the food against the protocones
and their ascending internal cingula ebove. The
adaptive significance of these minute features in the
comminution of the harder food which was probably
selected by this species is thus clearly brought out.
Figure 293. — Lower jaw of Telmatherium cuUridens
One-fourth natural size. Am. Mus. 1560. Twin Buttes, Bridger Basin, Wyo.;
upper part of Bridger formation (Bridger C or D).
It is seen also that the entoconid and paraconid
below press the food against the hypocones above,
that the tip of the hypocone fits squarely into the
antero-internal part of the basin of the trigonid, that
the single deuterocones of the superior premolars fit
into the posterior internal valleys (=talonids) of the
inferior premolars, and that the elongate p^ above is
effectively correlated with the elongate and secant
P2 below.
Millc dentition of fT. validum. — A pair of young
jaws from the Washakie Basin (Am. Mus. 2356)
exhibit dp2, dps, dp^, m', nr in place, with the true
premolars as well as nis still deeply buried in the jaw.
Each of the deciduous premolars is fully molariform,
with sharply defined double crescents; precocious
molarizatiou is, in fact, characteristic of mUk pre-
molars of titanotheres in general. Dps measures
(ap. by tr.) 19 by 11 millimeters, dp4 24 by 14. The
enamel is vertically crenulate on the outer surface.
An important fact is that this jaw is in a more ad-
vanced stage of evolution than the type of T. cul-
tridens, since the second permanent premolar (P2)
has the talonid V much better developed, and it
may therefore belong to T. validum.
Lower jaw of Telmatherium cultridens. — The par-
tially preserved type jaw (fig. 292) exhibits (1) two
mental foramina, the second indistinctly shown, the
larger and more anterior being below p^; (2) a gradual
increase in depth from 58 millimeters behind p2 to 65
behind m2 and 76 behind nia, with a thickness of 20
millimeters below mo.
A more perfectly preserved young jaw (Am. Mus.
1560, fig. 293) in which ms is not fully emerged
exhibits a long (91 mm.) and rather shallow (26 mm.)
symphysis and laterally compressed chin (47 mm.);
the ramus exhibits two mental foramina and gradually
increases in depth from 51 millimeters behind pa to 58
behind m2 (inside), the thickness being 19 millimeters
below m2. All these measurements would naturally
increase with advancing age. The depth of the angle
below the condyle is 134 millimeters; the coronoid
attains a free height of 49 millimeters and is regularly
hooked or recurved from base to tip. The angle is
thin but extended downward and backward very
decidedly, as in Manteoceras manteoceras. Compari-
son with Am. Mus. 12193 (Bridger C 3), which prob-
ably belongs with this species, brings out the differ-
ences due to age. In the younger jaw (Am. Mus.
1560) the chin and ramus are shallower, the whole
ascending ramus narrower, the angle less depressed,
the coronoid shorter, more delicate and recurved, less
truncate at top, and with the lateral flange much less
pronounced.
Measurements of Am. Mus. 12193 are as follows:
Pi-ms, 194 millimeters; pi_4, 68; mi_3, 125.
Telmatherium validum Marsh
Plate LXIV; text figures 93, 289
[For original description and type references see p. 160]
Type locality and geologic horizon. — Bridger Basin,
Wyo.; Bridger formation, Uintatherium-Manteoceras-
Mesatirhinus zone, level Bridger D.
Specific characters. — P'-m^, 195 millimeters. In
males, superior canines large, elongate (55 mm.),
lateral superior incisors subcaniniform; ectolophs of
superior premolars with sharply cingulate ridges; pre-
molars more progressive than in T. cultridens; trans-
verse measurements of p'-m' greater than in T. cultri-
dens. Frontonasal region unknown.
The only teeth definitely known are those of the
type in the Yale Museum (No. 11120), a male indi-
vidual first characterized by Marsh in 1872 and fully
discussed later by Earle. The animal is young, since
the crown of the last molar is entirely unworn, and all
the distinctive characters of the surfaces of the teeth
are still sharply defined. As noted above in the
description of Palaeosyops these surface characters
disappear rapidly by the wearing action of the food.
The type specimen represents a comparatively large
and powerful animal. The canines, relatively more
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
345
prominent than in T. ultimum, suggest a possible
affinity to T. altidens of Uinta C The lateral superior
incisors are almost as distinctly caniniform as in T.
ultimum. As compared with T. cultridens, distinctive
characters are the more progressive rectigradations,
seen principally in the premolar teeth, as enumerated
below, also the greater width of the premolars. From
M. manteoceras this animal is readily distinguished
by its very long and less curved canines.
A comparison of the detailed measurements of the
teeth is given in the following table :
Measurements of teeth in Telmatherium validum and T. cultri-
dens, in millimeters
T. validum, Yale
Mus. 11120 (type)
T. cultridens, Am.
Mus. 1560
Antero-
posterior
Transverse
Antero-
posterior
Transverse
26
18
24
18
23
34
39
42
13
IS
17
23
11
20
26
32
34
41
40
12
15
16
24
15
21
18
21
30
37
36
13
16
pl
9
P2
19
PS .. ---
22
p<_
26
Ml
M^ -
29
38
M3 . - __..__
40
D
9
V
12
13
15
In T. validum the opposite superior incisor series
converge slightly. I' has an angulate antero-exter-
nal face and sharply defined postero-internal cingulum;
i^ is a larger tooth, with a broad and sharply defined
cingulum; while in i' we have a subcaniniform crown of
very large size, with compressed anterior and posterior
edges and somewhat less prominent internal cingulum,
sloping downward and backward. The inferior in-
cisors are not Icnown.
The superior canine (55 mm.) is a powerful lance-
shaped tooth, larger but of the same form as in T.
cultridens, with a very convex antero-external and
more plane postero-internal face, bounded by the
sharply defined anterior and posterior ridges.
The superior grinding series, including the spaced
p', extends 195 millimeters anteroposteriorly, as com-
pared with 180 in T. cultridens (both males). In
general, the teeth are similar to those of T. cultridens,
but besides the larger size we note the following pro-
gressive features: (1) On p' the internal cingulum is
well defined, with the rudimentary posterior cusp
larger; (2) p^ is slightly broader and shorter (ap. 24
mm., tr. 20); (3) there are protoconules on p^"*. The
protoconules on the true molars are the only variable
or reversional characters.
P' is much larger than in T. cultridens. It is
separated by narrow intervals both from the. canine
(9 mm.) and from the second premolar. In p^ be-
sides the greater breadth we note the somewhat more
anterior position of the deuterocone and the more sub-
equal convexities of the protocone and tritocone on the
ectoloph, although the base of the protocone is still
much broader than that of the tritocone; p^ has a
rudimentary protoconule but no suggestion of a
tetartocone. In p' we have a more quadrangular
crown with a more elevated ectoloph than in T. cul-
tridens, and a more symmetrical development of the
deuterocone and tritocone, although the former is
still widely expanded at the base. In both p^ and
p* the internal cingulum is slightly less complete than
in T. cultridens. On p** there is a very striking ele-
vation of the ectoloph accompanied by greater promi-
nence of the antero-external style and greater sym-
metry of the deuterocone and tritocone convexities.
The molars represent a progression upon those of
T. cultridens, with sharp prominent styles, serrate
external cingula, elevated anterior cingula, pointed
protocones and hypocones, which are somewhat
flattened and vertically striated on their inner faces;
m^ has a prominent and serrate posterior cingulum
but no trace of a hypocone.
Telmatherium ultimum Osborn
Plates XVI, XLVI, LI, LV, LXV; text figures 126, 219, 223,
255, 256, 289, 294-298, 300, 508, 516, 592, 593, 647, 717, 745
[For original description and type references, see pp. 177, 184. For slteletal
characters see p. 653]
Type locality and geologic horizon. — White River,
Uinta Basin, Utah; Uinta formation {Diplacodon-
Protitanotherium-Epihippus zone, Uinta C 1).
Specific characters. — Skull very large (basal length
510 mm.), zygomata arching (zygomatic breadth 300
mm., estimated). Incisors and canines large, i' very
large, subcaniniform. P'-m^ 218 millimeters (type);
premolars progressive; p^ p* with the two outer cusps
subequal and externally convex; well-developed
internal and nearly complete external cingula; very
rudimentary tetartocone swellings; molars large
(m'-m^, 129 mm.) and progressive.
The discovery in Uinta C, by 0. A. Peterson, of a
female skull and jaw (Am. Mus. 2060) and the ante-
rior half of a male skull with well-preserved dentition
(Am. Mus. 2004) representing this species was a
most important one. The animal was at first sup-
posed by Osborn to be a terminal member of the
Palaeosyops series. Subsequently it was compared
point by point with Marsh's type of Telmatherium
validum and was found to exhibit the most striking
resemblances in the dentition and those parts of the
skull in which comparison could be made in both.
Highly distinctive is the premaxillary symphysis,
more deep and elongate than in Metarhinus, deeper
than in the type and paratype of Manteoceras
manteoceras, more abbreviate than in Dolichorhinus.
The region of the malars below the orbits is also
characteristic and similar to that of T. cultridens but
shows a more decided depression for the anterior
346
TITAJSrOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
portion of the masseter, and the posterior end of
the malar has a deep vertical flange. In dentition,
as enumerated above, T. ultimum is directly progressive
from T. validum.
Figure 294. — Type skull and lower jaw of Telmatherium ultimum
One-fifth natural size. Am. Mus. 2060, from Uinta C (true Uinta); White River, Uinta Basin, Utah
Measurements of Telmatherium ultimum, in millimeters
SkuU, basal length
Skull, breadth across zygomata.
Face, length
Cranium proper, length
Free nasals
Dental series, total length i'-m^.
Pi-m3
Pi-p^
Mi-m3
P, ap. by tr
P, ap. by tr
P, ap. by tr
C, ap. by tr
C, vertical
P', ap. by tr
P^, ap. by tr
P^, ap. by tr
PS ap. by tr
Ml, ap. by tr
M\ ap. by tr
MS ap. by tr
Am. Mus.
2060
(type);
Uinta C
510
"300
270
"250
85
303
218
90
129
14X14
15X15
19X18
19X12
20X19
21X28
26X35
38X40
46X49
46X51
Am. Mus
2004
(paratype);
Uinta C
320
229
95
137
15X14
15X17
22X20
25X22
44
19X12
20X22
23X28
27X35
40X37
49X44
50X48
568
In general comparison with Manteoceras the skull
of this species of Telmatherium presents very pro-
nounced differences: (1) the occiput differs widely in
its height and rounded summit and in the presence
of two large facets above the foramen magnum;
(2) in front of this, on the vertex of the skull, is a
relatively long, delicate sagittal crest without the
characteristic pit of Manteoceras and lacking the over-
hanging supratemporal ridges; (3) the nasals are
laterally recurved and distally truncate, as in Manteo-
ceras, but the free portion is relatively shorter; (4)
the zygoma resembles that of the Palaeosyops or
LimnoJiyops type — that is, it is without
the infraorbital shelf — and especially
parallels that of Palaeosyops in the de-
velopment of a deep flange on the lower
surface of the malars, which is an ad-
vance on the M. manteoceras condition.
The above table shows rather marked
differences in proportions of the teeth
between the type and paratype; the
cheek teeth in the paratype are all
relatively longer and narrower.
As a whole the skull is mesaticephalic.
Comparison of the outline dorsal and
palatal views of Telmatherium ultimum
and Manteoceras (figs. 296, 303) brings
out a large number of very distinctive
characters.
The horn rudiments are so inconspicu-
ous in both the male and female skulls
that they were not observed by the
author for a long time. In the female
they may be said hardly to exist, and in the male (PI.
XVI) they can be seen only by very close scrutiny.
As above noted, it is difficult to say whether they
are in a retrogressive or stationary condition. They
are certainly far less progressive than in Manteoceras.
Slcull of T. ultimum. — The cranium of this species is
represented by the type, a superbly preserved female
skull (Am. Mus. 2060), and by the anterior portion of
the paratype, a male skull (Am. Mus. 2004), in which
the youthful age is such that many of the sutures can
be made out. The skuU of the type is laterally crushed
in the anterior half, but the width across the zygomata
has probably not been greatly lessened. The general
proportions are mesaticephalic, the cranium being very
much longer than that of P. leidyi but much less
elongate and deeper vertically than that of D. hyogna-
thus. The measurements are, length 510 millimeters,
breadth 300 (estimated). It is readUy distinguished
from all other crania by the combination of the follow-
ing principal characters: (1) Nasals relatively short
(free length 78 mm.), the lateral downward extensions
being wholly covered by the maxUlaries; (2) prominent
narrow sagittal crest; (3) greatly elevated occiput;
(4) deeply extended malar and squamosal flanges of
the zygomatic arch; (5) premaxillary symphysis ver-
tically extended; (6) frontals with horn swellings rudi-
mentary— that is, consisting of convexities so slight
(paratype) that they are with difficulty observable.
In palatal aspect the brachycranial proportions
decidedly predominate over the dolichocranial in the
basicranial region of the skull as shown in the following
characters: (1) The shortness of the anteroposterior
measurements (as from glenoid facet to mastoid proc-
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
347
ess, from foramen ovale to condyle 100 mm.), as
compared with the transverse measurements (across
zygomata, 300 mm., estimated; across postglenoid
processes, 195; across mastoid processes, 180); (2) the
upward slant of the basisphenoid; (3) the shortness of
the distance (140 mm.) between the pterygo-alisphe-
noid wing and the condyle; (4) the postglenoid and
paroccipital processes greatly flattened or extended
transversely with very moderate anteroposterior diam-
eter.
More in detail: The posterior nares open immedi-
ately between m^ and m^, whereas in P. leidyi they
as compared with 38 in D. Tiyognathus. Between the
foramen lacerum medium and foramen lacerum
posterius the basioccipital forms a prominent, later-
ally compressed keel. The occipital condyles are
comparatively slender and widely separate below the
foramen magnum. This aspect of the skull illus-
trates admirably (1) the broadly transverse extension
of the articular facets for the condyle of the jaw, (2)
the broadening of the postglenoid processes, (3) the
separation of the postglenoid and post-tympanic,
which is much wider than in the brachycephalic P.
major but very much narrower than in the dolichoct -
•p.ty.sq.
^'pgl.sj.
Figure 295. — Type skull of Telamatherium uliimum
One-fourth natural size. Am. Mus. 2060. White River, Uinta Basin, Utah; base of Uinta C, true Uinta formation. Ai, Side view. The depth of
the sltull in the middle region and immediately in front of the orbit has been increased by lateral crushing. The double lines mark the plane of
the sections in Figiu-e 255, Bi, Bj. Aj, Front view. As, Occipital view.
open opposite the posterior half of m^, and in Oligocene
titanotheres they often open opposite the posterior half
of m'. The postnarial space is relatively deep, or ver-
tically extended, and short anteroposteriorly ; the
line of junction between the pterygoid wings of the
alisphenoids and the palatines can not be clearly made
out. Unlike those of M. megarJiinus or P. major the
pterygoids and lateral wings of the alisphenoids descend
abruptly. The foramen ovale is separated from the
foramen lacerum medium by a bridge, 24 millimeters
phalic D. hyognathus; (4) also the sharply produced
downward flange of the posterior portion of the malar.
The superior aspect of the skull (fig. 296) fails to
give the actual shape of the nasals owing to the
marked crushing at this point. The entire length of
the nasals is 219 millimeters, as compared with 520,
the entire length of the vertex.
Horn rudiments. — In the type female skull there is
no evidence of the existence of a horn swelling at the
junction of the frontals and nasals. In the paratype
348
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
male skull (Am. Mus. 2004), however, there is a low
swelling (PI. XVI) at the junction of the nasals,
maxillaries, and frontals, at a point above and some-
what in front of the anterior rim of the orbit, which
betokens the presence of a horn rudiment in an even
more incipient stage than that of M. manteoceras or
D. hyognathus .
The uncrushed skull was evidently rather broad
between the orbits. The supratemporal crests are
moderately defined anteriorly, but as they enter the
ing over of the downward lateral extension of the
nasals; (3) the wide interval (109 mm.) between the
antorbital border and the narial notch; (4) the anterior
extension of the malars below the orbit; (5) the clear
definition of the lacrimals, partly external to and
partly within the orbit; (6) the prominence of the
postorbital processes of the frontals and malars re-
spectively; (7) the gently rounded conformation of the
malar below the orbit, which most nearly resembles
that in T. cultridens; (8) the sharp downward or in-
ap/iij^
Figure 296. — Type skull of Telamatherium ullimum
One-fourth natural size. Am. Mus. 2060. White River, Uinta Basin, Utah; base of Uinta C, true Uinta formation. Ai, Palatal view; A2, top view.
Lateral crushing has narrowed the frontal region and distorted the zygomata.
parietals they become more sharply defined, leaving a
shallow groove between the summits of the short
sagittal crest. The superior border of the lateral
occipital crest is rather delicate ; in fact, the entire skull
is slender rather than broad and massive.
The lateral aspect of the skull is distinguished by
the following characters: (1) The relatively short (85
mm.) free portion of the nasals; (2) the elevation of
the maxillaries on the sides of the face, somewhat as
in M. megarJiinus and D. hyognathus, and the cover-
ferior flange of the malars beneath their junction with
the squamosals; (9) the moderate upward extension of
the zygomatic squamosal bar; (10) the presence of a
cranial depression at the point of junction between the
parietals and the frontals, or above the mid-cranial
region, the skull being gently arched upward in front
of this point.
The anterior aspect of the skull (fig. 295) exhibits the
relatively deep premaxdlary symphysis as compared
with that of P. leidyi, and the absence of the extremely
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
349
long and deep maxillary union so characteristic of
D. hyognathus. The nasals are much less thickened
and decurved at the sides than in D. hyognathus. This
aspect of the skull also exhibits the depth of the
zygoma, including the malar and squamosal portion,
compared with the extreme shallowness of this arch
in D. hyognathus.
The occipital view of the skull is still more charac-
teristic, owing to its great height (194 mm.) as com-
pared with its breadth (137 mm.), also to the pres-
ence of a pair of oval prominences on either side of
the superior portion of the foramen magnum, as in
Menodus giganteus. This view also illustrates the
breadth of the paroccipital and postglenoid processes.
The dentition is finely represented in the complete
type skull (Am. Mus. 2060) and in a somewhat more
progressive stage in the paratype, consisting of the
anterior portion of a skull (Am. Mus. 2004).
The incisors are superbly shown in Am. Mus. 2004
and 2060 (PL LV; figs. 294-297). The superior in-
cisors in the type are pointed, decidedly cingulate or
cupped posteriorly, and sharply convex anteriorly, and
increase in size rapidly from i' to i'; the lateral incisor
(i^) is more caniniform than incisiform; the crown of
i' measures 15 millimeters vertically, and a slight ridge
extends down the posterior face to the apex of the
basal cingulum which branches on either side to form
lateral depressions; the crown of i^ (measuring 19 mm.
vertically) is slightly larger and of exactly similar
form; in i^ the caniniform crown (measuring 28 mm.
vertically) is distinguished on its postero-internal sur-
face by a lanceolate face with sharply defined antero-
internal and postero-external ridges, which sweep at the
base into the low, broad cingulum, exactly as in the
canine. In Am. Mus. 2004 the superior incisors are
even larger and the posterior cingulum is more strongly
accented. All the cutting teeth, both incisors and
canines, bear a striking similarity to those of T. cultri-
dens and T. validum — in fact, they are almost directly
progressive upon them, the only difference being that
the posterior angles and cingula are a little less sharply
accented. The inferior incisors are not known.
The superior tusks, which are completely preserved
only in the paratype (Am. Mus. 2004), are much more
decidedly of the lanceolate, typical Telmatherium
type than those of M. manteoceras, although the
anterior and posterior ridges are not quite so promi-
nent and sharply defined as in T. cultridens, T. validum,
or D. hyognathus; the tusks are none the less long
(43 mm.) and transversely narrower (22 mm.) at the
base of the crown than those of M. manteoceras
(25 mm.); the ridges pass inferiorly into a strong
postero-internal cingulum, which also clearly dis-
tinguishes these tusks from those of other species
so far as observed. The superior molar-premolar
series as a whole is not only larger (229 mm. in Am.
Mus. 2004) but has a very marked individuality
throughout, so that every tooth in the series can be
distinguished by careful observation from those of
either M. manteoceras or D. hyognathus. The type is
distinctly telmatherioid, exhibiting peculiar pro-
gressive modifications upon the dental type of T.
cultridens and T. validum which partly anticipate
those seen in Menodus. The series in the type
(Am. Mus. 2060) is of somewhat smaller size and in an
earlier or less progressive stage than that in Am.
Mus. 2004. In both specimens the breadth of the
molars equals or slightly exceeds the length, whereas
in the more dolichocephalic D. hyognathus molars the
length decidedly exceeds the breadth. The dental
proportions are therefore mesaticephalic.
Premolars. — The superior premolars of the type are
so much worn as to obliterate certain of their rudi-
mentary progressive characters. The following de-
scription of these parts is accordingly based upon the
unworn premolars of the paratype specimen, which
appears to be in a somewhat more progressive stage of
evolution. A narrow diastema, 12 to 16 millimeters.
Figure 297. — Paratype skull of Telmatherium uUimum
One-fourth natural size. Am. Mus. 2004, reversed; White River. Uinta Ba.sin,
Utah; Uinta C, true Uinta formation.
separates the canine from p^ The premolar series,
measuring 90 (type) and 95 (paratype) miUimeters,
as compared with 82 in M. manteoceras, exhibits not
only increase in size but marked progression in pattern,
as seen in the following characters: (1) Slightly in-
creased compHcation of p^ (ap. 19 mm., tr. 12) in the
rudimentary internal cingulum and more decided
elevation of the tritocene; (2) in p^ to p* of the para-
type (No. 2004) the deuterocones consist of antero-
posteriorly elongate ridges, much more pronounced
than in T. cultridens, convex on the lingual and
flattened on the buccal surfaces, totally different
from the smooth-sided deuterocones of M. manteoceras
and from the apically compressed cones with faint
lateral ridges in D. hyognathus; this feature, it should
be added, is much more distinctly exhibited in the
Httle worn series of Am. Mus. 2004 than in the much
worn series of the type, Am. Mus. 2060; (3) this deu-
terocone ridge is destined to give rise to the tetarto-
cones by constriction, and in p^ p** a faint rudiment of
the postero-internal cingulum can be observed in the
unworn crown; (4) the internal cingulum is faintly
defined around the entire lingual surface of the deutero-
350
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
cones; (5) the ectolophs are greatly elevated and con-
sist of the two well-defined subequal protocone and
tritocone convexities with a rudimentary external
cingulum and pronounced antero-external style; (6) in
p^ the progressive broadening tendency is illustrated
by the fact that the breadth (22 mm.) equals the length
on the ectoloph, a marlied advance upon what is
observed in either M. manteoceras or D. hyognathus — in
fact, this tooth now resembles p' in general pattern,
although retaining a more elongate contour; (7) in
p' the breadth considerably exceeds the length and
the crown is broadened internally by the expansion
of the deuterocone; (8) in p* we have a still more
quadrangular and molariform tooth, the length being
27 millimeters and the breadth 35, but in this tooth
the deuterocone is not quite so sharply defined.
The superior premolars of the type (Am. Mus.
2060) are distinguished from the premolars of the
paratype (Am. Mus. 2004) by the following
characters: (1) The premolar series is somewhat
shorter (90 mm. as compared with 95 in Am. Mus.
2004); (2) p^ in the type is less advanced, in
that the deuterocone is smaller and placed
farther back and the tritocone is less subequal
with the protocone; (3) in p' also the deu-
FiGURE 298. — Lower jaw of Telmatherium ulLimum
One-fourth natural size. Am. Mus. 2060 (type). White River, Uinta Basin, Utah; base ot Uinta C, true
Uinta formation.
terocone and tritocone are somewhat less progres-
sive and the tetartocone ridge is barely suggested;
the cingulum also is slightly less developed; (4) in p^
the tetartocone is indicated by a low, obtuse swelling,
whereas in Am. Mus. 2004 it forms the distal spur of
a very prominent deuterocone ridge. These differ-
ences in the tetartocones can hardly be due entirely
to differences in degree of wear (the type being much
the older of the two), because in the paratype the ridges
in question are so strong that they would probably
show even in the worn stage. These differences seem
to indicate that the paratype is somewhat more ad-
vanced than the type in its premolar evolution.
The inferior premolars, as observed in the lower jaw
of the type (Am. Mus. 2060), exhibit the following
characters: The postcanine diastema is about 20 milli-
meters in length; behind pi is a shorter diastema of
11 millimeters in length; pi and p2 are represented
only by the alveoli; pa is much damaged but was
incompletely molariform.
P4 (ap. 27 mm., tr. 19) is submolariform, lacking
only the prominence of the postero-internal cusp,
which is analogous to the entoconid of the molars.
The trigonid is higher than the talonid, and its V less
sharply defined. A weak external cingulum appears
opposite the outer midvalley and festoons the external
slope of the hypoconid.
Molars. — The superior molars are a powerful series
of teeth measuring 129 (type) to 137 (paratype) milli-
meters, with extremely elevated or subhypselodont
cusps, the ectoloph of the least worn m^ reaching a
height of 35 millimeters and the protocone of the same
tooth 24. The external cingulum is more pronounced,
especially in the type, Am. Mus. 2060, in which it
prominently guards the outer valleys and begins to
encircle the styles, reminding us of the cingulum
development in Menodus giganteus; the internal
cingulum is similarly prominent, embracing the entire
inner side of the crown in m^ of the
same specimen. A marked peculiarity
which is an advance on both M.
manteoceras and T. cultridens is the
prominence of the anterior cingulum in
m' to m', which swells into a large median
cingule, comparable to the protostyle of
most species of Menodus. The posterior
cingulum is less prominent except in m',
in which it is free and exceptionally high
(type); in the paratype it is connected
with an incipient hypocone swelling.
The elevated ectoloph is accented by the
sharp development of the parastyles,
mesostyles, and metastyles. The hypo-
cones of m^, m^ of both type and para-
type are very large and prominent, an
advance upon the conditions in T.
cultridens and T. validum. A fur.
ther peculiarity is that in the unworn paratype the
buccal surfaces of the internal cones (protocone and
hypocone) — that is, the surfaces facing the ectoloph — ■
are somewhat flattened and vertically striate, instead
of rounded and smooth, as in M. manteoceras and D.
hyognathus, which is faintly prophesied in T. validum.
The only retrogressive elements are the conules,
which have disappeared. The conules are largest in
brachyodont titanotheres; with advancing hypso-
donty the base of the paracones and metacones ex-
tends linguad and either absorbs or crowds out the
conules.
The inferior molar series represented in the jaw
associated with the type (Am. Mus. 2060) is of large
size (155 mm.). A smaller jaw (Am. Mus. 2033),
formerly referred to this species but now referred
provisionally to Manteoceras uintensis, is somewhat
shorter (147 mm.). Although this specimen probably
represents another genus and species, the molar
characters are somewhat similar to those of the type ,
EVOLUTION OP THE SKULL AND TEETH OF EOCENE TITANOTHERES
351
including as progressive features (1) the prominence
and the backward extension of the paraconid into a
parastylid; (2) the variable but distinct metastylid
ridge; (3) the external position of the hypoconulid
on 1X1} and the prominent internal ridge on it, which
gives it a concave form internally; (4) the external
cingulum slightly more progressive than in M.
manteoceras and dipping somewhat into the valleys
but not so deeply as in D. hyognathus.
Lower jaw of T. ultimum. — The jaw of this specimen
is represented by that of the type (Am. Mus. 2060).
The type jaw retains the characters of T. cultridens
in the rather slender recurved coronoid process but
departs from them by its rapidly increasing depth
posteriorly — in fact, the whole j aw is relatively deeper
than in the ancestral species. The distance from
the condyle to the incisive border is estimated at
435 millimeters in the type. The chin is strongly
compressed laterally (54 mm.), and behind it the jaw
gradually broadens and deepens, the lower border
being more nearly straight than in M. manteoceras
and terminating in the slightly depressed and back-
wardly produced angle; the condyle exhibits two
marked peculiarities: the outer half of the rotular
facet extends broadly forward, whereas the inner half
has a straight anterior border and unites posteriorly
by a much broader union than in M. manteoceras
with the broad facet for the postglenoid process.
The coronoid process, perfect in the type, is rather
narrow and uniformly recurved. The striking resem-
blance to T. cultridens observed in the dentition of
this species is therefore not seen in the jaws, which are
relatively shorter, more massive, and deeper posteriorly
(below ms) than in T. cultridens, all of which are pro-
gressive characters.
A second jaw (Am. Mus. 2033) was at first doubt-
fully referred to the same species. In this jaw the
second premolar is spaced as in the type. In other
features, as in ps, in the coronoid process, and in the
proportions of ma, this jaw resembles those of members
of the Manteoceras phylum, to which this one is now
provisionally referred. (See Manteoceras uintensis,
below.)
A skull in the Carnegie Museum (No. 2339) differs
from the type and paratype in having a longer tooth
row but shows generic agreemeiit with T. ultimum
in the general form of the skull, especially of the
zygomata, occiput, and nasals.
Telmatherium altidens Osborn
Plate LXV; text figures 127, 299, 300
[For original description and type references see p. 184]
Type locality and geologic Tiorizon. — Uinta Basin,
Utah; Uinta formation, Diplacodon-ProtitanotJierium-
EpiMppus zone (Uinta C).
Specific characters. — Pj-ms, 313 millimeters; a wide
diastema (55 mm., estimated) behind the inferior
canines; canines in males elevated (76 mm., estimated)
and pointed; pi_2 laterally compressed, nonmolari-
form; p3_4 submolariform. Subdolichocephalic, upper
postcanine diastema elongate.
Materials. — As described in Chapter III, this animal
is known only from a single lower jaw (Am. Mus. 2025)
with no parts of the skull or skeleton associated.
Comparison. — The reference of this specimen to the
genus Telmatherium depends chiefly upon (1) the large
size and vertical elongation of the canines, as in
T. validum; (2) the very large size of the lower incisors;
(3) the exceptional elongation of the lower postcanine
diastema, which is incipient in T. ultimum but was
evidently carried to a much greater extreme in T. alti-
dens; (4) p2 much less molariform than in Protitano-
therium and hence more like the simple, laterally com-
pressed p2 of T. ultimum; (5) the very large size of the
lower molars (mi_3), the form of which indicates large,
broad upper molars, as in T. ultimum.
The skull when discovered may well prove that this
animal represents a well-marked new generic stage.
On the other hand, the very large incisors and lofty
canines, the pronounced diastema, the characters of
Pi, p2, and the large, broad molars, as noted above,
appear to indicate generic Idnship with T. ultimum.
Chief characters. — The exceptionally long mandibu-
lar symphysis and wide postcanine diastema, as fore-
shadowed in T. ultimum, distinguish this titanothere
as possessing a relatively elongated facial region.
This character, as well as the long, relatively shallow
jaws, the elongation of ms, and the wide space behind
ma, is evidence that the skull as a whole was sub-
dolichocephalic, although far less so than that of
Dolichorhinus. In common with T. ultimum, "T. inci-
sivum," Protitanotherium, and all other upper Eocene
and Oligocene forms, T. altidens had undergone a dif-
ferential elongation of the middle part of the skull,
which allowed the molars to become extremely large,
both absolutely and as compared with the premolars.
The elevated, piercing canines are also exceptional
among titanotheres; they exceed those of the ances-
tral species. Thus the animal is very readily distin-
guished from any of the known species of the contem-
porary Diplacodon and Protitanotherium. A fourth
feature is the simple, nonprogressive, elevated, and
somewhat laterally compressed form of p2, which is
decidedly more primitive than the corresponding tooth
in Protitanotherium.
The cracked and much weathered teeth of this male
individual (Am. Mus. 2025) fortunately include the
median incisors (ij) of the opposite sides, the left
canine, and the entire grinding series of the right side
in sufficient preservation to define the species sharply.
In detail the median incisors are much larger, with
more pointed tips than those of Protitanotherium
emarginatum, measuring 19 millimeters on the anterior
face, 20 anteroposteriorly, and 15 transversely. These
352
TIT.USrOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
teeth distinctly suggest the upper median incisors of
T. ultimum. The laterally compressed or convex
anterior faces, the smoothly sloping posterior faces,
the U-shaped posterior cingula also suggest the Pal-
aeosyops type of tooth, although this dolichocephalic
animal does not appear to present any affinity to that
genus. It is difficult to determine the precise form
and proportions of the canines, the fang measure-
ments (vert. 76 mm., estimated; ap. 31; tr. 26)
indicating a more laterally compressed or dolicho-
cephalic type of canine than in Protitanotherium emar-
ginatum. The height of this tall and slender canine
exceeds 76 milHmeters (estimated), as compared with
53 in the male P. emarginatum and 56 (estimated) in
the male P. superhum. The name T. altidens refers
to this feature, as the tusk is the most elevated and
Figure 299. — Hypothetical reconstruction of the skull of Telmatherium altidens
One-sixth natural size. Designed to show especially the long postcanine diastema,
generic resemblance to T. ultimum. The lower jaw is Am. Mus. 2025 (type of T.
Am. Mus. 2060 (type of T. ultimum).
piercing among all the known titanotheres, not except- '
ing the giant Menodus giganteus of the Ohgocene.
Faint anterior and posterior ridges can be detected
on the anterior and posterior faces of the crown,
distinguishing this tooth readily from the canine of
Palaeosrjops major, in which the posterior ridge is on
the internal face of the crown. The very wide total
diastema between the canine and p2 measures 70 milli-
meters, as compared with 40 in Protitanotherium
superium and 45 in Telmatherium ultimum, which is
approached only by the wide diastema (51 mm.) in
Dolichorhinus hyognathus. The grinding series as a
whole measures 313 millimeters, as compared with 304
in P. emarginatum and 318 in P. superium, the lower
grinding series being, therefore, slightly smaller than
in P. superhum.
Premolars. — Pi and p2 are not so much compressed as
in Telmatherium cultridens but are somewhat swollen
transversely. Pi (ap. 19 mm., tr. 12) is a simple,
gently compressed cone, with a small posterobasal
cusp rising from the posterior ridge. This cusp is less
advanced than in Protitanotherium superhum or P.
emarginatum. P2 (ap. 27 mm., tr. 15) is also less
advanced than in those species, its posterior lobe being
smaller, lower, and much less crescentic superiorly. It
has a faint paraconid, no metaconid, and very faint
anterior and posterior internal valleys. It is thus
much like P2 of Manteoceras manteoceras, T. cultridens,
and (so far as known) T. ultimum.
In striking contrast with this is the progressive
structure of pa (ap. 30 mm., tr. 18), especially its ex-
tremely prominent median cusp ( = protoconid); the
anterior and posterior crescents are correspondingly
more defined than in p^; the rudiments of the metastylid
and cusps appear, corresponding to the paraconid and
..-;: . entoconid in the molars. P4
is a decidedly larger tooth
(ap. 35 mm., tr. 23) with
prominent internal cusps ( =
paraconid, metaconid, meta-
stylid, entoconid).
Molars. — The true molars
measure 195 millimeters in
length, as compared with
214 in Protitanotherium su-
perhum. The measurements
(ap. by tr.) are, nij, 45 by 29
millimeters; m2, 59 by 32;
ma, 89 by 35. The very
large size of the molars as in
P. superhum and other Uinta
C titanotheres is thus note-
worthy. The much worn
grinders give an imperfect
picture of the distinctive
characters of these teeth, but
the elongated face, and the supposed
altidens). The skull is restored from it would appear that the exter-
nal cingulum and the meta-
stylid are faintly indicated and that in ma the hypoconu-
lid is placed more on the internal or lingual side of the
crown, as in Palaeosyops paludosus. The grinding
series, therefore, presents two resemblances to that
of P. paludosus — namely, the prominent internal cusp
on Pa and the more internal position of the hypoconulid
on ma — yet neither of these characters is believed to
indicate genetic affinity. The closest resemblances are
to the inferior dentition of T. ultimum, from which
this jaw differs, however, in its greater size, its rela-
tively larger canines, and the more internal position
of the hypoconulid. This last condition may be partly
due to crushing.
Jaw. — The jaw of T. altidens is readily distin-
guished from the jaw of Protitanotherium emarginatum
and that of P. superhum by its dolichopic characters,
the wide diastema between the canine (55 mm., esti-
mated) and Pi, and that between ma (54 mm.) and the
anterior border of the coronoid process. The sym-
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
353
physisis extremely long (200 mm.); it is both, actually
and relatively longer than in P. emarginatum (155) or
P. superhum (158, estimated). The jaws are de-
cidedly deep, measuring 107 millimeters below p2, 124
behind m2. The coronoid was probably elevated, ta-
pering, and recurved toward the summit. The thick-
ness of the rami in the type jaw has been reduced by
crushing; below mi it is 40 millimeters. This jaw
therefore represents a large but fairly slender and
active animal, which in some respects is suggestive of
relationship with species of the long-jawed genus Meno-
ently short, broad proximally, and tapering distally;
face concave in front of orbits; frontonasal "horn
swellings" not evident; sagittal crest deep and nar-
row; occiput low with thin crests; dentition extremely
macrodont; incisors relatively larger than in any other
known titanothere.
Historical notes.^ln describing the type species
{Sthenodectes incisivus) of this genus the author, Earl
Douglass (1909.1, p. 305), said: "I think that this skull
represents a different genus from Telmatherium, but I
prefer to place it provisionally here rather than estab-
FiGURE 300. — Lower jaws of Telmatherium ullimum and T. altidens
One-fourth natural size. A, T. ultimum, Am. Mus. 2060 (type), reversed; White River, Uinta Basin, Utah; base of Uinta C, true Uinta
formation. B, T. altidens, Am. Mus. 2025 (type); White River, Utah; Uinta C.
dus of the Oligocene but in other respects is very unlike
an ancestor of Menodus — namely, the excessively large
size of the incisors, the retarded condition of pi and
P2, and the very long postcanine diastema.
Sthenodectes Gregory
Plates LXV, LXVI; text figures 129, 130, 301
[For original description and type references see p. 190]
Generic cTiaracters. — Skull mesaticephalic or sub-
brachycephalic; cephalic index 62-65; malars with-
out infraorbital shelf or protuberance; nasals appar-
lish another genus." Through the courtesy of Doug-
lass, Gregory (1912.1) was enabled to compare this
type with the extensive material in the American and
Yale Museums and reached the conclusion that T.
incisivum represents a different genus or subgenus, to
which he gave the name Sthenodectes, in allusion to
the great power and development of the incisors and
canines. The following characters were assigned by
Gregory in the original description of the genus as
compared with Telmatherium: (1) The incisors are much
larger and more advanced in evolution; (2) the post-
354
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
canine diastema is reduced or absent; (3) the superior
premolars p^ p\ p* are more progressive than in T.
ultimum, having very heavy internal cingula and pro-
nounced external cingula; (4) the basicranial region
differs in many details.
The type skull of the species (Carnegie Mus. 2398) is
vertically crushed, a condition that led to some errors
in the original description of the species which a second
skull in the Field Museum (No. 12168) enabled Greg-
ory to correct and to reach the following conclusion
as to the affinities of this animal:
Relation to Telmafherium. — StTienodedes is sharply
separated from the Dolichorhininae and at the same
time allied with Telmatherium by the following char-
acters: (1) General contour of the skull in basal view,
wholly unlike MetarMnus and resembling Manteo-
ceras or Telmatherium; (2) complete absence of infra-
orbital protuberance, the infraorbital portion of the
malar more like that of either Manteoceras or Telma-
tJierium; (3) midportion of malar with deep vertical
flange as in Telmatherium (contrast Metarhinus) ;
(4) incisors and canines readily derivable from the
Telmatherium type (compare figures of side view,
crown view; compare premaxillaries); (5) dentition
extremely macrodont (microdont in Metarhinus,
TQ.B.CYoAontm Telmatherium); (6) premolars more ad-
vanced than in T. ultimum but derivable from the
Telmatherium type (cf . T. validum) by enlargement of
internal cingulum, filling out the internal contour of
p^; (7) referred lower jaw (Field Mus. 12168) de-
cidedly nearer to Telmatherium ultimum than to
Metarhinus, macrodont, especially molars, ramus mas-
sive and deep; (8) basis cranii with postglenoid, post-
tympanic, meatus, and basioccipital nearer the
subbrachy-mesaticephalic type of Telmatherium than
to the subdolichocephalic type of Metarhinus.
Through parallel evolution there are some marked
resemblances to the Dolichorhininae, as follows: (1)
Premolars (p^~*) with very heavy internal cingula and
crowns well filled out on the inner side; (2) incisors
cupped by upgrowths of heavy cingulum; (3) subhyp-
sodonli or elongate character of the molars of the
type specimen.
Effects of crushing. — To the vertical crushing of the
type skull is possibly due the wide displacement of
the lacrimal bones on both sides of the face, result-
ing in the false appearance of "lacrimal pits." To the
crushing is also due the union of the postglenoid and
post-tympanic processes, the depression of the occiput,
and the abbreviation of the nasals.
Sthenodectes incisivus (Douglass)
Plates LXV, LXVI; text figures 129, 130, 301
[For original description and type references see p. 185)
Type locality and geologic horizon. — About 3 miles
northeast of well 2, Uinta Basin, Utah; upper levels of
Eohasileus-Dolichorhinus zone (Uinta B 2).
Specific characters. — Skull, length 488 millimeters,
zygomatic breadth 305 millimeters, cephalic index 62.
Dentition, p^-m^ 207 millimeters; m^-m^ 125; p^
large (ap. 19 mm., tr. 22), very progressive, with
advanced tritocone and complete internal cingulum,
deuterocones of p^~* relatively elevated, internal
cingula heavy, complete, m' (ap. by tr.) 42 by 45
millimeters with reduced posterior metacone crescent.
Materials. — Besides the type skull in the Carnegie
Museum (No. 2398), on which the above specific
characters are based, there is a well-preserved skull in
the Field Museum (No. 12168), also a pair of lower
jaws (Field Mus. 12166). According to Riggs (1912.1,
p. 38) all three specimens are from the same locality
and belong to the same species. The skull and lower
jaws in the Field Museum were discovered by Abbott
in the lenticular sandstones near well 2 at the foot of
Coyote Basin. The skull, Field Mus. 12168 (PL
LXVI), is shorter (460-300 mm., cephalic index
65) than in the type of S. incisivus but has the
broad-crowned molars and massive incisors of that
species. At approximately the same geologic level but
half a mile distant was found the lower jaw (Field Mus.
12166) referred to this species (PL LXVI), which
belongs to an older individual, as is evident from the
worn molars and incisors. It is described below.
This short-headed, massive-jawed titano there ex-
hibits a remarkable combination of characters. It
exceeds all other known titanotheres in the size of the
incisor teeth, which are correlated with the massive
jaws and the relative abbreviation of the skull, the
general proportions of which suggest those of Manteo-
ceras. The abbreviation of the facial region consti-
tutes a differentiation directly the opposite of that
which was occurring in the line which gave rise to
Telmatherium altidens in Uinta C, in which the face,
judging by the wide postcanine diastema, was elon-
gated. The indices are significant.
Indices of Sthenodectes incisivus
Cephalic index
Faoiocephalic index..
Molar-cephalic index
Carnegie
Mus. 2398
(type)
Field Mus.
12168
(referred)
65
44
28
The grinding teeth are also proportionally very
large. The skull is at once separable from that of
Mesatirhinus and Dolichorhinus by the short basi-
cranial region and the stout, wide, spreading zygo-
mata, as well as by the heavy, short premaxillae and
the absence of a rounded infraorbital protuberance, or
shoulder. It also differs from any of these genera in
the proportions of the molar teeth, m^ and m^ being
wider, or more brachycephalic in type. It parallels
the true" Dolichorhininae, however, in the advanced
EVOLUTION OP THE SKULL AND TEETH OF EOCENE TITANOTHERES
355
condition of the premolars and in the cupping of the
incisors. It resembles MetarTiinus, especially M.
earlei, in the following characters: (1) Broad forehead;
(2) concavity of the face in top view; (3) certain fea-
tures of the premolars; (4) proportions of the occiput;
(5) thin, high sagittal crest.
From the contemporary species of Manteoceras,
namely, M. uintensis, it is distinguished by (1) the
characters of the incisors and canines; (2) the much
more advanced condition of the premolars; (3) the less
elongate m^; (4) the feebly constricted postcanine
region. It parallels Manteoceras in the general pro-
portions of the skull and in the form of the zygomatic
arches, except that the malar portion of the arch has
the deep flange characteristic of Telmatlierium.
Sthenodectes suggests Tehnatherium ultimum in cer-
tain features of the incisors, canines, and molars, in
the detailed characters of the basicranial region and
in the spreading zygomata; but it is distinguished from
that form by (1) the much larger size and higher
development of the incisors, (2) the more advanced
condition of the premolars and premolar cingula, (3)
the different form of m', (4) the lower occiput and
sharper sagittal crest, (5) the wider forehead, (6) the
more angulate section of the infraorbital portion of the
malars, (7) the sharply tapering nasal bones. From
the European genus Brachydiastematherium, which it
resembles in having three large incisors, Sthenodectes
is distinguished by the markedly lower evolution stage
of the premolars (p2-p4).
Side and top views. — The top of the type skull has
been crushed downwai'd, especially above and in
front of the orbits. The premaxillaries, though some-
what flattened by pressure, are of very large size, in
correlation with the exceptional dimensions of the
incisors. The nasals are imperfectly preserved at
the end but appear to be even shorter than in T.
ultimum; they converge rapidly in front, about as in
MetarTiinus, and proximally they spread rapidly and
widely, measuring 125 millimeters transversely at the
outer junction with the frontals. The latter were
somewhat flattened but were very wide across the
orbits (tr. 192 mm.). In front of the orbits there is a
prominent vertical facial concavity suggesting the
conditions in MetarMnus fluviatilis. The infraorbital
foramen is large and prominent, apparently more so
than in T. ultimum. Above this foramen and in
front of the orbit is a triangular depression, in the
position of the lacrimal bone, occurring on both sides
of the skull but much larger in the right, which is
referred to by Douglass as a vacuity. It now seems
probable that these vacuities resulted from the down-
ward crushing which has squeezed the lacrimals out
of place; they lie immediately below the region where
the horn swelling usually appears, but the presence of
the latter is but vaguely if at all indicated. The fore-
head, as already stated, is broad and flat, and the depth
of the skull appears to be less than in T. ultimum
The opposite postorbital temporal crests run backward
into a long sagittal crest, which is quite high and thin.
The occipital crests are thin, but the whole occiput is
much lower than in T. ultimum.
Palatal view. — In the inferior aspect of the skull we
are struck by the great size of the dentition as a whole,
the great size and spatulate outline of the incisor re-
gion, the prominent pointed canine tusks, the long,
straight tooth row, the virtual lack of a postcanine
diastema, the wide, very progressive premolars, the
relatively large, subhypsodont molars, the widely
arching zygomata, and the short basicranial region —
all these, with the exception of the prominence of the
incisors and canines, being characteristic of Oligocene
titanotheres. The infraorbital part of the malar is
like that of Manteoceras in that it did not flare out-
ward into an infraorbital protuberance; just behind
the orbit the malar was very massive, and its broad
inferior expansion shows an area for the attachment
of the masseter; the postero-inferior portion of the
malar is a deep vertical flange, as in Manteoceras and
T. ultimum. The squamosal portion of the zygoma
is very stout and broad anteroposteriorly; the post-
glenoid process is rather small. The prominent ex-
ternal auditory meatus of the type in side view appears
to be closed below by the appression of the postglenoid
and post-tympanic processes, but this is probably due
to crushing, as the Field Museum specimen shows these
processes widely separated. The palate is long, and
the anterior border of the posterior nares is between
m- and m".
Incisors. — The anterior incisor (i^) is very large
(ap. 22 mm., tr. 20), and closely appressed in the
median line to its fellow of the opposite side; its large,
blunt tip lies near the median line; back of this is a
wide, oval basin, or pit, bounded by the very heavy
posterior cingulum and by the external ridge; the
front face is vertically deep (26 mm.). The median
incisor (i^) has a low median tip and wide posterior
basin. It remotely resembles that of DolicTiorhinus but
is far larger (ap. 25 mm., tr. 26) even than that of T.
ultimum. The very large canine (ap. 27 mm., tr. 27),
as already observed, is long and piercing, with a verti-
cal crown length of 57 millimeters, as compared with
42 in the paratype of T. ultimum. Its transverse
diameter is 27 millimeters, as compared with 22 in T.
ultimum. It has similar antero-internal and postero-
external edges but is distinguished by its heavier poste-
rior basal cingulum.
Premolars. — The premolars are larger and wider
than in T. ultimum. There is little if any postcanine
diastema, p' being crowded in behind the base of the
canine. Its crown is not preserved, but it appears
probable that this was broader — that is, more ad-
vanced— than in T. ultimum. In p^, p^, p* the trito-
cones are nearly equal to the protocones, and both
356
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
external and internal cingula are extremely progres-
sive, the external cingula being well defined across the
base of the protocones and tritocones, and the inter-
nal cingulum forming a wide basal shelf extending
around the whole anterior as well as the posterior in-
ternal border of the crown. P^ is thus almost like p'
(except for its smaller size and relatively smaller
by the more advanced condition of the deuterocone
of p", of the tritocones of p^"*, and of the internal
cingula. At the same time the premolars of S. in-
cisivus simulate those of Dolichorhinus longiceps, espe-
cially in their advanced tritocones, but are distin-
guished from them by their greater breadth and far
heavier internal cingula. A still nearer resemblance
Figure 301. — Type skull of Sthenodectes incisivus
Ona-fourth natural size. Carnegie Mus. 2398 (type). About 3 miles northeast of well 2, Uinta Basin, Utah; upper levels of EobasUevs-DolichorMnus
zone (Uinta B 2). Ai, side view; Ai, top view; A3, palatal view.
transverse diameter), whereas in the type of T. ulti-
mum p^ is much simpler than p^ P'* (ap. 23 mm., tr.
32) and p* (ap. 23 mm., tr. 39) are correspondingly
advanced but unlike T. ultimum show no trace of
tetartocone ridges and swellings.
The premolar series is thus readily distinguished
from that of the contemporary Manteoceras uintensis
is with the premolars of Metarhinus earlei, in which
p^ is almost as progressive and p^ and p* have heavy
internal cingula.
Molars. — The molars of the type are distinguished
from those of T. ultimum by the greater minimum
transverse diameter of m\ m^, by the less prominent
hypocone on m", by the weaker internal cingula on
EVOLUTION or THE SKULL AND TEETH OF EOCENE TITANOTHERES
357
m^, m', and by the detailed form of m^ From those
of the contemporary Manteoceras uintensis they are
distinguished by their smaller size, by the greater rela-
tive breadth of m^, more quadrate contour, especially
of m^ relatively smaller parastyles and mesostyles,
sharper external cingula. M^ is wide anteriorly (53
mm.) and narrow posteriorly; the posterior V is rela-
tively small. The molars of MetarMnus earlei are
relatively longer anteroposteriorly.
Comparative measurements of Telmatherium and Sthenodectcs, in
millimeters
■ Basal length of skull
Zygomatic breadth of skull. .
Cephalic index
Length of dental series (i'-m^) .
P'-m3
P'-p*
M'-m3
I', ap. by tr
P, ap. by tr
P, ap. by tr
C, ap. by tr
C, vertical
P*, ap. by tr
Ml, ap. by tr
M^, ap. by tr
M^, ap. by tr
T. ultimum,
Am. Mus.
2060 (type)
500
300 +
60
305
218
89
130
14X13
15X14
19X17
* 25X23
* 42X?
25X34
39X36
45X46
44X51
Carnegie Mus. Field Mus.
2398 (type) 12168
490
"■310
63-65
295
207
84
125
22X21
25X26
25X27
27X27
67X?
23X39
40X44
45X49
41X46
211
132
" Estimated. ' Am. Mus. 2004.
Lower jaw. — A lower jaw (Field Mus. 12166), found
on the same geologic level as the skulls but at some
distance, belongs to an aged individual. The crowns
of the incisors are almost worn away. The following
description and measurements are from Riggs (1912.1,
pp. 38, 39):
The mandible is 10 millimeters shorter than would be required
to fit the skull, but the dentition matches closely. The molars
have the strength necessary to oppose the massive upper
series; the canines and incisors, though not so massive as those
above, show such wear as would be expected in this form. The
canines are worn away diagonally at the point of contact with
the third upper incisor, but very little from contact with the
upper canines. There is a short diastema between canines and
premolars. The mandible as a whole is titanothere-like — deep
through the ramus, broad at the angle, concave in the tooth
line, and tapering toward the anterior extremity. The coronoid
is short and recurved at the tip.
Lower jaw of Field Mus. 12166
Millimeters
Length, condyles to incisors 360
Height, condyles above angle 168
Length of molar-premolar series 215
Length of molar series 130
Length of crown of canine (estimated) 30
Diameter of crown of canine 19
Depth of ramus from base ofps 60
Depth of ramus from base ofms 84
THE MANTEOCERAS-DOLICHORHINUS GROUP (mANTE-
OCERAS, MESATIRHINUS, DOLICHORHINUS, SPHENO-
COELUS, METARHINUS, RHADINORHINUS)
Stages, series, and subfamilies. — This second great
group of middle and upper Eocene titanotheres is
characterized by precocious horn swellings above the
eyes and many other featui'es in common. The single
specimen of this group (Eometarhinus) , discovered in
the Huerfano B ( = Bridger A) horizon of the Huerfano
formation of Colorado, is the sole known forerunner.
With this exception, this group is of much later geo-
logic appearance than the first group {Palaeosyops,
Telmatherium), being found in the upper levels of the
Bridger Basin, in the Washakie Basin, and in the lower
and middle levels of the Uinta Basin. The group
commenced to flourish in the Bridger and Washalde
regions during the period of the decline of the Palaeo-
syops phylum and survived it for a very long period,
but it was contemporaneous with the Telmatherium,
phylum.
We find that this group radiates into four series, as
follows :
Series included in the Manteoceras-Dolichorhinus group
Progressively large and mesati-
cephalic to brachycephalio
Progressively large and dolicho-
cephalic
Arrested in size, mesati
ephalic to dolichocephalic
Nasals wide
Nasals pointed
Later stage
Earlier stage
Protitanotherium ...
Unknown
Metarhinus
Eometarhinus. . _ . ..
(?)•
Rhadinorhinus.
Besides the rudimentary horns there are very
numerous characters which tie the members of this
second group together and distinguish them from the
palaeosyopine group. These characters point indis-
putably to a common ancestor. An underlying unity
of descent is at once observed in the accompanying
figures (fig. 302) of the four types of skulls included
in this group, which are all reduced to the same scale.
The four series are grouped into subfamilies and
genera as shown below.
358
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Suhdiinsions of the Manteoceras-JDolicJiorMnus group
Subfamilies
Manteoceratinae (=Brontopinae)
Dolichorhininae
RhadinorMninae
Horns on frontals.
No infraorbital shelves.
Mesaticephalic to brachycephalic.
Nasals broad.
Horns chiefly on nasals.
Large infraorbital shelves.
Dolichocephalic to hyperdoliohocephalic.
Nasals broad.
Horns retarded.
Rudimentary infraorbital shelves.
Dolichocephalic. Facial region upturned.
Nasals pointed.
Genera
Protitanotherium (Eocene) .
Manteoceras (Eocene) .
Mesatirhinus (Eocene).
Dolichorhinus (Eocene) .
Metarhinus (Eocene) .
?Sphenocoelus (Eocene) .
Eonietarhinus (Eocene) .
Rhadinorhinus (Eocene) .
The phyletic position of the recently discovered
Eometarhinus, from Huerfano B ( = Bridger A), is
ancestral either to Metarhinus or to Rhadinorhinus.
Of these genera Manteoceras and Mesatirhinus rep-
resent phyla which appear contemporaneously in the
upper Bridger but which have already diverged from
each other toward brachycephaly and dolichocephaly,
respectively. As these subphyla diverge more and
more the resemblances which are observed between
the lower members of each series become fewer, and
the differences become greater. Thus Manteoceras
and Mesatirhinus are much nearer each other than the
forms to which they respectively gave rise, namely,
Protitanotherium and Dolichorhinus. The Rhadino-
rhinus phylum may prove to be a distinct one, and in
some characters it points toward the Oligocene Mega-
cerops (Symhorodon) .
SUBFAMILY MANTEOCERATINAE {=BRONTOPINAE) OSBOEN, EOCENE
ANCESTOES
A branch of the same stock as that of Mesatirhinus
and Dolichorhinus. Precociously horned animals,
known from the upper deposits of the Bridger Basin,
from the Washakie Basin, and from the Uinta Basin.
First referred to Telmatherium and subsequently de-
scribed as Manteoceras, or "prophet horn." In all
known characters more nearly central or ancestral to
the Oligocene titanotheres of the genus Brontops than
any of the Eocene forms thus far discovered.
Manteoceras
General structure and habits.- — The presence of the
rudiment of a horn above and in front of the eyes is
the most distinctive and interesting feature of the
middle Eocene Manteoceras, which is the earliest known
member of this subfamily. Many more characters
both of the skull and the teeth make this a prophetic
or ancestral form of great significance and interest,
worthy of the most thorough, detailed study. Alto-
gether more than fourteen such prophetic characters
have been found in these animals. In point of size
the known individuals are intermediate between the
largest tapirs and the smaller rhinoceroses, such as
Rhinoceros (Dicerorhinus) sumatrensis.
The skull in these animals is moderately elongate,
or mesaticephalic. The fluctuations are between
mesaticephalic and brachycephalic types. Female
skulls tend to be somewhat more long and narrow;
aged male skulls tend to be broader and more robust.
The parts of the limbs and feet which signify speed,
especially the humerus, femur, and manus, indicate
that the quadrupeds belonging to this genus were
swifter than Palaeosyops but slower than Mesati-
rhinus. They were brachypodal as compared with
Mesatirhinus but considerably longer footed than
Palaeosyops. The large tusks of the males and the
earlier development of horn rudiments as compared
with the palaeosyopine group indicate that these
quadrupeds were vigorous fighters. In a large per-
centage of the adult specimens the teeth are much
worn, indicating that the food was somewhat harder
and drier than that of Palaeosyops. As feeders these
animals were better equipped than the members of
the Palaeosyops and Limnohyops series, for their
grinding teeth were decidedly more trenchant or cut-
ting, but even in the later members of Manteoceras
the grinding teeth are somewhat less efficient than
those of the contemporary telmatheres, because the
molar ectolophs are a little shorter and the premolars
are less advanced in evolution.
History of discovery. — The discovery of these animals
was one of the turning points in the history of the evo-
lution of the titanotheres. In 1894 the American
Museum expedition was working under the direction
of Dr. J. L. Wortman in a layer of brown sandstone 3
miles north of the base of Haystack Mounain, in what
is now loaown as the Washakie A level. Here two
skulls (Am. Mus. 1569, 1570) were found, and as
partly exposed in the field they attracted the attention
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHEKES
359
of Doctor Wortman as seemingly different from any
previously discovered. He described them in a letter
written to Professor Osborn from the field as exhibit-
ing rudimentary horns at the junction of the f rentals
and nasals and suggested the generic name Manteo-
ceras or "prophet horn." On the arrival of these
skulls at the American Museum Professor Cope, the
writer, and others who examined them expressed
great doubt as to whether the tuberosities (PI. XVI;
figs. 305, 307) above the orbits could really be re-
garded as incipient horns. These doubts were soon
removed by the discovery of similar horns in Dolicho-
rhinus cornutus { = 'hyognathus) of the middle Uinta,
an^ Doctor Wortman's observation was thus verified.
As detailed in Chapter III (p. 151) the animal was
first identified by Osborn with the imperfect upper
CO type teeth of the species Palaeosyops (Telmatherium)
vallidens Cope, previously found by Professor Cope in
the Washakie Basin; but it was subsequently ascer-
tained that this species, now provisionally referred to
the genus DolichorTiinus, belongs in a higher level,
Washakie B, whereas the types of Manteoceras man-
teoceras were both found in Washakie A.
These animals {M. manteoceras) were first supposed
to be confined to the lower levels of the Washakie
Basin, but subsequent exploration of the upper
Bridger by the American Museum expeditions has
proved that they were still more numerous in the
Bridger Basin; altogether the remains of more than
twenty animals of the type species (M. manteoceras)
have been found by the American Museum parties,
including seven skulls in Bridger D and four skulls in
Washakie A. In the upper levels of horizon A of the
Washakie Basin a more advanced stage has been
found, M. washaJciensis. Thus far these animals have
not been found in the Uinta Basin in beds of level B,
deposited during a period when they undoubtedly
lived; but in the lower part of Uinta C the genus reap-
pears in the important species described by Douglass
as Manteoceras uintensis. In the lower part of Uinta
C an animal nearly related to Manteoceras, if not its
direct successor, was discovered by the Princeton
expedition in 1894 and was subsequently recognized
by Hatcher as probably a successor of Manteoceras,
and named by him ProtitanotTierium emarginatum.
Geologic distribution. — The geologic levels at which
the remains of these animals have been found are
shown in Figure 334, and as the remains are numerous
in the upper Bridger, levels C and D, and in the lower
Washakie, level A, they indicate that these deposits
are contemporaneous. As observed in the text on
Telmatherium, the advent of Manteoceras appears to
have been contemporaneous with the last stage in the
development of the Palaeosyops-Limnohyops phylum
and with the first appearance of the Mesatirhinus-
Dolichorhinus phylum. The abundance of remains of
these animals in the upper Bridger deposits is very
101959— 29— VOL 1 26
striking. It is possible that they are represented also
by skeletal remains in the lower Bridger.
Affinities to other Eocene titanotheres. — The resem-
blances and contrasts between Manteoceras and Tel-
matherium have been pointed out in some detail in the
JDolicTzorh in us
A^anteoceras
Figure 302. — Skulls of titanotheres of the
Manieoceras-Dolichorhinus group
One-eighth natural size. A, Manteoceras manteoceras, middle
Eocene of Bridger Basin. Wyo.; upper Bridger. B, Mesati-
rhinus petersoni, middle Eocene of Bridger Basin, Wyo.; upper
Bridger. C, Melarhinus earlei, middle Eocene of Washakie
Basin, Wyo.; summit of Washakie A. D, Bolichorhinus hyo-
Snaihus, middle Eocene of Uinta Basin, Utah; Uinta B 2.
descriptions of Telmafherium. They may also be very
clearly seen by comparing the crania of the types of
these two general (figs. 210, 219). To summarize
Manteoceras is distinguished from Telmatherium by (1)
deeper facial concavities; (2) much more prominent
360
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
and rugose frontonasal horns; (3) progressive anterior
flattening of the vertex of the cranium and recession of
the sagittal crest; (4) presence of a supraparietal pit
and strongly bifid sagittal crest; (5) broad and de-
pressed occiput; (6) oblique shelf-like suborbital part
of the malars; (7) widely arched zygomata with de-
scending flange of malar relatively shallow; (8) pro-
gressively more round-topped superior incisors, the
outer relatively smaller than in Telmatherium; (9)
shorter, heavier, and rounder superior canines, with
very heavy roots; (10) less progressive tritocones, deu-
terocones, and cingula on superior premolars; (11)
somewhat less pronounced hypsodonty of the grinding
teeth; (12) broader and more rounded mesostyles;
(13) the less deep and finally more elongate premax-
illary symphysis.
Comparisons with MesatirTiinus and DolichorTiinus. —
The general resemblances of these animals have been
enumerated above. A number of resemblances in
general conformation are seen by comparison of
similar views of the crania of Manteoceras and of
Mesatirhinus. These indicate a closer ancestral af-
finity to MesatirTiinus than to Telmatherium. De-
tailed points of resemblance between Manteoceras and
Mesatirhinus are seen in (1) the tendency to form a
suborbital shelf, which is more pronounced in Mesati-
rhinus than in Manteoceras; (2) the depth of the facial
concavities, giving prominence to the nasofrontal
horn rudiments (a distinction must be noted here,
however, that the horn rudiments in Mesatirhinus
and Dolichorhinus are borne rather by the nasals than
by the frontals, whereas in Manteoceras the reverse is
the case); (3) pronounced affinities in the foot and
limb structure.
The statement may be made very emphatically,
therefore, that Manteoceras and Mesatirhinus have
risen from a common stock.
The distinctive characters of Manteoceras lie prin-
cipally in the proportions of the skull, dentition, and
feet and in the divergent evolution of the premolar-
molar series. Manteoceras is mesaticephalic in skull
and tooth structure and subbrachypodal in foot
structure, while Mesatirhinus is progressively both
dolichocephalic and dolichopodal.
Incipient horns. — As observed above, a notable
characteristic of these animals is the precocious horn
development. The horn swellings are borne directly
over the frontonasal suture (Pis. XVI, XVII). They
involve very slight convexity and are slightly rugose
only in the more aged specimens. As they are
exhibited in various degrees in all the skulls known,
they were certainly present in both sexes, although less
prominent in the females. These horn swellings have
a different origin in Dolichorhinus, as well as in Mesati-
rhinus, for in these genera (PI. XVII, figs. B, C\ C),
although placed about as in Manteoceras, they are borne
chiefly on the nasals and partly on the frontals — that
is, in front of the frontonasal suture.
Facial concavities. — The second distinctive character
that is correlated with or lends itself to this precocious
development of the horns is the concavity in the side
of the face, in front of the orbit, beneath the nasal.
This gives a greater prominence to the horn rudiments
and in life would permit the warty epidermal swellings
that covered these rudiments to be used more ef-
fectively in butting. This overhanging frontonasal
suture showB a wide contrast to the condition seen in
Palaeosyops. The concavity of the face in front of the
orbit, beneath the horn, is a very prominent feature
also in the Oligocene titanotheres and in Sthenodectes
incisivus of level B of the Uinta Basin, Utah.
The vertex. — In lateral or profile view the skull is
convex above the brain region, concave in the mid-
cranial region, and convex again in the nasal region,
as in Mesatirhinus. The horn rudiments, or hornlets,
are thus thrown into considerable prominence both
laterally and superiorly. The concave midportion of
the skull is again a progression in the direction of the
saddle-shaped top of the titanothere cranium. When
viewed from above the cranium also exhibits a spread-
ing of the space beneath the supratemporal ridges in
such a manner that the sagittal crest proper is limited
to the posterior region. In the V-shaped space on
top of the skuU between these converging temporal
ridges (or bifid sagittal crest) it is especially interesting
to observe that a deep pit is developed in the more
progressive and older forms, because we shall find a
vestige or reversion to this pit on top of the large,
flattened crania of some of the Oligocene titanotheres
{Brontops, compare figs. 304, 307, 374).
Dentition. — The superior incisors form a more
A-shaped series than in Dolichorhinus, where they
tend to form a fl, and the inferior incisors are more
transverse in position. The grinders are less hypso-
dont on the ectoloph, and the protocone tips are more
blunt than in Mesatirhinus and Dolichorinus. The
premolars are less advanced than in Mesatirhinus and
Dolichorhinus, because the tritocones and deuterocones
are relatively smaller, the ectolophs less flat, and the
"ribs" on the external face of the outer cusps wider
at the base. The relative degree of progression of
the premolar ectolophs in Dolichorhinus and Manteo-
ceras is a very complex matter, but after careful
comparison it may be summarized as follows:
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TXTANOTHERES
Characters oj 2?^ in Manteoceras, Mesatirhinus, and DolichorMnus
361
Manteoceras
Mesatirhinus
Dolichoihinus
Very broad at base
Strong
Very broad at base, but "rib"
appearing.
Often pinched or riblike.
Gently rounded.
Still more anterior.
Filling out.
Tritocone nearly equal to
tocone.
Tritocone convexity. __ _
More anterior
More rounded
Tritocone relatively somewhat
larger.
Relative size of protocones and
tritocones.
Variable
pro-
In general p^ in DolichorMnus is in a much more
advanced stage than in M. manteoceras. In p', p*
these differences become more pronounced. The
premolars were thus evolving along divergent lines in
Manteoceras on the one hand and in Mesatirhinus and
Dolichorhinus on the other. The general subfamily
kinship of M. manteoceras with Mesatirhinus and
Dolichorhinus is shown especially in the comparison of
p^, p*, in M. washaJciensis and Mesatirhinus petersoni,
but the generic differences are still evident.
Jaw structure. — The jaws are prophetic of the Oligo-
cene type, especially in the posterior region, with an
elevated coronoid, and with the border sharply de-
pressed below the angle (fig. 310); the chin, however,
is weaker and the coronoid relatively much larger.
Sex characters. — Differences in sex are indicated by
the smaller size of the canines in the females, as
observed in M. manteoceras. It is difficult to deter-
mine positively whether the horns are also less promi-
nent in the females than in the males. One well-
preserved, very old Manteoceras skull (Am. Mus.
12678) from Bridger C 5 has small canines and appears
to be a female. In it the horns are hardly less promi-
nent than in the type male. The type of M. washa-
Tciensis has very minute horn swellings and might be
taken for a female, but its canines are of intermediate
size.
Mesaticephalic slcull proportions. — The skulls are in-
termediate in proportion, or decidedly broader than
those of Mesatirhinus and Dolichorhinus and much
longer and narrower than those of Palaeosyops, the
breadth being about three-fifths the length, and they
may thus be described as mesaticephalic. In the ear-
lier forms of M. manteoceras of the middle Eocene the
zygomatic arches are rather stout and well arched.
In the much later M. uintensis they are more slender
than in Telmatherium ultimum but diverge widely,
forming a decided angle with the glenoid region.
There is only a rudiment of the infraorbital shelf
that is so characteristic of most species of Mesatirhinus
and Dolichorhinus.
Detailed features. — Characteristic detailed features,
some of which trend progressively toward the Oligocene
titanotheres, clearly distinguish these animals from
Palaeosyops and in a less degree from Mesatirhinus:
(1) The premaxillary symphysis is long and firm as
compared with that of Palaeosyops but shorter than in
Dolichorhinus; (2) the nasals are very characteristic,
being relatively short and stout, decidedly truncate,
distally somewhat spreading and laterally much
recurved; (3) in the sagittal line of the skull the suture
between the frontals becomes obliterated in adults, as
in many other ungulates with large diploe; (4) the
occiput is low and broad (fig. 306), very distinct in
form from that of Palaeosyops, and in the more
advanced specimens {Manteoceras washaJciensis) it
exhibits the lateral pillars which are so characteristic
of the Oligocene titanotheres.
Summary of progressive characters of Manteoceras
toward Brontops and other Oligocene titanotheres. —
Hatcher, at the time of the discovery and description of
the animal now called Protitanotherium emarginatum of
Uinta C, pointed out the fact that Manteoceras is in or
near the main ancestral line of the Oligocene titano-
theres rather than Dolichorhinus, which Osborn had
supposed also in that ancestral line.
The progressive characters of Manteoceras toward
the Oligocene forms are naturally somewhat more
marked in old than in young specimens. The follow-
ing items relate chiefly to the species M. manteoceras,
which is the most fully known and seems to lead espe-
cially toward the Oligocene Brontops: (1) Middle
part of the skuU elongate, face never very long, elon-
gation becoming very pronounced in the Oligocene
titanotheres; (2) rudimentary frontal-nasal horns ap-
parently increasing in size with age and probably more
pronounced and more rugose in the males, as in all the
Oligocene forms; (3) concavities in front of the orbits,
causing the rudimentary horns to overhang the sides of
the face (very prominent in the later Oligocene gen-
era); (4) nasals broad distally, shorter than in Doli-
chorhinus and in M. uintensis, suggesting the Brontops
362
TITANOTHERES OP ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
and Menodus nasals; (5) premaxillary symphysis
deepened and keeled (compare Oligocene genera); (6)
middle or frontal portion of the cranium flattened, the
flattening being associated with the progressive ob-
literation of the suture between the frontals and with
the abbreviation of the sagittal crest (compare Oligo-
cene genera); (7) middle portion of the skull saddle-
shaped in the region between the frontal-nasal horns
and occipital crest, showing a tendency that becomes
extreme in the Oligocene forms; (8) overhanging
supratemporal crests or ridges characteristic of age,
a tendency observed also in Oligocene titanotheres; (9)
occiput broadened and lateral pillars above the con-
dyles incipient, a feature observed in all Oligocene
titanotheres; (10) incipient expansion of the zygo-
matic portion of the squamosals and flattening out of
the squamosal portion of the zygoma, as in later
titanotheres; (11) deep backward angulation and
depression of the angle of the jaw, a feature observed
in certain Oligocene genera; (12) incisors tending to
become round-topped {M. uintensis), a tendency that
becomes very pronounced in the Oligocene genera;
(13) crowns of the canine teeth abbreviated, with
stumpy recurvature, foreshadowing the Oligocene
Brontops and Brontotherium; (14) ectolophs of the
premolar and molar grinding teeth elongated vertically,
a character that becomes pronounced in all Oligocene
titanotheres; (15) premolar ectolophs, showing incip-
ient double convexities, a character that becomes well
marked in all Oligocene genera; (16) fourth premolar
showing a famt suggestion of the tetartocones {M.
wasliakiensis) ; (17) premolars retarded in development.
Despite the approaches of Manteoceras to the Oligo-
cene Brontops in these 17 characters, there are reasons
why none of the known species of Manteoceras, and
especially the best known, M. manteoceras, can be
considered the direct ancestor of any known Oligo-
cene titanothere. This species differs from the Oli-
gocene titanotheres notably in the sharp postcanine
constriction of the face, the shallowness of the malar
below the orbit, and the slenderness of the malar
behind the orbit; and it is not yet known whether these
are progressive tendencies leading away from the
Oligocene type or are characters that were lost during
the transformation into the early Oligocene types, such
as Brontops iracTiycepJialus.
Manteoceras Hatcher
Plates XVI, XVII, XXIX, XLVI, LI, LIII, LV, LXIII, LXVII;
text figures 27, 29, 33, 87, 113, 121, 131, 132, 210, 215, 219,
220, 255, 302-313, 323, 324, 380, 406, 408, 409, 483, 484, 508,
510, 512-517, 521, 551-557, 566, 641, 646-649, 661, 673, 674,
685, 686, 688, 690, 701, 709, 710, 712, 717, 720, 721, 723, 724,
733, 745
[For original description and type references see p. 177. For slceletai characters
see p. 630]
Localities and geologic horizons. — Bridger Basin,
Wyo., levels C and, chiefly, D; Washakie Basin, Wyo.,
level A; Uinta Basin, Utah, lower part of level C
{M. uintensis).
Generic characters. — Facial concavities pronounced;
horn rudiments borne chiefly on the frontals; suborbital
portion of malars flattened, with a rudimentary shelf.
Superior incisors moderately enlarged; canines robust,
pointed, progressively more obtuse. Grinding series
subhypsodont; molar conules vestigial or wanting;
molars broader than in DolicliorJiinus or its allies;
premolar evolution retarded as to tritocones, deutero-
cones, and cingula; ectolophs with two convexities in
tandem.
As described in detail m the revision of the nomen-
clature (pp. 177-178) the synonymy of the genus and
type species has been confused and complicated, but it
has now been definitely cleared up according to modern
principles. The honor of discovering this important
evolution stage of M. manteoceras and of first re-
cognizing its prophetic character belongs to Wortman,
who also invented the apt name Manteoceras (prophet
horn). Osborn in 1895, the first to publish a descrip-
tion of the skull of this animal, refrained from giving
it a new name on account of the general resemblance
in the teeth to the very imperfect type of Cope's
"Palaeosyops vallidens." Hatcher, later in the same
year, proposed the generic name and correctly de-
fined the genus. Still later Osborn in manuscript
referred to this form as Palaeosyops manteoceras; but
this specific name is technically to be credited to Hay,
who (1902.1) first fastened the specific name man-
teoceras to the previous description and figures of the
original "prophet horn" skulls, so that the name now
stands as Manteoceras manteoceras Hay (Osborn MS.).
It is very important to note that seven skulls and
uppfr dentitions from the upper deposits of the
Bridger Basin (mostly level D) equal or exceed in
measurement and progressive characters three skuUs
from level A of the Washakie Basin and thus afford
corroborative evidence of the simultaneous deposition
of those sediments.
Materials. — A somewhat detailed enumeration of
materials seems to be important in this case for
purposes of geologic correlation.
1. Manteoceras manteoceras
Bridger C 2: A fragmentary adult skull (Am. Mus. 12194),
from Burnt Fork post office (Henrys Fork). The sagittal crest
bifid with deep intermediate pit. This is in an early stage of
development.
Bridger D: A male skull (Am. Mus. 12683), from Sage Creek
Spring, is important as supplementing the characters of the
type. It probably belongs to a somewhat early stage and
presents certain resemblances to the type of Telmatherium
cultridens. The measurement of p'-m^ is 176 millimeters, as
compared with 181 in the type of M. manteoceras. The most
striking feature (see figs. 305, 307) is the depth of the preorbital
concavities, which throws the frontonasal horn ridges into
exceptional prominence. The horn surfaces are slightly pitted
or rugose.
Bridger C or D: A fuUy adult skull (Am. Mus. 1511) found
on Henrys Fork. Basilar length, 447 millimeters; p>-m', 184;
canines large. Probably a male specimen.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
363
Bridger C or D: Skull of a very old male (Am. Mus.) 1545
found on Henrys Fork. Basilar length, 523 millimeters;
pi-m', 197; large fanged, recurved canines. Horn rudiments
rugose.
Bridger C or D: Skull, jaw, and parts of skeleton of a male
adult, with open sutures (Am. Mus. 1587) from Henrys Fork.
Affords characters of the feet.
Bridger D 2: Skull, jaws, and parts of skeleton (Am. Mus.
12204). Basilar length estimated at 490 millimeters. Probably
a male, aged. Grinding teeth relatively small. Affords
knowledge of the femur and part of the feet.
Bridger C 5: Very old female skull (Am. Mus. 12678). The
first superior molar of both sides has dropped out. Canines
short, recurved, cingulate posteriorly. Basilar length, 500
2. Manteoceras washakiensis Osborn
Washakie A (upper levels) : Crushed but complete skull (Am.
Mus. 13165) from the base of Haystack Mountain, summit of
the brown sandstone, or upper part of Washakie A. A female
with relatively small, obtuse, recurved and posteriorly cingulate
canines. Horn rudiments slightly defined. Grinding series,
pi-m^, 200 millimeters.
3. Manteoceras sp.
Washakie B : Of the three specimens or cotypes described by
Cope as Palaeosyops vallidens, the jaws (Am. Mus. 5098) from
Mammoth Buttes appear to belong to Manteoceras sp. indet.
The other cotype (upper teeth) is referred to Dolichorhinus (see
below) .
Figure 303. — Skulls of Manteoceras manteoceras
One-fourth natural size. A, Top view; Am. Mus. 1587, Henrys Fork, Bridger Basin, Wye, upper (?) levels. B, Palatal view, chiefly from Am.
IVlus. 2353, south of Haystack Mountain, Washakie Basin, Wyo., lower beds; partly restored from Am. Mus. 1570, La Clede, Washakie Basin; some
details and sutures from a specimen now in the National Museum, formerly Am. Mus. 1545, Bridger Basin.
millimeters; zygomatic breadth, 294 (estimated); p'-m^, 187.
Equal in size to palate from the upper portion of Washakie A.
Washakie A: The type skull of the species (Am. Mus. 1569)
from the brown sandstones. Probably an aged male. Horn
rudiments prominent, slightly rugose. Estimated basilar
length 492 milhmeters; p'-m', 183.
Washakie A: The type skull (Am. Mus. 1570). Adult male
skull, occiput lacking. P>-m', 186 millimeters. Agrees closely
in size with Am. Mus. 1511, from Bridger C or D.
Washakie A: Skull with jaws (Am. Mus 2353) lacking pre-
maxillae. Zygomatic breadth, 274 millimeters. The grinding
teeth are of relatively small size (p'-m', 178 mm.).
4. Manteoceras uintensis Douglass
Uinta C (lower levels) : The anterior half of a skull (Carnegie
Mus. 2388), "from gray sandstone in red Uinta beds," the
type of M. uintensis. A large male; grinding series, 240
millimeters (see below).
Synopsis of progressive characters in the three succes-
sive species of Manteoceras. — 1 . M. manteoceras: Levels,
Bridger C 2 to D and Washakie A. Skull of medium
size (basilar length 447-500+ mm.). Face relatively-
short; zygomata stout; horn swelling prominent.
364
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
P'-m^ 176-186+ millimeters; postcanine diastema
short; i^ much larger than i^; p" with deuterocone and
tritocone poorly developed; p"* (ap. by tr.), 19 by
26 to 22 by 29 millimeters; m\ 28 by 29 to 32 by 33;
m^ 35 by 37 to 41 by 40; m^, 36 by 39 to 39 by 43.
2. M.washakiensis: Level, Washakie A, upper part.
Slvull somewhat larger (basilar length in supposed
female 490 mm., estimated). Face relatively short;
zygomata moderate; horn swelling inconspicuous (? 9 ).
P'-m', 200 milhmeters (estimated); p^ with deutero-
cone somewhat better developed; p* (ap. by tr.), 23 by
29 millimeters; m\ 35 by 38; m^ 41 by 43; m', 39
by 42.
3. M. uintensis: Level, Uinta C, lower part. Skull
larger (basilar length not known). Face relatively
long; female horn swelUng (?) absent; zygomata not
stout, in inferior view forming a marked angle in front
of the glenoid surface. Postcanine diastema long;
disparity of i^ over i^ less marked; p'-m', 240 milli-
meters; p^ with deuterocone slightly and tritocone
markedly more advanced; p* (ap. by tr.), 26 by 32
millimeters; m', 40 by 38; m^, 52 by 46; m', 45 by 50.
Range in size of Manteoceras manteoceras and two successive stages oj increase in size, in millimeters
Manteoceras manteoceras
M. wa-
shakien-
sis, 13165
M. uin-
12194
12678
12683
1511
12204
2353
1532 -
1569
1570
1545
2388
Skull:
500
'■294
240
255
187
447
»490
492
°310
°245
305
63
183
164
80
103
186
165
81
107
623
'350
"209
"320
-66
197
181
83
118
490
»290
Dentition :
Pi-m3 1
176
160
79
100
15X8
18X16
17X21
20X26
28X29
35X37
36X39
184
168
82
104
16X11
22X19
19X25
20X28
27X°30
38X37
39X39
29X31
33X38
36X38
178
159
77
102
17X 9
18X18
18X23
21X28
30X31
38X39
35X38
»200
-183
83
116
15X 9
19X17
20X25
23X29
35X38
42X43
39X42
240
p2-in3
219
Pi-p*
101
Mi-ms
P', ap. by tr
108
138
24X11
P-, ap. b3' tr _._ _ _
22X18
20X25
23X29
33X34
42X40
42X42
21X19
19X24
20X28
31X33
38X41
38X43
"22X29
32X33
41X40
39X43
21X20
P', ap. by tr ..
24X29
P*, ap. bv tr .
20X27
29X31
38X38
40X39
26X32
M', ap. bv tr . .
40X38
M-, ap. by tr
MS, ap. by tr
52X46
45X50
" Estimated. * Crushed.
Numbers at heads of columns are those of the American Museum except the last (2388), which is of the Carnegie Museum.
The geologic horizon and other facts concerning the specimens are given below:
12194. Intermediate molar proportions. Bridger C 2.
12678. Very old female. Intermediate molar proportions.
Bridger C 5.
126S3. Young adult male. Smallest molar proportions.
Bridger D.
1511. Male. Small molar proportions. Bridger (?).
12204. Very old female. Small molar proportions. Wa-
shakie D 2.
2353. Female. Small molar proportions. Washakie A.
The accompanymg table of measurements brings
out the following facts :
1. In M. manteoceras there is a very considerable
range in size: Am. Mus. 1545 is larger in total skull
length than the type of M. wasliaMensis , but the first
and second molars are smaller.
2. The molars in different specimens of M. manteo-
ceras are either microdont (Am. Mus. 12194, 12683,
1511, 12204, 2353) or macrodont (Am. Mus. 1545,
1570), but other measurements, especially the dimen-
sions of the premolars, do not confirm this division;
1532. Male. Large molar proportions. Washakie A.
1569. Type. Male. Large molar proportions. Washakie A.
1570. Paratype. Large molar proportions. Washakie A.
1545. Old male. Largest molar proportions. Bridger
D (?).
13165. Female. Skull medium, cheek teeth large. Washa-
kie A, upper levels.
2388. Type. Male. Skull and molars very large. Uinta
C, lower level.
it does not seem to be due to sex, nor, so far as known,
to imply specific differences.
3. M. wasliaMensis, from the upper levels of
Washakie A, is not much bigger in total skull length,
but it is more progressive in the relatively large size
of the molars.
4. M. uintensis, from Uinta C, is far more advanced
than either M. manteoceras or M. washaJciensis in total
skull length (inferred), length of face, and all dimen-
sions of the dentition; but the molars are relatively
more advanced than the premolars.
EVOLUTION OP THE SKULL AND TEETH OF EOCENE TITANOTHERES
365
Manteoceras manteoceras Hay
[Telmalotherium manteoceras (Osborn MS.) ; Telmatotherium
vallidens Osborn, not Cope]
Plates XVI, XVII, XXIX, XLVI, LI, LIII, LV, LXIII,
LXVII; text figures 27, 29, 33, 87, 113, 215, 219, 255, 302-311,
323, 324, 380, 406, 408, 483, 508, 512-514, 516, 517, 521,
551-554, 556, 557, 566, 646-649, 661, 686, 701, 709, 712, 721,
723, 724, 745
[For original descriptions and type references see p. 177. For skeletal characters
see p. 630]
Type locality and geologic horizon. — Washakie Basin,
Wyo.; Uintatherium- Manteoceras- Mesatirhinus zone
(Washakie A). The most abundant material is from
the Bridger Basin, Wyo., some from Bridger C but
more from Bridger D.
Specific characters. — Skull of medium size, basilar
length, 447-500+ millimeters; cephalic indices, 58 to
68, face relatively short; zygomata stout; horn swelling
prominent; p'-m', 176-186+ millimeters; postcanine
diastema short; i^ much larger than i^; p^ with deutero-
cone and tritocone poorly developed; p^ (ap. by tr.),
19 by 26 to 22 by 29 millimeters; m^ 28 by 29 to 32
by 33; m^ 35 by 37 to 41 by 40; m^, 36 by 39 to 39
by 43.
The sTceleton. — The skeleton of Manteoceras is by
no means so fully known as that of Palaeosyops. The
feet are more slender than those of Palaeosyops leidyi
but much more robust than those of Mesatirhinus.
Intermediate proportions are seen throughout between
those characteristic of Palaeosyops, the extremely
broad-headed titanotheres, and of Dolichorhinus, the
extremely long-headed titanotheres. In Manteoceras,
therefore, the moderate breadth of the skulls (mesa-
ticephaly, PI. LIII) is associated with moderate
breadth of the feet (mesatipody).
Progressive and specific characters: (1) Several muta-
tions, subspecies, or substages from several different
levels may be represented in the 20 or more specimens
that have been referred to M. manteoceras; (2) as
shown above, there is a considerable range in size
between the smallest specimen (Am. Mus. 12683) and
the largest; (3) some skulls have rather small grind-
ing teeth in transverse measurement and are thus
microdont; others have large grinding teeth and are
thus macrodont, and this is not a sexual character;
(4) in some the canines are more slender (figs. 308,
309), in others more robust, the form typical of the
species being represented in Figure 311; but it is cer-
tain that the canines in some lines become progres-
sively obtuse and posteriorly cingulate and thus ap-
proach the Oligocene types.
From the detailed list of the materials given above
it is seen that the known individuals from the upper
levels of the Bridger Basin and the lower levels of the
Washakie Basin are from a single geologic horizon —
the Vintatherium- Manteoceras- Mesatirhinus zone. This
horizon, however, represents a long period of time,
but, owing partly to the slow rate at which the pre-
molars in Manteoceras were evolving, the known
specimens, although probably representing several
different levels, do not present very marked progres-
sive differences, except that Am. Mus. 12683, from
Bridger D, is less advanced in the condition of the
deuterocone of p^.
Horns. — As shown in the carefully drawn detailed
figures (Pis. XVI, XVII), the rudimentary horn con-
vexity is borne chiefly upon an anterior spur of the
frontals; it thus presents exactly the same relations
as those observed in Protitanotherium emarginatum
(figs. 318, 319, 374). In D. hyognathus (PI. XVII) the
maximum horn convexity is on the posterior spur of
the nasals, and the same is the case in the very much
more rudimentary horn of Mesatirhinus petersoni. In
another M. manteoceras skull (Am. Mus. 1545) the
swelling and rugosity is shared partly by the nasals.
In this stage of evolution, therefore, the osseous horn
is, strictly speaking, a frontonasal horn. Some of the
more aged specimens (especially Am. Mus. 1569) show
a very faintly rugose condition of the surface of the
bone on these horn bases.
Proportions. — The width of these skulls (see table
of measurements) is increased by the great out-
ward arching of the zygomata posteriorly, the propor-
tions, as presented in Am. Mus. 1569, being, length,
condyles to incisive border, 492 millimeters, width
310. In other words, the zygomatic breadth is nearly
three-fifths of the skull length, whereas in Mesati-
rhinus petersoni the breadth is a little less than one-
half the length, and in Dolichorhinus hyognathus the
breadth is only a little more than one-third the length.
Additional specific characters. — The other chief fea-
tures of the cranium are as follows : (1) The rudimen-
tary frontonasal horn swellings above described; (2)
the widening of the nasals posteriorly; (3) the pos-
terior spreading of the frontoparietal region; (4) the
deep parietofrontal pit between the posterior portion
of the supratemporal ridges, which have now almost
replaced the sagittal crest; (5) the relatively broad,
low occiput.
In many details of structure, enumerated below,
this skull unmistakably exhibits subfamily afHnity
with Mesatirhinus petersoni, yet it differs from that
species in many important features — namely, (1) the
infraorbital ridge is incipient but not prominent, (2)
the zygomata are stout and the zygomatic width of
the skull is much greater than in Mesatirhinus, (3)
the. basioccipital region is relatively broader and less
elongate, and the same is true of the palate.
The sTcull. — The superior view of the skull (fig. 304)
shows several characters which are prophetic of the
Oligocene Brontops:
1. The nasals are slightly expanded at the anterior
extremities, measuring 63 millimeters (Am. Mus.
1569), then contracting slightly to 60 millimeters
and again steadily expanding to 112 millimeters at
366
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
the junction with the frontals. The midlength of the
nasals is 174 millimeters, as compared with 164 in the
much smaller M. megarJiinus skull, showing that while
the cranial portion of the skull has greatly increased
in length, the nasal bones have not increased in length
so fast as they have in MesatirMnus ; in other words,
a retardation in the progressive lengthening of the
nasals is observable and is more strongly expressed
in M. uintensis; and this points toward the transfor-
mation of the Manteoceras skull into the Oligocene
titanothere type. In the contemporary D. Jiyogna-
thus, on the other hand, which does not lead into an
Figure 304. — Type skull of Manteoceras
manteoceras
Top view. About one-fifth natural size. Am. Mus 1569.
Washakie Basin, Wyo., level A ( UMatherium-Manteoceras-
Mesatirhinvs zone). After Osborn, Am. Mus. Nat. Hist^
Bull., vol. 7, flg. 8, 1895. H, horn swelling.
Oligocene titanothere, the nasals attain the extraor-
dinary length of 290 millimeters.
2. The frontal horn caps in the Eocene as in the
Oligocene titanotheres overlap the outer sides of the
nasals, so that the horn bases present upwardly and
outwardly.
3. The orbits (Am. Mus. 1570) are 133 millimeters
apart, and from the prominent triangular postorbital
processes the narrow but distinctly rugose supratem-
poral ridges converge backward into a broad, laterally
expanded frontoparietal plate which prophetically
represents the flattened summit of the Oligocene
titanothere cranium. This plate flares laterally
over the temporal fossae, as in many other species
of titanotheres. Behind this point the supratemporal
ridges converge to form a deep midparietal pit, which
is apparently homologous with the vestigial pit ob-
served in several species of the Oligocene Brontops;
the supratemporal ridges again diverge, leaviag a
narrow groove between the paired sagittal crest,
which is from 18 to 29 millimeters in width.
4. A very important feature of the superior view,
seen also in MesatirMnus, is the comparatively oval
form of the openings left by the zygomatic arches,
and the great backward stretch of the floor of the
temporal fossa from the j miction of the zygomata
with the skull of the occiput.
The palatal view of the skull (fig. 303, B), best seen
in three specimens La the American Museum, Nos.
1545, 2353, 1570, exhibits the following principal
characters: (1) The base of the cranium, the mid-
cranial region (postglenoid to orbit), and the face are
all relatively longer than in Palaeosyops but shorter
than in MesatirMnus; (2) the posterior nares open
about the middle of the skull, opposite the interval
between m2, ms; (3) the hard palate is gently arched;
(4) the posterior narial space is elongate and nar-
rowed by the decided median convexities of the pala-
tines; (5) the pterygoids are greatly reduced as thin,
elongate plates; (6) the relations of basicranial bones
and foramina are as shown in Figure 303, B; (7) there
are paired rugosities on the basisphenoid for attach-
ment of the recti capitis muscles; (8) there is an
elongate bridge (35 mm.) between the foramen ovale
and foramen lacerum medium; (9) the inner portions
of the occipital condyles are borne on the basioccipitals;
(10) the paroccipital process is delicate.
The anterior view of the cranium (fig. 305, B)
clearly illustrates the decurved and thickened margins
of the nasals, the prominence of the frontonasal
horn, the deep lateral facial concavities, the character-
istic structure of the premaxillaries, and the deep
premaxillary symphysis.
In the lateral view (figs. 307, 308) the superior
profile is incipiently saddle-shaped, as in the Oligocene
titanotheres, and we note that the skull descends from
the occiput to the midparietal region, then arches
gently upward to a point directly above the orbits,
and then descends to the tip of the nasals. This
facial convexity, combined with the lateral preorbital
concavities, contributes to and is correlated with the
prominence of the frontonasal horn. The facial
concavity profile is similar to that of MesatirMnus
and is totally different from the transversely convex
preorbital section of Palaeosyops. Possible but doubt-
ful evidence of a progressive shortening of the face
is observed in the variable position of the infraorbital
foramen. In most of the skulls (Am. Mus. 1570, 1511,
1587, 1545) there is a broad bridge of bone over
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
367
the infraorbital foramen, as in M. petersoni; in skull
Am. Mus. 2353, on the other hand, in which the face
appears exceptionally short (a condition possibly
due in part to crushing), this bridge is abbreviated,
the foramen issuing directly in front of the malar-
lacrimal maxUlary bar. The maxillaries contribute
the anterior portion of this bar. The projecting
infraorbital shelf of M. petersoni is absent, being
replaced by a prominent, more or less sharply convex
Figure 305. — Skulls of Manteoceras manteoceras and Palaeo-
syops leidyi
Front view. One-fourth natural size. In Manteoceras (A) the horn region projects
laterally above the facial concavity. In Falaeosyops {,B) the horn region does not
so project and there is no facial concavity.
longitudinal ridge, quite different from the more
rounded suborbital bar of Palaeosyops or the broad,
gentle convexity of T. cultridens and T. ultimum.
Immediately below the orbit the malars are flat, and
behind the orbit they are slightly concave; they give
off the prominent postorbital processes and then
gently arch outward with a convex exterior and a
concave interior surface. The malars thus present
two very striking differences from the Oligocene
type: first, they are much shallower below the orbit
and concave instead of convex externally; second, they
are relatively slender and constricted behind the
postorbital process. The zygomatic portion of the
squamosal exhibits a wide superior as well as a broad
lateral expansion, foreshadowing the decided develop-
ment of this bone in the Oligocene titanotheres.
Dentition in general. — The dentition as compared
with that of MesatirJiinus and DolicJiorhinus agrees in
the following characters: (1) The incisors are short-
FiGUBE 306. — Skulls of Manteoceras manteoceras and' M.
washakiensis
Occipital and front views. One-fourth natural size A, M. manleoceras, from speci-
mens in the American Museum, chiefly No, 1570, La Clede, Washakie Basin,
Wyo., Washakie A. Nasals and occiput restored from No. 1669 (type) , Washakie
Basin, Washakie A; canines and incisors from No. 1511, Biidger Basin, and No.
12678, Hemys Fork Hill, Bridger Basin, Bridger C 5. B, M. manteoceras, Am.
Mus. 1587, Henrys Fork, Bridger Basin, level unknown. C, M. wasfialiensis,
Am. Mus. 13165 (type), base of Haystack Mountain, east end, Washakie Basin,
Washakie A. •
crowned and rounded rather than conical, and the
disparity of i' over i^ is less marked than in Telma-
therium; (2) the sublanceolate canines are broadly
obtuse at the base, taper rapidly at the summits, and
have faint anterior and posterior ridges; (3) the pre-
molars and molars are identical in general structure
368
TITANOTHERES OF ANCIENT WyOMING, DAKOTA, AND NEBRASKA
but are of the mesaticephalic or subbrachycephalic
type. The dentition differs sharply from that of
Mesatirhinus and DoIicTiorMnus in the relatively short
diastema behind the inferior canines and the retarded
Figure 307. — Skull of Manieoceras manieoceras
One-fourth natural size. Am. Mus. 1569 (type) and 1570. Washakie Basin, Wyo., level A
Side view,
view (reversed) of the tj^pe and paratype skulls
development of the deuterocones and tritocones of
the premolars.
Incisors. — The inferior incisors as seen in Am. Mus.
1566 approach the transverse or slightly arched posi-
tion observed in M. petersoni (Am. Mus. 1567) and
in DoIicTiorMnus. The perfectly preserved crown of
i2 indicates that this tooth is slightly larger than ii
and nearly if not quite as large as is; the posterior
face is smooth, with a faintly indicated basal cingulum.
Canines. — The canines of the same jaw in the male
are estimated as 40 millimeters in height, 21 antero-
posterior, 21 transverse; they are implanted by
stout fangs which cause the outer face of the
ramus to bulge; faint anterior and posterior
ridges bound the convex inner face of the
canines; the diastema is much shorter than in
DoIicTiorMnus (17 mm. in No. 1566). The
superior incisors, partly preserved in Am. Mus.
1511, 1545, and fully preserved in Am. Mus.
12683, from Bridger D, are pointed, with a
posterior basal cingulum slightly more promi-
nent than in DolicTiorTiinus; in Am. Mus. 12683
they increase regularly in size from i' to i' and
at first sight resemble those of TelmatTierium
culfridens, but, as shown in Figure 309, they
are distinguished by their slightly smaller size,
weaker posterior cingulum, and less caniniformi^
The superior canines are robust, sublanceolate (that
is, with anterior and posterior ridges), tapering and
recurved; height 39 millimeters, transverse 26, antero-
posterior 24; they are provided with very stout fangs.
The postcanine diastema is very short, not exceeding
7 millimeters.
Premolars. — The superior molar-premolar series,
although entirely broken away in the type specimen,
is superbly shown in five almost complete sets of
teeth. In general, as compared with the grinders of
MesatirMnus and DoIicTiorMnus, we observe the
brachycephalic influence, indicated, first, in the rela-
tive shortness and breadth of each of the teeth, and
second, in the arching or posterior divergence of
the series, which is much
more marked than in Mesa-
tirMnus. The series meas-
ures from 176 millimeters in
Am. Mus. 12683 to 197 in
the old male No. 1545, as
compared with 160 in M.
megarMnus, and is about 20
millimeters greater than in
DolicJiorMnus.
The chief distinctions of
the premolars from those of
MesatirMnus appear to be as
follows (Pis. LXVII, LXXII):
In M. manieoceras p^~* are relatively broader, the
deuterocones are not so wide anteroposteriorly and
lie farther backward, giving a more oblique contour
to the crown; the ectoloph is less hypsodont, its
anterior convexity is much broader, and its posterior
convexity is pronounced. The external cingulum,
while variable, is vestigial in Am. Mus. 1511, 2353,
and slightly indicated in Am. Mus. 1570, 1532. P'
is a simple, elongate tooth, bifanged; the tritocone
ectoloph is convex, and in certain specimens (Am.
Mus. 1511) the tritocone nearly equals the protocone
Composite side
After Osborn, Am. Mus. Nat. Hist. Bull., vol.V, flg. 7, 1895.
FiGUEE 308. — SkuU of Manieoceras manieoceras
One-fourth natural size. Am. Mus. 12683, Sage Creek Spring, Bridger Basin, Wyo., level D.
Ai, Anterior half of skull; A2, cross section through malar and m^, showing flattened external
face of malar.
in size; this tooth, nevertheless, still retains the
ancestral character of elongation and the marked
postero-internal position of the deuterocone. P^ is
slightly broader than long and exhibits various
degrees in the prominence and external convexity of
the tritocone; the external cingulum in the less worn
dentitions is well defined. No rudiment of the postero-
internal cusp can be discerned except in one very
large individual (Am. Mus. 1532), in which very faint
indications of this fourth cusp are seen in p^ and p*.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
369
In p^, a larger tooth, we observe a more constant
enlargement of the tritocone. The cingulum in all
these premolars (except Am. Mus. 1532, the most pro-
gressive specimen) fails to encircle completely the
inner sides of the obtuse deuterocones. Comparative
measurements of the premolars are given below.
The inferior premolars are represented in Am. Mus.
1566, 1563, 2353. Of these. No. 1566 represents the
least progressive stage. The premolar measurements
here given relate to this specimen. Pi (ap. 12 mm.,
tr. 9) is a simple, laterally compressed cone. P2
(21 by 11) has an elevated protocone, slightly hollowed
on its antero-internal border, with a rudimentary
cuspule (=paraconid of molars) and a low postero-
external cusp (=hypoconid of molars). In ps (18 by
12) the cusp analogovis to the metaconid in the molars
is beginning to be constricted off from the protoconid
and is defined by a prominent antero-internal con-
vexity, and the anterior (trigonid) and posterior
(talonid) crescents are beginning to be marked, but
the entoconid, as in all middle Eocene titanotheres,
is not yet developed. In p4 (20 by 14) we have a
submolariform tooth lacking only the cusps analogous
to the paraconid and entoconid but with a well-
developed cusp analogous to the hypoconid.
Molars. — The subquadrate sujJerior molars, which
vary in longitudinal measurement from 100 to 118
millimeters (a wide range of variation), as well dis-
played in five individuals, exhibit (1) progressive ex-
ternal cingula with internal cingula in the valleys; (2)
protoconules vestigial or wanting; (3) a marked
angulation of the postero-internal border of m^
accompanied by a prominent elevatipn of the cingulum
at this point, and in one specimen (Am. Mus. 1511) a
small, distinct hypocone, which, however, may
represent a metaconule. All the molars in the speci-
mens at hand are too much worn to give the height of
the ectoloph, which was undoubtedly elongate in
the unworn condition and which slopes strongly
inward.
Comparative average measurements {ap. by tr.) of molars of
Manteoceras, Mesatirhinus, and Dolichorhinus, in millimeters
Manteoceras
manteoceras
(6 skulls)
Mesatirhinus
petersoni
(4 skulls)
Dolichorhinus
hyognathus
(4 skulls)
Ml
M2
29X31
38X38
38X40
25X26
31X32
32X34
35X35
43X44
M3
45X43
The inferior molars, well preserved in Am. Mus.
1566, 2353, 1563, are uniform in character, measuring
from 111 to 118 millimeters, with an incomplete
external cingulum which fdls the valleys and in No.
1566 rises behind the talonid into a rudimentary
entostylid, or reduplication of the entoconid. Beside
this reduplication we observe an incipient metastylid
or reduplication of the metaconid. The teeth rapidly
increase in length as we pass from mi to ma, the
measurements in No. 2353 being mi 28 millimeters,
m2 35, ma 53. The lower molars show a strong sub-
family resemblance to those of Mesatirhinus but are
distinguished by their somewhat larger size (mi_3 110
mm. in Am. Mus. 1566, as compared with 106 in the
type of M. petersoni) and especially by their greater
breadth (ms (tr.) 23 mm.; in M. petersoni 18). In
Manteoceras also the cutting V's, as seen in crown
view, form more acute angles than in Mesatirhinus.
These differences are still more emphasized by com-
parison with Dolichorhinus hyognathus, which has
long, narrow molars, wide-angled V's, and a relatively
small hypoconulid on ms.
Jaws. — Many fine jaws belonging to this animal
have been collected by the American Museum ex-
peditions in the
Bridger and W a-
shakie Basins. Un-
fortunately only a
few are associated
with the skulls,
namely, Am. Mus.
1545, 1587, 12204,
2353. Another not
associated is Am.
Mus. 1566 (Bridger D).
As shown in the
plates and figures,
the jaws, like the
skulls, teeth, and feet,
show certain charac-
ters prophetic of the
Oligocene titano-
theres. These char-
acters are always
most clearly dis-
played in the aged
forms and are (1)
the depressed lower °^^'S!S^:
border of the angle;
(2) the pit in the an-
terior border of the coronoid behind ms; (3) the
breadth and vertical elevation of the coronoid proc-
ess and sharp recurvature at the summit; (4) the
greater prominence and convexity of the chin.
The coronoid process is more robust and less
strongly recurved than in Telmatherium cultridens;
it is decidedly more elevated and less falciform and
recurved than in Dolichorhinus hyognathus. It is thus
more of the type which we should expect to find as
ancestral to the coronoid processes of the Oligocene
titanotheres.
Figure 309. — Incisors and canines of
Manteoceras manteoceras
Am. Mus. 12683, Sage
Creek Spring, Bridger Basin, Wyo., level D;
supposed male. B, Am. Mus. 12678, Henrys
Fork Hill, Bridger Basin, level C 5; aged female.
370
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The peculiarities of the jaw of M. manteoceras (fig.
310, PI. LI) are as follows:
1. In most of the jaws there is a narrow postcanine
diastema (18 mm. in No. 1566).
3. The powerfully rooted canines involve a sharp
convexity of the bone opposite the alveoli, followed by
a sharp depression in the outer face just below the
postcanine diastema.
4. The chin, as seen from below, is
broad and gently convex anteroposte-
riorly and transversely, but on account of
the great variation in both species it is
difficult to express exactly the differ-
ences in the horizontal rami between
Manteoceras manteoceras and Mesati-
rhinus petersoni.
5. The principal constant difference
is the weaker chin of Mesatirhinus, in
old jaws of which the lower border of
the ramus is straighter than in Manteo-
ceras; the ramus also seems stouter and
deeper and rapidly increases in depth
posteriorly from 54 millimeters behind
Pa to 64 behind m2 and 87 behind ma.
6. In adult jaws the coronoid is of an
entirely different shape from that of
Palaeosyops, being elevated and broad
at the top with the summit gently
recurved, whereas in Palaeosyops it is
pointed and well recurved at the top.
It resembles that of Mesatirhinus but
lacks the posterosuperior prolongation,
and the superior portion of the process
is flattened anteriorly with sharply
angulate anterior external and internal
borders.
7. The angle is produced decidedly
downward and backward, and there is
a rugose outer border in the old males.
8. The posterior border of the jaw
between the angle and the condyle is
somewhat incurved in some specimens
but nearly straight in others.
Measurements of lower jaws of Manteoceras
manteoceras, in millimeters
Figure 310. — Lower jaw of Manteoceras
One-fourth natural size. A, M. manteoceras, Am. Mus. 1566, Bridger Basin, Wyo. level probably C or D.
B, M. manteoceras, Am. Mus. 1563, Bridger Basin, level probably D; very progressive lower jaw (front
part corrected from No. 1560). C, M. uintensist. Am. Mus. 2033 (reversed), White Elver, Uinta Basin,
Utah, Uinta C; doubtfully referred lower jaw.
Pi-mj
Pi-P4
Mi-m3
Condyle to incisive
border
Condyle to bottom of
angle
Depth of ramus below
m3
Length of symphysis-
Am. Mus.
1566; Bridger
C orD
182
73
110
360
165
86
109
Am. Mus.
2353; Washa-
kie A
195
76
117
96
2. The opposite incisor series are placed more trans-
versely than in the premaxillaries above, in which
they are more convergent.
Detailed cTiaracters. — Probably as a sex-
ual character the skull itself is not so large as that of the
largest Bridger specimens, the estimated basilar length
(490 mm.) being less than in the large male M. manteo-
EVOLUTION OF T£E SKULL AND TEETH OP EOCENE TITANOTHERES
371
greater than in M. manteoceras and with complete
internal cingula; superior canine in female (?) obtuse,
recurved, with heavy posterior cingulum; p'-m' 200
millimeters (estimated), p^ with deuterocone somewhat
better developed, p* (ap. by tr.) 23 by 29; m' 35 by 38,
m^ 41 by 43, m' 39 by 42. Face relatively short,
zygomata moderate, horn swelling inconspicuous.
The female type skull (Am. Mus. 13165) of this
species was found by the American Museum expedition
of 1906 at the base of Haystack Mountain, at the
summit of the exposures of level Washakie A. Its
decidedly progressive characters beyond those of M.
manteoceras, from Bridger D and Washakie A, per-
fectly accord with its somewhat higher geologic level.
These are displayed chiefly in the canine, premolar,
ceras (Am. Mus. 1545) from Bridger D, where the
length is 423 millimeters. Similarly, from sexual
causes the horn rudiments and facial concavities are
not pronounced. The posterior sagittal crests are
characteristically bifid, or deeply grooved superiorly,
terminating anteriorly in the parietal pit which is so
distinctive of this species. The occiput is distin-
guished by the very decided prominence of the
occipital pillars.
Dentition. — It is the teeth which afford the most
marked distinctions of this species. The Cannes
(ap. 23 mm., tr. 21) are abbreviate, measuring 24
millimeters in length, the tips being worn off. P^ is
a compressed, conical, bifanged tooth, measuring
(ap. by tr.) 15 by 9 millimeters. P^ (19 by 17)
exhibits marked external
convexities and a weak ex-
ternal cingulum. As there
is considerable variation in
the strength of the cingulum
in M. manteoceras it is un-
certain whether the weak
cingulum is progressive or
not. The deuterocone is
more advanced in develop-
ment than in the average
M. manteoceras. P' (ap. 20
mm., tr. 25) exhibits the
tetartocone fold somewhat
more conspicuously than ;'n
the most progressive
Bridger D specimens. P^
(23 by 29) is progressive in
transverse measurement
and in the development of a
low, barely perceptible te-
tartocone swelling. M' (35
by 38) exhibits a prominent
internal cingulum, which is
almost continuous around
the lingual side of the pro to-
cone. M^ (42 by 43) shows
a strong development of the cingulum (progressive), a and molar teeth. The horns are inconspicuous, prob
Figure 311. — Skulls of Manteoceras manteoceras and M. washakiensis
Side view. One-fourtli natural size. A, if. washakiensis , Am. Mus. 13165 (type); base of Haystack Mountain, east end,
Washakie Basin, Wyo., upper levels of Washakie A. B, M. manteoceras, Am. Mus. 12678; Henrys Fork Hill, Bridger
Basin, Wyo., level C 5; supposed female skull; back of skull slightly raised to correct the vertical crushing.
crenulation of the enamel, and an elongate ectoloph.
M^ (39 by 42) is slightly inferior in size to m^, the
cingulum is most pronounced, and there is a well-
developed hypocone ridge (progressive) but no dis-
tinct hypocone.
Manteoceras .washakiensis Osborn
Plate LXVII; text figures 121, 306, 311, 717
[For original description and type references see p. 182]
Type locality and geologic horizon. — Base of Haystack
Mountain, Washakie Basin, Wyo.; summit of Uinta-
therium- Manteoceras- MesatirMnus zone (Washakie A).
Specific characters. — Skull somewhat larger than in
M. manteoceras (basilar length in supposed female 490
mm., estimated). Superior molars and premolars
ably because the specimen represents a female, and
they are smooth rather than rugose. The small size
of the canines is also a sexual character, but the form of
the canines is very different from that of the specimens
in Bridger D and Washakie A in their approach to
the obtuse form characteristic of the Oligocene
Brontops.
Progressive characters. — (1) Canines short, obtuse,
recurved; (2) internal lobes of p^ and p^ broadening,
with shelf for development of deuterocone; (3) a
tetartocone spur observed in p', as in most progressive
Bridger specimens; (4) very distinct internal cingulum
on m'~''; (5) true molar series relatively longer as
compared with the premolar series than in M. manteo-
ceras, in which an average of six skulls gives the
372
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, , AND NEBRASKA
length pi"* as 76 per cent of that of m'~^ but in M.
washaJciensis p^"* is only 71 per cent; in other words,
the molar series is relatively longer, which might be
expected, as there is a steady general increase in
the relative length of the molar series as we pass
from the lower Eocene Eotitanops to the Oligocene
Brontofherium.
Jaws of Manteoceras washakiensis? from Washakie
A and B. — In the jaw from the base of Washakie A
(Am. Mus. 13176) doubtfully referred to M. wasTia-
Iciensis the dental measurements are all larger than
in the well-preserved jaw of M. manteoceras forming
Am. Mus. 1566 (p2-m3 182 mm. as compared with 168),
and the premolars are distinctly more progressive.
The large jaw from the Bridger (? level D) (Am.
Mus. 1563) referred to M. manteoceras agrees nearly in
size with Am. Mus. 13176.
It is noteworthy that in all the explorations of the
Washakie B and Uinta B levels few or no remains refer-
able to Manteoceras have been found. The single
exception, and this of doubtful character, is the jaw
employed by Cope as the first cotype of his species
" Palaeosyops" vallidens. This specimen (Am. Mus.
5098), consisting of the two incomplete rami, from
Mammoth Buttes, Bitter Creek, regarded as an
upper Washakie level, apparently represents the
Washakie B stage of Manteoceras. (1) The measure-
ment of p2-ni3 (187 mm.) agrees almost precisely with
that of a M. manteoceras jaw (Am. Mus. 2353) as-
sociated with a skuU; (2) the measurements of ms
(ap. 57 mm., tr. 23) indicate a slightly longer tooth
than that in Am. Mus. 2353 (54 by 23). The char-
acters of the teeth and jaws are otherwise the same.
Manteoceras uintensis Douglass
Plate LXIII; text figures 131, 132, 310, 312, 313
[For original description and type references see p. 186]
Type locality and geologic Jiorizon. — About 5 miles
northeast of well 2, Uinta Basin, Utah; gray sandstone
in lower part of Uinta formation {Diplacodon-Pro-
titanotJierium-Epihippus zone, Uinta C). The per-
sistence of Manteoceras in Uinta C is surprising because
that level also furnishes the next higher stage in the
phylum, namely, Protitanotherium. This persistence
is partly explainable by the fact that M. uintensis is
apparently an aberrant side branch with a decided
elongation of the muzzle and to some extent of the
tooth rows, or grinding series; in other words, it gives
certain indications of dolichopy, whereas the main
line {Protitanotherium) is mesaticephalic.
Specific characters.- — SkuU larger than in M. manteo-
ceras (basilar length not known) ; muzzle relatively long,
horn swelling (?) not larger than in M. manteoceras;
zygomata not stout, in inferior view forming a marked
angle in front of the glenoid surface. Postcanine dia-
stema long (28 mm.), postcanine constriction very
marked; superiority in size of i' over i^ appearing less
marked than in M. manteoceras; p'-m^ 240 millimeters;
p^ with deuterocone and tritocone more advanced than
in M. manteoceras; p* (ap. by tr.) 27 by 34; m\ 37 by
38; m', 49 by 52.
That this species should be referred to the genus
Manteoceras is indicated by the round-topped incisors,
the robust, recurved canines, the twin convexities of
the premolar ectolophs, the broad, subhypsodont m',
with large parastyles and mesostyles, and the widely
arched zygomata. From the Bridger and Washakie
Basin species of Manteoceras the present one is dis-
tinguished by its larger size, more dolichocephalic
appearance, shorter free nasals, slightly more progres-
sive premolars, and more pronounced postcanine con-
striction. All these characters serve also to dis-
tinguish M. uintensis from Sthenodectes incisivus
(Douglass), which has very large, "cupped" incisors,
long, lanceolate canines, and very advanced premolars.
Affinities with Protitanotherium emarginatum and P.
superbum are indicated by the general form of the
incisors and canines, short nasals, and broad molars;
but weU-marked differences from these forms are seen
in the more elongate face, the larger size of the
incisors, more pronounced postcanine constriction,
and absence or arrested condition of the horn swelling.
A remote analogy to Dolichorhinus is seen in the
lengthening of the face, of the postcanine diastema, and
of m^, as well as in the broad arching of the incisor
series. These dolichocephalic features in a mesati-
cephalic skull illustrate the subfamily kinship of
Manteoceras with Dolichorhinus.
The skull as a whole must have been large, for the
distance from the premaxillaries to the glenoid region
of the squamosal is given as 430 millimeters, as com-
pared with 335 for the same measurement in M.
manteoceras, an increase of nearly 25 per cent. The
premaxillaries in correlation with the large size of the
incisors are angulate superiorly; the free nasals are
short (85 mm.), a progressive feature; the face is
long (255 mm., as compared with an average of 184
in M. manteoceras). The horn swellings were certainly
not larger than in M. manteoceras — a surprising fact,
because the large canines and incisors indicate male
sex. The infraorbital canal is broad; the zygoma is
deep in its middle portion, but apparently the buccal
swelling was slight or absent. Just back of the orbit
the malar was slender but not so slender as in
M. manteoceras; in inferior view the infraorbital
portion of the malar formed a low ridge which was
somewhat more pronounced than in M. inanteoceras
but did not form a distinct shoulder as it does in
Mesatirhinus and Dolichorhinus. The posterior nares
open opposite the metacone of m^, whereas in M.
manteoceras they sometimes open opposite the para-
cone of m'.
Passing to the dentition, we note that the incisors
are intermediate in form between those of Manteoceras
manteoceras and of Protitanotherium emarginatum,
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHEEES
373
although nearer the former in shape and much larger
than in the latter. The anteroposterior and trans-
verse dimensions of the incisors are, i', 18 by 14
millimeters; i^, 19 by 17; i^ 22 by 20. The incisors
in general resemble those of M. manteoceras in their
rounded front faces and obtusely pointed tips but
differ in the form of the posterior cingulum, which is
now represented by a very large, obtuse basal rim.
The posterior face of i', i^ shows a median vertical
ridge. The posterior face of the large i' is much worn.
The canine is very stout, with widely protruding and
divergent fangs; crown measurements (ap. by tr.),
27 by 20 milhmeters ; it is followed by a large post-
canine diastema (26 mm.), which distinguishes
this species from M. manteoceras, and also by a
marked postcanine constriction.
Premolars. — P' (ap. 22 mm., tr. 11) seems to
be a somewhat more compressed, elongate tooth
than in the preceding species, in which the pos-
terobasal cingulum is produced upward along the
median line into a very low incipient cusp. In
p^ (25 by 21) the deuterocone is still confined to
the postero-internal corner of the tooth but is a
little more advanced than in M. manteoceras; the
tritocone is also slightly more developed but not ^
so much as in either StJienodectes incisivus or Doli-
chorMnus. The external cingulum opposite the
tritocone is rounded. In p^ (24 by 29) the deutero-
cone has a rather backward appearance as com-
pared with the same cusp in M. washakiensis, but
the internal cingulum is pronounced, although still
incomplete opposite the middle of the deutero-
cone; the protocone is still much larger than the
tritocone; the "parastyle" is pronoupced; the ecto-
loph convexities are marked; the external cingulum
is pronounced except where it "festoons " the proto-
cone convexity. In p^ (28 by 33) the deuterocone
seems less robust than in M. wasJialciensis; the cin-
gulum is robust but does not surround the deu-
terocone; the "parastyle" is prominent, and the
protocone convexity broad at the base; the trito-
cone is at least no bigger than in M. wasJialciensis; the
external cingulum is a broad, rounded ridge opposite
the tritocone.
Molars. — The molars (m^-m^, 138 mm.) are some-
what larger than those of M. wasJialciensis but other-
wise agree fairly well. The external cingulum is very
pronounced opposite the valleys. The molars show a
marked asymmetry of the external V's, the anterior
V being more widely open and the posterior having a
short posterior limb. The antero-internal cingula are
heavy but not complete around the inner sides of thepro-
tocones (contrast M. wasJialciensis). In m^ the postero-
internal corner is less angulate than in M. wasJiaJciensis.
The anteroposterior and transverse dimensions are,
m\ 37 by 38 millimeters; m^ 48 by 48; m', 48 by 52.
Other measurements of M. uintensis are given in the
table above.
Manteoceras uintensis?
From the base of Uinta C in 1895 Peterson obtained
a very large skull, apparently female (Am. Mus. 2029),
which unfortunately is too aged as well as. too much
crushed and imperfect to afford distinctive characters
for definition. It is somewhat smaller than the type of
M. uintensis. If more complete it might be found to
Figure 312. — Type skull of Manteoceras uintensis
One-fourth natural size. Carnegie Mus. 2388. "About 5 miles northeast of well 2,
Uinta Basin; from gray sandstone in red Uinta beds, lower portion of horizon C."
Ai, Side view, nasal region crushed; A2, palatal view, crushed laterally.
represent a female of M. uintensis (with which it
agrees in the retarded evolution of the premolars) or a
species transitional between Manteoceras wasJiaJciensis
and a higher stage of evolution. Size or metatrophic
characters are truly progressive and undoubtedly
correlated with other characters of distinct specific
value. Its total length, condyles to incisive border,
is about 555 millimeters, as compared with 523 in the
largest skull of M. manteoceras found in Bridger D.
The grinding teeth, p^-m^, measure 227 millimeters,
as compared with 203 in the largest of the specimens
of M. manteoceras from Bridger D and 240 in M.
uintensis. Its progressive zygomatic brachycephaly
is indicated by the widely arching zygomata, which
attain a transverse width of about 360 millimeters.
374
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
the relations of width to length thus being width 360,
length 555. The reference to Manteoceras is con-
firmed by the obliquely flattened form of the infra-
orbital portion of the malars, a character by which
this animal may readily be distinguished from the con-
FiGURE 313. — Upper canines and incisors of Manteo-
ceras uintensis
One-haU natural size. Carnegie Mus. 2388 (type). "About 5 miles
northeast of well 2, Uinta Basin; from gray sandstone in red Uinta
beds, lower portion of horizon C."
ttemporary Telmatherium ultimum,
though there is no infraorbital shelf.
The proportions of this part of the
skull indicate an elongation of the
facial region (as in M. uintensis), a
feature certainly not charateristic of
the line leading to Protitanoiherium.
The animal is undoubtedly a fe-
male, as indicated by the relatively
small canine teeth.
A lower jaw (Am. Mus. 2033, fig.
310, C) found in the same region
and at first referred to Telmatherium
ultimum (cf. above) agrees in size
with this specimen. It also belongs
to a female animal and is of the
mesaticephalic rather than dolicho-
cephalic or brachycephalic type. The canines are
relatively small and recurved. M3 is a relatively
short (63 mm.) rather than elongate tooth as in
Telmatherium ultimum (76 mm.).
Protitanotherium
General characters. — In their phylogeny the animals
known as Protitanotherium are among the most
interesting of the titanotheres that lived in the Uinta
Basin, because of their evident relationship to Man-
teoceras on the ancestral side and to Brontops and
other Oligocene titanotheres on the descendant side.
The profile figure of the horn region of Manteoceras
manteoceras, P. emarginatum, two young skulls of the
Oligocene Brontops hrachycephalus, Allops marshi,
and Brontops rohustus illustrates the resemblance
(fig. 712).
These upper Eocene animals are robust, massive,
vigorous, and well protected both by their powerful
tusks and by their rapidly developing horns, which
are far more prominent than those of any of the earlier
Eocene titanotheres. Therefore the suggestion of
Hatcher (1895.1, p. 1084) that this animal should be
called Protitanotherium was eminently appropriate.
Progressive characters. — The phyletic increase in
size of the animals of this series is best illustrated by
comparison of the adult jaws of M. manteoceras, P
emarginatum, and P. superbum (figs. 310, 315, 321).
The preservation of the ancestral phyletic character
in these jaws is certainly very striking. The change
is chiefly proportional, or quantitative. The differen-
tial or generic distinctions are to be found especially
in the teeth and in the horns.
Specific stages. — These protitanotheres are known
to include two stages — (1) P. emarginatum Hatcher,
originally described as Diplacodon emarginatum, found
'y
Figure 314. — Restoration of Protitanotherium emarginatum
By Charles E. Knight. About one-ninth natural size.
near the base of Uinta C, or the "Diplacodon beds"
of Marsh, and distinguished from the following stage
chiefly by its inferior size and more brachyodont
teeth, and (2) P. superbum Osborn, a much larger
animal, whose remains were probably found at a
higher geologic level.
Horns. — The bases of the horns preserve the Eocene
anteroposterior elongation. This elongate oval form
is in marked contrast to the obliquely oval, triangular,
or transversely elongate form of adult Oligocene
titanotheres. In very young Oligocene titanotheres,
however, the resemblance in the elongate oval horn
swellings to those of Protitanotherium is very noticeable
(fig. 374).
Size. — In the species P. superbum these animals
attain a size considerably surpassing that of many
of the smaller forms in the lower Oligocene levels.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
375
Protitanotherium Hatcher
Plates LXVIII, LXIX; text figures 24, 29, 87, 112, 128, 314^321,
371, 374, 375, 408, 409, 594-596, 647-649, 701, 712
[For original description and type references see p. 176. For skeletal characters
see p. 656]
Generic characters. — Horns relatively large, elliptical
LQ section, with anteroposterior diameter greatly
Comparison. — We at first note the incipient loss of
the piercing function of the incisor teeth. The lateral
superior incisors are still large, but the median incisors
are reduced. This may indicate that these animals
were given to browsing and that the tongue and lips
were increasingly used for the prehension of food, while
the incisors became functionless and gradually aborted.
Figure 315. — Lower jaws of Protitanotherium and Brachydiastematherium
le-fourth natural size. A, P. emaTginaium, Princeton Mus. 11242 (type); Uinta Basin, Utah, Uinta C; region of angle and ma supplied from
Am. Mus. 2028. B, P. superbum, Am. Mus. 2501 (type), reversed; Uinta C. C, B. tTansilvamcum, front type of lower jaw; upper Eocene (?) of
Andrashiza, Transylvania.
exceeding the transverse. The incisor series numeri-
cally typical, -1; canines relatively large, robust and
recurved; p^ submolariform but without entoconid;
P3, p2 transitional.
Materials. — These animals are known from three
specimens referred to P. emarginatum and three referred
to P. superbum. The lower grinding teeth are fully
known, but the upper grinding teeth are only partly
known.
101959— 29— VOL 1 27
We observe in comparison with Manteoceras that the
nasals have taken on the broad, quadrate character
which distinguishes the nasals of certain of the lower
Oligocene titanotheres, such as Brontops and Menodus.
The horns are intermediate in evolution between those
of Manteoceras and of Brontops. The incisor teeth
still retain the proportions observed in Manteoceras,
but the median upper incisor is acquiring the rounded
form characteristic of most OUgocene titanotheres.
376
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
FiGnRE 316. — Type skull of Protitanotherium emarginatum
Less than one-fourth natural size. Princeton Mus. 11242, Kennedys Hole, 8 miles north of White Elver and 25 miles east of Ouray Agency, Uinta Basin, Utah;
Uinta C. Provisional reconstruction of skull and lower jaw, front part directly from the type.
Figure 317. — Type skull of Protitanotherium emarginatum
One-fourth natural size. Princeton Mus. 11242, Kennedys Hole, 8 miles north of White Elver and 26 miles east of Ouray Agency,
Uinta Basin, Utah; Uinta C. Ai, Side view of front part of skull; Aj, front view of front part of skull and lower jaw.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
377
The canines are intermediate in form between the
Manteoceras and Brontops types, perhaps nearer
Brontops, and were evidently robust fighting weapons.
Unfortunately the form of the superior and posterior
parts of the skull is still unknown.
D
Figure 318. — Nasal region in three specimens of Protilanotherium
Top view. One-fourth natural size. A, X), P. emarginatum, Princeton Mus. 11242 (type); B
iuperbumf, Carnegie Mus. 2855; C, P. superbumf, Princeton Mus. 11213.
Protitanotherium emarginatum Hatcher
[Diplacodon emarginatus Hatcher, p. 177]
Plates LXVni, LXIX; text figures 24, 29, 87, 112, 314-320,
374, 375, 408, 648, 649, 712
Type locality and geologic horizon. — Kennedys Hole,
8 miles north of White River and 25 miles east of Ouray
Agency, Uinta Basin, Utah; Uinta formation {Dipla-
codon-Protitanotherium-EpiMppus zone, Uinta C).
Specific characters. — Pi-ms 294 millimeters (esti-
mated), pi with a small talonid; ii small, round-topped;
is large, bluntly pointed ; ia much larger than is. Lower
canine more erect, recui'ved, and abruptly swelling at
the base. Postcanine diastema short (27 mm.) but
relatively longer than in P. superbum. Lower pre-
molars and molars more brachyodont, with sloping
curves.
P. emarginatum appears to be closely allied to P.
superbum but is apparently a lower stage, distinguished
by its smaller size, more brachyodont premolars and
molars, and more erect recurved lower canines, which
swell more rapidly at the base.
Materials. — This species is represented by the
type (see below), by the finely preserved anterior
portions of a jaw in the Yale Museum
(No. 635 D), and by a fairly well preserved
jaw in the American Museum (No. 2028).
The jaw in the Yale Museum is especially
valuable because it includes, besides the
incisors, canines, and portions of the pre-
molars, a complete ma, a tooth which is
imperfect in the type.
Skull. — The type of the present species,
discovered by Hatcher himself, is the an-
terior portion of a skull and lower jaw
(Princeton Mus. 11242). Hatcher noted
the greater size of this animal than Dipla-
codon elatus as indicated by the length of
the grinding teeth (294 mm., as compared
with 244). So far as preserved the dorsal
surface of the skull is concave anteroposte-
riorly and suggests the broad, fiat frontal
region and flattened parietal vertex of the
Oligocene forms. The nasal openings are
high and deeply incised. The horns are
composed of the frontals overlapping the
nasals; they are placed longitudinally and
directed upward, outward, and forward.
The nasals are broad, strong, and rather
short, firmly coossified, concave inferiorly;
they measure in free length 107 millimeters
and in greatest breadth 123. The specific
name P. emarginatum, refers to the fact that
the nasals are "emarginate anteriorly," but
they are not deeply indented in the
midline, as in a larger pair of nasals also
described by Hatcher (fig. 318) which may
be referred provisionally to P. superbum.
The premaxillaries are well developed and separated
anteriorly by a deep median
notch, below which they are
firmly coossified. The maxilla-
ries are expanded at the base of
the canines and decidedly con-
stricted between these teeth and
p'. The postcanine diastema
measures 37 millimeters, and
back of this the maxillaries ex-
pand rapidly in order to ac-
commodate the large posterior
premolars and molars. The in-
fraorbital foramen was probably
situated just above p*, as in
Palaeosyops, Limnohyops, Tel-
matherium, and Manteoceras. In Mesatirhinus and
Dolichorhinus it is more nearly above m^
Figure 319. — Sections
of the nasals and
horns of Protitano-
therium einar gi-
natum
Princeton Mus. 11242 (type).
One-seventh natural size.
378
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Dentition. — The statements made below refer to the
type specimen unless otherwise indicated. The su-
perior incisors in general are of the M. manteoceras
type but exhibit several differences, which, on the
whole, are progressive toward the Oligocene titano-
theres of the genus Brontops. The median pair of
superior incisors (i^) are distinctly the smallest of the
series and are assuming the degenerate conical form
seen in the Oligocene species of Brontops,'^'' a change
that is effected especially by the reduction of the
posterior cingulum and the rounding together of the
anterior and posterior faces, with the consequent
loss of prehensile or cropping function. The inferior
series are all still pointed, jDosteriorly cingulate, and
functionally of the cropping type; 12 is the largest of
the series and ia is the most reduced. The superior
incisors are nearly in a transverse line, but i^ is slightly
behind i^. The incisors are separated from the
terior faces, and with low posterior U-shaped basal
cingula.
The median incisors (ii) are somewhat compressed
laterally (height 15 mm., ap. 14) but exceed in size
the lateral pair. The second incisors (i2) are much
larger (height 18, ap. 18) and the lateral incisors (ia)
are the smallest (height 12, ap. 12, tr. 12) and the
simplest, although still incisiform.
The canines are distinguished by robust, rounded
fangs, which diminish rapidly into forward-directed,
tapering, pointed crowns, again suggesting those of
M. manteoceras, especially by the posterior swelling at
the base and the sublanceolate, internally flattened
apex with faint anterior and posterior ridges. In the
superior canines the diameters at the base are trans-
verse 26 millimeters, anteroposterior 28; height 49.
The inferior canines exhibit much the same form (ap.
32 mm., tr. 28, height 52), the base of the crown sloping
Figure 320. — Lower jaw of Protitanotherium emarginatum
One-fourth natural size. Partial reconstruction of anterior part of jaw to ms from Princeton Mus. 11242 (type), Kennedys Hole, 8 miles
north of White River and 25 miles east of Oiu-ay Agency, Uinta Basin, Utah, Uinta C; ms and posterior part of jaw from Am. Mus.
2028, White River, Utah, Uinta C.
canines at the sides by a narrow diastema (9 mm.)
and separated in the median line by a diastema (10
mm.) similar to that observed in Mesatirhinus mega-
rhinus. The lateral incisors (i^) are large pointed teeth
(height 21 mm. side, 17 front; ap. 17) with oblique
posterolateral basal cingula and rather sharp lateral
cutting edges. The second incisors (i^) are much
smaller (height 12 mm., ap. 14) with heavy posterior
basal cingula connected with the apex of the tooth
by a median ridge which divides the somewhat concave
posterior surface. The median incisors (i') are still
smaller (height 10 mm., ap. 11), with subspherical
crowns and posterior median ridges rising to unite
with the apex of the crown and obscure the cingulum
except on the posterolateral side.
The transverse extent of the inferior incisors is 69
millimeters; they aU exhibit pointed crowns, with
uniformly convex anterior faces, more concave pos-
" In the Oligocene genera probably the median pair of superior incisors (iO and
he lateral pair of inferior incisors (ij) had disappeared, leaving i>, i' and ii, ia (see
p. 448).
backward into a deep, powerfully implanted fang, the
crown diminishing rapidly as it rises to a rounded, sub-
lanceolate apex.
The premolar-molar series are separated by a short
diastema from the canines in both jaws (27 to 30 mm.
above, 23 below). A postcanine diastema of varying
length is seen in Telmatherium ultimum, T. altidens,
Manteoceras uintensis, Diplacodon elatus, and Protitano-
therium superhum, an indication that it is independ-
ently preserved in different phyla. Of the superior
teeth unfortunately p^ only is preserved. This is a
bifanged tooth measuring (ap. by tr.) 20 by 11 milli-
meters, with a simple protocone, a sessile or rudimen-
tary posterior heel, and a postero-internal cingulum
anci concavity. The alveolus of p" shows that it was
a broader and much more advanced tooth than that
in Manteoceras manteoceras. The inferior grinding
teeth measure about 294 millimeters in length, as
compared with 180 to 192 in M. manteoceras.
In the type the lower premolar series measures 103
millimeters on the left side, in which pi is abnormal
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
379
There is a marked asymmetry of this tooth on the two
sides of the jaw, the right tooth being much smaller
than the left and of a form normal in upper Eocene
titanotheres. The abnormal pi is much swollen, a
tendency seen also in the other cheek teeth and even
more emphasized in the Oligocene titanotheres. The
normal pi (ap. 16 mm., tr. 11) is compressed, sub-
conic, with a posteriorbasal lobe that is better devel-
oped than in earlier forms. P2 (ap. 25 mm., tr. 16) is in
about the same stage of evolution as in P. superhum
but is more brachyodont; it exhibits a protoconid
relatively much more depressed than in M. manteoceras,
while the postero-external cusp (hypoconid) is rela-
tively more elevated, and a rudimentary internal cus-
pule ( = paraconid) begins to appear. This is therefore
a much more progressive tooth than the po of M.
manteoceras but is clearly derivable from it. In pg,
though the anterior crescent ( = trigonid) remains
larger than the posterior crescent ( = talonid), the
measurements (ap. by tr.) being 27 by 17 millimeters,
the antero-internal cusp begins to be well defined; no
median internal cusp corresponding to the metaconid
of the molars appears, the tooth being less progressive
in this respect than in P. leidyi but derivable from the
conditions observed in ps of M. manteoceras. P4 dif-
fers from the true molars chiefly in its smaller size (ap.
30 mm., tr. 21), in the somewhat greater elevation of
the anterior lobe ( = trigonid), and in the absence of a
distinct postero-internal cusp ( = entoconid) . The
external cingula are obsolete on p2, ps and much
reduced on P4.
The inferior molars of the type are characterized by
shallow internal valleys and an incomplete external
cingulum, which is inflected in the valleys, by a well-
defined posterior cingulum, by prominent paraconid
and metaconid, by a considerable elevation (hypo-
conid) of the crown (26 mm.) in m2. Metaconid folds
are present as in many other titanotheres. The meas-
urements (ap. by tr.) are mi, 46 by 26 millimeters;
m2, 57 by 34 (estimated) . M3 is incomplete in the type ;
in another specimen (Am. Mus. 2028) it measures
78 millimeters, and in a third specimen (Yale Mus.
635 D) 79. In Protitanotherium superhum this tooth
measures 98 millimeters.
Jaw of Protitanotlierimn emarginatum. — The type
jaw of P. emarginatum. (Princeton Mus. 11242) ex-
hibits the anterior half of the ramus and symphysis
The second jaw (Am. Mus. 2028) preserves the pos-
terior half but lacks the coronoid and condylar proc-
esses. The third jaw (Yale Mus. 635 D) includes
the symphyseal portion only.
The type jaw exhibits a very massive symphysis,
159 millimeters in length, 80 millimeters across the
narrowest portion of the chin below, with the charac-
teristic postcanine constriction seen in M. manteoceras.
The ramus is thickened (35 mm.) below mi and in-
creases very rapidly in depth from 92 millimeters
behind ps to 126 behind m2. The progressive increase
of the ramus in depth posteriorly is evident in this
series as well as in Telmatherium ultimum, T. altidens,
Palaeosyops, etc. It is more pronounced in forms with
relatively large molars. The mental foramen of the
type jaw is single and placed directly below the
posterior fang of p2.
In the American Museum jaw of P. emarginatum
(No. 2028) the depth is 90+ millimeters behind pa^
108 behind m,, and 144 behind ms; the distance from
the back of ms to the posterior border of the angle is
186 millimeters. The lower border is crushed, the
distortion concealing its natural contour, but there
appears to be a slight up curve below the coronoid,
with a broad downward and backward sweep of the
angle, and the posterior border of the angle appears
to rise to the condyle with a slight incurvature.
Protitanotherium superhum Osborn
Plate LXIX; text figures 128, 315, 318, 321, 371, 593, 6i7
701
For original description and typs references see p. 185. For skeletal characters
see p. 655]
Type locality and geologic horizon. — White River,
Uinta Basin, Utah; Uinta formation (Diplacodon-
Protitanotherium-EpiJiippus zone, Uinta C, probably
higher levels).
Specific cTiaracters. — Very large (pi-ms, 318 mm.).
Lower canines in males very robust, relatively nearer
to the midline than in P. emarginatum; pi distinctly
double-fanged; postcanine diastema abbreviated
(about 30 mm.); premolars in about the same stage of
complication as in P. emarginatum, but premolar and
molar cusps more steeply sided (that is, more hypso-
dont), p4 submolariform, ps, p2 transitional; external
cingulum on ps, p4 a little clearer; true molars very
large (mj-ms, 210 mm.); ms with hypoconulid sharply
constricted at base.
Materials. — This species is at present known from
the type jaw (Am. Mus. 2501), discovered by Peterson
in 1895. There are also two upper molar teeth re-
corded as belonging to the same individual, as well
as a pair of nasals in the Princeton collection (No.
11213). These nasals (fig. 318) are distinguished
from those of P. emarginatum by their superior size
and by a deep incision in the median line anteriorly.
Comparisons. — The relative measurements of P.
superium, P. emarginatum, and Teleodus avus are
shown below.
Measurements of Proiitanotherium and Teleodus, in millimeters
P. emargina-
tum (upper
Eocene)
P. superbum
(upper
Eocene)
T. avus
(lower
Oligocene)
P,-mj - --- - --
■294
-187
"■99
27
78
318
210
105
31
99
Mi-m3 - -
Pi-p, _
(p!-pO 106
Transverse posterior lobe
M. . --
99
380
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
P. superhum is distinctly of the Manteoceras and
Protitanoiherium phylum. It is, however, more pro-
gressive toward the Brontops stage in that it is larger,
that its canines are extremely robust, and that p2_4 are
a little longer but mi_3much longer than in P. emargi-
natum. We observe that Teleodus (Brontops) avus of
the Oligocene has grinding teeth similar in size to
those of P. superhum.
Inferior dentition. — The canines and grinding teeth
are preserved in the type jaw, but the incisor series is
wanting. They represent a specific progression upon
those of P. emarginatum in their greater size; in the
abbreviation of the postcanine diastema; in the
lengthening and broadening of the true molars, a
progressive tendency that may be followed from
Eotitanops through Manteoceras and Protitanotherium up
into the Oligocene titanotheres ; in the arrested length
of the premolar series; and in the more pronounced
hypsodonty of the cusps of the premolars and molars.
Considered in detail, the inferior canines are seen to
be relatively closer together than those of either 31.
manteoceras or P. emarginatum, which are presumably
near the ancestral types of P. superhum; but they
retain the characteristically robust fangs and the
stout recurved crowns with gen er all 3^ rounded section,
faint anterior and posterior cingula. The basal crown
measurements of the canines are anteroposterior 32
millimeters, transverse 31. The postcanine diastema
is relatively reduced and uneven on the two sides — ■
namely, 30 millimeters on the left, 20 (estimated) on
the right.
The premolar series (PI. LXIX) is closely continuous,
measuring 105 millimeters. Pi of the left side was
apparently somewhat smaller than pi of the right.
Nothing remains of the crown of this tooth except the
laterally compressed subsecant talonid. P2 (ap. 28
mm., tr. 16) is distinguished from that of M. manteo-
ceras by the deepening of the crescents and the more
decided accent of the internal cusps, which are analo-
gous to the paraconid, metaconid, metastylid, and
entoconid on the molars. P3 (ap. 28 mm., tr. 18) is a
slightly longer and decidedly broader tooth than in
P. emarginatum; it is little if any more progressive
in the development of the cusps analogous to the
paraconid, metaconid, metastylid, and entoconid of the
molars, which, being less worn down than in the type
of P. emarginatum, produce at first the impression that
they mark a higher stage of evolution, but the diJSfer-
ences seem to be due largely to difference in wear;
however, the entoconid of the right side only is some-
what better developed than in P. emarginatum. P4
(ap. 33 mm., tr. 21) is still more decidedly molari-
form, the median internal cusp ( = metaconid) being
much more prominent than in either of the preceding
teeth, a feature foreshadowed in M. manteoceras.
The true molars are decidedly longer and broader
than those of P. emarginatum. This progression in
size and especially in width points toward brachy-
cephaly. There is a somewhat stronger accent of the
paraconids than in P. emarginatum, but this may result
from the greater wear in the type of that species, which
would depress the metaconids and entoconids nearer
to the level of the paraconid. The metastylid ridges
do not appear very prominent. The external cingu-
lum tends to festoon the sides of the protoconids and
hypoconids in a faint line, whereas in P. emarginatum
the cingulum is straighter and is thus (by comparison
with other titanotheres) seemingly more advanced
than in P. superhum. The outer surface of the ecto-
loph is entirely smooth on the median portion of the
lobes — that is, the cingulum has disappeared. The
measurements, in millimeters, are as follows: Mi, ap.,
52; tr., 27 through trigonid, 30 through talonid. M2,
ap., 63; tr., 35 through trigonid, 36 through talonid.
M3, ap., 95 (estimated); tr., 40 through trigonid,
38 through talonid.
Superior molars. — The second and third superior
molars of the same individual (PL LXIX) are almost
certainly associated with the type lower jaw; they have
the proper dimensions to fit the lower molars, and in
this jaw, as in others, they show more dentine than the
lower teeth; they are quadrate in form, m- measuring
57 by 57 millimeters and m^ 62 by 62. They exhibit
imperfectly developed external cingula and an internal
cingulum , which faintly festoons the inner cusps of the
crown. The features of m^ are the complete wearing
out of the prefossette, the somewhat median position
of the protocone, and the somewhat detached and
anteroposteriorly compressed hypocone; it is note-
worthy that, as in other titanotheres, while the ex-
ternal crescents are extremely worn the internal cones
are very slightly worn, the protocone barely exhibit-
ing exposure of the dentine at its apex. M^ shows the
bottom of the prefossette, a very large protocone, and
an angulate hypocone region, in which, however, there
is a sessile cingulum but no rudiment of the hypocone
proper.
Comparison of teeth of Protitanotherium with those
of other genera. — Protitanotherium is at once distin-
guished from Teleodus avus of the lower Oligocene by
the character of the incisors, which in Protitanotherium
are large and more or less flat-topped and in Teleodus
smaller, with ovoid or hemispherical tops. The canines
of Protitanotherium are more robust, especially at the
base; those of Teleodus are more slender, erect, and
evenly sloping. The premolars of Protitanotherium
are in a lower stage of evolution — that is, they are less
molariform than those of Teleodus and Brontops.
The jaw. — The dimensions of the jaw of P. superhum
(type) considerably surpass those of Brontops hrachy-
cephalus, from the lower Titanotherium zone, the length
from condyle to incisive border being 580 millimeters
and the depth of the angle below the condyle 270, as
against 490 and 220, respectively, in a small B.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
381
brachycepJialus (Am. Mus. 1495). The general contour
of the jaw in these two species, however, is somewhat
similar; there is the same long, slightly convex chin;
the lower border in profile is convex below the grinders,
slightly concave below the coronoid, and is produced
downward and backward into the angle, a concave
border rising from the angle to the condyle; and the
stout, somewhat recurved coronoid processes have a
heavy anterior and gently convex anterior border.
This jaw, as compared with earlier forms, certainly
resembles in its main features and proportions those
vertically oval extension for articulation with the
postglenoid process. The broadening and downward
extension of this postglenoid facet on the inner side
of the condyle is a striking progressive feature, which
was probably acquired by all late Eocene and lower
Oligocene titanotheres. The anterior border of the
rotula extends nearly straight across, as in Manteo-
ceras and probably also as in other phyla. The
posterior border of the jaw arches gently forward below
the condyle and then suddenly expands backward into
the downward and backward produced angle.
Figure 321. — Lower jaw of ProUlanotherium superbum
One-fourth natural size. Am. Mus. 2501 (type), reversed; White River, Uinta Basin, Utah; Uinta C. The coronoid is somewhat altered by
crushing.
of P. emarginatum, Manteoceras manteoceras, and Mesa-
tirMnus megarhinus.
The symphysis is very massive, extending 158
millimeters anteroposteriorly, with 100 millimeters as
the least transverse measurement of the chin; it is
gently convex anteroposteriorly and somewhat more
decidedly convex transversely; the postcanine con-
striction is relatively less decided than in M. manteo-
ceras; the jaw increases rapidly in depth from 109
millimeters behind ps to 124 behind m2 and 163
behind m^. In view of the relatively short diastema
behind the canine and the rather rapid rise of the
coronoid border behind ms, it appears that this jaw
is progressively shortening and deepening. The
coronoid is stout, gently recurved, and relatively less
elevated above the postcoronoid border than in
M. manteoceras or Brontops hracJiycephalus.
The condyle is greatly extended transversely
(106 mm.) and narrow anteroposteriorly (24 mm.),
therefore exhibiting a rather sharply convex rotula,
except on the inner side, where it exhibits a deep,
Measurements of type loioer jaws of ProUlanotherium and
Brachydiastematheriiim, in millimeters
Is, anteroposterior
I3, transverse
C, maximum anteroposterior
diameter
C, maximum transverse diameter.
C, lieight of crown
Postcanine diastema, maximum. _
Postcanine diastema, minimum. _
Pi-p4, anteroposterior
Pi, anteroposterior
P2, anteroposterior
P2, transverse *
Pj, anteroposterior
P3, transverse '
Tt, anteroposterior
P), transverse '
Ml, anteroposterior
Ml, transverse ^
B. transil-
p. emar-
vanicum
ginatum
22
12
20
12
38
32
31
28
-40
52
12
35
3 +
25
107
"99
18
16
26
25
"17
16
31
27
'■22
17
38
30
-28
21
50
46
»30
26
105
« Estimated. ^ Transverse measurements are made through anterior lobe.
382
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
TRANSITIONAL TITANOTHERES IN THE EOCENE OF EUROPE
Brachydiastematherium Biickh and Maty
Plate LXX; text figures 100, 315
[For original description and type references see p. 166]
Geologic Jiorizon. — Recorded as lower Eocene, but
more probably middle Eocene (Abel), upper Eocene
(Osborn), or even lower Oligocene (Stehlin).
Generic and specific cliaracters. — Brachycephalic.
Size large, about that of Protitanoiherium superium.
Dentition: ly, Cx, P4, M^. Lower incisors large)
heavily cingulate posteriorly, is larger than 12; post-
canine diastema very short; canines stout, with short,
heavily cingulate crowns; pi compressed, P2-P4 pro-
gressive, submolarif orm ; premolar entoconids rela-
tively well developed.
Brachydiastematherium transilvanicum Bockh and Maty
Materials. — The type species, B. transilvanicum, horn
Andrashaza, in Transylvania, Hungary, is represented
by a fragmentary lower jaw. As shown by every
detail of the dentition this animal was unquestionably
a titanothere, much resembling Protitanotherium, and
not, as its describers supposed, a relative of Palaeo-
therium.
Geologic age. — The age was originally recorded as
lower Eocene, but it is probably upper Eocene. The
question of the geologic age of this specimen is one of
the most important in the chronology of the titano-
theres. Pavay, its discoverer, as well as Bockh, its
describer, and A. Koch, who studied the strata in
which the type was found, assigned a lower Eocene
age to the species, and their opinion was accepted by
Dep^ret, who placed the genus among the upper
Ypr^sien fauna, which is correlated by Osborn with
the lower part of the middle Eocene Bridger of North
America (Bridger B), which contains the relatively
primitive Palaeosyops pdludosus.
This animal was found in the same beds as Prohyra-
codon orientale Koch. In his monograph "Die Sauge-
thiere des schweizerischen Eocans" Stehlin (1903.1, p.
125, note) remarks: " Ich hege indess einen starken Ver-
dacht, dieser Prohyracodon mochte, wie Koch selbst
friiher annahm, oligocanen und nicht mitteleocanen
Alters sein." Schlosser (1901.1, p. 27) points out
that Prohyracodon is not, as Koch had believed, a
forerunner of Hyracodon but is closely related to the
aceratheres (hornless rhinoceroses). Abel (1910.1,
p. 24) appears to be doubtful as to the geologic age of
Prohyracodon but considers it the most primitive of
the European rhinoceroses, more primitive than
Meninatherium, which is Aquitanian. From Abel's
figure of Prohyracodon Matthew (letter, 1914) infers
that it is of lower Oligocene or at most of upper Eocene
age. It is closely allied, as Schlosser and Abel state,
to the earlier Oligocene aceratheres, and it is much
more progressive than Hyrachyus or Amynodon.
Brachydiastematherium should therefore be considered
of lower Oligocene or upper Eocene age.
Characters. — Every detail of the dentition shows
that, as compared with the American titanotheres, B.
transilvanicum is in an upper Eocene stage, closely simi-
lar to that of animals found in horizon C of the Uinta
Basin. The indications are that titanotheres migrated
from some northern center at about the same time
into eastern Europe and into North America.
Brachydiastematherium agrees with the upper Eocene
Protitanotherium superbum in general appearance and
in the dimensions of pi to mi. The canines have a
peculiar very heavy curved internal posterior cin-
gulum ridge, and the crown seems shorter and more
recurved than in P. superbum. All the premolars
appear to be in a slightly higher stage of evolution,
and the pronounced external cingula are horizontal
rather than festooned. The postcanine diastema is
much shorter, an indication of a higher specialization.
From P. emarginatum this species differs not only in
the more advanced evolution stage of the premolars
and shape of the canine but apparently also in the fact
that is is the largest of the series, if Bockh and Maty's
identification of this tooth as is is correct. The in-
cisors are also larger and more heavily cingulate
posteriorly.
Brachydiastematherium differs from Telmaiherium
altidens especially in the obtuse shape of the canines
and in the much more progressive ■pi-p4. The large
size of the heavily cingulate incisors suggests, however,
the possibility that the genus under consideration may
rather be an ofi^shoot of the Telmatherium than of the
Manteoceras-Protitanotherium series. From Stheno-
dectes incisivus, which it resembles in having three very
large incisors on each side of the jaw, Brachydiaste-
matherium differs in the much more progressive
evolution stage of the inferior premolars, p2-p4-
Derivation from Sthenodectes or from a nearly related
form is suggested by the general appearance of the
incisors, canines, and grinding teeth.
The incisors of B. transilvanicum, with their broad
i posterior cingula, parallel those of the Brontotherium
type among the American Oligocene phyla.
Brachydiastematherium is completely transitional
between the Eocene and Oligocene titanotheres, in
so far as it retains six large lower incisors of Eocene
type in company with very progressive lower pre-
molars of Oligocene type.
The measurements (estimated from the original
illustrations) in comparison with those of the nearest
American titanotheres show that the incisors, canines,
and p4 are all very large, the diastema very short,
the grinders broad.
On the whole, the evidence indicates that Brachy-
diastematherium represents a distinct European phy-
lum, which closely paralleled certain upper Eocene
American titanotheres in many respects but was
distinguished by the combination of three large lower
incisors; heavy, blunt canines; very short diastema;
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
383
very progressive premolars; and broad molars. The
skull was probably brachycephalic — that is, it had
broad zygomata.
subfamhy doiichoehininae riggs
PHYLA AND RELATIONS
These Eocene titanotheres branch from the same an-
cestral stock as that of Manteoceras. Some are doli-
chocephalic, and some are mesaticephalic. The horn
swellings are developed chiefly on the nasals, partly
on the frontals. Facial region laterally compressed;
elongate symphyseal union of premaxillaries; orbits
prominent; infraorbital processes more or less promi-
nent. First occurring (Mesatirhinus) on upper levels
of the Bridger Basin (C and D) and lower level of the
Washakie Basin (A), reaching a dolichocephalic cli-
max {DolichorMnus) in Washakie B and Uinta B 1
or terminating in dwarfed mesaticephalic forms
(Metarhinus) and other collateral branches.
Following Eometarhinus of Bridger A, Huer-
fano B, the subfamily Dolichorhininae contains
the following phyla or separate series:
1. Mesatirhinus; probably ancestral to Dolichorhinus.
2. Dolichorhinus; extremely dolichocephalic, becoming
extinct.
3. Metarhinus; aberrant, small to dwarfed; broad nasals.
4. Sphenocoelus; little known, perhaps a branch of
Mesatirhinus.
A closely related subfamily, Rhadinorhininae,
contains
5. Bhadinorhinus; nasals short, pointed; possibly ances-
tral to Megaceropinae.
In their common ancestral characters these
animals exhibit closer affinities to the Manteoceras
phylum than to either the Palaeosyops-Limno-
hyops phylum or the Telmatherium phylum.
They possess in common small canine tusks
and rudimentary but progressively developing horns
and thus do not appear to have been vigorous
fighters, their relation to other animals doubt-
less being defensive rather than aggressive. They
possess long and rather straight rows of grinding
teeth, usually narrow, and the zygomatic arches are
slender and not widely projected. They are thus
readily distinguished from their broad-headed con-
temporaries, such as Manteoceras.
The known members of the series geologically are
the two species of Mesatirhinus of Bridger C and D,
each of which gives rise more or less directly to the
extraordinarily large, long-headed Dolichorhinus of
the upper beds of the Washakie Basin and the middle
beds of the Uinta Basin. The known species of Meta-
rhinus are confined to the sandstone of the fluviatile
deposits of the Uinta Basin, a fact which suggests that
they may have been small aquatic animals. At the
other extreme stand the species of Bhadinorhinus,
readily distinguished by short, pointed nasals and the
absence of infraorbital shelves and exceptionally in-
teresting because of their apparent resemblances to
the great Megacerops phylum of the Oligocene.
A simple key to the skeletal characters of these mid-
dle and upper Eocene animals is as follows:
A. Nasals elongate, spreading laterally, decurved; prominent
infraorbital processes (Dolichorhininae); face bent down-
ward:
1. Mesatirhinus; primitive, dolichocephalic, horns rudi-
mentary, feet elongate.
2. Dolichorhinus; progressive, hyperdolichocephalic, horns .
prominent, feet abbreviate.
3. Metarhinus; mesaticephalic to subdolichocephalic, nasals
expanding, dwarfed in size.
4. Sphenocoelus; hyperdolichocephalic, basieranial pits.
B. Nasals abbreviate, pointed, no infraorbital shelves (Rhadino-
rhininae):
5. Rhadinorhinus; dolichocephalic, facial region upturned.
Figure 322. — Phjdogenetic relations of the species of Metarhinus,
Mesatirhinus, Dolichorhinus, and Rhadinorhinus
The author's theories (1919) as to the phylogenetic
relations of these five phyla are expressed in Figure
322.
HISTORY OF •DISCOVERY AND CLASSIFICATION
The separation of these five distinct phyla has been
an extremely long and difficult undertaking, beginning
with the work of Cope in 1872 and ending in 1919.
Even now the position of Sphenocoelus and the relation-
ship of Bhadinorhinus and Megacerops await elucida-
tion. The chronologic record follows:
1872. Cope describes, under the name " Palaeosyops
vallidens," fragmentary upper and lower jaws from the
Washakie Basin. This little-known species, probably
from Washakie B, is now provisionally regarded as a
primitive stage of Dolichorhinus, namely, D. vallidens.
1889. Scott and Osborn describe as "Palaeosyops
hyognathus" a large lower jaw from Washakie B.
This now proves to be Dolichorhinus hyognathus.
1891. Earle describes as "Palaeosyops megarhinus"
a small skull from Washakie A, recognizing, however,
384
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
its diflerences from the typical Palaeosyops. This is
now known as MesatirMnus megarhinus.
1894. Peterson explores horizon B 2 of the Uinta
Basin and discovers a remarkable long-skulled form.
1895. This long-skulled form is described by Osborn
as " Telmatotherium cornutum," which is now known
to be a synonym of DolocJiorMnus hyognatJius.
1895. Osborn also describes a smaller form from
Uinta B 1 as "Telmatotherium diploconum." This is
now known as Rhadinorhinus diploconus.
1895. Osborn also describes, from Uinta B 2, SpJieno.
coelus uintensis, a form that still remains problematical.
1895. Earle soon afterward points out the ances-
tral relationship of "Palaeosyops megarhinus" to " Tel-
matotherium cornutum," an affinity now recognized as
that of Mesatirhinus to Dolichorhinus.
1895. Hatcher recognizes " Telmatotherium cor-
nutum" as a new genus, namely, Dolichorhinus, pos-
sessing horns, but not directly ancestral to any of the
Oligocene titanotheres.
1894-1906. American Museum explorers in the
Bridger and Washakie Basins, under Peterson, Mat-
thew, and Granger, bring together good material of
the "Palaeosyops megarhinus" type.
1908. Osborn reviews the narrow-skulled or doli-
chorhine Eocene titanotheres . with the following
principal results:
{a) Hatcher's term DolicTiorhinus is adopted, and
D. cornutus is shown to be a synonym of D. hyognathus,
both occurring in Washakie B. A new species, Doli-
chorhinus intermedius, is described.
(h) Earle's Palaeosyops megarhinus from Bridger B
and Washakie A is made by Osborn the type of the
new genus " Mesatirhinus," ancestral to Dolichorhinus.
The new species Mesatirhinus petersoni is described
from Bridger C.
(c) A related group, including small Eocene tita-
notheres with slender limbs and relatively short,
narrow skulls, from Uinta B 1 and Washakie B, is
recognized by Osborn as the distinct genus Metarhinus,
including M. fluviatilis, M. earlei, and [?] Telmatothe-
rium diploconum.
(d) The opinion is expressed that Metarhinus and
Dolichorhinus represent the long-skulled form of the
same stock that gave rise to the relatively broad-
skulled Manteoceras.
1909. Douglas describes two new species of Doli-
chorhinus {D. heterodon, D. longiceps) from Uinta B 2.
1912. Riggs greatly extends our knowledge of the
dolichorhines of Uinta B 1 and revises and expands
the species Metarhinus, Mesatirhinus, and DolicJio-
rhinus, establishing the new subfamily Dolichorhi-
ninae and basing the new genus Rhadinorhinus on the
type R. abbotti, including also the " Telmatotherium dip-
loconum" of Osborn.
1919. Osborn describes Eometarhinus from the
upper part of the Huerfano formation, representing an
extremely primitive ancestor of Metarhinus.
The original and the present determination of the
synonymy of these species is thus as follows:
Palaeosyops vallidens = Dolichorhinus vallidens.
Palaeosyop.g hyognathus = Dolichorhinus hyognathus.
Palaeosyops megarhinus = Mesatirhinus megarhinus.
Telmatotherium cornutum = Dolichorhinus hyognathus.
Telmatotherium diploconum = Rhadinorhinus diploconus.
COMPARISON WITH MANTEOCERAS
There are in Manteoceras, Mesatirhinus, and Doli-
chorhinus resemblances which prove that these animals
sprang from the same stock. They appear specially in
the comparison of the skulls of M. manteoceras and
Mesatirhinus megarhinus; in other words, the ancestral
and atavistic characters of Mesatirhinus are those
which it has in common with Manteoceras, among
which are (1) preorbital concavities; (2) nasals long,
decurved, truncate distally; (3) posterior nares com-
pressed, or narrow space between the palatines and
pterygoid plates; (4) zygomatic arches shallow; (5)
occiput broadly depressed; (6) pit in the parietal
vertex of the Manteoceras skull represented by a long
slit in the Mesatirhinus skull; (7) angulation of the
malars of Manteoceras represented by the suborbital
shelf of Mesatirhinus. Their ancestral affinity to
Manteoceras is also seen in (8) the position of the
horns above the preorbital concavities; (9) the
elongate form of the horn rudiments. There is a
decided departure from the position of the horn rudi-
ments of Manteoceras (PI. XVII) — namely, in that in
the Dolichorhininae the horn swelling is chiefly a pro-
tuberance of the nasal bones, whereas in Manteoceras
the horn swellings are chiefly on the frontal bones, the
nasofrontal suture of the dolichorhines being pushed
back by the remarkable elongation of the nasals. (10)
Another distinction is that in Manteoceras the horn
swelling is decidedly in front of the orbit, whereas in
Mesatirhinus it lies more directly above the orbit.
Other differences appear in connection with the fact
(11) that the face is relatively longer in the dolicho-
rhines than in Manteoceras.
The face is relatively longer than in Manteoceras. ,
Correlated with this is the fact that in Metarhinus,
Mesatirhinus, and Dolichorhinus the grinders are
farther forward with reference to the orbit than in
Manteoceras — that is, in the members of these groups
the postorbital process of the malar lies above the
mesostyle of m^, and in Manteoceras it lies above the
parastyle of m^ Similarly in Dolichorhinus the
lacrimal lies above the mesostyle of m^, in Manteoceras
above the mesostyle of m'. In Dolichorhinus this
relation appears to spring as much or more from the
backward displacement of the orbit (especially of its
upper border) as from the forward displacement of the
molar series. This oblique backward displacement
of the orbit may have been correlated with the in-
creased size of the nasofacial muscles, and with this
factor may also have been correlated the hypertrophy
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
385
of the posterior end of the nasals, the reduction of the
anterior prong of the frontals, and exclusion of the
frontals from the horn swelling.
The conclusion is that the dolichorhines sprang
from the same stock as Manfeoceras but that they
dolichorhine phylum (MesatirJiinus-DolichorMnus)
afford the finest examples we have yet discovered, ex-
cepting only among the Equidae, of the changes in
both skull and teeth which are correlated with pro-
gressive dolichocephaly, accompanied by progressive
Figure 323. — Top view of the skull in the Manteoceras-Dolichorhinus group
One-eighth natural size. A, Manieoceras manteoceras; B, Metarhinvs earlei; C, MesaiiThmus petersoni: D, Dolichorltinus Jiyognatlius.
diverged and radiated along lines of their own into
persistent mesaticephalic and extreme dolichocephalic
types.
cyptocephaly," or bending of the face downward on
the cranial axis, as in many other grazing quad-
rupeds. Combined with this slowly acquired and
DOLICHOCEPHALY AND CYPTOCBPHALY IN THE MESATIRIIINUS-
DGLICHOEHINDS PHYLUM
Figure 324. — Palatal views of the skull in the Manteoceras-Dolichorhinus group
One-eighth natural size. (See fig. 323.)
only imperfectly attained cropping and grazing adap-
tation the cranium also converges toward the
Oligocene titanothere type, as observed in the flatten-
ing of the top of the skull {Dolichorhinus), the rise of the
horns {Mesatirhinus, Dolichorhinus), the incipient
metamorphosis of the premolar and molar tooth
Elongation of the crowns of the grinding teeth, or
subhypsodonty, generally characterizes the dolicho-
rhines, from which we infer that they fed on harder
kinds of food and may have adopted grazing as well as
browsing habits. The members of the central
" Cyptocephaly (Kuirru, ut^aXij) is a comparatively new term (first employed
by Osborn as cytocephaly)
386
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
pattern (RJiadinorJiinus) . Yet so far as we know
these long-headed animals of the Mesatirhinus-
DolichorMnus phylum were not destined to survive and
give rise to any Oligocene titanothere but rather to
terminate in an excessively specialized type.
CONVERGENCE OR KINSHIP CONTROL
These dolichorhines afford an illuminating illustra-
tion of kinship control in the simultaneous evolution
of different character groups. Some of these character
groups are predetermined or controlled by ordinal,
others by family, others by generic affinities with
related titanotheres. The general expression of kin-
ship control may be summarized as follows:
1. Perissodactyl or ordinal kinship is shown, among
other characters, in the independent progressive com-
plication of the premolar-molar teeth, three of the
premolars tending to acquire the pattern of the molars,
although this pattern is less perfectly developed in
other Perissodactyla.
2. Titanothere family kinship is indicated in the
independent progression of the development of naso-
frontal horns, the flattening of the top of the cranium.
FiGUBE 325. — Leidy's cotypes of Palaeosyops
( = Mesaiirhinus) Junius
Natural size. Crown view of premolar and molar in the
museum of the Acad. Nat. Sci. Philadelphia. Bridger (?)B;
level doubtful. A, Eight fourth lower premolar; B, posterior
part of third lower molar.
the obliteration of the sutures on the top of the
cranium, and the elongation of the middle portion of
the cranium.
3. The subfamily kinship to the Manteoceratinae is
indicated in the persistent subtriangular shape of the
horns, in the development of an infraorbital shelf, and
in the contracted posterior nares and broadening
nasals.
In their progressive dolichocephaly, a tendency
that strongly affects the middle region of the skull
between the orbits and the auditory meatus, these
animals (Mesatirhinus-Dolichorhinus) are partly inde-
pendent of their subfamily, family, or ordinal relations
and follow an extreme adaptive direction of their own
in the elongation of the midcranial region and of the
teeth.
In this special adaptation to their partial grazing
habits the dolichorhines further parallel certain of the
Equidae and other grazing animals, such as the cattle,
in their cyptocephaly. (See figs. 213, 214.) The
incisor teeth further acquire deep posterior pits, or
pockets, analogous to the pits that are developed in
the incisors of the upper Oligocene Equidae and that
tend to become typical cropping teeth.
DIVERGENT OR INDEPENDENT EVOLUTION OF CHARACTER
GROUPS IN THE DOLICHORHINES
The independent evolution of these four or five
groups of characters as observed in the skull alone
obviously affords only a partial picture of the play
and interaction between the vast number of contem-
poraneous processes that are involved in the evolution
of the members of this phylum. If we could similarly
compare all parts of the vertebral column and of the
limbs, we should probably discover many additional
illustrations of this law of the evolution of groups of
characters under the influence partly of kinship and
partly of independent adaptation.
The principle of independence or divergence is well
illustrated in the skull. In Figure 302 the skulls of
Manteoceras and the MesatirMnus group are compared
as seen from the side. The palatal view of Manteoceras,
Metarhinus, MesatirMnus, and DolichorJiinus (flg. 324)
brings out the resemblances and contrasts between
these four forms. The superior view (fig. 323) also
brings out the wide progressive divergences between
these undoubtedly related forms.
We may also compare superior views (PI. LXXX)
of the skulls of Metarhinus and Rhadinorhinus, show-
ing how the latter departs from the other members of
this dolichorhine group in the abbreviation and point-
ing of the nasals and in the reduction of the infraorbital
processes.
PROGRESSIVE DOLICHOCEPHALY IN MESATIRHINUS-
DOLICHORHINUS
The dolichocephaly, which is the chief progressive
character of the MesatirTiinus-Dolichorhinus phylum,
is beautifully illustrated in the accompanying series
of illustrations (figs. 339, 340).
A very important fact (see the following table)
is brought out by the cranial indices and ratios in this
series of species — namely, that while the skulls lengthen
and become relatively narrower, the facial portion is
not relatively abbreviated as it is in the Manteoceras
phylum, because the faciocephalic index is the same
in the terminal member of the series, D. JiyognatJius,
as in M. megarJiinus, the most primitive member of the
series. In other words, in Manteoceras the face is
abbreviated; in the dolichorhines it is not.
Cranial indices of MesatirMnus and Dolichorhinus
M. megarhinus
M. petersoni
D. superior
D. longioeps
D. intermedins
D. hyognathus ( = cornutus)
D. fluminalis
56-59
51-52
52
41-46
43-46
45
48
49
49
51-53
41
39
38
35-37
38
36
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
387
In other words, while the ratio of breadth to length
falls from 59 to 45 per cent, the ratio of the length of
the face to the entire length of the cranium, or facio-
cephalic index, remains at 48. This is a very impor-
tant distinction, because progressive abbreviation of
the face is characteristic of all the true Oligocene
titanotheres correlated with dolichocephaly but is not
found in these dolichorines. We observe other corre-
lated dolichocephalic changes in the skull — namely,
(1) the occipital condyles are set broadly apart; (2)
the external auditory meatus becomes widely open;
(3) wide spaces arise between the cranial foramina;
(4) the horn rudiments rise chiefly on the nasals and less
On the frontals; (5) the nasals have a long exposure on
the top of the skull; (6) there are correlated changes
in the teeth.
The teeth also show the following correlations with
doliochocephaly : (1) The opposite molar-premolar
series become parallel; (2) the palate is narrowed,
elongated, and arched; (3) the individual molar teeth
are elongated or laterally compressed; (4) the pre-
maxillary and mandibular symphyses become greatly
elongated; (5) the jaws become long and slender, and
there is an increasing distance between ms and the
posterior border of the jaw, the coronoid becoming
relatively low and backwardly recurved, its anterior
edge oblique rather than transverse, and the chin
shallow and sloping (hence the term hyognaihus, or
hog-jawed, applied by Scott and Osborn); (6) simi-
larly the inferior molars become elongate; (7) the
whole upper tooth row shifts forward with reference
to the orbit.
The recently discovered EometarJiinus is described
on pages 200, 419, 420.
Mesatirhinus Osborn
Plates XVII, L, LXXI, LXXII; text figures 26, 27, 33, 122, 211,
217, 219, 265, 302, 327-331, 333, 339-341, 483, 511-514, 516,
520-523, 526, 558, 559, 562-571, 586, 647, 656, 661, 686, 702,
713, 716, 724, 745
[For original description and type references see p. 182. For slieletal eliaracters
see p. 636]
Geologic horizon. — Bridger C and D and Washakie A.
Generic characters. — Middle Eocene titanotheres of
small but increasing size; basilar length, 354-485
millimeters. Mesaticephaly progressing to dolicho-
cephaly. Horns incipient, borne chiefly on the nasals;
prominent infraorbital malar shelf; nasals elongate,
laterally recurved; cranium profile convex; face de-
flected; sagittal crest gradually broadening; no sec-
ondary palate. Humerus short; tibia relatively long;
carpus and tarsus narrow; astragalus with elongate
neck; metapodials relatively elongate.
Geologic distribution. — There is the Eometarhinus of
Huerfano B (Bridger A), and the Mesatirhinus Junius
of Bridger B. In Bridger C and D, also in Washakie
A, there first appears a rich array of small titanotheres,
which are readily distinguished from the contemporary
species of Palaeosyops, Telmatherium, and Manteoceras
by the generic characters enumerated above. These
animals are related on the one side to Metarhinus and
on the other, by progressive changes, to Dolichorhinus,
and the phylum is therefore regarded as central. The
phylum is divided into the smafler, more primitive
species Mesatirhinus megarhinus and the partly con-
temporaneous, more progressive species M. petersoni.
These species are contemporaneous in Bridger C 5, and
both animals are found in Washakie A, which is
evidence that they are contemporaneous and not
successive species. At the summit of Uinta B 1 occurs
the larger and more progressive "Mesatirhinus"
superior, with partly flattened cranium. This animal
is here referred to Dolichorhinus.
In the Uinta region Metarhinus is so abundant in the
fluviatile sandstones of Uinta B 1 that the horizon is
named the Metarhinus zone. The animals disappear
at the summit of this zone in the " Metarhinus sand-
stones."
The synopsis of these species is as follows :
Mesatirhinus Junius (Leidy)?, Bridger B, a diminu-
tive animal. (See fig. 325.)
Mesatirhinus megarhinus (Earle), Bridger C and
Washakie A. Skull small (about 354 by 170 mm.);
cephalic index, about 53; faciocephalic index, 48;
palatal crests narrow; nasofrontal horns incipient;
premolars in less advanced stage.
Mesatirhinus petersoni Osborn, Bridger C and D and
Washakie A. Skull of intermediate size (about 425 by
205 mm.); cephalic index, 49; parietal crest narrow;
faciocephalic index, 48; premolars in more advanced
Mesatirhinus (Dolichorhinus) superior Eiggs, sum-
mit of Uinta B 1. Skull larger (485 by 240 mm.);
cephalic index, 52; faciocephalic index, 48; parietal
crest spreading; cranium flattened on top.
It is important to note that although these three
species succeed each other progressively and this pro-
gression leads directly to Dolichorhinus, there is no
proof of direct phyletic succession.
388
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Measurements oj Mesatirliinus megarhinus, M. petersoni, Dolichorhinus superior, and Metarhinus fluviatilis, in
millimeters
M. megarhinus
M. petersoni
D. supe-
rior, Field
Mus. 12188
(type)
Princeton
Mus. 10008
(type)
Am. Mus.
1514
Am. Mus.
12202
Am. Mus.
1523
Am. Mus.
1571
Am. Mus.
1509
Am. Mus.
1566
Am. Mus.
12184
(type)
tilis, Am.
Mus. 1500
(type)
Pmx— condyles -
"■354
°402
386
"■425
485
"355
Mx— condyles - -
342
"180
210
92
"ISO
183
'■212
205
"220
"255
89
250
"180
"200
Face ...
°170
0 195
223
178
157
140
91
19
23
205
"205
"170
202
"190
Nasal-postorbital frontal process _
P"-m3 . . ... ......
168
147
190
156
141
90
18
23
195
154
138
87
18
140
125
80
16
22
147
132
83
17
22
147
132
82
17
22
156
139
89
18
23
184
105
146
P2-m3 .
130
Mi-m3
16
85
P*, ap. - ... -
17
P*, tr
23
* Estimated.
10008. Washakie Basin.
1514. Washak;ie A.
12202. Bridger C 5.
1523. Bridger C or D.
1571. Washakie A.
1509. Bridger D.
1556. Bridger D.
12184. Bridger D 3.
12188. Uinta B 1.
1500. Uinta B 1.
The figures show that M. petersoni, most specimens
of which are from the higher levels of Bridger D, is
considerably larger in all measurements than M.
megarJvinus. Both are much larger than the type of
Metarhinus fluviatilis from Uinta B 1.
Mesatirhinus Junius (Leidy)
Text figures 91, 325
[For original description and type references see p. 159]
Type locality and geologic horizon. — The type lower
molar of M. Junius, according to Leidy's description,
was found near Fort Bridger, Wyo., at a geologic level
that Granger places in Bridger B. If this geologic
level is correct M. Junius is the oldest known species
in the Mesatirhinus phylum. Its geologic age must,
however, be regarded as indeterminate.
Specifl.c characters. — A doubtfully referred specimen,
imperfectly known. M'-m^, 69 millimeters. A dimi-
nutive Mesatirhinus or Metarhinus.
Materials. — The type specimen (Acad. Nat. Sci.
Philadelphia) is very fragmentary. The only other
material that may be referred even provisionally to
this species is a diminutive set of teeth, including
ml to m^ (Am. Mus. 12686), from level D 5 of the
Bridger. The teeth present generic resemblances to
those of Mesatirhinus megarhinus but are far smaller
than in any known upper Bridger, Washakie, or Uinta
titanothere, m'-m' measuring only 69 millimeters, as
against 85 in Metarhinus fluviatilis . The teeth are less
hypsodont than in allied species; m' is relatively very
small and m' very quadrate. A comparison of the
lower molar with the type ms of Palaeosyops Junius
Leidy leaves the specific identity doubtful. The com-
parative measurements are as follows:
Measurements of teeth of
species of Mesatirhinus, in millimeters
M.jimius?,
Am. Mus.
12686;
Bridger D 5
M. mega-
rliinus,
Am. Mus.
12202;
Bridger C 5
M. fluvia-
tilis, Am.
Mus. 1500
(type);
Uinta B 1
M. peter-
Mus. 1656;
Bridger D
M. peter-
soni, Am.
Mus. 1612;
Bridger D
Mi-ms
Ml, ap
Ml, tr
69
21
22
36
83
85
23
90
26
27
Ms, ap
46
Mesatirhinus megarhinus (Earle)
Plate LXXII; te.xt figures 106, 217, 324, 326, 328, 330, 331, 508,
510, 558, 560, 561, 685, 737
[For original description and type references seep, 170. For slieletal ciiaracters see
p. 637]
Type locality and geologic horizon. — Washakie Basin,
Wyo.; probably Washakie A. The species is also
recorded from Bridger Basin, Wyo., Uintatherium-
Manteoceras- Mesatirhinus zone, level Bridger C
( = Washakie A).
Specific characters. — P'-m^, 140-147 millimeters;
true molars, 77-83. Cephalic index, 56-59. Cranial
length, premaxillaries to condyles, 354 millimeters
(estimated); facial region rather short (about 170
mm.); faciocephalic inde-x 48 (estimated); occipital
condyles broad; premaxillary symphysis short; infra-
orbital shelf prominent. Tetartocones on p^, p* very
rudimentary. Nasofrontal horns incipient.
Materials. — The type (Princeton Mus. 10008) is
badly preserved and unfortunately lacks the occipital
condyles, which appear to be relatively broader in
Mesatirhinus than in Dolichorhinus. The geologic
level of the type is not definitely recorded, but is
probably Washakie A.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE •TITANOTHBRES
389
The material in the American Museum referred to
this species includes the following: From the Bridger
Basin, skull, No. 12202 (level C 5); palates, Nos. 12206
(level C 5), 1519, 1523 (level unknown); lower jaws,
Nos. 1520, 1551, 12207 (level C 5), 12199 (level C 5);
from the Washakie Basin (level A), palates, Nos. 1513,
1514; lower jaws, Nos. 1575, 1577.
Mesadrhinus petersoni Osborn
Plates L, LXXI, LXXII; text figures 26, 27, 33, 122, 211, 217,
219, 255, 302, 327-331, 333, 339-341, 483, 611-514, 516, 520-
523, 526, 558, 559, 562-571, 586, 647, 656, 661, 686, 702, 713,
716, 724, 745
[For original description and type references see p. 182. For
skeletal characters see p. 641]
Type locality and geologic Jiorizon. — ■
Cattail Springs, Bridger Basin, Wyo.;
Uintatherium-Manteoceras-MesatirMnus
zone (Bridger), level D 3. Also re-
corded from Bridger C (?) and
Washakie A.
Specific cliaracters. — P^-m^, 154-157
millimeters; true molars 87-90. Skull
length, premaxillaries to condyles (esti-
mated), 402-425 millimeters; cephalic
index, 51-52; preorbital facial region
(estimated), 195-205 millimeters; facio-
cephalic index 48. Other characters
as in M. megarhinus — that is, broad
occipital condyles, infraorbital shelf,
etc.
Materials. — The type skull (Am.
Mus. 12184) is from Bridger D 3 (fig
327). Comparison of this animal with
the type of M. megarhinus can leave
no doubt that we have to do here
with a more advanced stage of evolu-
tion. The skull is longer, the preorbital
region especially. The grinding teeth
occupy more space, and there is an
average advance in all the premolar
rectigradations, which prove that these
differences in form and size are not
merely due to fluctuations of size or
differences of sex.
Other specimens (in the American
Museum except as noted) referred to
this species are, from the Bridger, skulls Nos. 1509
(level D) and 1556 (level D; now in British Museum),
lower jaw No. 1567, lower jaw No. 12191 (level C 2);
from Washakie A, skull No. 1571 and lower jaws
Nos. 1512, 13178.
Of these No. 1571, from Washakie A, fortunately
has associated with it the fore foot, radius, ulna,
astragalus, and pelvis. Another valuable skeleton
(Am. Mus. 11659) is recorded from Bridger C 5,
and a well-preserved forearm and manus in the Prince-
ton Museum (No. 10013) came from Bridger C or D
of Henrys Fork, Wyo.
From Washakie A comes a very progressive doli-
chocephalic specimen (Am. Mus. 1651) consisting of
the three upper molars, which are strongly compressed
laterally and measure collectively 96 millimeters.
This specimen is provisionally referred to this species
and appears to be an important and interesting
transitional form leading into DolicJiorJiinus.
General characters of Mesatirhinus megarhinus and
M. petersoni. — It is impracticable to describe M.
Figure 326. — Type skull of Mesatirhinus megarhinus
One-fourth natural size. Princeton Mus. 10008, Washakie Basin, Wyo., level Washakie A?. A i. Side
view (reversed) ; As, palatal view; A3, top view.
megarhinus and M. petersoni separately, because it
would involve duplication of description.
Sexual characters: Unfortunately the imperfect
preservation of the canine teeth does not admit of the
sharp separation of males and females that is possible
for many of the series of skulls. Comparison of the
teeth in the more perfectly preserved jaws, however,
shows that the canines were decidedly smaller in the
females than in the males.
390
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Cranial elongation with age: There is considerable
evidence that cranial elongation is not only a progres-
sive but an age character — that is, one which appears
in advanced years and through the "law of accelera-
tion" will appear in earlier years of subsequent gener-
ations. For example, the space between the glenoid
fossa and m^ elongates with age, and correlated with
it is the elongation of the ramus of the jaw between ma,
the anterior border of the coronoid process, and the
condyle.
pas P"~
Figure 327. — Tj'pe skull of Mesatirhinus petersoni
One-fourth natural size. Am. Mus. 12184, Cattail Springs, Bridger Basin, Wyo.,
Bridger D 3. Ai, Side view; Ai, top view.
Slcull. — The general characters of the skull are as
follows: (1) Skull as a whole elongate, with consider-
able space between the glenoid process and the occipital
condyles, leaving the auditory meatus open, as con-
trasted with the contemporary Palaeosyops, in which
this space is abbreviated; (2) zygomatic arches rela-
tively straight, slender, and gently arched outward;
(3) on the malars an infraorbital shelf, which ap-
parently increases progressively; (4) sagittal crest
sessile and narrow as compared with LimnoJiyops
laticeps; (5) nasals long, expanding and decurved dis-
tally; (6) premaxillary symphysis more abbreviate
than in Telmatherium; (7) face moderately bent upon
cranium, parietals convex in side view; (8) postorbital
process of malar above posterior part of m^. On
comparing the side views of Telmatherium cultridens
and of Mesatirhinus petersoni we see that in the
former the premaxillary is stouter, vertically deeper
anteriorly, and extends posteriorly farther up on the
maxillary. In T. ultimum this is much more pro-
nounced. In Mesatirhinus, in correlation with the
smaller incisors and more slender maxUla, the pre-
maxillary is shallower vertically, and the sym-
physeal surface is more delicate.
The skull of members of Mesatirhinus petersoni ex-
hibits many marks of general affinity to those of their
collateral relative Manteoceras manteoceras. Among
these are (1) the shape of the symphyseal union of the
premaxillaries ; (2) the narrowing of the postnarial
space between the pterygoids; (3) the sutural rela-
tions of the nasals, frontals, maxillaries, malars, and
lacrimals, as seen in side view, with the exception of
the position of the horn rudiment; (4) the concavities
at the side of the face slightly above and in front of
the orbits; (5) just above these concavities the promi-
nent convexities of the nasals at their junction with
the frontals above the orbits, extremely interesting as
a very early stage of horn evolution and prophetic of
the distinct horn base of Dolichorhinus; (6) presence
of a long and narrow pit in the anterior portion of the
sagittal crest.
Among the most significant resemblances to Man-
teoceras are also the similarity in the base of the
craniima; the slender zygomata, constricted back of
the orbit; and the underlying similarity in the denti-
tion in spite of differences of proportion.
The skull differs markedly from that of Manteo-
ceras, however, in the presence of infraorbital shelves
and in its greater dolichocephaly. It also differs from
Manteoceras in the shape of the occiput, shape of the
skull top, and especially in the dentition. Its closer
affinities, therefore, are with Dolichorhinus.
The detailed characters of the teeth exhibit a direct
dolichocephalic adaptation of those of the Manteo-
ceras type. The community of type, again, is due to
a community of ancestry, the two lines running
together perhaps prior to Wind River and Huerfano
time.
More in detail: The superior view of the skull
(fig. 328) exhibits the characteristic anterior expansion
of the nasals, which measure transversely (Am. Mus.
1556, M. petersoni) anterior region 59 millimeters, mid-
region 43, posterior region 84; the total length is 167.
The nasals are separate anteriorly but firmly coalesced
posteriorly; the lateral convexity just in front of their
junction with the frontals (figs. 327, 328) represents the
rudimentary stage in the evolution of the horn. The
frontals are expanded above the orbits (91 mm., tr.),
gradually contract posteriorly, and are bounded by the
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
391
prominent supratemporal crests, which arise from the
postorbital processes and in some adult individuals
converge in the form of two broadly convex ridges
into the short and narrow sagittal crest (10 mm.)-
Between these ridges there is a median depression.
The suture between the frontals and
parietals can apparently be made out
in the Princeton skull (No. 10041,
Mesatirhinus petersoni?). The parie-
tals are best observed in the same
skull and in Am. Mus. 1509 (-M.
petersoni). In the superior view of
the skull as figured we observe also
the short symphyseal union (42 mm.)
between the premaxillaries (much
more abbreviate than in Dolicho-
rhinus), the prominent infraorbital
shelf on themalars, and the slender
section of the zygomatic arches.
In the inferior view of the skull of
M. petersoni (fig. 328) we observe
the converging incisive borders of
the premaxillaries, the relatively
narrow and transversely arched
palate, which measures 152 milli-
meters from the incisive foramen to
the posterior nares. The palatal
portion of the palatines measures
70 millimeters in the midline and
converges anteriorly; on either side
of the posterior nares the convex
inner surfaces of the palatines con-
verge, and on the inner side of the
narrowest portion of this postnarial
space are placed the slender ptery-
goids, which are well defined. The
conformation of this entire region is
very characteristic of this genus as
well as of M. manteoceras . The
basioccipital region is best exhibited
in the Princeton skull (No. 10041,
M. petersoni), a very distinctive
feature being the wide separation
(28 mm.) by a plate of bone
the foramen ovale and foramen
lacerum medium, the same plate
measuring but 17 millimeters in
the contemporary Limnohyops lati-
ceps. The conformation of this important region
of the skull, as well shown in Figure 333, includes
the following noteworthy features: (1) The deep
groove extending backward and inward on the inner
side of the postglenoid facets, believed to have lodged an
extension of the meniscal cartilage, as in the horse; (2)
the prominent basioccipital and basisphenoid; (3) the
narrow bridge of bone between the foramen condylare
101059— 29— VOL 1 28
and the foramen lacerum posterius; (4) the continuity of
the foramen lacerum medium and foramen lacerum pos-
terius; (5) the peculiar inward extensions of the con-
dylarfacets; (6) the general elongation of the basicranial
axis; (7) the open nature of the auditory meatus.
nf One-fourth natural size,
tain, Henrys Fork, B
FiGUEB 328. — Skull of Mesatirhinus petersoni
Ai, Side view (reversed); Britisli Mus. (formerly Am. Mus. 1556), Big Bone Moun-
idger Basin, Wyo., Bridger D; occipital region restored from Am. Mus. 1509, Big Bone
Mountain, Henrys Fork, Bridger Basin, upper Bridger; and Princeton Mus. 10041; Incisors and canines
from Am. Mus. 1571, Washakie Basin. Az, Top view; occipital region from Princeton Mus. 10041. As,
Palatal view; details of pterygoid region from Am. Mus. 1509 (see above); incisors from Am. Mus. 1571 (see
above) and 1514 (Af. mcgarhinns) , La Clede Meadows, Washakie Basin; suture between basioccipital and ex-
occipital from Am. Mus. 12202 (J/. mejarUnus) , Summers Dry Creek, Bridger Basin, Bridger C 5, lower level.
In the lateral view of the skull of M. petersoni
(figs. 327, 328), we observe especially the horizontal
suture connecting the maxillaries with the nasals, the
lateral compression of the sides of the face at this
point, the rudimentary horn convexities of the nasals,
the scalelike overlap of the nasals by the frontals,
the participation of the maxillaries in the anterior
portion of the infraorbital shelf, the vertical extension
392
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
of the lacrimals, the postorbital processes on the
frontals and malars, the slender malar portion of the
zygomatic arch, the comparatively slight upward
Figure 329. — Skulls of Mesaiirhinus
petersoni
One-fourth natural size. A, Front view of skull in
British Museum (formerly Am. Mus. 1566); Big
Bone Mountain, Henrys Fork, Bridger Basin,
Wyo., upper Bridger, level D, B, Occipital view
of Am. Mus. 1509; Big Bone Mountain, Henrys
Fork, Bridger Basin, upper Bridger, level D.
cm'vature of the squamosal portion of the zygoma, the
incipient arching of the parietal region, the greatly
elongate and not deeply vertical temporal fossa, the
wide space between the postglenoid and post-tympanic
processes, the relations of the frontals, parietals,
occipitals, and squamosals, and the formation of the
temporal fossa.
The occiput is relatively broad and low, measuring
(Princeton Mus. 10041, M. petersoni?) 107 millimeters
transversely by 85 vertically. There is a deep de-
pression in the superior portion of the occiput; the
occipital condyles are widely divergent superiorly
on either side of the foramen magnum.
The anterior view of the skull of M. petersoni (fig.
329, A) best illustrates the characteristic form and
symphyseal junction of the premaxillaries, the stout
lateral decurvature of the nasals, and the postero-
lateral horn rudiments on these bones.
Dentition; influence of dolichocephaly. — In general
the teeth show the dolichocephalic tendency, although
they have not reached the extreme of elongation seen
in the species of DolichorJiinus; they also are to be
contrasted with those of the more mesaticephalic
M. manteoceras. Thus it may be noted that Manteo-
ceras and Mesaiirhinus are separated by strong differ-
ences in the premolars and also in the molars. The
premolars of Mesatirhinus are distinguished from
those of Manteoceras as follows: (a) They are rela-
tively longer as compared with their width; (b) in
crown view p^-p"* appear more circular than in Man-
teoceras in consequence of the deuterocones being
farther forward and the postero-internal part of the
crown more evenly rounded out; (c) the tritocones are,
on the whole, relatively larger and flatter externally;
(d) the external cingula are better defined opposite
the tritocone; (e) the protocone ribs on the ectoloph
are more pronounced and narrowed; (/) the protoco-
nules and tetartocones are better developed. Between
typical members of M. megarhinus and M. petersoni
the differences are of a progressive character — that is,
in M. petersoni the deuterocones and tritocones are
larger, the tetartocones and ectoloph ribs are much
more pronounced.
Incisors. — The superior incisors (fig. 330) are ar-
ranged to form a forward-pointed arch — that is, the
opposite series are less parallel to each other than in
T. cultridens and less transverse in position than in
Palaeosyops. The series is short-crowned, with convex
anterior and convexo-concave posterior faces; the
posterior cingulum foreshadows the marked develop-
ment of the cingulum in DolicTiorhinus. A note-
worthy character is that i' is less caniniform than in
TelmatJierium.
Canines. — The canines are subround in section
rather than laterally compressed as in TelmatTierium.
The enameled crown area measures vertically 36 milli-
meters and in base diameter 18 millimeters in certain
specimens of M. petersoni. In the smaller specimens
of M. megarhinus the crown measures 26 millimeters
Figure 330. — Incisors, canines, and
prema.xillae of Mesatirhinus
One-balf natural size. A, Crown view of ^f.
megarhinus, Am. Mus. 1514, La Clede Mead-
ows, Washakie Basin, Wyo.; B, side view of
M. petersoni, Am. Mus. 1671, Washakie Basin.
in length. The inferior canines (Am. Mus. 1576,
1575) are more slender and rounder toward the tip,
with more feebly indicated anterior and posterior
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
393
ridges, a feature which at once distinguishes them from
the lower canines of T. cultridens, in which these ridges
are prominently marked.
Premolars. — The superior premolars especially ex-
hibit the progressive rectigradations in the new cus-
pules, which, as well as the progressive changes of
proportion, are subject to slight fluctuations. In the
more advanced American Museum specimens {M.
petersoni) the first superior premolar is separated from
the canine by a narrow diastema, whereas in the less
advanced Princeton Museum type of M. megarJiinus,
which represents a less dolichocephalic stage, it is
in actual contact with the canine. The detailed
characters are as follows: P' is bifanged, with a
simple protocone, strongly compressed laterally, in
which the proportions are typically 14 millimeters
anteroposterior by 9 transverse; in the more pro-
gressive specimens (M. petersoni) the tritocone (a
rectigradation) is seen as a rudimentary swelling of
the posterior base of the crown, which is less con-
spicuous in M. megarJiinus. P^ is a highly character-
istic tooth, suboval or slightly compressed trans-
versely, the proportions (ap. by tr.) varying in different
specimens from 15 by 13 to 16 by 16 millimeters;
the proportions of this tooth are those correlated
with mesaticephaly progressing into dolichocephaly;
it is typically tricuspidate (protocone, deuterocone,
tritocone); a generic feature is the excess of the large
conic protocone over the small, externally flattened
tritocone; the ectoloph is slightly cingulate (M.
petersoni) but lacks the strongly accented cingulum
around the base of the tritocone seen in T. cultridens
and T. validum. P^ is naturally a more progressive
tooth, the breadth exceeding the length (ap. 14
millimeters, tr. 18, M. megarhinus ; ap. 17, tr. 20,
M. petersoni), the tritocone and protocone com-
ponents of the ectoloph being more subequal, the
basal external cingulum opposite the tritocone being
more accented, and the deuterocone being more directly
internal in position. P* still further marks this
progression toward the molar type in its dimensions —
17 by 22 millimeters (ap. by tr.) in M. megarhinus, as
compared with 18 by 24 in M. petersoni. The external
cingulum, varying in both species, is either partially
indicated (Am. Mus. 1523, 1571) or extends across
the outer face of the crown (Am. Mus. 1514, 1556); the
less progressive individuals (Am. Mus. 1523, 1513, M.
megarJiinus) pass into more progressive stages (Am.
Mus. 1556, 1509, M. petersoni) in which a faint rudi-
ment of the protoconule is observed in p^, p^, and in
Am. Mus. 1556 {M. petersoni) even a faint elevation
of the tetartocone is observed (a rectigradation).
Similarly the convex external rib of the protocone
becomes more marked.
In comparing the premolar series in all these
specimens it is seen that the external cingulum
exceptionally almost or quite embraces the ectoloph,
but that the internal cingulum never completely
embraces the deuterocone, as in the type of MetarJiinus
fluviatilis. The premolar cingula are on the whole as
progressive or more progressive than those of T.
cultridens and D. vallidens. Another important pro-
gression is seen in the premolar ectolophs — namely,
in certain specimens, Am. Mus. 1556, 1509 {M.
petersoni) the tritocone ectoloph is flat, as in D.
vallidens (Cope), whereas in other specimens, Am.
Mus. 1571 (if. petersoni), 1513 {M. megarJiinus),
12184 (type of M. petersoni), the tritocone ectoloph is
more conic, as in M. manteoceras.
Comparative measurements of the superior pre-
molars are given in the table on page 388.
The inferior premolars are more or less perfectly
represented in six jaws in the American Museum
collection, none of which, however, are certainly
associated with skulls. Pi is a small, conic or slightly
flattened tooth, separated from the canine by a diastema
8 to 12 millimeters in length; a slight diastema (4
mm.) also separates it from P2; pi is a typically single,
rarely bifanged tooth, with a narrow, laterally com-
pressed, recurved, pointed crown (9 by 6 mm.). P2 is
a bilobed tooth and elongate, but relatively less so
than in T. cultridens — in fact, it is slightly more pro-
gressive than in that species; the typical measure-
ments are 18 millimeters anteroposterior and 9
transverse; anterior to the elevated protocone is the
beginning of the anterior valley and a rudiment of
the antero-internal cusp ( = paraconid) ; the much more
depressed talonid similarly consists of a shallow,
rudimentary crescent, opening inward. P3, like its
fellow in the upper jaw, shows more equal anterior
and posterior lobes, on which the crescents and
internal valleys and the cusps corresponding to the
paraconid and entoconid of the molars are more
accented; the typical proportions in M. megarJiinus
are 17 by 9 millimeters; in M. petersoni the typical
proportions of p2 are 19 by 10, but this tooth has
only a rudiment of the prominent internal cusp
corresponding with the metaconid of the molars.
P4 is still further advanced or submolariform, having a
prominent internal cusp corresponding to the meta-
conid in the molars, each lobe consisting of two fairly
defined crescents ; it differs from mi in its smaller dimen-
sion {M. megarJiinus 17 by 11 mm., M. petersoni 19 by
12) and in the nonelevation of the postero-internal cusp
(entoconid). The cingulum is practically rudimentary
or wanting in all these teeth.
Molars. — The MesatirJiinus or generic characters of
the superior molars (PI. LXXII) are seen in the fol-
lowing features: (1) The slight excess of anteropos-
terior over transverse diameters, especially in the more
dolichocephalic M. petersoni; (2) the high, sharply
pointed protocone (unworn height, 6 mm.); (3) the
high, elongate external cusps (height of unworn para-
cone, 23 mm.); (4) the very sharp para-, meso-, and
394
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
metastyles; (5) the reduced but still persistent proto-
conules (sometimes vestigial, Am. Mus. 1519, 1513,
M. megarhinus); (6) the anterior and posterior cingula;
FiGTjRE 331. — Lower jaws of Mesalirhinus
One-tourth natural size. A, if. megarhinus, Am. Mus. 1520, Bridger Basin, Wyo.,
level unknown, B, M. petersoni, Am. Mus. 1512, La Clede Meadows, Washakie
Basin, Washakie A; symphyseal region restored from Am. Mus. 1575 (.M. mega-
rhinus), La Clede, Washakie Basin; canine from Am. Mus. 1551 (M. megarhinus).
Twin Buttes, Bridger Basin. C, M. petersoni, Am. Mus. 1667, Bridger Basin,
level unknown; lower jaw of an old animal. D, M. petersoni, Am. Mus. 13178,
north of Haystack Mountain, Washakie Basin, Washakie A; lower jaw of an
old animal.
(7) the vestigial metaconule seen in m^ only in certain
specimens (Am. Mus. 1556), the majority exhibiting
no trace of this cusp; (8) the serrate external cingulum
at the bottom of the ectoloph valleys, especially in the
more progressive specimens.
The superior molar series measures from 87 to 91
mUluneters in M. petersoni and from 77 to 83 in M.
megarJiinus. The inferior molar series measures from
94 millimeters in M. megarhinus to 104 in M. petersoni.
The inferior molars are characterized by faint
serrate, noncontinuous external cingula, which follow
the curvature of the crown inward between the outer
lobes, as distinguished from the cingula in P. paludosus,
which form a straight line along the base of the outer
border of the tooth. In the long, narrow ma {M.
megarlhinus, ap. 43 mm., tr. 19; M. petersoni, ap. 46,
tr. 19) a serrate internal cingulum rises on the inner
side of the hypoconulid but does not ascend so
prominently as in T. cultridens. This hypoconulid is
progressively conic in form; it is slightly more conic,
more median in position, and less sharply crescentic or
cupped on the inner side than in T. cultridens; but in
certain specimens (Am. Mus. 1512, 1577) it has the
moie crescentic form of the T. cultridens type. In
some molars (Am. Mus. 1512, 1575, 1520) faint rudi-
ments of the metastylid fold are seen, but as a rule the
internal valleys are open and smooth. Other teeth are
too much worn to determine the presence or absence of
the metastylid fold.
Lower jaws of M. megarJiinus and M. petersoni. —
The lower jaw of these animals is represented by a
large number of separate jaws belonging to both
species (see below). These jaws taken together afford
very complete knowledge of the progressive, age, and
sexual characters. There is a very marked disparity
in size between the smallest (Am. Mus. 1520, M.
megarhinus) and the largest (Am. Mus. 1512, M.
petersoni).
Comparative measurements of Mesatirhinus and Metarhinus, in
millimeters
Pi-ms
Pj-ms
Mi-m3
Ms, anteroposterior
Incisive border to angle.
Mesatirhinus
megarhinus,
Am. Mus. 1520
(Bridger D?)
162
146
94
43
Mesatirhinus
petersoni.
Am. Mus. 1512
(Washakie A)
176
■160
102
46
325
Metarhinus
fluviatilis.
Am. Mus. 1946
(Uinta B 2)
161
157
102
46
The coronoid rises rather rapidly behind m3, with
a more or less rounded or angulate anterior border and
with nearly parallel anterior and posterior contours
until near its summit, when it suddenly curves back
into a decided posterior hook. (Am. Mus. 1512,
fig. 331.)
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
395
The condyle in M. petersoni is well raised (144 mm.)
above the lower border of the angle and extends 107
millimeters behind ms; it is more extended antero-
posteriorly and less transversely than in M. manteo-
ceras. The angle is very similar to that of Eotitanops
borealis on a larger scale, with a slender and slightly
incm-ved posterior border. The lower border of the
ramus is concave below the coronoid, convex below
the grinders, and rises gradually into a laterally com-
pressed chin gently rounded on the inferior surface.
The ramus increases in depth posteriorly. The
symphysis measures 69 to 80 millimeters, and as seen
from below the chin is sharply contracted to a width of
40 millimeters behind the canines. Below mi the
rami attain in males the width of 20 millimeters. On
the whole this is a progressive development of the E.
borealis type of jaw, the chief difference being the
broader coronoid.
Age characters. — In an aged, somewhat larger, more
elongate, and perhaps more progressive jaw (Am.
Mus. 1567), from the Bridger, there is a wider space
(130 mm.) between the condyle and the posterior
fang of ms, the condyle itself is wider (57
mm.) and less extended anteroposteriorly,
the gentle rounding of the posterior border
of the angle seen in E. borealis and the
typical M. megarJiinus changes into a more
decided, angulate projection of the postero-
inferior border. Seen from behind, the
border of the angle is marked by a sudden
sharp inflection about two-thirds of the
distance below the condyle
occipital condyles measure 98 millimeters transversely,
as compared with 86 in M. petersoni. The width
across the postglenoid processes is 183 millimeters, as
compared with 150 in M. petersoni.
This cranium may possibly belong to an animal
with a dentition such as that which we have referred
below to D. vallidens.
A progressive jaw from Washalcie A. — A specimen
(Am. Mus. 2355) from Washakie A at Glove Springs
consists of the rami incomplete posteriorly, but includ-
ing all the teeth. It belongs to the dolichocephalic
Mesatirhinus series. It is much larger than the most
advanced jaw of M. petersoni from Washakie A. The
measurements are compared below:
Figure 332. — Lower jaw of Mesatirhinus sp. with deciduous dentition
The COndvle One-half natural size. Am. Mus. 12211, Summers Dry Creek, Bridger Basin, Wyo., Bridger C.
is more transversely extended.
A small lower jaw (Am. Mus. 12211), from Bridger
C, has deciduous incisors 1 and 2 and three deciduous
premolars in place, with some of the replacing teeth
below them. The incisors are chisel-shaped, some-
what like the adult incisors of Lambdotherium. The
fourth deciduous premolar is more molariform than
its successor, especially in the somewhat better devel-
opment of the entoconid. The chin is very sloping.
(See fig. 332.)
Mesatirhinus sp.
Large progressive sTcuil (fig. 333). — There is interest-
ing evidence (Princeton Mus. 10041) of the existence
in Washakie B (?) of a much larger animal than M.
petersoni, progressive at least in size toward Doliclio-
rJihius vallidens. It differs from MesatirTiinus superior
in the narrow sagittal crest.
Unfortunately only the occiput is preserved. It ex-
hibits in the parietal profile a pronounced convexity;
the parietal crests are also broadly divergent anteriorly
and rounded, suggesting those of SpTienocoelus.
The superior dimensions are indicated by the follow-
ing comparisons: The occiput measures 90 millimeters
in height, as compared with 80 in M. petersoni; the
Measurements of Mesatirhinus and Dolichorhinus, in millimeters
Pi-ms-
Pz-ma-
Mi-ms.
175
164
103
192
177
112
Washakie B:
D. hyogna-
thus, Prince-
ton Mus.
10273 (type)
240
213
120
The postcanine diastema in this specimen is -long
(32 mm.). As in MesatirTiinus the canines are incurved
as well as recurved.
Mesatirhinus ( = Dolichorhinus) superior Riggs
Reference may be made here to the skull of M.
{ = Dolichor7iinus) superior, which is fully described
below (p. 405). This animal is intermediate in form
between the two genera Mesatirhinus and Dolicho-
rhinus, so that it might be placed in either genus with
equal propriety. The skull and hypocone on m' of
M. superior incline us to connect this skull with species
of Dolichorhinus.
396
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Dolichorhinus Hatcher
Plates XVII, XXIX-XXXII, XLVI, LII, LIII, LV, LXXI-
LXXVII; text figures 27, 33, 105, 110, 125, 133-137, 140
210, 214-220, 254, 255, 302, 322-324, 335-337, 339-353, 483'
508-511, 520, 521, 579-585, 588-591, 647, 661, 685-686, 71l|
724, 733, 737-740, 742, 743, 745
[For original description and type references see p. 177. For skeletal characters see
p. 645]
Generic cTiaraders . — Animals of relatively large
size, extremely dolichocephalic; cephalic indices 43 to
47; face relatively long; faciocephalic index 48 to 51;
postorbital process situated above m'; summit of
cranium broadly flattened; space above small brain
chamber filled with large air sinuses; occiput low and
broad; relatively prominent supraorbital horn swell-
ings on nasals; axis of face and middle part of skull
bent downward. Astragalus of the long-necked type.
Figure 333. — Imperfect cranium of Mesaiirhinus petersoni
One-fourth natural size. Princeton Mus. 10041, Washakie Basin, Wyo., Washakie, i
view; As, occipital view; As, top view; A4, basal view.
DolichorTiinus, appropriately named by Hatcher in
reference to the elongation of the nasal region, is a
titanothere in which we observe the dolichocephalic
extreme. The genotype species, D. Mjognathus , from
Uinta B 2, is connected by -transition forms in Uinta
B 1, such as Dolichorliinus longiceps and especially D.
superior, with the advanced structural stages of Mesa-
iirhinus, namely, M. petersoni, so that there is no
question that DolicliorMnus is a descendant of certain
species of Mesatirhinus. Besides the highly progres-
sive D. hyognathus and the more primitive D. longiceps
there are several species of Dolichorhinus less clearly
defined, a fact which indicates that this was a domi-
nant and highly diversified form during the period of
deposition of the river sandstones and flood-plain
deposits of the levels Washakie B and Uinta B 1 and
B 2 (see below).
History of discovery. — This animal first became
known through Cope's personal exploration of the
Washakie Basin exposures of 1872, which yielded his
cotypes of "Palaeosyops vallidens" ; this species ap-
parently represents a distinct stage of Dolichorhinus,
but unfortunately it is still known only from an im-
perfect lower jaw and some upper teeth. The next
discovery was that of Scott, Osborn, and Speir, of the
Princeton expedition of 1878, consisting of the large
lower jaw which in 1889 Scott and Osborn made the
type of the species "Palaeosyops hyognathus." The
third step was marked by Peterson's discovery in 1894
on behalf of the American Museum of Natural History
of several skulls and parts of skeletons
in the Uinta Basin. These skulls aroused
unusual interest because of the presence of
well-developed horn bases above the eyes,
in reference to which Osborn named the
animals Telmatotherium "cornutum." He
first considered that they represented a
direct progressive transition from "Tel-
matotherium vallidens" {= Manteoceras)
toward the Oligocene titanotheres, but,
as Hatcher pointed out in 1895, the horn
development in these animals is a paral-
lelism rather than a direct approach to
the Oligocene titanotheres, for accom-
panying these horns are other characters
which exclude the animals from such
ancestry. Hatcher accordingly separated
the species as a distinct genus, Doli-
chorhinus. It was long believed that
Dolichorhinus was confined to the Uinta
Basin level B 2, to which Osborn gave
the name Dolichorhinus cornutus zone.
The animal certainly occurs in Uinta B 2,
especially in the river-deposited sandstones,
A. Ai, Side jq very great abundance and may be con-
sidered as the dominant titanothere type
of this deposition because it so far outnumbers all
other types.
The next step in discovery was made by the Ameri-
can Museum expedition of 1906 in the Washakie
Basin, during which Paul Miller found a beautifully
preserved skull and jaws of a Dolichorhinus associated
with parts of the skeleton and specifically identical
with the type of D. cornutus prevailing in Uinta B 2.
This discovery, together with evidence previously
found, demonstrated the synchronism of the Washakie
B 2 and the Uinta B 2 deposits. Further comparison
of the jaws of this Washakie specimen with the type
jaw of "Palaeosyops hyognathus," also from Washakie
B, demonstrated that the species D. cornutus is a syn-
onym of the earlier-described D. hyognathus. Exact
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
397
study and comparison of all these specimens resulted
in the opinion that "Palaeosyops vallidens " also belongs
to the Mesatirhinus-Dolichorhinus group.
Among the crania that were at first included within
the single species D. cornutus are two stages — an earlier
stage, to which the name D. intermedius may be given,
and a later stage, D. hyognathus. Akin to and possibly
to be regarded as "mutations"
of these stages are the species
D. heterodon and D. longiceps
of Douglass.
Geologic range and faunistic
parallels. — The type of Dolicho-
rhinus cornutus { = 'hyognathus)
was found by Peterson in the
sandstone at the top of Uinta
B 2. The genus thus ranges
downward through 700 feet of
deposits to the type locality of
DolicJiorhinus longiceps, repre-
senting a long period of geo-
logic time, in which we should
expect considerable evolution
of structural type as well as
considerable changes in the con-
temporary mammalian life. In
Uinta B 1, for example, Doli-
chorhinus longiceps is associated
with MetarJiinus fluviatilis and
M. riparius. It is noteworthy,
however, that Dolicliorldnus sel-
dom occurs in the same sand-
stone with Mefarhinus, a fact
indicating that these animals
occupied somewhat different
local habitats. It is also a
striking fact that Manteoceras
does not occur at all in Uinta
B 1 or B 2 nor has it been
found in Washakie B, while its
relative MesatirJiinus occurs
quite abundantly. This would
appear to prove that DolicJio-
rhinus, MetarMnus, and Mesa-
tirliinus had different habitats
and habits from either Mante-
oceras or Telmatherium, and
that the conditions existing
during the period of deposition of Uinta B 1 and B 2
were particularly favorable to the preservation of
Dolichorhininae — namely, DolicTiorhinus , MetarMnus,
and Mesatirhinus. Among other ungulates no repre-
sentatives of the Equidae or Tapiridae are found
mingled with the dolichorhines. The hyracodont or
light-limbed division of the rhinoceroses is repre-
sented by rare remains of Triplopus. The amphib-
ious division of the rhinoceroses is represented by
quite abundant remains of Amynodon. Among the
Amblypoda, or giant quadrupeds, Eohasileus is very
abundant and characteristic of the Dolichorhinus zone.
Among the Artiodactyla the ancestral elothere
Achaenodon occurs in the lower levels close to Doli-
FlGTJRE 334.-
Geologic section of the Bridger formation in the Washakie Basin
chorJiinus, and Protelotherium occurs in the uppermost
levels. The giant flesh eaters Mesonyx and Harpago-
lestes are characteristic of this life zone.
In general, the occurrence of the majority of these
dolichorhine titanotheres in river sandstones associated
with the remains of other fiuviatile or river-border types,
such as Amynodon, Achaenodon, and possibly Eohasileus,
398
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
tends to favor the view that the doUchorhines which
frequented the river borders were subject to being
washed into the sandy deposits during periods of flood.
apparently indicating no increase in speed. Dolicho-
rhinus longiceps may be described as dohchocephahc
and brachypodal. (See p. 652.)
Figure 335. — Restoration of Dolichorhinus longiceps
By E. S. Christman, based on the mounted skeleton in the Carnegie Museum. One-fifteenth natural size.
The bodily proportions of the dolichorhines were similar
to those of the existing forest-living pigs of Africa.
Habits of Dolichorhinus. — We may compare Dolicho-
rhinus remotely with Hippidium, an aberrant South
American Pleistocene horse, in which an excessively
The muzzle was rather expanded, or truncate; the
face was not so long as that of other titanotheres.
The incisors were arranged in a semicircle and made
some approach in form to the cropping incisors of the
ruminant. These teeth were also partly cupped to
Figure 336. — Skull and lower jaw of Dolichorhinus hyognaihus
One-Iourth natural size. Skull, Am. Mus. 1851; lower jaw, Am. Mus. 1856. Both from White River, Uinta Basin, Utah, level Uinta B 2.
long skull is combined with exceptionally short meta-
podials, in contrast with those of typical horses.
So far as we can judge from very sparse evidence,
the feet of Dolichorhinus were surprisingly short,
facilitate prehension, as in the lower Miocene species
of the horse. The diastema behind the canine tooth
is longer than in other titanotheres, as in typical
herbivorous forms. The canines in the males were
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
399
moderately long, recurved, sharp edged, and may have
been used in fighting, as by the existing camels; the
offensive power of the front teeth was less, however,
than in the short-jawed Palaeosyops. The cheek
teeth, concerned in the comminution of food, were
relatively long crowned, with pointed cusps, and
constituted an elaborate cutting and triturating appa-
ratus. The movement of the more slender mandible
was partly vertical, partly oblique, since the wearing
of the cheek teeth gives evidence of an oblique shear-
ing action. Adapted to these conditions were the
length and proportions of the chewing muscles and
their angles of action. (See below for details.)
It is therefore probable that since the food evidently
required finer cutting and better trituration than
the food of Palaeosyops, Dolichorhinus was either
a browser on harder materials or a grazer, perhaps
coming out from the forests at night into the open
grassy places or searching for smaller twigs, like
the Indian rhinoceros {R. unicornis). The bend-
ing down of the facial upon the cranial axis is a
characteristic of many grazers, whereas the bend-
ing up of the facial axis is generally characteristic
of browsers.
Directing attention, on the other hand, "to the
progressive backward shifting of the hinder border
of the posterior nares to what is known as the "sec-
ondary palate," Riggs (1912.1, p. 36) has advanced
the hypothesis that DolichorMnus was a river-fre-
quenting form which perhaps fed upon submerged
plants, like the moose. The backward shifting and
closure of the hinder border of the palate is an ob-
vious advantage to animals feeding partly in the
water and is characteristic of many water-living
forms.
General characters of the genotype, D. hyognathus. —
The elongate skull, the broad, flattened, and suture-
less cranial region, the elongate nasofrontal horns
are characters partly of progressive dolichocepha-
ly, partly of family affinity to the Oligocene forms.
The features of the main line of Dolichorhinus are
the extreme narrowing and lengthening of the
skull and zygomatic arches, the convex upward
arching instead of a concave saddle shape of the
skull top, the broad infraorbital shelf, the shal-
low jaws, the parallel series of grinding teeth, and
especially the extremely long, narrow nasals. The
horns are borne chiefly on the nasals, as in Mesatirhi-
nus, in contrast with Manteoceras, in which they are
borne chiefly on the frontals. The occiput is low,
possibly in correlation with the bending down of the
cranium. In palatal view we observe the marked
backward extension of the posterior nares and the
formation of a secondary palate. The jaw is dis-
tinguished by its long, slender, recurved coronoid
process and its depressed angle. These characters
combine to constitute this animal one of the most
peculiar and distinctive of the whole titanothere
At a first glance the long skull suggests that of a
horse, but a closer examination shows that, although
both are dolichocephalic, the resemblance is entirely
superficial; the horse has a primitive short cranium
(brachycrany) and an enormously long face (dolichopy)
or preorbital region. Dolichorhinus has an elongate,
highly modified cranium (dolichocrany) and postor-
bital region and a relatively short face (brachyopy).
As compared in detail with the skull of a horse that of
Dolichorhinus furnishes an instructive minsrlins: of
craruzan.
FiGUBE 337. — Skulls of Dolichorhinus hyognathus (A) and modern
horse (B)
One-sixth natural size. Tliese show analogous and divergent adaptations to grazing
habits. A-A', Basicranial axis; B~B^, basipalatal axis.
convergent resemblances to other long-headed un-
gulates and divergent hereditary differences. Among
the convergent resemblances in Dolichorhinus are
(1) the lengthening of the whole skull, especially of the
face; (2) the bending down of the anterior half of the
skull; (3) the backward prolongation of the palate;
(4) the semicircular or cropping arrangement of the
incisors; (5) the prominence of the orbits; (6) the
forward extension of the masseter muscle, the anterior
slip in Dolichorhinus being attached to the infra-
orbital shelf.
Among the divergent hereditary differences charac-
teristic of the titauotheres and shown in Dolichorhinus
400
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
are (1) the lengthening of the middle part of the skull;
(2) the development of supraorbital horn swellings;
(3) the character of the teeth; (4) the shape of the
4o6
Dolichorhinus
fluminalis
Dolichorhinus
cornufus, type
Sfhenodectes
incisivus, iype
Dolichorhinus
heferodon
Dolichorhinus
hyognathus
(cornufus)
Sfhenodectes
incisivus
Rhadinorh/nus
diploconus
Dolichorhinus
longiceps, type
Dolichorfilnus
superior, type
Mefarhinus
riparius, type
Metarhinus
earlei
Telmalotherium
Dolichorhinus
longiceps
Rhadinorhinus
abboffi
Metarhinus
fluviafiiis, type
Sphenocoelus
uintensis, type
Metarhinus
cristatus, type
(?Dolichorhinus
longiceps)
Metarhinus
riparius
-" DiplojCodoTh
~f£r'~ zone
idZA.
ATrvynodorhskeL.Am.M:us.N9 J933
•nooLorv scutdstoThe^JAmynodon. irvt^rmedius
' yProtelotTierium, uintense
I EohCLSii^LLS
< Stylinodon.
I ProtylopiLs
. EohasUeus-
^°J)_olic/iorkinus
zone
s^tndstone^ Harpct^ol^stes
EobcLsileus uintensis, type
FieldMus. 12170
JHobasileiis
Triplopus
Mesonyjc obtuside/is
? Triplopus
In the comparison of numerous dental series we
observe that the male teeth are somewhat larger,
including the robust, sharp-edged canines, whereas
the female jaws are more slender and the ca-
nines smaller and rounder, with shorter enamel
caps. Sex apparently does not affect the de-
velopment of the osseous horns, which are
practically similar in the male and female
skulls.
Synopsis of species. — The following sum-
mary gives the principal features of the species
assigned to Dolichorhinus:
UPPER LEVELS
D. hyognathus (Osborn) = Z). cornutus Osborn.
Summit of Uinta B 2 and middle of Washakie B 2.
Cranium large, most progressive, broad and convex,
length 550 millimeters, breadth 240, cephalic index 46,
faciocephalic index 53; broad secondary palate; horns
well developed.
D. fluminalis Riggs. Summit of Uinta B 2. Dis-
tinguished by extreme backward prolongation of
secondary palate. Length, type skull, 520 millime-
ters, breadth 233, cephalic index 45, faciocephalic
index 48.
D. intermedius Osborn. Uinta B 2. A broad form
with elongate skull; length 485 millimeters, breadth
223 (estimated), cephalic index 45 (estimated), facio-
cephalic index 49; horns less prominent; secondary
palate less extended posteriori}'.
D. heterodon Douglass. Summit of Uinta B 2.
Similar to D. intermedius. Length 487 millimeters;
cephalic index not determined; faciocephalic index 50.
NO MAMMALS RECORDED
Figure 338. — Geologic section of the Eohasileus-Dolichorhinus and Meta-
rhinus zones in the Uinta Basin
lower jaw; (5) the broadening of the top of the skull;
(6) the relatively short crowns of the grinding series.
As a whole the grinding series is short (206 mm.)
in proportion to the length of the skull, the molar
index being 38. Correlated with molar dolichocephaly
the inner and outer cones of the molar teeth are
closely approximated and the crowns are elongated
and narrowed. In adaptation to harder kinds of
food the crests and cones are elongate or subhypso-
dont; both the parastyles and mesostyles are very
sharp and prominent.
The backward and downward prolongation of
the- bony palate is a very distinctive feature. In
early stages {D. intermedius) the secondary palate is
rudimentary and lies much above the plane of the
primary palate; in later stages it descends and lies
on the same plane as the primary palate, also extend-
ing very far backward (D . fluminalis) .
LOWER LEVELS
D. longiceps Douglass. Base of Uinta B 2. Very
abundant; more primitive; probably ancestral to D.
hyognathus. Horns incipient. Large size, length 545
millimeters, breadth 260; cephalic index of type 47;
faciocephalic index 48; cranial roof narrow, less arched.
D.? vallidens (Cope). Washakie B(?). Imperfectly
known teeth, more primitive than those of D. hyo-
gnathus.
D. (Mesatirhinus) superior (Riggs). Summit of
Umta B 1. Smaller and more primitive. Horns very
rudimentary. No secondary palate.
Summary of cranial indices in Dolichorhinus
Species
Cephalic
index
Faciocephalic
inde.t
D. hyognathus. Am. Mus. 13164, 9 _._
46
53
D. hyognathus. Am. Mus. 1851, 9
(type of Telmatherium cornutum)
43
51
D. intermedius, Am. Mus. 2001
»45
49
D. intermedius, Am. Mus. 1837 (type)__
-41
49
D. heterodon, Carnegie Mus. 2340 '
50
D. fluminalis. Field Mus. 12205 (type)__
45
48
D. longiceps, Carnegie Mus. 2347
47
48
D. longiceps, Am. Mus. 1852, 9
42
48
D. superior. Field Mus. 12188
52
48?
The extremes of these specific indices are also
presented above.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
Dolichorhinus vallidens (Cope)
Plate LXXIV; text figures 95, 341, 353
[For original description and type references see p. 362]
Geologic horizon. — Washakie Basin, Wyo., level
401
Washakie B.
than in D. hyognathus; hypoconulid of m^ elongate; in
general more primitive than either D. intermedins or
D. hyognathus.
As shown above, the imperfectly preserved upper
and lower jaws, the co types of Cope's original descrip-
tion, were not found together. Nevertheless it now
FiGUBB 339. — Skulls showing progressive dolichocephaly in the Mesalirhinus-Dolichorhinus phylum
Side view. Ono-fourth natural size. A, Mesatirhinus petersoni, British Mus. (formerly Am. Mus. 1556), Big Bone Mountain, Henrys
Fork, Bridger Basin, Wyo., Bridger D; B, Dolichorhinus superior, Field Mus. 12188 (type), Uinta Basin, Utah, Uinta B 1; C, D. longiceps,
Carnegie Mus, 2347 (type), Uinta Basin, Uinta B 2; D, D. hyognathus, Am. Mus. 1851, White River, Uinta Basin, Uinta B 2.
Specific characters. — P'-m^ 185 millimeters (esti-
mated); mj-m,,, 123; ectolophs of superior premolars
with a broad basal spreading of the protocone con-
vexities; tritocones more flattened than in D. inter-
medins; p^, p^ of same proportions asm D. hyognathus;
lower premolars less compressed and more primitive
appears probable though not certain that the lectotype
lower jaw (Am. Mus. 5098) and at least one of the
original upper dentitions (Am. Mus. 5097) do pertain
to the same species.
Doubtful reference. — The reference of these types to
Dolichorhinus is provisional; if the jaws are correctly
402
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
referred the cranium is apparently much less dolicho-
cephalic than that of D. hyognathus or D. longiceps.
Comparison with Dolichorhinus superior is also diffi-
cult and unsatisfactory; in D. superior the premolars
appear to be different in contour, also their cingula
are not so heavy; the measurements of the superior
teeth (p'-m^) in these two species are approximately
assigned this animal to Palaeosyops. Osborn at first
regarded it as belonging to the genus Manieoceras but
subsequently recognized the prevailing dolichoce-
phalic characters and placed the animal near Dolicho-
rhinus.
Lectotype lower jaws of D. vallidens (Am. Mus.
5098). — Comparison with typical lower jaws of D.
FiGTJKE 340. — Skulls showing progressive dolichocephaly in the Mesatirhinus-DoUchorhinus
phylum
Top and palatal views. One-eightli natural size. A, Ai, Mesatirhinas petersoni, British lilus. (formerly Am. Mus. 1556),
Big Bone Mountain, Henrys Fork, Bridger Basin, Wyo., Bridger D; B, Bi, Dolichorhinus superior, Field Mus. 12188
(type), Uinta Basin, Utah, Uinta B 1; C, Ci, D. Usngiceps, Carnegie Mus. 2347 (type), Uinta Basin, Uinta B 2; D, Di,
J), hyognathus, Am. Mus. 1851, White River, Uinta Basin, Uinta B 2. pn^. Primary border of the posterior nares;
pn', secondary border of the posterior nares.
the same — 185 millimeters (estimated) in D. vallidens
and 182 in D. superior. The hypocone on m^ prob-
ably absent in D. vallidens, is present and strong in D.
superior. The upper teeth of D. vallidens (paratype)
are structurally ancestral to those of Diplacodon, but
so also are the upper teeth of Mesatirhinus petersoni.
History. — The species was at first referred by Cope
(1885.1, p. 700) to the genus Palaeosyops. Earle also
hyognathus from both Washakie B 2 and Uinta B 2
shows that D. vallidens was a smaller animal and some-
what more primitive in the details of the lower pre-
molars. (See fig. 353.)
The chin is only partially preserved and with it the
root of the right canine, which is stouter than in sup-
posed females of D. hyognathus. The first lower pre-
molar, as indicated by the alveolus in Cope's drawing
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
403
(1885.1, pi. 52, fig. 3), appears to have had but one root.
P2 is shorter anteroposteriorly and less compressed
than in the typical D. hyognathus; its posterior V is
also smaller as compared with the protoconid and less
sharply developed. P3 is not preserved. P4 is also
less compressed, the posterior V lower and more
primitive in form than in D. hyognaihus. The true
molar series is considerably shorter, but the posterior
half of ms and especially the hypoconulid are relatively
longer and more compressed. The space between ms
and the ascending ramus was less. Comparative
measurements are as follows :
Measurements of Dolichorhinus vallidens and D. longiceps? , in
millimeters
Front of canine to hinder border of ma
Front of pi to hinder border of mj
Length of true molar series
P2, ap. by tr
P4, ap. by tr
M3 (anterior lobe) , ap. by tr
M3, length of hypoconulid
D. vallidens,
Am. Mus.
6098, lecto-
type jaw
230
146
123
19X11
23X13
55X20
14
D. longiceps?.
Am. Mus.
1852
275
165
139
23X11
25X15
60X23
15
The specimen under consideration is distinguished
from jaws of M. manteoceras by the longer molar
series and more elongate hypoconulid on ms.
Upper teeth of the paratype of Dolichorhinus vallidens
(Am. Mus. 5097). — The characters of the premolar
ectolophs are so constant in all the many specimens of
D. hyognathus that the marked differences which they
present in the paratype of D. vallidens, approaching as
they do the characters of the Mesatirhinus premolars,
appear to establish the specific separation.
The whole series of upper grinding teeth (p'-m^) of
D. vallidens is estimated at 185 millimeters, as compared
with 177 in D. intermedins and 206 in D. hyognathus.
Comparison with Dolichorhinus hyognathus. — As
noted above, the superior grinding series is shorter
than that of D. hyognathus (185 mm. (estimated), as
compared with 206), and the detailed anteroposterior
and transverse measurements of the crowns of the only
perfectly preserved teeth, p^, p'*, are practically identi-
cal with those of the average D. hyognathus, as shown
below :
Measurements of upper premolars in species of Dolichorhinus, in
millimeters
P^ anteroposterior
P^ transverse
P^, anteroposterior
P^i transverse
P', internal lobe, anteroposterior
P*, internal lobe, anteroposterior
D. inter-
raedius,
Am. IMus.
1837 (type)
D. valli-
dens. Am.
Mus. 5097
(paratype)
13
15
9
10
18
20
20
22
16
19
19
22
15
9.8
20
22
21
23
The linear ectoloph measurements of the true molars
are intermediate between those of D. intermedins
and D. hyognathus. The ectolophs of the premolars
of D. vallidens (fig. 341, B) afford the most distinctive
character — namely, the broad festoon and the basal
spreading of the convexities of the protocone, a primi-
tive character which relates these teeth to the Mesati-
rhinus stage. The deuterocones of p^"* are more
D
A
Figure 341. — Upper premolars of Mesatirhinus, Dolichorhinus,
and Metarhinus
Outer side view. Natural size. A, Mesatirhimis petersoni, British Mus. (formerly
Am. Mus. lo5G), Big Bone Mountain, Henrys Fork, Bridger Basin, Wyo.,
Bridger D; B, Dolichorhinus vallidens, Am. Mus. 5097 (paratype), reversed,
Mammoth Buttes, Bitter Creek, Washakie Basin, Wyo., Washakie B; C,
DolicliorJiinus hyognathus, Am. Mus. 1850, White River, Uinta Basin, Utah,
Uinta B 2; D, Metarhinus fiuviatilis. Am. Mus. 1946, White River, Uinta Basin,
Utah, Uinta B 1.
primitive and the tritocones are more flattened than
in' the D. intermedius type.
These characters tend to show that so far as indi-
cated by the paratype D. vallidens is somewhat more
primitive than either D. intermedius or D. hyognathus.
Comparison with M. petersoni and other forms. — The
superior grindiag teeth of this paratype resemble
those of certain specimens of M. petersoni on a larger
scale. The progressive distinctions are (a) the pres-
ence of a cingulum on the inner side of p'; (6) the
quite complete cingulum on the inner side of p^ and
p* and the somewhat more flattened and elevated
ectolophs of p^~^, which are less elevated, however,
404
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
than in D. Tiyognathus; (c) the external cingula of
the molar teeth are a shade more prominent; [d) the
protoconule and metaconule have nearly vanished, al-
though inconspicuous vestiges still persist.
The faint rudiments of the tetartocone folds of the
premolars are less marked than in certain specimens
of M. petersoni. The ectolophs of the superior
premolars are readily distinguished from those of
M. manteoceras by the flattening of the tritocones.
The ectoloph of p^ is simple and sharply convex.
That of p- consists of a prominent protocone convexity
which spreads downward into a broad cingulum bound-
ing the base of the tritocone (the same region in
D. Jiyognathus is strongly constricted); the tritocone
ectoloph is nearly flat or very slightly convex. In
p' the protocone ectoloph is a convex ridge spreading
toward the base into a cingulum, while the tritocone
ectoloph is very gently convex but without a rib.
P* has the convexity opposite the apex of the proto-
cone, while the outer face of the tritocone is slightly
convex and the basal cingulum is nearly continuous
across the ectoloph. The above-mentioned features
enable us to distinguish the upper premolars from
those of Manteoceras and of both D. intermedius
and D. Tiyognathus.
Summary. — In the lectotype lower jaw of D.
rallidens the premolars are decidedly more primitive
than in D. hyognathus. In the paratype upper
dentition of D. vallidens the premolars are somewhat
more primitive than in D. hyognathus. It is thus
not certain that the lectotype and paratype belong to
precisely the same stage of evolution; but, on the
other hand, there is no evidence that they are specifi-
cally distinct.
Comparative measurements, in millimeters, showing progressive proportions of skull and teeth of Dolichorhinus
Premaxillaries to
condyles
End of nasals to
middle top of occi-
put
Face, anteroposterior.
Cranium, anteropos-
terior
Transverse zygomata
Pi-m3
P2-m3
P2-p<
Mi-m3
P', ap. by tr
P^, ap. by tr
M', ap. by tr
M2, ap. by tr
M3, ap. by tr
D. supe-
rior,
Field
Mus.
12188
(type)
224
184
I), intermedius
Am.
Mus.
1837
(type)
475
230
236
'190
179
165
57
109
Am.
Mus.
2001
475
241
248
'■223
177
164
57
105
12X10
19X2120X25
31X32| 30X?
39X38
36X37 39X39
D. val-
lidens,
Am.
Mus.
5097
D. het-
erodon,
Carne-
gie
Mus.
2340
(type)
487
492
245
240
189
173
114
.22X25
.133X32
.|42X4I
.[39X38
D.flu-
minalis.
Field
Mus.
12205
(type)
520
233
171
105
D."lon
giceps,"
Carne-
gie
Mus.
2347
(type)
590
°270
-285
-264
197
178
60
115
15X10
D.lon-
giceps?.
Am.
Mus.
1852
573
270
283
230
202
■185
62
118
15X11
21X2920X26
35X?36X33
39X40J41X39
39X?i40X37
D. hyognathus
Am.
Mus.
13164
580
'290
250
205
186
65
119
18X11
23X29
36X33
43X42;
42X42
Am.
Mus.
1850
593
'260
320
208
185
61
120
16X9
22X25
36X30
41X38
43X35
Am.
Mus,
1851
550
570
280
288
»240
208
187
65
122
15X10
23X27
34X34
43X43
45X?
Am.
Mus.
1845
290
215
'565
Field
Mus.
12167
(D.
'cornu-
tus")
310
131
25X30|.
35X33|.
47X41.
47X41.
285
214
135
575
270
298
231
206
186
62
123
15X10
22X27
35X33
42X40
44X39
' Estimated.
12188. Uinta B 1.
1837. Female. Uinta B 2.
2001. Uinta B 2.
5097. Washakie B.
2340. Uinta B 2 (upper level).
12205. Uinta B 2.
2347. Uinta B 2 (low level).
1852. Female. Uinta B 2.
13164. Washakie B 2.
1850. Male. Uinta B 2.
1851. Female. Uinta B 2 (type of
Telmatotherium cornutum) .
1845. Uinta B 2.
1848. Uinta B 2.
12167. Uinta B 2.
The above table shows the dolichocephalic propor-
tions of the cranium proper and of the true molars
and the smaller dimensions of D. intermedius and the
intermediate proportions of the type of D. heterodon.
The type of D. longiceps and the type of D. "cornutus"
agree well in size with the skulls of D. hyognathus.
The well-preserved skull from the Washakie Basin
(Am. Mus. 13164), which is referred to D. hyognathus,
does not differ greatly in measurements from the type
of Dolichorhinus "cornutus" and the other Uinta B
specimens. The skulls of D. hyognathus, from Uinta
B, show a considerable difference in size, ranging from
the relatively small skull No. 1852 to the very large
skull No. 1845.
Measurements of the lower jaws of these species are
given on page 416.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
405
Dolichorhinus (Mesatirhinus) superior Riggs
Plates LXXV-LXXVII; text figures 137, 339, 340
[For original description and type references see p. 190]
Type locality and geologic Tiorizon. — Uinta Basin,
Utah, MetarMnus zone, top of the " Metarhinus sand-
stones," summit of Uinta B 1. D. superior comes
from a horizon 200 to 400 feet above that at which D.
longiceps is commonly found. The type was found
in the same ledge and associated with one of the more
advanced stages of MetarMnus {M. earlei). Thus D.
superior is contemporary with a more advanced stage
of development of DolicliorTiinus and with the last of
the Metarhinus phylum (Riggs).
Specific characters. — "Skull 485 by 255 millimeters,
molar series 182 millimeters, nasals free to a point over
last premolar, infraorbital process present, arches
slender anteriorly, nasals infolded at margins, sagittal
area expanded, canines small, p^ and p^ oblique to
axis of series. Molars relatively small, strong hypo-
cone on m^, posterior nares opening opposite the
anterior margin of last molar." (Riggs, 1912.1, p. 26.)
Materials. — The only specimen known is the type
skull in the Field Museum (No. 12188), described
below. This important form is transitional between
Mesatirhinus and Dolichorhinus. On the whole it
appears to be a primitive species of the genus Doli-
chorhinus. The original description by Riggs is as
follows :
This genus [Mesatirhinus], reported for the first time from
the Uinta formations, is apparently indigenous to the Bridger
and Washakie Basins. It is represented in the Field Museum
collections by a single specimen — an incomplete skull collected
by Mr. J. B. Abbott from the top of the Metarhinus sandstones
near gilsonite vein No. 2. The right arch is wanting, together
with the basioccipital and condyles. The dentition is anatomi-
cally complete excepting the incisors.
The skull presents striking similarities with the earlier
representatives of Dolichorhinus. From the dorsal view, the
nasals, facial, and supracranial regions appear very similar,
though the cranial region does not have the pronounced down-
ward curve characteristic of Dolichorhinus. In the palatal
view more marked differences are noticeable. The premolars
are more primitive, the molars smaller, and the posterior narial
opening is unmodified. In these characteristics the specimen
in hand resembles D. heterodon ^' from upper Uinta B more
closely. However, it differs from that species in having a
strong hypocone on the last molar and in the whole facial
profile. In our present knowledge of these many closely
related forms, this species may be regarded as the largest and
most highly specialized representative of Mesatirhinus.
This animal occurs geologically at the very summit
of Uinta B 1 (upper A of Riggs), fully 300 feet above
the first occurrence of Dolichorhinus longiceps. This
fact is important, because otherwise it would certainly
be considered the direct ancestor of Dolichorhinus ,
since it affords a complete structural transition to this
genus, as shown in the comparative outlines displayed
in Figure 339. This is another very interesting in-
'8 Douglass, Earl, Carnegie Mus. Annals, vol. 6, p. 310, 1910.
stance of the survival of a primitive stage side by side
with a progressive stage. We have an analogy in
existing nature in the survival of the hippopotami
of Liberia and the Nile regions of Africa, namely,
H. liheriensis and H. amphihius, the former extremely
primitive, the latter rather progressive.
Although the profile and the top views (figs. 339,
340) of the cranium of D. superior are closely similar
to those of D. longiceps, the palatal view is less similar
because of the entire lack of the secondary palate,
which in its various stages of development is so
characteristic of Dolichorhinus. In D. superior,
moreover, the horn cores are even more rudimentary
than in D. longiceps. There is a wide orbital-nasal
area, and a sharp downward curve of the nasals.
The species is also related to M. petersoni in its cephalic
index, which is 52 as compared with 47 in D. longi-
ceps— in other words, the skull is less dolichocephalic
than that of the typical Dolichorhinus.
The opening of the posterior nares is opposite the
margin of the second molar tooth, or in the same posi-
tion as the primary nares of Dolichorhinus. The
crowns of the molar teeth are somewhat shorter or
more brachyodont than in Dolichorhinus. The molar
cephalic index, or ratio of the length of the grinding
series to basilar length of skull, is estimated as 38, the
same as in D. hyognathus.
Dolichorhinus intermedius Osborn
Plate LXXIII; text figures 125, 342, 343
[For original description and type references see p. 184]
Type locality and geologic horizon. — Uinta Basin,
Utah; Eohasileus-Dolichorhinus zone (Uinta B 2).
Specific characters. — As compared with D. hyo-
gnathus, of inferior size; p'-m^ 179 millimeters; m'"',
109; length, premaxillaries to condyles, 462; trans-
verse zygomata, 190 (estimated); cephalic index of
type 41, of paratype 45; faciocephalic index 49.
Secondary palate present btit less developed than in
D. hyognathus; infraorbital shelf of malar relatively
narrow; premolars less progressive with subconic
deuterocones; all cingula less robust; nasals more
pointed or less expanded distally.
This species when described in 1908 was regarded
by Osborn as a structural ancestral stage, or ascending
mutation toward the typical D. hyognathus. It now
appears to be a dwarfed and somewhat more primi-
tive form, which thus coincides in some of its char-
acters with D. longiceps (the true ancestor of D.
hyognathus) except that the horn bases appear to be
more distinct. It might perhaps be regarded as a side
or dwarfed phylum related to or identical with the
D. heterodon of Riggs.
Materials. — The type is the skull Am. Mus. 1837,
representing the main characters of this species.
Another skull (Am. Mus. 2001) is somewhat less
typical. These skulls are recorded from Uinta B 2.
406
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The more exact level, however, is not stated. Only
from thehr less progressive condition does it appear
probable that they belong to a somewhat lower
geologic stage than the typical D. Jiyognathus.
Proportions. — These animals are smaller than those
referred to D. Jiyognathus. The type skull belongs to
a young adult female with canines proportioned as in
the females of the type species. The total length
(462 mm.) is somewhat inferior to that of the older
animal (Am. Mus. 2001), in which the length is
485 millimeters, as compared with an average of 550
in D. Jiyognathus. Similarly the superior grinding
series measures 179
millimeters, as com-
pared with 156 in
Mesatirhinus peter-
soni and 206 in D.
hyognathus.
Comparison with
D. (cornutus) hyo-
gnathus.— The crania
are of inferior di-
mensions throughout.
The nasals are nar-
rower anteriorly; the
horns are less prom-
inent and are borne
entirely on the nasal
bones; the flattened
vertex of the skull in
the parieto-occipital
region is relatively
narrow.
The incisors, as
shown by i', pre-
served in Am. Mus.
2001,are deeply pitted
or pocketed, posteri-
orly. P' is a small,
simple tooth, less
broadened anteriorly
than in any speci-
mens of D. hyogna-
thus. The common
characters as com-
pared with B. hyo-
1837 (type), White River, Uinta Basin, Utati, gnathuS in P^-p'' are :
(1) The crowns are
less hypsodont throughout; (2) the protocone con-
vexities on the ectolophs are more convex or less
sharply constricted; (3) the tritocone convexities, on
the other hand, are somewhat more prominent; (4)
the deuterocones are more rounded or conic, and there
is less prominence of the internal cingula and of the
protoconules. All these characters indicate a lesser
degree of progression.
Additional note on Doliehorhinus intermedius. — Three
skulls in the Carnegie Museum from Uinta B 2
(middle levels), Nos. 3094, 3095, 3096, collected by
Doctor Douglass, are referred to this species. The
principal dimensions of two of these are given below:
Measurements of skulls of Doliehorhinus intermedius, in milli-
meters
Figure 342. — Skull of Doliehorhinus
intermedius
Top view. One-tourth natural size. Am. Mus.
1837 (ty:
Uinta B 2.
Pmx-condyles
Transverse zygomata (estimated)
Pi-m3
Pi-p*
M"-m3
472
170
176
70
103
462
Doliehorhinus longiceps Douglass
Plates XXX-XXXII, LXXIII, LXXV-LXXVII; te.xt figures
135, 136, 254, 335, 339, 340, 343-346, 353, 589-591, 724
[For original description and type references see p. 188. For sjceletal characters see
p. 651]
Type locality and geologic horizon. — Uinta Basin,
Utah; Eobasileus-Dolichorhinus zone (Uinta B 2).
Geologic range 300 to 400 feet.
Specific characters. — Horn bases small; cranium
large, 530-550 millimeters; breadth, 264-240; cephalic
index 44-47; secondary palate in early stage of
development, lying above level of primary palate;
cranial vertex narrow posteriorly; premolars with
relatively feeble internal cingula.
Geologic distribution. — The type skull of this prim-
itive and clearly defined species, recorded by
Douglass as "700 feet below Uinta red beds," would
place the type well down in Uinta B 2. The four
skuUs (Field Mus. 12175, 12176, 12193, 12200)
collected by Riggs extend from the lower to the upper
portion of Uinta B 2 or the "upper Metarhinus beds"
of Riggs. These specimens are somewhat smaller
and less specialized than the type; they vary in length
from 525 to 560 miUimeters.
Type.-^^he. type skuU of Douglass has been dis-
torted from right to left and from above downward,
so that the left upper part is tilted and overhangs
the left temporal fossa and orbits. The right pre-
maxillo-maxillary rostrum is flat, and the general
wearing plane of the left tooth row is tUted toward
the right, while the parietofrontal vertex above the
squamosal region is squeezed up into a long antero-
posterior convexity.
This distortion makes it difficult to determine what
are the real structural differences from D. hyognathus,
but the judgment of Douglass in separating this species
is fully confirmed by the skuUs discovered by Riggs
in Uinta B 1 and B 2. One skuU (Am. Mus. 1852),
presumably that of a female in regard both to measure-
ments and to characters, appears to bridge over the
differences between this species and the type of
D. cornutus { = JiyognatJius), as shown in the following
measurements :
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
407
Measurements of Dolichorhinus longiceps and D. hyognalhus, in
millimeters
Tip of nasals to occipital crest
(lateral)
Premaxillary to condyle
Transverse zygomata
Face, anteroposterior (premaxil-
lary tp postorbital process,
frontal)
Cranium, anteroposterior (post-
orbital frontal to condyle)
Pi-m3
Pi-p<
M'-mS
P', ap. by tr
P^ ap. bytr
M', ap. by tr
M2, ap. by tr
M', ap. by tr
D. longi-
ceps, Car-
negie Mus.
2347 (type)
590
'555
264
"270
■■285
197
79
115
15X11
21X29
35X ?
39X40
39X ?
D. hyognnttius
Am. Mus
1852 (a
transitional
form)
'550
230
270
283
202
78
118
20X26
36X33
41X39
40X37
Am. Mus.
1851 (typo
of Tel-
matotlie-
rirua
cornutum)
577
550
'240
280
288
208
82
122
15X10
23X27
34X34
43X43
45X45
Thus these measurements indicate that as com-
pared with the type of T. cornutum { = D. Jiyognathus)
the type of D. longiceps is somewhat broader and its
tooth dimensions sUghtly less, except that p** is wider.
Field Museum sJculls. — The four skulls as described
by Riggs (1912.1, p. 33) are somewhat smaller, less
specialized than the type, and range in length from
525 to 560 millimeters. One of the largest, a finely
preserved skull, is shown in Plate LXXVI. There is
little evidence of incipient horn cores. The nasals
overhang the margins of the premaxillaries, which are
somewhat narrower than in the type of Douglass.
Compared with the type of D. intermedins, the smaller
D. longiceps skull (Field Mus. 12193) approaches
closely in size; the dental series is similar in length;
the premolars are more advanced in structure. In
the palate there is a ridge between m^ and m^ corre-
sponding to the primitive position of the posterior
narial border, which is bridged over by the outgrowth
of thinner plates from the lateral margin of the palatal
bones so that the nares have receded to a point behind
the hamular processes of the pterygoids; the plates
of this secondary palate are, however, so thin that
they are often broken through, so that the secondary
border of the posterior nares can not be precisely
determined. The secondary palate in this species is
pierced by a pair of foramina; its posterior extension
is an enfoliate process free from the lateral walls and
probably attached to the inferior margins of the vomer
(Riggs).
A mandible associated with the incomplete skull of
D. longiceps (Field Mus. 12200) is relatively strong,
curved in the ramus, and broad at the angle. The
skeleton of this same specimen, which was found near
the base of Uinta B 1, is described on page 651.
The detailed measurements of these skulls are given
in the following table:
Measurements of DolicJiorJiinus iy Riggs, in millimeters
D. longi-
ceps, Car-
negie Mus.
2347 (type)
D. cornu-
tus. Field
Mus. 12167
D. flumi-
nalis, Field
Mus. 12205
(type)
Skull:
Length, incisors to condyles
Breadth across arches
Breadth above orbits ;
Postorbital process to condyles
Last molar to condyles
Length of free nasals
Greatest breadth of nasals
Postglenoids to condyles (median line) .
Length of molar-premolar series
Length of molar series
Length of crown of canine
Diameter of crown of canine
Length of diastema
Narrowest point in sagittal area
Breadth of orbitonasal area
Mandible:
Length, condyles to incisors
Height, condyles above angle
Length of molar-premolar series
Length of molar series
Length of crown of canine
Diameter of crown of canine
Depth of ramus from base of ps
Depth of ramus froiia base of ma
'545
260
595
285
'310
300
152
81
197
115
140
214
135
40
22
16
550
247
'134
305
263
170
79
122
212
131
24
17
14
52
'560
255
130
305
280
160
530
240
'535
295
258
132
264
259
520
233
116
137
57
121
200
124
115
192
121
67
114
198
122
171
105
32
18
400
159
209
123
29
16
70
57
» Estimated.
101959— 29— VOL 1-
408
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 343. — Skulls of Dolichorhinus
From White River, Uinta Basin, Utah, level Uinta B 2. One-fourth natural size. A, D. iniermedius, Am. Mus. 1837 (type), reversed; B,
D. heterodon, Carnegie Mus. 2340 (type), reversed; C, D, longkeps, Carnegie Mus. 2347 (type), "from the lowest level at which fossils
were found in horizon B" (Uinta B 2).
Measurements of Dolichorhinus heterodon, D. longiceps, and D.
hyognathus, in millimeters
Pmx-condyles
Transverse zygomata
Mi-m'
P^ ap. by tr
M', ap. by tr
M2, ap. by tr
M3, ap. by tr
D. hetero-
don,
Carnegie
Mus. 2340
(type)
487
240
114
22X25
33X32
42X41
39X38
"485
«225
119
21X26
32X33
42X42
42X42
Carnegie
Mus. 2347
(type)
"555
»264
115
21X29
35X ?
39X40
39X ?
D. hyogna-
thus. Am.
Mus. 1851
(type of
Telmato-
therium
cornutnm)
550
-240
122
23X27
34X34
43X43
45X ?
Additional observations on Dolichorhinus longiceps. —
A skull in the Carnegie Museum (No. 2865) referred
by Mr. Peterson to D. longiceps is associated with a
complete fore limb and other parts of the skeleton.
It was found at a low level in Uinta B 1. It differs
from the type of D. heterodon in having a larger m^;
it appears to be smaller than the type of D. longiceps
in skull dimensions but somewhat larger in the
second and third upper molars.
Mr. Peterson's description (1914.3) of this skull,
with the mandible and hyoid bones, is in substance as
follows :
The specimen (No. 2865) consists of the greater portion of
the skull, the posterior part of the mandible of the left and frag-
ments of the right side, the hyoid arch, the cervical vertebrae,
two dorsal and two lumbar vetebrae, together with the fore
limb and foot practically complete.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
409
Cranium and mandible. — The cranium is somewhat smaller
than in the type of Dolichorhinus longiceps, the sagittal area of
the parietals is more compressed laterally, the zygomatic
portion of the squamosal is slenderer and less expanded laterally,
and the basicranial axis has a greater bend. These characters
together with the slightly larger teeth constitute the most
marked differences in the two crania compared, but that they
should be regarded as of specific value is rather questionable.
The base of the skull has received some crushing fore and aft,
a fact to which the greater curvature of the basicranial axis
may partly be due.
The sudden downward bend of the occiput of Dolichorhinus
heierodon, the flatter frontal region, the smaller preorbital
ledge, and the smaller and more delicate nasals seem to separate
that species more widely from the present specimen. Further-
more, the difference in the geological horizons in which D.
heierodon and the present specimen were found is to be consid-
ered. The former came from horizon "Lower C, " while the
latter was found in the lower part of horizon "Upper A" of
the Uinta sediments.
The high coronoid process and its sudden backward turn at
the top, so characteristic of the mandible of Dolichorhinus, is
well shown in this specimen. The angle is much compressed
laterally, the temporal fossa is located high up but is quite
deep, and the horizontal ramus has but small vertical diameter.
Measurements Milli-
meters
Length of skull from anterior border of the orbit to top of
occiput 365
Anteroposterior diameter of upper molar series 125
Transverse diameter of frontals at postorbital processes 145
Depth of mandible at ms 71
Length of stylohyal, approximately 168
Anteroposterior diameter of basihyal, median line 15
Hyoid arch. — The hyoid arch may best be compared with
that of the tapir, because in that genus there is apparently no
extended anterior appendix or process such as is seen on the
basihyal of the horse or the rhinoceros. However, the bone as
a whole, especially its anterior border, is relatively heavier than
in the tapir. The thyrohyal is unfortunately broken off on
both sides. This element was perhaps relatively less developed
than in Tapirus ierrestris. The ceratohyal is also unfortunately
broken off at the upper end, but its length was no doubt pro-
portionately equal to that of the American tapir, while the
shaft is less constricted anteroposteriorly. The epihyal is not
present; this bone no doubt was nodular in character, as is the
case in Tapirus terrestris. The anterior portion of the shaft of
the stylohyal is rounder in cross section than in the tapir or
the horse, but the upper end is flattened and terminates in
enlarged processes, the superior attached to the hyoidial portion
of the temporal bone and the inferior somewhat more obtusely
rounded, extending downward and outward. This riblike
upper end of the stylohyal is more suggestive of the rhinoceros
or the horse than of the tapir. (See figs. 344 and 345.)
Dolichorhinus hyognathus (Osborn)
[Telmatherium cornutus Osborn]
Plates XVII, XLVI, LII, LIII, LV, LXXI, LXXII; text
figures 27, 33, 105, 110, 215, 217-219, 254, 255, 302, 336, 337,
339-341, 346-353, 483, 511, 520, 521, 579, 580, 582-585, 588,
647, 661, 686, 737, 743, 745
[For original description and type references see pp. 169, 173. For skeletal characters
see p. 645]
Type locality and geologic Tiorizon. — Uinta Basin,
Utah; summit of Eohasileus-DolicTiorliinus zone (Uinta
B 2). This animal is very abundant within its known
geologic range through the upper 200 feet of Uinta B 2.
The type specimen of D. cornutus ( = 7iyognathus) and
most of the crania in the American Museum collection
were found by Peterson in the upper or " Amynodon
sandstones," at the summit of Uinta B 2, but the
animal has also been recorded by Peterson 150 feet
below the summit of B 2. It is not thus far recorded
in Uinta C A single specimen has been found in the
Washakie Basin, Wyo., on the 180-foot level of Washa-
kie B 2.
Specific characters. — Skulls large, 550 by 240 to 595
by 285 millimeters; relatively narrow, cephalic indices
46 to 43; face relatively long, faciocephalic index 53
to 51; grinding series p'-m^, average 206 millimeters;
molar-cephalic index 38; horn cores very prominent;
face decidedly bent down on cranium — that is, cypto-
cephalic; secondary palate broad and nearly on the
same plane with the primary palate; premolar ecto-
lophs more hypsodont ; premolar protocone convexities
sharply ridged; molars with prominent cones and
crests; vestigial protoconules; hypocones of m^ very
distinct.
Figure 344. — Skull referred to Dolichorhinus longiceps? , side
and top views
One-sixth natural size. After Peterson. Carnegie Mus. 2865.
This animal, from Uinta B 2, represents the most
advanced stage known of this series. It appears to be
a progressive descendant of D. longiceps from the base
of Uinta B 1. The crania are somewhat larger in all
dimensions than those of D. longiceps or D. inter-
medius, and the parieto-occipital vertex is broader.
Synonymy. — The male jaw, type of D. hyognathus
(Princeton Mus. 10273), was found by the Princeton
expedition of 1878 in Washakie B 2. When compared
with the female skull and jaw (Am. Mus. 13164)
found at the 185-foot level of Washakie B 2, it can
not be separated specifically. Thus they must both
be referred to D. hyognathus. These specimens in
turn closely resemble in form and measurement the
females in Uinta B 2 which were first referred to D.
cornutus. Thus D. cornutus can not be separated
specifically from D. hyognathus.
410
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Materials.— 1:^0 other Eocene titanothere is repre-
sented by so rich cranial material as this species.
The numerous skulls and jaws, although uniformly
recorded from the DolichorTiinus (cornutus) Tiyognathus
zone (Uinta B 2 and Washakie B 2), were undoubtedly
Figure 345. — Hyoid apparatus of Dolichorhinus longiceps?
(1, 3) compared with that of a modern tapir, Tapirus ter-
restris (2)
One-haU natural size. After Peterson. The two upper figures show a front view,
the three lower a side view. This almost unique fossil belongs with the skull and
jaws previously figured (fig. 34-1) and other bones comprising the specimen Car-
negie Mus. 2865. bh, Basihyal; th, thyrohyal; ch, ceratohyal; eh, epihyal; sh,
stylohyal. Compare the hyoid bones of Brontops sp. (Am. Mus. 518, fig. 425).
found at somewhat different levels and may represent
different stages of mutative progression, although it
seems impracticable to separate them into species.
These crania are enumerated below.
Washakie B 2:
Princeton Mus. 10273, type of D. hyognaihus; jaw of an
aged animal.
Am. Mus. 13164, skull and jaws of a young adult female;
m' slightly worn, associated with parts of skeleton, fore
limb, and parts of vertebrae.
Uinta B 2:
Am. Mus. 1850, skull of a young adult male; m^ just appear-
ing.
Am. Mus. 1845, skull of a young adult; m^ in place but
unworn.
Am. Mus. 1851, skull (type of Telmatotherium cornutum)
of aged female; m^ well worn.
Am. Mus. 1852, skull of young female; m^ slightly worn.
Am. Mus. 1848, very old skull; m^ greatly worn, sex inde-
terminate.
Am. Mus. 1843, anterior part of skuU of large size, asso-
ciated with complete backbone and humerus.
Am. Mus. 1849, parts of skull and fragments of skeleton
doubtfully recorded as of base of Uinta B 2.
Field Mus. "l2167, skull from Uinta B (1 or 2).
Jaws. — Besides the type of D. hyognaihus from
Washakie B 2, we have the jaws Am. Mus. 13164,
associated with a skull. Also from Uinta B 2 we have
thirteen jaws more or less completely preserved, in-
cluding Am. Mus. 1834, 1836, 1840, 1852, 1854, 1855,
1856, 1857, 1858, 1941, 2008.
STcull. — The afiinities of the skull of these animals
to that of Mesatirhinus petersoni are apparent in
many details of structure, but, as above noted, there is
a very marked progressive advance, which is bridged
over partly by the intermediate stages of D. inter-
medius and D. longiceps. The affinity to M. ( = Doli-
chorhinus) superior is still closer. The full descrip-
tion which follows is based principally on the female
skull Am. Mus. 1851 (type of Telmatotherium cornu-
tum) and the superb skull and jaws, also of a female
(Am. Mus. 13164), from Washakie B 2.
In the superior aspect (fig. 346) we are immediately
struck by the extraordinary elongation of the nasals,
which extend behind the line between the orbits and
occupy a little less than one-half of the entire length of
the skull. The longitudinal suture persists between
the nasals and is traceable a short distance back
between the frontals. These bones expand to 129
millimeters immediately above the orbits and are
convex both anteroposteriorly and transversely; the
line of junction between the frontals and parietals is
obliterated. The vertex is here arched both trans-
versely and longitudinally. The supratemporal ridges,
now wholly lateral in position, follow the superior
border of the supratemporal fossa; the top of the
cranium is expanded slightly to 108 millimeters and
then contracts to 62 millimeters just in front of the
junction with the occipitals. This flattened arching
and spreading of the vertex of the skull naturally differs
both according to age or growth and the progressive
stage of evolution which the skull represents. The
superior view also displays the comparatively long
and slender zygomatic arches, which reach a maximum
width of 245 millimeters and an average width of 231,
as compared with 550, the total length of the skull.
Horns. — The nasals diverge suddenly into the
osseous horns, which lie directly above the orbits,
whereas in M. manteoceras the horns lie in front of the
orbits; they present an outward-directed elongate-
oval convexity, to the posterolateral portion of which
only the frontals contribute. On the vertex between
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
411
the horns are gentle longitudinal valleys separated by
median longitudinal convexities; the horn bases thus
actually rise decidedly above the surrounding surfaces
and overhang the orbits. A biologic fact of interest is
that the horns appear to be developed as strongly in
the female as in the male skulls and are not at this
stage a distinctively sexual character; in none of the
crania are they distinctly rugose, as in some of the
male crania of M. manteoceras . Under these horn
swellings, which are 127 millimeters apart, the nasals
nent paired eminences, as in Am. Mus. 13164. The
extreme elongation of the posterior nares is unique
among perissodactyls. The pterygoid plates of the
alisphenoid are elongate and depressed on either side
of the long and narrow postnarial depression. The
palatines do not crowd into the postnarial space as in
M. manteoceras. The infraorbital malar plates con-
stitute a very prominent shelf, the anterior part of
which is shown by the sutures to be composed of the
maxillaries. To this prominent infraorbital shelf
Figure 346. — Skulls of Dolichorhinus
One-fourth natural size. A, D. longkcps, Carnegie Mus. 2347 (type), Uinta Basin, Utah, Uinta B
B, B. hyognatlius, Am. Mus. 1851, White Eiver, Uinta Basin, Utah, Uinta B 2.
narrow to 66 millimeters, then broaden again to 76 at
the widest point near their extremities.
Palatal aspect. — As seen from below (fig. 347), the
elongation of the palate, in which the palatine and
maxillary plates take about equal share, is a most
striking feature. The posterior nares open behind
. m^. A kind of secondary palatal plate is formed by
the backward and upward extension of the dorsal
surface of the palatine. In this compressed post-
narial chamber the maxilloturbinals appear as promi-
was probably attached an anterior slip of the masseter
muscle, as in many other mammals with weak zygo-
mata. Behind these projections the malars are seen
to present a long and comparatively narrow edge.
The lacrimals are larger and have a broader extension
on the face than in any other species. The lacrimal
tubercle is preserved in one skull, as in the Palaeosyops
series.
Among the most striking results of progressive
dolichocephaly are those seen in the conformation of
412
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
the articular facets for the condyles of the jaw. The
postglenoid processes are narrow and tuberous; the
glenoid facets are not transverse in position but
directed obliquely forward, as if their outer borders
were drawn out anteriorly by a stretching of the skull.
In M. manteoceras these glenoidal facets are more
directly transverse. The whole basicranial region is
Vplam.ms.plj
VlpfJUJO
Figure 347. — Skull of Dolichorhinus hyo-
gnathus
Palatal view. One-fourth natural size. Am. Mus. 1S51
(type of Telmaiotherium cornutum Osborn). Supple-
mentary details from Am. Mus. 1845. Both specimens
from White Kiver, Uinta Basin, Utah; Uinta B 2.
correspondingly elongate, the bridge of bone separat-
ing the foramen ovale and the foramen lacerum medium
now measures 42 millimeters as compared with 27 in
M. petersoni or 17 in the brachycephalic LimnoTiyops
laticeps.
The anterior aspect of the skull (fig. 348) exhibits
the extraordinarily long premaxillo-maxillary rostrum,
which extends horizontally backward into the floor of
the narial chamber, the total length being 144 milli-
meters. The infraorbital shelves are well shown.
The infraorbital foramina are deep and narrow; the
nasals are vertically decurved at the sides, so that
they form three sides of a square. The projection of
the nasal horns, although the animal is a female, is
admirably illustrated.
The lateral aspect of the skull (fig. 349) shows that
the midcranial concavity characteristic of Manteo-
ceras and the Oligocene titanotheres is replaced in
this species by the prominent frontoparietal convexity,
the lowest point of the cranium in the upper profile
being above the orbits between the horns. The suture
between the maxillary and the nasals is similar in
form to that in related species, the nasofrontal junc-
tion being above the orbit. The failure of the frontals
to send forward a spur overlapping the enlarged nasals
is well shown. A broadly concave space (68 mm.)
separates the narial notch and the orbit, and the
infraorbital foramen issues 33 to 40 millimeters in
front of the orbit. The maxillaries contribute to the
anterior portion of the infraorbital process, the chief
convexity being formed by the malars. Below the
orbits the malars are gently concave, as in the related
species of this genus. Other characteristic dolicho-
cephalic features are the limited vertical extent of the
zygomatic portion of the squamosal, the great fore
and aft thickening of the postglenoid processes, and
the widely open external auditory meatus.
The occipital view (fig. 348), best shown in Am. Mus.
1845, is highly characteristic. The height of the occi-
put, 142 millimeters from the basioccipital to the
supraoccipital crest, approximately equals the width
across the middle of the occiput. Above the foramen
magnum two prominent ridges diverge and terminate
in tuberous convexities in the upper lateral portions
of the crest.
The interior structure of the skull, including that of
the narial and cranial cavities, is shown in Figure 254.
Dentition in general. — The grinding teeth are more
or less perfectly preserved in most of the crania and
jaws; three specimens afford a complete knowledge
of the upper and lower cutting teeth.
Incisors. — The superior series has a semicircular
arrangement (Am. Mus. 1851); the median pair are
separated by a considerable diastema (12 mm.). The
incisors increase in size regularly from i^ to i'. They
exhibit convex anterior faces, more flattened posterior
faces, with a median convex ridge. The posterior
cingulum rises to form a distinct cup in i^ less marked
in i^ and i'. P is fully incisiform (in contrast to its
caniniform shape in the contemporary Telmatherium) ;
it exhibits a narrow antero-external cingulum besides
the postero-internal, obliquely sloping cingulum; it
differs from other incisors in its more elevated crown.
A narrow diastema (9 mm.) separates i' from the
canine.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
413
The inferior incisors are also arranged in semicircular
series, have obtusely pointed crowns and evenly convex
anterior faces, slightly recurved concave posterior
faces, and a pronounced median rib, which expands
into the basal cingulum. The transverse measure-
ments of the broadest part of the crown in Am. Mus.
ap. 18 mm., tr. 15) the anterior and posterior ridges
are less prominent, but a large, obtuse posterior
basal cingulum is observed. The lower canines are
more obliquely placed and recurved than in T. cul-
tridens. There is a conspicuous antero-internal cingu-
lum, but the lingual side of the base of the crown is
pglscf.
•p. fy. s^
FiGUKE 348. — Skulls of Dolichorhinus hyognathus
Front and occipital views. One-fourth natural size. A, Am. Mus. 1851 (type of Telmatotherium cornutum Osborn), front view. B, Am. Mus.
1815, occipital view. Both specimens from White River, Uinta Basin, Utah; Uinta B 2.
1856, a female, are respectively i', 14 millimeters;
i^, 17; i', 16. The entire breadth of these teeth in
this specimen is 72 millimeters.
Canines. — The sexes are sharply distinguished by
the size of the lower canines : in the males the enameled
crown of the tusks measures vertically 41 millimeters,
smooth in the middle basal portion only; elsewhere it
is cingulate. The posterior cingular ridge is slightly
less acute than in T. cultridens.
In the upper canines also the difference between
the sexes is sharply marked, the male tusks in Am.
Mus. 1850 measuring (ap. by tr.) 24 by 20 millime-
FiGURE 349. — Skull of Dolichorhinus hyognathus
One-fourth natural size. Am. Mus 1851 (type of Telmatotherium cornutum Osborn). White Eiver, Uinta Basin, Utah, Uinta
B 2. X, y, Section lines in Figure 255.
in the females only 27. The fine male tusks pre-
served in Am. Mus. 1850 (ap. 24 mm., tr. 21) are
laterally compressed, with sharply defined antero-
internal ridges and somewhat less prominent posterior
cutting ridges. No internal basal cingulum is observed
in this specimen. In the female tusks (No. 1856,
ters and having a vertical height of 42, whereas the
female tusks of the type (Am. Mus. 1851) measure 18
by 15 and have an estimated crown height of 27.
Upper premolar-molar series. — The distinctive ecto-
loph characters of p^-p* are a sharp vertical protocone
rib or ridge, a flattened to gently convex tritocone, and
414
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
a well-elevated ectoloph. In general this series is long,
averaging 202 to 208 millimeters in length. The true
molars are very strongly dolichocephalic in certain
specimens (as in Am. Mus. 1850), in which the antero-
posterior considerably exceeds the transverse measure-
ment of each tooth. In other specimens, however
(as in Am. Mus. 1851, type), the anteroposterior and
transverse diameters are more nearly subequal. In
most specimens the external cingulum is sharply
defined, although there is considerable variation in
this respect also. The ectoloph cusps and internal
cones are subhypsodont, or elevated (paracone 34
mm. high, protocone 23 mm.). The internal cusps
(the deuterocones of the premolars and the proto-
cones and hypocones of the molars) throughout the
series are rounded at the apices and are decidedly
Figure 350. — Upper incisors and canines of
DoKchorhinns hyognathus
One-half natural size. A, Am. Mus. 1S51, White River,
Uinta Basin, Utah, Uinta B 2, side view; B, Am. jMus.
1845, White Biver, Utah, Uinta B 2, crown view.
convex internally as compared with those in the
Telmatherium phylum. Vestigial protoconules appear
in the molars of Am. Mus. 1850. Small, more or less
cingulate hypocones appear variably in m^ A well-
defined and sometimes broad internal cingulum extends
around the lingual side of the premolars in the more
progressive specimens (Am. Mus. 1850, 1851, 1852).
P^-p* in crown view appear more subcircular in outline
than in Telmatherium.
Premolars. — Of the superior premolars, p' is sepa-
rated by a slight diastema (15 mm.) from the canine
and is a bifanged, narrow, laterally compressed tooth
(ap. 16 mm., tr. 9) with convex buccal and more concave
lingual faces. The posterobasal lobe is becoming well
defined. P^ is more subcircular than in M. petersoni,
its proportions being 20 by 19 millimeters, a condition
which is due to the more anterior position of the deu-
terocone. The protocone and tritocone are subequal,
, but the protocone is much more prominent externally.
Figure 351. — Lower incisors and
canines of Dolichorhimis hyogna-
thus
One-half natural size. Am. Mus. 1856, White
River, Uinta Basin, Utah, Uinta B 2; crown
view.
the tritocone being stdl nearly flat. A rudimentary
protoconule is observed (Am. Mus. 1850). In p'
(ap. 20 mm., tr. 21) the inner portion of the crown is
broader, a rudimentary protoconule is seen, a
slight spur foreshadowing the tetartocone extends
back from the deuterocone, the protocone exhibits a
narrow but sharply convex buccal face, the trito-
cone is slightly convex, with a basal cingulum. In p^
the ectoloph rises to 21 millimeters, the external
cingulum is more continuous, and the inner side of the
crown is relatively broader, the crown measuring 22
by 24 millimeters. In
these premolars (p',
p*) the tetartocone
rudiments are of the
faintest character; in
well-worn teeth they
are not perceptible at
all. 1>. hyognathus has
less progressive tetar-
tocones than T. ulti-
mum but is very highly
specialized in respect
to the peculiar sub-
circular form of p^-p^.
D. hyognathus presents a considerable advance
beyond D. vallidens and some advance beyond D.
intermedius in the elevation of the ectoloph as a whole;
the increased symmetry of the protocones and trito-
cones, especially in p^, which is a much more progres-
sive tooth than in D. vallidens; and the more nearly
subcircular form of p^-p^.
A specific dolichocephalic feature of the inferior pre-
molars is the spacing of pi in the midst of the long dia-
stema between the canine and p2, as
seen especially in the type of D.
hyognathus, in which this diastema
measures 52 millimeters, the diastema
in front of pi measuring 25 and
that behind 14. These diastemata
naturally increase as the individuals
advance in age, and they are affected
by individual growth, by the stage of
evolution, and by the sex. In the fe-
male (Am. Mus. 1856) the total dia-
stema between the canine and p2 is
42 millimeters, nearly in the center
of which lies pi. The lower pre-
molars are well preserved in Am. Mus. 1856, from
which the following descriptions and measurements
are taken: Pi is a laterally compressed tooth, measur-
ing (ap. by tr.) 15 by 8 millimeters, with the
posterobasal lobe well defined and bearing a dis-
tinct cusp. In p2 (ap. 23 mm., tr. 11) the protocone,
which is somewhat less prominent (16 mm.), rela-
tively, exhibits the antero-internal concavity and a
rudiment of the antero-internal cusp ( = paraconid);
the posterior lobe (=talomd) is more distinctly of the
Figure 352.-Left
upper canine of
Dolichorhinus
hyognathus
One-half natural size.
Am. Mus. 1850; ex-
ternal view.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
415
molar type and has a rudimentary fold analogous to
the metastyhd. In ps (ap. 23 mm., tr. 12) still further
progress is made, there being
quite a decided antero-in-
ternal valley and a well-
defined posterior basin. In
P4 (ap. 24 mm., tr. 15) we
find a submolariform tooth
including a high metaconid
and rudimentary paraconid
but lacking the distinct en-
toconid.
Molars . — The superior
molar series varies in length
from 118 to 131 millime-
ters and in addition to the
characters enumerated
above exhibits a very broad
and prominent antero-inter-
nal extension of the cingu-
lum, the crown of m^ being
broader in front (43 mm.
through parastyle) than it
is behind (32 mm. through
metastyle) . In some speci-
mens (Am. Mus. 1852) the
hypocone of m' rises as a
small but sharp and distinct
cusp but is not so prominent
as in the type of Rhadino-
rhinus diploconus. In other
specimens (Am. Mus. 1851)
it is less prominent and more
cingulate. The proportions
of the molars are given in
detail in the table (p. 416).
The inferior molar series
measures 138 millimeters in
the female (Am. Mus. 1856).
A very distinctive feature is
the infolding of the external
cingulum between the outer
lobes of the tooth, which is,
however, less mai-ked in the
type than in most other spec-
imens. In the Princeton
type of D. Tiyognatlius
(Princeton Mus. 10273) the
only molar preserved is mg,
which measures (ap. by tr.)
64 by 25 millimeters, agree-
ing almost exactly with
specimens in the American
Museum. There is no proof
of sexual inferiority in the
female grinding teeth (ma, ap. 62 mm., tr. 28), and
we should not expect it, because the females require
as much food as the males or more. In the type mg
there are rudimentary folds on the inner valleys, and
the hypoconulid has a crenulate internal cingulum.
Figure 353. — Lower jaws of Dolichorhinus
A, D. valUdens, Am. Mus. 5098, one of the cotypes of " Palaeosyops " mllidens Cope, here regarded as the lectotype; Mammoth
Buttes, Bitter Creek, Washakie Basin, Wyo.; Washakie B?. B, D. hyognaOim, Am. Mus. 1S56; White Eiver, Uinta
Basin, Utah, Uinta B 2; ooronoid from Am. Mus. 1852 {Z>. Jonskeps?), White River, Uinta Basin, Utah, Uinta B 2. C, D.
hyogmthus, Princeton Mus. 10273 (type); White Kiver, Uinta Basin, Utah, Uinta B 2; coronoid and angle restored from
Am. Mus. 1852 (D. longiceps?) . A one-half natural size; B and C one-fourth natural size.
In the more perfectly preserved mg of a female speci-
men (Am. Mus. 1856) the external cingulum is very
feebly marked except opposite the external valleys.
416
TITAJSrOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The internal valleys are more prominent and con-
spicuous, and the hypoconulid is reduplicate. In the
more worn m2 of the same specimen (ap. 43 mm., tr.
25) these internal valley folds have been worn away,
and the same is true of the still smaller mj (ap. 35
mm., tr. 20). The external cingulum is more sharply
accented in some specimens (as in Am. Mus. 1855)
than in others.
Jaw of DolichorTiinus hyognathus. — The somewhat
fractured and crushed type jaw (Princeton Mus.
10273) presents the distinctively long and shallow
ramus and shallow sloping symphysis that are char-
acteristic of this species. (See fig. 353.) The measure-
ments given below serve to determine the variations
due to age, to sex, and in a measure to the progressive
evolution of this type.
Measurements oj Dolichorhinus hyognathus, in millimeters
Washakie B 2
Am. Mus.
13164, V
Princeton
Mus. 10273,
<? (type)
Am. Mus.
1862, old ?
Am. Mus.
1856, 9
Condyle to incisive border
Depth, condyle to angle
Depth, ramus behind pz
Depth, ramus behind m2
Depth, ramus behind ma
Length of symphysis
Least width of chin
Free height of coronoid
Thickness of ramus below m2
Vertical thickness of symphysis posteriori}'.
Canine to ms
Pi-ms
Pj-ms
Mi-ma
Transverse diameter ofms
448
+ 135
58
69
74
109
59
445
65
79
95
115
57
65
75
95
114
62
■112
61
102
55
65
32
268
230
213
138
26
■300
"246
'218
•137
27
26
37
276
233
210
142?
27
270
223
194
127
(?)
430
■148
54
60
79
117
55
29
30
274
238
207
138
27
We observe that the specimen from Washakie B
(Am. Mus. 13164), a female, is practically similar in
its measurements to the females found in Uinta B 2.
(Am. Mus. 1852 may belong to the species D. longi-
ceps.)
The most perfectly preserved rami are those of the
females Am. Mus. 13164 (Washakie B) and 1856, from
which Figure 353 B is taken.
The symphysis is greatly elongated (114 mm.) in the
type of D. hyognathus. The ramus increases very
gradually in depth from 50 millimeters in front of p?
to 85 in front of ms, being of moderate thickness
(26 mm.) In the female jaw Am. Mus. 13164 (from
Washakie B) the angle is produced moderately down-
ward and backward, the condyle not being so greatly
raised (138 mm.) above the angle as in the brachy-
cephalic types. The most distinctive feature is the
coronoid, which is produced backward so as to over-
hang both condyle and angle.
Dolichorhinus heterodon Douglass
Plate LXXIII; text figures 133, 134, 343
[For original description and type references see p. 187]
Type locality and geologic horizon. — Uinta Basin,
Utah; horizon Uinta B 2 or C 1 (Douglass).
Specific characters. — A small, aberrant form, pos-
sibly related to D. intermedius; distinguished by some-
what larger size, p'-m^ 189 millimeters, faciocephalic
index 50. Secondary palate above level of primary
palate; large occipital condyles; premolars and molars
with heavy internal and external cingula; a distinct
mesostyle and a prominent parastyle on p*; premolar
ectolophs very oblique. Agreeing with D. intermedius
in general skull and tooth characters and especially in
the marked anteroposterior convexity of the parietal
vertex.
Materials. — Represented by the type specimen only,
in the Carnegie Museum (No. 2340). As shown by
the accompanying measurements this type represents
an animal distinctly smaller than D. hyognathus and
closely allied to D. intermedius, of which it may be a
somewhat more progressive successor.
Among its primitive features is the elevation of the
secondary palate above the plane of the primary
palate. The secondary palate is present in the type
specimen but has not yet grown do^vnward near the
horizontal plane of the primary palate as in D. cornutus.
Among its progressive features distinguishing it from
D. intermedius are the heavier internal and external
cingula of the upper premolars and molars, the well-
developed mesostyle on p*, and the prominent para-
style on p', p*. A very rudimentary mesostyle is
observed on p^ which is a rare feature among titano-
theres, as ordinarily mesostyles are not developed in
the premolars.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
417
Comparative measurements of Dolichorhinus, in millimeters
Skull:
Pmx to condyles
End of nasals to middle top
of occiput
Face, anteroposterior (pmx
to postorbital frontal)
Cranium, anteroposterior
(postorbital frontal to con-
dyles)
Transverse zygomata
Dentition:
Pi-m'
P'-p*
Mi-m3
P', ap. by tr
P^, ap. by tr
P', ap. by tr
P*, ap. by tr
Ml,, ap. bytr
M2, ap. by tr
MS, ap. by tr
heterodon,
Carnegie
II us. 2340
(type)
487
492
245
240
189
76
114
17X17
20X20
22X25
33X32
42X41
39X38
D. intermedius
Am. Mus. Am. Mus.
2001 1837 (type)
"485
473
247
250
"225
177
72
105
12X10
16X17
19X20
21X ?
31X ?
37X39
38X38
463
476
227
236
190
180
72
109
16X16
17X20
19X21
32X ?
40X38
35X37
Dolichorhinus fluminalis Riggs, 1912
Plates LXXV-LXXVII; text figure 140
[For original description and type references see p. 191]
Type locality and geologic horizon. — Uinta Basin,
Utah; "Amynodon sandstones," summit of Eohasileus-
Dolichorhinus zone (Uinta B 2).
Specific characters. — Extreme backward extension of
secondary palate. Skull rather small, 520 by 233
millimeters, cephalic index 45; faciocephalic index 48;
molar-premolar series 171 millimeters; molar cephalic
index 36. Nasals narrow and slightly tapering;
posterior nares open between hamular processes;
postorbital processes of jugal back of the last molar;
incipient horn cores in the form of narrow ridges.
Molar series relatively short; canines short and
recurved.
Materials. — The only specimen known is the type
skull in the Field Museum (No. 12205).
Specific relations. — The high geologic level, namely,
the summit of Uinta B 2, is to be especially noted —
that is, this animal occurs contemporaneously with the
most advanced specimen of D. hyognathus. Like D.
heterodon it appears to be related as a much more pro-
gressive form to D. intermedius, with which it agrees
in the angulate form of the narial recess (which is
rounded in D. hyognathus), in the tapering nasals, in
the position of the postorbital processes of the jugal
behind m^ (which process is in front of m^ in D.
hyognathus). The face is relatively short, the facio-
cephalic index being 48; in i*. hyognathus it is 53 to 51.
The molar series is relatively short, the index being
36, while in D. hyognathus it is 38. Moreover, the
skull and dentition is smaller than in D. hyognathus and
is highly specialized in the extension of the secondary
palate, the broad occipital vertex, straight tooth row,
and extreme dolichocephaly of the basicranium.
In his original description Riggs observes (1912.1,
p. 36):
D. fluminalis displays a high degree of specialization in the
postnarial characters. The nares are bridged over so as to oblit-
erate almost all evidence of their primary position. In this
process the opening has receded to a point back of the hamular
processes. This recession, noted in less degree in other species,
is evidence of a secondary adaptation to aquatic habits of
feeding. Other characters of the skull in this species would
not indicate that this animal was aquatic in its general habits.
Like many other terrestrial mammals it probably fed upon sub-
merged plants. The slenderness and delicate modeling of
the skull would suggest an animal lighter of limb and more
active than other species of this genus. In the development of
horns the type of this species is more advanced than the type
specimen of D. cornutus [hyognathus]. Its narrower sagittal
area, its strongly recurved canines, and much smaller molars
readily distinguish it from that species.
Sphenocoelus Osborn
Text figures 111, 354
[For original description and type references see p. 174]
Sphenocoelus appears to be referable to the sub-
family Dolichorhininae. This animal may be an
aberrant derivative of Mesatirhinus, but it is clearly
distinguished from Dolichorhinus. The type skull of
Sphenocoelus, so far as preserved, resembles that of
Metarhinus riparius Riggs in general form and in
many details. Sphenocoelus may therefore be closely
allied to that type.
Geologic horizon. — Uinta B 1.
Characters. — Of extreme dolichocephalic type. Base
of cranium with the basisphenoid laterally compressed
to afford space for a pair of bony pits in the roof of the
pharynx; a sessile sagittal crest; occiput low; occipital
condyles excessively broad; glenoid facets oblique as
in Dolichorhinus; external auditory meatus widely
open inferiorly.
Geologic distribution. — The geologic level of the
only loiown specimen (the type) is the middle portion
of Uinta B 1 or "upper Metarhinus zone" of Riggs,
where remains of this animal are found associated with
Dolichorhinus longiceps but especially with the Meta-
rhinus fluviatilis type. The type specimen (Am. Mus.
1501), consisting of the posterior part of the skull
only, affords further proof of the wide adaptive radia-
tion of the titanotheres. Although our present
knowledge is confined to the posterior half of the type
skull, it appears that this animal, although aberrant,
probably belongs not far from the Mesatirhinus-
Dolichorhinus phylum.
The name Sphenocoelus was applied to this form by
Osborn because of the presence of two cavities in the
alisphenoid bones on each side of the basisphenoid,
418
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
which thus appears to form a septum between two
elongate oval cavities, whose function is unknown.
It may be conjectured that they lodged diverticula
of the epithelial lining of the oral cavity or that the
"pits" were mere antra, or hollow spaces, the outer
A4t
Figure 354. — Skull of Sphenocoelus uintensis
One-fourth natural size.
m. Mus. 1501 (type); Wagon Hound Bend, White Eiver, CTinta Basin, Utah, Uinta B I.
Ai, Top view; A2, basal view; A3, occipital view; A4, side view.
wall of which became extremely thin or even partly
incomplete. Similar thin-walled or partly open cavi-
ties occur on either side of the basisphenoid in certain
specimens of the two-toed sloth {Chohej^us hoffmanni) .
When this peculiar skull was first described the
relationship of the animal to which it belonged was
entirely uncertain. Now, after more detailed study
of the Eocene titanothere skull in general and especially
of the skull of Dolichorhinus, it appears that Spheno-
coelus is certainly an aberrant titanothere and not a
chalicothere (suborder Ancylopoda), as at first sug-
gested. The chief consid-
erations in favor of this
view are the following: (1)
The large alisphenoid canal
relates it to the Perissodac-
tyla; (2) the wide space
between the foramen ovale
and foramen lacerum medi-
um (40 mm.) removes it
from affinity with all other
perissodactyl families ex-
cept the Brontotheriidae
and Equidae; (3) the basi-
cranial region and especially
the elongate, oblique form
of the glenoid facets for the
lower jaw is strikingly sim-
ilar to that of Dolichorhinus
hyognathus, although the
postglenoid processes are
somewhat different in form ;
(4) the detailed relations of
the foramen rotundum,
alisphenoid canal, foramen
ovale, tympanic fossa, fora-
men condylare, as well as
the form of the petrosal
bone and of the paroccipital
processes, point in the same
direction; (5) the form of
the temporal, sagittal, and
occipital crests, the arrange-
ment of the venous postpa-
rietal foramina, the spacing
of the postglenoid and post-
tympanic processes (see top
view in fig. 354), are all
clearly foreshadowed in the
brain case Princeton Mus.
10041 (figs. 713, 716), which
is provisionally regarded as
a very progressive Mesati-
rhinus.
Generic distinctions. —
Mingled with all these
points of resemblance to
Dolichorhinus are strildng
generic differences, such as the pair of pits already
mentioned, the short sagittal crest diverging into raised
temporal ridges, the excessively large and broad con-
dyles, and, as indicated in the widely separated postgle-
noid and post-tympanic processes, an extreme degree
of dolichocephaly.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
419
Sphenocoelus uintensis Osborn
Text figures 111, 354
[For original description and type references see p. 175]
Type locality and geologic horizon. — Wagon Hound
Bend, White River, Uinta Basin, Utah; Metarhinus
zone (Uinta B 1).
Specific characters. — Occipital condyles of striking
breadth (131 mm.); occiput of moderate
height (130 mm.), of considerable breadth
(117 mm.); width across zygomatic arches
(230 mm.) about the same as in D.
hyognathus. The sagittal crest short (89
mm.) and thin (8 mm.) in posterior cranial
region. Diverging, rounded supratemporal
ridges.
Materials. — The only specimen known
is the type skull in the American Museum
(No. 1501). The most striking fact about
Sphenocoelvs is that although more special-
ized in certain points than Dolichorhinus
it is geologically older than D. hyognathus,
as it comes from Uinta B 1. Sphenocoelus
may at present be regarded as a descendant
of a certain type of Mesatirhinus (such
as Princeton Mus. 10041, a brain case)
which retained the narrow occipital crest
but otherwise paralleled Dolichorhinus.
The sphenoid pit measurements are ap-
proximately as follows: Length 41 milli-
meters; width 13; depth 22. They are
distinctly roofed over dorsally with bone.
The function of these basicranial pits is
entirely conjectural. Nothing similar has
been observed in other Perissodactyla.
The occipital condyles exceed in width
(131 mm.) those of even the very broad
type of i>. hyognathus. The glenoid facets
exhibit a long, obliquely transverse exten-
sion, so characteristic oiD. hyognathus; the
postgienoid processes also have a peculiar
obliquity. In front and to the outer sides
of the postglenoids the squamosals are
deeply concave.
Conclusion. — The peculiar pits remove
this animal from any other known genus or
species; it is possible that they are not con-
stant characters. The various distinctive
characters, while somewhat extreme, ap-
pear to be paralleled or foreshadowed in
the member of the Mesatirhinus-Dolichorhinus series.
Eometarhinus Osborn
Text figures 156, 355
[For original description and type reference see p. 200]
This recently discovered Eometarhinus is recog-
nized as ancestral to Metarhinus and is thus the ear-
liest known member of the Metarhinus phylum.
Geologic horizon. — The type specimen was found 205
feet below the top of the Huerfano formation, in
Huerfano B {Eometarhinus-Trogosus zone).
Generic chai^acters. — Small; ancestral to Metarhinus;
with rudimentary frontonasal horn; nasals elongate;
overhanging premaxillaries, decurved as in Metarhinus;
no infraorbital shelf; characters apparently interme-
FiGUEE 355. — Type skull of Eometarhinus huerfanensis, from Huerfano B
One-half natural size. A, nasals, top view; Ai, anterior nasal sections; A2, posterior nasal sections; B,
0, palatal 1
side 1
' witli crown view of dentition.
1 rudiments.
diate between those of the Metarhinus and Mesati-
rhinus phyla.
Type species. — Eometarhinus huerfanensis. (See
below.)
Original description. — Osborn writes (1919):
This new genus and species from the upper Huerfano is
founded upon the anterior portion of a skull (Am. IMus. 17412)
420
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
representing an animal widely distinct in aU its characters
from the contemporary P. foniinalis and more closely related
to the other group of middle Eocene titanotheres to which
Metarhinus, Mesaiirhinus, and Dolichorhinus belong. The most
surprising character in an animal of this geologic antiquity is
the very rudimentary osseous horns at the junction of the
nasals and frontals, indicating the horn rudiment, a very ancient
character in this phylum. The nasals are long, arched, de-
curved, and revolute on lateral borders, thus resembling the
rhadinorhinal type in the metarhine group. The malars below
the orbits are prominent. There was probably no infraorbital
shelf, as in Rhadinorhinus. The comparative measurements of
Eometarhinus, of Mesaiirhinus megarhinus, and of the contem-
porary Palaeosyops foniinalis are shown below.
Eometarhinus huerfanensis Osborn
Text figures 156, 355
[For original description and type reference see p. 200]
Type locality and geologic horizon. — The type speci-
men was found near the Huerfano-Muddy divide, 3
miles west of Gardner, in the Huerfano Basin, Colo.,
205 feet below the top of the Huerfano formation, in
the EometarJiinus-Palaeosyops foniinalis zone (Huer-
fano B).
Specific characters. — Inferior in all measurements to
M. megarhinus. Premolars with small deuterocone.
Pl-m^ 124 miUimeters; p^-p*, 53; ml-m^ 72.
Materials. — This species is known from the type
skuU (Am. Mus. 17412) and from two referred
specimens — a fragment of a lower jaw (Am. Mus.
17013) with the first and second molars preserved
(Osborn, 1919.494, fig. 7, B), and the right and left
fourth lower premolars with a fragment of a canine
(Am. Mus. 17416). Both of these specimens are from
approximately the same level as the type and from
the same general locality. Doubtfully referred to this
species are three fragmentary upper molars (Am.
Mus. 17415), found 3 miles north of Gardner, on the
lowest level of the upper Huerfano (Huerfano B), or
200 to 300 feet below the level of the type.
General characters. — The principal skull characters
are noted above under the generic description. The,
dental formiila is normal. The premolars are small,
apparently very simple in pattern. The antero-
posterior diameters of the molars appear to exceed
the transverse; as they are in fractured condition, no
accurate measurements can be taken. Apparently a
hypocone on m^
Measurements of Eometarhinus, Mesaiirhinus, and Palaeosyops,
in millimeters
Eometartiinus,
Am. Mus.
17412
Mesatirhinus
megarhinus,
Am. Mus.
12202
Palaeosyops
fontinalis,
Am. Mus.
17425
Pi-ms
P2-m3
Mi-m3
?■•, anteroposterior
124
109
72
14
»18
21
147
133
83
17.5
23
25
26. 5
28
31
"146
"131
83
16.5
21. 5
M', anteroposterior
M', transverse
23
26
M3, anteroposterior
M^, transverse
25.5
29
34
A tibia, found in association with the type skull,
measures 275 millimeters (estimated) in extreme
length.
Metarhinus Osborn
Plates LII, LXXI, LXXIV, LXXVIII-LXXX; text figures
123, 124, 138, 139, 219, 302, 323, 324, 341, 356-361, 404, 407,
509, 521, 522, 573-578, 647, 745.
[For original description and type references see p. 183]
Animals of medium or small size, divergent in
structure, and probably different in habit and habitat
from members of the Mesatirhinus-DolicJiorhinus series,
hence the name Metarhinus; perhaps of fluviatile or
semiaquatic habit; skull with rudimentary horns,
elongate expanding nasals, orbits prominent; opposite
sides of the upper jaw firmly united, proportions
mesaticephalic; persistent sagittal crest and narrow
occipital condyles.
Geologic horizon. — This group of small animals is at
present found only on one geologic level — namely, Uinta
B 1 and Washakie B 1, which may be Icnown as the
Metarhinus zone. It originally sprang from the same
ancestral stock as Mesatirhinus megarhinus, but in
these "metarhines" dolichocephaly was arrested and
mesaticephaly persisted. The four or five known
species exhibit a considerable number of characters in
common which afford ground for regarding them as a
related natural group of the smallest titanotheres of
the period; they are truly dwarfed forms. The skull
is very broad across the orbits, which are notably
prominent; immediately in front of the orbits the face
contracts; the nasals are long and expand distally.
There is some evidence that the phylum divides into
two lines, consisting of broad-headed and narrow-
headed forms that run parallel throughout the period
represented by Uinta B 1.
The remains of these animals are very abundant,
and the genera Metarhinus and Dolichorhinus are
equally well represented (Riggs, 1912.1, p. 24) in
Uinta B 1. The remains are usually associated in
the same ledges and were apparently deposited under
the same conditions. In one ledge of sandstone a
large part of an articulated skeleton of D. longiceps
was so mingled with the skeleton of a young Meta-
rhinus that it was at first mistaken for a young animal
of that genus. In other ledges, however, Metarhinus
is very abundant and Dolichorhinus is absent. Other
genera occurring in the same life zone are the more
rare Rhadinorhinus, the giant amblypod Eohasileus,
the small hyracodont Triplopus, and two large creo-
donts, Mesonyx and Earpagolestes. Occasionally
crocodiles and numerous turtles are found, also beds
of fresh-water clams. Reeds, leaves, and branches of
trees are abundant in the upper sandstones of the
Metarhinus zone.
Geologic occurrence in channels. — It is evident that
our knowledge of this Metarhinus zone fauna is con-
fined to that of the intrusive sandstone ledges of stream
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITA.NOTHERES
421
origin, because fossils are rarely found in the shales
which alternate with these ledges (Riggs, 1912.1, p. 24).
Skulls are often embedded in the semigravelly layers
and have their narial or orbital cavities filled with
pebbles which could be carried only by rapidly
flowing water. Another evidence of stream action
lies in the complete dissociation of the various
skeletal elements; seldom are the lower jaws associated
with the skulls, or so many as two vertebrae found
articulated. In the exceptional instances where the
whole skeletons are but little disturbed they are found
embedded in the fine-grained homogeneous sandstone
apparently deposited in more quiet water, such as
deep pools or eddies.
This prevailing mode of occurrence supports Os-
born's theory that these animals were of semiaquatic
or fluviatile habits, as indicated by tne name given to
the type species of the genus, M. fluviatilis.
Generic characters. — Small titanotheres, basilar
length 355 to 415 millimeters. Persistently mesatice-
phalic; postcanine diastema short; orbits very promi-
nent, with projecting infraorbital shelves; nasals
expanding distally; narial opening deeply recessed
at sides; premaxillary symphysis greatly elongated;
snout moderately broad; rudimentary horns on fronto-
nasal suture; sagittal crest high and thin. Incisor
teeth small, cingulate; canine teeth small, pointed,
recurved; grinding teeth subhypsodont; pz'emolars
rather progressive; hypocone of m^ present or absent;
hypoconulid of ms small, conic.
Historical notes. — The type species of this genus was
recognized by Osborn (1908.318) from a specimen
found in Uinta B 1 , which had been confused previously
with Mesatirhinus megarhinus. This is a very small
animal, perhaps the primitive member of the series.
At the same time Osborn recognized in Washakie B a
second species, M. earlei, distinguished from M. fluvia-
tilis by its much greater size and the lesser prominence
of the orbits. After the successful expedition of
1910 in the Uinta Basin, Riggs (1912.1) added M.
cristatus, an animal of intermediate size but of the
same proportions as M. fluviatilis, also M. ripa?-ius,
an animal of larger size with an apparently longer
and narrower skull.
Osborn placed in the genus MefarJiinus the species
M. diploconus, which Riggs (1912.1) on excellent
groimds removed to the new genus RhadinorMnus.
It is possible that the Palaeosyops Junius of Leidy,
from Bridger B, a very diminutive form, represents
an ancestral form of this metarhine phylum.
Is Metarhinus diphyleticf — Riggs (1912.1, p. 27)
regards the genus as including two phyla, one contain-
ing M. fluviatilis and M. riparius, which were more
primitive and had longer heads and larger canines,
the other containing M. earlei and M. cristatus,
which were relatively shortheaded and had somewhat
more progressive teeth. The restudy of these forms
by cephalic indices partly sustains Riggs's opinion,
but indices can be depended upon only when a con-
siderable number of skulls can be measured, because
the effect of crushing is deceiving. The indices
actually taken are as follows: M. earlei, 63, 61, 60,
60, 60; M. cristatus, 60 (estimated); M. fluviatilis, 58,
56; M. riparius, 55, 51.
It would appear from these indices that M. earlei
contains the forms with broadest heads and that
M. riparius contains the forms with narrowest heads.
There are also other characters which may divide
these animals into two phyla, as shown below.
Phyla of Metarhinus
M. riparius
M. earlei, M cristatus, M. fluviatilis
More elongate skulls.
Broad-faced skulls.
Hypoeone on m' constant.
Hypocone on m^ variable.
Canines larger (? males).
Canines smaller (? females) .
Frontals narrow.
Frontals broad.
Supracranial areas lyre-shaped.
V-shaped sagittal area.
Smaller molars.
Molars relatively large.
The synopsis of these species in chronologic order
is as follows:
Metarhinus fluviaiilis Oahorn. Middle of Uinta B 1; skull
small, moderately broad (length 352 mm., breadth 205; cephalic
index 56-58); a hypocone on m^.
Metarhinus earlei Osborn. Summit of Uinta B 1 and Wash-
akie B; skull somewhat larger (length 405 mm., breadth 255,
or 338:245); somewhat broader (cephalic index 60-63);
rudiments of a secondary palate; no trace of hypocone.
Metarhinus cristatus Riggs. Lower level of Uinta B 1;
type skull of intermediate size (length 385 mm., breadth 240;
cephalix index 60); hypocone present, cingulate; similar to
M. fluviatilis.
Metarhinus riparius Riggs. Summit of Uinta B 1; skull
larger (length 406 mm., breadth 210, or 406 : 210; cephalic
index 51-55, estimated); apparently longer and narrower;
a hypocone on m^
Metarhinus fluviatilis Osborn
Plates LII, LXXI, LXXIV; text figures 123, 341, 356-358, 404
[For original description and type references see p. 183. For skeletal cliaracters
see p. 644]
Type locality and geologic Jiorizon. — White River,
Uinta Basin, Utah; Metarhinus zone (Uinta B 1);
abundant.
Specific characters. — Relatively short, broad skull,
length 352 millimeters, breadth 205, or 355: 200; cepha-
lic index 56 to 58. Eye socket small; circumorbital
ridges prominent; premaxillary symphysis elongate;
basicranial region short; sagittal crest high and
prominent. Grinding teeth subhypsodont, m^ with
a rudimentary "cingulum hypocone" in the type.
Geologic distribution. — Uinta B 1 is the horizon of
the type specimen (Am. Mus. 1500) and of the
referred specimen (Am. Mus. 1877). Similar but
somewhat more progressive forms showing the same
422
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
diminutive measurements and probably representing
higher mutations or species occur in Uinta B (Uinta B
2 of this monograph) according to Peterson's field rec-
ords (1893)— namely, Am. Mus. 1946, 1864, 2059 (a
tiny jaw), 1865. Riggs, however (1912.1, p. 21), reports
no species of Metarhinus in Uinta B (Uinta B 2 of
this monograph), nor have we found any specimens
referable to M. fiuviatilis in upper levels of Uinta B 1 .
It is possible that the line between B 1 and B 2 is not
drawn at the same point by these two observers.
General characters and Tiabits. — The type skull (Am
Mus. 1500) is supplemented by a crushed skull (Am.
Mus. 1877). Of the two skulls known neither in-
cludes the complete nasals nor affords a knowledge
of the nasofrontal horn. We can not therefore speak
positively as to this character. A further knowledge
of these animals is afforded by Am. Mus. 1946, from
Uinta B 1 (see geologic note above), which includes
a palate with superior teeth associated with a jaw
and parts of the pes and of the limbs. Also recorded
from Uinta B 1 we find a series of upper teeth, which
differ from those of the type in that the external
cingulum is absent and the tetartocones on p^ are
somewhat more advanced. From Uinta B 2 also are
recorded the American Museum jaws 1865 and 2059.
The latter is a very small jaw.
This animal is by far the most diminutive of the
known upper Eocene titanotheres — in fact, it may be
described as a dwarfed form. The specific name, M.
fiuviatilis, was assigned by Osborn on the ground that
the animal was probably a river-living animal.
In size it is appreciably smaller than the known
specimens of M. earlei, although the teeth are larger
than those of the doubtfully referred Metarhinus
Junius from the Bridger.
Specific distinctions. — Distinctions from Metarhinus
megarhinus are foimd in both the progressive and the
adaptive characters. In its progressive characters,
although it is an animal of inferior size, M. fiuviatilis
presents an advance upon M. megarhinus in the
increased hypsodonty of the molar teeth, the increased
strength of the external cingulum, the decidedly sharp
and compressed parastyle, the upward curvature of
the anterior.portion of the grinding series, a character
pointing toward the Oligocene brontotheres. The
protoconids or anterior crests of p2_4 are ridged, and
the cusps corresponding to the metaconid are better
developed. The tooth progression is thus parallel
with that of Dolichorhinus in certain respects, divergent
in others.
The adaptive characters are most interesting. We
observe especially that the narial openings are carried
very far back on the sides of the face, so that a very
narrow space is left between the orbits and the narial
notch (a feature observed also in M. earlei and
Rhadinorhinus diploconus). The region across the
orbits shows unusual breadth (partly attributable to
crushing), because the orbits as a whole are prominent
and the circumorbital ring appears to be elevated and
the eye sockets themselves are small. This feature
is analogous to that in the Oligocene brontotheres,
in which the orbits are small. We might therefore
conjecture that these animals had adopted aquatic
habits, because although the orbits are invariably
prominent, the eyes tend to become smaller in all
swimming ungulates. A pes provisionally referred to
M. fiuviatilis (from Uinta B 1) has slender metapodials,
which would indicate cursorial rather than amphibious
habits. The habitat must be left an open question
until the skeleton becomes definitely known.
Correlated with a relative shortening and broad-
ening of the skull is the narrowing of the occipital
condyles. It is difficult to discover the reason for
the elongation of the maxillary symphysis. The
symphysis appears to be elongate partly because of
the deep recession of the narial openings.
Detailed description. — The skull of this species
exhibits a great number of interesting characters,
because it presents a wide departure from the skull
of all other Eocene titanotheres except those of the
little group to which it belongs. It is a cm-ious mix-
ture of adaptive, progressive, and conservative charac-
ters. Among the last may be mentioned the relative
elongation and height of the sagittal crest. The skull
is sharply characterized specifically by the combination
of the above characters with a prominent infraorbital
shelf, a prominent antorbital bridge, and a prominent
postorbital process. The superior view of this peculiar
skull exhibits the elongate symphyseal union of the
premaxillaries, recalling that of Dolichorhinus, the
broadened posterior portion of the nasals, the deep
lateral depression of the face, or antorbital fossae in
front of the orbits, the prominent lateral projection
of the orbits, the relatively short supratemporal ridges,
which rapidly unite posteriorly, the prominent, narrow
sagittal crest. The general resemblance of this
aspect of the skull to that of Rhadinorhinus diploconus
is quite apparent. The palatal view brings out dis-
tinctive features. The posterior narial space is long
and narrow with parallel sides, the palatines not
projecting inward as in Manfeoceras manteoceras. As
compared with that of R. diploconus it appeal's to
afford the following important differences: In M.
fiuviatilis the basicranial region appears to be less
dolichocephalic, the zygomata are relatively heavier,
the infraorbital shelf (wanting in R. diploconus) is very
prominent. The anterior view of the skull exhibits
a prominence above the orbit which possibly represents
a supraorbital frontonasal horn element; it may,
however, be due to crushing. The lateral aspect of
the skull is equally distinctive, as it exhibits the deep
narial notch separated from the orbit by a very narrow
interval, the prominent antorbital bridge concealing
the infraorbital foramen, and the sharp and deeply
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
423
depressed infraorbital shelves. The peculiar elevation
of the front of the upper jaw may be partly increased
by crushing but appears to be correlated with a similar
upward flexure of the front part of the lower jaw (figs.
356, 358). The third molar is somewhat farther
back with reference to the orbit
than in Doliehorliinus.
Dentition of type and referred speci-
mens.— Some of the "referred"
specimens are recorded from Uinta
B 1 and may well belong to a more
progressive species than M. fluvia-
tilis. The following description of
the dentition includes the referred
Uinta B 1 specimens which may
belong to more progressive muta-
tions or species than the type of M.
fluviatilis.
The teeth of this diminutive
titanothere are represented by the
very much worn and somewhat
crushed series of the type (Am-
Mus. 1500, Uinta B 1) and by the
perfect grinding teeth of Am. Mus.
1946, Uinta B 1, with which is
associated the lower dentition ; also
by the complete dentition of an-
other jaw (Am. Mus. 2059, Uinta
B 1) and possibly by a more frag-
mentary upper jaw (Am. Mus. 1864)
from Uinta B 1.
Incisors : The six upper incisors,
as seen from the lower side, are
arranged in a semicircle, much as in
DolicTiorhinus. They have smooth,
gently convex anterior faces and in-
crease gradually in size from i'
to i', a tooth which is separated
from the canine by a narrow dia-
stema. The inferior incisors as
exhibited in Am. Mus. 2059 are
intermediate between the Mesa-
tirJiinus megarhinus and Dolicho-
rhinus types, pointed by wear, in-
creasing in size very gradually from
ii to is, and having smoothly con-
cave posterior surfaces bounded
by a sessile cingulum.
Canines: The superior canines
are also of the M. megarhinus form
and although much crushed in the type exhibit evi-
dence of smoothly rounded sides swelling toward
the base of the crown and accented by very deli-
cate anterior and posterior ridges. The coronal
measurements (ap. by tr.) are 17 by 16 milli-
meters. In Am. Mus. 1946, although probably a male,
101959— 29— VOL 1 30
the canines are still smaller (ap. 18 mm., tr. 16). The
inferior canines in jaw Am. Mus. 1946 are finely pre-
served except at the tips, measuring 15 by 14 millime-
ters ; they exhibit a strong antero-internal cingulum at
the base. In the other jaw (Am. Mus. 2059), obviously
FiGTJEE 356. — Skull of Metarhinus fluviatilis
One-fourth natural size. Am. Mus. 1500 (type); White River, Uinta Basin, Utah; Uinta B 1. Ai, Side
view (crushed downward); Aj, palatal view; As, top view; Ai, occipital view.
a female, the canines are much smaller (height 18
mm., ap. 13, tr. 12), with strong antero-internal
cingulum, slightly recurved apex, and somewhat ex-
panding base. The incisors and canines as a whole
seem to foreshadow the short swollen type seen in the
Oligocene Brontotheriinae.
424
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Superior premolar-molar series: The grinding series,
although of diminutive dimensions (145 mm. in the
type, which includes p', and 137+ mm. in the longi-'
tudinally crushed cotype), are really more progressive
in character than those of M. megarhinus in the fol-
lowing respects: (1) P' has a subquadrate instead of
elongate subtriangular crown, the trito- and deutero-
cones being subequal; (2) m^ has an elevated postero-
internal cingulum, which forms a subfunctional hypo-
cone in the type; (3) the ectolophs of p^ p*, as well
as of the molars, are relatively more elevated.
Premolars: The premolars of the type measure
(ap. by tr.), p^, 13 by 14 mm.; p^ 15 by 19; p*, 17
by 23. The internal cingula are well defined and
progressive, completely surrounding the deuterocone of
p^ and nearly surrounding the deuterocone of p*.
The external cingula of p^ p' are not well marked
across the protocone but are very pronounced opposite
the tritocone; in p* they are strongly developed oppo-
site both cusps, and the protocone "rib" was also
pronounced, all marks of relatively advanced speciali-
zation.
The coronal pattern is exhibited much better in
Am. Mus. 1946, in which the premolar measurements
Figure 357. — Right lower premolars (pi-pi)
of Metarhinus fluviatilis
Crown view. Natural size. Am. Mus. 1946; White River,
Uinta Basin, Utah, Uinta B 1.
(ap. by tr.) are, p', 15 by 17 millimeters; p^ 15 by 20;
p*, 17 by 23.
The most significant features of the premolars in
this specimen (No. 1946) are as follows: (1) The
deuterocones are large and give a well filled out
subquadrate inner contour to p^ p*; (2) the tritocones
are very large and progressive, especially in p', p*)
and have subflat ectoloph faces except in p^, which
has a more convex tritocone ectoloph; (3) the proto-
cones (antero-external cusps) have large, sharply
defined external ribs; (4) no tetartocones are .yet
present; (5) the external cingula are very advanced,
rising into prominent parastyles, faintly continuous
across the protocone base in p% p*, and better de-
fined opposite the tritocone; the external cingulum of
p* at the base of the tritocone surrounds a sharp
protuberance, emphasizing the gentle protuberance
seen here in M. megarhinus; (6) the internal cingula
are well defined but still incomplete opposite the
deuterocone, the posterior cingulum of p^, p* very
broad; (7) the internal cones of the premolars and
molars are relatively elevated, and they have very
thick enamel, these conditions causing the partly
worn tips to be sharply truncate.
Allowing for differences in wear and for some differ-
ences in level, we conclude that, from the evidence of
the premolars. No. 1946 is related to or referable to
M. fluviatilis.
The inferior premolars, measuring 59 millimeters in
Am. Mus. 1946, are also more advanced than those of
M. megarhinus. Pi is more advanced than in that
species; it is more elongate, has a larger posterobasal
swelling, a lower, rounder tip, and a faint rudiment of
the anterior valley. There is a faint external cingulum
on the anterior and posterior end. Its measurements
are 10 by 6 millimeters (ap. by tr.). P2 is also more
advanced than in Mesatirhinus, with a lower trigonid,
a somewhat higher talonid, and better-defined an-
terior and posterior valleys. Its measurements are
16 by 95 millimeters. In P2-Pi the protoconid forms
a high, blunt transverse ridge, extending internally
into the cusp analogous to the metaconid. In ps
(ap. 16 mm., tr. 10) the molariform tendency is still
more pronounced, the posterior lobe being stronger,
with rudiments of the internal styles appearing.
P4 (ap. 18 mm., tr. 13) exhibits a somewhat more ac-
cented external cingulum, and the cusp analogous to
the entoconid is much more elevated than in Mesa-
tirhinus and only less elevated than in the molars.
Molars: The lower molars do not differ greatly from
hose of M. petersoni, except that they are of slightly
smaller size (98 to 103 mm.), have the external cingula
a little better defined, are perhaps slightly more hypso-
dont, and have a more conic hypoconulid on ma. In
M. fluviatilis, as in M. megarhinus, the hypoconulid of
m3 is of small size and variable (or progressive) from
a crescentic to conic form.
Lower jaws of Metarhinus fluviatilis; type and re-
ferred specimens. — The type skull lacks the jaw. The
superior grinding series (Am. Mus. 1946) is associated
with the jaws, and they are also found in the nearly
perfect jaw of Am. Mus. 2059 and in the left ramus of
Am. Mus. 1865, which is from Uinta B 1 . The diminu-
tive jaw forming Am. Mus. 2059 barely exceeds in
length the jaws of certain specimens of Eotitanops
horealis, but the rami are relatively deeper, and the
dentition is of course far larger and more advanced.
Measurements of the lower jaws of specimens of
Mesatirhinus and Metarhinus are given in the ac-
companying table:
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHBRBS
425
Measurements of species of MesatirJiinus and MetarTiinus, in millimeters
[Specimens in Am. Mus. except 12195, which is in the Field Museum]
Incisive border to angle
Incisive border to condyle
Posterior border of jaw to ms-
Depth below ma
Pi-m3
Mi-m3
Mesatirhinus
M. mega-
rhinus,
Bridger
No. 1520
162
94
M. petersoni, Bridger
No. 1567 No. 1512
132
74
168
325
112
83
172
103
325
325
-90
79
169
106
M. sp.,
Uinta B,
No. 1859
"SSO
350
90
86
195
MIS
M. fluviatilis, Uinta B
161
102
M. riparius
Uinta B 1
(upper
Metarhinus
beds),
No. 12195
280
285
78
62
157
338
172
110
These jaws are distinguished by the laterally con-
stricted chin, correlated with the constricted premax-
illaries above; also by the slender rami with the lower
borders slightly downcurved below the grinders and
upcurved below the coronoid, extending back into an
angle which is produced posteriorly, as in Mesati-
rhinus. With the exception of the chin, the jaw, so
far as known, was of the Mesatirhinus type. The an-
terior border of the coronoid is decidedly angulate in
character; the coronoid itself rises with nearly parallel
anterior and posterior borders to a gently recurved
tip. These characters suggest those of a dwarfed col-
lateral of M. megarhinus. In the supposed female
(Am. Mus. 2059) the ramus appears much deeper
and more slender than in Am. Mus. 1946.
The type of Heterotitanops parvus Peterson — is it a
young Metarhinus? — A very young, perhaps a fetal
skeleton in the Carnegie Museum (No. 2909), the
type of Heterotitanops parvus Peterson (figs. 150-152),
may possibly represent Metarhinus fluviatilis. It was
found lower down in Uinta B 1 than any other known
mammalian remains. The type of Metarhinus flu-
viatilis was also found by Mr. Peterson in Uinta B 1.
According to W. K. Gregory its reference to Meta-
rhinus or to some closely allied genus is indicated by
the following facts :
1. The deciduous cheek teeth, in size and appear-
ance, bear to the adult dentition of Metarhinus fluvi-
atilis much the same relation as the deciduous denti-
tion of Oligocene titanotheres (PI. XXIII) bears to
the permanent dentition — that is, the deciduous molars
are more molariform than the permanent premolars,
and they are elongate anteriorly and have more widely
open external V's and less prominent mesostyles.
2. The large upper and lower grinding teeth that
have not wholly emerged from the alveoli are prob-
ably not m' and mi, as supposed by Peterson, but dp*
and dp4. The anteroposterior diameters as given by
Peterson are dp* 21 millimeters, dp4 25, dimensions
nearly as great as in m' and mi of the Metarhinus fluvi-
atilis type and apparently large enough for dp* and dp4
of that species. The resemblance of dp*, dp4 to m' and
mi of that species seems sufficiently striking to indicate
congeneric relationship. The measurements of dp^
Figure 358. — Lower jaws of Metarhinus
One-fourth natural size. A, M. fiumatilis. Am. Mus. 2059, White River,
Uinta Basin, Utah, Uinta B 1; B, M. earlei. Am. Mus. 13179, northwest
point of Haystack Mountain, Washakie Basin, Wyo., Washakie B.
are also not inferior to those of dp4 in Mesatirhinus sp.
(Am. Mus. 12211).
3. The deep lateral excavation of the anterior
nares, which leaves a very narrow bridge of bone be-
tween the narial sinus and the orbit, a feature char-
acteristic of Metarhinus and its near aUies, is strongly
marked also in the specimen under consideration.
426
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
4. In the side view the form of the lambdoidal
ridges of the occiput is not dissimilar to that of Meta-
7-hinus ; however, these ridges do not unite above into
a narrow median crest as they do in MetarTiinus
fluviatilis but form a flattened vertex which suggests
that of Dolichorhinus intermedins. Possibly the median
crest of the adult M. fluviatilis may be derived by the
Figure 359. — Lower jaw of Metarhinus? (Rhadinorhinus?) sp.
One-fourth natural size. Am. Mus. 1859; White River, Uinta Basin, Utah,
Uinta B 1.
upgrowth of these lambdoidal ridges toward the mid-
line. At any rate the form of occiput seen in Hetero-
titanops is not seen elsewhere outside the subfamily
Dolichorhininae. In brief the animal named Eeteroti-
tanops parvus may provisionally be regarded as the
young of Metarhinus fluviatilis.
The postcranial skeleton has been well described by
Peterson and is chiefly interesting as illustrating the
highly progressive or caenogenetic character of the
slceleton, which foreshadows the adult in the expansion
of the scapula, of the proximal end of the humerus, and
of the ilia, as well as in the great relative size of the
thorax and in the general proportions of the limb
bones.
Metarhinus earlei Osborn
Plates LXXVIII-LXXX; text figures 124, 219, 302, 358, 361,
407, 521, 522, 573-575, 577, 647, 745
[For original description and type references see p. 183. For slseletal characters
see p. 644]
Type locality and geologic horizon. — North side of
Haystack Mountain, Washakie Basin, Wyo.; Meta-
rhinus zone (Washakie B 1). Also abundant in Uinta
Basin, Utah, at the summit of the Metarhinus zone
(Uinta B 1), in the " Metarhinus sandstones" of Riggs.
Specific characters. — Skull (Am. Mus. 13166, type),
length 393 millimeters, breadth 240, or 388:245,
or 405:255; cephalic index 60-63. Occipital condyles
narrow (78 mm.), premaxillary symphysis elongate,
nasals elongate, spreading distally, prominent in-
fraorbital shelf . Type p'-m', 167 millimeters. Molar
series broad and low crowned, no hypocone on m^;
canines slender, diastema short.
This animal is readily distinguished from M.
fluviatilis by its superior size and by the lesser prom-
inence of the orbits.
Geologic distribution and materials. — The type of
this species is a skull (Am. Mus. 13166), fortunately
discovered by the American Museum expedition of
1906 in Washakie B 1. The nasals and the naso-
frontal horn region of the type are broken away.
A jaw belonging to another individual (Am. Mus.
13179) was also found in Washakie B 1 and agrees
approximately in measurement with this skull. It
may prove to belong to the same species.
In the Riggs collection of the Field Museum there
are two skulls (Nos. 12169 and 12187), also two lower
jaws (Nos. 12178 and 12189). These four specimens
are recorded from the uppermost " Metarhinus sand-
stones," or the Metarhinus zone (Uinta B 1).
Skull. — The first feature of note in the type skull is
the mesaticephalic proportions, the measurement being
240 millimeters across the zygomata and 393 from
the condyles to the symphysis — that is, the width
is a little less than two-thirds the length, whereas in
Mesatirhinus petersoni skulls the width is only a
trifle more than one-half the length. Correlated
with this feature is the relative narrowness of the
occipital condyles (78 mm.). The occipital region is
also distinctive because of a deep pit on the upper
part of the occiput and a pair of hooldike processes
ttu-ning inward on the borders of the occipital crest
(fig. 361); these falciform, incurved, overhanging
borders are quite distinctive. The sagittal crest is
short. The infraorbital shelf is prominent but
Figure 360.-
-Skull and deciduous teeth of type of
Heieroiiianops parvus
One-half natural size. Alter Peterson. Carnegie Mus. 2909 (type), White
River, Uinta Basin, Utah, Uinta B 1. Possibly a young Metarhinus.
A, Skull; B, right upper deciduous molars (dp^, dp^ dp^); the anterior
tooth is probably the permanent p'; C, left lower deciduous molar (dp().
slender. There is no hypocone on m^. We are espe-
cially struck by the prolongation of the premaxillary
symphysis, as illustrated in Figure 361, and the resem-
blance to Dolichorhinus. The well-preserved basi-
cranial region, which is also of mesaticephalic character,
indicates the affinities of this species to Mesatirhinus
megarhinus and more remotely to M. petersoni.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
427
The fine skull in the Field Museum, No. 12187,
illustrates the very peculiar constriction of the nasals
in front of the orbits and their distal expansion, also
the rudimentary horn bosses and curved nasals in
lateral view. Riggs notes that
the Uinta Basin representatives
of M. earlei are somewhat
broader headed and more mas-
sive than the Washakie Basin
type and that in both the Field
Museum skulls the hypocone
on m' is wanting; there is also
a slight elongation of thepostca-
nine diastema. (Riggs, 1912.1,
p. 30.)
Dentition. — A feature distin-
guishing this animal from both
M. megarJiinus and M.fluviatilis
is seen in the double convexities
(protocone and tritocone) of the
ectolophs of the superior pre-
molars, as in certain skulls of M.
petersoni. The superior molars
exhibit prominent ectolophs and
greatly elevated internal cones.
The internal cingula of p^, p* are
large and progressive, extending
completely across the internal
base of the protocone. The
external cingulum of p* also
extends nearly across the base
of the tritocone. The deutero-
cone of p^ is larger than in M.
petersoni. The crowns as a
whole are laterally compressed.
This hypsodont character is
an advance on the condition
observed in M. petersoni.
Jaw. — A jaw from Washakie
B (Am. Mus. 13179) provision-
ally referred to this species, as
shown in Figure 358; exhibits
proportions which are distinc-
tively mesaticephalic, like those
of the skull. It is of very su-
perior size, and the ramus is of
much greater depth than in the
jaw referred to M.fluviatilis (fig.
358). The canines are short,
rounded, and recurved. The
hypoconulid of ma is conic and
posteriorly cingulate, with a
detached cuspule on the base of
the inner side.
Additional ohservations on MetarJiinus earlei. —
An excellent skull in the Carnegie Museum, No. 3098,
found 190 feet above the bottom of Uinta B 1, has the
dimensions following.
Millimeters
Basilar length 410
Zygomatic width 243
Cephalic (zygomatic) index 60
P'-m3 162
pas. pa-'' '"^
Al
Figure 361. — Type skull of Metarhinus earlei
One-fourth natural size. Am. Mus. 13166 (type) ; north side of Haystack Mountain, Washakie Basin, Wyo., Washakie B 1.
Ai, Side view: A2, palatal view; A3, top view; A4, maxillo-premaxillary symphysis, dorsal view; As, occipital view.
pi-p< 63
M'-m3 98
The infraorbital shelf is almost as small as it is in
RhadinorMnus, but the nasals are broad.
428 TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Measurements of Metarhinus fiuviatilis, M. cristatus, M. earlei, and RhadinorJiinus diploconus, in millimeters
Am. Mus.
1600 (type),
Uinta B 1
Am. Mus.
1946, Uinta
B2
M. cristatus,
Field Mus.
12194 (type)
M. earlei,
Am. Mus.
13166 (type),
Washakie B 1
Am. Mus.
2055, Uinta
B2
Am. Mus.
1863 (type),
Uinta B 2
Pi-m3
p2-m3
Mi-s
P^, anteroposterior
P*, transverse
Transverse condyles..
Pmx to condyles
Transverse zygomata.
145
132
85
18
23
140
131
84
92
23
»385
240
167
156
102
-21
26
°76
393
240
19
168
156
103
20
26
"76
'' 440
* Specimen crushed.
This table shows the marked inferiority in size of
M. fiuviatilis as compared with E. diploconus and
M. earlei, as well as the close agreement in most
measurements between M. earlei and the type of E.
diploconus. M. cristatus is intermediate in size between
M. fiuviatilis and M. earlei.
Measurements of species of Metarhinus, in millimeters
M. fiuviatilis,
Am. Mus.
1500 (type)
M. riparius
Field Mus.
12186 (type)
M. cristatus.
Field Mus.
12194
Am. Mus.
13166 i.type)
Skull
Length, incisors to condyles
Breadtli across arches
Breadth between orbits
Postorbital process to condyles
Last molar to condyles
Length of free nasals
Greatest breadth of nasals
Postglenoids to condyles (median line) .
Breadth across condyles
Greatest depth of arch
Length of molar-premolar series
Length of molar series
Length of crown of canine
Diameter of crown of canine
Length of diastema
Narrowest point in sagittal area
Breadth of orbitonasal area
352
205
112
214
107
39
406
>> 210
<> 114
205
189
»128
68
87
79
55
160
93
24
18
11
10
41
406
220
107
212
193
115
68
89
82
51
155
93
29
20
10
385
240
145
215
195
390
237
137
198
182
388
245
142
405
255
220
220
90
75
92
169
104
Mandible
Length, condyles to incisors
Height, condyles above angle
Length of molar-premolar series .
Length of molar series
Length of crown of canine
Diameter of crown of canine
Depth of ramus from base of ps-
Depth of ramus from base of ms.
338
133
172
110
34
21
52
39
330
125
162
102
» 32
19
49
60
184
120
74
73
79
50
158
100
29
15
42
130
170
103
30
16
11
17
50
330
157
171
107
"315
135
168
110
340
148
170
105
' Specimen distorted.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
429
Metarhinus cristatus Riggs
[Compare M. fluviaiilis Osborn]
Plates LXXVIII, LXXIX; text figure 139
[For original description and type references see p. 191]
Type locality and geologic horison. — Uinta Basin,
Utah; Metarhinus zone (Uinta B 1), lower levels.
Specific characters. — Skull, estimated length 380
millimeters, breadth 240; molar series 94; frontal
region broad; sagittal crest long and high; zygomatic
arches relatively heavy. Molars short crowned, no
hypocone on m^.
Materials. — Known only from the type (Field Mus.
12194), a single skull lacking nasals and premaxillaries.
The type (fig. 139) was described by Riggs under
the impression that it came from a much higher
geologic level than that at which the type of M.
fiuviatilis was found. This geologic record has now
been corrected to show that M. cristatus was found at
nearly the same level as the type of M. fiuviatilis —
the lower half of Uinta B 1, or the Metarhinus zone.
M. cristatus may therefore be the same animal as
M. fiuviatilis. It is true that there is no hypocone
on m^ in M. cristatus, but this element is described as
a mere "cingulate hypocone" in M. fiuviatilis. The
disparity in size between the two types is not very
great, the breadth across the zygomatic arches being
205 millimeters in M. fiuviatilis and 240 in M. cristatus.
The character of the sagittal crest and of the orbital
region in the two types is quite similar, although it
would appear that in the M. fiuviatilis type the orbits
are somewhat more prominent. The difference may
be sexual, but the size of the skull of M. cristatus is
intermediate between that of M. fiuviatilis and M.
eirlei (see measurements above). Its breadth, or
cephalic index, is possibly greater, although its length
is merely estimated. It certainly contrasts sharply
with the long-headed M. riparius.
Metarhinus riparius Riggs
Plates LXXVIII, LXXIX; text figure 138
[For original description and type references see p. 191]
Type locality and geologic horizon. — Uinta Basin,
Utah; "Metarhinus sandstones" at summit of Meta-
rhinus zone (Uinta B 1). Abundant in both lower and
upper levels of Uinta B 1.
Specific characters (Riggs). — Skull long and narrow,
length 405 millimeters, breadth 210, cephahc index
55, 51; anterior cranial region expanded, sagittal
crest short; interorbital region relatively narrow and
rounded; rudimentary horn cores above orbits; man-
dible straight in the ramus ; canines large, lower canine
long and recurved ; molar series relatively short, 88-93
millimeters, hypocone usually present on m^.
Geologic distribution and materials. — The type of this
species (Field Mus. 12186) was found in the "Meta-
rhinus sandstones " at the summit of Uinta B 1 , but the
animal also occurs abundantly in the lower levels,
associated with the broad-headed M. cristatus and
with Dolichorhinus longiceps (see table, above). The
author describes it as the most common species in
this life zone. Four skulls, two associated lower jaws,
and one isolated jaw in the Field Museum collections
are referred to it — namely. Field Mus. 12174, 12183,
12191, 12195, 12196.
General characters. — The type of these species is a
laterally crushed skull, a condition which greatly
increases its apparent dolichocephaly; this character,
however, seems to rest substantially on other speci-
mens, the cephalic indices ranging from 51 to 53. It
is described as the long-headed Metarhinus riparius
and may possibly be ancestral to the aberrant animal
described above, known as Sphenocoelus, which it
resembles in the following characters: (a) Relative
dolichocephaly; (&) form of the occipital and especially
of the temporal crests; (c) wide separation of the post-
glenoid and post-tympanic processes. It is possible
that this represents a branch, M. riparius-Sphenocoelus
phylum, which may also be represented in the Prince-
ton occiput (Princeton Mus. 10041) from Washakie ?A.
The hypocone on m^ is not invariably present
although seen in the type and in the examples of the
species from higher levels. The canines (Riggs,
1912.1, p. 29) are said to be strong, whereas in the
broad-headed species, M. cristatus and M. earlei, they
are reduced in size.
Metarhinus? sp.
Text figure 359
A lower jaw (Am. Mus. 1859), from Uinta B 1, is
larger than that of any known Metarhinus or Rhadi-
norhinus but probably represents a relative of those
genera. It differs from the various species of Doli-
chorhinus in being stouter and in having a shorter
tooth row. Comparative measurements are given
above.
Another lower jaw (Am. Mus. 2355), from "Glover
Spring," level Washakie B, apparently represents a
large Metarhinus of uncertain specific reference. It
is probably not referable to Dolichorhinus vallidens.
SUBFAMILY RHADINORHININAE (=?MEGACEROPINAE)
Animals of medium size; divergent in structure
from members of the Mesatirhinus or Metarhinus
phyla. Skull with nasals contracted and pointed
distally; cranial profile concave instead of convex;
skull cyptocephalic ; facial region — that is, grinding
series — somewhat upturned; infraorbital processes
not prominent ; frontonasal horns wanting or retarded
in development; traces of terminal nasal horns in one
species. Molars subhypsodont; premolars slightly
progressive.
This group of small titanotheres is contemporaneous
with Metarhinus, Mesatirhinus and Dolichorhinus Ion-
430
TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
giceps in Uinta B 1 . It is a well-defined branch from
the more typical Dolichorhininae and, as first observed
by Gregory, has the right combination of characters
for a remote ancestor of the Megaceropinae. Such an
ancestral relationship, however, awaits confirmation
by discovery in Uinta C. RJiadinorJiinus certainly
does not lead either into Diplacodon or into Protitano-
therium. Riggs observes (1912.1, p. 41):
Rhadinorhinus apparently represents a side branch from the
early Metarhinus stem. The facial and cranial regions and the
zygomata are similar. The nasals indicate an early specializa-
tion in another direction. The dentition is more highly spe-
cialized than that of Dolichorhinus and in some respects
resembles it. The posterior nares open a little farther forward
than those of Metarhinus. The mere trace of an infraorbital
process also removes it somewhat from the long-nosed, low-
ground titanotheres.
The grounds for relating this phylum to that of
Megacerops { = Symborodon) are stated fully below.
DolicTiorhinine affinities. — Eemote relationship to
the dolichorhines {MesatirTiinus-DolichorMnus) rather
than to Manteoceras is indicated by resemblance in
the following characters: (a) Form of temporal and
sagittal crests; (6) form of zygomatic arches in top
view; (c) marked preorbital concavity and projecting
orbits; (d) slight projection of infraorbital portions
of malar; (e) similar conformation of palatal and
basicranial region; (/) rudiments of a secondary palate;
(g) incisors somewhat resembling the dolichorhiue
type; (h) premolars and molars in all features doli-
chorhine with added peculiarities.
These animals differ widely from the typical doli-
chorhines, however, in the relatively short, pointed
nasals and in the concave or saddle-shaped profile of
the cranium as seen from the side but even more
distinctly in the upbending of the grinding series,
which carries with it an upturning of the anterior
portion of the face, a feature very distinctive of
Brontoiherium and especially of Megacerops, as shown
in Figures 401-403.
Rhadinorhinus Riggs, 1912
Plates LXXIV, LXXX, LXXXII; text figures 109, 141, 322,
359, 362-364, 401-403, 405, 647
[For original description and type references see p. 192]
Geologic horizon. — MetarJiinus zone (Uinta B 1);
also Eohasileus-Dolichorhinus zone (Uinta B 2).
Generic characters. — Titanotheres with slender sub-
dolichocephalic skulls; cephalic indices 47-52; nares
deeply recessed laterally; orbits not prominent; nasals
tapering or pointed and abbreviate anteriorly; in-
fraorbital shelves rudimentary or wanting. Molars
with elevated hypocones; premolars somewhat pro-
gressives; hypocone of m' present or rudimentary.
History. — As shown above the first of these animals
known was originally described by Osborn in 1895 as
" TelmatotTierium diploconum," but it was subsequently
(1908.318) transferred by the same author to Meta-
rhinus diploconus. The type lacked the nasal bones.
The superior specimen found by Riggs, in 1910, a
skull with the nasals preserved, justified his creating
the new genus and species Rhadinorhinus abhotti,
belonging to Uinta B 1 , or a lower geologic level than
that of R. diploconus (Uinta B 2).
The species appear to be distinguished as follows :
Rhadinorhinus abhotti Riggs, Uinta B 1; skulls somewhat more
primitive and of smaller size; length 426 millimeters, breadth
224, cephalic index 52; hypocone on m' vestigial.
Rhadinorhinus diploconus (Osborn), skulls of somewhat larger
size, length 440 millimeters, breadth 210, cephalic index 47;
hypocone on m' large.
The indices of these species, which are of limited
reliance because based on single specimens, are re-
markably similar, as follows :
Indices of Rhadinorhinus
Cephalic
Facio-
cephaUc
Molar-
premolar-
cephalio
Molar-
cephalic
47
°52
43
44
38
38
23
24
Rhadinorhinus abbotti Riggs
Plate LXXX; text figures 141, 403
[For original description and type references see p. 192]
Type locality and geologic horizon. — Uinta Basin,
Utah; Metarhinus zone (Uinta B 1). Rare.
Specific characters. — Skull mesaticephalic, length 426
millimeters, breadth 224; cephalic index 52. P^-m^
164 millimeters; m'-m^ 104. Nasals shorter than
premaxillaries, thickened at suture, and tapering
toward a terminal rugosity; posterior nares opposite
m^; sagittal crest long and narrow. First and second
incisors with semioonical crowns; postcanine diastema
short.
Materials. — The type of this interesting species
(Field Mus. 12179) was discovered by Mr. J. B.
Abbott, in whose honor it was named by Riggs. No
other specimens are recorded from Uinta B 1 .
From the author's original description the following
citations may be made: In its general proportions the
skull of R. ahhotti (PI. LXXX) is similar to that of
Metarhinus riparius; somewhat longer and more
slender in the zygomatic arches; the skull suggests a
longer-limbed and more active animal; the skull is
slight in construction, the arches more slender; the
frontal region is rounded; the postorbital processes
elongate; the nasals are convex on the inferior surface
owing to the thickening at the sutural line; the facial
and palatal aspects differ widely from those of Meta-
rhinus riparius i the anterior narial opening is high and
the nasals are not infolded at the sides; they terminate
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
431
above the anterior margins of the canines so that the
premaxillaries are exposed when the skull is viewed
from above.
There is no antorbital fossa, as in M. fluviatilis; the
muzzle is broad and the canines evidently elongate.
The teeth are longer in the crowns (that is, more
hypsodont) throughout than in MetarMnus. The
author continues (1912.1, p. 37):
The molar teeth are long in the crown, having inner cones
nearly equal in height to the eetoloph. The molar-premolar
series is well preserved and unworn in the type specimen. The
canines are broken at the alveolus; half of the incisors are pre-
served. The dentition as a whole is more highly specialized
than that of Metarhinus. The first and second incisors have
short, rounded, semiconical crowns indented by cups on the
posterior surface. The third incisor has a more elongate crown
(23 mm.), the cup is suggested by a prominent cingulum on
the posterior margin. An interval of 6 millimeters separates
the third incisor from the canine. The latter is 18 millimeters
in diameter and circular at the alveolar section. The elongate
third incisor would indicate a long-crowned canine. The first
premolar is a simple, blunt cone with an internal cingulum and
emplanted by two roots. Premolars ^' ^' * increase steadily in
length of eetoloph and deuterocone; strong internal cingula
persist. The last three have taken on the subrectangular out-
line indicating a stage in dental specialization similar to that of
Sthenodectes. The molars are long-crowned; the protocone
increases steadily in height from first to third. The hypocone
is more prominent in the second and reduced to a vestige in the
cingulum of the third. The entire molar-premolar series is
more curved [that is, in a vertical plane] than in any other
Uinta titanothere.
Additional observations on RTiadinorhinus ahhotfi. — A
fine skull in the Carnegie Museum (No. 2866, Uinta
B 1) has the following measurements, which are very
close to those of the type:
Measurements of Rhadinorhinus abhotti, in millimeters
Pmx to condyles
Transverse zygomata
Postorbital process to condyles
Last molar to condyles
p'-m'
Mi-m3
Field Mus.
12179 (type)
Carnegie
Mus. 2866
426
424
224
220
240
220
196
200
164
168
103
101
The infraorbital protuberance is small but distinct ;
the nasals are not so sharply pointed as in the type.
The superior maxillary symphysis is very long. This
specimen, together with the type, clearly shows that
Rhadinorhinus is an early offshoot from the Meta-
rhinus stem (W. K. Gregory).
Measurements of skulls of Rhadinorhinus, Mesatirhinus, and Sthenodectes, in millimeters
Length, incisors to condyles
Breadth across arches
Breadth between orbits
Postorbital process to condyles
Last molar to condyles ^.
Length of free nasals ,__^j__^^
Greatest breadth of nasals
Postglenoids to condyles (median line) -
Breadth across condyles
Greatest depth of arch
Length of molar-premolar series
Length of molar series
Length of crown of canine .
Diameter of crown of canine
Length of diastema :
Narrowest point in sagittal area
Breadth of orbitonasal area
R. abbotti,
Field Mus.
12179 (type)
426
224
134
240
196
102
84
38
164
103
R. diploconus,
Am. Mus.
1863 (type)
440
'210
172
110
M. petersoni,
Am. Mus.
12184 (type)
435
215
115
220
210
95
165
105
-20
10
M. superior,
Field Mus.
12188 (type)
585
240
120
276
260
138
36
184
105
23
18
15
26
52
S. incisivus,
Carnegie Mus
2398 (type)
488
305
170
261
197
207
130
56
25
0
94
S. mcisivus,
Field Mus.
12168
460
300
160
262
190
101
211
132
47
27
0
o Estimated.
Measurements of lower jaw of Sthenodectes incisivus {Field
Mus. 12166)
Millimeters
Length, condyles to incisors 360
Height, condyles above angle 168
Length of molar-premolar series 215
Length of molar series 130
Length of crown of canine (estimated) 30
Diameter of crown of canine 19
Depth of ramus from base of ps 60
Depth of ramus from base of ma 84
Rhadinorhinus diploconus (Osborn)
Plates LXXIV, LXXXII; text figures 109, 362-364, 401, 402,
405, 647
[For original description and type references see p. 173]
Tyfe locality and geologic horizon. — White River,
Uinta Basin, Utah; Eolasileus-Dolichorhinus zone
(Uinta B 2), two specimens.
432
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Specific characters. — Skull mesaticephalic, length
440 millimeters; breadth 210 (estimated); cephalic
index 47. Facial region upturned; basicranial region
subdolichocephalic; premaxillary symphysis elongate;
infraorbital shelf absent; malar rounded; preorbital
region relatively abbreviate; occipital condyles rela-
tively narrow. P'-m', 168 millimeters ; large hypocone
on m' (?variable); marked upward flexure (cypto-
cephaly) of premolar series and of incisive border;
superior premolars somewhat progressive, with rudi-
mentary tetartocone rectigradations ; molars elevated,
with prominent protocones.
Materials. — This peculiar animal, according to
O. A. Peterson's record, is geologically more recent
than B. ahhotti, since it occurs in Uinta B 2 in the
Aj.
Figure 362. — Type skull of Rhadinorhinus diploconus
One-fourth natural size. Am. MuS. 1863 (type); White River, Uinta Basin, Utah; Uinta B 2. Ai, Side
view, as partly reconstructed in 1895 by H. F. Osborn and R. Weber; drawing reversed. A2, Top view.
Later reconstructions of this skull are shown in Figure 364.
Eobasileus-DolichorTiinus zone. The type skull (Am.
Mus. 1863) was discovered by Peterson in Uinta B 2
during the American Museum expedition of 1894.
Our knowledge is partly supplemented by another
skull (Am. Mus. 2055), also from Uinta B 2.
History. — The specific name R. [Telmatotherium)
diploconus was assigned by Osborn in reference to
the large hypocone on the last superior molar of the
type specimen, a character which is lacking in R.
ahhotti, also in the second specimen from Uinta B 2
(Am. Mus. 2055). There is, therefore, some doubt
whether the hypocone on m^ is constant. In the
original description it was also stated that the naso-
frontal did not possess a horn; there seem to be no
certain indications of a horn swelling in the nasofrontal
region. The extremities of the nasals are wanting.
The small size of the canines in circular section may
indicate that the type specimen was a female. The
premaxillary symphysis is decidedly longer and more
firmly united than in MesatirMnus megarhinus.
Distinctive features. — In contrast with MesatirMnus
and DolichorJiinus we observe that the frontoparietal
profile is concave instead of convex; associated with
this is the distinctively upturned facial region of the
skull. The mesaticephalic proportions of the skull are
indicated by the moderate transverse breadth of the
zygomata, with an estimated width of 210 millimeters
as compared with the total length of 440 millimeters
from the symphysis to the condyles. Correlated with
this is the relative narrowness of the occipital condyles
as compared with those of M. megarhinus. The primi-
tive elongation of the sagittal crest is
comparable to that in M.fluviatilis and
M. earlei. The animal also agrees with
these species decidedly in the narrow-
ness and abbreviation of the preorbital
region. Like these animals, it may
be described as narrow-snouted (a
characteristic of Megacerops). The
narial notch is also deeply recessed,
so that there is only a short space at
the side between the notch and the
anterior border of the orbits. The
postorbital processes of the frontals
are very large. Beneath the orbit is
found one of the most distinctive char-
acters— namely, the simple, rounded
form of the malars, which is in wide
contrast to the oblique shelf of
Manteoceras or the broadly project-
ing shelf of all the other species of
MesatirMnus and Metarhinus.
R. diploconus differs from Metarhinus
fl,uviatilis as foUows: (1) All the cheek
teeth are more elongate anteroposteri-
orly, hence the internal border of the
molars is less oblique than in M.fluvi-
atilis; (2) the internal cingulum of p* is
not complete; (3) the postero-external
shoulder on p"* is more prominent; (4) the skull top in
side view is broadly concave (flatter in M. fluviatilis) ;
(5) the zygomatic arch in side view curves downward
more sharply.
Variations. — Of the two skulls referred to R. diplo-
conus, one (Am. Mus. 2055) is smaller and has smaller
teeth than the type and lacks the hypocone on m'.
Cyptocephaly . — The upturned face of R. diploconus
suggests that of Megacerops of the Oligocene. It
would seem that the skull in correlation was some-
what saddle-shaped above, with the nasal region
more elevated than the frontal.
Features in detail. — The superior view of the type
skidl (Am. Mus. 1863) lacks the extremities of the
nasals; it exhibits the marked backward extension of
these bones, the great prominence of the orbital ring,
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
433
the gentle supratemporal ridges converging from the
very prominent postorbital processes into the long,
narrow sagittal crest, the gentle outward arching of
the zygomata. In the badly distorted palatal view of
the same skull a resemblance to Mesatirhinus is never-
theless evident. The distorted occiput exhibits an
elevation of 97 millimeters and a deep superior con-
cavity. In the lateral view (figs. 362, 364) we are
struck by the downward, midcranial curvature of the
profile (which appears more marked than in Mesati-
rJiinus and Manteoceras) , by the very prominent
conformation of the orbital
region above described, by
the narrow space, measuring
only 45 millimeters, between
the orbit and the narial notch.
The malar, as in other titano-
theres, forms the entire outer
portion of the infraorbital
bar, the maxillary entering
into its anterior rim, as in
MesatirTiinus and Manteo-
ceras. It appears to lack
the infraorbital shelf as noted
above. The malar passes
anteriorly into the narrow
and decidedly convex and
elevated bridge at the point
of union with the lacrimals.
It is narrow and flattened
below the posterior portion
of the orbit as it passes
backward; it is thin on the
inferior surface. The zygo-
matic portion of the squa-
mosal is moderately elevated
and expanded.
Dentition. — The materials
consist of the type skull (Am.
Mus. 1863) and of another
skull (Am. Mus. 2055) con-
taining alveoli of the cut-
ting teeth and the worn crowns of the grinding
teeth.
Incisors: The superior incisor alveoli have the
characteristic convergence of the opposite series, and so
far as can be judged from the alveoli, which increase
regularly from i' to i^, the canine fang (ap. 14 mm., tr.
12) is laterally compressed and is of small dimensions,
like those of a female.
The molar-premolar series are of the Mesatirhinus
type; they measure 169 millimeters in length, as com-
pared with 156 in M. petersoni. They are distinguished
by several features, among them the very prominent
pointed hypocone of m^, to which the specific name
refers but which is lacking in Am. Mus. 2055. P^, p'
exhibit the very rudimentary postero-internal or
tetartocone ridges which are observed in T. ultimum
and Dolichorhinus. There are other faintly indicated
progressive characters, especially the comparatively
deep medifossettes and postfossettes on the molars
and premolars, correlated with the greater hypso-
donty, and the nearly symmetrical protocone and
tritocone convexities on the ectoloph of p^"'*, as in M.
earlei.
Premolars: More in detail: P' is separated from
the canine by a narrow diastema (5 mm.) and is a
f:/,l.a.(n.
Type skull of Rhadinorhinus diploconus
One-fourth natural size. Am. Mus. 1863. White River, Uinta Basin, Utah; Uinta B 2. Partly straightened,
occipital crest has heen narrowed by crushing. Ai, Top view; M, palatal view.
bifanged tooth; p^ is in a much more progressive
stage of evolution than in M. petersoni, with its sub-
quadrate form, anteriorly placed deuterocone, sub-
functional tetartocone, well developed and slightly
convex tritocone, the dimensions (ap. by tr.) being 16
by 19 millimeters; p^ is of similar pattern (ap. 17
mm., tr. 21), with nearly symmetrical protocones and
tritocones, an anterior style, and a rudimentary pos-
tero-internal cusp or tetartocone. In the develop-
ment of the last-named cusp p^ is slightly more pro-
gressive than p^ (ap. 20 mm., tr. 25); p^ is of the same
pattern but entirely lacks any trace of the tetartocone.
An internal cingulum nearly surrounds these teeth ex-
cept opposite the deuterocone.
434
TITANOTHERBS OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Molars: M' (ap. 29 mm., tr. 27) is slightly longer
than broad and thus like the other molars is of marked
dolichocephalic type; the hypococone is very high and
prominent. M^ (ap. 37 mm., tr. 34) exhibits faint
external and incomplete internal cingula with a strong
antero-internal cingular basin, a progressive feature
of aU these molars. This basin, or "hypoconid fossa,"
indicates relatively advanced hypsodonty in the lower
molars, which we might infer also from the subhypso-
donty of the upper molars.
Peculiar also is the sharp furrow dividing the proto-
cones from the hypocones, and the prominence of the
internal cingulum midway between the protocones and
hypocones. M^ has an elevated ectoloph (28 mm.)
and is also longer (37 mm.) than broad (35 mm.);
(3) nasals small, pointed anteriorly; (4) malars and
antorbital bar rounded, with very short space in front
of antorbital bar; (5) orbits small, inset; (6) a deep
recession at the sides of the nares, and nasals high set;
(7) occiput with a deep concavity; (8) chin weak,
concave below, sloping up to a plane higher than that of
the grinding teeth; (9) ramus with coronoid process,
etc. (see Am. Mus. 2059, Metarhinus fluviatilis, a
diminutive copy of the Oligocene Megacerops type);
(10) first and second upper incisors {R. ahhotti Riggs)
with short, rounded subcorneal crowns; (11) canines
of somewhat obtuse form, with swelling posterior
cingulum; (12) lower premolars sloping upward anteri-
orly with strong internal cingulum; (13) upper pre-
molars relatively progressive, well-rounded contours ,
large tritocones and relatively
progressive tetartocone rudi-
ments; (14) tetartocone rudi-
ments in p^, p' {R. diploconus)
more advanced than in p*, pre-
molars sub quadrate in form;
(15) molars elongate or sub-
hypsodont.
Intermediate forms between
R. diploconus which may be
discovered in Uinta B 2 and
Uinta C will determine the
question whether this relatively
primitive form is ancestral to
the Megaceropinae.
successors to the manteo-
ceras-dolichorhinus group
(eotitanotherium, DIPLA-
codon)
subfamily diplacodontinae (=?meno-
DONTINAE, = ?BEONTOTHEKmf AE)
[Eocene phylum Diplacodon]
Including upper Eocene fore-
„ . , , , runners of the Oligocene genera
One-fourth natural size. Am. Mus. 1863. White River, Uinta Basin, Utah; Umta B 2. Straightened and recon- ° "^ .
structed, except nasals and condylar region. Ai, Side view (reversed); A2, front view, partly straightened, but JilenoduS, BrOntotlieriUm.
occiput probably too high and postorbital process of frontal too low; A3, occipital view, straightened. Primitivclv dolichocenhalic
Figure 364. — Type skull of Rhadinorhinus diploconus
its prominent hypocone constitutes a distinctive
feature, but from its absence in Am. Mus. 2055 it may
not be a valid and constant specific distinction.
Is Rhadinorhinus ancestral to Megacerops? — The
possible ancestry of R. diploconus to Megacerops first
suggested itself to W. K. Gregory in 1903 and was
carefully considered by him and by Osborn in sub-
sequent years. There are, in fact, many features in
which Rhadinorhinus diploconus appears to fore-
shadow Megacerops, chief among which are the fol-
lowing: (1) Saddle-shaped profile, high nasofacial
region, depressed frontoparietal profile; (2) facial
region relatively abbreviate, upturned — that is, an
upward flexure of maxillaries and premaxillaries;
progressively mesaticephalic. Precocious develop-
ment of horns. Nasals narrowing anteriorly, curved
downward at the tips. Second internal cones
on the superior premolars precociously developed.
(For progressive characters see Oligocene stages,
p. 467.)
The phyletic relations of the two animals now to be
described are not certainly determined; they are pro-
visionally placed in a separate subfamily (Diplaco-
dontinae), although it is possible that they belong
in one of the OHgocene subfamiUes (Menodontinae,
Brontotheriinae). They include the Diplacodon of
Marsh, discovered in 1873, and the Eotitanotherium of
Peterson, discovered in 1912.
EVOLUTION OF THE SKULL AND TEETH OP EOCENE TITANOTHERES
435
Eotitanotherium Peterson
[Diploceras Peterson; cf. Diplacodon Marsh]
Plate LXXXI; text figures 148, 149, 365-367, 372, 598-605, 647
[For original description and type references see p. 196. For skeletal characters see
p. 656]
Geologic Tiorizon. — Uinta B 2. In 1913 Peterson
(1914.1) announced the surprising discovery, in the
middle levels of the Uinta Basin, Utah, of a large
titanothere having a pair of prominent bony out-
growths above the orbits, which he proposed to name
Diploceras. Peterson later gave it the name Eotitano-
therium to replace Diploceras, which was preoccupied.
Generic characters. — Large upper Eocene dolicho-
cephalic titanotheres with very long, decurved nasals
and well-developed, anteroposteriorly oval horns.
Premolars decidedly progressive. P^, p* with large
distinct tetartocones and reduced external and inter-
nal cingula. The type of Eotitanotherium oshorni
Peterson (Carnegie Mus. 2859), so far as known,
conforms with the definition of Diplacodon as given
below:
Skull mesatlcephalic to dolichocephalic, zygomatic arches
slender; superior premolars with flattened ectolophs and double
convexities, p^-p"* progressive, quadritubercular — that is, with
tetartocones; molars of elongate or dolichocephalic type —
that is, laterally compressed.
E. oshorni Peterson further agrees with Diplacodon
elatus Marsh in the general small, obtuse form of the
canine and in the presence of a wide post canine
diastema.
On the other hand, the type of E. oshorni appears to
represent a distinct species or even a different genus
from D. elatus, for although it comes from a lower
geologic level (Uinta B 2) its premolars are decidedly
more progressive in character, p', p* having the
tetartocones larger and more separate from the deu-
terocones, the external and internal cingula reduced,
and the whole appearance of the crown more molari-
form than in D. elatus.
Comparative measurements of the two species are
given below.
Comparative measurements of Eotitanotherium and Diplacodon,
P>-m3
Pi-p4
M'-m'
P^, ap. by tr
M', ap. by tr
M2, ap. by tr_ ..
M', ap. by tr
Canine, vertical
Canine, anteroposterior
Postoanine diastema
Nasals, length
E. osborni,
Carnegie Mus.
2859 (type)
D. elatus, Yale
Mus. 10320
(type)
246
250
97
90
146
158
29X37
28X34
40X43
41X42
48X50
62X42
60X55
60X57
25
17
28
24
119
Comparative measurements of Eotitanotherium and Diplacodon,
in millimeters — Continued
Pi-ma
Pi-P4
P2-P4
Mi-ms
Postcanine to hyloid of ms
Postcanine diastema .
Depth below ms .
E. osborni,
Carnegie Mus
2859 (type)
255
95
79
158
283
34
93
D. elatus, Yale
Mus. 10320
(type)
Whether the paratype skull (Peterson, Carnegie
Mus. 2858) belongs with this species is somewhat
doubtful. The greater size of the horns and the asso-
ciated widening of the nasals in the paratype may
well be a sexual difference, the paratype being possibly
a male, the type a female. On the other hand, the
paratype differs radically from the type in having
the tetartocone of p* feebly developed, thus approach-
ing D. elatus.
Although Eotitanotherium, from Uinta B 2, is prob-
ably allied to the somewhat later Diplacodon (Uinta
C 1), it seems best to let it stand provisionally as a
distinct genus, especially in view of the progressive
character of p^ and p* in this older stage.
Eotitanotherium osborni Peterson
[Diploceras osborni Peterson]
[For original description and type reference see p. 195. For skeletal characters
see p. 656]
Type locality and geologic horizon. — Duchesne River,
near My ton, Uinta County, Utah; Eohasileus-Doli-
chorinus zone (Uinta B 2).
Specific characters. — P'-m^ 240 millimeters; m'-m^,
145. Median and lateral incisors small and "round-
topped," approaching Oligocene type, with heavy
posterior cingulum; lateral incisors more massive.
Canines short, subconic, without cingula, postcanine
diastema 28 millimeters. Premolars and molars with
little or no external cingula; internal cingula reduced;
p', p* highly progressive, submolariform, with large
tetartocones (type) and two well-marked external con-
vexities; dimensions of p* (ap. by tr.) 29 by 37 milli-
meters. Upper molars wider than in D. elatus,
m' with incipient tetartocones. Nasals long, tapering,
decurved. Horn swellings low, elongate, oval in basal
section. Lower jaw with deep ramus and deep
symphysis.
The type and paratype skulls of Eotitanotherium
{Diploceras) oshorni were thus described by Peterson
(1914.1, pp. 30-37):
SKULL
[Fig. 366]
In comparing the recently discovered material with the best
preserved remains of Protitanotherium (P. emarginatum Hatcher)
a number of important differences are at once observed. The
436
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
nasals of the new species are longer, thinner, somewhat narrower
(especially in specimen No. 2859); furthermore the lateral
borders of the nasals are much less thickened, and instead of
the broadly emarginated area at the free end of the nasals in
P. emarginatum, the termination of the nasal of the present
form has an abrupt downward turn resembling that of
Megacerops coloradensis Leidy, and its anterior margin is very
convex transversely, instead of concave, as is the case in P.
emarginatum. Upon the whole the nasals of the species we are
X,
^^»8««i«^
'■'*%
Figure 365. — Skulls of Eotitanotherium osborni
[ = fig. 365].) This varied development of the horn cores is no
doubt due to se.xual differences, or possibly to individual
variation. The premaxillaries extend well in front of the
maxillaries and are separated in front, forming a deep median
notch, as in P. emarginatum, so that the median pair of
incisors are wide apart, while farther back they are firmly
coossified and also solidly fused with the maxillaries. The
infraorbital foramen is also of large size as in P. emarginatum
and located above p^ as in the latter species. The maxillary
is on the whole very robust and shows
that it had advanced well toward
the condition found in Diplacodon
and Titanotherium. This is also true
of the ]'ugal, the prominent lower
border of which has the downward
and backward sweep in front of
and under the orbit, which is charac-
teristic of Titanotherium. The zygo-
matic arch, though widely expanded
behind, is, however, less robust than
in the Oligooene genus and agrees
better with the type of Diplacodon
elatum described by Marsh. The
postorbital processes on the frontal and
jugal are of large size, in this respect
unlike Titanotherium. The postorbital
process on the frontal of the latter
genus is usually located farther back
and is much smaller in proportion.
The external portion of the glenoid
cavity is preserved in No. 2858 and is
somewhat less convex in the antero-
posterior direction than in the latter
genus. As in Titanotherium the
anterior palatine foramina are small
round openings, which in the present
genus are situated farther back from
the alveolar border of the incisors. The
palate is of the deep concave form
usually met with in the titanotheres,
and the posterior narial opening ex-
tends approximately as far forward
as in the Oligocene genus, reaching to
the posterior portion of m^.
That the type of the skull was saddle-
shaped is very evident from the
material under study, but whether or
not the characteristically broad
superior aspect of the parietals and the
heavy and broad occiput seen in
Titanotherium had been attained to
the same degree as the similarity of
the anterior region in the two genera
suggests might have been the case will
not be completely known until the
posterior region of the skull of the Uinta
After Peterson. One-Iourth natural size. A, Type skull, with associated lower jaw, 9 , Carnegie Mus. 2859, Duchesne representatives of this phylum is
River near IVIyton, Uinta County, Utah, Uinta B 2; B, paratype skull, tf , Carnegie Mus. 2858 (reversed), same fo^nd. It is highly probable that the
locality and level as A.
describing extend further forward. There seems to be a con-
siderable variation in the development fo the horn cores; thus,
in skull No. 2858 this protuberance appears to have a develop-
ment comparable to that of some of the titanotheres found in
the Ohgocene, while in specimen No. 2859 these osseous bosses
are very much smaller, more conical, and in proportion more
like those of P. emarginatum, in spite of the fact that the skull
we are considering pertains to an old individual. (See PI. VII
similarity presented by the anterior
region will be preserved throughout the cranium, which will
then reveal more exactly the features of a true titanothere than
was anticipated. From the type of Protitanotherium emar-
ginatum at Princeton University, Hatcher ^o was apparently
able to determine that the sagittal crest is absent and that the
dorsal surface of the skull is probably slightly concave antero-
posteriorly.
!• Am! Naturalist, vol. 29, p. 1085, 1895.
EVOLUTION OP THE SKULL AND TEETH OF EOCENE TITANOTHERES
437
Measurements in millimeters
No. 2858
No. 2859
Skull
Diameter from incisors to posterior border
of glenoid cavity
470
"ISO
67
55
260
"93
-67
<• 290
Diameter from incisors to anterior border
of orbit _____
''179
-67
Diameter from incisors to anterior border
of posterior nares
Transverse diameter at the canines
Transverse diameter at diastema between
the canines and the premolars
Transverse diameter across the horn cores. _
Lower jaw
Total length of jaw fragment-
256
86
67
136
380
Diameter from incisor to pi _ _
67
»70
»80
92
" Approximate measurements.
MANDIBLE
[Fig. 3C5]
The lower jaw is somewhat depressed by crushing, but
allowing for this fact, it appears that the horizontal ramus of
Diploceras osborni is shallower than in P. emarginatum. Char-
acteristics which may further be noted are the more rounded
under surface of the symphysis and the constriction of the lower
jaws in the area between the canine and the premolars, which is
greater than in P. emarginatum. As in the latter, the symphysis
is strong and the mental foramen is large, located well down
on the ramus, directly below P2. The lower jaw is broken
off back of ms.
DENTITION
[Fig. 366]
The upper incisors and canines are well preserved, though
much worn in the two crania under description. The molar-
premolar series is less completely preserved in No. 2858, while
in 2859 the superior dentition is completely represented. The
lateral incisor and the canine of the right mandible and the
complete molar-premolar series of the left ramus are also
present in the latter individual.
As stated above, the median upper incisors are widely
separated by the deeply excavated median notch of the pre-
maxillaries. As seen in the illustration, the incisor series is
placed well in front of the canine and the arc of the circle,
which their arrangement represents, is more convex than in
P. emarginatum. Their crowns are nearly circular in outline,
covered with a heavy coat of enamel, bluntly conical, with a
prominent cingulum at their posterior bases. They perhaps
increase in size more gradually from i' to i^ than in P. emar-
ginatum. The canine is relatively smaller than in the latter
genus, which imparts a much lighter looking aspect not only
to this region of the dentition but also to the entire outline of
the anterior portion of the muzzle in the paratype. No. 2858,
as well as in the type. No. 2859. Furthermore the crown of the
canine (especially in No. 2859) is shorter, blunter, and the
lateral ridges are less developed in the present species than in
either P. emarginatum or Diplacodon elatum. D. elatum has
the canine more nearly of the same proportion as in P. emar-
ginatum. The diastema back of the canine is relatively longer
and its border much thinner than in P. emarginatum, in which
respect it is more nearly like Diplacodon elatum.
The crown of pi is so much worn that its characters can not
be made out. It is, however, of greater anteroposterior than
transverse diameter and undoubtedly had a simple structure
like that of P. emarginatum. P^ is also much worn, especially
Figure 366. — Nasals and horn swellings of Eotitanotherium
osborni
One-fourth natural size. Ai, Carnegie Mus. 2859 (type), Duchesne River near
Myton, Uinta County, Utah, Uinta B 2, top view; Aj, same, basal view; B,
Carnegie Mus. 2858 (paratype), locality and level same as A, top view.
along the external portion. The external face of the ectoloph
is subdivided by a deep vertical groove and is much convex both
anteroposteriorly and supero-inferiorly. This deep groove adds
greatly to the anteroposterior convexity of the proto- and trito-
cones. The general outlines of the tooth are less quadrate than
in Titanotherium, which is apparently due to the lack of devel-
438
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
opment of the antero-internal angle in the species under con-
sideration. In the type of Diplacodon datum p' is lost, while
the external portion of p^ is broken off. In the present species
the deuterooone of p^ is less ridgelike than in D. elatum, the
two interal tubercles being somewhat better indicated and the
ridge between them distinctly less developed. P' is more
quadrate in outline than the preceding tooth and has two dis-
tinct internal tubercles on the crown, which are separated by a
shallow groove, while in Diplacodon elatum these tubercles are
united into a solid internal ridge, revealing a distinct differen-
tiation from what is seen in the present species [compare PI.
LXXXll. On the other hand, p' both in the type we are
describing and in D. elatum are similar, there being two internal
tubercles, deutero- and tetartocones, the former considerably
the larger.^" The more important differences in the dentition
of the two forms, so far as they can now be compared, seem to
be in the proportion of the canines, the difference in the length
of the premolar series, and the detailed structure of p'. The
greater length of the premolar series is naturally to be expected
in a form from a lower geological level.
The detailed characters of the molar series of the genera here
compared present no differences of importance. The two Uinta
forms agree in the obscure or feeble development of the cusp-
like elevations on the anterior face of the molars near the inner
angle, more conspicuously developed in Titanotherium. At the
postero-internal angle of the cingulum of m' in the Oligocene
Figure 367. — Two upper molars of Eotitano-
therium {" Diploceras") osborni Peterson
Crown view. One-half natural size. Carnegie IVIus. 2860a.
These isolated teeth were found with the paratype, Carnegie
Mus. 2S60; Duchesne River near Myton, Uinta County,
Utah.; Uinta B 2.
forms there is sometimes a distinct tubercle, which is indicated
in the Uinta forms by only a slight swelling of the cingulum.
In proportion the inferior incisor dentition is further in ad-
vance of the canine than in P. emarginatum. Ii and i2 are repre-
sented only by a portion of their roots buried in the symphysis.
I3 has a very prominent cingulum posteriorly. Notwithstand-
ing the much smaller size of the specimen, its crown has very
nearly the same diameter as in P. emarginatum, which would
indicate that the inferior incisors were possibly larger in propor-
tion and more nearly equal in size. The crown of the canine
is injured, but its diameters appear to be equal to those in the
superior series, though relatively smaller than in P. emar-
ginatum. Pi has a single root and a simple conical crown,
which has not received any wear, due to its somewhat inferior
position. Pj is submolariform and in its general characters
does not differ from the same tooth in P. emarginatum. P3 is
quite molariform, while P4 has a complete molar pattern.
There is no difference in the general features of the lower
molars in the two genera here compared, and in turn the molars
of Diplacodon are on the whole quite similar in their detailed
structure to those of the Oligocene genus.
•» In No. 2858, the paratype, there is only one internal tubercle, the deuterocone
which may by some be regarded as of sufficient importance to constitute a specific
difference. For the present I prefer to regard this character as possibly representing
a reversion.
The proportion of the alveolar border occupied by the lower
premolars of this species is in accord with the upper series —
that is, of a greater anteroposterior diameter than in P. emar-
ginatum and D. elalum.^^ Judging from the type (lower jaw)
of Protitanotherium superbum Osborn, recently described,'^ that
species also has the same proportion of the molar-premolar
series as the two latter, while Telmatherium? altidens of the
same publication has a longer premolar series and more nearly
agrees with the present genus.
Measurements, in millimeters
Length of superior incisor series
1', anteroposterior diameter
1', transverse diameter
I^, anteroposterior diameter
I^, transverse diameter
I' , anteroposterior diameter
1', transverse diameter
Canine, anteroposterior diameter at the base.
Canine, transverse diameter at the base
Length of molar-premolar series
Length of superior premolar series
P', anteroposterior diameter
P', transverse diameter
P^, anteroposterior diameter
P^, transverse diameter
P', anteroposterior diameter
P^, transverse diameter
P*, anteroposterior diameter
P*, transverse diameter
Extent of superior molar series
M', anteroposterior diameter
M', transverse diameter
M^, anteroposterior diameter
M^, transverse diameter
M', anteroposterior diameter
M', transverse diameter
I3, anteroposterior diameter
I3, transverse diameter
Canine, anteroposterior diameter, approxi-
mately
Canine, transverse diameter, approximately
Length of inferior molar-premolar series
Length of inferior premolar series
Length of inferior molar series
Pi, anteroposterior diameter
Pi, transverse diameter
P2, anteroposterior diameter
P2, transverse diameter
P3, anteroposterior diameter
P3, transverse diameter
P4, anteroposterior diameter
P4, transverse diameter
Ml, anteroposterior diameter
Ml, transverse diameter
M2, anteroposterior diameter
M2, transverse diameter
M3, anteroposterior diameter
M3, transverse diameter
No. 2859 No
34
11
10
12
12
15
14
19
18
246
101
19
12
22
25
30
31
33
38
■146
38
45
52
54
67
51
14
12
17
14
255
94
160
14
10
24
14
28
18
29
20
38
26
49
30
78
32
" Professor IMarsh's measurement of the molar series of the type of D. elatum is
31 In remeasuring the molar series of Prof. IMarsh's type of Diplacodon elatum it
would seem that he was in error in regard to the measurement, which should read
167 instead of 152 millimeters.
3! Osborn, H. F., New and little-known titanotheres from the Eocene and Ohgo-
cene: Am. Mus. Bull., vol. 24, p. 615, 1908.
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
439
Diplacodon Marsh
Plates LXXXI, LXXXII; text figures 99, 368-371, 406, 594,
595, 597, 723
[For original description and type references see p. 166]
Geologic Jiorizon. — True Uinta formation (Uinta C).
History. — The original and only known remains of
Diplacodon elatus, an animal that has played a very
large part in titanothere literatm-e, consist of a
crushed palate with a full series of grinding teeth,
found by the well-known prospector Samuel Smith on
White River, Utah, August' 24, 1874. The geologic
level is now known to be the lower part of Uinta C,
a level higher than that of the species Eotitanotherium
osborni, which we have just been considering. The
specimen was described by Marsh in February, 1875,
as representing a new genus, which he named Diplaco-
don in reference to the double internal cones on the
superior premolar teeth. Marsh immediately recog-
nized the importance of this animal as a structiu-al
connecting link between "Palaeosyops" and "Bronto-
therium." He subsequently observed: "In the upper
Eocene, both [Limnohyops and Palaeosyops] have
left the field, and the genus Diplacodon, a very near
relative, holds the supremacy." (Marsh, 1877.1,
p. 31.) This animal was also regarded by Earle^
Scott, and Osborn as a direct connecting link between
the Eocene and Oligocene titanotheres.
Generic cTiaraciers. — Skull mesaticephalic to dolicho-
cephalic; zygomatic arches slender. Superior premo-
lars with flattened ectolophs and double convexities;
p^~* progressive, quadritubercular — that is, with
tetartocones; molars of elongate or dolichocephalic
type — that is, laterally compressed.
Horns. — In his original description Marsh stated
that this animal was distinguished from the Oligocene
titanotheres "by the absence of horns." As the
skull in the type and only known specimen of D.
elatus is badly crushed it is difficult to determine
whether or not the type animal possessed rudimentary
horns; if it had they were certainly not so large as
those of the Oligocene titanotheres. In the related
type of Eotitanotherium (Diploceras) oshorni Peterson
there are well-developed horns with elongate oval
bases.
Progressive molar characters. — The type skull is too
imperfectly known for us to distinguish all its con-
servative and progressive characters. In the dentition
the following progressive characters are observed:
(1) Premolars with well-developed tetartocones; (2)
complete internal cingula; (3) rudimentary fossettes
on the crown surface; (4) premolar tritocones large
and subequal with protocones; (5) p^ much more
progressive than in Telmatherium ultimum or Manteo-
ceras uintensis, with large, centrally placed deutero-
cones and subquadrate rather than triangular contour;
(6) molars with very distinct fossettes or pits on the
crown surface near the ectoloph.
101959— 29— VOL 1 31
All these molar teeth characters point toward the
Oligocene stages of the teeth of Menodus rather than
of Brontotherium; on the other hand, the structure of
the canines and incisors points rather toward Bronto-
therium than Menodus.
Diplacodon elatus Marsh
Plates LXXXI, LXXXII; text figures 99, 368-371, 406, 597
[For original description and type references see p. 166]
Type locality and geologic horizon. — Uinta Basin,
Utah; Diplacodon-Protitanotherium-Epihippus zone,
Uinta C 1, the true Uinta formation.
Figure 368. — Type skull of Diplacodon elatus
Palatal view. One-fourth natural size. Yale Mus. 10320; Uinta O I. Partial
reconstruction of the under surface of the slcuU based on the type materials.
Specific characters. — Skull mesaticephalic to doli-
chocephalic. P'--m', 242 millimeters; molars elongate
anteroposteriorly and subhypsodont; premolars short
and broad; tetartocone rudiment on p^, tetartocones
increasing in distinctness on p^ and p"*; m^ without
hypocone; canines small in females.
The type specimen (Yale Mus. 10320) is a female,
as indicated by its small, rounded, recurved canines.
The postcanine diastema is considerable, measuring
24 millimeters. The grinding teeth are laterally
440
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
compressed, and the zygoma is slender and decidedly
feeble, indicating that the skull was dolichocephalic.
The total dimensions of the grinding series are, pre-
molars, 89 millimeters; molars, 152; premolar-molar
series, 242. This is of smaller size as compared
with the inferior grinding series of Protifanotherium
emarginatum (304 mm.), of P. superbum (318), and of
Telmatherium altidens (313).
Progressive characters. — Among the details of pro-
gressive character may be mentioned : (1) The cingulum
pas. Tiis. tr Aj
Figure 369. — Type skull of Diplacodon elatus (female)
Upper jaw and zygoma. One-fourth natural size. Yale Mus. 10320; Uinta C 1.
zygoma partly reconstructed; si, S2, sections; As, top view of zygoma,
is continued on the inner sides of the crowns of the
premolar teeth; (2) the premolar teeth are rendered
quadrate internally by the expansion of the tetartocone
shelf, and p^-p"* exhibit the progressive development
of the tetartocone from before backward by constric-
tion or budding from the deuterocone crest; (3) the
fossette of the crown is seen in a rudimentary form in
the premolars and very distinctly in the molars, a
feature characteristic of all Oligocene titanotheres and
clearly foreshadowed in Rhadinorhinus diploconus of
the upper Eocene.
Canines. — The canines are small, laterally com-
pressed, and slightly recurved, but too much worn
and fractured to clearly represent their form.
Premolars. — As noted above, the deuterocone crest
is convex on the median or lingual side and flat or
concave on the outer or buccal side, as in Telmathe-
rium. Another distinctive character is seen in the
ectolophs, which are decidedly flattened, especially
the outer surfaces of the tritocones, reminding us of
the condition in Dolichorhinus; the external cingulum
arises and festoons the protocones. The detailed
proportions of the teeth (ap. by tr.) are, p^ 21 by 23
millimeters; p^ 25 by 29; j)\ 28 by 35. In p* the
deuterocone is large and elevated (17 mm.) and the
worn tetartocone is low (13 mm.) but almost entirely
distinct.
Molars. — In the molars the external cingula are
faint, but the internal cingula are wanting. A
characteristic feature is the antero-internal expansion
of the cingulum, which makes the entire anterior
border angular and prominent, especially as seen in
m^ This cingulum ridge bounds the fossa for the
metaconid and is clearly foreshadowed in Rhadino-
rhinus diploconus. The fossettes are nearly worn out
in m' and m^; both anterior and posterior fossettes
appear, and in m' the anterior fossette is a deep,
narrow pit. The proportions of the teeth (ap. by tr.)
are, mS 41 by 44 millimeters; m^, 55 by 55; m', 60
by 55. These proportions are decidedly different from
those in Protitanotherium. (Fig. 371.) The posterior
cingulum of m^ is elevated, and a low, distinct hypo-
cone swelling appears. The
ectolophs exhibit compressed
mesostyles. The ectolophs are
moderately hypsodont; the pos-
terior view of m^ shows that
the mesostyle terminates in a
horizontal ridge, as in many
little-worn or unworn titano-
there molars; it is especially
strong in Rhadinorhinus diplo-
conus.
Comparison of Diplacodon ela-
tus with middle Eocene titano-
theres.— Diplacodon combines
characters of Telmatherium and
of Dolichorhinus but does not agree with either in all
respects.
Comparison with Telmatherium: Diplacodon shares
with Telmatherium the tendencies toward dolichoce-
phaly and toward the development of tetartocones,
but it far outstrips Telmatherium in both features and
differs very radically in others, as follows: The canines
are small; p^ in Diplacodon is much more advanced
and of different type; the tritocones of the superior
premolars are flattened instead of rounded; the molars
Ai, Upper jaw and teeth, with
partly reconstructed.
Figure 370. — Third and fourth upper pre-
molars of Diplacodon elatus
Natural size. Yale Mus. 10320 (type, reversed). Uinta
CI. Outer side view.
are much more elongate and have flattened ectolophs;
the zygomata are much more slender.
Comparison with Protitanotherium: These animals
are readily distinguished from the mesaticephalic
Protitanotherium in all the dolichocephalic propor-
tions of the teeth; and from the radical differences in
EVOLUTION OF THE SKULL AND TEETH OF EOCENE TITANOTHERES
441
the teeth it is safe to infer that the proportions of the
skull also differed widely.
Comparison with DolicJiorhinus : Resemblances to
members of this genus are observed in the flattening
of the premolar ectolophs, especially of the tritocones;
A.M. 250i, type
^YaleMus.l03Z0,type
FiGtTBE 371. — Upper molars of Diplacodon and
Prolitanotherium compared
One-half natural size. Superimposed contours of the first and
second upper molars of the dolichocephalic D. elatus (heavy
line) {Yale Mus. 10320, type) and the brachycophalic P. su-
perbum (light line) (Am. Mus. 2501, type) .
in the slenderness of the zygomatic arches posteriorly;
in the elongate or dolichocephalic type of the grinding
teeth; and in the small size of the canines. The most
important progressive or divergent differences from
Dolichorhinus are the great development of the tetar-
tocones, especially on p*; the great breadth of p**; the
absence of a broad infraorbital shelf on the malars;
the apparent retardation of the horn rudiments.
Comparison with RhadinorJiinus: Some resemblances
between Rhadinorhinus and Diplacodon elatus led
to the doubtful view that the two were related.
These resemblances are seen especially in the propor-
tions of the molars, which are of dolichocephalic type,
and of the premolars, which are relatively broad.
Among other characters common to the two species
are the following: Molars subhypsodont; m^ of elon-
gate, compressed form; parastyles and mesostyles
sharp and delicate; hypocones of m', m^ set well in on
crown; external cingula delicate; deep fossettes
median and posterior. In the premolars we observe
that the crowns are relatively broad; the tetarto-
cones of p^, p' are somewhat progressive in B. diplo-
conus; the tritocone is very large on p^-p*; and the
medifossettes on p^-p* are deep.
Type skull oj Diplacodon elatus. — The fractured
skull affords only a few characteristic features. The
main indications are of a dolichocephalic type, with
slender zygomatic arches. The premaxillary sym-
physis is apparently deep, measuring 92 millimeters
from the incisive border to the lower border of the
nasal notch. The extent of the palatines upon the
hard palate was apparently rather narrow, the pos-
terior nares opening directly opposite the interval
between the second and third molars. There is the
characteristic rugosity at the junction of the basi-
sphenoid and basioccipital. As above noted, there is
no conclusive evidence regarding the horn rudiments
except that if present at all they appear to have been
not very prominent.
The most important feature by far is the slender
and simple structure of the zygomatic arch (fig. 369).
The malars give no evidence of the existence of an
infraorbital shelf; on the contrary, this region was
smooth, flattened, and not very prominent. Simi-
larly, the squamosal portion of the arch is shown,
giving a maximum depth of 30 millimeters and a
Figure 372. — Facial region of Eotitanotherium os-
borni and Brontotherium leidyi
One-fifth natural size. A, E.oshorni, Carnegie IVIus. 2859 (type), Uinta
B; B, B. leidyi, Nat. IVIus. 4249 (type), Chadron A.
width of 37. The very slender zygoma was apparently
nearly parallel with the sides of the skull, as in Doli-
chorhinus, but as in Rhadinorhinus diploconus it
lacked the infraorbital shelf; it also lacks the deep
vertical expansion seen in Telma.therium.
CHAPTER VI
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHEEES
SECTION 1. REVIEW OF THE ENVIRONMENT, GEO-
LOGIC SUCCESSION, AND GEOGRAPHIC DISTRIBU-
TION OF THE LOWER OLIGOCENE TITANOTHERES
In order to facilitate an understanding of the evolu-
tion of the skull and dentition of the Oligocene
titanotheres, a brief resume of the matter presented
in Chapter II is here given. The environment of the
lower Oligocene titanotheres, described in Chapter II,
was different from that of their Eocene ancestors. It
consisted mainly of the broad flood-plain region east
of the Kocky Mountains. The geographic range,
however, extended westward into the broad upland
plateaus west of the Rockies and northward over
British Columbia into Asia as far west as the eastern
part of Europe.
The conditions in the
Great Plains region appear
to have been eminently
favorable to the existence
of the titanotheres, because
the members of all the
numerous branches into
which this great family was
divided show an increase
in size, which is especially
conspicuous in the males.
Our ideas regarding the
geologic levels of the species
and the mutations of the
members of the four sub-
families are founded upon
the original observations of
Hatcher, who very care-
fully recorded the vertical
distribution of the types
and other specimens in his great collection for the
National Museum, which are fully enumerated else-
where in this monograph. To the records of these
specimens have been added records obtained from
other museum collections.
The faunistic subdivisions of the Titanotherium
zone, which forms part of the White Eiver deposits
and is of Chadron age (lower Oligocene), range in
thickness from 150 to 200 feet, as follows:
Feet
Chadron C (upper or Brontops robustus zone) 30-50
Chadron B (middle or Brontops dispar zone) 70-90
Chadron A (lower or Brontops brachycephalus zone) 50-60
The lower Oligocene deposits of the Chadron forma-
tion of the Great Plains were laid down on the irregu-
larly worn surface of the Pierre shale (Upper Cre-
taceous), which had been exposed to erosion for a long
time. Consequently the deposition of the Titanothe-
nwm-bearing beds was not uniform: it began at some
points earlier than at others, and the total thickness of
the Titanotherium zone accordingly ranges from 1 50 to
200 feet. There are also some discrepancies in the
records, which are doubtless due to irregularities of
deposition in the overflow and stream channel deposits.
The known areas where deposits of the Titanothe-
rium zone are exposed and where fossils occur are shown
on the accompanying map (fig. 373).
The change of form of the lower Oligocene titano-
theres was almost as great while this 200 feet of
Former land areas Former migration areas
Figure 373. — Map showing the areas in which remains of titanotheres have been found (solid
black) and areas in which, during Eocene and Oligocene time (oblique lines) titanotheres were
probably in migration
The general regions in which titanotheres have been found are the northwestern United States, the Gobi Desert (Mongolia),
Burma, and southeastern Europe.
sediment was being deposited as that of the Eocene
titanotheres while 2,000 feet of sediment was being
deposited. We infer that the average deposit of 200
feet of sediment in so many localities entirely deceives
us as to the length of lower Oligocene time. These
sediments were being laid down probably not for hun-
dreds but for thousands of years. During this long
period the titanotheres were certainly very abundant
over the entire western plains.
Without exception all the animals whose remains
are found at the base of the Titanotherium zone were
relatively small, and all had short and superficially
similar horns. The great increase in size observed be-
tween the animals of the beds of Chadron A and those
443
444
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
of Chadron C is made evident by a comparison of the
outlines of the skulls (figs. 389-393) and lower jaws
(figs. 395-397).
A2
Figure 374. — Comparison of upper Eocene and lower Oligocene titanotheres, showing
similar early stages in the evolution of the horns
Ai, A?, Reconstruction of the skull and jaw of an adult Protitanotherium emarginatum, Princeton Mus. 11242; upper
Eocene. Bi, Ba, Partly reconstructed skull ofa young Brontopsbrachycephalus, Nat. Mus. 4258; lower Oligocene.
One-eighth natural size.
In order to comprehend the extraordinary number
and the extremely varied forms of the titanotheres
that inhabited South Dakota in early Oligocene time
we must first imagine the existence of a vast conti-
nental region as the theater of evolution of these mi-
grating animals, a region far larger than any of the
comparatively small areas in which
the fossil remains have been col-
lected and which are shown by
the black areas on Figure 373.
A vast level or undulating
country, consisting of great flat
plains traversed by slow meander-
ing streams, bounded on the west
,,- — "V by mountain ranges, valleys, and
\ plateaus interspersed with active
/' volcanic peaks but allowing free
y' migration to the east, north, and
/A, south — such was the environment
""' ' I \ of the Oligocene titanotheres.
\^.'i; I'' J SECTION 2. INTRODUCTION TO
[\ ■■"'' THE ANATOMY OF THE SKULL
i AND THE DENTITION OF THE
i OLIGOCENE TITANOTHERES
I
\ HORNS: TRANSFORMATION, ELON-
1 GATION
Length oj the horns. — The grad-
ual evolution of the horns in the
Eocene titanotheres was followed
by their more rapid evolution in
the Oligocene titanotheres, until
they became the dominant and
central feature of the skull, con-
ditioning its entire architectiu-e.
With the development of the
horns as powerful weapons are cor-
related changes in the structure of
the nasals, of the zygomatic arches,
of the cranial vertex, of the occi-
put, of the vertebral spines and in
the entire anatomy of the anterior
portion of the body. The primary
divisions of the Oligocene titano-
theres as determined by length of
horn are as follows:
Menodontine group (short-horned) :
Teleodus, Brontops, Diploclonus,
Allops, Menodus.
Brontotheriine group (long-horned) :
M egacerops {" Symborodon"),
Brontotherium.
First stage of development. — The
horns in the Oligocene titano-
theres, as in the Eocene Dolicho-
rhinus and Protitanotherium, arise
at the junction of the frontonasal
suture, slightly in front of the
orbits, overhanging the sides of
the face (fig. 374). The primitive
horn section is an anteroposterior oval. The longest
diameter of all the earliest horn tips is parallel with
the long axis of the skull. The anterior edge of the
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
445
oval extends into the sides of the nasals (N); the pos-
terior edge of the oval subsides laterally toward the
frontals, lacrimals, and maxillaries {M).
Second stage of develop-
ment.— A low transverse
" connecting crest " arises
at the median bases of
the hornsand gives them
a more or less distinct tri-
lateral section consisting
of an antero-internal face,
an antero-external face,
and a posterior face.
These three faces are es-
pecially distinct in spe-
cies of Menodus, such as
M. trigonoceras . In Mega-
cerops, however, the con-
necting crest is not typi-
cally developed, so that no
internal angle (C) appears
{Megacerops acer, p. 545).
Sections of the horn. —
Each phylum takes on a
characteristic horn sec-
tion, which may be repro-
duced by bending a cop-
per or lead wire in the Figure 375
manner shown in Figure
376. The horn section is
always cut at right angles
to the perpendicular axis
of the horn and as near
the base of the horn as
practicable. Each horn
develops progressively a
characteristic section at
the base (fig. 399) and at
the tip. The four chief
types of horn sections
are as follows:
Short-horned:
Anteroposterior oval, subtriangular, oblique oval,
subtransverse oval : Brontops, Diploclonus.
Subtriangular oblique, subtriangular transverse:
Menodus, Allops.
Long-horned:
Subtriangular, suboval, transverse oval: Bronto-
therium.
Rounded, anteroposterior oval, flattened anteri-
orly, rounded posteriorly: Megacerops.
Forward shifting of horns. — The horns tend to shift
forward and absorb the nasals. Thus the general
correlation of horns and nasals is as follows:
Short, triangular, oval horns, elongated nasals; Bron-
tops, Menodus.
Long, oval or rounded horns, abbreviated nasals: Bron-
totherium, Megacerops.
-Sections at base of
horn in the six chief generic
types of Oligocene titanotheres
(B-G) and in the upper Eo-
cene Prolilanolherium emargi-
natum (A)
A, ProtitanoiheriumemaTginatum, Princeton
Mus. 11242; anteroposteriorly oval. B,
Brontops dispar, Nat. Mus. 4290; rounded,
obliquely oval. C, Diploclonus amplus,
Yale Mus. 12015a (type); rounded, trans-
versely widened, trihedral. D, Allops
serotinus, Yale Mus. 42.')1 (type); angu-
late, trihedral. E, Menodus giganteus.
Am. Mus. 505 (neotype); rounded, trihe-
dral. F, Megacerops acer, Am. Mus.
6348 (type): rounded, quadrate. G,
Brontotkeriitm plutyceras, Harvard Mus.
(type); transversely oval. All one-fifth
natural size.
In the extremely long-horned types, such as Bronto-
therium, the horns shift forward until they overhang
the anterior nares and finaUy the symphysis; they
thus absorb the nasals but retain their base of sup-
port on the greatly shortened maxillaries. Thus the
nasal angle (iV) disappears, and the horns acquire a
transverse oval section.
Horns in females. — In skulls from the higher geologic
levels of the Oligocene the difference between the
horns of the two sexes is rather marked; in skulls from
the lower Oligocene and from the Eocene the difference
is less. There is reason to believe that the horns were
at first alike in both sexes. In females the horns
exhibit an arrested stage of development. This fact
is most clearly shown by a comparison of two female
skulls of Brontotherium (Am. Mus. 1005, 1006) with
two male skulls (Am. Mus. 492, 1070). (See fig. 377.)
In many females the horns are imperfectly ossified at
the tips; in some they are pointed. In species of
Menodus the "connecting crest" is more constant and
more pronounced in males than in females, but in
females of Brontotherium the connecting crest appears
Figure 376. — Position of the standard sec-
tions and contours of Oligocene titanothere
skulls
N, Nasal contour; M,median section of nasals and connect-
ing crest; H, basal horn section; HN, oblique-longitud-
inal section, nasals to horn tip; P, section across parietal
verte.x; B, buccal section of zygoma.
to rise almost to the summit of the horn, as in the
female skull selected by Cope as the type of Menodus
peltoceras ( = Brontotherium curtum) (fig. 478). The
446
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
male skull of M. giganteus (Am. Mus. 505, 1066, 1067;
fig. 444) has stout triangular horns, whereas the female
skull (Am. Mus. 506; fig. 445), although a very large
specimen, has more slender horns, which are rounded
in section, and has very slender canines. Similar
sexual differences are observed in specimens of M.
trigonoceras.
Sport variations in the horns. — It is difficult to deter-
mine whether the internal "hornlet," or branch horn,
which appears on the inner side of the main horn in
specimens belonging to at least three different phyla
(Diploclonus, Menodus, Brontotherium) is a veritable
progressive character — that is, a rectigradation — or
merely a sport. This branching horn is believed to
be a generic character of the animal called Diploclonus
amplus by Marsh. Another example of internal
Figure 377. — Male and female skulls of Brontotherium
gigas
A, Am. Mus. 1006, 9 ; B, Am. Mus. 492, <? . One-twelfth natural size.
branching is that shown by a female of Brontotherium
gigas elatum (Am. Mus. 1006).
Effects of geologic crushing. — Vertical or lateral
crushing not only profoundly influences the entire
proportions of the skull but totally alters the shape
and angulation of the horns, as illustrated in two
examples of Menodus trigonoceras and Megacerops
copei (figs. 439, 394, G), one of which is crushed ver-
tically, the other horizontally.
NASALS; EXPANSION, ABBREVIATION
The hypertrophy of the horns and the compensating
atrophy of the nasals were pointed out by the writer
in 1887. Primitive nasals are invariably elongate, but
they show a fundamentally characteristic form, which
again distinguishes two groups. In the menodontine
group they are distally broad and truncate, as in
Menodus, primitively narrow and broadening distally,
as in Brontops, and moderately broad and laterally
decurved, as in Diploclonus; in the brontotheriine
group they are distally tapering and decurved, as in
Megacerops, and progressively abbreviate and pointed,
as in Megacerops and Brontotherium.
The shape of the nasals, however, does not sharply
distinguish all members of the two groups, as there is
more or less convergence between the members of
different phyla. In members of both groups the
tendency to shorten the nasals appears to be progres-
sive; it is less marked in Menodus and is carried to the
greatest extreme in Brontotherium.
In the Menodus group an age character is the distal
broadening, expansion of the nasals, as seen in the
comparison of young and old specimens of different
species of Brontops.
ZYGOMATIC ARCHES: EXPANSION, BUCCAL PLATES
Expansion oj the arches. — The progressive spreading
of the posterior portion of the zygomatic arches
(figs. 391-393) is a highly characteristic feature of
members of both the menodontine and brontotheriine
groups. The more primitive titanotheres in both
groups exhibit moderately expanded zygomata with-
out any rugose areas. The rugose development of
the zygomatic arch takes place pari passu with the
massive development of the horns. Thus the highest
degree of zygomatic expansion characterizes the great
brontotheres (figs. 392, 394) in which the horns reach
their maximum development. This becomes a sexual
character; the extremely robust and widely spreading
zygomatic arches of the more progressive species of
males present a contrast with the moderately expanded
arches that are associated with the feeble or imperfect
horns and small canines of the less progressive females.
In males and females of Menodus the same differences
are observed, but in a less marked degree than in
Brontotherium. The less expanded zygomata of the
female skulls give them a less brachycephalic and
more primitive character throughout the phylum.
Thus in the nasals, in the horns, and in the zygo-
matic arches the males always appear more progressive
and the females more primitive.
Zygomatic cephalic indices. — The expansion of the
zygomatic arches is so much more rapid than the
elongation of the skull as a whole that the breadth
across the zygomata nearly if not quite equals the
basilar length; thus a skidl which is really elongated,
like that of Brontotherium platyceras, presents a high
zygomatic index, whereas the brachycephalic general
character is less marked in the grinding teeth and in
the parietal vertex of the skull (compare fig. 390, B,
D,F).
OCCIPITAL PILLARS; AUDITORY MEATUS
The occiput. — Correlated with the progressive evo-
lution of the horns is the progressive transformation
of the occiput from the transversely convex contour
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
447
of the vertex to the deeply concave contour, corre-
lated with the development of the enormous lateral
pillars and supraoccipital rugosities to which are
Figure 378. — Occipital view of skulls in
different phyla of Oligocene titanotheres,
showing widening of the occiput and
development of its lateral pillars and
superior crests in the final stages
After Osboni. A, Menodus Jieloceras, Am. Mus. 6360
(type); Chadron A. B, Megacerops acer, Am. Mus.
6348 (type): ?Chadron C. C, Allops marski, Harvard
Mus.; ?Chadron B. D, Menodus trigonoceras, Am.
Mus. 1067,9; f^hadron C. E, Megacerops bucco, Am.
Mus. 6345a,d' (type); Chadron C. F, Brontotherium
gigas, Am. Mus. 492, .J ; Chadron C.
attached the powerful muscles and tendons of the
neck. Figure 378 illustrates these extremes of struc-
ture. In general the massiveness of the occiput is
directly correlated with the size, location, and function
of the horns. Thus in Megacerops acer (fig. 378, B)
poc.
.p.pe?:
p.oc.
B
» ^ft ~ TTLp.per.
Figure 379. — Influence of progressive brachy-
cephaly on the auditory region of perissodactyls
A, Dolichocephalic, Eguus caballus; B, mesaticephalic, Tapirus;
C, mesaticephalic, DiceroTMnus sumatrensis; D, brachycephalic,
Rhinoceros sondaicus. Note disappearance of mastoid portion
of periotic (m. p. per.) and inclosure of auditory meatus (e. a.m.)
inferiorly. Parallel changes occur in the titanotheres.
the occiput is narrow, slender, and slightly indented,
in keeping with the relatively slender horns, and
presents a very wide contrast to the broad, rugose.
448
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
deeply indented occiput of Brontoiherium gigas elatum
(fig. 378, F). _
Inclosed auditory meatus. — The progressive union of
the postglenoid and post-tympanic processes in the
titanotheres parallels that which we observe in a
comparison of the rhinoceroses, tapirs, and horses
(fig. 379). For example, in the skull of the primitive
Menodus Jieloceras (fig. 378, A) the external auditory
meatus is widely open below, paralleling the condition
which we observe in the Sumatran rhinoceros, R.
(Dicerorhinus) sumatrensis (fig. 379, C). In Bronto-
therium platyceras the auditory meatus is reduced to
a small foramen inclosed by a solid wall of bone,
paralleling the condition observed in the Javan
rhinoceros, R. sondaicus (fig. 379, D).
SEXUAL CHARACTERS COMMON TO All PHYLA
The following sexual characters, though common to
all phyla, are most conspicuous in titanotheres from
the higher geologic levels :
Male
1. Skulls larger.
2. Canines larger and more
robust.
3. Incisors larger, more con-
stant.
4. Nasals broader, more ro-
bust at tips.
5. Horns more powerful and
robust.
6. Connecting crest very
prominent.
7. Arches more widely ex-
panded.
8. Zygomatic-cephalic indices
higher, more brachy-
cephalic.
9. Occiput more robust, ex-
panding, and rugose.
10. Closure of cranial sutures
accelerated.
1. Skulls smaller.
2. Canines smaller, more
pointed.
3. Incisors smaller and more
variable.
4. Nasals narrower, less ro-
bust at tips.
5. Horns shorter, more point-
ed, less completely ossi-
fied at tips.
6. Connecting crest less
prominent.
7. Zygomatic arches less
widely expanded.
8. Zygomatic-cephalic indices
lower, more mesatice-
phalic.
9. Occiput less robust.
10. Closure of cranial sutures
retarded.
The incisors are apparently more variable and are
more likely to drop out of place in the females than
in the males, especially in individuals of BrontotJierium
(PL XIX). The smaller canines are among the most
persistent characteristics of the female (PI. XX).
Our observations do not confirm Hatcher's remark
that "a feeble internal cingulum" upon the premolars
is a female character. The entire grinding series ap-
pears to be relatively as large and as vigorously devel-
oped in females as in males. Between the females
and the males in the ascending series of BrontotJierium
there is a very marked and rather puzzling disparity
in the size of the skull.
TEETH: DISTINCTIVE FEATURES AND EVOLUTION
Incisors, superior and inferior, considered as phyletic
characters. — The strong or the feeble development of
the incisors and the presence or the absence of certain
members of the incisor series are two characters that
are distinctive of the phyla, genera, and species.
In contrast to the Eocene titanotheres, all the known
Oligocene titanotheres, except Teleodus, had only two
pairs of incisor teeth. As early as upper Eocene time
the reduction and the loss of incisors is foreshadowed
in ProtitanotJierium and Diplacodon by the hypertrophy
of certain pairs of incisors and the atrophy of others.
This liypertrophy, atrophy, and disappearance of the
incisors is graphically presented below:
0. 0. 0
Megacerops: „ „ „
, . , 0. 12. 13 , 0. 0. 0
Menodus: ~. — ^- to — 5 —
0 i^ i' 0 0 i'
Brontoiherium: -■ — ■ — 7; to „'.,-,
ii. 12. 0 0. 12. 0
0. i^. i^ 0. 0. i'
Broniops brachycephalus: . ' . ' ^ to
ii. ij. 0 0. 12. 0
Teleodus avus:
ProHlanotherium :
In the Broniops phylum the third lower incisor
(is) is apparently the first to disappear in the lower
jaw, and the third upper incisor (i^) the first to dis-
appear in the upper jaw. The jaw of Teleodus avus
(PL XIX, D) contains six incisors. The outermost
pair (is) have very short roots and insecure tenure,
so that further evolution in the same direction would
probably result in the crowding out of is. By far
the largest teeth with the longest roots are the second
incisors (12). Intermediate in size are the first in-
cisors (ii).
In the upper jaw of Brontoiherium (PL XIX, A) the
median pair (i') have apparently been lost; the per-
sistent teeth represent the second incisors (i^) and the
greatly enlarged third incisors (i^). In the lower jaws
of Brontoiherium a third pair of incisors (is) have
apparently been lost. The first pair (ii) have spread
apart, leaving a diastema in the midline; the second
pair (i2) remain large and usurp the position of the
third (is).
The form of the crown of the incisor teeth is also
highly distinctive. In members of the menodontine
group the incisor crowns are smoothly rounded or
conic, often laterally compressed, as in Teleodus
(PL XIX, D). In members of the brontotheriine
group the incisors are cingulate posteriorly.
Canines, superior and inferior, considered as sexual
characters. — The canines (PL XX) are highly dis-
tinctive of each phylum and of each genus, as they
differ widely in form and in function. The shape is
the same in both sexes, but those of the male are
always larger and much more powerful than those of
the female. For example, in Menodus giganteus
(Am. Mus. 505) the male tusks measure 62 by 34
millimeters, whereas the female tusks (Am. Mus. 506)
measure 40 by 21 millimeters. In Brontoiherium
elatum also, as shown by a comparison of five skulls,
the female tusks are about two-thirds the size of the
male tusks.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
449
All primitive members of the menodontine group
(such as Menodus or Brontops) have long, pointed
canines, whereas all primitive members of the bronto-
theriine group (such as Brontotherium leidyi) have short,
obtuse canines. Secondarily some progressive members
of the menodontine group (such as Brontops rohustus)
develop short, obtuse canines that are not readily
distinguished from the short, obtuse canines of the late
members of the brontotheriine group (such as Bronto-
therium gigas). This is an example of convergence.
Within each of these two groups the canines are
differentiated. The extremely elongate and powerful
canines of Menodus are readily distinguished from the
smaller canines of Brontops, as well as from the antero-
posteriorly compressed canines of Allops. Again, in
the brontotheriine group the robust canines of Bronto-
therium are readily distinguished from the diminutive
canines of Megacerops.
I Premolar evolution; retrogression, abbreviation. — In
i correlation with the abbreviation of the face (brachy-
i opy) seen in the titanotheres the premolars are sacri-
ficed to the evolution of the molars. This is observed
in the variability or loss of px, in the arrested molari-
zation of the premolars (or their failure to acquire the
complete molar pattern), and in the relative abbrevi-
ation of the premolar series as compared with the molar
series and as expressed in the premolar-molar index.
This evolution is just the reverse of that in the
dolichopic Equidae, in which the premolars evolve
more rapidly than the molars.
In the Menodus phylum the face is relatively
elongate (dolichopic) and the premolar index (50-53)
remains more constant. In the Megacerops and Bronto-
therium phyla the face is relatively abbreviate and the
premolar-molar index (42-46) is low and retrogressive,
although the premolars increase greatly in width.
Premolar-molar indices
Vm'-mV
Menodontine group (Menodus phylum)
Brontotheriine group (Brontotherium phylum)
Upper Titaiiotherium zone (lower Oligo-
Menodus giganteus
Menodus proutii
Telmatherium ultimum
Telmatherium cultridens
Eotitanops borealis
___ 50-53
49
.___ 61
61
63
Brontotherium curtum
42-46
47
Lower Titanotherium zone (uppermost
Eocene) .
47
Upper Wind River (lower Eocene)
Arrested molarization of
premolars. — In all Oligocene
titanotheres arrested molar-
ization is seen, first, in the
entire absence of a meso-
style on the ectoloph of the
permanent superior pre-
molars, although the meso-
style is present on the milk
premolars; second, in the
retarded development of
the tetartocones, especially
on p*. The retarded devel-
opment of the tetartocone
of p* may be adaptively cor-
related with the fact that
this tooth erupts much later
than p' or p^ (PL XXI,
figs. 405, 406; Carnegie Mus.
116). Nevertheless, p* is
nearly as broad as m\
whereas in Eocene titano-
theres, except those of the
very highest levels, p^ is much narrower than m'.
Figure 380. — Inferior as-
pect of chin in Manieoceras
A fragment from the upper Bridger,
Am. Mus. 1746, probably Manieo-
ceras manieoceras, showing large size
of canine roots, length of roots of ia,
and shortness of roots of ii and h-
One-half natural size.
The differential rate of molarization of the premolars is
one of the most characteristicdistinctions between phyla.
Each phylum has its own rate of molarization. In Bron-
tops the premolars transform very slowly. In Diploclo-
nus, Allops, and Menodus, respectively, they transform
with increasing rapidity. In Megacerops and Brontothe-
rium the premolars transform very rapidly. Thus in the
two extremes the retarded premolars of Brontops brachy-
cephalus with incipient tetartocones are readily distin-
guished from the progressive premolars of Brontotherium
leidyi with strongly developed tetartocones, although
both animals belong to the same geologic level.
Loss oj p\ with age. — The presence or absence of px,
which was much cited as a specific character by Marsh,
is rather an age character. This tooth is present in
many young skulls and absent in many old skulls, as
has been observed in specimens of B. brachycephalus,
B. dispar, B. robustus, Allops crassicornis, Menodus
giganteus, Brontotherium gigas. It lacks a firm hold
in its socket, and its root impinges against the obliquely
placed roots of the canines. This tooth comes into
use very early in Brontops, Menodus, and Bronto-
therium and tends to drop out early because all the
teeth protrude from their sockets as wear on them
450
TITANOTHEKES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
advances, and the roots of p^ soon lose their firm hold
in the alveolus. P^ is always a smaller and simpler
tooth than p^, which tends to crowd it out of place.
Molar evolution. — The Oligocene grinding tooth
evolves out of the primitive Eocene grinder and takes
its proportions from the skull. Thus we compare the
elongate, dolichocephalic grinding tooth of Menodus
FiGUEE 381. — Upper molars of Menodus giganteus and Alloys marshi
A, Menodus giganteus. Am. Miis. 496. First left upper molar, posterior view. This tootli was beginning
its eruption at the time of the animal's death; in life the tips of the cusps would soon have come into use.
The internal cusps (pr) are low; the outer wall of the tooth is produced into a long arc. This outer
wall wears down much faster than the inner cusps, and the tooth as a whole slowly rotates downward
and inward, the outer side moving faster than the inner side, so that in old animals the outer roots are
widely protruded and the surface of the crown is tilted inward.
B, AUops marshi, Am. Mus, 499. Second right upper molar, crown view. From the base of the metacone
a small projection, the "crochet, "runs forward into the deep medifossette; from the middle of the crown
opposite the mesostyle a second small projection, the "crista," runs inward and forward; from the base
of the paracone a third ridge, the "antecrochet," runs back toward the crochet and crista. Crochet,
antecrochet, and crista appear in both the deciduous premolars and the permanent molars of Oligocene
titanotheres along with the deepening medifossette. They are barely foreshadowed in the upper Eocene
Diplacodon and Rhadinorhinus.
Two-thirds natural size.
comparative measurements of the brachycephalic
Brontops robustus, the dolichocephalic Menodus gigan-
teus, and the brachycephalic Brontotherium gigas
elatum. (See p. 451.)
Upper molars; hypocone and cingulum. — ^The hypo-
cone of m^ appears as a low cusp on the posterior
cingulum in Brontops dispar and as a prominent cingu-
lum cone in Menodus (Diconodon,
Anisacodon montanus). In some speci-
mens of Menodus the hypocone of m^ is
separate and is surrounded by a cingu-
lum; in others it is small but distinct or
is confluent with the cingulum. Thus
the separation of the hypocone would
not appear to be a valid specific, much
less a valid and constant generic
character.
Cope observed that the strong or
feeble development of the cmgulum di-
vides the titanotheres into two parallel
groups (Cope, 1891.2, p. 9), which are
now recognized as follows:
Menodontine group {Menodus, Brontops,
etc.) : Cingulum strongly or distinctly
developed.
Brontotheriine group (Brontotherium,
Megacerops, etc.): Cingulum retro-
gressive, feebly developed, or wanting.
with the abbreviated, brachycephalic, transversely
spreading grinder of Brontotherium (fig. 382.)
Special characters. — The most exceptional character
is the vertical elongation of the ectoloph (figs. 227,
228, 381) which attains twice the height of theproto-
cone; thus a disharmonic crown is produced, hypsodont
on the outer side and brachyodont on the inner side;
this elongation of the ectoloph leaves a deep pit
(medifossette) in the central valley of the crown,
which is bounded by three secondary foldings of
enamel — anterior, median, and posterior — which are
comparable to but not homologous with the crochet,
antecrochet, and crista of the rhinoceros molar tooth.
These secondary folds are rectigradations which are
also slightly developed in certain upper Eocene
titanotheres. The medifossette is distinctly fore-
shadowed in the upper Eocene Diplacodon. A shallow
postfossette appears internal to the hypocone. On
the antero-internal border of the crown appears a
prominent cuspule which is comparable to the proto-
style; it never detaches itself from the protocone;
in the center of the crown are sometimes observed
vestigial or reversional traces of the protoconule and
of the metaloph. The terminology of the molars, as
compared with that in other perissodactyls, is set
forth in Chapter V (p. 263).
The correlation of dolichocephaly and brachy-
cephaly with tooth proportions is illustrated in the
In Menodus the cingulum is especially
strong; in Megacerops it is especially
feeble. The cingulum is thus a dis-
tinct phyletic or group character. It is not a
sex character, as Hatcher suggested (1893.1, p. 216).
Figure 382. — Extreme dolichocephalic (A) and brachyce-
phalic (B) types of upper premolar-molar series in Oligocene
titanotheres
A, Menodus trigonoeeras, Carnegie Mus. 3068, one-fourth natural size; B, Bronto-
therium gigas elatum. Am. Mus. 492, tooth row reduced to the same absolute length
as in A.
For example, in the female skull of Menodus giganteus
(Am. Mus. 506) the cingulum is quite as strongly
marked as in the male skulls (Am. Mus. 505, 1066,
1067). The cingtdum is less strong in Brontops and
Diploclonus than in Menodus and is almost obsolete
in Megacerops acer and Brontotherium, platyceras.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
451
Correlation of dolichocephaly and brachycephaly with propor-
tions of teeth
[MeasuremeDts in millimeters]
Brontops
robustos, Yale!
Mtis. 12048 I Field Mus.
(type) P 6927
Brontotherium
gigas. Am.
Mus. 492
Pmx to condyles
Width of skull across
zygomatic arches
Zygomatic index
Pi-m3
Dental index
Pi-p*
M>-m3
P*, ap. by tr
Sum of anteroposterior
measurements of m'-m^
compared with sum of
transverse measurements-
765
667
87
350
45
137
220
40X65
220 X 255
825
515
62
425
51
150
270
■ 50 X 68
» 249 X 246
830
740
89
353
42
130
241
47X72
241X277
• Measurement taken from Am. Mus. 605.
Lower molars. — The lower grinding teeth also
indicate either the brachycephalic or the dolichoce-
phalic proportions of the skull. This is especially
witnessed in the third lower molar (PI. XXII), in
which, for example, the relatively long, narrow form
in Menodus contrasts with the relatively broad,
robust form in Brontotherium. M3 is further dis-
tinguished clearly in the different phyla by the form
of the hypoconulid, or third lobe, which is more
lophoid in the Menodus group, more crescentic in the
Brontotherium group. Similarly the main crescents
are somewhat more open in dolichocephalic molars
and more closed or acute in brachycephalic molars.
The cingulum is strongly developed on the lower
grinders in members of the menodontine group and
feebly developed or obsolete in members of the
brontotheriine group.
The molarization of the lower premolars proceeds
step by step with the molarization of the upper
premolars. Thus the premolars acquire the molar
pattern slowly in the menodontine group and more
rapidly in the brontotheriine group.
The internal wall, especially of the third lower
molars, develops sharp crests (metacristid, entocristid^
fig. 383), which are similar in form and position to
those of certain other early perissodactyls, especially
the paleotheres and chalicotheres; but, with the
exception of Lamhdotherium, the titanotheres do not
develop the metastylid and entostylid, cusps which in
other perissodactyls arise by fissure of the metaconid
and entoconid respectively.
DEVELOPMENT OF THE SKUIL AND DENTITION
Stage 1. — The earliest known stage (PL XXIII) is
represented by a specimen in the Yale Museum,
which is a lower jaw containing the deciduous teeth
of a newly born animal. Three alveoli of deciduous
incisors and two deciduous premolars (dp2, dp3)are
in place; the latter are unworn. The third and last
deciduous premolar had not yet erupted and much of
it is buried in the jaw. The titanotheres, like many
other ungulates, apparently had but three deciduous
premolars on each side in the upper and the lower jaws.
The position of the incisor alveoli was the same as
in Teleodus avus (PL XIX, D) and suggests their
identification as dii, di2, dis. The opposite incisors
were separated in the midline, and dii lies much
below the plane of di2. Perhaps this indicates a
protrusile tongue. The first permanent premolar,
Pi, is just emerging. Possibly the deciduous canine
had been shed at an earlier stage. The deciduous
premolars (dpa, dps) have heavy external cingula.
The horizontal ramus of the jaw is very shallow; the
ascending ramus relatively very heavy.
Stage 2.- — Stage 2 is represented by a "calf" jaw
with alveoli for three deciduous incisors and for the
deciduous canines (Am. Mus. 510; PL XXIV, A),
which is provisionally referred to Menodus giganteus.
PROTOCONID
HYPOCONUL
•D ^ENTOCONID I METACONID PARACONID
\ I
METAC^RISTID |
Figure 383. — Third left lower molar of Bronto-
therium leidyi, showing the metacristid and
entocristid
Carnegie Mus. 93. One-half natural size
It includes the alveolus of permanent pi of the left side.
The deciduous premolars (dp2-dp4) are in place and
slightly worn. They are more molariform than the
permanent premolars that succeed them.
Stage S. — Stage 3 is represented by a "calf" jaw
of Menodus giganteus (Am. Mus. 509; PL XXIV, B).
The root of i2 (?) is in place; the remaining front teeth
are not preserved; the tip of the permanent canine is
embedded in the jaw, and behind it is a root that may
belong to pi; dp2-dp4 are in place; dp2 and dps are
considerably worn, but not dp4. Permanent pi is in
horizontal line with dp2, and although it is a very
small tooth it is probably the one that is present in
adult titanotheres. Mi lies nearly ready to cut the
gum.
Stage 4- — One of the yoimgest Imown titanothere
skulls (fig. 384) is in the Musemn of the University
of Wyoming (No. 4). It was collected by Mr. W. H.
Reed in HeU's Half Acre, Natrona County, Wyo.,
from a low level in the Titanotherium zone.
452
TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
The large size of the first true molar (ectoloph
ap. 57 mm.) indicates a larger animal than 5. hracTiy-
cephdlus; it is of a size equaled in B. dispar (Nat.
Mus. 4290). The horn swellings are small knobs,
not much larger than those of very old individuals of
Manteoceras manteoceras. The horn swelling on each
side consists of a protuberance on the anterior tip
A2
Figure 384. — Development of jaws and teeth, stage 4
Skull and teeth of female calf, Univ. Wyoming Mus. 4, referred to Bronlops iispar.
Skull and lower jaw, one-third natural size; Aj, upper teeth, one-half natural siz'
lower teeth, one-half natural size.
of the frontals, which is embraced in a corresponding
excavation of the expanded posterior border of the
nasals. The free portions of the nasals are short
and rounded distally, with sharply decurved lateral
borders. The sutures in the region of the horn — that
is, the nasofrontal, lacrimal, and maxillary sutures —
conform to the plan exhibited in the National Museum
skull 4258 and in skulls of other young titanotheres.
The upper jaw shows the permanent incisor i' (which is
exposed in the specimen but was possibly beneath the
gum in life) ; a tooth is just coming into place which is
apparently p' of the adult; three deciduous premolars
(dp^, dp', dp*) are in place. The lower jaw shows \2, (?)
pi, and dp2, dps, dp4. M' and mi are buried in the
jaws. Measurements of this specimen are as follows :
Measurements of deciduous teeth 0/ Brontops dispar {Univ.
Wyoming Mus. 4)
Upper jaw
Millimeters
P'-dp< 132
Pi, ap. by tr 18X18
Dp2, ap. by tr 29X25
Dp3, ap. by tr 37X32
DpS ap. by tr 44X33
Dp*, tr. (across mesostyle) 40
Permanent m', ectoloph, anteroposterior. 57
Permanent m', ectoloph, height of meta-
cone 45 +
Nasals to middle of horn 75
Nasals, free breadth (estimated) 55
Pmx to condyles (rough estimate) 305
Lower jaw
Front edge of symphysis to angle 284
Height condyle to angle 145
P,-dp4 128
Dpi, ap. by tr. (trigonid) 11X10
Dp2, ap. by tr. (trigonid) 31X15
Dp3, ap. by tr. (trigonid) 37X20
Dp4, ap. by tr. (trigonid) 49X25
Stage 5. — A more advanced stage is shown in
a young jaw (Carnegie Mus. 124; PI. XXIV, C)
referred provisionally to Brontops dispar. The
permanent incisors ii and 12 are just coming in,
while the deciduous incisors and canines have
probably been shed. The permanent canine lies
j ust below the surface. The tooth designated ? dpi
although associated with mUk teeth, appears to be
the permanent pi. The milk molars dpo-dp^ are
worn. Ml is just protruding. M2 lies buried in
the ascending ramus, below the coronocondylar
sinus.
Stage 6. — A later ontogenetic stage is illustrated
in a remarkably complete skull and jaw in the
Carnegie Museum (No. 116), which were de-
scribed by Hatcher in 1901 (1901.1; figs. 385,
386). The specimen was found on Warbonnet
Creek, Sioux County, Nebr., near the base of
the Titanotlierium zone. The reference to Bron-
tops hrachycepJialus is provisional.
In the side and top views (fig. 385) it is seen
that the horns are formed by the overgrowth of
the frontals upon the nasals, as in the Eocene Man-
teoceras. The lacrimal is expanded, and its outer
ridge is continuous with the external ridge of the horn.
The parietals extend forward upon the frontals. The
interparietal is apparently distinct. The occiput is
shown in Figure 386.
EVOLUTION OF THE SKULL AND DENTITION OE OLIGOCENE TITANOTHERES
453
The construction of the skull conforms in its under-
lying plan to that of Eocene titanotheres, differing
chiefly in the shortening of the face, the lengthening of
the midcranium, and the widening of the skull top.
shed. The deciduous premolars (dp-, dp', dp'') are well
worn. M' is fully in place. P^ p^ p' lie embedded;
permanent p| are apparently represented in the
well-worn teeth immediately in front of dpf. In
FiGUBE 385. — Development of jaws and teeth, stage 6
Broniopsf brachycephalus?, Carnegie Mus. 116: young skull and Jaw. One-fourth natural size. After Hatcher. Level near the base
of the Chadron formation (Titanotherium zone). Ai, Side view; An, top view.
Dentition. — The deciduous teeth were about to be
shed, and their roots protrude widely. In the upper
dentition the deciduous incisors, di-(?), di'(?), are
small and round topped. Behind them the permanent
canine lies buried. The deciduous canine has been
the lower jaw the milk incisors were probably in the
gums; at least their alveoli must have been shallow.
The permanent canine is beginning to come in. The
milk molars are well worn, and mi is in place. P2,
P3, p4 lie on descending levels in the jaw, the first
454
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Hearing eruption. M2 also lies embedded, and nis
is beginning to form.
Hatcher's original description of the skull is as
follows :
When viewed from above the frontals are much the more
conspicuous elements. They are bluntly rounded posteriorly
and are projected far backward beneath and between the
lateral anterior projections of the parietals. Anteriorly the
frontals are continued into two long lateral projections which
extend beyond the orbits, overlie the posterior and lateral
margins of the nasals, and give rise to the pair of horn cores
that form such characteristic features in the Titanotheridae.
The nasals are arched superiorly, concave inferiorly, with
rather long posterior extensions interposed between the frontal
horns. In the present specimen the nasals are very thin along
their inner margins but much thickened externally and posteri-
orly, where they give the chief support to the horns. They
are slightly shorter than the premaxillaries and somewhat
emarginate anteriorly.
The parietals are not so broad as the frontals. They are
deeply emarginate anteriorly and posteriorly, where they are
separated by the interparietal portion of the supraoccipital
Figure 386.
-Occiput of young skull of Brontops
brachycephalus?
Carnegie Mus. 116. One-fourth natural size. The exoccipitals meet above the
foramen magnum. The wide supraoccipital affords attachment to the power,
ful ligamentum nuchae, the recti capitis lateralis, complexus, and other neol;:
muscles.
much as in Equus. From the above description it will be seen
that the anterior border of the parietals overlies and incloses
laterally the posterior border of the frontals, while the anterior
border of the latter has a like articulation with the nasals, the
relative position of these bones being similar to that of the
shingles of a roof.
The zygomata are rather broadly expanded and are composed
about equally of the malars and squamosals.
When seen from the side the skull appears somewhat low,
with an abbreviated facial region and a rather long posterior
portion. The occipital crest and anterior frontal regions are
each somewhat elevated. The nasals appear rather deep and
send down the inferior and posterior projection, which articu-
lates by suture with the superior border of the maxillary. The
infraorbital foramen lies wholly within the maxiUary. The max-
illonasal suture is opposite the middle of the orbit. The lac-
rimal is rather large. The malar is long and thin; anteriorly
it has an extended contact with the maxiUary and posteriorly
with the squamosal portion of the zygoma. The squamosal
rises high above the external auditory opening and overlaps
the side of the parietal throughout most of its length. There is
a long, thin, transversely expanded postglenoid process and a
shorter and proportionately stronger post-tympanic process of
the squamosal. The tympanic is absent, having been lost from
the present specimen. The periotic is present, and its para-
mastoid portion appears externally between the post-tympanic
and parocoipital process. Just above this there is another
small bone which is continued into a long pointed process
inserted between the exoccipital and the squamosal and pa-
rietal; it probably became coossified later with the periotic,
but in the present specimen it is seen as a separate bone, as
shown in Figure I and in Plate VII. [See figs. 385, 386.)
The exoccipitals are rather large and support the paroccipital
process and the occipital condyles. The latter are ossified
from two distinct centers, the articular portions bearing distinct
epiphyses, as shown in the accompanying figures.
Seen from behind the occiput is low and broad. The condyles
are widely separated by the foramen magnum, which is much
broader than deep. The condyles are entirely supported by
the exoccipitals, which rise and meet in the middle line above
the foramen magnum, entirely excluding the supraoccipital
from any part in the formation of the superior border of that
opening. The supraoccipital is very broad and low. The
occipital crest is nearly flat above but broadly emarginate
posteriorly. (See fig. 1 of the text, and PI. VIII.)
Inferiorly the palate is seen to be formed anteriorly by the
very short premaxiUaries, for the most part broken away in the
present specimen, and by the maxillaries, between the posterior
lateral extremities of which are inserted the palatines. These
form the posterior median portion of the roof of the palate and
send backward on either side a lateral projection along the inner
sides of the maxillaries and pterygoids, which are continued
nearly to the posterior end of the basisphenoid.
The vomers are continued far back as a thin plate resting
upon the pre- and basisphenoids and sending downward a
thin, knifelike median bony septum. The basisphenoid in the
present specimen is entirely free from the basioccipital, the
suture being open, and the basioccipital had dropped out and
was lost before the specimen was found. The absence of the
basisphenoid and tympanic bones makes it impossible to describe
and locate the various foramina of this region of the skull.
Stage 7. — Still more advanced is the stage shown in
Am. Mus. 497 (PL XXV, A), deciduous and permanent
upper teeth of Menodus giganteus. The first tooth of
the series is the permanent pS as shown by its exact
agreement in measurements and in characters with
the first premolar n adult skulls of M. giganteus (Am.
Mus. 505, 506). PhylogeneticaUy this tooth may have
been forced into association with the deciduous pre-
molars dp^, dp' through the abbreviation of the muz-
zle region and the consequent crowding backward of
the permanent canine. In this specimen the perma-
nent canine, which is still buried in the jaw, lies closely
appressed against the second permanent premolar.
The serial homology of this tooth as p^ is also estab-
lished by its measurements, as compared with the adult
Menodus giganteus. In this as in many other mammals
there was probably no deciduous predecessor of p^
In the specimen here figured m' is coming into
place. The identity of these teeth is also established
by comparison with the adult M. giganteus, so that
there is no doubt that dp*, dp', dp^ are correctly
identified.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES 455
Comparative measurements, in millimeters, of deciduous and permanent dentition of species of Menodus and Brontops
[The measurements given in the first five columns were made on the
ectolophl
P>
pj
pS
Ml
Dp!
Dp>
Dp<
Dpi
M,
P>-dp<
Pi-dp4
22
23
32
35
45
46
72
71
77
76
M. giganteus, Am. Mus. 506, 9
63
45
46
»47
47
51
67
65
66
26
21
40
46
174
58
62
62
66
136
160
40
68
138
48
130
B sp ' 'Yale Mus
"135
B. sp.?, Univ. Wyo. 4 ._- -
18
57
29
39
47
49
57
132
128
SUMMARY OF THE REPLACEMENT OF THE TEETH IN
OLIGOCENE TITANOTHERES
The following is a summary of the order of succes-
sion of the upper and lower teeth as observed in the
six juvenile stages represented in Plates XXIII and
XXIV and in Figures 384 and 385.
1. The three deciduous incisors (di-^, f, f) have
the same relative position as the permanent incisors
(Ml i> f) ^ Teleodus avus. They were shed very
early.
2. The deciduous canines, known only from their
alveoli in one specimen (PL XXIV, A), were shed
perhaps even earlier than the deciduous incisors
(PI. XXIII, XXIV).
3. No evidence of deciduous predecessors of p' and
Pi has been observed either in Eocene or in Oligocene
titanotheres, and in these, as in other ungulates, there
were probably only three and not four deciduous pre-
molars.
4. The permanent p^ came into place soon after
the deciduous dpf , f and functioned with the decid-
uous series.
5. The first true molars (m^) came into place be-
fore the deciduous premolars had been replaced. At
later periods mf and mf came into place successively,
so that in old animals m-^ is greatly worn, whereas
mf is but little worn.
6. The fourth premolar (p|) follows the general
mammalian rule of coming in late.
In the Oligocene titanotheres, as in many other
mammals, the second, third, and fourth deciduous
premolars (dpf, f, f) are much more molariform
than the permanent premolars (pf, |, |) which re-
place them. Thus molarization of the deciduous
premolars is observed even in the middle and lower
Eocene titanotheres {LamMotherium, Palaeosyops)
and is equally characteristic of the Oligocene titano-
theres. In Menodus giganteus (Am. Mus. 497, PI.
XXV, A) the third deciduous premolars (dpf) are like
molars, dp' and dp* having prominent mesostyles and
large, distinct tetartocones.
101959— 29— VOL 1 32
Figure 387. — Stages of wear in the adult upper grinding teeth
of Oligocene titanotheres
The following specimens, all drawn to the same length, show the progressive degrees
of wear from the young adult (X) to the very aged (XV) ontogenetic stage of the
dentition: X, Allops crassieornis, Nat. Mus. 4289 (type); XII, Brontops roiustus,
(type), Yale Mus. 12048; XIII, Menodus giganteus. Am. Mus. 505; XIV, Allops
serotinus. Am. Mus. 620; XV, Brontops brachycephaJus, Nat. Mus. 4947. (See
p. 456 )
456
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
STAGES OF WEAR OF THE ADULT GRINDING TEETH
Seven early and adolescent stages in the ontogeny
of the dentition have been described above. In the
eighth and ninth stages (adolescent) the deciduous
premolars (dp^ dp^ dp'') are gradually replaced by
the permanent premolars p^, p^, p*. In the tenth
ontogenetic stage (X) of the young adult the internal
cusps of the second and third but not the fourth
premolars are beginning to show the dentine through
the enamel surface; the last molar is but little worn.
In the very aged fifteenth ontogenetic stage (XV)
XII. Twelfth ontogenetic stage: Broniops robustus
Yale Mus. 12048 (type).
X. Tenth ontogenetic stage: Alloys crassicornis,
Nat. Mus. 4289 (type).
AGE AND OTHER CHARACTERS COMMON TO BOTH SEXES
OF TITANOTHERES OF ALL STRATIGRAPHIC LEVELS
Age characters. — The age characters are naturally
much more conspicuous in males than in females. In
both sexes the adults of one geologic generation follow
the general law of anticipating the advanced muta-
tions or specific stages reached by adults of higher
geologic levels. Thus the variability of a tooth on a
Figure 388. — Skull contours showing extreme divergence between Menodus giganteus (A), a final term of
the menodontine series, and Brontotherium platyceras (B) , a final term of the brontotheriine series
In Menodus the opposite borders of the cranial roof diverge anteriorly, the horns are short and trihedral, the zygomata but little expanded,
and the occiput not greatly produced backward. In Brontotherium the opposite borders of the cranial roof are nearly parallel, the
horns very long and flattened, the zygomata widely expanded, and the occiput greatly produced backward. One-tenth natural size.
nearly the whole enamel surface of the crowns of
p'-m' inclusive has been worn away, so that the
dentine is very widely exposed; in m^ both the pro-
tocones and hypocones are much worn. The inter-
vening stages show intermediate conditions, as follows
(fig. 387) :
XV. Fifteenth ontogenetic stage: Brontops brachyce-
phalus, Nat. Mus. 4947.
XIV. Fourteenth ontogenetic stage: Allops serotinus,
Am. Mus. 520.
XIII. Thirteenth ontogenetic stage: Menodus giganteus,
Am. Mus. 505.
lower geologic level is prophetic of its absence on a
higher geologic level. This variability is especially
displayed in retrogressive structures such as the
degenerate incisor teeth in the Brontops series, as is
shown by the following formulas:
Juvenile incisors, I{i^ .
Adult incisors, I|^}- .
The incisive teeth tend to drop out in the adults, as
observed in the type of Diploclonus tyleri Lull.
Thus among the age characters are the foUowrng:
(1) Increasing size and rugosity of the skull, arches,
horns, and nasals; (2) distal expansion and rugosity of
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
457
the tips of the nasals; (3) loss of variable and vestigial
teeth, incisors and premolars, in old age; (4) union of
the anterior caudal vertebra with the sacrum to form
four sacrals (Hatcher, 1893.1, p. 217).
In general, growth is in a high degree differential —
that is, proportions constantly change, as, for example,
in the gain of width over length, in the rapid increase
of the horns in length, and in the relative gain in the
length of the nasals, although only in a few specimens
have we sufficient material to measure these differ-
entials in growth.
Abnormal sport and reversional characters. — Among
the abnormal characters may be observed the following :
(1) Reversional or abortive protoloph and metaloph
on the superior premolar teeth; (2) abnormal redupli-
cation of tetartocones on superior premolar teeth ob-
served in specimens of Brontops dispar, Allops cras-
sicornis, Menodus varians; (3) progressive rectigrada-
tional or anomalous reduplication of horns as observed
in specimens of Diploclonus, Menodus, Brontotherium.
SECTION 3. DIVISION OF THE OLIGOCENE TITANO-
THERES INTO GROUPS AND SUBFAMILIES
CHARACTERS OF THE SKUII AND TEETH OF THE MENO-
DONTINE AND BRONTOTHERIINE GROUPS
The following study of the characters of the skull
and teeth enables us to divide all the highly varied
forms of Oligocene titanotheres into two great groups,
the menodontine and the brontotheriine, which sepa-
rated from each other in Eocene time. (See p. 467.)
These characters, which are much more pronounced in
male than in female skulls, are seen to be the direct
and indirect results of one or other of the opposing
principles of skull and tooth transformation described
on pages 254-262, 450, namely:
1. Dolichocephaly and dolichopy versus brachy-
cephaly and brachyopy; cyptocephaly.
2. Differential transformation of canines (p. 448),
premolars, and molars. (See figs. 405, 406.)
3. Differential development of horns, nasals, orbits,
zygomata, etc. (See figs. 389-394.)
Primary groups oj the Oligocene titanotheres
MeDodontine group (Teleodus, Brontops, Diplo-
clonus, Allops, Menodus)
Brontotheriine group (Megacerops ("Symborodon"),
Brontotherium)
Canines -
Opposite grinding series as seen in psflate
view.
Upward flexure of premolars as seen in
side view.
Antorbital region of skull
Premolar series (length)
Internal cusps of upper premolars
Ectolophs of grinding teeth
Hypocone of m^
Internal cingula
External cingula
Zygomata
Horns
Malar-Iacrimal bridge over infraorbital
foramen.
Anterior narial cavity
Backward prolongation of occiput behind
zygomata.
Jaw
Pi
Convexity, top of parietals
Orbit
Skull vertex
Skull vertex, side view
Proportions of molar teeth
Primitively pointed, recurved, not closely
approximated toward median line.
Not strongly arched anteriorly toward
median line.
Slight to moderate
Relatively elongate
Relatively longer
Steep-sided (Menodus); moderately so
(Brontops) .
Somewhat more vertical
Often surrounded by cingulum
Sharp on premolar teeth
Usually pronounced
Slight to heavy, often deep
Short, diverging obliquely outward and
forward .
Usually very broad
Broad and open
Moderate
Angle sharply produced backward
Somewhat more elongate
Absent
Medium to large
Divergent anteriorly, convergent pos-
teriorly.
More concave
More elongate
Primitively short, bulbous, with swelling
posterior cingulum, closely approxi-
mated.
Strongly to very strongly arched toward
median Hne.
Very pronounced.
Abbreviated (Brontotherium) to very
abbreviated (Megacerops) .
Relatively shorter.
Very low, robust, subcircular.
Sharply depressed to crown.
Very heavy, triradiate, continuous with
cingulum.
Often less developed and rounder.
Reduced or absent.
Broad to extremely expanded and flat-
tened.
Long, usually more erect.
Usually very narrow.
Becoming very high and narrow.
Moderate to extreme.
Angle less produced backward, more
slender.
Somewhat abbreviate.
Pronounced.
Medium to small.
More or less parallel.
More convex.
More expanded transversely.
458
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Features of the jaws. — As most of the jaws found
were entirely dissociated from the skulls, it is difficult
to determine the generic and even more so the spe-
cific references of many separate jaws. Observation
should be directed first to the prevailing dolichoce-
phalic or brachycephalic proportions, to the shallow-
ness or depth of the ramus, to the slenderness or
massiveness of the angular region, and especially to
the depth or shallowness of the chin region. In
general, members of the menodontine group have a
deep symphyseal or chin region, whereas members of
the brontotheriine group have a shallow chin.
The generic diagnosis of remains that include a full
series of teeth is relatively simple. Specific diagnosis is
partly dependent on size. In order to associate a
lower jaw with a skull the measurement should be
Characters of the jaw that distinguish members qf the menodontine group from members of the brontotheriine group
taken from the glenoid cavity to the front of the
superior canine and from the mandibular condyle to
the posterior face of the inferior canine. If all the
other characters are properly determined and the
progressive stages of the superior and inferior pre-
molars correspond, this method of associating the jaws
with the skulls is reliable. This method, however,
can be used to advantage only on skulls and j aws that
are not distorted by crushing. Usually the most
practicable measurements are the following:
Upper: Front face of canine to middle of posterior fossa
of m^ (for hypoconulid of ma).
Lower: Rear face of canine to tip of hypoconulid of ma.
The following table shows the chief characters of
the jaw that distinguish the members of the two
groups :
Horizontal ramus
Symphyseal region in
side view below men-
tal foramen.
Region of angle
Incisors
Pi in fully adult jaws-
Diastema in front of pi
Canines __
External cingulum of
canines, premolars,
molars.
Upward flexure of
premolar series.
Menodontine group
Deep
Convex..
Barely pro-
duced.
I3
Present
Present
Slender.
Not sharply
defined.
Shallow
Very shallow..
Usually produced
downward.
Usually present
Present, wide -.
Slender to short, stout.
Present in early types;
lost in B. robustus.
Slight or moderate.
Shallow.
Shallow.
? Produced,
truncate.
Present
Present
Slender
to short,
stout.
Intermedi-
ate.
Slight or
moderate.
Allops
Intermediate -.
Intermediate to con-
vex.
Intermediate -
l2_l.— -.
Present or absent
Absent
Conic to compressed
anteroposteriorly.
Present-- -
Very slight
Deep...
Fuller..
Produced backward
and downward into
a convex elbow. Pos-
terior border obli-
que.
lo (typically)
Present or absent
Absent
Conic -
Strongly marked-
Brontotheriine group
Short, massive.
Broad, posterior
border vertical.
1 lo
?Absent..
7 Absent.
(?)
Brontotberium
Massive, deep poste-
riorly.
Very shallow.
Broad, posterior border
often vertical.
Present or absent.
Absent.
SwoUen at base with
massive p 0 s t e r i or
cingulum.
Absent.
Typically pronounced.
EVOLUTION OP THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
459
Allops walcotti
Figure 389. — Skulls of the menodontine group
Side view. One-twelfth natural size.
Subfamily Brontopinae: A, Brontops bracJiycepMlus, °, a very small and primitive member of the Brontops phylum, with mcipient
horn swellings, long nasals, and slender canines; Chadron A. B, Brontops dispar, an intermediate stage with moderately developed
horns; Chadron B. C, Brontops robustus, final stage of this phylum; a very massive brachycephalic skull with stout forward
directed horns, short, thick nasals, short, heavy occiput, and expanded zygomata; canines short and thick; Chadron C. D, Viplo-
clonus bicornutus, referred to this genus partly because of the accessory horn swelling (h'). The skull contour suggests that of
Menodus. E, Diploclonus amplus, a massive short skull recalling Brontops robustus but possessing an accessory horn swellmg and
verv short nasals. Occiput long.
Subfamily Menodonfinae: F, Allops walcotti, a primitive dolichocephalic form, supposed to be ancestral to the Allops phylum; Chadron
A G, Allops marshi, in general contour intermediate between Brontops dispar (B) and Menodus (H), thought to be related to
Allops serotinus and Allops crassicornis. H, Menodus giganteus, latest stage of the Menodus phylum; uolichocepnalic, horns trihedral
in basal section, nasals long, zygomata not expanded, canme long, cheek teeth with sharp external cingula; Chadron O.
In all members of this group the skull top in side view is deeply concave. The cheek teeth almost always have sharp external cingula.
460
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Megacerops copei "
Megacerops acer
Figure 390. — Skulls of the brontotheriine group
Side view. One-twelfth natural size.
Subfamily Brontotheriinae: A, Srontotherium leidyi, a primitive member of the Brontotkerium phylum with short horns transversely oval in
section, long nasals tapering and decurved, canines short and swollen, two stout incisors, and cheek teeth with but little external cingulum;
Chadron A. B, Brontotkerium gigas, a progressive brontothere with long erect horns transversely oval in basal section, nasals of intermediate
length, zygomata widely expanded, occiput long, and a marked parietal convexity; Chadron C. G, Brontotherium curium, a highly advanced
brontothere with very long horns far in front of the orbits and much flattened anteroposteriorly, nasals short, skull top long, and zygomata
much expanded; Chadron C. D, Brontotherium (.peltoceras) curtum, a female brontothere with skull short and massive, horns very thick, high
connecting crest, and nasals short.
Subfamily Megaceropinae: E, Megacerops copei, long horns thick at the base, nasals long and thin, canines very short and swollen, premolar
series upturned, and zygomata massive. F, Megacerops acer, horns of moderate length but very thick at the base, nasals short and thick,
premolar series upturned, zygomata heavy, midparietal swelling prominent.
Compared with the Menodontinae, members of this group usually have the skull top less deeply concave, and many have a parietal swelling.
The horns are commonly transversely oval in section, the premolar series upturned anteriorly, and the canines in males short and swollen.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHBRES
461
Diploclonus amplus
Allops serotinus
Menodus giganteus
Allops walcotti
Allops marshi
Brontops brachycephalus Brontops dispar
FiGUBB 391. — Skulls of the menodontine group
Top view. One-twelfth natural size.
Subfamily Brontopinae: A, Brontops brachtjcephalus, a primitive stage witli horns very short and nasals long and tapering. The specimen is a
female, and the zygomata are not much expanded; the skull top is rather slender, although the skull as a whole Is broad. B, BToniops
dispar, an intermediate stage with short horns rounded in section, nasals and frontals broad, and zygomata expanded. The midparietal
crest is constricted, as in many other members of this family. C, Diploclonus amplus, a highly specialized brachycephalic stage, resembling
Brontops robustus in general proportions. The horns are w ide and flattened at the base and pointed at the tips and bear accessory hornlets
on the antero-internal portion. The nasals are wide and short.
Subfamily Menodontinae: D, Allops walcotti, a primitive dolichocephalic type (thought to be ancestral to the Allops phylum) with small horns
elongate oval in section, nasals long and tapering, zygomata slender, and skull top narrow. E, Allops marshi, horns widely trihedral in
basal section and directed outward and upward, skull top fairly broad, zygomata gently expanded. F, Allops serotinus, a specialized stage
with long outward-directed horns, wide, short nasals, wide frontals, and moderately stout zygomata. G, Menodus giganteus, the terminal
stage of the Menodus phylum, with skull dolichocephalic, skull top long, zygomatic expansion moderate, nasals long and distally wide,
horns sharply trihedral in basal section, the connecting crest lying in the plane of the posterior face of the horns.
462
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
B i_^.XT/^XX:_/ A ic/^^\3 c
Brontotherium (peltoceras) curtum ?q Brontotherium leidyi ?? Brontotherium curtum
Figure 392. — Skulls of the brontotheriine group
Top view. One-twelfth natural size.
Subfamily Brontotheriinae: A, Broniotheriitm leidyi, a primitive stage with skull top long, small horns transversely oval at the tips, nasals
long and tapering, frontal region wide, zygomata not expanded; Chadron A. B, Brontotherium (peltoceras) curtum, skull top long but
zygomata arching widely, short massive horns confluent with the high connecting crest and pointed at the tips, nasals reduced and pointed.
C, Brontotherium curtum, a very large male skull with widely expanded zygomata, widely flaring flat-oval horn.s, very short nasals, wide
frontal region, occipital pillars widely expanded transversely.
Subfamily Megaceropinae: D, Megacerops acer, skull short with widely expanded zygomata, horns cylindrical with but little connecting
crest, nasals short and wide, skull top wide with parallel edges. E, Megacerops bucco, 9; resembles in general the preceding type but
has larger horns. F, Megacerops bucco, J , a large animal with greatly expanded zygomata and rather feeble cylindrical horns.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
463
^ Allops serotinus ■"■ Brontops brachycephalus " Brontops robustus
Figure 393. — Skulls of the menodontine and brontotheriine groups
Palatal view. One-tweirth natural size ■ ^ , „ i
Menodontine group: A, Brontops brachycephalus, brachycephalic, grinding teeth of moderate width. B, Bronlops robustus, skull very large
and brachycephalic, grinding teeth wide, incisors and canines massive. C, Allops serotinus, skull more elongate, grmding teeth of inter-
mediate proportions. D, Menodus giganteus, dolichocephalic, grinding teeth elongate, tooth rows rectilinear.
Brontotheriine group: E, Megacerops bucco, skull base very brachycephalic, grinding teeth broad, tooth rows curvilinear, fourth premolar sub-
molariform, canines small. F, Brontotlierium curtum, skull very large, skull base long but tooth rows curvilmear, grmdmg teeth wide,
fourth premolar submolariform.
464
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Brontops dispar
Diploclonus amplus
Allops walcotti
Figure 394. — Skulls of the menodontlne and brontotheriine groups
Front view. One-twelfth natural size.
Meuodontine group: A, Brontops dispar, horns short and thick, suhcurcular in section and pointed, nasals of moderate size, zygomata
moderately expanded, lateral incisors present, canines relatively long and conical. B, Brontops robustus, skull massive, horns comparatively
short and transversely expanded, with thick rugose tips, nasals short and robust, zygomata heavy, two large incisors, canines short, thick,
and conical. C, Diploclonus amplus, horns divergent, very wide at base and rapidly narrowing to the pointed tips, an accessory hornlet,
nasals small, zygomata expanded. D, Allops walcotti, skull small and slender, horns very small, nasals broad, zygomata deep and slender.
E, Allops serotinus, long divergent horns narrowing but little toward the tip, zygomata of moderate size, canines slender and pointed.
F, Menodus giganteus, long divergent horns narrowing rapidly to pointed tips and trihedral in section, nasals expanded distally, zygomata
relatively small and deep, canines (not shown) elongate pointed, incisive border edentulous.
Brontotheriine group: O, Megacerops copei, J, horns long, erect, and cylindrical, with no connecting crest, nasals thin, zygomata not widely
expanded (diminished by crushing), canines small, rounded, and close together, premolars upturned anteriorly. H, Brontotherium gigas,
cf , skull very wide, zygomata enormous, horns long and wide, nasals of intermediate length, two incisors on each side, canines short and
swollen. I, Brontotherium curium, 9?> horns short but very broad and massive, with high connecting crest, nasals short and narrow,
zygomata not widely expanded, canines swollen.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHBRES
465
CHAEACTERS AND EEIATIONS OF THE SUBFAMILIES
Each of the major groups described above divides
into two separate branches to which Osborn gives the
rank of subfamilies, as follows :
Menodontine group (short-horned) :
Brontopinae : Mesaticephalic (in females) to brachy-
cephalic. Incisor teeth persistent.
Menodontinae : Mesaticephalic, stenocephalic. In-
cisor teeth reduced or wanting.
Brontotheriine group (long-horned) :
Megaceropinae : Brachycephaho. Incisor teeth re-
duced or wanting.
Brontotheriinae : Mesaticephalic to brachycephalic.
Incisor teeth persistent.
Each of these subfamilies includes one or more
phyla, which in turn may embrace one or more genera.
PolypTiyly. — The fact that the Ohgocene titano-
theres separated into four subfamilies before the be-
ginning of the Ohgocene epoch has been shown both
by the distinct connection of two of these subfamilies
with different Eocene ancestors and by the marked
differences between titanothere remains that are
found at the very lowest levels of the Titanotherium
zone. Here the ancestors of the Menodontinae, of
the Brontopinae, and of the Brontotheriinae are found
to be quite separate and distinct. No Megacero-
pinae have yet been found at these low geologic levels
of the Oligocene.
Summary of distinctions. — The means of distin-
guishing the numerous branches and sub-branches of
the great family differ somewhat from those em-
ployed to distinguish the Eocene branches from one
another. The proportions of the head and of the
zygomatic arch, whether dolichocephalic or brachy-
cephalic, still remain a distinguishing characteristic.
Owing to the buccal expansion of the zygomatic
arches the males of Menodus are technically mesati-
cephaUc or even sub-br achy cephalic rather than
"dolichocephalic," in spite of the fact that the skuU
in general appearance and in form of the teeth is long
and narrow, especially in contrast with the wide skull
and grinding teeth of Brontops, Megacerops, and
Brontotherium. The narrower Oligocene skulls may
therefore be described as stenocephalic.
As the horns become the dominant feature of the
skull the main line of division first arises between the
short-horned and long-horned titanotheres. The typi-
cal shape of the horns, whether triangular, rounded,
oval, or flattened, also becomes of very great value.
With these weapons of offense are developed the
gigantic swellings of the zygomatic arches by which
the breadth of the skull as a whole is measured and
cephalic indices are determined. A further differen-
tiation is found in the presence or absence of incisor
teeth and in the shape, size, and offensive character
of the canine tusks. Wide divergence is seen also in
the process begun in upper Eocene time — namely, the
molarization of the premolars, or their transforma-
tion into the molar pattern. In some lines of descent
this process is accelerated, and in others it is retarded.
Another distinction is in the development of the
cingulum on the grinders.
FiGtTBB 395. — Lower jaws of the Brontotherium phylum
A, Brontotherium leidyi, Carnegie Mus. 93 (paratype); B, B. leidyi. Am. Mus.516;
C, B. hatcheri. Am. Mus. 1070; D, B. gigas, Yale Mus. 12009 (type); E, B. gigas
elatum, Yale Mus. 12061 (type of Titanops elatus); F, B. medium. Am. Mus. 1051.
All one-twelfth natural size.
466
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 396. — Lower jaws of the Brontops and Menodus phyla
A, Brontops brachycephalus, Am. Mus. 1495; B, B.dispar, Nat. Mus. 4941 (type); C, B. robustus, Prince-
ton Mus. 10061; D, B. robustus, Yale Mus. 12048 (type); E, Menodus torvus. Am. Mus. 6365 (type);
F, M. trigonoceras , Nat. Mus. 4745; Q, M. giganteus, Am. Mus. 506; H, M. giganteus. Field Mus.
P 5927. All one-twelfth natural size.
Figure 397. — Lower jaws of the Diplo-
clonus and Allops phyla
A, Allops walcotti (?), Nat. Mus. 4247; B, A. marshi.
Field Mus. P 6900; C, A. angustigenis, Ottawa Mus.
(cotype); D, Diplodonus bicornutus. Am. Mus. 1476
(type); E, D. tyleri, Amherst Mus. 327 (type). All
one-twelfth natural size.
Figure 398. — Heads of Oligocene titanotheres, showing proportions, lip structure, and horns
Modeled by Charles R. Knight. A, Brontops (bracbycephalic); B, Menodus (dolichocephalic); C, Megacerops (hyperbrachycephalio); D, Broniotherium (brachycephalic).
EVOLUTION OF THE SKULL AND DENTITION OP OLIGOCENE TITANOTHERES
Groups and subfamilies of Oligocene titanotJieres
467
[See fig. 697, p.
Menodontine group: Persistently short-homed titanotheres;
teeth with persistent cingula
i broad; grinding Brontotheriine group: Originally short-homed, progressively long-homed titano-
theres; nasals pointed, reduced; grinding teeth with cingula reduced
SUBFAMILY BBONTOPINAE
Descendants of the Eocene Manteoceras and Protitanotherium.
Subbrachy cephalic to hyperbrachy cephalic.
Horns suboval, subcircular, transversely oval.
One to two pairs of persistent incisor teeth.
Premolar transformation retarded.
Genera: Teleodus, Brontops, Diploclonus.
SUBFAMILY MENODONTINAE
Descendants of unknown Eocene ancestors, possibly related to
Tehnatherium.
Mesaticephalic to dolichocephalic and to brachycephalic.
Horns typically subtriangular.
Incisor teeth vestigial or reduced to one pair.
Premolar transformation not retarded.
Genera: Allops, Menodus.
SUBFAMILY MEGACEROPINAE
Descendants of unknown Eocene forms, possibly of Rhadinorhi-
nus.
Brachycephalic to hyperbrachycephalic.
Horns rounded and without connecting crest.
Incisor teeth reduced or vestigial.
Premolar transformation accelerated.
Genus: Megacerops.
SUBFAMILY BEONTOTHBRIINAE
Descendants of unknown Eocene forms.
Mesaticephalic to brachycephalic.
Horns transversely oval and progressively flattened.
Incisor teeth persistent.
Premolar transformation accelerated.
Genus: Brontotherium.
The members of these subfamilies are more or less
clearly connected with ancestral forms in the Eocene,
as shown in Figures 402-409. Further exploration,
however, is needed to bridge over securely these
lines of descent. Our present knowledge is about as
follows :
Genetic relations of Oligocene to Eocene titanotheres
Eocene titanotheres
OUgoceue titanotheres
Subfamily Palaeosyopinae. Becoming entirely extinct in
middle Eocene time.
Subfamily TelmatheriLnae. Exhibiting some resemblances
to the Oligocene Menodontinae but differing widely in the
presence of large incisor teeth.
Subfamily Manteoceratinae. Including two stages, Manteo-
ceras and Protitanotherium, which exhibit _ many resem-
blances to the Brontopinae but differ in the possession of very
broad nasal bones.
Subfamily Dolichorhininae. Composed chiefly of generic forms
that became extinct in the upper Eocene but early gave off one
branch.
Subfaraily Rhadinorhininae, which shows some resemblance to
Megacerops.
Subfamily Diplacodontinae. Including the genera Diplacodon
and Eotitanotherium, of uncertain relationships with the
lower Oligocene Menodontinae and Brontotheriinae.
No known descendants of the Palaeosyopinae.
Subfamily Menodontinae. Exhibiting some resemblances to the
Eocene Telmatheriinae; no known connecting forms.
Subfamily Brontopinae. Including Teleodus, Brontops, and
Diploclonus, which show many resemblances to the Eocene
Manteoceratinae.
Subfamily Megaceropinae. Including Megacerops, the "Sym-
borodon"of Cope, which shows certain resemblances to Rhadi-
norhinus and other very strong resemblances to Bronto-
therium.
Subfamily Brontotheriinae. Showing resemblances to the
Eocene Diplacodon, Eotitanotherium, Rhadinorhinus. Direct
Eocene ancestors unknown.
Brontopinae. — The more heavily built titanotheres
of the genus Brontops exhibit close resemblances to
the Eocene Manteoceras and Protitanotherium. They
pass through the lower Oligocene species of Teleodus
and exhibit a wide adaptive radiation into the genera
Brontops and Diploclonus. In general they have
short, robust limbs, very broad skulls, and short,
rounded or pointed horns and are provided with one
or two pairs of cropping teeth. The incisors have
rounded crowns; the canines are roundly pointed;
the premolars are very slow in developing the second
internal cones. Short limbs and broad, spreading
feet give them a singularly graviportal and brachy-
podal character, whtch is analogous to what is known
of the limb structure of Manteoceras, of the Eocene.
Menodontinae. — The long-limbed relatively cursorial
animals typified by the genus Menodus agree with the
Telmatheriinae in the very robust development of the
468
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
canines and of the grinding teeth but differ in the
marked degeneration of the incisor teeth. In the
loss of incisors they parallel the Megaceropinae.
In contrast to the Brontops series the limbs and
arches are slender, the bones of the girdles.are relatively
high and narrow, and the feet are of the elongate,
dolichopodal type. Thus the menodonts are relatively
cursorial in contrast with the typically graviportal
brontopines.
Brontotheriinae. — The gigantic titanotheres of the
subfamily Brontotheriinae appear to be related
either to the Diplacodon or to the Ehadinorhinus
with the elongation of the horns. The cheek teeth
are less elongate than in Menodus and have more
rounded crowns and nearly obsolete cingula, well
fitted for cutting and crushing coarse vegetation.
Megaceropinae. — The Megaceropinae are much
smaller animals than the brontotheres, distinguished
by skulls of intermediate proportions, with tall, rather
slender and cylindrical horns placed well forward
above the eyes. The incisors are absent. It appears
probable that the snout was narrow and terminated in
a pointed Up like that of the black rhinoceros of Africa.
The face was very short, narrow, and upturned, bring-
MENODUS
MEGACEROPS
BRONTOTHERIUM
bnuJiycefhalus
'brndiyc^Jtalus
MtngoruKems
FiGUKE 399. — Sections at base of horn in five principal lower Oligocene phyla of titanotheres, arrang
according to ascending geologic levels
Drawn to the same scale.
phylum of the upper Eocene.^^ They are distinguished
by the precocious development of the horns, the rapid
transformation of the premolar teeth, the stout,
obtuse canines, the retention of two pairs of upper
and lower incisor teeth that have cingulate crowns in
contrast to the smooth, rounded crowns in the Bron-
topinae. The horns early acquire at their extremities
a transversely oval shape, which finally extends down
to the base of the horn. The cranial vertex is ex-
tremely long and narrow, but the great buccal proc-
esses at the sides of the head develop pari passu
" See pp. 434, 441, 469, 560.
ing the mouth almost up to the level of the eye. The
small obtuse canines were brought close together
toward the median line. The cusps and cutting edges
of the cheek teeth were even more rounded than in
Brontotherium and entirely devoid of cingulum.
POSSIBLE EOCENE ANCESTORS OF THE BEONTOTHEEIINE
GROUP
We may again consider the evidence and theories as
to the transitions between Eocene and Oligocene
titanotheres. (Compare Diplacodon, Rhadinorhinus,
pp. 439, 441, 470-474.)
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
469
Diplacodon theory. — The imperfectly known skull
of Diplacodon elatus from the base of Uinta C exhibits
several resemblances to B. leidyi from Chadron A,
and Osborn was at first disposed (1915) to regard
Diplacodon as containing the ancestors of Brontothe-
rium. (See pp. 439, 441.)
Gregory at first regarded Diplacodon as intermediate
between the RJiadinorJiinus- Metarhinus group on the
one hand and Brontotherium on the other, but later he
was inclined to regard Diplacodon as possibly related
rather to the ancestors of the menodontine group, on
account of the resemblances in the premolar-molar
series between Diplacodon and the early members of
the Brontops and Menodus phyla. (See fig. 406.)
SECTION 4. OLIGOCENE GENERA ACCEPTED AS
VALID IN THIS MONOGRAPH
Generic name used
Valid generic name and type species by Osborn in 1902
Menodus Pomel (M. giganteus), 1849 Titanotherium.
Megacerops Leidy (M. coloradensis), 1870- Megacerops.
Brontotherium Marsh (B. gigas, jaw), 1873- Brontotherium.
Brontops Marsh (B. robustus), 1887 Megacerops.
Allops Marsh (A. serotinus), 1887 AIlops.
Diploclonus Marsh (D. amplus), 1890 Megacerops.
Teleodus Marsh (T. avus), 1890 Megacerops.
Menodus. — As is fully explained on pages 204-205,
Pomel's name Menodus giganteus (1849) was based
upon the first specimen of a titanothere made known
to the scientific world, a fragment of a lower jaw,
originally described by Prout in 1847. A comparison
Figure 400. — Restorations of lower Oligocene titanotheres of the four principal genera
A, Brontops rohustus; B, Menodus giganteus; C, Megacerops acen D, Brontotherium hatckeri. One-flftieth natural size.
RJiadinorJiinus tJieory. — In 1902 Gregory observed
that MetarJiinus fluviatilis and RJiadinorJiinus diplo-
conus of the middle Eocene foreshadow Megacerops and
BrontotJierium in the following characters: (1) Ante-
rior nares very deep, nasals becoming shorter; (2)
bridge over infraorbital foramen very rounded; (3)
upward flexure of premolar series in side view and
"curvilinear" effect in palate view; (4) canines small
and sometimes swollen at base, lower canines set near
each other; (5) premolars relatively advanced, tetar-
tocones set well in toward center of crown; (6) lower
jaw with spoutlike incisive region (cf. R. diploconus,
M. earlei, M. fluviatilis).
of a carefully drawn figure of this specimen given by
Leidy (1854.1, pi. 16, fig. 1) with more complete
material now available indicates that Menodus gigan-
teus is the same form as that which was later named
by Marsh BrontotJierium ingens. Accordingly, Meno-
dus giganteus Pomel has priority over Marsh's name
and has therefore been adopted in this work. Titano-
tJierium Leidy, although formerly used by the
present author instead of Menodus Pomel, is now
regarded as a synonym of that name, for reasons
given on pages 205-206.
Megacerops. — -Leidy's Megacerops coloradensis was
founded upon a fragment of a skull (including the
470
TITANOTHERES Or ANCIENT WYOMING, DAKOTA, AND NEBRASKA
coossified nasals and "horns" ; see p. 208) of a titanothere
that now appears to be congeneric with the skulls
subsequently named by Cope Symiorodon iucco. The
present author formerly used the name Megacerops for
the generic group called Brontops by Marsh, but a re-
examination of Leidy's above-mentioned type of
Megacerops coloradensis has convinced him that this
was an error. In the present monograph the name
Megacerops includes the forms called by Cope "Sym-
horodon bucco" and "S. altirostris." (See pp. 212, 215.)
Brontotherium. — Marsh's Brontoiherium gigas, the
genotype of Brontotherium, rested upon a certain lower
Brontops, Allops, Diploclonus, and Teleodus are all
now treated as distinct genera, although they were
formerly regarded by the author as referable to
Megacerops.
SECTION 5. THE MENODONTINE GROUP
SUBFAMILY BEONTOPINAE, INCLUDING THE PHYLA MAN-
TEOCERAS, PEOTITANOTHERIUM, TELEODUS, BRONTOPS,
AND DIPLOCLONUS
STRATIGRAPHIC LEVEL AND DISTINGUISHING FEATURES
The menodontine group consists of titanotheres of
upper Eocene to lower Oligocene age that reached a
Figure 401. — Skulls of Rhadinorhinus and Brontotherium
Palatal view. A, Rhadinorhinus iiploconus, Am. Mus. 1863 (type); White River, Uinta Basin, Utah, Uinta B 1; two-ninths natural
size. B, Brontotherium leidyi, Carnegie Mus. 93 (paratype); Chadron formation; one-sixth nattttal size.
jaw in the Yale Museum (see p. 210), which the present
author regards as congeneric with the flat-horned skulls
that were later named by Marsh Titanops elatus,
Titanops medius, Titanops curtus. Marsh, however,
erroneously referred his type skull of "Brontotherium"
ingens to the genus Brontotherium. "B. ingens"
proves rather to be a synonym of Pomel's Menodus
giganteus.
Symhorodon Cope is unfortunately a synonym of
Menodus Pomel, for the reason that the type species
Symiorodon torvus (see p. 211) was founded upon
lower jaws that appear to be congeneric with the type
jaw of Menodus giganteus Pomel.
climax in the upper levels of the upper Titanotherium
zone. Related to the Eocene Manteoceras. Dis-
tinguished by progressively broad heads (brachy-
cephaly), short-crowned teeth (brachyodonty), and
short or moderately proportioned feet (mesatipody) .
Horns short, progressively shifting forward, of
primitive trihedral section at the base, rounded to
oval at the summits, progressively transverse oval.
Nasals progressively reduced in length and broadening
at the extremities. Incisor teeth with rounded
crowns; one or two pairs persistent above and below.
Canine teeth pointed, of medium length, progressively
obtuse. Premolar evolution retarded. Zygomata,
EVOLUTION OP THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
47L
especially in males, progressively expanding into con-
vex buccal processes. Includes origiual contempora-
neous phyla known as Teleodus, Brontops, and
Diploclonus.
These short-horned, ; broad-headed titanotheres,
which have persistent incisor teeth, are very abundant
in the lower, middle, and upper levels of the upper
TitanotJierium zone. They present certain character-
istics which seem to be shared most nearly by the
animal that Marsh called Brontops dispar. This
animal is represented by a large nmnber of specimens,
which grade below into other specifically distinct forms,
such as Brontops irachycepTialus, and above into
larger forms, such as Brontops rohustus. It appears
wise to retain as generic or subgeneric names the names
The horns attain no very great length and are
usually circular in section at the summit. At the
base they are trihedral, rounded, or transversely oval
in section. The nasals, unlike those of Menodus and
Protitanotherium, are rounded anteriorly, progressively
shortened in ascending mutations, and in most old
individuals they expand at the extremities.
At the very base of the Titanotherium zone Hatcher
found a number of small, broad-skulled titanotheres
that certainly belong to the Brontops phylum and
that have been termed Brontops hrachycepTialus by
Osborn, because they are surprisingly broad-skulled.
Although they are perhaps not directly descended from
any known upper Eocene form, such as Protitano-
therium emarginatum, they have one striking feature
FiGUKE 402. — Skulls of Bhadinorhinus and Brontotherium
Side view. A, RhadiTwrhinus diploconus, Am. Mus. 1863 (type); White River, Uinta Basin, Utali; Uinta B 1;
two-ninths natural size. B, Brontotherium leidyi, Carnegie Mus. 93 (paratype); Chadron formation; one-
sixth natural size.
proposed by Marsh to distinguish members of the two
chief' phyla — namely, Brontops rohustus and Diploclo-
nus arnplus.
SUBFAIMILY CHARACTERS OF TELEODUS, BRONTOPS, AND
DIPLOCLONUS
COMPARISONS AND CONTRASTS
Very precise observation of the animals ranged
under the phyla Teleodus, Brontops, and Diploclonus
shows that they possess a large number of characters
in common which distinguish them more or less
clearly from the members of the far more readily
defined phyla Menodus, Brontotherium, and Mega-
cerops.
101959— 29— VOL 1 33
in common — the short or rudimentary horns are
placed directly above or slightly in front of the orbits
and exhibit an elongate oval section at the base, pre-
cisely like those of P. emarginatum. Such horns were
adapted to a lateral butting motion of the head, and
it is notable that they are always blunt, or have
elongate oval tips.
On the same low geologic level was also found a
lower jaw containing three incisor teeth (an Eocene
character), described by Marsh as Teleodus avus,
which may belong to the same group as B. irachy-
cephalus.
Features of the horns. — Horns thWt are transversely
oval at the tips belong to members of the Bronto-
472
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
tJierium phylum. The free portion of the nasal bones
is correspondingly elongate; the nasals contract in
width anteriorly. The canine tusks, especially in
males, are elongate, pointed, and slightly recurved.
It appears that all lower Oligocene titanotheres also
have elongate nasals. The extremely primitive char-
acter of the nasals and of the horns in the Teleodus-
Brontops-Diploclonus group is correlated with a re-
tarded stage in the evolution of the premolar teeth, a
very characteristic feature which sharply distinguishes
members of this group from members of the Megace-
rops and Brontotherium phyla. The premolars (fig.
406) are even more retarded or simpler than those of
iherium zone we note that the skulls referred to B.
hrachycephalus progress in size and in general evolution.
The horns shift forward somewhat on the face and
become elongate; the base of the horn becomes longer
in transverse diameter rather than in anteropos-
terior diameter. The free portion of the nasals
becomes shorter and spreads out distally. The pre-
molar grinding teeth gradually become somewhat
more complex in these stages of mutation, of change
of proportion, and of rectigradation, which are pro-
phetic of the next higher phase of evolution. It must
be remembered that many of these skulls are crushed
and distorted and represent differences in age and sex
Figure 403. — Skulls of Rhadinorhinus and Brontotherium
Top view. A, RJtadinorUnua abioiti. Field Mus. 12179 (type); Uinta B 1; two-nintlis natural size. B, Brontotherium
leidyi, Nat. Mus. 4249 (type); Chadron formation; one-sixth natural size.
Diplacodon elatus from the upper Eocene Uinta forma-
tion, a proof that D. elatus was certainly not the an-
cestor of Brontops. On the inner side of the superior
premolar crowns we see a large anterior cusp (deutero-
cone) followed by a low posterior ridge or small
rudimentary posterior cusp (tetartocone). The muta-
tions of B. IrachycepTialus in the lower beds were
therefore very characteristic and clearly separable
both from the upper Eocene forms and from succeed-
ing Oligocene forms.
Ascending mutations. — As we pass upward into the
middle A and upper A levels of the lower Titano-
as well as a number of progressive stages of evolution.
Among the animals specifically classed as B. hracTiy-
cephalus it is quite possible that ancestors of more
than one subsequent phylum may be foimd, such as
that leading to Diploclonus.
Prophetic characters among the aged individuals of
B. hracTiycepTialus point toward the much more robust
titanotheres of the middle beds to which Marsh gave
the names Brontops dispar and Brontops validus, the
former name having the priority. As the following
tables show, the Hatcher collection in the National
Museum is so rich in forms that it includes a series
EVOLUTION OP THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
473
of transition stages, almost unique in mammalian
paleontology, which connect the diminutive B. Ira-
chycepTialus from level A, the base of the Titanotherium
zone, with the large animals of level B known as B.
dispar, a form distinguished not only by its larger
size throughout but by rounded horns, shorter nasal
bones, more robust buccal processes of the zygomatic
arches, and somewhat more complicated premolar
grinding teeth. It appears that B. dispar is highly
characteristic and distinctive of the B levels, or middle
beds, but that a form of B. dispar passes up into the
base of the upper Titanotherium zone.
Diploclomis phylum. — The existence of a
second phylum is indicated by specimens that
are probably from the upper parts of the lower
and from the middle Titanotherium zone.
The stratigraphic records in regard to them
are incomplete. This is the species Diploclonus
bicornutus (Osborn), which is remotely related
to B. dispar and is distinguished, as the name
indicates, by a duplication of the horns on the
inner sides but especially by its very narrow,
elongate nasals and the straight outer contour
of the horns, as seen from the front. Nasals
of somewhat similar type are observed in the
animal from Assiniboia, Canada, which was
named Menodus selwynianus by Cope.
The most interesting biologic conclusion to
be drawn from this assemblage of varied types
is that in the lower and middle Titanotherium
beds there were many kinds of small and'J'mid-
dle-sized titanotheres more or less closely
related to Brontops. The group is certainly
diphyletic, possibly polyphyletic, and is there-
fore all the more difficult of analysis.
The type Brontops in the upper Titanotherium
zone. — On the lower levels of Chadron C, the
upper Titanotherium zone, are found the great
animals to which Marsh gave the name Sronfops
robustus, as well as the surviving members of
the Brontops dispar series. At first this animal appears
to be widely separated from Brontops brachycephalus
and B. dispar, but like B. dispar, which Marsh also
placed in the genus Brontops, B. robustus is represented
by a large number of skulls in different stages of growth,
and there can be no question that certain character-
istics of age, growth, and sex of B. robustus are close to
some of the variations in B. dispar, but no direct
ancestor to the type stages is known. The more
important ascending mutations will be expounded in
the more precise study of the phylum which follows.
Special distinctions from Menodus. — As already
stated, these animals belong to the same group as
the Menodontinae and are more closely related to
Menodus than they are to either Megacerops or
Brontotherium. It is therefore desirable to summarize
their distinctive characters: (1) The males show
brachycephaly, having a zygomatic index that ranges
in general from 79 to 91 (mesaticephaly) ; the females
have a zygomatic index that ranges from 64 to 69;
(2) all retain one or two pairs of upper and lower
incisors, which are full sized and round crowned; (3)
the canines in males are progressively shortened in
the ascending series, whereas in Menodus they are
Figure 404.-
-Lower jaws of Melarhinus fluviatilis and Brontotherium
hatcheri
, M. fluviatilis, Am. Mus. 2059; White River, Uinta Basin, Utah; Uinta B 1; two-ninths
natural size. B, B. hatcheri, Am. Mus. 1070; Hat Creek, Nebr.; Chadron formation; one-sixth
natural size.
persistently long and pointed; (4) the tetartocones
of the premolars are retarded in evolution, whereas in
Menodus they evolve rapidly; (5) as the horns shift
forward the nasals are progressively reduced in
length — they are not quadrate but broadly expanded
distally at their extremities; (6) the horns are typi-
cally suboval or cylindrical in basal section and have
rounded rather than trihedral tops as in Menodus;
(7) the face progressively shortens, whereas in Meno-
dus it remains persistently long; (8) in correlation
with the shortening of the face the tooth row becomes
bent upward anteriorly; (9) the zygomata expand
progressively.
474
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
A phylum somewhat intermediate between Brontops
and Menodus is Allops, from which members of the
Brontops phylum are distinguished by their retarded
tetartocones and by the fact that the horns in the
males are upright or slightly divergent, in contrast
with the depressed and widely divergent horns of
Allops.
Retarded evolution of the Brontops premolars. — The
retarded molarization of the premolars is a conspicu-
FiGXJEE 405. — Progressive evolution of the up-
per premolars in Brontotherium and its prede-
cessors
A, Eotitanops borealis, Am. Mus. 14887, Wind Eiver; B, Palaeo-
syops paludosus, Am. Mus. 13032, lower Bridger (althougli
this genus is not in the Brontotherium series, it illustrates a
stage of premolar evolution); C, JRhadinorhinus diploconus,
Am. Mus, 1863, Uinta B; D, Brontotherium leidyi, Nat.
Mus. 4249, Chadron A; E, Brontotherium gigas, Am. Mus.
492, Chadron C. All one-half natural size.
ous character of this phylum. The following facts
should be noted:
1. The exact stage of evolution of the tetartocones
is partly obscured by the degree of wear, so that much
worn teeth appear simpler in structure than unworn
teeth, and the greatest degree of complication appears
in the intermediate stages of wear.
2. All the specimens referred to the Brontops and
Diploclonus phyla show essentially similar premolar
characters; in the primitive forms the tetartocone of
p* is a concave spur from the deuterocone, continu-
ous postero-internally with the internal cingulum;
the spur becomes more convex on the buccal side,
less concave on the lingual side, and gradually loses
its connection with the cingulum; the point of con-
striction between the deuterocone and tetartocone
moves forward so that the tetartocone enlarges at the
expense of the deuterocone. In all species of Brontops
Figure 406. — Progressive evolution of the upper pre-
molars in Menodus and Brontops and their predecessors
A, Eotitanops borealis. Am. Mus. 14887 (neotype). Wind Eiver; B,
Manteoccras manteoceras. Am. Mus. 12683, upper Bridger (Telmatherium
cultridens illustrates this stage even better); C, Biplacodon elatus, YalQ
Mus. 11180, Uinta C; D, Brontops irachycephalus, Nat. Mus. 4258 (type),
Chadron A; E, Menodus giganteus, Am. Mus. 505, Chadron C. All one-
half natural size.
the tetartocone never appears entirely distinct from
the deuterocone, as it does in Brontotherium.
3. Within the species B. dispar there is considerable
range of evolution in the progressive development of
the tetartocone, possibly due to the crania having
been found on different levels.
4. An interesting fact is that occasionally there are
noticeable differences in the tetartocones on opposite
sides of the same individual.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
475
5. Comparison reveals the general kinship oi Br ontops,
Allops, and Menodus in the evolution of the premolars.
The most important biologic fact disclosed is that
the tetartocones show a somewhat different rate of
v-yv^'^/-..V.^ J
FiGUKB 407. — Progressive evolution of the lower pre-
molars in Brontolherium and its predecessors
A, EotUanops princeps, Am. Mus. 290 (type), Wind River; B, Palaeosyops
paludosus, Am. Mus. 11680, lower Bridger (although this genus is not
In the direct line, its premolars well illustrate this stage of evolution) ;
C, Metarhinus earlei, Am. Mus. 13179, Washakie B; D, Broniotherium
medium, Am. Mus. 1051, Chadron C. All one-half natural size.
evolution, or retardation and acceleration, in each
of these subphyla. A second fact of importance is
that Brontops in its later stages is convergent in many
characters with BrontotJierium.
Ascending mutations between species. — That species
referred to Brontops and Diploclonus are allied
generically is shown by numerous skulls that com-
bine features of two species, either in the same
Figure 408. — Progressive evolution of the
lower premolars in Brontops and its pred-
ecessors
A, EotUanops princeps, Am. Mus. 296 (type), Wind Riv-
er; B, Manteoceras manteoceras. Am. Mus. 1556, upper
Bridger; C, Protitanotherium emarginatum Princeton
Mus. 11242, Uinta C; D, Brontops brachycephalus. Am.
Mus. 1495, Chadron A. All one-half natural size.
or in different phyla, such as B. hracTiycephalus
and B. dispar (Nat. Mus. 4258, 1214) or D. bicor-
nutuS) D. amplus, and type of D. tyleri, Amherst
Museum.
ITie Brontops-Diploclonus phylum as represented in the Hatcher collection oj 60 sJculls and jaws from the Chadron
formation, in the United States National Museum
Catalog
No.
Sex
Genus and species
Specimen
Notes
4710
Female
Male
Male-
SkuU.- --
8733
Skull...
Extremely large male, exceeding in size the
type of B. robustus.
Very typical of the species.
8732
do -
Skull, anterior half
and jaws.
Skull- --
8766
Male-
do
8767
(?)
do
Skull . -
4943
Male -.
do -
Skull (typical)
Skull-.
8315
Male
do -
Showing transition from B. dispar (type)
to horns and canines.
8747
(?)-
do
Skull
Skull
Right ramus
Skull and jaws
Skull and jaws
Skull
Skull.--
Skull...
4696
1242
Male
Male
--.do--- ...-
:
do ..
series.
Transitional B. dispar to B. robustus.
Agrees in "tooth measurements with
type of B. robustus.
Jaw of old animal.
4941
Male
Male
Male
Male.
Type specimen.
1217
4245
4248
Brontops (serotinus) dispar Marshl :-.
Brontops dispar Marsh
do
Fine skull and jaws attached.
4253
(?)
do - -. -
Hyperbrachycephalic; old individual.
Fine skull.
4703
Male-
do
Skull. . -
4706
Male..
do - -.
Skull..
Do.
4738
Female
do
SkuU
476
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The Brontops-Diploclonus phylum as represented in the Hatcher collection of 60 sTculls and jaws from the Chadron
formation, in the United States National Museum — Continued
8316
7749
7642
4944
4942
4939
8742
8746
8725
8748
4290
8792
8769
8796
4948
1241
8319
8779
8784
8785
8764
4258 ]
8776
1233
1235
4259
4274
4261
8738
8754
8739
4940
4947
8750
1214
1237
4246
8320
8782
1862
Female
(?)
(?)
Female
(?)
Male
Female
Male
Female
Male-
Male-
(?)--
(?) —
Female-
Male,-.
Female -
Female-
(?)
Female -
(?)
Male--.
Female -
Male-
(?)-..
Male-
Female.
Male.
Male.
Male-
Male-
Male-
Male-
Female.
Genus and species
Brontops dispar Marsh
do
-do.
-do.
-do.
.do.
.do-
.do.
.do.
.do-
Brontops? dispar Marsh.
Brontops?
Female
(?)
Male
(?) Female
Male
(?)
(?)
(?)-
Brontops (brachycephalus) dispar Marsh.
Brontops dispar Marsh.
do
.do-
.do-
.do.
-do.
.do.
.do-
Brontops brachycephalus Osborn
Brontops (?dispar) brachycephalus Os-
born.
Brontops (brachycephalus) ?dispar
Marsh.
Brontops (dispar) ?brachycephalus Os-
born.
Brontops brachycephalus Osborn
do
Skull
Skull
Broken skuU
SkuU
Skull
SkuU
SkuU
SkuU
SkuU (part of skele-
ton).
SkuU, anterior half,
and jaws attached.
Skull
Pair of lower jaws
Posterior part of
skuU.
Right ramus and
symphysis.
Lower jaws
SkuU E, right jaw in
matrix.
Lower jaws. Young
Pair of lower jaws..
Pair of lower jaws..
Right ramus, ?AUops
marshi.
Pair of lower jaws..
SkuU
SkuU
Pair of lower jaws-
Large female (?) of progressive size.
Canines lacking; measurements agree well.
Equals Allops marshi?
Medium size, adult.
Measurements typical of female (rare) .
Vigorous male. Teeth poorly preserved.
Badly preserved and crushed.
Vigorous male; young, m' just coming in;
medium to small size.
.do-
.do-
-do.
.do-
.do-
Lower jaw
Pair of lower jaws-
SkuU.
SkuU.
SkuU
SkuU
Upper dentition.
Skull, lower jaws,
and femur.
SkuU
.do-
.do-
.do.
.do.
Brontops? brachycephalus Osborn .
do
Brontops brachycephalus Osborn .
SkuU
Upper dentition
SkuU
Lower jaw
Lower jaw
Lower jaws
Anterior half of lower
jaw.
Imperfect skuU
Very fine female. Transitional in meas-
urements from B. brachycepha us to B.
dispar.
Advanced stage, approaching B. dispar.
Advanced stage, approaching B. dispar.
Measurements agree p ecisely. Advanced
stage, approaching B. dispar.
Type specimen.
Measurements typical.
Superior dentition of both sides; advanced
in size.
Vigorous male skull; measurements typi-
cal.
Canine alveoli small; molars inferior to
type of B. brachycephalus.
Paratype.
Dentition of smaUer dimensions.
Juvenile, imperfect.
Measurements agree closely with B.
brachycephalus.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
477
Out of a multiplicity of specimens we are able to
select evidences of two or three directly successive
phyla, as follows:
I. Teleodus avus, an ancestral form to Broniops, from the lowest
geologic levels.
Teleodus primitivus, of doubtful relationship, preserving
three lower incisors.
II. Brontops phylum, occiput progressively prolonged back-
ward behind zygomata; rounded horns vertically placed:
C. Upper beds: B. dispar Marsh.
B. Middle beds: B. dispar Marsh.
A. Lower beds: B. brachycephalus (Osborn).
Ila. Brontops phylum, less directly successive; occiput less
extended backward, horns laterally overhanging maxiUaries,
obliquely to transversely oval in basal section:
C. Upper beds: B. rohustus Marsh.
III. Collateral phylum, less directly successive; occiput ex-
tended very far back, nasals narrow, progressively reduced, horns
with internal hornlets. Levels largely undetermined:
C. Upper beds: Diploclonus amplus Marsh.
B. Middle beds: D. tyleri (LuU).
B. Middle beds: D. bicornutus (Osborn).
§.3
UPPER
EOCENE
M. giganteus
ALLOPS MENODUS
' /
B. brachycephalus ' A. walcotti M. heloceras
\ t I ^
Teleodus avus ^^^ "^ *^
Protitanotherium
nasals progressively broadening and abbreviating;
horns moderately elongate, rounded, or transversely
oval; facial region progressively abbreviated.
Teleodus, a primitive stage.
Brontops, extreme brachycephaly, single-horned.
Diploclonus, brachycephaly to extreme brachycephaly,
duplicate-horned.
Figure 410. — Evolution of the horns in the Brontops phylum
a, Brontops tmchycephalus, Nat. Mus. 4261 (type), Chadron A 1; b, B. brachy-
cephalus, Nat. Mus. 1214, Chadron A 3; c, B. dispar, Nat. Mus. 4703, Chadron B 2;
d, B. dispar, Nat. Mus. 4941 (type), Chadron B 2; e, B. robustus, Nat. Mus.
4696, Chadron C 2. All one-fourth natural size. These outlines show progres-
sive increase in height and thickness of the horns; their gradual displacement
forward, in front of the orbits; progressive thickening of the nasals and the
confluence of their posterior upper border with the anterior border of the horns.
Figure 409. — Phyla of the Brontopinae and Menodontinae,
titanotheres of the short-horned group
CONSPECTUS OF CHARACTERS OF THE SUBFAMTLT BRONTOPINAE
Summary of general characters. — Titanotheres ex-
tending through lower Oligocene time; incisor teeth
2-1, the crowns rounded; canines of medium size,
pointed; premolars and molars cingulate; premolar
transformation retarded; skuU progressively brachy-
cephalic to hyperbrachycephalic, indices 71-91;
FiQUHE 41L — Basal section of the horns in the Broniops
phylum
A, Broniops brachycephalus, Nat. Mus. 4261 (type), Chadron A 1; B, B. brachy-
cephalus, Nat. Mus. 1214, Chadron A 3; C, B. brachycephalus, Nat. Mus. 4259,
Chadron B 1; D, B. dispar, Nat. Mus. 4703, Chadron B 1; £, B. dispar, Nat.
Mus. 4696, Chadron C 2; F, B. robustus, Yale Mus. 12048 (type), Chadron C 3.
All one-fourth natural size. These outlines show a progressive change in the basal
section of the horns from the obliquely oval section in B. brachycephalus through
the rounded trihedral section of B. dispar to the transversely oval section of B.
robustus.
Summary of special cTiaracters. — If^. Third upper
and second lower incisors the largest; second (or
median) upper incisor frequently shed in adult; circu-
478
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND- NEBRASKA
lar, smoothly rounded crowns ; incisors smaller
tlian in Brontoiherium. Canines cf usually of
medium size, anterior face 40, 42, 48 milli-
meters, recurved, pointed crowns, often com-
pressed anteroposteriorly; canines 9 more
slender, shorter, and in section rounder.
Opposite molar-premolar series rectihnear to
arched. Upward flexure of premolars in side
view progressive and especially pronounced
in hyperbrachycephalic skulls; dental (pre-
molar-molar, basilar) index cf 46-50.
Premolars, tetartocone of p* typically re-
tarded, often a concave spur; tetartocones
of p^, p^ more progressive, but less so than
in Menodus; external cingulum typically less
prominent than in Menodus. Molars with
internal cingula less prominent than in
Menodus, crowns less hypsodont than in
Menodus. Skull cf brachycephalic to hyper-
e'dc ha
Figure 412. — Progressive broadening of
the nasals in the Bronto-ps phylum from
B. brachycephalus, with narrow nasals,
through B. dispar to B. robustus, with
broad nasals
a, Brontops brachycephalus, Nat. Mus. 4947, Ghadron A
b, B. brachycephalus, Nat. Mus. 1214, Chadron, A; c,
B. brachycephalus, Nat. Mus. 1258, Chadron B; d
dispar, Nat. Mus. 4703, Chadron B; e, B. dispar, Nat,
Mus. 4696, Chadron C;/, B. robustus, Am. Mus. 1083,
Chadron C (?)
brachycephalic, zygomatic index 73-91, skull
9 with narrow buccal arches. Nasals primi-
tively elongate, progressively shortened and
swelling at the extremities; nasal index
61-140. Horns cf primitively low, elongate,
ovoid, becoming cylindrical {B. dispar), club-
shaped {Brontops robustus) or with a small
secondary branch {Diplodonus) , roundly tri-
hedral (B. dispar) or broadly oval (Brontops
rohustus). Face progressively abbreviate.
CONSPECTUS OF CHARACTERS OF SPECIES
The salient features of the several species
are set forth in the following summary:
Teleodus avus Marsh. The most ancient form re-
corded from the base of the lower beds. Readily dis-
tinguished by the three incisor teeth with rounded
crowns in the lower jaw, although the type is some-
what larger than B. brachycephalus. Phyletic position
somewhat doubtful.
Figure 413. — Lower jaws of Teleodus primitivus, Brontops brachycephalus,
and Allops ualcottif
A, Teleodus primitivus, Otta.wayias. (type); Cypress Hills, Saskatchewan; perhaps the smallest
and most prhnitive known jaw of an adult Oligocene titanothere, with horizontal ramus long and
moderately deep, chin gently convex, angle not projecting sharply backward, ascending ramus
relatively broad, coronoid of moderate height, three incisors, and a prominent postcanine
diastema. The external cingula of the cheek teeth are partly confluent with the ectolophs,
whereas in all other Menodontinae they are sharply defined.
B, Brontops brachycephalus. Am. Mus. 1495; compared with the preceding has a horizontal ramus
very shallow anteriorly, angle projecting downward and backward, coronoid high, and molars
relatively larger.
C, Allops walcottif, Nat. Mus. 4247; differs frona the typical Brontops in having sharply defined
external cingula and no diastema in front otpi (pi has dropped out). It differs from the known
Menodus in possessing incisors.
All one-fifth natui'al size.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
479
Teleodus primitivus (Lambe). Known from a jaw found in
the Cyp ress Hills, containing three lower incisor teeth, of which
i2 is the largest and ii the smallest.
Brontops brachycephalus (Osborn). Very abundant in the
lower beds, extending from the lower into the middle levels and
passing into transitional forms in the upper levels. Readily
distinguished by the broad, rounded skull, and very simple
premolar teeth.
Brontops dispar Marsh. Very abundant and characteristic of
the middle levels. Transitional in structure between B.
brachycephalus and B. robustus.
Brontops robustus Marsh. An enormous animal characteristic
of the lower levels of the upper beds. It presents in an extreme
degree the progressive characters of this genus — great width of
the skull combined with decided abbreviation of the horns as
well as of the nasals and surprisingly backward development of
the premolar teeth.
Diploclonus bicornutus (Osborn). Closely related to Brontops
dispar, from which it differs in its longer, narrower nasals and
internal hornlets on the horns. Geologic level unknown, prob-
ably the lower part of the middle beds.
Diploclonus lyleri (Lull). Intermediate between D. bicornutus
and D. amplus. Brachycephalic, with shortened nasals and
well-developed internal hornlets. Geologic level propably mid-
dle beds.
Diploclonus amplus Marsh. Extremely brachycephalic, with
short nasals; divergent horns with steep connecting crest.
Probably from the upper beds.
Conspectus of characters of jaw in the menodontine group
Character
Teleodus
Brontops
Diploclonus
Allops
Menodus
Horizontal ramus
Deep.
Shallow.
Shallow.
Intermediate.
Deep.
Symphyseal region in
Convex.
Very shallow.
Shallow.
Intermediate to con-
Fuller.
side view below mental
vex.
foramen.
duced.
downward.
truncate.
downward into a con-
vex elbow. Posterior
border oblique.
I3
IfT
I2T
IfT
1 0 (typically) .
Present or absent.
Pi in fully adult jaws
Present.
Usually present.
Present.
Present or absent.
Diastema in front of pi..
Present.
Present, wide.
Present.
Absent.
Absent.
Slender.
Slender to short;
Slender to
Conic to compressed
anteroposteriorly .
stout.
short; stout.
External cingulum of
Not sharply
Present in early
Intermediate.
Present.
Strongly marked.
canines, premolars,
defined.
types, lost in B.
molars.
robustus.
Upward flexure of pre-
Moderate.
Slight or moderate.
Slight or mod-
Very slight.
Absent.
molar series.
erate.
MEASUREMENTS OF THE BRONTOPS SERIES
The species of the genus Brontops, like the species of
Menodus, constitute a closely graded ascending series
beginning in the lowest and running up into the latest
beds of the Titanotherium zone. The range in the
chief measurements of the successive stages may
be summarized as follows:
Measurements of Brontops, in millimeters
Pi-m'
Pi-p*
Mi-m3
Pmx to condyles
Zygomatic index
Nasal length
Horn length
340-376
132-161
215-236
743-843
77-87
52-90
130-210
310-345
123-145
195-215
660-687
78-87
85-90
155-198
B. brachy-
cephalus, cf
265-297
101-123
160-180
580-610
72-?82
60-85
59-135
Hence the series of measurements exhibits small
gaps between B. brachycephalus and B. dispar in the
measurements p'-m', m'-m', in the basilar length
(pmx to condyles), and in the length of the horns; but
the two species are continuous or overlap in respect to
the length of the premolar series, in the zygomatic
index, and in the nasal length. On the other hand
B. dispar is continuous with or overlaps B. roiustus
in the measurements p'-m', p'-p*, m'-m^, as well as
in the zygomatic index, nasal length, and length of
horns; but there is a sharp gap between the two species
in the basilar length, B. roiustus being a far larger
type.
480
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Standard measurements in the Brontops phylum, in millimeters
Upper teeth
Skull
Jaw and teeth
a
1
S
1
o
1
p.
i
i
a
S3
a
a
i
1
o
1
•1
1
1
1
1
1
1
M
s
s
w
CM
a
1
o
i
1
1
'i
•i
a
B. robustus, Am. Mus. 1069, c?
376
355
363
376
350
363
345
340
147
144
145
149
137
151
139
132
236
220
225
232
220
223
215
220
237
215
203
207
210
207
203
205
195
203
185
195
200
179
180
813
743
755
"843
765
795
795
640
620
600
508
667
628
615
76
83
79
675
"640
87
52
90
153
126
133
123
130
180
47
47
37
34
B. robustus?, Carnegie Mus. 314
B. robustus, Yale Mus. 12048, c? (type).
B. robustus, Carnegie Mus. 315
42
36
87
79
77
635
76
149
207
140
174
210
385
136
250
40
— -
683
B. robustus, Princeton Mus. 11015, cf
48
41
26
385
137
240
43
B. robustus, Nat. Mus. 4696
"■85
77
125
124
-540
515
553
535
79
»80
78
730
250
»598
B. dispar, Nat. Mus. 4706, d'
345
"■315
»340
»340
330
"330
320
315
320
308
310
320
295
297
288
278
280
97 R
145
140
131
130
123
126
128
130
128
127
122
121
123
47
?7
665
B. dispar, Nat. Mus. skull V
687
B. dispar, Nat. Mus. 4248, &.. -- _
685
B. dispar, Nat. Mus. 4941, cf'Ctype)
B. dispar, Nat. Mus. 4245, d' .. . _.
48
45
94
92
102
110
188
155
«365
41
30
27
685
655
B. dispar, Nat. Mus. 4253
B. dispar?, Nat. Mus. 4738, 9
35
24
660
650
660
640
670
643
600
596
"560
605
545
93
108
B. dispar, Nat. Mus. 8742, 9 .. .
435
565
428
520
420
B. dispar, Nat. Mus. 4290, cf (type of
"validus"). - -
40
27
85
605
85
93
90
120
100
110
198
B. dispar, Nat. Mus. 8764, 9 .
B. dispar, Nat. Mus. 4703, cf - -
45
30
87
65
585
162
B. dispar?, Carnegie Mus. 123, 9
B. brachycephalus, Nat. Mus. 1214, c?.-
B. brachycephalus, Nat. Mus. 4274, cT..
B. brachycephalus. Am. Mus. 1495
B. brachycephalus, Nat. Mus. 8738, cf --
B. brachycephalus, Nat. Mus. 4259, cf ..
B. brachycephalus, Nat. Mus. 8739, cf--
39
- —
456
118, 171
118 171
— -
23
85
92
59
310
104
195
''505
441
-470
110
114
113
"104
101
105
+ 32
175
170
175
163
178
160
150
+ 24
— -
-500
60
93
135
648
610
B. brachycephalus, Nat. Mus. 4258, &..' 280
B. brachycephalus, Nat. Mus. 1862 i 273
B. brachycephalus, Nat. Mus. 4261, c?
(type) - 280
43
28
440
72
«520
65
113
40
----
'•580
«480
82?
85
102
81
143
B. brachycephalus?, Nat. Mus. 4940, 9 .1 265
«109
98
— ?
82
?
B. brachycephalus, Nat. Mus. 4947, 9 . 2.'i3
34
590
+ 31
380
+ 39
64
Percentage of change from B. brachy-
cephalus (type) to B. robustus (type)_.
+ 25
1
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES 481
Measurements of sJculls and jaws associated with or referred to the Brontops phylum, in millimeters
Brontops robustus
Brontops dispar
Brontops
brachy-
cephalus.
Am. Mus.
1495
Teleodus
avus, Yale
Mus. 10321
(type)
Teleodus
Am. Mus.
10690, cf
Yale Mus.
12048, d'
(type)
Princeton
Mus. 11015,
Nat. Mus.
1217, cP
Nat. Mus.
4941, tf
(type)
Ottawa
Mus.
(type)
Skulls and upper teeth :
Anterior canine to hypoconulid fossa of third
molar _
307
345
215
529
-397
385
255
665
193
"310
288
171
443
313
310
195
»505
129
P'-m3
376
236
597
345
215
196
»330
210
M'-mS
Symphysis to glenoid
Lower jaws and teeth:
Posterior canine to hypoconulid of third molar_ - _
320
"270
Pi-ma -_. --_-.._
385
240
230
"365
233
-610
o 273
Mi-ms _
183
Symphysis to condyle
" 450
Depth below third molar
99
SYSTEMATIC DESCRIPTIONS OF GENERA AND SPECIES IN
THE BRONTOPS-DIPLOCLONUS PHYLUM
Teleodus Marsh
Plate XIX, D>, D^; text figures 188, 204, 409, 413
[For original description and type references see p. 227]
Generic characters. — Related to or identical with the
lower Oligocene Brontops. Three lower incisors in
each jaw, instead of two as in other Oligocene titano-
theres; i2 the largest of the series; rounded or non-
cingulate incisor crowns.
Species. — Teleodus avus Marsh, Teleodus primitivus
(Lambe). As the most primitive Oligocene genus
known Teleodus should be distinguished from the most
progressive upper Eocene species, Profitanotherium
superhum, P. emarginafum, and other Eocene types.
Teleodus is distinguished from Protitanotherium by
three characters — (1) incisors smaller, round topped,
not flat topped; (2) canines erect, more slender; (3)
premolars more advanced in evolution.
Teleodus avus Marsh
{" Megacerops avus" Osborn, 1902)
Plate XIX, D; text figures 188, 409
[For original description and type references see p. 228J
Type locality and geologic horizon. — Chadron forma-
tion, lower Titanotherium zone. Exact locality not
published.
Specific characters. — Dental formula I-j, Cx, P-j,
M^. Of the three lower incisors in place on each
side the middle one is the largest; there is a short
diastema behind the lower canine, but no first pre-
molar. P^-p* 106 millimeters; molars incomplete.
Type, Yale Mus. 10321.
As already indicated, this species is known only
from the type lower jaw in the Yale Museum collec-
tion and is of exceptional interest because it retains,
either as a reversional or as a fixed generic or specific
character, the full number (6) of lower incisor teeth,
as in all the Eocene titanotheres, the formula thus
being I^ as against I^qj-
Consistent with this primitive morphologic condi-
tion the type jaw was found by Hatcher near the
base of the lower Titanotherium zone, and if the
structural indications are supported by other speci-
mens it may well remain as the type of a distinct
genus, Teleodus Marsh; yet other structural characters
relate it closely to Brontops and point rather to affinity
to this genus.
Dentition. — The summits of the incisors are ob-
tusely rounded and noncingulate, as in Brontops; the
lateral pair (I3) is so much reduced and crowded as
to furnish conclusive evidence that it is the lateral
incisors (I3) which first disappear in all the bron-
topine titanotheres. The second incisors (I2) are
relatively large — larger, in fact, than the median pair.
482
TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
which in turn are a trifle larger than the external
pair. The incisors of Teleodus avus thus differ widely
from those of Protitanotherium emarginatum, since
they are round topped and small instead of being
flat topped and large. The canines are of medium
size, rounded in basal section, with faint posterior
cingula. The presence of only three premolars is also
noteworthy. There is no alveolus for pi. P2 is
separated by a narrow diastema from the canine.
The combined measurement of the three premolars
(106 mm.) is only 5 millimeters greater than that of
the posterior molar, which shows in a marked way the
inferiority of the premolar to the molar series.
The jaw represents an individual slightly smaller
than the paratype of Menodus proutii Leidy and larger
than the jaws attributed to Brontops hrachycepJialus.
The canines show that this specimen was a male, and
the advanced wear of the molar teeth indicates that
the animal was fully adult. The external cingulum is
deficient opposite the widest part of the molar crowns,
whereas in Menodus and Brontops the cingulum tends
to surround the molars completely; in fact, through-
out the series the external cingulum is not very promi-
nent. The internal crest of the hypoconulid of ma
is moderately developed and crenulate.
Taken together the characters of the canines, of
the incisors, and of the grinding teeth suggest affinity
to Brontops, and for the present we may regard the
genus Teleodus as of subgeneric rank.
Teleodus primitirus (Lambe)
Text figures 204, 413, A
(For origina] description and type references see p. 235)
Type locality and geologic horizon. — Cypress Hills,
Saskatchewan; lower TitanotJierium zone.
Specific cJiaracters. — ly, P^. Second lower incisor
the largest, median lower incisor the smallest, crowns
rounded. Premolars, 103 miUimeters (estimated);
molars 183 (estimated). Lambe writes:
In the Cypress Hills specimen the crowns of the incisors are
of a depressed spherical shape, with a tendency to come to a
rounded central point above. The second incisor is the largest,
and the first is slightly smaller than the third, which is the
most upright. The first is more procumbent than the second.
Between the inner pair is a very decided interval, leaving a
space of 6.5 millimeters between the crowns of the two teeth.
The crowns of the canines are broken off, * * * and the
right first premolar is lost from its alveolus. [Lambe, 1908.1,
pp. 49, 50.]
This apparently valid type represents the smallest
and most primitive Oligocene titanothere known at
the present time (1914). It is from the Cypress
Hills of Saskatchewan. It differs from T. avus in its
inferior size, in the presence of four lower premolars
and in the proportions of the incisor teeth, in which
is is smaller than ii. The rounded form of the
crowns of these incisors relates them to those of B.
irachycephalus. Another Brontops affinity is seen in
the presence of an external cingulum on the grinding
teeth. Lambe's detailed comparison with " Menodus"
angustigenis Cope is as follows (op. cit., p. 50):
Compared with Megacerops angustigenis (Cope) * * *
M. ■primiiivus differs in the following respects: There are six
incisors instead of four, and the breadth of the jaw between the
canines, which are of smaller diameter, is relatively greater;
the diastema between the canine and the first premolar is twice
as long; the symphysis is of greater length, reaching back to a
point almost in line with the posterior edge of the fourth
premolar (in angustigenis in line with the anterior root of the
corresponding tooth) ; the exterior cingula are much less de-
veloped; the coronoid process is shorter. In both species the
premolars have reached about the same stage of development
toward the molar pattern, and the molars have very much the
same proportions. In angustigenis the internal cingula are
partially developed. In primitivus the mandible is shorter,
proportionately deeper, and not so thick in the neighborhood
of the alveolar border posteriorly.
Brontops Marsh
("Megacerops phylum," Osborn, 1902)
Plates XVIII, XX, XXI, XXIII, XXIV, XXXIII-XXXVI,
XLIII, XLVII, LXXXII-CIII, CLIII, CLX, CXCV-
CCXXIX; text figures 24, 28, 176, 179, 180, 195, 374, 375,
384-387, 389, 391, 393, 394, 396, 398-400, 406, 408-425,
482, 483, 510, 606-611, 616, 618, 622-024, 626, 628, 630, 635,
636, 639, 640, 650, 651, 653-655, 657-660, 665, 669, 673, 685,
686, 693, 699, 708, 712, 719, 727, 744, 746
[For original description and type references see p. 222. For skeletal characters
see pp. 674-6761
Geologic horizon. — From lower to upper TitanotJie-
rium zones.
Generic characters. — Upper and lower incisors 2-1.
Canines pointed, becoming obtuse. Premolars cin-
gulate; tetartocones retarded. Horns of an elongate
to transverse oval section, shifting forward. Skulls
progressively brachycephalic.
Incisors. — The Brontops phylum exhibits an assem-
blage of characters that sharply mark it off from the
main phyla of Brontotherium and Megacerops, as
follows : In young animals there are two upper incisor
teeth on each side (i^, i^), which in some specimens
tend to be reduced in old age to one on each side (i^)
by the loss of the median incisor; thus the incisive
formula, ffx) clearly distinguishes these animals from
Menodus and Megacerops, in both of which the in-
cisors are vestigial, the formula being f^. In Bronto-
therium the formula is the same (fEy,) but both upper
incisors, in the males at least, are strongly persistent.
Canines. — In the early stages of Brontops the canines
are less elongate and less pointed than in Menodus,
less robust and obtuse than in Brontotherium, but
somewhat larger than in the feebly armed Megacerops;
they are generally of a graceful, recurved form.
Premolars. — As noted above, the premolar grinding
teeth are decidedly retarded in complication — that is,
in the assumption of tetartocones and a molar pattern.
Thus the premolars of these animals from levels B
and A can readily be distinguished from those of
Menodus, which are advanced or progressive in com-
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
483
plication, as well as from those of Brontoiherium,
which are rapidly progressive. The presence of both
internal and external cingula is a character which
Brontops presents in common with Menodus.
Skull. — It is to be noted that the horns shift forward
and evolve at the expense of the nasals, as in Mega-
cerops and Brontoiherium, whereas in Menodus the
horns are stationary in their relative position. The
progressive brachycephaly of Brontops affects chiefly
Specific cTiaracters. — Skull small, premaxillaries to
condyles about 580 to 590 millimeters; males with
strongly expanded zygomata, zygomatic index 82;
transverse measurement about 480 millimeters.
Females more mesaticephalic, zygomatic index 64.
Horns very short ; outside length in transitional skulls
(male), 85 to 135 millimeters; horn section antero-
posteriorly elongate to roundly trihedral. Nasals
elongate, not spreading distally; free length (esti-
B
mm,
A
m
D
-Sections and contours of skulls of Brontops brachycephalus
Figure 414.-
A, Nat. Mus. 4947, 9 (cotype) and B, Nat. Mus. 4261 (type) ; botb from Chadron A. In these more primitive members the horns are small bosses, rounded to elongate oval
in basal section, the nasals are long and tapering, the parietal crest is narrow. C, Nat. Mas. 1214; upper levels of Chadron A; horns distinctly larger, rounded trihedral
in section with external, internal, and posterior faces; nasals thicker and relatively shorter. D, Nat. Mus. 4268; base of Chadron B; an immature animal of more pro-
gressive type with horns larger and oblique oval in section, nasals wider, zygomata stouter. All one-seventh natural size. (Of. flg. 376.)
not only the zygomatic arches, which spring strongly
outward, especially in the old males, but also the entire
vertex of the skull.
Brontops brachycephalus (Osborn)
{Megacerops brachycephalus Osborn, 1902)
Plates XXI, XLVII, LXXXIII-LXXXVIII; text figures 195,
374, 387, 389, 391, 393, 396, 399, 406, 408-419, 610, 611, 712
719, 744
[For original description and type references see p. 231. For skeletal characters
see p. 676]
Geologic horizon. — Lower Titanotherium zone (Chad-
ron A, A 1, A 2, A 3).
mated), 85 to 109 millimeters. Nasal index, 90 to
140. Grinding series, p'-m', in males 280 millimeters,
in females 253 ; dental index in males 48, in females 43 ;
true molars, in males' 178 millimeters, in females 150.
Canines not strongly compressed anteroposteriorly,
length of crown in males 34 to 40 millimeters, in
females 34. Tetartocone of p* rudimentary (a con-
cave spur from the deuterocone continuous with the
internal cingulum)^ tetartocones of p^ and p' small,
constriction separating tetartocone from deuterocone
in the hinder third of the crown. Backward pro-
longation of occiput moderate.
484
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Brontops iracJiycephalus is distinguished from its
successor, B. dispar, by the much smaller size of all
parts of the skull. From Allops marshi it is distin-
guished by its somewhat smaller dimensions. From
Menodus proutii it is distinguished especially by its
shorter molar series and relatively wider premolars,
with less progressive tetartocones. These differences
are shown below.
Measurements of skulls of Brontops, Allops, and Menodus, in
millimeters
B. braohycephalus
A. marshi
a"
9 (Nat.
Mus. 4947)
M. proutii
Pi-m3. ... _
265-297
101-123
160-180
590-610
72-82 (?)
253
105
150
590
64
300-335
119-135
185-205
630-675
64-71
333
pi-p4... .. .
135
Mi-m3 .. --
203
Pmx to condyles
Zygomatic index
628
(?)
A female skull (Nat. Mus. 4947) that is referred to
B. hrachycepTialus differs from the male skuUs in its
Figure 415. — Upper canines and incisors of Brontops braohycephalus
Nat. Mus. 4947, 9 1 canines subconic witti slender cingula; incisors (i', i-) small, with circular and
rounded crowns. Natural size.
smaller size and lesser width throughout. From the
type of AUops walcotti this female skull differs in its
smaller size, especially of p'-m^ (253:285 mm.). In
the males of B. bracliyceplialus the fourth premolar is
narrower than in the species of AUops.
Materials. — The specific characters of B. hracTiycepJi-
alus rest upon exceptionally complete material, most
of which is in the National Museum. The type of this
species is the adult male skull Nat. Mus. 4261 (PL
LXXXVII), from the base of the lower Titanotherium
zone. The paratype is the exceptionally aged female
skull (Nat. Mus. 4947; Pis. LXXXIII, LXXXVII)
foimd by Hatcher at the very base of the Titanotherium
zone only 14.4 feet above the Cretaceous Pierre shale.
A third specimen is skull Nat. Mus. 4940, also re-
corded from the lower beds but from the upper level
(A 3). A jaw (Am. Mus. 4247) agrees closely in size
with this skull but may belong with Allops walcotti.
A fourth specimen (Nat. Mus 1214, fig. 418) also
comes from the upper level of A, or the lower beds,
and it shows features of approach to B. dispar. A
fifth skull (Am. Mus. 1495) is distinguished by its
very small horns.
General characters. — These small, broad-skulled ti-
tanotheres are distinguished by the lateral arching and
relatively broad section of the zygomatic buccal proc-
esses. The nasals taper toward the extremities and
are slightly decurved. The horns rise above the skull
from 25 millimeters in the females to 60 in the male
skull, the corresponding outside measurements beiug
85 and 102. The basal section of the horns (fig. 414)
is highly characteristic, consistiug of a broad outer
maxillary face, lying in an oblique plane, and a broad,
highly convex inner face. The horns are obtusely
roimded at the top. The superior iacisors (fig. 415)
are two in number on each side in yoimg animals,
but the iuternal pair, I^ are variable and likely to dis-
appear in aged animals. The canines are elongate,
pointed, with a posterior cingulum only and a strongly
convex posterior face, varying in length from 33
millimeters in the females to 43 in the males. A
very distinctive character is the structure of the
premolars, the main internal portion of the crown
being composed of the deuterocones, while the tetar-
tocones are small and rudimentary, especially in p^ p*.
In the type male skull the tetartocone of p^
on the right is a prolongation of the deutero-
cone, concave and continuous with the cin-
gulum on the lingual or internal border; on
the left side the tetartocone is more circular.
In Am. Mus. 1495 the tetartocone on p* is a
concave spur from the cingulum; the tetar-
tocone on p' is concave on the lingual side,
convex on the buccal side, which is evidence
that the buccal side develops faster. The pre-
molar cingula are developed progressively in
ascending levels but are never so broad as
in Menodus. A primitive feature of m^ is
the aborted metaloph extending inward upon the
crown. There are incipient protostyles on the superior
molars.
Paratype jemale sTcull. — The very aged female skull
(Nat. Mus. 4947) in the eleventh stage of growth, dis-
covered by Hatcher in 1887 on the bottom level of the
lower Titanotherium zone, is the paratype of this
species. The palate is finely preserved (PL
LXXXVII). It is by far the smallest and one of the
most primitive skulls of the brachycephalic series; it
measures only 590 millimeters from the occipital con-
dyles to the premaxillae and 380 millimeters across
the zygomata, while the total premolar-molar series is
only 253 millimeters as compared with 320 millimeters
in the lower jaw of the type of Teleodus amis. The
nasals are rather elongate, contracting anteriorly.
The horns are small, rounded knobs rising 25 milli-
meters above the top of the skull from a point slightly
anterior to the orbits — in fact, immediately in front
of the preorbital malar bar; they are directed outward
and backward, rugose at the tips and set wide apart,
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
485
overhanging the sides of the maxillae; the horn sec-
tions are obliquely placed ovals that converge ante-
riorly.
Figure 416. — Reconstruction of crushed skull of
Brontops brachycephalus
Am. Mus. 1495; Chadron A; South Dakota. Ai, Skull fully
adult, less brachycephalio than the type of B. brachycephalits,
and one of the most primitive known of all Brontopinae. One-
sixth natural size. Very small horns (fundamentally similar
to those of the Eocene Manteoceras) consisting of swellings on
the frontals which overlap the nasals. The surface of the horn
swelling (Ai) is roughened and pitted as if for the attachment
of a tough or even horny epidermal cap. A2, Horn region.
One-third natural size. The upper teeth of this skull are shown
in Figure 417; the lower jaw in Figure 413, B.
The cranium proper is narrow, measuring
80 millimeters across the vertex; the occiput
is broken away; the zygomata arch outward
and give a brachycephalic character to the
skull, although the measurements show that
it is in reality mesaticephalic, the zygomatic
index being only 64. This mesaticephaly is a female
character. In section these arches are deep, thin
bars. The posterior nares are placed very far back,
opposite the posterior face of m^ The vomer ex-
tends well back on the basisphenoid with a median
keel in the portion lying between the pterygoids,
extending back almost as far as the basisphenoid
rugosity, as seen also in succeeding types. The ali-
sphenoid canal, foramen ovale, f. lacerum medium,
and f. condylare are clearly defined. The somewhat
rugose junction of the basisphenoid and basioccipitals
foreshadows the prominent projection for the rectus
capitis muscle, which is so characteristic of succeeding
species of Brontops.
Dentition. — Of the incisors two (i\ i^) persist on one
side and only one on the other. The premolars are
completely worn down, but so far as one can observe
the deuterocone constitutes almost the entire inner face
of the crown "of p*. The hypocone of m' is a cingule
connected with a very interesting abortive metaloph, as
seen in some other titanotheres. The inner contour of
the crown is very rounded, the primitive triangle being
scarcely altered by the addition of the slight protostyle
and hypocone.
Type male skull. — The upper surface of this aged
male skull (Nat. Mus. 4261, PI. LXXXVII, figs. 399,
410, 411, 414) is especially well preserved. Although
compared with the female it is a relatively large and
adult animal, in the ninth stage of growth, the horns
rise only 85 millimeters above the nasals (outside meas-
urement). They exhibit the primitive or original
position of these structures — namely, very far back and
immediately above the orbits, directed outward. The
very robust section of the horn base is a male character
distinguishing this from the above-described female
horn. This skull is very interesting, moreover, because
it illustrates the great breadth (480 mm., estimated)
attained by the zygomatic arches in this low geologic
level. This brachycephaly is a male character. The
dental formula is I^~, P*~'. The premolars are dis-
tinguished by narrow internal cingula and rudimentary
tetartocones; the teeth exhibit faint external and de-
cided internal cingula, but the most decided feature is
the sessile character of the tetartocones, which occupy
barely one-third of the inner face of the premolar
Figure 417. — Left upper grinding teeth of Brontops brachycephalus
Am. Mus. 1495. One-third natural size. These teeth agree well in form and measurements with
those typical of B. brachycephalus. They difler from those of Allops walcotii (type) in minor
points, and from those of Menodus proutii in being less elongate anteroposteriorly, especially
the molars, which also have the internal eingulum less pronounced. The premolar pattern
resembles that of aU the more primitive members of the menodontine group. (See flgs. 413, 416.)
crowns. The incisors are absolutely transverse in
position with perfectly smooth, rounded crowns; the
canines are quite large.
The American Museum specimen (No. 1495), the
skull of a yoimg animal, also exhibits the character-
486
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
istic features of this species; the horns are somewhat
smaller than in ProtitanotJierium emarginatum of the
upper Eocene. They present an elongate oval section,
with sloping anterior face. The zygomata are fairly
developed and typical in section. The anteroposterior
and transverse diameters of the molars are subequal.
The tetartocone of p* is a concave spur from the
the angle projects downward and backward; the
coronoid is high; the molars are relatively larger.
As arranged in the order of sequence from the base
to the summit of the lower zone the remarkable series
of skulls in the National Museum (see table on p. 480)
gives us a complete series of ascending mutations toward
the higher stages of B. dispar in the middle beds.
Proportions of teeth of Brontops brachycephalus,
in millimeters
Antero-
posterior
Transverse
c
pi
P2
PS
P^
Ml
M2
M3
23
23
28
32
38
49
59
61
20
20
32
43
54
49
59
63
Figure 418. — Skull (B) and horn region (A) of Brontops brachycephalus?
Nat. Mus. 1214; Chadron A 3. The horns are more progressive than in the type of this species and in their
trihedral section suggest Allops marsM, The dental measurements agree better with those of B, bra-
chycephalus, although the premolar series is unusually long. The large canines probably indicate male
sex. One-third natural size.
cingulum. P' is set on a line with p^ (outer wall) and
is separated by a diastema from the canine. The
canines are procumbent, as in Nat. Mus. 4261, with
faint external cingula. M^ has crenulate metalophs,
m3 a hypoconulid not greatly different from that of
Menodus. The jaw of this specimen is shown in Fig-
ure 396, A. As compared with Teleodus avus this
jaw has the horizontal ramus very shallow anteriorly;
Transitional stages from Brontops hra-
cJiycephalus to B. dispar. — A transitional
stage is represented by a rather primitive
skull (Nat. Mus. 4940), from the top level
(A 3) of the lower beds, which is interme-
diate in the length of its horns between
the type of B. brachycephalus and some
of the more primitive varieties of B.
dispar. This skull taken as a whole is
also somewhat more progressive in char-
acter than the male type and female par-
atype above described. The nasals are
thin, elongate, and contract anteriorly.
The horns present a decidedly oblique
oval form at the base but round up
toward the summit into a more cylin-
drical section; they rise 60 millimeters
above the midline of the skull, and since
this animal is well advanced toward the
eighth stage of growth, this is probably
the maximum development of the horns
in the males at this stage. The pos-
terior nares open somewhat more ante-
riorly than in the type of B. brachy-
cephalus.
There is but one incisor above on each
side and the first superior premolar is rep-
resented by its alveolus only on the right side, the alve-
olus being closed on the left. It is important to note
this reduction of the median incisor, or ii, and the shed-
ding of the first superior premolar. The remaining pre-
molars, p^-p*, show a slightly stronger development of
the tetartocones than in the B. brachycephalus type and
paratype, although in p* the deuterocone still constitutes
the main internal portion of the crown, as in the type
EVOLUTION OP THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
487
of B. hrachycephalus ; in m^ the abortive metaloph ex-
tends outward from tlie cingule without a distinct
hypocone.
A second transitional skull from upper A (A 3) (Nat.
Mus. 1214; fig. 418) may also be considered transi-
tional between B. iracJiycepJialus and B. dispar. It
is in the seventh growth stage. The median pair of
incisors (i|) has also dropped out entirely; the first
superior premolar persists on one side and has dis-
appeared on the other, so that the formula in both
somewhat similar to that in Leidy's type of Mega-
cerops; the horns show a more trihedral section at the
base than in the preceding specimens owing to the
incipient development of the connecting ridge (cr).
A transitional stage from lower B is represented by
the young male skull (Nat. Mus. 4258) found in B,
or the middle levels — namely, 71.45 feet above the
Cretaceous Pierre shale. It gives us valuable details
of the architecture of the antei'ior portion of the face.
Seen from above (Pis. LXXXIV, LXXXV) the
A
D
Figure 419. — Sections and contours of skulls of Brontops hrachycephalus (A) and B. dispar (B, C, D)
, Nat. Mus. 4259; B, Nat. Mus. 4941; C, Nat. Mus. 4703; D, Nat. Mus. 4290 (type of B. validus). All one-eighth natural size. The more progressive members of B.
brachycephalus (A) approach B. dispar but are distinguished by their smaller size, small horns, low connecting crest, and relatively slender zygomata. In B. dispar
(B, C, D), which is characteristic of Chadron B, the horns are much stouter with rounded basal section, the nasals thicker and shorter, and the zygomata more
massive. The more progressive members (D) approach B. roiustus. Note the sunilarity of the buccal section in D to that of Am. Mus. 518, 9 (flg. 424; see also
fig. 376, showing position of sections and contours).
specimens is I', P*~', demonstrating absolutely the
variability of p^ The marked primitive feature of
the premolars is that while in p^, p^ the tetartocones
are becoming more distinct and rounded, in p* the
deuterocone still forms almost the entire inner surface
of the crown. In this, as in the preceding types, the
metaloph extends from the hypocone cingule of m'
outward toward the metacone. The rugosity on the
basisphenoid is very prominent; the nasals are of
medium length in this specimen, expanding slightly at
the sides, which is an age character, and their form is
101959— 28— VOL 1 34
frontals cap the nasals and constitute the entire tips
of the horns. In side view, however (PI. LXXXIV),
the nasals are seen to support the main anterior por-
tion of the base of the horn, resting postero-inferiorly
on the lacrimals, elements which are not quite so ex-
tensive as represented in Plate LXXXIV, A^. Sub-
sequently the lacrimal coalesces with the malar, and
we speak of the ridge extending down from the side of
the horns as the malar or external ridge. In palatal
view the posterior nares open between m^ and m'.
The vomer (PI. LXXXVI) extends well backward.
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The transitional skull just described (Nat. Mus.
4258) and another skull (Nat. Mus. 4259) belong to
B. hrachyceplialus rather than to B. dispar, because,
as shown in the table of measurements above, they
agree far better with B. hracliycepJialus in tooth meas-
urements, the horn sections (figs. 414, 419) agree
more closely with those of B. hrachycephalus , and in
their tetartocone development the premolars are
more advanced than the type of B. hrachycephalus but
less advanced than in the type and referred specimens
of B. dispar.
Review of primitive and progressive characters oi
Brontops hrachycephalus. — Nasals: All the skulls ex-
hibit nasals having a median terminal notch. The
nasals are nari'owest distally in the fourth stage of
growth and expand toward the extremities in the
seventh stage; the progressive shortening and distal
widening of the nasals are distinctive features of the
Brontops phylum. In general, abbreviation of the
nasals is correlated with brachycephaly.
Horn growth: The sexual distinctions of size and
form in the horns are well marked. The horns not
only increase steadily in size in succeeding geologic
levels but in course of individual growth they re-
capitulate clearly the ancestral Eocene history of
this species. In young bulls the horns present an
elongate oval, as in the upper Eocene Protitanotherium
emarginatum; as the males become older the antero-
posterior diameter relatively decreases, the transverse
diameter increases, and the horn gains in height,
rising only 37 millimeters above the skull in the
fourth stage.
Skull growth: In the meantime the summit of the
skull, or parietal vertex, does not perceptibly increase
in width; the top of the occiput extends considerably
back of the zygomata. The zygomatic arch sections
are as highly distinctive as the horn sections, since
the transverse diameters nearly equal the vertical
diameters; the zygomatic arch is relatively broad
and shallow and contrasts with that of Menodus
heloceras, which is distinctly dolichocephalic.
Teeth; variability of incisors: There are two well-
developed incisors in the youngest skull transitional
to B. dispar (Nat. Mus. 4258); the median pair has
dropped out of another skull in the later part of
the seventh stage (Nat. Mus. 1214); but both in-
cisors persist in the still older type male (Nat. Mus.
4261). In all the male skulls the canines are promi-
nent and pointed, measuring from 40 to 43 milli-
meters in length in the males and resembling in
form those of B. dispar.
Pi'emolars: The first superior premolar is small;
it exhibits an imperfectly formed internal crest and
no tetartocone. The constitution of the remaining
premolars, p^-p* (Nat. Mus. 4258, 1214), is mter-
esting as exhibiting the relative rate of evolution of
the internal cusps and bringing out the fact that the
anterior premolars are more progressive than the
posterior; this condition is adaptively correlated with
the fact that the anterior premolars come into use
earlier than the posterior premolars, as shown by
Hatcher (1901.1, p. 261). Thus p^, although the
smallest tooth, exhibits the largest tetartocone, form-
ing a direct continuation of the crest of the deutero-
cone; p^ has its tetartocone distinctly budded off,
with conical sides; while in p* the tetartocone is a
mere cingule (figs. 406, 417). Whereas in the typical
B. hrachycephalus of the lower A levels the deutero-
cone constitutes almost the entire inner surface of
the crown, especially in p^"'*, and the tetartocones are
rudimentary, in the later geologic stages (Nat. Mus.
4258, 4259) the tetartocones evolve into small, low,
irregular-shaped cusps, leading into the condition
found in the type of B. validus (Nat. Mus. 4290).
The variability of the first superior premolar is shown,
as mentioned above, in one of the older B. hrachy-
cephalus skulls (Nat. Mus. 1214), in which it persists
on one side but has disappeared on the other; this
reduction and variability is seen also in B. dispar and
in Menodus. A very important character consists in
the fact that whereas the external cingulum on the
premolars is variable and feebly developed in some
skulls and more strongly marked in others (Am. Mus.
1495, Nat. Mus. 4258), the internal cingulum is con-
stant but never very broad, as it is in Menodus
trigonoceras .
Molars: The ratio of molar length to premolar
length in this evolution stage is as 175 to 114. In
the molar series m^ also offers a very distinctive
character: the hypocone is a cingule, and as in B.
hrachycephalus an abortive crenulate crest (metaloph)
extends from this toward the metacone; the strong
cingule in front of the protocone known as the "pro-
tostyle" is here a feeble, inward-directed ridge.
Brontops dispar Marsh
{Brontops validus Marsh; " Megacerops dispar" Osborn, 1902)
Plates XXIV, XXXIII, LXXXVIII, XCIV; text figures 24,
180, 375, 384-386, 389, 391, 394, 396, 399, 409-412, 419, 421,
423, 609, 616, 719, 744
For original description and type references see p. 223. For slceletal characters
see p. 664]
Geologic horizon. — Middle beds (Chadron B) to
lower portion of upper beds (C).
Specific characters. — Skull of intermediate size in
males; incisive border to condyles 660 to 685 milli-
meters, in female 610. Males brachycephalic, zygo-
matic index 73 to 87, average 80; zygomatic width
555 milhmeters (average). Females more mesati-
cephalic, zygomatic index 60-71, zygomatic width
390 millimeters. Horns progressively shifting anteri-
orly, more elongate, 162 to 200 millimeters. Basal
section roimded, summits cylindrical. Nasals pro-
gressively abbreviated (105 to 85 mm.), with for-
ward-shifting horns, spreading distally. Grinding
series 310 to 340 millimeters, average 329. Dental
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
489
index 46 to 49, average 47. Canines pointed and
more or less compressed anteroposteriorly. Pre-
molars more progressive; tetartocones p^"^, more
advanced than in B. bracliycephalus. Tetartocone
of p^ progressively developed from a concave spur
from the deuterocone, continuous with the cingulum,
to a rudimentary cusp. Backward prolongation of
the occiput behind the zygomata very pronounced.
This significant species connects closely with the
most progressive stages of Brontops hrachycepJialus
and with certain skulls which may be referred to
Brontops rohustus, although the typical B. dispar
does not appear to be directly ancestral to the typical
B. rohustus.
The range in size in skulls referred to Brontops
dispar is given above.
From Allops marsJii, as well as from Brontops
hrachycepJialus , the skull of this species is distinguished
especially by its greater robustness (zygomatic index
78-87, as compared with 64-71), larger horns, and
shorter, thicker nasals, as indicated in the following
measurements:
Measurements of Brontops dispar and Allops marshi, in
millimeters
Pi-m3
P'-p4
Mi-m3
Canine:
Vertical
Anteroposterior. .
Pmx to condyles
Zygomatic width
Cephalic index
Nasal length ■_
Nasal breadth
Horns, external length
! . dispar, Nat,
Mus. 4290
(type of
B. validus)
320
130
203
40
27
660
565
85
85
120
198
A. marshi,
Am. Mus. 501
(type)
310
122
191
34
25
673
450
67
100
105
105
From Menodus proutii we observe that B. dispar is
distinguished by its brachycephaly, by the greater
relative width of its premolars and molars, and by the
more retarded development of the premolar tetarto-
cones, especially in p*.
From Allops serotinus and A. crassicornis we observe
that B. dispar is usually distinguished by its higher
zygomatic index and by the lesser width of the pre-
molars.
The type of B. dispar, unfortunately a badly
crushed skull, has less robust horns and nasals than
the type of B. validus, but the general agreement in
measurements as well as in dental characters appears
to indicate that these two types belong either to the
same or to very nearly related species.
Geographic and geologic distribution. — According to
Hatcher's record, the typical B. dispar skulls were
found mainly in B or the middle Titanotherium zone,
but they extend also into the lower levels of the upper
beds, namely, lower C (Chadron C 1). The four
ascending skulls in the National Museum most ac-
curately recorded by Hatcher and Stanton in 1901
are from 55 to 71 feet above the Cretaceous Pierre
shale, as follows:
Skull 4258, Brontops bracliycephalus (transitional), 71.45 feet
above the Pierre.
SkuU 4259, B. brachycephalus (transitional), 55.6 feet above
the Pierre.
Skull 4290, B. dispar (type of B. validus), 62 feet above the
Pierre.
Skull 4703, B. dispar, 62 feet above the Pierre.
Geologic variation. — Inconsistent with the general
observation that the more progressive forms are found
on higher levels is the fact that the specimen first
named (Nat. Mus. 4258) is a very primitive form
which is distinctly related to B. hracTiycephalus, as
noted above. The type of Diploclonus tyleri, although
very advanced, was also found at a very low level —
namely, 35 feet above the Pierre shale, in 200 feet of
beds belonging to the Titanotherium zone (Lull).
These inequalities are certainly due to the uneven
surfaces of the Pierre shale on which the White River
was deposited.
Materials of B. dispar. — This species is represented
in Hatcher's collection for the Geological Survey by
twenty-one or more skulls and several jaws now in
the National Museum, including especially the follow-
ing: The juvenile transitional skull (No. 4259) from
lower B, or the middle beds, referred to B. brachyce-
phalus; the male type of B. dispar (Nat. Mus. 4941)
from middle B; four finely preserved male skulls —
namely. No. 4703 from middle B, No. 4290 from
middle B (this specimen is the type of Brontops
validus), and Nos. 4245, 4248 from B. The two skulls
last named are in the seventh and ninth stages of
growth. The female sex is represented by No. 4738
from lower B. There is also a complete skull and
ower jaw (Nat. Mus. 1217) from the upper zone
(level C).
This magnificent collection of well-recorded material
in the National Museum enables us to determine
positively the range of Brontops dispar from lower
Chadron B to the base of Chadron C. In this long
geologic range there are marked progressive advances
in evolution.
B. validus a synonym of B. dispar. — Careful com-
parison of the measurements and characters of the
type of B. dispar (Nat. Mus. 4941), from middle B,
with the type of B. validus (Nat. Mus. 4290), also from
middle B, together with the supplementary evidence
furnished by the collection of skulls in the National
Museum, shows that these nominal species are identi-
cal. They both represent the brachycephalic, short-
horned titanotheres of the middle levels. Although
the type of B. validus has on each side two upper
incisors and four premolars, a study of other materia
490
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
proves that both the median incisors and the anterior
premolars may be shed at an early age. In other
words, it is demonstrable in B. dispar, as in B. irachy-
ceplialus, that the median incisors (i') are variable,
that they are found in all young animals, that they
may or may not drop out in later life. The lateral
incisors (i^) are constant. The stage of evolution
of the premolars in the type of B. validus is identical
with that in the type of B. dispar and the sagittal
sections of the skull in the two type specimens are
closely similar. The horn and nasal sections in the
two type specimens are essentially similar; the
differences (fig. 419) are due to the lateral crushing
and less advanced age of the skull of B. dispar.
Chief distinctions from other species. — B. dispar is
thus readily distinguished from B. brachycephalus by
its greater measurements and more progressive
characters throughout; from Allops marshi by the
cylindrical section of its horns and by the great back-
ward extension of its occiput ; and from the geologically
succeeding B. rohvstus by the smaller size and cylindri-
cal section of the horns and the backward extension
of the occiput.
Detailed characters oj the type of Brontops dispar. —
The type of Brontus dispar consists of a male jaw
and skull (Nat. Mus. 4941) with the occipital region
entirely wanting (PL XCIII). The extreme lateral
crushing of the skull obscures its real characters and
actual resemblance to the type, which Marsh sub-
sequently named Brontops validus. As association
of the skull and jaw is rare and as this specimen rep-
resents one of the predominant species of the middle
zone, the type deserves full description and measure-
ments in addition to those given in the comparative
columns above.
Measurements of type of Brontops dispar, in millimeters
Inferior canine (partly worn) , anterior 41
Inferior canine, anteroposterior 30
Length of symphysis 159
Width of condyle - 137
Premolars and molars:
p2
p3
P<
ml
m}
m3
Superior series:
Anteroposterior. _
30
40
43
62
78
80
Transverse
39
60
60
72
80
90
pj
pi
Pi
mi
m2
mi
Inferior series:
Anteroposterior. .
31
39
45
63
77
108
Transverse
22
29
33
40
44
44
Teeth: Formula, 1\, P|f|. A stout lateral pair of
incisors persists in both upper and lower jaws, though
the animal is in the beginning of the eighth stage.
The median incisors are wanting in both jaws. The
canines are prominent and recurved, the superior
canine is embraced by a posterior cingulum, while the
inferior canine has a faint external cingulum sur-
rounding all except the inner side. The anterior pre-
molars (pj) have dropped out in both upper and
lower jaws, being represented only by large alveoli;
p^, p^ p* have moderately developed internal and
vestigial external cingula and tetartocones in the
middle stage of development. The hypocone is a
prominent cingule on m'. The second lower premolar,
P2, is an elongate simple tooth, the metalophid not
being defined at all, and the hypolophid forming an
incipient crescent; ps, p4 present double crescents and a
rudimentary metastylid. The lower molars exhibit
partly developed external cingula, reduced opposite
the protoconid and hypoconid ; they also exhibit rudi-
ments of the metastylid. The last inferior molar ip
highly characteristic of the Brontops- Menodus group,
a broad internal shelf extending inward from the
hypoconulid, with a trace of a posterior cingulum
beneath it. The much crushed jaw exhibits a shallow
symphysis, a large mental foramen beneath pa.
Plate XCIV gives its partly restored outlines.
Skull: The skull exhibits the short, truncate, and
cleft nasals, short horns with a subtrihedral section
at the base, rounded malar, nasal and connecting
ridges, and roimded faces. The posterior face is
broad and imiformly convex, as in the other advanced
members of this phylum. The horns have a charac-
teristic angle of the outer border, leadiag into a
rounded summit, as shown in the section. The
zygomatic arches are robust and convex, but they are
too much crushed for an accurate section.
Type of Brontops validus. — The type of the syno-
nymic species Brontops validus (Nat. Mus. 4290) is a
male in the seventh growth stage. It is well figured
in Plates LXXXVIII, LXXXIX of the present mono-
graph. The principal measurements are as follows:
Measurements of type of Brontops validus
Millimeters
Molar-premolar series 320
Condyles to premaxillary tips 660
Free portion of nasals, length 85
Free portion of nasals, breadth 120
Length of horn to nares 199
There are two incisors above and below ; the canines
are large and pointed, indicating that this specimen
was a male; the fourth premolar has a rather feeble
tetartocone without a distinct valley separating it
from the protocone; the hypocone of the third su-
perior molar is a shallow ridge arising from the cin-
gulum. The species is clearly characterized by the
short, square nasals and horns borne well forward;
but it can not be separated specifically from the pre-
viously described Brontops dispar.
General characters of B. dispar. — The measure-
ments of the skulls referred to B. dispar are singularly
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
491
uniform and constant, as shown in the table above.
In the males the superior premolar-molar series
measures 310 to 345 millimeters in length. The
basilar length, premaxillaries to occipital condyles,
ranges from 660 to 685 millimeters, the male zygo-
matic arches from 515 to 565, the horns from 155 to
200. In the females the dental and cranial measure-
ments exhibit the usual sexual contrasts of inferiority.
Main features of skull: The nasals are short but
broad and thick in section in the type of B. validus,
85 to 103 millimeters in length and 100 to 143 milli-
meters in breadth; in other males (Nat. Mus. 4245,
4703, 4706) they are somewhat long and narrow,
resembling those in B. hracTiycepTialus and thus afford-
ing a complete transition, as may be readily seen in
Figure 419. The adult or growth changes are
prophetic of the higher specific stages of this phylum.
Thus in early stages of growth the nasals are some-
what thinner and more narrow distally; in later stages
of growth they thicken and broaden near the ends,
while the median cleft deepens. The form of the
horns is highly distinctive of this species; seen from
in front (PI. XCIV) the lateral contours of the horns
exhibit a strong convexity above the middle portion;
from this point upward they suddenly contract into
rounded, rugose tips; this rounding and pointing of
the tips of the horns is a very distinctive feature as
compared with the broad oval tips in members of the
Brontotherium phylum. The basal section of the
horn in the type skull of B. dispar (figs. 399, 419)
approaches that of B. rohustus (fig. 421) in the round-
ing out of the "anterior" or "nasal" angle or ridge
and in the anteroposterior flattening; the posterior
faces of the horns of B. dispar, however, unlike those
in the type of B. rohustus, are directed obliquely
inward toward each other and present a long, uniform
convexity. Some horn sections are more trihedral
because of the prominence of the nasal ridge. In
lateral view the horns are seen to be slightly recurved
and to be united, especially in the old males, by a
deep and broad connecting crest. The zygomata
present prominent convex buccal processes, which
have a deep, thick section, similar to that of B.
rohustus but somewhat less heavy.
Details of male skull: The type skull of Brontops
validus, "skull K," Nat. Mus. 4290 (Pis. LXXXVIII,
LXXXIX, XCI, XCII; fig. 419), found at the 62-foot
level above the base of the Titanotherium zone, in
middle B, admirably exemplifies the distinctive
features of the skull. It is in the seventh stage of
growth. Seen from in front (PI. XCIV), the nasals
are supported by a vertical bony septum, which in
life may have been united by cartilage with a similar
bony septum that extends up from the premaxillae
(cf. Tapirella hairdi). The top of the skull is seen
to be a larger development of the B. hrachycepTialus
type (compare Pis. LXXXV and XCI), broad above
the orbits, narrowing at the vertex, with a rugose
supratemporal crest. In the middle of the parietal
vertex of certain specimens (Nat. Mus. 4703) a very
large central pit (PI. XCII) is observed. This is a
vestige of the space between the convergent supra-
temporal ridges and is similar to the median pit in
some of the Eocene forerunners of Brontops — namely,
Manteoceras. The occipital pillars are strong, but
the rugosities at the summits are somewhat narrower
than in B. rohustus. The palatal view (Pis.
LXXXVIII, LXXXIX) also illustrates the brachy-
cephalic character. The vomer extends back as a
prominent keel upon the basisphenoid, which exhibits
a robust, paired rugosity for the rectus capitis muscle
at its junction with the basioccipital.
The progressive size evolution of the skull is dis-
tinguished first by the slight increase in length and
great increase in width over that of B. hrachycepTialus,
owing partly to the rapid expansion of the buccal
processes but also to the broadening of the skull
itself; second, by the elongation and strengthening
of the horns as described above. Additional observa-
tions on this specimen are given above.
Dentition: The teeth show many characters in-
herited from B. hracTiycepTialus. The formula (If^,
P|^) exhibits hereditary variability, manifested in
the occasional loss of one incisor and one premolar.
The median incisor sometimes disappears before the
seventh stage (Nat. Mus. 4941, 4703), or it may per-
sist late in life (Nat. Mus. 4290, 4245). The canines
are relatively long and rounded in section at the base,
more or less compressed anteroposteriorly, foreshadow-
ing the Allops sublanceolate canine, less elongate than
in members of the Menodus phylum, and much less
obtuse and recurved than in members of the Bronto-
tTierium series. The first superior premolar is a fairly
large and persistent tooth, but it also sometimes
aborts and disappears early (Nat. Mus. 4703). The
ectoloph of p' is well in line with the ectoloph of p^, p^
(contrast BronfotTierium) . In the advanced stages of
B. dispar the transformation of the premolars has
proceeded much further than in any specimen of
B. hracTiycepTialus; the tetartocones of p^ and p^ are
relatively distinct, but in p* the tetartocone is still a
backward spur from the deuterocone, more rounded
than in B. rohustus, but never sharply rounded off or
prominent and always a much less distinct and sepa-
rate cusp than in p'. In the type of B. dispar, in fact,
the tetartocone of p* is only a little more advanced
than in the typical B. hracTiycepTialus. The external
cingulum of the superior premolars is fairly well
marked; the internal cingulum is progressive or
stronger than in B. hracTiycepTialus. On m^ the hypo-
cone is a prominent cingule.
Lower jaw (fig. 423): In the type jaw (Nat. Mus.
4941) vertical crushing has greatly diminished the
depth of the horizontal ramus, but the symphyseal
region was probably shallow. There was a diastema
between pi (which has dropped out) and the canine.
492
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The canine is stout and conical; the incisors are sub-
spherical; the external cingula are not as sharply
defined as in Menodus.
Female skulls: The females of this species (Nat.
Mus. 4738) are, according to the principle expressed
above, distinguished by smaller and much less robust
canines; the greatly reduced canines, the more slender
zygomatic arches, and the inferior dimensions of this
skull and teeth as a whole constitute the distinctive
sexual characters. The internal and external premolar
cingula are quite as strong as in the male representa-
tives of this species.
The skull forming Nat. Mus. 4738 (fig. 391, A)
presents rather slender zygomatic arches; the nasals,
although imperfectly preserved, indicate a slender,
unexpanded contour. This skull is in the eighth stage
of growth, yet it retains two superior incisors on each
Figure 420. — Restoration of Brontops rohustus
By Charles R. Knight. Ahout one-ninth natural size.
side. The canines measure only 35 millimeters an-
teriorly as compared with 47 in the largest males. The
protocones of m^, m' are rather low, are set well in
from the inner margin of the tooth, and have a slight
concavity on the lingual side near the top — an individ-
ual variation. A peculiar feature of this specimen is
the reduplication of the tetartocone of p"*, an anomalous
character (see explanation above) also observed in B.
hracTiycepJialus (Nat. Mus. 4259) and in the type of
Alloys crassicornis as well as in the type of Menodus
(Menops) varians. A character reversional to B.
hracTiycephalus is the abortive crenulate metaloph of
m', extending forward and inward from the hypocone.
In palatal view the less robust character of the basi-
sphenoid and the narrowness of the zygomata are well
illustrated. Another character, which may be an
individual variation, is the nonclosure of several of the
sutures, the median suture between the nasals and
frontals being still widely opened. Similarly the
external portion of the mastoid is still well defined
between the post-tympanic and the exoccipital proc-
esses. The post-tympanic and postglenoid processes
of the squamosal are not conjoined below.
Extension oj B. dispar into upper beds, or C. — ■
Although B. dispar is especially characteristic of the
middle Titanotherium zone, evidence of its occurrence
in the upper beds even to a point a little below middle
C is apparently indicated by skull No. 1217 in the
National Museum. This fine specimen with the lower
jaw attached (PI. XCIII) appears to represent a very
large male of B. dispar, the length from condyles to
premaxillaries being estimated at 702+ millimeters,
m^-m^ 196, mi-mj 230. The detailed measurements
of the grinding teeth, however, as in the large upper
zone brontotheres, are smaller through-
out than those of the large B. dispar,
Nat. Mus. 4696, from the middle beds.
Brontops robustus Marsh
{" Megacerops robustus" Osborn, 1902)
Plates XVIII, XX, XXXIII-XXXVI, XLIII,
XLVII, XCV-CIII, CXCV-CCXXIX; text
figures 24, 179, 387, 389, 393, 394, 396, 399,
400, 409-412, 420-423, 482, 483, 510, 608,
618, 622, 624, 626, 628, 630, 635, 636, 639,
640, 650, 651, 653-655, 657-660, 669, 685,
686, 693, 699, 712, 719, 727, 744
[For original description and type references see p. 222. For
skeletal characters see p. 666]
Geologic horizon. — Lower level of the
upper Titanotherium zone (C).
Specific characters. — Length of skull
765 to 813 millimeters, strongly brachy-
cephalic, zygomatic index, 77 to 87,
average 79; backward prolongation of
occiput slight or absent; horns shifted
forward and inclined forward, basal sec-
tion broadly transverse oval, summit
obtuse, outside length 207 millimeters (type); nasals
abbreviate, free length 76 millimeters, breadth 149,
nasal index 51 ; dental index 46, same as in B. brachy-
cephalus and B. dispar, total grinding series 345
to 376 milhmeters; molars 215 to 236, average
223; canines rounded in section, obtuse, stout, length
of crown 42 to 48 millimeters; tetartocones of
p^"^ somewhat less developed than in typical B.
dispar.
This species marks the end of the Brontops phylum,
toward the end of lower Oligocene time. It is related
to B. brachycephalus but does not directly succeed
any of the known mutations of B. dispar.
General characters. — This is the most fully known
species of titanothere. The fine type skull and skele-
ton in the Yale Museum (Yale Mus. 12048) represent
a middle-aged animal. Both were found in "Corral
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
493
Draw," near the base of Chadron C (upper Titanothe-
rium zone), 60 feet below the summit, as recorded by
Hatcher. This skeleton is so complete that almost all
its parts were figured for this monograph in plates and
woodcuts under the direction of Professor Marsh.
Brontops rohustus appears to have been the predomi-
nant animal during the deposition of the beds of the
lower C 1 levels of the upper Titanotherium zone in
South Dakota. It is a comparatively short-horned
type, contemporaneous Vv'ith the early members of the
that raises a doubt as to the invariable persistence of
the median incisor. These and other characters relate
this animal to Brontops hrachycepJialus and B. dispar,
in which the median superior incisor is also variable.
Convergence of Brontops rohustus to BrontotTierium. — ■
The typical B. rohustus is very easily distinguished,
but there are at least six other skulls that simulate
Brontotherium in the premolars, in the expansion of
the zygomata, and in the flattening of the horns. In
some of these also the skull vertex appears long, but
B C
Figure 421. — Sections and contours of skull of Brontops rohustus
Nat. Mus. 4696; middle levels of Chadron C; either a very large and progressive B. dispar or a primitive B. robustus, with stout cylindrical horns far
in front of the orbits, broadlv and roundly trihedral in section. Nasals short and vride. B, Am. Mus. 1083, referred to B. rohustus, although the horns
are roundly trihedral in section. C, Yale Mus. 12018 (type); lower levels of Chadron C; very thick horns incUned forward and far in front of the
orbits, in basal section very widely oval; nasals very short and thick, zygomata massive, and parietal crest wide. All one-eighth natural size.
long-horned series of brontotheres, such as B. gigas.
The long axis of the basal horn section tends to become
directly transverse, and the posterior face is somewhat
flattened; the nasals are still broad, resembling closely
those in some specimens of B. dispar. The presence
of two large upper incisors in the type skull of this
species was believed by Marsh to be its most distinc-
tive character; two upper incisors are observed also
in skulls of the same species in the Yale and American
Museums. The male skull (Princeton Mus. 11015),
however, carries but one lateral superior incisor, a fact
it never has the midparietal protuberance, and the
nasals and horns are nearer to B. rohustus than to
Brontotherium; also the internal cingula of the pre-
molars are well rounded, and the canines are pointed
at the tips.
Observations on the measurements of Brontops ro-
hustus.— The males of this species exhibit a con-
vergent resemblance to Brontotherium in the buccal
expansion of the zygomata, in the reduction of the
external cingula of the premolars, in the progressive
well-rounded premolar tetartocones of certain speci-
494
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
mens, and in the massive form of the canine. The
internal cingula of the premolars, however, are pro-
nounced and well rounded as in Brontops. The refer-
ence of these specimens to Brontops roiustus rather
than to Brontotherium is confirmed by the following
contrasting measurements :
Measurements of Brontops, Brontotherium gigas, and Bronto-
therium platyceras, in millimeters
Brontops
robustus
Brontotherium
gigas
Bronto-
therium
platy-
ceras,
Field
Mus.
12161
Car-
negie
Mus.
313
Yale
Mus.
12048
(type)
Am.
Mus.
492
Car-
negie
Mus.
341
pi-m3 . .
355
144
220
743
350
137
220
765
635
42
36
353
126
241
830
793
35
32
355
132
218
773
35
23
340
Pi-p* ... ...
120
M'-m' - - ....
223
880
Nasals to midvertex occi-
895
Canines:
47
37
36
Anteroposterior
29
In general Brontops rohustus is distinguished from
Brontotherium by the greater length of the premolar
series, by the shorter skull base and skull top, and by
the larger canines, which are also more pointed at
the tip. The horns are much shorter than in the flat-
horned species of BrontotTierium, and the connecting
crest is lower.
The range of size in skulls referred to this species is
indicated as follows:
Millimeters
SkuU, basilar length 743-813
Pi-m3 340-376
Pi-p^ 132-151
M'-m" 230-237
Zygomatic index 77- 87
Outside length of horns 130-210
The premolars and molars are wide compared with'
those of Menodus:
Brontops robustus
Menodus trigonoceras.
40X65 to 46X69
43X51 to 41X55
73X84 to 90X89
72X70 to 82X79
Geologic and geograpJiic distribution. — So far as re-
corded, all the known specimens of this species were
obtained from the Titanotherium zone (Chadron for-
mation) of South Dakota. It appears probable that
the geologic level recorded for the type specimens by
Hatcher, namely, lower C of the upper Titanotherium
zone, is characteristic.
Materials. — The type skull and skeleton in the Yale
Museum (No. 12048) afford the best knowledge of the
extreme characteristics of this type. In the American
Museum there are two fine skulls (Nos. 1083, 1069)
which confirm or supplement the characters observed
in the type. In the Princeton Museum there are two
male skulls, also found in "Corral Draw," South
Dakota (Nos. 11439, 11015), associated withmany
parts of the skeleton; also a valuable skull (No. 10061)
which is transitional between B. robustus and Allops
marshi in the disposition of the horns. In the Na-
tional Museum there is an extremely brachycephalic,
robust skull (Nat. Mus. 4253) of smaller size than the
typical B. robustus, which may represent a primitive
or transitional variety of this species. It agrees in all
its principal characters with the type skull, although
smaller and in many features more primitive.
Slcull. — The most prominent cranial characters of
this well-defined species are extreme brachycephaly,
extreme abbreviation of the occiput, marked depres-
sion of the midregion of the cranial vertex, rapid
elevation of the summit of the occiput, downward and
forward inclination of the zygomata. In female
specimens of B. dispar the length exceeds the breadth
by 200 millimeters; in the type of B. validus the length
exceeds the breadth by 100 millimeters. In this skull
also the length exceeds the breadth by only 100
millimeters. This feature stands in widest contrast
with the dolichocephaly of the contemporary Menodus
giganteus, in which the length exceeds the breadth by
over 230 millimeters. The type skull, with its
abbreviate occipital region, also contrasts with that
of the contemporary Brontotherium gigas, which shows
a decided backward extension of the occiput, especially
in the males. The age or growth characters of this
skull are similar to those in some of the collateral
ancestors of this phylum — namely, expansion and not
very marked elongation of the horns, which shift
progressively forward and reach an extreme over-
hanging position in the type. Old skulls exhibit a
thickening of the connecting crests between the horns,
also the distal growth, widening and rugosity of the
nasals, and thickening of the buccal swellings of the
zygomata. These buccal swellings are somewhat
crushed laterally in the Yale type specimen ; but in all
the skulls examined the buccal section appears to be
smaller and more concave above, and on the outer
border a less strong flaring out at the sides is observed
than in the skulls of Brontotherium medium or B.
curtum.
In general, the skull apart from its much greater
brachycephaly resembles on a grander scale that of
Allops marshi, both in its superior and inferior aspects
and in the form of the nasals, yet the retarded condi-
tion of the premolar tetartocones, the extremely wide
transverse sections of the horns, and, the relative
obtuseness of the canines do not point to direct rela-
tionship to A. marshi.
Horns. — Horns are found, however (as in Am. Mus.
1083, 1069, Princeton Mus. 10061), which are transi-
tional in position and basal section between those of
Allops marshi and of this species; progressive varieties
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
495
of B. dispar also show more or less transition to this
transversely flattened type of horn.
Correlated with the general brachycephaly, the
palate is broader and the grinding series more widely
arched. Similarly we observe great breadth in the
postglenoid and paroccipital processes. It is to be
noted that the postglenoid and post-tympanic proc-
esses unite late in life (Am. Mus. .1069), whereas in
the contemporaneous species of Brontotherium {B.
gigas and B. curtum) these processes unite early in life.
The vomerine ridge is prominent, and there is a down-
ward continuation, or septum, of the vomer tending
to unite with the palatine and thus divide the posterior
nares. This feature is observed also occasionally
in B. dispar. Rugosities for the attachment of the
recti capitis antici majores muscles are located at the
junction of the basioccipital and basisphenoid (type
skull and Am. Mus. 1069), another Brontops character.
The external bridge over the infraorbital foramen is
relatively narrow in contrast with its more primitive
width, correlated with the longer face in A. marshi.
The antorbital malar ridge is less rounded than in
B. gigas and its congeners. There is a pair of small
protuberances (k, k) on either side of the median line
of the upper portion of the occiput, which very prob-
ably were situated between the insertion areas of
the median ligamentum nuchae and of the paired
semispinalis capitis muscles. These protuberances
are also observed in A. marsJii, B. gigas, and B.
curtum. The great strength of the occipital pillars
and the broadly rugose flare of their summits contrast
with the narrow ridges observed in B. dispar.
The hyoid elements, represented in Figure 425,
include the stylohyals, 265 millimeters in length,
which are expanded superiorly and exhibit a long,
flattened shaft and a slight distal expansion terminat-
ing in cartUage; the ceratohyals, 80 millimeters in
length, expanded at both extremities. The epihyals
were probably almost vestigial, as in other peris-
sodactyls. In the horse they consist of a small
nodule of bone embedded in the cartilage that con-
nects the ceratohyal with the stylohyal.
Dentition. — The variability in the upper incisors
has already been mentioned. The formula is probably
If^. The incisors exhibit the strongly convex,
smoothly rounded crowns so characteristic of Teleodus
avus, B. hracTiycepTialus , A. marshi, and B. dispar.
The canines are quite distinctive of this species. In
the males they are large and robust, recurved, pointed,
and some specimens (Am. Mus. 1069) possess a
strong posterior cingulum, thus being readily distin-
guishable from the canines of the contemporary Meno-
dus and Brontotherium. The diastema behind the
canine of the Yale type skull is unusual, although it
is seen occasionally in the earlier Brontops hrachycepha-
lus and appears to be a somewhat distinctive specific
character, since it is also observed in the two male
skulls (Am. Mus. 1083, 1069). The premolars are
highly distinctive in their retarded stage of evolution;
p^ exhibits a rudimentary postero-internal cusp; the
tetartocones of the succeeding p^"^ are rather feebly
developed; in p* the tetartocone is a mere spur; the
external cingula are barely indicated, and the internal
cingula of the premolars are rounded and crenulate.
In the molars also the cingula are irregularly de-
veloped. The hypocone of m^ is variable in form,
either cingulate (Am. Mus. 1083) or a small, acute
cusp (Am. Mus. 1069). A reversional feature of
interest is a crenulate crest or abortive metaloph
connected with the hypocone, recalling the similar
structure in B. irachycephalus. Skull Am. Mus.
1069 exhibits on m^ a well-defined "crochet" and
" antecrochet, " and also a minute "crista." As
Figure 422. — Skull of Brontops rohusius
Top view. One-ninth natural size. Am. Mus. 1069; resembles
the type of B. robustus in its extreme brachycephaly. The
horns are shorter and thiclier at the tips. The midparietal
opening is seen also in certain skulls of B. dispar.
observed above, the grinding series tend to rise
anteriorly.
Additional details of the type sTcull. — In reference to
the characters noted above, it is important to observe
that the top of the type skull is somewhat crushed
from front to back, increasing the concavity along
the top line and the spread of the vertex. The
grinding series are in the ninth stage of wear. It is
in this skull that we observe extraordinary breadth
and relative shortness, extremely robust zygomata,
horns extremely far forward and oval in transverse
section, well overhanging the abbreviated nasals, and
the arching character of the opposite grinding series.
The chief measurements are as follows: Top of
occiput to tip of nasals 635 millimeters, transverse
496
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
A, BTontops dispar, Nat. Mus. 4941
(type) ; depth of horizontal ramus
greatly diminished by vertical
crushing, but the symphyseal re-
gion was probably shallow. A
short diastema between pi (which
has dropped out) and the canine,
which is stout and conical. In-
cisors subspherical, external cin-
gula not so sharply defined as in
Menodus.
B, B. Tobustus, Princeton Mus.
10061; symphyseal region crushed
laterally, ramus with long,
straight lower border, angle pro-
duced. Two well-developed in-
cisors, canine stout, diastema in
front of pi (which has dropped
out), external cingula not sharply
defined.
C, B. Tobustus, Yale Mus. 12048
(type); ramus long and shallow,
symphyseal region very shallow,
angle very large and produced
downward, coronoid high, verti-
cal truncate. Incisors stout, ca-
nines very stout and short,
diastema in front of pi, external
cingula reduced, crowns of pre-
molars and molars less hyp.^odont
than in Menodus.
All one-fifth natural size.
Figure 423. — Lower jaws of Broniops dispar and B. robustus
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
497
width of zygomata 667, transverse width of occipital
condyles 225, condyles to tip of incisors 765, molar-
premolar series 350, true molars 220, p'~* 137; anterior
enamel of canines, worn, 42; anterior diameter of
canines, 36.
Teeth (Pis. XX, C, CI). — The canines are relatively
_.,...._ robust, obtuse, and re-
curved, with narrow
posterior cingula. The
incisors have smooth,
rounded crowns of equal
size, two on each side.
The diastema behind the
canine is 20 millimeters
and forms a very charac-
teristic feature of the
type. The premolar cin-
gula are nearly obsolete
except on the inner
sides of p^~*; a cingu-
lum is faintly shown
on the outer side of
p^. The tetartocones
are feebly developed
throughout, espe-
cially on p*, in which
the deuterocone con-
stitutes the chief
inner portion of the
crown with a ridge-
like tetartocone ex-
tending posteriorly.
Lower jaw (PI. CII
and fig. 423).— The
j aw is very deep from
the condyles to the
bottom of the angle;
there is a robust,
depressed angle,
prominent, tall, rectangular coronoid process;
the mental foramen is very large; the symphy-
sis is long (20 mm.). There is a single lower
incisor on either side of the lower jaw, with the
vestigial median alveolus of a second incisor on the
right side. There is a deep median cleft in the
incisive border. The inferior canines are heavy and
laterally compressed (transverse diameter 25 mm.,
anteroposterior diameter 32). The anterior pre-
molars, pi, are small but bifanged on either side
and are ready to drop out; there is a single alveolus
on the left side. The external cingulum is reduced,
being marked only upon the valleys and anterior
face of the premolar-molar series, with the exception of
m2_3, in which the external cingulum is somewhat
more developed. In all titanotheres the inner side
of the molar crowns is entirely devoid of a cingulum.
The hypoconulid of ms is not very broad but has a
FiGUBE 424. — Sections and con-
tours of skull of Brontops sp.
Am. Mus. 518 (for mounted skeleton, see PI.
XXXV). The horn sections and zygomata
suggest those of Brontops robustus. One-
eighth natural size.
crenulate internal crest, which is entirely absent in
Menodus.
Transitional skull to B. rohustus. — In the Princeton
Museum is a fine skull (No. 10061, PI. XCV) which
in many respects is transitional to B. rohustus. It
belongs to a large male and was used in the first
published restoration of "Menodus" coloradensis bj
Scott and Osborn (1887.1). Although of smaller size
it agrees with B. rohustus in the presence of a pair of
smooth, rounded upper incisors, decided diastemata
behind the canines, and in the general conforma-
tion of the zygomata, retarded development of
the tetartocones on the premolars, absence of ex-
ternal cingula, and large upper and lower canines.
It is more primitive than B. rohustus in the presence
of two rounded lower incisors, in the more rounded
summits of the horns, and in the somewhat less ex-
panded zj^gomatic and nasal sections. The angle of
the jaw is produced backward rather than downward.
Figure 425. — Hyold bones of Broniops sp. (A) (Am. Mus. 518)
compared with those of the tapir (B), black rhinoceros (C), and
horse (D)
Oblique medial aspect of right side. sJi, Stylohyal; ch, ceratohyal; bh, basihyal; ft,
thyrohyal; glh, glossohyal. Compare hyoid bones of Golichorhinus longiceps (fig. 345).
The glossohyal is lacking in Dolichorhinus and possibly in all other titanotheres.
Ohservations on the sTcull of the mounted skeleton
referred to Brontops rohustus (Am. Mus. 518). — The
generic and specific determination of this skeleton is
difficult owing to the imperfection of the skull.
It was originally referred by Osborn to Titanothe-
rium rohustum, but later he referred it to Brontotherium
498
TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
gigas, 9 , chiefly because the carpus of this specimen is
slightly different from that of Marsh's type of Brontops
robustus. Unfortunately the specimen lacks all the
front teeth, as far back as p' and p2. The upper pre-
molars have the tetartocones well constricted, as in
Brontops, and very different from the circular tetar-
tocones of male brontotheres. A supposed female of
Brontotherium gigas (Am. Mus. 1006) also has the
resemblance to the jaws of brontotheres; from Bron-
tops robustus type it differs to some extent; perhaps
its nearest resemblance is to the type of Diploclonus
tyleri. The available measurements of the skull and
dentition, though few, are nearer to those of Brontops
robustus (especially Am. Mus. 1069) than to those of
large male brontotheres. The specimen is much
larger than the supposed female brontotheres of B.
B
A C
Figure 426. — Sections and contours of skulls of Diploclonus bicornulus and D. tyleri
A, Diploclonus hicornuius. Am, Mus. 1476 (type); horns much as in Brontops dispar but with an accessory hornlet, basal section roundly trihedral, nasals
long, zygomata little expanded. B, Am. Mus. 1081, referred to i>. tyleri: stout horns roundly trihedral in section, connecting crest high, zygomata
stout. C, D. tyleri, Amherst Mus. 327 (type); horns widely oval in section, accessory hornlets more pronounced than in hicornutus, zygomata expanded.
One-seventh natural size. Sections and contours of D. bicornuius and D. selwynianus (see fig. 185) indicate that these forms, with their peculiar narrow
nasals, are not ancestral in type to D. ampins, with its broad and abbreviate nasals.
tetartocones much restricted, but the reference of this
specimen to Brontotherium is very doubtful.
In Am. Mus. 518 the external cingulum of the upper
premolars and molars is absent, as in both Brontops
robustus and Brontotherium, so this character is not
decisive. The external cingulum of the lower pre-
molars is reduced, but on the whole the dentition ap-
pears closer to that of B. robustus than to that of
Brontotherium. The lower jaw presents no close
curtum and B. gigas. The sections and contours of
the horns and nasals are certainly different from those
of the supposed female brontotheres and still more
so from those of male brontotheres. The sections are,
in fact, closer to those of Diploclonus tyleri and Bron-
tops robustus (especially Am. Mus. 1083). The
manus has the magnum broader, more angulate than
that of B. robustus as figured by Marsh, but as a whole
the manus is similar in proportions to that of Brontops
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
499
rohustus. The hind limbs mounted with this skeleton
belong to other individuals.
On the whole the evidence indicates that the
mounted skeleton is not a female brontothere but is a
member of the Brontopinae and probably Brontops
rohustus.
Diploclonus Marsh
{"Allops" Marsh, Osborn, 1902)
Plates XXXVII, CIV-CX; text figures 185,
187, 196, 200, 201, 375, 389, 391, 394, 397,
409, 426-428, 630, 639
[For original description and type references see p. 227.
For skeletal characters see p. 675]
Generic characters. — A phylum col-
lateral to other Brontops phyla, appar-
ently distinguished by the budding off
of a second hornlet on the inner side of
the horns and hence loiown as Diplo-
clonus; also by the narrow and elon-
gate nasals in the middle members of
the series; hornlets rising from the
inner faces of the horns.
Incisors 2—1. Premolars about as
in B. dispar, but with external cingula
becoming obsolete. Nasals progres-
sively abbreviating. Bridge over in-
fraorbital foramen rather sharp and
narrow. Progressive brachycephaly ;
also backward prolongation of the
temporal fossae. Other characters as
in Brontops.
A tendency to give off an internal
branch, knob, or "hornlet" (fig. 427)
on the inner and anterior faces of the
main horns is observed in a number
of titanothere skulls which otherwise
show strong affinities to the Brontops
phylum and especially to the species
B. dispar. Marsh regarded this inter-
nal hornlet as of so much importance
that he assigned it generic rank in
proposing the genus Diploclonus, the
genotype of which is the species Diplo-
clonus amplus. The presence of a small
accessory horn swelling has also been
observed in certain skulls of Menodus
and of Brontotherium. It is doubtful,
therefore, whether this hornlet is of
generic value. D. amplus and D. tyleri
resemble B. rohustus in many features but may
represent a distinct subphylum. Subsequently the
same character was observed by Osborn in two skulls
in the American Museum (Nos. 1476, 1081) which
were made the type and paratype respectively of the
species Diploclonus {" Megacerops") hicornutus (fig.
a fine skeleton with skull was obtained by the Amherst
Museum and described by Lull as a third species Dip-
loclonus {" Megacerops") tyleri (Amherst Mus. 327).
The sum of progressive characters is as follows:
(1) Hornlets on the inner sides of the horns; (2) in-
creasing brachycephaly; (3) gradual or retarded
evolution of premolars, p* being the most retarded
tooth. The sum of retrogressive characters is as
A B
Figure 427. — Sections and contours of skull of Diploclonus amplus
A, Nat. Mus. 4710. In this supposed female the horns are relatively short, the connecting crest well devel-
oped, horn section trihedral, nasals broad, zygomata slender, B, Yale Mus. 12015a (type); a male, with
horns relatively long, directed forward and placed far in front of the orhits, basal horn sections trihedral,
connecting crest high, zygomata well expanded, nasals small and narrow. One seventh natural size.
follows: (1) Relative abbreviation of the premolar-
molar series, the index in D. hicornutus being 45, which
is less than that in D. amplus; (2) apparent reduction
or loss of median incisors.
A doubtful genus and phylum. — There is no question
whatever of the affinities of these animals to Brontops
426), the latter now referred to D. tyleri; and finally j and of their substantial proximity to this genus.
500
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
From the materials at hand, which are hmited to four
or five skulls at present, it is doubtful whether they
should be separated as a distinct phylum of generic
value. The question turns on whether the hornlet
is a sport character or a permanently progressive
character, which can be settled only by more extensive
material than is now available. On the whole, it
seems to be wisest at present to regard these species
as a subphylum collateral with the- main phyla of
B. dispar and B. robustus.
Affinities with Brontops dispar and Allops marsTii. — •
Among the chief resemblances to Brontops dispar are
(1) the marked brachycephaly, especially in the widely
expanding buccal processes of the zygomata, empha-
sized most distinctly in D. tyleri and D. amplus; (2) the
marked backward prolongation of the occiput behind
the widest portion of the zygomatic arches, which dis-
tinguishes these skulls at once from those of Allops
marshi and B. robustus and relates them to B. dispar;
(3) the short horns with rounded summits which par-
take of the general characters of those of B. dispar,
although a tendency to broaden and flatten becomes
marked in D. tyleri and extreme in D. amplus. The
nasals show progressive abbreviation: they are elon-
gate in D. hicornutus, more abbreviate in D. tyleri, and
extremely abbreviate in D. amplus. Features of
difference from B. dispar are not only the internal
hornlets upon the horns but the somewhat more
marked development of the connecting crest between
the horns.
Unfortunately the records as to geologic distribution
are very incomplete. The geologic level of the least
progressive stage, D. hicornutus, is not recorded. The
level of the intermediate stage, D. tyleri, is recorded
as 35 feet above the Pierre shale, 200 feet below the
summit of the Titanotherium zone; this would place
this animal in lower B or even in A. The level of the
most progressive species, D. amplus, is not recorded.
Distinctive characters of the species of Diploclonus. —
This is not a monophyletic series, like Menodus or
Brontotherium; it is diphyletic. The skulls at once
seem to be distinguished from those of Allops marshi
and Brontops rohustus by the great backward extension
of the occipital region behind the zygomata. The pit
in the center of the vertex relates them to other species
of Brontops. The progressive broadening of the
zygomata closely parallels that of the other collateral
phyla.
D. hicornutus (Osborn) is the most primitive form, with
relatively long and narrow nasals, rounded and more erect
horns, least expansion of the zygomatic arches, and least
depression of the angular border of the jaw.
jD. selwynianus (Cope) is a little-known animal from the Cy-
press Hills, Saskatchewan, represented only by the nasal bones,
which are intermediate in size between those of D. hicornutus
and D. tyleri; they resemble the nasals of D. hicornutus more
closely in their narrow, elongate, and laterally decurved form.
D. tyleri (Lull) is an intermediate stage of evolution, char-
acterized by relative abbreviation of the nasals, more anterior
position of the horns, greater prominence of the hornlets, more
widely arched zygomata, depression of the lower angular region
of the jaw. This animal is a more pronounced development of
the D. hicornutus type, the internal hornlets being larger, the
horns much broader and more decidedly projecting forward.
The geologic level of this animal is said to be 50 feet above the
Pierre shale. It is probably from the middle levels, B.
D. amplus Marsh is distinguished by very short and obtuse
nasals and a prominent internal hornlet on the horns, horns
broadly divergent and compressed anteroposteriorly, excessively
wide zygomatic arches, stout, recurved canines. The features
of this progression as seen in the superior view of the skulls are
displayed in Plate CVIII, A, and in Figure 391. This robust
animal shows an extreme development of the characteristics of
this subphylum. Massive skuH, very broad zygomatic arches,
very short horns, with a decided development of the internal
hornlets. Probably belongs on the levels of Chadron of C.
Standard measurements
in the Diploclonus phylum, in millimeters
Upper teeth
Skull
Jaw
CM
fM
i
o
II
i
a
Ph
675
683
s
i
1
615
468
"■680
a.
it
o
91
68
8
|l
il
665
15
'"g
Is
K
S
■2„
o
II
II
|i
1
D. amplus, Yale Mus. 12015a, cf (tj-pe).
D. amplus, Nat. Mus. 4710, 9
D. tyleri, Amherst Mus. 327, & (type) __
?302
307
363
133
119
136
207
205
227
40
»40
45
24
22
35
'■119
?140
115
130
104
98
?84
93
170
375
140
235
42
31
685
355
340
-11
127
138
-6
222
206
0
710
680
0
608
86
-690
635
192
"183
D. bicornutus, Am. Mus. 1476, &
(type)
232
»35
22
500
Percentage of change from Diploclonus
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES 501
Measurements of slculls and jaws associated with or referred to the Diploclonus phylum, in millimeters
Skull and upper teeth
Lower jaw an/i teeth
pi-m3
M'-m'
Symphysis
to glenoid
Posterior
canine to
hyloid of
ma
Pi-m3
Mi-m3
Symphysis
to condyle
Depth
below ms
D tyleri, Amherst Mus. 327 (type)
363
340
227
206
375
-343
235
232
685
500
500
»325
154
Diploclonus bicornutus (Osborn)
{"Megacerops" bicornutus Osborn, 1902)
Plates CIV, CVII; text figures 196, 389, 397, 409, 426, 428
[For original description and type references see p. 231]
Geologic horizon. — Undetermined, probably middle
beds (Chadron B).
Specific characters. — Skull index indeterminable from
type; total length, 680 millimeters; length of grinding
series, 340; index, 50. Nasals long and narrow (104
by 86 mm.). Horns subcircular (type) to transverse-
oval (Am. Mus. 1081, c? ) in basal section; with anterior
and internal hornlets; connecting crest moderately
developed; horns placed anteriorly to orbits. Preor-
bital bridge of malar with a sharp edge. Occiput
greatly produced backward.
This is the least specialized stage, one which is
closest to Brontops dispar except in the narrow form
of the nasals.
Observations on the measurements of Diploclonus
bicornutus. — Diploclonus is a less conservative and
less clearly distinguished phylum than Brontops,
Menodus, or Brontotherium. Aside from the presence
of a small secondary horn swelling, a character seen
also in certain other skulls which are not referred to
Diploclonus, the type of D. bicornutus presents a
mingling of the characters of Allops marshi, Brontops
dispar, and even in some features of Menodus. In
comparison with these species the measurements
of the type skull are given below:
Measurements of Diploclonus bicornutus, Brontops dispar, Allops marshi, and Menodus proutii, in millimeters
pi-m'
pi-p<
M>-m3 .
Pmx to condyles-
Zygomatic index.
Nasal length
Nasal breadth
Horn length
D. bicornutus,
Am. Mus.
U76, c? (type)
340
138
206
680
104
183
Nat. Mus.
4248, cT
340
140
207
685
78
94
102
188
A. marshi,
Am. Mus.
1445
(paratype)
335
135
203
675
69
105
100
113
Carnegie
Mus.
3063, 9
335
135
203
628
120
125
150
The table shows that these specimens, which are
referred to four genera, agree closely in dental meas-
urements but differ in the proportions of their nasals
and horns.
Characters of the type. — The type of D. bicornutus
was long supposed to be female, but the canines and
zygomata are stouter than in females of other species.
The specific characters above set forth are based upon
the type skull (Am. Mus. 1476). Associated with
this type in the original description by Osborn as a
paratype was the skull Am. Mus. 1081 (PI. CV),
which is now regarded as more nearly related to an-
other species, D. tyleri.
The following description is thus based solely
upon the type of D. bicornutus. The sex of this
specimen is somewhat uncertain, but the antero-
posterior diameter of the canine (506 mm.) indicates
that it was a male.
The lateral compression of the type skull and jaws
(Am. Mus. 1476) prevents the determination of the
proportions of the skull or of the zygomatic index.
The wearing of the grinders indicates that the specimen
represents the eighth stage of growth. Thus we
conclude that all the progressive characters are very
distinctly developed and indicated.
Skull. — The relatively long, narrow proportions of
the skull as seen from above (PI. CIV) are due partly
to lateral crushing. Even with allowance for some
deformation the skull is not very broad. As seen
in side view (PL CIV) it combines the long nasals of
A. marshi with the vertically elongate horns and
backward expanded occiput of D. dispar. An
anterior hornlet is plainly visible. The tuberosi-
ties on the inner sides of the horns are only faintly
developed, and were it not for comparison with the
specimens of D. tyleri they would hardly be noticed.
502
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Similar anterior tuberosities, or hornlets, are observed
in certain male skulls of M. giganteus (Am. Mus.
505). As seen ftom above (fig. 444) the parietal
vertex is moderately broad. A very distinctive
feature is a sharp preorbital malar bridge. There is
considerable expansion behind the orbits, as in B.
dispar; as in all aged individuals, the supratemporal
crests are sharp and overhanging. The summits of
the occipital pillars are rugose and greatly expanded.
In lateral view (PI. CIV) the postglenoid process is
seen to be relatively narrow and exhibits a short
surface of contact with the post-tympanic. On the
palatal surface of the skull the vomer apparently does
not extend back to overlap the parasphenoid. There
is only a slight basisphenoid rugosity.
This skull shows group resemblances to Menodus
trigonoceras as follows: (1) Nasals long, set very high;
(2) orbit very large; (3) tooth row very long (dental
index 50, 46-51 in M. giganteus); (4) zygomata not
expanded and in side view pitching sharply downward
and forward; (5) lower jaw resembling that of Menodus
in the well-developed chin and depressed angle.
Dentition. — The incisor alveoli are not well defined.
The canines are much worn; they exhibit crowns of
rounded form. The absence of external cingula on
the premolars may be due in part to the extreme wear.
The internal cingula are well developed. The tetarto-
cone developments on the premolars are approxi-
mately of the same stage as in B. dispar. In m^ the
hypocone is an elevated part of the cingulum.
Summary. — This skull may be that of an old male,
affording an example of a progressive oft'shoot of one
of the Brontops lower beds phyla.
Diploclonus selwynianus (Cope)
{Menodus s^elwynianus Cope, 1889; " Megacerops?" selwynianus
Osborn, 1902)
Text figure 185
[For original description and type references see p. 225]
Type locality and geologic horizon. — Swift Current
Creek, Saskatchewan; Titanotherium zone.
Specific characters. — Nasals long (free length about
115 mm.), narrow (free width 75 mm.), abruptly
rounded in front and sharply decurved at the sides,
with a deep longitudinal inferior cavity.
The chief ground for referring the uniquely narrow
nasals of the type of this species to Diploclonus is the
resemblance they bear to those of Diploclonus bicornu-
tus, which is closer than to that of any other form.
The resemblances in the nasal bones consist in (1) the
narrowness, (2) the deep longitudinal inferior cavity,
(3) the decurved sides, and (4) the massive and de-
curved tips.
Cope's description of this specimen from the Swift
Current Creek region is cited in full elsewhere (p. 226).
To this description may be added the following ob-
servations. The inferior view (fig. 185) exhibits the
paired cavities, or frontal antrum, at the junction of
the nasals, frontals, and maxillaries; the suture be-
tween the nasal bones has entirely disappeared.
These nasals are very characteristic and quite dif-
ferent from those of Menodus or Megacerops colo-
radensis. Lambe describes the type as follows
(1908.1, p. 47):
The coossified nasal bones of one individual constitute the
type of this species. They are long and narrow, abruptly
rounded in front, and bent downward at the sides. The lower
surface is deeply excavated in a longitudinal direction.
Diploclonus tyleri (Lull)
{Megacerops tyleri Lull, 1905)
Plates XXXVII, CV-CVIII; text figures 200, 201, 397, 409,
426, 428, 630, 639
[For original description and type references see p. 234. For slteletal characters
see p. 675]
Geologic horizon. — Recorded as 35 feet above the
Pierre Cretaceous, or 165 feet below the summit of the
Titanotherium zone. Regarded by Lull as in the upper
part of the lower Titanotherium zone, but more
probably belonging in the level of the middle beds.
Specific characters. — Brachycephalic. Zygomata
widely arching. Nasals abbreviated or broad; free
length, 140 millimeters. Grinding series, 363. Horns
shifted forward; distinct hornlets on the inner sur-
faces; horn section at the base trihedral to trans-
versely oval, with sharp outer angle. Canines stout,
recurved. Superior incisors 2-1.
General characters. — This animal is certainly a
member of a collateral branch of the B. dispar series,
agreeing with the aged specunens of the typical
B. dispar even more closely than the type of D.
bicornutus. The species appears to represent a stage
of phyletic evolution intermediate between that of
the more primitive D. bicornutus and the more pro-
gressive D. amplus. This conclusion, however, awaits
confirmation by additional evidence.
Materials. — The type of this specific stage or
ascending mutation is the fine skeleton and anterior
portion of the skull in the Amherst collection (No. 327) .
Apparently belongmg to the same stage is the finely
preserved skull and jaws (Am. Mus. 1081) originally
described by Osborn as a paratype of D. bicornutus
but presumably representing a more recent or pro-
gressive stage of evolution.
Observations on the measurements of Diploclonus
tyleri. — The two skulls referred to Diploclonus tyleri
differ from the type of D. bicornutus in having larger
molars and more massive horns. They differ from
the type of D. amplus in having larger molars. The
type skull and jaw in side and top views approaches
that of Brontops robustus, except that it has a well-
developed pair of accessory horn swellings. The
characters of the incisors, canines, postcanine diaste-
mata, and premolars further suggest affinity with
B. robustus; but the anteroposterior measurements
of the dentition differ from those of Brontops and agree
with those of Menodus trigonoceras, as shown below:
EVOLUTION OF THE SKULL AND DENTITION OP OLIGOCENE TITANOTHERES
503
Measurements oi Di-ploclonus tyleri and Menodus irigonoceras,
in millimeters
D. tyleri
M. trigonoce-
Amherst Mus.
327, (f (type)
Am. Mus.
1081, cf
ras, Nat.
Mus. 4291, J
Pi-m5
363
355
360
Pi-p' :
136
127
136
Mi-ms
227
222
224
680
770
!
This agreement with Menodus irigonoceras, taken
in connection with the relatively slender manus and
pes of the type of D. tyleri as compared with B.
roiustus, constitutes one of the many facts which
suggest the possibility of occasional hybridization
among the genera Bron-
tops, Menodus, and AUops
(W. K. Gregory).
Description of the
type. — Lull's specific defi-
nition may be cited in full
(Lull, 1905.1, p. 445):
Horns well in front of orbits,
directed somewhat forward
and outward, an elongate oval
in basal section with the long
axes in line, rounded oval at the summit. Hornlets quite
conspicuous, on the inner face of the horns midway between
tlie base and summit. Connecting crest low and inconspicuous.
Nasals broad, well rounded in front, and but slightly arched
beneath. Zygomata expanded and deep, with a well-rounded
outer face. Dentition: Superior incisors represented by the
deep and well-defined median alveoli and by the lateral teeth,
which remain in place and which have hemispherical crowns
which show little sign of wear. The canines are
lanceolate, with a well-developed postero-internal
cingulum. There is a short diastema in front of,
and a longer one behind, the canine. Premolars
with a smooth internal cingulum, less pronounced
in the middle of the tooth and with no external
cingulum. The deuterocone is well developed,
while the tetartocone, especially of premolar 4,
is inconspicuous.
The jaw is deep and robust, with the alveoli
of two incisors, probably of the second and third,
deep and distinct. There is no space between
the lateral incisors and the canine, though be-
tween the two median alveoli a considerable gap
occurs. There seems to have been a small
diastema behind the lower canines, which are
lanceolate, though with a less prominent cingulum and not so
strongly recurved as the upper ones.
The same author observes that the creature most
nearly resembles D. bicornutus (Osborn) and D. amplus
Marsh, having certain characters suggestive of each;
but there are enough important differences to render
it distinct and to warrant the erection of a new species
for its reception. The dental formula is If, C\, P|,
M|; the two median superior and all of the lower
incisors are represented by deep inclosed alveoli, as
101959— 29— VOL 1 35
if the teeth had dropped out after death. Toward
the base the horns are oval to triquetrous in section;
toward the summit they become rounder and rough-
ened at the extremities. "It would seem," observes
Lull, "from the similarity of the roughened patches
to those on the rhinoceros nasals, as though the entire
prominence had been clothed with skin, with two
rhinoceros-like horns, a larger one at the apex and
a smaller one on the summit of the hornlet."
Figure 428. — Lower jaws of Diploclonus bicornutus and D. tyleri
, Diploclonus bicornutus, Am. Mus. 1476 (type); animal very old and hence the angle is broad and
prominent (compare the aged type of M. torvus, fig. 437, A). Canines stout and conical, external
cingula not sharply separated from the eotoloph, B, D. tyleri, Amherst Mus. 327 (type); broadly
resembles S. robusius. A diastema in front of pi. One-fifth natural size.
Characters of referred specimen (Am. Mus. 1081). —
This specimen was employed by Osborn as a paratype
of D. hicornutus. Skull broad, index 85 ; tooth row elon-
gate (367 mm.) ; index 50. Condyles to incisive border
710 millimeters. Horns anteriorly placed and directed
obliquely forward. It is important to note that the
plaster restoration of the nasals (Pis. CV, CVII) and
of the anterior part of the zygomata prevents a true
determination of the characters of these parts of the
skull. This specimen is also of advanced age and
504
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
shows many senescent characters. The teeth are
extremely worn and reveal nothing decisive; the
cingula of the premolars are smooth and worn down.
The horn tips are extremely rugose, and the internal
hornlets are well marked (fig. 426, B). The occipital
vertex is rugose, with deep paired indentations and
knobs. The zygomata are greatly swollen trans-
versely. As in many aged specimens of Brontops,
a single large incisor is preserved in either premaxUla;
the superior incisive formula was thus 2-1.
Diploclonus amplus Marsh, 1890
{"Allops" amplus, Osborn, 1902)
Plates CVIII-CX; text figures 187, 376, 389, 391, 394, 409,
427
[For original description and type references see p. 227]
Geologic Jiorizon. — Titanotherium zone, level not
recorded.
Specific characters. — Skull extremely brachycephalic,
zygomatic-basilar index 91. Nasals greatly abbre-
viated and narrow, horns broadly oval in basal section,
directed upward and outward, with conspicuous
internal hornlets. Grinding series relatively reduced,
length 302 millimeters, index 45. Upward flexure of
premolar series anteriorly pronounced. Zygomata
extremely broad with convex buccal expansions.
General characters. — The nasofrontal section of the
type skull first suggested comparison with a form
transitional between Brontotheriwn gigas and B.
curtum, but numerous other characters forbid such a
phyletic reference, especially the long-pointed canines,
40 millimeters (the type was a male), the presence of
only one superior incisor on each side, the retarded
development of the tetartocones of the premolars,
the rounded tips of the horns. All these characters
remove the animal from relationship with Bronto-
therium, Menodus, or Megacerops and indicate its
affinity to Brontops; this is confirmed by more
searching study of the character of this type, but
especially by the existence of the more primitive and
transitional forms D. hicornutus and D. tyleri above
described.
Diploclonus amplus is by far the most progressive
species in this phylum, as demonstrated by (1) the
extreme abbreviation of its nasals, (2) the strong
development of the connecting crests between the
horns, (3) the greater prominence of the internal
hornlets, and (4) the transverse-oval expansion of the
basal horn section, which has now reached an extreme
stage, parallel to that observed in Brontops rohustus.
This species, however, is readily distinguished from
Brontops rohustus, in common with other members of
the phylum to which it belongs, by the marked
backward elongation of the occipital region behind
the zygomatic arches.
This specimen tends to confirm the hypothesis that
the species grouped under Diploclonus formed one or
more subphyla parallel with Brontops.
Materials. — The type skull of D. amplus in the
Yale Museum (No. 12015a) is undoubtedly a male.
A supposed female skull is found in Nat. Mus. 4710.
The type is the only specimen which we have yet
seen that certainly belongs to this species..
Observations on the measurements of Diploclonus
amplus. — The type and only known male skull of this
species is very brachycephalic, although crushing may
have contributed to the extremely high zygomatic
index (91).
The male skull approaches the paratype of Allops
marshi in the anteroposterior measurements of the
dentition, but in its general conformation it rather
suggests B. rohustus. The nasals are reduced in size.
Measurements of Diploclonus amplus and Allops marshi, in
millimeters
Pi-m3
Pi-p'
M'-m3
Pmx to condyles
Zygomatic index
D. amplus,
Yale Mus.
12016a (type)
(?)302
133
207
675
(?)91
A. marslii.
Am. Mus.
1445 (para-
type)
335
135
203
675
69
Detailed description of the type. — As seen from above
the skull is broad and short in proportion, the index,
91, expressing its marked brachycephaly. We are
struck by the strong divergence of the horns, a feature
which is probably intensified by vertical crushing.
They are slightly convex on the anterior surface, with
very prominent external ridges. They are united by a
very deep connecting crest, as exhibited in the section.
The prominent tuberosity or hornlet employed by
Marsh as a generic character is 14 centimeters below
the tip of the horns; it is especially well developed on
the left side, measuring 37 millimeters anteroposteriorly
and 35 transversely. A very distinctive feature of the
horn is the sharpness of the preorbital crest. As seen
from above (PI. CVIII, A) the zygomata project
widely outward. A slightly distorted section of the
buccal processes is shown in the diagram. The upper
surfaces are somewhat flattened. In the middle of the
vertex is a shght tuberosity which probably indicates
the vestige of the pit so characteristic of members of
the Brontops phylum. The occiput is deeply exca-
vated with stout lateral pillars and rugose summits.
It is considerably produced backward behind the zygo-
mata. In inferior aspect the posterior nares are seen
to open just opposite the third molar. The pterygoid
wings of the aHsphenoid are short, rugose, flaring
strongly outward; the presphenoid is keeled by the
EVOLUTION OF THE SKULL AND DENTITION OP OLIGOCENE TITANOTHERES
505
v^omer, and there is a very prominent rugosity at the
junction of the basisphenoid and basioccipital, as seen
m Brontops and Menodus.
Dentition. — Since the animal is of advanced age, we
find just within the canine a large alveolus which cer-
tainly contained a large incisor tooth. In early life
there were probably two incisors, separated in the
median line by a diastema. The canines are long
(40 mm.) and pointed, with a postero-internal
cingulum. The first premolar is a relatively stout,
persistent, bifanged tooth, thrust closely against the
canine, partly by crushing; the tetartocone of p^ occu-
pies a very small part of the inner face of the crown, as
in B. dispar. Illustrating the mechanical ineffective-
ness of the internal cones of the molar teeth is the fact
that whUe the ectolophs are worn out of proportion in
m'~^, the internal cones of m^, m^ are still unaffected
by wear. The hypocone of m' is quite prominent and
still connected with the cingulum.
The abbreviation of the premolar-molar series is a
retrogressive character which this phylum apparently
shares with the Brontotherium phylum. The molar
series, measuring 302 millimeters, is actually shorter
(index 45) than that in the much less specialized
form D. hicornutus, in which the teeth measure 340
millimeters with an index of 50. Similarly, in the
great skulls of Brontotherium the tooth row is actually
shorter in the highly specialized B. platyceras than in
the ancestral species B. gigas.
Supposed jemale skull (Nat. Mus. 4710). — The
result of searching comparison and measurements is
the reference of this skull as a female of Diplodonus
amplus. The morphologic difference is about as great
as between the type of D. hicornutus and that of D.
tyleri.
The affinities to Brontops in this female skuU are
shown in the foUowing points: (1) Small lateral in-
cisors persistent; (2) canines of rounded form; (3)
premolars with retarded tetartocones, rounded in-
ternal and fainter external cingula; (4) horns of small
size, with long, flat external face and rounded top
section; (5) orbits with broad postorbital processes;
(6) a broad zygomatic shelf, and downward extension
of the occiput as in B. dispar; (7) rugosity on the
basisphenoid, with the vomerine bridge carried well
back.
The special resemblances of this supposed female to
D. amplus are (1) 'the general similarity of the horns
in respect to their position, basal sections, and con-
necting crests; (2) the smaller zygomatic section; (3)
the sharp preorbital malar bridge.
The chief points of difference between this specimen
(Nat. Mus. 4710) and the Yale type of D. amplus,
which is a male, are (1) absence of hornlets, perhaps
a sexual distinction; (2) greater width, flatness, and
slenderness of the nasals, which may be attributed
in part to the nondevelopment and lack of forward
advancement of the horns; (3) smaller buccal processes
of the zygomata, which, however, in section suggest
those of D. amplus on the inner and inferior faces
especially. (See fig. 427.)
The reference of this skull to D. amplus is therefore
still provisional.
Subfamily MENODONTINAE
Titanotheres chiefly of lower Oligocene age, reaching
their climax in the upper levels of the Titanotherium
zone (Chadron formation). Distinguished by narrow
heads (mesaticephalic) diverging into phyla with long
heads (dolichocephalic) and broad heads (brachyce-
phalic). Horns short, slightly shifting forward, broadly
trihedral in basal section and widely divergent at
the summits. Nasals typically elongate, broad and
square distally, secondarily abbreviate. Incisor teeth
reduced or vestigial. Grinding teeth with prominent
cingula. Premolars with accelerated tetartocones.
This is the second branch of the short-horned ti-
tanotheres, which in many respects is closely related
to the first branch, the Brontopinae; in others it
appears to present original characteristics of its own.
It is typified by the genus Menodus, also known as
Titanotherium, and includes the related genus Allops.
Between the two are transitional forms of subgeneric
rank described as Menops and Anisacodon by Marsh.
The resemblances of the original species of these two
main phyla of Menodus and Allops, namely, Menodus
heloceras and Allops walcotti, both occurring in the
very base of the Titanotherium zone, are relatively
close to Brontops hrachycephalus. Yet it would appear
that the Menodontinae diverged from the Brontopinae
before the end of Eocene time. While the Menodus
phylum maintains its mesaticephalic and dolicho-
cephalic form throughout, members of the Allops
phylum become brachycephalic and converge toward
members of the Brontops phylum. Thus Allops crassi-
cornis resembles Brontops rohustus in its proportions.
The ancestors of Menodus and of Allops, although
not readily distinguishable in the base of the Titano-
therium zone (Chadron A) , indicate that the divergence
took place in late Eocene time, when the following
main distinctions were established:
Menodus phylum
Primitively mesaticephalic, progressively dolichocephalic.
Incisors §;§, extremely vestigial, buried beneath the gums.
Canines very prominent, rounded, or compressed transversely.
Nasals typically broad and elongate.
Skeleton tall.
Feet stilted, of mediportal type.
Allops phylum
Primitively mesaticephalic, progressively brachycephalic.
Incisors reduced,
with rounded summits.
Canines prominent, compressed anteroposteriorly, expanded
transversely.
Nasals progressively abbreviating.
Skeleton little known.
Feet broader.
506
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
SYSTEMATIC DESCRIPTIONS OF GENERA AND SPECIES
IN THE ALLOPS PHYLUM
AUops Marsh
Plates XX, XXI, XXXVIII, XC, CXI-CXXII, CXXXII;
text figures 184, 189, 197, 207, 375, 378, 381, 387, 389, 391,
393, 394, 397, 399, 409, 413, 429-434, 608, 612, 615, 712
[For oi-iginil description and type references see p. 224. For ske:etal characters sec
p. 678]
Generic characters. — Mesaticephalic, progressively
brachycephalic. Lateral pair of superior incisors
persistent; canines compressed anteroposteriorlj^, flat-
tened on posterior face; grinding teeth with moder-
ately distinct cingula; premolars simpler than in
Menodus but with progressive tetartocones; molars
with elevated and pointed cusps, the transverse
slightly exceeding the anteroposterior diameters.
Nasals broad and intermediate in length, abbreviat-
ing in progressively brachycephalic types; horns short,
broadly trihedral in section, directed obliquely out-
ward.
This genus was estabhshed in 1887 by Marsh, who
selected the species A. serotinus as the genotype. In
1891 this author strengthened the genus by adding the
species AUops crassicornis, a more advanced form.
In 1902 Osborn described the species Megacerops
marshi, named in honor of Professor Marsh, which
was later referred to the AUops phylum. Finally, in
1917, Osborn added the most primitive species AUops
walcotti, named in honor of the former Director of
the United States Geological Survey. Although these
species are grouped within a single genus, they ap-
parently do not constitute a direct phyletic succession.
There are also puzzling affinities to Menops varians,
the type of the genus Menops.
Progressive brachycephaly is characteristic of these
scattered and loosely related species, as shown in the
following ascending series in the Titanotherium zone:
Zygomatic
index
Upper beds: AUops crassicornis (type), brachycephalic. _ 76
Upper beds: UTenops wonons, subbrachycephalic 73
Svimmit of middle beds: AUops serotinus (type), sub-
brachycephahc 72
Middle beds, lower levels: AUops marshi, mesa,ticepha.\\c- 64-69
Base of lower beds: AUops walcoUi (type), mesatice-
phalic (?)
Geologic distribution of AUops. — AUops walcotti
Osborn: A. walcotti, from the lower levels of the
lower Titanotherium zone, is a small, very primitive
form, distinguished by a narrow head, long, broad
nasals, and relatively long face, horns with elongate
oval section, and two superior incisors.
AUops marshi (Osborn): A. marshi, from the upper
levels of the lower beds, is distinctly mesaticephalic.
Nasals long and broad, horns subtrihedral in section,
premolars more progressive than in A. walcotti. This
animal is finely represented by eleven skulls in several
museums, which present a series of ascending muta-
tions.
AUops serotinus Marsh: In A. serotinus the nasals
are still elongate and the horns are slender, elongate,
subtrihedral, preserving the section characteristic of
A. marshi. The premolars are still in a retarded stage
of evolution. This species is represented by five speci-
mens, two of which appear to present transitional
stages toward A. crassicornis, as follows: Skull Am.
Mus. 520 appears to be in a transitional stage between
A. serotinus and the more brachycephalic species A.
crassicornis, for the horns are in an intermediate stage
of development and the dental measurements have
the degree of development of the^premolar tetartocones
coincident with those in A. serotinus. Transitional
skull Nat. Mus. 4938 nearly equals in some of its meas-
urements the A. crassicornis type, but the premolar
tetartocones are still decidedly retarded.
AUops crassicornis: The type specimen of the species
A. crassicornis is distinguished both by the more
massive proportions of the skull, the obtuse, short and
massive horns, the abbreviation of the nasals, and
especially by the more advanced development of the
tetartocones upon the premolars. It is by no means
certain that this massive, broad-headed animal is a
descendant of the types- named above.
Stratigraphic position of species of AUops
Stage
C
Level
Species
Upper.
Middle.
Lower.
?A. montanus.
?[Menops varians].
?A. crassicornis.
B
Upper.
Middle.
Lower.
A. serotinus.
A. marshi.
A
Upper.
Middle.
Lower.
A. marshi.
A. walcotti.
EVOLUTION OF THE SKULL AND DENTITION OP OLIGOCENE TITAN OTHERES
507
Progressive pJiyletic characters. — Members of the
Allops phylum are known from the whole Titanothe-
rium zone. In general, the skull and teeth are inter-
mediate in character between the typical Brontops
and typical Menodus. The skull in males, originally
mesaticephalic, shows a marked progression toward
brachycephaly, the zygomatic indices rising from 64
to 76. The nasal bones in the males progressively
shorten as in members of the Brontops phylum. The
broadly trihedral basal section of the horns connects
these elements with Menodus rather than with Bron-
tops. The face is relatively longer than in Brontops
and more abbreviate than in Menodus. As in Bron-
tops the incisors are round topped with a formula of
2-1. One of the most distinctive features of Allops is
the transversely lanceolate form of the canine teeth
which enables us to connect A. vmlcotti with this series
rather than with the Menodus series. The opposite
grinding series are rectilinear, or nearly parallel, as in
Menodus. While the grinders approach those of
Menodus in having elongate or subhypsodont crowns
they are less dolichocephalic in proportion — that is,
the transverse diameters of the molar teeth slightly
exceed the anteroposterior diameters, whereas in
Menodus the reverse is the case. The dental index
is high — in males 46-47, in females 45-50. The
premolars exhibit pronounced external cingula as in
Menodus.
Briefly, these animals resemble Menodus in the
trihedral shape of the horns and in numerous other
characters. They differ from the true Menodus in
the progressive abbreviation of the nasals, in the
brachycephaly of the zygomatic arches, in the reten-
tion of at least one pair of upper incisor teeth. Thus
they are provisionally assigned an intermediate
phyletic position.
Several of the more advanced or upper-level speci-
mens of ^. serotinus were discovered in the overflow de-
posits of the upper Titanofherium beds rather than in
the main sandstone or channel deposits. This may
afi'ord some clue to the rarity of these crania.
Characters of the genotype. — In describing the
genotype, Allops crassicornis , in 1887 Marsh charac-
terized it as a skull resembling in general that of
Menodus giganteus but as differing in the possession
of a single superior incisor tooth. The type possesses
a pair of well-developed second incisor teeth which
are always vestigial in Menodus but present in the
subgenus Menops. There are other still more im-
portant differences, which will be enumerated in the
description of this species.
Affinities of Allops. — Allops marshi has a very low
zygomatic index (64-69) in comparison with Allops
serotinus (72) and A. crassicornis (76), a fact which
suggested the theory that A. marshi may represent
the females of some other species such as B. dispar.
Some of the smaller skulls referred to Allops marshi
are with difficulty distinguishable from females of
Brontops hrachycephalus ; others approach B. dispar,
with which they agree in dental measurements (see
above); many are also strongly suggestive of Menodus
trigonoceras in the characters of the premolars and
molars and in the horns. All known skulls of A.
marshi are distinctly inferior in size to those of A.
serotinus. As shown by the detailed characters of
the skull and dentition and by the tables of measui'e-
ment, A. serotinus and A. crassicornis combine the
characteristics of Brontops and of Menodus in a
remarkable manner. They exhibit the characteristic
horn sections and distally squared nasals of Menodus,
the sharp premolar and molar cingula of Menodus,
combined with the broader skidl and broader grinding
teeth, expanded zygomata, shortened free nasals,
and retained incisors of Brontops. The form of the
canine also is more or less intermediate between the
conical canine of Brontops and the sharp-edged canine
of Menodus. The type and only known specimen of
Menops varians resembles Allops crassicornis in the
thick outwardly divergent horns, in the wide skull
(index 73), in the wide premolars and molars, in
the reduplicate tetartocone of p*. On the other hand
it approaches Menodus giganteus in the anteropos-
terior dimensions of the dentition and skull and in
the nasal and horn sections, so that it is placed in
the genus Menodus.
These facts suggest the possibility that some of
the forms called Allops may represent a hybridization
between species of Brontops and of Menodus.
Measurements, in millimeters, of lower jaws and teeth correlated
with and referred to Allops
Posterior canine to hypoconulid of ma
Pi-m3
M ,-m:i
Symphysis to condyle
Depth below ms
A. marshi,
Nat. Mus.,
Gidley's skull
320
335
225
375
94
?A. walcotti,
Nat. Mus.,
4247
303
298-1-
200
510
108
508
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Standard measurements in tJie Allops 'phylum,"- in millimeters
XJpper teeth
Skull
Jaw and teeth
1
"p.
a
a
>
§1
s
i
o
T
o
8
S
(1,
1
ft
5
.1
1
1
a
1
1
i
a,
i
h
3
l|
SB
as
D
"ft
a
>>
j
A. crassicornis, Nat. Mus. 4289, cf (type) . 370
150
148
142
131
220
217
213
9,10
45
44
20
30
750
''570
76
143
146
72
210
A. serotinus, Nat. Mus. 4938,?
335
330
330
330
38
36
26
40
23
20
739
720
665
705
+ 702
675
"600
565
430
525
72
78
64
74
695
710
648
690
96
68
115
81
137
106
118
133
133 206
A. serotinus, Nat. Mus. 2151,9
A. serotinus, Nat. Mus. 4251, cf (type)__
"^A (dispar) serotinus'^ Nat Mus. 1217
i-UO
140
196
195
196
203
203
198
205
200
193
179
194
183
192
191
185
186
185
169
+ 30
240
230
A. marshi, Am. Mus. 1445, cf (para-
335
323
323
32C
320
31S
313
317
315
''31C
31C
31C
30?
30C
28e
135
133
129
126
130
133
132
129
131
126
122
?80
124
119
112
119
+ 34
37
"22
466
69
615
105
100
113
35
23
24
663
660
665
427
64
620
545
625
99
93
112
105
108
140
A marshi'' Nat Mus. 4738
430
64
366
127
240
31
28
A marsW Nat. Mus. 4942
A marshi? Nat. Mus. 8314
34
22
662
655
675
i'640
673
656
643
630
640
448
420
465
67
64
69
336
113
225
35
20
375
A marshi' Nat. Mus. 1213, ?
— -
98
33
34
23
25
310
115
207
33
20
520
A. marshi, Am. Mus. 501, 5 (type)
A marshi, Field Mus. 6900
450
"440
?476
450
"340
67
66
?74
71
625
620
696
100
103
105
103
105
126
97
330
A. marshi?, Carnegie Mus. (?) 341,9 .__
A. marslii?, Nat. Mus. 1215, 9
27
34
19
20
A. walcotti, Nat. Mus. 4260, d' (type) ._ _
A walcotti Nat Mus 8753
686
105
98
100
34
Percentage of change from Allops wal-
+ 3C
+ 17
+ 67
-11
+ 40
+ 240
....|____
o Allops, like BiploclonuSj is a less clearly consecutive and distinguished phylum than Brontops, Menodus, or Broniothenum. For observations on the measurements
see p. 507. ' Estimated.
The Allops phylum as represented in the Hatcher collection of 26 sTculls and lower jaws from the Chadron formation
in the United States National Museum
Genus an
d species
Catalog No.
Material
A
crassicornis
Do
Marsh
4289,
4709,
8740,
2117,
4938,
2161,
4261,
1226,
8318,
4945,
8731,
8777,
1216,
8769,
4252,
8737,
&
&
cT
9
9
&
&
9
9
&
9
9
&
cf
Skull. Type.
Skull. Measurements agree with type. Canine not typical.
Do
Skull, right half of jaw. Agrees in size with type. Nasals, upper canines, and incisors
Do
lacking. Specific reference doubtful.
Skull.
A
serotinus M
Do
Skull. Horn like that of A. crassicornis.
Skull.
Do
Skull. Type.
Do
Anterior part of skull.
Do -
Skull. Measurements agree with A. serotinus, female. Canine not typical. Specific
Do
reference doubtful.
Skull. Same size as large Brontops dispar. Specific reference doubtful.
Do
Anterior half of skull and jaws.
A.
A.
? serotinus Marsh
marshi (Osborn)
Do-
Pair of jaws.
Skull and teeth. Measurements agree with A. marshi. Larger than B. brachycephalus.
Skull. Larger than any specimen of A. marshi; approaches A. serotinus in molars.
Do_
Skull. Resembles paratype of A. marshi.
Do
Skull and jaws, left humerus, and pelvis. Close to type. Molars same length as those
of Brontops dispar.
EVOLUTION OF THE SKTJLL AND DENTITION OF OLIGOCENE TITANOTHERES
509
The Allops phylum as represented in the Hatcher collection of 26 skulls and lower jaws from the Chadron formation
in the United States National Museum — Continued
Genus and species
Catalog No.
Material
A.
marshi (Osborn)
Do - - _- --
1213, ?
8798
8317?
4738
4942
4778
4254, cf
4260, &
4247, ?
8753
Skull. Tooth measurements agree with A. marshi. Slender zygomatic arch.
Right lower jaw and symphysis.
Do --
Skull. Generic reference doubtful in absence of canines and nasals.
A
? marshi? (Osborn)
Do
Skull. Agrees in measurements with other skulls of A. marshi.
Upper teeth.
Do
Skull. Very young individual. Generic and specific reference doubtful.
Do
Skull. Vestigial lateral superior incisors. Generic and specific reference doubtful.
^
Skull. Type advanced in age; see paratype (No. 8753).
Do
Jaws.
Do
Nearly complete upper dentition (i^-m^). Paratype. Palate slightly smaller than in
type. Canine measurements very characteristic of the genus.
Allops walcottl Osborn
Plates XX, XXI, CXI, CXII: text figures 207, 389, 391, 394,
397, 409, 413, 429, 430
[For original description and type references see p. 241)
Geologic horizon. — Lower Titanotherium zone (Chad-
ron A) of South Dakota.
Specific characters. — Premolars with small tetarto-
cones; p'-m' 285 millimeters. Incisors %. Horns
elongate oval, no connecting crest. Mesaticephalic.
Nasals elongate, broad. Face relatively elongate.
The type skull of this species (Nat. Mus. 4260) from
Chadron A is narrow and elongate, partly owing to
lateral crushing. This feature conceals its resem-
blance to Allops marshi, which is apparent in other
features — namely, (1) primitive, long nasals, (2) horns
primitively short and obliquely oval, (3) large lateral
incisor (i2) and small first (ii) or median incisor,
(4) premolars accelerated, tetartocones more advanced
than in Brontops rohustus of level C.
Observations on the measurements of Allops walcotti. —
The type of this species exhibits the following measure-
ments in comparison with skulls of B. hrachycephalus
and Menodus heloceras, which shows that the type of
Allops walcotti has relatively large premolars and small
molars.
Measurements of Allops walcotti, Menodus heloceras, and Bron-
tops hrachycephalus, in millimeters
Pi-m3
Pi-p<
M"-m3
Pmx to condyles
Nasal length
Horn length
PS ap.Xtr
MS ap.Xtr
A., walcotti,
Nat. Mus.
4260 (type)
285
112
169
640
105
100
35X51
60X61
M. heloce-
ras, Am.
Mus. 14576
265
170
603
132
70
B. brachycephalus
Nat. Mus.
4940, 9
265
101
160
102
32X51
62X70
Nat. Mus.
4261, cT
(type)
280
»104
178
580
85
33X53
68X73
' Estimated.
The skull is crushed laterally, but probably had a
low zygomatic index — that is, it was mesaticephalic.
While its reference to Allops requires confirmation, its
nearer affinities appear to be with this genus rather
than with Brontops or Menodus. The external cin-
gula of the premolars are not as sharply defined as in
other primitive members of the menodontine group.
Geologic and geographic distribution. — This species
is represented at present by a single skull, the type
(Nat. Mus. 4260), which is recorded as probably from
the lower levels (A) of the lower Titanotherium zone
of South Dakota.
Comparison with other species. — This animal should
naturally be compared with other titanotheres from the
lower beds. It is readUy distinguished from B.
brachycephalus by a number of characters as follows : (1 )
The skull is much more doHchocephalic, a feature that
is intensified by lateral crushing (see below); (2) the
nasals are long and subquadrate in form; (3) the
individual measurements of the grinding teeth show
that the series throughout is somewhat narrower than
the grinding series in B. brachycephalus.
These contrasts with the broad-skulled B. brachy-
cephalus naturally suggest comparison of this animal
with primitive members of the long-skulled Menodus
phylum, such as M. heloceras, remains of which from
the lower Titanotherium zone are sparsely known.
Comparison with M. heloceras shows that A. walcotti
possesses the following distinctions: (1) Horns more
elongate-oval in section, less trihedral; (2) no con-
necting crest between the horns; (3) alveoli for two
incisor teeth of considerable size (incisors are extremely
vestigial in the Menodus phylum).
There remains the comparison with Allops marshi,
the form to which, on the whole, this cranium seems
to present the largest number of resemblances. The
animal differs from A. marshi in the inferior dimensions
of the dental series as well as in the detailed propor-
tions of the teeth and the characters of the premolars,
but it appears to be an older and more primitive form
which belongs to the A. marshi phylum. In _ the
horn section, in the proportions of the nasals, in the
510
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
shape of the canine teeth it strongly resembles A.
marshi.
At the same time Allops walcotti presents certain
resemblances to Menodus Tieloceras and was long re-
garded by the author as ancestral to Menodus, which
it foreshadows in its apparent dolichocephaly and in
the strongly developed external cingula of the pre-
molars as well as in the somewhat accelerated pre-
molar tetartocones.
Description of the type sTcull. — The type skull, No.
4260, as figured on Plates CXI and CXII of this mono-
The skull is in the seventh stage of growth. It
exhibits a slender but strongly indented occiput in
the center of which is a median ridge which projects
from the superior border. The cranial vertex is ex-
ceptionally long and narrow, with a decided lateral
crest overhanging the supratemporal fossa; in the mid-
region of this supratemporal crest we note (PI. CXI)
two lateral projections which are also observed both
in Allops serotinus (Nat. Mus. 4251) and in several
skulls of the Menodus phylum. In front of these are
the supraorbital projections. The horns are lateral
B
C
D
Figure 429. — Sections and contours of skulls of Allops walcotti and A. marshi
A, Allops walcotti, Nat. Mus. 4260 (type); lower levels of Chadron A; a very primitive stage, approaching both Brontops hrachycephalus and Menodus Jieloceras in the
character of its sections. Low horns elongate in basal section and placed not far in front of the orbits, nasals long and narrow, zygomata slender. No very definite
marks of affinity with Allops are revealed by the sections. B, A. marshi, Am. Mus. 501 (type) ; horns elongate trihedral in basal section and placed considerably in
front of the orbits, nasals broad distally, zygomata little expanded. C. A.marslii, Am. Mus. 1145 (paratype); the sections differ little from those of the type, although"
the skull as a whole is broader. D, A. marshi. Harvard Mus.; differs somewhat from the type in the steeper profile of the horns, basal section obliquely trihedrali
zygomata little expanded, parietal vertex narrow. One-eighth natural size.
graph, was referred mistakenly by Marsh to Bronto-
therium gigas notwithstanding its marked inferiority
in size to the type of that species, which properly
belongs to the Brontotherium phylum. In studying
the plate and figures emphasis must be laid on the
fact that the skull is extremely crushed laterally,
and that its original mesaticephalic proportions are
artificially increased into dolichocephaly. In other
respects the plate as well as the accompanying figures
and sections give an admirable idea of the formation
and proportions of this primitive titanothere.
in position, overhanging the sides of the face and
of an elongate-oval basal section anteroposteriorly,
distinct from that of the type of Allops marshi. The
long axis of the section is anteroposterior, whereas
in A. marshi it is oblique; the horn rises only 70
millimeters above the side of the narial aperture,
40 millimeters above the vertex of the skull. The
nasals also resemble those in the type of A. marshi;
they extend very far forward, contracting slightly,
and cleft at the tips. The relative elongation of the
face is a decided feature (PL CXI, A'), the bridge over
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
511
the infraorbital foramen being very broad, even
broader than in Menodus giganfeus; there is a pre-
orbital knob on the lacrimal, which is seen also in
M. trigonoceras . The zygomatic arches, as shown in
all three views of the skull, are deep and narrow, not
expanding widely, and resembling those of the Menodus
type. There is a narrow contact between the post-
tympanic and postglenoid processes, and the par-
occipital process is narrow in palatal view (PL CXII),
wherein the dolichocephalic structure is again ap-
parent but is somewhat exaggerated in this type by
lateral crushing.
Dentition. — There are two incisor alveoli which show
that these teeth had not undergone the degeneration
observed in members of the Menodus phylum. The
right canine so far as preserved shows somewhat less
anteroposterior compression than in the type of A.
marshi. The premolars of A. walcotti exhibit pro-
nounced internal and faint external cingula and feebly
developed tetartocones (fig. 430). The tetartocones,
however, are better developed than in the contem-
porary B. hrachycephalus , since the constriction sepa-
rating the tetartocone from the deuterocone in p-~^ is
a little more anterior in position. Furthermore, the
tetartocone of p* is not a concave spur from the
cingulum as in B. h'acJiycepJialus
or B. dispar. Characteristic
features are the crenulation of the
internal face of the deuterocones
and the broad internal cingula.
In regard to the proportions of
the grinding teeth, comparison
with the average anteroposterior
and transverse diameters of the
grinding teeth of five skulls of B.
iracJiycepJialus shows that in this
specimen the teeth are less mark-
edly brachyodont, a fact in accord-
ance with its supposed affiliations
with the Allops phylum. The
teeth are, in fact, as elongate as in
members of the Menodus phylum,
but this may be due partly to the
lateral crushing. Detailed meas- "'
urements of the tooth proportions indicate, however,
that they are closer to B. hracJiyce2)halus than to those
of M. trigonoceras .
Characters oj lower jaw. — A lower jaw (Nat. Mus.
4247, fig. 413) agi'ees exactly in size and may be pro-
visionally associated with this species. It presents
the following characters: The coronoid is elevated;
the angle projects downward and backward; the inen-
tal foramen is below ps. ly, P^. The canines are
pointed, erect, with an incomplete internal cingulum
and somewhat flat inner face; pi is entirely wanting,
a variable character; the cingulum on the remaining
grinders is moderately developed; nis exhibits an in-
cipient crenulate internal crest of hypoconulid.
Measurements of jaw of Allops walcotti, Nat. Mus. 4B47
Millimeters
Angle to symphysis .525
Condyle to angle 235
Canines, anterior enamel 34
Canines, anteroposterior 21
Ms, anteroposterior 88
M3, transverse 38
Pa-mj 290
This lower jaw has long, conical canines; it should
be compared with the lower jaw of Am. Mus. 1495.
Allops marshi (Osborn)
(Megacerops marshi Osborn, 1902)
Plates XXXVIII, CXII-CXVI; text figures 197, 378, 381, 389,
391, 397, 399, 409, 429, 431, 432, 615, 712
[Foi- oi-iginal description iiud type refersDces see p. 233. For slceletal characters
see p. 678]
Geologic horizon. — The geologic levels of the type
and paratype of Allops marshi in the American
Figure 430. — Upper teeth of Allops walcotti
. Mus. 8753 (paratype); m^ is missing. One-half natural size.
Museum are not recorded, but they are probably from
upper A or lower B. A fine skull (Field Mus. P 6900)
is recorded as from .50 feet above the Pierre shale and
50 feet below the level of a skull of Brontotherium
hatcheri — that is, probably the upper levels of A or the
lower levels of B. A female skull (Nat. Mus. 1215)
slightly smaller than A. marshi is recoi'ded from
middle B.
Specific characters. — Skull longer than in Brontops
hrachycephalus (645 to 675 mm., average 665), but
shorter than in B. dispar, of mesaticephalic proportions ;
zygomatic index, 64 to 69, average 66, and thus similar
to females of B. hrachycephalus. Horns (105 to 140
mm.) longer than in B. hrachycephalus ; basal section
512
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
trihedral, more or less pointed or trihedral at the
summits. Nasals elongate, free length 98 to 105 milli-
meters. Incisors 2-1, large i^, small i^ Grinding
series, length 310 to 335 millimeters, average 319;
molars average 192; dental index same as in B. hrachy-
cepJialus and B. dispar, namely, 47. Canines strongly
compressed anteroposteriorly, length of crown 34 to
37 millimeters. Premolars with tetartocones of p^
better developed than in B. hrachycepJialus, tetarto-
cones of p^ and p^ better developed than in B.
IracJiycepTialus and similar to progressive members of
B. dispar. Occiput not greatly prolonged back of
zygomatic arches.
Observations on the measurements oj Allops marshi. —
From Brontops dispar the skulls of A. marsJii are dis-
tinguished above all by their very low zygomatic
index, 64 to 69, as compared with 77 to 87 in B. dispar.
This marked narrowness, together with the small size
of the canines (vertical diameter 34 mm., as compared
with 40 in B. (validus) dispar), has led to the suspicion
that Allops vfiarsTii may be the female of B. dispar. A
comparison of the paratype of Allops marsTii with the
type of B. validus is given below:
Measurements of Brontops (validus) dispar and Allops marshi, in
millimeters
P'-mS
Pi-p<
Mi-m3
Canines :
Vertical
Anteroposterior
Pnix to condyles
Zygomatic index
B. (validus)
dispar, Nat.
Mus. 4290
(type)
. raarstii,
Am. Mus. 1445
(paratype)
320
130
203
40
27
660
85
335
135
203
37
22
675
64
The relative widths of p* and m' also appear not to
differ very clearly in the two forms:
Anteroposterior and transverse measurements of p* and m^ in
Brontops (validus) dispar and Allops marshi, in millimeters
pi
M3
Ap.
Tr.
Ap.
Tr.
B. dispar (type of validus)
A. marslii (type)
40
39
60
61
72
70
75
73
A. marslii is decidedly smaller than A. serotinus and
A. crassicornis, and apparently no known skulls bridge
over this gap. It is larger in all measurements than
B. hrachyceplialus.
Materials. — This species is represented by 15 or
more skulls, including the type (Am. Mus. 501, Pis.
CXIV, CXV), a well-preserved skull; the paratype
(Am. Mus. 1445, Pis. CXIV, CXV) ; a well-preserved
skull (Brit. Mus. 4446 M, PL CXIV) whose geologic
level is not recorded; an exceptionally perfect skull
(Field Mus. P 6900), associated with the lower jaw and
parts of the skeleton, from the upper levels of A or
the lower levels of B; a fine skull in the Museum of
Comparative Zoology, level not recorded, originally
described by Scott and Osborn (1887.1, p. 158) as
" Menodus coloradensis"; a cast of an unknown skull
(Carnegie Mus. 289) ; a skull, recorded from upper A,
probably a female (Nat. Mus. 1213); a skull from B
(Nat. Mus. 1215).
Comparison of Allops with members of the Brontops
phylum. — Are these specimens females of B. dispar?
There is some evidence that the specimens attributed
to A. marshi are female forms of B. dispar, consisting
chiefly of the following items: (1) The dental indices
are the same; (2) the tetartocone development is prac-
tically identical; (3) the cephalic index is mesaticepha-
lic, like that of the females of B. dispar; (4) the speci-
mens of A. marshi are found on lower levels of the
range of B. dispar. On the other hand, the evidence
against regarding A. marshi as the female form of B.
dispar is somewhat stronger, as follows: (1) The skulls
of A. marshi are generally recorded from lower geologic
levels; (2) the skulls in the Field and Harvard museums
have larger canines, indicating that they are males;
(3) the occiput of A. marshi is not greatly prolonged
behind the greatest width of the zygomatic arches, in
contrast with B. dispar, in which the occiput is greatly
produced posteriorly (see diagram); (4) the horns of
A. marshi are more triquetrous or transversely oval,
while the horns of B. dispar are rounded; (5) the nasals
of A. marshi are much more elongate.
Although the specific and phyletic distinction of A.
marshi from B. dispar thus appears certain, the
"group" affinity of the animals is very strong. We
observe (1) the pit in the vertex of the cranium, as
in B. brachycephalus , B. dispar; (2) the features in
which A. marshi differs from B. dispar tend to relate
A. marshi more closely to the succeeding form B.
rohustus.
Among the similarities between A. marshi and B.
roSttsius are the following : (1) The form of the canine,
which tends to obtuseness; (2) the abbreviation of the
occiput behind the zygoma; (3) the broad, rugose
summit of the occipital pillars; (4) the breadth and
squareness of the nasals; (5) the presence of two knobs
on either side of the median line of the occiput for
the recti capiti muscles. The horns in their prophetic
growth do not acquire the trihedral section seen in
the Menodus phylum but tend to become more oval.
Thus in the paratype of ^. marshi (Am. Mus. 1445) we
find an approach to the transversely expanded horns
of B. robustus. Among the more primitive characters
of the skull distingmshing A. marshi irom B. rohustus
are the smaller size and less robust structure in general,
the shortness and obliqueness of the nasal section of
the horns, the greater length and slenderness of the
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
513
nasals, the greater breadth of the malar bridge over
the infraorbital foramen, the narrower contact between
the postglenoid and post-tympanic processes.
Against the theory that A . marshi is directly related
to B. rohustus is, however, to be noted the important
fact that in the premolar teeth the tetartocones are
more advanced than in B. rohustus, in which they are
singularly retarded in development. The transversely
expanded canines constitute another clear distinction.
Specimens referred to Allops marsM seu crassicornis. —
A skull in the British Museum of Natural History
collection, London (No. 5743 M), may be regarded
provisionally as a very advanced or progressive stage
in the evolution of this species, although it exhibits
some characters which lead us to regard it as aberrant
from the typical A. marshi, especially the somewhat
flattened superior section of the horns, which suggests
resemblance to Brontotherium leidyi; but the internally
placed tetartocones of the premolars differentiate this
type from any member of the genus Brontotherium, in
which the tetartocones are invari-
ably externally placed — that is, to-
ward the buccal side of the crown
of the teeth rather than toward
the lingual side, as in this speci-
men. The female sex of this spec-
imen is apparently indicated by the
small size of the horns and the
slenderness of the canines and zy-
gomata. The nasals are somewhat
long and delicate, cleft distally; the
short horns point obliquely out-
ward, giving evidence of having been ,
subflat posteriorly. They are other- '
wise of the general type seen in
A. marshi. The zygoma, although
crushed, was apparently deep, with
slight buccal expansion. The ex-
ternal auditory meatus was a large,
round opening, as in Menodus and
Allops — that is, of mesaticephalic
type. The measurements (see table,
p. 508) agree better throughout with
A. marshi than with any other type, although even
in the paratype of A. marshi the tetartocones of
the premolars are not so strongly developed as in
this specimen. Two well-developed upper incisors
are retained on the right side, i' being much smaller
than i^.
The lower jaw, with its flat, deep ramus, pronounced
chin, slim, pointed canine (slightly flattened on the
internal face), weak, noncrenulate hypoconulid of va^,
resembles the primitive type of the Menodontinae in
general and to a less degree that of the A. marshi type.
Our conclusion is that this animal corresponds more
nearly with a very progressive stage of A. marshi than
with any other known species. The extremely
advanced condition of the tetartocones of the pre-
molars may, however, entitle it to distinct specific
rank. The skeletal parts which are thought to be
associated with this skull are described below.
STcull in the Harvard Museum of Comparative Zool-
ogy.— This skull (fig. 431) agrees with the type in
the specific characters of the horns and nasals and in
the possession of round-topped incisors.
Progressive characters and ascending mutations in
specimens referred to Allops marshi. — The type of
Allops marshi (Am. Mus. 501) exhibits a zygomatic
index of 67. It differs from the paratype (fig. 429)
in the shape of the nasals and in the more retarded
premolar tetartocones. There is a trace of the cir-
cular pit in the occipital vertex. The tooth row is
short (310 mm.).
Figure 431. — Skull of A Hops marsh
Harvard Mus. Front and side views. One-sixtli natural size.
The paratype, probably a male (Am. Mus. 1445),
represents a more progressive stage: (1) it is the broad-
est skull referred to this species (zygomatic index 69),
the skull being broadly depressed with stout zygo-
mata; (2) the tetartocones are very advanced; (3)
the occiput is extremely short behind the zygomata;
(4) the nasals are thin and do not spread distally; (5)
there are traces of a circular pit in the occipital vertex ;
(6) the tooth rows are longer than in any other speci-
men in the series, namely, 335 millimeters; (7) the
same may be said of the molars (average 205 mm.);
the grinding teeth are also slightly more elongate;
(8) the tetartocones are more advanced than in B.
rohustus — in fact, more advanced than in several speci-
mens attributed to B. dispar.
514
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The British Museum specimen (No. 5743 M)
above described agrees well with the type and para-
type in all measurements except in the zygomatic
index, which is low and agrees with that of a supposed
female (Nat. Mus. 1213) from the lower beds.
The horns are more erect than in the paratype-
The tetartocone development is about the same as in
the type.
The Field Museum specimen (No. P 6900) agrees
well with the type in measurements but differs in the
somewhat more backwardly prolonged occiput; the
canines, which are large, indicate that this specimen
is probably a male. It is more fully described below.
Figure 432. — Lower jaws of Allops marshi and Allops? sp.
A, Allops marshi, Field Mus. P 6900; ramus sweeping gently forward with lower border
nearly horizontal and angle not produced downward, chin slightly convex, canine appressed
to P2, external cingula not sharply defined. B, AUops'i sp., Ottawa Mus.; Cypress Hills,
Saskatchewan. This specimen (one of Cope's cotypes of M, ang-ustigenis) differs from the
typical Menodus in the presence of incisors (as indicated by the alveoli) and in the somewhat
less hypsodont form cf the grinding teeth; the angle is not produced backward. The refer-
ence to Allops is provisional. About one-si.\th natural size.
The skull in the National Museum (No. 1213)
agrees closely with the type of A. marshi in measure-
ments and is somewhat more progressive in premolar
evolution than the type of B. hrachyceplialus.
Another skull (Nat. Mus. 1215) is recorded from
the middle levels of B and may pertain to this species,
although the occiput is more prolonged backward than
in the type; the premolar tetartocones might represent
either this species or B. dispar.
Equally interesting is the skull Nat. Mus. 1214,
referred to B. hrachycej>7iahis but possibly an ancestor
oi A. marshi.
Detailed characters of Allops 7narshi. ^-The. detailed
characters of the type (Am. Mus. 501) are as follows:
The skull, apparently belonging to a female, is in the
seventh growth stage. The incisive border exhibits
alveoli for a large lateral and small median incisor.
The canines have short, obtuse, lanceolate crowns, a
posterior and a slight lateral cingulum. The first su-
perior premolar is bifanged, close behind the canine;
it exhibits no tetartocone. The tetartocones are less
advanced than in the contemporary 31. trigonoceras,
moderately developed on p' and p^ and forming a de-
pressed loop on p*; the external and internal cingula are
clearly defined. The molars exhibit partial cingula on
the ectoloph and rudimentary cingula on the
inner side. The hypocone of m'^ is connected
with the cingulum (in the paratype, however,
the hypocone is separated from the cingulum as
a low cusp) ; there is a rudimentary metaloph,
especially on m^. The postorbital process of the
frontal crest forms a rudimentary sharp angle;
the orbit is large; the malar bridge is broad,
with a low ridge; the buccal expansions of the
zygomata are slight in top view; the edges of
the frontal crest are seen to diverge widely,
running to the malar ridges of the horns. The
nasals are of medium length, deeply convex
above and concave below, projecting beyond
the premaxillaries anteriorly.
The paratype skull (Am. Mus. 1445), prob-
ably that of a male, is, as above noted, more
progressive. In p* the tetartocone is more
distinctly budded off. As in the type the
canines have the short, sublanceolate section
and posterior lateral cingulum quite distinct from
the more pointed canines of Brontops dispar.
The Field Museum skull (No. P 6900) as-
sociated with a lower jaw and a manus of
Allops marshi, from Phinney Springs, S. Dak.,
has been skillfully reconstructed by Mr. Riggs
and is now one of the least distorted titano-
there skulls of all that are known. It agrees
closely in measurements with the other skulls
referred to^. marshi and shows corresponding
specific characters.
Detailed measurements of the upper grinding
teeth prove that the transverse measurements
exceed the anteroposterior measurements throughout.
This proportionate excess of transverse diameter is
progressive in this brachycephalic line.
Ascending mutations. — This species, therefore, in-
cludes specimens which represent several "ascending
mutations" in increase of size, in the development of
the premolar teeth, in separation of the tetartocones,
and in expansion of the zygomatic arches. Of these
the geologically oldest perhaps is the National
Museum specimen No. 1214, which is distinctively a
B. hrachycephalus in its measui'ements. Next in order
of evolution comes the female specimen (Nat. Mus.
1213), distinctly an A. marshi in its measurements.
EVOLUTION OF THE SKULL AND DENTITION OB^ OLIGOCENE TiTANOTHERES
515
On a higher plane but still in an intermediate stage
of development is the female type skull of A. marsJii
(Am. Mus. 501). The Carnegie Museum skull No.
123, formerly referred to this species, is more probably
a Brontops dispar (Am. Mus. 1445). The paratype
is the largest and most progressive form known.
This range of progressive evolution in the materials
at hand afi'ords strong additional proof of continuity.
Characters oj the lower jaw. — The characters of the
lower jaw of this species are illustrated in Figure 432,
A. The chin is gently convex, not angulate; the angle
is but little produced downward; the lower border
of the ramus is nearly horizontal.
Allops serotinus Marsh
("Allops serotinus" Osborii, 1902)
Plate.s XCIII, CXVII-CXX, CXXXII; text figures 184, 375,
387, 391, 393, 394, 399, 409, 433, 434
[For original description and typs reterencas see p. 225]
Geologic horizon. — Titanotherium zone of South
Dakota; summit (?) of middle beds (Chadron B).
Specific characters. — Skull mesaticephalic. Length
705 millimeters (c?), breadth 525. Zygomatic index
74. P'-m^, 330 millimeters. In males nasals broad
(133 mm.) and short (81 mm.). Horns subtriangular
elongate, widely divergent. Incisors 2-1, external
incisor large, median incisor reduced or wanting.
Superior canines, males, 41 millimeters. Premolars
with deuterocones forming main internal portion of
crown. Tetartocones much smaller, especially on p''.
External cingula defined on premolars, more or less
continuous on molars.
Measurements of Allops crassicornis and A. sero-
tinus.— The skulls referred to Allops serotinus and
Allops crassicornis form an ascending series, in which
p'-m' rises from 330 to 370 millimeters. The three
skulls referred to A. crassicornis differ from those re-
ferred to A. serotinus, especially in the greater length of
the premolar and of the molar series. The contrasts
in measurements with Brontojjs dispar and with Meno-
dus trigonoceras are shown in the accompanying table.
Measurements of Allops crassicornis, A. serotinus, Brontops dis.
par, and Menodus trigonoceras, in millimeters
[All specimens male]
P'-m3
P'-p*
iVIi-ms
P*, ap. by tr_
355-370
131-150
210-220
46X68
1S43, ap. by tr 80X82
Pmx to condyles
Zygomatic index
Nasal length
Nasal breadth
330
13.3-140
195-206
43X59
68X72
750i 705-739
76^ 72-78
72[ 68-81
146 106-137
320-345
122-145
200-215
40X60
72X75
643-665
78-87
85-94
102-120
335-360
127-136
220-225
43X51
72X70
to 82X79
670-770
66-79?
115-140
120-135
From this table it will be seen that Allops crassi-
cornis and A. serotinus are distinguished by the great
extent of p'-p'', by the relatively great width of p* and
m^ by their high zygomatic index as compared with
that of Menodus trigonoceras, and by their relatively
short, wide nasals. In general the skulls of Allops
serotinus and A. crassicornis combine the character-
istics of Brontops and of Menodus.
General characters. — This species includes crania that
immediately suggest Menodus giganteus, but on close
examination they are seen to be less powerful and less
robust. It is noteworthy that the supraoccipital crest
is less rugose and was probably provided with less
powerful muscles.
Materials. — The type skull (Nat. Mus. 4251) was
originally recorded by Hatcher from the top of the
Titanotherium zone of South Dakota. This record,
however, was subsequently revised by Hatcher as
possibly due to secondary erosion and deposition of
the overlying Brule formation {Oreodon zone) below
the true upper level. Thus we may provisionally re-
gard Allops serotinus as typically of the uppermost
levels of the middle beds. It is noteworthy that the
matrix is a fine clay, and the skull has a brilliant
yellowish-white surface.
Another skull (Nat! Mus. 2151) is recorded as
foimd by Hatcher is the same whitish clay matrix,
about 15 feet away from the type, and entered as
80 feet above the Cretaceous Pierre shale. This fur-
ther confirms the view that these animals belong in
the upper portion of the middle beds, or B, and not
in the upper Titanotherium zone (C).
Two other skulls are known, namely, Nat. Mus.
4938 and Am. Mus. 520.
Description of the type sTcull. — In addition to the
general characters noted above in which this skull
differs from that of the typical Menodus or the typical
Brontops, it is important to record the following
detailed characters:
The type skull (Pis. CXVII, CXIX) is vertically
crushed, and owing to this the lateral spread of the
horns is unnaturally increased. It is a male speci-
men. Although in the seventh stage of growth, the
skull is decidedly delicate in its structure and pro-
portions. It represents an animal about a fourth
smaller than the majority of the specimens referred
to the great animal Menodus giganteus. The inferior
aspect of the skull is well represented in Plate
CXVIII, Am. Mus. 520. The nasals of the type
skull (Nat. Mus. 4251) are much shorter (81 mm.)
than in Menodus giganteus and relatively as broad
(133 mm.). The rugose expansion of the tips is an
age character. The horns are long (240 mm.), ex-
panding directly outward at the sides, with a low
but broad connecting crest; the basal section (fig.
433) is peculiar in the incurvation of the outer border
between the nasal and the malar ridges. The gen-
eral form is similar to that in M. giganteus. Seen
516
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
from in front the inferior contour of the horns is
straight, as in M. giganteus; the angles of the basal
portion are more sharply trihedral or defined than in
the type of A. crassicornis. As seen from above we
observe a decided midtemporal as well as supraciliary
overhang of the supratemporal crest, as in the type
of A. walcotti (Nat. Mus. 4260). The supratemporal
crests exhibit, in fact, a prominent flare or projection
a short distance behind the orbits. The occiput is
deeply excavated, but the occipital pillars are rela-
than in the large type Menodus" giganteus. The distinc-
tive anteroposteriorly compressed form of canine is well
represented in Plate CXXXII, C. The postero-
internal cusps of the premolars are much less devel-
oped than in A. crassicornis or in M. giganteus — in
fact, in p* the tetartocone is a feebly developed ridge,
and the deuterocone forms the most internal portion
of the crown; in other words, this tooth is in a primi-
tive stage of evolution. The tetartocone of p^ is a
mere spur. The hypocone of m' is an elevated cin-
m
A
B
Figure 433. — Sections and contours of skulls of AUops serotinus and A. crassicornis
A, AUops serotinus, Nat. Mus. 4231 (type); upper levels of Chadron B; stout horns directed outward (although here represented as vertical) with a well-
defined trihedral basal section (as in Menodus), nasals short and wide, zygomata relatively slender. B, A. serotinus, Nat. Mus. 4938; lower levels
of Chadron C; horns directed outward and roundly trihedral in basal section, nasals short, thick, and very wide, zygomata expanded. C, A.
crassicornis, Nat. Mus. 4289 (type); lower levels of Chadron 0; horns stout and very thick, basal section roundly trihedi'al, zygomata not much
expanded. All one-eighth natural size.
tively light and are surmounted by a simple rugose
flare. As in Brontops and Menodus the occiput is
decidedly produced behind the zygomata. As in
Menodus the zygomatic arches are relatively deep
and narrow, and the parietal vertex is also relatively
more constricted than in BrontotJierium and Megacerops.
Dentition. — -The dental characters of the type are
highly distinctive. A large single incisor persists on
each side, relatively more robust than that in A.
crassicornis. The canine crowns measure 41 millime-
ters anteroposteriorly and are thus slightly smaller
gule, contrasting with the cone which is so distinctive
a feature of Menodus.
In its premolar evolution A. serotinus is more re-
tarded than A. crassicornis.
Female .skull Nat. Mus. 3151.— The skull (PI.
CXVII, D) found not far from the' type of A. serotinus
is believed to be a female of the same species. It
differs from the type in several structural characters
which may be attributed partly to the differences of
sex, as follows: (1) It is of smaller size; (2) the horns
are more sessile, resembling those of A. marshi; (3) the
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHEEES
517
nasals are more elongate; (4) the zygomata are less
widely expanded; (5) the canines are of feebler dimen-
sions. In the grinding teeth the dental index, 50,
is remarkably high, but the actual linear measurements
of p'-m^, p'-p*, m'-m^ are the same as in the type
male skull. This is in accordance with the general
principle which we have found throughout titanotheres,
that the females while inferior in other characters
frequently present grinding teeth equal in size to
those of the males.
Transitional skull or ascending mutations. — The
American Museum skull No. 520, as seen from above
(PL CXVII, B), resembles the type skull of A. sero-
tinus in many respects. As seen from below (PI.
CXVIII) it has the true subgeneric characters of
Allops, especially in the peculiar transverse, lanceolate
form of the canine teeth and in the roimded form of
the single incisor tooth. On the other hand, it differs
from the type of A. serotinus and appears to be transi-
tional toward a higher type in the decidedly greater
Figure 434. — Coossified nasals and proximal part of
horns of Allops f serotinus?
Specimen from Cypress Hills, Saskatchewan, Canada, in tlie Ottawa
Museum, referred by Lambe to Megacerops assiniboiensis, One-
tbird natural size.
abbreviation of the nasals. In top view the horns
are seen to be intermediate between those of A.
serotinus and A. crassicornis. Similarly, while the
nasals are shorter than in A. serotinus they are nar-
rower than in A. crassicornis. The linear measure-
ments of the grinding teeth (330 mm.) agree precisely
with those of the male and female specimens of A.'
serotinus, and a still more conclusive resemblance is
seen in the decidedly retarded development of the
tetartocones, so far as one can judge from their worn
condition.
Skull Nat. Mus. 4938. — This is another skull which
combines the primitive premolar structure of A.
serotinus with the more abbreviated nasals and more
massive form of A. crassicornis. The detailed meas-
urements of the teeth in this skull agree more closely
with those of A. serotinus than with those of A. crassi-
cornis. On the other hand, in both size and shape
of the sections of the horns the skull appears to agree
more closely with A. crassicornis. Together with the
skull above described (Am. Mus. 520) this skull might
be cited to prove the existence of transitions between
the two successive stages.
Summary. — The type (Nat. Mus. 4251) and the three
other skulls provisionally referred to A. serotinus —
namely, the female (Nat. Mus. 2151), the transitional
form (Am. Mus. 520), and the second transitional
form (Nat. Mus. 4938) — all agree in the retarded
state of evolution of the tetartocones, which are little
further advanced than in the species Brontops dispar.
The characters m which they disagree with B. dispar
are found in the horns and nasals, and these either
represent ascending mutations of the A. serotinus type
or more probably progressive stages toward the next
higher species, A. crassicornis.
Allops crassicornis Marsh
{"Allops crassicornis" Osborn, 1902)
Plates CXIX-CXXII; text figures 189, 387, 409, 433, 608, 612
[For original description and type references see p. 228. For skeletal characters
see p. 6791
Geologic horizon. — Titanoiherium zone of South
Dakota, middle or upper level (B or C).
Specific characters. — Skull proportions of males
more robust than in A. serotinus. Skull length 750
millimeters, width 570 (estimated). P'-m^ 370
millimeters. Nasals broad and abbreviate, horns
massive and obtuse. A small pair of lateral incisors.
Canines as in A. serotinus. Superior premolars with
distinct tetartocones and well-defined external cin-
gula. Tetartocone on p' distinct but somewhat
smaller than the deuterocone. Hypocone wanting
on m', replaced by rudimentary metaloph. Zygo-
mata widely arched outward, with buccal expansions.
General characters. — The type (Nat. Mus. 4289) of
this species resembles that of A. serotinus (1) in the
peculiar lanceolate form of the canines, which are elon-
gate, compressed anteroposteriorly, and flattened on
the posterior face, perhaps a further development of
a condition seen in A. marshi; (2) in the marked
development of the cingulum on the premolars;
(3) in the proportions of the premolars; (4) in the
trihedral basal horn section. The type differs dis-
tinctly from A. serotinus (1) in the far greater develop-
ment of the tetartocones; (2) in the larger and more
robust proportions of the skull; (3) in the short
obtuse horns. Its structural character, like that of
A. serotinus, is intermediate between those of the
Brontops and Menodus main phyla, and in its extreme
size it perhaps presents an advanced stage of evolution
of the subgenus Allops — that is, of the Allops col-
lateral phylum — in the same manner that Diplo-
clonus amplus presents an extreme evolution of its
collateral phylum. It is noteworthy that this skull
differs from all the preceding male and female skulls
518
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
referred to the subgenus AUops in the following
points :
1. Greater skull length, 750 millimeters; others 640 to 730.
2. Greater length of grinding series, 370; others 310 to 335.
3. Greater length of premolars, 155; others 131 to 140.
4. Greater length of molars, 220; others 192 to 215.
5. Greater length of canine; anterior face 45; others 33
to 40.
This uniform increase in size in all measurements
implies that A. crassicornis comes from a higher
geologic level than any of the other forms and is in
thorough accord with the advanced condition of the
tetartocones.
Geologic and geographic distribution. — The geologic
level recorded for this specimen by its discoverer,
Hatcher, is the upper part of the middle beds; in other
words, it is the same as that assigned to the type and
the female of A. serotinus. A marked progression in
the evolution of the premolars is, however, posi-
tive proof that this animal belongs to a very much
higher geologic level than A. serotinus, though there
is little doubt as to its phyletic descent from that
species.
Materials. — Four skulls m the National Museum
(Nos. 4289 (type), 2117, 4709, 8740) are referred to
this species.
Description oj type. — The type skull (Nat. Mus.
4289) is undoubtedly that of an old male. The nasals
are square and heavy, considerably longer than in
Brontops rohustus but much shorter than in Menodus
giganteus. The horns are set very wide apart, and
it is a noteworthy fact that they entirely lack the con-
necting crest which is so distinctly developed in speci-
mens of Menodus. The basal section reveals the
affinity to A. serotinus; it is distinctly trihedral, and the
long axis is oblique rather than transverse as in Bron-
tops rohustus. In proportions this skull is dolicho-
cephalic; ratio of length(750 mm.) and width (570 mm.)
gives a zygomatic index of 75; the measurements are
rendered less exact, however, by the artificial spreading
of the left side of the type skull. The proportions are
intermediate between the mesaticephalic and the
dolichocephalic types.
Dentition. — We are first impressed by the reduced
size of the superior lateral incisors (PI. CXXI), which
are more vestigial than in any member of the true
Brontops series. The canines are a very distinctive
progressive development of the A. serotinus form,
attaining a length of 45 millimeters. The fourth
supei'ior premolar has a low but sharp and distinct
tetartocone. The second and thii'd pi'emolars also
have well-defined tetartocones. The third superior
molar has a peculiar sharp elevation of the cingulum
but no hypocone. The premolars and molars either
parallel or indicate their affinity to those of Menodus
in the decided development of the external cingula;
P^i P') P^ also exhibit strong external cingula.
THE MENODUS MONOPHYLUM
(" Titanotherium phylum," Osborn, 1902; " Symborodon" torvus
(jaw), Menodus, Menops, Diconodon)
Like Brontotherium, the genus Menodus includes a
nearly continuous series of ascending mutations from
the base to the summit of the Titanotherium zone,
which may be considered nearly if not quite m'ono-
phyletic. The lines of separation between so-called
"species" are wholly arbitrary.
Briefly stated, the distinctions of the Menodus
phylum as a whole are as follows: (1) Skulls dolicho-
cephalic, indices 62-70; (2) premolars rapidly pro-
gressive in molarization; (3) superior incisors aborted;
(4) horns abbreviated, triangular, not shifted forward;
(5) high dental index; (6) canines large, piercing.
Phyletic characters. — Large, long-limbed or cursorial
titanotheres, long skulled and short horned. Ranging
from the lower to the upper Titanotherium zone.
Vestigial incisor teeth and long, pointed canines.
Grinding series elongate, with a high dental index.
Crowns of grinders subhypsodont and with promi-
nent cingula.
These menodonts constitute one of the most sharply
defined of all the titanothere phyla. The elongation
of the feet and of the limbs indicates that of all the
titanotheres Ivnown they were the most rapid travelers.
While sparsely known in the lower Titanotherium
zone and somewhat more abundant in the middle
Titanotherium zone, they were relatively numerous
in the upper zone, an indication of the more favorable
conditions which this region presented toward the
end of the Titanotherium period for animals of this
type. They appear to have divided the honors with
the great long-horned brontotheres, although their
i-emains are much more rare.
The short, stout, pointed horns stood in wide con-
trast, however, with the great recurved horns of the
brontotheres. While the canine tusks may have
served as minor weapons of defense, it appears probable
that the menodonts, as the swiftest members of this
great group, had recourse to flight when attacked by
herds of carnivores.
To the anatomist the menodonts are extraordinarily
interesting in presenting extremes both of dolicho-
cephaly and of dolichopody — that is, of long-headed,
long-limbed, and long-footed development, in contrast
with the broad headed. The group affinities of these
animals with members of the Brontops phylum are in-
dicated by the less intermediate characters of the
Allops phylum. Whereas all other titanotheres ex-
hibit progressive abbreviation of the nasal bones, in
these animals the nasals retain the broad, elongate
form first seen in the upper Eocene Protitanotherium
emarginatum.
As compared with the more or less intermediate
genus Allops these true titanotheres are more dolicho-
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
519
cephalic, the zygomatic index being 62 to 70. They
are distinguished also by the extreme reduction
within the incisive alveoli of vestiges of the upper
incisor teeth, a fact which is confirmed in every speci-
men, whereas in Allops we find two small but per-
sistent upper incisors which pierce tha gum. In the
lower jaws the incisors appear to be wanting entirely,
although the evidence is less conclusive. As distin-
guished from Brontops the premolar teeth are very
directed, and adapted to lateral motions of the head in
defense or attack; (3) persistently elongate nasals,
the only phylum in which this character occurs; (4)
vigorous development of the grindiag teeth, including
a high dental index, a strong development of the cin-
gula, and a marked vertical elongation or subhypso-
donty of the crowns; (5) the elongate and piercing
character of the canine tusks, which project so far
above and below the line of the grinders that it is
A B
Figure 435. — Sections and contours of skulls of Menodus heloceras and M. trigonoceras
A, M. heloceras, Am. Mus. 6360 (type) ; small horns trihedral in basal section, parietal crest narrow, zygomata very slender. B, M. trigonoceras, Am. Mus. 6355
(lectotype or type?); pointed pyramidal horns roundly trihedral in basal section, connecting crest pronounced, nasals long and wide, parietal vertex
wide, and zygomata very slender. C, AT. trigonoceras (or giganieusf), Nat. Mus. 1219; horns longer with trihedral section, nasals very wide and distally
expanded. One-seventh natural size.
rapidly progressive in the evolution of the postero-
internal cusps or tetartocones. This character is also
shared by Allops.
Characters oj proportion. — -The six great distinctive
characters of Menodus are correlated with its length
of limb, height of body, and relatively cursorial habit,
as follows: (1) Dolichocephaly or length of skull con-
trolling all the parts both of the bones and of the teeth;
(2) short, triangular, and pointed horns, outward
101959— 29— VOL 1 36
reasonable to regard them as weapons of defense and
offense; (6) the entire absence of incisor teeth, a
feature in which Menodus parallels Megacerops and
Brontotherium. Of these sLx characters the elonga-
tion of the skull (mesaticephaly or dolichocephaly) is
the most dominant in correlation with all parts of the
skull and teeth.
The dolichocephaly of Menodus is demonstrated in
every measurement in the table below. In length
520
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
the skull almost equals the giant Brontoiherium, but
in width across the zygomata it is far inferior to this
animal; in the length of its grinding series it is far
superior to either Brontops or Brontotherium; in the
ratio of its grinding series to the entire length of the
skull it is again superior, as well as in the actual
measurements both of its premolar and of its molar
series. Its fourth premolar is actually longer and
narrower than that of Brontotherium. The sum of the
transverse measurements of its grinding teeth is
decidedly less than that in Brontops or BrontotJierium.
The table on page 523 shows that dolichocephaly, a
distinctive phyletic character of Menodus, is impressed
not only upon the skull but upon the teeth.
The horns preserve their triangular or trihedral
shape because the connecting crest as seen in the
upper view of the type of Menodus giganteus (Pi.
CXXXVIII, fig. 391) is strongly developed, as well as
the anterior ridge which extends downward into the
sides of the nasals and the external ridge, the latter
extending laterally into the antorbital bar. In the
beautifully preserved male sicull from the summit of
the Titanotherium zone in the Field Museum (PI.
CXL) the pointed apices of the horns are clearly
displayed. As observed in the side view of the same
skull (PI. CXXXIX) the facial or preorbital portion
of the skull is relatively elongate, and there is a broad
bridge across the infraorbital foramen as well as a
very deep anterior junction of the premaxillaries.
The zygomatic arches are strengthened by depth
rather than by breadth; it is true that a moderate
expansion is observed in this and other old males,
such as the type of T. "ingens," but the enormous
buccal swellings characteristic of Brontops and Bron-
totherium are not developed. In the auditory region
a highly dolichocephalic character appears — namely,
the relatively open condition of the external auditory
meatus due to the noncontact or retarded contact
of the postglenoid and post-tympanic processes.
Similarly, the occipital condyles project widely at
the back of the skull. The occiput is relatively high
and ascends rapidly above the condyles. The jaw is
also highly distinctive, with its long, straight lower
border, its backward rather than downward project-
ing angle, and its well-defined chin. It is totally
different from the jaw of Brontotherium.
In Menodus giganteus the dental series attains the
finest proportions known in any member of the order
Perissodactyla. The vigorous development of the
teeth, exhibited in the large canines, in the pronounced
cingula, in the length of the grinding series as a whole,
in the height of the crown, in the excess of anteropos-
terior over transverse measurements of the grinders,
and in the progressive dental index, affords one of the
most distinctive characters of this phylum as a whole.
Harmonic evolution. — In progressive increase in size
all parts of the skull and dentition share alike between
the stages M. heloceras and M. giganteus — namely,
about 50 per cent. The horns increase in length 190
per cent as compared with an increase of 250 per cent
during the same period in Brontotherium — that is, they
somewhat more than share the general increase in
length of the skull, but they do not dominate in de-
velopment to the detriment of other features as in
Brontotherium.
Sexual characters. — The differences between the
males and females are very marked, as clearly shown
in the contrast between two male and two female
skulls in the American Museum. The female skulls
are smaller in most of their dimensions, less rugose on
the surfaces for muscular attachment. In the males
the horns are more robust, more decidedly triangular
rather than rounded, the triangular form being sharply
defined especially at the base. In their vigorous
growth they sometimes exhibit the anterior tuberous
branching, as seen in skull Am. Mus. 505. In the
females the horns are relatively slender, with less
anteroposterior diameter at the base, and more pointed
at the summit. While in both sexes the incisors are
vestigial and certainly do not pierce the gum, there
is marked disparity in the canines. In the males
(PL CXL) the canines are formidable weapons, the
anterior enamel face of one specimen (Am. Mus. 505)
being estimated at 70 millimeters in length, whereas
in females the canines are far more slender, the an-
terior face being 45 millimeters. On the grinding
teeth the cingulum is equally robust in both sexes — •
in fact, the most pronounced development of this
structure is observed in the female skull (Am. Mus.
1067), which is evidence that development of the cin-
gulum is not a sexual character.
Progressive specific stages. — Three species are defi-
nitely distinguishable as characteristic of the lower,
middle, and upper zones respectively. There are
also two transition species and a closely connected
series of "ascending mutations." (See table on p. 523.)
Menodus heloceras (Cope) of Chadron A: The type
of M. heloceras (Cope) was at first placed as the ances-
tor of this phylum, but with some hesitation because
of its imperfect preservation. So far as known it
conforms in skull structure, although in a very prim-
itive stage of development. Its ancestral position
in the Menodus phylum was confirmed (1910) by
Granger's discovery of a second skull at the base of
the Titanotherium zone.
Menodus {" Symborodon") torvus (Cope) is an inter-
mediate stage between M. heloceras and M. trigonoceras ;
it is known only from the type jaw, which is the
genotype of the genus Symhorodon.
Menodus {"Titanotherium") proutii Leidy: Leidy
founded the genus Titanotherium on two specimens,
the first was Pomel's type of Menodus giganteus; the
second specimen, which Leidy fully described and
measured, becomes the type of his species proutii;
it belongs to an animal intermediate in size between
M. torvus and M. trigonoceras.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
521
The Menodus pTiylum as represented by the Hatcher collection of 25 sTculls of species of Menodus from the Chadron
formation in the United States National Museum
Species
M. giganteus PomeL
Do
Do
Do
Do
Do
Do.
Do.
Do.
Do.
Do.
Do.
M. (trigonooeras) giganteus
Pomel.
M. trigonoceras (Cope)
Do
Do
Do
Do
M.? trigonoceras (Cope)
M. trigonoceras (Cope)
Catalog No.
M. proutii (Owen, Nor-
wood, and Evans) .
Do
Do .
M. torvus (Cope)
M. heloceras? (Cope)
1220, cf
1212
8745, &
8741, d'
8735, V
8761, cf
8756
8751
8765, cf
8781
8778
4745, 9
8793
1219, cf
4291, cf
4707
8760, ? 9
8768
1246
4257
4701, ?
8736
113
8799
8744
Material
Skull. Old male, fuU size. Nasals typical; also horns.
Posterior part of skull.
Skull. Nasals and canines typical; also horns. Size medium (="B. ingens" stage).
Skull. Measurements typical of smaller males; not so large as type.
Skull. Slender canines.
Anterior half of skull. Measurement of p'-p* agrees with that in smaller males of M.
giganteus. Specific reference uncertain.
Skull. Poor.
Anterior half of upper milk dentition. May be large M. trigonooeras.
SkuU.
Complete left ramus and part of right.
Last half of lower jaw and molars 1, 2, and 3 of upper right.
Lower jaw.
Upper milk dentition (dp'-dp^) and m', large animal.
Anterior part of skull. An old male, approaching M. giganteus stage.
Skull. Very fine progressive type; almost at M. giganteus stage.
SkuU.
Crushed skull. Inferior specimen.
SkuU. Poorly preserved.
Posterior parts of both rami.
Skull from level B 3. Long nasals; vestigial incisor and measurements remove it from
B. dispar.
SkuU from level C 1 and left fore foot.
SkuU.
Left ramus of jaw. Type.
Pair of lower jaws.
Skull. Very young, m" just protruding; elongate proportions of Menodus.
specific reference.
Doubtful
Menodus trigonoceras (Cope), of Chadron B: It is
interesting to note that Cope's description of M.
trigonoceras immediately followed that by Marsh of
M. giganteus. M. trigonoceras, which is probably
characteristic of B, or the middle Titanotherium zone,
is distinguished from M. giganteus, first, by its inferior
dimensions throughout; second, by the shape of the
horns, the horn section being an equilateral triangle,
as the distance between the malar, nasal, and con-
necting ridges is practically equal. The skull of
M. trigonoceras is directly intermediate in size between
that of M. heloceras and M. giganteus.
Menodus giganteus Pomel, of Chadron C: In
Menodus giganteus, a superb titanothere, specifically
equivalent to Menodus {" Brontotherium") ingens
Marsh, absolutely determined as characteristic of the
upper Titanotherium zone, and even of the uppermost
levels (Chadron C 3), we reach the climax of this
monophyletic series, which is distinguished by the
extreme development of all the distinctively phyletic
characters as compared with the inferior stages. Its
relative abundance indicates that it was capable of
holding its own in the struggle for existence between
the numerous phyla of Brontops and Brontotherium.
Remains of species of Menodus have been found at
the stratigraphic levels indicated below:
Upper Titanotherium zone: M. giganteus Pomel. Skulls large
(type 755 by 553 mm.). Buccal processes of zygomata |
strongly developed. Tetartocones of p'-p* distinct. Hypo-
cone of m' usuaUy separated from cingulum. Horns large,
directed obliquely outward and upward. Connecting ridge
prominent.
Middle Titanotherium zone: M. trigonoceras (Cope). Skulls
of medium size (type 678 by 490 mm.). Buccal processes of
zygomata moderately developed. Tetartocones of p'"^
distinct. Hypocone of m^ sharp and distinct or a sharp
cingule. Horns of medium size; basal section equUateral;
connecting ridge.
Lower Titanotherium zone: M. heloceras (Cope). Skulls small
(width of type across zygomata 392 mm.; nasals to occiput
(estimated) 545 mm.). Horns smaU, subtrihedral; basal
section subtriangular; internal angle rounded; no con-
necting ridge.
Ohservations on the measurements of the Menodus
series. — The species of this genus constitute an as-
cending series extending from the lowest to the highest
levels of the Titanotherium zone. The range in the
chief measurements of the successive stages may be
epitomized as follows:
Measurements of species of Menodus, in millimeters
Pi-m3
Pi-p*
Mi-mS
Pmx to condyles.
Zygomatic index.
Horn length
385-465
141-176
246-285
760-825
62-70
150-290
333-360
127-136
203-224
628-770
66-?79
132-214
300-313
265
115-119
190-192
170
?655
603
?74
?79
70
522
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
From this it will be seen that in our present collec-
tions there are considerable gaps between the suc-
cessive stages so far as regards the measurements
p'-m^, p'-p^, and m'-m', but that the other measure-
ments overlap. The nasals in this phylum remain
very long, whereas in other phyla they are reduced in
length.
This is a remarkably consecutive and distinct
phylum. Additions to the already large number of
skulls may give a continuous series of measurements
in each column.
Geologic level. — Unfortunately, the records of the
geologic level of members of this phylum are by no
means so exact as those of members of other phyla.
No member of this phylum is certainly recorded from
the lower Titanotherium zone of South Dakota,
although the primitive characters of the oldest known
species, M. heloceras, indicate that it belongs on this
lower level. In Wyoming one skull of M. Jieloceras
(Am. Mus. 14576) has been recorded from the base
of the Titanotherium zone. Only one skull (Nat.
Mus. 4257), belonging to the species M. trigonoceras,
is definitely recorded as from the middle TitanotTierium
zone (level B). Another skull, referred to M. proutii
(Nat. Mus. 4701), is tabulated from the lower level
of C. This uncertainty ceases, however, in the upper
beds, in which we have two positive records — namely,
Nat. Mus. 4291, entered as from the upper Titano-
tTierium zone (C), and a skull in the Field Museum
(P 5927), recorded as found near the top of the upper
Titanotherium zone, or upper C. It is thus probable
that Menodus persisted to the very end of the great
titanothere epoch.
SYSTEMATIC DESCRIPTIONS OF GENERA AND SPECIES IN
THE MENODUS PHYLUM
Menodus Pomel, 1849
(Tiianoiherium Leidy , 1852; Symborodon Cope (<S. tonus, jaw
only); "Titanotherium Leidy," Osborn, 1902)
Plates XVIII, XX, XXII, XXIV, XLVII, XLIX, LXXXII,
CXXIII-CXLII, CLVII-CLIX; text figures 24, 159, 160,
166, 168, 172, 175, 181, 227, 228, 375, 378, 381, 382, 387-389,
391, 393, 394, 396, 398-400, 406, 409, 435-447, 613-619, 630,
639, 640, 642, 701, 708, 713, 715-719, 744, 746
1 For original description and type references see p. 204. For skeletal characters see
p. 6811
Generic characters. — Characters 3-9, 14, 21, 23, 24,
26 (see below) are expressions of the general tendency
to dolichocephaly.
Dentition. — (1) I|^. Incisors vestigial. (2) Ca-
nines, (? large, elongate (42-70 mm.), with slight
anterior and strong postero-internal cingulum; pos-
terior face rounded; 9 slender, small. (3) Opposite
grinding series rectilinear (not arched). (4) Upward
flexure of premolar series, as seen in side view, slight.
(5) Length of premolar-molar series about equal to
one-half that of the skull, from premaxillary tips to
occipital condyles (dental index 48 to 51). (6) Pre-
molar series long. (7) Internal cusps of grinding
teeth with sides progressively steep, ectolophs more
nearly vertical than in Brontops, external crescents of
molars more open. (8) Anteroposterior- diameter of
m^, m' usually greater than transverse diameter. (9)
Cingula present between grinders. (10) P|;f; p^
with worn crown rounded or pear-shaped in outline,
outer wall overlapped posteriorly by ectoloph of p^.
(11) Premolar tetartocones exhibiting rapid progres-
sive development. (12) Premolars with pronounced
internal cingula, edge rounded to sharp, external
cingula sharp, well marked. (13) Molars with partial
internal cingula pronounced, external cingula pro-
nounced. (14) Hypocone of m' often separate, and
surrounded by cingulum.
Skull. — (15) Skull proportions dolichocephalic. (16)
Facial portion of skull elongate, with premaxillaries
not reduced. (17) Cranial part of skull very elon-
gate. (18) Preorbital malar bridge broad to very
broad, with median (malar) ridge low and obtuse
{M. heloceras, M. trigonoceras) or weU rounded {M.
giganteus) . (19) Infraorbital foramen very conspicuous
in side view. (20) Malar below postorbital process
subflat, deep. (21) Free portion of nasals persistently
elongate, with parallel sides; but sometimes expanding
distally in old individuals. (22) Horns of small to
medium size, originating over preorbital malar ridge,
shifting forward progressively;^* basal section pro-
gressively trihedral, obliquely placed; summit of horn
round to trihedral. (23) Zygomata slightly arched;
buccal expansions slight or moderate, in section deep
rather than broad. (24) Occiput moderately pro-
duced backward behind zygomata; vertex of skull
broad posteriorly; pillars flaring slightly to moder-
ately; occiput indented; with median knobs slight or
absent. (25) Basisphenoidal rugosity usually present;
vomerine septum present. (26) Jaw deep, elongate
ramus, pronounced chin; coronoid uniformly broad,
elevated, square at summit; angle depressed or back-
ward extended.
The contrasts in proportions of the terminal mem-
bers of the Brontops, Menodus, and Brontotherium
phyla are shown in the following table:
Measurements of typical forms of Brontops, Brontotherium, and
Menodus, in millimeters
Pmx to condyles
Zygomata, transverse,
P>-m3
Dental index
Pi-p«
M'-m3
P', ap. by tr
M3, ap. by tr
Brachy-
cepbalic:
Brontops
robustus,
Yale Mus.
12048 (type)
766
667
350
46
137
220
40X65
81X92
Dolicbo-
cephabc:
M. giganteus,
Field Mus.
P6927
(typical)
Hyperbracby-
cepbalic:
Brontotherium
gigas elatum,
Am. Mus. 492
(typical)
825
515
425
51
150
270
- 58X73
"■100X78
830
740
353
42
126
241
47X72
91X99
1 Univ. Wyoming Mus.
" A surprising fact is that the horns in the unerushed M. gigantetis (Field Mus.
P 5927) are almost as far forward as in the unerushed B. gigas. In the unerushed
B. platyceras and B. ramosum the forward displacement appears extreme.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
The dolichocephalic character of Menodus in its
typical species M. giganteus is demonstrated by every
measurement of the accompanying table, in contrast
with the brachycephalic Brontops rohustus and
hyperbrachycephalic Brontotherium gigas.
523
the
Standard measurements of slcuU, jaw, and teeth in the Menodus phylum, in millimeters
Upper teeth
Skull
Jaw and teeth
!
&
1
S
o
1
l|
i°
o
1
i
g
S
1
t
o
1
1
1
1
1
1
a
P4
a
f
S
>
s
.g
§
o
II
g|
■-3
1
■a
8
1
1
M. giganteus group
M. giganteus, Univ. Wyo., cf
465
176
285
60
35
815
M. giganteus Pomel (type)
"280
M. giganteus ("ingens"), Yale Mus.
12010, cT . .....
428
162
266
553
755
755
750
695
712
125
150
175
105
120
125
140
140
145
M. giganteus, Nat. Mus. 1220, cf
798
825
760
777
800
!'805
770
178
290
M. giganteus, Field Mus. P 5927, cf .._
"Menops" varians, Yale Mus. 12060, cf
425
410
393
390
"■385
360
150
155
153
145
'■141
135
270
255
250
245
"246
223
52
"■70
33
35
31
515
555
545
590
62
73
70
400
125
280
47
690
M. giganteus, Am. Mus. 505, cf (neotype)
195
M. giganteus, Nat. Mus. 8741, cf -
M. giganteus. Am. Mus. 506, 9
»40
45
20
27
"770
738
122
"115
404
347
364
109
115
"260
238
242
"651
M. proutii-trigonoceras group
M. giganteus (trigonoceras) , Am. Mus.
1066.-
546
70
592
M. giganteus (trigonoceras), Am. Mus.
1007 -
590
M. giganteus (trigonoceras), Nat. Mus.
4291, c?
360
355
136
133
224
225
47
44
26
25
770
725
"510
'440
490
66
60
"720
"670
130
127
137
"125
'135
135
145
132
150
214
132
M. giganteus (trigonoceras), Am. Mus.
1067, 9
360
126
237
39
— -
610
M. trigonoceras, Am. Mus. 6355 (lecto-
M. trigonoceras, Munich Mus. (mounted
"355
345
"690
M. trigonoceras. Am. Mus. 6356 (cotype)_
129
220
— -
26
140
356
"120
245
32
M. trigonoceras, Nat. Mus. 1219
M. trigonoceras, Nat. Mus. 4257
M. trigonoceras, Carnegie Mus. 3068, 9 _,
M. proutii, Am. Mus. 9335
345
335
333
136
127
221
220
203
192
42
33
39
'■21
22
115
120
120
190
150
670
628
535?
79?
655
313 119
M. proutii, Nat. Mus. 8736
680
335
234
M. proutii, Nat. Mus. 4701, 9
300
115
190
32
----
"655
485
»74
....
120
125
M. ("Symborodon") torvus, Am. Mus.
6365 (type). .
310
100
210
"535
M. heloceras
M. heloceras. Am. Mus. 6360 (type)
392
"480
7
"■79
"545
"620
M. heloceras, Am. Mus. 14576
"265
75
170
67
"603
37
132
33
— -
70
30
Percentage of increment from M. helo-
ceras to M. giganteus
1
524 TITANOTHEKES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Measurements of sTcuUs and jaws associated with and referred to the Menodus phylum, in millimeters
Skulls and upper teeth
'Lower jaws and teeth
Interior
canine to
hypoconu-
lid of ma
pi-ni3
Mi-ms
Symphy-
sis to
glenoid
Posterior
canine to
hypoconu-
lid of ms
Pi-ma
Mi-in3
Symphy-
sis to
condyles
Depth
below ma
M . giganteus Pomel (type) _
»280
-260
262
237
242
245
238
-651
625
610
590
172
M. giganteus. Am. Mus. 506, ?_
"404
"385
"246
"■ess
'387
385
369
-404
393
360
364
356
347
176
M. giganteus, Nat. Mus. 4735 (?4745)
137
M. giganteus (trigonoceras), Am. Mus.
1067, 9
M. giganteus (trigonoceras). Am. Mus. 1007
370
355
225
543
144
160
M. giganteus (trigonoceras). Am. Mus.
1066, 9 . --_ ._- - -
360
345
223
220
592
M. trigonoceras, Am. Mus. 6356 (cotype)..
M proutii, Nat. Mus. 113 (type)
370
■■514
234
"227
«214
''208
»515
°«535
i 522
163
M. trigonoceras. Am. Mus. 6345b, ? 9
"323
= 312
<i 295
"■323
' 316
<i261
120
M. ("Symborodon") torvus, Am. Mus.
6365 (type)
124
M. ("Symborodon") torvus, Am. Mus.
6365 (type)
124
290
-265
170
443
» Estimated.
Menodus heloceras (Cope)
{Megaceratops helocerus Cope, 1873; "Titanotherium helocerus"
Osborn, 1902)
Plates XLVII, C XXIII, CXXIV; text figures 168, 378, 399,
409, 435, 436, 613, 639
[For original description and type references see p. 212. For skeletal characters see
p. 681]
Type locality and geologic horizon. — Cedar Creek,
Logan County, Colo.; lower Titanotherium zone.
Specific characters. — Skull small, premaxillaries to
condyles estimated about 603 millimeters, tip of
nasals to top of occiput about 545. Dolichocephalic
(index unknown), width across zygomata 392-425
millimeters. Horns small, "elevation 50 millimeters"
(Cope), basal section trihedral (with flattened, subequal
external, internal, and posterior faces), tip rounded.
Horns above antorbital malar ridge. Nasals squared
distally, free length and breadth equal (100 by 100
mm. fide Cope). Buccal swelling of zygoma incipient.
Premolar-molar series 265 milhmeters (estimated).
Materials. — The type skull was found on Cedar
Creek, Logan County, Colo., the same geographic
region which yielded the various Symborodon skulls.
The characters based upon the very imperfect type
skull (Am. Mus. 6360) are supplemented by those of
the much more perfectly preserved skull (Am. Mus,
14576) found in Wyoming; the latter is in a very
primitive stage of development (fig. 436), especially
in regard to the small size of the molars, m'-m^,
which measure 170 millimeters.
' Eight side . ■< Left side.
Description of the type. — This is the smallest and
apparently also the most primitive member of the
Menodus phylum; it approaches even the Eocene
titanotheres in its slender, narrow occiput (Pis.
CXXIII, CXXIV) and its slender zygomata. In
lateral view we note the weU-marked postorbital
processes of the frontals and the open condition of the
auditory meatus. All these characters point toward
affinity with Menodus. The specimen, as indicated
in Plate CXXIV, A\ is in fragmentary condition,
the facial region being largely wanting. The nasals
have been lost, but as measured by Cope they were
equal in length and breadth (100 by 100 mm.), a point
of significance in the consideration of the affinities of
thi.s animal because in M. trigonoceras and in the type
of M. giganteus the breadth also equals the length,
whereas in all stages of Brontops the breadth exceeds
the length. Cope's other measurements are as
follows :
Measurements of Menodus heloceras
Millimeters
Least width of paiietal plane 104
Elevation of horn core 50
Length of free nasal bones 100
Width of free nasal bones at base 100
The rudimentary condition of the horns also indi-
cates that this animal belongs in the lower Titano-
therium zone. The horn section (figs. 399, 435) con-
sists of the relatively acute anterior angle, an external
angle, and a gently rounded internal angle, the last
attributable to the absence of any connectmg crest
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
525
between the horns; the horn m its sub trihedral shape
and the position of its long axis is similar to that in
Menodus giganteus and is quite distinct from that of
the early members of either the Brontops or the Bronto-
therium phylum. Similarly the zygomatic section is
relatively deep and narrow, agreeing again with that
of M. trigonoceras. In consonance with dolichoce-
phaly, the postglenoid, post-tympanic, and paroc-
cipital processes are relatively short and simple. We
observe a well-defined postmastoid foramen, as in
other titanotheres. The posterior nares open between
the opposite m'. The basisphenoid is smooth. Un-
fortunately the teeth are so fractured that they yield
only one important fact — namely, that this is a very
old individual, and that its simple characters are not
juvenile but specific.
Measurements of two specimens of Menodus heloceras, in
millimeters
Basilar length
Breadth, zygomata
Height, temporal fossa.
Width, parietal plane..
Free length, nasals
Free width, nasals
Am. Mus. 6360
(type)
(Colorado)
395
105-
104
100
100
Am.Mus.l45"6
(Wyoming)
603
450
110
114
132
111
<• Estimated.
Slcull referred to Menodus Jieloeeras, from the hase of
the Titanotherium zone, Wyoming. — The type skull is
so poorly preserved that its measurements yield little
of interest. A skull which is provisionally referred
to this species (Am. Mus. 14576, fig. 436) comes from
the very base of the Titanotherium zone at Beaver
Divide, near Hailey, Wyo. It exhibits the following
comparative measurements:
Measurements of Menodus heloceras, Alloys walcotti, and
Brontops brachycephalus, in millimeters
M. heloceras,
Am. Mus.
14676
A. walcotti,
Nat. Mus.
4260 (type)
B, brachycephalus
Nat. Mus.
4940
Am. Mus.
1495
F-m3
'•265
285
112 +
169
640
105
100
265
101
160
288
pi-p«
118
M'-m^ .-
170
0 603
132
70
171
Pmx to condyles
Nasals, free length
Horn length
»560
102
Hence the skull referred to M. heloceras is distin-
guished by small molars, long nasals, and short horns,
which are all primitive characters. The form of the
horns and nasals suggests ancestral relationship to
Menodus or Allops.
The upper molars, though poorly preserved, appear
to resemble those of Brontops hrachycephalus and differ
from those of Menodus heloceras in bekig relatively
wider.
Menodus torvus (Cope)
(Symhorodon torvus Cope, 1873; not "Symborodon torvus Cope,"
Osborn, 1902)
Plates CXXV, CXXVII, CXXXII, CLVIII, CLIX; text figures
166, 396, 4.37
[For original description and type reierences see p. 210]
Type locality and geologic horizon. — Northeastern
Colorado, Titanotherium zone.
Specific characters. — Lower jaw of Menodus type,
slightly -smaller than that of M. proutii; pi-ms, 310
millimeters; premolars much crowded (pi-p4, 100
mm.); incisive border apparently edentulous; an ex-
ternal cingulum on the canines, premolars, and molars;
Pi very small and closely crowded between the canine
and P2.
The jaw (Am. Mus. 6365), chiefly described by
Cope, is the lectotype of the species "Symhorodon"
torvus, and this species is the genotype of his genus
Symborodon. This jaw proves to belong to the Meno-
dus phylum and to be unrelated to the "Symborodon"
of previous descriptions.
The measurements of Am. Mus. 6365 given by
Cope are more or less inaccurate but serve to identify
the type. The following may be substituted:
Measurements of Menodus torvus and M. trigonoceras, in
millimeters
M.(" Symborodon")
torvus, Am. Mus.
6365 aectotype)
M .(" Symborodon ")
trigonoceras, Am.
Mus. 6345, 9°
Angle to front of canine
545
555
Depth of angle below condyle.
225
263
Length of symphysis
155
142
Depth of ramus immediately
behind ms
124
'120
Molar-premolar series
310
'333
P2-P4 . ..
100
210
100
'233
« Associated by Cope with type skull of S. bucco. ' Estimated.
Measurements of premolars and molars in Menodus torvus {type) ,
in millimeters
P2
P3
Pi
M,
M2
Ms
27
20
37
27
43
30
51
36
67
44
97
43
Type lower jaw of M. torvus. — The ramus ascends
with a gentle curvature to the incisive border, which
is absolutely devoid of any traces of teeth or alveoli,
being thin and smooth with a small pit on its lower
surface (Pis. CXXV, B; CXXVII; fig. 437). The
incisor formula is therefore highly distinctive: !<>.
The canines were evidently small; the fangs measure
23 millimeters anteroposteriorly, 20 transversely; the
526
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
crowns are at present broken away but could not
have projected much above the level of the other
teeth. There is absolutely no space for the first
premolar; the second is placed directly behind the
canine and has an elevated anterior, noncrescentic
lobe and a posterior crescent; the third and fourth
premolars exhibit both anterior and posterior cres-
Menodus proutii (Leidy)
{Titanotherium proutii Leidy, 1852)
Plates LXXXII, CXXV, CXXVI; text figures 160, 409, 438
(For original description and type references see p. 205]
Geologic Tiorizon. — Precise geologic level unknown,
probably middle level of the Titanotherium zone,
Chadron A 3 or B 1.
Figure 436. — Skull of Menodus heloceras
, Side view; As, top view; Aj, palatal view. One-sixth natural size. Am. Mus. 14576; found at the very base of the Titanotherium
zone immediately overlying beds probably equivalent to Uinta C, Beaver Divide, near Hailey, Wyo. This very primitive skull
agrees with the type of Menodus heloceras in the trihedral basal horn section, elongate proportions, and slender zygomata. It is
also structurally allied to Allops walcotti and Brontops brackycephalus. The upper molars, though poorly preserved, resemble those
of Brontops brackycephalus rather than those of Menodus proutii.
cents, no internal cingula, faint external cingula. The
vertical arching of the premolar series is very pro-
nounced. The molars also are devoid of internal
cingula but present more or less complete external
cingula.
Specific cJiaracters. — Size of type jaw slightly
smaller (mi-nis 234 mm.) than that of jaws referred to
M. trigonoceras. Referred upper teeth (p'-m^ 313
mm.) also smaller than in M. trigonoceras. Upper
and lower premolar-molar series with cingula.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
527
Characters of the type (Jedotype) lower jaw. — The
reasons for regarding this lower jaw fragment (Nat.
Mus. 113) as the type, or lectotype, of this species
are given on page 210 of this monograph. The speci-
men consists of part of the left ramus of the lower
jaw containing the posterior alveolus of ps, both
alveoli of p4, and the three molars, which are much
worn and somewhat damaged in certain places. The
molars exhibit the strongly de-
veloped external cingula that are
characteristic of the true Menodus,
and the measurements show that
we have here a small member of
the Menodus phylum, distinctly
larger than the very primitive
M. heloceras but somewhat smaller
than the typical M. trigonoceras of
the middle Titanotherium zone.
Measurements oj Menodus
proutii. — In the type lower jaw of
this species the true molar series
(234 mm.) is about 5 to 10 milli-
meters shorter than in jaws re-
ferred to M. trigonoceras. The
specific name proutii is accordingly
applied to the smaller members of
the "trigonoceras" group. If we
adopt the trinomial system of
nomenclature the specific name
proutii might include M. proutii
proutii for the smaller skulls, and
M. proutii trigonoceras for the
larger skulls.
The molars of the type are 20
millimeters longer than in the type
of Symhorodon torvus Cope, which
is also a menodont, and the jaw as
a whole was considerably larger
than those referred to AUops
marshi.
In a National Museum skull
(No. 4701, ? ) referred to M.
proutii the premolar-molar series
is 45 millimeters shorter than in
the typical M. trigonoceras, the basilar length is also
much shorter, and the zygomatic index (74, estimated)
is relatively high.
Description of the neotype of M. proutii. — The upper
jaw selected by Osborn as a neotype is a specimen
(Am. Mus. 9335) from Lance Creek, Wyo., recorded
from the middle Titanotherium zone.
The upper molars of the neotype are of conformable
size with the lower molars of the type. In fact, the
well-worn lower premolar-molar series of the type
jaw has been found to articulate fairly closely with
the less worn upper premolar-molar series of the
neotype maxilla (fig. 438). The neotype upper grind-
ers exhibit the characteristic internal and external
cingula of Menodus, lofty ectolophs, tetartocones
strongly developed on p^, p^, and p*, a prominent tri-
angular hypocone on m', and molar crowns distinctly
elongated anteroposteriorly; they also exhibit rudi-
ments of the crochet and antecrochet.
Figure 437. — Lower jaws of Menodus {Symhorodon) torvus and M. trigonoceras
Menodus torvus, Am. Mus. 6366 (type) ; a very aged animal with angle very prominent and truncate, cheek teeth
with sharply defined external cingula, and incisive border edentulous as in Menodus. B, M. tngonoceras, Nat.
Mus. 4745; a very characteristic jaw, showing rather full symphyseal region, angle produced downward and
backward, canines and cheek teeth sharply cingulate and subhypsodont. One-fifth natural size.
Additional measurements of Am. Mus. 9335 (neo-
type of M. proutii) are given below.
Millimeters
P2, ap. bytr 26X28
Millimeters
Canine, vertical (esti-
mated) 39
Canine, anteroposterior.- 24
P'-m3 313
pi-p4 119
M'-m' 192
P^ap. bytr 34X41
F\ ap. by tr 38X49
Ml, ap. by tr 59X04
M2, ap. bytr 69X62
M3, ap. bytr 69X67
Dolichocephaly is strongly marked in m^~'. The
measurements show that this specimen is smaller
than Carnegie Mus. 3068.
528
TITANOTHERES OF AJSTCIENT WYOMING, DAKOTA, AND NEBRASKA
Transitional skulls jrom M. proutii to M. trigono-
ceras. — A skull (Carnegie Mus. 3068 or 558?) assigned
to M. trigonoceras proutii appears to be distinguished
subspecifically from the more progressive stage M.
trigonoceras by the retarded development of the
tetartocones of p*, by the more primitive size and
section of the horns, by the hypsodont characters of the
external cusps of the molars, and by the correspond-
ingly deep fossettes. This apparently occupies an
intermediate position between the M. Tieloceras of the
lower beds, M. torvus, and the M. trigonoceras of the
middle beds. The sex is apparently female. To
this may be attributed the fact that the horns are
somewhat more primitive, elongate-oval in section.
The skull is that of a young animal in the sixth stage
of growth, the protocone of m' being barely worn.
As shown in the comparative measurements, the
superior grinding teeth are of strikingly dolichocephalic
type, the measurement p'-m^, inclusive, being 320
millimeters.
Figure 438. — Upper teeth of Menodus proutii
Am. Mus. 9335 (neotype); Chadron B, Lance Creek, Wyo.; exhibits the generic characters of Menodus.
Canines conical with sharply defined cingulum, both external and internal cingula of the premolars and
external cingulum of the molars sharply defined, molars elongate anteroposteriorly. Premolar pattern the
same as in Brorttops hrachycepJialus. The measurements of these teeth indicate reference to Menodus
prfmtii. One-third natural size.
In this skull, as in dolichocephalic skulls generally,
the bridge over the infraorbital foramen is broad and
flat, and the external auditory meatus is widely open
below. A wide space separates the postglenoid and
paroccipital processes. The dolichocephaly of the
dental series is also shown in the spacing of the teeth,
which is so considerable that there is a slight diastema
between the first superior premolar and the canine;
cingula appear between the grinding teeth. P' is a
large, actively functional tooth, elongate antero-
posteriorly. The canines have the highly charac-
teristic form of Menodus (Titanotherium) , being long
and pointed, with anterior and posterior cingula; at the
same time, their slender section indicates that they
belong to a female. Other distinctively Menodus
characters are the crenulate internal faces of the
deuterocones and the crenulate and strongly developed
external and internal cingula with sharp edges. In
the grinding teeth the hypocones are very prominent,
and both the internal cones and the external crescents
exhibit deep slopes; this elongate or subhypsodont
character of the grinders, distinctive of all true
Menodus teeth, is very striking.
Menodus trigonoceras (Cope)
{Symborodon trigonoceras Cope, 1873; "Titanotherium trigono-
ceras" Osborn, 1902)
Plates XX, XXII, CXXVIII-CXXXII, CXXXIV, CXXXV,
CLVII; text figures 172, 378, 382, 396, 399, 409, 435, 437, 439,
440, 613-616, 630, 639
[For original description and type references see p. 213. For skeletal characters see
p. 683]
Type locality and geologic Tiorizon. — Titanotherium
zone, level Chadron B.
Specific characters. — Skull larger than in M. Tielo-
ceras. Premaxillaries to condyles 670 to 725 milli-
meters, tip of nasals to top of occiput 670 to 738.
Mesaticephalic, width across zygomata 485 to
535 (?) millimeters. Index 74 (?)'. Horns, out-
side length 132 to 190 millimeters, basal section
triangular, tips pyramidal; horns a little in front of
preorbital malar ridge. Nasals squared distally,
free length 115 to 140 millimeters,
breadth 125 to 135 (?). Buccal
swelling of zygoma more decided.
Premolar-molar series 333 to 360
millimeters, premolars 127 to 136,
molars 203 to 225, canines, anterior
d', 42 to 47. Premolar tetartocones
more progressive, tetartocone of p"*
more pronounced.
This animal is directly successive
to Menodus proutii and connects
this species by a series of "ascending
mutations" with M. giganteus.
on the measurements of Menodus
trigonoceras. — In the typical skulls the premolar-
molar series is 83 millimeters shorter than in the type
of M. {" Brontotherium") ingens. Several skulls (Am.
Mus. 1066, 1067, 1007; Nat. Mus. 4291) which were
formerly referred to M. giganteus have the premolar-
molar series over 60 millimeters shorter than in the
type of M. giganteus and appear to belong rather with
M. trigonoceras. Between the largest skull now
referred to M. trigonoceras (Am. Mus. 1066) and the
smallest referred to 31. giganteus there is a marked
difference in the length of the true molar series (23
mm.). Hence in our collections at present, with
reference to the longitudinal dimensions of the
grinding teeth, M. trigonoceras does not quite overlap
M. giganteus but is separated by a small but distinct
interval.
M. trigonoceras may be distinguished from Allops
serotinus by the greater relative anteroposterior
measurements of the true molars, by the lesser width
of p*, and by the great length of the nasals, as follows :
Observations
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
529
Measurements of Menodus trigonoceras and Allops serotinus, in
millimeters
Pi-m3
Pi-p<
Mi-m^
Pmx to condyles
Zygomatic index
Nasal length
P*, ap. by tr
M^ ap. by tr
M. trigono-
ceras, Nat.
Mus. 4291, d
A. serotinus.
Am. Mus.
620
360
330
136
133
224
206
770
720
66
78
130
68
"41X55
43X65
<■ 79X73
78X82
" Am. Mus. 6356.
Materials. — This important intermediate stage is
represented by the type and paratype skulls in the
American Museum (Nos. 6355 and 6356, Cope collec-
tion) ; also by two skulls in the National Museum
(Nos. 4257, 1219). Nat. Mus. 4257 is definitely
recorded by Hatcher as from the upper level of B, the
middle Titanotherium
zone, a fact of extreme
importance, as helping
to determine the geo-
logic level of this spe-
cies. Another skull
(Nat. Mus. 4701, ? )
agrees in measurement
with M. proutii but is
recorded from the base
of C; it thus may be
a female of M. trigono-
ceras. There is no jaw
positively associated
with M. trigonoceras;
a referred jaw is Am.
Mus. 1007, which is transitional to M. giganteus in
its measurements.
If measurements are reliable as indications of specific
affinity we should have to add to this species the skulls
Am. Mus. 1066, cr', 1067, ? , and the jaw Am. Mus. 1007,
which were formerly included under M. giganteus.^
Characters oj the type and paratype of M. trigono-
ceras.— Cope's type (Am. Mus. 6355) fortunately is a
skull in which all the superior portions are fairly pre-
served, as shown in Plates CXXVIII and CXXIX;
the principal missing parts are the premaxdlaries, maxil-
laries, and occipital condyles. A single molar tooth,
m^, shows that this type skull is young or in an early
stage of growth, while the paratype skull (Am. Mus.
6356) is entering the seventh stage of growth.
This disparity of age explains some differences be-
tween these two specimens, especially in the zygomatic
arches, which are wholly attributable to growth or age.
The type skull (No. 6355) belongs to a young indi-
vidual; the nasals are long and rather thin, with
nearly parallel sides, truncate distally. The horns are
short, widely divergent, with trihedral section and a
well-defined intermediate connecting crest, which is
placed at the back part of the horn. The cranial
vertex is of moderate width and extends backward
into a slightly expanded occiput with lateral pillars,
extending into rugose summits; the zygomatic arches
are slender in vertical section, but they show marked
expansion. The third superior molar bears a small,
very sharply defined hypocone, a distinctive phyletic
character of this series.
The paratype skull consists of the anterior portion
(Am. Mus. 6356) of a fully adult individual in the
seventh stage of growth. The more advanced age
presents us with the adult characters of this species,
especially the thickening and rugose expansion of the
nasals distally, the thickening and moderate expansion
of the zygomatic arches, the well-developed preorbital
and postorbital processes.
Dentition. — The specific identification of the para-
type with the type is rendered more positive by the
Figure 439. — Skull of Menodus trigonoceras
Side view. Tiiis skull (now in the Munich Museum, formerly Am. Mus. 1066) is vertically crushed, so that the nasals are tilted
upward, the horns crushed downward, and the vertical diameters of the orbit and parietal region lessened. One-sixth natural size.
close similarity in the structure" of the third superior
molar in both skulls. This tooth has a sharp and dis-
tinct hypocone. Other dental characters have weighty
specific value (Pis. CXXXI, CXXXII). The techni-
cal formula, I^, C^, P^, M^, fails to convey an idea of
the vestigial character of the incisor teeth, the crowns
of which barely break through the bone of the incisive
border and certainly did not cut through the gum;
thus the premaxilla of this animal should be described
as functionally edentulous. The sex of the paratype is
apparently female, the canine being rather slender and
elongate. The maxillary teeth are rather progressive
in development; the first premolar exhibits a rudi-
mentary tetartocone, or four cusps in all. The tetarto-
cones on p^, p'' are well developed but still much
inferior in size to the deuterocones; the premolars are
further distinguished by pronounced internal and
external cingula. Throughout the premolar-molar
series we observe a deep pit in the midvalley, just
internal to the ectoloph, which is homologous with the
530
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
medifossette of the rhinoceros and horse molar. Just
internal to this pit two ridges enter the median valley.
These are apparently homologous with the "crochet"
and " antecrochet " of the rhinoceros and horse molars.
The molar teeth also exhibit well-developed external
cingula; the internal cingula, as in all other titano-
theres, are incomplete opposite the protocones.
Skulls of M. trigonoceras. — Of the six skulls in the
National Museum collection referred to Menodus
™
ri
.-^-^
Menodus Mffonoceras ^
Col. /fus., Denver- fl
k
jg^' ^^^B
',
d
HM^^^gtt^^c^
SlxJ
p
f
Vlv^
lj
. ,|j! ,■
l^s
r%
fi^
Figure 440. — Skull of Menodus trigonoceras
In the Colorado Museum, Denver. The nasals and horns were detached when
the skull was found, but there is no question as to the association.
trigonoceras one (Nat. Mus. 4257) is recorded from the
upper levels of the middle Titanotherium zone. It
belongs to a small male in the seventh stage of growth.
The anterior measurement of the canines is 42 milli-
meters (estimated). Two extremely small vestigial
incisors persist upon one premaxUla only. In the
first superior premolar the tetartocone is connected
by a confluent crest with the deuterocone, as in other
species of this genus. The tetartocone cusp on p^,
p^ is very distinct, especially in p''.
From the vertex to the tips of the nasals the skull
measures 665 millimeters; the free length and the free
width of the nasals are nearly equal, namely, 120 by
125 millimeters. The horns measure 150 millimeters
from the edge of the anterior nares to the tips. On the
inferior surface of the skull the vomer extends back
from the presphenoid as a distinct narrow keel.
In another skull, an old male (Nat. Mus. 1219), as in
the more advanced or eighth stage of growth, the horns
measure 195 millimeters as compared with 150 in the
specimen just described. This is a natural growth
increase. It is important to note that the nasals are
practically the same length as in the younger speci-
mens, namely, 115 millimeters; but they are greatly
increased in breadth (160 mm.), also in the rugose
expansion of their tips, which support a pair of
lateral tuberosities. This animal is evidently a very
old male, and its senescent character is emphasized by
the entire disappearance of the vestigial upper incisors.
It is also a very large individual. The premolar-molar
series measures 345 millimeters.
In contrast with the two skulls described above
there is a third skull (Nat. Mus. 4701) belonging to
a young female which is recorded by Hatcher from
level C 1 of the upper Titanotherium zone. The
premolar-molar series measures 300 millimeters. The
female sex is indicated by the small, pointed ca-
nines (34 mm.). Inconsistently with its high geo-
logic level as recorded, the skull exhibits a relatively
low stage of development in its premolar teeth, the
tetartocones being less clearly separated off than in the
old male above described. Although the animal is still
young, the incisors have entirely disappeared, as well
as all traces of their alveoli. The internal cingulum
of the premolars is quite as prominent in this female
as in the males, again proving that the cingulum is
not a sexual character. There is a large hypocone on
the last superior molar, which thus appears to be a
specific if not a generic character.
The Ottawa Museum skull jrom the lower Oligocene
of Swift Current River, Assinihoia, Canada. — The frag-
mentary cranium described and figured by Cope
(1891.2, p. 10, pi. 6) as "Menodus americanus" very
probably belongs to M. trigonoceras.
Menodus giganteus Pomel
(" Brontotherium" ingens Marsh, 1873; "Titanotherium ingens"
Osborn, 1902)
Plates XVIII, XX, XXIV, XLVII, XLIX, CXXXIII,
CXXXV-CXLI; text figures 24, 159, 227, 228, 375, 381,
387-389, 391, 393, 394, 396, 399, 400, 406, 409, 441-446, 617-
619, 630, 640, 642, 701, 713, 715-719, 744
[For original description and type references see p. 204. For skeletal characters
see p. 687]
Geologic horizon. — Upper Titanotherium zone.
Specific characters. — Skull very large, premaxillaries
to condyles 770-825 millimeters, tip of nasals to top
of occiput 712-755+ . Dolichocephalic, width across
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHEEES
531
zygomata 515-553 millimeters, index 62-70. Horns,
outside length 150-290 millimeters; basal section tri-
angular, tips pointed; horns much in front of antorbital
malar ridge. Buccal swelling of zygoma about as in
M. trigonoceras (that is, moderate). Premolar-molar
series, 360-428 millimeters; premolars 133-162, molars
223-270; dental index 51 (typical). Premolar tetar-
tocones very progressive but still connected with
deuterocones by a narrow strip, tetartocone of p* very
pronounced. Canines c? very long (70 mm.).
Menodus giganteus was certainly a contemporary of
its long-horned rival Brontotherium gigas and probably
extended up to the period of the still more specialized
Brontotherium platyceras. While far inferior in length
of horn to those brontotheres, it was certainly superior
to them in the possession of a superb series of grinding
teeth with connecting crowns and more elevated crests
and cones.
The sexual characters of M. giganteus are very
clearly brought out by comparison of two male skulls,
Nos. 1066 and 505, and two fe-
male skulls with jaws, Nos. 1067
and 506, in the American Mu-
seum collection. A fine skull in
the National Museum (No. 1220)
also belongs to a male. The
skull Nat. Mus. 4291 is recorded
from the highest level of the
upper TitanotJierium zone. The '
male skulls exceed the female in
the width of the zygomatic
arches, thus proving that, as in
the brontotheres, the buccal proc-
esses are partly secondary sexual
characters. The comparatively
slender, pointed horns, and
pointed canine teeth of the fe-
males are well exhibited in
Plate CXXXV, B. This figure
shows incidentally the extreme
effects of lateral crushing as con-
trasted with the extreme effects
of vertical crushing and deformation of the skull.
The type specimen of the species "B. ingens" was
erroneously employed by Marsh to complete the
characters of the genus Brontotherium, but we now
know that the genotype of Brontotherium — namely,
B. gigas — is a very different animal in many ways.
The type specimen oi"B. ingens " in the Yale Museum
is a fine representative of the species, as described in
detail below. The absence of the premaxillary bones
in this specimen led to many errors regarding the
condition of the incisor teeth; we now have conclusive
evidence that in this species, as in M. trigonoceras, the
incisors are atrophied, vestigial, or entirely wanting.
The evidence as to the character of this species afforded
by the type specimen is now supplemented by that of
an exceptionally fine series of skulls in the American,
National, and Field Museums.
Observations on the measurements oj Menodus gigan-
teus.— The true molars of the fragmentary type lower
jaw agree in measurement and other characters with
jaws referred to this species.
In the referred skulls the grinding teeth range from
385 to 465 millimeters in length, a difference of 80
millimeters. The largest skull (in the University of
Wyoming) has a longer dentition than any other
known titanothere (p'-m^ 465 mm.).
The extreme contrasts between the narrow mesati-
cephalic skull and dentition of Menodus giganteus and
the broad (brachycephalic) skull of Brontotherium
platyceras are shown in the table below.
Figure 441. — Restoration of Menodus giganteus
By Charles R. Knight. About one-ninth natural size.
Measurements of Menodus giganteus, Brontotherium platyceras,
and B. gigas, in millimeters
P'-m3
Pi-p*
M'-m3
Molar index
P'', ap. by tr
M', ap. by tr
Pmx to condyles-
Zygomatic index.
Nasal length
Horn length
425
150
270
32
825
62
175
290
58X73
100X78
B. platy-
ceras, Field
Mus. 12161
340
120
223
25
80
38
■390
47X72
91X99
' Estimated.
532
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
From Allops crassicornis we observe that M. gigan-
teus is distinguished by the greater length of p'-m^
and especially of the true molars {M. giganteus 246-285
mm., A. crassicornis 220), by the greater length of the
nasals, and by the much lower zygomatic index.
Figure 442. — Skull of Menodus giganteus
Yale Mus. 12010, type of Broniotherium ingens Marsh. Front view.
One-eighth natural size. This view shows well certain features
of Menodus, as follows: Horns trihedral and pointed, nasals dis-
tally broad, zygomata deep with small buccal expansions, canines
large, cheek teeth with sharp external cingula.
From Brontops robustus M. giganteus is readily
distinguished as follows:
Measurements of Menodus giganteus and Brontops robustus, in
millimeters
M. giganteus
B. robustus
Pi-m3 .. _ __ --
385-465
141-176
246-285
50X68 to 58X73
93X89 to 100X78
777-825
62-73
105-175
340-376
pi-p«. ..
139-151
Mi-m'. . -
220-236
P*, ap. by tr _
43X55 to 46X69
M', ap. by tr
Pmx to condyles
Zygomatic index
Nasal length
73X84 to 90X89
743-813
76-83
52-90
Geologic and geographic distribution. — The geologic
range of this species as recorded by Hatcher is from
the summit of B, the middle zone, to the summit of C,
the upper Titanotherium zone.
Materials. — The species is represented by Pomel's
type jaw, which has been destroyed and is known
only from Leidy's figure and measurements; also
by Marsh's excellent type skull of "B. ingens" in the
Yale Museum (No. 12010), found in Colorado; by
three skulls in the American Museum (Nos. 505, c? ;
1066, c? ; 1067, ? ); by six fine skulls in the National
Museum (No. 1220, etc.) ; and by the superb skull asso-
ciated with cervical and anterior dorsal vertebrae
(Pis. CXXXIX, CXL) in the Field Museum of Chicago
(No. P 5927). The largest skull of all is that in the
University of Wyoming Museum.
According to the measurements given in the table
above, the skulls Am. Mus. 1066 and 1067 and Nat.
Mus. 4291, which are all referred here to M. gigan-
teus, are much closer to M. trigonoceras and are sepa-
rated from the true M. giganteus skulls by a consider-
able gap.
Characters of Marsh's type of "B. ingens." — The
principal characters of the type skull as given by
Marsh with measurements are entered in the table
above. Additional characters are as follows: This
animal is in the eighth stage of growth, advancing
toward the ninth. The tips of the internal cusps of
m'^ are slightly worn, barely showing the dentine.
The top of the skull is uncrushed but somewhat
depressed just behind the connecting horn crest.
The right horn and right half of the nasals are partly
restored, although figured as complete in Plates
CXXXVII and CXXXVIII. In the left horn, which
is more perfect, the triangular shape is very marked.
Figure 443. — SkuU of Menodus giganteus
Yale Mus. 12010, type of Broniotherium ingens Marsh. Palatal view.
One-eighth natural size. Drawing made by Berger under the
direction of Marsh. This view shows Menodus characters as
follows: Skull dolichocephalic, nasals long and distally broad or
squarish, horns trihedral in section and pointed, tooth rows rec-
tilinear, cheek teeth with sharp internal and external cingula,
premolars with large postero-lnternal cusps, molars elongate an-
teroposteriorly, zygomatic expansion moderate.
The anterior angle of the horn runs directly into the
nasals. The premolars exhibit very broad internal
cingula and pronounced external cingula; the first
premolar displays an internal crest formed of the
deuterocone and tetartocone. The second and third
premolars have the tetartocones more distinct, ele-
vated, and sharply separated off from the cingula;
the deuterocones and tetartocones are subequal,
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
533
thus being a progressive advance beyond the M. tri-
gonoceras condition. The hypocone of m' is tubercu-
late, with a narrow crest extending off toward the
metacone; this tooth exhibits a sessile "crochet"
and "antecrochet."
Additional measurements of the type of " B. ingens"
Millimeters
Summit of occiput to tip of nasals 755
Extreme spread of horns 490
Free length of nasals 125
Width of nasals 125
Molar-premolar series, estimated 428
True molar series 265
Dentition. — The incisors either consist of extremely
reduced crowns in depressed alveoli, resembling the
vestiges seen in M. trigonoceras, or are entirely
obsolete. Of the male skulls in the American Mu-
seum No. 1066 shows two vestigial alveoli and No. 505
shows one small alveolus; similarly, in the National
Museum No. 4291 shows two alveoli partly closed.
In one of the female skulls (Am. Mus. 1066) there
is a single vestigial incisor on one side, lying in its
alveolus below the surface of the jaw.
Canines: Contrasting with this atrophy is the
hypertrophy of the canines, which are readily dis-
tinguished from those in members of the Bronto-
therium phylum by their long and pointed crowns,
circular to suboval in section. Almost completely
encircling the base of the crown is a prominent
cingulum; this is a very exceptional and distinctive
character; in the females the cingulum is perhaps
somewhat less prominent. In the males the largest
canines measure about 70 millimeters; in the females
the canines measure about 40 millimeters. The chief
distinction of the female tusks is their much more
slender and pointed character.
Premolars: The persistence of the first premolars in
both jaws is a general characteristic of this species,
although not observed in the specimen in the Field
Museum. A peculiarity of p' is the presence of a
distinct tetartocone, which is even stronger in "B.
ingens " than in M. trigonoceras. In the succeeding
premolars, p^-p*, the tetartocones are more or less
well defined and distinct from the deuterocones, either
connected by a low ridge or altogether separate, the
proportion between the size of the deuterocone and
tetartocone being as 5 to 3. This acceleration of the
tetartocone is a progressive character; it is less ad-
vanced in the American Museum skulls than in the
Yale Museum type of "B. ingens," which was prob-
ably found on a very high geologic level. As a rule
the tetartocones increase in distinctness as we pass
backward from p' to p^, but as in all other titano-
theres the tetartocone in p* is generally less distinct.
A mesostyle ridge is faintly developed on p* in some
specimens of Menodus. It is best developed in cer-
tain specimens of Brontofherium and Megacerops in
which p* is more molariform than in Menodus.
Both molars and premolars have deep pits or
"medifossettes" in the midvalley, just internal to the
ectoloph; this "fossette" is bounded internally
by folds homologous with the "antecrochet" and
"crochet" of rhinoceros molars. The molars and to
a less extent the premolars are proportionately nar-
rower and longer than in the brachycephalic phyla.
Molars: The "fossette" just mentioned is correlated
with the vertical elongation of the ectoloph, which
now measures 80 millimeters in height, while the pro-
tocone measures only 28 millimeters; the outer wall of
the tooth is thus between two and three times as high
as the inner wall. As in the preceding species of this
phylum, the internal cusps also of the grinding teeth
have steep slopes.
The hypocone of m^ shows individual variability:
it is either large and distinct (Am. Mus. 1067), or
small and distinct (Nat. Mus. 4291), or confluent
with the cingulum (Am. Mus. 1066).
The cingulum is certainly the most conspicuous
and distinctive character of the grinding teeth; not
only does it encircle the canine, but it is sharply
defined upon the outer faces of the entire superior
and inferior premolar-molar series (Pis. CXXXVI,
CXXXVII, CXXXIX). The superior premolars are
readily distinguished by the broad shelf-like internal
cingulum, unlike that in any other phylum. The
superior molars are also cingulate upon the inner
sides. The only part of the entire dental series which
entirely lacks the cingulum is the inner side of the
inferior premolar-molar series, where no titanothere
displays a cingulum.
The grinding series as a whole, therefore, is distin-
guished by its great length, by the proportionate
length of the individual teeth, the anteroposterior
slightly exceeding the transverse diameter (in the
molars), and by the pronounced development of the
cingulum.
Skull. — The chief distinction of the skull Ues in its
proportions; it is relatively long and narrow, the
average ratio of two male skulls being, length 773
millimeters, breadth 545 millimeters, and consequent
index 69, whUe in the type of Brontops rohustus the
ratio is 765 to 667, and the index is 87. In all the
skulls the nasals are large and quadrate, measuring
from 127 to 175 millimeters in length and from 125
to 145 and more in breadth. As in M. trigonoceras,
advancing age is accompanied by a rugose develop-
ment of the extremities, with a lateral expansion of the
tips and a deepening of the median cleft. Other
growth characters of the skull are the increasing length
of the horns from 150 to 290 millimeters, the increasing
rugosity and breadth of the occiput, the widening of
the buccal expansions of the zygomata. These buccal
expansions, however, remain limited in extent and do
not involve a large part of the zygomatic arch as in
Brontotherium; immediately below and slightly anterior
to the buccal expansion is a very characteristic,
534
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
slightly downward projection. Other conspicuous
characters are the relatively large orbits (as compared
with those of Brontotherium) , the bridge over the
infraorbital foramen, the steep slope of the occipital
vertex (contrast Brontotherium). Additional features
of advancing age are the development of the marked
supraciliary expansion of the lateral frontal parietal
crest (Nat. Mus. 1220), the obliteration of the suture
between squamosal and malar.
Horns. — The distinctive characters of the horns are
not only the connecting crest, which, as shown in the
section, is situated very far back, but the normal
triangle with the external angle at the vertex. (See
figs. 399, 444, 445.) The skull has other distinctive
features: First, we observe the width of the bridge
over the infraorbital foramen; as in the Palaeosyopinae
and also in the related Brontops phylum this bridge
extends into a thin plate anteriorly, so that the fora-
men is distinctly seen on the side of the face, whereas in
Brontotherium and Megacerops the malar bridge is
narrow and convex and the foramen is not seen in
side view. This elongate condition of the infraorbital
canal and breadth of the malar bridge is correlated
with the dolichocephaly existing throughout the
A
B
Figure 444. — Sections and contours of skulls of Menodus giganteus and M. varians
M. giganteus: A, Am. Mus. 505, cf (neotype), and B, Yale Mus. 12010 (type of Brontotherium ingens); relatively long-pointed horns witli a prominent
anterior swelling suggestive of the accessory hornlet in Diploclonus, horns trihedral in hasal section, connecting crest near the plane ot the posterior
face of the horns, nasals long and wide, zygomata little expanded. C, M. varians, Yale Mus. 12060 (type); horns thiclser (especially at the top), basal
section widely trihedral, nasals broad and thicli, zygomata somewhat expanded. One-ninth natural size.
direction, which is chiefly outward and slightly up-
ward. A very distinctive character in front view is
the straight inferior contour of the horn (see fig. 442);
all members of the Brontotherium and Megacerops
series have a curved or convex lateral inferior contour
as seen from in front. Correlated with the lesser
strain of the horns and the great breadth of the
anterior nares the maxillary pillars on either side of
the anterior nares are much thinner than in the stout-
horned species. The basal horn section as compared
with that of M. trigonoceras has already been described
and is very distinctive; briefly it consists of an isosceles
skull. In the midline of the parietal crest there is
observed a median ridge clearly shown in the section,
there is also occasionally a conspicuous knob in the
midparietal region. In the occipital region we observe
a mastoid foramen; the postglenoid and paroccipital
processes are proportionately narrower and deeper
than in the Brontotherium series, another fact in keep-
ing with the dolichocephalic structure of the skull.
Again there is narrower contact between the post-
glenoid and post-tympanic, a more open auditory
meatus, and a wider space behind the ear than in the
brachycephalic types.
EVOLUTION OF THE SKULL AND DENTITION OP OLIGOCENE TITANOTHERES
535
Lower jaw. — Dolichocephaly is also influential in
the jaw structure, as among the long-skuUed rhinoc-
eroses. The rami of the jaw are long and deep, the
lower border being carried horizontally forward until
it terminates in a well-marked angulation of the chin;
from this point it ascends sharply to the incisive
border; the angle is broad, with a marked backward
extension. This decided angulation of the chin is
quite distinct from the convex lower border and
shallow, sloping chin in the Brontotherium series.
Another distinctive feature is the greater size of the
mental foramen (Am. Mus. 1067) and its more pos-
terior position beneath the fourth premolar. In the
Brontotherium series the foramen is usually smaller
and more anteriorly placed. The angle has a smooth
border, is rather delicate and narrow but none the
less prominent, flaring obliquely backward with an
even curve at the postero-inferior border, in contrast
with the Brontofherium angle, which is broader, with
a rugose border and truncate apex. The coronoid is
broad, has a smooth, rounded anterior edge, and is
not strongly curved backward. The smooth backward
extension of the angle, the straight lower border, and
the definition of the chin are characters exactly
paralleled in the dolichocephalic rhinoceroses.
Variation in size. — This is not only distinctive of
sex, the females being more slender in all their pro[)or-
tions, but of the stages of growth which are recorded
in the measurements of the skulls, and also of the
stages of vertical succession or evolution which are
recorded in the measurements chiefly of the premolar-
molar series, as shown in the tables on page 523.
Final stage of evolution. — In the Field Museum,
Chicago, there is a superb skull (No. P 5927), found
near the top of the upper TitanotJierium zone at
Phinney Springs, S. Dak. With it were discovered
the lower jaws, cervical and anterior dorsal vertebrae.
(See p. 686.) The skull is in a beautiful state of pres-
ervation and has been very accurately reconstructed
by Mr. E. S. Riggs. It belongs to a fully adult
animal, with well-worn teeth, and strongly accentuates
all the characteristic features of the type of M. {"Bron-
totherium") ingens. The lateral and anterior views
(Pis. CXXXIX, CXL) accordingly give the most per-
fect idea of the extreme development of this phylum.
Noteworthy peculiarities in the dentition are these:
(1) PMs present on the left side of the skull but absent
on the right. In most other specimens p' is constant;
this, however, proves that in Menodus as in Brontops
p' is a variable tooth. (2) Pi is also absent in this
specimen, the canine being closely crowded against
P2. (3) The third inferior molar exhibits a very small
hypoconulid which lacks the internal cup or crescent.
Noteworthy progressive features are the presence
of an anterior swelling or incipient hornlet on both
the horns, somewhat similar to that observed in the
type of Diploclonus hicornutus, and the very anterior
position of the bases of the horns, which are almost
101959— 29— VOL 1 37
as far in advance of the line of the orbits as in the
type of Brontotherium gigas. This feature does not,
however, interfere with the great length of the nasals,
which protrude far beyond the vertical line of the pre-
maxillaries. Notable, also, are the relatively great
length of the horns (outside measurement 290 mm.).
The detailed measurements of this important skull
are presented in the table on page 523.
Observations on the Menodus giganteus of the Uni-
versity of Wyoming. — The largest dentition among
known titanotheres is exhibited in a specimen of
Menodus giganteus in the
University of Wyoming,
from the upper Titano-
therium zone of Bates
Hole, Carbon County,
Wyo. The premolar se-
ries (p'-m^) measures 465
millimeters, as compared
with 410 in the type.
The premolars are very
large (176 mm.) and highly
progressive They con-
form in all details to the
Menodus type. They
show very clearly the
enamel folds and ridges
which are called "crista,"
"protoloph," and "meta-
loph." Their tetarto- „ aa^ a ^- j
^ Figure 445. — Sections and con-
cones, although large, are tours of skull of Menodus gi-
still connected by a bridge ganteus
with the deuterOCOneS, as Am. Mus.SOG, 9 (of. fig. 446). in females
in the Menodontinae ^en- °^ '^'^ species the horns are slender but
11 rpu J" '^ ha.ve a relatively high connecting crest,
erally. Ihe dimensions the tasal section is obUquely and roundly
of the premolars and mo- t"hedral, the nasals are long and are de-
. ,, curved at the tip. One-eighth natural
lars are as loilows: size.
Millimeters
Pi, ap. by tr 25X26
P2, ap. by tr 40X47
ps 51X59
P^ (estimated) 58X73
Ml (estimated) 81X77
M2 (estimated) 100X80
M^, ectoloph, anteroposterior 114
M-, transverse (across mesostyle) (estimated) 95
M^, height of ectoloph (paracone) 88
MS, ap. by tr. (estimated) 100X78
M^, transverse (across niesost3'le) 90
Menodus varians (Marsh)
{Menoys varians Marsh, 1887; " Titanotheriiim ingens Marsh,"
Osborn, 1902)
Plate CXLII; te.xt figures 181, 409, 444
[For original description and type references see p. 223]
Geologic horizon. — Geologic level not known, prob-
ably upper Titanotherium zone.
Specific characters. — The skull is relatively shorter
or less extremely dolichocephalic than that of M.
giganteus. Skull length, premaxillaries to condyles
536
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
760 millimeters, nasals to top of occiput 695, width
across zygomata 555, zygomatic index 73 (that is, sub-
brachycephalic). Nasals decidedly shorter than in
M. giganteus (free length 105 mm., breadth 140).
The horns are outwardly directed but of similar section
to those in M. giganteus. Two vestigial upper inci-
sors retained on each side. Premolar-molar series
(410 mm.) relatively very long (dental index 54).
Premolar tetartocones very distinct. P* with redupli-
cate tetartocone. Hypocone on m' large and distinct
from the cingulum.
This animal appears to be specifically distinct from
M. giganteus and to present a form which tends to
bridge over the structural gap between Menodus and
Allops, since while exhibiting all other characters of
the protocone and the hypocone of m^ being well worn.
In its superior aspect the skull is very similar to
Marsh's type of ingens, although less dolichocephalic;
the horns are connected by the characteristic low trans-
verse crest. In the plane of the posterior faces they
are crushed downward and outward in such a manner
as not only to change their normal elevation but to
decrease the acuteness of the trihedral section and
Figure 446. — Lower Jaws of Menodus giganteus
A, Am. Mus. 500, ? (cf. flg. 445); B, Field Mus. P 5927, a large male. Both jaws have the ramus elongate, the
into a rounded elbow, and the teeth sharply cingulate. One-fifth natural size.
ngle produced posteriorly
M. giganteus, it possesses the short nasals character-
istic of Allops. That this skull may be an Allops is
further indicated by its sub-brachycephaly (index 73)
as in Allops serotinus, also by the proportions of its
horns and nasals.
Materials. — The species is known only from the
type skull in the Yale Museum (No. 12060), which is
represented in Plate CXLII.
Description oj the type. — The type skull is that of an
old animal in the eighth to ninth stage of growth, both
make it difficult to express in cross section (fig. 444, C)
their actual form. The nasals had the spread but not
the length characteristic of Menodus. In inferior
aspect of the skull the nares open immediately behind
the second molar, as in M. ingens. There are traces
of a median vomerine keel and a very prominent
rugose projection at the junction of the basisphenoid
and basioccipital, which is apparently broken away in
the type of M. ingens. As in M. ingens the orbit is
directly above the posterior portion of the first molar.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
537
Measurements of Menodus varians
Millimeters
Length of skull, vertex to tip of nasals 695
Basilar length, occipital condyles to premaxillary 760
Transverse width across zygomata 555
Molar-premolar series 410
True molar series 255
Premolar series 155
Canine crowns, anteroposterior diameter 35
Free length of nasals 105
Free width of nasals 140
Dentition. — The alveoli for two vestigial iacisors
upon each side persist, with a narrow median diastema
between them. There is a vestige of a third alveolus
on one side. The alveoli are exceptionally shallow,
and these teeth were undoubtedly vestigial. The
canines are so robust as to indicate that this was a
male animal. The cingulum is continuous around
the anterior face; the crowns are broken off. Close
behind the canines are the bifanged, well-developed
first premolars; these teeth are distinguished as in
M. giganteus by very broad, crenulate internal cingula
and prominent external cingula; in p^ and p^ the
tetartocones are very prominent. P* exhibits an
interesting example of correlated bilateral variation
in the double conical summits of the tetartocones;
this, however, is probably due to the advanced con-
dition of the cingulum. The hypocone is strongly
developed and entirely distinct from the cingulum
upon m^, as in Marsh's type Diconodon montanus and
in several specimens of M. trigonoceras.
Relationship to Menodus. — In describing this sup-
posed genus, Menops, Marsh observed: "The present
genus is most nearly related to Diconodon and in its
molar teeth agrees with that form. It differs in the
presence of two upper incisors on each side." The
entirely vestigial character of the incisor teeth natu-
rally forbids our assigning them generic value. This
animal presents so many points of close similarity
with the type of M. ingens that the single decisive
specific character which can be selected is the abbre-
viation of the nasals and the somewhat less extreme
dolichocephaly. The nasals are less quadrate than in
M. ingens, the free length (105 mm.) being less than
the free width (140 mm.). Even this character is
possibly attributable to individual variation, since
some of the typical specimens of M. giganteus and M.
trigonoceras show a relative abbreviation of the nasals.
A similar broadening and abbreviation of the nasals
occurs in the subgenus Allops. The inferior contour
of the horns is less straight, and the horns apparently
diverge more widely than in M. ingens, a condition
partly attributable to the downward crushing of the
skull or perhaps indicating affinity with the fiat,
outwardly directed horns of Allops.
The reduplication of the tetartocone on p'* is another
feature observed in Allops serotinus. A more impor-
tant difference is that the skull is proportionately
shorter and less dolichocephalic than in M. ingens^
the proportions being, length 760 millimeters, breadth
545, as compared with length 825, breadth 550. The
hypocone of ni'' is even larger and more distinct than
in most specimens of M. giganteus, resembling that
in the type of Diconodon montanus. This may be a
specific or progressive character.
We observe the similarity to M. giganteus in many
other details, such as the sections of the hdrns, the
shape of the zygomatic arches, the presence of small
alveoli for the incisors, the antorbital knob on the
molars, the comparative length of the face, the ap-
parent exposure of the mastoid bone, the sharp and
horizontal shelf of the top of the occipital pillars,
the bifanged, well-developed first premolars, the ex-
treme cingulate development of the canines and grind-
ing teeth, the pi-ominence of the tetartocones and of a
hypocone on m^
Our conclusion is that the genus Menops is probably
equivalent to Menodus, while the species M. varians
may be regarded provisionally as valid and distinct.
Menodus montanus (Marsh)
{Anisacodon montanus Marsh, lS75; Diconodon montanus
Marsh, 1876; " Symborodon montanus" Osborn, 1902)
Text figures 175, 447
[For original description and type references see p. 217]
Type locality and geologic horizon. — Recorded as
from "northern Nebraska"; Titanotherium zone.
Specific characters. — Nasals narrow, relatively ab-
breviate, so far as preserved, resembling those of
Allops serotinus, grinding teeth with transverse diam-
eter exceeding the anteroposterior, thus resembling
Allops. Incisors reduced or vestigial, as in Menodus,
m^ with a distinct hypocone, prominent external
cingula on all grinding teeth, and a strong internal
cingulum on the premolars. M'-m^ 218 millimeters.
General characters. — The genotype (Yale Mus.
10022) of Anisacodon montanus is an incomplete
fragment of a skull, together with the complete
molar series of both sides, portions of the left
maxillary, of the left zygomatic arch, the ex-
tremities of the nasals, and the left superior pre-
molars. There are paired alveoli for p^ As in
Menodus the incisive border is narrow, sharp, and
functionally edentulous, but it exhibits two vestigial
alveoli. The premaxillary contains two caniae al-
veoli. The internal cingulum of the premolars is
prominent and rounded; the external faces of both
premolars and molars exhibit a sharply serrate cin-
gulum. The third superior molar presents a conical
hypocone quite distinct from the cingulum. To this
the generic name Diconodon, originally applied by
Marsh, refers. The grinding teeth are of about the
same size as those of Menops varians, the molar
fossettes are very deep.
538
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Measurements of Menodus monianus
Millimeters
Canine alveoli, anteroposterior diameter 31
Canine, transverse diameter 19
Nasals, transverse 96
Nasals, longitudinal (so far as preserved) 87
Second superior molar, transverse 88
Second superior molar, anteroposterior 76
Molar series, superior 218
Affinities to Menodus. — As shown in the above de-
tailed description this animal agrees with M. "ingens,"
and especially with M. varians, in the vestigial char-
FiGURE 447. — Teeth and nasals of Menodus montanus
valid species, although more perfect material may
relate it still more closely or even specifically to M.
giganteus.
SECTION 6. THE BRONTOTHERIINE GROUP
We now consider the brontotheriine group as dis-
tinguished from the menodontine group. It contains
two phyla, closely affiliated in their Eocene origin
but widely diverging in their Oligocene evolution — ■
namely, the Megaceropinae and Brontotheriinae.
GROUP CHARACTERS
Lower Oligocene titanotheres of me-
dium to larger size. Horns progres-
sively elongating in the males until
they attain great size at the expense of
the nasals, which are atrophied. Face
abbreviated, or brachyopic. Skulls with
broadly spreading zygomatic arches and
brachycephalic characters in the audi-
tory region, in the occiput, in the broad
proportions of the upper grinding teeth,
and in the arching of the opposite series
of grinding teeth. All grinding teeth
devoid of external cingula in males.
Premolar grinding teeth precociously
becoming more molariform than in
other groups. Orbits small.
This group includes all the known
long-horned titanotheres, both the
smaller megaceropines and the larger
brontotheres. Of the two the mega-
ceropines are less formidable animals,
without incisor teeth, and with a less
powerful action of the horns. The
brontotheres are hj far the most formid-
able of all the titanotheres known, ani-
mals of gigantic size, with powerful
horn action, hence appropriately termed
by Marsh "thunder beasts." The ana-
tomical resemblances between these
animals are by no means confined to
the superficial similarities but extend to
all parts of the skull and teeth, as we
have seen in the introduction of this
Yale Mus. 10022 (type of Diconodon monianus Marsh). A, Third right upper molar. The elongate pro-
portions and hypocone surrounded by a cingulum are seen also in certain other specimens of Menodus
(of. Am. Mus. 1067). B, Fourth upper premolar and first and second true molars of the left side. The chapter, and pomt tO a COmmOU anCCStry
elongate proportions and sharp internal and external cingula are seen as in J/enoii«5. C, Alveoli of the . -p„„p„„ time
upper incisors and canines. The upper incisors were vestigial, as in Menodus; the canines were large. -^^ -^"*- . ' t .-i u
implying male se.x. Di, Top view of distal portion of nasals. The paired projections are seen as in The phylctlC characters 01 the brOUtO-
Menodus. Di, Front view of nasals. One-half natural size.
acter of the incisors, the bifanged premolar, the pres-
ence of a distinct hypocone on the last superior molar,
the marked external cingula on both premolars and
molars, and the strong internal cingula on the pre-
molars. It apparently differs from Menodus in the
narrow nasals and the more transversely extended
proportions of the molar teeth.
It seems probable that this fragmentary specimen
should be considered provisionally as the type of a
theriine group common to Megacerops
and Brontotherium are indicated below:
1 . There is a general increase in size of the skull and
skeleton.
2. The postorbital region of the skull shows rapid
elongation, the preorbital region rapid abbreviation,
occiput widely prolonged behind the zygomata, or-
bits small, and never a prominent postorbital process.
In uncrushed skulls the orbits do not appear as if
closed posteriorly.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
539
3. The zygomatic arches spread greatly in males,
less in females; the zygomatic index becomes very
high in extreme forms; the postglenoid unites with
the post-tympanic process to inclose the external audi-
tory meatus; the transverse measurements of the su-
perior grinding teeth exceed the anteroposterior
measurements throughout; the occiput greatly in-
creases in breadth and slowly in height; the opposite
grinding series become arched and strongly bent up-
ward, the opposite canines thus being brought near
together. One of the most distinctive and constant
features of the skull is the anteroposterior convexity
on top of the parietals and the general flatness of the
skull top in contrast to the deep concavity of this
region in the Menodontinae.
4. The dental index appears to be on the average
somewhat lower than in the menodontine group.
The variability of the dental index is probably due to
crushing. The grinding series does not increase in
length so rapidly as the skull but increases greatly in
width. The internal cingula tend to disappear or
degenerate.
5. The canines are obtuse or recurved, massive in
males, small in females, never elongate and pointed.
6. The premolars are distinguished by the acceler-
ated development of the tetartocones, which are
placed farther in toward the center of the crown —
that is, away from the internal or lingual border, a
common distinctive feature; the anterior premolars,
p-^, are very progressive, with distinct tetartocones,
but nevertheless are thrust inward toward the lingual
line and tend to drop out in old age.
7. The internal cingula of the premolars are massive
and bluntly crenulate, gradually becoming more or less
confluent with the base of the crowns.
8. The dominant feature of the skull is the hypertro-
phy or elongation of the horns and the corresponding
atrophy or abbreviation of the nasals, a compensa-
tory character. The horns are arrested in develop-
ment in the females, but the nasals are abbreviated
in both sexes, though less abbreviated in females than
in males.
The brontotheriine group possessing these common
characters early subdivided into two very distinct
phyla which we term respectively the genera Mega-
cerops and BrontotJierium, animals which enjoyed an
independent simultaneous development from the
base to the summit of the Titanotherium zone.
Members of these two phyla inherited a number of
ancestral characters and also a number of predisposi-
tions to a similar evolution, which are enumerated
in the phyletic and family definitions above. Thus in
both phyla the horns progressively increase in size,
the teeth undergo similar changes.
Megacerops, however, is readily distinguished from
Bi'ontotherium in many parts of the skull and teeth
and probably also in the skeleton, as we shall un-
doubtedly demonstrate when the skeleton becomes
fully known. The most conspicuous points of differ-
ence are shown in the accompanying table.
Comparison of features of memhers of the Megacerops and Brontotherium pJiyla
Megacerops phylum (subfamily Megaeeropinae)
Brontotherium phylum (subfamily Brontotheriinae)
1. Animals of small to medium size, either slowly increas-
ing or arrested in size.
2. Skulls mesaticephalic to brachyeephalic.
3. Horns rounded in section, vertical in position; placed
rather above the orbits and not greatly shifting for-
ward, with the connecting crest small or absent.
4. Narial aperture high and narrow.
5. Nasals thin and progressively reduced in lengtli.
6. Incisor teeth usually vestigial; incisive borders edentu-
lous in males and females.
7. Canine teeth small and closely approximating to one
another.
1. Animals of small (B. leidyi) to extremely large size, in-
creasing to the largest size.
2. Skulls dolichocephalic to brachyeephalic, finally attaining
a high zygomatic index.
3. Horns transversely oval to flattened in section, widely
divergent, shifting forward and progressively developing
a high connecting crest.
4. Narial aperture constricted by the heavy buttresses sup-
porting the horns.
5. Nasals thick and rapidly reduced to short, obtuse knobs.
6. Superior incisor teeth invariably persistent, 2-1 in males.
7. Canines large, obtuse, arrested in growth, separated from
one another.
So far as is indicated by the breadth of the cheek
teeth, the curvilinear premolars, the upward flexure of
the premolars, and the broad zygomatic arches (for
example, M. hucco), Megacerops appears to be even
more brachyopic than Brontotherium. Megacerops
is less extreme in the horns than Brontotherium but
more extreme in the somewhat closer approximation
of the canines and stronger arching of the premolar
series.
The grinding teeth of Megacerops are of the same
type as those of Brontotherium; the molars are not
readily distinguishable; the premolars of Megacerops
are smaller with reduced internal cingula. While the
incisors are usually absent in the adults there is
evidence that they were present in young animals.
A skull in the National Museum which resembles
Megacerops in its horn and nasal structure exhibits
large superior incisors.
540
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
SEXUAL CHAEACTERS
In the females of both series we observe a far more
marked distmction from the males than that observed
in the menodontine group. Not only are the canine
teeth smaller, but there is a pronounced diEference in
the proportions of the horns. There are certain other
detailed characters which so far as our observations
are valid seem to separate the females from the males.
A summary of these contrasting sexual characters in
Brontotherium is given below.
Males
Horns long; connecting crests
large and prominent; nasals
decidedly short. Canines
larger; incisors more con-
stant, formula f;! or f;-?-.
Buccal processes of zj'go-
mata greatly produced; occi-
put greatly extended behind
the line of the zygomata.
Females
Horns relatively short; con-
necting crests less promi-
nent. Nasals relatively long.
Canines smaller; incisors less
. persistent, sometimes want-
ing. Buccal processes of
zygomata less prominent;
occiput not so widely ex-
tended behind zygomata.
From some specimens it would appear also as if the
premolars of brontotheres were somewhat less pro-
gressive in evolution in the females than in the males.
This would controvert the general principle observed
in Menodus that the grinding series of the teeth,
which are so essential to the nutrition of the females,
is the one character in which the sexes do not differ.
Specimens of the female sex are smaller in size through-
out, as seen in the detailed table of measurements.
SUBFAMIIY MEGACEEOPINAE
Relatively small, long-horned titanotheres, known
chiefly from the middle Titanotherium zone. Horns
precociously evolved, vertical in position, placed
above the orbits, with little or no connecting crest.
Incisor teeth much reduced or actually vestigial,
canines very small, placed close together, thus tend-
ing to contract the premaxUlaries.
Geologic Jiorizon and geographic distribution. — So far
as known these animals are of medium size or rela-
tively small and are recorded chiefly from the middle
Titanotherium zone of Cedar Creek, Colorado, and
from the lower portion of the upper Titanotherium
zone of South Dakota, possibly also from Assiniboia,
Canada.
Four skulls of Megacerops were found in lower
Oligocene (Chadron) deposits at the levels indicated
below, chiefly according to the records of J. B.
Hatcher:
Level C:
?M. acer (type).
M. copei, Nat. Mus. 4711 (type;.
M. acer, Univ. Wyo. Mus.
Level B:
M. bucco, Nat. Mus. 4705, 5 . (Level rather
doubtful.— J. B. Hatcher.)
Distinguishing features. — The males develop long
horns, which, imlike those of the typical Brontotherium.,
are placed only slightly in front of the orbits, so that,
as shown in Charles R. Knight's models and restora-
tions (PI. XVIII, C; fig. 454), the eyes appear almost
directly below the horns. As in the brontotheres the
eyes were small. The horns are directed almost
vertically upward, with the long basal axis placed
obliquely, the basal section never tending to become
transversely oval, as ia Brontops roiustus and Bronto-
therium, or triangular, as in Menodus. The reason for
this is that the horns are supported or braced prin-
cipally in a fore and aft direction, instead of across
the skull by means of the connecting crest, indicating
that they were used largely in a vertical or tossing
motion of the head rather than in the lateral motion
characteristic of the brontotheres.
The narrowness of the chin and premaxillary
region and the entire absence of incisor teeth indi-
cates that these animals were provided with a narrow
and pointed prehensile upper lip, contrasting with
the somewhat broader lips in the brontotheres.
A decided and highly characteristic feature is the
marked abbreviation (brachyopy) of the facial por-
tion of the skull, correlated with the cyptocephaly
or upward flexure of the anterior grinding teeth.
Materials. — Remains of Megacerops are compara-
tively rare. Their existence was first made known
by Leidy in 1871 (see p. 210) through the type spe-
cies Megacerops coloradensis ; secondly, by Cope in
1873 from the skulls from Colorado described as
Symborodon bucco, S. acer, S. altirostris. Cope's
specimens are in the American Museum of Natural
History. A lower jaw ia the American Museum
(No. 6364) appears to represent a very small new
species of this genus, known as M. riggsi. Two ex-
ceptionally complete skulls from South Dakota are
in the National Museum, one of which is referred to
a third species, M. copei, related to M. bucco. The
first (Nat. Mus. 4705, skull O') is small horned, ap-
parently belonging to a female of M. bucco, and is
recorded by Hatcher from the top of Chadron B ; the
second (Nat. Mus. 4711, skull V'), the type of M.
copei, apparently a male, is recorded by Hatcher from
the middle beds and was known by the collectors as
the "rabbit skull," because of the resemblance of the
flaring horns to the ears of the jack rabbit.
The Megacerops phylum as represented in the Hatcher collection
of six skulls and lower jaws from the Chadron formation, in
the United States National Museum
Genus and species
Catalog No.
Materia]
M
bucco (Cope)._
4705, ?
Fine skull. Typical.
Do
4700, ^
Skull. Close to M. acer Cope.
Differs from typical Mega-
cerops in having single up-
per incisors. Horns and ca-
nine of Megacerops type.
M
copei (Osborn).
4711, &
Skull. Type.
M
riggsi? Osborn^
1236, cf
Lower jaw.
M
riggsi Osborn _ .
5412, cT
Right jaw, with symphysis;
contains molars.
M
sp
S786, ?
Lower jaw. Poor.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TXTANOTHERSE
541
Our knowledge thus rests chiefly on six skulls and
two lower jaws as representing this genus. There
are also isolated horns and the top of a cranium in
the American Museum collection and fragmentary
skeletal material (figs. 625, 629, 638). In the IJni-
versity of Wyoming there is an excellent skull of
Megacerops acer.
General characters. — These specimens agree in the
exceptionally small size of the canines, in both the
males and the females. The nasals are thin in two of
the species, M. hucco and M. copei. M. acer is some-
what divergent in the thickening of the nasals and
in other characters.
It is thus apparent that Megacerops is a separate
collateral phylum, resembling Brontotherium in the
elongation of the horns and paralleling Menodus in
the degeneration of the incisors; but it differs from
both phyla in the shape and position of the horns
and in the approximation of the canines toward the
median line.
STceleton. — The skeleton is known only from a few
portions associated with M. acer in the American
Museum, which indicate that the animals of that
species were rather small.
While generally of smaller size and differing widely
from Brontotherium in the entire absence or vestigial
character of the incisor teeth, the position and basal
section of the horns, and some other characters, these
animals present many resemblances to Brontotherium,
especially in the elongate horns — in some cases oval
in top section — in the backward prolongation of the
occiput, the broad contact between the postglenoid
and post-tympanic, the roundness of the malar bones,
the suddenly projecting buccal expansions of the
zygomatic arches, bluntness of the canines, round-
ness and bluntness of the internal cusps of the pre-
molars; also the abbreviation of the premolar series
and reduction of the cingula, the strongly recurved
coronoid, and the general contour of the jaw.
The common characters of these Megacerops species
are clearly perceived in a comparison of the transverse
sections of the horns and of the nasals, which are
highly characteristic. The main features of corre-
spondence between Megacerops copei and Menodus
torvus are, first, that the nasals are very thin, even
in the portion between the horns; second, that the
horns are set widely apart at the base. (This char-
acter is obscured by lateral crushing in the male
type skull of M. copei.) Reference to the detailed
descriptions of the skulls in these two species
shows that they are closely related but that M.
copei is more primitive, especially in the retarded con-
dition of the tetartocones, the section of the buccal
processes, and the persistence of the reduced in-
cisors.
SYSTEMATIC DESCRIPTIONS OF GENERA AND SPECIES
IN THE MEGACEROPS PHYLUM
Megacerops Leidy, 1870
(Megaceratops Cope, 1873, in part (M. acer); Symborodon Cope,
1873, in part (5. hucco, S. altirostris) ; "Symborodon" Osborn,
1902)
Plates XVIII, CXLIII-CLX, CLXXXVI; text figures 24, 164,
167, 169, 375, 378, 390, 392-394, 398-400, 434, 448-456, 625,
629, 638, 640, 719, 744, 746
[For original description and type references see p. 208. For skeletal characters
see p. 691]
Generic characters.— Incisors, typically vestigial. Ca-
nines small, obtuse. Grinding teeth without cingula,
deflected upward. Premolars with progressive tetar-
tocones. Skull brachy cephalic to hyperbrachyce-
phalic (zygomatic index 84), brachyopic. Nasals
slender, narrow, decurved, abbreviated progressively.
Horns set vertically, typically without connecting
crest; placed above orbits; rounded in section.
The genotype of Megacerops (1870) is the species
M. coloradensis Leidy, represented by nasals and
horns (fig. 448). The genotype of "Symborodon"
(1873) is the species S. torvus Cope, which is repre-
sented by a lower jaw. (See p. 211.) This jaw does
not belong to the same phylum as Megacerops because
it proves to be that of a Menodus.
General characters. — Dentition: I^°. Incisors re-
duced, vestigial, or wanting; canines reduced, obtuse;
opposite grinding series arched to strongly arched;
upward flexure of face and premolar series as seen in
side view extreme; length of premolar-molar series
generally less than two-fifths that of the skull, from
premaxillary tips to occipital condyles; premolar series
very short; internal cusps of grinding teeth low, robust,
well rounded, ectolophs sharply depressed to the
crowns of the teeth; anteroposterior diameter of m-
and m' less than transverse diameter; no cingula be-
tween grinders; P||; p' subc[uadrangular, outer wall
not overlapped posteriorly by ectoloph of p^; premolar
tetartocones exhibiting early and pronounced develop-
ment; premolars with internal cingula blunt, reduced
or absent, external cingula variable; molars without
internal cingula, external cingula faint or absent;
hypocone of m' prominent, triradiate.
Skull: Skull proportions mesaticephalic to brachy-
cephalic; facial portion of skull much abbreviated;
premaxillaries contracted; cranial portion of skull
elongate; anterior narial aperture high and narrow;
preorbital malar bridge very narrow, mainly com-
posed of the median ridge, which is very prominent,
subcylindrical, in side view concealing the infraorbital
foramen almost entirely; anterior portion of malar
stout, rounded; malar below postorbital process
strongly convex; free nasals tapering, progressively
abbreviated; horns of medium to large size, forward
shifting slight or wanting, basal section deep antero-
posteriorly, with antero-external or maxillary face flat.
542
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
oblique, and strongly convex inner face; summit
rounded to oval in section; zygomata strongly arched,
buccal expansions finally extreme, in section broad
rather than deep; occiput moderately produced back-
ward behind zygomata; basisphenoidal rugosity absent,
vomerine septum (?) absent.
Jaw: Jaw robust rather than deep, with shallow or
concave chin, coronoid strongly recurved. Angle
deep, rugose.
Incisors : In most adult specimens the incisors have
been shed and are represented only by alveoli. Thus
the type of M. acer has medium-sized superior incisor
alveoli. A specimen (Nat. Mus. 4700) closely re-
sembling M. coloradensis in horn structure has per-
sistent superior incisors.
Oiservations on the measurements of the Megacerops
series. — The members of this series are distinguished
from Brontotherium chiefly by the cylindrical horns and
by the feeble development of the connecting crest.
The basilar length (pmx to condyles) is shorter than in
any species of Brontotherium except B. hypoceras and
B. leidyi. The premolars as a rule are shorter and
smaller than in Brontotherium, as shown below. The
molars, however, in proportion to the basilar length
of the skull, are sometimes relatively larger than in
Brontotherium, as shown in the following tables:
Molar index in species of Megacerops and Brontotherium
M. acer, Am. Mus. 6350, 9 (^)
M. acer, Univ. Wyoming (|f|)
M. bucoo, Am. Mus. 6345a (type) (Iff)
M. copei, Nat. Mus. 4711 (type) (H^)
B. ?tichoceras, Nat. Mus. 8313 (Mi)
B. platyceras, Field Mus. 12161
B. curtum, Yale Mus. 12013 (type)
B. gigas, Am. Mus. 492
B. leidyi, Nat. Mus. 4249 (type)
B. leidyi, Carnegie Mus. 93
Anteroposterior and transverse dimensions of p* and w? in species
of Megacerops and Brontotherium, in millimeters
M. acer?, Am. Mus. 6350 (type of M.
altirostris)
M. acer, Univ. Wyoming ,
M. bucco. Am. Mus. 6345a (type)
M. bucco, Am. Mus. 6353
M. bucco, Nat. Mus. (skull A)
B. gigas hatcheri, Nat. Mus. 4262
B. gigas, Am. Mus. 492
B. gigas (hatcheri), Carnegie Mus. 341
83
80
■74
99
84
o Estimated.
The internal cingula of the premolars are usually
more reduced than in Brontotherium.
The close kinship of Megacerops to Brontotherium
is revealed in many details of the incisors, canines,
premolars, and molars, in the great expansion of the
zygomata, and in the possession of a midparietal
protuberance.
Measurements of species of Megacerops and Brontotherium, in
millimeters
Stage 1:
B. leidyi
Stage 2:
B.?ticho-
ceras,
Nat.
Mus.
8313 (A)
Stage 3
Car-
negie
Mus.
93
Nat.
Mus.
4249
(type)
M. acer
M.
Type
(c?)
Univ.
Wyo.
rostris"
(type)
Pi-m3
300
190
665
66
114
104
290
186
665
118
107
313
196
695
76
90
111
330
Mi-m3
60
290
200
646
84
70
242
215
Pmx to condyles —
Zygomatic index —
Nasal length
Horn length
630
44
165,??
Standard measurements in the Megacerops phylum, in millimeters
Upper teeth
Skull
Jaw and lower teeth
?
a
313
ft
ft
&
"126
1
"196
>
1
D
34
0.t-.
|i
8.2
D
33
8
a
695
h
N
538
'c'
g
l|
i
o
76
1
'750
1
i
f
W
111
'175
165
290
242
'185
190
349
S
122
s
232
11
o
38
S
jl
■a-
o
33
g
ft
a
Brontotherium ?tichoceras, Nat. Mus.
8313, &
90[ 126
'110 '127
593
M. coloradensis (type) _.
M. ("Symborodon") acer. Am. Mus.
6350, 9 (typeofM. "altirostris") ---
330
127
215
'630
610
'640
'690
750
'640
44
60
70
=90
117
114
115
130
M. ("Symborodon") acer. Am. Mus.
6348, cf (type)..
M. acer, Univ. Wyoming 2, cf (?)
300
'300
315
317
«294
99
noo
116
119
106
200
200
204
198
195
646
'665
645
543
«665
'520
84
'81
M. ("Symborodon") bucco. Am. Mus.
6345a, d' (lectotype)
M. bucco, Nat. Mus. 4705, 9 .. .. _._
22
29
28
20
27
80 135
M. bucco. Am. Mus. 6353, cf -
M copei, Nat. Mus. 4711 (type)
'620
=601
80 125
=300
M. riggsi. Am. Mus. 6364 (type)
282
85
194
'465
1
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
543
Sequence qf species. — The following summary affords
the principal characters by which the chief species
so far recognized may be distinguished from one
another:
1. Megacerops coZoradensis Leidy is the genotype. The type
is lost and is known only from the figure of the coossified nasals
and horns, which indicates an animal about the size of M. bucco.
Titanotherium zone, Colorado.
2. Megacerops (" Symborodon") bucco (Cope) is a more pro-
gressive species. It is decidedly brachycephalic. It still retains,
however, the long thin nasals. The horns have shifted to a more
anterior position. The buccal processes of the zygomata now
become very prominent, as in the brontotheres. The occipital
pillars begin to expand, the parietal crest is broader, and as a
distinctive, progressive character, the tetartocones of the pre-
molars are more developed. It is important to note that this
extreme type probably belongs to the upper beds. Colorado
and South Dakota. Titanotherium zone, level Chadron B.
3. Megacerops acer Cope is represented by the male type
skull and by a female skull which Cope made the type of his
species M. altirostris. It is also mesaticephalic and differs from
M. copei in the thick, short nasals, in the divergence of the
horns, in the somewhat more anterior position of the horns,
while it resembles M. copei in the small size of the tetartocones
of the premolars. The occiput (PI. CLVII, C; fig. 451) is
readily distinguished from that of any brontothere by its slender
characters, indicating that the muscles of the neck were not
so robustly developed in these animals. Colorado and South
Dakota. Titanotherium zone, upper beds.
4. Megacerops copei (Osborn), named in honor of Professor
Cope, appears to belong to the middle Titanotherium zone (Chad-
ron B). In proportions the skull is mesaticephalic; the horns
are vertical, elongate as seen from the front, and their form
suggested the unique name "rabbit's ears," which is applied
to this skull. They are placed typically directly above the
orbits, yet the nasals are very thin, and, as shown in the section
(fig. 450), there is no connecting crest. Colorado and South
Dakota. Titanotherium zone, upper beds (Chadron C).
5. Megacerops assiniboiensis Lambe is an animal of small
size, known only by the jaw. Saskatchewan, Swift Current
Creek. Titanotherium zone.
6. Megacerops riggsi Osborn is distinguished by its especially
short massive jaw. Colorado. Titanotherium zone.
Synopsis of specific cJiaracters qf Jour species of Megacerops
M. copei, Nat. Mus. 4711
M. altirostris, Am. Mus. 6350 '
(type)
Skull
Nasals
Horns
Horns, malar ridge
Horns, top section
Horns, connecting crest
Horns, position
Internal flange of buccal
process.
Occiput and pillars
Occiput pits
Occiput parietal crest
Premolars
Mesostyle p'
Mesaticephalic
Thin
Vertical
Prominent
Strongly oval
None
Over orbits
Not prominent
Narrow
Absent?
Narrow
Tetartocones smaller.
Absent
Extremely brachy-
cephalic.
Thin, cf long, 9 medium
Divergent
Absent
Rounded to oval
None
c? anterior to orbits
9 ?intermediate
Prominent
Expanding
Absent?
Broad
Tetartocones more de-
veloped.
Present
Mesaticephalic?
Thick, short
Divergent
Intermediate
Oval
Well-defined
[Anterior to orbits
(cT?
( 9 not prominent
Narrow
Pronounced
Narrow
Tetartocones smaller.
Mesaticephalic?
Thick, short.
Divergent.
Faint or absent.
Rounded to oval.
None.
Intermediate.
Narrow.
Absent?.
Intermediate.
Tetartocones smaller.
Present.
o A small female.
Jaw cTiaracters. — The jaws of Megacerops are read-
ily distinguished by the abbreviation of the anterior
portion corresponding with the extreme abbreviation
or brachyopy of the facial region, with which the ab-
breviation of the lower premolar series is also corre-
lated. A second distinction is the narrowness of the
chin and the small size and approximation of the
*• Regarded as the female of M. acer.
canines, correlated with the reduction of the lower
incisor teeth. The anterior part of the face in Mega-
cerops is contracted and the lips were probably narrow
and pointed in contrast with the broad, massive lips
of Brontotherium.
Details of the contrasts in the character of the jaw
are as follows:
Jaw characters in Megacerops and Brontotherium
Megacerops
Brontotherium
Short, massive
Shallow
Broad, posterior border vertical
?It
Absent?
Symphyseal region in side view
Very shallow.
Broad, posterior border often vertical.
Pi in fully adult jaws
Diastema in front of pi
Absent.
?
Swollen at base with massive posterior cin-
gulum.
Typically pronounced.
Upward flexure of premolar series
544
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Megacerops coloradensis Leidy
Text figures 164, 448
[For original description and type references see p. 20S|
Tyfe locality and geologic Tiorizon. — Colorado; Ti-
tanotlierium zone. Exact locality and level unknown.
Specific characters. — Nasals long (110 mm., esti-
mated), sharply decurved and tapering distally; horns
cylindrical with slightly flattened tips, external length
175 millimeters (estimated). Connecting crest very
low.
Materials. — The type specimen, consisting of the
coossified nasals and horns, has been lost, and the
characters of the genus rest upon Leidy's original de-
scription and figures and on the model (Am. Mus.
9018, fig. 448) made from them.
Transitional referred specimen, Brontotherium ticho-
ceras. — In the National Museum there is a beautifully
preserved skull (No. 8313,
with associated lower j aw)
which in horn structure
recalls the type of M.
coloradensis, except that
the nasals and horns are
somewhat shorter. This
skuU is of extraordinary
interest inasmuch as it
combines the nasals, horns,
and absence of connec-
ting crest which are char-
acteristic of Megacerops
with the large incisors
and canines which had
T? AAo a„+;„„„.,„^ hitherto been regarded
Figure 448. — Sections and _ *=
contours of nasals and horns as characteristic only of
of Megacerops coloradensis. BrontotJierium. The pre-
Ara. Mus. 9018, a model made trom molars are relatively Small,
Leidy's figures and from measurements . " .
of iiis type, wiiich lias been lost. One- and the internal cmgula
seventh natural size. j^^^^ ^^^^^^^ completely
disappeared, much more than in BrontotJierium and even
more than in most Megacerops skulls. The measure-
ments of this skuU approach those of the type of
Megacerops hucco except that the basilar length is
greater, the zygomatic width and horn length are
less. It should also be compared with Brontotherium
hatcheri, but it differs from that type in its shorter
horns, absence of connecting crest, and absence of in-
ternal cingulum on the premolars.
Megacerops bucco (Cope)
{Symhorodon bucco Cope, 1873; "Symhorodon iorviis" Osborn,
1902)
Plates CXLIV, CXLV, CLIII-CLVI; text figures 169, 170, .378,
392, 393, 399, 449, 719, 744
[For original description and type references see p. 2121
Type locality and geologic horizon. — Cedar Creek,
Logan County, Colo.; Titanotherium zone.
Specific and generic characters. — I^l^j P^. Superior
incisors probably absent; premolars with cingula
vestigial or wanting, tetartocones nearly as large as
deuterocones and more distinct then in M. copei,
a faint mesostyle upon p*. Skull shows progressive
increase in size over M. copei, length nasals to occiput
750 millimeters, premaxillaries to condyles 665 (esti-
mated); extremely brachycephalic, index 100 (esti-
mated); nasals thin, medium in length, broad (90 by
130 mm.); no connecting crest between horns; horns
c? short (185 mm., estimated); zygomata broadly ex-
panded; occipital pillars expanding, with wide superior
rugosities.
SlcuU and jaw selected hy Cope as types of Symhorodon
hucco. — The type skull of S. hucco Cope includes a
fairly preserved skull (Am. Mus. 6345a) with enor-
mous zygomatic expansions. The jaw placed with
this skull by Cope appears to belong to another
phylum. The premaxillaries are wanting. All the
maxillary teeth on both sides are preserved with the
exception of p^ The horns and the anterior portions
of the orbits have required much restoration but
serve to afford some very distinctive characters. The
skull is readily distinguished as a Megacerops by the
entire absence of the connecting crests between the
horns, which are set widely apart but so damaged
that a perfect basal section can not be made. The
nasals are very thin and of medium length. The
zygomata arch very widely and exhibit a flattened
section which is even more extreme than that of
Brontotherium gigas and quite distinct from the con-
vex section of the skull Am. Mus. 6346, also referred
by Cope to S. hucco but regarded by us as pertaining
to a species of Brontotherium, probably B. curium.
Posteriorly the cranium broadens out into the occiput,
but it is important to note that this breadth and the
peculiar zj'gomatic section are altered by crushing.
Dentition. — The animal is in the eighth stage of
growth, all the internal cones of the teeth being worn
except upon m^. The external cingulum is obsolete
throughout the grinding series, as in M. acer. The
internal cingulum is entirely lacking on p^ and is
very slightly marked on p^. It is also wanting on the
inner sides of the molars. As a marked progressive
character, the internal cusps of the premolars are
robust and well roimded. The hypocones on m^ m-
are prominent and bidge on the lingual side. The
hypocone is well developed on m^, triradiate in form,
and connected with the cingulum by only a slight
posterior ridge.
Female slcull of M. hucco in the National Museum. —
The skull Nat. Mus. 4705 is that of an aged female in
the ninth stage of growth, the horns of which are set
very wide apart (PI. CXLV, B; figs. 392, E, 449, B).
The well-worn canines measure only 25 millimeters
anteriorly and are proportionately the smallest we
have observed in any species of titano there. The
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
545
evidence regarding the incisors is not positive, but
there were apparently two vestigial teeth, which
certainly could not have been functional. The
premolar-molar series measures 315 millimeters. The
internal cingula on the premolars is vestigial or en-
tirely wanting. The skull appears to be brachy-
cephalic, the estimated index being 81; the measure-
ment from condyles to symphysis is 645 millimeters,
while the width across the zygomata is estimated at
525 millimeters, an excess of length over breadth
of only 120 millimeters. The nasals are very broad
but at the same time short, the free length being only
80 millimeters. The horns are short, measuring 190
millimeters on the outer side, and exhibit at the base
the section so characteristic of this genus, which is
due to the flatness of the anterior or maxillary face
and the convexity of the posterior buttress.
Megacerops acer Cope
(Megaceralops acer Cope, 1873; Symborodon altirostris Cope,
1873; "Symborodon acer Cope," Osborn, 1902)
Plates XVIII, CXLVI-CLII, CLVII; text figures 167, 170,
375, 378, 390, 392, 399, 400, 450-453, 625, 638
[For original description and type references see p. 211]
Type locality and geologic Tiorizon. — Cedar Creek,
Logan County, Colo.; Titanotherium zone, level un-
known.
Specific and generic characters. — I^^, P^^. Incisors
greatly reduced but more persistent than in M. hucco;
canines small; p^ small, rounded, p^-p^ with tetarto-
cones distinct but smaller than in Menodus torvus,
a prominent mesostyle upon p^ (No. 6350), narrow
internal cingula; hypocones of m^-m^ large, projecting
on lingual side, hypocone of m' triradiate. Nasals
thick, short, and moderately broad (60 by 114 mm.);
horns c? 290 millimeters, 9 165, basal section
typical, summits transverse oval; cranial vertex
dolichocephalic; occiput high, narrow, not deeply
indented superiorly. Zygomata 9 with buccal proc-
esses flattened or slightly concave above, size mod-
erate, nasals to occiput 640 millimeters. This differs
from M. copei in the form and direction of the horns
and presence of a connecting crest in the male, in the
long, narrow parietal vertex, and in the thiclaiess of
the nasals.
Materials. — This was the second species of "Sym-
horodon" described by Cope from Colorado in 1870.
The type specimen (Am. Mus. 6348) is the skull of a
male titanothere of medium size but with long horns,
from the Titanotherium zone of Cedar Creek, Logan
County, Colo. The exact geologic level is not known.
Subsequently Cope proposed the name "Symborodon"
altirostris for a skull of the same species (Am. Mus.
6350), a female, as first pointed out by Osborn in
1896. In addition to these two skulls there is for-
tunately a third (Am. Mus. 6349), probably also a
female, in which the right zygoma is preserved.
Relationships to other species of Megacerops. — There
can be no question as to the generic relationship of this
animal to Megacerops hucco and M. copei. It presents
a number of interesting and significant points of con-
trast and agreement. The affinity is seen especially
in the form and position of the horns in the type, which
indicate that they were used in the manner character-
istic of other members of the genus. They point up-
m^.
B
Figure 449.-
-Sections and contours of skull of Megacerops
bucco
A, Am. Mus. 6345a (type); horns directed upward and forward, their basal section
rounded with a flattened external face, the section of the upper part of the horn
transversely oval, nasals wide, zygomata widely expanded. B, Nat. Mus. 4705.
In this supposed female the horfls, as in all other members of this genus, are sub-
cylindrical, their basal section has a flat external face, and the remainder of the
section is well rounded; nasals rather short; zygomata moderately expanded.
One-seventh natural size.
ward and slightly outward, and as they are strength-
ened posteriorly their bases do not lie very far in
front of the orbits. This relatively long-horned titano-
there was, therefore, a true Megacerops, a conclusion
which is further supported by the resemblance of the
basal horn section to that of "Menodus torvus " and by
its wide contrast to that of any species of Menodus, such
as M. trigonoceras, or of Brontotherium, such as B. gigas.
546
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Seen from in front the horns are placed close together
at the base and diverge more at the summits than in
the other species, although the divergence is less than
in B. gigas, for example. That this divergence was
correlated with the development of a shallow trans-
verse connecting crest in the males, which was also
present in certain females (Am. Mus. 6349) but nearly
wanting in the supposed female Am. Mus. 6350 (PI.
millimeters and are relatively much shorter than in
M. bucco and M. copei. This stage is therefore parallel
to that of B. curium among brontotheres. The
crania of these small animals are rather mesatice-
phalic than brachycephalic. The name altirostris,
given to the female skull by Cope, referred to the
elevation of the nasals above the premaxillae when
seen from in front.
B
A
Figure 450.-
c
-Sections and contours of skulls of Megacerops copei and M. acer
A, Megacerops copei, Nat. Mus, 4711 (typo) ; horns long and subvertica], their basa] section thiclc, the external face flat, and the remainder of
the section well rounded; the section of the upper part of the horn widely oval, with a recurved external angle; no connecting crest; nasals
larger and broader than in ^f. acer and narrow at the end.
B, M. acer, Am. Mus. 6348 (type); horns long and slightly inclined forward, their basal section roundly quadrate with a concave external
anterior face; the section of the upper part of the horn transversely oval; a connecting crest; nasals short. C, M. acer, Am. Mus. 6350,
2 (type of Symborodon altirostris Cope, now referred to M. acer): horns relatively short and directed forward, basal section roundly
quadrate, with a long concave external face, section of the upper part of the horn transversely oval, nasals short and thick.
All one-seventh natural size.
CL), proves that it was adapted to resist the lateral
strains to which the horns were subjected. Yet in M.
copei and M. hucco the horns are divergent and the
nasals are as thin between the horns as toward the tips.
Nasals. — Correlated with the somewhat more an-
terior position of the horns and partly by compen-
sation of growth there is a very marked abbreviation
of the nasals, which now measure only 60 by 114
Occiput. — The occiput resembles M. copei also in
the form of the top of the occiput (fig. 451), which is
not expanded and, in fact, is peculiarly narrow and
simple. Instead of the knobs (fig. 378, F) seen in
Brontotherium there is a pair of pits (fig. 451) on either
side of the center of the occipital crest. The occiput
lacks the broad superior flare, which is correlated with
brachycephaly and the lateral motions of the skull.
EVOLUTION OP THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
547
A very distinctive feature is the fact that it is not
incurved in the median line superiorly.
The buccal expansion of the zygoma is preserved
only in one of the female skulls (Am. Mus. 6349), in
which it is seen to be broad and flattened above, as
in M. hucco, paralleling that of Brontotherium, gigas.
The portion of the malar preserved in No. 6350 is
stout and rounded.
Teeth. — That the superior incisors were greatly
reduced is demonstrated by the small alveoli in the
female skull Am. Mus. 6350, which, however, are
not so small as in Menodus. The rounded first pre-
molar resembles that of M. copei; the slightly defined
cingulum on the inner side of the premolars and the
rather retarded development of the tetartocones dis-
tinguish these teeth from those of M. hucco. There
is a well-marked mesostyle on p''. As in M. hucco
and M. copei, the distance from the anterior edge of
m' to the premaxillary border and the proportionate
length of the premolar series arc excessively short, com-
pared with Menodus, Allops and Brontops, and even
compared with Brontotherium (Pis. CLI, CLII, and
CLXV). The upward flexure of the premolar series
and elevation of the canines and incisors is very
marked and in connection with the elevated position
of the nasals and perpendicularity of the horns must
have imparted a peculiar retrousse character to the
face (PI. CXLVI).
Sexual characters. — The males are well distinguished
from the females by the length of the horns, which is
290 millimeters in the male (Am. Mus. 6348) as com-
pared with 177 in one female (Am. Mus. 6350) and
138 in the second female (Am. Mus. 6349). The basal
section of the horns is substantially similar, and the
summits are transversely oval in section; but in one
of the female skulls (No. 6350) the horns project
forward without recurvature, while in the other they
are recurved as in the male skull. There is a strong
convexity of the cranial vertex in the frontoparietal
region (a family character) in the male No. 6348 and
the female No. 6350, which is feebly developed in
the female No. 6349. The horn of the latter has a
malar ridge.
Additional ohservations on skulls of Megacerops acer. —
The type of M. acer is a relatively long and slender
male skull (Am. Mus. 6348) lacking all the teeth,
the zygomatic arches, the maxUlaries and premaxil-
laries. The skull is sharply distinguished from an}^
previously described by its obtuse and short but
tapering nasals, long, recurved horns with deep an-
teroposterior basal section and oval tips, a narrow
vertex, and narrow, simple occiput. In addition to
these characters there appears to be a strong convexity
in the central portion of the top of the cranium not
altogether due to crushing, which is characteristic of
Brontotheidum and Megacerops. The basal section of
the horns indicates that they are strengthened not only
by a connecting crest but by a posterior ridge passing
backward above the orbits, which is separated by a
flat surface from the malar ridge.
The skull Am. Mus. 6350 resembles that of M.
acer in the abbreviation of the nasals and narrowness
of the vertex and the occiput and differs from M.
acer in characters which Cope thought to be specific
but which are now found to be sexual. These female
characteristics are the relatively short horns and com-
paratively low connecting crest, as shown in the section
(fig. 450); the only difference of note is the convexity
between the malar and posterior ridges of the horns.
The anterior teeth of this type have been broken
away and lost since the original description. The
superior incisors are represented by two small alveoli,
placed upon a nari'ow border, indicating that these
teeth were small and disappeared in old age. The
formula was, therefore, I^"^, P-^. The crowns of the
canines have been destroyed. These teeth were of
small size, apparently as in M. hucco. The first pre-
molar is a small tooth with three cusps (protocone,
deuterocone, and tritocone); the second, third, and
fourth premolars e.xhibit four well-developed and dis-
FiGURE 451. — Upper part of occiput of
Megacerops acer
Am. Mus. 6351, showing pits tor the ligamentum nuchae
and rugosities for the recti capitis laterales. One-fourth
natural size.
tinct cusps, including a convex tetartocone. A fea-
ture which may be specifically characteristic is the
distinct mesostyle upon p*. A crenulate and not
sharply defined cingulum is observed on the inner
side but is entirely wanting on the outer side of the
premolars. This more or less vestigial condition of
the cingulum is also a distinctive character of the
series to which this animal belongs, as well as of the
old males of the species of Brontotherium of the upper
beds. The molars are very broad. On m' the hypo-
cone is triradiate, well developed, but not distinct
from the cingulum. Another well-marked feature is
the prominence and roundly blunt character of the
hypocones of m'-m'.
A female skull (Am. Mus. 6349) fortunately pre-
serves the right zygomatic arch, indicating that this
is broad and somewhat flat superiorly, having a sec-
tion similar to that of Brontotherium gigas but less
robust. A distinctive character is the breadth of the
postglenoid process. The occiput has the relatively
slender proportions seen in the type and cotype. The
rugosities on top of the occipital pillars are much less
extensive than in Megacerops rohustus.
548
TITANOTHERES OF ANClfiNT WYOMIMG, DAKOTA, AND NEBRASKA
Referred, skull, Megacerops acer, University of Wyo-
ming Museum. — ^A very fine skull (figs. 452, 453) in
the University of Wyoming, collected by Mr. W. H.
Figure 452. — Skull of Megacerops acer
Dniv. Wyoming Mus. 2. Side (Ai) and top (Az) views. One-sixth natural size. This sliiill combines the
characters of several supposed species of Megacerops. In the form and length of the horns it approaches
the type of M. acer; in the character of the nasals it resembles the type of Symborodon attirostris; in the
general form of the skull top it resembles especially the supposed female of M. biicco (Nat. Mus. 4705);
and in the dimensions of the grinding teeth it agrees with the lectotype of M. bucco (Am. Mus. 6345a).
Eeed in the upper Titanoiherium zone near Alcova,
Natrona County, Wyo., has somewhat smaller horns
than the type. Its nasals recall those of the type of
S. altirostris. The premolars, as in other members of
the genus, have the internal cingula nearly obsolete
and the tetartocones subequal to the deuterocones;
the width of the molars exceeds the length. The zygo-
matic index (84) is high. In addition
to the measurements of this skull
given above (p. 542), we record also
the following:
MilUmeters
Diastema behind canine 20
PS ap. by tr 34X45
M2, ap. by tr 73X84
M3, ap. by tr 78X83
Molar inde.x 29
Megacerops copei (Osborn)
{Symborodon copei Osborn, 1908)
Plates CXLIII-CXLV; text figures 24, 203,
390, 394, 399, 450, 454, 640
[For original description and type references see p. 235]
Geologic horizon — Middle Titano-
iherium zone. South Dakota.
Specific and generic characters. —
JZiO.^ p4 Incisors (type) persistent
but greatly reduced; canines very
small, reduced, c? 28 millimeters;
premolars with cingula reduced or
absent; tetartocones connected with
deuterocones by a longitudinal ridge.
Skull: Nasals thin, short and broad
in proportion, 80 by 125 millimeters;
horns S 300, no connecting crest,
transverse oval near summit; buccal
processes of zygomata c? stout and
convex; malar in front of buccal
process very deep, beneath postorbital
process stout, convex; occipital pillars
medium, not greatly expanded at the
summits. Size small, premaxillaries
to condyles 620 millimeters.
Materials. — This species is known
only from the type skull (Nat. Mus.
4711, skuU V), which is finely illus-
trated in Plates CXLIV, CXLV. It
is from the middle Titanotherium zone,
probably from the upper levels or
even lower, and impresses one with
the excessive thinness of the nasals,
which in these animals are very broad
posteriorly (120 mm.), while not of
very great free length. The animal
is in the eighth stage of growth
and has thus attained its fully adult
characters. The small size of the canines (28 mm.)
might lead us to regard it as a female, but all the
specimens belonging to this genus are characterized
by small canines, and the great length of the horns
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
549
(301 mm.) proves that it is a male. The incisors
were apparently very small and shed early, being
indicated solely by a single alveolus on each side.
Figure 453. — Skull of Megacerops acer
CJniv. Wyoming Mus. 2. Palatal view. One-sixth natural size. Shows striking
resemblance to the supposed female of M. bucco (Nat. Mus. 4705) and to the type of
Symborodon aUirosins, Affinity with Broittotlierium is also indicated in the form
of the canines, the highly progressive submolariform premolars, and the broad
molars.
The canines are similar to Brontotherium canines
but of much smaller size. This degeneration of
the anterior teeth was shared by the first premo-
lar, which is a small tooth placed close to the
canines; in spite of its small size, however, it very
probably had a well-developed tetartocone as in
all other Brontotheriinae, indicated by the angu-
late postero-internal border of the much worn
tooth. The succeeding premolars, p^-p"*, exhibit
progressive development of the tetartocones (Pis.
CXLIII, CXLV); these cusps are still connected
with the deuterocones by a low crest, however,
and are not clearly distinct; the internal cingula
are feebly developed and obsolete opposite the
deuterocones. M^ has a heavy triradiate hypo-
cone. The grinding series as a whole measures
300 millimeters, exactly as in M. hucco, but the
skull of this specimen is 130 millimeters shorter
than in M. hucco. It would appear that in this
species, as in Brontotherium, the dental series does
not increase proportionately with the skull.
Skull. — The lateral view of the type male skull
(PL CXLIII) at first sight strongly suggests that of
B. gigas; but a close comparison reveals the unpor-
tant difference that the horns are placed very much
farther back; they are almost directly above the
orbits with buttresses extending backward at the
base over the orbits. The horns thus obtained a
firm support posteriorly, a mechanical adaptation
that compensates for the entire deficiency of a trans-
verse connecting crest. When seen from in front
(PI. CXLIV) the horns are chiefly vertical and parallel,
although this is somewhat exaggerated by lateral
crushing; they certainly do not flare outward as in
B. gigas. This unique position correlated with the
narrowness of the occiput indicates chiefly a vertical
motion of the skull in using the horns and explains
the absence of a connecting crest, which would be of
great service in protecting the nasofrontal region
from fractures occasioned by side blows. The horns
differ from those of M. acer in the stronger develop-
ment of the transverse oval section near the summits
caused by the sharp development of the malar ridge.
Megacerops assiniboiensis Lambe
Text figures 205, 434, 455
All that is known of this animal is fully stated on
pages 239-240. The lower jaw is shown in Figure 455.
Megacerops syceras (Cope)
(Menodus syceras Cope, 1889)
Text figures 186, 456
[For original description and type references see p. 226]
This species is known only from the coossified
nasals and horn cores. It is probably a member of
this phylum.
Figure 454. — Restoration of Megacerops copei
By Charles R. Knight. About one-ninth natural size.
550
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Megacerops riggsi (Osborn)
Plates CLVIII-CLX; text figures 208, 455_
I For original description and type references see p. 242]
Type locality and geologic horizon. — Horsetail Creek,
northeastern Colorado ; Titanotherium zone.
Specific characters. — Of small size, smaller than any
Imown individual of Megacerops or Brontotherium . A
Figure 455. — Lower jaws of Megacerops assiniboiensis and
M. riggsi
A, Megacerops riggsi, Am. Mus. 6364 (type). A small titanothere having a massive
short jaw, a short chin, and a swollen ramus. (See PI. CLVIII, A.) The cheek
teeth lack cingula. B, if. assiniboiensis, Ottawa Mus. (type). A small short
jaw provisionally referred to Megacerops. The grinding teeth lack e.xtemal
cingula. One-flfth natural size.
very massive jaw with a small coronoid process and
a very short symphysis. Premolar series greatly
abbreviated (85 mm.). Premolars and molars with
reduced external cingula.
The type of this species, named in honor of Mr.
E. S. Riggs, of the Field Museum of Natural History,
is a jaw in the Cope collection (Am. Mus. 6364) which
was wrongly referred by Cope to his species "Symhoro-
don" acer. It represents a highly specialized and
small form of Megacerops.
SUBFAMILY BRONTOTHERIINAE
Titanotheres attaining the largest size, chiefly of
the upper Titanotherium zone, although known from
the lower beds upward. The horns progressively
longest, most broadly oval, and flattened, shifting
forward with the extreme reduction of the free portion
of the nasals; very prominent connecting crest, pro-
gressively increasing size. Two pairs of persistent
superior incisor teeth in males; canines large, obtuse.
Brachycephaly expressed in the great horizontal
expansion of the buccal processes, in the proportions
and arching of the grinding teeth, and in the inclosure
of the auditory meatus in the males. Occiput much
produced behind the zygomatic arches. Vertex of
skull in superior view elongate.
The ancestry of this great Oligocene phylum may
possibly be found in the upper Eocene, perhaps in
species of Diplacodon or of Eotitanotherium, described
on pages 439, 441.
In these huge animals the titanothere family reached
a climax. The generic name "thunder beast," based
on the genotype species Brontotherium gigas, is highly
appropriate because it applies to the most robust and
most massively horned not only of the titanotheres
but of all the known Perissodactyla.
Marsh mistakenly associ-
ated with the type jaw of B.
gigas the skull of Menodus, "B.
ingens," to define the generic
characters of Brontotherium,
and he assigned the actual
skuU of B. gigas to a different
genus, naming it Titanops
elatus; but we have found that
the lower jaw of B. gigas is in
a stage slightly antecedent to
that of Titanops elatus Marsh.
We have also discovered that
this great animal Brontothe-
rium gigas {elatum) is a central
form, whose ancestors {B. leidyi)
extend down to the base of the
Titanotherium zone and whose
successors {B. platyceras) extend
up to the very summit of Chadron
C (the upper Titanotherium zone).
Thus in the present memoir
Brontotherium is shown to embrace
a most remarkable and nearly
monophyletic series or succession
of species, eight or possibly nine
of which are now known from the
lower to the higher levels, present-
ing certain common generic char-
acters throughout. Modified by a
progressive increase in size and by a series of remarkable
stages in the evolution of the horns, in the recession of
the nasals, and in the expansion of the buccal processes
of the zygomata, they culminate in a unique type of
perissodactyl skull, that of B. platyceras, which appears
to be the last of its great race. This species is cer-
tainly in the last stage of evolution along its line.
Affinities. — The nearest allies of Brontotherium are
members of the genus Megacerops, but the true
brontotheres (the males at least) are readily dis-
tinguished by the presence of two pairs of upper and
lower incisor teeth, by their robust canines, the
Figure 456. — Sec-
tions and contours
of skull of Mega-
cerops? syceras
Ottawa Mus. (type), provi-
sionally referred to Mega-
cerops partly because the
basal section of the horns
is roundly quadrate and
has a flat external face.
Nasals of moderate length,
connecting crest low.
One-seventh natural size.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
551
flattened oval section of the horns, the progressive
connectuig crests between the horns, and ia general
by their superior if not dominant size. The jaw of
B. gigas (PL CLXXI) is very characteristic of the
genus, with its recm-ved coronoid process, shallow
chin, robust mandibular section, two pairs of incisor
teeth, and obtuse rounded canines. The finely pre-
Horns. — The males of all these animals exhibit
progressively broad and flattened horns, transversely
oval from base to summit, diverging widely at the
summits, with the great connecting crest rising be-
tween them (unlike the typical Megacerops), to resist
lateral strains, and evolving at the expense of the
constantly diminishing free portion of the nasals
(fig. 457).
Skull. — The top view of the cranium is very long
(fig. 459), owing to the extreme forward position of
the horns and the great backward extension of the
occiput behind the zygomatic arches. The cranium
proper is dolichocephalic; the expansion of the buccal
processes of these arches is, however, so broad that
the total breadth of the skull finally equals the total
Figure 457. — Composite sections showing tlie evolution of the horns and reduction of the free nasals in the
Brontotherium phylum
a, Broniothenum leidyi, Nat. Mus. 4249 (type), Chadron A 2; &, Brontoihenum hypoceras, Nat. Mus. 4273, Chadron A 2; c, BTonlotherium hypoceras,
Nat. Mus. 4702, Chadron A 37; d, Brontotherium hatcheri. Am. Mus. 1070, Chadron; e, Brontotherium gigas, Yale Mus. 12061 (type of B. elatm),
upper (?) Chadron; f, Brontotherium gigas, Am. Mus. 492, Chadron C; g, Brontotherium curtum, Yale Mus. 12013 (type), Chadron C; h, Brontothe-
rium curtum, Nat. Mus. 4946, Chadron C 3; i, Brontotherium ramosum. Am. Mus. 1447 (type), Chadron C; j, Brontotherium platyceras. Am. Mus.
1448, Chadron C. All one-fourth natural size. In the earliest stage, B. leidyi, the horns are small and placed near the orbits, there is no connecting
crest, and the nasals are long and slender; in the latest and most specialized stage, B. platyceras, the horns are extremely long and placed far in
front of the orbits, the coimecting crest is very high, and the free portion of the nasals is practically vestigial. Between these extremes lie a con-
siderable number of intermediate stages.
served type skull of B. (Titanops) elatus (PL CLXXV)
enables us fully to characterize the genus. The suc-
ceeding and stni more advanced stages in the evolu-
tion of this phylum were originally named as foUows:
Menodus dolichoceras Scott and Osborn, Titanops
curtus Marsh, Titanops medius Marsh, TitanotJierium
ramosum Osborn.
101959— 29— VOL 1 38
length, the zygomatic index in B. platyceras being
110, or hyperbrachycephalic. This condition is best
imderstood when the skuU is viewed from below
(fig. 393). The whole structure of the skull, including
the broad and spreading occiput (fig. 378) and the
excessive buccal expansion, is adaptively adjusted to
the development of the horns, which from their feeble
552
TITANOTHERBS OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
development in the females (as in the type of Menodus
peltoceras Cope {=lBrontoiherium curtum, figs. 459,
477, 478) are judged to have been fighting weapons
in the males.
FHarv.Mu5..type
EYaleMu5.l20l3,type
D Yale Mu5. 12061
CAM. 1070
B Nat.Mus.427Jn«)type
ANat.Mu5.4249,t^pe I -^ ^y nhypoceras AZJi.n^^otyf^
lleidyi N.M. 4249, type
Figure 458. — Basal sections of the horns in the Bi-ontotherium phylum
A, Brmlotherium Icidyi, Chadron A 2; B, £. hypoceras, Chadron A 2; C, B. Jiatcheri, Chadron; D, B. gigas, Chadron
?C; E, B. curtum, Chadron C; F, B. platyceras, Chadron. These sections show the progressive change from
the small anteroposterior oval of B. Uidyi to the very wide transverse oval of B. platyceras.
increase in size of the skull as a whole (24 per cent).
The total length of the premolar series diminished,
but in B. gigas the individual molars are enormous —
for example, m^ (ap. by tr.) 91 by 94 millimeters, as
Figure 460.- — Contrast in contour of horns and
nasals between male and female brontotheres
The female (a), Am. Mus. 1006 (referred to B. gigas), differs
from the male (b) , Am. Mus. 492 (referred to B. gigas),
chiefly in having short horns and nasals, combined with a
high connecting crest.
Figure 459. — Skulls cf male and female brontotheres
Top view. About one-ninth natural size. A, Broniotherium platyceras. Am. Mus. 1448 (cotype) . In this male skull vertical crushing has
emphasized the width of the zygomata. B, B. curtum (peltoceras). Am. Mus. 1006. This female skull is small, the horns are short,
and the zygomata are unexpanded.
Grinding teeth. — The grinding teeth suffered prob-
ably from the highly "selective value" placed upon
the horns and were retrogressive in many respects;
the total increase of length in the grinding series
(21 per cent) thus did not fully share in the progressive
compared with M. giganteus (Am. Mus. 505), m^ 87
by 84. The breadth of p* was also exceptionally great,
and nowhere else do the tetartocones attain such
vigorous development as in B. gigas, not excepting
Menodus giganteus.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
553
Sexual characters of brontotheres
Horns elongate.
Connecting crest very high.
Occiput very robust and back-
wardly produced.
Zygomatic arches very broad.
Incisors persistent.
Canines larger.
Horns abbreviate.
Crest less developed.
Occiput less robust and less
produced backwardly.
Zygomatic arches less broad.
?Incisors less persistent.
Canines two-thirds size of
male canines.
The contrast between the males and females gener-
ally is well brought out in the accompanying figures
of the male and female skulls of B. gigas. It is obvi-
ous that a number of correlated characters disappear
in the nondevelopment of the horns in the females,
especially the marked width of the zygomatic arches,
which may have been partly defensive structures, and
the nonextension of the occiput posteriorly, which was
probably designed in the males for the support of the
great muscles of the neck.
Standard measurements in the BrontotJierium 'phylum, in millimeters
Upper teeth
Skull
Jaw and lower teeth
1
ft
1
1
S
I
1
a
a
§
s|
1
o
i
3
e
1
tsi
13
1
D
3
.ft
1
ft
1
i
£
1
1
■i
S
w
?
S
■3
1
■a
S
3.
0
.1
ft
s|
1
0
s
1
1
B. platyceras, Am. Mus. 1448, cf
337
»123
221
40
30
'■728
^815
....
'■693
20
81
"433
MOO
-390
'399
380
365
365
355
160
B. platyceras, Field Mus. 12161, cf
B. ramosum. Am. Mus. 1447, cf (type).-
B. curtum, Yale Mus. 2013, c? (type).
340
350
350
120
°137
223
216
'■228
36
29
880
741
"790
710
'775
-610
80
?78
895
665
795
38
40
52
167
95
°105
34
— -
B. curtum, Nat. Mus. 4946, &
348
128
224
218
222
228
211
235
197
197
■"
<'840
'■780
'■673
620
»670
545
74
— -
65
105
B. curtum Nat. Mus 1211, cf
B. curtum, Am. Mus 1005, ?
345
345
130
129
27
20
-80
635
44
101
B. curtum, Nat. Mus. 1232 cf
R. ciirtiim?, Rrit. Mns. iifi2f)
34
38
31
700
825
563
80
"620
34
45
85
110
163
320
B. medium, Nat. Mus. 4256, cf (type)..
B. medium?, Nat. Mus. 4699
365
293
305
138
117
106
B. medium?, Nat. Mus. 4716, 9
B. dolichoceras, Harvard Mus., cf (type)
B. gigas. Am. Mus. 492, cf
49
87
90
115
310
365
353
356
343
348
126
132
133
"■133
241
218
218
227
35
35
33
32
23
24
830
773
770
775
"SIO
740
640
87
82
793
B. gigas (hatcheri), Carnegie Mus. 341..
B. gigas (hatcheri), Nat. Mus. 4262, cf --
B. gigas, Nat. Mus. 4244
70
67
90
110
107
105
275
170
350
350
»121
235
33
28
625
657
84
728
805
B. gigas, Yale Mus. 12061 (type
260
258
365
117
34
30
600
B. gigas?, Am. Mus. 1006, $
335
330
127
217
»37
22
715
660
800?
79
65
85
90
100
105
72
74
114
118
56
88
130?
102
126
120
107
118
105
0
180
B.? tichoceras. Harvard Mus. (type)
B. hatcheri. Am. Mus 1070 cf
132
118
118
192
198?
35
31
265
200
220
250
143
'■165
104
107
260
375
-120
263
35
29
fiOO
B. hatcheri, Field Mus. P 5926, cT
»310
320
-710
530
646
620
540
— -
760
B. hatcheri?, Univ. Wyoming 1 cf
B. hatcheri, Nat. Mus. 1216, c? (type)
-330
»116
219
37
28
630?
75
700
B. hypoceras?, Nat. Mus. 4702
B. hypoceras, Nat. Mus. 4273 (neotype).
B. leidyi, Carnegie Mus. 93, c? (paratype)
B. leidyi, Nat. Mus. 4249 (type)
Percentage of increment, B. leidyi to B.
350
305
300
290
20
133
-123
120
114
12
212
189
190
186
22
715
34
30
665
665
26
440
66
615
625
27
335
127
213
33
31
565
40
....
554 TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Measurements, in millimeters, of sTculls and jaws associated with and referred to members of the Brontotherium phylum
Skull and upper teeth
Jaw and lower teeth
Anterior
canine to
hypoconu-
lid of m3
pi-ms
M'-ms
Premaxil-
lary to
glenoid
Posterior
canine to
hypoconu-
lid of m3
Pi-ms
Mi-ms
Symphy-
sis to
condyle
Depth
below
ma
B. platyceras. Am. Mus. 1448, cT
362
»372
337
350
350
365
221
216
240
232
»483
B. ramosum, Am. Mus. 1447 (type).
B. curtum, Yale Mus. 12013, cf (type).
B. medium, Nat. Mus. 4256, c?.. . .. ._
450
400
393
365
285
272
730
632
?
B. medium, Am. Mus. 1061
181
R gigas, Am. Mus. 492, r?
380
353
241
595
B. (Titanops) elatum, Yale Mus. 12061
260
263
233
219
220
222
213
195
231
B. hatcheri, Am. Mus. 1070, cT
360
360
375
358
"330
332
338
335
284
"352
600
633
630
538
522
565
"465
690
179
B. hatcheri, Nat. Mus. 4262
125
B. hatcheri, Univ. Wyo., c?
320
198
R. hati^hfiri, Am. Mus. 1068, (f
345
330
166
R. leirlyi, Am. Miis. 516
133
300
190
"275
353
100
M. coloradensis, Nat. Mus. . .
351
'313
!'192
517
116
Observations on the measurements of the Brontotherium
series.- — This phylum is as consecutive and as distinct
from other phyla as that of Menodus. The separation
of certain "ascending mutations" as "species" is very
arbitrary. Yet B. leidyi is very far removed from
B. platyceras.
The generic contrasts in measurements which have
been given above indicate that in comparison with
members of the menodontine group {Brontops, Allops,
Menodus) the male individuals of the typical spe
cies of Brontotherium from the upper beds are distin-
guished by very short tapering nasals, extremely long
flattened horns, long skull top, widely expanded zygo-
mata, very short wide premolars, and massive broad
molars.
The progressive allometric evolution from Bronto-
therium leidyi to B. curtum and B. platyceras may be
epitomized as shown in the accompanying table:
Allometric evolution from Brontotherium leidyi to B. curtum and B. platyceras
[Measurements in millimeters]
Pi-m3
Pi-p<
M'-mS
Pi, ap. by
tr.
M', ap. by
tr.
Pmx to
condyles
Zygomatic
index
Nasal
length
Horn
length
B. platyceras:
Field Mus. 12161, S
340
120
223
880
80
38
390
Am. Mus. 1448
42X67
84X93
B. curtum:
Yale Mus. 12013 (type) .
350
"228
790
78?
52
380
45X63
72X84
Nat. Mus. 4946
348
365
128
138
224
235
'■840
825
°810
830
770
710
74
87
65
45
90
87
70
90
105
74
114
365
B. medium, Nat. Mus. 4256, <? (type)
320
B. gigas (elatum), Yale Mus. 12061. . .
350
B. gigas:
Am. MiiR. 492, f7
353
343
-310
126
132
118
241
218
192
47X72
45X57
91X99
365
Nat. Mus. 4262
275
B. hatcheri:
Field Mus. P 5926 . .. _
200
Nat. Mus. 1216 (type) ... ..
250
B. hypooeras, Nat. Mus. 4273, 9
B. leidyi, Carnegie Mus. 93, <? .. .. .
305
300
"123
120
189
190
°165
665
66
104
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERBS
555
As thus arranged the skulls form a series in which
the horn length and the zygomatic width increase,
whUe the free portion of the nasals becomes extremely
abbreviated. But the length of the whole grinding
series in the later species does not increase pari passu
with the size of the skull and with the length of the
horns. From B. leidyi to B. medium it increases
rapidly from 300 to 365 millimeters, but after that,
in B. curium and B. platyceras, the measurement
p'-m^ falls to or below 350. Meanwhile the true
molars also, which had increased from 190 to 241
millimeters, fall to 223. If this apparent falling off in
the increase of the grinding series as a whole shall be
confirmed by much more extensive material it may
indicate that the excessive increase in the horns was
detrimental to further increase in the size of the grind-
ing series (Osborn). The lengthening of the premolar
series is arrested by the shortening of the face, but the
widening of the premolars as well as of the molars
affords a compensatory increase in grinding area. The
length of the true molars as compared with the basal
length of the skull changed very little, as shown below:
Molar index in species of Brontoiherium
B. platyceras:
Field Mus. 12161 (basilar length extreme) 25
Am. Mus. 1448 (basilar length shortened by crushing). 30
B. ramosum, Am. Mus. 1447 (type) 29
B. curtum:
Yale Mus. 12013 (type) 29
Nat. Mus. 4946 26
B. medium, Nat. Mus. 4256 (type) 28
B. gigas:
Am. Mus. 492 29
Nat. Mus. 4244 29
B. "hypoceras," Nat. Mus. 4702 29
B. leidyi:
Carnegie Mus. 93 28
Nat. Mus. 4249 28
The table of measurements above needs extension
from additional material and revision with reference
to the length of nasals and of horns.
SYSTEMATIC DESCRIPTIONS OF GENERA AND SPECIES IN
THE BRONTOTHERIUM PHYLUM
Brontotherium Marsh, 1873
{Titano-ps Marsh, 1887;
Plates XVIII-XXII,
CXXIV, CXXXII,
CCXXXV; text figures
174, 177, 178, 182, 191,
375, 377, 378, 382, 383,
481, 515, 519, 620-640,
690, 707-710, 719-727,
[For original description and ty
'Brontotherium Marsh," Osborn, 1902)
XXXIX-XLII, XLVII, LXXXII,
CLVII, CLXI-CXCIV, CCXXX-
■ 10, 18, 21, 24, 25, 27, 29, 33, 87, 165,
193, 194, 198, 199, 202, 212, 229, 372,
388, 390, 392-395, 398-405, 407, 457-
643, 648, 649, 652, 661, 662, 668, 688,
740, 744, 746
pe reterences see p. 209. For skeletal characters
Generic cTiaracters. — Incisors in males persistent,
with large posterior cingula, (?) variable in females;
canines large, obtuse; grinding teeth with retrogressive
cingula; premolars with progressive tetartocones.
Skull mesaticephalic to brachy cephalic (zygomata).
Horns diverging laterally, typically with connecting
crest, shiftuig forward, oval to flattened in transverse
section.
General characters. — Characters 3-9, 14, 15, 16,
18, 19, 21, 22, 23, 24, 28, 29, below, are correlated
with brachycephaly. (1) 1 14- cf, incisors c? large,
flat-crowned, with stout posterior cingulum; 9 ?one
or both pair absent. (2) Canines d' short, robust,
obtuse, with stout posterior cingulum; 9 slender,
small, with posterior face fiat. (3) Grinding series
arched (curvilinear). (4) Upward flexure of pre-
molar series as seen in side view decided. (5) Length
of premolar-molar series 290-365 millimeters, dental
index 42-46. (6) Premolar series short, 114-140
millimeters. (7) Internal cusps of grinding teeth
low, robust, well rounded, ectolophs sharply depressed
to the crowns of the teeth. (8) P*, m^~^ very broad,
anteroposterior diameter of m^ and m' less than trans-
verse diameter. (9) No cingula between grinders,
premolar cingula sessile or absent. (10) P|^, p^
perhaps more frequently absent than in Menodus
and Brontops, subquadrangular, often with well-
developed tetartocone, outer wall not overlapped
posteriorly by ectoloph of p^. (11) Premolar tetarto-
cones, cf exhibiting early and pronounced develop-
ment, 9 moderate development, tetartocones large
and rounded, set well in toward the center of the
crown. (12) Premolars with internal cingula, c?
blunt, reduced or absent, 9 as in Brontops, external
cingula variable. (13) Molars without internal
cingula, external cingula faint or absent. (14)
Hypocone of m^ prominent, sometimes triradiate.
(15) Basilar length of skull, 665-830 millimeters;
proportions mesaticephalic to hyperbrachycephalic
(index 66-110). (16) Facial portion of skull abbrevi-
ated, with premaxUlaries reduced. (17) Cranial
part of skull elongate. (18) Preorbital malar bridge
narrow, with median ridge prominent and well
rounded. (19) Infraorbital foramen iaconspicuous
in side view. (20) Malar below postorbital process
subflat to round. (21) Free nasals tapering, progres-
sively abbreviated and finally vestigial. (22) Horns
of medium to extreme size, shifting forward progres-
sively, finally extreme, basal section progressively
becoming a transverse oval, summit of horn becoming
broad, flattened anteroposteriorly, oval in section.
(23) Zygomata strongly arched, buccal expansion
progressive, in section finally excessively broad and
flattened (c?). (24) Midparietal convexity pro-
nounced. (25) Occiput cf extremely produced back-
ward behind zygomata, 9 much less produced.
(26) PUlars flaring moderately to extremely, occiput
strongly indented, with median Icnobs. (27) Basi-
sphenoidal rugosity variable, vomerine septum vari-
able. (28) Postglenoid cf very broad. (29) Jaw
robust with shallow more or less concave chin, coro-
noid rather slender, tapering, anterior border rounded
or square in section, strongly recurved or hooked,
pointed at extremity, angle depressed and rugose (d^).
Incisors reduced in females. — There is some evidence
for the absence or reduction of the upper and lower
incisors in females; this evidence rests chiefly on the
556
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
supposed female skull referred to B. curtum (Am.
Mus. 1005). The permanent incisors were protruded
very early and have weak implacements in the crowded
premaxillae. Hence they are sometimes shed in old
animals.
Ancestral brontotJieres in the lower Titanoiherium
zone. — The Eocene ancestry of this subfamily is still
in some doubt; its earliest known members may prove
to be the species Diplacodon elatus Marsh and Eoti-
tanotherium osborni Peterson of Uinta C (upper
Eocene).
Relatively small and short-horned ancestors of the
brontotheres have fortunately been determined by
the writer as occurring according to Hatcher's geologic
record in the very base of the lower Titanotherium zone,
definitely proving that the Brontotherium phylum early
separated from the Megacerops phylum and at a much
earlier datef r om the Brontops- Menodus phylum. These
primitive brontotheres are comparatively rare in the
lower beds. They include two species as follows:
Brontotherium leidyi Osborn is a comparatively
small animal with short horns, which is determined as
a brontothere, however, by two very distinctive
characters — first, the transversely oval section of the
summit of the horn; second, the very progressive
condition of the premolar grinding teeth, in which the
tetartocones are well developed. This animal is of
extraordinary interest by comparison with B. platy-
ceras in showing the extreme transformation in the
horn region which took place in the course of the
deposition of 200 feet of sediment. It is represented
by a number of admirably preserved specimens.
Brontotherium hypoceras (Cope). — This is an animal
first named Symhorodon hypoceras by Cope from the
tip of a horn and some other fragments of the skull.
He was struck by the transversely oval section of
this horn tip. This very imperfect type (PL CLXVI,
fig. 174) was fortunately compared with the skull in
the National Museum and determined by Osborn as
a most interesting transitional stage between B.
leidyi and the species of the higher levels. Like B.
leidyi it has rounded grinding teeth with low cusps,
well-developed tetartocones, and obtuse canines.
There is no question as to its phyletic position.
Brontotheres of the middle Titanotherium zone.—
Brontotherium- hatcheri Osborn, from the middle Titano-
therium zone (Chadron B) includes animals of inter-
mediate size, named in honor of J. B. Hatcher, the
chief explorer of the Titanotherium-he&Ting beds,
which connect B. leidyi and B. hypoceras of the lower
zone with B. gigas and its successors of the upper zone.
Brontotherium tichoceras (Scott and Osborn) possibly
belongs geologically to the upper zone, but both its
geologic and its phyletic position are somewhat un-
certain, although there is no doubt whatever as to its
general affinity to Brontotherium.
Brontotheres of the upper Titanotherium zone. —
Brontotherium gigas Marsh, a giant form, was probably
characteristic of the lower part of the upper zone,
namely, Chadron C 1. The evolution of the special
characters of the phylum, the horns, and the zygo-
matic arches now appears to be accelerated. As shown
in the accompanying table six succeeding stages,
species, subspecies, or mutations can be distinguished,
to which specific names may here be given. While
the evolution of the dominant characters of the horns,
connecting crests, and buccal processes of the zygomata
is progressive, and while the premolars follow the
universal law of increasing complication by the devel-
opment of the tetartocones, the series of grinding teeth
as a whole is partly arrested and in some respects re-
trogressive in development. On account of the
shortening of the face the premolar series is shorter
than in Menodus, but the premolars have the most
advanced tetartocones known in any genus except
Megacerops, and both molars and premolars are ex-
tremely broad, so that in basal view the dentition
appears enormous. It is true that in B. platyceras
(Field Mus. 12161) the premolar-molar series is
slightly shorter than in B. gigas elatum (Am. Mus.
492). The incisors, on the contrary, in the males at
least, are remarkably persistent, and one of the
readiest means of distinguishing Brontotherium is by
its reduced cingula on the superior incisors, which are
quite different from the smooth, rounded incisors of
Megacerops.
Stratigraphic horizons of Brontotheres
"a
O
a
o
X)
6
o
a
(O
O
M
o
1
C
3. Upper
2. Middle
1. Lower
Brontotherium platyceras
B. ramosum
B. curtum
B. medium
B. dolichoceras
B. gigas elatum
B
3. Upper
2. Middle
1. Lower
B. gigas
?B. tichoceras
B. hatcheri
A
3. Upper
2. Middle
1. Lower
B. hypoceras
B. leidyi
ilocene
I Basin
Horizon C 1
Diplacodon elatus
^
h3
3 .S
3*5
Horizon B
Eotitanotherium osborni
EVOLUTION OP THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
557
Geologic succession of lower Oligocene species of Brontotherium
C. or LARGE SIZE
Upper Tilanotherium zone:
Sixth stage, B. platyceras: Grinding teeth 337-340 milli-
meters; nasals vestigial; horns extremely broad and flat.
Fifth stage, B. ramosum: Grinding teeth 350 millimeters;
nasals vestigial; horns planoconvex at base, expanding
at summits.
Fourthstage, B. curium: Grindingteeth 346-350 millimeters:
nasals shorter; horns slightly convex to plane posteriorly.
Third stage, B. medium: Grinding teeth 365 millimeters;
nasals shorter; horns with prolonged malar ridge; con-
cave antero-external faces.
Second stage, B. dolichoceras: Nasals shorter; horns with
incipient malar ridge.
First stage, B. gigas: Animals of maximum size; grinders
330-353 millimeters with cingula; horns longer, oval, re-
curved, 275-360 millimeters in height; nasals narrow,
more reduced.
B. OP INTERMEDIATE SIZE
Middle Tilanotherium zone, upper levels:
B. tichoceras (level unrecorded) : Grinding teeth 330
millimeters; horns obliquely oval; phyletic position
somewhat doubtful.
B. hatcheri: Grinding teeth 310-320 millimeters with
cingula more or less developed; nasals longer than in
B. gigas; horns 250 millimeters in height.
A. OF SMALL SIZE
Level doubtful, either lower or middle Tilanotherium zone;
B. hypoceras: Grinding teeth as in B. leidyi (305 mm.);
nasals intermediate in length; horns longer (140 mm.),
more anterior in position.
Upper levels of lower Tilanotherium zone:
B. leidyi: Grinding teeth 290-300 millimeters, typically
smooth and rounded, cingulum faint or wanting; nasals
elongate: horns rudimentary, placed above orbits.
The Brontotherium phylum as represented in the Hatcher collection of 44 skulls and jaws from the Ghadron
formation in the United States National Museum
B. platyceras (Scott and Osborn) .
Do
B. ramosum (Osborn)
B. medium (Osborn)
B.? medium (Osborn)
B. curtum Marsh
Do
Do
Do-...
Do
Do
B.? curtum Marsh
B. gigas? Marsh
B. gigas Marsh
B.? gigas Marsh
Do
B. gigas Marsh
B.? gigas Marsh
Do
Do
Do
Do
B. gigas? medium?
B.? gigas Marsh
B. gigas (hatcheri) Marsh
B. sp .
B. medium (?tichoceras)
B.? tichoceras (Scott and Osborn)
B. sp "-.
?B. sp
B. sp
B. hatcheri Osborn
B. medium? (hatcheri)
?B. hatcheri Osborn
B. hatcheri Osborn
B. (hatcheri) hypoceras (Cope)-..
B. (hatcheri) hypoceras
B. hypoceras (Cope)
Do
Do
Do
B. leidyi Osborn
Do
B. (leidyi?) (hypoceras?)
Catalog No.
8730
8729, d'
1243, 9?
4256, c?
8772
8726
4946, cf
1232, c?
1211, c?
8727, cf
8728, 9?
8743, cf
4244, 9?
8752
8755
8758, 9
8762, 9
4903
8773, cf?
8774
8783
8787, cf ?
8791, c?
8800, 9
4262, c?
8805
4699, cT?
8313, c?
1227, 9
8734
8775, cf
1216, cf
4716, 9?
4255
4704, 9
8780, cf
8789, &1
4273, c?
4702, 9?
8763, 9?
8757
4249, 9?
4250
8795
Material
Skull.
Nasals and horns.
Horns.
SkuU. Type.
Pair of jaws.
Fine skuU and horns.
Very fine skull. Right horn broken and healed during life.
Left upper dental series.
Skull.
Anterior half of skull.
Skull and jaws.
SkuU, nasals, horns (partly destroyed).
Skull. Horns imperfect. Agrees with B. gigas in dental measurement.
Upper dentition. Premolars and molars smaller than in large B. gigas.
Upper dentition.
SkuU.
Skull and parts of lower jaws. Horns small and feeble.
Parts of both rami.
Pair of lower jaws lacking posterior part of left. Aged individual.
Pair of lower jaws (young).
Left ramus.
Pair of lower jaws.
Pair of lower jaws.
Left ramus.
SkuU, jaws, and part of skeleton.
Lower jaws and anterior dorsal (young).
SkuU.
Skull and jaw (good).
Lower jaw (immature).
Indeterminate skull lacking nasals and horns. Poor.
Pair of jaws (small).
SkuU. Type.
SkuU. No canine, horns, or nasals. Genetic reference doubtful; B. hatcheri
in measurement.
Skull with front part of jaw.
Skull. Small female; horns with connecting crest.
Pair of lower jaws.
Right ramus.
Skull. Neotype. Important male skull.
SkuU.
SkuU.
Pair of jaws; horn cores; nasals.
SkuU. Type.
Part of jaw.
Pair of jaws.
558
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Brontotherium leldyi Osborn, 1902
Plates XIX-XXII, XLVII, CLXI-CLXV; text figures 27, 29,
33, 87, 198, 199, 372, 383, 390, 392, 395, 399, 401-403, 405,
457, 458, 461, 462, 464, 519, 620, 621, 624, 626-629, 635-637,
661, 668, 690, 709, 720, 727
[For original description and type references see p, 234. For skeletal characters
see pp. 691, 697]
Geologic Tiorizon. — Lower Titanotherium zone of
South Dakota, middle level (Chadron B).
Specific characters. — P'-m' 290-300 millimeters,
m'-m^ 114-120, dental index 45. Basilar length of
therium zone, middle level (B 1), and by a beautifully
preserved skull and lower jaws (No. 93) in the Car-
negie Museum at Pittsburgh, associated with a
humerus, radius, ulna, tibia, and atlas.
General description. — This species is positively re-
corded from the lower Titanotherium zone and is of
great interest because Brontotherium leidyi, while very
primitive and approaching the primitive members of
the other phyla in certain respects, is yet readily dis-
tinguished from its contemporaries — first, from Bron-
WMM0
D
A B
Figure 461. — Sections and contours of skulls of Brontotherium leidyi and B. hypoceras
Brontotherium leidyi, Nat. Mus. 4249 (type); horns small and placed not far in front of the orbits, their basal section ovoid with the narrow end external, the
section of the upper part of the horns roundly oval, connecting crest slight, nasals long and lender, zygomata slightly expanded. B, B. leidyi, Carnegie Mus. 93,
a referred specimen with horns longer, basal section obliquely oval, zygomata little expanded, nasals long. C, B. hypoceras, Nat. Mus. 4273 (neotype), horns
much longer than in B. leidyi and placed far in front of the orbits, their basal section oval, nasals short. Thus B. hypoceras is a much more advanced stage
than B. leidyi. D, B. hypoceras, Nat. Mus. 4702. One-seventh natural size.
skull 665 mUlimeters, zygomatic breadth 440, index 66
(dolichocephalic or mesaticephalic). Nasals elongate
(free length 118 mm.), broad proximally (105 mm.),
tapering distaUy. Horns low (height above narial
sinus 107 mm.), placed above preorbital foramen,
recurved, with gently sloping anterior contour, sum-
naital section a transverse oval. Zygomata with
buccal swelling slight.
Materials. — The species is represented by the type
skull (Nat. Mus. 4249, skull R) from the lower Titano-
tops hrachycephalus by the greater length and size of
the skull as a whole; second, from the type of Allops
walcotti (Nat. Mus. 4260) by the more pronounced
development of the tetartocones of the premolars,
especially of p4, by the reduction or absence of internal
and external cingula on the premolars and molars, by
the rounded, obtuse contours of the premolars, and
especially by the transverse oval section of the horns
from base to summit. In all these characters it antici-
pates the well-marked B. gigas and B. hypoceras types.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
559
Other features distinguishing B. leidyi from A.
walcoUi are no median vertical ridge on occiput; no
lateral projections from midtemporal crests; great
breadth opposite supraorbital processes; horns slightly
more anterior to orbit; no lacrimal knob; broad con-
tact of postglenoid and post-tympanic processes.
Oiservations on the measurements of Brontotherium
leidyi. — The type and hypotype specimens agree fairly
well in measurements, and this primitive species is very
clearly defined. The zygomatic index is extremely
low, owing to the nonexpansion of the zygomata; as
compared with B. platyceras the nasals are about three
times as long, while the elevation of the horns above the
nasals is only about one-fourth as great.
The type skull (Nat. Mus. 4249) is in the beginning
of the ninth growth stage. The zygomata, although
much crushed, give evidence of a
sudden flattening and buccal expan-
sion posteriorly, similar to that which
we observe on a larger scale in B.
gigas. The occiput shows narrow
lateral pillars; it was deeply cleft
above. The nasals are remarkably
long, narrowing anteriorly, and de-
curved. The bridge over the infra-
orbital foramen is rather broad. The
infraorbital foramen is partly exposed
on the side of the face. The horns
rise a very short distance above the
vertex and, unlike those in A. walcotii,
are markedly recurved and broadly
oval at the top. Judging by the
alveoli, there is a pair of large lateral
incisors with a vestigial alveolus for a
median incisor on the right side. The
dental formula is thus I^^, P^. The
premolars have rounded inner and
outer surfaces and vestigial or even
no trace of cingula and in this respect
resemble those in certain specimens
of upper Titanotherium zone species
of this series. The tetartocones are
strongly developed in p^, p^, p*, occu-
pying about two-fifths of the inner
surface of the crown in p^, p^ and
being large and prominent even on p*.
The medifossettes are much deeper
than in Brontops hrachycephalus. The
hypocone is cingulate on m^ and
presents a very strong crest or aborted metaloph
extending in toward the metacone; m^ shows the
so-called antecrochet and crochet.
The beautifully preserved specimen in the Carnegie
Museum (No. 93, Pis. CLXI-CLXV, fig. 462) differs
from the National Museum type in the following
points: (1) The median incisor is persistent but
reduced; (2) the premolars have sessile, crenulate
internal cingula; (3) the tetartocones are set even
farther in toward the middle of the crown; (4) the
preorbital malar bridge is narrower and rounder. It
is thus in every respect a characteristic Brontotherium.
It shows very large, flat-crowned, functional incisors;
exceptionally heavy, blunt, posteriorly cingulate
canines in the upper and lower jaws; p^ preserved
but small, tetartocones prominent, well rounded;
cupped and crenulate hypoconulid on ms, occipital
Figure 462. — Lower jaws of Brontotherium leidyi
A, Carnegie Mus. 93; ramus crushed vertically but teeth well preserved, chin shallow, angle not produced, canine
very stout, no diastema in front of pi, grinding teeth without external cingula. B, Am. Mus. 516; provisioually
referred to B. leidyi on account of the near agreement in measurements. Incisors stout, canine short and
swollen, a small diastema in front of pi, cheek teeth almost lacking external cingula, premolar series curved
upward anteriorly. One-fifth natural size.
knobs (PL CLXV), inconspicuous infraorbital fora-
men and rounded malar bridge, nasals thin except
along median inferior ridge, chin sloping, coronoid
hooked. Taken in connection with the type it proves
that Brontotherium must have acquired its generic
characters very early, especially since both these
specimens are from the lower zone.
Jaw of B. leidyi. — The jaw in this stage of evolution
is represented by a specimen (Am. Mus. 516) which
560
TITAJSrOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
was formerly (Osborn, 1896.110, p. 181, fig. 5) errone-
ously described and figured as pertaining to Menodus
trigonoceras. It exhibits the convex lower border, the
small mental foramen below the fang of ps, the absence
of external and internal cingula upon the inferior
molar-premolar series, the reduced first lower pre-
Brontotherium? rumelicum (Toula)
(Menodus? rumelicus Toula, 1892)
Text figures 193, 463, 464
[For original description and type references see p. 230)
Type locality and geologic horizon. — Lower Oligo-
cene(?) of eastern Rumelia, Balkan Peninsula, Europe.
Figure 463. — Two lower molars and symphyseal region of Brontoiherium? rumelicum
Type, paratype, and referred specimen of "Menodusf rumelicus" Toula; geologic level, "Belvidereschotter"; locality, Eajali, northwest of Burgas,
eastern Rumelia (fide Toula). A, Type, third right lower molar, crown view. Bi, B2, Paratype, second right lower molar (probably of another
animal), external and internal views. Ci, C2, Referred specimen, symphyseal region (probably of another animal), external and inferior views.
One-half natural size. The type offers no very distinctive generic characters, but the referred symphyseal region agrees nearly with BroTitotheTium
in general contour and in the proportions and position of the premolars. These specimens, together with the fragment named Titanotherium
bohemkum and the type lower jaw of BracJiydiasiematherium transilvankum, are the sole known European representatives of the titanotheres.
molar, the short, obtuse canines, with a postero-
internal cingulum, the well-developed pair of lower
incisors, the crenulate crest of the hypoconulid on
ma, all characteristic of Brontotherium. The chin
is a little more decided than in B. gigas; the ui-
cisors are not evenly rounded but have a posterior
cingulum.
Generic and specific characters. — -Symphyseal region
and premolars of paratype approaching Brontotherium
leidyi in characters and measurements. Premolars
with reduced external cingula. Symphysis massive,
flattened below.
The second (paratype) and third lower molars seem
to offer no decisive specific characters.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
561
Comparative measurements of Brontotherium? rumelicum, in
Ms, ap. by tr_
M2, ap. by tr.
Pi-p,
'94X43
■ 75X42
= 123
B. leidyi,
Nat. Mus.
4249 (type)
96X41
68X40
119
102X48
67X48
113
Menodus
trigono-
ceras,
Am. Mus.
1067
108X45
74X43
123
B. robus-
tus, Yale
Mus.
12048
(type)
110X47
77X46
133
"Type. 'Paratype. ■Hererred.
These measurements offer no very decisive evidence,
but a direct comparison of the cast of the referred
lower jaw reveals a strong resemblance to the jaws of
Brontotherium.
Comparative measurements of third inferior molar of
Brontotherium? rumelicum, in millimeters
Materials. — Referring to the technical description of
Toula (Toula, 1892.1, p. 612) we note that the speci-
mens consist, first, of the type, a portion of a ramus
containing ms (fig. 463), and secondly of the paratype,
an isolated lower molar. This lower molar is cer-
tainly a second molar (m2) rather than a first molar
(mi) as identified by Toula — first, because this tooth is
less worn than mj; second, because it is larger than mi.
Both of these specimens were found at Kajali, north
of Burgas, eastern E,umelia, near the Black Sea, north-
B. ru-
melicum
(type)
B. leidyi,
(Carne-
gie Mus.
03)
Men
trigon
Am.
Mus.
1007
odus
oceras
Am.
Mus.
1067
Total anteroposterior
Transverse, anterior lobe.
Transverse, posterior
94
43
39
27
96
41
108
48
46
27
108
45
Transverse, third lobe
Of the highest importance and interest is
the question, How many branches of the ti-
tanotheres invaded Europe and Asia? There
is no doubt that this family was chiefly North
American in evolution, but the presence near
the Black Sea of the animal described by Toula
as Menodus rumelicus, of the previously de-
scribed BracJiydiastematherium, and of the
newly described Menodus ioJiemicus affords
positive proof that at least three branches of
the great family of titanotheres actually in-
vaded Asia and eastern Europe. The titano-
theres discovered in Mongolia by the Asiatic
expedition of 1922-23 are described in the ap-
pendix (pp. 913, 942).
Geologic level. — It is a remarkable fact that the
geologic records in the present case as well as in that
of Brachydiastematherium conflict with those in
America. The BracJiydiastematherium is said to have
been found in beds of lower Eocene or Ypresian
[lower] age. The present type of M. rumelicus, on the
contrary, is recorded as of extreme upper Miocene age,
or as equivalent to Pikermi. Possibly M. rumelicus
represents a survivor of the titanotheres in western
Asia and eastern Europe after the period of their
extinction in North America; but this is rendered
improbable by the fact that the present type is in a
lower Oligocene stage of evolution.
FiGUBB 464. — Lower jaws of Brontotherium? rumelicum and B. leidxji
A, Symphyseal region referred to B. rumelkum: B, lower jaw of £. leidyi. One-sixth natural size.
west of Constantinople. Four years later Toula
(1896.1, p. 922) described a portion of a mandibular
symphysis which was found in the same locality
(figs. 463, 464).
Affinities. — The resemblances of these specimens to
the Ancylopoda {Chalicotherium and Macrotherium)
were pointed out by Toula, who, however, finally con-
cluded by referring all these specimens to Menodus.
We note the following marked differences from the
Chalicotheriidae: (1) Presence of a strong third lobe
on ms, which is absent in all chalicotheres; (2) a mas-
sive symphysis and chin, which is reduced and degen-
erate in all chalicotheres; (3) presence, so far as we can
562
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
judge by alveoli, of large canine teeth, which are re-
duced in all chalicotheres; (4) absence of the meta-
stylid, in contrast with the presence of the metastylid or
reduplication of the metaconid in all chalicotheres.
Thus in spite of the authority of Von Zittel, who
also suggested reference of the types to the Chali-
cotheriidae, the anatomical evidence is absolutely
against the affinity to the Ancylopoda. On the other
hand, comparison of casts of three specimens kindly
presented to the American Museum by the late
Professor Toula shows a close resemblance to the large
Oligocene titanotheres in the structure both of the
teeth and of the jaw. The chief or distinctive char-
acters are as follows:
Symphysis massive, canines probably large, four
lower premolar teeth, lower molars without internal
or external cingula, hypoconulid ms without strong
internal crest.
Apparently this type is very similar to Brontotherium
in its measurements and in the structure of the jaw so
far as preserved but especially in the structure of the
grinding teeth. The entire absence of internal and
external cingula on the grinding teeth forbids reference
to Menodus and relates the animal certainly to
Brontotherium, in which the cingula are reduced. The
massive symphysis of the lower jaw and the reduced
crest of the hypoconulid on ma are, it is true, features
which suggest a member of the Menodontinae; but on
the whole the prevailing characters relate this animal
certainly to a branch of the Brontotherium phylum.
Brontotherium hypoceras (Cope)
{Symborodon hypoceras Cone, 1874; " Brontotherium hypoceras "
Osborn, 1902)
Plates CLXVI, CLXX, CLXXXIII; text figures 174, 399, 457,
458, 461
[For original description and type references see p. 216]
Geologic horizon. — Middle Titanotherium zone
(Chadron B) ; possibly also in lower beds of the upper
zone (Chadron C).
Specific characters. — Skull larger than B. leidyi
(premaxillaries to condyles 715 mm., estimated), free
nasals shorter (74 mm.). Horns placed anteriorly,
oval from base to summit, outside length 140 milli-
meters; buccal processes prominent. M'-m' 305 milli-
meters (Nat. Mus. 4273) — that is, somewhat longer
than in B. leidyi. A single superior incisor in adults (?).
The specific name apparently refers to the erroneous
opinion of Cope that there was a second pair of horns,
consisting of low protuberances, behind and on a lower
level than the first pair.
Materials. — The materials consist of the type and
of two imperfect skull specimens in the National Muse-
um. Cope's imperfect type of this species (Am. Mus.
6361) consists principally of a horn tip, the bridge
over the infraorbital foramen, and parts of a zygoma,
which fortunately offer characters of diagnostic value.
The type horn apparently belonged to a young adult or
female; it was taken to Washington and closely
compared with two relatively complete specimens in
the National Museum, with the gratifying result that
specific identity was established. These two National
Museum skulls are from the upper levels of the lower
Titanotherium zone, or the lower and middle levels of
the middle zone, and are distinguished by short,
transversely oval horns, similar in section to those of
B. gigas, but about only one-third as long and as
massive.
The first skull (Nat. Mus. 4702, skull k') was recorded
from the lower beds of upper level C, but Mr. Hatcher
regards the record as doubtful. This skull is a B.
hypoceras in its horn and nasal development, but it has
the dental size and development of B. gigas (see table
above). It is therefore somewhat problematic. Un-
fortunately it lacks the canines, so we can not deter-
mine the sex positively. The very marked develop-
ment of the tetartocones on the premolars indicates an
affinity to B. gigas; it also proves that this is a suc-
cessor to B. leidyi or of more recent geologic age. In
p* the tetartocone is fully as large as the deuterocone
and entirely separate; the crowns of the premolars are
exceptionally smooth, rounded, devoid of cingulum,
with obtuse cusps. The molars are also devoid of
cingulum. In m^ the hypocone is a prominent
cingule. The nasals are thin and resemble those in
B. gigas. The horns are very small and have the
characteristic oval B. gigas section from base to sum-
mit; they are placed remarkably far forward. In
superior view the skull resembles that of B. leidyi on a
larger scale. The top of the cranium narrows grad-
ually, with the occipital pillars apparently slender
(although this part is largely restored), the occiput
being deeply indented; the zygomata have flattened
posterior expansions which recall in contour those of
B. gigas, although less robust.
The second, the neotype skull (Nat. Mus. 4273,
skull 1) is correctly recorded from the middle Ti-
tanotherium zone, level Chadron B 2. The skull (con-
sisting of only the anterior portion) is small, although
it belongs to a very old animal in the tenth stage of
growth. It corresponds to the preceding skull, but
the horns are longer and placed equally far forward;
as shown in the section (fig. 461, C) there is a low
connecting ridge.
Dentition. — In the neotype skull (Nat. Mus. 4273)
the dental measurements are rather close to those of
B. leidyi, save that the premolar series is longer, but
the "free nasals" are much shorter and the horns
longer. Extremely valuable evidence is afforded
by the short, obtuse canines with strong posterior
cingula, which resemble those of B. gigas on a small
scale. In B. leidyi a reduced upper median incisor
was observed; but in this specimen no trace of the
upper median incisor is found, there being only one
pair of well-developed lateral incisors. The loss of
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
563
i ' may be an age character. The crowns are absolutely
smooth, oval, and without cingulum. This incisor
reduction may also indicate female sex; if not, it
would appear to place B. Tiypoceras out of direct
ancestry to the B. gigas-platyceras line. Since all
specimens of B. gigas show two well-developed incisors
the presence of but a single incisor appears to be a
distinctive and specific character in B. Jiypoceras.
The characters of the premolars and molars and es-
pecially the development of the tetartocones are
about the same as in the first skull, except that there
is a sessile crenulate cingulum on the inner sides of
P2, P4, and the tetartocone on p4 is somewhat smaller.
Brontotherium hatcheri Osborn, 1908
Plates XIX, XXXIX-XLII, CXXIV, CLXVII-CLXX,
CLXXII; text figures 202, 395, 399, 400, 404, 457, 458, 465-
468, 470, 630, 632, 639
[For original description and type references see p. 235. For skeletal characters
see p. 695]
Geologic Jiorizon. — Middle Titanotherium zone of
South Dakota.
Specific characters. — If, P|. Nasals moderately
long (97 mm.), thin at the edges. Horns, 250 +
millimeters, two-thirds the length of the B. gigas
horns. Skull length (premaxiUaries to condyles)
710 millimeters (estimated), width (across zygomata)
530 (estimated).
This species was named in honor of J. B. Hatcher,
paleontologist and explorer.
Materials. — The species is represented by four speci-
mens, namely, the type skull (Nat. Mus. 1216, skull
a) and the referred material, including in the National
Museum a second skuU (No. 4255, skull Q), in the
American Museum the anterior portion of a skull,
No. 1070, associated with a lower jaw, and in the
Field Museum, Chicago, a complete uncrushed skull
(No. P 5926).
The type is the large skull Nat. Mus. 1216, com-
plete except the premaxiUaries, recorded by Hatcher
from the uppermost levels of the middle Titano-
therium zone. Although a fully adult animal, in the
ninth stage of growth, as compared with B. gigas
it has shorter horns (250 mm.) of rounded section, a
relatively lower connecting crest, and longer nasals
(97 mm.). It appears to represent an early phase
of evolution of B. gigas, the type brontothere. The
horns are very round or convex in section and have a
well-defined external ridge on the lower outer portion ;
the connecting crest is relatively shallow, and the
nasals are thin; the zygomata are broad and flat.
The premolars are well advanced, the tetartocone of
p* being quite distinct.
A skull in the National Museum (No. 8313) de-
scribed below under Brontotherium tichoceras should
also be compared with the type of this species but
differs in the absence of a connecting crest.
A second skull of B. hatcheri (Nat. Mus. 4255),
equally primitive, is also recorded by Hatcher from
the uppermost level (Chadron C) of the middle Titano-
therium zone. This animal, in the seventh stage of
growth, exhibits intermediate horns (280 mm.) with a
still lower connecting crest than in the type of B.
hatcheri; the basal horn section has a more prominent
nasal angle than in the type. The nasals and horns
both in form and in section might be described as
intermediate between Brontops rohustus and Bronto-
therium gigas.
In the American Museum is a skull (No. 1070) which
may represent either an earlier stage of evolution, such
as B. hatcheri, or a variety of B. gigas. This skull is that
of a bull in the seventh stage of growth and affords
valuable supplementary knowledge of the dental and
mandibular structure of an advanced B. hatcheri
(Pl.CLXIX). It proves that the formula is If, Pf. P'
may drop out in old ani-
mals, just as in B. gigas.
The inferior lateral inci-
sors are robust teeth with
a crenulate posterior
cingulum; the median
pair are smaller.
Degeneration of the
cingulum. — The premo-
lars of this specimen
(Am. Mus. 1070) have
an especial interest as
showing several stages
in the decline of the
internal cingulum,
which, as we have ob-
served, is very feebly
developed in B. leidyi
and B. hypoceras. On
p^, right, the cingulum
is a blunt ridge slightly
crenulate; on p^, right,
the cingulum is in-
complete, the crenula-
tions being much more
prominent and more separate; on p', left, the
line of the cingulum is very faint, the crenula-
tions are very prominent, separate, and appar-
ently in process of being absorbed into the bases
of the internal cones. The occurrence of the cin-
gulum in this imdoubted bull coincides with other
evidence against the hypothesis that the cingulum is a
sexual character.
A peculiar feature of the enamel of the lower and
upper teeth is the fine horizontal striation of structure,
which is also observed in succeeding members of this
series.
In this animal the horns are rounder and shorter
than in the type of B. gigas, but the connecting crest
FiGTJHE 465. — Sections and
contours of skull of Bron-
totherium hatcheri
Nat. Mus. 1216 <type). One-seventh
natural size.
564
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
is much deeper than in the type of B. Jiatcheri. This
specimen of B. Jiatcheri, moreover, tends to bridge over
Figure 466. — Skull of Brontotherium hatcheri
followed by a depression; (8) in the gentle slope of the
occipital vertex in side view; (9) in the characters of
the cheek teeth, which have very
large circular tetartocones set well in
toward the middle of the crown,
vestigial external cingula, and low,
rounded internal cingula and
rounded external faces; (10) in the
minute p'. In general this skull also
supports the specific separation of B.
hatcheri as a lower stage of B. gigas.
Comparative measurements are given
above.
University of Wyoming skull of B.
hatcheri. — A skull (figs. 466, 467) and
associated jaw (fig. 468, A) which are
provisionally referred to this species
are in the University of Wyoming.
They were discovered by Mr. W. H.
Reed in the upper Titanotherium zone
in the northeastern corner of Carbon
County, Wyo. The absence of a con-
necting crest, the subcylindrical form
of the horns, and the form of the nasals
at first suggest the type of Megacerops
coloradensis ; but the skull differs in
many points from that of Megacerops
acer, and the measurements are on the
whole closer to those of B. hatcheri.
The fourth upper premolar measures
41 by 52 millimeters (ap. by tr.); the
lower jaw from front of canine to back
Univ. Wyoming Mus. 1; Chadron C (flde W. H. Eeed); northeast corner of Carbon County.Wyo. Side of an^lc 598' Other measurements are
¥iew. One-sixth natural size. The horns and measurements are close to those of S. hatcheri, the basal • ^ u '
section of the horns recalls B. tichoceras, and the associated lower jaw is characteristic of Bronlotherium. glVen aDOVe.
the gap between B. leidyi and B. gigas, because while
approaching the latter in the connecting crest between
the horns and in its larger size, it resembles the older
form of B. leidyi in the following respects: (1) The
canines, incisors, and grinding teeth are similar, (2)
the horns are intermediate in position between those
in B. leidyi and B. gigas, (3) the top view of the skull
is similar to B. leidyi.
Field Museum sTcull of B. hatcheri. — In the Field
Museum, Chicago, there is a beautiful and nearly
uncrushed skull of this species from Phinney Springs,
S. Dak., recorded from the middle zone, at a level 75
to 100 feet above the Pierre shale. The skull is a
typical Brontotherium in every respect and contrasts
sharply with the Brontops-Allops- Menodus group
while agreeing with Megacerops and Brontotherium,
especially (1) in the shape of the nasals, which are
gently tapering and distally decurved in both top and
side views; (2) in the anterior narial opening as seen in
side view; (3) in the upward flexm-e anteriorly of the
tooth rows; (4) in the very narrow, rounded bridge
over the infraorbital foramen; (5) in the relatively
small orbit; (6) in the shape of the connecting crest
and horns; (7) in the midparietal convexity, which is
Figure 467. — Skull of Brontotherium hatcheri
Univ. Wyoming Mus. 1. Front view. One-sixth natural size. This view
shows well the resemblances to B. hatcheri. (Compare Pis. CLXVII,
CLXIX, A2, CLXX, B.) The specimen is much crushed and distorted.
Additional observations on B. hatcheri. — This form is
not very clearly separated from B. gigas, since it was
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
565
founded on a skull (Nat. Mus. 1216), while B. gigas
rests upon a lower jaw. The type skull has the nasals
longer and the horns shorter and less flattened than
in the type of Titanops elatus, and the same is true of
the other three skulls which stand nearest to the type
(Univ. Wyo. Mus. 1, Field Mus. P 5926, Am. Mus.
1070). The range in the principal measurements of
these skulls as compared with B. gigas, the succeeding
stage, is as follows:
Range in measurements of Brontotherium gigas elatum and B.
hatcheri (gigas?), in millimeters
pi-m'
Pi-p4
Mi-m3
Pmx to condyles
Zygomatic index
Nasal length
Horn length
348-353
126-133
218-241
770-830
82-87
70-87
275-365
« 310-320
118-132?
192-?
7 10-?
85?-105
200-265?
Skull Am. Mus. 1070 has the horns a little longer
and more slender than in the type of B. hatcheri, but
its nasals are as short as in the typical B. gigas elatum
and its premolars are as long as in that form. Hence
it appears to connect these two stages.
Comparative measurements of jaws of B. hatcheri and B. gigas,
in millimeters
Symphysis to angle
Condyle to symphysis. _.
Depth, condyle to bot-
tom of angle
Depth of jaw below mj-.
Molar-premolar series-..
Premolars
Molars
M3, transverse
Mj, longitudinal
Canine enamel, anterior.
Canine enamel, antero-
posterior
Length of symphysis
B.hatcberi,
Am. Mus.
1070
605
600
300
100
375
■120
263
48
103
35
29
B. gigas,
Yale Mus.
12009 (type)
620
634
305
105
365
117
250
47
118
34
30
173
B. gigas
(hatcheri?),
Nat. Mus.
4262
623
" Titanops
elatus,"
Yale Mus.
12061 (type)
625
285
304
101
115
350
121
235
260
44
°53
100
118
33
28
182
" Estimated.
The lower molar grinding teeth of B. hatcheri
(263 mm.) exceed in linear dimensions those of either
the B. gigas type specimen (250) or those of the
T. elatus type specimen (260); consequently the pro-
gressive characters of B. gigas elatum are to be seen
in the elongation of the horns, the elevation of the
connecting crest, and the abbreviation of the nasals
(figs. 465, 470), rather than in the increasing length
of the grinding series.
The jaws of B. hatcheri, B. gigas, and B. elatum
(figs. 468, 471) are influenced so much by age, sex,
and individual variation that it is diQacult to separate
them specifically.
Brontotherium tichoceras (Scott and Osborn)
{Menodus tichoceras Scott and Osborn, 1887; " Megacerops
tichoceras" Osborn, 1902)
Plate CLXXXVII; text figures 177, 469
[For original description and type references see p. 219]
Geologic horizon. — Titanotherium zone of South
Dakota. Level unrecorded.
Specific and generic characters. — Skull about as large
as in B. gigas. I^, P^. Superior canines and both
pairs of incisors well developed ; grinding teeth without
cingula, premolars with large, distinct tetartocones;
nasals of medium length and thickness, contracting
anteriorly; base of horns of stout, obliquely transverse
section. Buccal processes very prominent, slightly
concave superiorly; postglenoid and post-tympanic
widely conjoined. Grinding series 330 millimeters —
that is, about the same as in smaller skulls of B. gigas.
Materials. — The species is known only from the
type, in the Museum of Comparative Zoology, Har-
vard University, a skull which lacks the horns.
The absence of the summits of the horns and of the
occiput and the unfortunate loss of the cutting teeth
(which has occurred since the original description of
this skull was published) prevent us from placing
this species phyletically wit i certainty. It assuredly
does not belong to Menodus but is rather a member of
the brontotheriine group, as shown by the large
rounded tetartocones and midparietal convexity. It
resembles Megacerops in the general aspect of the non-
cingulate grinders, the smooth basisphenoid, and some-
what also in the basal horn section and the midparietal
protuberance and convexity; it radically difl^ers from
this genus in the much greater development of the
cutting teeth, or incisors, and the wide separation of
the canines. It resembles Brontotherium in the form
and size of the cutting teeth (so far as the characters
of these parts can be judged from the original figures),
the wide union of the postglenoid and post-tympanic
processes^ and the narrowing of the nasals anteriorly.
On the whole, therefore, B. tichoceras appears to stand
closest to the Brontotherium phylum, although it lacks
especially the characteristic transverse basal horn
section, which seems to be merely an exaggeration of
the basal section of the type of B. hatcheri. If the
section were preserved at a little higher point on
the horn it might show some of the characteristic
flattening.
Additional ohservations on the type. — A reexamination
of the type in 1902 afforded opportunity for additional
measurements and sections of the horns and zygomatic
arches. The canines in the type, although now lost,
were large and indicate that this was probably a male
566
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 468. — Lower jaws of Brontotherium hatcheri and B. gigas
A, Brontotherium hatcheri, Univ. Wyoming Mus. 1 (reversed); B, B. hatcheri. Am. Mus. 1070; C, B. gigas, Yale Mus. 12009 (type). One-
fiftb natural size. In these jaws the chin is shallow, almost concave, as in typical brontotheres, the ramus stout and massive, the
canines spoon-shaped, the incisors well developed, the external cingula wanting, no diastema in front of pi (if that tooth is present) .
EVOLUTION OF THE SKULL AND DENTITION OP OLIGOCENE TITANOTHERES
567
specimen. There are large, distinct tetartocones on
the superior premolars. Cingula on all the grinding
teeth are wanting. The nasals are of medium length
and thickness, contracting anteriorly and decurved
slightly at the tip. The horn pillars are stout, con-
tracting the anterior nares, but to a less extent than
in B. dolicJioceras. The postglenoid and post-tympanic
processes are widely conjoined. These characters
concur with the position of the horns, the relative
shortness of the nasals, the breadth of the zygomata,
as indicating that this species belongs on the upper
levels of the TitanotJierium zone.
Measurements of Brontotherium tichoceras
Millimeters
Molar-premolar series 330
Free length of nasals 65
Free breadth of nasals 130
Tips of nasals to connecting crest 130
Occipital condyles to tips of nasals 800?
Transitional skull (Nat.Mus. 8313). — Measurements
of this skull will be found under Megacerops (p. 542).
Additional observations on the measurements of Bron-
totherium tichoceras. — The relationships of the type
and only known skull of this species are somewhat
doubtful. It possibly belongs near B. gigas and B.
hatcheri, although the detailed dimensions of its grind-
ing teeth do not support this view but suggest, on
the contrary, possible relationship with Megacerops,
as follows:
Comparative measurements of teeth of Brontotherium tichoceras,
in millimeters
B. tichoceras
Megacerops
(S. altirostris)
B. gigas, Am.
Mus. 492
B. "hatcheri,"
Am. Mus. 1070
Ap.
Tr.
Ap.
Tr.
Ap.
Tr.
Ap.
Tr.
P'
17
26
37
41
66
76
74
20
34
48
61
64
80
85
?19
27
34
43
69
77
76
17
40
48
59
69
79
82
20
30
34
39
57
77
81
19
39
49
60
66
72
81
20
29
39
43
67
25
P2
40
ps
63
P*
63
Ml
70
M2
M'
The deep basal section of the horns is also consistent
with Megacerops affinities.
Brontotherium gigas Marsh, 1873
(Titanops elatus Marsh, 1887; "Brontotherium gigas" ,OsboTn,
1902)
Plates XX, LXXXII, CXXXIII, CLXX, CLXXI, CLXXIII-
CLXXV, CLXXXIV, CXC, CCXXX-CCXXXV; text fig-
ures 29, 165, 229, 377, 378, 382, 390, 394, 395, 405, 467, 458,
460, 468, 470-472, 516, 622, 623, 625-631, 633, 634, 639,
662, 662, 690, 719, 721, 724, 744
[For original description and type reterences see p. 209. For skeletal characters see
pp. 690-694]
Geologic horizon. — Upper Titanotherium zone, lower
to upper levels.
101959— 29— VOL 1 39
Specific characters. — P'-m' 330-353 millimeters;
m'-m^ 216-241; dental index typically 42, rising to 46.
Basilar length of skull S (typical) 830 millimeters, 9
760, zygomatic breadth c? 740, index 89 (brachy-
cephalic). Nasals rather short, free length (? 87 milli-
meters, free breadth c? 115. Horns very large, typi-
cal <? 330-365 millimeters, 9 180 but with high
connecting crest as in cf , basal section in males trans-
versely oval but not flattened, more flattened at top
but less so than in succeeding species, horns much in
front of orbits and above premaxillaries. Buccal
expansion of zygoma very
broad in males, slight or
moderate in females. Ca-
nines, incisors, premolar
tetartocones, and cingula
much as in B. leidyi and
other brontotheres.
The characters of Bron-
totherium gigas elatum as
determined from the type
jaw of B. gigas and the
type sknU and jaws of B.
{Titanops) elatum represent
the next stage beyond B.
hatcheri. This advance is
seen not in the dimensions
of the grinding teeth, which
are no greater than in B.
hatcheri (see table), but in
the more advanced evolu-
tion of the horns and con-
necting crest and in the
further reduction of the
nasals (figs. 468, 470).
General characters of
Brontotherium gigas.— y^ith
the species B. gigas, the
"giant thunder beast," we
pass to the long-horned tita-
notheres of the upper Tita-
notherium zone, recorded
by Hatcher from the lower,
middle, and upper levels.
The imposing animal B.
gigas was apparently in
the line of ascent to the
still more remarkable B. curtum, B. ramosum, and B.
platyceras of the summit. The largest bulls of B. gigas
(Am. Mus. 492) greatly exceeded those of Brontops
rohustus in size, the skull measuring 32.6 inches
(830 mm.) from the premaxillaries to the occipital
condyles by 29.1 inches (740 mm.) across the zygo-
matic arches, as compared with 765 by 667 milli-
meters in the aged B. rohustus skull. The expansion
of the zygomata renders the brachycephaly, as mea-
sured on the palatal surface of the skull, still more
marked, the length exceeding the breadth by only 3.5
Figure 469. — Sections and
contours of skull of Bron-
totherium? tichoceras
Harvard Mus. (type of Menodus
iichoceras). One-ninth natural size.
The extremities of the horns are un-
fortunately missing but must have
been massive. The basal section is
very large; the external face is flat-
tened, and the internal and posterior
faces are well rounded. Connecting
crest low, nasals massive and wide,
zygomata much expanded, parietal
crest wide and convex. The generic
reference is uncertain; although these
sections suggest those of Megacerops
bitcco, the animal may be related to
Brontoihenum hatcheri.
568
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
inches (90 mm.), and the zygomatic index rising to
89 as compared with 62 in a typical Menodus giganteus.
The proportions of the huge skeleton belonging to the
same large bull (Am. Mus. 492) show corresponding in-
crease in the length of the ribs, height of the dorsal
hump, and width of the pelvis (see above).
Materials. — Our knowledge of this species is de-
rived from exceptionally rich materials, including es-
pecially the type jaw of B. gigas, the splendid type
skull and jaw of B. {Titanops) elatum in the Yale
Museum (No. 12061), and the superb skull of B. gigas
in the American Museum (No. 492), broadly recorded
from the upper Titanotherium zone. Associated with
(allowing for a slight difference of age), in form of jaw,
canines, and molars (especially ms), in the ratio of
premolars to molars. It differs only in the presence of
Pi, which has dropped out in B. gigas, and in the
reduction of the cingulum on the premolars. The
relation of these species is further confirmed by speci-
mens in the National and American Museums which
A
B
Figure 470. — Sections and contours of skulls of Bi-ontotherium hatcheri and B. gigas
A, BrontotJierium hatcheri, Am. Mus. 1070; borns shorter than in B. gigas and connecting crest lower; hasal section of horns'roundly
trihedral, not so wide as in B. gigas. B, B. gigas, Yale Mus. 12061 (type of Titanops elatus)', horns long, erect, and placed
far in front of the orbits, their basal section roundly trihedral, section of the upper part of the horn a transversely extended
oval, nasals fairly long, parietal crest wide. C, B. gigas. Am. Mus. 492; horns stouter, basal section more rounded oval and
upper section less flattened, nasals shorter, and zygomata widely expanded. One-ninth natural size.
the skull of B. gigas are a pelvis, forearm, manus, and
dorsal vertebra with ribs. There are also two skulls
in the National Museum — No. 4262, from the lower
levels of the upper Titanotherium zone, a young male
in the fifth stage of growth, associated with consider-
able portions of the skeleton, and No. 4244, with
imperfectly preserved horns, recorded from the top
levels of the upper zone. There is also the skull of a
small but aged female (Am. Mus. 1006).
Relation of Brontotherium gigas and Titanops ela-
tus.— A comparison of the type jaw of B. gigas with
that of Titanops elatus shows that the former repre-
sents a younger, the latter a somewhat older and more
robust individual of the same species; the type T.
elatus jaw (see description below and measurements
above) resembles the type B. gigas jaw in size
are described below. The additional generic char-
acters of Brontotlierium must, therefore, be derived
not from the skull which Marsh referred to B. ingens,
but from the skulls which he referred to Titanops
elatus and Titanops curtus.
Our conclusion is that the type skull and jaws of
B. (Titanops) elatum belong to a somewhat larger
male animal than the type jaw of B. gigas and may
represent a more progressive ascending mutation,
which might be termed B. gigas elatum. The skull
and jaws of B. (Titanops) elatum are certainly more
progressive than those of B. liatcheri.
Ohservations on the type jaw of Brontotherium gigas. —
The type lower jaw was figured by Marsh (1876.1, pi.
12) and is well represented in Plate CLXXI of this
monograph.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
569
Measurements of Brontotherium gigas
Millimeters
Symphysis to angle 620
Symphysis to condyle 600
Depth of jaw, condyle to bottom of angle 305
Depth of jaw below first molar 105
Molar-premolar series 365
True molar series 258
Premolar series 117
Ml, transverse 37
Ml, anteroposterior 59
M2, transverse 41
M2, anteroposterior 78
Ms, transverse 47
Ms, anteroposterior 118
Canine enamel, anterior 34
Canine, anteroposterior 30
Length of symphysis 173
the fourth premolar and then ascends rapidly to the
incisive border with a slightly convex chin, but there
is an interval between the chin and the canine that is
concave and somewhat hollowed out beneath. The coro-
FiGURB 471. — Lower jaws of Brontotherium gigas and B. medium
A, Brontotherium gigas, Yale Mus. 12061 (type of Tiianops elatus). This animal has the measurement of mi-ma longer than in the type
of is. gigas and may represent a slightly higher stage. The angle is very prominent and rugose, as in old animals. The molars have
reduced external cingula. B, B. medium, Am. Mus. 1051; the largest known titanothere jaw. It conforms in all important respects to
the Brontotherium type. One-fifth natural size. ,
The jaw is of the true bronto there type. The men-
tal foramen is at a point vertically below the posterior
fang of the third premolar. Above and behind it are
two smaller foramina. The jaw is swollen beneath
nary process as figured is restored at the tip ; the ante-
rior border presents a smooth, rounded ridge antero-
externally, which is brought into further prominence
by a depression running just back of and parallel with it.
570
TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Teeth: The incisors are represented by two pairs of
rather small alveoli with a slight diastema between;
the lateral alveolus is slightly larger than the median;
the canines are low-crowned, obtuse, flattened, and
cingulate on the postero-internal surface and probably
indicate a male; there is a narrow diastema (11 mm.)
between the canines and second premolars and no in-
dication of pi. The premolars exhibit no external or
internal cingula; there is an external crenulation only
on the valleys of the molars ; the grinding teeth there-
fore are decidedly noncingulate. M2 and ms exhibit
a rudimentary metastylid; ms has a broad-cupped hy-
poconulid with a slightly crenulate internal crest.
The presence of lower incisors and the shape of the
canines differentiate this jaw from that of the animal
later described by Cope as Symhorodon tonus. The
looped structure of the hypoconulid of ma is different
from that in the Menodus series. The closest resem-
blances in the jaw, canines, absence of cingula, and in
ms are to the animal later termed by Marsh Titanops
elatus.
Ohservations on the type skull oj B. {Titanops)
elatum considered as pertaining to B. gigas. — The
cranial vertex is somewhat crushed and moderately
broad, with a rugose crest overhanging the orbital and
temporal fossae; the nasals are rather narrow, of me-
dium length; the external auditory meatus is com-
pletely coalesced below, the angle of the jaw is
slightly depressed, and also extended backward.
The age of the skull is determined as in the ninth
stage, all the internal cusps of the grinding teeth
being worn except upon the last superior molar.
It should be noted that the zygomatic arches, the
premaxillaries, and the anterior portion of the jaw
of this skull are wanting.
Teeth: The superior premolars exhibit a distinct
internal cone, the tetartocone, which is well marked
off and separate from the deuterocone on p', p*.
P* also apparently exhibits a mesostyle. The hypo-
cone is fairly prominent but partly connected with
the cingulum on m^ The internal cingulum of the
superior premolars is sessile and crenulate. The
lower premolars and molars exhibit a vestigial or
basal cingulum except on nis, in which the cingulum,
as in the type of B. gigas, is slightly more decided.
Ms exhibits a broad hypoconulid and crenulate
internal crest, a characteristic phyletic character.
Measurements of the type jaw are given above.
The type skull is unfortunately incomplete, so that
the chief dental measurements are lacking, but the
basilar length and the dimensions of the nasals and
horns are approached by a finely preserved skull in
the American Museum (No. 492). This is remark-
able for its very large true molars (241 mm.), its
great basal length (830 mm.), and its high zygomatic
index (87). The horns are less flattened than in
B. medium and B. curtum but more flattened than in
B. hatcJieri.
Two other skulls provisionally referred to this
stage (Carnegie Mus. 341, Nat. Mus. 4262) agree
closely in measurements with each other but have
the true molars (218 mm.) and the horns (275 mm.)
considerably shorter than in Am. Mus. 492; another
skull (Nat. Mus. 4244) has much shorter horns
(170 mm.). Thus these skulls seem to connect
B. gigas elatum with B. Tiatcheri and serve to illus-
trate the variability in length of the horns and of the
nasals.
The female skull (Am. Mus. 1006) which is referred
to this species approaches the male skulls in the
dimensions of the cheek teeth as well as in the sec-
tions of the horns and nasals. But it differs from all
typical BrontotJierium skulls in having well-defined
internal cingula and retarded tetartocones in the
premolars; so that relationship with Brontops rohustus
might be suspected were it not for the much closer
agreement in dental measurements with B. gigas.
Fine male slcuU (Am. Mus. 492). — The fully adult
characters of the males are admirably shown in the
nearly perfect American Museum skull No. 492,
which is in the seventh stage of growth. Two smooth,
rounded incisors are preserved on one side above; the
canines measure 35 millimeters anteriorly as compared
with 41 in B. robustus; they are shorter and obtuse
(Pis. XX, B, CLXXIII, CLXXIV). There is no
diastema behind the canine. The greatly reduced
first premolar is pressed close to the canine and tends
to drop out; the premolars can readily be distinguished
from those of B. rohustus; the tetartocones although
low and obtuse are about two-thirds as large as the
deuterocones and quite distinct throughout. The
premolar cingula, however, have retrogressed as in
B. leidyi, being practically vestigial on the outer side
and less distinct on the inner side than in B. ro-
hustus; the molar cingula also are nearly obsolete.
On m' is a sessile hypocone connected with the
cingulum.
The grinding teeth, considered so important to the
welfare of large herbivorous quadrupeds such as this,
are in certain respects in a condition of retrogression.
In the type of B. gigas the bluntness and crenulation
of the internal cingula precede degeneration. In
this very large skull (Am. Mus. 492) they measure
353 millimeters, only 3 millimeters more than in the
type skull of B. rohustus. This contrast between the
arrested development of the teeth and the pro-
nounced evolution of the dominant protuberances of
the skull has its parallel among the Dinocerata.
The inverse relations of the nasals and horns illus-
trate the law of compensation of growth, the free
nasals being here absorbed and reduced in compen-
sation for the great elongation and expansion of the
horns, which are now strengthened by a very promi-
nent crest, attaining a vertical thickness of 140 milli-
meters as compared with 78, the maximum thick-
ness in B. rohustus. Behind the horns the skull
EVOLUTION OF THE SKULL AKD DENTITION OF OLIGOCENE TITANOTHERES
571
slopes into a broad saddle shape, more uniform in
width anteroposteriorly than in B. rohustus, partly
owing to lateral crushing, but the vertex is much
longer, especially as seen in the backward extension
of the occiput behind the zygomata.
The parietal vertex is very wide, with overhanging
supraorbital and supratemporal crests, terminating
in a powerful rugose occiput which is broader and
more shelf-like superiorly than in M. rohustus. The
buccal processes are flatter superiorly, though the
area of the section is not greater than in the large
specimen Am. Mus. 6346, which is provisionally re-
ferred to B. curtum; as in B. leidyi there is a flange
extending inward from the center of the zygoma
(fig. 470, c). The pillars of the occiput are very
powerful, with a pair of tuberosities (fig. 378, F, k, k)
projecting upward on either side of the median line.
We have seen these tuberosities in Megacerops and in
prophetic form in B. leidyi. They represent the
development of special muscular fasciculi or tendons
for the semispinalis capitis muscles and the liga-
mentum nuchae; these tuberosities are rudimentary
or absent in certain skulls of Menodus and are re-
placed by two pits in Megacerops acer. There is a
wide union between the post-tympanic and post-
glenoid processes; each pterygo-alisphenoid wing is
cleft on its extreme lower border for the insertion of the
pterygoid muscle; the pterygoid wings are somewhat
shorter than in Brontops rohustus. As in B. leidyi
and in contrast to B. rohustus, there are no paired
protuberances at the junction of the basioccipital
and basisphenoid for the attachment of the recti
capitis antici majores. The vomerine ridge is reduced.
Juvenile stage of growth. — A relatively young male
skull in the fifth stage of growth (Nat. Mus. 4262),
with which are associated the backbone and limbs
(see above), illustrates the form of the horns, nasals,
and zygomata of this species in the young condition
and serves to strengthen the identification of Titanops
elatus with B. gigas. It was found at Hat Creek,
Wyoming, and is recorded by Hatcher as from the
lowest level of the upper Titanoiherium zone (Chadron
C). The horns measure 275 millimeters. The for-
mula is If, Pf; the upper incisors are small and
conic without posterior cingulum; the lower incisors,
of which the outer one is preserved, have a crenulate
posterior cingulum; the superior canines measure 33
millimeters and are short and obtuse, with a charac-
teristic posterior slope and cingulum; the inferior
canines have a postero-internal, less prominent
cingulum. The tetartocones on p^-p* are very dis-
tinct; p* has a mesostyle as in the type specimen.
The jaw resembles that of the type of B. gigas in the
chin and angle; the chin is extremely shallow.
Character of the female slculls. — The determination
of a small skuU (Am. Mus. 1006) as a female of this
species is rendered probable by its discovery on the
same level as the large B. gigas (Am. Mus. 492), as
well as by the sections of the nasals, horns, and zygo-
mata. The disparity in size between this female
skull and the male skull is very marked. The animal
is well advanced in the seventh stage of growth.
The deep connecting crest and shape of the nasals
resemble those of B. gigas, but the horns are very
short and obtuse, and the buccal processes are much
less robust. The malar bridge is broader than in
the male skull, as in B. curtum female, and the median
ridge is obtuse and low. The post-tympanic has a
characteristic broad, flattened union with the post-
glenoid; the palatine aspect also resembles that of
B. gigas on a small scale, with a smooth basisphenoid.
Sexual characters are observed not only in the
extremely small canines but apparently also in the
reduction of the incisors, so far as we can judge from
the alveoli and fangs. Careful examination reveals
the reduced fang of a lateral incisor on the right side.
The canines differ from those of the male not only in
the much smaller size but also in lacking the swollen
appearance; they are, in fact, recurved, conic. The
premolars differ from those of the male in the lesser
size of the tetartocones and in the pronounced non-
crenulate cingulum.
A very large skull in the National Museum (No.
4244), in the fourth stage of growth, either a large
female or a juvenile male with imperfectly developed
horns, is recorded by Hatcher from the top level of
the upper zone. The horns measure 196 millimeters,
with the anterior ridge more prominent than in the
type. Other characters conform closely to the type —
namely, I^, distinct tetartocones on p*, m^ with a small
but distinct hypocone, buccal processes flattened. A
variation is the loss of p'.
Female characters in general. — A comparison of the
supposed female skulls of Brontotherium gigas and of
B. curtum shows that since the male progressive
characters are largely correlated with excessive size
and horn development the females actually appear
to be more primitive — namely, the horns are shorter,
the zygomata are slighter, the occiput is less extended
backward, the malar bridge is wider, the orbit is
larger, the external auditory meatus is somewhat more
open.
The apparent loss or reduction of the incisors, the
somewhat retarded condition of the tetartocones, and
the pronounced development of the internal and
external cingula are, however, verj^ difficult facts to
explain as sexual characters.
Additional observations on the supposed female of
Brontotherium gigas. — A female skull in the American
Museum (No. 1006) which has been referred to this
species presents a very puzzling specimen. The sections
of the horns and nasals offer resemblances and differ-
ences to those of the type of B. (Titanops) elatum which
have been assumed to indicate a female of this species.
572
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
On the other hand, the specimen differs from male
brontotheres and resembles the Menodontinae, espe-
cially Dijjloclonus amplus, in the following features:
(1) The narrow well-defined
internal cingula; (2) the
dimensions of p* (ap. by
tr., 40 by 61 mm.) and m^
(84 by 82) are closer to
those of a skull referred to
Diploclonus amplus (Nat.
Mus. 4710), which are re-
spectively 42 by 61 and
82 by 81 millimeters,
than they are to those of
male brontotheres, which
usually have larger p* and
wider m'; (3) the canine
(PI. CXC) is not of the
swollen Brontotherium
type; by analogy with
Menodus the female canine
in Brontotherium should
be somewhat like that of
the male, only much more
slender; (4) the cranial
vertex lacks the midpa-
rietal convexity or emi-
nence of BrontotJierium and
Megacerops. The antero-
posterior measurements of
the dentition (see above) do
not offer decisive evidence
of relationship with Bronto-
tJierium. In brief, the sys-
tematic position of this
skull at present appears
doubtful and it may pos-
sibly belong near Diploclonus amplus. (See fig. 472.)
Figure 472. — Sections and
contours of skull of Bronto-
therium gigas?
Am. Mus. 1006, 9, referred to B. gigas on
account of the characteristic form of
these sections. It seems, however, to
be too small to be a female of that
species (W. E. Gregory). One-eighth
natural size.
Brontotherium dolichoceras (Scott and Osborn)
{Menodus dolichoceras Scott and Osborn, 1887;
"Brontotherium dolichoceras" Osborn, 1902)
Plate CLXXXVI; text figures 177, 473, 474
[For original description and type references see p. 220J
Geologic horizon. — TitanotJierium zone
of South Dakota, level not recorded.
Specific characters. — Nasals, free length'
49 millimeters, breadth 90 millimeters —
that is, more reduced than in B. gigas,
less reduced than in B. curtum. Horns
elongate (310 mm.), about as in B. gigas;
horn section transversely oval, more progressive than
in B. gigas, with anterolateral depression and incipient
external ridge. Size of skull about as in B. hatcheri.
Premolar cingula reduced, which perhaps is an indi-
vidual variation.
"^■■^
As described the type of this species is an adult,
not aged male preserved in the Museum of Compara-
tive Zoology at Harvard. It is in the seventh stage of
growth, with the internal
cones of p^ completely worn
and the protocone and hy-
pocone of m' worn. As
shown by the measurements
given below, the skxill was
of medium size, or about as
large as the type skull of B.
hatcheri.
This species, named "the
long-horned brontothere " by
Scott and Osborn, repre-
sents a next higher stage in
the evolution of the bron-
tothere phylum, although it
appears rather as a collat-
eral branch than as a mem-
ber of the main line of as-
cent. The type skull is the
only material that can be re-
ferred certainly to this spe-
cies. The nasals now project
only 49 millimeters in front
of the anterior bases of the
horns, and in sagittal section (fig. 473) there is a uniformly
convex curve from the summit of the connecting crest to
Figure 473. — Sections and
contours of skull of Bron-
totherium dolichoceras
Harvard Mus. (type) elongate horns
subo val in basal section, nasals short.
One-ninth natural size.
Figure 474. — Skull of Brontotherium dolichoceras
Harvard Mus. (type). The right horn is longer and more slender than the left, possibly an effect of
crushing. (See PI. CLXXXVI, C.) One-eighth natural size.
the tips of the nasals, as in the fine specimen of B. curtum
(Nat. Mus. 4946) described below. The horn section
on the right side is the more perfect, that on the left
being crushed laterally; the horns are crushed in upon
the antero-external faces, a feature which exaggerates
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
573
the flatness of the anterior face of the horn section
(fig. 473). At the sides of the horns there is a short
extension of the external crest, and in this individual
the malar, anterior, and intei'nal angles are not very
sharply defined. Characteristic of the horns are
the external ridge extending from the sides of the
nasals upward and the very marked hollowing out
of the maxillary face. As observed in the original
description the great size of the maxillary horn pillars
reduces the opening of the anterior nares to a narrow
aperture. The connecting crest is fairly prominent.
The vertex of the skull has the very characteristic
longitudinal uniform breadth. The occipital crests
are wanting, but the paired knobs (fig. 378, F) are
preserved. As in Brontotherium and Megacerops
the basisphenoid is without rugosity. As in brachy-
cephalic skulls generally the postglenoid and post-
tympanic processes are widely conjoined, the auditory
meatus being very small.
Teeth. — The upper premolars are distinctly of the
Brontotherium- Megacerops type, with well-rounded,
distinct tetartocones and reduced cingula.
Conclusion. — This skull differs from the type of
B. (Titanops) medium (Marsh), in its smaller size
and in the absence of the basisphenoid rugosity. It
exhibits some resemblances to Megacerops acer but is
distinguished by the more marked divergence of the
horns. In the absence of knowledge of the occiput and
of the condition of the cutting teeth, incisors, and
canines, the phyletic position of this species can not
be considered as absolutely determined at present.
It appears possible, however, that in spite of its
somewhat smaller size (compare the range in size in
skulls referred to B. gigas) this species may be the
same as that later termed B. medium by Marsh.
Brontotherium medium (Marsh)
(Titanops medius Marsh, 1891)
Plates XIX, XX, CLXXII, CLXXVI, CLXXX; text figures
25, 191, 395, 407, 471, 475
[For original description and type references see p. 228]
Geologic horizon. — Upper Titanotherium zone of
South Dakota.
Specific characters. — Size about that of a large B.
gigas; basilar length of skull 825 millimeters; grinding
series elongate (365 mm.), dental index 43 (about as
in B. gigas); nasals more reduced (free length 45 mm.,
free breadth 110); horns long (320 mm.), becoming
flattened in section, connecting crest in type shal-
lower than in B. gigas or in B. curtum.
So far as we can observe this species appears to be
in a stage of evolution very similar to that of the
previously described B. dolichoceras.
The name "medius^' was probably assigned by
Marsh in reference to the intermediate length of the
nasals between those of B. gigas and B. curtum.
Materials. — Our knowledge of this evolution stage
is afforded by the type specimen of B. (Titanops)
medium, a large and finely preserved male skull in the
United States National Museum (No. 4256). It is
recorded as from the top levels of the upper Titano-
therium zone. The grinding series in the type speci-
men of B. medium is exceptionally elongate, namely,
365 millimeters; consequently we have associated
with this species a very large lower jaw (Am. Mus.
1051).
Slcull. — We notice especially the persistence of two
upper incisors and the prominence of the well-rounded
tetartocones on the premolars, that upon the right, p"*,
being sharply distinct. The incisors differ from the
typical brontothere form in exhibiting smooth, rounded
crowns, apparently noncingulate. The canines are
short (38 mm.), heavy, and blunt, as in B. gigas, with
robust posterior cingula. The anterior premolar, p',
is triangular and exceptionally small, with a single
much-worn internal cusp. The hypocone of m^ is
small and cingulate, but from this cusp there extends
completely across the crown the abortive metaloph
so characteristic of many members of this and other
phyla. The animal is still young, being intermediate
between the fifth and sixth stages of growth ; the only
internal cones which are worn are those of m.\ yet the
horns attain an outside measurement of 320 millime-
ters, proving the rapid individual development of these
important organs. The horns are distinguished by
the sharp carrying up of the external and anterior
ridges to a high point. Thus the outer or maxillary
face is concave below and flattened above. This fea-
ture is also seen in a less conspicuous manner in the
type of B. dolichoceras. The nasal section and the
length of the nasals (45 mm.) are also identical in the
two specimens.
Lower jaw. — Possibly belonging to this species is the
enormous jaw in the American Museum collection
(No. 1051). As shown in Plate CLXXII, B, and in
Figure 471, B, this has the characteristic form of the
canines, the posterior molar, angle of the jaw, etc., of
B. gigas, but the lower incisors show some reduction
and lack the decided development of the cingulum
observed in No. 1070 (B. hatcheri). Moreover, the
great length of this jaw (730 mm.), measured from the
condyle to the symphysis, favors its reference to
B. medium rather than to B. ramosum or B. platyceras,
in which the jaw is relatively shorter.
Conclusion. — B. medium is one of several stages of
the upper zone resting on a single type and tending
to connect B. hatcheri, B. gigas, B. dolichoceras,
B. medium, B. curtum, B. ramosum, B. platyceras in a
more or less continuous phjdum.
Additional observations on Brontotherium medium. —
The type skull is remarkable for the large size of the
574
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
premolars and molars and for the shortness of the
nasals. The other two skulls provisionally entered
under this species above (p. 553) have much smaller
dental measurements, about as in B. gigas or Jiatcheri,
and may belong in either species. The female skull in
the British Museum, referred provisionally to B. pel-
toceras, approaches the female skull referred to B.
curtum (Am. Mus. 1005) except that the nasals are
smaller. Future investigation may settle whether
B. medium is distinct from B. gigas elatum.
very distinct tetartocones ; external cingula feeble;
internal cingula reduced. Nasals greatly abbrevi-
ated, free length 52 millimeters. Horns greatly
elongated in males (355-380 mm.), in females abbrevi-
ate (160 mm.); basal section of horns strongly con-
vex anteroposteriorly, slightly convex to plane pos-
terioi'ly, breadth moderate. Zygomata with buccal
expansions broad, plane above, bulging at the sides,
plane below. Skull mesaticephalic, basal length cf
790-840 millimeters, zygomatic breadth c? 620 milli-
A B C
Figure 475. — Sections and contours of skulls of Broniotherium medium and B. curtum
A, Brmtoiliermm medium, Nat. Mus. 4256 (type), short nasals approaching in form those of B. curium, but horns not so long; basal section of horns has a
flattened external face. B, B. curtum, Nat. Mus. 4946. C, B. curtum, Yale Mus. 12013 (type). In B. curtum nasals are very short, horns long, basal
section well rounded externally and flat posteriorly, zygomata well expanded. One-ninth natural size.
Brontotherlum curtum (Marsh)
{Titanops curtus Marsh, 1887; fMenodus peltoceras Cope, 1891;
" Brontotherlum curtum" Osborn, 1902)
Plates XX, XLVII, CLVII, CLXXVII-CLXXX, CLXXXIII-
CLXXXVI, CXC, CXCII-CXCIV; text figures 182, 390,
392-394, 399, 457-459, 475-478, 709, 719, 744
[For original description and type references see p. 224]
Geologic horizon. — Upper Titanoiherium zone of
South Dakota.
Specific and generic characters. — Molar-premolar
series 350 millimeters (estimated). Premolars with
meters; zygomatic index 73-77. Vertex of skull
elongated by expansion of occiput back of the zygo-
matic arches and by forward shifting of the horns.
Female skull much smaller (nasals to top of occiput
635 mm.).
The chief advances in this stage beyond B. gigas
and B. medium are in the still further abbreviation
of the nasals, the more flattened posterior face of the
horns, the more backward extension of the occiput,
all progressive characters which are bridged over more
or less fully by transitional types. The disparity
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
575
between the males and females is now still more
marked, the latter being apparently represented by
specimens referred to the species B. {Menodus)
peltoceras Cope.
Materials. — Besides the noble type skull of Titanops
curtus in the Yale Museum (No. 12013) this species
is represented by a pair of horns in the Harvard
Museum and by a large, fully adult skull (Nat.
Mus. 1211) and the skull of an old bull (Nat. Mus.
4946) in the National Museum. Both these skulls in
the National Museum are recorded from the top
level of the upper TitanotJierium zone (Chadron C).
In the American Museum there is fortunately a
small female skull, No. 1005. Cope's type of
Menodus peltoceras (Am. Mus. 10719) consists simply
of the great connecting crest supporting a pair of
low horns.
Detailed description of the type sJcull. — The type
male skull (Yale Mus. 12013, Pis. CLXXVIII-
CLXXX) is in the seventh stage of growth. The
horns have therefore not attained their maximum
length, and the connecting crest is comparatively low
and rounded. The horns are placed well forward
and vertically overhang the canines. The external
ridge has increased, while the anterior ridge has
practically disappeared (PL CLXXIX; figs. 457, G;
475-478). The section of the horns is now almost
planoconvex, the inferior face being strongly convex
and the posterior face almost plane; at the summits
the horns exhibit a rugose lateral expansion suggestive
of that which is so strongly marked in B. ramosum.
Another progressive feature is the very sudden expan-
sion of the zygomatic arches into a broad and relatively
shallow convexity (Pis. CLXXVIII, CLXXX). The
vertex of the cranium is bounded by a lateral crest
with a rugose border overhanging the temporal fossa.
Dentition. — The superior canines are short, re-
curved, with a broad posterior cingulum. The pre-
molars exhibit well-separated internal cones, es-
pecially on p' and p^, there being a distinct cleft
between the deuterocones and the tetartocones. The
first superior premolar is indicated by an alveolus on
the left side. In the molar-premolar series the
enamel is crenulate on the sides of the internal cones.
On m' the hypocone is fairly prominent but not
separate from the cingulum.
Age and growth characters. — As above noted the type
male skull, being in the seventh stage of growth, does
not exhibit either the maximum length of the horns
or the maximum development of the connecting crest.
In the fully adult male skulls in the National Museum
(Nos. 4946 and 1211), which are in the eighth stage of
growth, we have finely illustrated the skull characters
of the old bulls. The connecting crest of No. 1211,
which is the oldest and most progressive skull, is now
extremely deep and descends by an almost straight
line to the tips of the nasals, which are now only 40
millimeters in length. The horns are very long
(355 mm.), recurved, with the characteristic strongly
convex section in front and slightly convex section
behind. The second old bull in the National Museum
(No. 4946) is somewhat less progressive, the horn sec-
tion (fig. 475, B) being very convex anteriorly and less
convex posteriorly. The horns are relatively broader
and more flattened posteriorly, and the nasals are still
more abbreviate (65 mm.). The horn of this speci-
men has been broken off and partly regenerated during
life, a fact which appears to show the value of the
horns in contests between the males.
A specimen in the Harvard Museum (No. 1004)
represents a much smaller animal in an earlier stage of
evolution, in which the horns are still convex on the
posterior surfaces. All these specimens agree with
B. gigas in the backward extension of the occiput
behind the zygomata, in the inward flange of the
zygomata, and in numerous other details of character.
Other features in the National Museum skull (No.
4946) are the following. It appears that skull growth
continued even after the teeth were much worn down.
Added to the progressive feature of the lengthening
and flattening of the horns and the abbreviation of the
nasals, marked in this bull, is the incipient develop-
ment of a narrow ridge on the outer side of the horns,
an exaggeration of the "external or malar ridge"; it
does not extend very far but is marked in old indi-
viduals. This "malar ridge" is prophetic of the con-
tinuous external ridge down the entire side of the horn
in the higher phyletic stage, B. platyceras. The
structure of the zygomatic arch is especially interesting;
immediately behind the orbit it consists of a vertically
compressed plate. A variation of interest is the
small tuberosity on the inner side of the horn, sug-
gestive of the internal hornlet of Diploclonus.
Carnegie Museum specimen. — A fine pair of horns
in the Carnegie Museum (No. 560) which are referred
to this species have an external height of 365 milli-
meters and a basal anteroposterior diameter of 103
millimeters. They are of special interest as showing
a pair of secondary horns or hornlets, which consist of
elongate oval swellings 78 millimeters long at the base
and about 22 millimeters high, located on the internal
border of the main horns.
Male slcull referred hy Cope to " Sy7nhorodon" iucco
(Am. Mus. 6346). — As we have seen above, the actual
type of " Symhorodon" hucco is a skull (Am. Mus. 6345)
which belongs to Megacerops hucco. Another speci-
men described and figured by Cope as "Symhorodon"
hucco (Am. Mus. 6346) is probably referable to B.
curtum. The buccal processes are very broad (656
mm., estimated), arching suddenly outward, deep and
biconvex in section. The occipital crest is backwardly
extended. There is a median vertical ridge over the
foramen magnum, and a pair of rugosities on either
side of the median line. The basisphenoid is smooth.
576
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
It is probable that this skull belongs to Brontotherium
and not to Megacerops, because it has prevailing
Brontotherium characteristics. There are, however,
Figure 476. — Horns of Brontotherium curium
Carnegie Mus. 560. Posterosuperior view of horns showing rugose tips
and accessory swellings or hornlets. (After Peterson.) One-eighth
natural size.
some differences. The zygomatic buccal sec-
tion is somewhat more convex than that in
the type of B. curium, which is flatter. From
the evidence now at hand we might regard
this skull as representing an intermediate or
connecting stage between B. gigas and B.
curium, and we should expect that if com-
plete it would be found to possess horns and
nasals also intermediate between those of the
two species. We shall await subsequent dis-
coveries with great interest.
Female sliull of B. curtum. — A small skull
in the American Museum (Am. Mus. 1005)
was formerly referred by the present author
to a female of B. gigas; but the nasals are
so excessively short and the sagittal section
so strongly resembles that of B. curtum that
this small female skull is more probably ref-
erable to this species. Like the female origi-
nally referred to B. gigas, it illustrates
afresh the great disparity in size between the
cow and bull titanotheres at this period
of evolution. The apparently primitive but
actiially sexual characters exhibited in this
skull have been noted alreadj^. As in the
female skull of B. gigas the premolar tetarto-
cones are less distinctly circular than in the
males. The internal cingulum is less reduced,
the occiput is not so greatly prolonged back
of the zygoma, and the canines are short
and recurved rather than bulbous. This list
of sexual characters appears like a summary
of primitive characters. The premaxillaries
are edentulous, or toothless, confirming the
evidence afforded by the B. gigas female skulls,
that in the brontotheres the upper incisors were re-
duced or wanting in the females — a conclusion, how-
ever, which requires final substantiation by additional
evidence. The malar bridge over the infraorbital
foramen is broader than in the male. While the
frontal section closely resembles that of B. curium,
the sexual disparity is illustrated not only by the light
zygomata and narrow occiput, but by the irregularly
osseous summits and by the form of the horns, which
are less flattened posteriorly than in the male type of
this species (Pis. CXC, B; CXCII; figs. 459, B; 477, B).
Of interest is the vertical septum extending from
the under surface of the nasals to meet a similar
septum rising from the premaxillaries, which is seen
also in other skulls in the different phyla. WhUe in-
complete, this septum illustrates a tendency toward
the formation of an intranarial septum similar to that
in the rhinoceroses.
A
B
Figure 477. — Sections and contours of skull of Brontotherium curtum
A, British Mus. 5629. One-eighth natural size. This specimen has short horns, a high con-
necting crest, and moderately expanded zygomata, as in the supposed females of Bronto-
therium, but the canines are larger than in females and suggest that this is a male. It
resembles Cope's type of " Menodus peltoceras." The basal section of the horns is flat-
tened to concave posteriorly and has a flattened external face. The nasals are small and
pointed. B, Am. Mus. 1005,9. The horns are even shorter than in the preceding speci-
men, and the connecting crest is equally high. The basal section of the horns is wide,
flattened posteriorly and antero-externally; the zygomata are moderately expanded.
Another character which parallels especially the
brachycephalic rhinoceros is the flattening out of the
lower lateral portion of the occiput.
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHERES
577
Additional observations on the supposed female oj
Brontotherium curium. — The female skull described
above (Am. Mus. 1005) has been referred to this
species because of the abbreviated nasals, high
connecting crest, and flattened horns. The shortness
of the horns has been assumed to be a female char-
acter. The systematic position of this skull, however,
is hardly free from doubt, and the remarks made above
with reference to the supposed female of B. gigas may
possibly apply here, especially in view of the dimen-
sions of p* (ap. 40 mm., tr. 61) and m^ (84 by 82),
which suggests reference to the Menodontinae rather
than to the Brontotheriinae. However, the premolar
external cingula in the present specimen are obsolete,
the tetartocones are better developed, and the antero-
posterior measurements of the dentition are also close
to those of male skulls of B. curtum, with which species
this skull may be left for the present.
Female sTcuU, Cope's type oj Menodus peltoceras
(Am. Mus. 10719). — This type, described in detail
in Chapter III, page 230, confirms the evidence afforded
by the American Museum specimen No. 1005 as to
the shortness of the horns in the females of these
upper-level brontotheres. It resembles Am. Mus.
1005 in the extreme abbreviation of the nasals, in
the steepness and size of the connecting crest, and
in the marked prominence of the vertical ridge on
the outer side of the horn. In sagittal and basal
section the specimen agrees best with the female skull
referred to B. curtum, but the still more extreme
reduction of the nasals is equaled only in the species
B. platyceras, to which this fragment may possibly
belong.
Conclusion. — The type specimen of M. peltoceras
probably represents a female of one of the long-
horned species of brontotheres, probably B. curtum.
A finely preserved skull in the British Museum,
(No. 5629) is represented in Plates CXCIII, CXCIV
and Figure 477, A.
It presents a puzzling character in the large size
of the canines, which resemble those of a male; in
all other measurements it agrees with the male speci-
mens of B. hatcheri and B. gigas, but in the con-
formation of the peculiar shield in front of the skull
and of the abbreviated nasals this certainly resem-
bles the supposed female of B. curtum (Am. Mus.
1005). It also exhibits a still stronger resemblance
to the type of Menodus peltoceras Cope just described.
The canines and grinding teeth in form are those of
a true Brontotherium. It is difficult to interpret this
specimen satisfactorily. If it is actually a male it
may indicate that B. peltoceras was a distinct, peculiar
species of bronto there with short horns; or this may
be an aberrant specimen, either a female in which
the canines are exceptionally large, or an aberrant
male in which the horns are exceptionally short.
This specimen certainly raises some doubt as to
our interpretation of the female sex of the skulls
described above.
Brontotherium ramosum (Osborn)
{Titanolherium ramosum Osborn, 1896; "Brontotherium ramo-
sum" Osborn, 1902)
Plate CXCI; text figures 194, 457, 479
[For original description and type references see p. 231]
Geologic horizon. — Upper Titanolherium zone of
South Dakota.
Specific characters. — I^-, C^, P^, M^. Premolars
with obsolete external and reduced internal cingula;
with two distinct internal cones. Molar-premolar
series 350 millimeters. Skull brachycephalic. Horns
elongate, in males 399 millimeters; gently plano-
convex at the base. Very broad, gently biconvex and
laterally expanded at the summits, with a very deep
connecting crest. Nasals greatly abbreviated. Zygo-
mata expanded into two wide, flat plates.
Figure 478. — Left horn and nasals of Brontothe-
rium curtum? (female)
Am. Mus. 10719 (type of "Menodus peltoceras"); White River,
Colo. Front view. One-fourth natural size. This fragment
represents the left horn and coossified nasals of a brontothere
allied in form to British Mus. 5629. It may have been a
female of one of the long-horned, short-nosed types {B.
curtum, etc.).
This species or ascending mutation appears to be
a further evolution of the B. gigas, B. curtum, B.
medium phylum. It is related in many characters
to B. curtum and, on the other hand, is transitional
toward B. platyceras, as shown in the sections of the
horns, nasals, and zygomatic arches (fig. 479).
This "branching" or " spreading-horned " bronto-
there is represented only by the type skull of a very
old bull in the American Museum collection (No. 1447)
in the tenth stage of growth, and by a pair of horns
with connecting crest in the National Museum
(No. 1243).
From this somewhat scanty material the validity
of the stage entitled B. ramosum awaits confirmation.
As a proof of extreme age even the cingulate hypocone
of m' is well worn, a rare occurrence. Note also that
578
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
the skull is covered with exostoses, which are partly
age or even senescent characters and may represent
extreme development of the B. curium type of skull.
Among these exostoses are the branching internal
expansions of the tips of the horns, to which the name
ramosum refers. Remarkable exostoses are observed
on the outer and inner portions of the buccal zygo-
matic processes and on the rugose tips of the
occipital pillars.
The type skull is vertically crushed; this interferes
with the natural
position of the
horns, which are
pitched strongly
forward, whereas in
life the horns must
have been vertical
or even slightly re-
curved, as in well-
preserved speci-
mens of B. gigas and
B. platyceras. Zy-
gomatic expansion
has now reached a
still more extreme
stage; the total
length of the skull
along the basal line
is closely estimated
at 741 millimeters,
while the breadth
across the zygo-
matic arches now
attains 775, the
breadth thus ac-
tually exceeding
the length. The
appearance of
breadth is, how-
ever, greatly in-
FiGUBB 479. — Sections and contours of creased by crush-
skull of Brontotherium ramosum mg. The molar-
Am. Mus. 1447 (type). One-ninth natural size. The premolar SCricS is
marked forward pitch of the horns, the extreme hqw shortened tO
flatness of their basal section, and the extreme „-„ .i,-
width and flatness of the zygomata have all been oOV millimeters,
emphasized by crushing, but even before crushing -wJiick is IcSS than
the anterior face of the horns was probably convex, .
thenasalswereextremely short, and the zygomata m ■^' 7Yie(t%U7Yl.
were much expanded. The horns are still
longer than in the preceding types, measuring 399
millimeters (15.7 mches) as compared with 355 in
the oldest male of B. gigas and 380 in the type of
B. curium.
A clear separation from the B. gigas type is seen in
the frontonasal or sagittal section, which resembles
very closely that of the aged B. curium, attaining a ver-
tical depth of 160 millimeters. The horn section
near the base is, however, quite different from that of
B. curium, being relatively broader, less convex
anteriorly, and quite as flattened or even slightly
concave posteriorly. The apical section of the horns
is far more flattened and broadened, being very slightly
convex on both sides, whereas the apical section of
the B. curium type is more similar to the basal section.
The external ridge, unlike that in B. plaiyceras,
extends down two-thirds of the side of the horn but
does not reach the malars in front of the orbits. The
nasals are of about the same length as in B. curium.
The buccal processes are extraordinarily broad and
flat; they exhibit the internal projection toward the
temporal fossa, also seen in B. gigas. At the back of
the skull the post-tympanic forms a broad and close
union, practically a synostosis, with the postglenoid,
narrowing down the auditory meatus to a small tube.
In the median upper portion of the occiput we do not
observe the small pair of tuberosities (fig. 378, F)
which characterize B. gigas. The tops of the occipital
pillars are grooved by the upgrowth of the rugosities.
Unfortunately the premaxillaries are broken, and
the maxillaries lack the first premolars, which are
represented by alveoli. A very striking progressive
feature is the complete separation of the tetartocones
from the deuterocones in p^-p*, so that although
well worn down an enamel isthmus still separates the
two areas of dentine. The external cingula of these
teeth are obsolete; the internal cingula are ill defined.
B. ramosum therefore represents a very advanced
stage of evolution but does not reach the climax
attained in this remarkable series of skulls by the
succeeding stage known as B. platyceras.
The type skull is badly crushed, giving a false ap-
pearance of extreme width. In dental measurements
it is closer to certain specimens of B. liatcJieri than it
is to B. plaiyceras.
Brontotherium platyceras (Scott and Osborn)
{Menodus platyceras Scott and Osborn, 1887; "Brontotherium
platyceras" Osborn, 1902)
Frontispiece; Plates XVIII, XIX, CLXXXI, CLXXXII,
CLXXXVII-CLXXXIX; text figures 10, 18, 24, 27, 33
178, 375, 388, 399, 457-459, 480, 481, 620, 640, 643, 648,
649, 661, 707, 725, 726
[For original description and type references, see p. 221]
Geologic Tiorizon. — Upper Tiianotherium zone of
South Dakota.
Specific characters. — Dental formula in males I-,
C-, P^^, M^. Incisors with rounded, smooth, oval
crowns, long axis of crowns anteroposterior, canines
(? 42 millimeters. Premolars with external cingula,
obsolete, internal cingula reduced; tetartocones fairly
distinct. Premolar-molar series 337 millimeters. Skull
brachycephalic, 730 by 850 millimeters. Nasals vesti-
gial, 20 millimeters. Horns extremely flattened trans-
versely, slightly convex posteriorly; deep connecting
crest, continuous malar ridge. Zygomata with broad,
deep buccal processes.
EVOLUTION OF THE SKULL AND DENTITION OP OLIGOCENE TITANOTHERES
579
Materials. — This grand stage, the "flat-horned"
brontothere, which represents the climax of the
evolution of the long-horned titanotheres, is compara-
tively rare. It is represented by the type horns in
the Harvard Museum, also by another pair of horns
in that museum. The well-preserved skull of an
old male in the American Museum (No. 1448) was
discovered by one of the expeditions under Mr.
J. B. Hatcher; it agrees closely with the type and may
be taken as a neotype. The finest skull known is
backward. The nasals are now mere pendent tuber-
osities from the base of this plate and the external
nares open directly forward. At the sides of the head
there were the great cheek processes. As this was the
face of an animal in the seventh stage of growth, we
may imagine the grotesque appearance of an old bull
in the tenth stage. The breadth of the skull has now
reached its extreme, the width of the zygomatic
arches being about 32 inches (815 mm.), exceeding
the length (distance from tips of the premaxillaries
to the occipital condyles) by 3.35 inches (85 mm.).
It is evident when we compare these correlated pro-
gressive characters of the skull with the compara-
tively stationary characters of the teeth that the chief
force of evolution or of selection was directed toward
the development of the horns and buccal processes;
while the teeth, so essential to the vitality of this
Figure 480. — -Restoration of Brontotherium plalyceras
By Charles E. Knight. About one-ninth natural size.
that in the Field Museum (No. 12161), discovered
by an expedition under Mr. E. S. Riggs.
Brontotherium platyceras is one of the most re-
markable and exceptional imgulates known to science.
Both in front and in side view the head presents a
unique and extraordinary appearance.
The connecting crest between the horns has grown
to such a height and the horns to such a breadth that
in front view the head terminates in a great vertical
plate 6.75 inches deep and more than 18 inches broad
at the widest part. From the sides of this plate the
horns extended upward and were gently recurved
species, were stationary. It is difficult to account for
the sudden extinction of this apparently prosperous
and numerous race of brontotheres; but the mechanical
imperfection of the teeth may have been one of the
factors in the extinction as fully discussed in a sub-
sequent chapter.
Teeth. — An exception to the general retrogression
of the teeth is the persistence of the two upper in-
cisors, as in all previously described males of the
Brontotherium phylum. As preserved in the neotype
skull (Am. Mus. 1448) the summits of the superior
incisors are smooth and oval, with the long axis
580
MTANOTHERES of ancient WYOMING, DAKOTA, AND NEBRASKA
anteroposterior. It is difficult to imagine of what
service these small incisors could have been to the
animal and why they should have been preserved in
this phylum while they have degenerated in the
others. But the fact of their persistence proves that
they were used in some manner, at least in the males.
The canines have degenerated; they are smaller than
those of B. gigas and very much smaller than those of
B. ramosum. The premolars have less distinct internal
cones than in B. ramosum and are seen to correspond
M^
A
B
Figure 481. — Sections and contours of skull of Brontotherium plaiyceras
A, Harvard Mus. (type); B, Am. Mus. 1448 (neotype). One-ninth natural size. In tliis species the very long
horns have a much flattened section and are placed far in front of the orbits. The free portion of the nasals is
almost vestigial, and the zygomatic expansions are very massive.
rather with those of B. gigas. The cingula on the
posterior molars are obsolete.
STcuU. — As we have seen in the type the elevation
of the connecting crest between the horns is so great
that it now connects them at the base, forming a
broad, continuous crest (PI. CLXXXVII) subcon-
cave posteriorly, with sharp edges.
The sections of the horns in the American Museum
skull (No. 1448, fig. 481, B) were taken just above
the crest. They are extremely broadened, with the
anterior faces slightly convex and the posterior faces
plane or slightly concave. The posterior concavity
of the horns, observed in the Harvard type specimen
(fig. 481), is here rendered plane by crushing. This
plane section is carried nearly to the summits, which
do not expand as in B. ramosum. A comparison of
the sagittal sections with those of the previously
described species also illustrates the extreme thin-
ning, anteroposteriorly, and the vertical face of the
connecting crest. The fact that this crest and the
horns incline forward is probably
due to vertical crushing.
The nasals are reduced to 20
millimeters, one-half the length
observed in B. ramosum and B.
curtum. The entire outer edges of
the horns are now composed of
the sharp external ridges, which
reach from the summit to the ante-
rior superior rim of the orbit, where
they are grooved horizontally by
the lacrimal ducts. The comple-
tion of this sharp outer face is the
fulfilment of a growth tendency
which we have seen in various de-
grees of development from 2 to
many inches in length in preced-
ing stages.
The entire anterior portion of
the neotype skull is distorted by
pressure; the posterior portions are
entirely wanting and have been re-
stored in plaster. The external au-
ditory meatus was entirely closed
by a deep union of the postglenoid
and post-tympanic processes. The
enormous zygomata were fortu-
nately preserved, especially that
upon the right side, which is accu-
rately represented in the section.
The mass of this buccal process
is as great as in B. curtum and
apparently exceeds that of B.
ramosum. This section and the
less progressive structure of the
premolars furnish two reasons for regarding B. plaiy-
ceras as derived from the B. gigas datum and B.
curtum types rather than from the B. ramosum type.
According to this view B. ramosum would represent a
contemporary or collateral species rather than one in
the direct line of ascent.
Additional observations on the measurements of Bron-
totherium curtum and B. plaiyceras. — The five male
skulls referred to B. curium exhibit the following range
in measurements as compared with B. plaiyceras:
EVOLUTION OF THE SKULL AND DENTITION OF OLIGOCENE TITANOTHEEES
581
Measurements of Brontotherium curtum and B. platyceras, in
millimeters
P'-m^
Pi-p*
M'-m^
Pmx to condyles-
Zygomatic index -
Nasal length
Horn length
B. curium B. platyceras
345-350
128-130
218-228
780-840
74-?78
52-65
355-380
337-340
120-123
221-223
728-880
80
20-38
390
Thus B. curtum is a little larger than B. platyceras
in the dental measurements but has longer nasals and
shorter horns.
In B. platyceras the type consists only of a pair of
horns and hence affords no dental measurements.
The first referred skull (Am. Mus. 1448) is crushed,
so that the length of the horns has probably been in-
creased, the basilar length perhaps lessened. The
true molars are not as large as in certain specimens of
B. gigas, but the ratio of molar length to the basal
length of the skull is at least not less than in other
brontotheres. In basilar length the first specimen of
platyceras is smaller than curtum, the second speci-
men is much larger.
The second referred skull in the Field Museum of
Chicago (No. 12161) is the largest titanothere on
record and is superbly preserved (Pis. CLXXXI,
CLXXXII). Its flattened spreading horns combine
features of B. platyceras and B. ramosum. Its molar
index (25) is lower than in typical brontotheres
(28-29). The premolar series is short.
101959— 29— VOL 1-
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XX
UPPER AND LOWER CANINES OF OLIGOCENE TITANOTHERES
A, Brontotherium gigas, supposed female (Am. Mus. 1006), right superior canine, external view. B, Bro>itothe>-iu7rt Sigas^ male (Am. Mus. 492),
right superior canine, external view. C^, Brontotherium leidyi (Carnegie Mus. 93), right superior canine, external view^; C^ the same,
left upper canine, internal view. D, Brontops robuiiws, male (Am. Mus. 1083), right superior canine, external view. £>, Menodus giganteus,
male (Am. Mus. 505), right superior canine, external view; E^, the same, left superior canine, internal view. F, Allops walcottit type (Nat.
Mus. 4260), right inferior canine, external view. G, Menodus heloceras (Carnegie Mus.), right superior canine, external view. H, Bronto^
therium medium. (Am. Mus. 1051), right inferior canine, external ^/ie.vv. Ii, Brontotherium. leidyi (Am. Mus. 516), right inferior canine,
external view; I^, the same, left inferior canine, internal view. Ji, Menodus trigonoceras (Nat. Mus. 4745), left infe;-ior canine, external
view; J^, the sam.e, right inferior canine, internal view. All natural siz;e
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XXI
A, Brontops hrachycephalu
LEFT UPPER PREMOLARS OF OUGOCENE TITANOTHERES
male (Nat. Mus. 4258). B, AUops walcotti, male, type (Nat. Mus. 4250).
leidyi, type (Nat. Mus. 4248). All natural sise
C, Hrontotherium
^'3
< £
o
a i
U. S. GEOLOGICAL STJHVET
tif}^
dps dp4
MONOGRAPH 65 PLATE XXIII
cd.
As
''^>
Ai
JUVEiNiLb JAW Kbi-bRRED BY MARSH TO BRONTOPS
One-half natural size. (See p. 455.) Important because it
ikeleton. Ai, External view of the left ramus; A;, internal
view; A^, superior view of the left ramus. cd. Condyle; di,
, alveolus for second deciduous incisor; dij, alveolus for third
ning in with deciduous series and probably belonging with it);
Generic and specific refereni
exhibits the chara(5ters of the juvenile ;
view of the right ramus; A3, anterior
alveolus for fir^ deciduous incisor; di2
deciduous incisor; p\^ fir^ premolar (coi
dp2, second deciduous premolar; dps, third deciduous premolar; dpi, fourth deciduous premolar
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XXIV
?^Z
JUVENILE JAWS AND TEETH OF OLIGOCENE TITANOTHERES
A, Menodus giganteus (Am. Mus. 510), superior vie-wr. B, Menodus giganteiis (Am. Mus. 509), superior view. Ci, Brontops dispar Carnegie Mus.
124), external view of right ramus; C2, the same, superior view of dentition. All one'half natural sise
V. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XXV
SUPERIOR DECIDUOUS AND PERMANENT GRINDING TEETH OF MENODUS GIGANTEUS
Ai, Am. Mus. 497, deciduous premolars and firA permanent molars of left side, crown view^; A2, the same, showing permanent teeth
embedded in the alveolar region. Both t^vo'thirds natural si^e
CHAPTER VII
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
SECTION 1. METHODS BY WHICH THE TITANOTHERE
SKELETON HAS BEEN STUDIED
PRINCIPIES OF THE EVOLUTION OF THE IIMBS OF
HOOFED ANIMALS
The postcranial skeleton of the titanotheres, al-
though less fully known than the skull, is no less
significant in its bearing on our knowledge of the
evolution of these animals and of that ancient West
which was their habitation or the scene of their migra-
tions. To understand that ancient West we must try
to bring its animals and plants back to life. The
attempt to restore the titanotheres as living and
migrating animals has led to the establishment of four
new principles in the evolution of the limbs of the
hoofed mammals generally, principles that have been
worked out by the author in cooperation with Dr.
William K. Gregory, who has published (Gregory,
1912.1, pp. 267-294) a preliminary study which
includes many independent observations of his own
on the mechanics and adaptations of limb movement.
These principles were discovered through comparison
of the skeleton and musculature of all the perissodac-
tyls and of the proportions of the upper and lower
limb segments in a large number of ungulates, in-
cluding perissodactyls, artiodactyls, amblypods, and
proboscideans. These four principles of limb evolu-
tion, which had been only partly or incompletely
recognized previously and which are fully described
in Chapter IX, are briefly as follows:
1. The relative length and the angulation of the
upper and lower segments of the limbs and the planes
of the articular facets furnish a means of elucidating
the adaptations to speed and to weight in all the
hoofed mammals, living and extinct. Thus by de-
termining the relative lengths and proportions of the
limb segments among living forms in which the
speed, weight, and general limb movements are loiown,
we may estimate the adaptations to similar functions
and habits in the titanotheres and other extinct forms.
2. Apart from their ancestral paleotelic adapta-
tions, all ungulates, in their bony and muscular
systems, show secondary cenotelic adaptations to
similar mechanics of speed and weight, which form
closely analogous or convergent groups and are exhib-
ited in the form and the proportions of the limbs and
of the shoulder and pelvic girdles.
3. Within each of the nine families of perissodactyls
that are more or less closely related to the titanotheres
analogous or convergent adaptation produces closely
similar limb and shoulder-girdle forms from more or
less dissimilar ancestral forms.
4. Between the primitive, light-limbed, subcursorial
Lambdoiherium type and the ponderous Brontotherium
type, the titanotheres pass through four stages of limb
types (figs. 685, 686). From a light body and limb
type {Lambdotherium) , analogous to that of the primi-
tive cursorial horses, they enter a medium limb stage
(Eotitanops) like that of the tapirs, then pass through
something near a primitive rhinoceros stage {Mesa-
tirJiinus), and ultimately attain the final titanothere
stage {Brontotherium), which is in some respects
similar to that seen in the elephants.
The titanotheres and other hoofed mammals that
exhibit these four stages in the development of the
limbs are broadly designated as follows:
1. Subcursorial digitigrades, partly perfected in
swift limb movements, including PJienacodus (condy-
larth), EoMppus (horse) , Lamhdotherium (titanothere),
primitive types of ungulates of lower Eocene time, in
which limb proportions are inherited from ancestral
unguiculates and show evidence of remote ambulatory
and even of still more remote arboreal adaptation
(Matthew, Gregory). In these animals the radius
and tibia are relatively long; the metapodials, typified
by Mts III and Mtc III, are relatively short.
2. Mediportal digitigrades, of medium weight and
speed, with moderately heavy body and limbs and
clumsy motion (digitigrade), such as Tapirus (tapir),
MesatirMnus (titanothere). Most middle-sized quad-
rupeds of middle Eocene time have limb proportions
intermediate between the cursorial and graviportal
extremes. These proportion ratios survive in the
existing Tapirus, which, however, in its body and
limb proportions is more cursorial than the Eocene
titanotheres.
3. Subgraviportal digitigrades, partly transformed
into or prophetic of the weight-bearing (graviportal)
type, such as Palaeosyops (titanothere).
4. Graviportal digitigrades, of perfected weight-
bearing type, with angulate limbs capable of more or
less rapid movement according to the length and the
angulation of the limbs, such as Opsiceros (black
rhinoceros), Brontotherium (titanothere). These ani-
mals are all digitigrades — that is, the feet rest partly
on terminal hoofs, partly on pads beneath the phal-
anges. They are heavy or bulky forms, such as the
rhinoceroses and titanotheres, which retain the
digitigrade type of foot, although some (Menodus)
have straight hind limbs.
Besides the four types of limbs or limb movement
that are exemplified in the titanotheres there are three
other types, which are not yet known in animals of
the titanothere group, as follows:
583
584
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
1. Primitive ambulatory (plantigrade), such as
Pantolamhda (amblypod) ; also the opposite extreme.
2. Cursorial unguligrade, with swift type of limb
and secondarily elongated feet, such as NeoJiipparion,
Equus (horse), Gazella (gazelle), all light, long-footed
Figure 482. — Evolution of the skeleton in titanotheres
A, First stage (subcursorial): EotitaTiops borealis^ lower Eocene. B, Middle stage
(mediportal): Palaeosyops Uidyi; middle Eocene. C, Final stage (graviportal);
iJroniops ro6«siw5; lower Oligocene. Scales approximate; A too large. CSeefig.483.
types, in which the limbs terminate in single hoofs and
the pads beneath the phalanges are reduced or wanting.
3. Rectigrade-graviportal, straight-limbed, weight-
bearing type, with special pillar-like or vertical
weight-bearing disposition of the limbs and short
gravigrade feet, such as ElejpJias, Mastodon (elephants) ;
also such clumsy, slow-moving forms as Coryphodon
and Uintatherium (amblypods), with gigantic or heavy
bodies supported on straight or column-like limbs in
which the terminal phalanges and hoofs are reduced
and the limb is supported on a heavy pad.
In general, the stages
represented by these
types indicate that two
main divergent direc-
tions were taken in
limb adaptation —
namely, into cursorial
or speed types or into
graviportal or weight
types. Bridging over these extremes are types
that combine speed and weight.
In the titanotheres there is an evolution of
constantly changing proportion both in the mass
or weight of the muscles and bones and in the
length of the limb segments, adapted to con-
stantly changing habit of speed and weight,
passing through the subcursorial, mediportal,
subgraviportal, and graviportal types and ap-
proaching a rectigrade-graviportal type in the
huid limbs. Thus, while the bones of the head
are constantly changing, every muscle and bone in
the skeleton is also constantly changing.
SIZE AND PROPORTIONS OF EOCENE TITANOTHERES
As shown in Figures 483 and 661, the members of
the six or seven phyla of middle Eocene titanotheres
differ much less in the height of the shoulder and of
the hip than in weight and speed.
The height of a quadruped depends upon the total
combined length of the segments — the scapula, hu-
merus, radius, manus — modified, by the flexure or
angidation at the four joints — shoulder, elbow, wrist,
and phalangeal joints.
Quadrupeds differ widely in respect to angulation :
heavy animals have straighter limbs than light,
swift-moving animals. The normal angles at the
shoulder, the elbow, and the phalangeal joints may
be determined precisely by a study of the planes of
the articular facets (figs. 518, 664, 666, 667).
Naturally the best method of ascertaining the height
of the animal is to make a restoration (Palaeosyops,
fig. 536), if a sufficient number of parts are available,
laying out the limbs in their proper angulation, as
indicated by the planes of the articular facets and
measuring the net height directly.
It is seldom that the material is so complete that
all the limb segments and articular facets can be
measured and determined for the purpose of estimat-
ing height, so that we may be forced to make estimates
based on the ratios of the length of the parts preserved
to those of similar animals of loiown height.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
585
Broniups
.^rontops
robzistzics'i
¥
maepsyops-^yogruiffmsYi-''^'^^
EoUtanops
KesatirMnus Manteoceras -
Jrianteoceras
?
SoHtanops
— princeps
A
E
F
H
J
K
=10-m..
iS.Ocm. 66c7n. 80.7cm. 89c7rv. 90cm,. 92cTrv. 201cm. I03cnv. JOicm. JBOcnv. 250.2om,
'ISin,. ^2ft.2m,. =Zfb.7\m. =2fi.Min. '2ft.ll^m. =3ft. ^3ft.3%m. ^Sft-Hiin. =3ft.5in. =6ft.3m.^8ft.2\iiu
FiGTJRE 483. — Estimated height at the shoulder (to top of scapula) of Eocene and Oligocene titano-
theres (see below), compared with that of the tapir {Tapirus indicus)
A, Lamhdotherium popoagkum: B, Eoiitanops gTegoryi; C, Eotitanops pTinceps: D, Eotitanops major; E, Mesatirhinus petersoni; F, Manteoceras
manteoceras (female); Q, Tapirus indicus; H, Palaeosyops leidyi; I, DolichorUinus hyognathus; J, Manteoceras manteoceras (male); K,
Brontops rohustus? (female, Am. Mus. 518); L, Brontops robustus (type, male). One twentieth natural size.
Estimated Jieights of Eocene titanotheres
Length of
humerus
(centimeters)
Height to top of scapula
Designation
in Figure 483
Lower Eocene:-
Lambdotherium popoagicum
Eotitanops gregoryi
Eotitanops princeps
Eotitanops major
Middle Eocene:
Mesatirhinus petersoni
Manteoceras manteoceras, ?
Tapirus indicus (modern tapir) .
Limnohyops monoconus?
Palaeosyops leidyi
Manteoceras manteoceras, cf
Upper Eocene:
Dolichorhinus hyognathus
Lower Oligocene:
Brontops robustus
Brontops robustus '
20.3
26+
'^29
25
29.3
32.5
35
53
6L 5
32
35.6
38
39.3
38.6
78
82
-38
45.6
''66
80.7
"^90
92
•92
101
/104
"103
M90
250.2
15
A
18
B
26
C
31.8
D
35
E
35K
F
36
G
36
39M
H
41
J
40J^
75
983^
« Scapula and humerus conjectural.
t Humerus known; scapula and radius conjectural. Restored from known relative length of humerus, scapula, and radius in allied forms.
<: Forearm and manus Princeton Mus. 10013, humerus and scapula supplied from other individuals.
'* Femur known; other segments unknown. Height computed by assuming same relative length of hmb segments as in Palaeosyops leidyi.
' Manus, forearm, and humerus known, scapula computed.
f Femur known, other segments unknown. Height computed by assuming same relative length of limb segment as in Palaeosyops leidyi.
0 Composite restoration (fig. 579).
'' Measured from mounted skeleton (Am. Mus. 518).
• Type skeleton (Yale Mus. 12048).
586
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
DIVERGENCE AND CONVERGENCE IN THE SKELETON OF
POLYPHYIETIC SERIES
DIVERSE ADAPTIVE TYPES OF LIMB STRUCTURE
Our study of the locomotor skeleton of the titano-
theres strongly supports the polyphyletic theory that
we derived from the study of the skull and the teeth —
namely, that there were two larger branches or groups
of Eocene titanotheres, the palaeosyopine and the
Manteoceras-DoUchorhinus, each of which was divided
into two or more smaller branches or subfamilies.
The interpretation of the structure of the feet and
limbs would be simple if the feet and limbs of all these
six or seven genera were invariably divergent, but the
law of convergence again comes into play, and long
after one subfamily has diverged from another there
remain or arise within each subfamily forms which,
through convergent adaptation and inheritance, par-
allel in foot and limb adaptation forms in the other
subfamilies.
To make this double convergence clear we recall the
existence in Eocene time of five distinct groups or sub-
families of titanotheres with respect not only to geo-
logic level but to divergence as to speed and to weight,
namely:
Palaeosyopine group:
Lower Eocene:
Lambdotheriinae (cursorial types).
Eotitanopinae (subcursorial types).
Middle Eocene:
Palaeosyopinae (mediportal and subgraviportal types) .
Manteoceras-DoUchorhinus group :
Middle and upper Eocene:
Manteoceratinae (subgraviportal types).
Dolichorhininae (mediportal types).
The Lambdotheriinae (fig. 486) are primitively
cursorial, like the early Eocene Equidae and Lophio-
dontidae.
How far the Eotitanopinae may be directly ancestral
to the Palaeosyopinae or Manteoceratinae we do not
know.
The known Eotitanopinae (figs. 484, 492) are light-
limbed, or subcursorial — that is, the feet are relatively
shorter (more mediportal) than those of the tapir and
less light in structure than those of the primitive
Equidae.
In the various Palaeosyopinae we find a condition
somewhat transitional between the tapir and the
heavier types, like the rhinoceros. These animals
include more heavy-limbed (Palaeosyops) and light-
limbed (Limnohyops) types.
Similarly within the Manfeoceras-DolicTiorJiinus
group there are two series of forms, mediportal {Mesa-
tirhinus) and subgraviportal (Manteoceras). Thus we
discover mediportal Palaeosyopinae (such as Limno-
hyops) and mediportal Dolichorhininae (such as Mesa-
tirMnus), which are somewhat alike in their limb
adaptations and can be separated only by careful
scrutiny of certain less conspicuous features of ances-
tral separation, which bring out the subfamily char-
acters. There are also graviportal Palaeosyopinae
(such as Palaeosyops), which resemble graviportal
manteoceratines (such as Manteoceras) in some respects
but differ in others.
It has taken a vast amount of study of the scattered
and often unassociated limb materials to determine
the I'eal subfamily relationships in the limbs of these
different forms, which are concealed by the veneer of
similar adaptation or ceno telle resemblance; but it
may now be demonstrated positively that in each
subfamily of the middle Eocene titanotheres there
were relatively light-limbed and relatively heavy-
limbed forms. The differences between these adaptive
extremes of relatively rapid-moving and slow-moving
forms may be observed and measured in every single
bone of the limbs, and especially in the small bones of
the carpus and tarsus.
To illustrate how a single bone may be highly
distinctive, an outline of the central bone of the carpus,
the magnum, as it appears in different forms, is given
herewith (fig. 485). On the left is the magnum of the
subcursorial Eotitanops, relatively high and narrow;
on the right that of the subgraviportal Palaeosyops
rohustus, relatively broad and low. In general, high
and narrow proportions of the magnum characterize
cursorial forms and low and broad proportions char-
acterize graviportal forms. Thus the transformation
of this single bone reflects the kind of allometric change
which prevailed in all parts of the skeleton.
Or, to take the foot as a whole (fig. 484), the manus
of MesatirJiinus represents the mediportal extreme
among the titanotheres analogous to that of Tapirus,
while the manus of Palaeosyops represents the sub-
graviportal extreme analogous to that oi Hippopotamus.
Both types of feet occur at the same geologic levels and
in the same geographic regions; they belong to con-
temporaneous titanotheres, one perhaps seeking forest
ground like the habitat of Tapirus, the other the bor-
ders and possibly the waters of rivers and lakes like
the habitat of Hippopotamus.
There are thus, as shown in Figures 502, 503, 510,
515, 620, wide adaptive radiations among the Eocene
titanotheres in modes of locomotion on different kinds
of soil.
All these adaptively diverse types of limb structure
appear to have been derived from a small and primitive
type, which was a subcursorial, light-limbed, slender-
footed, and relatively speedy animal, well adapted for
escape rather than combat.
The adaptive transformation of the limbs of titano-
theres has not yet been traced so continuously as that
of the skull or of the teeth; the scattering of parts of
limbs has rendered the association and identification
of many separate bones exceptionally difficult. We
are still in doubt as to the limb structure in certain
phyla, especially in Telmatherium. Much remains to
be discovered through further exploration in the field.
EVOLUTION OP THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHEEES
587
This deficiency of material renders all the more
valuable the broader investigation and comparison
with other Perissodactyla which will be presented in
Chapter X. The new and precise system of compara-
tive measurements of the limbs establishes certain
subphyla (genera). Members of these six or more
phyla were separated by the weight, shape, and pro-
portions of the body, feet, and limbs, such separation
being indicative of a considerable range of size, speed,
and migrating ability in search of food. This demon-
FiGUEE 484. — The phyla of Eocene titanotheres, as represented by the manus
A, Lambdotheriinae (,Lambdotherium); B, Eotitanopinae (Eotitanops); C, Palaeosyopinae (Limnohyops); D, Palaeosyopinae iPalaeosyops);
E, Manteoceratinae (.Manteoceras)] F, T>o\iehorhininae (Mesatirhinu^). One-third natural size.
laws of limb adaptation and enables us to estimate
fairly closely the speed and weight-bearing powers of
the various kinds of titanotheres.
1. Existence of many pliyla. — The chief conclusions
drawn from the limb bones of titanotheres, moreover,
confirm those drawn from the skull — namely, the exist-
ence of six or more distinct phyla in the lower Oligo-
cene. That there were six or more contemporaneous
but more or less independent phyla, characterized by
differences of limb structure and modes of locomotion,
is clearly shown, as well as the fact that these include
two main phyla each of which embraces at least three
strates the existence of two early radiations (sub-
families) and of subsequent branch radiations (genera).
FiQUKE 485. — Progressive broadening of the magnum
in Eocene titanotheres
A, Eotitanops (subcursorial) ; B, Mesaiirkinus (mediportal) ; C, Lim-
nohyops (mediportal); D, Palaeosyops (subgraviportal) . One-balf
natural size.
2. Adaptive characters superposed on ancestral. — •
Similar limb and foot proportions, like similar skull
588
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
proportions, were independently developed in unre-
lated phyla and are often misleading as to real rela-
tionships. Thus the subfamilies duplicate each other
in adaptation; there arose in both subfamilies short-
footed and long-footed forms, as summarized in vari-
ous tables.
TERMS USED IN DESCRIBING THE SKELETON OF THE
TITANOTHERES
The following list includes most of the terms used in
this monograph in describing the postcranial skeleton
of the titanotheres:
Terms used in describing the 'postcranial sTceleton of the titanotheres
English terms
Shoulder girdle and fore limb
Scapula
Superior or suprascapular border
Anterior or prescapular border
Posterior, postscapular, axillary border
Spine
Tuberosity of spine
Supraspinous or prescapular fossa
Infraspinous or postscapular fossa
Internal subscapular surface
Neck of scapula
Coracoid process
Coracoid border
Glenoid cavity
Humerus
Head
Greater tuberosity
Lesser tuberosity
Bicipital groove
Deltopectoral crest
Deltoid tuberosity
Tuberosity for the teres major or latissimus dorsi-
Winding (brachialis) surface
Ectocondyle
Extensor (supinator) crest
Entocondyle
Distal articular surface
"Capitellum" for radius
"Trochlea" for ulna
Olecranal, anconeal (posterior) fossa
Supratrochlear (anterior) fossa
Radius
Head
Internal process
External process
Radiohumeral or trochlear facet
Biceps tubercle
Extensor groove
Brachialis rugosity
Styloid process
Radiocarpal facet
Ulna
Olecranon
Olecranal fossa
Ulnohumeral trochlea
External facet of the ulnohumeral trochlea
Coronoid process
External process
Lesser sigmoid fossa
Ulnoradial facets
Ulnocarpal facet
Styloid process
Carpals
Proximal row, scaphoid, lunar, cuneiform, pisiform
Posterior tuberosity of lunar
Scapula
Margo suprascapularis
Margo prescapularis
Margo axillaris
Spina scapulae
Tuberositas spinae scapulae
Fossa prescapularis (supraspinata)
Fossa postscapularis (inf raspinata)
Facies subscapularis
Collum scapulae
Processus coracoideus
Margo coracoideus
Ca vitas glenoidalis
Humerus
Caput humeri
Tuberculum maj us
Tuberculum minus
Sulcus bicipitalis
Crista deltopectoraUs
Tuberositas deltoidea
Tuberositas teres
Facies spiralis
Ectepicondylus
Crista " supinatoria " (extensoria)
Entepicondylus
Facies distalis
capitellum
trochlea humeri
Fossa olecrani
Fossa supratrochlearis
Radius
Caput radii
Processus internus capitis radii
Processus externus capitis radii
Facies proximalis (radiohumeralis) radii
Tuberositas bicipitalis radii
Sulcus extensoria
Rugositas brachialis
Processus styloideus radii
Facies distalis (radiocarpalis) radii
Ulna
Olecranon
Fossa olecrani
Incisura semilunaris (facies ulnohumeralis)
Facies externa incisurae semilunaris
Processus coronoideus (anconaeus) ulnae
Processus externus ulnae
Fossa sigmoidea minus
Facies ulnoradialis
Facies distaUs (ulnocarpalis)
Processus styloideus ulnae
Carpalia
Os scaphoideum, os lunare, os cuneiforme, os
pisiforme.
Tuberositas posterior ossis lunaris
cb.
t. sp.
fas. sup. sp.
fos. inf. sp.
glen,
h.
cap.
t. maj.
t. min.
s. bic.
t. del.
t. te.
brack, ant.
ectep.
cr. "sup."
eniep.
cptl.
trch.
fos. ol.
pr. in.
pr. ex.
cptl. h.
tbclm. rad.
s. ex.
br. ant.
pr. sty.
ol.
fos. ol.
inc. sml.
fac. ex.
pr. anc.
pr. ex. u.
fos. sig. min.
pr. sty. ul.
sc, lu., cu.,
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
Terms used in describing the postcranial skeleton of the titanotheres — Continued
589
English terms
Shoulder girdle and fore limb — Continued
Carpals — Continued.
Distal row, trapezium, trapezoid, magnum, unci-
form.
Posterior tuberosity of magnum
Metacarpals II-V
Phalanges, 1st, 2d, 3d, or ungual
Sesamoids
Sacrum, pelvic girdle, hind limb
Sacrum
Sacrals 1-4
Sacro-iliac rugosity
Pelvis
Right and left innominate bones
Ilium
Superior or supra-iliac border
Sacral process
Lateral process
Iliosacral rugosity
Dorsum or dorsal surface (gluteal fossa) _
Ventral surface (iliac fossa)
Neck of ilium
Rectus tuberosity
Pubis
Symphysis pubis
Pubi-ischiadic symphysis
Pectineal tuberosity, pubic spine
Obturator (thyroid) fenestra
Ischium
Obturator (thyroid) fenestra
Tuberosity
Ischial spine
Femur
Head
Ligamentum teres pit
Digital fossa
Great trochanter
Second or lesser trochanter
Third trochanter
Linea aspera
Plantaris fossa
Internal tuberosity
External tuberosity
Entocondyle, inner femorotibial trochlea.
Ectocondyle, outer femorotibial trochlea-
Patellar trochlea, surface
Inner keel, outer keel
Patella
Tuberosity for the quadriceps femoris
Tuberosity for the patellar ligament
Tibia
Tibiofemoral trochlea
Entocondylar surface
Ectocondylar surface
Spine
Cnemial crest, tubercle
Supero-external tuberosity
Supero-internal tuberosity
Popliteus fossa
Distal posterior process
Tibiocalcaneal trochlea
Internal malleolus
Carpalia — Continued.
Os trapezium, os trapezoideum, os magnum, os
unciforme.
Tuberositas posterior ossis magni
Metacarpalia II-V
Phalanges
Ossa sesamoidea
Sacrum
Vertebrae sacrales 1-4
Rugositas iliosacralis
Pelvis
Os innominatum, dextrum et sinistrum
Ilium
Margo superior
Tuber sacrale
Tuber coxae
Rugositas iliosacralis
Facies dorsalis
Facies ventralis
CoUum ilii
Processus antero-inferior
Pubis
Ramus superior
Symphysis pubi-ischiadica
Tuberositas pectinea
Fenestra thyroidea
Ischium
Fenestra thyroidea
Tuberositas ischii
Spina ischii
Femur
Caput femoris
Fovea capitis
Fossa trochanterica (digitalis)
Trochanter major
Trochanter minor
Trochanter tertius
Linea aspera
Fossa plantaris
Tuberositas interna
Tuberositas externa
Entocondylus
Ectocondylus
Trochlea patellaris
Carina interna, carina externa
Patella
Tuberositas tibiae
Tuberositas ligamenti patellae
Tibia
Facies proximalis
Facies entocondylaris
Facies ectocondylaris
Spina
Crista cnemialis (tuberculum)
Tuberositas supero-externa
Tuberositas supero-interna
Fossa poplitealis
Processus distalis posterior
Facies articularis inferior
Malleolus internus
tz., id., mg., unc.
II, III, IV, V.
tu. sac.
tu. cox.
pr. ant. inf.
pu.
ram. sup. pb.
f. thyr.
tu. is.
sp. is.
/.
cap.
lig. ter.
tr.\ t.'
tr.'^, t."
tr.^, t."
t. in.
t. ex.
con. in.
con. ex.
trch.
cr. en.
t. ex.
t. in.
mal. int.
590
TITANOTHEBES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Terms used in describing the postcranial sTceleton oj the titanotheres — Continued
Sacrum, pelvic girdle, hind limb — Continued
Fibula
H ead
Fibulotibial facet
Shaft
Fibulocalcaneal facet
External malleolus
Calcaneum
Tuber calcis
Sustentaculum
Sustentacular facet
Ectal facet
Inferior facet
Cuboid facet
Astragalus
Astragalotibial trochlea
H ead
Neck
Navicular facet
Cuboidal facet
Tarsals
Navicular
Entocuneiform
Mesocuneiform
Ectocuneiform
Cuboid
Metatarsals II-IV
Phalanges
Fibula
Caput fibulae
Facies articularis superior
Corpus fibulae
Facies articularis inferior-
Malleolus externus
Calcaneum
Tuber calcis
Sustentaculum tali
Facies sustentacularis
Facies ectalis
Facies inferior
Facies cuboidea
Astragalus
Trochlea astragali
Caput astragali
Collum astragali
Facies navicularis
Facies cuboidalis
Tarsalia
Os naviculare
Os entocuneiforme
Os mesocuneiforme
Os ectocuneiforme
Os cuboideum
Metatarsalia II-IV
Phalanges
fb.
mal. exi.
cal.
tu. cal.
sus.
ect.
inf.
ch.
as.
cr. ex., ci
(nav.)
icb.)
ch.
ml. II-IV.
ph.
SECTION 2. THE POSTCRANIAL SKELETON OF
LOWER EOCENE TITANOTHERES
SUBFAMILY LAMBDOTHEEIINAE ,
Lambdotherium
Lambdotherium is a small, swift, slender-limbed,
relatively abundant animal, less abundant than the
contemporary horses but more abundant than the
small lophiodonts (Heptodon). The largest Lamb-
dotherium measures about 14 inches (350 mm.) at the
shoulders, as compared with the smallest Eotitanops
(E. gregoryi), which measures about ISJ^ inches (456
mm.) at the shoulders. It exceeds in size most of the
contemporary species of Eohippus and equals the large
EoMppus robusius of the lower Wasatch.
Lambdotherium popoagicum
The materials of L. popoagicum are extremely
scanty; in only two specimens (Am. Mus. 4880, 14903)
are parts of the skeleton associated with the teeth,
and these parts are very fragmentary. (See figs. 486-
489.)
In the atlas (fig. 487) the vertebrarterial canal
pierces the anterior portion of the base of the trans-
verse process; that is the primitive condition in tita-
notheres and m Perissodactyla generally.
The restoration of the fore limb (fig. 488) is highly
conjectural, because the lengths of the bones are un-
known. The proportions are heavier than in Eohippus.
The ratio of the radius to the humerus is estimated
at 90 per cent, the same as in Eohippus.
The scapula (fig. 487) presents an elongate neck
(tr. 19 mm.) and the lower part of the postspinous fossa
is very narrow; the lower part of the prespinous fossa,
which is not all preserved, is slightly broader; the
spine descends rather low. The distal end of the
humerus indicates the presence of a small entocondyle,
and its narrow rotula accords with the deep, laterally
compressed proximal articular surfaces of the ulna
and radius. The length of the radius is estimated at
103 millimeters; the shaft is rather narrow and sharply
convex. The ulna is elongate and has a high, deep,
laterally compressed olecranon process; the proximal
portion of the shaft behind the radius is deep.
The carpus (fig. 489) is correlated with this slender
structure, being relatively high and narrow but
broader than in Eohippus; the scaphoid rests
chiefly on the magnum anteriorly, and the high and
narrow lunar exhibits a broad supporting unciform
facet and an oblique nonsupporting magnum facet.
The magnum, wanting in this specimen, was probably
small; it is relatively broad in Eohippus. The cimei-
form is rather high and narrow.
The manus exhibits four digits; the ends of the
metapodials show a tendency to functional tridac-
1 tylism, because the median digit (III) is larger than
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENB TITANOTHERES
591
the adjoining digits. This enlargement of D. Ill
is indicated not only by its diameter but by the
enlargement of its proximal phalanx, which is much
broader than the proximal phalanges of D. 11 or D.
IV. This indicates a rather pronounced tendency
to mesaxonic structure, whereas the middle Eocene
titanotheres exhibit either a primary or secondry
paraxonic and finally isotetradactyl condition. D.
V is somewhat shorter and more slender.
The structure of the astragalus of this specimen
(Am. Mus. 4880), as inferred from the distal end of
the tibia which is preserved, indicates that the keels
of the astragalus were quite sharp.
The tarsus is compared with that of the less cur-
sorial Eotitanops and the more cursorial Eohippus in
Figure 502. An astragalus associated with Am. Mus.
2991 exhibits a high, long neck and a high, narrow
sustentacular facet, showing that the astragalo-
cuboidal facet was very narrow (figs. 491, 503).
Figure 486. — Reconstructed skeleton and restoration
of Lambdotherium popoagicum Cope
Made by E. S. Christman under the direction of W. K. Gregory.
One-tenth natural size. This provisional reconstruction is based on
the following American Museum specimens from the Wind River
Basin, Wyo.: 14007 (Alkali Creek, Wolton), maxilla, malar, top of
cranium; 14903 (Alkali Creek, Buck Spring), squamosal, basicranial
region; 14899 (Alkali Creek, Buck Spring), lower jaw; 4880, distal end
of scapula, fragments of humerus, radius, ulna, and manus; 14921
(Wolton), astragalus and calcaneum. Hemaining parts conjecturally
modified from Eohippus and other contemporary perissodactyls.
SUBFAMILY EOTITANOPINAE
Eotitanops
Our knowledge of the skeleton of the species of
Wind River Eotitanops is relatively small; we know
that there were gradations of size from smaller and
101959— 29— VOL 1 41
fac.ar±.post
lighter to larger and heavier, subcursorial forms, the
general evolution being in the direction of the laiown
middle Eocene titanotheres. The pes especially gives
us valuable records of this progressive increase in
size and weight.
As compared with the middle Eocene titanotheres
of the Bridger formation these lower Eocene forms
were small and light limbed
and had high, narrow ankle
and wrist joints, resembling B
most nearly the light-footed -'
members of middle Eocene age,
such as Mesatirhinus . In the
skeleton, as in the skull, there
are general primitive features
that accord with the other
lower Eocene perissodactyls,
such as the lophiodonts (Hep-
todon) and horses (EoMppus).
The details of the carpus and
tarsus remind us also of these
lower Eocene perissodactyls,
although Eotitanops is decidedly
larger and less distinctively
cursorial than either Heptodon
or Eohippus. The abbrevia-
tion of the fifth or outer digit
(Mtc V) of the manus forbids
our describing the animal as
functionally tetradactyl; taken with the extreme dis-
placement of the lunar on the unciform (a ti-idactyl
specialization) this reduction tends to support the
idea of Gregory that these animals, after having
entered an evolution toward tridactylism, reversed the
process and reentered tetradactylism in middle Eocene
time.
Measurements, in millimeters, of teeth and limb bones referred to
Eotitanops, compared with Lambdotherium and Tapirus
Figure 487. — Atlas and
scapula of Lambdothe-
rium, popoagicum
\m. Mus. 4880, Wind Eiver
Basin, Wyo., Wind Eiver for-
mation. A, Distal portion of
left scapula, outer side view; B,
rear view of incomplete atlas.
Two-thirds natural size.
Lower grinding teeth (p^-ms).
Median metatarsal, length
Median metacarpal, length
Humerus, length
Femur, length
64
78. 4
98
85
Lower grinding teeth (pj-ma)-
Median metatarsal, length
Median metacarpal, length
Humerus, length
Femur, length
E. princeps E. major
105
87
205
■250
118
114
113
° Estimated.
592
TITANOTHEKES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
tb.rn/j
Figure 489. — Forearm and man us of Lambdotherium popoagicum
Am. Mus. 4880, Wind River Basin, Wyo., Wind River formation, level B. Incom-
plete left forearm and manus. The length of the radius and of the metacarpals
is conjectural. Ai, Front view; Aj, outer side view of forearm. One-half natural
size.
Figure 488. — Fore iimb of Lambdotherium popoagicum
Am. Mus. 4880, Wind River Basin, Wyo., Wind River formation, level B. Incom-
plete forearm, external view. The length of the elements is conjectural. One-
half natural size.
Figure 490. — Left manus of Lambdotherium
Eotitanops
A, Lambdotherium (Am. Mus. 4880); B, Eotitanops (Am. Mus, 296), showing the
ancestral type of titanothere. The length of the metacarpals in Lambdotherium
is conjectural. One-half natural size.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
593
The association of limb bones with teeth of the
same individual is a very rare occurrence. Thus the
reference (see table on p. 585) of several of the hind
feet is inferred only from the size and degree of slender-
ness of the limbs.
Eotitanops gregoryi Osborn
There is reason to believe that Eotitanops gregoryi
was both small and slenderly proportioned; that E.
Figure 491. — Astragalus of
Lambdotherium popoagicum
Am. Mus. 2991, Wind River Basin, V^yo.,
Wind River formation, level B. Left
astragalus, front and rear views. Two-
thirds natural size.
horealis and E. princeps were still of light construc-
tion, corresponding with the delicately constructed
skull of E. horealis ; and that the littl e-known E. major
was somewhat more massive in its proportions. The
whole range of adaptive radiation of the limbs is
parts may be conveniently described under a single
heading.
Atlas of Eotitanops horealis
The atlas of E. horealis is preserved in the neotype
skeleton (Am. Mus. 14887, fig. 495). In that speci-
men, which is the earliest known, we observe a broad
resemblance to the atlas of the middle Eocene Mesa-
tirhinus and of the existing Tapirus, both probably
representing the primitive perissodactyl type. The
characters are (1) centrum relatively elongate, (2)
facets for axis in obliquely convergent planes, (3)
vertebrarterial canal entering slightly above posterior
rim of transverse process and issuing on lower median
face of same. Comparison with the atlas of Mesa-
tirhinus (fig. 560) indicates that we have here the
ancestral titanothere type, which is highly modified
in some of the progressive forms.
Cervical and dorsal vertebrae of Eotitanops princeps (type)
The two posterior cervicals (C. 6, C. 7), three scat-
tered dorsals, and one caudal of the type (Am. Mus.
296, fig. 496) show the following characters: (1)
Neck relatively short as in Palaeosyops; C. 1 to C. 7
estimated at 180 millimeters as compared with 320,
Figure 492. — Restorations of the lower Eocene titanotheres of the Wind River formation
A, Lambdotherium popoagicum: B, Eotitanops princeps; C, Eotitanops gregoryi. One-thirtieth natural size.
subcursorial or less truly cursorial than either Lamh-
dotherium, Heptodon, or Eohippus. Estimates of the
heights of these animals, which, it will be recalled,
were not successive but partly contemporaneous, are
as follows:
Height of Lambdotherium and Eotitanops
the estimated basal length of the skull; (2) cervicals
with broad depressed centra, facets oblique (length
Type
Estimated
height at
shoulder
.i2
= -2
J3
Smallest and most cursorial
350
450
660
800
14
Second smallest true titanothere
known (Eotitanops gregoryi)
Of intermediate size (E. princeps).
Of largest size (E. major)
18
26
32
ABC D
Figure 493. — Metatarsals of Eotitanops
Median metatarsals of E. gregoryi (A), E. iorealis (B), E. princeps (C), E. major (D). Natural size.
Eotitanops borealis and E. princeps
The species and the mutations of E. horealis and
E. princeps are so closely related that their skeletal
of centra, C. 6, 25 mm.; C. 7, 26); (3) three scattered
dorsals laterally compressed, centra elongate, meas-
uring, ?D. 3, 26 millimeters; ?D. 4, 25; ?D. 10, 29.
594
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
nd lunar of type of Eotitanops borealis | Fore feet of Eotitanops princeps (type)
The radius (Am. Mus. 4892) associated with the
type upper molars is incomplete; it exhibits an
asymmetrical proximal articular surface, and a
prominent internal distal projection (styloid process)
of the distal articular surface.
This type (Am. Mus. 296), originally described as
E. iorealis and now referred to the mutation E.
princeps, is the only individual in which portions of
the fore and hind limbs, vertebrae, and jaws are
known, enabling us to establish the proportions.
Figure 494. — Reconstructed skeleton and restoration of Eotitanops borealis
Made by E. S. Christman under direction of W. K. Gregory. About one-tenth natural size. The provisionally recon-
structed skeleton is based on the following American Museum specimens from the Wind River Basin, Wye: No. 14887
(E. borealis, Dry Muddy Creek 12 miles above mouth), skull, pelvis; No. 296 (.E. princeps),, lower jaw, humerus, manus,
femur; No. 14888 (£. borealis, Alkali Greek, Davis ranch), part of scapula, pes. Remaining parts conjectural.
The lunar is a very distinctive element in the lower
Eocene titanotheres, especially in the fact that in the
front view it rests mainly on the unciform and retains
a narrow vertical facet anteriorly for the magnum,
although in the back view it rests almost equally upon
the magnum and the unciform. This is an adapta-
tion to tridactylism or functional reduction of digits
I and V.
The manus is numerically tetradactyl, but a strik-
ing feature is the slenderness of the fifth digit (Mtc V).
The carpus is narrow and deep — transverse measure-
ment 41 millimeters, vertical 34 (through cuneiform
and unciform). The lunar presents anteriorly a
superior transverse measurement of 17 millimeters as
compared with a total vertical measurement of 22; it
presents inferiorly a broad, oblique unciform facet
EVOLUTION OF THE SKELETON OP EOCENE AND OLIGOCENE TITANOTHERES
595
and a narrow, nearly vertical magnum facet, similar
to but relatively smaller than that seen in Mesatirhinus
megarJiinus; posteriorly it rests half on the magnum, half
on the unciform. As this'bone, therefore, practically
distributes three-fourths of
its front weight to the
unciform, it exhibits an ex-
treme displacement which
appears to be a character
common to the early Eocene
Perissodactyla generally.
Similarly the scaphoid is
narrow (13 mm.) and deep
anteroposteriorly (24 mm.)
with a vertical measure-
ment of 19 millimeters; it
covers the entire superior
face of the magnum. The
magnum is a very distinc-
tive bone as seen in front,
being small, vertically ex-
tended (12 mm.) and later-
ally compressed (9 mm.).
The imciform is a vertically
deep element measuring 22
by 19 millimeters trans-
versely. The metacarpal
displacement is also extreme,
Mtc II and Mtc III abut-
ting widely against the
magnum and unciform respectively. The metacar-
pals are distinguished by the small size of Mtc V,
Figure 495. — Atlas of
Eotitanops borealis
Am. Mus. 14887 (neotype atlas asse-
dated with skull; cf. flg. 250). .^i.
Anterior or condylar surface; A2,
posterior or cervical surface; A3, dor
^sal; Ai, ventral; As, side view. OnC'
third natural size.
by 7 anteroposterior, being, therefore, relatively large-
Mtc III is still larger; it measures 85 millimeters
vertically and 13 transversely. Mtc IV measures
73 millimeters vertically and 11 transversely. Thus
Figure 497. — Radius of
Eotitanops borealis
Am. Mus. 4892, fragments of the right
radius associated with the type upper
teeth. Ai, Distal part; A2, proximal
part; A3, proximal view. One-third
natural size.
Figure 498. — Lunars of
Eotitanops
Left lunar of JE. borealis (.\m. Mus.
4892, type; Ai, front view, A2, top
view) and E. princeps (Am. Mus.
296, type; Bi, front view, B2, top
view).
Figure 496. — Vertebrae of Eotitanops princeps
Cervical (Ce, C7) and dorsal (D) vertebrae associated with the type lower jaw
Mus. 296). One-half natural size.
in which the shaft measures only 6 millimeters antero-
posteriorly by 9 transversely. Mtc II has a length of
79 millimeters and a shaft diameter of 12 transverse
Mtc IV is decidedly smaller than Mtc II. This lack
of symmetry in the digits on either side of the third
digit is characteristic of the early Eocene titanothere
manus; it prepares the way for the secondary parax-
onic condition in which Mtc II, III, IV, V become
more nearly subequal.
Humerus and femur
The two limb bones known, the humerus and femur,
are distinguished by the
lateral compression of the
shafts in harmony with
the compression of the
manus; they indicate
that E. princeps as com-
pared with the middle
Eocene titanotheres
was narrow chested and
slender limbed, that its
proportions most nearly
resemble those of Mesa-
tirhinus of the middle
Eocene, and that the
corresponding elements
are much more primi-
tive.
The humerus (fig. 500)
is relatively elongate, a
km. primitive character; the
great tuberosity is less
developed than in more
recent types. The shaft in the upper third measures
20 millimeters transversely, 35 anteroposteriorly, as
Figure 499. — Manus of
Eotitanops princeps
m. Mus. 296; Wind Eiver Basin,
Wyo.; Wind River formation, level
B. Eight manus of type. One-half
natural size.
596
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
compared with the total length, 205; the proximal
measurement across the tuberosities is 49, while the
measurement of the head to the front of the great
tuberosity is 67; the moderately prominent deltoid
crest extends 88 millimeters below the head, and the
supinator ridge rises 60 millimeters above the interior
border of the radial trochlea.
The femur (fig. 500) has only partly assumed the
distinctive titano there family characters; the shaft has
not yet flattened; it more closely resembles that of
below the trochanters, by the elevated patellar facet
facing anteriorly and inferiorly.
Pelvis of Eotitanops borealis (neotype)
The proportions of the left innominate bone pre-
served in the neotype of Eotitanops horealis (Am. Mus.
14887) indicate a subcursorial type of pelvis, slen-
derly buUt, with the following characters: (1) Neck
of ilium relatively narrow (tr. 29 mm.); (2) tuber-
coxae and external border of ilium partly concave;
(3) pubo-ischiadic fenestra elongate or vertically
i ma/
Figure 500. — Humerus and femur
Eotitanops princeps
Am. Mus. 296, Wind River. Right humerus (Ai, A2)
andtemur (B) of type. One-third natural size.
other subcursorial Eocene perissodactyls, such as
Hyrachyus, rather than the mediportal middle Eocene
forms. Among the middle Eocene titanotheres its
closest resemblances are to the femur of MesatirMnus.
The head is lacking. From the great trochanter to
the bottom of the internal condyle the shaft measures
250 millimeters. The bone is further distinguished
from that of some of the more recent or middle
Eocene titanotheres by the laterally compressed and
deeply recurved great trochanter, by the very promi-
nent second trochanter, by the lateral compression or
convexity of the shaft anteriorly between the trochan-
ters, a character which it shares with MesatirMnus,
by the flattening of the shaft posteriorly between and
Ai
Figure 501. — Pelvis of Eotitanops borealis
A, Am. Mus. 148S7' (neotype), incomplete pelvis associated with skull, ventral
(Ai) and outer side (A2) views; B, Am. Mus. 14888, fragment of left ilium,
indicating a somewhat wider ilium than that in the preceding specimen, asso-
ciated with teeth and other parts, ventral view. One-third natural size.
compressed; (4) plane of ilium at angle to that of
pubis; (5) proportions of subcursorial rather than of
mediportal type.
Pes of Eotitanops
The pes as compared with those of the other lower
Eocene Perissodactyla (fig. 502) — namely, Eohippus
(Equidae) and Heptodon (Lophiodontidae) — again
exhibits Eotitanops as approaching the relatively large
and slow-moving perissodactyls of the period, although
possessing many characters in common with the con-
temporary lighter-limbed forms.
In the comparative outlines of Figure 502 it wUl
be observed that Eotitanops horealis stands midway
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
597
between EoTiippus venticolus and Hyrachyus agrarius,
as a transition between cursorial and mediportal
adaptation. The changing proportions of the astrag-
alus, calcaneum, and ectocuneiform are especially sig-
nificant. Attention should be called to Eotitanops as
isotridactyl — that is, the three metapodials are more
uniform in size than those of either Eohippus or
Hyrachyus.
Peculiar features of the Eotitanops pes are (1)
elongate neck of astragalus; (2) a vertically elongate
the distal ends of the metapodials. These gradations
correspond broadly with those we have observed in
the dental series and with the successive increases in
Eotitanops major Osborn
Type.^Am. Mus. 14894, a left median metatarsal
(fig. 493, D) ; also the distal end of the right tibia.
Specific cTiaracters. — Of superior size; Mts III 104
millimeters longitudinal, 16 transverse, index 15.
JSofiippus
Jleptodo,
Figure 502. — Left pes of cursorial and subcursorial Eocene Perissodactyla
A, EoMppus venticolus, a primitive hippoid with narrow, slender foot and enlarged median metatarsal; B, Beptodon calciculus, a primi-
tive lophiodont with side toes not reduced; C, Eotitanops borealis, a primitive titanothere with broader tarsals and metatarsals
D, Hyrachyus agrarius, a primitive cursorial rhinoceros. One-half natural size.
cuboid (inferential); (3) elevated ectocuneiform; (4)
Mts III with cuboid facet narrow or wanting; (5)
Mts IV with broad ectocuneiform facet; (6) meta-
podials keeled posteriorly; (7) proximal median
phalanges relatively short; (8) inferior astragalo-
calcaneal facets usually small, entirely separate from
sustentacular. The gradations in size of the as-
sociated or referred foot bones of Eotitanops are well
shown in the accompanying series (fig. 503) displaying
This ill-defined species indicates the existence in
Wind Eiver time of a relatively large, slender-footed
titanothere, which is possibly ancestral to some of
the short-footed middle Eocene types.
A comparison of the measurements of E. major with
those of E. horealis, from the Wind Eiver formation
(Am. Mus. 14888), indicates an animal of much larger
size but of the same foot proportions as shown by the
index. There is a small cuboidal facet.
598
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Measurements of Eoiitanops major and other species, in
millimeters
E.
borealis
E.
major
Mesati-
rhinus
peter-
soni
Tapiras
tetrestris
Median metatarsal. III, length
Width of shaft _
86
13
15
21
26
104
16
15
25
33
120
21
17
114
21
Index
Tibio-astragalar facet, transverse-
18
bones of the carpus and tarsus, or in the shape of the
small bones on the sides of the manus, such as the tra-
pezium and pisiform, and of the pes, such as the
entocuneiform. There are also certain proportions in
the respective width or depth of the elements of the
carpus and tarsus which remain highly distinctive.
Every segment of each limb develops as a unit
independent^ of every other segment. Allometric
adaptation of each segment follows the general or
uniform laws that are observed in other quadrupeds.
These laws bring about the convergence or parallelism.
Eokzppiis B^eptodon. -LamMd/}-
N
-EotiioTiops
Figure 503. — Astragalus and calcaneum of cursorial and submediportal lower Eocene Perissodact^la
Front and distal views. A, Eohippus sp., with narrow astragalus (in distal view the astragalus barely touches the cuboid); B,
Heptodon calcicidus; C, Lambdotkeriiim popoagicum; D, Eotiianops gregoryi; E, Eoiitanops borealis; F, Eoiitanops sp. The
cuboid facet (cb) on the astragalus is better developed in the later stages, though it is still much narrower than in middle
Eocene titanotheres. Two-thirds natural size.
SECTION 3. MIDDLE EOCENE GROUPS AND PHYLA
The titanotheres of middle Eocene time may be
grouped as follows:
Palaeosyops (subgraviportal and brachj'podal) .
Limnohyops (mediportal and mesatipodal) .
Manteoceras and Dolichorhinus (subgraviportal and
brachypodal) .
Mesatirhinus (mediportal and mesatipodal) .
DOUBLE PARALLELISM IN THE PALAEOSYOPINE AND
MANTE0CERAS-D0LICH0RHINU3 GROUPS
The double parallelism of graviportal and medi-
portal proportions is so strong that it may again be
said that truly ancestral (paleotelic) characters prove
to be less conspicuous, obvious, or apparent than
adaptive (cenotelic) characters. We must search
for real subfamily relationships in inconspicuous parts
of the limbs, as in the form of the facets between the
Allometric change may progress entirely independ-
ently of affiliation to remote ancestral stock or
syngenesis; descendants of primitive mesatipodal
forms may remain mesatipodal or may become either
brachypodal or dolichopodal.
Figure 504. — Astragalocalcaneal facets in lower
Eocene Perissodactyla
A, Eohippus sp.; B, Heptodon calciculus; C, Systemodon primac-
vus; D, Lambdoiherium popoagicum. Two-thirds natural size.
Like brachycephaly and dolichocephaly in the skull,
an allometric tendency toward brachypody [or
dolichopody generally proceeds to an extreme, '.but
not invariably.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
599
Figure 505. — Left astragalus
and calcaneum of Eoiita-
nops sp.
FAMILY AND SUBFAMILY CHAEACTERS OF SKELETAL
PAKTS IN MIDDLE EOCENE TITANOTHERES
VERTEBRAE, ATLAS
Palaeosyopinae . — In the atlas of LimnoJiyops, as in
Palaeosyops, the spinal nerves (see fig. 508, A) issue
nearer the median line than in Telmaiherium (fig. 508,
B); the neural arch and the bridge over the vertebral
artery are narrow. The
axis is rather narrow, with
a slender odontoid process.
The postero-inferior tuber-
osity is moderately devel-
oped.
In the atlas of Palaeo-
syops (Am. Mus. 1580, fig.
508, A), as in LimnoJiyops,
the spinal nerves issue
nearer the median line than
in Telmafherium. Unlike
LimnoJiyops the neural arch
is broad and powerful; the
inferior vertebrarterial
bridge is wide; the pleura-
pophyses expand broadly
Am. Mus. 14895, Wind River Basin, • u„i-u j;..„„t;„„„ T^kn
Wyo.; Wind River formation, level '" both du-ections. The
B. Facet for tibia (/ft). Two-tiiirds postcro-inferior tuberosity
°^'""' '"■'■ is moderately strong.
TelmatJieriinae. — In the atlas of Telmatlierium ulti-
mum (Am. Mus. 2060; fig. 508, B) the spinal nerves
are well separated, the pleurapophyses expand lat-
erally but do not flare, the inferior vertebrarterial
bridge is broad, but the canal itself is contracted, fore-
shadowing the exclusion of the vertebral artery from
the canal, which is a very distinctive character of
some of the Oligocene titanotheres. The occipital
cotylus is broad with massive borders; the odontoid
facet is exceptionally broad and open (thus distin-
guished from that of Manteoceras) . The postero-
inferior tuberosity is moderately strong.
Manteoceratinae. — The atlas of Manteoceras (Am.
Mus. 12204, fig. 508, E) has the characters we should
expect to find in a broad-skulled member of the
Manteoceratinae: the vertebrarterial opening is very
large posteriorly; the bridge is narrow (16 mm.) and
rounded; the articular facets for the axis form a widely
open angle; the postero-inferior tuberosity, which
underlies the axis, is stout and prominent, as in the
Oligocene titanotheres.
Dolichorhininae. — In Mesatirhinus (Am. Mus. 1523.
fig. 508, C) the pleurapophyseal wings are not known;
the atlas presents an approximation to that of Doli-
chorMnus: (1) the centrum is relatively elongate, (2)
the spinous foramina are nearer together, (3) the ver-
tebrarterial canal is covered by a bridge of medium
length, (4) the facets for the axis are approximated and
obliquely face each other, (5) the cotyli for the occip-
ital condyles are relatively deep, (6) there is a reduced
postero-inferior tuberosity.
In DolicJiorhinus (fig. 508, D, Am. Mus. 1844, 13164
associated with skull), the body is relatively elongate,
the occipital cotyli are very broad and deep, the
spinous nerve notch is very deep and narrow, the
pleurapophyses flare widely, the vertebraterial canal
is more open than in Telmatlierium ultimum. As
in Mesatirhinus prominent bony processes connect the
transverse ligament above the odontoid process.
Summary. — The atlas of Palaeosyops and Limno-
Jiyops conforms to the brachycephahc types of skull,
that of DolicJiorJiinus to the dolichocephalic type;
those of Telmatlierium and Manteoceras are interme-
diate between these extremes.
From the limited materials in our possession {Palaeo-
syops, DolichorJiinus, MetarMnus) we observe that the
scapula of the Eocene titanotheres is subject to wide
adaptive range from the more elongate mediportal
type of the supposed MetarMnus to the broader sub-
graviportal type of Palaeosyops. The scapula is, how-
ever, quite distinctive in its family or syngenetic
form as compared with that of other Perissodactyla,
being characterized as follows: (1) a vertically elon-
gated supraspinous fossa, which is equally broad above
and below, giving a relatively straight anterior border;
(2) a relatively large infraspinous fossa of subtriangu-
lar form, in which the border rapidly contracts toward
Figure 506. — Metatarsal and tibia of Eotitanops
major
Am. iVIus. 14894 (type), Wind River Basin, Wyo.; level B.
Median metatarsal: Back (A>), front (A^), distal (A'), and
proximal (A') views. Distal end of left tibia: Front (B') and
distal (B-) views. One-iialf natural size.
the neck — on this border is a distinct "angle" which
becomes sharply marked in later types; (3) a broad,
short neck, or collum scapulae. This is the graviportal
prototype from which the scapula of the heavy Oligo-
cene titanotheres is readily derived. On the whole,
this middle Eocene scapula is analogous to that of the
mediportal Tapirus, but it is distinctly tending toward
and even modified in the graviportal direction.
600
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Only three scapulae are known, as follows : (1) Medi-
portal, Dolichorhinus longiceps (figs. 509, 582); (2)
subgraviportal, Palaeosyops (figs. 509, 545), with.
In Mesatirhinus we should expect to find the scapula
proportioned somewhat like that in Tapirus; unfor-
tunately this scapula is unknown.
i'lGUUE 5Ui. Restoration of Eotilanops borealis
Modeled by E. S. Christman. About one-twelfth natural size
height 345 millimeters, breadth 260; (3) supposed
MetarUnus (figs. 509, 576, Am. Mus. 1873, Uinta B),
elongate compared with that of Palaeosyops — namely,
length 265 millimeters, breadth 165 (estimated). The
supposed MetarJiinus may be taken as the mediportal
type; it is less expanded toward the upper border.
HUMERUS
We hare observed (fig. 510) that the characters of
the humerus of the titanothere family are rapidly
intensified by adaptation. Each genus exhibits dis-
tinctive ratios of length of the humerus and radius, as
shown in the accompanying table.
Measurements of fore limh of certain titanotJieres compared with tapirs, in millimeters
Eadio-
humeral
ratio
Metacarpo-
humeral
ratio
Tapirus terrestris
Tapirus indicus
Eotitanops princeps (borealis), T^m. Mus. 296
Limnohyops? monoconus. Am. Mus. 11689
Palaeosyops leidyi, Am. Mus. 1544
Palaeosyops copei?. Am. Mus. 12205
Dolichorhinus h3'0gnathus, Am. Mus. 13164..
Menodus trigonoceras, Munich Mus
Brontops robustus, Yale Mus. 12048 (type)..
Brontops sp., Am. Mus. 518
205
250
203
293
325
340?
■315
620
608
528
177
228
228
235
237
'284
520
504
478
77
72
69?
81?
83
82
90
106
120
85
109
113?
106
50
48
41
37
34?
30
240
230
214
38
37
40
It wUl be observed that ia Palaeosyops and DolicJio-
rliinus the humerus is much longer than the radius.
These are subgraviportal types, but even in the medi-
portal Limnohyops the humerus is somewhat longer
than the radius.
It is, however, a very marked distinction of the
titanotheres, already pointed out, that as they become
heavier the radius elongates more rapidly than the
humerus. Thus the radiohumeral ratio rises from 72
in Palaeosyops leidyi to 90 in Brontotherium gigas.
The rugose muscular attachments of the humerus
rapidly assume progressive graviportal characters,
which even in the lower Bridger enable us to distin-
guish readily the titanothere humerus. These char-
acters are strengthened in the Oligocene titanotheres
and were undoubtedly correlated with certain dis-
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
601
tinctive motions of the fore limbs as follows: (a) The
great tuberosity of the humerus expands into a broad,
elevated plate; the deltoid crest descends into a dis-
tinct tuberosity; the supinator crest secondarily
increases in size and prominence. Of the above
characters the platelike great tuberosity (fig. 510) is
Primitively (Eotitanops iorealis) the deltoid tuber-
osity of the humerus is near the upper part of the
shaft; secondarily it is extended downward. Prim-
itively the great tuberosity is a sessile prominence
divided by a shallow median notch; secondarily it
becomes very prominent and platelike and is divided
fo.intei^v.
E4
FiGUKE 508. — Atlas of Eocene titanotheres
A, Palaeosyops rohustus, Am. Mus. 1580, upper Bridger; B, Telmatherium ullimum, Am. Mus. 2060 (type), Uinta C; C, Mesatirhinus megarhinus. Am.
Mus. 1523, upper Bridger; D, DoUchorhinus sp., Am. Mus. IS44, Uinta C. A-D, Dorsal views. E, Manieoceras manteoceras, Am. Mus. 12204, upper
Bridger: Ei, Dorsal view; Ej, ventral view; Ej, anterior view; Ei, posterior view. One-third natural size. The arrows indicate the course of the first
spinal nerve and vertebral artery.
the most distinctive, (b) Distally the radio-ulnar
articulation (rotula and capitellum) is decidedly
asymmetric; this asymmetry persists in the titano-
theres; its significance is fully explained on page 602.
(c) The ectocondylar or supinator crest is already quite
prominent, (d) The entocondyle is less prominent.
by a very deep median notch. Primitively the ecto-
condylar crest, which is small, is distinctly defined,
rising somewhat on the shaft (Eotitanops); second-
arily it rises and widely expands. A distinctive fea-
ture of the humerus as compared with that of the
Hyracodontidae and Amynodontidae is the prominent
602
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
rugosity (fig. 500) on the inner side of the shaft for
the tendon of the latissimus dorsi muscle.
Figure 672 gives a comparison of the humeroradial
articulations of Tapirus, Palaeosyops, and Rhinoceros.
We observe that the asymmetry of the trochlea and
capitellum persists in graviportal types, whereas the
cursorial Equus acquires a more symmetrical form.
Although this titanothere elbow joint has points of
similarity with that of the tapirs, paleotheres, and
other primitive perissodactyls, the form as a whole
appears to be somewhat distinctive. Viewed from
below, the trochlea is much broader than the capitel-
lum; the trochlea is nearly plane, the capitellum
sharply convex. In the Oligocene titanotheres (fig.
510) the capitellar area widens out so that there is less
marked asymmetry with the trochlea.
Correlated with this is the asymmetry of the upper
end of the radius as seen from the front, the trochlear
Figure 509
, Metark:
-Types of scapula in middle Eocene titanotheres
s, Am. Mus. 1873 (mediportal) ; B, DoUchorhinus sp., Am. Mus. 1833 (mediportal)
C, Palaeosyops robusius, Am. Mus. 1680 (subgraviportal).
portion being depressed, the capitellar portion ele-
vated. A series of comparative views of the proximal
end of the radius in several genera (fig. 511) brings
this point out clearly.
In the primitive ulna {Lambdotherium) the olecranon
is decidedly erect and truncate at the top ; it becomes
more depressed and pointed in the progressive weight-
bearing forms. In Limnohyops and Mesatirhinus is
seen an intermediate mediportal condition (fig. 511).
In Mesatirhinus it is rounded and moderately rugose
at the extremity; in Limnohyops more heavily rugose.
In the graviportal forms {Palaeosyops, fig. 511) it be-
comes pointed and heavily rugose at the extremity,
prophetic of the condition in the heavy Oligocene
titanotheres. In Dolichorhinus it is incurved, abbre-
viated, and depressed. The shaft of the ulna is of the
flattened, trihedral form. This olecranon process is
cleft by a faint groove superiorly, which becomes very
distinct in the Oligocene forms.
Evolution of tJie manus. — The ancestral form of
manus and pes is, as shown above, subcursorial. The
principles of convergent mediportal and graviportal
adaptation, observed in the arches aiid limbs, also
dominate the foot structure; the divergence, in fact,
is still more conspicuous. On the principles set forth
above (pp. 583-584), vertical elongation of every
element of the carpus and tarsus as a rule points to
speed, while depression and transverse extension
point to weight.
The general trend of the evolution of the manus
and pes in the titanotheres is from the relatively high
and narrow subcursorial type {Eotitanops) through
an intermediate or mediportal type {Mesatirhinus)
into a relatively broad and low graviportal type
{Palaeosyops).
From the conspicuous adaptive divergences either
in the mediportal or the graviportal direction,
it is first necessary to select the relatively
obscure family or universal characters of the
manus and pes. The digital formula is that
of all primitive perissodactyls. The manus
has four digits (mesaxonic to subparaxonic);
the pes has three digits (mesaxonic). One
peculiar feature of the titanotheres is that (as
in the amynodont rhinoceroses alone among
all other perissodactyls) the primitive four
digits persist in the manus. The pes in the
titanotheres, as in all other perissodactyls,
because it is connected with the chief loco-
motor organ, the hind limb, is more reduced
in size and more progressive in form than
the manus.
Primitive manus. — The manus of the ear-
liest types known {Lambdotherium, Eotitan-
ops) is mesaxonic — that is, the third, median
digit (D. Ill) is distinctly the largest, as in Eohippus,
Heptodon, and Lophiodon. The retention and expansion
of the fifth digit (D. V) to share the weight appears
to be partly secondary.
Adaptive manus. — The titanothere family characters
maybe summed up as follows: (1) Persistent tetra-
dactylism, with considerable adaptive divergence
among the genera in the relative length of the four
digits, especially of D. V, which is relatively shorter
in some forms {Palaeosyops, Limnohyops) and longer
in others {Mesatirhinus, Manteoceras) ; (2) in the front
view of the metacarpals, Mtc II, while supporting
mainly the trapezoid, articulates broadly against the
magnum; Mtc III, while mainly supporting the
magnum, also articulates broadly against the unciform;
(3) the trapezium is likewise present (in Eocene time)
and articulates with the side of the trapezoid, also
with Mtc III, and often (especiallj' in Manteoceratinae)
with the scaphoid; (4) the proximal phalanges are of
EVOLUTION' OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
603
Figure 510. — Types of fore limb in Eocene and Oligocene titanotheres
Subcursorial: A, Lambdotherium; slightly less than one-third natural size. Mediportal: B1-B3, Limnohyops; Ei, £2, Mesathhrnu
natural size. Subgraviportal: C1-C3, Palaeosyops; Di, X>2, Manteoceras; Fi-Fs, BoUchorhinus; slightly less than one-eigl
G1-G5, Brontops; slightly less than one-twelfth natural size.
slightly 1(
h natural
ss than one-eighth
size. Graviportal:
604
TITANOTHEEES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
relatively broad type (mediportal and graviportal),
never elongate (cursorial) as in the primitive horses;
(5) the median phalanges are short; (6) the distal
phalanges expand at the extremities and exhibit deep
median clefts, indicating the attachment of broad
horny hoofs like those of the rhinoceroses rather than
narrow or appressed hoofs of the equine or hyracodont
type. The phalanges differ considerably in length
and in breadth in different genera.
Manus oj the palaeosyopine group. — The palaeosyo-
pine group includes three genera, TelmatJierium,
Limnohyops, Palaeosyops; the manus of the last two
named is known. Of these Limnohyops is more
primitive and mediportal in its proportions and articu-
lations and is thus analogous to the primitive and
mediportal Mesatirhinus of the Dolichorhininae.
-Characteristic details of radius and
upper Eocene titanotheres
ulna in middle and
Lower row, proximal end of left ulna, outer view. Middle row, distal end of left radius, front
view. Upper row, proximal end of left radius, front view. A, Limnohyops monoconus?;
Limnohyops laticeps; C, Palaeosyops cf.
tirhinus petersoni: G, Dolichorhinus hyognathus.
leidyi; D, Palaeosyops Tobustus;
One-sixth natural size.
Figure 512 shows how similar these two forms are, yet
a very close examination of the details of structure in
the carpals and tarsals proves beyond question that
they belong in separate subfamilies and diverged from
each other at a remote period.
The following dry descriptive details are of philo-
sophic interest to the comparative anatomist because
they demonstrate the assertion made above that we
can discover a distinct syngenetic (common origin)
character in each element of the carpus when closely
examined and compared. They prove that the rela-
tively light-limbed and more rapidly moving Limno-
hyops retains more of the ancestral form and propor-
tion (thus analogous to Mesatirhinus) while the short-
footed Palaeosyops diverges most widely in the gravi-
portal direction (thus analogous to Manteoceras) .
Limnohyops characters. — Of mediportal proportions;
mesatipodal; carpus more elevated and compressed
than in the subgraviportal Palaeosyops; metacarpals
and digits relatively longer than in Palaeosyops but
relatively broader than in Mesatirhinus; digit V rather
short; scaphoid and cuneiform relatively high; lunar
high, resting mainly on unciform, obliquely on mag-
num; magnum high, laterally compressed; second
phalanges abbreviated; terminal phalanges cleft and
spreading distally as in Mesatirhinus.
There are two subtypes of Palaeosyops manus.
(See figs. 537, 549.)
Palaeosyops characters (Am. Mus. 12205). — Of
graviportal proportions; brachypodal; carpus broad
but less depressed than in Palaeosyops frohustus;
digits more abbreviated and spreading than in Palaeo-
syops leidyi; trapezium large, with no scaph-
oid articulation apparent; magnum relatively
high, rather broad, with but five distinct
faceted angles; the lunar facet in front view
is continuous with the unciform; terminal
phalanges irregular, rounded distally.
Manus oj the Manteoceras-Dolichorhinus
group. — In this subfamily Mesatirhinus is the
primitive and mediportal type analogous to
Tapirus and Limnohyops, while Manteoceras
is the modified subgraviportal type analogous
to Palaeosyops, although less extreme. Close
comparison of the carpals of Mesatirhinus
and Dolichorhinus demonstrates the manteo-
ceratine affinity and divergence from the
palaeosyopine type.
Mesatirhinus. — A mediportal carpus or
relatively high, narrow fore foot, all the
elements being vertically elongated but less
so than in Tapirus terrestris. Mesatipodal;
total breadth of carpus of a typical specimen
{M. megarhinus) 65 millimeters, total depth
43 ; scaphoid less deep anteroposteriorly than
in Palaeosyops, appearing relatively high and
narrow; lunar very high, with oblique magnum facet,
thus in front view bearing principally on unciform;
trapezium narrow, with scaphoid, trapezoid, and Mtc
II facets; trapezoid relatively large; magnum high,
relatively narrow, and subquadrate, with high facets,
hook of magnum more symmetrical, broadly spatulate,
with deep median groove; unciform with its longest
diameter oblique instead of horizontal, as in Palaeo-
syops; Mtc V elongate, manus consequently more
tetradactyl; metacarpals decidedly slender but less so
than in Tapirus terrestris; distal phalanges elongate,
cleft, and broadly expanded distally.
Dolichorhinus. — The manus of Dolichorhinus repre-
sents a sub-brachypodal specialization of the Mesati-
rhinus type, paralleling that of Manteoceras. (See
p. 606.)
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
605
Figure 512. — Manus of lower and middle Eocene titanotheres
A, LamhdotheTium popoagicum; B, Eotitanops princeps; C, Limnohyops monoconus: D, Palaeosyops leidyi; E, Mante\
F, MesaiirMnus petersoni. One-tliird natural size.
, ,, ^si_S^ \j:^ ABC
Bi W^ Ci Dr
Figure 513.— Comparison of the right scaphoid in middle Eocene Figure 514.— Terminal phalanges of the manus in middle Eocene
titanotheres titanotheres and amynodonts
Lower row, front view; upper row, top view. A, Palaeosyops leidyi: B, Limnohyops Generic differences in the terminal (ungual) phalanges of the third digit. A,
monoconus; C, MesatirUnus petersoni: D, Manieoceras manteoceras. One-half nat- Mesatirhinus petersonit: B, Manteoceras manteoceras: C, Limnohyops monoconus?:
jifal gijg_ D, Palaeosyops sp.; E, Amynodon sp. One-halt natural size.
606
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Manteoceras. — A graviportal, or short,- broad foot;
digits and metacarpals, however, not spreading
apart distally as in Palaeosyops. Brachypodal; de-
tailed proportions and facets of carpals resembling
a bi'oad, flat face, subquadrate as seen from in front,
hook of magnum more symmetrical, broadly spatulate;
second phalanges abbreviate; terminal phalanges dis-
tinctly abbreviate, expanded distally, with a pro-
D
Figure 515. — Progressive graviportal adaptation in the pelvis of Eocene and Oligocene titanotheres
A, Eotitanops borealis (subcursorial); B, Palaeosyops major (inediportal); C, Manteoceras sp. (subgraviportal); D, Brontotheriumlsp. (graviportal).
One-eighth natural size.
those of Mesatirhinus rather than those of the Palaeo-
syopinae; lunar high, resting chiefly on unciform with
a very oblique magnum facet; magnum relatively high
and narrow, with five facets in front view, acuminate
above, with lunar unciform facet continuous, forming
nounced median groove. From the above description
it appears that Manteoceras resembles Mesatirhinus in
the proportion of the carpals. It differs widely in
the abbreviation of the metacarpals and especially of
the terminal phalanges.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
Summary of contrasting cliaracters of the carpus and manus in middle Eocene titanotheres
607
Palaeosyopine group.
Manteoceras-Dolichorhinus group
Typical genera, Palaeosyops (subgraviportal), Limnohyops (mediportal)
Typical genera, Mesatirhinus (mediportal), Manteoceras, Dolichorhinus (sub-
graviportai)
1. Carpus broad, composed of relatively broad, low, and deep
(anteroposteriorly) elements, correlated with broader
spreading digits and spreading or rounded hoofs.
2. Scaphoid broad, deep anteroposteriorly, outer lateral face
prolonged backward, radial facet rounded.
3. Cuneiform (Palaeosyops) broad, chief diameter horizontal.
4. Trapezium large, articulating with trapezoid, with Mtc II,
but not generally with the scaphoid.
5. Trapezoid relatively large, subquadrate in form in front view.
6. Magnum large, with five to six facets, hook of magnum
strongly asymmetrical, pointed.
7. Unciform more horizontal.
8. Metacarpals miore mesaxonic — that is, Mtc III relatively
elongate; Mtc V abbreviate.
9. Metacarpals with deep proximal facets for the carpals at the
extremities.
10. End phalanges rounded or spreading at the extremities.
1. Carpus composed of relatively -high elements, less deep
anteroposteriorly, hoofs truncate and spreading dlstally.
2. Scaphoid relatively high, more shallow anteroposteriorly,
radial facet flat.
2a. Lunar high and laterally compressed, resting chiefl.v on
unciform.
3. Cuneiform relatively high and laterally compressed.
4. Trapezium large and deep vertically, articulating with
trapezoid, with Mtc II, also with scaphoid (Mesatirhinus) .
5. Trapezoid relatively small.
6. Magnum high, relatively narrow, subquadrate, with five
facets, hook of magnum more symmetrical.
7. Unciform more oblique.
8. Metacarpals more paraxonio — that is, Mtc V well devel-
oped and podium more tetradactyl.
9. Metacarpals with proximal facets for carpals truncated
rather than deep, with distal facets less globose.
10. End phalanges broadly expanded distally.
The pelvis of the middle and later Eocene titano-
theres has passed beyond the mediportal Tapirus
stage and is in a stage between the subgraviportal
and graviportal, conforming with the principles of
the evolution of the pelvis set forth in Chapter IX
(pp. 743-745). Although associated and complete
pelves are rare, the material available throws a great
deal of light on the characteristic form and develop-
ment of this important organ.
The chief materials and the geologic horizons at
which they were collected are as follows:
Eotitanops horealis, Am. Mus. 14887 (fig. 501): Wind River
A or B.
Limnohyops laticeps, Yale Mus. 11000 (fig. 532) : Bridger C or D.
Limnohyops ? sp., Am. Mus. 2348 ' (fig. 538) : Washakie A.
Palaeosyops frobustus, Princeton Mus. 10232 (fig. 539) : Bridger
Cor D.
Palaeosyops fmajor, Am. Mus. 13116 (fig. 533): Bridger B.
f Manteoceras manteoceras, Am. Mus. 2358 (fig. 555) : Washakie
B.
Dolichorhinus hyognathus, Am. Mus. 1843 (figs. 579, 580):
Uinta B.
f Dolichorhinus ?hyognathus, Am. Mus. 1860: Uinta B.
Of the above the pelvis of Eotitanops (fig. 501)
shows a mingling of subcursorial and prophetic medi-
portal characters. The complete associated pelvis of
the type of Limnohyops laticeps in the Yale collection
belongs to a juvenile individual, which may partly
explain the fact that it is in a more primitive stage of
development than any other middle Eocene titano-
' Specimen mounted with skeleton of P. leidyi in American Museum.
101959— 29— VOL 1 42
there pelvis known; the superior border of the iliac
crest is partly thin and concave, representing the
crista iliaca between the rugose borders of the tuber
sacrale and tuber coxae.
A similar pelvis in a somewhat more advanced
stage (fig. 538) is that from Washakie A (Am. Mus.
2348), which has been mounted with the skeleton of
Palaeosyops leidyi in the American Museum. This
represents a slightly more advanced stage, which
nevertheless retains the thin, slightly concave crista
iliaca between the rugose areas of the tuber coxae and
tuber sacrale.
A pelvis certainly belonging to Palaeosyops, from
Bridger B (Am. Mus. 13116, fig. 533), is distinguished
by the elongate os innominatum and by the uniformly
convex superior crest of the ilium.
Of this type also is the pelvis (fig. 539) in the Prince-
ton Museum (No. 10232), attributable to P. leidyi or
P. rohustus, distinguished by a much broader ilium,
with a uniformly convex superior border.
Referable to Manteoceras is the finely preserved
pelvis (Am. Mus. 2358) from Washakie A or B,
characterized by very great iliac breadth (fig. 555)
and the broad expansion of the supra-iliac border.
This is an advanced subgraviportal type, prophetic
of the form of the pelvis in the Oligocene titanotheres.
The pelvis associated with Dolichor'hinus hyognathus
from Uinta B (Am. Mus. 1843) is again elongate
(fig. 580), with a uniformly convex superior border
of the ilium.
The comparative measurements of these pelves are
shown in the accompanying table.
608
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Measurements oj pelves of Eocene titanoiheres compared, luitTi tapirs, in millimeters
Tapirus
indicus
7Palaeo-
syops,
Princeton
Mus.
10232
Palaeo-
syops
major,
Am. Mus.
13116
Limno-
hyops
laticeps,
Yale Mus.
11000
(type)
Limno-
hyops,
Am. Mus.
2348
?Man-
teoceras,
Am. Mus.
23S8
Dolicho-
rhinus,
Am. Mus.
1843
Dolicho-
rhinus,
Am. Mus.
1860
TDiplaco-
don
elatus
TProtitano-
therium
Total length of os innommatum.._
Total length of ilium
408
258
169
430
230
100
415
270
190
444
285
168
406
203
440
290
175
500
280
175
430
285
140
530
290
185
143
70
470
300
175
610
335
Total width of pelvis across ilia
Total width of iUum (superior
°665
265
235
253
340
Anteroposterior pubo-ischiadic
Total breadth across ischiadic
Breadth of peduncle of ilium
40
60
53
73
65
A comparison of these figures and measurements
demonstrates that the main features of the progressive
evolution of the titanothere pelvis in graviportal
adaptation are the following:
1. Relative expansion of the iliac crest.
2. Reduction of the thin crista iliaca and expansion of
the rugose superior borders of the ilium.
3. Relative abbreviation of the os innominatum.
4. Uniform rugose convexity of the superior crest of the
ilium.
The above are all progressive graviportal characters.
It is demonstrated that even as far back as middle
Eocene time the early weight-bearing or subgraviportal
type of ilium was well established among the titano-
theres, and that from the evidence afforded by the
ilium alone these animals were heavier bodied and
slower moving of limb than the modern tapirs. Our
analysis (see below) of the graviportal adaptation in
the ilium makes entirely clear the general functional or
adaptive stages through which the pelvis of these
Eocene titanotheres is passing. We observe that the
relatively elongate ilium of Limnohyops is in the first
stage, that Palaeosyops with its shorter ilium and
heavier body has passed beyond this, and that
Manteoceras possesses the fully developed graviportal
type of ilium.
The principal family characters of the pelvis appear
to be as follows: (1) The metapophyses of the posterior
lumbar vertebrae articulate with the anterior border
of the ilium, as in Equus; (2) in LimnoTiyops the
first and second and half of the third sacral vertebrae
expand to articulate with the Uium, whereas in
Tapirus only one and half of another unite with the
ilium; similarly in Manteoceras three sacrals enter
into the iliac union; (3) the ilium is much longer than
the ischium; (4) the superior border of the ilium is
slightly indented (Limnohyops) or uniformly convex
{Palaeosyops, Manteoceras); (5) the peduncle of the
ilium is broad and short, its dorsal border presenting
a sharp ridge; (6) the ischia are not separated or cleft
posteriorly as in the Amynodontidae, Rhinocerotidae,
and other Perissodactyla; (7) the suprasacral area
(tuber sacrale) progressively expands; (8) the rugose
borders (tuber sacrale and tuber coxae) expand and
unite to obliterate the thin intermediate crista iUaca.
ILIUM
The ilium of Eocene titanotheres exhibits a con-
siderable range of generic characters, as follows:
UIUM OF THE PAIAEOSYOPINE GROUP
Limnohyops. — Ilia moderately expanded, with in-
dented superior border.
Palaeosyops. — Os innominatum elongate, iliac crest
uniformly convex, expanding into a broad border in
Bridger C or D stages.
UIUM OF THE MANTEOCEEAS-DOLICHORHINUS GROUP
Mesatirhinus. — The fragmentary specimen Am.
Mus. 1571 exhibits a slender peduncle of the ischium.
The ilium is unknown.
Manteoceras. — The pelvis provisionally associated
with this genus exhibits a very broad, uniformly
convex superior border of the ilium; os innominatum
relatively broadening and abbreviated.
Dolichorhinus. — Superior borders of ilium mod-
erately expanded or narrower than in the supposed
Manteoceras; os innominatum moderately elongate.
The detailed description of these various types of
pelves is given in the succeeding section of this memoir.
Materials. — A very large number of femora are
preserved, many of which have the tibiae associated.
On these it is possible to establish the femorotibial
type of the titanotheres, which from the beginning
to the end of their evolution is very distinctive.
The femur of even the ancestral Eotitanops (fig. 500)
displays some of the characteristic titanothere fea-
tures, although it retains the flexed knee, distal
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
609
patella, and many of the proportions and characters
of its primitive subcursorial ancestors.
The femora of the middle Eocene palaeosyopines
(Palaeosyops, Limnohyops) and manteoceratines
(Mesatirhinus, Manteoceras) are seen from Figure 516
to display many special titanothere characters in
common.
The chief characters of the femur of the titanotheres
are the following, the comparisons shown being made
chiefly with the femur of the contemporary Amyno-
dontidae, which are animals of the same size.
posterior side is recurved, surrounding a deep, sharply
characteristic pit. The second trochanter, t", is
relatively less prominent than in primitive rhinoc-
eroses; the third trochanter, t'", is placed high on
the shaft, as in all the Equidae and in the primitive
Rhinocerotidae (Hyrachyus) ; in Eocene titanotheres it
never extends down to the middle of the shaft as in
the progressive Rhinocerotidae; the third trochanter
is less extensive than in Hyracodon or Amynodon.
4. The second and third trochanters are more
nearly opposite each other, as in the primitive Equi-
FiGURE 516. — Femora and tibiae of middle Eocene titanotheres
C, Mesatirhinus petersoni; D and E, Manteoceras manteoceras; F, Telmatherium uliimum.
One-sixth natural size.
1. The femur is longer than the tibia. This differ-
ence is seen in the early mediportal types and increases
with the graviportal adaptation.
2. The head of the femur is primitively more
spherical but becomes progressively flatter or less
spherical.
3. The shaft is rather straight and flat and has a
slight forward curvature inferiorly. The great tro-
chanter, t' , is not very high, being but slightly raised
above the level of the head and relatively smaller
and less prominent than in Amynodon. Its inner
dae. This is a peculiar and rather characteristic
feature.
5. The external patellar ridge is smaller and less
produced superiorly; the distal condylar surfaces for
the tibia are somewhat flatter; the internal one
is larger.
A most characteristic feature is the patellar trochlea.
In the Palaeosyopinae the knee is straighter; in the
Manteoceratinae it is more flexed. In the palaeosyo-
pine subgraviportal types (Palaeosyops, Limnohyops)
the patellar trochlea (fig. 518) is more vertical and
610
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
thus presents much more anteriorly than in the early
rhinoceroses {Amynodon). This indicates that the
femur was carried vertically at an early period. In the
manteoceratines the patellar trochlea is more distal
and oblique; thus in the more light-limbed Mesati-
rhinus the patellar facet presents more obliquely
downward and forward, and the femur was carried
more obliquely, as in the rhinoceroses {Hyrachyus
and Amynodon). Even in Manteoceras it was more
downward than in Palaeosyops. Distally the internal
and external tibial condyles are nearly subequal. In
the amynodonts the external condyle is much more
prominent, and the corresponding external tibial facet
is produced backward.
Femoi'otibial ratios. — In all the titanotheres the
tibia is much shorter than the femur. This disparity
is progressive (tibiofemoral ratio, 77 to 54) in the
course of graviportal adaptation.
Length of hones of Mnd limbs oj titanotheres and of tapir, in millimeters
Tapirus indicus
Eotitanops princeps, Am. Mus. 296 (t3'pe)__
Palaeosyops major, Am. Mus. 1316
Palaeosj'ops leidyi, Am. Mus. 1544 (type)..
Limnohyops sp., Am. Mus. 11689
Manteoceras manteoceras, Am. Mus. 1587.-
Mesatirhinus "petersoni," Am. Mus. 11659.
Dolichorhinus hyognathus, Am. Mus. 13164
Menodus trigonoceras, Munich Mus
Brontops robustus, Yale Mus. 12048 (type).
Brontotherium gigas, Am. Mus. 519
320
250
433
370
355
390?
358
386
770
812
780
258
332
290
285
272
285
120
86?
137
110
111
37
34?
31
30
31
430
448
427
118
119
220
212
200
33
30
28
26
20
Contrasts with Amynodontidae. — The amynodonts
are contemporary semiaquatic rhinoceroses. In Amyn-
odon the great trochanter is more elevated ; the patellar
facet is subhorizontal at the distal extremity of the
shaft.
TIBIA
The tibia (fig. 516) is invariably shorter than the
femur, the ratios in the various genera being as in the
FiGTiRK 517. — Distal end of the femur in a middle
Eocene titanothere and an upper Eocene amyno-
dont
A, Manteoceras manteoceras; B, Amynodon sp. One-third nat-
ural size.
above table. So far as the relative abbreviation of
the tibia is indicative of speed and weight, Manteoceras
is relatively the slowest and Mesatirhinus relatively
the swiftest of the middle Eocene titanotheres.
Among the special titanothere characters are the
following: (1) External tuberosity of cnemial crest
more prominent than internal tuberosity; (2) cnemial
crest concave superiorly and deeply excavated on the
external side; (3) shaft deeply trihedral in midsection
and broadly flattened inferiorly. Characters 2 and 3
persist in the Oligocene titanotheres.
As compared with the large contemporary Rhinoc-
erotoidea (amynodonts and hyracodonts) of middle
Eocene time, some of which attained the same size
as titanotheres, we observe the following distinctions:
(] ) In Palaeosyops the cnemial crest runs more
obliquely inward, crossing the shaft; (2) the crest is
broader at the summit and does not extend so far
down the shaft as in Hyrachyus and Hyracodon; (3) the
tibia of Palaeosyops and Manteoceras in proximal and
distal views is shallow anteroposteriorly.
As compared with the tibia of Amynodon, the titano-
there tibia is more robust and lacks the posterior
prolongation of the femoral condyle.
The fibula was slender (but relatively less so than
in Amynodon), subcylindrical, broadly expanded above
for the articulation of the posterior inferior surfaces of
the external head of the tibia, and closely appressed
with the tibia below by a broad articulation, articu-
lating broadly also with the astragalus and exhibiting
a postero-inferior facet for the calcaneum during the
extreme extension of the foot.
In the pes, as in the manus, we observe certain
syngenetic family characters which are peculiar to all
titanotheres but which in various genera are more or
less concealed by adaptations to speed or to weight.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
611
This law of the dominance of teleogenetic (adaptive)
over syngenetic (ancestral) character is in force in
every single element of the pes as in the manus.
For example, the articulations between the astra-
galus and calcaneum exhibit both a mediportal type
resembling that of tapirs, of light-limbed rhinoceroses,
and of paleotheres and a graviportal type resembling
that of other graviportal perissodactyls. We are
therefore again compelled to examine minor and less
cuboid, respectively. In one specimen of Palaeosyops
Mts IV abuts against the ectocuneiform, but this is
unusual. Similarly Mts III occasionally doe.-i not
abut against the cuboid.
In the two middle Eocene groups, namely, the
Palaeosyopinae and the Manteoceras-Dolicliorhinus
group, it is possible by very careful study to discover
distinctive generic characters by which we may sepa-
rate every bone of the manus and of the pes, although
Figure 518.-
Angulation of the knee joint: relation of patellar facet to long axis of femur
Amynodon; Cae,
A, C, D, and E, one-fourth natural size; B, one-half natural size. Lm, LimnoJiyops; Hy, Hyrachyus; Ms, MesatiThinus; A
Caenopus.
conspicuous characters in order to discover the real
syngenetic family resemblance to be found in each
element.^"
Among family characters of the pes of the titano-
theres are the following:
1. The small fibulocalcaneal facet, as in the Equidae,
the fibula barely passing upon the calcaneum in the
extreme extension of the foot. A median pit on the
astragalus checks the flexion of the tibia by receiving
its posterior process.
2. The entocuneiform is very large, articulating on
the inner side of the mesocuneiform, of Mts II, and
of the navicular, and freely projecting backward from
the pes like a pisiform.
3. The mesocuneiform is invariably a very small
bone as in primitive mammals.
4. The ectocuneiform is a large element.
5. Mts II and III typically abut on the outer
proximal facets against the ectocuneiform and the
'• The significance of the astragalocalcaneal facets as family characters in Perisso-
dactyla was first pointed out by Osborn in the article "Evolution of the ungulate
foot" (Scott and Osborn, 1890.1, pp. 531-569).
through convergence and inheritance the bones are
often brought to resemble each other closely. The
h<^
Figure 519. — Inner side view
of left fibula
A, Palaeosyops sp.; B, Limnohyops
monoconusf ; C, Brontotherium leidyl.
One-sixth natural size.
chief subfamily distinctions in typical genera are
stated on the following page.
612
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Contrast in features of tJie pes in the middle Eocene groups
I. Palacosyopine group: Palaeosyops, Limnohyops
1. Tarsus broad, composed of anteroposteriorly deep elements,
correlated with more or less spreading digits.
2. Calcaneum with somewhat rounded or oval shaft of tuber
calcis moderately expanded at the summit.
3. Cuboid relatively broad.
4. Proximal facets on the metatarsals deeply extended antero-
posteriorlv.
II. Manteoceras-Dolichorhinus group: Manteoceras, Mesatirhinus, Dolichorhinus
1. Tarsus more shallow anteroposteriorly; more elevated in all
its elements. Digits less spreading distally.
2. Calcaneum with laterally compressed and distally truncate
tuber calcis.
3. Cuboid high, relatively narrow.
4. Metatarsals with proximal facets relatively shallow. Distal
facets more flattened.
PES OF THE PALAEOSYOPINE GEOTJP
Limnohyops (figs. 520, 530). — The adaptive charac-
ters of the pes of this animal, like those of the manus,
are primitive, mediportal, and mesatipodal. Digits
broader than in Mesatirhinus and stouter and heavier
than in Tapirus — that is, of less swift type. Astraga-
lus with elongate neck and vertically elongate sus-
tentacular facet, unlike the rounded facet of Palaeo-
syops. Calcaneum with tuber deep and expanding
at summit. Cuboid deep (shallow in Palaeosyops).
Entocuneiform very large. Mesocuneiform very small.
Ectocuneiform very large, quadrilateral (not vertically
extended) as in Mesatirhinus. Metatarsals of medium
length; Mts III abutting against cuboid, Mts II
abutting against ectocuneiform. Terminal phalanges
cleft and expanding distally somewhat as in Mesati-
rhinus but more robust.
Palaeosyops. — The most robust and graviportal pes
(figs. 520, 540) known among Eocene titanotheres.
Digits stout and widely spreading; prominent muscu-
lar rugosities indicating powerful flexor and extensor
attachments; astragalus readily distinguished by its
abbreviated neck, broad, shallow tibial trochlea, and
especially by the large and oval sustentacular facet for
the calcaneum, as well as by the broad cuboid facet;
calcaneum with obliquely placed tuber calcis and ex-
panded summit. Cuboid in two specimens very
short, thus Mts IV articulating with ectocuneiform;
the broad, abbreviated cuboid apparently a constant
character. Other 'elements of tarsus relatively broad
and low, deep anteroposteriorly; Metatarsals broad
and stout, broadening at the lower extremities;
ectocuneiform large and triangular; mesocuneiform
very small, with rounded edges; entocuneiform large,
of irregular shape. Terminal phalanges obtusely
rounded, very distinctive.
In the passage from Limnohyops to Palaeosyops we
again note the transition from a rather heavy medi-
portal type, of the proportions of the tapir, to a much
heavier (subgraviportal) type with broad, spreading
feet adapted to supporting the animal in swampy
ground; the reduction of the terminal phalanges points
to reduced hoofs, a possible amphibious adaptation.
PES OF THE MANTEOCEEAS-DOIICHOKHINlrS GROUP
As in the manus, we may distinguish the pes by the
relatively high and narrow proportions of the podial
elements. The mediportal Mesatirhinus presents a
wide contrast with the subgraviportal Manteoceras.
Mesatirhinus. — This mediportal pes (figs. 520, 569,
572) is readily distinguished by its relatively high and
narrow proportions. The angles and facets are all
sharply defined and sculptured. The bones of the
second row of tarsals and proximal facets of meta-
tarsals deep anteroposteriorly. Astragalus with elon-
gate neck, narrow, vertically produced sustentacular
facet for calcaneum; a narrow cuboidal facet. Cal-
caneum very distinctive, with its deep, laterally
compressed tuber and truncate distal extremity.
Navicular relatively deep and narrow. Cuboid, also
high, narrow and compressed in median line. Meso-
cuneiform and ectocuneiform distinguished by their
elevated and sharply quadrate form.
Manteoceras (fig. 557). — Astragalus only known.
Tarsus apparently of the same type as in Mesatirhinus
but relatively broader, judging from the astragalus.
Astragalus with a sharply defined tibial trochlea,
subelongate neck, vertically elongate, straight-sided
sustentacular facet for the calcaneum, of same type
as in Mesatirhinus but relatively broader; a broad
cuboidal facet (unlike Mesatirhinus).
Dolichorhinus (PI. XXXI). — Mesatipodal — that is,
metapodials relatively abbreviated, as in Manteoceras.
A full description appears upon a subsequent page.
Comparing these three forms with respect to the'
elongation of the metapodials, Mesatirhinus is mesati-
podal, Dolichorhinus is stiU mesatipodal but transi-
tional, whde Manteoceras approaches the brachypodal
condition. Another type of more elongated manteo-
ceratine foot (Am. Mus. 2352) has been discovered and
is described in full below. It is omitted here because
its generic association is somewhat doubtful.
SYSTEMATIC DESCRIPTIONS OF MIDDLE EOCENE
TITANOTHERES
SUBFAMILY PALAEOSYOPINAE
Mediportal and progressively graviportal titano-
theres of the lower and upper Bridger deposits and the
lower deposits of the Washakie Basin. Feet mesa-
tipodal to brachypodal. Ungual phalanges truncate
to rounded. Astragalus progressively widening.
Limnohyops
The skeleton of Limnohyops, so far as laiown, is
readily distinguished in all its parts from that of
Palaeosyops by its mediportal type — that is, by its
lighter construction. Yet the body was heavier and
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
613
A--V !
Figure 520. — Comparison of pes in four species of middle Eocene titanotheres
A, Limnohyops monocorms; B, Palaeosyops leiiyi: C, Mesatirhinus petersoni; T>, Mesatirliinust (Seep. 643.)
One-tbird natural size. The pes of TelmatJierium is unknown except in the upper Eocene T. ultimwrri'
that of Manteoceras is linown only from the astragalus.
614
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
the motions were slower than in either the American
or Asiatic species of Tapirus. Palaeosyops entered a
graviportal line of evolution, but Limnohyops was
more conservative. Its feet become mesatipodal
Limnohyops? monoconus, Am. Mus. 11699 (manus, radius, and
ulna, figs. 525 B, 527) ; Bridger B 2.
Lhnnohyo-ps? monoconus, Am. Mus. 11690 (complete hind limb,
fig. 529); Bridger B 1.
Measurements of limb hones of Limnohyops laiiceps
and L. monoconus, in millimeters
L. lati-
ceps,
Yale
Mus.
11000
(type)
Humerus, length
Radius, length I 230
Radius, breadth, I
proximal I 53
Radius, breadth.
distal
"Ulna, length
Carpus, width
Mtc II, height
Mtc III, height
Mtc III, distal width
(maximum)
Mtc IV, height
Mtc V, height
Femur, length
Tibia, length
60
307
L. ?monoconus
Am.
Mus.
11689,
Bridger
B2
295
230
58
55
312
"75
99
109
33
357
285
Am.
Mus.
11690,
Bridger
Bl
55
308
103
34
97
79
387
297
Figure 521. — Astragali of Eocene titanotheres
A, Lambdother'mm popoagkum; B, EotUanops borealis; C, Limnohyops monoconus: D, Palaeosyops
robustiis; E, Manteoceras manteoceras; ¥ and G, Mesatirhinus petersoni; H, Metarhinus cf. M.
earlei; I, DoUchorhinus hyognaikus. Ai, Bi, etc., front view; A2, B2, etc., back view. Astragalc-
calcanealfacets: ectal (ed), sustentacuiar (stw), and inferior (in/); astragalon»vicalar facet (m);
astragalocuboidal facet (c6). One-third natural size.
rather than brachypodal, and the limbs throughout
are narrower and more slender than those of Palaeo-
syops. The hand and foot were relatively elongate —
that is, they were appressed rather than spreading.
The ungual phalanges are truncate, expanding dis-
tally, rather than rounded and obtuse, as in Palaeo-
syops. These animals present many adaptive
resemblances to Mesatirhinus in the manteoceratine
subfamily, yet so far as known the parts are somewhat
heavier and more robust throughout.
The materials of Limnohyops are as follows:
Limnohyops laticeps, Yale Mus. 11000 (type skuU and parts of
skeleton, PL LVII, figs. 261 B, 264, 531, 532); Bridger D (?).
Limnohyops? monoconus, Am. Mus. 11689 (vertebrae, fore and
hind limbs, figs. 525 A, 530) ; Bridger B 2.
Two skulls of Limnohyops have been found
in Bridger B, namely, L. prisons and L.
monoconus, the latter an animal of consider-
able size. Parts of three skeletons in the
American Museum collection (Nos. 11689,
11699, 11690) belong, possibly, to L. mono-
conus. The generic reference to Limnohyops
is well assured; the specific reference is
doubtful. The ground for the generic refer-
ence of the skeletal specimens to Limnohyops
is their agreement in many details with the correspond-
ing bones of the type skeleton of L. laticeps, namely, (1)
the distal end of Mtc III of the type agrees with Mtc
III of Am. Mus. 11699; (2) the radius of Am. Mus.
11689 agrees in length and in details of character with
the radius of the type; (3) the ulna of Am. Mus.
11689 agrees closely with that of the type of L. laticeps,
especially in the rugose area of the olecranon.
Limnohyops monoconus?
Incomplete skeleton from Bridger B 2, Grizzly Buttes, Bridger Basin, Wye,
Am. Mus. 11689
The material includes the atlas, a cervical, several
anterior dorsal vertebrae, the humerus, radii, ulnae,
incomplete manus and pes, femur, tibia (figs. 510,.
EVOLUTION OF THE SKELETON OF EOCENE AND QLIGGCENE TITANOTHERES
615
Figure 522. — Calcanea of Eocene titanotheres
A, Lambdotkerium popogagicum; B, Eotitanops borealis; C, Livinohyops monoco-
nusf; D, Palaeosyops robustus; E, Mesatirhinus peter sonif; F, Metarhinus cf. M.
earlei; G, f M esatirhinus sp. (Washakie). Astragalocalcaneal facets: ectal (ect),
sustentacular (sus), and inferior (inf). One-third natural size.
Figure 524. — Principal measurements of the carpus and tarsus
A, Tarsus, transverse and vertical; astragalus, vertical (interior and median); calca-
neum, vertical; metatarsal III, vertical and transverse, maximum v/idth near
distal end. B, Carpus, transverse and vertical; lunar, transverse and vertical;
metacarpal III, height and maximum width near distal end.
C
Figure 523. — Left ectocuneiform tarsi of lower and middle
Eocene titanotheres
A, Eotitanops borcaUs; B, Mesatirhinus petersoni; C, Limnohyops monoconus; D,
Palaeosyops rohustus. Facets for second metatarsal (II), third metatarsal (III),
navicular (nf), cuboid (c&).and mesocuneiform (cn^). One-half natmal size.
Figure 525. — Humerus, radius, and ulna of Limnohyops
monoconus?
Am. Mus. 11689. Ai Left humerus, radius, and ulna, outer side view; A2, left
radius, and As, left humerus, both front view; B, right humerus, radius, andulna,
the last two partly restored from Am. Mus. )1699, front view. One-sixth natural
size. (Compare fig. 527.)
616
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
511, 513, 514, 520-522, 685). Measurements of the
limb bones are given above.
Adaptive distinctions from Palaeosyops. — This mate-
rial enables us to further distinguish LimnoTiyops from
Palaeosyops. Manus and pes mesatipodal rather
than brachypodal; all limb bones more slender;
humerus relatively shorter; ulna more curved, with
distinctive olecranon process; manus narrow; lunar
Figure 526. — Left manus, radius, and ulna of
Mesatirhinus pelenioni (doubtfully referred)
Princeton Mus. 10013, upper Bridger, placed here for com-
parison with the manus of Limnohyops (fig. 527). Aj,
Front view of manus; As. top view of carpus; A3, inner
side view of carpus; Bi, distal view of radius and ulna; Ba,
proximal view of radius. One-third natural size.
more acutely wedge-shaped distally. Magnum not
so wide; metacarpals more slender, femur with more
slender, curved shaft. Astragalus more slender,
with narrow sustentacular and cuboidal facets; cal-
caneum with narrow sustentaculum; metatarsals
narrow, Mts V more curved, metatarsals not expand-
ing distally; ungual phalanges truncate and square
distally.
Palaeosyopine syngenetic characters. — Among the
more obscurely but syngenetically important char-
acters tending to ally this animal to Palaeosyops are
an astragalus having in common the following
peculiarities: A pit for a ligament on the internal or
tibial face just below the trochlear keel, a rather
sharp extension of the superior edge of the navicular
facet, a prominent protuberance near the distal end
of the tibial face, an inward projection of a sinus or
fossa tending to separate the sustentacular facet
from the well-developed facet for the tibial sesamoid.
Other resemblances with the Palaeosyopinae are seen
in the subglobose shape of the distal facets of the
metapodials, in the marked anteroposterior depth of
the scaphoradial and other carpal facets, in the
depth of the proximal metapodial facets, and in the
femur with patellar facet facing anteriorly.
Distinctions from the Manteoceras-Dolichorhinus
group. — Limnohyops is separated from Mesatirhinus
by the following characters: Limb bones of more
graviportal type; humerus longer, radius stouter,
more curved; ulna more curved and with larger
olecranon; manus somewhat broader and shorter
throughout; femur flatter, femur and tibia a little
stouter but of about the same relative length; astra-
galus broader, sustentacular facet farther in toward
the middle of the bone; sustentacular facet broader,
not so straight sided, forming a much more open angle
with the navicular facet, inferior astragalocalcaneal
facet smaller; a pit on the internal or tibial face of the
astragalus just below the trochlear keel; metatarsals
a little shorter and broader. Notwithstanding these
differences there are many general adaptive resem-
blances to Mesatirhinus, especially in the femur,
tibia, and metatarsals. The differences, however,
appear to indicate generic separation.
Limnohyops is analogous to Manteoceras, especially
in the general characters and proportions of the
Figure 527. — Manus, radius, and ulna of Limnohyops mono-
conus (doubtfully referred)
Am. Mus. 11699, Bridger B 2. Ai, Right manus, front view; A2, right carpus, out-
side view; As, phalanges of median digit; At, distal view of radius and ulna. One-
third natural size. (Compare fig. 525.)
humerus, tibia, astragalus, and pes — so much so, in
fact, that it was formerly referred provisionally to
that genus, from which, however, it is distinguished,
so far as known, by the following characters: Scaph-
oid deeper anteroposteriorly, trapezoid facet more
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
617
oblique, magnum facet narrower, lunar facet narrower;
distal or phalangal facets of metacarpals and metatar-
sals more globose, less flattened, proximal facets of
same widely truncate posteriorly; tibia relatively
longer with narrower proximal end; astragalus with
narrower cuboid facet, navicular facet not so deep
anteroposteriorly, inner keel of astragalotibial trochlea
sharper, prominence on the distal part of the inner or
tibial surface not ending in a budlike tuberosity,
sustentacular facet less straight-sided.
Forearm and manus from Bridger B 2, Grizzly Buttes west, Bridger Basin, Wye;
Am. Mus. 11699. Limnohyops ref.
.Figure 527. (For measurements see above)
This well-preserved specimen resembles in general
appearance the preceding one (Am. Mus. 11689) but
differs in the following particulars: The radius is a
little more slender; the olecranon of the ulna is deeper;
the manus is smaller but agrees in so many close
details with the other manus (Am. Mus. 11689) that
there can scarcely be any doubt that the two belong
to the same genus — namely, Limnohyops.
The manus accordingly offers some further points
of contrast with that of Manteoceras — namely, in
the magnum the front face is not so sharply polyg-
onal, the posterosuperior head or eminence is much
narrower and extends obliquely backward and down-
ward, whereas that of Manteoceras is subtruncate pos-
teriorly and sharply ridged superiorly; the posterior
hook of the magnum is much more slender and ends
postero-inferiorly in a bluntly oval pitted surface;
the facet for the third digit, Mtc III, is deeper antero-
posteriorly and narrower posteriorly, that of Manteo-
ceras being subrectangular in general outline; the
unciform is narrower transversely and deeper verti-
cally, its supero-external or cuneiform facet is less
extensive transversely and narrower externally, it is
less concave in front view, its supero-internal or lunar
facet is less wide, especially posteriorly, all the ridges
between adjacent facets are less angulate; the postero-
external protuberance of the unciform is much more
pointed posteriorly; the cuneiform carpi are much
narrower transversely than in Manteoceras and reveal
many striking detailed differences. Some, or even
most, of the foregoing characters of the manus in
question may be primitive characters which have
been lost in Manteoceras in adaptation to the progres-
sive broadening of the manus; but if the manus in
question belonged to a direct ancestor of Manteo-
ceras manteoceras of the succeeding horizon we should
expect it to foreshadow that form a little more defi-
nitely, and the very well-marked differences indicate
again that it belongs to some other genus.
Analogy to Mesatirhinus. — As compared with Mesa-
iirJiinus (fig. 526) this manus exhibits a rather strik-
ing general resemblance; but the carpals are broader,
the scaphomagnum articulation is more oblique, the
metacarpals are broader, and the ungual phalanges
are larger and not so sharply flaring and truncate
distally.
Conclusions. — This manus, although it is more
slender than that of Palaeosyops, shares with it certain
palaeosyopine characters in which it contrasts with
Mesatirhinus and Manteoceras, such as the greater
anteroposterior depth of the scaphoid, the angular
antero-internal border of the cuneiform, the pointed
rather than spatulate hook of the magnum, the
broader carpals, metacarpals, and phalanges.
Hind limb from Bridger B 1, lower Cottonwood Creek, Bridger Basin, Wyo.;
Am. Mus. 11690. Limnoiiyops ref .
Figures 516, 518, 519, 523, 686. (For measurements see above)
The material includes only the rather well pre-
served left hind limb, namely, the femur, patella,
tibia, fibula, and pes.
As compared with the hind limb of Am. Mus.
11689, described above, the present specimen differs
in its larger size and somewhat stouter proportions,
especially of the femur; the metatarsals are noticeably
longer and have straighter sides; in the astragalus the
sustentacular facet ex-
tends a little further dor-
sad, the ectal facet is less
deeply concave, the distal
calcaneal facet is larger,
and corresponding differ-
ences occur in the calcane-
um. These differences are
accompanied by so many
detailed resemblances
(in contrast with other
genera) that it appears
probable that the hind
limb in question belongs in
the same genus (Limnohy-
ops) with Am. Mus. 1 1689,
11699, above described,
though possibly to a differ-
ent species of that genus.
As compared with other Palaeosyopinae the present
specimen offers many resemblances and a few rather
decisive differences: the femur, though smaller than
that of P. major, does not differ greatly in its propor-
tions and offers no clear-cut distinctions; the pes is
noticeably higher and more slender; the astragalus
is narrower with a narrower neck, more straight-
sided sustentacular facet, narrower cuboid facet
and sharper internal keel on the trochlea.
Distinctions from Manteoceras-Dol ichorhinus
group. — From Mesatirhinus the specimen differs in its
larger size, more robust femur, tibia, and pes, distally
diverging vertical sides of Mts III. The astragalus
is wider, its ectal facet shallower and not produced
downward at the lower internal corner, the susten-
tacular facet was probably broader, not so straight-
FiGURE 528. — Right scaphoid
of Palaeosyops sp. (A) and
Limnohyops monoconus (B)
Front (Ai, Bi) and upper (As, B2) sur
faces. One-half natural size.
618
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
sided, and it formed a more open angle with the
cuboid facet; the calcanea show corresponding differ-
ences. But these differences are accompanied by
many general resemblances that indicate the annec-
tent or primitive character of the pes in question.
Comparison with Manteoceras is made difficult by the
scantiness of the material referable to that genus. The
femur is of nearly the same length as in the young male
Manteoceras (Am. Mus. 1587), but the third trochanter
Figure 529. — Left hind limb of Limnohyops monoconus (doubt-
fully referred)
Am. Mus. 11690; Bridger B 1. A', Front view; A^, outer side view. One-sixth
natural size.
seems smaller; the tibia is absolutely and relatively
longer, its proximal end is narrower, its distal end is
broader; the patellar facet is decidedly longer and
more anterior than in Manteoceras. The astragalus,
as compared with that of the young Manteoceras (Am.
Mus. 1587), offers the following differences: The inter-
nal trochlear ridge is angulate rather than broadly con-
vex; the ectal facet is shallower, less produced on its
lower internal border; the sustentacular facet is much
longer: the cuboid facet is decidedly smaller; the navic-
ular facet not so deep anteroposteriorly. As compared
with the astragalus of the old female Manteoceras?
(Am. Mus. 12204) most of the above-described differ-
ences also hold good, save that the siz(3 in Am.
Mus. 11690 is considerably greater and the navicular
facet seems proportion-
ately deeper rather than
shallower.
Conclusion. — In brief,
the hind limb under
consideration appears
to belong neither with
Palaeosyops, Mesatirhi-
nus, nor Manteoceras.
On the other hand, it
appears not to differ gen-
erically from Am. Mus.
11689, described above,
and together with that
specimen it may repre-
sent a Bridger B mem-
ber of the Limnoliyops
phylum.
Limnohyops laticeps
Many parts of the
skeleton of i. laticeps are
fortunately preserved
with the type skull in
the Yale Museum (No.
11000). The geologic
level is probably Bridger
C or D, at Henrys Fork
Hill, Bridger Basin, Wyo .
Cervicals. — The axis and four other cervicals (C. 3,
C. 5, C. 6, C. 7) are preserved; the fourth cervical is
missing. The epiphyses are mostly detached, indicat-
ing a juvenile condition.
The axis (C. 2) and third
vertebra (C. 3) especially
have large inferior keels or
hypapophyses. The third
to sixth cervicals exhibit
broadly depressed inferior
lamellae. The seventh
(C. 7) is irnper forate
exhibiting narrow, rodlike
transverse processes with-
out inferior lamellae.
Dorsals. — The median
dorsals exhibit centra
angulate but not keeled
inferiorly. In one of the
posterior dorsals the cen-
trum is slightly keeled and
the spine is elevated (112
mm.) above the base of the centrum.
Lumhars. — The lumbars preserved are apparently
the second and the fourth (L. 2, L. 4); the first, third
Figure 530. — Right pes of lAm-
nohyops monoconus?
.\m. Mus. 11689; Bridger B 2. Ai, Front
view; A2, outer side view of tarsus; A3
phalanges of median digit. One-third
natural size.
FiGuiiE 531. — Ventral surface
of sacrum of Limnohyops
laticeps
Yale Mus. 11000 (type); Bridger C or
D. One-third natural size.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
619
and fifth lumbars are missing. They exhibit increas-
ingly broad centra and moderately broad transverse
processes. The zygapophyses are vertically placed,
and unlike those of Palaeosyops are slightly if at all
revolute.
Sacrals. — The most distinctive character of the
sacrals is that the sacrum includes five vertebrae, the
fifth (S. 1) being due to the coalescence of an anterior
caudal. They measure 171 millimeters anteropos-
teriorly. The first and second sacrals and the anterior
portion of a third sacral enter into union with the ilium.
Pelvis. — The innominate bones are preserved almost
entire (fig. 532). The left os innominatum measures
406 millimeters anteroposteriorly. The crest of the
ilium measures 203 millimeters transversely; the
superior border is thin and indented in the median
portion or crista iliaca. The rugose tuber sacrale
is narrower (80 mm.) than the rugose tuber coxae
(140 mm.). The elongate proportions of the innomi-
nate bones are partly due to the juvenile nature of
this individual.
Fore limb oj type. — The proximal and distal portions
of the right humerus are preserved. The head and
Figure 532. — Right os innominatum of Limno-
hyops laticeps
Yale Mus. 11000 (type); Bridger C or D. One-.sixth natural size.
greater tuberosity measure 122 milHmeters (antero-
posterior); the head measures 65 (transverse); there
is a wide and deep bicipital groove ; the proportions are
somewhat altered by crushing; the distal end of the
humerus measures 68 (transverse) ; the total width of
the distal articular surface is 55.
The radius and ulna are complete. The radius is
distinguished posteriorly by a characteristic median
groove; it is 230 millimeters long; its humeral facets
measure 53 millimeters transversely and 27 antero-
posteriorly.
The ulna is 307 millimeters in length; its distinctive
character (see fig. 511) is the downward extension on
the outer side of the olecranon of the rugose crest for
the triceps muscle.
Palaeosyops
The Eocene titanothere skeleton is best known in
members of the genus Palaeosyops, which was the
largest and the most heavily built titanothere of that
time, adapted to slow locomotion on soft ground along
water borders and in marshes and swamps. We have
the nearly complete skeleton of P. leidyi from the
upper Bridger and parts of the more progressive P.
rohustus and of the ancestral P. major. A titanothere
family likeness is seen throughout the axial and
appendicular parts, but generic and subfamily differ-
ences, closely shared with Limnohyops, are apparent
throughout.
Up to the end of the lower Bridger deposition we
may safely select the largest and most massive titano-
FiGUKE 533. — Pelvis of Palaeosyops major
Am. Mus. 13116; Bridger B 3. A, Ventral aspect: B, dorso-superior
aspect. One-sixth natural size.
there bones as belonging to Palaeosyops. The gravi-
portal tendency is especially manifest in the hind
limb, with its elongate femur and abbreviate tibia
and slight angulation at the knee joint. The broad
and spreading foot bones of both manus and pes
are especially distinctive. The manus is functionally
tridactyl rather than tetradactyl, owing to the reduced
condition of the fifth digit, which is set apart from the
other digits in a manner peculiar to this genus. The
620
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
/!
terminal phalanges are obtusely rounded and not
deeply cleft distally, indicating the presence of
imperfect hoofs. Additional evi-
dence of water-loving or semiaquatic
habits is found in the weak spines of
the dorsal and lumbar vertebrae,
which are analogous to those in
Coryphodon and Metamynodon. The
dentition points to habits of feeding
on the succulent plants which are
characteristic of the borders of
streams. The skeleton as a whole
shows far more aquatic adaptations
than that of the tapirs.
Palaeosyops major
Palaeosyops major is represented
by some well-preserved limb bones
and a pelvis associated with the large
skull (Am. Mus. 13116) from Bridger
B 3. The large size of these bones,
coming as they do from so low a
level, is very noticeable. They even
exceed in total length those of P.
leidyi (Am. Mus. 1544), but the long
bones are more slender. The most
characteristic feature of the pelvis
(fig. 533) is the narrowness and ap-
parently uniform convexity of the
superior or anterior border of the
ilium as compared with its expansion
in subsequent stages of the develop-
ment of the titanotheres. This nar-
rowness is a primitive character, cor-
responding with the low geologic
level (Bridger B 3) at which this
specimen was found. Other chief
generic characters exhibited in this
skeleton are the following: Radius
(fig. 510) strongly arched forward
and having a deep groove for the ex-
tensor carpi radialis muscle; ulna
(fig. 511) without the incurved olecranon of MesatirJii-
nus; femur with a
straight, flattened
shaft, and a pa-
tellar groove pre-
senting anteriorly ;
tibia with an out-
curved cnemial
crest. The short-
necked astragalus
and the stout cal-
caneum are like-
wise of Palaeo-
syops type.
The detailed
measurements of
this important skeleton as compared with the com
posite P. leidyi (fig. 536) are as follows:
<fi)-\
\.maLv
(as)
Figure 534. —
Right femur and
tibia of Palaeo-
syops major
Am. Mus. 131 16; Bridger
B 3. One-sixth nat-
ural size.
Figure 535. — Astragalus and calca-
neum of Palaeosyops major
Am. Mus. 13116; Bridger B 3. A, Left astragalus
and calcaneum; B, left calcaneum with astragalus
removed. One-third natural size.
Measurements of skeletal parts of Palaeosyops major and P.
leidyi, in millimeters
Femur, length
Femur, distal breadth
Femur, breadth across head and great
trochanter
Tibia, length (middle)
Radius, length
Ulna, length
Astragalus, vertical inner face
Astragalus, proximal width
Calcaneum , vertical length
Calcaneum, greatest width
Mts II, length
Mts III, length
Mts III, distal width, transverse
Pelvis, total length of os innominatum.
Pelvis, width of crest of ilium
p. major,
Am. Mus.
13116
(Bridger B 3)
435
100
127
335
280-290
378
61
53
119
63
126
136
43
448
"220
P. leidyi.
Am. Mus.
(Bridger C, D)
370
93
132
290
235
315
57
49
97
62
102
110
42
'46&
'265
» Estimated. ' Am. Mus. 2348.
Palaeosyops leidyi
Type locality and geologic horizon. — Bridger Basin,
Wyo.; Bridger formation, level Bridger C and D.
Slceletal characters. — The first description and prelim-
inary restoration of the skeleton of this species (under
the name "Palaeosyops paludosus") was published by
Earle in his memoir of 1892 (1892.1, p. 314). In this
restoration the forefoot of Mesatirhinus was errone-
ously associated. (See fig. 86, p. 151.)
The mounted skeleton of Palaeosyops leidyi in the
American Museum of Natural History (No. 1544) is a
composite one, reconstructed from a number of speci-
mens belonging to difl'erent individuals of at least two-
species {P. leidyi, P. robustus) and two geologic levels,
collected by the American Museum expedition of 1893
under Dr. J. L. Wortman. The well-preserved skull
and lower jaws, the cervical and first two dorsal verte-
brae, and the fore limbs down to the manus are from
one individual, an old male, No. 1544, the type of
P. leidyi. The scapulae, ribs, dorsals 4-11, 14-17,
caudals 8-10, and four sternals were supplied from
No. 1580. The right carpus, belonging to the prin-
cipal specimen (No. 1544), warrants the completion
of the rest of the feet from other specimens, cliiefly
No. 1550. The left hind limb was made up from
Nos. 1582 and 1562; certain of the dorsals and lum-
bars were supplied from Nos. 1593, 5177, 5158. The
foregoing specimens, after careful study, were found to
agree very closely in size and other characters, and
are probably all referable either to P. leidyi or to the
closely allied P. robustus. But the sacrum and pelvis,
No. 2348, from the Washalde Basin, are of somewhat
doubtful reference to this genus and may belong to
the genus Limnohyops. Certain remaining parts
(such as caudals 1, 2, 11-18, the right femur, the tibia
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHEEES
621
and fibula, the right cuboid and cuneiforms, most of
the unguals, and parts of many ribs and vertebrae)
were restored in plaster. (See fig. 536.)
The formula for the vertebral column probably but
not certainly is cervicals 7, dorsolumbars 21?, sacrals 4.
A careful study indicates that there were 17 dorsals
and at least 3 lumbars. The presacral region is rather
short, the neck decidedly so, the back moderately so.
In adaptation the short neck, high anterior dorsal
spines, rather slender mid-dorsal spines, elevated but
not very broad lumbar spines, and long ribs indicate
a very deep-chested animal of about the same pro-
portions as Tapirus indicus, but considerably larger
(perhaps one-fifth) and slower in its movements, with
decidedly less development of the extensor muscles
of the back, as indicated by narrower dorsal spines
throughout.
lamina of neural arch, thin, elevated neural spine re-
stored at the summit. Transverse measurement of
neural spine near the base 11 millimeters, anteropos-
terior measurement 23.
All the upper portions of the neural spines of the
anterior dorsals are restored except that of D. 4,
which is complete. The base of the spine of D. 1 is
laterally compressed and much more extended ante-
roposteriorly than that of C. 7. Measurements, 29
millimeters anteroposterior, 21 transverse (at base).
In D. 2, D. 3 the spines are very stout and transversely
extended at the base. In the succeeding dorsals the
neural spines become gradually less expanded at the
base as they recede in height; but the basal trans-
verse width (35 mm.), the deep posterior concavity,
and the moderately limited anteroposterior diam-
eter (24 mm.) are decidedly the characteristic
Figure 536. — Composite mounted skeleton of Palaeosyops leidyi
Including skull, neck, and parts of the fore limb of the type of P. le
eitlier to P. leidrji or to the closely related P. roiustus.
The atlas and axis belonging with skull No. 1544
exhibit the following characters: Atlas with large ver-
tebrarterial canal traversing inferior portion only of
transverse process (47 mm. broad), which is indented
but not perforated to the front as in LimnoJiyops
laticeps; first cervical nerve issuing from anterior por-
tion of the neural arch; broadly rugose keel on lower
posterior portion of the centrum; transverse measure-
ment of anterior cotyli 102 millimeters, vertical meas-
urement 40, transverse atlas (estimated) 223. Axis
also with broadly rugose inferior keel, neural spine
moderately extended anteroposteriorly, abutting an-
teriorly against upper portion of neural arch of atlas.
Cervicals 3-6 with neural spines and pleurapophyses
(cervical ribs) partly restored; pre- and postzyga-
pophyses placed in decidedly oblique planes facing
inward and outward respectively. C. 7 with narrow
dyi (Am. Mus. 1544) and material from other individuals referred
Upper Bridger C and D. One-fifteenth natural size.
features of all the dorsal spines as far back as
D. 12 inclusive; in D. 13 the neural spine begins to be
more laterally compressed or flattened and of some-
what greater anteroposterior extent. In D. 4 and D. 10
we have a typical spine complete to the tip. These
proportions of the neural spines somewhat resemble
those of Amynodon and are quite distinct from those
in the Oligocene titanotheres Megacerops and Meno-
dus, as well as from those observed in the horse,
tapir, and rhinoceros, in all of which the neural spines
are more extended anteroposteriorly than transversely.
The anterior dorsal spines are also higher and less
elongate anteroposteriorly than in Dolichorhinus.
The planes of the zygapophyses are significant as
follows: Decidedly oblique planes of the prezyga-
pophyses and postzygapophyses characterize D. 1 and
D. 2, both vertebrae belonging with the skull (No. 1544)
622
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
of P. leidyi ; this is a point of generic distinction between
the Oligocene titanotheres. In D. 3 the planes of the
pre- and postzygapophyses suddenly become more
horizontal; this horizontal position also characterizes
D. 4-D. 11. In D. 14-D. 17 and in all the lumbars
the postzygapophyses become cylindrical or partially
revolute and are firmly embraced by the deeply con-
cave prezygapophyses as first observed by Earle.
The relations of the vertebrae to the capitula and
tubercula of the ribs are clearly shown in the vertebrae
belonging with the skull No. 1544, and partly in those
belonging to the other individual, No. 1580. All the
ribs articulate by prominent and more or less separate
anterior and posterior capitular facets with the pos-
terior and anterior vertebral faces respectively. Ribs
Figure 537. — Manus of Palaeosyops leidyi
Composite fore foot from Am. Mus. 1544 and 1550 (mounted with 1544).
Bridger C or D. Ai, Front view; A2, phalanges of the third digit. One-
third natm-al size.
1-13 exhibit projecting tubercula; in ribs 14-17 the
tubercula become more sessile and reduced. The con-
cave to flat parapophysial facets for the tubercula are
decidedly prominent in the anterior dorsals but become
less prominent and distinct or pediculate in the pos-
terior ones, the tubercular facet of the ribs being
widely separated from the capitular facets. The para-
pophysial processes for the articulation with the tuber-
cular of the ribs are very stout and downwardly ex-
tended, more or less triquetrous, in vertical section,
the anterior face being rounded and widely extended
outward; in the anterior dorsals the transverse meas-
urement across the parapophyses is 117 millimeters
in D. 1, 114 in D. 4, and 99 in D. S. These processes
gradually subside (to 60 mm. tr. in D. 17) in the
posterior dorsals.
The series of ribs associated with the scapulae (Am.
Mus. 1580) indicate a deep and powerful chest; the
anterior ribs are broad and flattened externally, the
seventh rib assumes a more rounded or trihedral
section, which also characterizes the eighth, ninth, and
tenth; the eleventh to the seventeenth are smaller,
with a flattened oval section; the strength of the chest
is attested by the presence of prominent tubercular
facets on the posterior ribs. The middle ribs ver-
tically measure 502 millimeters, not allowing for
curvature; the last ribs measure 260; the depth of the
first rib is approximately 260.
Four sternals (No. 1580) are completely preserved
and a portion of the most anterior, there being five
preserved in all, though the exact number is not known.
The xiphisternum (st. 5) is very shallow, expanding
anteriorly and posteriorly, with a linear measurement
of 84 millimeters and a transverse measurement pos-
teriorly of 51. The three midsternals are laterally
compressed, with the following measurements: St. 4,
42 millimeters (anteroposterior) by 22 (transverse) ;
St. 3, 45 by 28; st. 2, 50 by 22. The presternum (st. 1)
is also laterally compressed; the anterior portion is
not preserved. The midsternebrae have the same
form in Oligocene titanotheres.
The centra of C. 2-4 exhibit broadly rugose pos-
terior keels, which die out in C. 5. The centra of
C. 5-D. 2 are slightly convex inferiorly. From this
point backward the centra become somewhat laterally
compressed, especially toward the lumbars, and are not
keeled inferiorly; D. 5 measures 37 millimeters, D. 7,
39; D. 9-D. 15, about 42; D. 17, 48; L. 1, 40; L. 2, 45.
The lumbar vertebrae are estimated in the restora-
tion as three in number, but there may have been
four lumbars and but sixteen dorsals. The trans-
verse processes are estimated at 174 millimeters; they
are of only moderate breadth.
The sacrals are not preserved. They are present,
however, in a specimen referred to Palaeosyops sp.
(see below).
The three anterior caudals are restored, the neural
arch disappears with the supposed ninth, the centra
of the fourth, fifth, sixth, ninth, tenth, eleventh
(estimated) are preserved (Am. Mus. 1544).
Comparison with the vertebral column of Brontops
so far as known serves to emphasize the inference
based upon the study of the skull that there are few
resemblances except in family characters and many
differences, and that Palaeosyops is not in the line of
ancestry of this Oligocene genus.
The structure of the scapula in this species is known
from that of another individual, Am. Mus. 1580,
associated with this mount. The general propor-
tions of the scapula are breadth 260 millimeters,
height 345, and, like the proportions of the limbs,
denote a slow-moving animal; the total areas of the
prescapular and postscapular fossae are approximately
equal; the prescapular fossa (origin of supraspinatus
muscle) is subrectangular and exhibits a marked down-
ward extension of the anterior border, an especial
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
623
characteristic of all titanotheres, to a point within
115 millimeters of the glenoid border; at this point
the fossa measures 104 millimeters in front of the
spine; the postscapular fossa (origin of infraspinatus,
teres major, teres minor muscles), on the other hand,
is triangular, the broadest point, 131 millimeters, being
near the superior border; the posterior border rises
thus in a nearly straight line from the glenoid to the
superior border and then sweeps around in a uniformly
convex curve to the point above the anterior notch.
This scapula conformation is remotely paralleled by
that of Tapirus indicus, which, however, differs in the
greater depth of the supracoracoid notch; it also
resembles that of certain early rhinoceroses {Caenopus
occidentalis) , which, however, lacks the anteroposterior
breadth; it is also somewhat similar in type to that
of Rhinoceros sondaicus, differing again in the greater
breadth and the absence of retroversion of the spine.
These comparisons are naturally valuable not as in-
dicative of relationship but as pointing to analogous
development of the shoulder muscles and gaits of these
animals; they are distinctively noncursorial. The
breadth above the glenoid, including the coracoid
process, is 87 millimeters, the width of the neck is 80.
The spine rises gently from above the neck with a
re trover ted border reaching its widest point 210 milli-
meters above the glenoid and 144 millimeters below
the superior border; at this point the spine rises 50
millimeters above the postscapular fossa.
The humerus of P. leidyi found with the skull (No.
1544) is intermediate in its robust proportions, which
are between those of Tapirus indicus and those of
Rhinoceros (Dicerorhinus) sumatrensis. In general
the muscular processes are about as strongly devel-
oped as in Tapirus indicus. The proportions of the
bones of the fore limb are accurately determined from
bones of both sides found with the skull. The upper
arm (humerus) (310 mm.) is considerably longer than
the forearm (ulnoradius) (235 mm.), as it usually is in
slow-moving, animals. The proportions of the humerus
are similar to those in Tapirus indicus, the approxi-
mate breadth across the great tuberosity being 103
millimeters and the total distal breadth 92; the lesser
tuberosity (insertion of subscapularis') and the greater
(insertion of supraspinatus and infraspinatus) are
moderately prominent; the greater tuberosity rises
into a platelike crest very characteristic of the titano-
theres; it sends in an internal hook; the bicipital groove
is thus broad and shallow; the highest point of the
deltoid crest (insertion of pectoralis major and del-
toid) is 119 millimeters below the head, it is slightly
retro verted; below this point it subsides gradually
into the shaft; the supinator ridge is sharply defined,
but not hooked, its summit being 109 millimeters above
the trochlea; the entepicondyle or internal eminence is
relatively low and obtuse; distally the radial trochlea
of the humerus is divided unequally into a smaller,
i01959— 29— VOL 1 i3
less convex internal portion and a larger, more de-
.cidedly convex external portion (the tuberculum).
Earle (1892.7, p. 357) has directed attention to the
fact that this conformation of the trochleae differs
somewhat from that of other Perissodactyla; it is
correlated with the more elevated and plane internal
himaeral facet and the more depressed and concave
external humeral facet of the radius.
The elevation of the inner side of the front face of
the radius is characteristic of titanotheres but is seen
also to a considerable extent in Amynodon. The width
of the humeral trochlea is 63 millimeters. There are
broad and deep anconal and supratrochlear fossae
but no perforation; the upper end of the radius
measures 63 to 67 millimeters transversely.
The shaft of the radius arches well away from the
ulna and expands distally to a width of 62 to 67 milli-
meters. There is a distinct extensor groove in the
anterior distal face.
The ulna has a stout olecranon and straight tri-
hedral shaft well separated from that of the radius,
with a broadly transverse section, of which one face
presents directly forward and is hollow or grooved
above and flattened below.
The right carpus belongs to the same individual as
the skull (Am. Mus. 1544). The specific character of
the carpus is its great breadth (91 mm.) as compared
with its maximum depth (54 mm.) measured from the
lunar to the bottom of the unciform. Similar broad
and low proportions characterize each element. The
lunar is especially distinctive, measuring 36 milli-
meters transversely, 31 vertically, with an exception-
ally broad facet for the magnmn (fig. 537); a small
trapezium was present but is not preserved in this
specimen. In this, as in all other titanotheres, there
were large facets on the inner side of the magnum and
unciform for Mtc II and Mtc III, respectively.
The metapodials (Am. Mus. 1544, 1550) are rela-
tively broad and distally expanded; the lengths or
vertical diameters are, Mtc II, 107 millimeters;
Mtc III, 112; Mtc IV, 97; Mtc V, 75. Mtc III has
a vertical diameter of 112 millimeters and a transverse
diameter distally of 46; the shaft of each has a rela-
tively broad and shallow flattened section, this being
a characteristic family feature. The metapodial dis-
placement is very marked, Mtc II and Mtc III
abutting widely against the magnum and unciform,
respectively. The proximal portion of the shaft of
Mtc II presents a small facet to which the trapezium
was attached. The three phalanges of D. Ill meas-
ure 39, 26, and 23 millimeters in length, respectively.
The distal phalanges are broadly rugose (tr. 35
mm.) with a deep medial cleft; the upturned distal
facets of the second phalanges indicate that when
the foot was drawn backward the unguals were more
sharply extended on the second phalanges. Even
with this reservation the fore foot was broad and
624
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
spreading distally, the transverse measurement across
the inner and outer toes resting on the ground being
240 millimeters. Palaeosyops thus presents a very
wide contrast to the relatively narrow fore foot of
MesatirMnus. The contemporary species of rhinoc-
eros {Amynodon) is also distinguished by the height
Figure 538. — Pelvis of LimnoMjops (doubtfully referred)
Am. Mus. 2348, superior view. Washalde ?A. One-sixth natural size.
and lateral compression of all the elements of the
carpus and metacarpus and by the greatly reduced
terminal phalanges.
The pelvis and sacrum in this mounted specimen
(Am. Mus. 2348) are from the Washakie Basm, Wyo.
As noted above it may belong to Liimiohyops, because
the supra-iliac border does not exhibit the uniformly
convex arch seen in Palaeosyops. The sacrum, how-
ever, differs from that of the type of Limnohyops in
being composed of four vertebrae, the last two of
which are clearly caudosacrals. The sacral plates ex-
tend deeply down the inner sides of the ilia, the total
oblique measurement from the summit of the sacral
spines to the bottom of these plates being not less than
170 millimeters. The greatest width of the plate an-
teriorly is 136 millimeters as compared with the total
length of the four sacrals, 159 millimeters; the spines
are moderately broad, thin, and coalesced at the sum-
mits. The fourth or posterior sacral extends the plate
well beyond the posterior superior crest of the ilium.
In general the pelvis (Am. Mus. 2348) is decidedly
elongate as compared with that of the Oligocene titano-
theres. The total length (465 mm.) is only 23 milli-
meters less than the total breadth (488 mm.). The
superior crest of the ilium, actually measuring 265
millimeters transversely, has a uniformly concave
transverse surface with a moderately thickened su-
perior border slightly sigmoid — that is, convex in its
outer half and straight or slightly concave in its inner
half — as it rises to its convex junction with the sa-
crum. The distinctive feature of this pelvis is the
slightly concave "crista iliaca" between the broadly
rugose tuber coxae and tuber sacrale. The conforma-
tion of this border is analogous to that of Rhinoceros
sondaicus and of other quadrupeds the ilia of which
are transitional between the purely cursorial type of
dium seen in Equus and the weight-carrying type with
a uniformly convex superior border seen in Uinta-
therium, ElepJias, and Menodus. The neck of the
ilium measures 67 millimeters transversely. The me-
dian border of the posterior pelvic opening is marked
by an elongate sharp ridge constituting the spine of
the ischium (origin of gemellus superior, coccygeus,
and levator ani muscles) opposite the acetabulum,
behind which is the well-defined lesser ischiadic notch.
Behind this the ischiadic border is rounded until it
passes into the prominent ischiadic tuberosity. The
vertical depth of the ischium at this point is 117 milli-
meters. The pubo-ischiadic symphysis measures 174
millimeters anteroposteriorly. The obturator fora-
mina are elongate, oval, measuring 87 millimeters
(anteroposterior) by 49 millimeters (transverse), sepa-
rated by 19 millimeters, the narrowest point of the
symphysis.
The hind limb has a total length from the head of
the femur to the tip of the median phalanx of D. Ill
of 940 millimeters as compared with 810, the total
length of the fore limb, and 500, the expansion of the
ilia. These figures give an idea of the chief propor-
tions of the animal.
The femur (Am. Mus. 1582) measures 370 milli-
meters as compared with 290 of the tibia, this exces-
sive length of the upper limb being correlated with
Figure 539. — Pelvis of Palaeosyops of. P. leia
Princeton Mus. 10232, superior view. After Earle. One-sixth natural size.
relatively slow movements. The chief characters of
the femur are as follows: The long, straight shaft
flattened superiorly on the posterior surface; great
trochanter moderately elevated; breadth across head
and great trochanter 132 millimeters (estimated)
apex of third trochanter 143 millimeters below head
apex of second trochanter 132 millimeters below head
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
625
both trochanters somewhat elevated and nearly
opposite each other (an important point of distinction
from the rhinoceroses); shaft narrowing, with a
rounded anterior and flattened posterior surface
below trochanters; transverse measurements, distally
93 millimeters, across condyles 83; the internal and
external condyles are equally elevated but the internal
condyle is somewhat smaller; patellar facet moderately
elongate, vertically placed — that is, looking forward
and very slightly downward; vertical measurement
63 millimeters, transverse 40. This is the titanothere
type of femur, which, although varying in proportion
in the longer-limbed forms, is readily distinguishable
from that of Amynodon and of other contemporary
Perissodactyla.
The tibia (Am. Mus.
1562) is distinguished by a
moderate development of
the cnemial crest; posteri-
orly the popliteal space is
bounded by a high internal
and a low external border;
the posterior surface of the
midshaft is gently convex,
the inner surface flattened;
anteriorly the cnemial crest
subsides into the angulate
anterior face of the shaft.
The fibula is restored
from other specimens. It
has a very elevated postero-
external facet on the tibia,
and inferiorly it articulated
not only with the astrag-
alus but also in extreme ex-
tension of the pes it barely
touched the calcaneum. In
the contemporary aquatic
rhinoceros Amynodon the
tibia is relatively shorter,
the superior head of the
fibula is more inferior in
position, and distally the
fibula barely, if at all, touched the calcaneum.
The pes is known principally from the associated
feet of Am. Mus. 1550 and metapodials and tarsus of
Am. Mus. 1582, all finely preserved. It is distin-
guished generically by the relatively short neck of the
astragalus, by the broad and relatively shallow cuboid,
navicular, and cuneiforms, and by the moderate
elongation of the metatarsals.
The total height of the calcaneum is 107 millimeters
as compared with 63, the maximum width; the tuber
calcis is suboval in section, the long diameter being
obliquely placed; in the extreme extension of the tibia
in some individuals this bone passes over posteriorly
onto the calcaneum; the fibula also has a calcaneal
facet.
The astragalus exhibits characters of family value in
the arrangement of the ectal, sustentacular, and
inferior facets as shown in Figure 541. The cuboidal
facet has a broadly oval external border. The astrag-
alar trochlea measures 50 millimeters transversely,
while the depth of this bone on the internal side is
58; the trochlear groove is of moderate depth; on
the outer side the trochlear surface thins out poste-
riorly, becoming confluent with the actual astragalo-
calcaneal facet, allowing for the passage of the tibia
upon the calcaneum above mentioned; distally the
inner side view of tarsus. Am. Mu
One-third natural size.
Figure 540. — Left pes of Palaeosyops leidyi
Am. Mus. 1589: Ai, Front view of pes with the phalanges foreshortened; A2, upper view of phalanges of the median digitfAa,
11682: Bi, Front view of pes; B2, upper view of phalanges of the median digit.
neck of the astragalus measures 48 millimeters trans-
versely and 11 vertically.
On its anterior face the cuboid measures 30 milli-
meters transversely and 23 vertically. The navicular
measures 44 transversely and 14 in its deepest part
vertically. The entire breadth of the second row of
the tarsi is 65, as compared with 55 in BTiadinorhinus
diploconus. The entocuneiform is well developed on
the postero-internal side of the tarsus, the mesocunei-
form is small, while the ectocuneiform is large, with
an internal facet for Mts II and an external facet for
626
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Mts IV, a very distinctive feature of this pes, which
enables us to distinguish it from that of Amynodon.
The metatarsals, like the metacarpals, are moder-
ately long, have flattened shafts, and are widely
expanded distally. The vertical measurements are,
Mts II, 103 millimeters; Mts III, 110; Mts IV, 103;
Mts III, distal breadth, 43. Mts II shows a postero-
lateral facet to which a vestige of Mts I was probably
attached. As regards displacement Mts IV and III
are "serial," articulating solely with the cuboid and
ectocuneiform respectively; Mts II, on the contrary,
is "displaced," with a prominent internal process and
abutment against the ectocuneiform.
The spread of the phalanges distally as moimted is
181 millimeters. The proximal phalanges are much
'?/!«k; /cuSI — ' ^-^ (cull) (navj
Figure 541. — Relations of the ectal, sus-
tentacular, and inferior facets of the as-
tragalus and calcaneum in Palaeosyops
One-third natural size.
the largest and longest in D. Ill; the first phalanx
has a length of 39 millimeters; the second, which
exhibits a sharply upturned distal facet, has a length
of 25; the third is restored, but the distal phalanx of
D. IV, measuring 33 millimeters in length and 35
in breadth, is spreading and distally cleft.
Measurements of the composite mounted skeleton of Palaeosyops
leidyi {Am. Mus. 1544)
Skeleton: Millimeters
Total length, pmx to tail drop 2, 007
Total length, pmx to ischial tuberosity 1, 980
Total height, top of third dorsal spine 1, 090
Total height to top of scapula 1,010
Total length along vertical column 2, 083
Skull: Millimeters
Total length, incisors to occipital condyle 415
Total breadth, transverse zygomata 310
Vertebral column, length:
27 presacral vertebrae 1, 230
7 cervicals 320
Midcervical centrum 34
17 dorsals 820
Fourth dorsal centrum 38
Fourth dorsal, height of spine 163
3 lumbars 148
Second lumbar centrum 47
4 sacrals 159
19 caudals (partly restored) 615
Ribs:
Fifth rib, outer curve 545
Eighth rib, outer curve 620
Scapula :
Length 345
Greatest width 220
Pelvis :
Os innominatum, total length 465
Width across ilia 488
Fore limb, total length 810
Humerus, total length 325
Radius, total length 235
Ulna, total length 315
Carpus, top of lunar to bottom of magnum 50
Carpus, width 96
Manus, length, lunar to tip of D. Ill 250
Digit III, length of metacarpal 113
Digit III, breadth of metacarpal 45
Hind limb, total length 940
Femur, total length 370
Tibia, total length 290
Pes, total length, os calcis to tip of D. Ill 340
Tarsus, height, calcaneum to ectocondyle 76
Astragalus, height, inner face 58
Astragalus, breadth, distal end 48
Calcaneum, total length 100
Calcaneum, total width 63
Mts III, length 110
Palaeosyops robustus
The characters of the postcranial skeleton of P.
robustus, so far as known, seem to differ but little
from those of P. leidyi. In the composite skeleton
of P. leidyi (Am. Mus. 1544) many of the vertebrae,
ribs, and parts of the feet and limbs seem to agree
in size with the remaining parts associated with the
P. leidyi type skull and yet are themselves associated
with skulls and teeth referred to P. robustus.
The seven vertebrae (fig. 544) associated with a skull
of P. robustus (Am. Mus. 1580) and used in restoring
the vertebrae of the skeleton P. leidyi agree with those
of Manteoceras but are nevertheless remarkable in this
respect — that the flange on the pleurapophysis of
C. 6 was much weaker than that in the rhinoceroses,
tapirs, horses, and most placental mammals. The
pleurapophysial flange on C 5 was expanded; it is
incompletely preserved, but it does not appear to
differ greatly from that of Manteoceras or of Doli-
chorMnus. In C. 4-C. 7 the prezygapophyses and
postzygaophyses face, respectively, obliquely inward
EVOLUTION OF THE SKELETON OP EOCENE AND OLIGOCENE TITANOTHERES
627
and outward, whereas in D. 1 and in the succeeding
dorsals they face respectively upward and downward.
This oblique inward and outward facing of the pre-
ih.inter-t/.
FiGUBB 542. — Atlas of Palaeosrjops robiistus
.■vm. Mus. 1580 Bridger formation, Bridger Basin, Wyo. Top view. One third
natural size.
zygapophysial and postzygapophysial facets,
respectively, is seen not only in Mnnteoceras,
Dolichorliinus, and the Oligocene forms but in
ungulates generally and is associated with the
vertical movement of the neck.
In C. 5-C. 7 the posterior face of the cen-
trum approaches a transverse oval form, the
front face is wider at the top and narrow at the
bottom. Comparison with Manteoceras is shown
in Figure 552.
The neural spines of the cervicals and dorsals
are narrow anteroposteriorly and deeply exca-
vated posteriorly. They increase rapidly in
height, from 55 millimeters in C. 4, 70 in C. 5,
and 135 in C. 7 to 201 in D. 1. The centrum
of C. 5 (posterior face) measures 43 millimeters verti-
cally, 55 transversely; that of C. 6 begins to assume
the narrower and deeper form characteristic of the
dorsals. The lamellae of C 4 and C. 5 are broadly
The scapula of P. robustus is represented by a
referred specimen (Am. Mus. 1580). It is decidedly
wider in proportion to its height than the scapula
of Dolichorhinus. The humerus associated with the
skull (Am. Mus. 1580, fig. 546) exhibits a length of
345 millimeters (estimated). It is
thus slightly longer and much more
massive than the humerus of Palaeo-
syops? sp. (Am. Mus. 12205), from
Bridger D, and it is much longer than
the humerus of Manteoceras from
Bridger D (Am. Mus.
12204), which meas-
ures 260 millimeteis
The distal bieadth
Figure 544. — Cervicals and dorsals of Palaeosyops robustus
Am. Mus. 1580; Bridger D (?). Cervicals 4-7, dorsals 1-3. One-third natural size.
trochlea is 60 millimeters. The radius (fig. 546) is
extremely broad (71 mm.) both proximally and dis-
tally, a marked progressive character of the Palaeo-
syopinae. The ulna also sharply expands distally;
the olecranon rises to a
pointed rugose knob.
The associated carpus
agrees closely with that
of P. leidyi but is of
somewhat larger dimen-
sions. Mts IV measures
FiG"0KE 543. — Atlas and axis of Palaeosyops leidyi?
Am. Mus. 12201. Bridger C 4. Inferior surface. One-third natural size.
expanded downward, while those of C. 6 and C. 7
exhibit a transverse rodlike expansion, those of C. 6
measuring 134 millimeters transversely. These parts
are not very well preserved in the P. leidyi skeleton.
The disposition of the lamellae and of the spines
is different in details from that in the skeleton of
Dolichorhinus.
115 millimeters; Mts III
expands to 49 distally.
Other characters of the
Palaeosyops rohustus
manus are as follows : The
most massive type, exhib-
iting graviportal and per-
haps aquatic adaptation; i-^-^fAT
brachypodal; carpus very Figure 545. — Left scapula of
broad (94 mm.) in upper Palaeosyops robustus
Bridger specimens, deep Am, mus. isso; Bridger c or d (?).
, ± • 1 / A r, One-sixth natural size.
anteroposteriorly (40 »
mm.), measuring vertically from summit of lunar to
bottom of magnum 50 millimeters; metacarpals robust,
shafts widely spreading distally ; scaphoid very deep
anteroposteriorly and laterally (thus differing from
628
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
.'(qMnh.h) '^
l^prex.uL. Tjic^eTwXh-)
.prex.Tia. (cptl^hj-
prni\
prst.ul.
prst.
Figure 546. — Bones of forearm of Palaeosyops
A, Palaeosyops robustus. Am. Mus. 1580; Bridger C or D; left humerus, front and distal views. B, Palaeosyops leidyi (type), Am. Mus. 1544; Bridger
C or D; right humerus and forearm; outer side view, with section of radius (r) and ulna (u). C, Right radius of same, front view. D, Palaeo-
syops sp., Am. Mus. 12386; Bridger C; left humerus, front and distal views. Ei, Palaeosyops major, Am. Mus. 13116; Bridger B 3; left radius
and ulna, front view. Ea, The same, outer side view. F, Palaeosyops robustus, Am. Mus. 1580; Bridger C or D; left ulna, front view of proximal
end. G, Palaeosyops robustus, Princeton Mus. 10360; Bridger Basin; left radius and ulna, with sections. H, Palaeosyops robustus, Am. Mus.
1580; Bridger C or D; left radius. All one-fourth natural size.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
629
that of the manteoceratine group); lunar deep antero-
posteriorly and broad laterally, resting broadly on
magnum as well as on unciform; trapezium small,
broadly articulating with trapezoid, and barely articu-
lating with Mtc II, no
facet apparent for scaph-
oid; trapezoid broad;
magnum very broad,
with six distinct faceted
angles ; large lunar facet
Figure 547. — Left astragalus of appearing in front view,
Palaeosyops copei? hook of magnum asym-
Am. Mus. 12205a; Bridger D 1; front (Ai) and metrical, pointed ; Unci-
rear (A2) views. One-third natural size. !■„„,„ u „J^„ ■U„^4
form broadly hori-
zontal, supporting more than half of the lunar; ter-
minal phalanges cleft and somewhat rounded rather
than spreading distally; the carpus while not ancestral
is in general analogous to that of the heavy Oligocene
titanotheres, namely, Megacerops and Bronfotherium.
Figure 548. — Fore limb of Palaeosyops copei?
Am. Mus. 12205; Bridger D 1. Ai, Front view of left fore arm and manus, one-sixth
natural size; A2, outer view of same, with humerus, one-sixth natural size; A3, right
ungual phalanx of median digit, inferior view, one-sixth natural size.
Palaeosyops copei
Two skeletons found close together on level D 1 of
the Bridger Basin (Am. Mus. 12205 and 12205a) are
especially valuable because they give us the propor-
tions of the limbs and the complete structure of the
manus of an exceptionally short-footed type of Palaeo-
syops, the specific determination of which is doubtful;
it may be provisionally referred to P. copei. One of
these specimens (No. 12205a) belongs to a younger
and smaller individual, the other (No. 12205) to an
older and larger individual.
Skull and skeleton of the younger and smaller individual (No. 12205n)
Slcull. — The skull belongs to a rather young indi-
vidual. It exhibits the following especially important
characters (fig. 281): (1) The nasals taper slightly
anteriorly, much less so than in Palaeosyops leidyi;
(2) the nasals are not deeply decurved at the sides as in
LimnoTiyops; (3) the nasals exhibit V-shaped prolonga-
tions on the sides of the face as in Palaeosyops; (4) there
Figure 549. — Left manus of Palaeosyops copeif
Am. Mus. 12205; Bridger D 1. Ai, Front view of manus; A2, upper view of
phalanges of median digit; A3, inner side view of carpus; Ai, upper or proximal
view of carpus. One-third natural size.
are very slight and smooth prominences on the sides of
the face at the junction of the nasals and frontals, occu-
pying the same position as the rudimentary horns which
have been observed in Palaeosyops; (5) there is a
narrow and apparently deep sagittal crest.
Fore limb. — The chief upper limb character is that
the humerus is longer (335 mm., estimated) than the
radius (223 mm.), indicating slow speed, and we are
surprised to find that the manus is shorter than in
either Palaeosyops leidyi or P. roiustus. The humerus
measures about 63 millimeters across the distal
trochlea. The ulna and radius are much less massive
than in the type of Palaeosyops leidyi and slightly less
630
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
massive than in the type of L. laticeps; the ulna,
including the olecranon, measures 288 millimeters;
the radial shaft 220. Distally the combined radio-
ulnar facets for the carpus measure 70 millimeters.
A part of the left manus is also preserved.
Hind limb. — The hind lunb from the head of the
femur to the ankle joint measures 642 millimeters.
The pes is unfortunately unknown. The femur is a
long (376 mm.), rather slender bone, greatly exceed-
ing the tibia (270
mm.) in length, as in
all slow-moving ani-
mals. Both these
bones are readily
distinguished from
those of Palaeosyops
leidyi by their more
slender proportions.
The femur has the
long, straight form
characteristic of
titanotheres gener-
ally, with the second
and third trochan-
ters nearly opposite
each other in the
upper portion of the
shaft; distally the
patellar trochlea
points obliquely for-
ward. Thepatellais
a relatively smaller
element than in
Palaeosyops leidyi
(transverse 40 mm.,
vertical 56). The
tibia also exhibits a
slender shaft. In
general these limbs
are distinguished
from those of Pal-
FiGURE 550.— Right hind limbs of aeosyops leidyi by
their greater slen-
derness, from those
Palaeosyops major and P. copei?
A, P. major, Am. Mus. 13116; Bridger B 3; femur
and tibia. B, P. copei?, Am. Mus. 12205; Bridger
D 1; femur, tibia, and fibula. One-sixth natural of LimnohyopS lati-
^'^'^' ceps by their inferior
length. The astragalus (fig. 547) is of palaeosyopine
type.
Another more adult skeleton (Am. Mus. 12205)
Another skeleton (Am. Mus. 12205) was found with
the one described above, at Lone Tree, Henrys Fork
Bridger Basin, Wyo., level Bridger D 1. It is larger
and belongs to a fully adult individual, the tibia
measuring 300 millimeters as compared with 270 in the
specimen just described.
Fore limb. — By far the most important part of this
skeleton is the fore limb. The total length of the ulna
is 326 millimeters, that of the radius 235; the transverse
radio-ulnar carpal facets measure 85.
Manus. — Three very distinctive characters are
found in the manus as compared with that of Palaeo-
syops leidyi — (1) the abbreviation of Mtc V; (2) the
relatively elongate form of the distal phalanges as
compared with those of Palaeosyops; (3) the broader
displacement of the lunar on the unciform and its
narrower facet on the magnum. These features are
correlated with narrower hoofs, somewhat greater
speed, and slightly more mesaxonic disposition of the
metacarpals. The transverse measurement across the
top of the carpus is 89 millimeters. The scaphoid
rests on the trapezoid and magnum only, because the
trapezium, though well developed, articulates with
Mtc II and the trapezoid only. The lunar is narrower
(32 mm.) superiorly than that of P. leidyi and in-
feriorly is readUy distinguished by its narrow and
nearly vertical facet for the magnum and relatively
broad and horizontal facet on the unciform. The
cuneiform measures 40 millimeters superiorly. The
magnum is a much smaller bone than in Palaeosyops
leidyi, subquadrate in form, with a short (16 mm.)
oblique continuous facet for the lunar and unciform.
The metacarpals measure, Mtc II, 96 millimeters;
Mtc III, 106; Mtc IV, 93; Mtc V, 74. This animal
was therefore decidedly short-footed ; the median meta-
carpal in the contemporary Palaeosyops leidyi meas-
ures 116 millimeters. The form of the distal pha-
langes is somewhat more elongate and less expanded
distally than in Palaeosyops.
Femur. — The femur exhibits the lesser and third
trochanters directly opposite each other; distally it
measures 90 millimeters across the condyles.
Tibia. — The tibia is finely preserved. The femoral
facets measure 87 millimeters (tr.), the astragalar
facets 43 (tr.) the shaft 31 (tr.) by 33 (ap).
Measurements of limb bones referred to Palaeosyops, in Tnillimeters
P. leidyi,
Am. Mus.
1544 (com-
posite
slceleton),
Bridger
C(?), D
Humerus, length
Radius, length
Radius, breadth, proximal end
Radius, breadth, distal end
Ulna, length
Carpus, width
Mtc II, height
Mtc III, height
Mtc III, greatest width, distal.
Mtc IV, height
Mtc V, height
Femur, length
Tibia, length
P. copei.
Am. Mus.
12205
(adult),
Bridger
Dl
p. copei.
Am. Mus.
12205a
(young),
Bridger
Dl
335
233
223
62
54
67
56
325
290
95
86
96
93
105
103
39
35
93
89
75
73
376
370
293
273
325
235
63
67
315
96
106
113
45
102
75
370
290
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
631
Palaeosyops copei?
From an uncertain level in the Washakie Basin
comes an imperfect pes (Am. Mus. 5097) associated
with other limb fragments, which Cope erroneously
referred to his "Palaeosyops vallidens" but which is
certainly a member of the Palaeosyopinae, whereas
Cope's P. vallidens is now referred to LclidiorJiinus
vallidens, a member of the ilanteoceras-BolicJiorhinus
group. Possibly this pes is referable to P. cojpei.
The astragalus in question agrees in most characters
with those referred above to Palaeosyops copei (Am.
Mus. 12205, 12205a) but is distinguished by the
wider sustentacular facet, deeper navicular facet, and
lower internal trochlear keel. The broad cuboidal
facet also rounds off into the distal calcaneal facet;
the pit on the inner face of the astragalus, below the
internal trochlear keel, is very deep, and the internal
distal protuberance for the lateral ligament is also
very prominent; the depression for the tip of the
fibula is wanting. The cuboid (absent) was elongate,
and Mts IV does not articulate with the ectocuneiforra
(cf. P. leidyi, above).
Another palaeosyopine Washakie specimen from
the Cope collection (Am. Mus. 5105) is an astragalus
associated with fragmentary limb bones and bearing
the same museum number as the upper dentition which
was referred to above as allied to Palaeosyops copei.
The characters of the astragalus do not support this
association; it is about one-fourth smaller than that of
P. leidyi and has a relatively narrower neck and
narrower ectal and sustentacular facets, the ectal
facet being shallow. It rather resembles a small
Limnoliyops.
SUBFAMILY MANTEOCERATINAE
Graviportal titanotheres of the upper deposits of
the Bridger Basin, Wyo., the lower deposits of the
Washakie Basin, Wyo., and the upper deposits of the
Uinta Basin, Utah. Feet brachypodal. Ungual pha-
langes truncate. Tibia very short. Astragalus wide.
Manteoceras
General features. — Judging from its limb proportions,
M. manteoceras was a slow-moving animal, much less
alert than the tapir (T. terrestris) and less swift than
its congeners of Mesatirhinus. All its known skeletal
parts were found in Bridger D, so that it was con-
temporaneous with the larger and more massive species
of Palaeosyops. These parts tend to confirm the view
that Manteoceras was allied to Mesatirhinus , DolicJio-
rhinus, and the Oligocene titanotheres. The Manteo-
ceras of this period was a short, low-bodied animal.
The skeleton, like the skull, is in many features
prophetic of the Oligocene titanotheres; it is more
paraxonic and tetradactylous, D. 5 being relatively
longer than in Palaeosyops; the tibiae are relatively
shorter (x%^ of the femur) than in any other Eocene
titanothere; the humerus is intermediate in length and
in its tuberosities foreshadows that of Oligocene type,
there are broad flangelike pleurapophyses in the
posterior cervicals. Many of the adaptive analogies
ally it to Palaeosyops, although the deeper paleotelic
resemblances ally it to Dolichorhinus.
Generic characters. — Atlas narrower than in Palaeo-
syops but broader than in Mesatirhinus; axis with a
high spine; neural canal of cervicals and anterior
dorsals rounded rather than angulate superiorly;
anterior faces of cervical centra subcircular, lateral
flange on C. 6 large, spines of cervical and dorsal ver-
tebrae abbreviate as in Palaeosyops, second dorsal
with prezygapophysis forming an angle with the neural
spine. Humerus intermediate to short; manus mesa-
tipodal; superior facets of carpals and metacarpals
flatter, relatively wider posteriorly; scaphoid shallower
anteroposteriorly; lunar with subvertical magnum
facet; cuneiform flatter; trapezoid with facet for trape-
zium continuous with scaphoid facet. Magnum rela-
tively narrow, with scaphoid facet subvertical, pos-
terior hook spatulate; unciform relatively narrow,
with subquadrate lunar facet. Metacarpals longer,
narrower and more straight-sided than in Palaeosyops,
but broader than in Mesatirhinus; proximal facets
wide posteriorly; metacarpal V relatively longer than
in Palaeosyops; phalanges smaller, shorter, and
broader; distal phalanges broadly expanded, truncate,
and deeply cleft. Femur of intermediate length.
Tibia both relatively and absolutely short, with broad
proximal end. Astragalus intermediate, rather broad,
with broader convex tibial keel of the trochlea and
broad cuboid facet, but also with straight-sided sus-
tentacular facet and deep navicular facet.
Materials. — Material referable to this genus is rare.
There are, in fact, only two specimens in which parts
of the postcranial skeleton are certainly associated
with the skull and dentition — Am. Mus. 1587, a young
adult male from Bridger D, the skull of which is shown
in Figure 303, and Am. Mus. 12204, a crushed skuU of
an old animal, possibly a female, from Bridger D 1
or D 2. A third specimen (Am. Mus. 12216), con-
sisting of an incomplete manus, is not associated with
a skull or with dentition but agrees so closely with
Am. Mus. 1587 that it may be referred with confi-
dence to Manteoceras.
From these three specimens we learn the principal
characters of the cervical vertebrae, anterior dorsals,
humerus, manus, femur, tibia, and astragalus. A
pelvis is provisionally referred. The remaining verte-
brae, ribs, sternals, and most of the pes are practically
or wholly unknown.
Manteoceras manteoceras
A close examination of the vertebrae of Manteo-
ceras in comparison with those of Palaeosyops and
Dolichorhinus shows that the differences relate mostly
to rather minor details and do not lend themselves
to broad characterizations. It may be said, how-
632
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
ever, that in a general way the vertebrae of Manteo-
ceras seem to be intermediate in form between those
of Palaeosyops and of DolicTiorhinus, sharing with
those of Palaeosyops the high spine of the axis and the
anteroposteriorly narrower spines of the anterior
dorsals and sharing with those of DolicTiorhinus the
larger pleurapophysal flange of C. 6, the marked angu-
lation between the prezygapophysis and the neural
spine in D. 2, the inferior keels on D. 1, D. 2.
A3/
A2
Figure 551. — Atlas of Manteoceras manteoceras
Am. Mus. 12204; upper Bridger. Ai, Posterior view; As, anterior
view; A3, ventral view; Ai, dorsal view. The arrows indicate the
course of the first spinal nerve and vertebral artery. One-third
natural size.
The atlas of Manteoceras has a larger vertebrarterial
canal, the anterosuperior border of the cotylus is more
deeply concave, the superior openings for the spinal
nerve are nearer the lateral borders. The median
hypapophysis is acuminate rather than peg shaped.
The posterior root, or proximal portion of the pleura-
pophysis is thicker vertically. The atlas, on the
whole, is intermediate in form between those of
Palaeosyops and of DolicTiorhinus, it being broader
than that of Dolichorhinus, but sharing with it several
of the characters already mentioned.
The axis probably had a taller spine than in Palaeo-
syops; the crura of the neural arch appear shorter
anteroposteriorly; the postzygapophyses were larger
and less inclined downward; the neural canal was
more arched — that is, less angulate superiorly; the
posterior face of the centrum was not so wide.
The third to seventh cervical vertebrae (see fig. 552)
differ from those of Palaeosyops (Am. Mus. 1580,
1562). The neural canal is smaller and more roundly
arched superiorly; the anterior faces of the centra
are more circular instead of being broadly flat-
tened at top; the posterior faces of the centra are
also rounder superiorly and shallower vertically; the
prezygapophyses and postzygapophyses are possibly
larger. The neural spines seem to be of about the
same relative size in the two genera, and lateral
flanges or pleurapophyses appear on C. 3 to C. 6 in
both genera, but the flange on C 6 seems to have
been larger in Manteoceras than in Palaeosyops.
The first dorsal had a spine of nearly the same rela-
tive size and proportions as in Palaeosyops; the neural
canal is smaller and more arched superiorly, the front
face of the centrum is rounder at top, and the rear face
of the centrum, including the facets for the capitulum
of the second rib, is more transversely oval. The
oblique external buttress or column formed by the con-
joined pedicles of the prezygapophysis and pleura-
pophysis in Manteoceras is shorter; the prezygapoph-
ysis seems more horizontal and broader anteropos-
teriorly. In the second dorsal both the anterior and
posterior faces of the centrum are more transversely
oval, the pleurapophyses spring from the sides of the
centra at a lower level, the rib facets are smaller;
the prezygapophyses face upward and form an open
angle with the base of the neural spine, whereas in
Palaeosyops the prezygapophyses of this vertebra face
forward and upward and are nearly continuous with
the base of the spine. The remaining vertebrae are
too imperfectly known to warrant description.
The scapula (Am. Mus. 1587) is too insufficiently
known to afford diagnostic characters.
The humerus (Am. Mus. 12204, fig. 553, B) is of
intermediate length (290 mm.) between the long
humerus of Palaeosyops (325) and the relatively short
humerus of DolicTiorhinus (255 to 315); it agrees nearly
in length and general characters with the humerus of
the Limnohyops skeleton Am. Mus. 11689, from
Bridger B 2 (see above) ; it is massive proximally, with
a heavy deltoid ridge; it apparently had a platelike
crest for the infraspinatus muscle and stout tuber-
osities for the supraspinatus and deltoid muscles; the
supinator crest is relatively short. The radius and
ulna are not known.
The manus (Am. Mus. 12204, 1587, 12216) repre-
sents a broad-footed phase of the same stock that gave
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
633
rise to Mesatirhinus and Dolichorhinus, and although
broader than in Mesatirhinus is separated from that
of Palaeosyops by many trenchant characters, the
resemblances to that form being chiefly homoplastic or
analogous. The width of the carpus (79 mm., esti-
mated) is intermediate between that of MesatirMnus
(65 to 72) and of Palaeosyops (95), but nearer the
former than the latter. The superior facets of the
carpals and metacarpals are flatter than in Palaeosyops
and relatively wider posteriorly. More in detail, the
scaphoid (Am. Mus. 12204) in superior view is more
shallow anteroposteriorly, being suboval, with broad
A2 ^- — ^-^ Ba
Figure 652. — Seventh cervical vertebra of Manieoceras man-
teoceras compared with that of Palaeosyops leidyi
Ai, A2, M. manieoceras, Am. Mus. 12204; anterior and posterior views. Bi, B2
P. leidyi, Am. Mus. 1562; anterior and posterior views. One-third natural size.
ends, whereas in Palaeosyops it is elongate anteropos-
teriorly, with pointed ends; the facet for the magnum
is relatively broader, that for the trapezoid more hori-
zontal than in Palaeosyops. The lunar (No. 12204)
has the facet for the magnum more vertical and that
for the unciform more horizontal than in Palaeosyops;
its posterior face is relatively deeper and much nar-
rower, but not so much so as in MesatirMnus. The
cuneiform (Am. Mus. 12216) differs from that of
Palaeosyops in the following particulars: The superior
(radial) facet is not produced antero-internally into a
ridged prominence, is deeper anteroposteriorly near the
external end, and is separated from the broad pisiform
facet by a high keel; of the two facets for the lunar
the upper one is comparatively deep vertically, the
lower one is very shallow; the unciform facet is deeper
anteroposteriorly. As compared with that of Mesati-
FiGUEE 553. — Left humerus of Manieoceras 7nan-
ieoceras
A, Am. Mus. 12384; Bridger C 3; front view. B, Am. Mus. 12204;
Bridger D 2; front (Bi), and outer side (Ba) views. One-si-xth
natural size.
rJiinus the cuneiform is much broader, but a subfamily
agreement is seen in the detailed characters of the
facets. The pisiform is not preserved. In the trape-
zoid (Am. Mus. 1587, 12216) the trapezium facet is
confined to the postero-external part of the bone and
554. — Right manus of Man-
ieoceras manieoceras
Chiefly from Am. Mus. 1587; Bridger C or D. Scaph-
oid, lunar, and magnum chiefly from Am. Mus.
12204. The magnum itself and the scapho-magnum
contact as represented are somewhat too broad to
fit well in 1587. Ai, Front view of manus; A?,
phalanges of median digit; As, outer side view of
metacarpal V; Ai, second phalanx of digit V. One-
third natural size.
is broadly and roundly continuous above with ihe
scaphoid facet, whereas in Palaeosyops it extends
nearer to the front face of the bone and meets the
scaphoid facet supero-anteriorly at an acute angle.
In Manieoceras also the front face of the trapezoid
is not carried obliquely downward and inward
634
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
into an angulate process, and in inferior view the
bone is more oval, whereas in Palaeosyops it is more
or less rhomboid. In most of these characters the
trapezoid is, on the whole, nearer to that of Mesati-
rJiinus than to that of Palaeosyops. The trapezium
is not known, but to judge from the character of the
trapezium facets on the trapezoid it probably ap-
proached nearer to the pear-shaped trapezium of
Mesatirhinus than to the broad ovoid trapezium of
Palaeosyops. The magnum (Am. Mus. 1587, 12204)
has a relatively narrower front face than ia Palaeosyops,
and its scaphoid facet is more vertical than horizontal;
the facet for Mtc II is larger, the facet for Mtc III
in inferior view is more straight-sided, broader pos-
teriorly, and less broad anteriorly; the posterior
hook of the magnum is broadly spatulate instead of
posteriorly pointed; the capitellum, or posterosupe-
rior head for the scaphoid and lunar, in Am. Mus.
relatively wider posteriorly, except in Mtc V; Mtc II
is a little shorter than in Palaeosyops, but Mtc IV and
V are as long or longer. The dimensions of the
metacarpals compared with those of the supposed
Palaeosyops copei?. Am. Mus. 12205, are as follows:
Measurements of metacarpals in Manteoceras manteoceras and
Palaeosyops copei?, in millimeters
Figure 555. — Pelvis of Manteoceras?
.'.ntero-inferior aspect of pelvis. Am. Mus. 235!!; Washakie Basin, level B 1. One-
sLxth natural size.
12204 appears more globose than in Palaeosyops, but
this character is not so well shown ia Am. Mus. 1587.
The unciform (Am. Mus. 1587, 12216) is proportion-
ately narrower than in Palaeosyops but broader and
deeper vertically than in Mesatirhinus; in top view
the lunar facet is subquadrate, whereas in Palaeosyops
it is more rhomboid, being produced postero-externaUy
and antero-internally; the cuneiform facet is not so
wide supero-externally, being thus of more even diam-
eter than in Palaeosyops, in which it is wide supero-
internally and narrow supero-externally. The tuber,
or posterior process of the unciform, on its internal
face meets the main body of the bone at right angles,
whereas in Palaeosyops it slopes gently down and
meets the posterior face on an open angle.
The metacarpals (Am. Mus. 12216, 1587, 12204 in
part) are relatively longer, narrower, and more
straight-sided than in Palaeosyops; their distal (pha-
langeal) facets are transversely subcylindrical rather
than subglobose; their proximal or carpal facets are
M. manteoceras
II, length
II, pro.ximal ■nidth (front)
II, distal width (maximum) _ _
III, length '
III, proximal width (front) __
III, distal width
IV, length !
IV, proximal width (front)
IV, distal width (maximum) _
V, length I
V, proximal width (front)
V, distal width (maximum) _ _
104
27
27
P. copei,
Am. Mus
12205
96
35
42
107
38
93
31
40
75
24
35
The phalanges are much smaller, relatively shorter,
broader distally, and deeper vertically than in Palaeo-
syops:
Comparative measurements of proximal phalanx of digit HI in
Manteoceras manteoceras and Palaeosyops copei?, in millimeters
Length
Transverse proximal
Vertical proximal
Transverse distal
M. manteo- P. copei,
ceras, 1 Am. Mus.
.\m. Mus. 1587 12205
36
35
24
27
The distal phalanges are short and widely expanded,
truncate, and deeply cleft distally, in contrast to the
longer, distally rounded to subpointed, rather feebly
cleft unguals of Palaeosyops. These differences,
considered in connection with the narrower, straight-
toed manus, ia contrast to the broad spreading-toed
manus of Palaeosyops, poiat to tapir-hke rather than
hippopotamus-lilve habits and tend also to confirm
not only the hypothesis that Palaeosyops was semi-
aquatic but also the hypothesis that Manteoceras was
allied to Mesatirhinus and to the Oligocene titano-
theres.
A well-preserved pelvis (Anr. Mus. 2358) froai the
Washakie Basin is provisioaaUy referred to Man-
teoceras. The measurements are as follows: Pelvis
transverse 530 millimeters, anteroposterior 450 ; pubo-
ischiadic symphysis 170. Only three sacral vertebrae
are preserved out of the probable four. This speci-
men illustrates the deeply revolute character of the
prezygapophyses of the first sacral vertebra, the trans-
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
635
verse extent (170 mm.) of the sacral arcade, the
flattened anterior (inferior) faces of the ilia, the vertical
keel below the pubo-ischiadic symphysis, the deep
acetabular notch. Its graviportal adaptation is indi-
cated by the uniformly convex superior border of
the broadly expanded ilia, in contrast with the
indented border of the pelvis of Limnohyopsf (Am.
fo&paC
a
IB
TTialu
FiGTJHE 556. — Femora and tibiae of
Manieoceras manteoceras
A, Left tibia and distal end of femur, Am. Mus.
12204, Bridger D 2, front view; B, left femur and
tibia, Am. Mus. 1587, Bridger C or D, front
view. One-sixth natural size.
Mus. 2348) mounted in the skeleton of P. leidyi
(see above).
The femur (Am. Mus. 1587, 12204) is not very well
preserved. Its length (400 mm.) is somewhat less
than that (435 mm.) of a large Palaeosyops major
(Am. Mus. 13116) from Bridger B 3; it is relatively
stouter than in the large Mesatirhinus petersoni
(No. 11659); the third trochanter is large; the distal
condyles are more sharply keeled than in Palaeosyops
major.
The tibia (Am. Mus. 1587, 12204) is much shorter
(length 265 mm.) than in Palaeosyops major (325 mm.) ;
it is somewhat shorter and much stouter than in
Mesatirhinus, and the proximal end is relatively
broad (87 mm).
The astragalus (Am. Mus. 1587, 12204) parallels
that of LimnoJiyops and Palaeosyops in the following
characters: Trochlea broad with very convex tibial
keel, neck relatively broad, cuboid facet broad, sus-
tentacular facet not very long vertically. It differs,
however, from that of the Palaeosyopinae and shows
the subfamily kinship with Mesatirhinus in the fol-
lowing: Neck not so broad in proportion to the total
height, navicular facet deep anteroposteriorly, sus-
tentacular facet straight-sided, forming with the
cuboid facet a broad L, its internal or tibial edge
Figure 557. — Left astragalus of Man-
teoceras manteoceras
Am. Mas. 1587; Bridger C or D; front and rear
views. One-third natural size.
set nearly flush with the internal face of the bone;
depression on internal face beneath the tibial keel not
forming a deep pit; process on internal face near distal
end forming a marked protuberance; articular surface
of the trochlea extending down antero-internally on
to the pedicle of the trochlea. The astragalus is
distinguished from that of Mesatirhinus not only by
its greater breadth and stoutness throughout, but
especially by the greater breadth of the cuboid facet,
the more convex tibial keel of the trochlea, the shorter
and broader sustentacular facet. Of the two astragali
Am. Mus. 12204 is much the smaller and difl'ers
from Am. Mus. 1587 in minor details, differences
which may be connected with the fact that No. 12204
is an old and possibly female animal, whereas No. 1587
is a very large young male.
The rest of the pes is unknown.
636 TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Comparative measurements of limh hones referred to LimnoTiyops, Manteoceras, and MesatirTiinus, in millimeters
Limnohyops monoconus?
Manteoceras manteoceras
Mesatirhinus petersoni
Am. Mus.
11689,
Bridger
B2
Am. Mus.
11699,
Bridger
B2
Am. Mus.
11690,
Bridger
Bl
Am. Mus.
1587,
Bridger
D
Am. Mus.
12204,
Bridger
D2
Am. Mus.
12216,
Bridger
D
Am. Mus.
11659,
Bridger
C5
Am. Mus
1571,
Wastialtie
A
Princeton
Mus. 10013,
Bridger
Humerus, length, head to interior condyle. _
295
52
228-233
58
56
74.
290
230
55
57
74
305
258
233
49
49
72
300
103
25
106
24
85
20
245
50
60
Radius, circumference of shaft, just above
75
Ulna, length 298-308
-310
72
310
'■79
65
Mtc II height
99
31
99
32
103
34
97
29
79
24
10]
25
Mtc III height
109
33
98
29
105
118
31
88
358
283-287
72
53
40
41
95
46
110
23
121
29
111
112
Mtc III, width, maximum distal
Mtc IV, height . ..--
28
387
297
83
55
42
44
108
55
114
24
126
34
97
30
82
22
'■395
268
56
49
44
91
29
95
"SI
83
98
Mtc IV, width, maximum distal
Mtc V, height . ... . _
22
82
! 26
19
"357
285
71
53
39
44
98
53
Tibia length
265
87
50
42
40
Astragalus, height, inner face
Astragalus, width, proximal end
Calcaneum, width (including sustentacu-
Mts II length
1
Mts III length
112
32
105
27
1
Mts IV length
97
30
91 1 __
28
1
a Estimated.
SECTION 4. THE POSTCRANIAL SKELETON OF UPPER
EOCENE TITANOTHERES
SUBFAMILY DOLICHORHININAE
Slender to mediportal titanotheres of the upper
deposits of- the Bridger Basin, Wyo., the deposits of
the Washakie Basin, Wyo., and the deposits of the
Uinta Basin, Utah. Feet mesatipodal to brachypodal.
Ungual phalanges truncate. Tibia slender to short.
Astragalus narrow.
The general adaptations of the limbs and other
parts of the skeleton of the Dolichorhininae have
already been stated. The main features of the adap-
tive radiation of the skulls and feet and the geologic
succession are as follows:
Mesatirhinus: Mesaticephalic; mesatipodal; Bridger C and D,
Washakie A.
Metarhinus: Mesaticephalic; mesatipodal; Washakie B, Uinta
B 1.
Dolichorhinus: Dolichocephalic; brachypodal; Washakie B,
Uinta B 2.
In the treatment of these forms it is convenient to
begin with the tapir-like Mesatirhinus, on the whole
the most primitive and central, and then to describe
MetarJiinus, a dwarfed, aberrant form. This will be
succeeded by the description of the extremely long-
headed Dolichorhinus, which is further distinguished
from the above animals by the possession of short feet.
Mesatirhinus
GENERAL FEATIJKES
The parts of the skeleton of Mesatirhinus are
readily distinguished by the collector and student as
belonging to the most slender-limbed of the Imown
middle Eocene titanotheres.
It is a striking proof of the generally heavy-bodied
proportions of the Eocene titanotheres that even
these most light-limbed members of the family are
somewhat heavier in their proportions than the
modern tapirs, animals which we are accustomed to
think of as rather heavy-bodied, forest-living, and
certainly not cursorial. In brief, Mesatirhinus in
limb structure is proportioned much as Tapirus in-
dicus, but the fore and hind feet were somewhat
broader and flatter, the back was more arched, the
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
637
neck was relatively shorter, the head more elongate,
thus bringing the lips closer to the ground.
Yet, as compared with the other middle Eocene
titanotheres, MesatirJiinus is mediportal, and all parts
of the skeleton which are known are relatively narrow
and deep in their proportions, including the atlas, the
Mesatirhinus petersoni, Princeton Mus. 10013, Figures 563, 565,
566; Bridger D.
Mesatirhinus petersoni?, Am. Mus. 11659 (fore and hind limbs
not associated). Figure 564; Bridger C 5.
Mesatirhinus petersoni, Am. Mus. 1571; Wasliakie A.
Mesatirhinus megarhinus, Am. Mus. 1523 (associated), Bridger
C?
■■^■'^"U"i n f'"! n r
Hind limbs // ::^y"' li
A.M. 11659 ^""^ yj^ fy
I'iGURB 558. — Restoration of the skeleton of Mesatirhinus petersoni
A provisional reconstruction of this light-limbed titanothere of Bridger D, based on specimens in the American Museum pertaining to various
individuals. The backbone, scapula, and pelvis are restored in broken lines from BolichoThinus: the ribs from Palaeosyops. One-twelfth
natural size.
imbs, the carpals and tarsals, and the metapodials.
We observe at once (figs. 512, 520) that the digits of the
manus are more compressed laterally and at the same
time more paraxonic and tetradactylous — that is, the
fifth digit of the manus is relatively longer than in the
above-described Palaeosyops and LimnoTiyops.
The above individuals include only a single vertebra,
the atlas. The scapula and pelvis are still unknown.
When found they will furnish us with very significant
characters.
In the following description of the skeleton it seems
best to describe together the parts belonging to the
Figure 559. — Restorations of Mesatirhinus petersoni (left) and Palaeosyops leidyi (right)
By Mrs. B. M. Fulda. Bridger C and D. About one-thirtieth natural size.
The above diagnosis is important because this type
of skeleton is broadly ancestral to the mediportal and
brachypodal DolichorJiinus.
The skeletal material of MesatirJiinus, like that of
Manteoceras, is rarely associated with parts of the
skull or teeth, so that identifications are difficult;
parts only of two skeletons have been found asso-
ciated. The specimens listed below have been col-
lected.
different species of Mesatirhinus. The geologically
oldest specimen named in the above table is M. mega-
rhinus, from the Bridger formation, level unknown
(Am. Mus. 1523), consisting of a skull with incomplete
axis. From Spanish John's Meadows, Bridger Basin,
Wyo. (level Bridger C or D), came the very well-
preserved radius, ulna, and manus (Princeton Mus.
10013) that were describedand figured by Earle (1897.1,
pp. 358-364) as belonging to Limnohyops laticeps.
638
TIT.\NOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 560. — Atlas of Mesati-
rhinus megarhinus
These unassociated bones agree very well with those of
Am. Mus. 1571. In Bridger C 5 was found Am. Mus.
11659, including the radius, ulna, manus, femur, tibia,
and pes, which afford a knowledge of the relative pro-
portions of the fore and hind limbs. These bones are
considerably larger than
those of any of the other
specimens and appar-
ently pertain to a large
individual of M. petersoni.
In beds of the same age
(Washakie A) was found
the leading associated
specimen (Am. Mus.
1571), including the front
Am. Mus. 1523; upper Bridger. The ar- . r .i 1 11 A
rows indicate the course of the first spinal part 01 the SKUll and. Up-
nerve and vertebral artery. Dorsalview. pgp dcnjtition of MeSOti-
One-third natural size. , . , • j j.i
rh'inus petersom, together
with the radius, ulna, incomplete manus, astragalus,
and fragments of the pelvis.
GENERAL CHAKACTERS OF THE SKELETON OF MESATIRHINUS
AS COMPARED WITH OTHER EOCENE TITANOTHERES
Atlas relatively narrow and deep, elongate, facets
for axis facing inward. Humerus relatively shorter,
radius relatively longer than in Palaeosyops or Man-
teoceras; radius long, slender, with narrow extremities
proximal, and distal articular facets shallow antero-
posteriorly. Ulna slender, with erect, subtruncate
olecranon. Manus high and narrow, straight-sided
parallel metacarpals, and short phalanges; carpals in
general narrow, vertically deep. Scaphoid shallow
anteroposteriorly, broad posteriorly. Lunar narrow,
lower end sharply wedge-shaped; magnum facet sub-
vertical; cuneiform facet ^^ ,.-., ,.-....
broad; cuneiform narrow,
not extended postero-exter-
nally. Pisiform with prox-
imal end sharply con-
stricted from the shaft.
Trapezium small, pear
shaped, articulating with
scaphoid. Trapezoid wide
anteroposteriorly, shallow,
truncate posteriorly. Mag-
num small, scaphoid facet j^' cpom. ^^^
broad. Unciform narrow, ^.^^^^^ 561.-Humerus of
vertically deep, lunar facet Mesatirhinus megarhinus
quadrate. Metacarpals Am. ivius. 12385; Bridger c or d. a,,
with carpal facets shallow
and truncate (rather than
extended backward as in Palaeosyops), elongate,
straight -sided; fifth metacarpal long and narrow,
proximal end embracing unciform externally. Prox-
imal phalanges relatively shorter than in Limno-
hyops and Palaeosyops, longer than in Manteoceras;
distal phalanx of D. Ill widely spreading and sharply
Front view; As, outer side view.
One-sixth natural size.
truncate distally, longer than in Manteoceras, median
cleft fairly marked. Femur slender with prominent
trochanters, with patella facet oblique to long axis of
shaft. Tibia long and slender, about y^ of the length
of the femur, with deeply concave facet for astragalus.
Pes with high tarsals, slightly divergent digits, narrow,
straight-sided metatarsals. Astragalus relatively high
and narrow, with narrow trochlea and elongate neck,
cuboid facet narrow, sustentacular facet narrow, ver-
tically elongate. Calcaneum elongate with narrow
sustentaculum and deep, laterally compressed tuber.
Navicular relatively deep vertically, forming a quad-
rant in superior view. Entocuneiform vertically oval,
pointed at top. Mesocuneiform anteroposteriorly
elongate, narrow. Ectocuneiform relatively deep,
shallow anteroposteriorly, and narrow transversely.
Cuboid deep vertically, shallow anteroposteriorly,
with subquadrate superior and inferior facets. Meta-
tarsals long, straight sided,
distal facet of Mts III
transversely cylindrical,
proximal facets truncate
posteriorly (not deep as in
Palaeosyopinae). Phalanges:
Proximal phalanx of Mts III
relatively . long, narrow;
middle phalanx wide and
shallow ; distal phalanx rela-
tively long, very broad
distally, sides obliquely
truncate.
The atlas (Am. Mus. 1523)
shows (fig. 560) a subfamily Figure 562
agreement with that of
Manteoceras and with that
of DolicJiorJiinus and con-
trasts with that of Palaeo-
syops in the following char-
acters: The cotylus is narrow transversely (93 mm.)
but deep vertically (40 mm.); its superior border
is deeply concave anteriorly; the neural arch is
produced above into a circular rather than an elon-
gate hillock; the neural tunnel is more roundly
arched above; the postero-median inferior process
below the odontoid is acuminate, not peg-shaped;
the axis facets do not diverge so widely — that is,
they face more inward; the pleurapophysial flanges
are not preserved; the vertebrarterial canal was large.
The atlas is distinguished from that of Manteoceras by
its relative narrowness and depth, both anteroposte-
riorly and vertically.
A humerus (Am. Mus. 12385) provisionally referred
to Mesatirhinus megavMnus from Bridger C 3 agrees
in most characters with that of Manteoceras (Am. Mus.
12204) but is smaller and more slender. In length
i (260 mm.) it is far shorter than the humerus of
-pr.sti'.rd,
Xz
Radius and
ulna of Mesatirhinus
petersoni
Am. Mus. 1571; Washakie A. Ai,
Right ulna, outer side view; A2,
right radius and ulna, front view.
One-sixth natural size.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
639
Pdaeosyops leidyi (325 mm.). The supinator crest is i Palaeosyops sp. (35). The external proximal facet for
relativeh" shorter. It is sharply separated from the
Ai Az
FiGUKE 563. — Left forearm and manus of Mesa-
tirhinus petersoni?
Princeton Mus. 10013; Bridger C or D. Ai, Front view; A2,
outer side view. One sixth natural size.
humerus of the contemporary cursorial rhinoceros
Hyrachyus by the marked asymmetry of the radial
facet, the marked down-
ward extension of the
deltoid ridge, the stouter
supinator crest, and the
heavier proximal end. It
is smaller than even the
smallest (Am. Mus. 1571)
of the three forearms
referred to Mesafirhinus
petersoni.
The radius of M. peter-
son (Am. Mus. 1571,
Princeton Mus. 10013) is
remarkable for its length
(233-245 mm.), its slen-
derness (circumference of
shaft just above middle,
72-75), and the narrow-
ness of its proximal end
(tr. 49, 50) and distal end
(tr. 49, £0). By way of
comparison, in Palaeo-
FiGURE 564. — Right manus and syops copei (Am. Mus.
fragments of radius and ulna 12205) the corresponding
of Mesaiirhinus petersoni measurements are length
Am. Mus. 1571; Washakie A. A, Right 907 -iv ± -u jxi,
manus; B, distal end of radius and ulna. ^'^' nUJuniCters, Dreaatll
Facets for scaphoid (sc), lunar (lu), and proximal end 62, distal
cuneiform (cu) . One-third natural size. j „►, • r r
end 67, circumference of
shaft 87. The proximal end of the radius of M.
petersoni is also shallower (ap. 30 mm.) than in
101059— 29— VOL 1 14
the capitellum of the humerus is much shallower
anteroposteriorly (20 mm.), more deeply concave
Figure 565. — Left manus, radius, and ulna
of Mesaiirhinus petersoni
Princeton Mus. 10013; upper Bridger. Ai, Front view of
manus; Az, top view of carpus; As, inner side view of car-
pus; Bi, radius and ulna, distal view; B2, radius, proximal
view. One-third natural size.
anteriorly, and continues externally into a high
anteroposterior ridge which is higher than that in
Palaeosyops. This ridge is supported interiorly by the
angulate external border
of the shaft, which slopes
downward and inward at
a gentle angle, whereas in
Palaeosyops the rounded
external border rapidly con-
tracts into the shaft so that
the proximal end of the
bone expands widely. The
shaft of the radius arches
forward a little less strong-
ly than in uncrushed speci-
mens of Palaeosyops, but,
as already noted, is much
more slender. The distal
end is much less expanded
transversely and relatively
deeper anteroposteriorly;
the internal distal process
is sharply prolonged down-
ward. The radius of M. petersoni (Am. Mus. 11659)
is badly crushed but differs from those above de-
FiGURE 566. — Right scaphoid
of Mesatirhinus and Man-
teoceras
Front and top surfaces. Ai, As, Mesa-
iirhinus petersoni, Princeton Mus.
10013; Bi, B2, Manieoceras manteo-
ceras, Am. Mus. 12204. One-half
natural size.
640
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
scribed in its greater length (258 mm.) and heavier
shaft.
The ulna of M. petersoni (Princeton Mus. 10013) is
slender (length 310 mm.) and slightly more curved
anteroposteriorly than in Palaeosyops. The olecranon
is relatively thicker transversely and is subtruncate
rather than pointed posterosuperiorly ; the dorsal
extension of the humeral facet is narrow. The antero-
external face of the shaft faces more obliquely out-
ward; this is partly because the external longitudinal
ridge is much lower than in Palaeosyops. The distal
end of the shaft curves backward more sharply. The
distal cuneiform facet is nearly at right angles to the
main axis of the shaft, and the external or styloid
process is not pronounced, whereas in Palaeosyops the
cuneiform facet is more inclined to the shaft, and the
styloid process is more pronounced.
Manus (Am. Mus. 1571, Princeton Mus. 10013):
The chief characteristics of the manus in comparison
with those of Palaeosyops are its narrowness and
vertical height, the straight-sided character of the
metacarpals, and the shortness of the phalanges, so
that it represents the extreme dolichopodal stage
known among the Bridger titanotheres. It shares
several of these characters with Manteoceras, as well
as many of the more detailed characters of the carpals
and metacarpals, and differs from that form chiefly
in its greater narrowness. The general measure-
ments are given on page 636. In the measurements
given below more in detaU the first always refers to
Princeton Mus. 10013, the best-preserved specimen,
and the second to Am. Mus. 1571. The scaphoid is
relatively deeper vertically (25, 27 mm.), shallower
anteroposteriorly (36, 35), and broader posteriorly
(23) than in Palaeosyops (Am. Mus. 12205); the radial
facet is flatter, the trapezoid and scaphoid facets
form a more open angle (Princeton Mus. 10013 only);
the facet for the capitellum of the magnum is wider;
there is a close general agreement with the scaphoid
of Manteoceras (Am. Mus. 12204), save that the bone
is narrower and the trapezoid facet smaller. The
lunar is relatively narrow (27 mm.) and deep (32, 33
mm.) on both the anterior and posterior faces; the
inferior end is more sharply wedge shaped, the mag-
num facet being sub vertical in front view; the lower
facet for the cuneiform is broader. In Palaeosyops
the lunar is separated in front from the cuneiform
by the dorsal ridge of the unciform. The cunei-
form is narrow (26 mm.), not extended postero-
externally, with ulnar and unciform facets com-
paratively deep anteroposteriorly (Princeton Mus.
10013, crushed in Am. Mus. 1571); facet for pisiform
less elongate; cuneiform narrower than in Manteo-
ceras (Am. Mus. 12216). The pisiform contrasts in
many characters with that of Palaeosyops: the distal
end of its tuber is less expanded vertically and thicker
transversely; its ulnar facet is concave and triangular,
that of Palaeosyops is convex and with rounded con-
tour; inferiorly its cuneiform facet is rounded, that of
Palaeosyops is deeply angulate; the head, or proximal
end, is sharply constricted from the shaft, that of
Palaeosyops rises gently from the shaft. The trape-
zium is small (greatest length 22 mm.), pear-shaped
(resembling a small patella) rather than broadly
ovate {Palaeosyops); a distinctive feature is that it
articulates with the scaphoid. Trapezoid relatively
wide (tr. 19 mm.), flat, anteroposteriorly shallow
(16 mm.), and more truncate posteriorly, not pro-
duced anteroposteriorly into an oblique projection;
also shallower anteroposteriorly than in Manteoceras.
Magnum (Princeton Mus. 10013) rather small,
scaphoid facet broad, flat, and subhorizontal; posterior
hook broadly spatulate (Am. Mus. 11659); facet for
Mtc II large and sharply ridged; magnum thus
agreeing in general with that of Manteoceras but
smaller and with narrower capitellum. The unci-
form is narrow (extreme width 37 mm.) and deep
vertically (diameter at right angles to long axis
25 mm.); lunar facet subquadrate, ridge separating
lunar from cuneiform low and not greatly produced
posteriorly; postero-external process with relatively
slender base and subpyramidal top, agreeing in facets
with unciform but entire bone narrower. As noted
above, the metacarpals (Am. Mus. 1571, Princeton
Mus. 10013) are elongate, straight-sided, subcylin-
drical rather then flattened, and more nearly parallel
with each other than in the spreading manus of
Palaeosyops; the distal facets are somewhat flatter
(less subglobose); the fifth metacarpal (Mtc V) is
relatively much longer and narrower. Distinctions
from Manteoceras are found chiefly in the greater
narrowness and in the obliquely triangular rather
than the posteriorly truncate broad proximal facet of
Mtc II and III.
Considered more in detail: The first metacarpal,
as in other perissodactyls, is entirely wanting, imless
it is represented possibly by the distal part of the
trapezium. The second metacarpal (length 110 mm.,
maximum distal width 25) has the trapezoid facet
deeply concave in front and produced postero-
iaternally into a blunt tip, imlike both Palaeosyops
and Manteoceras; the trapezium facet, as in Manteo-
ceras, is small and confined to the postero-external
border; the facet for the magnum forms an elongate
rectangular, nearly plane sin-face rather than an irreg-
ularly warped band, it is also more shallow posteriorly
than in Manteoceras; the facet for Mtc III is quite
small (contrast Palaeosyops). The third metacarpal
has the proximal facet (for the magnum) pointed
posteriorly instead of roundly truncate, as in Manteo-
ceras and Palaeosyops; the facet for Mtc II is very
small; the facet for the unciform is broadly triangular.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
641
whereas in Manteoceras it is intermediate; the facet
for Mtc IV is relatively smaller than ia Palaeosyops
and faces more downward than outward (Princeton
Mus. 10013); in Manteoceras it is intermediate. Mtc
IV has the proximal facet very different from that in
Palaeosyops; the unciform is flatter on top, and its
posterior part is not decurved so sharply; posteriorly
this facet is not so broad; the facet for Mtc III is
nearly divided into two triangular facets, whereas ia
Palaeosyops it forms a broad half ring; the facet for
Mtc V is much shallower; ia all these characters
Mtc IV approaches that of Manteoceras. The fifth
metacarpal offers a very wide contrast to that of
Palaeosyops; it is actually much longer (82 mm. as
compared with 75) while only about half as wide
(19 as compared with 36); the proximal end is pro-
duced externally into a high ridged prominence, which
embraces the unciform externally and causes the
unciform facet to face obliquely upward and inward;
the facet for Mtc IV is relatively narrow. In all
these characters except the extreme slenderness the
fifth metacarpal, lilve the fourth, approaches Manteo-
ceras manteoceras.
Although the third metacarpal is considerably
loager than in Palaeosyops the first phalanx of the
same digit is only about two-thirds as long (24 mm.)
as that in Palaeosyops (37); it is, however, nearly as
broad (23, estimated) as it is long (24) and therefore
has about the same proportions as ia Palaeosyops
(ap. 37, tr. 36); this phalanx is thus proportionately
longer than in Manteoceras. The distal phalanx of
the same digit is widely spreading and sharply trun-
cate distally, with a fairly marked distal .cleft; in
these features it approaches the corresponding phalanx
in M. manteoceras but is longer in proportion to its
distal breadth (ap. 19 mm., tr. 28, as compared with
17 by 33 in No. 1587, M. manteoceras). The remain-
ing phalanges call for no special remark.
Mesatirhinus petersoni?
The manus Am. Mus. 11659, a part of the larger
skeleton from Bridger C 5, differs from those of
Am. Mus. 1571 and Princeton Mus. 10013 chiefly
in its larger size, as shown in the table of measure-
ments. It is pretty badly crushed but agrees well
in most details of the facets, etc.
The hind limb is preserved only in the larger skele-
ton Am. Mus. 11659. The femur approaches that
of HyracTiyus, first, in the large size of the third tro-
chanter, which is relatively a little farther down the
shaft than ia Palaeosyops major; second, ia the position
of the patellar facet, which is more nearly at right
angles to the long axis of the bone, whereas ia Palaeo-
syops major it is prolonged upward and backward
and becomes nearly parallel to that axis. But these
characters are approached in the femur of Manteoceras
aad besides being associated with a manus and an
astragalus of Mesatirhinus type, the femur itself is
separable from that of HyracTiyus by various differ-
ences in the shape of the head, great trochanter, and
distal end. The femur (length 358 mm.) is shorter
than in Palaeosyops leidyi (370 mm.), but its relative
length as compared ^with the tibia (T 79, F 100) is
the same as in Palaeosyops. The chief distinctions
from the femora of Palaeosyops, Telmatherium, and
Manteoceras lie in the greater slenderness of the shaft
and probably also in the position of the patellar
facets (see above).
The tibia, though somewhat crushed, was long
(283 mm.) and slender; its distal third was relatively
steeper anteroposteriorly and narrower transversely
than in Palaeosyops major; a marked difference is
seen in the region of the astragalar facets; in front
view the facet for the internal keel of the trochlea is
deeply incised and bounded
internally by a promiaent
vertical malleolar process; in
iaferior view the same facet
is broad posteriorly, whereas in
Palaeosyops it is narrow pos-
teriorly; the facet for the ex-
ternal half of the trochlea is
deeper anteroposteriorly and
less produced antero - exter-
nally.
The pes (Am. Mus. 11659),
like the manus, is of the long,
narrow type, with high tar-
sals and straight-sided meta-
tarsals.
The astragalus (Am. Mus.
11659) is considerably larger
than the one that is associated
(Am. Mus. 1571) with teeth of M. petersoni type; the
navicular facet is also relatively deeper anteroposteri-
orly; the cuboid facet, the sustentacular facet, and the
neck all seem relatively a little wider. But notwith-
standing these differences, generic affinity is indicated
by the following characters in common, which serve to
separate these two astragali from those of other
genera. As compared with that of Palaeosyops
the whole bone is long (vertically high) and narrow,
with relatively narrower trochlea and neck; internal
or tibial keel of trochlea sharply rather than roundly
convex, inner slope of external trochlear keel flatter,
ridge bounding navicular facet superiorly not sharply
projecting, navicular facet shallower anteroposte-
riorly, cuboid facet narrower, less sharply inclined to
the long axis of the navicular facet; sustentacular
facet narrow, straight-sided, lying on the extreme
internal (tibial) side of the posterior face, broadly
Figure 5 6 7. — Right
manus of Mesatirhi-
nus petersoni?
Am. Mus. 11659; Bridger C 6.
One-third natural size.
642
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
continuous below with the cuboid facet, with which it
forms a well-defined L. The astragalus is distin-
guished from that of Manteoceras by the characters
noted above, chiefly the greater narrowness of the
bone as a whole, the sharper internal keel of the
trochlea, the narrower cuboid and sustentacular
facet. The astragalus approaches those of the
forerunners of LimnoJiyops from Bridger B (Am.
Mus. 11689, 11690), the principal differences noted
above being the relative narrowness of the bone as a
whole, the truncate inner border of the navicular
Figure 568.^Left
femur and tibia
of Mesatirhinus
peiersoni?
Am. Mus. 11659; Bridger
C.5. Front view. One-
sixth natural size.
Figure 569. — Left pes of Mesatirhinus
petersonif
Am. Mus. 11659; Bridger C 5. Ai, Front view;
As, inner side view; A.i, phalanges of digit Ill-
front view; A*, ungual phalanx of digit III,
top view (a different individual from A3).
One-third natural size.
facet, the somewhat narrower and straighter susten-
tacular facet.
The calcaneum (Am. Mus. 11659) in its total length
(93 mm.) is nearly as long as that of P. leidyi (101 in
Am. Mus. 1589), but its total width across the susten-
taculum is only 47 millimeters as compared with 60
in P. leidyi; the tuber calcis as seen from above is
more straight-sided and its distal end less expanded
and rugose than in Palaeosyops; the sustentacular
facet is elongate-oval (vertical diameter 32 mm., tr.
13) rather than broadly oval, as in Palaeosyops.
The navicular is nearly as deep vertically as in
Palaeosyops but is much smaller both anteropos-
teriorly and transversely; in superior view it forms a
quadrant.
Aj (cH
Figure 570. — Left astragali
of Mesatirhinus petersonif
A, Am. Mus. 11659; B, Am. Mus. 1571.
Ai, Bi, Back view; A2, B2, front
view. Astragalocalcaneal facets;
ectal (ect), sustentacular (sus), and
inferior (inf). One-third natural size.
The entocuneiform (Am. Mus. 11659) forms a vertical
oval (vertical 29 mm., tr. 22), which is broad below,
pointed at top, and obliquely truncate anterosupe-
riorly by the facet for the navicular; in Palaeosyops
(Am. Mus. 1589) this bone is much wider below, not
so high vertically, and ends
above either in a sharp
angle or in a low, rounded
hillock. The navicular
facet is subcircular, espe-
cially at its upper end,
whereas that of Palaeosyops
is either large and ovoid
(Am. Mus. 1589) or broadly
rounded (Am. Mus. 11682);
the facet for the meso-
cuneiform forms a decided
angle with the navicular
facet, while in Palaeosyops
it is more nearly in the same
plane. The lower end of
the posterior or internal
face shows two prominent
features — an oval facet for Mts II near the anterior
border, and near the posterior border a large, rounded
protuberance, probably for the attachment of a liga-
ment; this protuberance is wanting in Palaeosyops and
at first adds to the difficulty of homologizing the widely
different borders and facets in the entocuneiform of
the two genera.
The mesocuneiform (Am. Mus. 11659) is a small,
anteroposteriorly elongate (20 mm.), narrow (10 mm.)
bone, in top view differing widely from the triangular
mesocuneiform of Pal-
aeosyops.
The ectocuneiform
(Am. Mus. 11659,
fig. 523) as compared
with that of Palaeo-
syops is deep vertically
(18 mm.; 19 in P.
leidyi), shallow antero-
posteriorly (33 mm.;
40 in P. leidyi), and
narrow transversely
(21 mm.; 25 in P.
leidyi) ; there are cor-
responding differences
s. Am. Mus. 1550; B, M. peter- . t r J i
soni.'. Am. Mus. 11669. Ai, Bi, Inner side ID- the laCCtS, and the
view; A., B., outer side view. One-half antcro-external f a C C t
natural size. ...
for the cuboid is lacking.
The cuboid (Am. Mus. 11659) as compared with
that of Palaeosyops is also deep vertically (31 mm.)
and much more shallow anteroposteriorly (27 mm.);
the superior and inferior facets are subquadrate rather
than anteroposteriorly elongate. As compared with
that of Limnohyops from Bridger B (Am. Mus.
A 2
(TTzisH)
Figure 57 L — Left entocuneiform
tarsi of Palaeosyops and
Mesatirhinus
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
643
11690) the cuboid is narrower transversely, with some-
what differently shaped facets and a larger postero-
external process.
The metatarsals are long and straight-sided. The
measurements show that Mts III, although several
millimeters longer than in P. leidyi (Am. Mus. 1589),
is only -f^,; as broad near its distal end; the proximal
facet (for the ectocuneiform) is more truncate pos-
teriorly; as in the manus the distal facet of Mts III
is more transversely cylindrical and less convex
transversely than in Palaeosyops, but the distal facets
of Mts II and IV, being narrower, are more convex
or subglobose transversely.
The proximal phalanx of Mts III as in the manus
is relatively longer (ap. 30 mm.), narrower (tr. 28 mm.),
and vertically shallower (17 mm.) than in Palaeosyops.
The middle or second phalanx, on the contrary, is
relatively wider (tr. 25 mm.) and much shallower (13
mm.). The distal or ungual phalanx of Mts III is
again rather long (27 mm.), narrow proximally (19
mm.), and very broad distally (28 mm.), the extreme
tip being transversely and the sides obliquely truncate.
This phalanx therefore differs widely from that of
Palaeosyops leidyi (Am. Mus. 1550), which is shorter
(23 mm.) anteroposteriorly and roundly spatulate
distally.
Comparative measurements of the pes in DolicTiorJiininae, in millimeters
Mesatirhimis
sp., Am. Mus.
2352;
Washakie
(B?)
Mesatirhinus
petersoni?.
Am. Mus.
11659;
Bridger C 5
Metarhinus
sp., Am.Mus.
2058;
Uinta B 2
Dolichorhinus
hyognathus,
Am. Mus.
13164;
Washakie B
Dolichorhinus
hyognathus,
Am. Mus.
1845;
Uinta B 2
Astragalus, height of inner face
Astragalus, breadth of trochlea
Astragalus, height of sustentacular facet.
Aftragalus, width of cuboid facet
Calcaneum, height
Calcaneum, width across sustentaculum..
Mts II, length
Mts II, distal width
Mts III, length
Mts III, distal width
Mts IV, length
Mts IV, distal width
65
50
39
11
128
54
53
40
35
9
94
48
110
69
55
30
12
147
37
136
26
120
28
110
120
37
108
21
« Estimated.
MesatirhJnus?
A large unassociated pes from the Washakie Basin
(Am. Mus. 2352), of uncertain level, resembles in
many ways that of the referred MesatirMnus petersoni
(Am. Mus. 11659) described above but is much larger,
as shown by the measurements given below. The
proportions tend to dolichopody. As compared with
that specimen the astragalus is a little broader in
proportion to its height, and the cuboid facet is a
little more prominent, both progressive characters;
the sustentacular facet agrees with that of Mesa-
tirhinus and contrasts with those of Dolichorhinus and
Metarhinus in its long, straight-sided character; but
it is sharply separated from the cuboid facet, a very
exceptional condition. The calcaneum has the stout
neck, unexpanded head, and narrow sustentaculum
characteristic of Mesatirhinus. The third and fourth
metatarsals and the second phalanges of D. Ill and
D. IV agree closely with those of Mesatirhinus save
in the larger size. The pes is thus distinguished from
those of Dolichorhinus hyognathus and Metarhinus by
the greater length of the metatarsals and by the char-
acter of the sustentacular facet of the astragalus; it
is also distinguished from that of Manteoceras by the
characters of the astragalus, in which the internal
keel of the trochlea is narrowly rather than broadly
convex, the cuboid facet is relatively narrower, the
sustentacular facet vertically longer, and the neck not
so short.
The pes, therefore, appears to represent a large
species of Mesatirhinus perhaps allied to Dolichorhinus
vallidens.
The question of the genei'ic reference of this pes is one
of great morphologic interest, which must be finally
determined by the discovery of another skeleton.
Its detailed characters (see fig. 572) are as follows:
(1) General proportions high and narrow; (2) Mts II
measures 145 millimeters, as against 118 in Dolichorhi-
nus hyognathus; (3) Mts IV measures 136 millimeters
ascompared with lOSinD. hyognathus; (4) the astrag-
alus exhibits an exceptionally high and narrow sus-
tentacular facet separated inferiorly from the small
inferior facet; (5) the calcaneum exhibits an elongated
tuber calcis (138 mm. as compared with 114 in D.
hyognathus).
Another large pes from Washakie B (Am. Mus.
13175) is of the same dolichopodal type and appar-
ently of the same species. It is certainly a member
644
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
of the Manteoceras-Dolichorhinus group rather than
of the palaeosyopine group. It does not belong to
the short-footed DolichorMnus Jiyognathus, but like
the foregoing (Am. Mus. 2352) may represent a long-
footed representative of this genus or of Mesatirhinus.
Metarhinus?
Provisionally referred pes from Uinta B 1. — To
Metarhinus are referred provisionally three hind feet
doubtfully recorded from the Metarhinus zone (Uinta
B 1), as follows: Am. Mus. 1950, a plaster cast from
a lost original; Am. Mus. 1947, including chiefly the
Figure 572. — Pes referred to Mesatirhinus
Am. Mus. 2352; Washakie B (?) . One-third natural size.
astragalus, calcaneum, and navicular; Am. Mus. 2058,
including among other fragments the astragalus and
metatarsals II and IV, from the top of horizon B.
These feet are referred to Metarhinus because no
skulls of Mesatirhinus but abundant skulls of Meta-
rhinus have been found at the geologic levels in which
they occur.
The astragalus represents an advance upon the
Mesatirhinus type, from which it differs in the follow-
ing progressive characters: Trochlea a little more
flattened on the anterior face and a little wider in
proportion to the length (vertical diameter) of the
bone; process for ligament on posterosuperior border
of internal face very pronounced; cuboid facet wider
(Am. Mus. 2058), sustentacular facet vertically
shorter and narrowing instead of broad superiorly,
ectal facet shallower. All these characters are seen
also in Dolichorhinus , in which, however, the astrag-
alus is larger, the sustentacular facet shorter and
Figure 573. — Pes of Meta-
rhinus cf. M. earlei
Am. Mus. 1950. Cast of left pes, partly
restored. One-third natural size.
Figure 574. — Astragalus, cal-
caneum, and navicular of
Metarhinus cf. M. earlei
Am. Mus. 1947. A, Front view of right
astragalus, calcaneum, and navicular;
B, distal view of astragalus and calca-
neum. One-half natural size.
even more pointed above, and the cuboid facet wider.
The calcaneum (Am. Mus. 1947) has a long neck,
narrow sustentaculum, and in general resembles that
of Mesatirhinus, but the sus-
tentaculum is smaller and more
oblique and the top of the
prominence bearing the ectal
facet also shows a considerable
facet for the tibia; the ectal
facet itself is flatter; in most
points except size it resem-
bles the calcanea referred to
Dolichorhinus.
The second metatarsal (Am.
Mus. 2058, 1950), closely resembles that of Mesati-
rhinus, and the same is true of the third (No.
1950) and the fourth (No. 2058). In the shape
of the facets they also resemble Dolichorhinus, but
they are distinguished by their slenderness. From
this close similarity in the pes to that of Mesati-
rhinus and also from the similarity in the dentition
and other parts we are led to expect that the
manus of Metarhinus will also be found to resemble
that of Mesatirhinus.
Figure 575. — Astragalus
of Metarhinus cf. M.
earlei
Am. Mus. 1947. Rear view
(drawing from kft astragalus
reversed). One-third natural
size.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
645
Comparative measurements oj the supposed hind feet of Metarhinus
sp., in millimeters
Metarhinus?
Mesa-
tirhinus
peter-
soni?,
Am.
Mus.
11669,
Bridger
C5
Am.
Mus.
1950
(cast),
Uinta
Bl
Am.
iMus.
1947,
Uinta
Bl
Am.
Mus.
2058,
top of
Uinta
Bl
Astragalus, height of inner face
Astragalus, breadth of trochlea
Astragalus, height of sustentaou-
lar facet-
55
40
53
39
29
8
87
44
57
44
32
10
53
40
35
Astragalus, width of cuboid facet
9
88
45
108
21
94
Calcaneum, width across susten-
48
Mts II, vertical length _
110
Mts II, distal width.
Mts III, vertical length.
120
Mts III, distal width
I
28
Mts IV, length- __. _ ..
106
22
110
Mts IV, distal width
1
Juvenile skeleton. — A skeleton of a newly born or
fetal animal, from Uinta B 1 (formerly called Uinta
upper A) was described in 1914 by Peterson (1914.2)
under the name Heterotitanops parvus. It consists
(fig. 578) of the greater part of the skeleton (Carnegie
Mus. 2909) including the skull and lower jaw, lacking
only the feet. As noted in Chapter VI (p. 426) the
skull and deciduous dentition of this animal present
important indications of relationship with some of the
smaller Dolichorhininae, presumably Metarhinus.
The vertebral formula, according to Peterson, is
approximately as follows: Cervicals 7, dorsals 16 or
17, lumbars 3 (?), sacrals 4 or 5, caudals 14 or 15.
This may therefore be practically the same as in Doli-
chorhinus — namely, cervicals 7, dorsals 17, lumbars
4 (?), sacrals 4.
The anterior face of the sacrum is quite even with
the supra-iliac border of the pelvis, a characteristic of
the titanotheres generally. The thoracic cavity was
of large size, as indicated by the rather long ribs.
There are apparently six bones in the sternum. The
scapula is titanotheroid in its general outline, the
spine being less overhanging than usual, which is
probably a juvenile character (Peterson). The other
limb bones are in a very immature condition but so
far as preserved suggest the limb proportions of Uinta
Basin titanotheres (Peterson).
Dolichorhinus
GEOLOGIC HORIZON AND GENERAL FEATURES
These peculiar long-skulled, short-footed animals
are known only from the Eohasileus-DolichorTiinus zone
(Washakie B 2 and Uinta B 2). They are readily
distinguished by their very long skulls, and so far as we
know they had short necks and relatively short, heavy
limbs. As they are partly adapted in the skull and den-
tition to grazing habits we should expect to find them
long-legged, or subcursorial, but they were not. The
fortunate discovery in Washakie B 2 of a specimen of
D. hyognathus (Am. Mus. 13164), in which the skull
and parts of the skeleton are associated, proves that
the hind foot of Dolichorhinus was brachypodal
(fig. 585). The contemporary perissodactyls of similar
size are Manteoceras and Sphenocoelus (an aberrant
titano there). Another contemporary is the peculiar
rhinoceros Amynodon, which is readily distinguished
by its long, slender feet.
SKELETONS REFERRED TO DOLICHORHINUS HYOGNATHUS
Materials. — Our knowledge of the skeleton of D.
hyognathus is very slight; it is based chiefly on remains
of two individuals. The first American Museum
Figure 576. — Left scapula of
Metarhinus? sp.
Am. Mus. 1873; Uinta B 1. One-sixtli
natural size.
Ai '^'A2
Figure 577. — Left radius and
ulna of Metarhinus earlei ?
Am. Mus. 2363; Wasliakie B 1. Ai,
Outer side view; A2, front view. One-
sixth natural size.
specimen (No. 1843) was found by Mr. O. A. Peter-
son in 1894, in horizon B 2 of the Uinta Basin, Utah.
It consists of a nearly complete vertebral series, a
part of the pelvis, the anterior part of the skull, the
humerus, and one rib. The second specimen (Am.
Mus. 13164) was found by Mr. Paul Miller in 1906, in
Washakie B; it consists of the finely preserved skull
already described, with which were associated parts
of the atlas and axis, one lumbar, parts of the scapula,
the humerus, the proximal half of the ulna-radius, the
femur, two metatarsals, and isolated foot bones.
Another specimen also found in Uinta B 2 (Am. Mus.
1836) consists of parts of the radius and ulna, associ-
ated with the jaw. An atlas (Am. Mus 1837) was
found associated with the skull of D. intermedius.
A number of other bones not associated with cranial
material have been referred to this genus, especially an
atlas (Am. Mus. 1844); also a part of the scapula
(Am. Mus. 1833) and a radius and ulna (Am. Mus.
1831). The manus, unfortunately, is not known.
The materials above enumerated enable us to make a
646
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
partial and provisional restoration (fig. 579) of this \ Peterson (Carnegie Mus. Mem., vol. 9, pt. 4), is based
peculiar animal. This tentative restoration is un- on much better material and is doubtless more
doubtedly incorrect in many details, especially in the j accurate.
Figure 578. — Skeleton of a newly born animal, provisionally identified as Metarhinus sp.
C.irnegie Mus. 2909; Uinta B 1; tjT)e of Beteroiiianops parvus Peterson. After Peterson. One-fourth natural size.
Figure 579. — Provisional restoration of the skeleton of Dolichorhinus hyognathus
One-fifteenth natural size. Based on the following specimens in the .American Museum of Natural History: 1843, Uinta B 2,
anterior half of skull with lower jaw, vertebral column, sacrum^ and part of pelvis; 13164, Washakie B, remaining parts of
skull, humerus, portions of radius and ulna, femur, Mts III, IV; 1833, Uinta B 2, scapula. The remaining parts, which
are more or less hypothetical, are based on Mesathhinus, with modifications supplied by fragments from Uinta B 2. The
number of dorsolumbar vertebrae shown in this restoration (19) is incorrect, for the last two dorsal vertebrae are omitted.
The complete vertebral column of Dolichorhinus longiceps in the Field Museum, Chicago, has 17 dorsals and 4 lumbars.
The manus as restored is too high and slender. (Compare PI. XXXII.)
limbs, which were very incompletely known at the
time the restoration was made. The tibia as restored
is too long, the femur as preserved in No. 13164 is
much shorter than in D. longiceps. The restoration
of the skeleton of D. longiceps, figured in 1924 by
General proportions as displayed in the composition of
the two principal skeletons (fig. 579). — The total length
of the animal with the head outstretched — that is,
measured from the premaxillaries to the ischium — is
estimated at 2.02 meters (6 feet S inches), the height
647
at the shoulder 1.09 meters (3 feet 6 inches). Alto-
gether the vertebral column is mechanically superior
in the strength of its muscular attachments to that of
Palaeosyops. We observe certain analogies to the
spinal column of Equus. The backbone is adapted to
the support of the long, depressed head; the broad
neural spines of the dorsal vertebrae serve for the
attachment of the ligaments and muscles supporting
the cranium. The spines in the lumbar region were
also deep, indicating the presence of powerful running
muscles. This apparent adaptation of the backbone
in the lumbar region to a strong running action is to
our surprise not correlated with length of limb or of
foot, because both the humerus and the hind feet are
relatively abbreviated. The median metatarsal meas-
ures only 120 millimeters. Other striking characters
of the restoration are the broad inferior lamellae of
FiGUHE 581. — Atlas referred to Dolichorhinus sp.
Am. Mus. 1844; Uinta C; dorsal view. Ttie arrows indicate the
course of the first spinal nerve and vertebral artery. One-third
natural size.
the cervical vertebrae (C. 3-C.
ments in Am. Mus. 1843 are:
6). The measure-
Actual iength curve of back, axis to spine of posterior
1. 35
Length of neck lacking axis — that is, C. 2-C. 7 .30
Vertebral column. — The vertebral formula as pre-
served in Am. Mus. 1843 appears to be cervicals,
7; dorsals, 15-1-; lumbars, 4; sacrals, 3-4. The exact
number of dorsals is not revealed by this specimen.
In the specimen of Dolichorhinus longiceps described
by Eiggs (1912.1, p. 31), the formula is, dorsals, 17;
lumbars, 4; sacrals, 4. In Palaeosyops the number
is not certainly known. In two genera of Oligocene
titanotheres {Brontotherium, Menodus) the number
of dorsolumbars is believed to be 20.
The atlas (fig. 581) is partly known from a specimen
in the American Museum (No. 13164) and fully known
by comparison of this specimen with a much larger
atlas (Am. Mus. 1844). It is moderately elongate; its
breadth is 203 millimeters. The vertebrarterial canal
traverses the base of the transverse processes. The
atlas is narrower transversely than that of Palaeosyops
and it has narrower pleurapophyses (201 mm. in
Metarhinus as compared with 240 in P. rohustus) ; the
superior border of the cotylus is deeply concave
anteriorly, the dorsal prominence (neural spine) is
larger, the cotyli are larger, and the articular sur-
faces for the axis form a more deeply concave sinus.
648
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
As compared with Palaeosyops the cervicals had
sHghtly longer and relatively smaller centra, shghtly
longer prezygapophyses and postzygapophyses and
perhaps larger pleurapophysial flanges on C. 1- C.
6; the spines are broken off but may have been
more slender at the base. The anterior dorsals had
relatively somewhat smaller centra with pronounced
inferior keels; the neural spines were broader antero-
posteriorly; the anterior zygapophyses of D. 2 faced
upward and inward, making a decided angle with the
neural spine as in Manteoceras, whereas in Palaeosyops
the zygapophyses were nearly continuous with the
spine.
The neck, which measures only 320 millimeters in
length, is short, especially when compared with the
remarkably long skull, which is estimated at 540
millimeters.
The remainder of the column all belongs to one
animal (Am. Mus. 1843). The axis (fig. 580) does not
exhibit so high a spine as in Palaeosyops or Manteoceras,
but we recall the fact that BolichorMnus has a low
occiput. There is little evidence of high spines on
C 3-C. 5; the powerful ligamentum nuchae was sup-
ported by the high and extensive spines of D. 1-D. 8.
Cei'vicals 1-6 are characterized by widely expanding
and actually overlapping inferior lamellae, distinct in
C 6 from the pleurapophysis above (thus unlike
Palaeosyops). The centra are quite deeply opistho-
coelous and laterally compressed. The zygapophyses
face vertically and obliquely outward and inward from
C 3 to the anterior face of D. 1 ; they face horizontally
downward and upward in D. 1 to D. 12; beginning
with the posterior face of D. 12 to L. 4 the zygapophyses
are vertically placed, facing outward and inward and
more or less sigmoid or revolute in curvature, as in
certain lumbars of Palaeosyops. The neural spines
from D. 1 to L. 4 are extended anteroposteriorly in
marked contrast to the feeble spines of Palaeosyops.
The metapophyses are unusually broad; that of L. 4
articulates with the front border of the ilium as in
Eguus. The centra throughout are relatively deep;
the depth equals the height in the posterior cervicals
and anterior dorsals, but in the lumbars the height
slightly exceeds the depth. The centra are com-
pressed, or keeled inferiorly. There are apparently
but three true sacrals in this specimen, but the number
can not be ascertained positively.
The upward curvatm-e of the dorsolumbar region
of the column is greater than that represented in
Figure 580 but is correctly indicated in the restoration.
Figure 579.
Arches and limb hones. — A scapula (fig. 582) is doubt-
fully associated with Dolichorhinus; it presents rather
high and narrow proportions and measures 320 milli-
meters vertically.
The humerus is known from two specimens. It is
& highly characteristic and progressive bone, closely
resembling that of the large lower Oligocene titano-
theres, such as Brontotherium leidyi. The most dis-
tinctive Oligocene titanothere character is the very
high, thin, and platelike great tuberosity (tub. maj.)
with an erect anterior process. The humerus asso-
ciated with Am. Mus. 1843 (fig. 583) is short; it
measures 285 millimeters to the tip of the great
tuberosity, while the length of the shaft is 255 mUh-
meters. The humerus of the other specimen (Am.
Mus. 13164) belongs to an animal of larger size;
length of shaft, head to distal extremity, 315 milli-
meters; extreme distal width 95. The disparity in
size of the two humeri as compared with the approx-
imate equality in size of the respective skulls is a
puzzling feature.
The length of the forearm or ulna and radius can
only be estimated; it certainly is considerably less
than that of the humerus (estimated humeroradial
ratio 81). This indicates (see p. 733) that Bolicho-
rMnus was an animal capable of more speed than
Palaeosyops but of less speed than Mesafirhinus, in
which the humerus and radius are more subequal.
The most highly characteristic feature of the ulna
is the form of the olecranon process, which is obtuse,
highly rugose, and incurved, as shown in Figure 584, a
character which relates this animal to Mesatirhinus.
The manus of D. longiceps, figured by Peterson (Car-
negie Mus. Mem., vol. 9, pt. 4, pi. 54), is remark-
ably like that of Mesatirhinus but relatively some-
what shorter and more massive.
Little is certainly known of the pelvis beyond the
outline indicated in Figure 580.
The femur (Am. Mus. 13164, fig. 579) exhibits a
total length of 387 millimeters as compared with the
length of the humerus, 315, and with the total basUar
length of the skull, 540. The femur has the character-
istic straight shaft of the titanotheres generally and is
readily distinguished from that of the contemporary
Amynodon (fig. 518) by the more vertical position of
its patellar facet (in Amynodon these facets are placed
very obliquely) and the lesser development of the third
trochanter, which is very prominent in Amynodon, as
in aU other rhinoceroses.
Mts III and IV (Am. Mus. 13164, fig. 585) are of
almost brachypodal proportions, being no longer than
those of Mesatirhinus petersoni but much broader
(dimensions are given above). They show syngenetic
resemblances to Mesatirhinus and Metarhinus.
COMPARISON OF THE FOKE LIMB OF DOLICHORHINUS AND
AMYNODON
Before it was learned that Dolichorhinus was brachy-
podal a finely preserved fore limb (Am. Mus. 1961,
from Uinta B 2, figs. 586, 587) was provisionally referred
to that genus. It differs from Dolichorhinus, however,
and agrees with Amynodon in the detailed characters
of the upper arm and forearm, especially in relative
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
649
length and slenderness, form of the great tuberosity
and supinator crest of the humerus and of the ole-
cranon. The manus has the deep carpus, relatively
long metacarpals, and extremely abbreviate ungual
phalanges of Amynodon; metacarpal III is enlarged,
the foot being functionally mesaxonic.
UNASSOCIATED ASTRAGALI PBOVISIONALLY REFERRED TO
DOLICHORHINUS AND MESATIRHINUS
The tarsus is known from unassociated specimens
only. The structure of the tarsus is uncertain.
Three types of relatively large astragali are found in
Figure 582. — Left scapula
of Dolichorhinusf hyogna-
Am. Mus. 1833; Uinta B 2.
sixth natural size. Compar
ure 590.
One-
:Fig.
Figure 583. — Humerus of
Dolichorhinus hyognaihus
A, Am. Mus. 13164, Washakie B,
right humerus, front view; B, Am.
Mus. 1843, Uinta B 2, left humerus,
front (Bi) and distal (B2) views.
One-sixth natural size.
the Eoiasileus-DolicJiorhinus zone (Washakie B 2 =
Uinta B 2). These are sketched in Figure 588. B
(Am. Mus. 1845) is said to be associated with part
of a Dolichorhinus skull. It differs from Mesatirhinus
in the short sustentacular facet, deeply grooved above;
also in the short neck. Of this type another astraga-
lus (Am. Mus. 1838, fig. 588, C) has a broader cuboidal
facet and is wider. A (Am. Mus. 2352, from Wash-
akie B 2), which is not associated with other remains,
agrees with Mesatirhinus in the long neck and long
sustentacular facet. D (Am. Mus. 1962), although not
associated with other remains, possibly belongs to Tel-
matherium. (See above.) If B belongs to Dolicho-
rhinus it is certainly distinctive.
SKELETONS REFERRED TO DOLICHORHINUS LONGICEP8
Three partial skeletons that have been referred to
this species are loiown. Two are in the Carnegie
Museum at Pittsburgh. They were discovered in
1912 by Peterson (1914.3) in the upper levels of horizon
B 1 of the Uinta Basin Eocene, on White River, Utah.
The skull, mandible, and hyoid bones of this specimen
have been described in Chapter V; the vertebrae,
limbs, and feet, as described by Peterson," are noticed
below. The third skeleton referred to this species is
in the Field Museum at Chicago and was discovered
in 1910 by Riggs (1912.1) in the "upper Metarhinus
beds" (upper level of Metarhinus zone = Uinta B 1,
formerly included in Uinta A). A photograph (PI.
XXXII) and numerous measurements (see below) of
this skeleton were kindly supplied for this monograph
through the courtesy of Mr. Riggs.
SKELETON OF DOLICHORHINUS LONGICEPS COMPARED WITH THAT
OF EOTITANOTHERIUM (DIPLACODON?) OSBORNI
The description by Peterson (1914.1, pp. 132-137)
of the vertebrae and limbs of the specimen of Dolicho-
rhinus longiceps in the Carnegie Museum (No. 2865)
may be abstracted and restated as follows:
The atlas. — In comparing the atlas with that of Eotitanothe-
rium osborni Peterson, it is at once observed that the bone is
proportionally higher and longer but of a less transverse di-
ameter, which is due chiefly to the shorter transverse process in
the present genus. The anterior cotyle is on the whole very
nearly as large as but is deeper than in Eotitanoiherium, and
its inferior surface is more distinctly separated. The odontoid
process of the axis is proportionally longer and reaches nearly
through the inferior arch of the atlas, while in Eotitanoiherium.
Figure 585. — -Metatarsals
of Dolichorhinus hy o-
gnathus
Am. Mus. 13164; Washakie B.
Right metatarsals III and IV.
Ai, Front view; .Aa. proximal
view. One-third natural size.
Figure 584. — Radius and ulna of
Dolichorhinus hyognathus
Ai, Outer side view of proximal end of left
radius and ulna. Am. Mus. 13164, Wash-
akie B; A2, front view of same, combined
with distal end of radius. Am. Mus. 1831,
Uinta B 2; A3, rear view of left olecranon
shown in Ai. One-sixth natural size.
it does not. The articulation for the axis is much deeper than
in Eotitanotherium and not nearly as broad, in this respect more
nearly suggesting the condition found in some Oligocene rhi-
noceroses {Diceratherium) than the horned titanotheres. The
transverse process is pierced by a large foramen, unlike Eoti-
tanoiherium, in which this canal is smaU, or completely absent.
The axis. — The body of the axis is possibly somewhat longer
than in Eotitanotherium, the anterior opening of the arterial
canal located farther back, and the postzygapophysis is smaller
and less rounded in outline, while the neural spine and the
ventral keel have approximately the same general proportions.
" Peterson's final description of these skeletons (Carnegie Mus. Mem., vol. 9,
No. 4, 1924) was received too late for extended notice in this monograph.
650
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The other cervical vertebrae present no characters of sufficient
importance to mention in this connection.
The dorsal vertebrae. — The first dorsal has a short depressed
centrum and a prominent keel. The spine and transverse
processes are broken off. The other dorsal vertebra belongs
therium. The coracoid border above the notch is more curved
forward, as is also the glenoid border. The general outlines
of the scapula are on the whole more suggestive of the Rhino-
cerotidae than the titanotheres.
The humerus. — The humerus is short and heavy. The bone is
comparatively shorter than in Eotitanotherium. Unfortunately,
the greater tuberosity is broken on the posterolateral face, but
near the deltoid groove the superior face is complete and indi-
cates very plainly that the tuberosity is not as high as in
Eotitanotherium. The lesser tuberosity accords more nearly
with that shown in the latter genus. The deltoid groove is also
of about the same size in the two genera here compared. The
deltoid ridge is less prominent in Dolichorhinus, while the distal
end of the bone is quite nearly alike in the two genera.
A
Figure 586. — Manus of Amynodon and Mesatirhinus compared
A, The amphibious rhinoceros A. intermedius?, Am. Mu-s. 1961; B, the terrestrial
mediportal titanothere M. ■petersoni?, Princeton Mus. 10013. One-third natural
size.
well back in the series and has a higher and more evenly rounded
centrum, without ventral keel, but with the indication of a
heavy neural spine.
The lumbar vertebrae. — The two last lumbar vertebrae are
present; the body of the last being depressed, as is usual in the
case of the last lumbar, and has also the neural spine suddenly
reduced in the fore-and-aft direction. The transverse process
of the same vertebra is quite heavy and projects outward and
forward. Near the base of the process on the posterior face
there is a heavy and rounded process, which possibly came in
close contact with a similar process on the anterior face of the
pleurapophysis of the first sacral vertebra.
When the vertebrae described above are compared with the
vertebral column of Dolichorhinus hyognathus it appears that
the neural spine of the atlas of the specimen in New York is
more prominent, while the position of the transverse process
and the anterior exit of the vertebrarterial canal of the a.xis
appear to be the same in the two specimens. The cervical
series as a whole appear to be slightly shorter in the specimen
preserved in New York (D. hyognathus).
Measurements of Dolichorhinus longiceps
Millimeters
Atlas, greatest anteroposterior diameter 105
Atlas, greatest transverse diameter, approximately 180
Atlas, greatest vertical diameter 88
Axis, anteroposterior diameter of centrum, odontoid process
included 95
Axis, height, including neural spine 125
Cervical region, total length, approximately 395
The scapula. — The scapula is little if any shorter than in
Eotitanotherium, as figured by Peterson, but its general out-
lines differ from those shown in that genus. The lower portion
of the coracoid border is more deeply notched than in Eotitano-
FiGURE 587. — Left fore limb of the amphibious rliinoceros
Amynodon intermedius?
Am. Mus. 1961, Uinta B 2. Formerly referred to Dolichorhinus. Ai, Outer
side view; A2, front view of forearm and manus; B, front view of humerus.
One-sixth natural size.
The radius and ulna. — The radius and ulna are much shorter
than in Eotitanotherium and proportionally also much heavier.
There is a tendency to coossification of the two bones in the
present specimen, the shaft is rounder, and the articulation for
the humerus is less deeply excavated than in Eoliianotherium.
In comparing the ulnae of the two genera in more detail, it is seen
that there is a less developed tubercle on the outer margin of the
tendinal groove of the olecranon process in Dolichorhinus than
in Eotitanotherium. In consequence the groove is not as well
defined in the genus under description, though the termination
of the olecranon process is fully as well developed. In Dolichc-
rhinus there is a greater constriction of the olecranon between
the upper border of the great sigmoid notch and the termina-
tion of the process than is seen in Eotitanotherium. Otherwise
the ulna is cjuite similar in the two genera.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
651
The manus. — The manus of the specimen under description is
complete with the exception of the ungual phalanges and the
proximal phalanges of digits III and IV, which were not re-
covered. The foot as a whole is short and broad, and, when
compared with the manus of Eoiitanotherium, it may be said to
be heavier. In comparing the carpal elements of the two
genera it is at once observed that they are all of greater height in
the present genus than in Eotilanotherium.
Left astragali of Dolichorhinus and allied types
Front (Ai, etc.) and rear (.A2, etc.) views. A, MesaliTMnusf, Am. Mus. 2352, Washakie B? (of. figs.
569, 670, 572) ; B, Dolichorhinus hyognathus?, Am. Mus. 1846, Uinta B 2 (probably associated with
a skull of Dolichorhinus sp.) ; C, Dolichorhinus?, Am. Mus. 1838, Uinta B 2; D, Tdmatheriumf,
.-im. Mus. 1962. Uinta B 2. One-third natural size.
The metacarpals, in proportion to the carpals, are shorter than I
in Eotiianoiherium. The metapodial keel of Mtc II is less |
oblique to the long axis of the
bone than that in Eoiitanoiherium,
otherwise the differences between
these two genera are slight. The
liead of Mtc III differs from that
in Menodus by having the ulnar
portion more squarely truncated
and by the much smaller size of
the facet for Mtc II on the radial
angle. Mtc IV presents only
slight differences from the corre-
sponding bone in Oligocene titano-
theres. In its general details Mtc
V is quite similar to the same
bone in Eoiitanotherium but pro-
portionally shorter.
There are considerable variations in the length of the limb of
the genus Dolichorhinus. The humerus and the radius and
ulna of specimen No. 1961 ^s in the American Museum very
nearly agree in general length with those of No. 2865 in the
Carnegie Museum, while the fore foot of the former specimen is
considerably longer than in the latter.^' On the other hand, the
specimen No. 13164 (Am. Mus.) from the ?Washakie B level
indicates that the humerus is relatively longer and the fore foot
shorter than in the fore limb of Dolichorhinus in the
Carnegie Museum, which is described in this paper.
In conclusion this Carnegie Museum skeleton shows
that Dolichorhinus had a relatively shorter and wider
manus than that of Mesatirhinus — that it was in fact
almost brachj'podal.'"'
SKELliTON OF DOLICHORHINUS LONGICEPS IN THE
FIELD MUSEUM
The skeleton of Dolichorldnus longiceps in
the Field Museum (No. 12200), from Riggs's
"upper Metarhinus beds" ( = upper part of
Uinta B 1), comprises a well-preserved skull
and jaws associated with a vertebral column
which is complete from the atlas to the last
sacral vertebra; together with the left scapula,
both humeri, the left radius and ulna, and the
complete pelvis.
This important skeleton furnishes the proof
that in Bolichorhinus longiceps there were 21
-Ai
Figure 589.-
^2 ^3
-Cervical vertebrae of Dolichorhinus longiceps?
Carnegie Mus. 2865. Ai, Left side of atlas; Aj, anterior view of atlas; As, left side of axis.
natural size.
Measurements of Dolichorhinus longiceps (by Peterson)
Millimeters
Total length of scapula 337
Total length of humerus, head to distal end 285
Total length of ulna 340
Total length of radius '. 295
Total length of manus, approximately 200
Height of carpus at unciform and cuneiform 59
Transverse diameter of carpus at proximal row of carpals 90
Greatest length of Mtc II 116
Greatest length of Mtc III 124
Greatest length of Mtc IV 109
Greatest length of Mtc V 95
As in Eotitanoiherium and the titanotheres generally, the
phalanges are short, broad and depressed. In comparing
Osborn's restoration of Dolichorhinus (fig. 579) with the above-
described fore limb it appears that the foot of the present speci-
men is shorter, while the radius, ulna, and scapula are longer.
dorsolumbar vertebrae, 17 dorsals and 4 lumbars.
In the preliminary reconstruction of Dolichorhinus
33 The dimensions of the metacarpals of this specimen in millimeters as measured
by W. K. Gregory are a»s follows;
Length of
middle of
shaft
Breadth
above dis-
tal condyle
Mtc II
112
117
105
91
33
Mtc Hi
Mtc IV-
29
MtcV
23
'^ The manus referred to, Am. Mus. 19G1 (fig. 587), is probably referable to Amyno-
don sp. Its resemblances to Dolichorhinus, however, are remarkably close.
■"' Peterson's final illustrations of the manus of D. longiceps (Carnegie Mus. Mem.,
vol. 9, pi. 54, 1924) show that it is strikingly similar to that of Mesatirhinus, only
somewhat broader. Mtc III is relatively longer and narrower than Mts II. The
manus as a whole is of the compressed, straight-sided type, presenting a great
contrast to the spreading manus of Palaeosyops.
652
TITANOTHERES OP ANCIENT "WYOMING, DAKOTA, AND NEBRASKA
hyognatJius by Osborn and Gregory only 19 dorso-
lumbars were assigned to this animal, but prob-
ably the last two dorsal vertebrae were missing in
the specimen upon which this restoration was based
(Am. Mus. 1843, figs. 579, 580).
Without placing the specimens of D. Jiyognathus
and the more ancient D. longiceps side by side it is
difficult to determine from a comparison of the illus-
trations what are the really significant differences
between these two sets of vertebrae. In the D. lon-
giceps specimen the spine of the axis appears to be
more acute posteriorly, the spines of dorsals 6, 7, 8,
and of the posterior dorsals and lumbars seem to be
wider anteroposteriorly.
Measurements of the skeletons of Dolichorhinus longiceps and
D. hyognathus, in millimeters
Distance from tip of nasals
to last vertebra, as
mounted, on straight line;
right side
Length of skull, pmx to eon-
d3'les
Length of skull, nasal tip to
lambdoid crest
Width of skull across zj'go-
mata
P'-m'
Mi-m'
Pi-mg
Lower jaw, incisors to angle. .
Length of neck as mounted-.
Scapula, length
Humerus, length
Radius, length
Ulna, length
Os innominatum, right, length
from crest of ilium to tu-
berosity of iscliium
Number of dorsals
Number of lumbars
Number of sacrals
Pield
Mus.
12200;
Uinta
B 1
Carnegie
Mus.
2865
1,910
533
"570
''210
184
122
195
423
360
375
275
287
330
290
17
4
4
Am. Mus.
1843;
Uinta
B2
Am. Mus.
13164;
Washakie
337
285
295
340
248
C?)15-
542
580
250
205
119
230
420
'308
" Estimated. ' Modified by distortion.
SUBFAMILIES TELMATHERIINAE, BRONTOPINAE?, AND
DIPLACODONTINAE
Nothing was known of the skeleton of Uinta C
(true Uinta) titanotheres until the publication of the
memoir "The Mammalia of the Uinta formation,"
by Scott and Osborn, in 1890 (Osborn, 1890.51).
Parts of four skeletons were described in this memoir —
Princeton Mus. 10393, 10395, 10396, 10396a. These
were all attributed to Diplacodon elatus. It is now
evident that this specific reference is somewhat
doubtful, because, as we have already seen (p. 96),
Diplacodon elatus is only one of
four or five kinds of titano-
theres that were living contem-
poraneously during the period of
deposition of Uinta C. Since the
original Princeton expedition
the Princeton, American, and
Carnegie Museums have sent
parties into the same field,
which discovered parts of many
additional but very incomplete
skeletons. Unfortunately not
any of these skeletons are asso-
ciated with teeth or skulls in
such a manner as to enable us to
determine them specifically or
generically. It may be said,
therefore, that the skeletal char-
acters of the Uinta C titano-
theres are assigned only pro-
visionally.
As there were no less than five
kinds of titanotheres in Uinta
C, such skeletal remains as are
found there may belong to any
one of the following phyla: (1)
Successors of the Telmatherium
phylum; (2) members of the
true Diplacodon phylum, in
which the skull is long and
Peterson. One-tenth natural slender, rescmbliug in some re-
spects that of the Menodontinae,
in others that of ancestors of Brontotherium; (3)
descendants and mem-
bers of the Manteoceras
phylum, so far as
known only in the
lower beds of Uinta
C; (4) members of
the Protitanoiherium
phylum, animals of
robust and large size,
skeleton progressive
upon that of Manteo-
ceras, believed to be
ancestral to Brontops
of the Oligocene; (5)
descendants or rela-
tives of EotitanotJie-
rium, a progressive
titanothere so far
loiown only from
Uinta B 2.
With these precau-
tions clearly in mind,
I we may now describe the parts as provisionally referred.
Figure 590. — Right fore
limb of Dolichorhinus
Carnegie Mus. 2865. After
Figure 591. — Manus of Dolicho-
rhinus longiceps?
Carnegie Mus. 2865. Front view. After Peter-
son. One-third natural size.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
653
SUBFAMILY TEEMATHEEIINAE
Telinathedum ultimum (referred)
The hind limb of Telmatherium ultimum (Am. Miis.
1942), from Uinta C, indudes the femur, tibia, and
pes. It is evidently of palaeosyopine rather than
manteoceratine affinity. Its association with the
genus Telmatherium of the palaeo-
syopine group is provisional and
rests mainly on the characters of
the femur, tibia, and pes, especially
on the short neck and facets of
the astragalus, which are of the
palaeosyopine rather than of the
manteoceratine type. The tibio-
femoral ratio is 70.
The entire length of the hind
limb from the head of the femur
to the extremity of Mts III is
estimated at 965 millimeters,
distributed as follows: Femur,
crushed, estimated, 400; tibia,
actual length, 323 ; pes, estimated,
240.
The femur exhibits a very prom-
inent third trochanter, which is
considerably lower down on the
shaft than the second trochanter.
The palaeosyopine character is
seen especially in the broad patellar
facet, which is elongate and faces
anteriorly. The tibia presents a
slight anterior depression in the
cnemial crest.
The pes (fig. 593) is more dis-
tinctive. It is of an advanced
type in many details; its refer-
ence to Telmatherium rather than
to any of the Manteoceratinae is
Figure 592. Hind ^ue to the character of the astrag-
limb referred to Te?- alus, which is of the short-necked
matherium tilHvium type, with a broad oval susten-
Am. Mus. 1942; Uinta c. tacular facet; the very large tibio-
one-sixth natural size. ^stragalar trochlea is broad and
flat, presenting anteriorly; the navicular facet spreads
well on the front face of the astragalus; the cuboidal
facet is narrower than in Protitanotherium. The navic-
ular and ectocuneiform are also low and broad
elements of graviportal type.
All these bones are deep anteroposteriorly, and all
facets are subhorizontal or subvertical rather than
oblique. The metatarsals are likewise broad and in
detailed characters suggest those of the Bridger
Palaeosyopinae on a larger scale. The proximal
facets of the metatarsals are deeply extended antero-
posteriorly, as in the Palaeosyopinae, and unlike those
of Dolichorhinus, which are shallow, as in the Manteo-
ceratinae.
Skeletal remains originally referred to Diplacodon elatus by
Scott and Osborn
In 1890 parts of four separate skeletons from Uinta
C 1 (Princeton Mus. 10393, 10395, 10396, 10396a)
were described by Scott and Osborn (1890.1) and were
all attributed to Diplacodon elatus. It has been dem-
onstrated in Chapter V that Marsh's type of D. elatus
shows affinities to both Menodus and Brontofherium.
The elongate proportions of the skeletons now to be
described, moreover, resemble those of the Oligocene
Menodus rather than those of the Oligocene Brontops;
consequently it appears improbable that they belong
to Protitanotherium, which in cranial characters resem-
bles Brontops. The generic and specific references are
therefore uncertain.
The principal characters are as follows: (1) Short
neck, cervical vertebrae flattened and abbreviated; (2)
long limbs, pelvis and feet elongated as in Menodus; (3)
prominent and recurved hook on the humerus; (4) meta-
podials of manus elongated, stilted, functionally tetra-
dactyl in proportion; (5) radius long and slender (Am.
Mus. 2035); (6) tibiashort and massive(Am. Mus. 2056).
Cervicals. — It is noteworthy that the cervical centra,
although belonging to an animal nearly double the size
of Palaeosyops in height and breadth, are only a trifle
longer, a fact which points to the
progressive abbreviation of the
neck. The axis (Princeton Mus.
10396a) exhibits a broad spine
overhanging the postzygapoph-
yses, laminae very slightly
notched, postzygapophyses of
elongate oval form, transverse
processes hooked and perfo-
rated at the base, centrum with
a sharp inferior keel. The re-
maining cervicals and dorsals
(Princeton Mus. 10396) prob-
ably belong to a single indi-
vidual. The cervicals are prob-
ably the third, fourth, fifth, and
sixth. Of these C. 5 is the most
complete, spine pointed, verti-
cally placed and grooved pos-
teriorly, zygapophyses very
stout, facets placed at angles
of 45°, vertical diameter of
the centra much greater than
the transverse, opisthocoelous,
transverse processes not ex-
tending below the level of the centrum. The centra
of seven dorsals are preserved. Figure 594 repre-
sents one between the seventh and tenth exhibiting
the following characters: Centrum opisthocoelous
and considerably excavated at the sides; anterior,
posterior, and transverse diameters about the same;
lower half of the spine triangular, indicating a con-
siderable elongation and strongly oblique inclination;
Figure 693.— Pes of Tel-
matherium? ultimum?
Doubtfully referred left pes, Am.
Mus. 1942; Uinta C. One-third
natural size.
654
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
zygapophyses almost horizontal. The single lumbar
centrum preserved is considerably longer than the
above-described dorsal vertebrae and exhibits a stout
keel; spine broad and grooved posteriorly; zygapoph-
yses rounded and vertically placed.
midsection the spine much expanded along the border
and overhanging the postscapular fossa. The parts
preserved indicate that the scapula was lofty, with a
rounded suprascapular border unlike the somewhat
angular border of the Oligocene Menodus scapula.
Figure 594. — Vertebrae and fore limb of Diplacodon or Prolitanoiherium
Princeton Mus. 10390; Uinta C. After Scott and Osborn. Ai, A2, Front and side views of a cervical vertebra; B, a dorsal
vertebra (D. 7-10?), side view; Ci, Cj, side and rear views of a lumbar vertebra; D, incomplete left scapula; E, rear view
of left humerus; Fi, left humerus, radius, and ulna, outer side view; F, left radius and ulna with manus, front view,
oblique perspective. One-sixth natural size.
The scapula of the same individual (Princeton
Mus. 10396) is preserved, its total length being esti-
mated at 600 millimeters (fig. 594). The characters
are as follows: Coracoid process a stout tuberosity;
glenoid fossa elongate, a rather shallow oval; spine
ascending gradually from the neck and passing without
an acromion process into a deep, recurved ridge; in
The humerus (fig. 594) of the same specimen (Princeton
Mus. 10396) lacks the head, lesser tuberosity, and
bicipital groove. Its chief characters are as follows:
Stout deltoid ridge, terminating in a prominent
recurved hook; the inner length is estimated at 450
millimeters; shaft twisted upon itself, as in Rhinoceros;
supinator ridge less distinctly marked and less promi-
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
655
nent than in Menodus; pronator ridge rugose but not
projecting; supratrochlear fossa deeply excavated,
and trochlear process somewhat oblique to main axis
of shaft.
The ulna and radius are complete in the same
specimen (Princeton Mus. 10396, fig. 594) and taken
together indicate a rather long and slender forearm.
Other characters are as follows: Relative diameters
' (nv.J {cb)
-Astragalus and oalcaneum of Dipla
codon or Prolitanolheriuvi
Princeton Mus. 10396; Uinta C. Left astragalus and calcaneum,
crushed (?). Ai, Front view; As, outer side view. One-third
natural size.
of distal facets of radius and ulna about as 5 to 2,
ulna with a very stout, rugose olecranon, with the
posterior border presenting a single concave curvature,
with shaft triangular in midsection, and a deep groove
on the anterior face. The radius in midsection of the
shaft is suboval anteriorly and flattened posteriorly.
Facet for the entocondyle of the humerus with a deep
anteroposterior diameter. The manus (Princeton
Mus. 10396, fig. 594) lacks the carpus. The meta-
carpus exhibits a high, stilted tetradactyl or digiti-
grade type, the distinctive feature of the foot consist-
ing in the subequal size of the second and fifth
metacarpals, which brings the working median axis
of the manus between the third and fourth digits
instead of through the middle or third digit. This
is a decided progression upon any known Bridger type.
The measurements are as follows: Mtc III, length
180 millimeters; Mtc V, 130. Other measurements
may be taken from the figure.
The femur belongs to another animal (Princeton
Mus. 10395). The original drawing represents bones
belonging to two individuals of the same size drawn
in combination. This bone presents the essential
characters of the femora of titanotheres.
The tibia of the same animal (Princeton Mus.
10395) is about five-sevenths the length of the femur;
the cnemial crest is moderately prominent; there is
the usual triangular section of the shaft just below
the crest passing into an oval section in the lower
third; the proximal and distal faces are too much
worn to admit of description.
101959— 29— VOL 1 45
The somewhat crushed astragalus and calcaneum o
a tarsus in the Princeton Museum (fig. 595) are of
great interest. The principal feature of the calcaneum
is the extremely narrow, deep, and elongate tuber
calcis, which has an unusually flattened section. There
is also a distinct fibular facet, and the calcaneum forms
part of the tibial trochlea. The astragalus rests upon
over one- third of the upper surface of the cuboid;
the three astragalar facets — the ectal, sustentacular,
and inferior — are entirely distinct.
An ilium (Am. Mus. 2084) may also be referred
to this species. The measurements (estimated) are as
follows: Transverse width across ossa innominata 665
millimeters, total width of superior border of ilium
340, length of anterior border to acetabulum 335.
The superior border is well rounded.
A tibia (Am. Mus. 2056), rather short and massive,
having a total length of 325 millimeters, is provision-
ally referred to the same species, P. emarginatum.
SUBFAMILY BEONTOPIFAE?
Bones provisionally referred to Protitanotherlum superbum
A radius (Am. Mus. 2035) exhibits a length of 425
millimeters and perhaps may be referred to P. super-
lum. The length of the radius in the Princeton
Museum provisionally referred to JDiplacodon elatus
is 350 millimeters.
A large astragalus (Am-. Mus. 2030) has the breadth
(width across trochlear keels, 700 mm.) that seems
appropriate for this species. The cuboid facet is very
progressive (fig. 596).
Figure 596.— Left astragalus of ProlUanoOierium
superbum
Am. Mus. 2030; Uinta C. Front and rear views. One-third
natural size.
SUBFAMILY DIPLACODONTINAE
Pelvis referred to Diplacodon elatus
The only portions of the skeleton certainly associated
with the type of D. elatus are the cervical vertebrae
associated with the type skull in the Yale Museum
(No. 10320).
The pelvis (Princeton Mus. 10393, fig. 597) origi-
nally referred to Diplacodon elatus by Scott and Osborn
in "The Mammalia of the Uinta formation" (1890.7,
pp. 516, 517) may still be referred provisionally to
that species. The marked characteristic of this
pelvis is the great length of the ossa innominata as
compared with their breadth. The ilia did not expand
656
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
immediately above the acetabulum, but there is a long
and rather slender neck of the Uium beyond which the
borders begin to expand. The acetabular border of
the ilium (a) presents a short, sharp curvature and is
relatively much shorter than in Palaeosyops. The
ischiac border (b) is much longer, with a gradual
In Eotitanotherium oshorni the anteroposterior diameter of
the atlas is rather small, while transversely it is proportionally
greater than in the Oligocene forms. This is due in a great
measure to the longer transverse process of the Uinta form.
The cotyle for the occipital condyle is also deeper and the groove
for the odontoid process of the axis extends farther forward on
the inferior arch, due probably to the proportionally longer
odontoid in Eotitanotherium osborni.
The axis of the type is represented by a portion of the centrum,
the complete neural arches, and the spinous process. The arch
is somewhat depressed by crushing, but it is evidently of rather
large size. The vertebra as a whole possibly has a smaller
anteroposterior diameter than is the case in most of the titano-
theres of the Oligocene; the articulating surface for the atlas is
Figure 597. — Incomplete ilium and ischium of Diplacodon
elatiis
Princeton Mus. 10393; Uinta C; provisionally described. After Scott and Osborn.
One-sixtli natural size.
curvatme. It follows that the supra-iliac border
apparently presented outward more than directly
upward and forward. Below the acetabulum the
ischium has a triangular section, then expands in a
plane directly perpendicular to that of the ilium.
Eotitanotherium Peterson
Although the type specimens of Eotitanotherium
oshorni Peterson were found in the upper part of hori-
zon B 2 of the Uinta Basin, Utah, rather than in
horizon C (true Uinta formation), this animal is in a
stage of evolution in many respects similar to that of
the titanotheres of horizon C and therefore may be
conveniently described with them.
Peterson's description of the skull and dentition
of the type and paratype and comparison with the
Princeton specimen of Diplacodon or ProtitanotTierium,
figured above, have been cited in Chapter V; his
description of the rest of the skeleton (1914.1, pp.
37-51), with modifications in the nomenclature, is as
follows :
The atlas of the type (No. 2859) is quite complete. There is
also the greater portion of an atlas with the paratype (No.
2860). With regard to the posterior division of the arterial
canal it may be said that there appears to be some variation
in the Uinta species. Thus it is seen that in the type the base
of the transverse process is pierced by a small foramen (see fig.
598), while in the paratype there is no evidence of this foramen
on the posterior face of the transverse process. Of the later
Uinta forms there is apparently no atlas known. In comparing
the Oligocene titanotheres with the Uinta specimens before us,
there is a corresponding variation. The atlas of the Oligocene
types further varies in the anteroposterior diameter and in the
prominence of the neural spine and the transverse processes.
FiGUBB 598. — Atlas and axis of Eotitanotherium oshorni
Carnegie Mus. 2359 (type); Uinta B 2. Ai, Anterior view of atlas; As, posterior
view of atlas; Bi, posterior view of axis; Bj, lateral view of axis. One-third
natural size.
located more laterally, and the postzygapophysis has a greater
vertical obliquity and a more nearly rounded outline than in
the latter. In the Princeton specimen <' it is seen that the
arterial canal is located back of the posterior edge of the articu-
lation for the atlas, while in Eotitanotherium osborni the fora-
men is, on a direct side view, partially hidden by the backwardly
<i Scott, W. B., and Osborn, H. F., The Mammalia of the Uinta formation:
Am. Philos. See. Trans., vol. 16, pt. 3, p. 614, pi. 9, fig. 15, 1889.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHEEES
657
extended process of the articulation. I judge that the axis as
a whole, in the present form, is relatively shorter than in the
Princeton specimen. In more minute details the description of
Scott and Osborn (op. cit., p. 514) agrees well with the parts
preserved, in the specimen before me — that is, the heavy spine
pared. The second, third, fourth, and fifth dorsals have their
spines very nearly complete. In proportion they agree quite
well with those of the Oligocene genus but are more strongly
inclined backward. As in Menodus the transverse processes are
not extremely heavy and the capitular facets for the ribs are of
large size, while the sides of the centra are deeply concave. The
latter are deeper than broad, and the inferior borders, especially
the posterior ones, are distinctly more keeled than in Menodus.
Back of the eighth dorsal there is a break in the vertebral
column and a number of bones are lost. A second block which
was found, together with the one just described, contains por-
tions of six poste-
rior dorsals and
three lumbar verte-
brae. [See fig. 599.]
The neural spines
of the dorsal series
are prominent and
quite lumbar -like
in tlieir general
character. The
zygapophyses are
also of the inter-
locking lumbar type, and there
are prominent metapophyses.
The centra are somewhat mu-
tilated, but enough is preserved
to indicate that they are deep
of comparatively small
transverse diameter.
There are, as stated, three
lumbar vertebrae present in the
paratype (No. 2860). These
bones are fortunately found in
position succeeding the last dorsal vertebra, and for the
first time apparently furnish data as to the correct num-
ber of the lumbar vertebrae of the titanotheres. That
Figure 600. — Scapula of
Eotitanotherium osborni
Carnegie Mus. 2859 (type); Uinta
B 2. After Peterson. One-sixth
natural size.
Figure 599. — Vertebrae of Eotitanotherium osborni
Carnegie Mus. 2860 (paratype); Uinta B 2. After Peterson. A, Last
cervioal and dorsal vertebrae, right side; B, posterior dorsal and lumbar
vertebrae, left side. One-sixth natural size.
overhanging the postzygapophyses, the inner turn of the
transverse process, and a prominent inferior keel.
The succeeding four cervical vertebrae in the paratype
(No. 2860) are represented only by fragments. They
appear to have short opisthocoelian centra, as in Dipla-
codon described by Marsh and Osborn, and a prominent
ventral keel.
The seventh cervioal vertebra is completely worked out
in half relief and shows the chief characteristic features
[fig. 599]. The long and pointed spinous process is well
shown, as is also the neural arch and the centrum. The
pre- and post-zygapophyses are, as in the axis, located
quite laterally and face directly upward and downward as
in Menodus. The transverse process shows a tendency
to develop the broad round termination found in Brontops
validus of the Oligocene.
There are eight dorsal vertebrae, which are worked
out in half relief and rest on the original block of sand-
stone on which they were found. The neural spine
of the first dorsal is broken off about 10 centimeters
above the neural arch, but judging from the size of the
fracture, the spinous'] process attained a length equal,
and perhaps even proportionally greater, than was the case
in B. validus with which the Uinta remains have been com-
Ai A2 Bi B2
Figure 601. — Humerus, radius, and ulna of Eotitanotherium osborni
Carnegie Mus. 2860 (paratype); Uinta B 2. Ai, Anterior view of humerus; Ai, posterior view
of humerus; Bi, lateral view of radius and ulna; B2, anterior view. One-sixth natural size.
the last one of this series is the last lumbar vertebra there
is but little or no doubt, inasmuch as the neural spine
658
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
is very suddenly reduced in its fore-and-aft dimension and
also shows the presence of the very heavy transverse process
and the well-expanded postzygapophysis to meet the corre-
spondingly broad surfa,ces of the sacrum. Unfortunately the
greater portion of the centrum is weathered away, but from
what remains it appears that it was more depressed than
The sacrum is not represented. The oaudals appear to be
short and heavy and in other respects like those of the Oligooene
forms.
Figure 602. — Manus of Eotitanotherium
osborni
Carnegie Mus. 2860 (paratype); Uinta B 2. Ai, Supe-
rior view of pisiform; A2. lateral view; both one-third
natural size. B, Dorsal view of manus, one-sixth
natural size. After Peterson.
are those in front of it. Of the first and second lum-
bars the centra are large, sharply keeled, and the
transverse processes, though generally broken off, are
seen to have been prominent, though attenuated.
There are large metapophyses, and the neural spines
are high and of great anteroposterior diameter.
Measurements of Eotilanoiherium dsborni, in millimeters
Figure 603. — Femur, tibia, and pelvis of Eotitanotherium osborni
i, Distal end of femur, Carnegie Mus. 2860 (paratype); B, dorsal view of tibia, Carnegie
Mus. 2862 (paratype); C, lateral view of pelvis, Carnegie Mus. 2859 (type). After
Peterson. One-sixth natural size, a, Acetabulum.
Atlas
Greatest anteroposterior diameter
Greatest transverse diameter
Greatest transverse diameter of articulation for
occipital condyle
Vertical diameter of articulation for occipital
condyle
Axis
Greatest height
Greatest transverse diameter
Transverse diameter of postzygapophyses
Length of centrum of a median cervical vertebra.
Depth of centrum including inferior keel, approx-
imately
Seventh cervical, greatest height when vertebra
is in position
Seventh cervical, length of spine
Seventh cervical, anteroposterior diameter of
centrum
Second dorsal, greatest height when vertebra is
in position
Second dorsal, length of spine
Seventh dorsal, greatest height when in position.
Seventh dorsal, length of spine
Last dorsal, greatest height when in position
Last dorsal, length of spine
Second lumbar vertebra, greatest height when
in position
Second lumbar vertebra, length of spine
Caudal belonging to middle region of tail, length.
No. 2859 No. 2860
90
250
140
60
«138
158
70
95
250
138
60
37
45
195
120
70
300
■325
200
165
165
90
165
"95
29
' Approximate.
Figure 604. — Pes of Eotitanotherium osborni
Carnegie Mus. 2360 (paratype). Ai, Dorsal view of pes; Aj, posterior
view of astragalus. After Peterson. One-third natural size.
The ribs are represented only by a few fragments, and there
are no sternebrae.
The greater portion of the scapula is represented with No.
2859. The upper and lower ends were found separately
embedded in the sandstone ledge, but in working out the two
portions it is seen that they pertain to the same side of two
individuals. The bone as a whole, so far as comparison may be
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
659
660
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
made, presents characters not unlike those in the Princeton
specimen referred to Diplacodon elatus (see above) . However, in
the specimen under description (possibly a female) the coracoid
is seen to be relatively smaller than in the latter. The groove
between the base of the coracoid and the border of the glenoid
cavity is larger in proportion than in Menodus, and the excava-
tion on the coracoid border, immediately above the coracoid,
has a less abrupt curvature. This is due to the smaller develop-
ment of this angle in Eoiitanotherium. The coracoid border is
otherwise quite straight as in Menodus. The superior portion
of the glenoid border is broken off, but in the region of the break
there is a similar broad extent of the superior portion of the
blade. The spine is damaged, but it was apparently overhang-
ing like that in Diplacodon described by Osborn, and thus less
extended over the postscapular fossa than in Menodus.
In comparing the humerus of the present form with that of
Brontops validus, the difference most noticeable is the relative
robustness and the length. In the Oligocene form the bone is
short and very heavy, while in the present genus the bone is
longer in proportion and also lighter. Superiorly the greater
tuberosity extends higher above the head than in Menodus but is
not so robust, the proximal end as a whole being more delicately
proportioned. The bicipital groove is deep and well defined, as
in the Oligocene genus. On the other hand the deltoid ridge,
though very prominent, does not terminate in the heavy
recurved process as in B. validus but descends much more
gently toward the supratrochlear fossa. Distally there is less
variation between the two forms here compared. The anconeal
fossa in the species under description is relatively broader, and
the supinator ridge is less rugose. The trochlea is slightly
deeper but not more obliciue than in B. validus.
The humerus as described and figured by Osborn holds an
intermediate position between the Oligocene genus and the
present form. This is especially shown in the development of
the deltoid ridge, which in the Princeton specimen is consider-
ably more developed than in the genus under description.
Measurements of humerus of Eoiitanotherium, in millimeters
No. 2860
No. 2861
Length from head to distal end
Transverse diameter of lower part of deltoid ridge_
Transverse diameter at broadest portion of supi-
365
-108
115
365
"110
113
78
" The shafts of the two bones are more or less crushed, and the measurement is
only approximately correct.
Both radii and ulnae are represented in No. 2860. A third
radius was also found in the same sandstone ledge in close
proximity to the spot where Nos. 2858 and 2859 were found.
The radius and ulna are long and relatively slender when
compared with those of the Princeton specimen of Diplacodon
and the Ohgocene genus B. validus. Thus the forearm of the
new genus is actually a little longer than in Diplacodon and is
very nearly as long as that of B. validus, notwithstanding the
much smaller size of the Uinta form of which we are speaking.
Another striking difference between the forms here compared is
the lateral expansion of the pro.ximal and distal ends of the
radius. In the Oligocene form the shaft of the radius is more
rounded in the middle region, while more proximally and distally
a sudden expansion takes place, which is also well displayed in
the Uinta specimen described and illustrated by Scott and
Osborn. In Eoiitanotherium osborni the shaft is flatter, more
uniform throughout, and the proximal and distal ends compara-
tively little expanded.
The proportions of the ulna conform to the radius, and it is
consequently slenderer and proportionally longer than in
Diplacodon and Menodus. In detail the bone is otherwise quite
similar to that in the two latter genera, including the well-
defined tendinal groove on the anterior superior angle of the
olecranon process so characteristic of the ulna of Brontops
validus, but apparently less developed in the Princeton speci-
men, judging from the illustration (PI. IX, figs. 10-10 '^).
Measurements of No. 2862
Millimeters
Radius, greatest length 380
Radius, transverse diameter at middle of shaft 40
Radius, transverse diameter of head 78
Radius, transverse diameter of distal end 77
Ulna, length of olecranon process 100
The forefoot of No. 2860 is represented by the scaphoid, pisi-
form, trapezoid, Mtc II, IV, and V, and one or two phalanges.
No. 2859 has also Mtc IV and V represented.
As might be anticipated from the description of the limb, it
is found that the foot is higher than in B. validus of the Oligo-
cene. Thus the scaphoid is higher in i)roportion and narrower
than in the latter species but is of considerable fore-and-aft
diameter. In detail there are only such differences as one might
expect from the general outlines described — that is, the different
articulating surfaces of the distal face are narrow and long,
while the articulation for the radius is less concave antero-
posteriorly than in the Oligocene form. The pisiform has a
similar long attenuated shaft terminating in an obtuse tuberosity
of considerable vertical diameter but transversely rather thin.
Besides the greater height of the trapezoid, the small posterior
superior facet for the magnum, which is characteristic of
Menodus, is practically wanting in the present form. Judging
from the facet on the posteroradial angle there is present in the
new Uinta genus a trapezium of considerably larger size.
Mtc II is long, quite broad, but of small anteroposterior diam-
eter, which is in part due to crushing. The proximal end is
partly broken off, so that the different facets can not be accu-
rately compared. The shaft is of quite uniform width until the
distal articulating surface is reached, where there is on the
radial face a sudden expansion. This character is less apparent
in the Oligocene forms and also apparently less than in the
metacarpus of the Princeton specimen from the Uinta, as figured
by Scott and Osborn. Mtc IV is, as stated, represented by
fragments in both type and paratype and displays no features
of especial importance.
Mtc V is longer and slenderer than the same element in B.
validus and that referred to Diplacodon (PI. IX, fig. 13). Prox-
imally and distally the bone is expanded much as in Menodus,
and the shaft, though relatively longer, is of a similar cylindroid
character. The facet for Mtc IV is located more laterally than
in the Oligocene genus, and the dorsal and ulnar faces are less
deeply grooved for muscular attachments. Near the distal end
is a flange on the postero-ulnar angle, which is similar to that
already described on Mtc II and is not generally present in the
Ohgocene titanotheres.
There is apparently more inequality in size between Mtc II
and Mtc V than represented in the figure of the manus of
Diplacodon by Scott and Osborn. This is very probably due,
to some extent, to the crushing of Mtc II of the specimen in
the Carnegie Museum. In the specimen at Princeton the com-
plete length of Mtc V is apparently represented. Its measure-
ments appear to be only about 13 millimeters longer though
nearly one-third broader than that of the specimen before us.
The phalanges are short, broad, and in every respect titano-
theroid.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
661
Measurements of No. S860
Millimeters
Scaphoid, vertical diameter 35
Scaphoid, transverse diameter 33
Scaphoid, anteroposterior diameter 53
Pisiform, total length 60
Trapezoid, vertical diameter 20
Trapezoid, transverse diameter 26
Trapezoid, anteroposterior diameter 36
Mtc II, greatest length 153
Mtc II, transverse diameter of head, approximate 37
Mtc II, transverse diameter of middle of shaft, approxi-
mate 30
Mtc II, transverse diameter of near distal end, approxi-
mate 42
Mtc V, greatest length 125
Mtc V, greatest transverse diameter of head 36
Mtc V, greatest transverse diameter of middle of shaft 20
Mtc V, greatest transverse diameter of near distal end 33
Proximal phalanx, length 31
Proximal phalanx, transverse diameter of proximal end 29
Proximal phalanx, transverse diameter of distal end 26
The pelvis of No. 2859 is represented by the greater portion
of the ilium. It is quite broad across the gluteal surface, but
the point of the iUum probably did not project laterally as
much as in B. validus. The constricted portion of the neck is
actually longer than in the latter species and also longer than
in the Princeton specimen of Diplacodon as represented on
Plate VIII in Scott and Osborn's worli. The pelvis as a whole
was consequently proportionally longer and probably narrower
than in the Oligocene genus. The ischium and pubis are not
represented.
In No. 2860 the lower half of the femur is present. The tibial
and dorsal faces of the shaft are convex, while posteriorly it
presents a flat surface. On the fibular angle may be seen the
lower portion of the prominent ridge below the third trochanter,
which decreases in prominence in its downward course. Near
the distal end the fibular border presents a roughened area for
muscular attachment, back and below which is the rather shallow
supracondylar fossa. Distally the condyles are rather well
separated by the deep and broad intercondylar fossa. The
lateral sides of the distal end (especially the fibular) is well
marked by the rugose attachment for muscles. The rotular
trochlea is proportionally deeper and narrower than in Menodus,
and the fossa immediately above it is much deeper and better
defined. In this respect the present genus agrees better with
Figure 5 on Plate VIII of Scott and Osborn's publication.
Measure7nents of femur of No. 2S80
Millimeters
Total length of the fragment 280
Transverse diameter of shaft about the middle region of the
fragment 60
Transverse diameter of distal end 108
Anteroposterior diameter of distal end 110
The greater part of the tibia is represented in the paratype
No. 2860, but it is badly crushed. Another individual, No. 2862,
has both tibiae present and is approximately of the same size
as the individuals we are describing. The bone is very nearly
as long as in B. validus. The ends are not expanded as in the
latter form, while the shaft is flatter, due in part to crushing.
The superior end carries a heavy and bifid spine, while the
upper anterior extremity displays the broad groove for the
patellar ligament as in Menodus. The cnemial crest, though
prominent, does not descend low on the shaft, another feature
recalling what may be observed in B. validus and in the Uinta
specimen figured by Scott and Osborn.''^ The anterior border
*2 If the illustratioa on PI. VIII, Fig. 6, in Scott and Osborn's publication is
one-fifth of nature, as is that of the femur in the same plate, the tibia of that form is
actually shorter than that in the genus here described.
of the distal trochlea was found weathered off, but the posterior
surface is complete and presents a very prominent descending
process on the median ridge of the articulating trochlea very
similar to what is seen in the later Uinta form and in Menodus.
From the material in hand it is shown that the hind hmb of
Eolitanotherium osborni corresponds well in length with the
fore limb.
Measurements of tibia of No. 2862
Millimeters
Greatest length, approximate 415
Transverse diameter of head 100
Transverse diameter of shaft, middle region 48
Transverse diameter of distal end, approximate 75
The hind foot of No. 2860 is represented by the calcaneum,
the astragalus, and the second and fourth metatarsals.
When compared with the Princeton specimen from a higher
Uinta level and also with the Oligocene genera, the tuber of the
calcaneum in the present form is seen to be as long in proportion
and compressed laterally to the same extent, while that portion
carrying the sustentacular facets is longer. The fibula also
■ apparently articulates with the calcaneum, but the posterior
portion of the tibial trochlea did not touch the calcaneum as in
Diplacodon and in Menodus. The astragalus is higlier and
narrower, and the metatarsals are longer and much slenderer
than in the latter genera.
When compared in more detail there are a number of differ-
ences between the genera here compared. On the calcaneum
of the genus under description the proximal astragalar facet
is not raised as high above the surface as in Menodus. The
greater process of the distal end extends lower down and the
facet for the cuboid is more oblique than in Menodus. As
already stated, the astragalus is higher and narrower, the
trochlear groove is deeper, with the articular surfaces of the
two condyles steeper, and the neck separating the distal end
from the trochlea longer than in the astragalus of the Oligocene
form and also somewhat longer than in Diplacodon as figured
by Scott and Osborn. Furthermore, the distal end of the
astragalus of the present form is more unequally divided by the
navicular and cuboid facets than in the Oligocene genus. These
facets of the astragalus in Menodus are more nearly subequal in
size, the cuboid facet having increased in size as well as being
located more distally on the bone, while in Eotitanotherium
this facet occupies a comparatively narrow area on the fibular
angle and is placed laterally.
The most noticeable difference of the astragalus of Eotitano-
therium osborni and that of the Princeton specimen as figured
(PI. VIII, fig. 86) seems to be in the three distinct astragalar
facets (viz, ectal, sustentacular, and cuboidal) of the latter,
while in the present form the ectal, besides extending higher,
unites with the cuboidal facet without distinct separation,
the two forming a perfect right angle apparently similar to
that in Mesatirhinus.*^
Aside from the greater proportionate length the metatarsals
differ from those in Menodus by being arched forward to a
greater degree. The shaft of Mts IV is more cylindrical,
and the facet for the cuboid more oblique.
Measurements of No. 2860
Astragalus: Millimeters
Total length 71
From lower end of external condyle to distal end 26
Greatest transverse diameter 68
Transverse diameter of trochlea 56
Calcaneum:
Greatest length 124
Length of tuber 64
Vertical diameter of tuber 45
Transverse diameter of tuber 22
Transverse diameter at sustentaculum 70
<3 Osborn, H. F., Am. Mus. Nat. Hist. Bull., vol. 24, p. 68, 1908.
662
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Mts II: Millimeters
Length 150
Transverse diameter at head 28
Transverse diameter of shaft, median region 21
Transverse diameter of distal end 26
Mts IV:
Length 41
Transverse diameter of head 38
Transverse diameter of shaft, median region 22
Transverse diameter of distal end 34
The restoration of Eotitanotherium osborni here attempted
(PI. X) is obtained from the material described in the preceding
pages, and it is chiefly based on two individuals. As previously
stated, the front of the skull, the lower jaws, atlas, axis, pelvis,
and a few fragments of the feet pertain to one individual, the
type, while the rest of the vertebral column, a few ribs and
limb bones, as well as a number of foot bones, belong to a
second individual, one of the paratypes.** The dotted hnes
represent estimated diameters and are consequently conjectural
as to proper contour outlines. This is especially true of the
posterior portion of the skull, the sacrum, the ischium, the
upper half of the femur, and the caudal region. There are
inserted two cervicals, two dorsals, the sacrum, and the greater
part of the caudal region. The vertebral formula as repre-
sented in the illustration is the same as that of the articulated
skeleton of Menodus from the Oligocene now in the Carnegie
Museum. The vertebral formula of Eotitanotherium osborni is
in part therefore tentative and is as follows: Cervicals 7, dorsals
17, lumbars 3, sacrals 4, caudals 18. The ribs are conjectural.
Tlie illustration is effected for the purpose of ascertaining at
a glance the general proportions of the animal. Each part
represented by the solid lines is drawn directly from the bones
themselves, by the assistance of the pantograph, and the illus-
tration as a whole is fairly reliable.
Measurements
Centimeters
Total length of the vertebral column from the premaxillary
to the end of the sacrum, all curves of the backbone in-
cluded 252
Height of skeleton at fore limb 138
Height of skeleton at hind limb 114
SECTION 5. THE POSTCRANIAL SKELETON OF
OLIGOCENE TITANOTHERES
Five factors have conspired to limit our knowledge
of the postcranial skeleton of Oligocene titanotheres
and to make it unsatisfactory.
First, owing to the conditions of preservation and
deposition, the skulls of Oligocene titanotheres are
rarely found in association with the rest of the skeleton,
so that only a few skeletons have been mounted.
Marsh's type of Brontops robustus in the Yale Museum
and the skeleton of Menodus trigonoceras in the Munich
Museum are nearly complete. The skeleton of Bron-
tops dispar in the Carnegie Museum is nearly com-
plete, but it lacks the skull. The skeleton of Bron-
tops roiustus in the American Museum includes the
skull, fore limbs, and thorax of one individual, but the
pelvis and hind limbs are supplied from another
individual. "'' A few skulls were found in association
" There was no other material found with the remains of Eotitanotherium described
in the preceding pages, except a few fragments of turtles. All the material was found
within a radius of about 20 feet.
"« Since this was written the Colorado Museum of Natural History, at Denver,
Colo., has mounted an unusually fine skeleton of Megacerops acer, and the U. S.
National Museum, at Washington, has mounted'a partial skeleton of Brontottterium
haicheri.
with parts of the limbs or of the backbone, but by
far the greater part of the collections consist of unas-
sociated skulls, jaws, and parts of skeletons.
Second, the natural scarcity of well-associated
material was intensified by the hasty methods of
early collectors, who, with certain exceptions, through
eagerness to secure the conspicuous and highly
prized skulls, neglected the opportunity to collect
skulls and skeletal parts in association or failed to
make careful records of such association. Still further
damage was done through crude methods of packing
and transportation.
Third, an unfavorable condition arises from the
crushing and distortion suffered by the skeletal parts
of titanotheres during the structural alteration and
displacement of the ancient strata in which they are
entombed. A skeleton in the American Museum of
Natural History (No. 1064), in which there is a good
association of dentition, vertebrae, and limbs, is
rendered practically useless for systematic study
through the crushed condition of the bones. Such
distortion often alters the natural contours and pro-
portions in a deceptive manner, as noted by the late
J. B. Hatcher (1902.1), who showed that in the Car-
negie Museum skeleton certain limb bones on one
side were much lengthened, whUe on the opposite side
the corresponding elements were shortened, the former
having been crushed laterally, the latter longitudi-
nally. As a result of distortion, added to the gr^at
difficulty of securing uniformity in the methods of
measuring such large skeletons, it follows that many
of the measurements given below would hardly be a
safe basis for precise quantitative distinctions between
supposedly different species.
Fourth, a difficulty arises from the wide differences
between males, females, and young of the same species
and at different ages, which even in the study of well-
preserved skulls causes considerable uncertainty as to
the correct identification of certain specimens. Be-
sides this it is probable that fully adult male titano-
theres of the same species would show considerable
variation in minor characters and measurements of
the vertebrae and limbs, as in the case of modern
rhinoceroses.
Fifth, the postcranial skeleton of titanotheres has
not hitherto been used to any extent by authors in
defining the genera and species, so that, from a
systematic point of view, the vertebrae and limbs have
not been regarded as sufficiently important to require
the degree of intensive study which has been made of
the skull and dentition.
As a result of these unfavorable conditions we have as
yet only an incomplete knowledge of the generic char-
acters of the vertebrae and limbs of Oligocene titano-
theres, while we have hardly begun to recognize specific
differences. The following studies are preliminary and
should be supplemented by a more systematic and
intensive study of the material available comparable
with that which has been made upon the skulls.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
Chief skeletal characters of the Oligocene phyla
663
Brontops phylum
AUops phylum
Menodus phylum
Brontotherium phylum
Progressively graviportal, bra-
chypodal. Scapula and pel-
vis less broad than in Bronto-
therium. Tall and ponderous
animals.
Relatively small,
short-limbed.
Relatively long, slender-
limbed. Scapula and
pelvis of high, slender
type. Neck relatively
long.
Progressively increasing in size and weight.
Skulls very large. Necks relatively short
and massive. Scapula and pelvis ex-
tremely broad, graviportal type. Limbs
very broad and massive. Feet relatively
short as compared with Menodus.
STceletons and parts of skeletons of Oligocene titanotheres here described "
Genus and species
Specimen
Museum
Brontops ?brachycephalus
Referred partial skeleton with skull.
Mounted.
Victoria Memorial Museum, Ottawa.
Brontops dispar
Referred skeleton without skull.
Mounted.
Carnegie Museum, Pittsburgh, No. 92.
Brontops robustus
Type skeleton with skull. Mounted-
Peabody Museum, Yale University, No. 12048.
No
skull. Mounted.
518.
Diploclonus tyleri
Type incomplete skeleton with skull.
Amherst Museum, No. 327.
AUops marshi
Field Museum of Natural History, Chicago, No. P6900.
Mounted.
Allops crassicornis
Referred partial skeleton with skull
British Museum (Natural History), London,
No
5743M.
Menodus trigonoceras
Referred skeleton with skull.
Mounted.
Munich Palaeontological Museum.
Do
Referred manus and pes
American Museum of Natural History, New York,
1079.
No.
Referred pes
American Museum of Natural History, New York,
1080.
No
Referred pelvis and hind limbs _
University of Nebraska, Lincoln, No. 3296.
Do
Referred ulna and radius.
Carnegie Museum, Pittsburgh, No. 120.
Do
Ref erred skull and cervicals
Referred partial skeleton with skull
Field Museum of Natural History, Chicago, No. 59
Carnegie Museum, Pittsburgh, No. 93.
27
Do
do
Carnegie Museum, Pittsburgh, No. 114.
American Museum of Natural History, New York,
1047.
American Museum of Natural History, New York,
1443.
American Museum of Natural History, New York,
Referred manus and pes
No
No
Do
Referred partial skeleton with skull
No,
Brontotherium gigas hatoheri.
Referred partial skeleton with skull.
Mounted.
492.
U. S. National Musevim, Washington, No. 4262.
« In addition to those listed there are two well-mounted skeletons of titanotheres (Metiacerops acer) in the Colorado Museum of Natural History, Denver, Colo. C1928).
(See PI. CCXXXVI.)
664 TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Comparative measurements oj the skeletons of lower Oligocene titanotJieres, in millimeters
i
o
II
|a
a
n
ii
S
<;
Is
Is
a
<!
If
f
%
i|
11
a
-a
§§
.So
a
<
i
ft"
<
i •
Is
II
1
is
is D
•oa
1
at
O
II
§
1
® cS
at
O >
as
1
|i
ei
i
a"_
jl
s
a
<!
a",
as
1
■S3
la
!2 ft
Is
Height at shoulder (top of third
1,855
2,502
765
667
460
580
690
2,311
"2, 230
"768
-''2, 120
655
465
637
700
770
830
740
440
612
657
....
320
275
305
Second dorsal, length of spine
600
600
....
690
....
570
"640
475
....
693
407
Humerus, length (head to ento-
545
404
615
495
• 170
680
200
220
74
225
88
210
84
1,220
900
820
430
230
110
86
144
180
530
455
165
610
207
425
365
450
350
557
525
380
306
93
410
— -
490
....
475
— -
460
168
642
84
535
167
65
155
1^05
"620
170
74
475
142
144
595
....
Carpus, width, proximal row (fac-
ets)
50
149
65
214
155
59
196
168
160
62
200
80
207
90 89
Mtc II, height
212
Mtc II, maximum width near
Mtc III, height
173
72
163?
59
935 +
809
= 686
« 432
218
250
155
....
233
199
187
160
Mtc III, maximum width near
76
Mtc IV, height
204
138
— -
226
188
174
210
Mtc IV, maximum width near
distal end
Pelvis, width
830
1, 170
820
902
762
432
Femur, length
785
446
208
105
590
350
147
70
560
345
140
82
703
430
Tibia, length...
300
Calcaneum, length
155
152
Astragalus, width
Astragalus, height (fibular side)..
80
120
Tarsus, width (astragalus and
calcaneum) _ _ _
90
100
106
173
126
.- -
Mts II, length...
Mts II, width near distal end. _
Mts III, length... . .
164
208
96
205'
156
200
170
Mts III, width near distal end
^ After correcting position of scapul
SUBFAMILY BRONTOPINAE
Lower Oligocene titanotheres, progressively gravi-
portal and brachypodal; trapezium lost in later forms.
Brontops dlspar Marsh
Materials. — Our knowledge of Brontops dispar is
based chiefly upon the remarkably complete skeleton
in the Carnegie Museum (No. 92), which lacks the
skull and three anterior cervicals. The skeleton
was discovered by J. B. Hatcher in Sioux County,
Nebr., in 1900 and described by him in 1902 (1902.1).
The locality is Warbonnet Creek; the geologic level
is near the base of the TitanotTierium zone as exposed
at that point, about 30 feet above the Pierre shale;
the matrix fine clays. The skull had been destroyed
by weathering; the skull and jaws mounted with the
specimen belong to another animal.
Limb distortion by crusMng. — An interesting feature
connected with the discovery is the distortion due to
the various positions of the several bones as they lay
embedded. The right femur, which was embedded
vertically, is some 6 inches shorter than the left, which
was embedded horizontally. The right and left
humeri similarly differ widely in proportion. This
striking discrepancy shows what extraordinary care
must be taken to avoid conclusions based merely
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
665
on measurements. The calcanea of the opposite sides
are of entirely different shape, although there is little
external sign of crushing. Had these femora been
found in different localities they would certainly
have been referred to different species if not to
different genera.
Comparison with Brontops rohustus. — As Hatcher
observes, not only is this skeleton smaller and some-
what less robust than that of B. rohustus, but there
are primitive characters of importance. (1) Com-
pare the slender, pointed neural spine of the first
dorsal vertebra, which although much longer resembles
that of the last cervical in form, with the broad, heavy
neural spine of the first dorsal in B. rohustus or in
Menodus. (2) In the carpus of B. dispar a well-
developed trapezium is present, while it is absent in
B. rohustus. In this connection Hatcher observes that
in all titanotheres from the lower Titanotherium
zone the trapezoid shows a trapezium facet.
The measurements are given in the table above.
Summary oj primitive characters. — This skeleton pre-
sents the following primitive characters as compared
with Brontops rohustus: (1) Less development of
neural spines on cervicals and dorsals, smaller neural
spines on cervicals; (2) relatively more elongate
scapula and innominate bones; (3) slightly less elevated
humeral tuberosity; (4) a trapezium; (5) lunar resting
more subequally on magnum and unciform; (6)
tarsus less compressed vertically; (7) less displace-
ment in manus and pes; (8) limited metacarpal dis-
placements— that is, Mtc II without large magnum
facet; (9) Mtc III with small ectocuneiform facet.
Size and proportions. — The total linear proportions
of this skeleton to that of the Brontops rohustus?
skeleton in the American Museum (No. 518, fig. 606)
can not be given owing to the absence of the skull,
but a few comparisons can be made:
Comparative measurements of Broniops dispar and B. robustusf
Shoulder height, to spine
of third dorsal
Breadth of pelvis
Length of fore limb,
including scapula
Length of hind limb
Presacral vertebrae
Length of femur
Millimeters
1,855
935
1,743
1,500
2,275
686
Ft. in.
6 1
3 h
5 9
4 11
7 5M
2 3
MillimeteTS
2,300
1,220
2,050
1,670
2,640
790
6 9
5 6
The skeleton of B. rohustus is thus throughout
between one-sixth and one-eighth larger.
Vertehrae. — The vertebral formula is C 7, D. 17,
L.?, S. 4, C. 18. The atlas (partly restored) exhibits
a hypophysial backward spur beneath the axis.
More in detail, the fourth to seventh cervical ver-
tebrae resemble those of B. rohustus in the lateral
processes but differ markedly in the comparatively
small spines on C. 6 and C. 7. D. 1 differs also in
its relatively slender spine; in D. 2-5 the spines are
approximately equal, that of D. 2 being the broadest
and somewhat the longest; in D. 7-14 the spines
gradually diminish in height; in D. 15 to L. 2 the
spines are relatively lower and broader than in
B. rohustus; in D. 3-12 the zygapophyses are hori-
zontal; in D. 13-17 the zygapophyses become obliquely
vertical. In L. 2 (the only lumbar preserved) the
transverse processes are rather narrow; the spine is
relatively broader and larger than in B. rohustus.
Twelve caudals are preserved, of which the third
bears a chevron bufnot the second.
Rihs. — The first rib is less massive than in B. rohus-
tus. The ribs increase steadily in length from K. 1
to R. 8; R. 8 to R. 10 are the longest and of nearly
equal length; R. 11-17 steadily diminish in size.
Sternals. — The manubrium is restored. S. 2 and
S. 3 are laterally compressed. The xiphisternum is
partially restored.
Scapula and fore limh. — As compared with that of
Brontops rohustus the scapula is relatively higher and
narrower, the anterior and posterior borders being
more nearly parallel, and the general form more
quadrate than triangular; this is due to the relatively
larger supraspinous fossa, the long, straight anterior
border for the attachment of the levator anguli scapu-
lae, the inferior angle of which above the supracoracoid
notch is placed on a lower level than in B. rohustus;
similarly the infraspinous fossa for the infraspinatus
muscle is relatively smaller; as characterized by a very
prominent rugosity at the posterior superior angle
(latissimus dorsi), the superior or vertebral border
is less arche.d than in B. rohustus.
The humerus exhibits a highly convex great tuber-
osity (supraspinatus and infraspinatus insertion),
which is relatively as elevated as in B. rohustus; the
deltoid ridge and tuberosity are equally prominent
and the external or supinator crest rises very promi-
nently, the shaft being deeply excavated between.
The internal face of the radius is somewhat less ele-
vated than in B. rohustus; its total length (404 mm.)
is double the circumference (202). The ulna exhibits
a broadly flattened anterior face; the olecranon lacks
the deep vertical cleft at the summit so characteristic
of the upper Oligocene titanotheres.
The carpus shows the following characters: Breadth
167 millimeters, lunar resting broadly on the magnum
facet (28 to 32 mm.) as weU as on the unciform facet
(34 to 40 mm.), more primitive or less displaced than
in B. rohustus, much more primitive than in Menodus.
The trapezium is present; it is attached above to
the scaphoid but reaches only halfway down the inner
face to the trapezoid; it is not in contact at all with
Mtc II. Magnum broad; Mtc II with a narrow
magnum facet (20 mm.); Mtc III with a relatively
666
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
narrow unciform facet (14 mm.) (this facet is broader
in B. roiustus); Mtc IV occupying main portion of
unciform; Mtc V articulating chiefly with the outer
side.
The linear measurements of the metacarpals are as
follows: Mtc II, 155 milUmeters; Mtc III, 176;
Mtc IV, 174 (?); Mtc V, 137. The proximal phal-
anges are very broad and large; the median phalanges
are short. The distal phalanges are relatively smaller
than in B. roiustus, expanded transversely and of
short linear dimensions; the widths are, Ph II, 43
millimeters; Ph III, 45; Ph IV, 46; Ph V, 38.
Pelvis and hind limb. — Massive as the pelvis is, its
intermediate character is indicated by the fact that
the total length of the innominate bone is 809 milli-
meters, as compared with the total breadth of 925
millimeters across the two ilia; the posterior opening
is correspondingly elongate vertically. The hind
limb as compared with the fore limb is long and
slender, with the following principal characters:
Femur with projecting head, feebly marked third
trochanter directly opposite second trochanter, and
both well down on the shaft, inner patellar facet
much more elevated than outer, patella vertically
elongate, inner condyle of the femur much the largest.
Tibia relatively elongate and less massive than in B.
roiustus, shaft of fibula much reduced superiorly.
Pes as a whole decidedly slender; tarsus having a
breadth of 110 millimeters across cuboid and navic-
ular; tuber calcis not placed transversely; astragalu.i
with a relatively long neck; navicular and cuboid
relatively deep; metatarsals of medium length and
slenderness, surprisingly light in structure; Mts II
not abutting against ectocuneiform as in B. roiustus;
Mts III with a very oblique lateral facet for cuboid
(differing widely from the broad facet in B. roiustus) ;
proximal phalanges long and rather narrow; distal
phalanges moderately broad.
Brontops robustus Marsh
Type specimen. — Our knowledge of Brontops roiustus
is based on the superb type specimen in the Peabody
Museum at Yale University (Yale Mus. 12048) which
was discovered and unearthed by Mr. H. C. Clifford
in 1875 in the upper Titanotherium zone near Chadron,
Nebr. It was appropriately made the type of Brontops
roiustus by Marsh, and it is very fully illustrated in
Plates XCVI-CIII, CXCV-CCXXIX of this mono-
graph. As noted in detaQ below, certain of the plates
and the very remarkable restoration by Berger under
Marsh's direction (PI. CCXXIX) include bones that
were derived from other specimens. In 1916 this skel-
eton was mounted under the direction of Prof. R. S.
Lull, who gives the following measurements: Height
at shoulder, 8 feet 23^ inches ( = 2.502 meters)*";
« In Professor Lull's judgment this measurement is probably 3 or 4 inches too
great — that is, the backbone as mounted is too high.
length over all, following vertebral column, 15 feet
2}/^ inches ( = 4.635 meters); length between per-
pendiculars to base of tail, 11 feet ( = 3.353 meters).
Of the presacral vertebrae 26 are preserved — 7
cervicals, 17 dorsals, and 2 lumbars — in a continuous
series, the third or posterior lumbar being apparently
missing.
The characters of the cervicals in lateral view are
accurately represented in the restoration forming
Plate CCXXIX and in Figure 610, A.
The atlas (PI. CXCV, figs. 1-5) measures 465 milli-
meters transversely, 158 vertically, 278 across the
axis facets; the neural arch is perforated at the side
for the exit of the first cervical nerve, which passes
down anteriorly in a deep notch between the junction
of the condylar cup with the transverse process (figs.
2, 4); the transverse processes (ribs) are moderately
robust, not very widely expanded anteroposteriorly,
somewhat truncate or square distally; the neural
spine is sessile and slightly grooved but not bifid
posteriorly. The axis (PI. CXCVI) measures 270 milU-
meters transversely, 290 vertically, the outside
measurement of the posterior face of the centrum
being transverse 137, vertical 119; the superior bor-
der of the spine is acutely convex and slopes obliquely
backward, slightly overhanging the centrum pos-
teriorly; the antero-inferior face is abruptly trun-
cate; the odontoid is short; the atlanteal facets are
continuous, the median area being transverse, the
lateral areas oblique and flaring. In C. 3-7 (PI.
CXCVII) the neural spines increase regularly in height,
that in C. 7 being pointed but entirely different in
character from the spine of D. 1 ; the prezygapophyses
and postzygapophyses are flattened and face upward
and inward, and downward and outward; the lami-
nae of the neural arches increase gradually in width
as seen from above in C. 3-C. 7. The cervical
transverse processes or ribs (pleurapophyses) in C 3
exhibit a broad, thin plate; in C. 4 a rugose superior
and small flat inferior lamella; in C. 5 rugose and
subequal superior and inferior lamellae; in C. 6 a
rugose superior lamella and widely expanded, antero-
posteriorly flattened inferior lamella; in the im-
perforate 'C. 7 the inferior lamella is entirely wanting.
Proportions of presacral centra. — The comparative
measurements on page 667 bring out the following char-
acters in the proportions of the centra. The short,
deeply opisthocoelous cervical centra (C. 3-7) range
from 650 to 700 millimeters in length, from 107 to 114
in width, and from 106 to 113 in height, practically
as high as broad. The dorsal centra are less deeply
opisthocoelous; they are longer than the cervicals,
ranging from 73 to 88 millimeters in length; in the
anterior dorsals (PI. CXCVIII) the vertical exceed
the transverse measurements, correlated perhaps with
the spines; in the middle and posterior dorsals the ver-
tical are less than the transverse measurements.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
667
Comparative measurements of cervicals, dorsals, and lumbars, in
millimeters "
C. 1 (atlas)
C. 2 (axis) .
C. 3
C. 4
C. 5
C. 6
C. 7
D. 1
D. 2
D. 3
D. 4
D. 5
D. 6
D. 7 _.
D. 8
D.9=
D. 10
D. 11
D. 12
D. 13
D. 14
D. 15
D. 16
D. 17
L. 1
L. 2
Anteropos-
terior
Transverse
69
245
110
267
70
108
65
107
58+
108
66
112
70
114
73
103
75
92
77
96
76
98
76
100
77
114
77
111
80
108
"90
103
75
94
82
''96
77
85
75
90
84
93
88
87
83
96
77
92
87
88
SB
"^92
107
112
106
111
113
114
"118
117
113
100 +
109
103
103
''107
90
«81
84
• The longitudinal measurements are taken along the inferior line of the centra
from between the edges of the articular facets; the vertical and transverse measure-
ments are taken on the anterior faces of the centra.
<> Height of spine, 695 millimeters.
•= The position of this vertebra in the series is somewhat doubtful.
' Estimated.
• Crushed.
Dorsals. — There are 17 rib-bearing or dorsal verte-
brae. The first dorsal (D. 1) marks a very abrupt
transition from the cervicals by the sudden elevation
of its very broad spine to a height of 590 + millimeters;
the laterally flattened upper portion of the spine rests
upon the relatively short triangular lower portion.
In the succeeding dorsals, D. 2-D. 10 (PI. CXCIX),
the triangular basal portion of the spine, deeply exca-
vated posteriorly (fig. 4), becomes relatively higher
until in D. 7 it extends nearly to the top of the
spine. The spine of D. 2 (PI. CXCVIII) appears
to have been the stoutest and the most elevated ; it is
more backwardly directed than that of D. 1. From
D. 3 to D. 17, inclusive, the spines gradually diminish
in height and become more slender and more back-
wardly directed. In L. 2 (PI. CC) the spine is short
and obliquely directed backward.
The zygapophyses are characteristic: the post-
zygapophysis of D. 1 and prezygapophysis of D. 2
are transversely oblique or face respectively outward
and inward, whereas in the American Museum
skeleton referred to Brontops they are subhorizontal.
The postzygapophyses of D. 2 face directly down-
ward and backward — that is, in a more nearly hori-
zontal plane. Similarly, the postzygapophysis of D. 2
and the prezygapophyses of D. 3-D. 15 are antero-
posteriorly oblique or subhorizontal, facing downward
and upward respectively rather than outward and
inward. In D. 14 a transition occurs to the outward
facing of the postzygapophyses and the inward facing
of the prezygapophyses. In D. 17 the articulation
is very slightly revolute, convexo-concave. In L. 1
the articulation is not preserved; in L. 2 (PL CC)
the postzygapophyses are flat and face obliquely
outward and downward.
Rib jacets. — Throughout the series the capitular
facets are the largest on the posterior sides of the
centra, attaining a very great size, for the especially
large third, fourth, and fifth ribs. In D. 12-D. 17
the capitular facets rise from the base of the neural
arches, and even in the posterior dorsals, D. 15-D. 17,
the tubercular facets are relatively large and well
developed.
Lumbars. — The lumbars are imperfectly preserved,
one of these vertebrae being lost entirely. The lum-
bar exhibits the following distinctive characters: Spine
recumbent, zygapophysis facing obliquely outward and
inward, not revolute, transverse processes rather feeble,
with broadly rugose and sessile inferior keel.
Sacrals. — There are four sacrals (Pis. CCXVIII,
CCXIX) rather imperfectly preserved; they rapidly
diminish in size, S. 1 being very much larger than the
succeeding sacrals.
Oaudals. — An unusually perfect series of 16 caudals
(PI. CCI) is preserved in sequence behind the last
sacral. The relatively small size of the most anterior
caudal (figs. 1-3) is correlated with the small size of
the posterior sacral and a small tail. The centra
increase in length from C. 1, which measures 43
millimeters, to C. 9, which measures 70, and dimin-
ish to C. 16, which measures but 35. Neural spines
and rapidly diminishing zygapophyses were present
on C. 1-C. 8. C. 3 (figs. 6, 7, 8) exhibits a chevron
inclosing the haemal artery; this element may also
be preserved in C. 4 (figs. 9, 10) and C. 5 (figs. 11, 12,
13). The transverse processes rapidly diminish behind
C. 1 and finally disappear in C. 11. The centra are
biconvex. Beginning at C. 8 the centra become
laterally compressed, and the posterior centra, C. 12-
C. 16, are decidedly compressed laterally.
i?i6s.— Nine perfect ribs (Pis. CCII, CCIII) are
preserved and fragments of others. On the whole they
are more slender and rounded in section than in the
skeleton of Brontops in the American Museum, de-
scribed below.
Sternebrae. — Three of the midsternebrae are pre-
served, measuring collectively 290 millimeters; the
anterior (restored in PI. CCXXIX) and one or more pos-
terior sternebrae are missing. The supposed second
sternebra is broadly depressed in section, measuring
but 42 millimeters vertically, 123 anteroposteriorly,
668
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
and 70 transversely; the flattening of the body of this I
element is partly due to crushing. The supposed
third sternebra is deeper, measuring 35 millimeters
vertically, and shorter (ap. 95 mm.) and somewhat
narrower (tr. 64 mm.). The supposed fourth
sternebra, which was largely cartilaginous at the
extremities, measures 39 millimeters vertically, 70
anteroposteriorly, and 72 transversely.
Scapulae. — As partly restored, the scapulae of the
type measure 690 millimeters vertically and 530
transversely; they are characterized by the narrow
rectangular prespinous fossa, the large, broadly
triangular postspinous fossa, and the rugose retro-
verted acrominal spine.
Humerus. — The humerus measures 622 millimeters
from the head to the inner condyle. As admirably
shown in Plate CCV, it exhibits (1) the deep bicipital
groove (fig. 5), (2) the elevated platelike tuberosity
(fig. 4) indented on the anterior superior border
(fig. 5), (3) the prominent deltoid ridge terminating
in the characteristic prominent and somewhat up-
turned knob (figs. 1, 3), (4) the rugose supinator
ridge, (5) the very marked upward extension on the
shaft of the internal radial facet (fig. 1).
Radius.— The radius (PL CCVI), measuring 495
millimeters, exhibits a transverse diameter of 170
millimeters for the humeral facets (fig. 5) and of
118 for the scaphoid and lunar (fig. 6). The rugose
insertion of the brachialis anticus muscle is placed
about the middle of the shaft (fig. 1). The shaft is
well rounded in midsection (fig. 2 a), the transverse
diameter being 78 mllimeters.
Ulna.— The ulna (PI. CCVII) , having a total linear
measurement of 680 millimeters (olecranon to unciform
facet), exhibits the characteristic groove (figs. 2, 3)
in the top of the olecranon; a stout trihedral shaft.
Manus. — The right and left manus are both
preserved; the unfortunate absence of both lunar.s
renders it difficult to ascertain the exact characters
of this important element. The total width of the
proximal row of carpals (fig. 1) across the facets
is 200 millimeters (estimated), of the distal row
202; the vertical depth through the anterior faces of
the scaphoid and magnum is 92; the length of Mtc
III is 210; the height from the scaphoid to the extrem-
ity of the median metacarpal is 322; this manus,
therefore, may be described as moderately broad and
short. (See PI. CCXXVIII.)
The detailed structure of the carpals as figured in
Plates CCVIII-CCXI is important. The maximum
measurements of the scaphoid (PI. CCVIII, fig. 5 r)
are, vertical 65 millimeters, transverse (radial face)
73, anteroposterior 82; the magnum facet (fig. 4)
is exceptionally narrow (30), and the scaphoid is
therefore relatively narrow transversely, deep verti-
cally, and extended anteroposteriorly. The radial
facet of the missing lunar is estimated at 170 milli-
meters, its magnum facet at 25, the unciform facet at
53; this bone had a relatively narrow, oblique facet
on the magnum and a much broader facet on the
unciform. The cuneiform (PI. CCIX, figs. 1-6) is
a smaller element, moderately deep (52 mm.)
vertically, with a rectangular ulnar facet (fig. 5)
which measures 57 millimeters transversely and 48
anteroposteriorly. The pisiform (PI. CCIX, figs.
7-12) is large, measuring 125 millimeters antero-
posteriorly. The trapezoid (PL CCX, figs. 1-6)
exhibits no evidence of a trapezium facet; it is rather
shallow (ap. 37 mm.) with a moderately broad
scaphoid facet (fig. 5 s), 53 millimeters transversely, 67
anteroposteriorly. The magnum (PL CCX, figs.
7-12) measures vertically 40 millimeters and trans-
versely 65 on the anterior face (fig. 7), with a broader
scaphoid (fig. 7 s) (35) and a narrower (20) lunar
facet (fig. 7 I). The unciform (PL CCXI, figs. 1-6)
is a very large element, its maximum breadth being
114 millimeters, height 58, depth of cuneiform facet
(ap.) 53 (fig. 1 p); superiorly (fig. 5 Z) it exhibits a
broad (55) lunar facet and a slightly broader cunei-
form facet (fig. 5 p) (60); on the lower surface (fig.6)
it has a narrow (12) abutment against the magnum
(fig. 6 m) and a broader facet for Mtc III (27) (fig.
6 mc III).
The metacarpals (Pis. CCXII-CCXV) on the
anterior face to the distal facet measure, Mtc II,
220 mniimeters; Mtc III, 225; Mtc IV, 210; Mtc V,
180. Their characters are so admirably shown in the
figures that they require no further description. The
proximal phalanges are figured in Plate CCXVI;
the median and distal phalanges and sesamoids are
figured in Plate GCXVII. It is noteworthy that the
distal phalanges are of moderate size and breadth,
very much smaller than in the American Museum
skeleton referred to Brontops, somewhat broader than
in Menodus.
Pelvis. — The pelvis is decidedly broad, the entire
width across the ilia being 1,220 millimeters (esti-
mated), as compared with 908, the length of the left
innominate bone (ilium to ischium). The pelvic
outlet measures 373 millimeters vertically (pubic
symphysis to sacrum) and 350 transversely. The
pubo-ischiadic symphysis measures 357 "millimeters
longitudinally. The extreme breadth of the iliac
plates is 407 millimeters. A more detailed discussion
of the characters of the pelvis is continued below.
The association of the fore and hind limbs in this
superb specimen is especially important as giving us
the proportions.
Femur.— The femur (PL CCXX) measures 820
millimeters and the tibia 465. The shaft of the femur
is somewhat crushed proximally, the great trochanter
extends far (395 mm.) down the outer side, the width
across the condyles distally is 187, the patellar trochlea
is distinguished by the prominence of the inner facet
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
669
and by the downward and forward projection of the
trochlea. In the distal view (fig. 5) we are struck by
the great disproportion between the greatly enlarged
inner condyle (fig. 5 i) and the much reduced external
condyle (fig. 5 e). This asymmetry of the end of the
femur is naturally correlated with the asymmetry of
the proximal faces of the tibia. The patella measures
165 millimeters vertically, 105 transversely, 100
anteroposteriorly; it is very stoutly convex.
Tibia.— The proximal face of the tibia (PI. CCXXI),
exhibits an elongate internal facet for the inner
condyle of the femur (fig. 5 c) and a shorter external
facet for the outer condyle (fig. 5 e). The facets con-
verging into the wide and distinctly paired intercon-
dylar spines — that is, the internal and external
facets — -are entirely separate; the cnemial crest is
obtusely rugose (fig. 1). The shaft (fig. 1 a) is tri-
hedral in midsection, the transverse diameter being
81 millimeters and the anteroposterior diameter the
same. Distally there is a shallow trochlea for the
astragalus (fig. 6).
Fibula.— The fibula (PI. CCXXII), articulating by a
flat proximal extremity (fig. 8) to the outer face of
the tibia, extends into a rounded or subtrihedral shaft
to form the external malleolus with facets (fig. 7 t', a,
c, and fig. 5 a) for the tibia, astragalus, and calcaneum.
Pes.— The left pes (PL CCXXVIII, fig. 2) is perfectly
preserved except the phalanges. The maximum width
of the astragalus and calcaneum combined is 144 mil-
limeters, the depth from the top of the astragalus to
the extremity of Mts III 320, to the extremity of the
phalanx of D. 3 420 (estimated). The calcaneum (PI.
CCXXIV) is distinguished by moderate length (230
mm.), marked width (85 mm.) of the shaft of the tuber
calcis (figs. 1-3), the long axis of which is obliquely
transverse. The anterior or astragalar face (fig. 1) ex-
hibits tibial (t), fibular (f) (32 mm.), ectal (a) (78 mm.),
sustentacular (a') (55 mm.), and inferior (a") (32 mm.)
facets, the last being strikingly reduced; distally
(fig. 5) the cuboidal facet is divided into two parts,
a larger anterior and a very much smaller posterior
part; it also shows the ectal (a') and sustentacular (a")
calcaneal facets. The astragalus (PI. CCXXIII) is
distinguished by the small astragalo-calcaneal inferior
facet, the very broad (44 mm.) cuboidal facet (fig.
3 c"), the moderate elongation of the neck (fig. 1);
the total width of the proximal trochlea (fig. 5) is 112
millimeters. The cuboid (PI. CCXXV, figs. 7-12) is dis-
tinguished especially by its broad abutment (fig. 12,
mt III; 35 mm.) for Mts III; superiorly the cuboido-
calcaneal facet (fig. 11; 35 mm.) is double (c, c")
and somewhat narrower than the cuboido-astragalar
facet (fig. 11a; 36 mm.) ; it unites with the ectocunei-
form (fig. 8, ec) by two separate facets, the posterior
of which is not represented in the drawing; its maxi-
mum vertical depth (fig. 7) is 46 millimeters. The
navicular (PI. CCXXV, figs. 1-6) is shallow (fig. 1; 35
mm.), presenting inferiorly (fig. 6) a broader (42 mm.)
ectocuneiform (ec) and a narrower (30 mm.) mesocu-
neiform (m) facet. The distinctive feature of the
ectocuneiform (PI. CCXXVI, figs. 7-12) is its anterior
depth (fig. 7; 27 mm.) and double, internal lateral
abutment facets (fig. 8, mt. II; fig. 12, mt. II, mt. II')
for Mts II; the mesocuneiform (PL CCXXVI, figs. 1-6)
is consequently shallower (fig 2) vertically (20 mm.),
measuring superiorly (fig. 5) 35 millimeters (tr.) by 48
(ap.). Mts II (PL CCXXVIII), measuring vertically
180 millimeters, abuts laterally above, against the ecto-
cuneiform, a primitive feature; the shaft with a sharply
convex ridge in front is deeply concave behind. Mts
III (PL CCXXVIII) , measuring 208 miUimeters verti-
cally, is distinguished by its moderately broad (42 mm.)
ectocuneiform, an unusually broad cuboidal (35 mm.)
facet; the shaft is broadly convex anteriorly and hollow
posteriorly. Mts IV (PL CCXXVII, figs. 1-7), measur-
ing 170 miUimeters vertically, exhibits a shaft subrec-
tangular in section superiorly and flattened and
rounded in its middle portion (fig. 7), with a rugose
line for muscular attachment, extending obliquely
downward and inward across the anterior face (fig.
1); the distal face (fig. 6) exhibits the large share
taken by the sesamoids (ss) and the limited share
taken by the phalanges {pJi).
Mounted skeleton in the American Museum of Natural
History provisionally referred to Brontops
A mounted skeleton and skull (No. 518) in the
American Museum of Natural History is now pro-
visionally referred to the genus Brontops. It exhibits
many specific if not generic distinctions from the
type skeleton of B. robustus just described. Yet the
characters of the skull and teeth relate it to Brontops
and clearly separate it from either Menodus or
Brontotlierium.
The skeleton is a composite. Its anterior part as
far back as the ilium belongs to one individual (Am.
Mus. 518) which was discovered by the American
Museum expedition of 1892, sent out under the direc-
tion of Dr. J. L. Wortman, who was assisted by Mr.
O. A. Peterson. It was found near the head of Corral
Draw in the Big Badlands of South Dakota. Accord-
ing to N. H. Darton, who in 1901 determined the level
of this specimen, it was found 32 feet below the 3-foot
sUiceous limestone layer at the top of the Titanotherium
zone (level Chadron C). Expeditions in two subse-
quent years, aided by the Princeton expedition,
resiflted in the discovery of the remains of other
animals of simUar proportions, which were used in
the mounted skeleton — namely, the pelvis (Am. Mus.
1065), the left tibia (Am. Mus. 1075), fibula (Am.
Mus. 1071), pes (Am. Mus. 1073, 1076), two femora
(Am. Mus. 1442, 1443). A few parts in the feet are
restored in plaster. The collocation of these hinder
parts with the leading specimen is probably not
certainly correct.
670
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
s p.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
671
As mounted with great skill by Mr. Adam Hermann,
the skeleton is about 14 feet long, 8 feet high, and 4
feet broad. The sex can not be positively determined,
as the canine teeth were wanting and are restored in
plaster. The teeth are well worn, the protocones of
m", m' being slightly abraded. The animal was
therefore in the tenth stage of growth, as defined
below, a fairly old adult.
An interesting feature of the skeleton, which may
bear upon the question of the sex, is the exostosis and
false joint in the center of the seventh rib on the right
side (fig. 606); this was undoubtedly an after repair of
a fracture, which may have been incurred in fighting.
This would support the idea that the skeleton is that
of a male, although it is known that the bulls of larger
quadrupeds sometimes charge upon females which
refuse their advances. As the cranial characters are
decidedly those of a female the latter supposition is
more probably the correct one.
The generic and specific determination of this skele-
ton is extremely difficult. It was originally referred by
Osborn to Titanotherium (Brontops) robustum, but
later he referred it to Brontotherium gigas, 9 , chiefly
because the carpus (Am. Mus. 518) diff'ers in important
characters from that of Marsh's type of Bronto-ps
robustus.
Unfortunately the specimen lacks all the front teeth,
as far back as p^ and pa. The upper premolars have
the tetartocones retarded and well constricted, as in
Brontops, and very difi'erent from the progressive cir-
cular tetartocones of male brontotheres. A supposed
female of Brontotherium gigas (Am. Mus. 1006) also
has the tetartocones much constricted, but the refer-
ence of this specimen is very doubtful. The external
cingulum of the upper premolars and molars is absent,
as it is in both Brontops robustus and Brontotheriwn, so
this character is not decisive. The external cingulum
of the lower premolars is reduced. On the whole the
molar-premolar dentition appears closer to that of
B. robustus than to that of Brontotherium.
The lower jaw as a whole presents no close resem-
blance to the jaws of brontotheres; from the type of
Brontops robustus it differs in minor characters; per-
haps its nearest resemblance is to the type of Diplo-
clonus tyleri. The significant measurements of the
skull and dentition, though few, are nearer to those of
the type and referred Brontops robustus (especially
Am. Mus. 1069) than to those of large male bronto-
theres. The skull is relatively larger than those of the
supposed female brontotheres of Brontotherium curium
and B. gigas and differs from them in many characters.
The sections and contours of the horns and nasals are
certainly different from those of the supposed female
brontotheres and still more from those of the male
brontotheres. The cranial sections are, in fact, closer
to those of Diploclonus tyleri and Brontops robustus
(especially Am. Mus. 1083).
101959— 29— VOL 1 46
The manus has the magnum much broader, more
angulate than that of B. robustus as figured by Marsh.
The lunar and magnum are also wider and more
angulate than in the supposed Brontotherium gigas
manus in the National Museum (No. 4262).
Figure 607. — Three views of mounted skeleton of
Brontops robustus, female
Chiefly Am. Mus. 518. After Osborn and Wortman (1895.105). The
skull and whole anterior part of the skeleton as far back as the pelvis
belong to one individual. The pelvis and hind limbs are supplied
from other specimens. (See fig. 606.)
As the hind limbs mounted with this skeleton belong
to other individuals they do not assist in the deter-
mination of the principal specimen.
On the whole, the evidence indicates that the
mounted skeleton is not that of a female brontothere
but is a member of the brontopine group, probably a
phase in the evolution of Brontops.
(372
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The vertebral formula is cervicals 7, dorsals 17,
lumbars 3(?), sacrals 4, caudals 20 + . In detail
the vertebrae resemble those of B. rohustus more than
those of Menodus giganteus, but they are clearly dis-
tinguished in many characters from both species.
Adaptive provision for the insertion of. a powerful
ligamentum nuchae correlated with the heavy skull
is seen in the great rise of the four anterior dorsal
spines, which are subequal in length, with their very
rugose superior extremities and triangular basal por-
tions fitted to resist the strain of the neck muscles.
The neck is short but moderately flexible; the back is
short, the tail greatly reduced.
The detailed characters of the presacral vertebrae
are best illustrated in Plates CXCV-CC, which were
drawn with such accuracy that most of the measure-
ments can be taken from them. In the following
description comparison is made throughout with the
type of Brontops rohustus.
The massive atlas measures 440 millimeters across
the pleurapophyses, which are prominently convex
(unlike the type of B. rohustus) at the extremities;
inferiorly the vertebrarterial canal traverses a broad
(54 mm.) bridge; the hypapophysis is tuberous;
the neural spine is prominent and bifid. The axis
exhibits atlanteal facets 235 millimeters in width;
the vertebrarterial canal traverses the central portion
of the platelike pleurapophysis (the lower portion in
B. rohustus) ; the superior border of the spine is some-
what indented anteriorly and broadly tuberous
posteriorly. In cervicals 3-7 the pre- and postzyga-
pophyses face obliquely inward and outward, re-
spectively, as in B. rohustus; the superior and inferior
laminae of the transverse processes differ considerably
in detail from those of B. rohustus, as seen by compar-
ison of Figure 618 and Plate CCXXIX. In the ribless
C. 7 the spine suddenly increases in height to about the
same proportions as that of B. rohustus. The cervical
centra are deeply opisthocoelous, the transverse and
vertical diameters being about ecjual.
The first dorsal is distinguished from that of the
type of B. rohustus by the fact that the postzyga-
pophyses are subhorizontal as in the succeeding dor-
sals, whereas in B. rohustus the postzygapophyses of
D. 1 and the prezygapophyses of D. 2 are transversely
oblique, resembling those of the cervicals. The neural
spines of dorsals 1-17 are characteristic; the spine in
D. 1 is greatly elevated and transversely compressed;
the spines of D. 2, D. 3, D. 4 are subequal in height,
gradually expanding into broad rugosities at the tips;
the triangular basal portion of the spine, with its pos-
terior groove, rises rapidly from D. 1 to D. 4; at the
same time the spines diminish rapidly in anteropos-
terior diameter, and this diminution proceeds as far
back as the spine of D. 14, behind which the spines
increase again in anteroposterior diameter into the
lumbar region. The zygapophyses, beginning with
the postzygapophyses of D. 1, lie in subhorizontal
planes as far back as D. 11, thence shift to transversely
oblique planes D. 12-D. 14, and into nearly vertical
planes, D. 15-D. 16. The zygapophysial facets of D.
17 and L. 1 are distinguished from all others by being
concavo-convex but not revoliite as in Palaeosyops.
The diapophyses, or attachment of the tubercles of
the ribs, pass from broadly expanded plates in D. 1
backward into tuberosities of diminishing size in D. 11,
and above these tuberosities they rise in D. 12-D. 17
into distinct vertical processes. The capitular facets
are borne largely on the posterior faces of the centra
and in a less degree on the anterior faces, throughout
as in B. rohustus, rising gradually to the upper sides
of the centra.
Three lumbars, although somewhat crushed later-
ally, afford all the principal characters. They exhibit
stouter spines and transverse processes than in B.
rohustus. The postzygapophyses of L. 1 and the
prezygapophyses of L. 2, although vertically placed,
have a convexo-concave articulation like that between
D. 17 and L. 1; in L. 2 and L. 3 the zygapophysial
facets are flattened or plane and placed in oblique
planes.
The caudals are added from another individual, and
neither the specific determination nor the enumera-
tion is certain. C. 2 is represented with a chevron,
and this, as Hatcher has pointed out, may be an
error; the first chevron may occur below C. 3, as
in B. rohustus. The centra are typically biconvex,
increasing in length from 40 millimeters in C 1 to
60 in C. 8. The bifid spines and the vestigial zyga-
pophyses persist in C. 1-C. 9; the transverse processes
subside in the same vertebrae.
The ribs are admirably preserved in a continuous
series on both sides. The general characters are as
follows: Ribs 1-5 progressively increase in length
and width of section anteroposteriorly; ribs 5-8 are
of approximately equal length but slightly decrease
in anteroposterior measurement; ribs 9-17 decrease
in length, also in anteroposterior diameter or width;
they increase in the convexity of the outer surface,
passing from a more flattened to a more convex,
lenticular section. The measurements may readily
be taken from the restoration except those of R. 1
(445 mm.), R. 2 (555 mm.), and R. 3 (625 mm.).
As compared with the Eocene titano there PaZaeosyops
the tubercles are less widely separated from the heads,
this being correlated with the less extreme transverse
extension of the diapophyses. The large size of the
ribs, the persistence of tubercles posteriorly, the
abbreviation of the lumbar region all indicate an
enormously capacious chest and abdominal cavity.
Two of the mid-sternebrae are preserved. They
are laterally compressed, with flaiing articular surfaces,
concave superior and convex inferior surfaces; the
most anterior, which may represent the second
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
673
sternal, measures 94 millimeters anteroposteriorly and
65 transversely; the succeeding sternal measures
85 millimeters anteroposteriorly and 48 transversely.
The scapula is perfectly preserved on both sides.
(The scapulae are placed too low on the side of the
thorax; cf. corrected placing
in fig. 606.) It exhibits a
vertical diameter of 657 milli-
meters, as compared with
690 millimeters in B. robustus
(type). Its maximum hori-
zontal diameter is 572 mil-
limeters. The postspinous
fossa is much more expanded
than the prespinous fossa, the
posterior border presenting a
concave backward and up-
ward surface to the thick-
ened rugose suprascapular
border, which rises with a
convex arch to a point in
front of the spine, at the
same time thinning gradu-
ally; the anterior border is
extremely thin in the mid-
section, forming a deep rec-
tangle, passing below into
the coracoid process; the di-
ameter of the glenoid border
is 160 millimeters. The tu-
berous elevated apex of the
spine is considerably below
the center of the scapula, in
contrast with that in Palaeo-
syops major, which is above
the center; the spine is still
more elevated at this point
than in B. rohustus.
The proportions of the
bones of the fore limb are
indicated by the relative
lengths of the humerus (550
mm.), radius (460), and lu-
nar to tip of middle digit
(365).
The humerus is 50 milli-
meters shorter than that of
Marsh's type of B. rohustus.
Its greater tuberosity and
deltoid crest are substan-
tially similar to those in the type of B. rohustus
(PI. CCV); its lesser tuberosity is much more
prominent than in that type. These tuberosities,
together with the great supinator crest, indicate
an enormous development of the supraspinatus, infra-
spinatus pectoral, subscapularis, deltoid, and other
muscles. The anterior trochlea of the humerus, as
in that of B. rohustus, presents a decided upward
extension of the internal face, correlated with the
elevation of the internal half of the radius. This
is a distinctive character of the titanothere forearm.
Figure 608. — Scapulae of Oligocene titanotheres
A, Brontops robuDtus, Yale Mus. 12048 (type); B, Brontops, Am. Miis. 518; C, AUops crassicornis, British Mus. 5743 M;
D, Menodus Irigonoceras, Munich Museum. One-eighth natural size.
The proportions of the radius may be judged from the
total length (460 mm.) as compared with the total
proximal breadth (165 mm.) and the distal breadth
(162 mm.), figures which should be contrasted with
those of the radius of Menodus giganteus. The prin-
cipal characters are the elevation of the inner side of
674
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
the proximal humeral facet; the presence of a promi-
nent rugosity for the insertion of the brachialis anticus
near the middle of the shaft; a deep median groove
180 millimeters below this rugosity for the common
extensor tendon; and a larger transverse diameter of
midshaft (72 mm.) as compared with the anteropos-
terior diameter (65 mm.)-
The ulna is highly characteristic, measuring 600
millimeters over all. It is distinguished by the bifid,
prominent and deeply cleft olecranon; the concave
anterior surface of the shaft with very rugose inter-
nal and external borders.; and the reduced inferior
extremity.
The manus (fig. 609) is highly characteristic of this
specimen and may be readily recognized by its breadth
its width is 89 millimeters, its vertical anterior meas-
urement 54. The lunar is broad and depressed, with
a much broader magnum facet than in B. robustus;
its width is 90 millimeters, vertical depth (anterior)
59; it also rests widely upon the unciform. The
cuneiform is characterized by a broad external
tuberosity for muscular attachment. Width 64 mil-
limeters,, vertical height 47. The trapezoid measures,
scaphoid facet, anteroposterior 74 millimeters, vertical
33. The magnum (tr. 76, vert. 47) is the most highly
characteristic element, being readily distinguished
from that of either Menodus or Brontops rohustus
(type) by (1) its vertical wedgelike extension upward
between the scaphoid and lunar, (2) its exceptionally
large size and breadth, and (3) its narrow contact with
Figure 609. — Manus of Brontops? sp. and Brontops disparf
A, Am. Mus. 518, a mounted skeleton referred to Brontopsf; B, Carnegie Mus. 92, a mounted skeleton referred to Brontops dispar?
are somewhat distorted by crushing. One-fourth natural size.
The carpals of this specimen
and massive proportions; the peculiar, irregular form
of the magnum, which is unlike that of either Menodus
or Brontops rohustus (type); the comparatively
broad and distally spreading terminal phalanges; and
the secondary carpal displacement, as shown in the
very broad lunar-magnum facet and the relatively
small size of the lunar-unciform facet. The propor-
tions are shown by the following measurements:
Width of proximal carpals (across facets), 207 milli-
meters; height of Mtc III, 218. These measurements
indicate that the manus is relatively broader, or has
less vertical depth, than that of the type of Brontops
robustus.
The carpus, considered in detail and compared
especially with the carpus of Brontops robustus (PI.
CCXXVIII), shows these principal features: The
scaphoid has a broader displacement on the magnum;
the unciform coi'related with the marked spreading
of the proximal end of Mtc III. The unciform has a
less transverse extent as seen from in front than in
either Menodus or Brontops robustus, correlated with
the less extreme displacement of the proximal carpals.
The metacarpals exhibit the following linear meas-
urements: Mtc II, 212 millimeters; Mtc III, 218; Mtc
IV, 204. They are larger and somewhat more mas-
sive throughout than in Brontops robustus. A striking
feature is the great proximal width of Mtc III and
its elongate upward extension against the unciform
between the magnum and Mtc IV; this process is
much longer than in Menodus or in Brontops robustus
type.
The terminal phalanges are highly distinctive,
being double the size of those of Menodus and much
larger than those of Brontops robustus; they are
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
675
rugose and expand distally, especially on D. 2-4,
somewhat as in the rhinoceros. The terminal phalanx
of D. 3 measures 55 millimeters transversely.
As the pelvis and hind limb of this mounted skele-
ton are not associated with the leading specimen and
not certainly determinable as to genus and species a
detailed description of these parts may be omitted.
These specimens bear the following American Mu-
seum numbers: Pelvis, 1065; left femur, 1443; right
femur, 1442; left tibia, 1075; right tibia and fibula,
493; left fibula, 1071; pes, 1073 and 1076.
Mounted skeleton of Brontops brachycephalus? in the Victoria
Memorial Museum, Ottawa
The author has had no opportunity of studying
this specimen, but a figure of it has been supplied
through the courtesy of Mr. Harlan I. Smith and Mr.
Charles M. Sternberg, of the Victoria Memorial
Museum. (See figs. 610, 611.)
Manus provisionally referred to Brontops robustus?
A large manus in the American Museum (No.
1046''^) is provisionally referred to this genus and
species. As compared with the manus of the mounted
skeleton already described (Am. Mus. 518) its meas-
urements are as follows:
Comparative measurements of manus referred to Brontops
robustus, in millimeters
.\m. Mus.
518
Am Mus.
1046 (right)
Am. Mus.
1046 (left)
Breadth _ __...._
207
100
89
54
90
59
64
47
74
33
47
76
68
92
212
56
218
58
205
100
83
52
89
°55
58
-45
76
39
41
"70
56
105
»194
»104
"70
55
69
50
Trapezoid , anteroposterior
67
37
40
65
56
57
Diploclonus Marsh
Skeleton of Diploclonus tyleri Lull
In the Amherst College Museum is an important
though incomplete skeleton (No. 327), the type of
Diploclonus tyleri Lull, which was found in Spring
Draw Basin, about 10 miles from the mouth of Bear
Creek, a tributary of Cheyenne River, S. Dak. The
« Parts of two individuals were included under the number 10!6, there being one
right and two left manus.
specimen was found lying 35 feet above the base of
some 200 feet of Titanotherium-he&img beds. The
skull and dentition are described on page 503.
The material includes a skull and lower jaw, the
atlas and axis, two cervical vertebrae, nine dorsals,
thirteen ribs, and the greater part of the fore and
hind limbs. Lull's description (Lull, 1905.1, pp. 448-
456) of the vertebrae and limb bones is as follows:
The atlas. — The atlas is a broad, heavy bone, with wide
articular facets and expanded transverse processes. The spine
is extremely low, and the short truncated hypapophysis extends
bacliward. Of the foramina, only that for the dorsal root of
the first cervical nerve is present, the ventral one, well shown
in Palaeosyops," being here represented by a deep notch as in
the rhinoceros, which our specimen also resembles in the lack
of a vertebrarterial canal and in the relative widths of the
anterior and posterior facets.
The dimensions of the atlas are:
Millimeters
Total width 320
Width across atlar-occipital facets 204
Width across atlar-axis facets 255
The axis. — The axis is a massive bone with a high neural
arch, the spine being an equilateral triangle in midsection. On
its posterior face a shallow groove arises between the zygsk-
pophyses which fades out about two-thirds of the -wsiY to the
summit. The prezygapophyses overhang the atlas in front
but present no articular facets. The odontoid process is a
truncated cone and is not so prominent relatively as in
Palaeosyops, being about one-third the length of the centrum
measured along its inferior face. The latter exhibits a low
longitudinal ridge below but is not deeply excavated on either
side, as in Palaeosyops. The transverse processes of the speci-
men are broken away, but the bases of its two supports are
seen, indicating the position of the vertebrarterial canal, which
is placed rather high on the centrum, though not on a line
with its upper surface, as is Palaeosyops.
The postzygapophyses look downward and outward; their
horizontal axes, if continued, would intersect at an angle of
90°. Altogether both atlas and axis resemble those of a
rhinoceros, much more than those of Palaeosyops.
The measurements of the axis are as follows:
Millimeters
Total height to summit of spine 295
Greatest breadth 241
Length of centrum including odontoid 133
The remaining cervicals are distinctly opisthocoelous, with
zygapophyses which widely overlap one another. With the
exception of the sixth, they are quite poorly preserved, and the
sixth is so badly crushed as to make measurements very
unreliable.
Of the dorsals, nine only are referable to the type specimen,
though three others are added in the mount. The opistho-
coelous centra are preserved, but the spines and transverse
processes are lacking.
The ribs. — Portions of thirteen ribs from both sides of the
body are preserved. In general form they are quite rhinoceros-
like, being somewhat widely expanded in the shaft. The
capitulum is nearly spherical in most of the ribs preserved, and
the two facets are separated from each other by a deep groove.
In an anterior rib, the second or third, the tubercular facet,
while mainl}' on the posterior side, arches over so as to lie in
part on the anterior face. The other ribs have the tubercular
facet entirely on the posterior face. The resemblances again
are with the rhinoceros rather than with Palaeosyops.
•' Earle, Charles, Acad. Nat. Sci. Philadelphia Jour., 2d ser., vol. 9, p. 294, 1892
676
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The appendicular skeleton
Fore limh. — While both scapulae are incomplete, the_v supple-
ment each other so that our knowledge of them is fairly perfect.
The proximal half of the left with its spine is well preserved,
while of the right nearly the entire distal border is present.
The glenoid is deeply concave anteroposteriorly and is broadly
elliptical in outline. The coracoid process is conical, somewhat
downwardly curved at the tip, separated by a deep notch from
the glenoid border, and not arising directly from it, as in
Palaeosyops, but separated by an interval of 38 millimeters.
The spine is high in the middle, with a broad roughened border.
It lowers insensibly into the general level of the scapular face
above and below, with no indication of an acromion. The tuber-
osity is not very pronounced, and the distal border is nearly
straight.
Millimeters
Breadth of shaft 85
Fore-and-aft diameter of shaft 77
The radius. — That of the left side is well preserved, except
that its distal end is somewhat weathered. It is not notably
heavy and has a well-rounded shaft, bat sUghtly compressed
fore-and-aft at the distal end. The radioscaphoid facet is pro-
longed upward on the posterior face, indicating a considerable
range of flexion of the wrist.
The principal dimensions of the radius are:
Millimeters
Length 490
Anteroposterior diameter of mid shaft 60
Lateral diameter of mid shaft 65
Lateral diameter of lower end 110
Figure 610. — Mounted skeleton of Brontops hrachycephalus?
la the Victoria Memorial Museum, Ottawa, Canada. Collected on Sage Creek, Niobrara County, Wyo., probably
from the lower Titanotherium zone (Chadron A). One complete individual except the left femur, right radius and
ulna, and a few foot bones. Maximum height 6 feet 6 inches. Courtesy of Charles M. Sternberg. Less than one-
twentieth natural size.
The dimensions of the scapula are:
Millimeters
Total length (estimated) 690
Width of superior border (estimated) 405
Fore-and-aft diameter of glenoid fossa 133
Height of spine 95
The humerus. — The distal portions of both humeri are pre-
served, but of the proximal portions that of the right only,
and as there is a portion of the shaft missing, the length can
not be measured. The distal end is broad and heavy, the ex-
ternal condyle being especially prominent and roughened for
muscular attachment. The inner trochlear is much the larger
and is higher than the outer one, thus indicating an outward
flexing of the elbow joint. The aconeal fossa is large and deep,
but there is no foramen. The breadth of the extremity meas-
ured at right angles to the axis of the shaft is 210 millimeters.
The ulna. — The entire left and fragments of the right are
preserved, except for the distal end of the former, which is
badly weathered. The ulna is notable for its huge compressed
olecranon, which widens out distally into a heavy roughened
tubercle.
The ulna measurements are as follows:
Millimeters
Length (estimated) 620
Anteroposterior diameter of olecranon from
the humeral facet 170
Lateral diameter of olecranon tubercle 140
Fore-and-aft diameter of mid shaft 80
Lateral diameter of mid shaft 80
The maniis (PI. IV, figs. 1-3). — The general proportions, well
shown in the figure, are somewhat broad rather than slender
and in direct correlation with the proportions of the skull.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
677
As has often been stated, the manus shows some distinctly
artiodactyl features, the naost notable being the retention of
four digits with tlie main axis between digits 3 and 4, rather
than lying in the third itself. Another remarkable feature is
the extreme flexibility of the carpus, esiseciaUy in the develop-
ment of a true ginglymoid joint between the proximal and
distal row of carpals. * * * AH of the elements are present
in the carpus, with the exception of the trapezium, of which
the last vestige has disappeared. The proximal facets are
shown in Figure 1 of Plate IV, though the limits of the radial
and ulnar areas are not with certainty definable.
The scaphoid articulates with the lunar by two facets sepa-
rated from each other by a roughened trough. The superior
scapholunar facet is long and narrow, its short axis vertical
and straight, while its longer axis sweeps to the rear in a gentle
convexity. It lias the same anteroposterior extent as the
scaphoradial facet above. The inferior scapholunar facet
is much smaller, having but half the fore-and-aft extent of the
superior. Distally the scaphoid articulates with the trapezoid
and the magnum and together with the lunar forms the deep
groove into which the pulley-like pivot of the magnum fits.
The lunar is a somewhat larger bone than the scaphoid,
articulating distally both with the magnum and the unciform.
The articulation between the lunar and cuneiform is again
double, the two facets being separated by a well-defined chan-
nel, which runs backward and slightly upward. The two
lunar-cuneiform facets are about equal in area. One can form
a very good idea of the distal lunar facets by the study of their
complementary facets figured in Plate IV, Figure 2.
The cuneiform is about half the bulk of the lunar and pre-
sents two facets on its inner face in every way the complements
of the lunar-cuneiform. On the pro.ximal face there is a
large, saddle-shaped facet for the ulna, and a smaller, semi-
circular cuneiform-pisiform facet in the rear, set almost at
right angles with the plane of the first. Distally there is a
large cuneiform-unciform facet, having the general form of an
equilateral triangle with rounded angles. It is again saddle-
shaped, concave in its fore-and-aft axis.
The pisiform is lacking from the right manus, but that of
the left is present and well preserved. It is much compressed
laterally, with a deep vertical expansion of the distal end, which
is decidedly rugose. The bone presents a gentle, sweeping
curve through an arc of nearly 90°. Proximally it bears two
well-defined contiguous facets for articulation with the cunei-
form and ulna respectively.
Of the distal row of carpals the trapezoid is absent, having
been replaced in the mount by that from another individual.
It is not a precise fit, there being some variation between its
facets and those of the original bone.
The articular faces are well shown in the figure, and it
will be noted that lateral movement is impossible, while a
remarkable range of flexion is indicated.
The magnu7n has on its lower face facets for the articulation
of metacarpals 2 and 3, that for 2 being rectangular, about
four times as long as wide. The pivot of the magnum is high
and prominent, as indicated in the figure.
The unciform is the largest bone in the carpus, with the
possible exception of the lunar. Distally it bears two facets
for metacarpals 4 and 5, while on the radial side there is one
which articulates both with the magnum and with meta-
carpal 3, the limits of the two articulations not being discernible.
A study of the distal carpal facets and the proximal meta-
carpal ones gives evidence again of more or less fore-and-aft
movement, but in the case of the median metacarpal no lateral
movement at all. The lateral metacarpals, on the contrary,
were capable of lateral as well as fore-and-aft movement, so
that, while the foot would spread somewhat when the creature's
weight was borne upon it, it was all in the lateral bones.
This would seem to be stiU further evidence that the true axis
of the foot was between digits 3 and 4, as in the artiodactyls.
The principal dimensions of the manus are:
Millimeters
Width of proximal facets 170
Width of distal carpals 170
Depth, lunar to summit of metacarpal 3 80
Length of metacarpal 3 250
The hind limb
The entire limb is figured in Plate IV, Figure 4. There was no
trace of the pelvis found associated with No. 327, though the
limbs are in excellent preservation and give but little evidence
of distortion by crushing.
The femur. — This is a fine bone, notable for its extreme flat-
ness, which indicates the pillar-like posture of the bone, as in
the elephant, as the shaft would not have been sufficiently
rigid to withstand springing had the thigh been flexed.
Another interesting feature is the absence of a third trochanter — ■
a character given by Marsh in his definition of the genus
Megacerops. There is a ridge on the outer side of the femur
1
1
1
t^
'JUf
!p^
i^^^H
1
<.isll
1
■li^^^^^l
J^v.
^
IH
Figure 611. — Mounted skeleton of Bron-
tops brachycephalus?
Oblique front view. In Victoria Memorial Museum,
Ottawa (shown also in fig. 610). About one thirty-
fifth natural size.
continuous above with the great trochanter, which probably
represents the vestige of the third. The measurements are:
Millimeters
Length 785
Width of proximal end 236
Width of distal end 204
Width of mid shaft 117
Depth of mid shaft 60
The tibia. — The general form of this bone is well shown in
the figure and calls for no special comment. The measure-
ments of the tibia are:
Millimeters
Length 446
Width of proximal end 200
Depth of proximal end 132
Width of mid shaft 80
Depth of mid shaft 77
The fibula is quite slender with expanded articular extremi-
ties; length, 395 millimeters.
The pes. — The general proportions are in keeping with those
of the manus. All of the tarsal elements are represented, with
the exception of the entocuneiform, which is entirely lacking.
678
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
The calcaneum. — The tuberosity is rather long and very
rugous at its distal end, and with a much flattened shaft which
is about one-half as wide as long. The calcaneum bears facets
for articulation with the cuboid, the astragalus, and on its
upper outer face a small one for the articulation with the
fibula.
The calcaneo-astragalar facet is somewhat saddle-shaped, its
fore-and-aft axis being a reversed curve, first concave, then con-
vex. The sustentacular facet, however, is deeply concave, the
transverse axis curving through an arc of 90°, while the fore-and-
aft axis is straight. Below there is but one facet, the calcaneo-
cuboid, somewhat semilunar in shape, extending about half the
width of the bone. Except for the articulation with the fibula,
tliere is little evidence of movement between the calcaneum
and the adjoining bones.
The astragalus presents a beautiful hourglass-sliaped astrag-
alotibial facet, bearing on its outer face a clearly defined fibula
facet. The range of flexion and extension in the tibiotarsal
joint is considerable. Distally two facets are indicated, that
for the navicular being by far the larger and somewhat flat-
tened, and with a small, prominent, downwardly projecting
process, which effectually limits anj' fore-and-aft motion. A
prominent ridge divides the two facets, that for the cuboid
being an elongated triangle, first convex and then concave
from before backward.
The navicular is very flat and presents two distal facets, that
for the ectocuneiform being the larger and somewhat triangular
in shape, while that of the mesocuneiform is semilunar, the
line of demarcation between the two being almost straight.
The cuboid has a thickness equal to that of the navicular and
ectocuneiform combined and exhibits proximally two contig-
uous facets, the lesser for the calcaneum and the greater for the
astragalus. Distally there are two facets, the external, the area
of which is about four times the greater, being for metatarsal
4. This is somewhat saddle-shaped, while the other, that of
metatarsal 3, is nearly flat.
The ectocuneiform is absent in the right pes, though present in
the left, being rej^laced in the former by that of another indi-
vidual. It articulates distally with metatarsals 3 and 2, though
the latter articulation almost fades out anteriorly, broadening
as one goes to the rear. This is markedly different from most
titanothere feet which the author has seen, in wliicli a wide line
of contact is indicated on the face of the tarsus. There is,
however, no possibility of contact between the miesocuneiform
and metatarsal 3, as the former articulate distally with meta-
tarsal 2 only.
The whole pes is remarkably rigid when compared wath the
manus, as there is little indication of any intertarsal movement,
none between the tarsus and the median metatarsal, and no
lateral and but little fore-and-aft play between the tarsus and
the lateral metatarsals.
The principal dimensions of the pes are:
Millimeters
Width of astragalar facet. 105
Length of calcaneum . 208
Width of the distal row of tarsals 140
Height, astragalus to proximal end of the
third metatarsal 108
Length of the third metatarsal 205
Conclusion
The general proportions of the skeleton would indicate a
huge animal, 7 feet 4 inches in height to the withers and some-
thing over 12 feet in length, somewhat rhinoceros-like in aspect,
but with more massive, pillar-like limbs, which, as Professor
Osborn has shown, are correlated with great weight. The
extreme flexibility of the carpus seems to indicate an elephant-
like habit of kneeling on the wrists when rising and lying down.
The creature was hardly adult, as indicated by the unossified
vertebral epiphyses, though probably of full stature, and it
indicated a form in the middle stage of evolution.
The numerous resemblances in both the skull and
the skeleton of Diplodonus tyleri to Brontops robustus
strengthen the conclusion that the Diplodonus phylum
is an offshoot of the Brontops phylum. The resem-
blances extend to the principal measurements of the
skull and skeleton in the end members of the two
genera. The chief differences are seen in the wider
and more specialized manus of Brontops rohustus and
in its longer femur. (Gregory.)
SUBFAMILY MENODONTINAE
Ailops Marsh
Two very incomplete skeletons are associated with
skulls of the genus Ailops, one in the Field Museum
at Chicago referred to Ailops marshi, the other in the
British Museum (Natural History) referred to Ailops
crassicornisf. This meager material has so far
yielded but few clear generic characters of the post-
cranial skeleton; it merely indicates that the smaller
species of Ailops have short limbs as compared with
both Brontops rohustus and Menodus trigonoceras.
Ailops marshi
An incomplete skeleton in the Field Museum (No.
P6900) comprises six dorsal vertebrae, 21 caudals,
part of the pelvis, and much of the limbs. The skull
of this specimen is described on page 514. This
skeleton as mounted in the Field Museum, under the
direction of Dr. E. S. Riggs, is figured below.
Vertebrae. — The dorsals are strongly opisthocoelous.
The centra measure from 48 to 50 millimeters on the
midventral line. The 21 caudals measure 111 milli-
meters in length, the .tail being relatively long. The
midcaudals increase in length; the posterior caudals
diminish, as shown by the following measurements:
Millimeters
Millimeters
Millimeters
1_-
40
40
8 ...
67
15
16... _
53
2
9
64
51
3
41
10___
65
17
46
4
50
11_._
63
18
41
5
60
12..^
63
19
36
6.:
63
13___
14_._
61
59
20-_..
21___.
31
7
65
25
Fore limh. — The height of the fore limb at the
shoulder is estimated at 1,285 millimeters.
The right scapula is nearly complete, except at the
top. Its height is estimated at 500 millimeters, as
compared with 690 in Brontops robustus and 640 in
Menodus trigonoceras.
The humerus measures 425 millimeters from the
head to the distal trochlea as compared with 615 in
Brontops robustus (type) and 557 in Menodus trigo-
noceras. The crest of the great tuberosity is mod-
erately developed, much lower than in B. robustus; it
is continued antero-internally into the usual incurved
EVOLUTION OP THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
679
hooklike process for the msertion of the supraspinatus
muscle; this process is much less expanded than in
B. rohustus. The deltoid tuberosity, on the lower
outer part of the deltopectoral crest, is also less devel-
oped than in B. robustus. The circumference of the
humerus at the nan'owest part of the shaft is 230
millimeters. The middle of the shaft is rather slender,
but the proximal end is wide. The rugosity for the
brachialis anticus, on the anterior face, is prominent.
The radius is only 360 millimeters in length, as
compared with 495 in Brontops robustus (type) and
525 in Menodus trigonoceras. The ratio of the length
of the radius to the basilar length of the skull, in per
cent, is as follows: Allops marshi, 53; Brontops
robustus (type), 64.8; Menodus trigonoceras, 75.
The ulna also is short, 475 millimeters in length,
as compared with 680 in Brontops robustus (type) and
595 in Menodus trigonoceras. The olecranon is long
and less expanded than in B. robustus (type).
The manus is small, measuring only 142 millimeters
across the proximal carpals as compared with 200 in
Brontops robustus (type). At the same time it is
relatively wider than that of Menodus trigonoceras,
which has about the same absolute width (149 mm.)
but a far longer median metacarpal. The carpals
present nothing remarkable. The combined trans-
verse width of the scaphoid and lunar is 95 milli-
meters; that of the cuneiform 49. The depth of the
carpus from the top of the lunar to the summit of
Mtc III is 63 millimeters. The median metacarpal
(Mtc III) is relatively and absolutely short and wide,
155 millimeters in length as compared with 225 in
Brontops robustus (type) and 233 in Menodus tri-
gonoceras.
Hind limb. — The hind limb is quite short, the length
of the limb from the head of the femur to the ground
being only 1,150 millimeters (estimated), even with
the limb fully extended.
The crest of the ilium is relatively narrow, measuring
530 millimeters in width.
The femur (length 590 mm.) is relatively shorter
than in Brontops robustus (type), the ratio of the
length of the femur to the basilar length of the skull,
in per cent, being as follows : Allops marsJii, 90; Brontops
robustus (type), 107; Menodus trigonoceras, 100.4.
The circumference of the shaft of the femur is 210
millimeters.
The tibia is short (length 350 mm.), the compara-
tive ratios of the length of the tibia to the basilar
length of the skull, in per cent, being as follows:
Allops marshi, 53; Brontops robustus (type), 56;
Menodus trigonoceras, 60.
The pes is small, the length of the calcaneum being
only 147 millimeters as compared with 230 in Brontops
robustus. The width of the tarsus across the navicular
and cuboid is 90 millimeters, the width of the astraga-
lus 70, and that of the cuboid 42. The tuber calcis is
oval in section. The facet for the fibula on the calca-
neum is prominent.
Allops crassicornis (referred)
An incomplete skeleton in the British Museum
(5743 M) belongs with the skull described in Chapter
VI. The poster anial skeleton is represented by the
atlas, axis, two cervicals, two dorsals (all more or less
crushed or imperfect), the right humerus, radius,
and manus (lacking cuneiform and two distal rows
of phalanges), the left femur, tibia, and astragalus,
the right pes (lacking meso- and entocuneiform and all
the smaller phalanges), the right and left patellae.
Vertebrae. — As compared with B. robustus the trans-
verse processes of the atlas were proportionately
smaller, projecting less, rounded instead of truncate
distally; the facets for the occipital condyles were less
oval in shape and tapering more at the bottom; in top
view the atlas lacks the median groove in the neural
arch seen in B. robustus. In the inferior view there is
a decided median process from the posterior haemal
surface.
Measurements of the atlas
Millimeters
Extreme width 275
NTeural arch, maximum anteroposterior 87
■ Width across cotyli 180
In comparison with B. robustus (type) the verte-
brarterial foramen of the atlas is larger, the dorsal
expansion of the neural arch is lighter, the odontoid
less spout-shaped, the condyles for the atlas more ver-
tical, narrower transversely and deeper vertically; on
the inferior surface the haemal ridge seems more pro-
nounced.
Measwfiments of the axis
Millimcteis
Width across cond^des (estimated) 180
Odontoid to posterior border of centrum 93
Fore limb. — The scapula is well preserved, 570 milli-
meters in length and 290 in width, less expanded
laterally than in B. robustus. The humerus (length
450 mm.) is slightly longer than that of the Field
Museum specimen of A. marshi (425). The circum-
ference of the shaft measures 225 millimeters. The
radius is similar to that of A. marshi, its length being
350 millimeters, circumference of shaft 160, breadth of
distal end 120, of proximal end 117, least width of
shaft 48. The manus is noteworthy for the wide dis-
placement of the scaphoid on the magnum, of the
lunar o n the unciform, so that there is only a very
narrow lunar-magnum facet. Perhaps in correlation
with this character the median metacarpal (Mtc III)
appears to be relatively somewhat wider than in typi-
cal Oligocene titanotheres.
680
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 612. — Parts of skeleton of Allops crassicornis?
British Mus. 5743 M, consisting of parts of skeleton, including skull and lower jaw. Ai, Atlas, front view; As, atlas, top view; Bi, axis, side view; Bs, axis, front
view; C, scapula; Di, right humerus, outer side view; Ds, right humerus, front view; E, right manus; F. left femur, front view; G, left tibia, front view.
One-sixth natural size.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
681
Measurements of manus
Millimeters
Manus, lunar to D. Ill, phalanx 3 240
Transverse, across carpals 150
Scaphoid, breadth 65
Scaphoid, height 45
Lunar, breadth 50
Lunar, height 40
Trapezoid, breadth 40
Trapezoid, height 29
Magnum, breadth 48
Magnum, height 31
Unciform, breadth 77
Unciform, height (maximum) 41
Hind limb. — The total height of the hind limb from
the femur to the bottom of Mts III is estimated
Lower Oligocene titanotheres of dolichopodal (rela-
tively narrow-footed) and swift-limbed type.
Menodus heloceras? (Cope)
The specific reference of the finely preserved pes of
Menodus in the American Museum (No. 1080) is
somewhat doubtful, but there can be no question
that it belongs to one of the small and primitive
species of that genus {M. heloceras or M. proutii).
The family characters are well marked — namely,
(1) large fibulocalcaneal facet; (2) large tibio-
calcaneo-astragalar facet; (3) widely separated "sus-
FiGURE 613. — Pes of Menodus trigonoceras, referred, and M. heloceras
Ai, FroDt view of pes of M. trigonoceras (Am. Mus. 1079), with ectocuneiform and Mts IV restored. (This pes was
associated with the manus shown in fig. 614, A.) Aj, The same, phalanges of D. Ill, front view. B, Front
view of pes of M. heloceras (Am. Mus. 1080, reversed). One-fourth natural size.
as 1,170 millimeters. The femur is short, length
560 millimeters, circumference of shaft 230. The
tibia also is very short (345 mm.).
Measurements of the pes
Millimeters
Total length, os calcis to D. Ill, phalanx 3 320
Width across astragalus and caleaneum 100
Calcaneum, length 140
Caleaneum, greatest breadth 93
Astragalus, length (vertical) midline : 60
Astragalus, breadth 82
Cuboid, vertical height, midline 27
Cuboid, breadth 60
Navicular, vertical height 20
Navicular, breadth 63
Mts III, height 155
tentacular, " " ectal, " and "inferior " astragalocalcaneal
facets; (4) displacement of Mts III against cuboid
and of Mts IV against ectocuneiform, displacements
that are very slight and that indicate the primitive
character of this pes.
The generic characters of Menodus are equally well
marked — namely, (1) pes relatively long and narrow,
or elevated, in all its elements; (2) long axis of
tuber calcis extended obliquely anteroposteriorly, not
transversely as in Brontops, the tuber terminating in
a rounded suboval head; (3) sustentaculum of calca-
neum slender and downwardly directed; (4) cuboid
relatively narrow, vertically extended, with slight
lateral abutment against Mts III; (5) mesocuneiform
682
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
distally shorter than ectocuneiform; (6) enlargement
of Mts IV and reduction of Mts II indicating that
the main weight is carried upon the third and fourth
digits (D. 3 and D. 4), the second digit (D. II) being
relatively reduced; (7) corresponding reduction of Mts
II and sharply convex ridge on its anterior face, as
in Brontops rohustus (type); (8) striking narrowness
or reduction of terminal phalanges, which are even
narrower than the middle phalanges.
The specific characters of this pes (those of M.
proutii or M. heloceras) are (1) simple, oval shaft
defined facet for Mts III, the displacement being
rather incipient; (6) similarly ectocuneiform deep-
ened (vert. 23, tr. 35), with a characteristic pair of
ectal facets for the cuboid and a double pair of ental
facets for the mesocuneiform and Mts II, respectively;
(7) mesocuneiform correspondingly shallow (vert. 16,
tr. 23, ap. 36) for the support of the reduced Mts II;
(8) the shaft of Mts IV considerably exceeds that of
Mts III in stoutness and greatly exceeds that of Mts
II; (9) Mts III (length 175 mm.) is distinguished
by a deep antero-internal groove in the superior
Bi "^^^ ^ Ai
Figure 614. — Manus of Menodus trigonocerasf
A, Am. Mus. 1079 (compare fig. 613): Ai, Front view, cuneiform restored; Aj, ptialanges of digit III; Aj, top view of carpus. B, .\.m. Mus. 515;
Bi, Manus, front view (lunar and magnum incorrectly restored); Bi, phalanges of digit III. One-fourth natural size.
of the tuber calcis, the entire length of the calcaneum
being 158 milUmeters; (2) relatively narrow (22 mm.)
displacement of the astragalus upon the cuboid;
(3) total width of 191 millimeters of the combined
distal facets of the calcaneum and astragalus; (4) navi-
cular relatively deep, measuring 22 millimeters
vertically, 58 transversely, with broad (35) ectocune-
iform and relatively narrow (19) mesocuneiform
facets; (5) cuboid relatively deep (vert. 35, tr. 45)
with broader calcaneal than cuboidal facets prox-
imally, an extremely large facet (tr. 37, ap. 47)
for the enlarged Mts IV, and a relatively small, ill-
portion of the shaft, and Mts II (length 156) has a
subtriangular section, owing to the anterior ridge and
grooving and flattening of the outer side facing toward
Mts III.
Among the chief primitive and specific characters
of this pes are, therefore, the relatively narrow dis-
placement of the astragalus and cuboid and of Mts II
and Mts III against the cuboid and ectocuneiform
respectively.
The proximal phalanges of D. 2 and D. 4 are
relatively elongate and are laterally compressed, a
striking; feature of this re2;ion being that the middle
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
683
phalanges are carried on the plantar side of the prox
imal, there being apparently a sharp flexure between
them. The distal phalanges are very narrow and
but slightly expanded at the extremities, the trans-
verse measurements being, II, 30 millimeters; III,
30+ ; IV, 28.
Menodus trigonoceras
Referred manus and pes
The reference of the pes of Menodus trigonoceras?
(Am. Mus. 1079) to the genus Menodus is confirmed
more primitive stage of M. Tieloceras and relate it to
the more progressive stage which we suppose to be that
of M. trigonoceras. All the dimensions of the pes are
larger, and the bones of the tarsus are decidedly
broader and flatter; the summit of the tuber calcis
is of more flattened or elongate oval section; the
cuboid is relatively broadened and flattened, and
Mts III is broadly articulated with it by displace-
ment, and the proximal phalanges, especially those of
D. 3 and D. 4, are shorter.
Figure 615. — Restorations of Menodus trigonoceras (A) and Allops marshi (B)
Not drawn to scale. A is baseiS on the mountei skelston in the Munich Museum and is about one twenty-second natural size (6 feet 5 inches).
by the presence of all the generic characters already
enumerated, as found in the pes referred to M.
lieloceras or M. proutii — namely, the slender, rounded
shaft of the tuber calcis, the narrow, obliquely directed
sustentaculum, the reduced cuneiform, the relatively
slender Mts II with the sharp ridging and grooving of
the superior portion of its shaft, the small size of the
terminal phalanges, the laterally compressed cuboidal
facet. We note, however, several important progres-
sive characters in this pes, which remove it from the
The calcaneum measures 155 millimeters vertically;
the combined distal astragalar and calcaneal facets
measure 105 millimeters transversely; the navicular
is decidedly more flattened (vert. 22, tr. 53) ; the cuboid
also is more flattened (vert. 36, tr. 57), but it still
exhibits a narrower astragalar (27) than calcaneal (42)
facet; the mesocuneiform is distinctively small (tr.
23, ap. 43). Mts II is readily distinguished by its
anterior median ridge bordered ectally by a groove;
it measures 173 millimeters vertically; it exhibits a
684
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
narrow, proximal ectocuneiform facet. Mts III, a
large and elongate bone, measures 200 millimeters
vertically and exhibits ar elatively broad (20 mm.)
proximal abutment against the cuboid.
compared with those of M. heloceras are broad and
strong; the terminal phalanges are broader than in
M. Tieloceras but still retain the small dimensions
characteristic of this genus; the marked expansion of
Figure 616. — Mounted skeletons of Brontops dispar? and Menodus trigonoceras
A, B. dispar?, supposed female, Carnegie Mus. 92; Warbonuet Oreek, SioiLx County, Nebr.; lower Titanottieriitm zone, perhaps 30 feet
above the Pierre shale (Hatcher). Adapted from a photograph published by Hatcher. The scapula and fore limb are here placed
higher up on the thorax than in the skeleton as mounted. The skull, destroyed by weathering, is here restored in outline from
supposed females of B. dispar. B, Skeleton of M. trigonoceras in the Munich Museum; Hat Creek badlands, Sioux County, Nebr.
(compare fig. 615, A). In the mounted skeleton the thorax is placed too high in relation to the scapula, but this error is corrected in
the drawing. Both figures ore twenty-second natural size.
As shown in Figure 613, in the relatively small D. 2
the proximal phalanx is more primitive — that is,
narrow and elongate; the distal phalanx is small; on
D. 3 and D. 4 the proximal and median phalanges as
the phalanges on D. 2, D. 3 is in keeping with the
tendency prophesied in M. heloceras or M. proutii
toward the support of the greater weight of the limb-
on the third and fourth digits.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
685
The manus of Menodus (Am. Mus. 1079, fig. 614),
fortunately associated with the pes above described,
is no less readily distinguished in all its parts from
the type of Brontops robustus. Its most striking dis-
tinctions lie (1) in the high, narrow proportions of the
manus as a whole, but especially of the carpus; (2) in
the extreme displacement illustrated especially in the
broad abutment of the lunar on the uncifoi'm with the
correspondingly reduced narrow oblique lunar magnum
facet; and (3) in the small simple form of the magnum
which readily distinguishes it from that of Brontops.
The proportions are shown in the following measure-
ments: Proximal breadth of carpus, 155 millimeters,
as compared with length from lunar to extremity of
D. 3, 340 millimeters; thus the carpus is relatively
narrower from side to side and deeper anteroposte-
riorly in proportion to its length than that of Brontops.
The extreme displacement of the carpus is first
noticed in the relatively wide extension of the scaphoid
on the magnum, in the correspondingly small weight
the lunar transmits to the magnum, as compared with
its broad surface resting on the unciform. The an-
teroposterior measurement of the radial face of the
lunar (73 mm.) decidedly exceeds its transverse
measurement (59 mm.), whereas in Brontops rohustus
(type) the anteroposterior and transverse measurements
of the radial facet of the lunar are subequal. The
radial face of the scaphoid measures 55 millimeters
transversely by 59 anteroposteriorly. The cuneiform
is wanting in this specimen. The trapezoid, which
exhibits no face for a trapezium facet, measures 28
millimeters vertically, 38 transversely, 48 antero-
posteriorly. The magnum measures 25 millimeters
vertically by 54 transversely ; it is of simpler and more
quadrate form than in Brontops owing to the oblique
ectal facet for the lunar above and the unciform below.
The unciform is decidedly broad, constituting one of
the most characteristic features of the Menodus carpus
with its broad abutment against Mtc III, measuring
transversely 82 millimeters. The metacarpals are
elongate but not extremely so, the vertical measure-
ments through the middle of the shaft being, Mtc II,
196 millimeters; III, 199; IV, 188; V, 172. The shafts
are relatively much more slender than in Brontops
rohustus. Mtc III exhibits a short, broad process
abutting against the unciform, like that seen in Bron-
tops.
Mounted skeleton in the State paleontologic collection at Munich
This specimen (fig. 616, B) from the Hat Creek bad-
lands of Nebraska comprises the greater part of the
skeleton. It was presented to the Museum by Com-
merzienrath Th. Stiitzel in 1897. The skull is a typical
Menodus trigonoceras. (See p. 528.) The limbs are
notably slender and elongate, in comparison with those
of Brontops, Allops, and Brontotherium, the apparent
slenderness of the limbs having been increased by
crushing. The following description is based on the
observations of the author and of Dr. Ernst Stromer
von Reichenbach.
Mounting. — The skeleton as mounted has the
scapulae placed too low on the sides of the thorax,
the curvature of the backbone is not quite correct, the
intervertebral spaces are not wide enough, so that the
backbone is somewhat too short. Certain parts of
the skeleton are more or less restored in plaster,
especially the symphysis of the lower jaw, the outer
side of the first three right upper premolars, the middle
of the upper border of the occiput, both wings of the
atlas, the spinous process of the atlas, the posterior
half of the tail, the greater part of the left ilium, both
hind feet (except the upper and lower ankle bones
and the second right metatarsus).
Dimensions. — The total length of the skeleton from
the tip of the nasals to the drop of the tail is estimated
at 3 meters. The height at the withers, to the tip of
the second dorsal spine, is 2.28 meters.
Figure 617. — Left astragalus of Menodus giganteus
Am. Mus. 505, Chadron. Front and rear views. One-third natural size.
Vertebrae. — The vertebral formula of the specimen
as mounted is cervicals 7, dorsals 17, lumbars 3,
sacrals 4(?), caudals 21 (restored).
As compared with Brontops robustus (type) the
spine of the atlas is higher and its lateral process
thicker, the spine of the axis is longer and lower, the
centra of the cervical vertebrae are longer and their
spines lower, except the spine of C. 7, which is higher;
the spines of the first two dorsals are subequal in
height and of fairly uniform diameter from base to
summit; the spines of the dorsals D. 4 and D. 5 are
straight, less concave anteriorly; the parapophyses
appear to be lower. The spine of the third dorsal
(D. 3) is the highest; behind that the spines decrease
gradually in height to the caudals; all appear thin
transversely and relatively high as compared with
those of B. robustus. The spines of D. 1-D. 4 are
remarkably thin and lack the deep posterior grooves,
except at the base. The articular facets of the pre-
zygapophyses in the third cervical vertebra present
upward and somewhat inward; in the succeeding ver-
tebrae they gradually turn more strongly inward,
especially in the midthoracic vertebrae. On the last
dorsal and first lumbar the articular processes are
weaker than usual. The last dorsal bears a shallow
facet for the tubercle of the rib. The lumbar vertebrae
bear flattened diapophyses just above the base of the
neural arches. In D. 1 and D. 2 the posterior zyga-
pophyses are nearly horizontal.
686
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
FiGTJRB 618. — Cervical and first four dorsal vertebrae of Brontops robustus and Menodus giganteus
A, B. robustus, vertebrae of type (Yale Mus. 12048); B, M. giganteus, vertebrae associated with skull in the Field Museum (P 5927).
Both one-eighth natural size.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
687
Longitudinal axial measurements of vertebrae
MillimelDrs
7 cervicals, as mounted, anteroposterior 600
17 dorsals, as mounted, anteroposterior 1, 270
3 lumbars, anteroposterior 230
Sacrals (crushed) , anteroposterior 150
21 caudals, anteroposterior 1,000
Atlas, ventral arch, anteroposterior 65
Axis, length (ventral, omitting the odontoid tip) 110
Sixth cervical, ventral length of centrum S6
First dorsal, height of spine 43
Rihs. — The ribs are long and slender, and the chest
was probably narrow. The sixth rib is the longest.
The estimated length of some of the ribs on the right
side is as follows:
Millimeters
R 1.
112.
R5-
Millimeters
445
530
795
R6.-
R8_-
R 17-
870
515
Fore limh. — As mounted the height from the top of
the scapula to the ground is 2,400 millimeters, the
lengths of the successive segments being, scapula 640,
humerus 557, radius 525, manus 350. Tlie scapula is
very distinctive in type, being long and narrow ante-
roposteriorly. The glenoid is narrow, and so is the
coracoid process. The spine is but little expanded,
and the postscapular fossa is not greatly extended
posteriorly. The humerus, altliough considerably
crushed, is profoundly different in proportional char-
acters from that of Brontops rohustus, being long and
slender, with a small crest on the great tuberosity and
less developed deltoid and supinator crests. The
radius and ulna also are long and slender; the ole-
cranon is very deep, laterally compressed, and with a
deep terminal groove. The manus is high and narrow
with long metacarpals and a narrow carpus. The
scaphoid is narrow, with a deep concave facet for the
trapezoid, which was vertically high. The lunar is
narrow, with a fair-sized facet for the magnum.
The magnum also is rather narrow. The unciform is
deeply extended downward on the outer side, for
Mtc V. The terminal phalanges are narrow.
Measurements of fore limb
Millimeters
Scapula, height (middle of glenoid to dorsal border) 640
Scapula, width (at right angles to preceding measure-
ment) 460
Humerus, length (head to capiteUum) 557
Humerus, right, extreme length 620
Radius, right, midlength 525
Radius, width of head 145
Radius, width of distal end 125
Ulna, length 660
Manus, length 350
Width across distal end of radius and ulna 170
Carpus, width 149
Carpus, height, lunar to summit of Mtc IV 80
Scaphoid, width 52
Scaphoid, height 47
Lunar, width 65
Lunar, height 72
Trapezoid, anteroposterior 47
Mtc II, height 214
101959— 29— VOL 1 17
Millimeters
Mtc III, height 233
Mtc IV, height 225
Mtc V, height (estimated) ^ 191
Terminal phalanx of D. II, right, width 34
Terminal phalanx of D. Ill, right, width 43
Terminal phalanx of D. IV, right, width 38
Terminal phalanx of D. V, right, width 29
Pelvis and Jiind limb. — The hind limb also is long
and slender, the total height of the hind limb as
mounted being 1,430 millimeters; the height from the
top of the ilium to the ground 1,900. The pelvis is
elongate and narrow (partly increased by pressure),
the total length being 820 millimeters, and the
breadth as mounted 830. The length of the outer
Figure 619. — Manus referred to Menodus
giganteus
Yale Mus. 12012. Figure prepared by Berger under direc-
tion of Marsh. Ob^er^e the small terminal phalanges
and the relatively high, narrow lunar. Slightly less
' than one-flfth natural size.
part of the crest of the ilium is 580, of the inner part
290. The femur is extraordinarily long and slender.
The hind feet are poorly preserved and do not give
much character.
Measurements of hind limb
Millimeters
Femur, midlength 703
Femur, least width of shaft S5
Tibia, length 430
Astragalus to Mts III 302
Menodus giganteus
The postcranial skeleton of this species is positively
known chiefly from a scapula and a series of vertebrae
in the Field Museum (No. 5927), which is associated
with a superbly preserved skull. (See p. 535.) An
astragalus associated with the fine skull Am. Mus.
688 TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
505 is shown in Figure 617. Referred material of , . ^^ ' Millimeters
,,., . -iiviAT Cervical 7, total height 585
M. giganteus comprises a manus m the i ale Museum p^j-sal 1 total height . ... 730
(No. 12012, fig. 619), a radius and ulna in the Came- Dorsal 4^ length of centrum 78
gie Museum (No. 120), and a pelvis with both hind
limbs in the University of Nebraska Museum (No. Radius and uina in the camegie Museum
3296) The radius (Carnegie Mus. 120) is very long and
Cervical and dorsal vertebrae in tiie Field Museum slender
The vertebrae in the Field Museum (No. 5927) Radius length . 475
include the seven cervicals and the first five dorsals. Radius, width at top 123
The centra of all the vertebrae are relatively elongate. Ulna, length 595
Figure 620. — Restorations of Brontotherium leidyi (A) and B. platyceras (B)
About one-thirtieth natural size.
and the first two dorsals have very long and broad
spines. (See fig. 618.)
Measurements of vertebrae
Millimeters
Length of six cervicals (without atlas) 575
Length of seven cervicals (with atlas) (estimated) 660
Axis, length of centrum (excluding odontoid process) 150
Cervical 3, length of centrum 83
Cervical 4, length of centrum 80
Cervical 5, length of centrum 78
Cervical 7, length of centrum 76
Pelvis and hind limbs in the Nebraska M
A pelvis with both hind limbs in the Museum of the
University of Nebraska at Lincoln (No. 3296) is
referable either to M. trigonoceras or to M. giganteus.
All the elements are relatively long and narrow.
Millimeters
Pelvis, length 902
Pelvis, width (crushed) 952
Femur, length 762
Tibia, length 432
EVOLUTION or THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
689
Manus in the Yale Museum
A manus in the Peabody Museum of Natural His-
tory at Yale University (No. 12012) may pertain to
Menodus giganteus; there is no proof of association
with Marsh's type of Brontotherium ingens ( = Meno-
The manus (fig. 619) is of the high, narrow type seen
in Menodus. The phalanges are bent back in the
rock so as to lie against the back of the hand. In the
figure they are represented a little too small and
appear to taper too suddenly.
Figure 621. — Atlas and axis of Brontotherium leidyi
t, Carnegie Mus. 93; atlas, top view. B, Carnegie Mus. 114; atlas and axis proTisionally referred to
B. leidyi. Bi, Atlas and axis, side view; Bj, atlas, top view; B3, front view. One-eighth natural size.
BE C
Figure 622.— Vertebrae of Broniops rohustus (A, B, C), Yale Mus. 12048 (type),
compared with those of Brontotherium gigas (D, E, F), Am. Mus. 492
A, D, Third cervical vertebra; B, E, third dorsal vertebra; C, F, second (?) lumbar vertebra. The third
cervical vertebra of Brontops has a much longer centrum and a stouter neural arch. (The neck of
Brontotherium was shorter.) The third dorsal vertebra of Brontops has the spine curved anteroposteriorly
and the lateral process less elevated. The second (?) lumbar vertebra of Brontops has a longer centrum
and more recumbent neural arch. One-eighth natural size.
dus giganteus), as the manus was received at the
Museum in 1874, whereas the skull was received in
1873, although both came from the same general
region and from the same collector (Devendorf).*^
*8 Information kindly supplied by Prof. R. S. Lull.
SUBFAMILY BEONTOTHERIINAE
The Brontotheriinae include lower Oligocene titano-
theres, extremely graviportal and brachypodal, espe-
cially in the pes.
690
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Brontotherium Marsh I Brontotherium? sp.. Am. Mus. 1443, manus; Am. Mus. 1047,
manus and pes (specific reference doubtful).
Iq the genus Brontotherium the only postcranial This material may now be described as a whole, the
elements that are certainly associated with identified several parts being compared with those of Brontops
skulls are the following: 1 rohustus (type).
FiGUEE 623. — Scapulae of Oligooene titanotheres
Figures prepared by Berber under the direction of Marsh. A, Brontotherium gigas halcheri, Nat. Mus. 4262: A', Outer
side; A', inner side; A', distal view. B, BrontopsT: B', Outer side, B', inner side. One-eighth natural size.
Brontotherium leidyi, Carnegie Mus. 93, skull, lower jaw, atlas,
scapula, humerus, radius and ulna, femur, tibia.
Broniotherium gigas, Am. Mus. 492, sliull, pelvis and sacrum,
second dorsal vertebra with ribs, ulna, lunar.
Brontotherium gigas hatcheri, Nat. Mus. 4262, sliull, lower jaw,
the greater part of both fore limbs and feet, parts of both
hind limbs and feet, and probably the pelvis. Parts of two
other individuals were mixed with this skeleton but have been
separated by J. W. Gidley.
Vertebrae referred to Brontotherium leidyi, Carnegie Museum
(Figs. 621, 622]
The atlas of B. leidyi (Carnegie Mus. 93) is com-
paratively small, measuring only 305 millimeters in
greatest transverse diameter. It differs from that of
Brontops rohustus (type) in being proportionately
higher, less extended transversely, with less expanded
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHEKES
691
lateral processes and with a higher neural protuber-
ance; the posterior ventral process is produced sharply
backward. The axis of B. leidyi (Carnegie Mus. 114),
apart from its much smaller size, differs from that of
B. robustus chiefly in the far less backward prolonga-
tion of the neural spine, in the more vertical position
Figure 624. — Humeri of Brontops robustus and Brontotherium
leidyi
A, Brontops robustus, Yale Mus. 12048 (type). B, Brontotherium leidyi, Carnegie
Mus. 93, associated with skull. Ai, Bi, Left humerus, front view; A2, B2, left
humerus, outer side view. In Brontotherium leidyi the crest of the great tuberosity
is relatively larger and more widely expanded and the deltoid process is relatively
smaller. One-eighth natural size.
of the posterior zygapophysial facet and the less
transverse position of the anterior condylar facets for
the atlas.
Vertebrae referred to Brontotherium gigas, American Museum
This material is associated with a skull (see p. 570)
and a pelvis (see p. 692).
The third (?) cervical vertebra (B. gigas, Am. Mus.
492) has the centrum much shorter anteroposteriorly
(55 mm.) than in B. rohusfus (type) ; its neural arch is
very small and slender; the lateral transverse process
is more broadly expanded distally; and the zygapo-
physial facets, both anterior and posterior, appear to
face more vertically than in B. robustus. Accordingly
the neck of B. gigas appears to have been shorter and
deeper than that of B. robustus.
The third dorsal vertebra (B. gigas, Am. Mus. 492)
is much stouter than that of Brontops robustus (type).
Figure 625. — Humeri of Mcgaceroj.
acer? and Brontotherium gigas?
A, M.i acer?. Am. Mus. 0351; B, B. gigas, Am.
Mus. 1062. Distal views. One-eighth natural
size.
Its centrum is larger and deeper, the neural arch
longer, straighter, and wider. The swelling above the
facet for the tubercle of the rib is much larger.
FiGUHB 626. — Radii of Brontops robustus, Brontotherium leidyi,
and Brontotherium gigas
A, Brontops robustus, part of type skeleton, Yale Mus. 12048; B, Brontotherium
leidyi, Carnegie Mus. 93, associated with skull; C, Brontotherium gigas. Am.
Mus. 492, associated with skull. The two ends of the radius in Brontotherium
appear to be wider, and the external contour more deeply concave; the styloid
process is more acute. One-eighth natural size.
The second(?) lumbar vertebra (B. gigas, Am. Mus.
492) has a less elongate centrum (ap. 87 mm.) than
that of Brontops robustus (type) (96 mm.); the neural
spine is apparently longer and more vertical and the
postzygapophysial facet more oblique; the anterior
face of the centrum is more convex.
Fore limb referred to Brontotherium leidyi, Carnegie Museum
This fore limb is from the Chadron A levels in
which the species B. leidyi occurs.
The scapula of B. leidyi (Carnegie Mus. 93) is less
expanded transversely than that of B. gigas (Nat.
Mus. 4262). The generic differences from the scapula
692
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
of Brontops robustus are not yet apparent. The
humerus {B. leidyi, Carnegie Mus. 93) is a relatively
very small bone (mid-length 380 mm.) with an
enormously expanded crest of the great tuberosity.
FiQUBE 627. — Radius and ulna of
Brontotherium
K, B. leidyi, Carnegie Mus. 93, outer side view
of radius and ulna; B, B. gigas, Am. Mus. 492.
One-eightli natural size.
Its distal end is relatively narrower and higher than
in Brontops rohustus (type). The radius {B. leidyi,
Carnegie Mus. 93) is likewise very short (mid-length
is markedly different from that of B. rohustus (type; ;
the styloid process is produced farther downward;
the external contour is more sharply concave; the
internal contour is straighter; the middle of the head
is less angulate. The ulna both in B. leidyi (Carnegie
Mus. 93, length 410 mm.) and in B. gigas (Am. Mus.
492, length 642 mm.) has the sigmoid notch much
less deeply concave in side view than in Brontops
rohustus; the shaft of the ulna appears to be more
slender, especially at the lower end. In B. gigas the
tuberosity of the olecranon (for the main mass of
the triceps) is widely expanded transversely.
Manus of Brontotherium, U. S. National Museum
The manus of Brontotherium is known from the
lunar of B. gigas (Am. Mus. 492) and from a mixed
lot of specimens in the National Museum (No. 4262;
see p. 690).
This manus (fig. 631), which probably belongs with
the skull of B. gigas (Nat. Mus. 4262), is much smaller
Figure 628. — Ulnae of Brontops robustus, Brontotherium leidyi, and Bronto-
therium gigas
A, Brontops rohustus, Yale Mus. 12048 (type); B, Brontotherium leidyi, Carnegie Mus. 93; C,
Brontotlierium gigas. Am. Mus. 492. The proximal part of ttie ulna of Brontops is wider and tlie
olecranon is perhaps shorter than in Brontotlierium. One-eighth natural size.
B C
-Olecrana of Brontotherium and Mega-
cerops?
End view. A, B. gigas, Am. Mus. 492; B, B. leidyi, Carnegie Mus.
93; C, Megacerops? sp.. Am. Mus. 351 (6351?). The olecranon of
the largest animal is greatly swollen. One-eighth natural size.
than that referred to B. gigas (Am. Mus.
492), but it agrees in measurements and in
characters with another manus (Am. Mus.
1047) which is assigned provisionally to B.
Jiatcheri. Hence it seems probable that the
manus shown in Figure 631 represents a
female of either B. Jiatcheri or B. gigas. The
differences between the manus of Brontothe-
rium and that of Brontops rohustus are ex-
hibited in Figure 630. In general the manus
appears to be more compact and less spread-
ing, especially at the lower end, than that of
Brontops rohustus; the magnum is wider and
the lunar magnum articulation more extended;
there are also detailed differences in the
other carpal elements, but at present it is
not known whether any of these are constant
or how far they may be individual rather
than generic differences.
305 mm.). Even in B. gigas (Am. Mus. 492, mid-
length 460 mm.) the radius is relatively shorter than
in Brontops rohustus (type) (mid-length 500). In both
species the whole contour of the radius in front view
Pelves and hind limbs referred to B. gigas
Pelvis. — The pelvis of Brontotherium is known from
three specimens — a well-preserved pelvis of B. gigas
(Am. Mus. 492), associated with a skull and some
EVOLUTION OF THE SKELETON OE EOCENE AND OLIGOCENE TITANOTHEBES
693
Figure 630. — Manus of Oligocene titanotheres
A, Brontops robustus, Yale Mus. 12048 (type); B, Brontops robustusf, Am. Mus. 518; C, Diplodonus tyUri, Amherst Mus. 327 (type); D, Menodus
trigonoccTas, Am. Mus. 515; E, Menodus trigonoceras, Am. Mus. 1079; F, Menodus giganteus, Yale Mus., after Marsh; G, BTontotherium hatcheHt
Am. Mus. 1047 (two middle digits restored); H, Brontothenum gigasf, Am. Mus. 1443 (two middle digits restored); I, BTontotherium gigas, Am.
Mus. 492 (digits I, II restored). All one-eighth natural size. In general the manus of Menodus is high and narrow and has small distal
phalanges; that of Brontotherium is short and wide and has wide distal phalanges. The manus of Brontops appears to be of intermediate type.
694
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
other parts of the skeleton; a well-preserved pelvis
referred by Marsh to B. gigas and figured in Plate
CCXXXI (Nat. Mus., no number) ; and a pelvis in the
National Museum (No. 4262) probably associated with
skull and limbs referred to B. gigas.
The pelvis of B. gigas differs from that of Brontops
robustus in its greater relative width and in having
the outer iliac crest produced more downward and
less directly outward.
Femur. — The femur of B. leidyi (Carnegie Mus. 93)
lacks the proximal end, but the bone appears to be
relatively broader than in Brontops rohustus (type).
The characters of the bone in B. gigas are not known.
Tihia. — The tibia of B. leidyi (Carnegie Mus. 93) is
relatively wider at the proximal end than that of B.
roiustus.
Pes. — The pes of Brontotherium is supposed to be
represented by three metatarsals (PI. CCXXXII) and
FiGUKE 631. — Manus and pes referred to Brontotherium gigas hatcheri
Nat. Mus. 4262. One-fourth natural size.
Measurements of pelvis in Brontotherium gigas and Brontops
robustus, in millimeters
B . gigas.
Am. Mus.
492
Bronjops
robustus
Yale Mus.
12048 (type)
Width. -
1, 170
820
142
1,220
Length iliac crest to ischial tuberosity
Pelvic index, ^^ ^^ X 100
length
900
134
some phalanges (PI. CCXXXV), which Marsh referred
to B. gigas; by an incomplete pes (Am. Mus. 1047);
and by a partial pes which may be associated with the
skeleton of B. gigas (Nat. Mus. 4262). The pes, as a
whole, is extremely short and wide, the third meta-
tarsal measuring 160 millimeters in length and 76 in
maximum width near the distal end, as compared with
225 in length and 88 in width in B. rohustus (type).
The index of the third metatarsal is thus 47 in B.
gigas and 40 in B. roiustus
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
695
A2
Figure 632. — Manus and pes referred to Broniotherium hatcheri
Am. Mus. 1047. Ai, Manus, as restored. Most of digits .III and IV and part of the lunar are restored. A2, Pes, as restored. Much of the digits
and a part of the navicular are restored. One-fourth natural size.
Figure 633. — Manus referred to Broniotherium gigas, as restored
Am. Mus. 1443, reversed. Digits III and IV and the phalanges of digits II and V restored.
One-fourth natural size.
696
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Figure 636. — Tibiae of Brontops
robust us and Brontotherium
Figure 634. — Pelvis of Brontotherium gigas hatcheri
Nat. Mus. 42C2. Associated with skull and limbs. After Marsh. About one-eleventh natural size
A, Brontops robustits, Yale Mus. 12048
(type); B, BTontothenum leidyi, Carnegie
Mus. 93. One-eighth natural size.
Figure 637. — Tibia and fibula of
Brontotherium leidyi
A, Carnegie Mus. 93, outer side view or
right tibia. B, Carnegie Mus. 114, inner
side view of left fibula. One-eighth nat-
ural size.
Figure 635. — Femora of Brontops robuslus
and Brontotherium leidyi
A, Brontops robuslus, Yale Mus. 12048 (type); B, Bronto-
iherium leidyi, Carnegie Mus. 93. One-eighth natural
size. The Brontotherium femur appears to be wider in
proportion to its length.
Figure 638. — Femora of Megaceropsf and Brontotherium?
A, Megaceropsf acerf, Am. Mus. 6351. B, Brontotherium? sp.. Am. Mus. 6347. Prosimal view.
One-fourth natural size.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
697
Figure 639. — Pes of Oligocene titanotheres
A, Broniops rohustus, Yale Mus. 12048 (type); B, Diploclonus tyleri, Amherst Mus. 327 (type); C, Menoiusf heloceras, Am. Mus. 1080; D,
Menodus Irigonoceras, Am. Mus. 1079; E, BTontotherium hatcherif, Am. Mus. 1047; F, Brontotheriumf gigas, Yale Mus., after Marsh. One-
eighth natural size. The pes of Menodus is very narrow, that of Broniothenum is very short and broad; the others are of intermediate type .
698
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
BIBLIOGRAPHY FOR CHAPTERS III, IV, V, VI, VII
Abel, Othenio.
1910.1. Kritische Untersuchungen tiber die palaogenen
Rhinocerotiden Europas: K.-k. geol. Reiohs-
anstalt Abh., Band 20, Heft 3, pp. 1-42, pis. 1, 2.
BoCKH, JOHANN.
1876.1. Brachydiasiemalherium transilvanicum Bockh et
Maty, ein neues Pachydermen-Genus aus den
eocanen Schichten Siebenburgens: K. geol.
Anstalt Mitt, aus Jahrb., Band 4, pp. 125-150,
pis. 17-18.
Cope, Edward Drinker.
1872.1. Second notice of e.xtinct vertebrates from Bitter
Creek, Wye: Paleont. Bull. 7, Aug. 22, 1872
(part) ; Am. Philos. Soc. Proc, vol. 12, pp.
487-488 (in full).
1872.2. Second account of the new Vertebrata from the
Bridger Eocene: Paleont. Bull. 2, Aug. 3, 1872;
Am. Philos. Soc. Proc, vol. 12, pp. 466-468.
1873.1. On some Eocene mammals obtained by Hayden's
geological survey of 1872: Paleont. Bull. 12,
Mar. 8, 1873.
1873.2. Second notice of e.xtinct Vertebrata from the Ter-
tiary of the Plains: Paleont. Bull. 15, Aug. 20,
1873.
Mioiasileus, type M. ophryas.
1873.3. Synopsis of new Vertebrata from the Tertiary of
Colorado, obtained during the summer of 1873,
Washington, October, 1873. (Said, on title
page, to be an e.xtract from U. S. Geol. and Geog.
Survey Terr. Seventh Ann. Rept., but never
published in tliat volume.)
1873.5. On the new perissodactyls from the Bridger
Eocene: Paleont. Bull. 11, Jan. 31, 1873;
Am. Philos. Soc. Proc, vol. 13, pp. 35, 36
(condensed).
1873.6. On the extinct Vertebrata of the Eocene of Wyo-
ming observed by the expedition of 1872, with
notes on the geology: U. S. Geol. and Geog.
Survey Terr. Sixth Ann. Rept., 1873, pp.
545-649, pis. 1-6.
1874.1. On some extinct types of horned perissodactyls:
Am. Assoc. Adv. Sci. Proc, 22d meeting, 1873,
pp. B 108-B 109.
Symborodon torvus, Miobasileus ophTtjas, Symborodon acer,
SymbOTOdon heloceras.
1874.2. Report on the vertebrate paleontology of Colo-
rado: U. S. Geol. and Geog. Survey Terr. Ann.
Rept. for 1873, pp. 427-533, pis. 1-8.
1878.1. On some characters of the Miocene fauna of Ore-
gon: Paleont. Bull. 30, Dec 3, 1878; Am.
Philos. Soc. Proc, vol. 18, pp. 63-78.
1879.1. On the extinct species of Rhinoceridae of North
America and their allies: U. S. Geol. and Geog.
Survey Terr. BuU., vol. 5, pp. 227-237; Am.
Naturalist, vol. 13, pp. 771a-771j (slightly
different form, with 8 cuts).
Chalicotheriidae Cope.
1880.1. The Badlands of the Wind River and their fauna:
Am. Naturalist, vol. 14, No. 10, pp. 745-748.
Lambdotherium popoagkum Cope.
1881.1. The systematic arrangement of the order Peris-
sodactyla: Am. Philos. Soc. Proc, vol. 19, pp.
377-401, May 14, 1881.
Menodontidae Cope.
Cope, Edward Drinker — Continued.
1881.2. On the Vertebrata of the Wind River Eocene beds
of Wyoming: U. S. Geol. and Geog. Survey
Terr. BuU., vol. 6, pp. 183-202.
1885.1. The Vertebrata of the Tertiary formations of the
West, Book I: U. S. Geol. Survey Terr. Rept.,
vol. 3, XXXV-M009 pp., pis. l-75a.
1886.1. The Vertebrata of the Swift Current Creek region
of the Cypress HiUs: Canada Geol. and Nat.
Hist. Survey Ann. Rept., new ser., vol. 1,
Rept. C, Appendix, pp. 79-85.
1887.1. The Perissodactyla: Am. Naturalist, vol. 21,
pp. 985-1007, November, 1887; pp. 1060-1076,
December, 1887.
1889.1. The Vertebrata of the Swift Current River, II:
Am. Naturalist, vol. 23, pp. 151-155, March,
1889.
Genus Haplacodon established for Menodus angustigenis.
1889.2. Vertebrata of Swift Current River, III: Am.
Naturalist, vol. 23, pp. 628-629, July, 1889.
1889.3. The mechanical causes of the development of the
hard parts of the Mammalia: Jour. Morphology,
vol. 3, pp. 137-277, pis. 9-19.
1891.1. On two new perissodactyls from the White River
Neocene of Nebraska: Am. Naturalist, vol. 25,
pp. 47-49.
Menodus peltoceTos Cope.
1891.2. On Vertebrata from the Tertiary and Cretaceous
rocks of the Northwest Territory, I, The species
from the Oligocene or Igwer Miocene beds of the
Cypress Hills: Canada Geol. Survey Contr.
Canadian Paleontology', vol. 3, pp. 1-25, pis. 1-14.
Cotter, G. de P.
1916.1. See Pilgrim, Guy E., 1916.1.
Doderlein, Ltjdwiq.
1890.1. See Steinmann, Gustav, 1890.1.
Douglass, Earl.
1909.1. Preliminary descriptions of some new titanotheres
from the Uinta deposits: Carnegie Mus. Annals,
vol. 6, pp. 304-313.
Earle, Charles.
1891.1. On a new species of Palaeosyops, Palaeosyops
megarhinus sp. nov.: Am. Naturalist, vol. 25,
pp. 45-47.
1891.2. Palaeosyops and allied genera: Acad. Nat. Sci.
Philadelphia Proc, vol. 43, pp. 106-117.
1892.1. A memoir upon the genus Palaeosyops Leidy and
its allies: Acad. Nat. Sci. Philadelphia Jour., 2d
ser., vol. 9, pp. 267-388, pis. 10-14.
Evans, John.
1850.1. See Owen, David Dale, Norwood, J. G., and
Evans, John, 1850.1.
EwART, J. Cossar.
1907.1. On skulls of horses from the Roman fort at
Newstead, near Melrose, with observations on
the origin of domestic horses: Roy. Soc. Edin-
burgh Trans., vol. 45, pt. 3, pp. 555-587, pis. 1-3.
Flower, Sir William Henry.
1876.1. On the relation of extinct to existing Mammalia,
with special reference to the derivative hypoth-
esis: Nature, vol. 13, pp. 307-308,327-328,350-
352, 387-388, 409-410, 449-450, 487-488, 513-
514; vol. 14, p. 11.
1885.1. An introduction to the osteology of the Mammalia,
3d ed., revised with the assistance of Hans
Gadow, London, Macmillan & Co.
1891.1 (and Lydekker, Richard). An introduction to the
study of mammals living and extinct, London.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
699
Granger, Walter.
1908.1. A revision of the American Eocene horses: Am.
Mus. Nat. Hist. Bull., vol. 24, pp. 221-264,
pis. 15-18.
1911.1. See Sinclair, William J., and Granger, Walter,
1911.1.
1912.1. See Sinclair, WiUiam J., and Granger, Walter,
1912.1.
Gregory, William King.
1912.1. Note on the upper Eocene titanotheroid Tel-
matheriumf incisivum Douglass, from the
Uinta Basin: Science, new ser., vol. 35, No.
901, p. 545.
1912.2. Notes on the principles of quadrupedal locomo-
tion and on the mechanism of the limbs in
hoofed animals: New York Acad. Sci. Annals,
vol. 22, pp. 267-294, pi. 34.
1916.1. Studies on the evolution of the primates. Part 1,
The Cope-Osborn "theory of trituberculy "
and the ancestral molar patterns of the pri-
mates: Am. Mus. Nat. Hist. Bull., vol. 35, pp.
239-257, pi. 1.
Hatcher, John Bell.
1893.1. The Titanotherium beds: Am. Naturalist, vol. 27,
pp. 204-221.
1895.1. On a new species of Diplacodon, with a discussion
of the relations of that genus to Telmatotherium:
Am. Naturalist, vol. 29, pp. 1084r-1090, pis. 28-40.
1901.1. On the cranial elements and the deciduous and per-
manent dentition of Titanotherium: Carnegie
Mus. Annals, vol. 1, pp. 256-261, pis. 7-8.
1902.1. A mounted skeleton of Titanotherium dispar Marsh:
Carnegie Mus. Annals, vol. 1, pp. 347-355, pis.
14^18.
Hay, Oliver Perry.
1899.1. Notes on the nomenclature of some North Ameri-
can vertebrates: Science, new ser., vol. 10, No.
243, pp. 253-254, Aug. 25, 1899.
1902.1. Bibliography and catalogue of the fossil Vertebrata
of North America: U. S. Geol. Survey Bull.
179, ii + 868 pp.
Hills, Robert C.
1888.1. The recently discovered Tertiary beds of the
Huerfano River basin, Colo.: Colorado Sci. Soc.
Proc, vol. 3, pp. 148-164, map.
Kiehnik, E.
1913.1. O nowym gatunku Titanotherium, ein neuer Titano-
theriumfund in Europa: Internat. Acad. Sci.
Cracovie Bull., ser. B, No. 10, pp. 1211-1225,
pi. 13.
Lambe, Lawrence M.
1908.1. The Vertebrata of the Oligocene of the Cypress
Hills, Saskatchewan: Canada Geol. Survey Contr.
Canadian Paleontology, vol. 3, pt. 4.
Lankester, Sir E. Ray.
1902.1. On Okapia, a new genus of GirafEdae, from central
Africa: Zool. Soc. London Trans., vol. 16, pt. 4,
pp. 279-314, pis. 30-32.
Leidy, Joseph.
1852.1. Description of the remains of extinct Mammalia
and Chelonia from Nebraska Territory, col-
lected during the geological survey under the
direction of Dr. D. D. Owen; in Owen, David
Dale, Report of a geological survey of Wiscon-
sin, Iowa, and Minnesota and incidentally a
portion of Nebraska Territory, pp. 553-572,
pis. 9-15, Philadelphia.
Prout's specimen
Leidy, Joseph — Continued.
1852.2. On a new species of rhinoceros from Nebraska,
Rhinoceros americanus: Acad. Nat. Sci. Phila-
delphia Proc, vol. 6, p. 2.
1853.1. On a collection of fossil Mammalia and Chelonia
from the Mauvaises Terres of Nebraska: Acad.
Nat. Sci. Philadelphia Proc, vol. 6, pp. 392-
394.
1854.1. The ancient fauna of Nebraska, or a description of
remains of extinct Mammalia and Chelonia from
the Mauvaises Terres of Nebraska: Smith-
sonian Contr. Knowl., vol. 6, art. 7, pp. 1-126,
pis. 1-24. (This article published in June, 1853.)
1854.2. Synopsis of extinct Mammalia the remains of which
have been discovered in the Eocene formations
of Nebraska: Acad. Nat. Sci. Philadelphia Proc,
vol. 7, pp. 156-157.
1869.1. The extinct mamalian fauna of Dakota and Ne-
braska, including an account of some allied forms
from other localities, together with a synopsis of
the mammalian remains of North America: Acad.
Nat. Sci. Philadelphia Jour., 2d ser., vol. 7, pp.
1-472, pis. 1-30.
1870.1. Remarks on Megacerops coloradensis; Acad. Nat.
Sci. Philadelphia Proc, vol. 22, pp. 1-2.
1870.2. On fossils from Church Buttes, Wyoming Terr.:
Acad. Nat. Sci. Philadelphia Proc, vol. 22, pp.
113-114.
Palaeosyops paluiosus, Microsus cuspidatus, Notharclos
tenebTosus.
1871.1. Remarks on fossil vertebrates from Wyoming:
Acad. Nat. Sci. Philadelphia Proc, vol. 23, pp.
228-229.
1872.1. On some new species of Mammalia from Wyo-
ming: Acad. Nat. Sci. Philadelphia Proc, vol.
24, pp. 167-169.
Palaeosyops Urimilis,
1872.3. On fossils from Wyoming: Acad. Nat. Sci. Phila-
delphia Proc, vol. 24, p. 277.
Palaeosyops Junius Leidy.
1873.1. Contributions to the extinct vertebrate fauna of
the Western Territories: U. S. Geol. Survey
Terr. Rept., vol. 1, pp. 14-358, pis. 1-37.
Palaeosyops Junius Leidy.
Loomis, Frederick B.
1907.1. Origin of the Wasatch deposits: Am. Jour. Sci.,
4th ser., vol. 23, pp. 356-364.
Lambdotherium primaevum.
Lull, Richard Swann.
1905.1. Megacerops tyleri, a new species of titanothere
from the Badlands of South Dakota: Jour.
Geology, vol. 13, pp. 443-456, pis. 3-4.
Lydekker, Richard.
1889.1. See Nicholson, Henry A., 1889.1.
1891.1. See Flower, Sir William Henry, 1891.1.
Marsh, Othniel Charles.
1870.1. Professor Marsh's Rocky Mountain e.xpedition,
discovery of the Mauvaises Terres formation
in Colorado (letter to J. D. Dana dated Aug.
12, 1870): Am. Jour. Sci., 2d ser., vol. 50, p.
292.
1871.1. Notice of some new fossil mammals from the
Tertiary formations: Am. Jour. Sci., 3d ser.,
vol. 2, pp. 35-44.
Palaeosyops minor Marsh
700
TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA
Marsh, Othniel Charles — Continued.
1871.2. Notice of some new fossil mammals and birds
from the Tertiary formation of the West:
Am. Jour. Sci., 3d ser., vol. 2, pp. 120-127.
Canis montanus.
1872.1. Preliminary description of new Tertiarym ammals,
Part I: Am. Jour. Sci., 3d ser., vol. 4, pp.
122-128; erratum, p. 504.
Palaeosyops lalkeps Marsh.
1873.1. Notice of new Tertiary mammals: Am. Jour.
Sci., 3d ser., vol. 5, pp. 407-410, 485-488.
Brontotheridae.
1874.1. On the structure and affinities of the Bronto-
theridae: Am. Jour. Sci., 3d ser., vol. 7, pp.
81-86, pis. 1-2.
1875.1. Notice of new Tertiary mammals, IV: Am.
Jour. Sci., 3d ser., vol. 9, pp. 239-250.
1876.1. Principal characters of the Brontotheridae: Am.
Jour. Sci., 3d ser. vol. 11, pp. 335-340, pis. 10-13.
1877.1. Introduction and succession of vertebrate life
in America (vice president's address. Am.
Assoc. Adv. Sci., 1877): Am. Jour. Sci., 3d
ser., vol. 14, pp. 337-378.
1880.1. List of genera established by Prof. O. C. Marsh,
1862-1879, 12 pp.
1887.1. Notice of new fossil mammals: Am. Jour. Sci.,
3d ser., vol. 34, pp. 323-331.
1889.1. Restoration of Brontops robustus, from the
Miocene of America: Am. Jour. Sci., 3d ser.,
vol. 37, pp. 163-165, pi. 6.
1890.1. Notice of new Tertiary Mammalia: Am. Jour.
Sci., 3d ser., vol. 39, pp. 523-525.
1891.1. Notice of new vertebrate fossils: Am. Jour. Sci.,
3d ser., vol. 42, pp. 265-269.
Allops crasskoTnis, Brontops dispar, BrOTitotJierium medium.
Matthew, William Diller.
1897.1. Development of the foot in the Palaeosyopinae:
Am. Naturalist, vol. 31, pp. 57-58.
1899.1. A provisional classification of the fresh-water
Tertiary of the West: Am. Mus. Nat. Hist.
BuU., vol. 12, pp. 19-75.
1909.2. See Osborn, Henry Fairfield, 1909. 321.
Merrill, George P.
1906.1. Contributions to the history of American geol-
ogy: U. S. Nat. Mus. Rept., for 1904, pp. 189-
734, pis. 1-37.
Nicholson, Heney Alltne.
1889.1 (and Lydekker, Richard). A manual of palaeon-
tology, London, Blackwood.
Norwood, J. G.
1850.1 (with Owen, D. D., and Evans, John). See
Owen, David Dale, Norwood, J. G., and Evans,
John, 1850. 1.
OsBORN, Henry Fairfield.
1878.3 (with Scott, W. B., and Speir, Francis, jr.).
Palaeontological report of the Princeton scien-
tific expedition of 1877: E. M. Mus. Geol. and
Archeol. Princeton Coll. Contr., No. 1, pp. 1-
106, pis. A, 1-10.
1887.30. See Scott, William B., and Osborn, H. F., 1887.1.
1890.50. Prehminary account of the fossil mammals from
the White River and Loup Fork formations,
contained in the Museum of Comparative Zoo-
logy, Part II, The Perissodactyla: Mus. Comp.
Zoology Bull., vol. 20, pp. 87-100, pis. 2-3.
1890.51 (and Scott, W. B.). The Mammaha of the
Uinta formation, pt. 3, The Perissodactyla;
pt. 4, The evolution of the ungulate foot: Am.
Philos. Soc. Trans., new ser., vol. 16, pt. 3, pp.
541-569, pis. 7-9. (For entire article, see
Scott, William B., and Osborn, H. F., 1890.1.)
OsBORN, Henry Fairfield — Continued.
1892.67 (and Wortman, J. L.). Fossil mammals of the
Wasatch and Wind River beds, collection of
1891: Am. Mus. Nat. Hist. Bull., vol. 4, pp.
81-147.
1895.98. Fossil mammals of the Uinta Basin, expedition
of 1894 (geologic levels by O. A. Peterson) :
Am. Mus. Nat. Hist. Bull., vol. 7, pp. 71-105.
1896.110. The cranial evolution of Titanotherium: Am.
Mus. Nat. Hist. Bull., vol. 8, pp. 157-197.
1897.126. The Huerfano lake basin, southern Colorado,
and its Wind River and Bridger fauna: Am.
Mus. Nat. Hist. BuU., vol. 9, pp. 247-258.
1898.143. The extinct rhinoceroses: Am. Mus. Nat. Hist.
Mem., vol. 1, pt. 3, pp. 75-164, pis. 12a-20.
1900.192. Phylogeny of the rhinoceroses of Europe (Rhi-
noceros Contributions No. 5) : Am. Mus. Nat.
Hist. Bull., vol. 13, pp. 229-267.
1902.207. Dolichocephaly and brachycephaly in the lower
mammals: Am. Mus. Nat. Hist. Bull., vol. 16,
pp. 77-89.
1902.208. The four phyla of Oligocene titanotheres : Am.
Mus. Nat. Hist. BuU., vol. 16, pp. 91-109.
1907.294. Tertiary mammal horizons of North America:
Am. Mus. Nat. Hist. Bull., vol. 23, pp. 237-253.
1907.301. Evolution of mammalian molar teeth to and
from the triangular type (edited by Dr. W. K.
Gregory), 250 pp., New York, MacmUlan.
1908.318. New or httle known titanotheres from the
Eocene and Oligocene: Am. Mus. Nat. Hist.
BuU., vol. 24, pp. 599-617.
1909.321 (and Matthew, W. D.). Cenozoic mammal
horizons of western North America, with appen-
dix, Faunal lists of the Tertiary Mammalia of
the West, by WiUiam Diller Matthew: U. S.
Geol. Survey BuU. 361, 138 pp.
1912.368. SkuU measurements in man and the hoofed
mammals: Science, new ser., vol. 35, No. 902,
pp. 595-596, Apr. 12, 1912; New York Acad.
Sci. Annals, vol. 22, pp. 341-342, 1913.
1912.372. The continuous origin of certain unit charac-
ters as observed by a palaeontologist (Harvey
lecture): Am. Naturalist, vol. 46, pp. 185-206;
249-278.
1912.382. Craniometry of the Equidae: Am. Mus. Nat.
Hist. Mem., new ser., vol. 1, pt. 3, pp. 57-100.
1913.400. Lower Eocene titanotheres, genera Lambdo-
therium, Eotitanops: Am. Mus. Nat. Hist. Bull.,
vol. 32, pp. 407-415.
1913.401. The sku]l of Bathyopsis, Wind River uintathere:
Am. Mus. Nat. Hist. BuU., vol. 32, pp. 417-
420, pis. 64^66.
1914.409. Recent results in the phylogeny of the titano-
theres: Geol. Soc. America BuU., vol. 25, pp.
403-405.
1914.416. Origin of single characters as observed in fossil
and living animals and plants: Am. Naturalist,
vol. 49, pp. 19.3-239.
1916.433. Tv/o new Oligocene titanotheres: Am. Mus.
Nat. Hist. BuU., vol. 35, pp. 721-723.
1918.473. Equidae of the OUgocene, Miocene, and PUo-
cene of North America, iconographic type
revision: Am. Mus. Nat. Hist. Mem., new ser.,
vol. 2, pt. 1, pp. 1-217, pis. 1-54.
1919.494. New titanotheres of the Huerfano: Am. Mus.
Nat. Hist. BuU., vol. 41, pp. 557-569.
Owen, David Dale.
1850.1 (with Norwood, J. G., and Evans, John). Notice
of fossil remains brought by Mr. J. Evans from
the Mauvais Terres or badlands of White
River, 150 miles west of the Missouri: Acad.
Nat. Sci. Philadelphia Proc, vol. 5, p. 66.
EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHERES
701
Owen, David Dale — Continued.
1852. L Report of a geological survey of Wisconsin, Iowa,
and Minnesota, and incidentally of a portion of
Nebraslca Territory, made under instructions
from the United States Treasury Department,
pp. v--xxxviii, 41-638, pis. 1-19, maps, geologic
sections, Philadelphia, Lippincott, Grambo
& Co.
Palmer, T. S.
1904.1. Index generum mammalium — A list of the genera
and families of mammals: U. S. Dept. Agr.
North Am. Fauna 23.
Peterson, O. A.
1909.1. Revision of the Entelodontidae: Carnegie Mus.
Mem., vol. 4, pp. 41-156, pis. 54-62.
1914.1. A new titanothere from the Uinta Eocene: Car-
negie Mus. Annals, vol. 9, pp. 29-52.
1914.2. A small titanothere from the lower Uinta beds:
Carnegie Mus. Annals, vol. 9, pp. 53-57.
1914.3. Some undescribed remains of the Uinta titano
there Dolichorhinus: Carnegie Mus. Annals
vol. 9, pp. 129-138.
1914.4. A correction of generic name [Eotitanolherium to
replace Diploceras]: Carnegie Mus. Annals, vol.
9, p. 220.
Pilgrim, Gut E.
1916.1 (and Cotter, G. de P.). Some newly discovered
Eocene mammals from Burma: India Geol.
Survey Records, vol. 47, pp. 42-78, pis. 1-6.
1925.1. The Perissodactyla of the Eocene of Burma: India
Geol. Survey Mem., new ser., vol. 8, No. 3,
pp. 1-14, pis. 1-2.
Titanotheriidae: Skatitanops, n. gen., Pondaung horizon-
Pakokku district, Burma; Sivatitanops cotteri Pilgrim, n. sp.;
Siratitanopa birmankum Pilgrim and Cotter; Sivatitanopsf
Tugosidena Pilgrim, n. sp.; Eotitanolherium Peterson; Eotitano-
theriumf lahirii Pilgrim, n. sp.
POMEL, A.
1849.1. Description d'un os maxillaire fossile de Palaeoihe-
rium par Hiram Prout, Am. Jour. Sci. and Arts
by Silliman's [sic] and J. Dana, 2" s6rie, vol. 3,
No. 8, p. 248: Bibhotheque univ. Gen&ve
(Suppl.), Arch. sci. phys. nat., vol. 10, pp. 73-75.
Pkout, Hieam a.
1846.1. Gigantic Palaeotherium: Am. Jour. Sci., 2d ser.,
vol. 2, pp. 288-289.
1847.1. Description of a fossil maxiUary bone of a Palaeo-
therium from near White River: Am. Jour.
Sci., 2d ser., vol. 3, pp. 240-250.
1860.1. On a tooth found in Virginia: Acad. Sci. St.
Louis Trans., vol. 1, pp. 699-700.
Leidyotherium.
RiGGS, E. S.
1912.1. New or little known titanotheres from the lower
Uintah formations: Field Mus. Pub. 159, Geol.
ser., vol. 4, No. 2, pp. 17-41, pis. 4r-12.
Mesatirkinua superior Riggs; Metarhinus riparius Biggs;
Metarhinus cristatua Eiggs; Dolkhorhinua fiuminalis Riggs;
RhadinoThinua Riggs.
RiJTIMEYER, L.
1882.1. Studien zu der Geschichte der HirschfamiUe, 1,
Schadelbau: Naturforschende Gesell. Basel
Verh., Band 7, pp. 3-61.
ScHLossER, Max.
1901.1. Zur Kenntnis der Saugethierfauna der bohm.
Braunkohlenformation: Deutsch. naturwiss.-
medicinisch. Verein f. Bohmen "Lotos" Abh.,
Band 2, Heft 3, pp. 1-43, pi. 5.
PTohyTacodon.
ScHtrcHERT, Charles.
1905.1. Catalogue of the type specimens of fossil inverte-
brates in the department of geology. United
States National Museum, Introduction: U. S.
Nat. Mus. Bull. 53, pt. 1.
Scott, William B.
1878.3. See Osborn, Henry Fairfield, and Scott, W. B.,
1878.3.
1887.1 (and Osborn, H. F.). Preliminary account of the
fossil mammals from the WTiite River forma-
tion contained in the Museum of Comparative
Zoology: Mus. Comp. Zool. Bull., vol. 13,
No. 5, pp. 151-171, pis. 1, 2.
1890.1 (and Osborn, H. F.). The Mammalia of the Uinta
formation: Part I, The geological and faunal
relations of the Uinta formation, by WiUiam B.
Scott; Part II, The Creodonta, Rodentia, and
Artiodactyla, by William B. Scott; Part III,
The Perissodactyla, by Henry Fairfield Osborn;
Part IV, The evolution of the ungulate foot,
by Henry Fairfield Osborn (see Osborn,
1890.51): Am. Philos. Soc. Trans., new ser.,
vol. 16, pt. 3, pp. 461-572, pis. 7-9.
1892.1. The evolution of the premolar teeth in the mam-
mals: Acad. Nat. Sci. Philadelphia Proc, vol. 44,
pp. 405-444.
Sinclair, William J.
1911.1 (and Granger, Walter). Eocene and Oligocene of
the Wind River and Big Horn Basins: Am. Mus.
Nat. Hist. BuU., vol. 30, pp. 83-117.
1912.1 (and Granger, Walter). Notes on the Tertiary
deposits of the Big Horn Basin: Am. Mus.
Nat. Hist. Bull., vol. 31, pp. 57-67.
Spier, Francis, jr.
1878.1. See Osborn, Henry Fairfield, Scott, W. B., and
Spier, Francis, jr., 1878.3.
Stehlin, H. G.
1903.1. Die Siiugetiere des schweizerischen Eocaens,
Critischer Catalog der Materialen, Teil 1, 2, 3:
Schweizer paljiont. Gesell. Abh., vol. 30, pp.
1-153, pis. 1-3, 1903; vol. 31, pp. 1.54^445, pis.
4-9, 1904; vol. 32, pp. 447-595, pis. 10, 11, 1905.
Steinmann, Gust.av.
1890.1 (and Doderlein, Ludwig). Elemente der Palaon-
tologie, 848 pp., Leipzig.
Palaeosyopinae (p. 777)
Thomas, Oldfield.
1893.1. Suggestions for the more definite use of the word
"type" and its compounds, as denoting speci-
mens of a greater or less degree of authenticity:
Zool. Soc. London Proc. for 1893, pp. 241-242.
Toula, Franz.
1892.1. Zwei neue Saugethierfundorte auf der Balkan-
halbinsel: Akad. Wiss. Wien Sitzungsber.,
Band 101, Abt. 1, pp. 608-615, 1 pi.
Menodus rumelicus.
1896.1. Ueber einen neuen Rest von Lepiodon (?) (Titano-
therium'i) rumelicus Toula spec: Deutsch. geol.
GeseU. Zeitschr., Band 48, pp. 922-924.
Wohtman, Jacob L.
1892.1. See Osborn, Henry Fairfield, 1892.67.
von Zittel, Karl Alfred.
1893.1. Handbuch der Palaeontologie, 1. Abt., Palaeozoo-
logie, Band 4, Mammalia, Unter Mitwirkung
von W. Ph. Schimper und A. Schenk, Olden-
bourg, Miinchen, und Leipzig.
n. S. GEOLOGICAL SURVEY
igj I
MONOGRAPH 55 PLATE XXVI
Ai
TYPE SKELETON OF EOTITANOPS PRINCEPS
Am. Mu8. 296; Wind River formation. Wind River Basin. Ai, Lower jaw and dentition, one-half natural si?e; Aj, left lower
teeth, three-fourths natural size; B, cervical, C, dorsal, and D, caudal vertebrae, aU three-fourths natural size; E, right
humerus, three-eighths natural size; F, right manus, one-half natural size; G, right femur, three-eighths natural size
101959— 29— VOL 1 48
B i
0 1
<5 -S
0 ^
X 3
d ii
*\
"'^•SUiifc
'Aa
'-^^2
Z o
£ s
Bi -
0 m
o S
o 3
Q h
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XXXlll
A. BRONTOPS DISPAR, FEMALE
In Carnegie Museum. The skull belongs to another individual. Slightly i
of Dr. W. J. Holland
than one-twentieth natural si2;e. Courtesy
■
1
IK
^
n
V
■
1
■
M
i
^v\\
5^'
Vi
/H
^1
1
^1
mf/
}
W/t
^
^
H
1
^1
BP^Hi
Yim
II
i
HH^^)^
T^-"*^^
!
9
^^^^;
m
t
1
H
ll
1
■
QM^H
H
1
1
H
^^^^^^w^
^^1
''JsaBtJ'**
1
mi
*
i^B
^
gij
■■
^^^^
B. BRONTOPS ROBUSTUS, TYPE
Yale Mas. 12043. Oblique side view. About one twenty-fourth natural size. Maximum height 8 feet 21/2 inches to top of
dorsal spine. Courtesy of Prof . R. S. Lull. (Compare PI. XXXIV)
MOUNTED SKELETONS OF BRONTOPS
U. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XXXIV
1
Ilik mm
P^K
^91
^\
■;%
»i If
WL Ij^i
1*
^^m ^^''
i'^
T M
^
■;. _rfj«a^:-v --^
;lliipF
MOUNTED SKELETON OF BRONTOPS ROBUSTUS, TYPE
Yale Mus. 12048. Courtesy of Prof. R. S. Lull. A. Oblique front view; B, side view. (Ckimpare PI. XXXIII, B)
U. 8. GEOLOGICAL 6DRVEY
MONOGRAPH 55 PLATE XXXVI
VERTEBRAL COLUMN OF BRONTOPS
, Brontopi rohuflus? {Am. Mu8. 518). The c
B, Brontops robuilus, type (Yale Mub. 12048),
ence to represent the vertebrae as if they wei
This figure was obtai
led by copying the carefully
(101959—29)
XT. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XXXVII
MANUS AND HIND LIMB OF DIPLOCLONUS TYLERI LULL
Amher^ Mus, 327, type. (After Lull.) Ai, Right manus; As^ proximal aspedt of di^al carpals; A3, proximal aspei5t of proximal
carpals; all one-fourth natural size. B, Right hind limb, one-eighth natural size
s a
O B
2 E
Pi a
^ i
Q a
0 a
2 3
TT. S. GEOLOGICAL SURVEY
MONOGRAPH 55 PLATE XLI
MOUNTED SKELETON OF BRONTOTHERIUM HATCHERI, FRONT VIEW
Nat. Mu8. 4252. (Compare Pis. XXXIX, XL, and XLII)
n. S. GBOLOaiCAL SURVEY
MONOGRAPH 55 PLATE XLII
MOUNTED SKELETON OF BRONTOTHERIUM HATCHERI, BACK VIEW
Nat. Mus. 4262. (Compare Pis. XXXIX, XL, and XLI)
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11