(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "The titanotheres of ancient Wyoming, Dakota, and Nebraska"

LIBRARY 
Connecticut Agricultural College 



Vol. 



3^^^n 



Y.% 



Class No. /T ^ O (JaV 



Cost 






Date 



U, S. C^itU- 



:Jf :ld. i93f 







— -,_g 




the 

university of 

Connecticut 

ibraries 




^^ 






NOT -t» «» TAKEN 



Digitized by the Internet Archive 

in 2009 with funding from 

Boston Library Consortium IVIember Libraries 



http://www.archive.org/details/titanotheresofan02osbo 



Department of the Interior 

Ray Lyman Wilbur, Secretary 



U. S. GEOLOGICAL SURVEY 
George Otis Smith, Director 









Monograph 55 



THE TITANOTHERES OF ANCIENT WYOMING, 
DAKOTA, AND NEBRASKA 



BY 



HENRY FAIRFIELD OSBORN 



VOLUME 2 




DOCUMENTS DEPARTMENT 
RECE/VED 

JAM 2 m? 

Wilbui Cross Library 
Unzversity of Ccnnecticut 



UNITED STATES 

GOVERNMENT PRINTING OFFICE 

WASHINGTON : 1929 



Note. — Monograph 55 is issued in two volumes. Volume 1 contains Chap- 
ters I-VII and Plates I-XLII; volume 2 contains Chapters VIII-XI, Plates 
XLIII-CCXXXVI, an appendix, and the index to both volumes. 



X^^^d 



CONTENTS 



Page 

Chapter VIII. The muscular anatomy and the restoration of the titanotheres, by W. K. Gregorj' 703 

Section 1. Muscles of the face and jaw 703 

Section 2. Restorations of the heads of titanotheres 706 

Section 3. Muscles of the neck and back 709 

Section 4. Muscles of the limbs of the Eocene titanothere Palaeosyops (subgraviportal type) 714 

Section 5. Muscles of the limbs and vertebrae of Oligocene titanotheres 725 

Section 6. Restoration of the musculature and body form of Broniops robusius, an Oligocene titanothere 725 

Chapter IX. Mechanics of locomotion in the evolution of limb structure as bearing on the form and habits of the titano- 
theres and the related odd-toed ungulates 727 

Section 1. Adaptation to locomotion in the limbs of the fleet (cursorial) and the ponderous (graviportal) types of 

titanotheres and other hoofed quadrupeds 727 

Mechanical and physiological principles governing the proportions and angulation of limb segments in graviportal 

ungulates 727 

Researches made and principles established 727 

Principles of leverage and muscular action, by W. K. Gregory 727 

Straightening of the limbs and arches in adaptation to great weight (graviportal type) 731 

Indices and ratios of limb segments in cursorial and graviportal ungulates (aUometric adaptations) 733 

Results of comparative studies 733 

Ratios showing evolution from primitive to graviportal and cursorial types 735 

Primitive ungulate types 735 

Phyletic progression (Eocene) to graviportal type 736 

Primitive cursorial condylarths and perissodactyls 736 

Mediportal or intermediate group 736 

Phyletic progression to cursorial type in the Equidae 737 

Cursorial light-bodied artiodactyls 737 

Subcursorial deep-bodied artiodactyls 738 

Cursorial heavy-bodied artiodactyls 738 

Graviportal short-limbed digitigrades 738 

Graviportal long-limbed digitigrades 738 

Graviportal rectigrades 739 

Summary of cursorial and graviportal proportions of segments of limbs of ungulates 739 

Features considered 739 

Forms of scapula 740 

Forms of humerus 741 

Analogous adaptation in humerus and femur 742 

Forms of ilium 743 

Length of ilium and of ischium and ratio of these lengths to total length of os innominatum 746 

Ratios of length of scapula and ihum to that of humerus and femur, respectively 746 

Section 2. Systematic comparison of the pectoral and pelvic arches and of the limb bones in eight famihes of peris- 
sodactyls 747 

Tapiridae and Lophiodontidae 747 

Palaeotheriidae 747 

Rhinocerotidae, Hy racodontidae, Amynodontidae 750 

Equidae '^53 

Brontotheriidae — the titanotheres 754 

Chapter X. Theories as to the origin, ancestry, and adaptive radiation of the titanotheres and other odd-toed ungulates.. 757 

Section 1. Origin and relationships of the Perissodactyla 757 

Nature and habitat of the ancestral perissodactyl 757 

Hypothetical origin of the Perissodactyla from the Condylarthra 758 

Eocene Condylarthra nonperissodactyl 759 

Principal characters of the ancestral perissodactyls 760 

Tetradactyl manus of primitive Perissodactyla 762 

The skull of primitive Perissodact3da 764 

Section 2. Origin and phyletic radiations of the titanotheres and other Perissodactyla 768 

Skull of the primitive titanothere 768 

Structure of the foot in the titanotheres and in other perissodactyls 769 

Mechanics of the perissodactyl manus 774 

Summary of the evolution of the perissodactyl families 776 

Convergence in habitat and habit 778 

Descent of the bunoselenodont families 779 

Descent of the lophodont families 779 

Survival and extinction of the Perissodactyla 780 

Phyletic branching of the titanotheres ■ 780 



IV CONTENTS 

Chapter X — Continued. Tage 

Section 3. Summarj' of the cranial and skeletal evolution of the titanotheres 780 

General conclusions reached 780 

Disharmonic evolution in length and breadth of skull 783 

Evolution of the skull in correlation (coadaptation) with that of the teeth and horns 784 

Are the proportions of skull and teeth adaptive? 786 

Radiation and divergence in Eocene skuUs 786 

Evolution of the skull in Eocene time partly prophetic of that in Oligocene time 786 

Independent evolution of the skuU in Eocene time nonprophetic of that in Oligocene time 786 

Zygomatic indices 787 

Differential evolution in skull proportions in the several phj'la of titanotheres j__ 788 

Palate and shifting posterior nares 788 

Rudiments of horns arising independently in Eocene phyla 788 

Modes of origin of the horns of the titanotheres 790 

Phyletio divergence in time of origin of horns 790 

Accelerated direct evolution of horns in the "prophet horn " phylum 790 

Retarded evolution of horns in the Palaeosyopinae and Telmatheriinae 790 

Divergence in the form of horns of the Oligocene titanotheres 792 

Correlation (coadaptation) of horns with cranial and dental characters 792 

Correlation of horns with sex 793 

Evolution of the brain 793 

Dental mechanism 794 

Gradual change of diet 795 

Harmony of proportions of head and grinding teeth 795 

Arrested evolution in the teeth 796 

Foramina of the atlas vertebra in the titanotheres and other perissodactyls 797 

Progressive adaptations of the pectoral and pelvic arches, the limbs, and the feet of the titanotheres 798 

Ancestral titanotheres suboursorial 799 

The titanothere pes 800 

Progressive stages of the manus 800 

The titanothere magnum 800 

Evolution of the astragalus 801 

Section 4. Bibliography for Chapters VIII-X 802 

Chapter XI. Causes of the evolution and extinction of the titanotheres 805 

Section 1. Modes and causes of the origin and evolution of new adaptive characters (rectigradations) and new propor- 
tions (allometrons) in old characters 805 

Distinction between modes and causes 805 

Speculation as to the causes of evolution 806 

Distinction between invisible germ evolution and visible bodily evolution in the titanotheres 806 

Principles of tetraplasy in individual development; ontogeny 807 

Principles of single character (biocharacter) evolution 808 

Separability and coordination of biocharacters 809 

Initiation of biocharacters : Is it environmental, ontogenetic, or germinal? 809 

Differing hereditary velocities; rates of motion of biocharacters 810 

Modes of biocharacter evolution actually observed 811 

Mammalian phyla distinguished by different ontogenetic and phylogenetic velocities in the same biochar- 
acters 811 

Distinction between rectigradation biocharacters and allometron biocharacters 812 

Resemblances between rectigradations and allometrons as biocharacters 813 

Contrasts between rectigradations and allometrons as biocharacters 813 

Principles of rectigradation 814 

Horns arise as typical rectigradations — orthogenetic , continuous 814 

Phylogenesis and initiation of horns in titanotheres 816 

Dental rectigradations 817 

Rectigradations of Osborn contrasted with mutations of Waagen 818 

" Mutations" and "species" 819 

Heritage separabiUty of rectigradation dental cusps and folds 819 

Rectigradation cusps on titanothere teeth are unlike saltations 819 

Rectigradations influenced by degree of zoologic affinity 820 

Principles of proportion 820 

Proportion biocharacters (allometrons) 820 

Terminology indicating proportions of the skuU and skeleton , 822 

Proportional evolution in the typical brachycephal Palaeosyops 823 

Contrast between evolution and growth 823 

Differential (disharmonic) allometrons in Palaeosyops 823 

BrachycephaUc allometrons affect all biocharacters; harmonic 823 

Harmonic proportion trend constitutes the generic character of Palaeosyops 824 

Fourfold modes of allometric brachycephaly 824 

Differential allometrons of specific value 825 



CONTENTS V 

Chapter XI — Continued. 

Section 1 — Continued. 

Principles of proportion — Continued. Page 

Harmonic allometry exceptional 826 

Dolichorhinus: adaptation of the lengthened head to the supposed habit of grazing 827 

Dolichocephalic increments of Mesaiirhinus and Dolichorhinus 827 

Slow but differential evolution in breadth of skull of Dolichorhinus 828 

Differential evolution in the grinding teeth of Dolichorhinus 828 

Contrasts of differential evolution in brachycephalic and dolichocephalic skulls 828 

Summary of harmonic and differential aUometrons in the skulls and feet and an interpretation of the phylogeny 

of the titanotheres, by W. K. Gregory 828 

History of research 828 

Application of the proportional reversal principle to the titanotheres 830 

Irreversible and reversible evolution of allometrons 833 

Separability and correlation of biocharacters 833 

Separabilitj' of allometrons in heredity 833 

Correlation, coordination; compensation of rectigradations and allometrons 834 

Theoretic causes of the evolution of new characters and new proportions 834 

Theories advanced to explain the origin of rectigradations and allometrons -^ 834 

Analysis of the evidence on the modes of origin of variation as considered in Darwinism, Lamarckism, and 

tetrakinesis 835 

Observed principle of tetraplastic development of body form; theoretic principle of the tetrakinetic evolution 

of the germ 835 

Quantitative increment of the four separable factors in development and evolution 837 

Analysis of the modes of variation; theoretic importance of initiation 838 

Bearing of saltation versus continuity on the Lamarckian, Darwinian, and tetrakinetic theories 839 

Darwin's hypothesis of fortuitous saltation and fluctuation 839 

Most saltations in mammals abnormal 840 

MendeUan discontinuity in heredity 841 

Truth and error in Johannsen's pure line saltation principle 842 

Experiments in the artificial selection of variations of proportion 842 

Theoretic and experimental causes of the evolution of allometrons 843 

Germinal allometrons arising by continuous or gradual change 843 

Germinal allometrons arising by sudden changes (saltations); interaction theory .! 844 

Certain germinal allometrons uninfluenced by the direct action of environment 844 

Germinal allometrons apparently influenced by direct action of environment 845 

Germinal allometrons, fluctuating and nonfluctuating 845 

Certain germinal allometrons of high survival selection value 845 

Other germinal allometrons of apparently no survival selection value; predetermination 846 

Ontogenetic allometrons experimentally influenced by changes of habit 846 

Harmonic or conflicting influence of the four factors of evolution 846 

Ontogenetic allometrons experimentally influenced by changes of environment 847 

Germinal allometrons: Prenatal, adolescent, adult 847 

Ontogenetic allometrons influenced by glandular internal secretions, enzymes, and other organic catalyzers. 848 

Germinal (?) allometrons influenced by organic catalyzers 849 

Summary of theoretic causes of evolution 849 

Evidence against the Lamarckian principle 849 

Evidence against the theory of selection of minute variations 849 

Evidence favorable to the selection of certain fluctuations 849 

Unknown causes of the origin of rectigradations 849 

Necessity of experiment on the tetrakinetic principle 849 

Bibliography of literature relating to the theories of evolution cited in section 1 850 

Section 2. Natural selection in mammals; causes of the extinction of the titanotheres and other quadrupeds 852 

Extinction of faunas in the age of mammals 852 

Gradual or sudden extinction 853 

Extinction of both the adaptive and the inadaptive 853 

Phases of extinction 854 

Multiple causes of extinction 854 

The law of natural selection 855 

History of opinion 856 

Cataclysmal hypotheses 856 

Unif ormitarian theories 857 

Lyell on extinction 857 

Balance of nature (Lyell, Darwin, Wallace) 858 

Environmental causes of extinction 860 

The physical environment 860 

Physiographic changes 860 

Changes of climate 861 



VI CONTENTS 

Chapter XI — Continued. 
Section 2 — Continued. 

Environmental causes of extinction — Continued. Page 

The living environment 867 

Plants 867 

Insects and Protozoa 869 

Epidemics 871 

Birds 874 

Mammals 874 

Contrasts between external (environmental) and internal causes of extinction 877 

Internal causes of preservation and extinction 877 

Immunitj- and adaptation 877 

Bulk not inherently inadaptive 878 

Value of single organs in survival or extinction 880 

Inadaptation of extreme specialization 883 

Survival of the unspecialized 883 

Irreversible evolution 883 

Inadaptation of dominant organs 883 

Selection of sexually dominant organs 884 

Causes of overdevelopment 884 

Rates of breeding and extinction 884 

Self-extinction through arrested variation 886 

Special features of the extinction of the titanotheres 886 

Conclusions regarding the theory of natural selection of Darwin and Wallace 887 

Summary of conclusions 888 

Bibhography of literature relating to the extinction of faunas cited in section 2 889 

Appendix. Eocene and Oligocene titanotheres of Mongolia 895 

Index 947 



ILLUSTRATIONS 



Plate Page 

XLIII. Probable arrangement of muscles of occiput and neck in Brontops rohustus 726 

XLIV. Musculature of the fore and hind limbs of Palaeosyops leidyi 726 

XLV. Ontogenesis of the torns of domestic cattle 894 

XLVI. Models of heads of Eocene titanotheres, showing brachycephaly, mesaticephaly, and dolichocephaly 894 

XLVII. Disharmonic differential evolution of characters in skulls and teeth of titanotheres 894 

XLVIII. Extinction or survival of 23 families of artiodactyl ungulates 894 

XLIX. Brain of Eocene mammals compared with that of modern and other mammals 894 

L. Type skulls of Palaeosyops leidyi and Mesaiirhinus petersoni 894 

LI. SkuUs of Telmatherium idlimum and Manteoceras manteoceras 894 

LII. Skulls of Dolichorhinus hyognathus and Metarhinus fluviatilis 894 

LIII. Skull form in Eocene titanotheres (brachycephalic, mesaticephalic, and dolichocephalic) 894 

LIV. Upper and lower grinding teeth of Lamhdotherium and Eotitanops 894 

LV. Comparison of incisors and canines in Eocene titanotheres 894 

LVI. Lower grinding teeth of Limnohyops and Palaeosyops 894 

LVII. Upper dentition of Limnohyops 894 

LVIII. Upper dentition in four species of Palaeosyops 894 

LIX. Upper dentition in three species of Palaeosyops 894 

LX. Upper dentition of Limnohyops and Palaeosyops 894 

LXI. Type skull of Palaeosyops leidyi, palatal view 894 

LXII. Upper and lower premolars of Limnohyops and Palaeosyops 894 

LXIII. Grinding teeth of Telmatherium and Manteoceras 894 

LXIV. Upper dentition of Telmatherium validum 894 

LXV. Dentition of Telmatherium and Sthenodectes 894 

LXVI. Skull and jaw referred to Sthenodectes incisivus 894 

LXVII. Upper dentition of Manteoceras 894 

LXVIII. Incisors and canines of Protitanotherium emarginatum 894 

LXIX. Upper and lower teeth of Protitanotherium, progressive stages 894 

LXX. Lower dentition of Brachydiastematherium transilvanicum 894 

LXXI. Lower dentition of Mesatirhinus, Metarhinus, and Dolichorhinus 894 

LXXII. Upper dentition of Mesatirhinus amd Dolichorhinus 894 

LXXIII. Upper dentition of Dolichorhinus 894 

LXXI V. Upper dentition of Dolichorhinus, Metarhinus, and Bhadinorhinus 894 

LXXV. Skulls and jaw of Dolichorhinus 894 

LXXVI. Skulls of Dolichorhinus, side view 894 

LXXVII. Skulls of Dolichorhinus, palatal view 894 



CONTENTS VII 

Plate Page 

LXXVIII. Skulls and jaws of Metarhinus 894 

LXXIX. Skulls of Metarhinus 894 

LXXX. Skulls of Rhadinorhinus and Metarhinus 894 

LXXXI. Upper dentition of Eoliianoth erium osborni and Diplacodon elatus 894 

LXXXII. Comparison of upper teeth of Rhadinorhinus, Diplacodon, Menodus, and Brontotherium 894 

LXXXIII. Skull of Brontops brachycephalus (paratype) 894 

LXXXI V. Skull of Brontops brachycephalus (referred) , top and side views 894 

LXXXV. Skulls of Brontops brachycephalus (referred), top view 894 

LXXXVI. Skull of Brontops brachycephalus, front and palatal views 894 

LXXXVII. Skulls of Brontops brachycephalus, palatal view 894 

LXXXVIII. Skulls of Brontops brachycephalus and B. dispar 894 

LXXXIX. Skulls of Brontops dispar, palatal views r 894 

XC. Skull of Brontops dispar, top, side, and palatal views 894 

XCI. Skull of Brontops dispar, top and side views 894 

XCII. Skulls of Brontops dispar, top view 894 

XCIII. Skulls of Brontops dispar, side view . 894 

XCIV. Skulls of Brontops dispar, front view 894 

XCV. Skull of Brontops robustus?, side view 894 

XCVI. Type skull and jaw of Brontops robustus, oblique side view 894 

XCVII. Type skull and jaw of Brontops robustus, side view 894 

XCVIII. Type skull of Brontops robustus, top view 894 

XCIX. Type skull and jaw of Brontops robustus, front view 894 

C. Type skull of Brontops robustus, palatal view 894 

CI. Dentition of type skull of Brontops robustus 894 

CII. Lower jaw of Brontops robustus (type) 894 

cm. Type skull and jaw of Brontops robustus 894 

CIV. Type skull of Diploclonus bicornutus 894 

CV. Skull related to Diploclonus tyleri 894 

CVI. Type skull, jaw, and \eeXh oi Diploclonus tyleri 894 

CVII. Naso-frontal region of skulls of Diploclonus 894 

CVIII. Type skulls of Diploclonus amplus and D. tyleri 894 

CIX. Type skull of Diploclonus amplus 894 

ex. Skull referred to Diploclonus amplus? 894 

CXI. Type skull of Allops walcotti 894 

CXII. Type skull of Allops walcotti and skull of Allops? marshif 894 

CXIII. Skull of Allops marshif, side and front views 894 

CXIV. Skulls of Allops marshi, side views 894 

CXV. Skulls of Allops marshi, palatal view of t3'pe, top view of paratype 894 

CXVI. Skulls of Allops marshi? 894 

CXVII. Skulls of Allops serotinus 894 

CXVIII. Skull of Allops serotinus, palatal view 894 

CXIX. Front views of skulls of Allops 894 

CXX. Skulls of Allops crassicornis and A. serotinus 894 

CXXI. Type skull of Allops crassicornis 894 

CXXII. Skull, jaw, and teeth of Allops crassicornis? seu marshi 894 

CXXIII. Type skull of Menodus heloceras 894 

CXXIV. SknWs oi Menodus heloceras (type) and. Brontotherium hatcheri 894 

CXXV. Lower molar and fragment of jaw of Menodus proutii and jaw of Menodus torvus 894 

CXXVI. Jaw referred to Menodus proutii (trigonoceras) 894 

CXXVII. Type jaw of Menodus torvus 894 

CXXVIII. Type skull of Menodus trigonoceras, top view 894 

CXXIX. Type skull of Menodus trigonoceras 894 

CXXX. Skull of Menodus trigonoceras (cotj'pe) and a radius doubtfully referred 894 

CXXXI. Palate of Menodus trigonoceras 894 

CXXXII. Upper molar series of Menodus and Allops 894 

CXXXIII. L'pper teeth of Menodus and Brontotherium contrasted 894 

CXXXI V. Upper and lower dentition of Menodus trigonoceras 894 

CXXXV. Skulls of Menodus trigonoceras and Menodus giganteus 894 

CXXXVI. Skull of Menodus giganteus, side view 894 

CXXXVII. Skull of Menodus giganteus, side view 894 

CXXXVIII. Skull of Menodus giganteus, top view 894 

CXXXIX. QknWoi Menodus giganteus, si&ewiew 894 

CXL. Skull of Menodus giganteus, front view 894 

CXLI. Skull of Menodus giganteus, palatal view 894 

CXLII. Type skull of Menodus varians 894 

CXLIII. Type skull of Megacerops copei 894 

CXLIV. Skulls of Megacerops copei (type) and M. bucco, front view 894 

CXLV. Skulls of Megacerops copei (type) and M. bucco, palatal view 894 



VIII CONTENTS 

Plate Page 

CXLVI. Skulls of Megacerops acer (type) and doubtfully referred female 894 

CXLVIi. Skulls of Megacerops acer (type) and doubtfully referred female, side view 894 

CXLVIII. Skulls of Megacerops acer (type) and doubtfully referred female, top view 894 

CXLIX. Skulls of Megacerops acer (type) and doubtfully referred female, top view 894 

CL. Skulls of Megacerops acer (type) and doubtfully referred female, front view 894 

CLI. Female skull of Megacerops acer? (t3'pe of Sijmborodon altirostris), palatal view 894 

CLII. Female skull of Megacerops acer? (type of Symborodon altirostris) , palatal view 894 

CLIII. Type skull of Megacerops {"Symborodon") bucco and jaws of Brontops? 894 

CLIV. Type skull of Megacerops {"Symborodon") bucco, palatal view 894 

CLV. Type skull of il/effacerops {"Symborodon") bwcco, top view 894 

CLVI. Skull of Megacerops bucco?, top view 894 

CLVII. Skulls of Menodus, Brontotherium, and Megacerops and a fragment of a humerus 894 

CLVIII. Type jaws of Megacerops riggsi and Menodus torvus, side view 894 

CLIX. Type jaws of Megacerops riggsi and Menodus torvus, top and basal views 894 

CLX. Jaws of Brontops sp. and Megacerops riggsi (type) 894 

CLXI. Skulls of Brontotherium leidyi, side view 894 

CLXII. Skulls of Brontotherium leidyi, top view 894 

CLXIir. Skulls of Brontotherium leidyi, front and palatal views 894 

CLXI'V. Skull and jaw of Brontotherium leidyi, side view 894 

CLXV. Skull of Brontotherium leidyi, palatal view 894 

CLXVI. Type and referred specimens of Brontotherium hypoceras 894 

CLXVII. Type skull of Brontotherium hatcheri, palatal view of skull, front view of horns 894 

CLXVIII. Type skull of Brontotherium hatcheri, top view 894 

CLXIX. Skull referred to Brontotherium hatcheri 894 

CLXX. Skulls of Brontotherium hypoceras?, B. hatcheri, and B. gigas, front view 894 

CLXXI. Type jaw of Brontotherium gigas 894 

CLXXII. Jaws of Brontotherium hatcheri? and B. medium? 894 

CLXXIII. Skull oi Brontotherium gigas elatum, side view 894 

CLXXIV. Skull of Brontotherium gigas elatum, front view 894 

CLXXV. Skull of Brontotherium gigas elatum (type of Titanops elalus), side view 894 

CLXXVI. T^'pe skull of Brontotherium medium 894 

CLXXVII. Skull of Brontotherium curtum 894 

CLXXVIII. Type skull of Bronioi/ieriMm {"Titanops") cwr^Mm, palatal view 894 

CLXXIX. Type skull of Brontotherium {"Titanops") curtum, side view; front view of horns 894 

CLXXX. Type skulls of Brontotheriiim medium and B. curtum 894 

CLXXXI. Skull of Brontotherium platyceras, side view 894 

CLXXXII. Skull of Brontotherium platyceras, top view 894 

CLXXXIII. Skull of Brontotherium curtum and horn of B. hypoceras 894 

CLXXXIV. Skulls of Brontotherium gigas and B. curtum (type) 894 

CLXXXV. Skulls referred to Brontotherium curtum 894 

CLXXXVI. Skulls of Brontotherium curtum (type), B. dolichoceras (tjrpe), and Megacerops? sp 894 

CLXXXVII. Type skull of Brontotherium tichoceras and type horns of B. platyceras 894 

CLXXXVIII. Skull of male Brontotherium platyceras, front view 894 

CLXXXIX. Skull of male Brontotherium platyceras, side view 894 

CXC. Skulls of female Brontotherium? gigas? and B. curtum 894 

CXCI. Type skull of male Brontotherium ramosum 894 

CXCII. Skull of female Brontotherium curtum, palatal view 894 

CXCIII. Skull of female Brontotherium curtum, front and top views 894 

CXCIV. Skull of female Brontotherium curtum, palatal view 894 

CXCV. Atlas of the type of Brontops robustus Marsh 894 

CXCVI. Axis of the type of Brontops robustus Marsh 894 

CXC VII. Fourth cervical vertebra of the type of Brontops robustus Marsh 894 

CXCVIII. Second dorsal vertebra of the type of Brontops robustus Marsh 894 

CXCIX. Tenth dorsal vertebra of the type of Brontops robustus Marsh 894 

CC. Second lumbar vertebra of the type of Brontops robustus Marsh 894 

CCI. Caudal vertebrae of the type of Brontops robustus Marsh 894 

CCII. Second and tenth left ribs of the type of Brontops robustus Marsh 894 

CCIII. Fourth rib of the type of Brontops robustus Marsh 894 

CCIV. Left scapula of the tj'pe of Brontops robustus Marsh 1 894 

CCV. Left humerus of the type of Brontops robustus Marsh 894 

CC VI. Left radius of the type of Brontops robustus Marsh 894 

CCVII. Left ulna of the type of Brontops robustus Marsh 894 

CCVIII. Left scaphoid of the tj'pe of Brontops robustus Marsh 894 

CCIX. Left cuneiform carpi and right pisiform of the type of Brontops robustus Marsh 894 

CCX. Left trapezoid and left magnum of the type of Brontops robustus Marsh 894 

CCXI. Left unciform of the tj-pe of Brontops robustus Marsh 894 

CCXII. Second left metacarpal of the type of Brontops robustus Marsh 894 

CCXIII. Third left metacarpal of the type of Brontops robustus Marsh 894 



CONTENTS IX 

Plato Page 

CCXIV. Fourth left metacarpal of the type of Brontops robustus Marsh i 894 

CCXV. Fifth left metacarpal of the t3'pe of Brontops robustus Marsh 894 

CCXVI. Proximal phalanges of left manus of the type of Brontops robustus Marsh 894 

CCXVII. Phalanges and sesamoids of the type of Brontops robustus Marsh 894 

CCXVIII. Pelvis and sacrum of the tj'pe of Brontops robustus Marsh 894 

CCXIX. Pelvis and sacrum of the tj'pe of Brontops robustus Marsh 894 

CCXX. Left femur of the type of Brontops robustus Marsh 894 

CCXXI. Left tibia of the type of Brontops robustus Marsh 894 

CCXXII. Left patella and left fibula of the type of Brontops robustus Marsh 894 

CCXXIII. Left astragalus of the type of Brontops robustus Marsh 894 

CCXXIV. Left calcaneum of the type of Brontops robustus Marsh 894 

CCXXV. Left navicular and left cuboid of the type of Brontops robustus Marsh 894 

CCXXVI. Left ectocuneiform and left mesocuneiform of the type of Brontops robustus Marsh 894 

CCXXVII. Left fourth metatarsal of the type of Brontops robustus Marsh 894 

CCXXVIII. Left manus and left pes of the type of Brontops robustus Marsh 894 

CCXXIX. Restoration of the skeleton of Brontops robustus Marsh 894 

CCXXX. Left lunar and right trapezoid referred by Marsh to Brontotherium gigas 894 

CCXXXI. Pelvis and sacrum referred by Marsh to Brontotherium gigas 894 

CCXXXII. Left second metatarsal referred by Marsh to Brontotherium gigas 894 

CCXXXIII. Left third metatarsal referred by Marsh to Brontotherium gigas 894 

CCXXXI V. Left fourth metatarsal referred bj' Marsh to Brontotherium gigas 894 

CCXXXV. Phalanges of left second, third, and fourth digits referred by Marsh to Brontotherium gigas 894 

CCXXXVI. Skeletons of Megacerops acer, male and female 894 

Figure 

640. Heads of four titanotheres 703 

641. Facial musculature and nasal cartilage of Manteoceras 704 

642. Facial musculature and nasal cartilage of Menodus giganteus 705 

643. Facial musculature and nasal cartilage of Brontotherium platyceras 706 

644. Relation of the contour of the head to the skull and jaws in living perissodactyls 707 

645. Model of half of skull and head of Palaeosyops leidyi 708 

646. Models of half of skuU and head of Eotitanops borealis and Manteoceras manteoceras 709 

647. Heads of titanotheres of middle and upper Eocene time 710 

648. Heads of Eotitanops, Manteoceras, Protitanotherium, and Brontotherium, front view 711 

649. Heads of Eotitanops, Manteoceras, Protitanotherium, and Brontotherium, side view 711 

650. Occiput of Brontops robustus, showing probable location of muscles 712 

651. Probable position of muscles of dorsal vertebrae in Brontops 713 

652. Right scapula of Brontotherium, showing principal muscle attachments 714 

653. Left humerus of Brontops robustus (type) , showing probable musculature 715 

654. Left radius of Brontops robustus (type) , showing probable musculature 716 

655. Left ulna of Brontops robustus (type), showing probable musculature 716 

656. Arrangement of the muscles of the flexor surface of the manus of Mesatirhinus petersoni 717 

657. Left femur of Brontops robustus (type) , showing probable musculature 719 

658. Left fibula and left tibia of Brontops robustus (type) , showing probable musculature 720 

659. Skeleton and restoration of Brontops cf. B. robustus 722 

660. Restoration of the superficial muscles of Brontops cf. B. robustus 723 

661. Restorations of nine species of titanotheres, Eocene and Oligocene 724 

662. Restoration of Brontotherium gigas 725 

663. Angles of muscular insertion and "parallelogram of forces" 728 

664. Angles of insertion of muscles to pelvis and femur in horse and mastodon 729 

665. Graviportal adaptation in the astragalus 729 

666. Cursorial and graviportal adaptations in horse and elephant 730 

667. Cursorial adaptation in fore limb and hind limb of horse 731 

668. Angulation of the fore limb in graviportal types 732 

669. Angulation of limb bones at shoulder and hip joints in Equus, Brontops, and Mastodon 733 

670. Hind limbs of cursorial and graviportal types 734 

671. Five types of scapula, cursorial to graviportal 740 

672. Forms of distal articular surface of humerus 742 

673. Seven forms of ilia, showing cursorial and graviportal adaptations 743 

674. Adaptive forms of ilia and sacral attachment 744 

675. Limb structure of perissodactyls: Fore limbs of Heptodon and Tapirus 745 

676. Limb structure of perissodactyls: Hind limbs of Heptodon and Tapirus 747 

677. Limb structure of perissodactyls : Fore limbs of paleotheres 74s 

678. Limb structure of perissodactyls: Hind limbs of paleotheres t 749 

679. Limb structure of perissodactyls: Fore limbs in four American rhinoceros subfamilies 750 

680. Limb structure of perissodactyls : Fore limbs in four Old World rhinoceros subfamilies 751 

681. Limb structure of perissodactyls: Hind limbs of rhinoceroses found in North America 751 

682. Limb structure of perissodactyls: Hind limbs of rhinoceroses found in the Old World 752 



X CONTENTS 

Figure Page 

6S3. Limb structure of perissodaotyls: Fore limbs of Equidae 753 

684. Limb structure of perissodactyls : Hind limbs of Equidae 754 

685. Limb structure of perissodactyls: Fore limbs of titanotheres 765 

686. Limb structure of perissodactyls : Hind limbs of titanotheres 756 

687. Carpus of Creodonta, Amblypoda, and Condylarthra, showing "alternating" and "serial" types 758 

688. Evolution of the upper molar pattern in condylarths and titanotheres 759 

689. Generalized carpus of insectivore-creodont type 760 

690. Progressive stages of structural evolution in the skull and molar teeth of titanotheres 761 

691. Astragali of Phenacodus and Heptodon 762 

692. Manus of Heptodon, Lambdotherium, and Eotitanops 762 

693. Fore and hind feet in perissodactyls 763 

694. Skulls of Eocene titanotheres, other perissodactyls, and one condylarth 765 

695. Skulls of Eocene condylarth and perissodactyls 766 

696. Skulls of Lophiodon leptorhynchus 768 

697. Family tree of the titanotheres 769 

69S. Family tree of the Perissodactyla 770 

699. Fore and hind feet of paraxonic and mesaxonic ungulates 771 

700. Carpus of Eocene perissodactyls 772 

701. Evolution of the astragalus in the titanotheres 773 

702. Back view of the carpus of a middle Eocene titanothere 775 

703. Left magna of an Eocene titanothere and two chalicotheres 775 

704. Left magna of Heptodon and Eotitanops . 776 

705. Phylogenetic tree of the Perissodactyla 777 

706. Family tree of the Perissodactyla , 777 

707. Outlines of body form of perissodactyls 781 

708. Disharmonic evolution shown in characters of Brontops, Menodus, and Brontotherium 783 

709. Evolution of the skull in the titanotheres 784 

710. Relative proportions of the skulls of Manteoceras and Brontotherium 785 

711. Proportions of skulls of Palaeosyops, Dolichorhinus , and Eotitanops 786 

712. Evolution of the frontonasal horn swelling in Manteoceratinae and Brontopinae 791 

713. "Brain casts" of titanotheres compared with the brain of a recent rhinoceros 792 

714. Relative size of brain and skuU in titanotheres and other Eocene perissodactyls, an artiodactyl, and an amblypod 793 

715. Brain of Menodus, Rhinoceros, and other quadrupeds 795 

716. " Brain casts " of titanotheres 796 

717. Evolution of the upper molar tooth in titanotheres 796 

718. Pattern of an upper molar of five perissodactyls 797 

719. Evolution of proportions in grinding teeth of Oligocene titanotheres 797 

720. Atlas and axis of titanotheres and other perissodactyls, showing probable course of arteries 798 

721. Evolution of the pelvis in titanotheres 799 

722. Three stages in the evolution of the manus in titanotheres 801 

723. Six stages in the evolution of the manus in titanotheres. ^ 801 

724. Evolution of the magnum in titanotheres 801 

725. Restorations of Eotitanops borealis and Brontotherium platyceras 805 

726. Relations of heredity germs of Eotitanops and Brontotherium 806 

727. Continuity of heredity germ from Eotitanops to Brontops 808 

728. Germinal origin and disappearance of characters 810 

729. Contrast between the ontogenetic and phjdogenetic velocity of a biocharacter 810 

730. Limbs of four embryos, showing ontogenetic velocity of limb bones in horses 811 

731. Relative velocities of a series of biocharacters 811 

732. Rectigradations in the teeth of Eocene ungulates 814 

733. Rectigradations and allometrons in skuUs, teeth, and feet of titanotheres 815 

734. Independent appearance of horn rudiments in different phyla of titanotheres 816 

735. Cusp rectigradations in Telmatherium 817 

736. Separability of rectigradation biocharacters in hybrid offspring of ass and horse 820 

737. Harmonic and disharmonic brachycephaly and dolichocephaly 821 

738. Disharmonic and distinctive increment in biocharacters of Palaeosyops and Dolichorhinus 822 

739. Continuous origin of allometron biocharacters in man and titanotheres 822 

740. Proportions and flexure of skull in horses and titanotheres 823 

741. Skulls of three species (probably successive) of Palaeosyops 824 

742. Outlines of skulls of Palaeosyops, Dolichorhinus, and Eotitanops 825 

743. Skull of Dolichorhinus hyognathus, type of " Telmatotherium cornutum" 829 

744. Differences in proportions of premolars and molars corresponding to differences in proportion of the skull 830 

745. Skulls of Eocene titanotheres, showing changes in proportion and in development of horns 831 

746. Models of heads of Oligocene titanotheres 832 

747. Separability and imperfect blending of allometric biocharacters in the facial bones of the horse, ass, and mule 832 

748. Contrasts between hypotheses of Lamarck, Weismann, and Osborn regarding the causes of evolution 834 

749. Cumulative or favorable influence of heredity, ontogeny, and environment 837 

750. Diagram illustrating the tetraplastic theory -- 837 



CONTENTS ■ XI 

Pigure Page 

751. Diagram illustrating the principle of tetraplasy 838 

752. Accelerated elongation of the limbs in the young zebra and guanaco 846 

753. Femur of dog, normal and as modified by amputation or congenital absence of fore limb 847 

754. Brachydactyly and dolichodactyly 848 

755. Examples of dwarfing due to removal or abnormal functioning of certain glands 848 

756. Theoretic selection of fluctuations of proportion 849 

757. Extinction of archaic mammals 853 

758. Affinities and duration of nine families of Perissodactyla 854 

759. Influence of secular desiccation on archaic and modern orders of mammals 866 

760. Brain proportions in Eocene perissodactyls, an artiodactyl, and an amblypod 882 

761-797. See Appendix. 



CHAPTER VIII 



THE MUSCULAR ANATOMY AND THE RESTORATION OF THE TITANOTHERES « 

By William K. Gregory 
SECTION 1. MUSCLES OF THE FACE AND JAW 



The skeleton of Eocene titanotheres resembles 
that of other perissodactyls in so many points that 
the general position and course of the principal 
muscles may be inferred by comparison with the 
musculature of existing perissodactyls — the horses, 
tapirs, and rhinoceroses. 



and Windle and Parson's work on the muscles of the 
Ungulata (Windle, 1901.1, 1904.1) were also con- 
sulted. Other authorities referred to are Cuvier, 
Cunningham, Ellenberger, Haughton, Schmaltz, Still- 
man, and Weisse. (See Bibliography, p. 802.) 

A study of existing musculature enabled the writer 
(Gregory, 1920) to infer the general anatomy of the 




Figure 640. — Heads of four titanotheres 

Representatives of the four principa genera of lower Oligocene time, restored by Charles R. Knight under the direction of the author. 
A, Brontops roiusius; B, Menodus gigantms; C, Megacerops copei; D, Brontotherium platyceras. About one-seventeenth natural size. 



The facial muscles of the horse, of the tapir, and 
of many other mammals have been described and 
superbly figured in a great monograph by J. F. V. 
Boas and S. PauUi (Boas, 1908.1). The facial muscles 
in the Sumatran rhinoceros have been briefly described 
and figured by Beddard and Treves (Beddard, 1889.1). 
Murie's dissection of the Malayan tapir (1872. 1) 

" The conclusions here presented are the results of studies made in 1912, 1917, 
and 1918 (Gregory, 1920.1). The drawings from which the figures were reproduced 
were made by Erwin S. Christman. 



facial muscles in the extinct Tertiary ancestors of the 
tapirs, horses, and rhinoceroses and especially the 
Eocene and Oligocene titanotheres. 

In the Eocene titanotheres, such as Manteoceras 
(fig. 641), the smooth anterosuperior border of the 
premaxillae and the adjacent region of the maxillae, 
as well as the front end of the lower jaw, doubtless 
lay beneath the orbicularis oris muscle, which sur- 
rounded the mouth. This muscle in ungulates,' 

703 



704 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



according to Boas and PauUi (1908.1), is merely 
the front part of the buccinator, which also extended 
backward alongside the grinding teeth and between 
the molars and the malar. 

To the corner of the mouth was doubtless attached 
the zygomaticus muscle, which arose from the front 
part of the zygomatic arch, beneath the orbit. 

To the surface of the upper lips at the sides was 
attached the wide nasolabialis, which passed obliquely 



side, as in the horse and tapir. It very probably 
ran above the huge nasal cartilage and beneath the 
lower rim of the nasal bone back to its origin in the 
wide depression in front of the orbit, which was 
bounded above by the nasals and behind by the 
raised anterior rim of the orbit. 

The anterior end of the upper hp and the side of 
the nose were very probably retracted^by the inferior 
branch of the maxillolabialis, which ran back along 




Figure 641. — Facial musculature and nasal cartilage of an Eocene titanothere 



Superficial layer of muscles; A2, relations of nasal chamber and surrounding muscles to skull, 
natural size. 



upward and backward, overlying both branches of the 
maxillolabialis (levator labii superioris alaeque nasi), 
to its origin on the curved rim of the nasal and 
maxillary, beneath the horn swelling and in front of 
the orbit. 

The anterior tip of the lip was raised by the superior 
branch of the maxillolabialis. This was perhaps 
joined by ligament with its fellow of the opposite 



the side of the face to its tendinous origin beneath 
the orbital portion of the zygomatic arch. The 
very massive nostrils were doubtless compressed by 
the action of the dilator naris and lateralis nasi 
muscles, which surrounded the back part of the 
nasal chamber and its diverticulum, which together 
filled the space below the nasals and above the 
maxilla. 



THE MUSCULAR ANATOMY AND THE RESTORATION OF THE TITAN OTHERES 



705 



inserted on the rather small coronoid process. The 
internal and external pterygoids were as well developed 
as in other perissodactyls. They doubtless assisted 
greatly in oblique and lateral movements of the 
mandible in feeding. 



As thus interpreted, the Eocene titanotheres differed 
from the tapirs in having the upper branch of the 
maxillolabialis placed more directly anteroposteriorly 
instead of having it run more obliquely upward and 
forward over the nasal cartilage. This arrangement is 
clearly implied by the shape of the nasals 
which are not pointed and retracted as in the 
tapir but long and distally spreading. Thus 
the shape of the nasals in Manteoceras proves 
that this animal could not have had a pro- 
boscis. Both branches of the maxillolabialis 
in Eocene titanotheres probably occupied their 
primitive position, immediately in front of the 
orbit, as shown by the raised preorbital border 
and by the depression in front of it. 

In Oligocene titanotheres the great deep- 
ening and shortening of the preorbital part of 
the skuU was accompanied by corresponding 
changes in the facial muscles. The area for 
the nasal chamber and for the dorsal branch 
of the maxillolabialis was greatly increased 
transversely but shortened anteroposteriorly, 
the muscle being bounded postero-externally 
by the sharp external ridge running down from 
the horn to the front end of the zygomatic 
arch. The inferior branch of the maxillo- 
labialis may well have covered the region below 
the infraorbital foramen, and probably its pos- 
terior tendon was attached to the rough surface 
below the orbital rim, along with the anterior 
end of the fascia of the masseter. 

The progressive abbreviation of the nasals 
in certain phyla, Brontops, Allops, Megacerops, 
together with the great widening of the area 
of origin of the maxillolabialis and the narrow- 
ing of the end of the lower jaw, points to the 
progressive development of an enormous nasal 
chamber and highly protrusile lip-nostril com- 
plex of great vertical thickness. 

In the Menodus phylum, on the other hand, 
the persistently wide nasals and the general 

configxiration of the face are somewhat more 

suggestive of the rhinoceroses; probably these ,,/////^''i0' 

cursorial titanotheres were more square-lipped Figure 642. — Facial musculature and nasal cartilage of a lower Oligocene 

and the upper lip was less pendulous. titanothere {Menodus giganteus) 

In Eocene titanotheres the masseter and Ai, superficial layer of museles; A2, relations of nasal chamber and surrounding muscles to 

its fascia extended over the whole lower 




skull. One-sixth natural size. 



border and part of the inner side of the malar, 
as indicated by the presence of deep scars in this 
region, analogous to similar scars among other 
perissodactyls. The masseter was inserted along 
the outer rim of the expanded angle of the 
mandible. 

In Oligocene titanotheres the masseter was relatively 
of enormous size. The temporal muscle doubtless 
filled a great part of the temporal fossa and was 



SECTION 2. RESTORATIONS OF THE HEADS OF 
TITANOTHERES 

Preliminary restorations of the heads of a series of 
eight Eocene and Oligocene titanotheres were made 
some years ago by Charles R. Knight under the direc- 
tion of Professor Osborn (figs. 220, 640). Afterward it 
was felt that the mouth and nose of these models called 
for further study, and a series of life-size head models 
was projected. 



706 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



In preparation for this work and as a part of our 
studies of the muscular anatomy of the entire titano- 
there skeleton, Gregory and Christman made a 



form, muscle form, and external appearance were 
observed not only by this means but in living rhinoc- 
eroses, tapirs, and other animals in the New York 



comparative study of the head muscles of the tapir, | Zoological Park and elsewhere. These observations 

were gradually applied to the study of 
the muscular anatomy of the heads of 
titanotheres ; and the probable location 
of the bony areas of muscular origin 
and insertion were determined and illus- 
trated in the series of diagrams here 
presented. 

Next a series of reconstructions of 
the skulls of five Eocene and Oligocene 
titanotheres was carefully made. Each 
model in this series forms a synthesis 
of the best-preserved specimens of parts 
of the skull of the species it represents, 
for no one fossil skull is perfectly pre- 
served. In these models also the distor- 
tion and crushing of the originals were 
corrected. The muscles of the jaw, 
neck, and face were modeled first upon 
one side, and after the models had been 
photographed in this stage the other 
side was modeled. 

The general shape of the muscles of 
the jaw and neck in recent perissodactyls 
is plainly indicated by the form and 
extent of the temporal fossae, lower 
jaw, and occiput, and our models of 
titanothere heads are doubtless fairly 
accurate in these parts. The precise 
shape of the nose and lips is more diffi- 
cult to determine, but the models here 
figured represent the result of a series of 
carefully considered hypotheses based 
upon a wide range of comparative ana- 
tomical facts. 

In this series of restorations we have 
taken into account the origin, evolution- 
ary trend, and systematic relations of 
all the animals studied. First we con- 
sidered the structure, the habits, and 
the probable appearance of the earliest 
known and generalized members of the 
titanothere group and of the related 
families of perissodactyls — the earliest 
"tapirs," "rhinoceroses," "horses, "etc. 
Second we studied the diverse and highly 
specialized end members of these fami- 
lies, and third we studied the interme- 
diate types. In the order Perissodactyla 
unusually complete material for such 
study is available. We know the skull and the den- 
tition of many genera of primitive lower Eocene 
perissodactyls; we know the features of a still greater 
number of their divergent descendants in the middle 




Figure 643. — Facial musculature and nasal cartilage of a lower Oligocene titano- 
there {Brontotherium plaiyceras) 
Ai, Superficial layer of muscles; A2, relations of nasal chamber and surrounding muscles to skull. One- 
sixth natural size. 



horse, elephant, and other mammals, as elaborately 
figured by Boas, Murie, Schmaltz, and others. Obser- 
vations were also made of freshly dissected heads of 
horses and zebras. The correlations between bone 



THE MUSCULAR ANATOMY AND THE RESTORATION OF THE TITANOTHERES 



707 



and later Tertiary ; and we still have surviving the very 
diversely specialized end members of three families — 
the tapirs, rhinoceroses, and horses. The divergent 
trends of evolution in food habits and structure from the 



These divergent trends of evolution in the peris- 
sodactyls are especially evident in the region of the 
nostrils and lips. In the earliest true titanothere, 
Eotitano'ps, the incisor and canine teeth and the whole 





Figure 644. — Relation of the contour of the head to the skull and jaws in living perissodaotyls 

A., Ehineroceros indicus (Indian rhinoceros); B, Opsiceros bicornis (black rhinoceros); C, Ceratoikerium simum (white rhinoceros); D, 
JSquus (horse); E, Tapirus terrestris (tapir). In E the restriction of the nasals is associated with a marked outgrowth of the protrusile 
proboscis. The angle of the mouth usually extends behind the anterior premolars, but in some species, as in the white rhinoceros, 
it is in front of the tooth row. The nostrils are generally in line with the middle of the narial sinus. 



primitive Eocene perissodactyls to their latest descend- 
ants are accordingly fairly well known, and we have 
attempted to illustrate them in the series of restorations 
of Eocene and later perissodactyls shown in Chapter IX. 
101959— 29--VOL 2 2 



front part of the jaws differed only in minor characters 
from the same parts in the contemporary primitive 
tapirs, rhinoceroses, and horses. Our restoration 
of the head of Eotitanops resembles our previous 



708 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



restorations of the head of the eariiest "horse" 
(EoMppus) and of the earUest cursorial "rhinoceros" 
{Hyracliyus), because the skulls themselves show an 
underlying similarity. However, as the incisors, 
canines, and bony jaws of Eotitanops already exhibit 
a step toward those of the titanotheres of the later 
Eocene, we felt justified in making the muzzle and 
lips heavier than those of the contemporary horses. 



improbable, because the entire front part of the skull 
is widely different from those of tapirs and elephants. 
If Brontotherium had any form of proboscis we should 
expect to find that the fossae for the origin of the 
levator muscles of the nose and lips and for the nasal 
chamber would be either more tapiroid or more 
elephantine in form than they are. The head seems 
to liavc been carried so low that a heavy, square, 




Figure 645. — Model of half of skull and head of an Eocene titanothere {Palaeo syops leidyi) 
Modeled by E. S. Christman under the direction of William K. Gregory. About one-eighth ntitural size. 



In the middle Eocene Palaeosyops (fig. 645) the 
structure of the jaws and the nasal region, in compari- 
son with those of tapirs and horses, gives very definite 
evidence that the nostrils were nearly terminal and 
the muzzle partly protrusile. 

The restoration of the muzzle and lips of the final 
member of the brontothere phylum, namely, Bronto- 
therium platyceras, was more difficult. The hypothe- 
sis that this kind of titanothere had a proboscis lip is 



protrusile muzzle, rather than a true proboscis, would 
be sufficiently long to reach the ground. 

As noted above, the extreme and abrupt narrowing 
of the nasal tips in the advanced stages of Bronto- 
therium and the expansion of the areas for the muscles 
of the nose and lips seem to indicate that the muzzle 
of these animals was very large and that the Hps were 
extremely muscular, thick, and protrusile — not at all 
proboscis-like but more like those of the black rhi- 



THE MUSCULAR ANATOMY AND THE RESTORATION OF THE TITANOTHERBS 



709 



noceroses. The construction of the jaws and grinding 
teeth indicates that their main food was very coarse 
herbage, rather than grass, so it is reasonable to sup- 
pose that the mobile lips would be adapted rather to 
browsing and tearing up shrubs than to cropping 
short grass, and consequently that the median part 
of the lips would be pointed and very protrusile, 
somewhat as in the black rhinoceros, rather than 
truncate, as in the grazing lips of the white rhinoceros 
and horse. 

Such being our conceptions of the first and of the 
last known members of the family, it seems reasonable 
to assign the middle Eocene genus Manteoceras (in 
which the bones and teeth were structurally interme- 




SECTION 3. MUSCLES OF THE NECK AND BACK 

In the following pages we have attempted to deter- 
mine what muscles and ligaments were attached to 
the principal parts of the skeleton; hence the subject 
matter is classified primarily according to the topog- 
raphy of the skeleton, secondarily according to 
musculature. 

The vertebrae of titanotheres are sufEciently like 
those of the horse to indicate that the axial muscula- 
ture was also essentially similar. The following de- 
scription is based largely on Schmaltz's (1909.1) figures 
of the cervical and dorsal musculature of the horse 
supplemented from photographs by Herbert Lang of 
a dissected white rhinoceros. 





Figure 646. — Models of half of the skuU and head of Eoiitanops borealis (A) and Manteoceras 

manteoceras (B) 
Modeled by E. S. Christman under the direction of William K. Gregory. One-eighth natural size. 



diate between these extremes) to an intermediate stage 
in the development of the nose and lips (figs. 646-649). 

The restoration of the fore part of the head of the 
upper Eocene Protitanotherium emarginatum is based 
upon the well-preserved facial bones and lower jaw of 
the type and is probably fairly accurate, but the 
restoration of the back part of the head is hypothetical. 

In Figure 647 we give another series of restorations, 
showing the heads of members of the principal genera 
of middle and upper Eocene titanotheres, except the 
Palaeosyopinae, and showing also the contour of the 
skull. 



The occiput (fig. 650), especially in the later titano- 
theres, bore highly rugose crests, which probably 
separated the jaw muscles (temporals) from the neck 
muscles and were covered with thick, tough skin. A 
deep median pit marks the insertion of the ligamentum 
nuchae. The semispinalis capitis (complexus) was 
probably inserted just beneath the occipital crest and 
lateral to the ligamentum nuchae. The splenius was 
lateral to the semispinalis and probably formed a 
wide sheet just behind the lateral occipital ridge. 
Below the splenius and behind the auditory meatus 
was the insertion for the tendon of the longissimus 



710 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 




Figure 647.— Heads of titanotheres belonging to eight genera of middle and upper Eocene time 

Hectored in accordance with the skull contours by Mrs. E. M. Fulda under the direction of William ^'^''ZLofneriZ^n^^ 
superbum: B, Tdmalherium uliimum: C, SMdinorhinus iipUconus: D, ilanteocems manteoceras: E, Sotttamtnermm osborm, i, 
Metarhinus earlei; G, DolidLorhinus hyognathus: H, Mesalirliinus peiersoni. One-twelfth natural size. 



THE MUSCULAR ANATOMY AND THE RESTORATION OF THE TITANOTHERES 



711 




Figure 648. — Heads of four titanotheres, showing 

progressive stages of development 

Front view. Modeled and drawn by E. s. Christman under the Figure 649. — Heads of four titanotlieres, showing progressive stages of development 

direction of William K. Gregory and H. F. Osborn. A, Eotitanops 

borealis. lower Eocene; B, Manteoceras mantcoceras, middle Eocene; Side view. Modeled and drawn by E. S. Christman under the direction of William K. Gregory and H. P. 

C, Proiitanotlierium emarginatum, upper Eocene; D, Brontotherium Osborn. H, Eotitanops borealis, lower Eocene; G, Manteoceras manteoceras, middle Eocene; F, Protilano- 

platyceras, lower Oligocene. One-ninth natural size. therium emarginatum, upper Eocene; E, Brontotherium platyceras, lower Oligocene. One-twelfth natural size. 



712 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



capitis, and below this in turn, near tlie lower end of 
the mastoid process, was the insertion of the tendon 
of the cephalohumeralis (cleidomastoideus). Beneath 
the semispinalis and splenius is a wide nuchal surface 
for the obhquus capitis superior, rectus posticus major, 
and rectus posticus minor. The relations of all 
these muscles to the cervical vertebrae are shown in 
Plate XLIII. 

The rugose tips of the dorsal and cervical spines 
served for the origin of the ligamentum supraspinale 
and the ligamentum nuchae, the latter being inserted 
into the tips of the cervical vertebrae and into the 



to the longissimus atlantis. Both had short branches 
arising from the outer surface of the prezygapophyses 
of successive vertebrae. Below the zygapophyses and 
above the transverse processes were the intertransver- 
sarii muscles, which also extended from one transverse 
process to the next. The outer rugosity of the trans- 
verse processes also bore the slips of the scalenus brevis 
(primae costae), levator scapulae ("serratus cervicis"), 
and cephalohumeralis. 

The infero-anterior processes of the cervical verte- 
brae gave rise to slips of the longus colli, which, joining 
with similar slips from the inferior faces of the centra. 




Figure 650. — Occiput of an Oligocene titanothere {Brontops robustus) , showing the probable general location 

and planes of the neck muscles 

The extremely rough occipital and lambdoidal crests were probably covered with tough skin, as in recent perissodactyls. The precise limits 
of the insertion areas can not be determined. Two-ninths natural size. 



middle of the occiput (iig. 650). On the anterior edges 
of the neural spines of the cervicals were inserted the 
shps of the spinalis dorsi (cervicis). The sides of the 
cervical spines carried the multifidus spinae, which 
ran obliquely downward and backward to be inserted 
above the posterior zygapophyses of the next succeed- 
ing vertebrae. The outer dorsal rims of the post- 
zygapophyses bore the slips of the semispinalis capitis, 
which ran upward and forward, converging toward 
and being inserted on to the nuchal surface, just 
below the occipital crest and on either side of the 
midline (fig. 650). Along the line of the prezyga- 
pophyses and postzygapophyses ran both the longissi- 
mus capitis, which was inserted into the posterolateral 
margin of the lambdoidal crest behind the auditory 
meatus, and the longissimus atlantis, which was 
inserted into the edge of the transverse process of the 
atlas. The longissimus capitis lay immediately dorsal 



ran forward and were inserted on the hypapophysis of 
the atlas. 

The principal attachments of the muscles and liga- 
ments of the presacral vertebrae are summarized 
below. It is assumed that every homologous part in 
the horse and the titanothere would carry a homo- 
logous muscle and ligament attachment. The very 
full data for such attachments in the modern horse 
given by Schmaltz (1909.1) accordingly form the 
chief basis of the determinations given. 

Principal attachmenis of the muscles and ligaments in the pre- 
sacral vertebrae in the horse and the titanothere 

Muscle attachments of the atlas. (See PI. XLm) 

Neural spine (s) : rectus capitis posticus minor. 
Transverse process (t), dorsal surface: obliquus capitis inferior. 
Transverse process (t) , ventral surface : obliquus capitis superior. 
Transverse process (t), e.xternal border: obliquus capitis 
superior. 



THE MUSCULAR ANATOMY AND THE RESTORATION OF THE TITANOTHERES 



713 



Transverse process (t), posterior border: longissimus atlantis. 
Inferior median spine: longus colli. 

Inferior arch, ventral surface: rectus capitis anticus lateralis, 
rectus capitis anticus minor. 

Ligament attachments of the atlas 

Inferior arch: ligamentum occipito-atloideum ventrale. 
Superior arch: Mgamentum occipito-atloideum dorsale. 

Muscle attachments of the axis. (See PL XLIII) 

Neural spine, median line: rectus capitis posticus major et 

medius. 
Neural spine, lateral surface: obliquus capitis inferior. 
Neural spine, lateral surface, posteriorly: multifidus spinae. 
Postzygapophysis, dorsal surface: semispinalis capitis (com- 

plexus) . 
Base of neural arch, side of centrum, and posterior portion of 

transverse process : intertransversales. 
Transverse process, posteriorly: oleidomastoideus. 
Inferior median keel, laterally: longus colli. 

Ligament attachments of the axis 

Neural crest, posteriorly: ligamentum nuchae. 

Neural spine, anteriorly: ligamentum interspinale. 

Neural arch, anterior border: ligamentum atlo-axoideum dorsale. 

Odontoid process, anteriorly: ligamentum apicis dentis. 

Centrum, inferiorly: ligamentum atlo-axoideum ventrale. 

Muscle attachments of cervical vertebrae 

Neural spine, anterolaterally : spinalis dorsi et cervicis. 

Neural spine, posterolaterally: multifidus spinae. 

Prezygapophysis and postzygapophysis, superficially: semii- 
spinalis capitis (complexus). 

Prezj'gapophj'sis and postzygapophysis, middle layer: longis- 
simus capitis et longus atlantis. 

Prezj'gapophysis and postzj'gapophysis, deep laj'er, superior: 
multifidus spinae. 

Prezygapophysis and postzygapophysis, deep layer, inferior: 
intertransversales. 

Pedicle of neural arch above transverse process: intertrans- 
versales. 

Transverse process, antero-inferior ventral surface: longus colli. 

Transverse process, posterosuperior: intertransversales, lon- 
gissimus cervicis, scalenus costarum, serratus anterior, rectus 
capitis anticus major, oleidomastoideus, splenius capitis. 

Inferior spine, laterally: longus colli. 

Ligament attachments of cervical vertebrae 

Neural spine, tip: ligamentum nuchae. 

Neural spine, anterior median : ligamentum interspinale. 

Neural spine, posterior median : ligamentum interspinale. 

Neural arch, laterosuperiorly : ligamentum subflavum. 

Centra, dorsaUy, forming floor of neural canal: ligamentum 

commune dorsale. 
Centra, ventraUy: ligamentum commune ventrale. 

Muscle attachments of second dorsal vertebra. (See fig. 651) 

Neural spines, near summit, laterally: splenius capitis, semi- 
spinalis capitis (complexus). 

Neural spines, anterolaterally: spinalis dorsi et cervicis. 

Neural spines, posterolaterally, upper part: spinalis dorsi. 

Neural spines, posterolaterally, lower part: multifidus spinae. 

Transverse process, dorso-externally : semispinalis capitis, 
multifidus spinae, longissimus cervicis, longissimus dorsi, 
levator costae. 

Ligament attachments of second dorsal vertebra 

Similar to those of the cervical vertebrae (see above) ; also on 
tip of neural spine: ligamentum supraspinale 



Muscle attachnents of tenth dorsal vertebra. 

Neural spine, near tip : spinalis dorsi or longissimus dorsi. 
Neural spine, posterolaterally: multifidus spinae. 
Transverse process, dorsal: multifidus spinae, longissimus 
dorsi, levator costae. 




Figure 651. — Probable position and attachments of the | 
principal muscles of the second and third dorsal verte- 
brae of a lower Oligooene titanothere (Broniops cf. B. 
robustus) 

Insertion areas in thinner black lines; directions of muscles in heavy blacli 
lines. One-eighth natural size. 

Ligament attachments of tenth dorsal vertebra 

[See PI. CXCIX] 
Muscle attachments of second lumbar vertebra 

Neural spine, on either side of tip: longissimus dorsi. 
Neural spine, posterolaterally: multifidus spinae. 
Prezygapophysis, antero-externally : multifidus spinae. 
Side of centrum concealed by psoas minor. 

Ligament attachments of second lumbar vertebra 

Chief ligaments as in preceding vertebrae; also, to postero- 
external corner of centrum: ligamentum iliolumbale. 



714 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



SECTION 4. MUSCLES OF THE LIMBS OF THE EOCENE 
TITANOTHERE PALAEOSYOPS (SUBGRAVIPORTAL 
TYPE) 

The limb bones of the middle Eocene titanothere 
Palaeosyops are so much like those of existing tapirs 
(Tapirus) that it is highly probable that the positions 
and courses of the muscles were also essentially smiilar, 
the differences being chiefly due to the more robust 
development of certain muscles ia Palaeosyops. Both 
animals are tetradactyl-tridactyl. Tapirus is typically 
mediportal; Palaeosyops is subgraviportal. 

In the following study Murie's description (1872.1) 
of the myology of the tapu- and Windle and Parson's 
studies (1901.1; 1903.1) on the muscles of the ungu- 
lates were constantly used, and comparisons were 
made with the myology of other mammals, especially 
the horse, as figured by Schmaltz (1909.1). 




tendon of the trapezius, which covered the 
shoulder muscles of the neck and back. To the lower 
border of the spine was attached a slip of the deltoid, 
which lay outside the broad infraspinatus muscle. 

The infraspinatus muscle passed downward below 
the spine and above the teres minor, which in turn was 
probably attached to the ridge for the teres minor, 
obliquely traversing the lower third of the postspinous 
fossa. Above and behind this streak, near the angle 
of the scapula, is a rough streak which is probably for a 
second slip of the deltoid. 

Along the axillary border and running below into a 
rugose axillary prominence, as well as into a curved 
area behind the ridge for the teres minor, is the origin 
of the long head of the triceps, the largest muscle of 
the forearm. The posterior angle of the scapula is 
roughened for the teres major. 



Serraius cervicis 
[Levaior- scan.) 



PecioraliS 
profundus 



Tr-apezius 

Pectoralis profundus 
Delioideus 



■Biceps 

Coraco brach-ial, 



Cap-ui 
lonp'um. 
Triceps 



Figure 652. — Right scapula of a lower Oligocene titanothere (Broniotherium cf. B. gigas), showing the principal 

muscle attachments 
A', inner side; A', outer side. (Compare flg. 623.) 



MUSCIES OF THE FOKE LIMB 

[See PI. XLIV, A] 
MUSCULAR ATTACHMENTS OF THE SCAPULA 

The large scapula is about •j-Vo" ^-s long as the 
humerus and afforded a broad base for the massive 
shoulder muscles. The superior border is roughened 
for the rhomboideus; the anterosuperior continuation 
of this roughening may have given attachment to the 
serratus cervicis (levator scapulae) subclavius and, in 
its antero-inferior part, to the pectoralis profundus 
(pars ascendens). 

The coracoid process gave origin anteriorly to the 
single-headed biceps and inferiorly to the coracobra- 
chialis. 

The supraspinous fossa and its muscle area for the 
supraspinatus are broad and well defined, and the same 
is true of the postspinous fossa, which is the area 
chiefly for the infraspinatus. The spine of the scapula 
is reflected and bears a'prominent tuberosity for the 



The medial or internal surface of the scapula in its 
upper third bears a large triangular roughening for the 
serratus magnus; below this nearly the whole surface 
was filled by the subscapularis. 

Near the axillary border the medial surface is raised 
into a long, gently convex eminence or column running 
parallel and internal to the postspinous fossa and 
culminating above in the surface for the serratus. 
This subscapular column greatly strengthens the blade 
against the pull of the serratus, infraspinatus, sub- 
scapularis, and caput longum of the triceps. 

MUSCULAR ATTACHMENTS OP THE HUMERUS 

The great tuberosity, expanded into a broad flange, 
gives insertion on its anterior hook to the supraspina- 
tus; a branch of this muscle forks over the biceps and 
is inserted into the lesser tuberosity. The superior and 
external surfaces of the great tuberosity form the 
insertion area for the infraspinatus. The bicipital 
groove is filled by the single-headed biceps. 



THE MXJSCULAR ANATOMY AND THE RESTORATION OF THE TITANOTHERES 



715 



The high deltopectoral crest gives attachment to the 
teres minor and below this is produced into a large 
reflected protuberance for the deltoid. The rough 
anterior border of the deltopectoral crest probably 
gave origin for a branch of the cephalohumeralis. 
Windle and Parsons (1901.1, p. 679) state that this 
muscle is a combination of the anterior part of the 
trapezius, the cleidomastoid, and the clavicular part 
of the deltoid. 

The lower extension of the deltopectoral crest very 
probably gave attachment to the pectoralis super- 
ficialis (p. major) and below this to the cephalohume- 
ralis. The posterior upper border of the deltopectoral 
crest probably gave attachment to the external head 
of the triceps. 



behind and below it was attached the medial head of 
the triceps. 

From the entocondyle arose the medial or humero- 
radial collateral ligament and the several flexors of 
the forearm. 

Above the olecranal fossa is an area which was 
probably for the so-called anconeus; this muscle in 
ungulates appears to represent a fourth head of the 
triceps. 

MUSCULAR ATTACHMENTS OF THE RADIUS 

The head of the radius was flattened and appressed 
to the ulna, and the radio-ulnar facets permitted only 
a very slight degree of supination. In correlation 
with this the tubercle (fig. 654, tub.) was greatly 
reduced and no longer served as the chief insertion 




Figure 653. — Left humerus of Broniops robusius (type), showing probable position and attachments of 

principal muscles 
Ai Front view; Ar, back view; A3, outer side view; Ai, inner side view. One- eightli natural size. 



The extensive winding surface for the brachialis 
anticus on the back of the humerus passes from above 
downward and forward; it lies behind the deltoid 
crest. Below the brachialis anticus area and above 
the supinator crest arose the powerful supinator 
longus and extensor carpi radialis. 

From the region of the ectocondyle arose the humeral 
slip of the extensor communis digitorum, the external 
collateral ligament, and the extensor carpi ulnaris, 
which in ungulates serves as a flexor of the forearm. 

On the inner or costal side of the humerus we observe 
especially the attachment for the subscapularis (into 
the lesser tuberosity) and the latissimus rugosity for 
the common tendon of the teres major and latissimus 
dorsi. Below and in front of the latissimus rugosity 
is the insertion area for the coracobrachialis, while 



point for the tendon of the biceps; this tendon was 
probably inserted chiefly on the front of the head .of 
the radius and may also have extended around to 
the inner side, as in Tapirus and Equus. 

On the inner or medial surface of the radius about 
2 inches below the proximal end is a rugose area which 
is probably for the insertion of the medial collateral 
ligament. Below this, in the middle third of the 
shaft and partly on the front surface, is a vertically 
extended rugosity which was probably, as in Rhino- 
ceros, the attachment area of the brachialis anticus. 
At first sight the rugosity in question seemed to be for 
the pronator radii teres, as in carnivores, but in the 
tapir and apparently also in Palaeosyops the tendon 
of this muscle extended down the whole inner edge of 
the radius. (Murie, 1872.1; Turner, 1850.1.) 



716 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



MUSCULAR ATTACHMENTS OF THE ULNA 

The massive olecranon has a very rough area for the 
long head of the triceps; to the olecranon were also 
attached the external (lateral) and medial (internal) 




Figure 654. — Left Tudius oi Bronlops robustus (type), showing probable position 

and attacliments of principal muscles 

Ai, Front view; As, back view; As, inner side view; Ai, outer side view. One oiglitli natural size. 

heads of the triceps, also the fourth head, or "an- 
coneus." The inner side of the olecranon probably 
gave attachment to the palmaris longus and to branches 
of the flexor digitorum profundus and flexor carpi 
ulnaris. 

The antero-external face of the ulna, together with 
the adjacent side of the radius, probably lodged 
the extensor ossis metacarpi pollicis (extensor meta- 
carpi obliquus), and the rough outer border of the 
ulna may have been for a slip of the extensor com- 
munis digitorum. 

The lower end of the ulna, closely appressed to the 
radius, retained but little independent movement, 
yet more than in the tapir. To the outer side of the 
shaft near the lower end was attached the lateral 
ulnocarpal ligament. Possibly the groove near the 
end of the ulna, above the cuneiform, lodged the 
tendon of the extensor digitorum lateralis. 

On the antero-internal edge (or interosseous ridge) 
of the ulna was stretched the strong interosseous 
membrane which separated the extensor from the 
flexor surfaces of the forearm. 

MUSCULAR ATTACHMENTS OF THE CARPUS 



rows moved apart vertically but preserved their 

transverse alinement. 

On the posterior aspect of the carpus the pisiform 

gave insertion superiorly to the flexor carpi ulnaris and 
possibly also to a slip of the extensor 
carpi ulnaris, inferiorly to the abductor 
minimi digiti; it was also no doubt em- 
braced by the annular ligament. The ex- 
tensor carpi ulnaris, according to Windle 
and Parsons (1901.1, p. 700), in most 
ungulates serves as a powerful flexor of 
the carpus. According to these authori- 
ties the slip attached to the pisiform 
(which appears to the present writer to 
belong rather to the true flexor carpi 
ulnaris) is present in Equus and many 
artiodactyls but is barely or not at all 
developed in Tapirus, Rhinoceros, Hyrax, 
Elei^has, Canis, Homo. Hence this 
pisiform slip may also have been absent 
in PalaeosyojJS. 

The tuberosities on the posterior aspect 
of the carpus gave attachment for a set 
of ligaments radiating from the middle 
bones (lunar, magnum) to the surrounding 
elements. The posterior tuberosity of the 

lunar also gave attachment to the interosseus dorsalis 

primus, abductor^minimi digiti, and adductor minimi 

ccip -yried. 



The proximal row of carpals, like those of Equus, 
was bound together transversely by a series of liga- 
ments running across the front face of the carpus and 
connecting respectively with the inner and outer distal 
portions of the radius. The distal row of carpals 
was similarly bound together by a second series of 
ligaments so that during flexion the proximal and distal 




Figure 655. — Left ulna of Brontops robustus (type), showing 

probable position and attachments of principal muscles 

Ai, Front view; A2, inner side view; As, baclc view. One-eightli natural size. 

digiti. The posterior tuberosity of the magnum, the 
form of which is very characteristic of perissodactyls, 
gave attachment not oidy to the radiating ligaments 



THE MUSCULAR ANATOMY AND THE RESTORATION OE THE TITAN OTHERES 



717 



above mentioned, but also to the adductor minimi 
digiti, the adductor indicis, and the adductores digi- 
torum medii et quarti. The spaces between the proxi- 
mal ends of adjacent metacarpals gave rise to the dorsal 
interossei, while the depressions on the posterior aspect 
of the proximal portion of the metacarpals served for 
the origin of the palmar interossei which were inserted 
■ into the sesamoids, the latter being attached by liga- 
ments to the proximal phalanges. The inner and 
outer sides of the metacarpals show distally deep 
lateral pits, one on either side of the facets for the 
phalanges; these pits were for the collateral ligaments 
that passed to the first row of phalanges and were 
inserted on their proximal inner and outer expansions. 

The palmar surfaces of the first row of phalanges 
were covered by the flexor digitorum profundus. The 
lateral protuberances at the proximal ends probably 
served for the collateral ligaments, while pairs of 
proximal palmar rugosities probably served for the 
ligaments of the sesamoids. At the sides of the 
proximal phalanges may have been attached the ad- 
ductors of the digits and the lumbricales (accessory 
flexors arising from the tendon of the flexor sublimis). 
Toward the distal end the lateral pits were probably 
for the collateral ligaments, whUe the protuberances 
above them probably gave attachment to the flexor 
sublimus. 

The second row of phalanges (fig. 656) gave insertion 
to the forked flexor profundus in the midpalmar surface. 
The dorsal surface of the second phalanges gave in- 
sertion to the extensors of the digits. The distal or 
ungual phalanges gave insertion on their dorsal sur- 
faces to the extensors and on their palmar surfaces to 
the flexor sublimis. The distal roughening for the 
hoof indicates a relatively small hoof in comparison 
with the tapir and rhinoceros. 

The whole palmar surface of the manus was prob- 
ably covered superficially by the flexor brevis manus, 
as in the tapir. 

MUSCLES OF THE HIND IIMB 

[See PI. XLIV, B] 
MUSCULAR ATTACHMENTS OF THE SACRUM 

The sacrum was securely fastened to the flium by a 
system of iliosacral ligaments probably similar in 
essentials to those in Equus as figured by Schmaltz 
(1909.1, Taf. 6) and consisting of (1) a dorsal ligament 
(called by Schmaltz the ligamentum sacro-iliacum 
dorsale breve) running along the tips of the sacral 
spines and being inserted anteriorly into the dorsal 
border of the sacral flange of the ilium; (2) a broad 
ligament running from the lateral edge of the trans- 
verse lamina of the sacrum obhquely upward and 
forward to the posterior border of the sacral flange 
of the ilium (ligamentum sacro-iliacum dorsale 



longum, Schmaltz); (3) a broad sheet of ligament 
extending between the transverse lamina of the 
sacrum, obliquely downward and backward to the 
dorsal border of the ilium above the acetabulum 
(spina ischiadicum) and to the dorsal border of the 
ischium as far back as the ischial tuberosity (liga- 
mentum sacro-spinosum et tuberosum. Schmaltz) ; (4) 
a number of oblique ligaments springing from the 
anterior dorsal edge of the sacral flange of the ilium 
and running to the spines of the lumbar vertebrae. 
The hgaments above described are of great impor- 
tance both in tying the pelvis securely in place and in 




Figure 656. — Arrangement of the muscles of the flexor 
surface of the manus of Mesatirhinus peiersoni 

Restoration of muscles based chiefly on dissection of the forefoot of a tapir 
by Murie after the removal of the palmaris longus, lumbricales, and 
parts of the flexor tendons (cut at x). One-third natural size. 

affording origin to several of the great muscles of the 
loins, croup, and buttocks, as follows: No. 4 (above) 
affords partial insertion to the longissimus dorsi; Nos. 
1, 2, 3 conjointly afford origin for the gluteus medius, 
gluteus maximus, and "biceps 1" (femorococcygeus) . 

MUSCULAR ATTACHMENTS OF THE PELVIS 

The pelvis of Palaeosyops is of the subgraviportal 
type in which the ilium spreads widely and has its 
superior border well rounded out rather than concave. 
In correlation with this adaptation the area for the 
gluteus medius on the dorsum of the ilium and that 
for the iliacus on the venter of the ilium are both 



718 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



wide ; it is also highly probable that the gluteus medius 
did not extend over the dorsal edge of the ilium as 
it does in cursorial animals and therefore was not 
inserted into the fascia of the longissimus dorsi 
(Gregory, 1912.1, p. 292). The superior border or 
crest of the ilium gave attachment to the superficial 
fibers of the longissimus dorsi and also to the obliquus 
abdominis externus and obliquus abdominis internus; 
the outer part of the superior border gave origin to 
the tensor fasciae femoris and to the inguinal ligament. 
The gluteus maximus may possibly have barely 
touched this border. The area for the gluteus mini- 
mus immediately above the acetabulum is roughened 
and streaked, probably indicating that the muscle 
was divided into a number of twisted and parallel 
strands as in Equus. Immediately in front of the 
acetabulum is a deep pit for the rectus femoris (the 
latter, according to Windle and Parsons, homologous 
with the reflected head of the rectus in man). On 
the neck of the ilium, on the antero-internal surface, 
is a pit for the psoas parvus; below this pit the sartorius 
probably took rise from the fascia of the ilio-psoas. 

The antero-internal surface of the ilium (corre- 
sponding to the iliac fossa of man) lodges the wide 
iliacus; between the area for the iliacus and the pit 
for the psoas parvus is a gently depressed surface 
homologous with a similar region in the horse and 
likewise bearing a nutrient foramen (arteria circum- 
fiexa femoris lateralis. Schmaltz). Dorsomedial to 
the iliac fossa and below the articular surface for the 
sacrum the ilium exhibits a flattened elongate surface 
which as in the horse and tapir probably lodged an 
anterodorsal extension of the obturator internus. . 

The lower part of the pelvis (that is, the pubis, 
ischium, and acetabular borders) in a general way 
gives origin to two principal groups of locomotor 
muscles — to the abductors of the femur, which are 
inserted either in the digital (trochanteric) fossa or 
immediately below it, and to certain flexors of the 
femur — the ischial head of the biceps femoris ("biceps 
2"), the semitendinosus, and the semimembranosus. 
Immediately behind the ischial spine and the area for 
the gluteus minimus the superior ramus of the ischium 
is smoothly rounded, this indicating the place where 
the obdurator internus came over the ridge on its 
way toward the femur. Immediately below this 
place, above and behind the acetabulum, is a small 
roughened area which probably indicates the position 
of the gemelli (PI. XLIV). 

The ventral border of the symphysis pubis probably 
gave origin to the closely intermingling adductor 
magnus and gracilis, while the lower borders of the 
thyroid (obturator) foramen, as well as parts of the 
symphysis, gave origin to the obturator externus. 

Near the posterior end of the ischium and on or 
near the dorsal external border is a long, low promi- 



nence for the true biceps femoris (caput longum of 
Homo); this contrasts with that of Equus, which has 
grown outward and downward into a great horizontal 
flange (the crista tuberis ischii, Schmaltz). Although 
the bicipital prominence in Palaeosyops and the 
bicipital flange in Equus are so different in appear- 
ance that we do not at first recognize their homology 
with each other, yet they are clearly homologous, 
because the rhinoceroses Hyrachyus and Caenopus, 
EoJiippus and other extinct perissodactyls, as well as 
Tapirus, afford intermediate conditions. Cuvier and 
Laurillard's plates also show that in many mammals 
the long head of the biceps femoris arises from the 
same region we have assigned to it in Palaeosyops. 
Immediately below the prominence for the biceps 
femoris is the probable location of the ischial head of 
the semitendinosus, while below and in front of the 
latter is the probable origin of the quadratus femoris. 
The region of the ischial tuberosity very probably 
served for the ischial head of the semimembranosus. 

The anterior ramus of the pubis probably gave 
origin to the pectineus; on its anterior border there 
is little if any iliopectineal eminence, the latter being 
very prominent in Tapirus and Equus and serving in 
part for the origin of the pectineus (Schmaltz, 1909.1, 
Taf. 7, 51). 

The obturator (thyroid) foramen on its inner side 
and around its border was probably filled by the 
obturator internus, which also sent forward and 
upward a -branch lying above and medial to the area 
for the iliacus. By comparison with Equus and Tapi- 
rus it seems very probable that in Palaeosyops the 
obturator internus arose inside the true pelvis, as in 
the Perissodactyla generally as well as in the Hyra- 
coidea and the Proboscidea, whereas in the Artio- 
dactyla it arises outside the pelvis (Windle and 
Parsons). 

MUSCULAR ATTACHMENTS OF THE FEMUR 

The femur as a whole in Palaeosyops and still more 
in Oligocene titanotheres is much more elephantine 
than horselike. The shaft is flattened, the patellar 
trochlea faces more anteriorly, the limb is straighter 
at the knee. The long flattened subvertical femur of 
the graviportal elephant is inclosed in front and at 
the side by a long vastus externus, of great transverse 
diameter; the short round femur of Equus and of the 
kangaroo is covered by a balloon-shaped vastus exter- 
nus of great anteroposterior diameter. The femur of 
Palaeosyops has the appearance of having been flat- 
tened by the pressure of the vastus externus in front 
and by the lateral strain of the great gluteus maxi- 
mus along the outer side. The long and transversely 
wide vastus externus of the femur contributes to the 
great weight-lifting power and feeble jumping or 
saltatorial power of the limb; while the balloon- 



THE MUSCULAR ANATOMY AND THE RESTORATION OF THE TITANOTHERES 



719 



shaped vastus of the horse and kangaroo contributes 
to the far greater saltatorial power. 

On the front face of the femur the massive but low 
great trochanter gave attachment to the gluteus 
medius. The region between the great trochanter and 
the head is widened and exhibits laterally the upper 
limits of the area for the vastus externus; in the mid- 
line is a roughened line separating the area of the 
vastus externus from that of the vastus internus, as 
in Tapirus. The area for the vastus externus (which 
in Tapirus was probably a very large muscle covering 
the entire length of the shaft of the femur) is 
not visibly separated below from the area for the 
crureus, which probably covered 
the front and lower outer part of 
the shaft. 

Outer edge of the femur: The 
great trochanter bears on its upper 
and external faces the area for 
the gluteus medius. The third 
trochanter bounds externally the 
area for the vastus externus and 
on its outer crest gives insertion 
to the deep slip of the gluteus 
maximus; the low ridge leading 
down from the third trochanter 
to the shaft may also have given 
insertion to fibers of the gluteus 
maximus. 

About 3 inches above the 
ectocondyle is the ridge for the 
external head of the gastroc- 
nemius. Immediately below the 
gastrocnemius ridge is the deep 
plantar fossa facing postero- 
externally and giving origin to 
the tendon of the plantaris (flexor 
sublimis). 

The external tuberosity of the 
femur (above the ectocondyle) 
bears on its outer face a deep pit 
for the collateral femoro tibial 
ligament. Lying between the 
last-named pit and the patellar 
trochlea is the pit for the tendon of the extensor 
longus digitorum, while behind this in turn and 
above the condylar surface is a circular deep pit 
for the tendon of the popliteus. (Possibly also the 
femoral portion of the peroneus longus may have 
arisen here. Compare Windle and Parsons, 1901.1, 
p. 278.) 

The posterior surface of the femur shows especially 
the following structures: (1) The great trochanter, 
bearing the posterior portion of the area for the gluteus 
medius; (2) the digital or trochanteric fossa, lodging 
the obturator internus, gemeUi, and obturator externus ; 
the quadratus femoris may have been inserted below 



these; (3) the second trochanter (trochanter minor), 
giving insertion both to the psoas and to the iliacus; 

(4) the third trochanter, for the gluteus maximus; 

(5) the area for the pectineus, lying toward the inner 
border and some distance below the second trochanter. 
A foramen for a branch of the femoral artery lies below 
the last-named area and in about the middle of the 
internal posterior edge of the femur. (6) The linea 
aspera (plainly shown in the femur, Am. Mus. 11690) 
lies in the middle third of the posterior face of the 
femur near the internal border and below the area for 
the pectineus; in this region bounded by the linea 
aspera internally and by the downward prolongation 




A3 
FiGUKE 657. — Left femur of Broniops robushis (type), showing probable position and 
attachments of principal muscles 
Ai, Front view; A2, inner side view; Aj, back view. One-eighth natui-al size. 

of the third trochanter the adductor longus was prob- 
ably inserted. 

Inner face of the femur: The principal topographic 
features of the inner face of the femur are as follows: 
(1) The pit for the ligamentum teres; (2) the promi- 
nent second trochanter, roughened for the attachment 
of the iliacus and psoas; (3) the area for the vastus 
internus along the inner shaft; (4) the foramen for the 
femoral artery; (5) near the lower end the area for 
the internal head of the gastrocnemius; (6) above the 
internal condyle a vertical ridge for the adductor and 
semimembranosus; (7) below this the pit for the liga- 
mentum collaterale' laterale. 



720 



TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA 



MUSCULAR ATTACHMENTS OF THE PATELLA 

The more or less pear-shaped patella (morphologi- 
cally the sesamoid of the quadriceps femoris) is on the 
whole more elephantine than horselike, in correlation 
with the straighter limb. It differs from that of 
Tapirus in having the dorsal prominence for the 
rectus sessile instead of hooklike; the upper surface is 
anteroposteriorly deep and fiat; the lower end for the 




lexlj 



Ai 



A2 



A3 



(AJi.) 



A4 



<^^, 




Figure 658. — Left fibula and left tibia of Broniops robustus (type), showing 
probable position and attachments of principal muscles 



Ai, Fibula, outer side; Ai, inner side; A3, rear view; A 
view; B3, outer side; Bi, inner side. 



I, front view. Bi, Tibia, front view; B 
One-eightb natural size. 



ligament of the patella is prolonged and pointed; 
the inner side is not produced. Correlated with the 
latter structure is the small size of the inner patellar 
keel on the femur, this keel not being globose above, 
as it is in Equus and to a less extent in Tapirus. All 
this implies that, unlike Equus, Pdlaeosyops could not 
rest the patella on top of the inner patellar keel of the 
femur but stood with straighter limb. 



MUSCULAR ATTACHMENTS OF THE TIBIA 

The tibia is rather similar to that of Tapirus. On 
the cnemial crest the surface for the ligamentum 
patellae (which was probably tripartite, as in the 
horse) faces more upward than in Tapirus, in correla- 
tion with the straighter limb, but is otherwise similar. 
The cnemial crest is somewhat less reflected than 
it is in Tapirus. It terminates above in a supero- 
external tuberosity, which not only 
gave origin to a part of the liga- 
mentum patellae but probably also 
gave insertion to some of the deeper 
strands of the biceps femoris (ischial 
head). Immediately postero-external 
to this bicipital tuberosity is a large, 
deep vertical groove for the tendon 
of the extensor longus digitorum. 
The tendon is attached to the femur 
immediately above the groove. 

Immediately below the extensor 
groove and behind the cnemial crest 
is the fossa for the tibialis anticus, 
covering the inner face of the tibia 
but not defined below and nowhere 
as sharply defined as in Tapirus. 
The front of the cnemial crest served 
chiefly for the inferior prolongation 
of the ligamentum patellae, but the 
lower end of the crest bears a promi- 
nent scar, into which was inserted a 
deep strand of the semitendinosus 
which passed across the inner face 
of the tibia on its way to the fascia 
of the shank. 

On the inner surface of the tibia, 
behind the cnemial crest, the area, 
for the gracilis is not defined, but 
no doubt that muscle was inserted 
in this region. The inner tuber- 
osity, a low prominence behind the 
middle line on the proximal border, 
together with the pit behind it, pos- 
sibly gave insertion to a deep slip of 
the semimembranosus and also to the 
ligamentum collaterale tibiale. 

Most of the inner or medial sur- 
face of the tibia was free of muscular 
attachments (planum subcutaneum. 
Schmaltz). Distally the inner face 
bears near the front border a sharply flattened promi- 
nence for the ligamentum collaterale breve running 
obliquely to the astragalus and calcaneum. Some dis- 
tance behind this prominence is a vertical groove for the 
flexor longus digitorum. The back of the tibia below 
the popliteal notch is taken up by a wide and deep 
fossa for the popliteus. This muscle may also have 
extended around on to the inner surface of the tibia. 



THE MUSCULAR ANATOMY AND THE RESTORATION OP THE TITANOTHERES 



721 



but only to a slight extent. The popliteus area is 
bounded below by the indistinct popliteal or oblique 
line, here indicated chiefly by a low prominence 
extending from the upper external edge obliquely 
downward and inward across the middle third of the 
back of the shaft. To this oblique line may very 
possibly have been attached remnants of the soleus 
(which in Tapirus has lost its sheathlike character), 
a muscle degenerated into an elastic band and shifted 
in origin upon the popliteal surface of the femur 
(Murie). 

The lower third of the back of the shaft terminates 
below in a posterodistal process which in the standing 
pose of the horse juts down between the back of the 
astragalus and the sustentaculum of the calcaneum. 
Postero-externally to this process the flexor longus 
hallucis and probably the tibialis posticus passed 
downward. 

The flexor longus hallucis and the flexor longus 
digitorum have been confused by certain authors. 
In man the flexor longus digitorum passes through 
the groove behind the internal malleolus of the tibia, 
while the flexor longus hallucis passes down behind 
the sustentaculum of the calcaneum (Cunningham ; 
1903.1, pp. 229, 236, 178, fig. 170), and the same 
is true of the horse (Schmaltz, 1909.1, Taf. 56, fig. 1); 
and yet Murie mistakenly calls the muscle which 
"glidesln the groove behind the malleolus of the tibia " 
"flexor longus hallucis," while the tendon which lies 
"behind the os calcis" he calls "flexor longus digi- 
torum," thus reversing the implied homologies with 
man, as very clearly shown in his figures. Even 
Windle and Parsons seem to be wrong, for they 
homologize with the flexor longus hallucis the muscle 
which "winds round the internal malleolus" (1903.1, 
p. 281). The flexor longus hallucis in the tapir joins 
the flexor longus digitorum, the latter forming the 
perforating tendons (flexor profundus) of the three 
digits. The tibialis posticus is not recognizable in 
the tapir (Murie, 1872.1, p. 166) nor in some artio- 
dactyls and Hyrax (Windle and Parsons, 1903.1, p. 
282), but is probably fused with the flexor longus 
hallucis. In the horse it is large and distinct 
(Schmaltz, 1909.1, Taf. 59, 53, 56). In the Sumatran 
rhinoceros it is fused with the flexor longus hallucis 
and flexor longus digitorum, these three together form- 
ing the flexor communis digitorum (Beddard, 1889.1, 
p. 22). In the elephant it is distinct. It seems not 
unlikely that in Eocene graviportal ungulates such as 
Palaeosyops the muscle was present and more or less 
appressed to the flexor longus hallucis. It may then 
have had an area on the back of the tibia just below 
the popliteal line and possibly another on the back 
of the fibula, behind the flexor longus hallucis. 

The soleus is another muscle of somewhat doubtful 
form and attachments. In man this broad muscle 



lies beneath the gastrocnemius and is broadly attached 
to the back of the flbula and to the popliteal or oblique 
line of the tibia; it is inserted into the tendo Achillis 
and assists in lifting the heel. In many ungulates, 
however, the soleus is of small size (Windle and 
Parsons, 1903.1, p. 280) and more or less cordlike; it 
probably contributes by its elasticity to the quick 
backward jerk or kick of the foot. In the tapir the 
soleus is very small and cordlike; it has shifted its 
origin up to the femur, in the popliteal space, behind 
the gastrocnemius area, and is here closely associated 
with the cordlike tendinous plantaris (Murie, 1872.1, 
p. 163 and pi. 10, fig. 13). In this extreme modifica- 
tion of the soleus the tapir is more specialized than 
either the horse, in which a slender belly of the soleus 
remains attached to the tibia behind the head of the 
fibula, or the rhinoceros, in which as figured by 
Beddard (1889.1, p. 21) the soleus is fleshy and 
apparently arises below the popliteus. In the ele- 
phant the soleus is well developed. It therefore seems 
likely that Palaeosyops and the Oligocene titanotheres 
avoided the extreme degeneration of the soleus which 
is characteristic of the tapir and retained a fleshy 
soleus, attached to the oblique or popliteal line of 
the tibia and possibly to the back of the upper part of 
the fibula. 

MUSCULAR ATTACHMENT OF THE FIBULA 

As compared with that of Tapirus the fibula of 
Palaeosyops has both proximal and distal ends larger 
and not so flat; the middle of the shaft is flatter. To 
the outer side of the head of the fibula was attached 
the peroneus longus, which extended down the front 
outer face of the shaft and passed to the foot through 
a deep peroneal groove. The top of the fibula (caput 
fibulae) behind the surface for articulation with the 
tibia shows an expanded smooth prominence, postero- 
internal to which was a groove; over both prominence 
and groove slid the tendon of the popliteus. Below 
the popliteal prominence the flexor longus hallucis 
probably took up most of the posterior surface of the 
shaft. The inner or medial face of the fibula was also 
largely fifled by the flexor longus hallucis. The front 
apex of the external malleolus was probably attached 
to the "ligamentum collaterale longum" running to 
the calcaneum. 

MUSCULAR ATTACHMENTS OF THE PES 

The distal rugosity of the tuber calcis served as the 
insertion area for the foUowing tendons: (1) The true 
tendo Achillis, which consists of the paired tendons of 
the gastrocnemius; (2) the complex of tendinous 
sheaths which descend from the biceps, gracilis, and 
semitendinosus and enwrap the tendon of the gastroc- 
nemius; (3) the tendon of the plantaris. This tendon 



722 



TIT.\NOTHERES OF ANCIENT WYOMING. DAKOTA, AND NEBRASKA 



arises beneath the gastrocnemius, but lower down near 
the heel it comes to the surface, curves around the ten- 
don of the gastrocnemius, then out, "spreading over 
the surface of the os calcis, partly attached to either 



The sustentaculum calcanei on its posterior surface 
bears a long vertical groove, lying next the shaft of 
the calcaneum; this groove lodged the tendon of the 
flexor longus hallucis. The medial tip of the susten- 





FiGURE 659. — Skeleton and restoration of Broiitops cf. B. 



Based on the mounted skeleton in the American Museum of Natural History and on the studies summarized in the text. One twenty-flfth 

natural size. 



side it continues into the sole of the foot." (Murie, 
1872.1, p. 163.) 

The back of the tuber calcis is roughened for the 
"ligamentum tarsi plantare longum" (Schmaltz,) 
which was inserted into the large tuberosity on the 
back of the cuboid. 



taculum is roughened for the ligamentum tarsi tibiale 
breve, which passed across to a superior postero- 
internal process on the astragalus. 

On the front face of the calcaneum the coracoid 
process (above the ectal facet) bears on its superior face 
two deep pits; the inner and larger one receives the 



THE MUSCULAR ANATOMY AND THE RESTORATION OF THE TITANOTHERES 



723 





FiGTJEE 660. — Restoration of the superficial muscles of the body of an Oligocene titanothere (Brontops of. B. robushis) after the 

removal of the panniculus 

Drawn by Erwin S. Christman. About one twenty-fifth natural size. 
101950— 29— VOL 2 3 



724 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



posterior tip of the fibula in extreme flexion of the 
foot; the outer smaller pit is for the "ligamentum 
tarsi fibul are breve" (Schmaltz), which passes to the 
outer side of the calcaneum. The lower fourth of the 
shaft of the calcaneum above the cuboid facet is 
externally roughened, apparently for the "ligamentum 




Lambdotheriijm, popoa/jicum EolUanons pri/iceps EotiUmops oreaorvi 



Figure 661. — Restorations of nine species of titanotheres from the 

lower, middle, and upper Eocene and lower Oligocene 

Drawn by Mrs. E. M. Fulda. About one-flftieth natural size 

tarsi fibulare longum" and perhaps also for the ex- 
tensor brevis digitorum. The external keel of the 
astragalar trochlea bears externally an articular band 
for the fibula. Partly surrounded by this band and 
lying behind it is a deep pit for the "ligamentum 
tarsi fibulare breve." 



On the inner side of the astragalus the posterosu- 
perior corner bears a large process for the "ligamentum 
tarsi fibulare breve." The same ligament was probably 
continued into the pit which lies below and in front of 
the last-named process. Also on the inner side of the 
astragalus near the lower end, above the navicular 
facet, is a high prominence, apparently for the 
ligamentum tarsi tibiale longum. Immediately 
above this prominence passed the tendon of the 
tibialis on its way to insertion on the proximal 
end of the second metatarsal. 

The roughened anterior faces of the navicular, 
cuneiforms, and cuboid doubtless betoken the 
ligaments connecting these bones with the as- 
tragalus and tibia. The roughened protuberance 
on the back of the cuboid served not only for 
the insertion of the ligamentum tarsi plantare 
longum as stated above but also for the origin 
of the interossei muscles on the sole. The large 
prominence on the back of the entocuneiform 
gave origin to the first interosseus. The front 
faces of the three metatarsals were probably 
overlain by the several bellies of the extensor 
brevis digitorum, which in the tapir passes down 
from the front of the calcaneum, from the ankle 
ligament and bones beneath, to the proximal 
phalanx of the middle toe, lateral slips being 
given off to the proximal phalanges of digits II 
and IV. 

The proximal end of Mts II is roughened 
on its medial face for the tendon of the tibialis 
anticus and probably for the "ligamentum tarsi 
tibiale longum." The outer surface of Mts IV 
is roughened near the proximal end, probably 
for a branch of the plantaris ligament. 

The shafts of the metatarsals on each side in the 
proximal third are roughened for the interossei 
and adductors, which as in Tapirus may have 
consisted of three pairs. (Murie, 1872.1, p. 166.) 
The sides of the proximal phalanges, as in the 
manus, are roughened proximally for the interos- 
sei and probably to the lumbricales. The latero- 
distal pits are for collateral ligaments, the dorsal 
face, proximal end, for the extensor brevis 
digitorum. On the plantar side the proximal 
phalanges were overlain by the perforated ten- 
dons of the plantaris, which were inserted into 
the sides of the second phalanges. 

In the ungual phalanges the dorsal face, proxi- 
mal portion, gave insertion to the extensor longus 
digitorum; the plantar face gave insertion to the flexor 
communis digitorum (flexor perforans). 

The lumbricales, which in Tapirus are very fully 
developed, were attached to the sides of the flexor 
tendons and were inserted on the proximal sides of 
the first row of phalanges. 



THE MUSCULAR ANATOMY AND THE RESTORATION OF THE TITANOTHERES 



725 



SECTION 5. MUSCLES OF THE LIMBS AND VERTE- 
BRAE OF OLIGOCENE TITANOTHERES 

The more robust development and marked broaden- 
ing of many of the limb bones of the Oligocene 
titanotheres as compared with those of the Eocene 
titanotheres indicate corresponding changes in the 
muscles. The principal attachments of the limb 
muscles are indicated in Figures 652-660. 

SECTION 6. RESTORATION OF THE MUSCULATURE 
AND BODY FORM OF BRONTOPS ROBUSTUS, AN 
OLIGOCENE TITANOTHERE 

The restoration of Brontops rohustus is based, first, 
upon a detailed study of the musculature of each 



pose finally adopted represents the animal in rapid 
forward motion with the head lowered, as if for attack. 
The position assigned to the limbs is that commonly 
assumed under such conditions by graviportal types, 
such as the elephant and rhinoceros. The scapula is 
placed well above the ribs, as in ungulates generally, 
and not so far down as it is in the mounted skeletons 
of titanotheres in the museums. When placed in this 
pose the fore part of the skuU is still some distance 
above the ground. In order to lower the mouth to 
the ground the animal possibly extended one fore 
limb far forward and the opposite fore limb backward. 
After the skeleton was drawn in this pose (fig. 659) 
an outline tracing of this drawing was used as a basis 




Figure 662. — Restoration of Broniotherium gigas 

Modeled by E. 8. Christmau under the direction of W. K. Gregory. Based on the studies of the musculature and skeleton described in Chapter 
VIII. The relative length of the skull and forearm and the depth of the thorax at the shoulder were determined from Am. Mus. 492. About 
one thirty-second natural size. 



bony element of the skeleton, as described above, and 
second, upon a comparative study of the general 
appearance of the musculature of the body as a whole 
in recent perissodactyls, as figured by several authors. 
Some valuable hints were obtained from photographs 
of partly dissected specimens of the white rhinoceros, 
taken in the field by Herbert Lang, and from figures 
of the Indian elephant by Cuvier and Laurillard. 

The first step toward insuring a lifelike pose in the 
restoration of the living animal was to obtain a correct 
pose for the skeleton. A cardboard manikin based 
upon the mounted skeleton in the American Museum 
was drawn to scale; in this manikin each movable 
element of the skeleton was fastened to fiexible copper 
wire, so that the pose could be readily changed. The 



for the placing of the muscles. There can be little 
doubt as to the general position of the chief muscles 
as seen after the removal of the panniculus and 
platysma layer; but the precise form and size of each 
muscle are of course more or less doubtful. In every 
case, however, we are guided, so far as possible, by 
the conditions observable in recent perissodactyls and 
to some extent in the elephant. 

In the life restoration the skin is of the pachyderm 
type, intermediate in character between those of 
elephants and rhinoceroses but somewhat more hairy. 
Possibly the color was more or less uniformly dark 
gray. The feet are intermediate in character between 
those of the rhinoceros and the elephant, as shown by 
the skeleton. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE XLIII 




PEOBABLE ARRANGEMENT OF THE MUSCLES OF THE OCCffUT AND NECK IN AN OLIGOCENE 

TITANOTHERE (BRONTOPS ROBUSTTJS) AFTER THE PLATYSItfA, SPLENIUS, RHOMBOn)EUS, 

TEAPEZroS, AND LEVATOR SCAPULAE HAVE BEEN EEMOVED 

One-sixth natural size. Blue lines indicate deep muscles, red lines superficial muscles, and black lines the liga- 
mentum nuchae ; x (blue) indicates a cut end of a muscle, and x (red) the origins of slips of the serratus cervicis 
(levator scapulae) . To the uttermost rim of the lambdoidal crest was attached the splenius, and below this, on 
the back of the mastoid region and paroccipital process, were the longissimus capitis and cephalo-humeralis 
(cleido-mastoid). Beneath the splenius was the complexus (semispinalis capitis). The nuchal surface of the 
occiput was filled by the obliquus capitis superior, the rectus capitis posticus major and minor, and the 
Ugamentum nncbae. {After Gregory and Christman. ) 




MONOGRAPH SS PUTE XLW 













A2 "" "■'' ''•i^ if '/»■•»'■") 




i:.v(,-.is...- tr,i...< <i.j.(. 



P/a..( fr.v/o,-ai„sC 




^ flex, loncrhiilj 



Pla.,( (f,yfcn.lu,j 



MUSCULATURE OF THE FORE AND HIND UMBS OF PALAE08Y0PB LEIDYI 



CHAPTER IX 



MECHANICS OF LOCOMOTION IN THE EVOLUTION OF LIMB STRUCTURE AS BEARING 
ON THE FORM AND HABITS OF THE TITANOTHERES AND THE RELATED ODD-TOED 
UNGULATES 



The facts presented in this chapter should be 
studied in connection with those given in Chapter 
VIII on musculature and bone structure. Together 
these sections afford a complete interpretation of the 
mechanical evolution of the titanotheres. 

SECTION 1. ADAPTATION TO LOCOMOTION IN THE 
LIMBS OF THE FLEET (CURSORIAL) AND THE 
PONDEROUS (GRAVIPORTAL) TYPES OF TITANO- 
THERES AND OTHER HOOFED QUADRUPEDS 

MECHANICAI AND PHYSIOLOGICAL PRINCIPLES GOVERN- 
ING THE PROPORTIONS AND ANGULATION OF LIMB 
SEGMENTS IN GRAVIPORTAL UNGULATES 

RESEARCHES MADE AND PRINCIPLES ESTABLISHED 

The laws of limb adaptation here presented are the 
result partly of joint and partly of independent re- 
searches made for this monograph. In 1900 Osborn 
(1900.181) independently established the relations of 
the articular facets and joints to speed and weight, 
and in 1911 he planned a comparative measurement 
and investigation of the limbs and arches in the nine 
families of perissodactyls and other ungulates. Osborn 
and Gregory jointly established the laws of limb 
proportion in adaptation to speed and weight, re- 
spectively, which were in part published in advance 
by Gregory (1912.1). Gregory (1912) independently 
investigated and applied the mechanics of muscular 
limb action and leverages in relation to speed and 
weight. The text of this chapter has been jointly 
prepared by Osborn and Gregory. 

The principle that the limbs of quadrupeds are 
compound levers and that the relative lengths of the 
upper, middle, and lower segments are adapted to 
specific muscular powers, loads, and speeds is applied 
in the present monograph to the elucidation of the 
adaptive contrasts between cursorial and graviportal 
ungulates. 

The limb movements of living animals have been 
investigated by Borelli (1680.1), Marey (1874.1), 
Stillman (1882.1), Haycraft (1900.1), Luciani (1905.1), 
Muybridge (1907.1). From the evolutionary point of 
view there are the studies of Cope (1889.3), Osborn 
(1890.51, 1895.99, 1900.181), Gaudry (1906.1), Greg- 
ory (1912.1). Cope made some suggestions as to the 
relative length of limb segments in saltatorial, cursorial, 
and graviportal types. Osborn especially developed 
the subject of angulation of the limbs in relation to 
great weight and to the articular planes of the proximal 
and distal limb facets. Gaudry pointed out the limb 



convergence between such straight-limbed, ponderous 
ungulates as Elephas, Uintatherium, and Pyrotherium 
as well as the convergence between numerous curso- 
rial animals {Equus, Theosodon). Matthew (1909.1, 
pp. 429-432) discussed the ratios of the limb seg- 
ments in various carnivores and ungulates, in adap- 
tation to cursorial and "rectigrade" modes of locomo- 
tion. Riitimeyer and Allen (1876) discussed the 
relative lengths of the different segments of the fore 
and hind limbs in the extinct and living species of 
the bison and introduced the ratio method for taxo- 
nomic purposes. This method was effectively used 
by Gidley (1903.1) in comparing the different genera 
and species of Equidae. A review of the mechanical 
principles of quadrupedal locomotion with special 
reference to the limbs of ungulates is given by 
Gregory (1912.1, pp. 268-269). The chief works are 
cited in the bibliography for this chapter. 

PRINCIPLES OP LEVERAGE AND MUSCULAR ACTION 
By William K. Gregoky 

The meaning of contrasting limb proportions and 
angulation in cursorial and graviportal animals be- 
comes clear if we keep in mind the underlying mechan- 
ical and physiological facts and principles. 

Observation and photographs of animals in motion 
at once reveal the simple fact that the limbs by sud- 
denly straightening out and opening the angles at 
the joints raise the whole body and not only enable 
it to fall forward but alternately accelerate and check 
the action of gravity. An animal running on a 
treadmill reveals the almost equally important fact 
that the limbs when extended anteriorly tend to drag 
the ground backward and to pull the body forward. A 
standing horse might give the impression that the fore 
limbs were relatively passive props and that the hind 
limbs were the chief propellers, but . a consideration 
of the action of the massive muscles of the fore limbs 
and study of a horse running on a treadmill show that 
the fore limbs also exert great strength in dragging the 
body forward, in raising the fore part of the body, and 
even in pushing it forward when the fore limb is ex- 
tended backward. In most ungulates the fore limbs 
sustain the greater loads. 

When the limbs are moving quickly they seem to 
swing with the regularity of a penduhmi, and it has 
often been noted that in cursorial animals the limb 
muscles are bunched or concentrated toward the proxi- 
mal or upper end of the limb, the muscles being pro- 

727 



728 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



longed distally as slender tendons; in this way the 
center of gravity of the limb is comparatively high and 
we get the rapid oscillation of a pendulum in which the 
bob or weight is placed near the pivot. 

In graviportal animals, on the contrary, the limb 
muscles are more evenly distributed down the limb 
and the slender tendinous portions are shorter, so that 
we have the slower oscillation of a pendulum in 
which the weight is moved down the shaft. But the 
free swing of the limbs is alternately accelerated and 




FiGUBE 663.- 



.'-'C 



-Angles of muscular insertion and "parallelogram of forces. 
After Gregory, 1912.1 




inversely proportioned to the amount of inert con- 
nective tissue that is interspersed with the striped 
muscle fibers. Muscles when stretched serve as 
ligaments and help to prevent dislocation of the joints, 
as well as to transmit the pull of other muscles. Thus 
the biceps muscle and tendon assist in preventing 
dislocation of the shoulder; the quadriceps femoris 
and its tendons prevent dislocation of the knee joint; 
the serratus and its allies help to sling the body between 
the shoulders. 

In general, long and slender muscles, such 
as the sartorius, exert a small force over a 
long range, while short and thick muscles, such 
as the gluteal mass in many ungulates, exert 
a great force through a short range. When 
the muscle fibers are set obliquely to the 
long axis of the muscle, as in the gastrocnemius, 
the muscle contracts slowly but with increased 
power. The contractile force is highest when 
a muscle is stretched to its full "physio- 
logical length" (that is, the greatest length 
it ever assumes during life), and the greatest 
force and velocity of contraction are developed 
when the movement of the muscle is checked 
during the initial stages and when the resist- 
ance is suddenly diminished. These prin- 
ciples operate constantly in the limbs, espe- 
cially in the straightening of the knee joint, 
and in many cases where there is a sudden 
snap or jerk, as in the action of the flexors 
of the digits in a running horse. 

It is a familiar dictum of elementary 
mechanics that in considering the action 
of levers "what is gained in speed is lost 
in power, and what is gained in power is 
lost in speed." In the skeleton these recip- 



If the muscles contract at equal rates and the angle of insertion is acute (ABD) the extremity of a 
bone is moved through a wide arc (CC), with feeble propulsive power {BE) and great lifting 

power (.-li?). As the angle of insertion increases (a6d) the extremity of the bone is moved through j^QCal relations of " POWCr " and "sDCed" 
a smaller arc (cc') with greater propulsive power (iir) and reduced lifting power (a6). If the muscles rrxj i. lu+1 1+' J" 

contract at unequal rates a slowly contracting muscle (.BD) inserted at a small angle may move ^-^e aiiected not Only by the relative QIS- 

the insertion point slowly over a wide arc (B£'). whereas a more rapidly contracting muscle (6(i) tancCS between the fulcruni (joint) the forcC 

may move the insertion point more quickly but through a smaller arc (bb') . zci i\ i.iii '■, 

(of the muscle), and the load or resistance, 



retarded by the actions of the muscles, and the fore- 
and-aft movement is complicated by tw-isting, ab- 
duction, and adduction of the various segments of the 
limb. 

When isolated for laboratory experiments, muscles 
react if stimulated according to relatively simple 
mechanical and physiological laws. An isolated mus- 
cle will contract upon stimulation to a varying fraction 
of its own length, from one-fifth to one-third. The 
force of a muscle — that is, its ability to overcome 
inertia at a given instant — is proportional to the num- 
ber of muscle fibers cut in a cross section of the 
muscle; but the work that a muscle can perform 
during the entire period of its contraction is propor- 
tional both to the area of the cross section and to the 
length of the muscle — that is, the total work performed 
is proportional to the mass of the muscle; it is also 



but also by the "angle of insertion" of the muscle. 
If a man pulls a heavy door directly toward him 
(at right angles to the door) he will naturally find 
that it wUl move much more easily than when he 
puUs it at an oblique angle; but in the second case, 
although the effort required is greater, the movement 
of the door is much faster. The "angle of insertion" 
of a muscle is formed by the long axis of the muscle, 
its point of insertion, and the line between the inser- 
tion point and the fulcrum or joint. (See fig. 664.) 
If the angle of insertion is large, as in the gastrocnemius 
of graviportal animals, great power is secured. If 
tlie angle of insertion is small, as in the flexors and 
extensors of the digits, the movement of the insertion 
point may be very rapid. 

From the diagram above referred to we learn that 
if two muscles contract at the same rate the lesser the 



MECHANICS OF LOCOMOTION 



729 



oL.ucci. 



angle of insertion the greater will be the distance tion points enable us in some measure to conceive 
traversed by the insertion point in one second of the cooperative action of many muscles of varying 
time. From this it follows that if muscles contract lengths, speeds, and insertion angles, which may all be 
at different rates the angles of insertion may be so inserted at different points on a single bone and yet 

complete one cycle of movement in nearly the 
same time. It is also clear that in some muscles, 
as in the flexors of the digits, great speed may be 
developed at the beginning of a movement and 
great power with slow speed at the end, and vice 
versa. 

Certain of the limb muscles (such as the coraco- 
brachialis) are wrapped around spirally warped sur- 
faces, and many of the joints between the segments 
(notably the astragalotibial joint) form segments of 
spirals, so that the mechanical advantages of the 
spiral wedge, or screw, are secured. By the proper 
regulation of the pitch of the screw either great 
"power" or great "speed" can be attained. For ex- 
ample, in extreme graviportal types the upper artic- 
ular surface of the astragalus is very gently inclined 
downward and forward, corresponding to a wedge or 
screw of low pitch. In this way the very heavy load, 
represented by the pressure of the tibia upon the 
astragalus, is easily pushed up the gently sloping 
articular surface. In cursorial types, on the other 
hand, the upper articular surface of the astragalus 
faces more directly forward, and the pitch of the 
screw is relatively high. Under these conditions a 
relatively great expenditure of muscular energy is 
required, but the movement of the tibia upon the 
0, 7) and in a typically graviportal form (Ib, the astragalus is correspondingly rapid. 
with narrow angles of insertion («', 0', y'). The entire complex of the skeleton, of the muscula- 





FiGURE 664. — Angles of insertion of certain extensor muscles 
to the pelvis and femur in the standing pose of a typicallj' 
cursorial form (A, the horse) with relatively wide angles of 
insertion {< 
mastodon) 



After Gregory, 1912.1 ture, and of the organs for developing and directing 

The heavy black lines represent the general direction of the muscles. The broken loCOmotivC energy is adapted in CUrSOrial auimals for 

leaping and fast running, in graviportal animals for 

the evenly maintained forward sweep of the walk and 

amble. The elongate feet and bent limbs of cursorial 

forms when suddenly straightened out act more or 



lines represent the radii of movement of the insertion points. In B the angle ( 
is probably too small — W. K. Q. 



adjusted as to keep constant the distance traversed in 
one second by the insertion point. The angle of inser- 
tion and c o n s e- 



quently the recip- 
rocal relations of 
speed and power 
change as the mus- 
cle contracts. In 
the case of the gas- 
trocnemius the angle 
of insertion de- 
creases as the heel 
is raised from the 
ground; in the case Figure 665 



Planes of astraga/o-t/'b/a/ facets 




P/anes of 

astraga/a - 

na\//cc//ar faceis 

-Graviportal adaptation in the astragakis. Backward shifting of the plane of the astrag- 



alotibial facet from an angle of 82° in Ursui to an angle of 26° in Elephas, the plane of the 
astragalonavicular facet being taken as horizontal. After Osborn, 1900.148 

Ur, Ursus: Ph, Phenacoiiis: Co, Coryphoion: Br, Brontops; Ui, Uintalherium: El, Blephis. 



of the flexors of the 
digits the angle of 
insertion and conse- 
quently the "power" increases and the speed decreases 
as the muscles contract. 

These variable relations of the rate of contraction, 
of the angle of insertion, and of the speed of the inser- 



less after the manner of a catapult in hurling the 
relatively light body forward. The very short feet 
and straight limbs of graviportal forms act as long- 
armed levers and low-pitched screws, which slowly 



730 



TIT.^><'OTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 





FiGUEE 666.- 



-Cursorial adaptation for the trot, run, and gallop in the horse (Ai-Aj) ; graviportal adaptation for 
the walk and amble in the elephant (Bi-B,). After Gregory, 1912.1 



Ai, The gallop of a horse, showing the leaping, catapult action of the slender limbs in a cursorial type. 

Bi, The amble of an elephant, showing the steady sweep and great weight-lifting power of the ponderous limbs in a graviportal type. 

As, B2, Extremes of cursorial adaptation in a Miocene horse, Hipparion (Neohipparion) (A2), and of graviportal adaptation in a mastodon (Bs). 

A3, B3, Angulation of limb segments and muscles in a horse (A3) and a mastodon (B3): 



Cursorial (A3) 
Hium and femur ar right angles in the standing pose. 
Muscles inserted at open angles, giving great propulsive power. 

Head of femur low, at right angles to shaft. 
Longissimus dorsi works in tandem with gluteus medius. 



Graviportal (Bj) 

Ilium and femur more nearly vertical. 

Muscles inserted at more acute angles, giving great vertical or weight- 
lifting power. 

Head of femur high, at top of shaft. 

Longissimus dorsi cut off from gluteus medius. 
Cmrsorial hind foot (Hipparion, Neohipparion), for springing and rapid lifting. B4, Graviportal hind foot (Mastodon), for powerful, slow 
lilting. Fulcrum at B', b'; power at B, b; resistance at A, a. If the pull of the calf muscles (M,m) and the weight at the ankle joints (W,w) 
were respectively equal, the long, cursorial foot would have great "moment of resistance" (WXB'A) and small "moment of power" (MXBB'), 
with high speed; whereas the short, graviportal foot would have relatively small moment of resistance (wXba) and great moment of power 
(mX&60, with slow speed. 



MECHANICS OF LOCOMOTION 



731 



lift, push, and pull forward the ponderous body. In 
more technical terms, the straightness of the limbs 
in graviportal animals is believed to have been 
evolved pari passu with the short rectigrade feet and 
with an ambling gait, while the bent or angulate 



ascertaining the position and angulation of the limbs 
in extinct quadrupeds. The term angulation refers 
especially to the angles which exist between the main 
proximal and distal segments of the fore and hind 
limbs, namely, between the humerus and ulnoradius 



I 



I 
I 




iencion, biceps 
to meiacarb- 




teyisor 
fasciae 



Resultant p-ull, tendons^ 

of knee and- ankle: 

») Flexio-n: tendo femoro- 

tarseus 

Extension: tendo plautari^ 



iendo accessoritt'S 
from biceps and 
semitendi'^csus 



Figure 667. — Cursorial adaptation in tlie fore limb (A) and hind limb (B) of the horse 

Upper segments short, angulate, with muscle insertions at open angles (power) ; lower segments elongate, muscle 
insertions oblique (speed) . In the standing pose the opposing flexors and extensors are stretched through 
the weight of the animal and act more or less passively as ligaments. The equilibrium is maintained also 
by means of special tendons on opposite sides of the limbs. After Grejory (1912.1); adapted from Schmaltz 
(1905.1). 



character of the limbs in cursorial animals is correlated 
with the very long, slender unguligrade feet and with 
a rapid running, bounding, galloping, or trotting gait. 

STRAIGHTENING OF THE LIMBS AND ARCHES IN ADAP- 
TATION TO GREAT WEIGHT (GRAVIPORTAL TYPE) 

Angulation. — The author (Osborn, 1900.181), was 
led some years ago to study the various means of 



and between the femur and tibiofibula. All primitive 
ambulatory and subcursorial ungulates exhibit sharp 
angles at the elbow and at the knee which are inherited 
from the unguiculate ancestors of ungulates. This 
acute angle is retained in aU the cursorial and medi- 
portal ungulates, but it tends to open out as the 
limbs become more vertical in the graviportal digiti- 
grades and graviportal rectigrades. 



732 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



From this we deduce the following law, which was 
partly formulated by the author in 1900 (op. cit., p. 94) : 
The straightening of the ungulate limb is a secondary 
adaptation designed to transmit the increasing weight 
through a vertical shaft and thus relieve the muscles 
and Hgaments from strain. Correlated with the 
straightening of the limb are the shifting of the 
proximal and distal articular facets of the humerus and 
■of the femur into direct lines of pressure, and the 
alteration of the sections of the articular surfaces 
from an oblique to a right angle, or a horizontal 
angle with relation to the vertical shaft. 

This law is clearly illustrated in the accompanying 
figures, based upon full-size vertical sections of the fore 



shaft in Elephas. By these means in the straight- 
limbed Proboscidea, Dinocerata, and Pyrotheria the 
proximal articular facet comes to lie almost directly 
across the top of the shaft, while the distal articular 
facet similarly lies across the bottom of the shaft. 

Angles. — In the progressive adaptation of the titano- 
theres to increasing weight we observe a similar modi- 
fication of the primitive angulation; both fore and hind 
limbs become more vertical. Even in the middle 
Eocene forms, such as the subgraviportal Palaeosyops, 
the hind limb is exceptionally vertical. 

Patella. — The above adaptation is accompanied also 
by a shifting of the position of the patella. In the 
primitive, highly flexed ungulate limb the patella is 




Figure 668. — Angulation of the fore limb in graviportal types, as shown by longitudinal 
sections of the humerus, ulna, and radius 

A, Hhinocnos unicornis; B, Brontotherium leidyi; C, Eobasileus cornuius: D, Elephas indicus. After Osborn. A, B, 
Incompletely graviportal adaptations, with the humeroradial angle more acute and the proximal and distal 
articular facets more oblique to the long axis of the shaft; C, D, advanced graviportal adaptations, with the 
humeroradial angle more open and the proximal and distal articular facets more horizontal to the long axis of 
the shaft. 



limbs of four types of quadrupeds which exemplify the 
progressive straightening of the limb, step by step, with 
progressive increase in bulk and size. (See fig. 668.) 

Axes. — The main axis of each limb bone is regarded 
as the line a-a, which passes through the center of the 
proximal facet and behind the center of the distal 
articular surfaces. It is noteworthy that the edges 
(6-&) of these proximal and distal articular surfaces 
are in parallel planes. It will be observed that the 
angle at the elbow is constantly lessening until in the 
extreme extension of the elbow joint in the elephant it 
becomes relatively small. 

Facets. — Another important change takes place: 
the proximal and distal facets, which are out of the 
line of the main axis of the shaft in Rliinoceros, are 
observed to shift more nearly to the extremities of the 



placed at the extremity of the shaft; the patellar facet 
forms a sharp angle with the axis of the shaft. In the 
graviportal femur the patella shifts forward to the 
front of the shaft, and the patellar facet becomes nearly 
parallel with the long axis of the shaft. 

Head of the femur. — In graviportal animals the head 
of the femur is sessile upon the top of the shaft, so that 
the weight of the body, in so far as it is represented by 
the pressure of the acetabulum upon the head of the 
femur, is transmitted directly through the shaft of the 
femur (H. F. Osborn). This arrangement, together 
with the subvertical position of the ilium, the rela- 
tively narrow transverse diameter of the pelvis across 
the acetabula, and the inwardly tilted tibial facet 
of the femur, brings the feet close to the midline and 
directly beneath the acetabulum (W. K. Gregory). 



MECHANICS OF LOCOMOTION 



733 



INDICES AND RATIOS OF LIMB SEGMENTS IN CUR- 
SORIAL AND GRAVIPORTAL UNGULATES (ALLOMETRIC 
ADAPTATIONS) 

RESULTS OF COMPARATIVE STUDIES 

The comparative studies pursued by Osborn and 
Gregory, based upon limb segment measurements 
taken by Gregory for this monograph in about sixty 
quadrupeds, bring forth some new and significant laws 
of limb proportion in addition to those already adum- 
brated by various students of animal locomotion and 





types we have, in the fore limb, humerus long, radius 
shorter, manus very short; whereas in cursorial, swift- 
moving types we have humerus short, radius long, 
manus very long. Similarly in the hind limb we have, 
in graviportal types, femur long, tibia short, pes very 
short, whereas in cursorial types we have femur short, 
tibia long, pes very long. 

During the course of evolution small subcursorial 
types have frequently diverged on the one hand into 
gigantic graviportal types and on the other hand into 






Figure 669. — Angulation of the limb bones at the shoulder and hip joints in the standing 
pose of cursorial (A) and graviportal (B, C) ungulates 

A', A^ Equus scotti, a Pleistocene horse; B', B^, Brontops robiistus, an Oligocene titanothere; C, C, Mastodon ameri- 
canus, a Pleistocene proboscidean. In the cursorial tj'pe the humerus is nearly at right angles with the scapula 
and the femur is nearly at right angles with the ilium. In the graviportal types all the elements become more 
nearly vertical in position. 



of the evolution of quadrupeds. Valuable as have 
been the previous studies of Marey, Osborn, Gaudry, 
and others on miscellaneous tj^pes, it is only through 
the measurement and comparison of all the ungulate 
types in continuous phyletic series, such as are now 
afforded through American paleontology, that these 
laws can be positively formulated as fundamental and 
universal principles of progressive evolution in the 
limbs. 

Antithetic proportions. — In general these measure- 
ments show that in extreme graviportal, slow-moving 



Weight allometry 
Lengthen. 
Shorten. 
Shorten. 



true cursorial tj^pes. The divergent evolutionary 
changes may be summarized thus: 

Speed allometry 

Humerus and femur Shorten. 

Radius and tibia Lengthen. 

Mtc III and Mts III Lengthen. 

Between these extremes we find many intermediate 
forms. Taken together, the living and extinct quad- 
rupeds may be arranged with reference to primitive 
or ancestral proportions and to either graviportal or 
cursorial adaptations into four general groups of types, 
adapted for various speeds and paces as follows. 



734 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 

Groups of types of quadrupeds adapted to different speeds and paces 



Type of locomotion and speed 



Type of limb and toot structure 



Examples 



1. Ambulatorj- : Small, primitive, low- 

bodied animals, which walk, shuffle, 
or run clumsily. 

2. Mediportal : Speed and pace of inter- 

mediate type, tending to become 
graviportal. 



3. Graviportal : Gigantic, massive animals, 

which swing the limbs ponderously 
in a walk or amble. 

4. Cursorial: Swift, light-limbed animals, 

which leap, gallop, and trot. 



Subplantigrade and spreading; feet short, 
limbs sharph' angulate. 

Digitigrade or partly unguhgrade; feet 
secondarilj' widened; digits of hind foot 
often reduced in number (from an earlier 
subcursorial phase), limbs more or less 
angulate. 

Rectigrade; feet very short and wide; pha- 
langes very short, terminal phalanges 
usually reduced; limbs straight or post- 
like. 

Unguligrade; feet very long, narrow, with 
marked digital reduction; terminal pha- 
langes large; limbs strongly angulate. 



Pantolambda, Hyrax, and 
sectivores and carnivores. 



many in- 



Tapirs, titanotheres (Mesatirhinus, 
Manteoceras), Oligocene rhinoceroses 
(Caenopus) . 



Oligocene titanotheres (Brontothe- 
rium), Amblypoda (Coryphodon,. 
Uintatherium) , Proboscidea. 

Most horses, antelopes, deer, titano- 
theres (Lambdotherium) . 



In comparing members of different races or the 
successive evolutionary stages of a single race we can 




FEMDR, 



Sliorieiring 

■witJv 

increasinff speed 

27.1 % of total leTTfftJt 



Lengtherixng 
■witlv 
increasing Weight 
48. 6 % of total lengt7T, 




TIBIA. 

Length remainijig relatively constant 



31.7% 




PES 



Lengthening 
■with, 
increasing speed 
41.1Yo of total lenffth. 



34.3% 



Shortening 
■with 
increasing ■weight 
17.1% of total length 



A 

Figure 670.- 



Hind limbs of cursorial and graviportal types, showing adaptive changes 
in the length of pro.ximal and distal segments 
A, Increasing length from proximal to distal segments in a cursorial limb (Neohipparion); B, decreasing length 
from proximal to distal segments in a graviportal limb (Mastodon) . 



find in the body no stationary or nonevolving bony 
segment which will serve as a norm, or fixed stand- 
ard of comparison: every bone 
is in a state of change; no 
segment is constant in length 
or in proportion to any adjoin- 
ing segment or to the body as 
a whole. Nevertheless if we 
arbitrarily assume that the 
proximal segment of the hind 
limb is of the same length (100) 
in all the animals measured 
we can then show the relative 
lengths of the remaining seg- 
ments of the limb in terms of 
the length of the femur. 

Indices. — These are the pro- 
portions of a single limb bone, 
like the femur, or of a single- 
arch element, like the scapula. 
For our purposes the least 
breadth of a bone may sim- 
ply be divided by the great- 
est articular length and the 
result expressed as a percent- 
age. 

Ratios. — These are obtained 
between two elements of a limb, 
as, for example, the length of 
the tibia divided by the length 
of the femur gives what may be 
called the tibiofemoral ratio. 
This ratio method assumes that 
every femur or humerus respec- 
tively, in being taken as a norm, 
is reduced to the same abso- 
lute length (100). Examples of 
such ratios are the following: 




MECHANICS OF LOCOMOTION 



735 



1. Tibiofemoral ratio = length of tibia -^length of femur. 

2. Radiohumeral ratio = length of radius-^ length of humerus. 

3. Metatarsofemoral ratio = length of Mts III-^ length of 
femur. 

4. Metacarpohumeral ratio = length of Mtc Ill-f- length of 
humerus. 

Extreme examples of the adaptive range of these 
ratios in graviportal and cursorial types expressed as 
percentages, are the following: 

Tibiofemoral ratio: Lowest (53) in Uintatherium mirahile, 
a slow, heavy type; highest (125) in Gazella dorcas, a swift, 
cursorial type. 

Metatarsofemoral ratio: Lowest (10) in Uintatherium mirahile, 
a graviportal type; highest (103) in Antilocapra, a cursorial 
type. 

Radiohumeral ratio: Lowest (52) in Pyrotherium, a graviportal 
type; highest (130) in Neohipparion whitneyi, a cursorial 
type. 

Metacarpohumeral ratio: Lowest (18) in Mastodon, a gravi- 
portal type; highest (144) in Gazella dorcas, a cursorial type. 

The widest recorded contrasts in limb ratios are as 
follows: 





Tibio- 
femoral 


Meta- 
tarso- 
femoral 


Radio- 
humeral 


Meta- 
carpo- 
humeral 




118 
53 


135 
10 


160 
70 


142 


Uintatherium (graviportal) 


19 



The contrasts between representatives of the chief 
types of limb adaptation may be further summarized 
as follows: 





Tibio- 


Meta- 


Radio- 


Meta- 




femoral 


femoral 


humeral 


huineral 


Primitive ancestral type, such as 










Phenacodus primaevus 


84 


31 


87 


42 


Mediportal intermediate type, 










such as Palaeosyops leidyi 


78 


30 


72 


34 


Graviportal digitigrade type, such 










as Hippopotamus amphibius 


67 


26 


68 


38 


Graviportal rectigrade type, such 










as Elephas indicus 


60 


13 


80 


22 


Cursorial unguhgrade type, such 












125 


81 


126 


144 







The extremes give us what we may popularly desig- 
nate speed ratios, weight ratios, mediportal ratios, 
and primitive ratios. 

Through convergent evolution animals of different 
ancestry may come to closely resemble one another in 
limb ratios, as follows: 



Cursorial: 

Perissodactyla, Hipparion 

(Neohipparion) 

Artiodactyla (Antilocapra) _ 
Graviportal: 

Edentata (Lestodon) 

Amblypoda (Coryphodon).. 
Brontotheriidae (Brontops 

robustus) 

Amynodontidae (Metamyno- 

don) 

Rhinocerotidae (Teleoceras) . 
Toxodontia (Toxodon) 



117 
123 



Meta- 
tarso- 
femoral 



101 
103 

12 
14 

26 

24 
25 
17 



130 
123 



Meta- 
carpo- 



116 
130 



39 
37 
38 



We observe that the tibia is relatively long in the 
primitive ungulates, probably, as observed above, 
because of their derivation from unguiculate ancestors, 
in which the tibia is long. The tibia abbreviates in 
most heavy, slow-moving types; it elongates in all 
rapidly moving types. 

The radius either remains stationary or abbreviates 
to a less degree than the tibia in slow-moving types'. 
In swift-moving types the radius appears to elongate 
as rapidly as the tibia. 

The metacarpals and metatarsals progressively 
abbreviate in graviportal types; they progressively 
elongate in cursorial types. 

RATIOS SHOWING EVOLTTTION FROM PRIMITIVE TO 
GRAVIPORTAL AND CURSORIAL TYPES 

The contrasts in the limb segment ratios of these four 
groups having been shown, it is now necessary to point 
out some of the common characters of primitive 
ungulates and to demonstrate the fixed laws of evolu- 
tion into the graviportal and cursorial extremes as 
well as the limb segment ratios which prevail in the 
mediportal types. This may be done in the following 
comparison of a series of groups of types. 

Primitive ungulate types. — The primitive ungulates 
listed below are partly ambulatory, partly subcursorial, 
or even cursorial; yet they have in common, as prim- 
itive or ancestral characters, the elongate tibia and 
radius and the primitively abbreviate metatarsus and 
metacarpus — that is, such animals as Euprotogonia 
and Eoliippus were undoubtedly cursorial, although 
they lack the elongate metapodials that are charac- 
teristic of modern cursorial types. 



736 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 

Measurements, in ■millimeters, and ratios of limbs of ungulates of primitive types 



Ambulatory : 

Pantolainbda bathmodon 

Meniscotherium terraerubrae — 

Subcursorial: Phenacodus primaevus 

Cursorial: 

Euprotogonia puercensis 

Eohippus sp 



149 
100 
234 

105 
162 



114 

91 

198 

107 
162 



Tibio- 
femoral 
ratio 



76 
91 

84 

101 
100 



Meta- 

tarso- 

femoral 

ratio 



124 

82 

167 



Radio- 
humeral 
ratio 



82 

58 

146 



110 



Meta- 
carpo- 



35 
37 
13 



Phyletic progression (Eocene) to graviportal type. — 
An extreme instance of progression into the heavy- 
bodied, slow-moving (graviportal) type is that found 
in the comparison of the three Amblypoda, Panto- 
lainbda of the basal Eocene, CorypJiodon of the lower 
Eocene, and Uintatherium of the upper Eocene. The 



relative abbreviation of the tibia, of the metatarsus, 
and of the metacarpus is in contrast with the slight 
relative elongation of the radius. This illustrates a 
frequently repeated principle of graviportal adaptation, 
namely, that the radius is not abbreviated step by step 
with the tibia but may actually be elongated. 



Measurements, in millimeters, and ratios showing progression of Eocene ungulates from the ambulatory to the 

graviportal type 





Femur 


Tibia 


Tibio- 
femoral 
ratio 


Mts III 


Meta- 
tarso- 

femoral 
ratio 


Humer- 
us 


Radius 


Radio- 
humeral 
ratio 


Mtc III 


Meta- 

carpo- 

humeral 

ratio 




149 
423 
692 


114 
260 
360 


76 
61 
53 


36 
62 
70 


34 
14 
10 


124 
363 
540 


82 
240 
380 


66 
66 
70 


31 

70 

106 


35 




19 




19 







Primitive cursorial condylarths and perissodactyls. — 
Even in lower Eocene time the condylarths and 
perissodactyls embrace many primitive swift and 
light-limbed types. These cursorial types, which 
include a lower Eocene titanothere, had what may be 
called high-speed ratios for the tibia and radius, but 



low-speed ratios for the metatarsus and metacarpus. 
The most primitive Eocene titanothere, Eotitanops, 
was subcursorial, and Lambdotherium was cursorial in 
limb proportions, but the material available for these 
genera does not afford the required indices. 



Measurements, in millimeters, and ratios of limbs of primitive cursorial condylarths and perissodactyls 



Tibio- 
femoral 
ratio 



Meta- 
tarso- 

femoral 
ratio 



Radio- 
humeral 
ratio 



Meta- 

carpo- 

humeral 

ratio 



Phenacodontidae: Euprotogonia puercensis. 

Lophiodontidae : Heptodon calciculus 

H yracodontidae : Hyrachy us agrarius 

Equidae: Eohippus sp 



105 
175 
254 
167 



107 
175 
243 
162 



101 

100 

95 

100 



45 
»75 
110 
'■82 



115 
197 
121 



114 
197 
110 



99 

100 

90 



5S 
47 
53 



Mediportal or intermediate group. — The mediportal 
group includes the middle Eocene titanotheres as well 
as the modern tapirs and some of the older types of 
rhinoceroses, all animals of intermediate size with an 
intermediate rate of speed. Here we observe a strik- 
ing uniformity in the limb segment ratios. As com- 
pared with the primitive ungulates shown in the first 
table above, the mediportal titanotheres were all 
apparently abbreviating the tibia in connection with 



their increasing weight. The metatarsals and meta- 
carpals were also abbreviating more rapidly in adapta- 
tion to weight. The radius, however, in relation to 
the humerus, remains relatively long; in several other 
phyla which are progressing toward the weight-bear- 
ing type the radius yields a less constant weight index 
than the tibia. As indications of speed the tibial and 
radial are much less reliable than the metapodial 
indices. 



MECHANICS OF LOCOMOTION 
Measurements, in millimeters, and ratios oj limbs of mediportal ungulates 



737 



Tibio- 
femoral 
ratio 



Meta- 
tarso- 

femoral 
ratio 



Radio- 
humeral Mtc III 
ratio 



Meta- 

carpo- 

humeral 

ratio 



Amynodontidae: Amynodon intermedius 

Tapiridae: 

Tapirus terrestris 

Tapirus indicus 

Titanotheriidae : 

Palaeosyops major 

Palaeosyops leidy i 

Manteoceras manteoceras 

Mesatirhinus petersoni 

Limnohyops monoconus 



262 
320 

433 

370 

390? 

358 

387 



'300 

208 

258 

332 
290 
272 
283 
283 



71 



79 
80 



77 
78 



108 
120 



137 
110 



205 
250 



177 
228 



91 



106 
120 



325 



235 



113? 
107 



118 
123 



293 



228 



50 

4S 



34?- 



Phyletic -progression to cursorial type in the Eguidae. — 
The above "primitive cursorials" constitute the type 
from which the "modern cursorials" evolved. The 
Equidae in their progression into the "modern cur- 
sorial" type present exactly the reverse of the gravi- 
portal progression seen above in the Amblypoda. 
They show a relative elongation of the tibia, radius, 
metatarsus, and metacarpus. After attaining the 



highest speed ratios in the lower Pliocene Hipparion 
(Neohipparion) wMtneyi we observe a retrogression, 
toward the more heavy-bodied modern Equus type 
and a still greater retrogression into the short-limbed 
and slow-moving Hippidion neogaeum of South 
America and in the still shorter-limbed Hyper- 
Jiippidium. This is one of the most significant series 
of comparative ratios which we have obtained. 



Measurements, in millimeters, and ratios showing cursorial progression 



Femur 


Tibia 


Tibio- 
femoral 
ratio 


Mtsm 


Meta- 
tarso- 

femoral 
ratio 


Humer- 
us 


Radius 


Radio- 
humeral 
ratio 


Mtc III 


162 


162 


100 


"82 


50 


121 


110 


90 


64 


170? 


170 


100? 


»98 


57? 


115? 


122 


106? 


80 


178 


193 


108 


121 


68 


136 


"136 


100 


92 


278 


277 


100 


218 


78 


205 


260 


127 


203 


249 


293 


117 


252 


101 


187 


244 


130 


218 


313 


310 


99 


277 


88 


237 


302 


127 


238 


392 


363 


93 


288 


73 


305 


363 


119 


240 


370 


330 


88 


263 


71 


289 


342 


118 


243 


340 


305 


89 


214 


63 


273 


287 


105 


198 



Meta- 
carpo- 



Primitive : Eohippus 

Intermediate: 

Mesohippus bairdii 

Mesohippus sp 

Hypohippus osborni 

Neohipparion whitneyi 

Equus kiang 

Equus caballus " Elmer Weeks" 

Equus scotti 

Retrogressive: Hippidion neogaeum. 



53 

69? 

68 

99 
116 
100 

78 

84 

72 



Cursorial light-bodied artiodactyls. — These animals 
are convergent with or analogous to the Equidae, 
yet in the elongation of the tibia they surpass even 



the Hipparion {Neohipparion) ratios and attain in 
Gazella and Antilocapra the highest speed ratios 
known. The list is as follows: 



Measurements, in millimeters, and ratios of limbs of cursorial artiodactyls 



Femur 


Tibia 


Tibio- 
femoral 
ratio 


Mts in 


Meta- 
tarso- 
femoral 
ratio 


Humer- 
us 


Radius 


Radio- 
humeral 
ratio 


Mtc III 


Meta- 

carpo- 

humeral 

ratio 


148 


142 


96 


72 


52 


118 


100 


84 


68 


57 


94 


103 


109 


62 


66 


74 


62 


83 


42 


56 


253 


295 


116 


255 


100 


198 


223 


112 


208 


105 


183 


223 


121 


183 


100 


133 


168 


126 


180 


135 


210 


260 


123 


218 


103 


164 


202 


123 


213 


130 


140 


176 


125 


132 


81 


93 


118 


126 


134 


144 



Primitive : 

Eotylopus reedi 

Tragulus napu 

Forest-living: Odocoileus hemionus. 

Plains-living: 

Antilope cervicapra 

Antilocapra americana 

Gazella dorcas 



738 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Stibcursorial deep-bodied artiodadyls. — There are I similar ratios. These are the OUgocene Oreodon and 
two types, ancient and modern, which have strikingly 1 the modern Sus. 

Measurements, in miUimeters, and ratios of limhs of subcursorial artiodadyls 





Femur 


Tibia 


Tibio- 
femoral 
ratio 


Mts III 


Meta- 

tarso- 

femoral 

ratio 


Humer- 
us 


Radius 


Radio- 
humeral 
ratio 


Mtc III 


iV'Ieta- 

carpo- 

humeral 

ratio 




161 

248 


142 
215 


88 
80 


62 
86 


38 
34 


138 
208 


113 

168 


81 
80 


57 
77 


41 




37 







Cursorial heary-bodied artiodadyls. — These relatively 
heavy-bodied animals were for the most part derived 
from light-limbed and light-bodied cursorial and 
subcursorial ancestors. This is positively known to 
be true of the'ancestry of the camels. They therefore 
are to be compared with the heavy-bodied cursorial 



Equidae as representing a retrogression from the high- 
speed ratios of the ancestral light-bodied forms. The 
excessively long limbs and high limb ratios of the 
giraffe are doubtless an adaptation to the peculiar 
feeding; habits. 



Measurements, in millimeters, and ratios of limbs of cursorial artiodadyls 



Femur 


Tibia 


Tibio- 
femoral 
ratio 


IMtsIII 


Meta- 

tarso- 

femoral 

ratio 


Humer- 
us 


Radius 


Radio- 
humeral 
ratio 


Mtc III 


470 


400 


80 


325 


66 


363 


455 


83 


330 


369 


355 


96 


243 


65 


290 


293 


101 


198 


466 


550 


118 


630 


135 


435 


698 


160 


618 


430 


454 


105 


350 


71 


334 


358 


107 


342 



Meta- 

carpo- 

humeral 

ratio 



Plains-living : 

Camelus arabicus 

Bison bison 

Girafia camelopardalis 

Torest-living : Cervus megaceros. 



56 

60 

143 

103 



Graviportal sJiort-limbed digitigrades. — These ani- 
mals are directly derived from ancestral mediportal 
digitigrades through increasing weight of body and re- 
duction in the length of the limbs. They are not 



short-footed but medium-footed, or mesatipodal. 
Three of them (Teleoceras, Metamynodon, Hippopo- 
tamus) are or probably were aquatic or amphibious. 



Measurements, in millimeters, and ratios of limhs of graviportal digitigrades 



Toxodon sp 

Teleoceras fossiger 

Metamynodon planifrons. 
Hippopotamus amphibius 



Femur 


Tibia 


Tibio- 
femoral 
ratio 


Mts III 


Meta- 
tarso- 

femoral 
ratio 


Humer- 
us 


Radius 


Radio- 
humeral 
ratio 


Mtc III 


577 


325 


56 


101 


17 


387 


298 


77 


147 


408 


233 


57 


105 


35 


°305 


238 


78 


114 


480 


280 


58 


118 


34 


393 


320 


81 


153 


498 


332 


67 


130 


36 


395 


270 


68 


152 



Meta- 

carpo- 

humeral . 

ratio 



38 
37 
39 

38 



" Estimated. 

Graviportal long-limbed digitigrades. — The Oligo- 
■cene titanotheres are the only known quadrupeds of 
this type. They are distinguished by being relatively 
long-hmbed in contrast with the short-limbed group 
just described and with the mediportal and graviportal 
rhinoceroses. A striking fact is that the limb ratios 
.are very similar to those in the low-bodied, short- 



limbed class above described. Thus the ratios remain 
the same in low-bodied and high-bodied heavy forms 
of moderate speed. It is noteworthy that the hind 
limb of the titanotheres parallels that of the elephant 
or rectigrade group and is quite unlike that of the 
other graviportal digitigrades. 



MECHANICS OF LOCOMOTION 

Measurements, in millimeters, and ratios of limbs of Oligocene titanotheres 



739 





Femur 


Tibia 


Tibio- 
femoral 
ratio 


Mts III 


Meta- 
tarso- 

femora] 
ratio 


Humer- 
us 


Radius 


Radio- 
bum eral 
ratio 


Mtc III 


Meta- 

carpo- 

bumerai 

ratio 


Brontotherium gigas _ 


780 
770 


427 
430 


54 
55 


200 
"220 


20 

38 


528 
620 


478 
520 


90 

83 


214 
240 


40 

38 


Menodus trigonoceras _ _ . 





Gramportal rectigrades. — The graviportal rectigrades 
are heavy-bodied, long-limbed, short-footed (brachy- 
podal) quadrupeds, typified by the elephants. The 
type has been called by Gaudry "rectigrade." We 
do not yet know the stages of graviportal progression 
in the evolution of the Proboscidea. Our knowledge 



of these stages must be gained by a comparison of the 
ancestral series, which is still incomplete. The 
measurements given below show the relatively short 
tibia and radius and the very short metatarsus and 
metacarpus, which are paralleled by those of the 
graviportal Amblypoda and Pyrotheria. 



Measurements, in millimeters, and ratios of limbs of graviportal rectigrades 



Tibio- 
femoral 
ratio 



IVIeta- 
tarso- 
femoral 
ratio 



Radio- 

bumeral 

ratio 



American mastodon (Mastodon americanus) 

Indian elephant (Elephas indicus) 

African elephant (Loxodonta africanus) 

Uintatherium mirable (Eocene) 

Pyrotherium sp. (Eocene) 



° 1, 020 

1,020 

1,050 

695 

622 



705 
618 
755 
360 
351 



117 

138 

144 

70 



540 

462 



670 
685 
870 
380 
238 



165 
183 
205 
106 



18 
33 
30 
19 



SUMMARY OF CTTRSORIAI AND GRAVIPORTAL PROPOR- 
TIONS OF SEGMENTS OF IIMBS OF UNGULATES 

FEATURES CONSIDERED 

The foregoing principles of evolution of the ungulate 
limbs are again illustrated in the widely varying types 
of locomotion comprised within the nine typical 
families of the great order Perissodactyla. These 
families range from the extreme cursorial to the ex- 
treme graviportal adaptation. A still more remarka- 
ble fact is that within several single families both 
extremes of adaptation are developed. The habits 
and habitats of these families and subfamilies are 
treated in Chapters I (Preliminary), II (Adaptive 
radiation), X (Origin and evolution), and VIII 
(Musculature of titanotheres and other Perissodac- 
tyla) of this monograph. 

We may now examine the perissodactyls in respect 
to the length, proportions, and other adaptive features 
of the scapula, humerus, radius, ulna, ilium, femur, 
tibia, and fibula. 

The student who desires to acquaint himself with 
the muscles which control the shifting rugose areas of 
origin and insertion should consult Chapter VIII. 

The foregoing summaries and comparisons of limb 
ratios in the ungulates generally help to a compre- 
hension of the adaptations of the arches and limbs of 
101959— 29— VOL 1—4 



the titanotheres and other Perissodactyla, because 
these animals naturally parallel members of other 
groups of ungulates. 

The arch and limb elements of the Perissodactyla 
are displayed in the following synoptic figures: 

Figures 671-673. Condylarthra and Perissodactyla. 

Figure 674. Artiodactyla and Perissodactyla. 

Figures 675, 676. Lophiodontidae (Helaletinae) (subcursorial) . 

Figure 676. Tapiridae (mediportal) . 

Figure 678. Palaeotheriidae (cursorial to mediportal) . 

Figure 679. Hyracodontidae (subcursorial). 

Figure 679. Amynodontidae (mediportal to graviportal) . 

Figures 679-682. Rhinocerotidae (subcursorial to graviportal). 

Figures 683, 684. Equidae (cursorial). 

Figures 685, 686. Brontotheriidae (mediportal to graviportal) . 

A comparative survey of these figures bone by bone 
or segment by segment demonstrates afresh that in 
each family respective adaptations to mediportal, to 
graviportal, or to cursorial locomotion results in con- 
vergent, parallel, or homoplastic forms, proportions, 
ratios, and indices. 

Under this masterful influence of teleogeny, or 
analogous adaptation, which masks, dominates, or 
completely conceals the syngenetic or ancestral char- 
acters, there still remain two causes of distinction or 
separation between members of different phyla: 



r40 



TITAXOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



(1) Certain features of original, or syngenetic 
resemblance survive; (2) the adaptation of structure 
is rarely exactly analogous, because the functions or 
movements of the limbs in two unrelated forms are 
rarety, if ever, exactly analogous. 

For example, the mediportal titanothere Mesnti- 
rhinus may imitate the mediportal Tapinis; the result- 
ing limb proportions and forms are closely but not 
exactly similar. Likewise the femur of the gravi- 
portal Brontotherium is not exactly similar to that of 
any other known graviportal form. The differences 
are due to differences in the gait, step, or mode of 
progression, as well as to the syngenetic influence of 
the ancestral forms from which these types respec- 
tively converged. 



FORMS OF SCAPULA 

That the proportion and form analogies, if not 
exact, are very close, is demonstrated over and over 
again in almost any bone we may select. For example, 
in the scapula there are three chief forms, which are 
produced repeatedly in difl^erent families — namely, 
the "mediportal," the "graviportal," and the "cur- 
sorial."' 

The primitive scapula of Phenacodiis is a suboval 
type from which that of Tapirus, Mesatirhinus , and 
other primitive mediportal scapulae are readily de- 
rived. A feature of the primitive scapula is the 
prominent acromial process of the spine, which is re- 
tained in the amblypod Uintatherium but lost in most 




ABC D 

Figure 671. — Five types of scapula 

A, Cursorial (.Mcsohippus); B, primitive (.Phenacodus); C, mediportal (.Tapirus); D, heavy mediportal {Palaeoiherium); all one-sixth 
natural size. E, Graviportal (Uintatherium); one-eighth natural size. 



The general contrast in the form and proportions of 
the limb segments between cursorial and graviportal 
types may be summarized as follows: 

Contrast between cursorial and graviportal types 



Arch and limb elements 


Cursorial 


Graviportal 


Neck of scapula ._ . 


Contracted. 

Lengthened. 

jNIore horizon- 
tal. 

More horizon- 
tal. 

Raised. 

Abbreviated. 

Elongated. 

Elongated. 

Abbreviated. 

Elongated. 


Expanded. 


Blade of scapula 


Axis of pelvis. 




Axis of humerus and femur 

Muscular insertions on hu- 
merus and femur. 


More vertical. 

Lowered. 

Elongated. 




Mts III and Mtc III 

Tuberosit}' of humerus 

Great trochanter on femur 


Abbreviated. 

Elongated. 

Abbreviated. 



Perissodactyla, in which we observe a prominent 
tuber spinae. 

The mediportal scapula, as illustrated in Tapirus, 
with graviportal variations seen in Palaeotherium mag- 
num and in Rhinoceros unicornis, is ovate, elongate, with 
rounded superior border and with equally developed 
fossae for the supraspinatus and infraspinatus muscles. 

The graviportal scapula as illustrated in Metamy- 
nodon, Teleoceras, and Brontotherium is, on the contrary, 
trihedral, wdth a broad angulate extension of the pos- 
terior border for the infraspinatus muscles. Its 
analogies are with the scapulae of Elephas, Uintathe- 
rium, and other large graviportal animals. 

The cui'sorial scapxila as illustrated in Neohipparion, 
Antilocapra, Hyracodon, of speed type, is long-necked, 
fan-shaped — that is, expanded above — with progres- 
sive enlargement of the infraspinous fossa and corre- 
sponding diminution of the supraspinous fossa; the 
superior border is truncate rather than rounded or 
angulate. 



MECHANICS OF LOCOMOTION 



741 



The evolution of the scapula is analogous to that of 
the ilium, chiefly in the following respects: (1) Length 
of the neck or peduncle; (2) distance between the areas 
of origin and insertion of the muscles; (3) breadth or 



contraction of the superior border; (4) support of 
thorax by scapula as compared with support of viscera 
by ilium. 



Comparison of mediportal, graviportal, and cursorial types of ungulates 



Mediportal 


Graviportal 


Cursorial 


Scapula elongate, suboval or quadrate; 


Scapula depressed, subtriangular; upper 


Scapula vertically elongated triangle; 


upper border rounded. 


border acute. 


upper border truncate. 


Neck elongate but not constricted. 


Neck sessile or abbreviate. ■ 


Neck elongate and constricted. 


Tuber spinae persistent. 


Tuber spinae expanded. 


Tuber spinae reduced. 


Prespinous and postspinous fossae of 


Postspinous fossa triangular or expanded 


Prespinous fossa reduced. 


ecjual size. 


superiorly. 




Examples: Palaeosyopinae, Tapiridae, 


Examples: Brontotheriinae, Amynodon- 


Examples: Equidae, Hj'racodontidae, cur- 


Palaeotheriidae, Oligocene Rhinocero- 


tidae, Rhinocerotidae (teleocerine) . 


sorial Palaeotheriidae (Paloplotherium 


tidae. 




minus), subcursorial Lambdotheriinae. 



FORMS OF HUMERUS 

The distal end of the humerus is a highly distinctive 
region in the ungulates because of the profound 
changes of function in the passage from the ancestral 
unguiculate type, in which the manus has powers of 
pronation and supination, to the specialized ungulate 
type, in which the manus is restricted to flexion and 
extension. The progress of this change of function 
is reflected in the structure of the distal articular 
surface of the humerus (rotula, capitellum) and of 
the inner and outer rugosities for the muscular attach- 
ments, which are loiowij as the entocondylus and 
ectocondylus. The ancestral ungulate arrangement 
of these condyles is as follows : 

Region of entocondyle, for the attachment of 
internal ligament, pronator teres muscle (reduced), 
flexor carpi radialis muscle, flexor profundus digi- 
torum muscle, flexor sublimis digitorum muscle, pal- 
maris longus muscle. 

Region of ectocondyle, for the attachment of exter- 
nal ligament, supinator longus muscle,^" extensor 
communis digitorum muscle, extensor minimi digiti 
muscle, extensor carpi ulnaris muscle. °' 

In the passage from the unguiculate to the ungulate 
type we observe the following general laws of trans- 
formation of the ectocondyle, entocondyle, and rotula: 

(1) All primitive ungulates (such as Periptychus, 
Phenacodus, Meniscotherium, Corypliodon, Uintatlie- 



rium) inherit a large entocondyle and a relatively 
small ectocondyle. 

(2) In all higher ungulates this condition is exactly 
reversed: the entocondyle is reduced or wanting, the 
ectocondyle is greatly enlarged. 

(3) In quadrupeds of intermediate size, or medi- 
portal types (such as Tapirus), the entocondyle and 
ectocondyle are equally developed. 

Thus the greatest muscular rugosities of the primi- 
tive graviportal Corypliodon are entocondylar, or on 
the inner side of the lower end of the humerus, whereas 
the greatest muscular rugosities of the progressive 
graviportal Rhinoceros are ectocondylar. In other 
words, there is a transfer of the balance of muscular 
power from the inner pronator and flexor side of the 
limb to the outer supinator and extensor side, together 
with a change of function of the extensor carpi ulnaris 
into a flexor of the forefoot (Windle and Parsons, 
1901.1, pp. 700, 701). 

In the meantime the distal articular extremity of 
the humerus, which is primitively composed of the 
rotula for the ulna and capitellum for the radius, an 
asymmetric arrangement, becomes a more symmetric 
joint. This symmetry of the inner and outer sides 
is attained only in the highly specialized cursorial 
elbow joint of Equus, which is strictly limited to 
flexion and extension. 

These four types of humeral adaptation may be 
summarized and contrasted as follows: 



Types of humeral adaptation 



Intermediate mediportal 
(e. g., Tapirus) 



Cursorial (e. g., Equus) 



Entocondyle 

Ectocondyle 

Rotula and capitellum 



Medium. 
Asymmetric. 



Medium. 
Medium. 
Asymmetric. 



Greatly reduced. 
Greatly enlarged. 
Asymmetric. 



Reduced. 
Reduced. 
Symmetric. 



5" Arises from the supinator crest above the ectocondyle. The supinator brevis is absent or vestigial in ungulates. It originally arose partly from the ectocondyle 
(Windle and Parsons, 1897.1, p. 402; 1901.1, p. 701). 

51 Serves as a flexor of the carpus in ungulates (Windle and Parsons, 1903.1, p. 391). 



-42 



TITAN OTHEKES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



The distal end of the humerus is thus highly dis- 
tinctive as between cursorial or graviportal function. 
It exhibits in the mediportal Tapirus a balanced con- 




FiGTJRE 672. — Forms of distal articular surface of humerus 

A, Primitive (,Phenacodus); B, primitive (Pantolamida); C, mediportal (Tapirus); D, 
subgraviportal (Limnohyops); E, graviportal digitigrade iShinoceros); F, graviportal 
rectigrade ( Coryphodon); G, cursorial (Equus). All one-third natural size. 

dition of muscular attachment upon the entocondyle 
and ectocondyle; similarly the deltoid crests are 
moderately developed and the tuberosities are high. 



A marked characteristic of graviportal adaptation 
is the rapid development of the ectocondylar crest 
and of the extensor carpi radialis muscle; this develop- 
ment is initiated in Palaeosyops and is carried to an 
extreme in Rhinoceros and the later titanotheres. 
The entocondylar process diminishes, and the 
humeral trochleae are oblique, asymmetrical. 

In cursorial types like the Equidae neither the 
entocondyle nor ectocondyle relatively increases, 
obviously because the lateral or pronating and 
supinating movements are almost wholly replaced 
by orthal or extensor and flexor movements. In 
cursorial forms the humeral trochlear convexities 
become vertical; in graviportal forms they remain 
oblique. 

ANALOGOUS ADAPTATION IN HUMERUS AND FEMUR 

Reviewing the following series of comparative 
figures and descriptions of the limbs of the ungulates, 
and especially of the perissodactyls, we deduce 
certain laws of transformation and transposition of 
areas of muscidar attachment that are almost as 
constant as the allometric linear changes of pro- 
portion set forth in section 1 of this chapter. 

Analogous principles of cursorial and graviportal 
adaptation operate in general in the fore and hind 
limbs, but there are some exceptions. The most 
striking contrast is observed in the muscular attach- 
ments at the upper end of the humerus and femur 
respectively. 

In graviportal progression the tuberosities of the 
humerus are raised and expanded and the analogous 
great trochanter of the femur is lowered and de- 
pressed. Conversely, in cursorial progression the 
tuberosities of the humerus are lowered and de- 
pressed and the great trochanter of the femur is 
raised and expanded. 
The chief divergences and parallelisms between the 
transformations of the humerus and femur are shown 
in the accompanying table. 



Divergences and parallelisms in the transformation of the humerus and femur in ungulates 



Adaptation 


Humerus 


Femur 


Mediportal 


Of medium length; equal to or exceeding the radius. 
Great tuberosity above level of head. 
Deltoid crest on upper third of shaft. 
Entocondylar and ectocondylar crests subequal. 
Elbow flexed. 


Of medium length; equal to or exceeding the tibia. 

Great trochanter above level of head. 

Third and second trochanters on upper third of shaft. 

Knee flexed. 


Graviportal 


Oblique. 

Elongate as compared with radius. 

Great tuberosity expanding and elevated. 

Deltoid crest expanding and lowered on shaft. 

Ectocondylar crest enlarged (entocondylar crest 

relatively reduced) . 
Olecranon of ulna depressed. 
Elbow less flexed. 


Oblique or vertical. 

Elongate as compared with tibia. 

Great trochanter depressed. 

Third trochanter expanding and lowered on shaft. 

Second trochanter sessile. 

Knee less flexed or straight. 


Abbreviate as compared with radius. 
Great tuberosity not elevated. 

Cursorial | Deltoid crest rather high on shaft. 

Ectocondyle and entocondyle reduced. 
Elbow fie.xed. 


Abbreviate as compared with tibia. 

Great trochanter erect, elevated. 

Third and second trochanters high on shaft. 

Knee flexed. 



MECHANICS OF LOCOMOTION 



743 



FORMS OF ILIUM 

In the evolution of the ihum, as in the evolution of 
the scapula, teleogenetic or analogous adaptation to 
weight or to speed is far more potent and conspicuous 
than ancestral influence. Of the ilium, as of the 
scapula, there are four main types, the primitive (such 
as that of PJienacodus) , the mediportal {Tapirus), the 
graviportal {Titanoiherium, Elephas), and the cursorial 
(NeoMpparion). These four types are fashioned by 
the demands of the musculature and the movements 
of the limbs. 

The ilium of titanotheres of the early graviportal 
type (Manteoceras) and of the advanced graviportal 
type (Brontops) is analogous to that of the heavy, 



Five types of ilia are illustrated in A, B, C, D, and 
E of Figure 673 : 

1. The primitive hypothetical ancestral form is shown in A. 

2. The superior border of the ilium of all the early Eocene 
perissodactyls, so far as known, exhibits the subcursorial form 
shown in B. 

3. The Eocene titanotheres fall into the incipient or sub- 
graviportal form E, which represents a considerable advance on 
the ilium of Tapirus. It is an intermediate form, with greatly 
broadened ilia, but still exhibits the interrupted superior 
border &. 

4. The Oligocene titanotheres fall into the extreme gravi- 
portal form F, with greatly expanded ihac border. This is the 
extreme weight-bearing type. 

5. The Equidae exhibit an extreme cursorial form (D) in 
Equus. 




Figure 673. — Seven typical forms of ilia, showing cursorial and graviportal adaptations 

A, Primitive subcursorial (Phenacodus), one-sixtli natural size; B, cursorial (Hyrachyus), one-sixth natural size; C, cursorial 
(Mesohippm), one-fourth natural size; D, extreme cursorial (Equus), one-eighth natural size; E, subgraviportal (Man- 
teoceras), one-eighth natural size; F, graviportal (Brontops) , one-t'^eUth natural size; G, graviportal (Teleoceras), one- 
eighth natural size. 



slow-moving rhinoceroses {Metamynodon, Teleoceras). 
It presents wide differences from the ilia of extremely 
cursorial types, such as Hyrachyus, Mesohippus, and 
Equus. 
The general contrast is shown below: 



Cursorial characters 


Graviportal characters 


Supra-iliac border narrow, 


Supra-iliac border broad, con- 


concave. 


tinuously arched, convex. 


Tuber coxae and tuber sacrale 


Tuber coxae and tuber sacrale 


separated by long, thin 


continuous, rugose. 


crista iliaca. 




Ilium relatively narrow and 


Ilium short, spreading. 


long. 




Iliac peduncle narrow, long. 


Iliac peduncle broad, short. 



To understand the evolution of the ilium we must 
distinguish between the crests (cristae) or borders 
and the fossae areas or surfaces, starting with the primi- 
tive mammalian ilium (A) and following the progressive 
changes in the main attachments for the abdominal 
muscles and the flexors and extensors of the limbs. 
The actual transitions between this primitive mam- 
malian (A) of condylarthrous or unguiculate type and 
the oldest mediportal ungulate type are not known, 
so that the comparisons made below are somewhat 
conjectural. 

It is noteworthy that whereas there is a wide con- 
trast in the scapulae of Tapirus (mediportal) and of 
Equus (cursorial), the ilia of the same animals are 
not greatly different in form (fig. 674). 



744 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



1. In the mediportal ilium of Tapinis cc (tuber 
coxae) corresponds broadly with the primitive anterior 
border of the ancestral ilium of the unguiculates and 
condylarths (A). It serves chiefly for the attach- 
ment of the obliquus abdominis externus and o. a. 
internus muscles. 



3. As compared with the even border of the unguicu- 
late, creodont, or condylarth ilium the raised border 
ss (tuber sacrale) appears to be a neomorph, or 
secondary upgrowth of the suprasacral border of the 
ilium, designed on the dorsal fossae for the expansion 
of the gluteus medius and longissimus dorsi muscles, 




Figure 674. — Adaptive forms of ilia and sacral attachment 
Ai-A', Mediportal (Ai, A^, Tapirus terrestris, juv.; A', the same, adult; AS T. indicus. adult); B, cursorial (,Eguus); C, subgraviportal (Manteoceras); 
C, the same with sacrum; D, graviportal {Bhinoceros); E, cursorial (.Eangifer). S, Sacral attachment; S', suprasacral fossa; ss, tuber sacrale; 
cc, tuber coxae; b, region between ss and cc. All two-filteenths natural size. 



2. The smooth, thin intermediate crista cs broadly 
compares with the dorsal border of the primitive ilium 
and is mainly for the insertion of the fascia of the 
longissimus dorsi, a muscle which is much more 
important in the ungulates than in the unguiculates. 



and on the ventral side for the ligamentous fixation of 
the pelvis to the sacrum. The gluteus medius, with 
its profound insertion in the great trochanter (f) 
and third trochanter (t'") of the femur, is one of the 
most important propellers of the heavy ungulate body. 



MECHANICS OF LOCOMOTION 



745 



4. The antero-inferior or ventral triangular surface 
(fossa iliaca) serves mainly for the attachment of the 
iliacus muscle, a flexor of the femur which is inserted 
into the second trochanter (t")- 

Starting with this interpretation of the adaptation 
of the bony surfaces of the ilium to its muscular func- 
tions we observe the following simple principles: 
(1) The expansion of the borders cc and ss is greater 
or less according to the respective development of 
weight or of speed; (2) thus in cursorial types such 
as the Equidae the borders cc and s.s remain rela- 
tively narrow, while the thin crista iliaca between 
them remains relatively broad; (3) the expansion of 
the border ss and of the adjacent suprasacral fossa 
for the attachment of the gluteus medius muscles 
is a universally progressive character 
in mediportal and graviportal ungu- C%N 

lates and is seen in different degrees \| .( 

in Tapirus, Equus, Rhinoceros, Pal- '^Ly 

aeosyops, and other Perissodactyla, 
but it is relatively far more marked 
in the graviportal Artiodactyla (such 
as Hippopotamus) than in the Peris- 
sodactyla; (4) the expansion of the 
border cc is also universal in medi- 
portal and graviportal quadrupeds 
{Amynodon, Rhinoceros, Brontothe- 
rium) with a heavy abdomen and 
large abdominal muscles; (5) thus 
the primitive concave superior bor- 
der cs or crista iliaca is contracted 
{Rhinoceros) and then obliterated by 
the expansion of the rugose borders 
cc and ss, which finally become con- 
fluent and form a continuous arch. 

All the most ancient perissodac- 
tyls, such as Eohippus, Hyrachyus, 
and even Amynodon, exhibit a flat or 
concave crista iliaca between the 
tuber coxae (cc) and the tuber 
sacrale (ss). In such cursorial and 
subcursorial types of ilium with con- 
cave or interrupted superior border 
(crista Uiaca) there is an extension 
of the gluteal muscles forward to 
unite with the fibers of the longis- 
simus dorsi, making possible the gal- 
lop. In primitive artiodactyls also 
this flat or concave superior border, 
or crista iliaca, is older than the con- 
vex superior border, because the 
subcursorial stage always appears to antedate the 
graviportal stage. 

The rounding out of the superior border (crista 
iliaca) of the ilium which accompanies the expansion 
of the gluteal abdominal muscles is proved to be a 
progressive or secondary character by the comparison 
of a great number of mediportal with succeeding 



graviportal types of animals. For example, we may 
compare the lower Oligocene Caenopus with the upper 
Miocene Teleoceras (fig. 673, G). Again, we may 
compare the mediportal upper Eocene Amynodon 
with the graviportal Oligocene Metamynodon, or the 





A B C 

Figure 675. — Limb structure of perissodact3'ls: Fore limbs of Heptodon and Tapirus 
A, Heptodon calzkaliM (family Lophiodontidae), one-fourth natural size; B, Tapirus lerreslris, and C, Tapirus 

indicus (family Tapiridae), one-sixth natural size. Showing the contrast between a primitive lower Eocene 

cursorial type {Heptodon) and a persistent modern mediportal type (.Tapirus). 

incipient graviportal Eocene Palaeosyops with the 
Oligocene Brontotherium. Thus the broad spreading 
and convex superior border is correlated with the 
vertically placed pelvis and elongate femur of such 
graviportal forms as Brontotherium and the rectigrade 
Elephas; it is correlated with the slower gait and with 
heavy abdominal muscles. 



746 



TIT.\NOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



LENGTH OF ILIUM AND OF ISCHIUM AND RATIO OF acetabulum, the length of the ischium, and the ratio 

THESE LENGTHS TO TOTAL LENGTH OF OS INNOMI- of these lengths to that of the OS innominatum in 

NATUM certain mediportal, cursorial, and graviportal types 

The absolute length of the ilium measured from the are shown in the following table : 

middle of the superior crest to the center of the 

Lengths of ilium and of ischium and their ratio to length of os innominatum in certain Perissodactyla and 

Artiodactyla 

[Measurements in millimeters; ratios in percentage] 



Ilium 


Ischium 


Length 


Ratio 


Length 


Ratio 


182 


56 


141 


44 


200 


56 


158 


44 


258 


60 


169 


40 


290 


60 


190 


40 


330 


62 


200 


38 


285 


67 


140 


33 


56 


50 


55 


50 


134 


50 


129 


50 


84 


63 


74 


47 


146 


53 


129 


47 


118 


55 


95 


45 


370 


55 


300 


45 


207 


61 


130 


39 



Length of 
OS inno- 
minatum 



PERISSODACTYLA 

Tapirus terrest ris, j uvenile (mediportal) 

Tapirus terrestris, adult (mediportal) 

Tapirus indicus (mediportal) 

Equus caballus (cursorial) 

Rhinoceros (Opsiceros) bicornis (graviportal) 

Manteoceras manteoceras (Am. Mus. 2358) (graviportal) 

ARTIODACTYLA 

Tragulus (subeursorial) 

Sus (subeursorial) 

Cervulus 

Auchenia lama (cursorial) 

Dicotyles (subeursorial) 

Hippopotamus amphibius (graviportal) 

Camelus dromedarius, juvenile (cursorial) 



323 
358 
427 
480 
530 
425 



111 
263 
158 
275 
213 
670 
337 



This table shows that we can not at present estab- 
lish any law or make any deduction as to these ratios 
with reference to speed and weight, but it brings out 
two facts of interest: First, the ilium as compared 
with the ischium is proportionately longer in the Peris- 
sodactyla than in the Artiodactyla; second, the ilium 
of the titanotheres is proportionately longer than in 
other Perissodactyla. 

The true cursorial, mediportal, and graviportal 
indices are obtained by dividing the entire length of 
the pelvis (from crest of ilium to extremity of ischium) 
by the entire breadth (across the widest part of Uium). 
In the graviportal type the pelvis is short and broad; 
in the cursorial type it is long and narrow. 

RATIOS OF LENGTH OF SCAPULA AND ILIUM TO THAT 
OF HUMERUS AND FEMUR, RESPECTIVELY 

From the constancy of the limb-segment ratios in 
graviportal and cursorial forms it was expected that 
somewhat similar ratios would be discovered between 
the arch elements and the proximal limb segments. 
For example, in the graviportal Proboscidea the ilium 
shortens and expands while the femur lengthens; 
conversely in the cursorial Equidae the scapula 
lengthens while the humerus shortens. 

To ascertain whether there is any law underlying 
these allometric changes Gregory made a test series 
of comparative measurements and obtained the fol- 
lowing ratios, expressed as percentages (scapula -?- 



humerus = scapulohumeral ratio ; ilium -J- femur = ilio 
femoral ratio) : 

Ratio of length of scapula and ilium to length of humerus and femur 



Phenacodus primae vus (primitive) 

Phenacodus wortmani (primitive) 

Coryphodon lobatus (graviportal) 

Mastodon americanus (graviportal, rectigrade) _ _ 

Elephas indicus (?) 

Elephas africanus (graviportal, rectigrade) 

Toxodon (graviportal, digitigrade) 

Palaeosyops leidyi (mediportal to graviportal) _ _ 
Palaeos3'ops major (mediportal to graviportal) _ _ 
Brontotherium gigas (mediportal to graviportal) _ 

Hyrachyus agrarius (cursorial) 

Hyracodon nebrascensis (cursorial) 

Amynodon intermedins (mediportal) 

Metamy nodon planif rons (graviportal) 

Teleoceras f ossiger (graviportal) 

Rhinoceros indicus (graviportal) 

Mesohippus (cursorial) 

Hypohippus (cursorial) 

Neohipparion (cursorial) 

Equus scotti (cursorial) 

Equus caballus (cursorial) 

Hippidion (subeursorial) 

Cervus megaceros (cursorial, heavy-bodied) _ . 

Bison bison (cursorial, heavy-bodied) 

Camelus arabicus (cursorial, heavy-bodied) __. 
Antilocapra (cursorial) 



Scapulo- 
humeral 
ratio 



109 
104 



150 
107 



121 
114 



118 
108 
120 
131 
118 
120 
132 
140 
140 
113 
112 



Ilio- 
femoral 
ratio 



74 
66 
66 
52 



73 
61 
78 
74 
60 
66 
77 
75 
60 
77 
83 



MECHANICS OF LOCOMOTION 



747 



This table appears to show that there is no constant 
significance in the ratios of length between the scapula 
and the humerus. In the Equidae, for example, the 
scapula appears to be elongating with speed as we pass 
from Mesohippus to Neohipparion, but the scapula of 
the slow-moving Eippidion exceeds in length that of 
the swift Neohipparion, and both are surpassed by that 
of Toxodon. Similarly the abbreviation of the ilium 
as compared with the elongation of the femur reveals 
no contrast of speed or weight ratios. 

The indices, or relations of breadth to length, of the 
ilium and scapula respectively afford the really sig- 
nificant and important figures. 

SECTION 2. SYSTEMATIC COMPARISON OF THE 
PECTORAL AND PELVIC ARCHES AND OF THE 
LIMB BONES IN EIGHT FAMILIES OF PERISSO- 
DACTYLS 

TAPIEIDAE AND LOPHIODONTIDAE 

THE TAPIRS 

In limb proportions the modern tapirs (Tapirus) 
are typical persistent mediportal types. Even at 
the present time they are in an earlier stage of 
evolution than the light-limbed Eocene titanotheres 
{Mesatirhinus) ; and they are also capable of more 
rapid motion, as indicated by the greater length of 
the metapodials. 

The scapula, with its distinctively tapiroid notch 
in the anterior border, is intermediate in contour 
between the cursorial and graviportal digitigrade 
types. 

The fore-limb mediportal characters of Tapirus 
are (a) rounded rather than angulate superior 
border of scapula; (6) prominent tuber spinae of 
the scapula; (c) no acromion scapulae; {d) promi- 
nent great tuberosities of the humerus. A primi- 
tive mediportal character of the humerus is the 
equal development of the muscular attachments 
on the inner and outer sides of the distal ends of 
the humerus, as reflected in entocondyle and 
ectocondyle of equal size. A primitive character 
of both Tapirus and Heptodon is the asymmetry 
of the distal rotula and capitellum of the humerus. 

Throughout all its parts the species T. terrestris 
is slightly more cursorial in type than T. indicus. 
This difference is observed in both the fore and hind 
limbs; the olecranon of the ulna alone exhibits a 
marked difference of character. 

The hind limb of Tapirus is of the primitive medi- 
portal type. The femur is somewhat longer than the 
tibia, the great trochanter is prominent, the second and 
third trochanters are less elevated on the shaft than 
in Heptodon. As in the fore limb, T. indicus is seen to 
be more graviportal than T. terrestris, especially in the 
weight of the shaft of the femur and prominence of 
the great trochanter. 



THE LOPHIODONTS 

Heptodon is to be regarded as a precociously cursorial 
lower Eocene relative of the heavy-bodied Lophio- 
dontidae of Europe. Its radius equals the humerus in 
length and the tibia equals the femur, whereas in 
Tapirus the corresponding proportions are the reverse, 
the humerus exceeding the radius, the femur exceeding 
the tibia. This is the primitive mediportal condition. 
The radius and ulna are relatively elongate, and the 
olecranon of the ulna is erect. The tuberosity of 
humerus is moderately developed. A primitive char- 




B C 

-Limb structure of perissodactyls: Hind limbs of 
Heptodon and Tapirus 

A, Heptodon calckulus (family Lophiodontidae), one-fourth natural size; B, Tapirus 
terrestris, and C, Tapirus indicus (family Tapiridae), one-sixth natural size. 
Showing the contrast between a primitive cursorial tjiie (Heptodon) and a persistent 
mediportal type ( Tapirus) . 

acter of Heptodon is the persistent acromial rugosity 
of the spine of the scapula. 

The cursorial characters of the hind limb of Heptodon 
are seen in the elongate tibia and the elevation of the 
second and third trochanters on shaft. 

PALAEOTHEEIIDAE 

The adaptive radiation of the paleotheres (Palaeo- 
tTierium, PaloplotJierium) into animals of greater or 
less speed and weight was observed by Cuvier. The 
bones of the fore limb, here figured after De Blainville 



748 



TITANOTHERES OF ANCIE>fT WVOMING, DAKOTA, AND NEBRASKA 










A B CD 

Figure 677. — Limb structure of perissodactyls: Fore limbs of paleotlieres 

Arch and limb segments showing contrasts. A, Palaeotherium medium (subcursorial); B, Palaeotherium sp., and C, P. crassum (mediportal) ; 
D, P. latum, and E, P. magnum (subgraviportal). After De Blainville, 1839.1. One-fourth natural size. 



MECHANICS OF LOCOMOTION 



749 



(1839.1, pis., vol. 4, Palaeotherium, pis. iii and iv), 
are too incomplete to afford allometric ratios, but they 
conform strictly to the universal principles of speed 
and weight allometry. The elongate radius of Paloplo- 
therium minus parallels that of the lightest cursorial 
types of Equidae. In Palaeotherium medium the 
radius exceeds the humerus in length, a subcursorial 



parallels that of the mediportal rhinoceroses like 
R. (Dicerorhinus) sumatrensis; the elongate scapula 
of P. magnum parallels that of the graviportal (black 
rhinoceros) type, namely, R. (Opsiceros) hicornis. 

In the hind limb, similarly, the tibia of Palaeo- 
therium medium apparently exceeds in length the 
femur and presents a "subcursorial" ratio. In the 




FiGUEE 678.- 



C D 

-Limb structure of perissodactyls: Hind limbs of paleotheres 



Limb segments showing contrasts. A, Palaeotherim 
(subgraviportal). 



medium (subcursorial); B, P. latum; and C, P. crassum (mediportal); D, P. magnum 
After De Blainville, 1839.1. One-fourth natural size. 



ratio. Weight-bearing progression from P. medium 
to P. crassum and P. latum is indicated in the broad- 
ening of the collum ■ scapulae and shortening of the 
radius. 

The fan-shaped scapula of Palaeotherium medium 
parallels that of the earlier cursorial horses like 
Mesohippus; the ovate-shaped scapula of P. crassum 



femur of P. latum, P. crassum, and P. magnum the 
third trochanter is very low down on the shaft, as in 
graviportal forms; this is paralleled in the modern 
graviportal rhinoceroses such as R. (Dicerorhinus) 
sumatrensis. The knee is strongly flexed, the patellar 
facet being very oblique with reference to the shaft 
of the femur, as in early amynodonts and rhinoceroses. 



750 



TIT.iNOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASB:A 



EHINOCEROTIDAE, HYRACODONTIDAE, AMYNODONTIDAE 
The extensive series of riiinoceros limb segments 
shown in Figure 679 ilkistrates all the principles of 
contrasting speed and weight progression. These 
animals have no genetic relationship; they belong to 
fom" entirely distinct subfamilies. The following 



cursorial (EyracJiyus) , the mediportal (Caenopus), the 
lobate or vertically elongate graviportal-digitigrade 
(Rhinoceros (Opsiceros) bicornis), and the triradiate 
graviportal {Metamynodon, Teleoceras). (4) Contrast- 
ing the limbs of Triplopus and Metamynodon, we observe 
the two extremes of cursorial and graviportal allometry. 




Figure 679. — Limb structure of perissodactj'ls: Fore limbs in four American rhinoceros subfamilies 

Limb segments of (A) Triplopus cubitalis, (B) Hijrachyus agrarius, (C) Hyracodon nebrascemis, (D) Caenopus occidentalis, (E) Caenopus tri- 
dactylus, (F) Amynodon intermedius, (G) Metamynodon planifrons, and (H) Teleoceras fossiger, showing the contrasts between subcursorial 
types (Hyrachyus, Hyracodon, family Hyracodontidae), a mediportal type (Caermpus), and graviportal types (Metamynodon, faTiily 
Amynodontidae; Teleoceras, subfamily Teleocerinae). A, about two-sevenths natural size; B, C, about one-seventh natural size; D-H, 
about one- ninth natural size. 



points appear: (1) It is noteworthy that the Eocene 
rhinocerotoid types (Triplopus, Hyrachyus, Hyra- 
codon) and even the Oligocene forms (Caenopus) are 
either subcursorial or cursorial. (2) A syngenetic 
resemblance between all the rhinoceroses is seen in 
the abbreviated spine of the scapula and in the ele- 
vated position of the tuber spinae. (3) The scapulae 
represent four adaptive types, namely, the elongate 



We first observe a very pronounced difference 
between the trihedral graviportal scapulae of Meta- 
mynodon and Teleoceras and the elongate lobate 
scapula of all the Old World modern rhinoceroses. 

The graviportal progression, as illustrated in these 
rhinoceroses, advances through the following changes: 
(1) The scapulohumeral ratio decreases in the gravi- 
portal American rhinoceroses (Metamynodon, Teleo- 



MECHANICS OF LOCOMOTION 



751 




A DBF 

Figure 680. — Limb structure of perissodactyls: Fore limbs in four Old World rhinoceros subfamilies 

Aceratheriinae, Ceratorhinae (Sumatran), Rhinocerotinae (Indian), and Dicerotinae (African), represented by (A) CeTatorhinus sumatrensis, (B) C. lepiorhinus , 
(C) Aceratkerium incisivum, (D) Ceraiotkerium simum, (E) Opsiceros (Diceros) bicornis, (F) Hhinoeeros indicus, (G) M. javanicus. Somewhat less than one-tenth 
natural size. 





B C D E F 

Figure 681. — Limb structure of perissodactyls: Hind limbs of rhinoceroses found in North America 

A, Hyrachyus agrarius: B, Hyracodon nebrascemis; both one-sixth natural size. C, Caenopus occidentalis: D, Amynodon tntermedius: E, Melamynodon planifron 
F, Teleoceras fossiger; all one-eighth natural size. These represent the four subfamilies of Hyracodontinae, Aceratheriinae, Amynodontinae, and Teleocerinae. 



752 



TITANOTHERES OF ANCIENT Wl'OMING, DAKOTA, AND NEBRASKA 



ceras), in which the scapula shortens and the humerus 
elongates; not so in the Old World rhinoceroses {R. 
(Opsiceros) bicornis, R. unicornis), in which the 
scapula exceeds the humerus in length; (2) the hume- 
roradial ratio increases in Teleoceras and Metamynodon 
onh' — that is, the radiohumeral ratio decreases; (3) in 
the Old World rhinoceroses the radius remains rela- 
tirelj^ of the same length as the humerus; (4) the neck 
("collum") of the scapula broadens; (5) the postero- 
superior border of the scapula becomes depressed 
(American graviportal genera only); (6) the infra- 



This series parallels the progression characteristic 
of graviportal titanotheres in certain features but not 
in others. 

This series represents a sequence of adaptive types 
but not a phylogenetic sequence. It includes mem- 
bers of four subfamily types, namely, hyracodont 
(subcursorial), aceratheriine (mediportal), amyno- 
dont (graviportal-digitigrade), and rhinocerotine (gra- 
viportal-digitigrade). The comparison between the 
cursorial, mediportal, and graviportal limbs exhibits 
the following laws of progression from subcursorial 







Figure 682. — Limb structure of perissodactyls: Hind limbs of rhinoceroses found in the Old World 
A, CeratoThinus sumatrensis: B, Opsiceros (Diceros) iicornis; C, Hhinoccros indkus: D, E.javanicus. One-tenth natural size. 



spinous fossa of the scapula increases {Metamynodon, 
Teleoceras); (7) the supra- and infraspinous fossae are 
equal or balanced in the heavy Old World types 
(Rhinoceros, Opsiceros); (8) the tuber spinae of the 
scapula increases and points upward; (9) the great 
tuberosity of the humerus rises above the head; (10) 
the deltoid crest is lowered to the middle of the 
shaft and rises in prominent tuberosities; (11) the 
ectocondylar crest of the humerus rises on the shaft, 
the entocondylar crest disappears; (12) the ulna 
olecranon becomes recumbent and expanded distally. 



and mediportal to graviportal adaptation: First the 
tibia relatively shortens, thus the femorotibial ratio 
increases, or the tibiofemoral ratio decreases; the 
shaft of the femur broadens, flattens, and straightens 
{Amynodon, Teleoceras, Rhinoceros, Opsiceros); the 
great trochanter, which is narrow and elevated in 
cursorial hyracodonts, becomes broad and depressed 
in the graviportal amynodonts, in Teleoceras, and in 
Rhinoceros; the third trochanter shifts from above 
downward to the middle part of the shaft and becomes 
very prominent (the rhinoceroses are unique in this 



MECHANICS OF LOCOMOTION 



753 



character); the second trochanter remains on the 
upper part of the shaft and becomes less prominent; 
the femoropatellar facet becomes more anterior 
(Metamynodon) as the limb straightens out; the 
patellar facet rises on the front of the shaft of the 
femur as the knee becomes less flexed and the hind 
limb more vertical; the same facet becomes more 



which show signs of adaptation to increased weight. 
The cursorial maximum is reached in the lower 
Pliocene Hipparion {NeoJiipparion) , after which there 
is a retrogression to the more heavy-bodied Equus 
and Hippidium. 

It is striking to observe how early in geologic time 
certain cursorial characters are attained in the horse 




Figure 683. — Limb structure of perissodactyls : Fore limbs of Equidae, showing their cursorial adaptation 

A, Orokippus osbornianiis, two-sevenths natural size. B, Mesohippus hairdi, one-fifth natural size. C, Hypoliippns osborni; T>, Neohipparion 
xoliitneyi; E, Hipparion gracile; all one-seventh natural size. F, Equus kiang; G, Eguus burckdli: H, Eguus caballus; all one-ninth natural 
size. Scales approximate. After De Blainville, 1839.1, and Osborn. 



. asymmetrical (a progressive syngenetic but not dis- 
tinctively a graviportal character). 

EQUIDAE 

The accompanying figures (683, 684) represent 
individuals selected from polyphyletic series of Eciuidae 
beginning with the primitive Eocene cursorial and 
light-bodied types, including the Oligocene cursorials 
and the extremely cursorial Neohipparion, the Plio- 
cene Hipparion, and the modern heavy-bodied Equus, 



family. Among the fore limb cursorial adaptations 
of Eocene equines are (1) scapula with relatively small 
supraspinous fossa (as in cursorial Artiodactyla) ; (2) 
with narrow "collum scapulae"; (3) with fan-shaped 
blade; (4) ulna with elongate vertical olecranon. A 
persistent primitive character is the acromial rugosity 
of the spine of the scapula. 

Additional cursorial adaptations acquired after 
Eocene time are (1) reduction of tuber spinae scapu- 
lae; (2) scapulohumeral ratio increase; (3) humero- 



754 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



radial ratio decrease; (4) radiohumeral ratio increase; 
(5) metacarpohumeral ratio increase; (6) tuberosities 
of humerus become less elevated or prominent, 
whereas they become more prominent in graviportal 
digitigrades; (7) entocondj'lar crest of humerus re- 
duced, ectocondylar crest slightly expanded; (8) spine 
of scapula abbreviated. 

The speed ratios and speed contours are progres- 
sively acquired up to and including NeoMpparion, 
the most light-limbed horse known. The retrogression 
of these speed ratios through Equus asinvs and E. 
cahaUus and into the very slow-moving Hippidium 
is due to the increasing weight of these animals. 
In this graviportal retrogression (1) the olecranon of 
the vdna becomes more robust and recumbent; (2) 
the neck of the scapula broadens; (3) the deltoid and 
ectocondylar crests become more prominent. 



BRONTOTHERHDAE— THE TITANOTHERES 

The adaptive progression from left to I'ight, as shown 
in the accompanying diagrams (figs. 685,. 686), is 
from subcursorial {Eotitanops) through mediportal 
{Mesatirliinus) to graviportal {Brontothernim) types. 
In the scapula Palaeosyops, a heavy type, shows 
incipient trihedral form; Brontops becomes markedly 
trihedral with prominent tuber spinae. 

Adaptation to graviportal progression in the fore 
limb is as follows: (1) Elevation of the great tuber- 
osity of the humerus into a broad plate, (2) lowering 
on the shaft and great prominence of deltoid crest, 
(3) production of deltoid tuberosity, (4) elevation of 
ectocondylar crest, (5) obliquity and asymmetry of 
humeroradial trochlea, (6) depression and thickening 
of olecranon of ulna. 




FiGUBE 684. — Limb structure of perissodactyls: Hind limbs of Equidae, showing their cursorial adaptation 

A, Orohippus osbornianus, one-fourth natural size. B, Mesohippus bairdi, one-fifth natural size. C, Hypohippus osbornj; D, Neohipparion whitneyi; 
E, Hipparion gracile; C-E one-seventh natural size. F, Equus kiang; G, Equus asinus; H, Equus burchelli; J, Equus caballus; F-J one-ninth 
natural size. Scales approximate. After De Blainville, 1839.1, and Osborn. 



A distinctive and syngenetic character of the 
Equidae is the doubling of the bicipital groove by the 
development of a median convexity. The degenera- 
tion of the ulna and fibula inferiorly is a special equine 
and also a general cursorial character (ulna reduced 
and fibula disappears in cursorial ruminants) . 

In the hind limb the chief cursorial progression is 
as follows: (1) Femorotibial ratio decreases; (2) tibio- 
femoral ratio increases; (3) third trochanter rises on 
shaft; (4) second trochanter relatively decreases in 
prominence. The increasing asymmetry of the femoro- 
patellar facets may not be regarded as a distinctively 
cursorial progressive character because it is also 
observed in the graviportal rhinoceroses. 



The platelike form of the great tuberosity of the 
humerus is a distinctive titanothere character. An- 
other distinctive feature is that in the titanotheres the 
radius increases in length as compared with the 
humerus, whereas in graviportal rhinoceroses the 
radius either remains of the same length or abbre- 
viates {Metamynodon, Teleoceras). 

A relatively straight or vertical hind limb, with 
little flexure of the knee joint, anterior patellar facet, 
and relatively short (graviportal) tibia, appears to have 
characterized the titanotheres from the middle Eocene 
onward. The graviportal progression, as seen in 
Figure 686, results in the following changes: (1) Reduc- 
tion in size of the crest of the great trochanter (rather 



MECHANICS OF LOCOMOTION 



755 




FiQUEB 685. — Limb structure of perissodactyls : Fore limbs of Eocene and Oligocene titanotheres belonging to several phyla 

These limb segments (proximal elements above, distal below) show the range of adaptation, which includes subcursorial (A), mediportal (B-F), and graviportal 
(Q) types. A, Lambiotherium, one-third natural size. Bi-Bj, Limnohyops: C1-C3, Palaeosyops; Di, D2, Manteoceras; Ei, Ej, Mesatirhinus; F1-F5, Dolicliorhinus; 
B-F one-eighth natural size. G1-G5, Bronio-ps, one-twelfth natural size. Scales approximate. 

101959— 29— VOL 2 5 



756 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



than its depression as seen in other graviportal types) ; 

(2) lowering of the lateral rugosity of the great tro- 
chanter to unite with that of the third trochanter; 

(3) depression and lowering of the third trochanter 
from the upper to the middle portion of the shaft 
(paralleling Astrapofherium and differing from the 
typical rectigrades, such as Mastodon, in which the 



noteworthy in Manteoceras and extreme in Bron- 
totherium.; (6) straightening, flattening, and elongation 
of the shaft of the femur to parallel the graviportal 
rectigrade type (Elephas) far more closely than the 
graviportal digitigrade type of the rhinoceroses. 
Both the syngenetic and teleogenetic characters of the 




A B C D E F G 

Figure 686. — Limb structure of perissodactjds: Hind limbs of Eocene and Oligocene titanotheres belonging to 

several phyla 

These limb segments (proximal elements above, distal below) show the range of adaptation, which includes subcursorial or primitive mediportal 
(A), mediportal (B-F), and graviportal (G) types. A, Eoiiianops princeps: B, MesatirJiinus petersoni; C, Palaeosyops major; D, Limnohyops 
monoconus: E, Manteoceras vianteoceTas: F, DoUchorhinus hyognathus; G, Brontops robustus. A-F, one-ninth natural size; G, one-fourteenth 
natural size. Scales approximate. 



second and third trochanters become sessile); (4) low- 
ering of the second trochanter toward the middle por- 
tion of the shaft (thus differing from other graviportal 
digitigrades, such as Rhinoceros, in which the second 
trochanter remains high up on the shaft); (5) great 
elongation of the femur and abbreviation of the tibia. 



femur are seen to be quite distinct from those of other 
graviportal perissodactyls and are very distinctive 
throughout the series. The only analogy with mem- 
bers of other families is seen in the comparison of the 
hind limb of MesatirJiinus and Tapirus terrestris, but 
this is not very close. 



CHAPTER X 



THEORIES AS TO THE ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION OF THE 
TITANOTHERES AND OTHER ODD-TOED UNGULATES 



SECTION 1. ORIGIN AND RELATIONSHIPS OF THE 
PERISSODACTYLA 

NATURE AND HABITAT OF THE ANCESTRAL 
PERISSODACTYL 

As shown in Chapter I, the branching off of the 
titanotheres from the other lower Eocene perissodactyl 
families — the horses, tapirs, and lophiodonts — points 
to the very great antiquity of the titanothere stock, 
to basal Eocene or closing Cretaceous time, when there 
probably existed a stem perissodactyl, doubtless a very 
small animal. This little quadruped was probably 
an inhabitant of the plateau regions of Asia, from 
which its descendants diverged through adaptive 
radiation into the ancestral nine great perissodactyl 
families that entered Europe and North America. 
It is certain that this adaptive radiation took place 
long before the beginning of Eocene time, for there is 
evidence that at the beginning of the Eocene epoch 
at least five of these perissodactyl families had acquired 
their distinctive family characters, especially those 
of the teeth. 

The grounds on which are based the theory of the 
derivation of the perissodactyls from a single perisso- 
dactyl stem are twofold. First, the members of each 
of the families preserve certain persistent primitive 
(paleotelic) ancestral characters, inherited in common 
from the perissodactyl stem form; second, each of 
these families exhibits in common certain potentialities 
(cenotelic) of evolutionary development, expressed 
in the progressive or advancing characters that each 
exhibits. 

This twofold affinity displayed in similar ancestral 
character and in similar progressive characters is a 
dominant feature in classification, which has been 
exemplified repeatedly in the preparation of this 
monograph. It is now seen to apply to orders as 
well as to families, genera, and species. 

The common characters of members of the perisso- 
dactyl families are listed below. 

Persistent primitive ancestral characters 

1. A complete series of upper and lower incisor and canine 

teeth, as in all primitive Ungulata, Condylarthra, and 
Amblypoda. 

2. Upper and lower grinding teeth originally with six rounded, 

conical cusps, as in the Condylarthra and Artiodactyla, 
but differing from those of the Amblypoda. 

3. Base of the skull with an alisphenoid canal for the trans- 

mission of the external carotid artery, as in Condy- 
larthra and many Creodonta; reduced or absent in Artio- 
dactyla. 



20 or 21, as compared with 
19-20, and Artiodactyla, 



4. Dorsal and lumbar vertebrae : 

Condylarthra, which have 
which have 20. 

5. Femur with a third trochanter, as in Condylarthra; reduced 

or absent in Artiodactyla. 

6. Manus with only four digits; poUex absent or much reduced; 

five digits in Cond3'larthra; five digits in primitive Artio- 
dactyla. 

7. Pes with distinctive perissodactyl ankle joint, quite distinct 

from that of Condylarthra or Artiodactyla. 

Potential new progressive characters 

8. Premolars tending to transform into the molar pattern, a 

very rare character in Artiodactyla. 

9. Manus originally mesaxonic, secondarily paraxonic; in the 

Artiodactyla originally and persistently paraxonic. 

10. Tarsus losing the ancestral middle ankle joint by flattening 

the articulation between the astragalus and the navicular; 
perfecting the tibiotarsal joint by mechanical play 
between tibia and astragalus. 

11. Fibula gradually losing its calcaneal facet; fibulocalcaneal 

facet persistent and specialized in Artiodactyla. 

12. Progressive specialization and adaptive radiation of the 

brain, skull, teeth, trunk, limbs, and feet for a great 
variety of habits and habitats in the nine different 
families. 

The characters indicated in this list debar the 
Perissodactyla from community of descent with the 
Artiodactyla (see Gregory, 1910.1, pp. 385-386), as 
was believed by Cope when he proposed the super- 
order Diplarthra to include the perissodactyl and 
artiodactyl stems. In the opinion of Osborn they also 
debar the Perissodactyla from descent from the Con- 
dylarthra, typified by Phenacodus of the lower Eocene 
and Euprotogonia of the basal Eocene, as believed by 
Cope in regarding the perissodactyl order as an off- 
shoot from the Condylarthra. Structurally, as Greg- 
ory has shown (op. cit., pp. 387-397), the ancestral 
perissodactyl resembled certain more primitive con- 
dylarths (Euprotogonia) in many features, although it 
differed from them fundamentally in the ankle joint; 
but in their potentiality of progressive evolution in 
the characters shown in Nos. 8 to 12 above, the 
momentum of the Perissodactyla can not be derived 
from the inertia of the Condylarthra. We thus can 
not agree with Gregory (op. cit., p. 397) that "the 
derivation of the perissodactyl order from the general 
insectivore-creodont-condylarth group of placentals 
seems fairly well established." It appears more 
probable that the perissodactyls sprang from some 
entirely unknown progressive placental source — large- 
brained, plastic — to which Osborn has given the name 
Ceneutheria, as distinguished from the archaic non- 
757 



758 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND' NEBRASKA 



progressive Meseutheria, which includes the Am- 
blypoda and Condylarthra, theoretical branches of 
relatively archaic and conservative stock (see Greg- 
ory, 1910.1, p. 458). This question is, however, still 
an open one and is of so much importance and inter- 
est in connection with the whole problem of the 
evolution of the hoofed mammals that it is desirable 
to review it in more detail in the light of our present 
knowledge. The final solution must await the discov- 
ery, probably in Asia, of the actual Upper Cretaceous 
and basal Eocene ancestors of the perissodactyls. 



CARPUS OF CREODONTA, AMBLYPODA, CONDYLARTHRA 

Our present (1918) knowledge of the structure of 
the carpus in these primitive orders of mammals is 
displayed in the accompanying Figure 687. All are 
pentadactyl, possessing five more or less spreading 
digits. The creodont Dissacus, the amblypod Panto- 
larribda, and the primitive condylarths Euprotogonia 
and Meniscoiherium are more or less plantigrade. The 
condylarth Phenacodus is secondarily digitigrade and 
tending toward tridactylism. 




Figure 687. — Carpus of Creodonta, Amblypoda, and Condylarthra, showing the primitive "alternating" 
placental type and the specialized "serial" type 

Pentadactyl 
Creodonta: 

A, Dissacus carnifex. Four-ninths natural size. A primitive placental type. Magnum small, supporting centrale and halt oi lunar; lunar 

wedge-shaped, resting on magnum and unciform. 
Amblj'poda: 

B, Pantolambda bathmodon, Am. Mus. 2547. Natural size. Magnum small, supporting centrale and part of lunar; lunar wedge-shaped, 

resting on magnimi and unciform. 
Condylarthra; 

C, Mmiscotherium sp. Natural size. Magnum small, supporting centrale and lunar; lunar flattened, resting on magnum and unciform. 

D, Euprotogonia. Four-thirds natural size. Magnum small, truncate, partly supporting scaphocentrale; chiefly supporting lunar; lunar 

truncate, retaining contact with unciform. 

Mesaxonic 

E, Phenacodus. Two-thirds natural size. Serial type. Magnum enlarged, truncate, wholly supporting lunar; lunar truncate, resting 

chiefly on magnum, with small contact with unciform. 



HYPOTHETICAI ORIGIN OF THE PERISSODACTYLA FROM 
THE CONDYLARTHRA 

HISTORY OF OPINION 

Cope (1884-1898) regarded the Condylarthra, 
typified by Euprotogonia and Phenacodus, as repre- 
sentives of the long sought five-toed ancestors of the 
perissodactyls. Osborn (1890.51) at first accepted 
this view and with Cope regarded the manus of 
Phenacodus as close to the primitive foot of the Peris- 
sodactyla and Ungulata; but in 1893 (1893.82, p. 89) 
he abandoned Cope's view and proposed the theory 
that the Condylarthra were part of an archaic ungu- 
late radiation not directly ancestral to the perisso- 
dactyls. Subsequently Osborn developed the idea 
(1894.89) that the source of the Perissodactyla was to 
be sought far back of the Condylarthra. 



The serial carpus of Phenacodus, as first observed by 
Matthew, is obviously secondary when compared 
with that of Euprotogonia. This important fact 
compels us to abandon Cope's great hypothesis that 
the serial or taxeopod type was ancestral to that of all 
the Ungulata; certainly the Amblypoda and Con- 
dylarthra did not spring from a "serial" type. 

This fact also modifies many of the conclusions 
which Osborn reached in his "Evolution of the un- 
gulate foot" (1890.51, p. 531), a treatise based upon 
the adoption of Cope's theory that the common 
protungulate foot was serial. 

Matthew (1897.2, pp. 309-310) showed that the 
taxeopod, serial foot of Phenacodus is not primitive 
but secondary, because the foot of its ancestor 
Euprotogonia has partly alternating or displaced 
carpals. He reached the following conclusion (op. 
cit., pp. 300, 308, 309): 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



759 



While we are unable to consider Euproiogonia the direct 
ancestor of the Perissodactyla, yet it had many characters 
approximating it to them. It is probable that the common 
ancestor of the typical Ungulata was thoroughly an unguiculate 
and that the first separation into the phyla of phenacodonts, 
perissodaotyls, and artiodact3'ls accompanied or preceded the 
development of hoofs from claws [p. 300]. * * * This 
evidence, though not conclusive, points unmistakably toward 
an alternating carpus as the primitive one. In Phenacodus 
the carpus is more nearly serial, although it varies in different 
individuals, and there is always a lunar-unciform contact. 
But if Phenacodus is a direct descendant of Euprotogonia 
this serialism must be secondary [pp. 308, 309]. We must go 
somewhat lower down than the Torrejon [that is, Thanetian] to 
find the junction of the equine and phenacodont phyla. 

Osborn (1898.148, p. 174) accepted Matthew's 
conclusion and asreed that the 

•^ puraconc 

primitive ungulate foot was 
interlocking, or alternating 
(lunar on unciform, scapho- 
centrale on magnum), deriv- 
able from that of such a 
creodont type as Dissacus, 
from which the carpus of the 
Amblypoda and Condylarthra 
and possibly of all Ungulata 
may be derived. 

Gregory (1910.1, p. 385) 
observes : 

The question of the derivation of 
the perissodactyls as an order and 
of their relationship to the Con- 
dylarthra is still open. Cope saw 
in Phenacodus the atavus of practi- 
cally all the hoofed orders. Osborn 
holds the contrary opinion that 
Phenacodus is a hoofed offshoot of 
the Creodonta and a member of the 
Meseutheria, or small-brained Cre- 
taceous-basal Eocene orders, and 
that the perissodactyls have sprung 
from some entirely unknown 
"Caeneutheria" (p. 457). * * * 

StUl another view may be adduced — that although neither 
Phenacodus nor Euprotogonia were the ancestors of the Peris- 
sodactj'la yet they resemble those forms more nearly than do 
any other known mammals; and that the basal Eocene ancestors 
of the Perissodactyla would, if discovered, fall under the 
superorder Protungulata as defined above (p. 383). 

The Protungulata are then defined as typified by 
the condylarth Euprotogonia. Gregory reaches the 
conclusion (op. cit., p. 396) that 

The ancestral perissodactyl as thus conceived resembled the 
more primitive eondylarths such as Euproiogonia in many 
features (p. 397). * * * To conclude, the derivation of 
the perissodactyl order from the general insectivore-creodont- 
condylarth group of placentals seems fairly well estab- 
lished. * * * On the other hand, genetic derivation from 



any well-known eondylarths {Phenacodus, Euprotogonia, 
Meniscotherium) is almost equally improbable. But with 
regard to many important dental and osteological features it 
is obvious also that Euproiogonia and Phenacodus bridge 
over the structural gap between the perissodaotyls and the 
lower unguiculate orders and, in brief, that the stem of the 
perissodactyls would very likely fall under the Condylarthra, as 
redefined by Matthew (1897). 

After summing up all the characters of the stem 
perissodactyls and comparing them in detail with the 
characters of Euprotogonia Gregory concludes (op. 
cit., p. 389): 

It is not denied that the ancestral perissodactyl had a larger 
brain case and better brain than Euprotogonia. It is merely 
inferred that there was a considerable variation in brain charac- 




■Evolution of the upper molar pattern in eondylarths and titanotheres 



In the earliest known titanotheres, Lambdotherium (D) and Eotitatiops (E), the upper molars had already attained the 
bunoselenodont pattern, having two low, subcorneal internal cusps and two very large external V-shaped cusps. The 
probable mode of origin of this pattern is, however, furnished by certain of the Eocene eondylarths (A, B, C). 

A, Protogonodon, a basal Eocene condylarth. In this stage the molar is nearly triangular and the primitive tritubercular 
pattern is still evident. All the cusps are low; the protoconule and metaconule are well developed. The hypocone 
(hy) is just growing up from the cingulum. 

B, EtipTotogonia, another basal Eocene condylarth. In this animal the hypocone is more progressive, so that the tooth is 
becoming subquadrate. 

C, Edocion, a lower Eocene condylarth. This stage shows the assumption of the V-like form of the crest of the 
paracone and metacone, the reduction of protocone and metaconule, the origin of the mesostyle, and the ineipiently 
V-shaped form of the hypocone. 

D, Lambdotherium, a lower Eocene aberrant titanothere. The mesostyle is now well developed, and the crests of the 
paracolic and metacone are completely V-shaped. The parastyle is strong, the metaconule is much reduced, and 
the hypocone is now subequal to the protocone. The paracone and metacone have also shifted inward toward the 
center of the crown. All the cusps have retained their low or bunoid form. 

E, Eotitanops, a lower Eocene true titanothere. The molar pattern is now definitely titanotheroid. The protoconule and 
metaconule are vestigial, but the main outer cusps (pa, me) are still low. 

F, Manieoceras, a middle Eocene titanothere. In this stage the anteroposterior diameter of the molar is considerably 
increased, the tips of the paracone and metacone are shifted well in toward the middle of the crown, the whole outer 
wall of the tooth has become considerably deeper, the protoconule and metaconule have practically disappeared, 
and the protocone and hypocone are low, rounded cusps. 

G, Brontotheriam, SilowQr Oligocene titanothere. Marked intensification of the tendencies noted under F. Paracone 
and metacone tips at the center of the crown, outer wall of tooth much elongated, anteroposterior diameter for the 
molar increased. This completes the transformation of a tritubercular molar, with simple conical cusps, into a 
quadrangular and bunoselenodont quadiitubercular molar, with W-shaped ectoloph and low, conic internal cusps. 

ters in the Condylarthra and that the perissodactyls sprang 
from some unknown, possibly Asiatic, larger-brained form, 
which in all its dental and skeletal characters would faU under 
the Condylarthra as here defined. 

EOCENE CONDYLARTHEA NONPERISSODACTYL 

Thus in the opinion of Matthew (1897) and Gregory 
(1910) the Perissodactyla sprang from unknown 
primitive Condylarthra of Cretaceous age. 

The opinion of Osborn (1894-1918) is that since 
the known Condylarthra have specialized in a dif- 
ferent direction from the Perissodactyla, the latter 
probably sprang independently from a different 
ungulate stock. The evidence for Osborn's opinion 
is in part as follows: 



760 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Euprotogonia , besides its very numerous resem- 
blances to what we may suppose to have been the 
stem perissodactyl type exhibits the following dif- 
ferences: (1) Humerus with large entepicondyle and 
entepicond.ylar foramen; (2) carpus with lunar resting 
on magnum and unciform; (3) astragalus with round 
or convex navicular joint; (4) brain (by inference 
from Plienacodus) extremely small and smooth, with 
large olfactory lobes and small frontal lobes, or 
cerebrum. 

There are thus strong differences between the most 
primitive perissodactyl humerus, carpus, and tarsus 
and those of Euprotogonia, Phenacodus, and Menis- 
cotherium, which tend to support Osborn's view that 
the Condylarthra represent a separate specialization 
or radiation of their own, while the Perissodactyla 
sprang from an independent primitive stock, Perisso- 
dactyla primitiva, of unknown derivation but probably 
related to the insectivore-creodont group. 

This opinion is further supported by many of the 
observations presented by Gregory in his "Orders of 
mammals," and especially by his demonstration 
(op. cit., p. 393) that the carpus of the ancestral 
Perissodactyla differs sharply from that of the loiown 
Condylarthra, a point which is fully treated below. 




Figure 6S9. — Generalized carpus of 

insectivore-creodont type 

After Gregory, 1910.1. 

In the condjdarth the head of the astragalus 
(navicular facet) is hemispherical and there is no facet 
for the cuboid; the sustentacular facet is more cen- 
tral in position; there is a definite constriction, or 
neck, between the head and the proximal end; the 
trochlea is wide, with low keels. In the primitive 
perissodactyl, on the other hand, the head, or navicular 
facet, is concave in the front view and convex in the 
side view; the sustentacular facet is vertically ex- 
tended and shifted toward the middle side; there is a 
distinct facet for the cuboid, which meets the susten- 
tacular facet at an open angle; the trochlea has high, 
sharp keels. 

CONCLUSION 

The remote ancestry of the perissodactyl stem, 
whether from "Condylarthra primitiva" or from an 



independent "Perissodactyla primitiva" placental 
stock, must be left an unsettled problem awaiting 
future discovery. 

Skull and dental indices nf certain ungulates 



Phenacodus wortmani._ 
Systemodon priinaevus_ 

Eohippus venticolus 

Isectolophus 

Hyraohyus 

Triplopus 

Eotitanops 

Lophiodon, c? 

Lambdotherium 

Paloplotherium 



Zygomatic- 
cephalio 
(bueco- 
cephalic) 



39? 

47 

41 

46? 

46 

50? 

52 

41 



Facio- 
cephalic 



51 

52 

60 

56 

62 

53 

59 

49 

65? 

62 



Premolar- 
molar- 
cephalic 



26 

30 

30 

39? 

32 

30 

26 

30 

27? 

35 



Note.— Premolar-molar cephalic index obtained from p^-m'^^Iength, premaxil- 
lary to condyle. 

PRINCIPAI CHARACTERS OF THE ANCESTRAL 
PERISSODACTYLS 

The following features characterize the ancestral 
perissodactyls: 

1. Ungulates of small size and cursorial locomotion, alert 
and quick in movement. 

2. Elongate limbs, narrow and stilted feet, of perfect mech- 
anism. 

3. Tetradactyl manus, tridactyl pes with reduced or vestigial 
fifth digit, terminal phalanges narrow and cleft distally. 

4. SkuU dolichocephalic, with facial region longer than 
cranial region. 

5. Brain, visual and auditory organs, and senses well de- 
veloped. 

6. Eutherian dentition, |-:-j-;-4;|; teeth throughout similar to 
those of Euprotogonia. 

7. Diet probably succulent, tender shrubs, herbage, berries, 
tubers, etc. 

8. Limb skeleton, humerus, carpus, and astragalonavicular 
joint all of perissodactyl type. 

Tetradactylism in the manus of Perissodactyla is 
in contrast to the pentadactylism in the manus of the 
Condylarthra. Not a single trace of the poUex has 
been actually observed in any primitive Perissodac- 
tyla. That the "Perissodactyla primitiva" once 
possessed a pollex is attested by the presence of the 
extensor ossis metacarpi pollicis muscle in the tapir 
and horse. The homology of this muscle with that 
in pentadactyl mammals can not be doubted. 

Tridactylism in the pes is accompanied by an 
occasional vestige of Mtc V (Eohippus). 

For details of perissodactyl structures see descrip- 
tion of skull and carpus below, and Gregory's outline 
(1910.1, pp. 387-395) of the principal characters of 
the stem Perissodactyla. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



761 





Protitajiotheriunv eTTtoT-c/inaiuTn 

Upper Eocene (upperUinfa) 



3fa7iteoc£ras TTzaTiteoceras 

MiddleEocene (upperBridger) 




Jjimnohyops priscus 

MiddleEocene (lowerBridger) 




EotUfxnops horecdis 

Lower Eocene(Wind River) 

Figure 690. — Progressive stages of structural evolution in the skull and molar teeth of titanotheres 
In Eotitanops the facial part of the skull is longer than the brain case (cranium). In Brontotherium the face is very short and the brain 
case 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, and the inner cusps (protocone, hypocone) retain their low, conical form. All the lower molars retain the W-shaped 
crown, which increases considerably in depth. 



762 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



TETRADACTYI MANUS OF PRIMITIVE PEEISSODACTYIA 

Tetradactylism.— The manus of the primitive peris- 
sodactyls (see fig. 692, A, B, C) differs widely from 
that of the primitive Amblypoda (fig. 687, B) and 




-(sicsi) 



dist. 




which have since been discovered reveal the more truly 
primitive perissodactyl manus, which is alternating, 
not serial. 

This displacement is partly a cursorial adaptation. 
The first to comment on the real significance of the 
primitive displaced lower Eocene perissodactyl foot 
was Gregory (1910.1, p. 393), who pointed out that 
the manus of Eeptodon (see fig. 692) has an extremely 
"displaced" carpus, the scaphocentrale completely 
covering the magnum, while the lunar rests entirely on 
the unciform. Subsequent discovery has confirmed 
Gregory's observation and proved that the primitive 
perissodactyl carpus, observed in the chalicotheres, 
tapirs, and titanotheres, as well as in the helaletids, to 
which Eeptodon belongs, is completely displaced. 

A functional classification and summary of these 
foot types is as follows : 

Primitive tetradactyls. — The primitive condition of 
the perissodactyl carpus is well illustrated in Figure 
700, A-E — namely, (1) magnum very small; (2) scaph- 
oid with elongate oentrale process resting on magnum; 
(3) lunar with a very slight magnum contact, thus 
resting chiefly or entirely on the unciform. This type 
is seen in the helaletids (A), tapirs (B), titanotheres 
(C), chalicotheres (E). In the most primitive 
horses (N) the magnum has a considerable lunar 
contact. 

Functional isotetradactyls. — In these perissodactyls 
(fig. 700, D, F-J) the fifth digit is functional— that is, 
all four digits are used in these forms typified by Eoti- 



FiGtTRE 691. — -Astragali of a condylarth, 
Phenacodus. -primaevus (A), and of a 
primitive perissodactyl, Heptodon (B) 

primitive Condylarthra (fig. 687, C, D, E) in being 

tetradactyl and in the arrangement of the carpal bones. 

Both Amblypoda and Condylarthra — in fact, all other 

Ungulata, including the Artiodactyla — are pentadactyl 

in their most primitive forms. It had long been sup- 
posed, because of the opinions of Marsh on EoTiippus 

and of Deperet (1903.1) on LopModon, that 

there was adequate evidence for pentadac- 

tylism in the most primitive Perisso- 

dactyla. Deperet, in 1903 (op. cit., p. 33), 

observed : " La patte anterieure du LopModon 

etait certainement pentadactyle." This 

conclusion may perhaps rest on the presence 

of a small facet on the internal side of Mtc 

II, a facet that probably served for articu- 
lation with the trapezium, which descends 

secondarily on the internal side of Mtc II 

in the absence of Mtc I. In the most primi- 
tive manus loiown of EoTiippus Granger 
finds a similar facet but fails to find any 
evidence of Mtc I either as a splint or as 
a vestige. Such a vestige of Mtc I may 
yet be discovered, but in the absence of 
positive evidence the primitive perissodactyl 
manus must be described as essentially 
tetradactyl. 

Primitive, alternating, displaced car pals. — 
Osborn's error in regard to the origin of the 
perissodactyl foot sprang from regarding as primitive j tanops (D), HyracTiyusiG,!!), LopModon (X),m.A\>j the 
the partly displaced feet of the middle and upper j upper Eocene tapir Isectolophus (J); the magnum and 
Eocene perissodactyls; it now appears that these j lunar expand so that there is increasingly broad 
feet are mostly modified. The feet of the lower magnum-lunar contact; the lunar becomes wedge- 
Eocene Sparnacian ( = part of Wasatch) perissodactyls | shaped distally, with equal magnum-unciform facets. 





Figure 692. — Manus of Heptodon, Lambdotherium, and Eotitanops 

, Heptodov, Am. Mas. 294, a primitive perissodactyl manus of cursorial type. B, Lambdotherium, 
Am. Mus. 4880, and C, Eotitanops, Am. Mus. 296 (type), illustrate the ancestral' titanothere type 
of manus. One-half natural size. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



763 



This type of carpus seems to be characteristic of the 
isotetradactyl perissodactyls. It is in wide contrast 
to the isotridactyl condition. 

Isotridadyl perissodactyls. — In these animals (fig. 700, 
K, L, M) the fifth digit (D. V) is nonfunctional, re- 
duced, or vestigial, while the remaining three digits, 
Mtc II-IV, equally divide the weight among them. 
The triplopodines, paleotheres, and helaletids develop 
this type of foot. It is noteworthy that in all 
these animals the scaphocentrale entirely covers 
the magnum; the lunar retains little or no mag- 
num contact but rests mainly on the unciform; 
the magnum may be vertically expanded; the 
scaphoid is closely approximated to the unci- 
form, so that these bones are almost or quite in 
contact. These isotridactyl types thus retain 
the primitive scaphocentrale magnum contact. 

Monodactyl perissodactyls (fig. 700, N-P). — 
Even in the earliest Sparnacian (Wasatch) horses 
monodactylism is indicated by the expansion of 
Mtc III, accompanied by a broadening of the 
magnum, which, as seen in EoMppus (N), partly 
supports the lunar; as the magnum expands and 
flattens (0, P) the magnum-lunar contact in- 
creases. This is a condition exactly the reverse 
of that in the isotridactyl types, such as 
Colodon (M). 

Evolution of the magnum. — This primitively 
small element (fig. 724, A, C, E) of the carpus is 
very significant in the variety of its shapes. It 
expands in all directions in the functional tetra- 
dactyl types, vertically in the isotridactyl types, 
and horizontally in the monodactyl type. Its 
hooklike posterior process (fig. 704) is for attach- 
ment of a tendon connected with the flexor pro- 
fundus digitorum muscle (a process lacking in 
Phenacodus). The magnum is thus one of the 
most distinctive bones of the carpus, as stated 
by Gregory (1910.1, pp. 393-395). It under- Figure 693.- 
goes a special and very characteristic evolution Lower 
in the titanotheres. As seen from the back the 
lunar always rests largely on the magnum. 

The manus, summary. — The discovery that in 
the earliest Eocene perissodactyls the scaphoid 
rests chiefly on the magnum while the lunar rests 
chiefly on the unciform proves that the Cope "taxeo- 
pod" theory of the del'ivation of the perissodactyl 
carpus from a "serial" type like Phenacodus must be 
abandoned. The primitive perissodactyl type is a dis- 
placed type; secondary types may become more or 
less serial. This reversal of the old and apparently 
well-established "taxeopod" theory has several impor- 
tant results. First, it tends to support the view that 
Euprotogonia and Phenacodus are not to be taken as 
ancestral types of the Perissodactyla but as inde- 
pendent offshoots of an insectivore-creodont stock; 
second, it brings out the new conception that most 



of the early perissodactyls were rather light-limbed, 
narrow-footed, cursorial animals. To this agility may 
have been due their wide geographic distribution, 
in open, semiarid country, and the acquisition of 
those psychic characteristics associated with speed and 
alertness of movement which gave them a decided 
advantage in competition with the small-brained 
Condylarthra. 




-Fore and hind feet in odd-toed ungulates (perissodactyls) 

pes, upper row manus. A, Brontops roftasius, a titanothere; B, Tapirus terrestris, A 
modern tapir; C, JRhinoceros (Opsiceros) bicornis, a rhinoceros; D, Equus, the Taodem horse. 
In primitive titanotheres the main a-Kis of weight, or symmetry, passes through the middle 
(third) digit both in the fore and the hind feet. In the later titanotheres, through spread- 
ing of the foot, the main axis in the foot as a whole presents a superficial resemblance to 
that of the hippopotamus, among artiodactyls. In all recent perissodactyls, however, it 
remains in the third digit, which finally becomes extremely large, while digits II and IV 
diminish or dwindle away. 

The mesaxonic or anisotridactyl condition of the 
manus — the condition in which the median toe was 
somewhat larger than the two lateral toes — was ac- 
quired very early. This condition is seen in the early 
helaletids, tapirs, titanotheres, and chalicotheres. The 
functional tetradactyl or isotetradactyl condition, in 
which D. II-V are all functional, is more or less a 
secondary one, as observed in Menodus and Amyno- 
don. The isotridactyl condition, in which the three 
median digits, Mtc II, III, IV, divide the weight, as 
observed in Triplopus, Palaeotherium, Colodon, is also 
secondary. 



764 



TITAXOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



THE SKULL OF PEIMITIVE PERISSODACTYLA 

As the vast diversity in the structure of the modern 
perissodactyl skull, described in Chapter I, was evi- 
dently acquired during the Tertiar.v period, the 
uniformity of skull structure in lower Eocene time is 
remarkable; horses, tapirs, phenacodonts, titanotheres 
all look alike when seen from above or from the side. 
(See figs. 694, 695.) As in the feet, this uniformity 
points irresistibly to a common ancestral form, which 
would be a composite of skulls A, C, D, E. The palatal 
and occipital views of these skulls are also similar. 

Historical notes. — There is little literature on the 
primitive perissodactyl skull, because the materials 
have hitherto been so scanty. 

Deperet (1903.1) made a special study of the skull 
of LopModon, based upon L. Jeptorhynchus, from the 
gres de Cesseras deposits of the Lutetien stage, equiv- 
alent to our lower Bridger middle Eocene time. 
This skull, while retaining seven primitive charac- 
ters, correctly enumerated by Deperet (op. cit., p. 28), 
also contains several specialized characters, namely* 
broadening of the zygomatic arches, abbreviation of 
the face, and elongation of the cranium. In these 
features it is a specialized form as compared with the 
much more ancient lower Eocene Sparnacian skulls 
here examined. Deperet concluded (op. cit., p. 49) 
that the skull of the primitive Perissodactyla had 
primitive features in common with those of the Hyra- 
coidea, the Condylarthra, and especially the Ambly- 
poda; but he traced the origin of the perissodactyl 
skull back of these groups to common protungulate 
ancestors in Cretaceous time. 

Gregory (1910.1, pp. 389-391) enumerated many 
primitive characters of the perissodactyl skull in 
addition to those enumerated by Deperet, all of which 
are noted below, with the exception of the designa- 
tion of the "muzzle" as heavy. After a close com- 
parison with the skull of the Condylarthra, he con- 
cludes (p. 391): "The preceding skull characters may 
confidently be assigned to the stem perissodactyl and 
are all inherited from an insectivore-creodont-condy- 
larth plan." 

The primitive perissodactyl skull is certainly closer 
to the condylarth plan than the primitive perissodactyl 
foot. 

General cJiaracters. — As seen from above, the skulls 
of certain Insectivora, of the Condylarthra, and of 
most primitive Perissodactyla are strikingly uniform, 
or analogous, namely: 

(1) Narrow, elongate proportions, constricted ante- 
riorly; (2) small craniocerebral region; (3) capacious 
temporal fossae; (4) limited zygomatic arches parallel 
to sides of skull; (5) moderately broad orbital region, 
orbits open posteriorly; (6) abruptly constricted pre- 
orbital region of face, similar to that of Solenodon; 
(7) occiput narrow and overhanging the condyles. 



Seen from the side, we observe a striking similarity 
of proportion and in the contour of the cranial 
profile, except in the specialized skull of Heptodon and 
of Helaletes, namely, (8) superior cranial profile sim- 
ple, arched, with sagittal crest; (9) orbit placed mid- 
way between occiput and premaxillary symphysis; (10) 
face sloping downward toward anterior nares; (11) 
occiput slightly overhanging condyles; (12) temporal 
fossae large, with slender zygomatic arches; (13) area 
occupied by grinding teeth limited. 

As to proportions, the ancestral skull is dolicho- 
cephalic — that is, the basilar length greatly exceeds 
the zygomatic width — and it is orthocephalic — that 
is, the palatal and basicranial regions are in parallel 
horizontal planes, whereas among modernized Perisso- 
dactyla many skulls are cyptocephalic — that is, the 
face is upturned or downturned. Also the face 
slightly or considerably exceeds the cranium in length, 
a condition technically known as proopic dolicho- 
cephaly, or dolichopy. 

Adaptation. — The constricted anterior portion of the 
face, including the small terminal anterior nares and 
premaxillaries uniting with the nasals, indicates the 
presence of a long, narrow tongue, a short upper lip, 
limited prehension by lips, small lateral-terminal 
nostrils. This is in contrast to the specialized tapir- 
like narial region of Heptodon and Helaletes (fig. 694, 
B, G), in which the anterior nares are relatively more 
widely open or receding, v/ith premaxillaries not reach- 
ing the nasals, distinctly lateral nostrils, and a more or 
less prehensile upper lip; obviously a specialized 
condition. 

The skull suddenly expands (fig. 695, A, B, C, D, E) 
opposite the orbits, which a,re usually large and widely 
open, indicating alert visual powers. The senses were 
probably keen in these animals, although no auditory 
bullae or osseous tympanic tubes are observed until 
the beginning of the upper Eocene {Eomoropus, 
Triplopus). 

It does not appear that the narial chamber was 
large, that the narial respiratory duct was very 
capacious, or that the olfactory chamber was large. 

The sagittal crest was prominent, arched, forming 
the highest point of the cranium, terminating in the 
high, narrow occiput, which indicates that the tem- 
poralis muscle constituted the chief musculature of the 
jaw. The masseters, attached to the rather slender 
zygomatic arches, were relatively feeble. The supe- 
rior borders of the zygomatic arches do not so continu- 
ously nor so distinctly connect with the lateral portions 
of the occipital crest, as indicated in LopModon by 
Deperet. 

The construction of the ear region is very primitive, 
the external auditory meatus being widely open; the 
osseous portion of the tympanic, if present, was loosely 
attached and seldom preserved. The mastoid is 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



765 




A, PlieTiacodus -wortmani 
(Phenacodontidae), lower 
Eocene (Wasatch forma- 
tion). 

B, Hepiodon calcicitjus (He- 
laletidae), lower Eocene 
(Wasatch rormation). 

C, Eohippus sp. (Equidae), 
lower Eocene (Wasatch 
formation) . 

D, Systemodon sp. (Tapi- 
ridae), lower Eocene 
(Wasatch formation). 

E, Lamhdotlic.rium popoagi- 
cum (Brontotheriidae), 
lower Eocene (Wind 
River formation). 

F, Eotitanops horeoJis (Bron- 
totheriidae), lower Eocene 
(Wind Eiver formation). 

G, Helaleles sp. (Helaleti- 
dae), middle Eocene 
(Bridger formation). 

H, IsectolopMs sp. (Tapi- 
ridae). middle Eocene 
(Bridger formation). 

I, Hyrachyvs sp. (Hyrachyi- 
nae), middle Eocene 
(Bridger formation). 

J, Eomoropits amarOTUTn 
(Chalicotheriidae), upper 
Eocene of Uinta Basin, 
Utah. 

K, Lophiodon leptorhynchus 
(after Deperet) (Lophio- 
dontidae), upper Eocene 
of Uinta Basin, Utah. 

L, Triplopus cuUtaUs (Tri- 
plopodinae), upper Eocene 
of Uinta Basin, Utah. 

M, Pnloplotherium (Paloplo- 
theriinae), upper Eocene 
of Uinta Basin, Utah. 



Figure 694. — Skulls of Eocene titanotheres and other perissodact3-ls and one condj'larth {Phenacodus) 
Originals or reconstructions in the American Museum of Natural History. Side view. .All figures two-fifths natural size. 



766 



TITANOTHERES OF .VNCIENT WYOMING, DAKOTA, AND NEBRASKLl 




C V^^x^^* D 

Figure 695. — Skulls of Eocene oondylarth (Phenacodus) and perissodactyls 
Reconstructed from materials in the American Museum of Natural History. Top views. All figures two-fifths natural size. 

A, Phenacodus (Phenacodontidae), lower Eocene. E, Isedolophvs (Tapiridae), middle Eocene. 

B, EoMppus (F.quidae), lower Eocene (first stage). 



Systemoion (Tapiridae), lower Eocene (first stage). 
D, Eotilanops (Brontotheriidae), lower Eocene. 



F, Hyrachyus (Hyrachyinae), middle Eocene. 

G, Eomoropus (Chalieotheriidae), upper Eocene. 
H, Triplopus (Hyracodontidae), upper Eocene. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



767 



exposed. It is only in upper Eocene perissodactyl 
skulls that we have found an osseous and tubular 
tympanic bone, as noted above. 

The brain case of the cranium as seen from above 
(fig. 695, A, C, D) does not appear to be much larger 
relatively in the primitive horses and tapirs (B, C) 



than in the condylarths (A); the sharp constriction of 
the brain anteriorly seems to be as marked in Systemo- 
don as in PTienacodus. The actual brain structure can 
not be known until a natural cast can be obtained. 

A summary of the skull characters as observed in 
the various types is as follows: 



Thirty primitive cranial characters possessed in common ly two species oj Eocene Condylarthra and ly fourteen 

species oJ Perissodactyla 





3 

1 

& 

1 


1 
1 

PL, 





1 

p. 
3 


a 

3 

s 

I 

-a 

w 


1 

1 

s 

1 


i 

1 


3 
1 

> 

a 


I 

1 


3 

1 


t3 


s 

>* 

H 


.a 
-a 

s 


.2 

1 

g 
1 

B 




B 

3 


Proportions: 

1. Dolichocephalic to mesaticephalic__ _ 


X 


X 
39? 

X 


X 

47 
X 


X 
41 
X 


X 


X 


X 
52 
X 


X 


X 


X 

46? 

X 


— - 


X 
46 
X 


X 
41 










50? 






3. Face longer than cranium_ __ _ _ _ 


X 


— - 


X 


X 


-— 








X 
49 












.^1 


52 
X 

X 
X 


60 

X 

X 
X 


X 
X 
X 

X 


65? 
X 

X 
X 


59 






56 


— - 


62 


53 


62 






X 

X 
X 


X 

X 
X 








Facial characters: 


X 
X 


i 


— - 


X 
X 


X 
X 


X 
X 


X 
X 




















X 
X 


V 




























X 




X 

X 
X 
X 
X 
X 

X 


X 

X 
X 
X 
X 
X 

X 


X 

X 
X 
X 
X 
X 

X 
X 

X 
X 
X 


X 

X 
X 
X 
X 
X 

X 
X 

X 

X 


X 

X 
X 
X 
X 


X 

X 
X 
X 
X 
X 

X 
X 

X 

X 


X 

X 
X 
X 
X 
X 

X 
X 

X 
X 


X 


X 
X 


X 


X 






12. Premaxillo-maxillary symphysis com- 








13. Opposite incisor series in convergent rows. 




X 




— - 


— - 




X 
X 
X 

X 
X 

X 
X 


X 
X 

X 
X 

X 
X 


X 
X 
X 






X 


X 






16. Lacrimals small, exposed, with tubercles^ 
Cranial characters: 








— 


X 


X 

X 
X 


X 
















19. Zygomatic arches short, slightly expand- 


X 
X 
X 


X 
X 
X 


-— 


X 
X 












X 








21. Brain case small, contracting anteriorly __ 
















X 


X 


X 


-— 


X 




















































X 










X 






Palatal characters: 
































X 


X 


X 


X 






















































X 
X 


X 
X 

26 


X 
X 

30 


X 
X 

30 





X 
X 

27? 


X 


V 


X 
X 


X 
X 

39? 
















X X 


-— 


X 
32 


X 
30 








Dental characters: 


26 




30 


35 













From the above general descriptions and table 
it is clear that the prevailing or primary type of 
perissodactyl skull is that approached by the 
horse (Eohippus), the tapirs {Systemodon, Isectolo- 
phus), the titanotheres {Lambdotherium, Eotitanops). 
In this primary type we observe that the face. 



including the orbits, is always longer than the cra- 
nium. We may regard this as a primitive char- 
acter. In the middle Eocene Lophiodon and all 
the middle Eocene titanotheres the cranium is longer 
than the face, a character which thus appears to 
be secondary. 



768 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



The constricted antorbital region of tlie face and 
terminal anterior nares is also distinctive, being ob- 
served in Plienacodus, Systemodon, LamMotlierium, 
Eohippus, LopModon, Isectolophus. This is a primitive 
character. The proboscis-bearing, receding nares of the 
specialized lophiodonts, Heptodon and Helaletes, are 
secondary analogies with the tapirs, closely similar to 
the condition observed in Profapirus of the lower 
Oligocene. 

This primitive perissodactyl skull is closely analo- 
gous or is actually related to the skull of the 
Condylarthra. In his detailed enumeration of the 
cranial characters of the primitive Perissodactyla 
Gregory points out (1910.1, p. 390) that this skull 
is of the insectivore-creodont-condylarth plan. 




FiGUHE 696. — Skulls of Lophiodon leptorhynchus 
A, Side view; 23, top view (of a different individujil). After Deperet. One-fomth natural s 

SECTION 2. ORIGIN AND PHYLETIC RADIATIONS 
OF THE TITANOTHERES AND OTHER PERISSO- 
DACTYLA 

The comparison of perissodactyls here made con- 
tinues that begun in Chapter I with a preliminary 
description of the relations of the titanotheres to the 
chalicotheres, paleotheres, and horses in the bunosele- 
nodont structure of their grinding teeth, as composing 
a great bunoselenodont branch in contrast to the great 
lophodont branch, which includes the tapirs, lophio- 
donts, and rhinoceroses, as presented in the accom- 
panying diagrams (figs. 697, 698), showing the general 
and detailed phylogeny of the Perissodactyla. 

SKULL OF THE PEIMITIVE TITANOTHERE 

Comparison of the skull of the primitive titano- 
theres {Larnbdotherium, Eotitanops) with that of the 
Condylarthra {Phenacodus) and, on the other hand, 
with that of all the primitive Eocene perissodactyls 



in which the skull is known, as shown above, reveals 
a striking general resemblance to the skulls of Eocene 
horses, paleotheres, tapirs, lophiodonts, the primitive 
rhinoceroses, and the chaHcotheres, which is attribut- 
able to two causes, namely, similar ancestry and 
similar habits. 

First, none of these skulls has diverged very far 
from the common ancestral forms; the orbit is 
situated midway of the head, and the face and the 
cranium are equal in length; the grinding teeth are 
short-crowned, brachyodont, and no special provision 
of the face is necessary for the accommodation of 
elongated teeth; the incisors and canines are relatively 
uniform in size, presenting no very striking enlarge- 
ment or reduction. Second, the similarity of the skulls 
is due to similarity of function and adaptation. 
These skulls are all adapted to the browsing 
habit; they are the type belonging to animals 
of cursorial, subcursorial, and mediportal gait, 
the muscles of mastication occupying similar 
areas on the zygomatic arches and on either side 
of the brain case, as exhibited both in the lateral 
and superior views of the skulls. 

In brief, the similarity in the skulls points 
both to similar ancestry and to analogy in 
habit. Nevertheless, we detect in the lower 
Eocene titanotheres evidences of incipient diver- 
gence, especially in the adaptation through 
greater or less I'ecession of the nasal bones for 
the prehensile function of the upper lip. While 
the nasal bones are full and elongate in the 
titanotheres (Eotitanops) and in the larger 
of the lophiodonts {Lophiodon) they recede both 
in the smaller cursorial lophiodonts (Helaletes) 
and in the cursorial paleotheres (Anoploihe- 
rium). The purpose of these adaptations for 
the prehensile function of the upper lip, so 
characteristic of the modern tapirs, is to subserve 
the browsing function, as described in the com- 
parison of the skull and mouth parts of the rhinoc- 
eroses in Chapter I (p. 32). This is the open-nostril 
type. The closed-nostril type is well illustrated in the 
skull of LopModon leptorhynchus, figured above, an 
appropriately named species, in which the nostrils 
are terminal and there is no room for the retractor 
muscles of the prehensile upper lip. In general the 
titanotheres conform to the latter type, with terminal 
nares and long nasals, but certain upper Eocene 
titanotheres (RhadinorJiinus) evolve the open nasal 
structure. In the two lower Eocene forms known, the 
cursorial Larnbdotherium and subcursorial Eotitanops, 
the skull is of the prevailing primitive perissodactyl 
type, similar in proportions to that of the horse 
Eohippus, yet it forms a starting point of evolution 
which finally results in a type fundamentally different 
from that of the horse in all its proportions. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



769 



STRUCTURE OF THE FOOT IN THE TITANOTHERES AND 
IN OTHER PERISSODACTYLS 

Fundamental characters. — The structure of the fore 
foot of the titanotheres is highly distinctive. It differs 
from that of the fore foot of all other Perissodactyla 



existingfamUies of perissodactyls, the tapirs, rhinocer- 
oses, and horses, the main weight passes directly through 
the center of D. Ill; they are mesaxonic. As shown in 
Figure 699 the mesaxonic extreme is the foot of the 
horse (Eqtms) , while the paraxonic extreme is exempli- 




FiGUKE 697. — Family tree of the titanotheres 

Showing the relation between the branches (phyla) , subfamilies, and genera known to science in 1919. The shaded areas show connections 

that are well established; the dotted lines show gaps that remain to be filled by future discovery, especially in the Uinta formation, Utah. 



except Amynodon in its paraxonic character, owing to 
secondary tetradactylism and to the enlargement of the 
functional fifth digit (D. V). Thus the main axis of 
weight, indicated by arrows, passes partly between D. 
Ill and D. IV, as in the Artiodactyla, whereas in all the 



fied in the deer (Cermis). In the tridactyl pes of the 
titanotheres the mesaxonic condition is exactly like 
that of the tapirs and rhinoceroses, presenting a uniform 
contrast to the paraxonic condition of the pes in all 
the Artiodactyla {Oreodon, Sus, Dorcatherium, Cervus). 



770 



TIT.\NOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 




3N3003 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



771 




101959— 29— VOL 2 6 



772 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 











ABC D 

Figure 700. — Carpus of the Eocene perissodactyls (horses, tapirs, lophiodonts, paleotheres, and rhinoceroses) 



Primitive tetradactyl; magnum small, covered by large "centrale" process of 
scaphoid; little or no contact with lunar in front; contact normal in back: 

A, Hcptodon calciculus, Am. Mus. 294 (Lophiodontidae), lower Eocene. 

B, Systemodon primaevum. Am, Mus. 15807 (Tapiridae). lower Eocene. 

C, Lambdotherium popoagkum. Am. Mus. 4880 (Brontotheriidae). lower Eocene 

D, Eotitanops princeps, Am. Mus. 296 (Brontotheriidae), lower Eocene. 

E, JSomoropus amaroTum, Am. Mus. 5096 (Chalicotheriidae), middle Eocene. 
Functional tetradactyl; magnum increasing, "centrale" process of scaphoid becom- 
ing reduced, lunar-magnum contact increasing: 

F, Isectolophus sp.. Am. Mus. 12219 (Tapiridae), middle Eocene. 

G, Isectolophus annectens. Princeton Mus. (Tapiridae), middle Eocene. 



H, Hyrachyus, Am. Mus. 12664 (Hyracodontidae), middle Eocene. 

I, Hyrachyus, Am. Mus. 1602 (Hyracodontidae). middle Eocene. 
Isotridactyl; magnum vertical, narrow, lunar-magnum contact diminishing: 

J, Triplopus cubitalis, Am. Mus. 5095 (Triplopodinae), middle Eocene. 

K, Colodon, Am. Mus. 658 (Helaletidae), middle Eocene. 

L, Lophiodon, after Dep6ret, Lyons Mus. (Lophiodontidae), middle Eocene. 

M, Palaeotherium, after De Blainville (Palaeotheriidae), middle Eocene. 
Monodactyl; magnum broadening, lunar-magnum contact increasing: 

N, HyTacotheriitm (Equidae), lower Eocene. 

O, Epihippus (Equidae), middle Eocene. 

P, Mesohipptis (Equidae), lower Oligocene. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



773 



The structure of the titanothere pes is typically 
perissodactyl, whereas the structure of the titanothere 
manus is comparable rather to that of the artiodactyl 
(even-toed) type. 

The retention of D. V in the titanotheres is probably 
due to the early evolution of a graviportal stage, out 
of more primitive mediportal and subcursorial stages. 

Classification hy foot structure. — In devising the 
nomenclature of the hind foot to signify the uneven 
number and uneven arrangement of the digits the 
perissodactyl branch of the Ungulata, including the 
titanotheres, has been variously termed the Peris- 
sodactyla by Owen (1848.1), the Imparidigitata by 
Riitimeyer (1865.1) and by Burmeister (1867.1), 



WIND RIVER 



gested by Marsh, because while in the hind foot 
the main weight passes through the middle toe and is 
therefore mesaxial, in the fore foot of many but not 
all of the titanotheres the main weight is distributed 
between the third and fourth toes and is therefore 
paraxial, although not so distinctly so as in the hippo- 
potamus. 

Among the amynodonts (amphibious rhinoceroses) 
we also observe the four-toed structure of the fore foot 
as in certain titanotheres. To reach the real meaning 
of this four-toed structure of the fore feet we compare 
the titanotheres with the most primitive laiown 
direct ancestors of the horses, tapirs, and rhinoceroses. 
We find that these also are four toed and more or less 



UPPER UINTA 



CHADRON 




Figure 701. — Evolution of the astragalus in the titanotheres 
Upper row, back view; lower row, front view. 

A, Lambdotherium sp., Am. Mus. 14921, lower Eocene, Wind River formation. 

B, EotUanops borealis, Am. Mus. 14888, lower Eocene, Wind River formation. 

0, Limnohyops monocomts, Am. Mus. 11089, middle Eocene, lower part of Bridger formation. 

D, Manteoceras manteoceras, Am. Mus. 1587, middle Eocene, upper part of Bridger formation. 

E, ProtUanothcrJum superbum, Am. Mus. 2030, upper Eocene, Uinta C, true Uinta formation. 

T, Menodus gigajiieiis, ,\m. Mus. SO.'i, lower Oligocenc, upper Titanotherium zone (Chndron C) . 

This series shows the progressive widening of the astragalus in the transition from cursorial to graviportal habits. The earlier members have the troch- 
lear keels high and angulate, the neck well defined, the cuboid facet verj* small, the sustentacular facet long and narrow. 'The latest members have 
the trochlear keels low and rounded and the head sessile, the constriction, or neck, being nearly obliterated. The cuboid at this stage is very wide, 
the sustentacular facet short and wide. 



after the principles of classification of the French 
paleontologist De Blainville (1816.1), and the Mesa- 
xonia by Marsh (1884.1). 

In seeking the relations of the hoofed animals 
(Ungulata) Richard Owen, who followed the distinc- 
tions in foot structure first pointed out by De Blain- 
vdle between the "pachydermes a doigts pairs" 
(number of digits even) and the "pachydermes a doigts 
impairs" (number of digits odd), might have found the 
isolated fore foot of an Oligocene titanothere a 
difficult subject. De Blainville's distinction between 
an odd or even number of digits at first appears to 
fail, as the fore foot of the titanotheres is four-toed, 
or paired, broadly resembling that of the hippo- 
potamus. It is no less difficult to apply to the fore 
feet the distinctions mesaxonic and paraxonic sug- 



paraxonic in the fore feet, which proves that the fore 
feet of the titanotheres are in one sense a persistent 
primitive element, because, they retain the fully func- 
tional outside toe of the primitive five-toed foot. 

The primitive titanothere pes is very similar to 
that of the other primitive Eocene perissodactyls. 
All these animals are in a cursorial or subcursorial 
stage of evolution. The progressive titanothere pes 
shown in Figure 701 is in the graviportal stage, in 
which the astragalus is flattened and rests broadly 
on the cuboid. It thus resembles the astragalus of 
the graviportal rhinoceros rather than that of the 
mediportal tapir, in which the astragalus has a 
narrow footing on the cuboid, and bears no resem- 
blance at all to that of the horse, in which the as- 
tragalus has entirely lost its anterior footing on the 



774 



TIT.\NOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



cuboid. The cursorial, mediportal, and graviportal 
proportions and articulations of the astragalus consti- 
tute an invariable feature of the adaptive radiation 
of the perissodactyls into these respective types of 
locomotion. They are correlated with the cursorial 
types of limb, pelvis, and shoulder girdle structure, 
which have been fully described in Chapter IX in the 
summary of cm'sorial and graviportal proportions of 
limb segments in imgulates (p. 739). 

In his studies on the evolution of the ungulate 
foot Osborn (1890.51) pointed out that the shape of 
the facets between the astragalus, calcaneum, and 
cuboid are distinctive in each of the perissodactyl 
families. The original distinctions, however, tend 
to be obscured by the convergent influence of the 
respective adoption of the cursorial, the mediportal, 
or the graviportal gait. The primitive arrangement 
of these facets in the Perissodactyla is possibly 
derived from the arrangement seen in the Condylarthra 
(Phenacodus), which, in turn, is an inheritance from 
clawed ancestors in which (a) the digits spread apart, 
(6) the astragalus and calcaneum diverge distally, 
and (c) there is no contact of the astragalus upon the 
cuboid and no astragalocuboid facet (W. K. Gregory, 
1905, MS.). In the primitive Perissodactyla the 
progressive development of cursorial habits caused 
the long axes of the calcaneum and astragalus to 
become more parallel, so that these bones articulated 
distally as well as proximally. 

The subsequent evolution of these astragalocal- 
caneal facets depends upon the following conditions: 
(1) Reduction of lateral digits in the monodactyl 
Eqims makes for immobility between the astragalus 
and calcaneum; (2) the tibio-astragalar joint is more 
mobile than in the succeeding mediportal and gravi- 
portal forms; (3) the extent of the astragalocalcaneal 
facets depends partly upon the shape of the tibia and 
fibula above and of the supporting elements, the 
cuboid and navicular, below; (4) the size and direc- 
tion of pull of the ligaments of the limb and foot; (5) 
the shape and arrangement of the astragalocalcaneal 
facets and adjacent parts condition the degree of 
rotary, vertical, and horizontal motion between the 
astragalus and calcaneum; (6) two general types are 
observed — (a) the astragalus with oblique distal ex- 
tremity, correlated with movabdity, (6) the astragalus 
with truncate distal end, correlated with less mova- 
bility. 

In general the perissodactyl astragalus is much 
more movable upon the calcaneum than the artio- 
dactyl astragalus. The horse alone has developed a 
comparatively rigid astragalocalcaneal union, by a 
system of interlocking joints and by the arrangement 
of the facets in three planes, more or less at right 
angles. This rigid astragalus is adapted to the elon- 
gate cannon bone (as in Artiodactyla), also to the 
reduction of the lateral digits. 



MECHANICS OF THE PERISSODACTYL MANUS 

Steps in evolution of the manus. — In the stem 
Perissodactyla we observe that (1) elevation from 
the plantigrade to the digitigrade position of the 
manus has already taken place in the ancestral 
stages, and there remains the transformation into the 
extreme unguligrade position shown in the Equidae; 

(2) atrophy or transposition of the muscles of rotation, 
of pronation, and of supination has also taken place, 
the muscles involved in fore-and-aft movement again 
reaching their highest development in the Equidae; 

(3) extreme reduction of the pollex has already taken 
place, for not a vestige of Mtc I has yet been dis- 
covered, although there is some evidence that it may 
have existed; (4) reduction of the trapezium follows, 
although this element is still retained in all primitive 
perissodactyls; (5) coalescence of the centrale with 
the scaphoid has already taken place, followed by a 
reduction of the centrale process of the scaphoid; (6) 
no "displacement" or slipping of the first and second 
row of carpals takes place, as Cope believed; (7) 
there is either growth and enlargement or reduction 
of several elements of the carpus, correlated with 
the expansion of the radius and the reduction of the 
ulna above and with the distribution of the weight 
through a diminishing number of the metacarpals 
below; (8) at each stage of reduction the functional 
tridactyl, isotridactyl, and monodactyl condition has 
its distinctive type of correlated carpal structure; (9) 
displacement of Mtc II against the magnum, of Mtc 
III against the unciform, the "alternating type" of 
displacement of Osborn, is a universal feature of the 
perissodactyl carpus. 

The mechanics of reduction and displacement, as 
worked out by Ryder (1877.1), Cope (1887.1), and 
Osborn (1890.51), require complete restudy in view 
of the fact that the primitive perissodactyl carpus is 
of the displaced type. 

The law enunciated by Osborn in 1890 (op. cit., p. 
568) is, however, probably in the main correct, 
namely : 

The direction and degree of intercarpal displace- 
ments are adapted to the gradual alteration of the 
major axes in the bones of the forearm and of the 
metapodimn, respectively, as brought about by en- 
largement and reduction, and tend to maintain these 
proximal and distal axes in the same vertical line. 

Carpus of cursorial, mediportal, graviportal types. — 
Like the bones of the tarsus, every bone of the peris- 
sodactyl carpus has its peculiar form, adapted to 
graviportal, mediportal, subcursorial, or cursorial 
locomotion. It is notable that the ancestors of each 
of the lower Eocene families have the cursorial type 
of carpus, in which the lunar rests mainly upon the 
unciform, with small, lateral contact on the magnum. 
Thus in Eotitanops we observe an arrangement of the 
carpals somewhat similar to that in the subcursorial 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



775 



Hyrachyus; yet from this type has been derived the 
heavy graviportal manus of Brontops, ia which the 




FiGUKE 702. — Back view of the carpus 
of a middle Eocene titanottiere 

Mesatirhinus petersoni, Princeton Mus. 10013. 
Showing the lunar as the keystone of the proximal 
row of carpals and of the magnmn and tinciform, 
on which it rests. In front view it is largely 
displaced on the unciform. Each bone of the 
second row of carpals (for example, the magnum) 
forms the keystone between one of the proximal 
row and one or more of the metacarpals. 

lunar rests broadly on the magnum. Similarly, the 
carpus of Lamhdotherium is of the cursorial type, all 



little is known of the primitive tapir (Sysfemodon) 
shows a similar subcursorial structure, while the 
primitive lophiodont {Heptodon) shows a markedly 
cursorial carpus, in which the lunar has no contact 
with the magnum in front. The adaptive mechanical 
principles involved are explained by Osborn in his 
"Evolution of the ungulate foot" (1890.51). In 
Figures 722, 723 it is shown that as the animal passes 
from the cursorial into the mediportal condition 
{Tapirus, Limnohyops , Palaeosyops) the lunar broadens 
and gains a facet on the broadening magnum. In 
extreme graviportal forms (Brontotherium, Rhinoceros) 
the carpus is very broad, the magnum is flattened and 
supports half the weight of the lunar. 

Adaptation in the magnum carpi. — Each carpal, 
like each tarsal, mirrors the primitive cursorial, medi- 
portal, or graviportal locomotor stage of the limb. 
As shown below, the magnum alone mirrors the entire 
locomotor evolution of the titanotheres from Eotiianops 
to Brontotherium. The cursorial adaption at the 
back of the magnum is the deep hook forming the 
long arm of a lever in flexing the carpus, well developed 
in cursorial forms, also in the mediportal tapir, in 
Mesatirhinus, and in the long-footed rhinoceroses. 
This hook is progressively reduced as an adaption to 
weight in the graviportal brontotheres. In the back 




Aoo/l- 



FiGTJEE 703. — Left magna of an Eocene titanotliere and two chalicotheres 
A, Palaeosyops hidyi, a titanothere; B, Moropus sp., and C, Macrotherium sp., chalicotheres. Top row, front view; middle row, outer 
side; lower row, inner side. In the chalicotheres the front face of the magnum becomes extremely high and compressed. The 
magnum becomes very deep anteroposteriorly, and the posterior process becomes confluent with the dorsal facet. In spite of these 
differences the ordinal kinship of titanotheres and chalicotheres is only partly disguised in the magnum. 



the elements high and narrow, the lunar resting by 
very small contact on the small magnum. What 



view of the carpus of the middle Eocene titanothere 
Mesatirhinus petersoni the magnum with^its long 



776 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



hook is seen to support half the lunar, while in the 
front view of the same carpus the lunar rests mainly 
on the unciform. In further special studies of the 
magnum Gregory has compared the magna of Palaeo- 
syops and of two chalicotheres {Moropus and Macro- 
therhim) which display family differences and ordinal 
resemblances, as shown in Figure 703. 

Reduction of digit V in the manus. — To the primi- 
tive cursorial habit of the perissodactyl ancestors is 
probably attributable the early reduction of D. I, 
no vestige of which has thus far been discovered in 
any perissodactyl, living or fossil, as pointed out 
above. Digit V suffers incipient reduction as the 
manus is raised into the unguligrade position, in which 
the main weight rests upon the horny sheaths of the 
terminal phalanges, supported by the posterior foot 
pad, as shown in Chapter I (p. 33). This reduction 
does not advance very far in the titanotheres because 
of the early transformation of the foot into the medi- 
portal, weight-bearing type, in which D. V becomes 
useful and therefore is retained. Nevertheless, some 



(cf) (H 



Pcicr 




Figure 704. — Left magna of two lower 
Eocene perissodactyls 

A, Heptodon cakkulus (Cope), Am. Mus. 294; B, 
Eotitanops princcps (Osborn), Am. Mus. 2% (type) . 
The small end inQQS anteriorly; the bones are seen 
from the outer side. Three-halves natural size. 
lei. c. r.. Tuberosity for ?extensor carpi radialis; ?/?. 
c. T., tuberosity for ?(]exor carpi radialis; (ce.), facet 
for hook of the scaphocentrale; (In.), facet for lunar; 
(unc), facet for unciform; (III), facet for Mtc III. 

of the mediportal titanotheres (such as Mesatirhinus) 
exhibit a more slender D. V and a tendency toward 
tridactylism, a secondary mesaxonic condition of the 
manus, which in course of further evolution would 
end in a tridactyl manus. All other perissodactyls 
except the graviportal amynodonts {Amynodon, aqua- 
tic rhinoceros) rapidly reduce D. V and pass into the 
isotridactyl condition, with the more or less rapid 
reduction of D. V conditioned hj the cursorial or the 
mediportal habit. Thus D. V persists in the medi- 
portal AceratJierium tetradactylum of the Miocene of 
Europe, also in the modern mediportal Tapirus. 
although it disappears in the mediportal rhinoceroses, 
all of which become typically isotridactyl. Several 
cursorial branches of the perissodactyl families — the 
Paloplotheriinae, the Triplopodinae, the lophiodont 
Helaletidae, the cursorial rhinoceroses (Hyracodon- 
tinae), the subcursorial horses of the forest-living 
Hypohippus branch — all retained three digits closely 



compressed into a narrow, solid foot, which reaches an 
extreme in Oolodon, the terminal member of the 
lophiodont Helaletidae. 

Reduction of terminal phalanges. — Whereas in all 
cursorial types the terminal phalanges support rela- 
tively large horny sheaths adapted to rapid progres- 
sion over hard ground, the mediportal and gravi- 
portal types tend to the enlargement of the foot pad 
and reduction of the terminal phalanges. Conse- 
quently in the large graviportal titanotheres, as in the 
Proboscidea, the terminal phalanges are greatly 
reduced in size, becoming almost vestigial even in 
such relatively swift-moving forms as Menodus, which 
has much more elongate digits and limbs than its 
bulky contemporary BrontotheTiurn. 

SUMMARY OF THE EVOLUTION OF THE PERISSODACTYL 
FAMILIES 

The closest rivals of the titanotheres were the other 
perissodactyls, but the titanotheres outstripped them 
all until the end of lower Oligocene time. It has been 
noted in Chapter I (p. 24), as shown in the accom- 
panying diagram (fig. 705), that the nine typical peris- 
sodactjd families had already diverged from one 
another in lower Eocene time, and that by the begin- 
ning of Oligocene time they were widely separated in 
their dental and osteological structure. In section 1 
of the present chapter (p. 768) it has been shown that 
when the titanotheres, horses, tapirs, and lophiodonts 
first appeared in America they were so similar in struc- 
ture of skull and feet that they can be separated only 
by careful analysis of the structure of their grinding 
teeth. It is not surprising that paleontologists and 
zoologists of the last century (Flower, Cope, Gill) were 
disinclined to separate them into distinct families. 

The separation of the Perissodactyla according to 
their fundamental divergence in the structure of the 
molar teeth was partly suggested by Schlosser (1886.1) 
in his "Beitriige zur Kenntnis der Stammesgeschichte 
der Hufthiere und Versuch einer Systematik der 
Paar- und Unpaarhufer." In 1892 it was more fully 
developed by Osborn (1892.67, pp. 90-94) in his thesis 
"The classification of the Perissodactyla," in which it 
was shown that the main desiderata of classification 
are, first, clearness of phyletic relationships; second, 
convenience; third, structure, since it appears that in 
this order the teeth are more fundamental than the 
feet. 

At this time Osborn (op. cit.) divided the Perisso- 
dactyla into 9 families and 19 subfamilies. In "The 
rise of the MammaUa in North America" (1893.82) 
Osborn extended this system, and in "The extinct 
rhinoceroses" (1898.143, p. 79) he showed that the 
Perissodactyla may be primarily divided by the fun- 
damental pattern of the upper grinding teeth into four 
superfamilies, namely, Titanotheroidea, Hippoidea, 
Tapiroidea, Rhinocerotoidea, to which, he observed. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



777 




Figure 705. — Phylogenetic tree of the Perissodactyla, showing the superfamily, family, and subfamily branches 




' Adaptive 

radiaiiorL of the 

subfamUies of 

Peri^sodcLcfyls 

IIcd)Lts an<L fhcd)itais 



MEDIPORTAL ({"'Km GRAVI portal ©"B-^ 



Figure 706. — Family tree of the Perissodactyla, showing the convergence of different branches through 
adaptive radiation in similar habitats and adaption to similar habits 



778 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



should probably be added the Chalicotheroidea as an 
aberrant supei'faniily with molar teeth related to the 
titanothere pattern and perissodactyl feet provided 
secondarilj' with claws. This arrangement is sub- 
stantially that which has been adopted in this mono- 
graph — that is, they have been divided into 5 super- 
families (-oidea), 9 families (-idae), 35 subfamilies 
(-inae). A subfamily may embrace one or more gen- 
era; it pursues an independent line of development 
and either terminates in an extreme or gives off a gen- 
eralized branch from which another subfamily may 
arise. 

Subsequent discovery (Osborn, 1913.398) has 
brought the chalicotheres more intimately within the 
order Perissodactyla, so that they are with some 
certainty now placed as a branch mtermediate between 
the Titanotheroidea and the Hippoidea. Thus accord- 
ing to our present knowledge (1918) the tabular 
classification of the Perissodactyla is as follows: 

Classification of the order Perissodactyla 

A. Bunoselenodont suborder or branch: 

I. Superfamily Titanotheroidea (titanotheres) : 

1. Family Brontotheriidae : 

Subfamily (phylum) : 

1. Lambdotheriinae. 

2. Eotitanopinae. 

3. Palaeosj'opinae. 

4. Telmatheriinae. 

5. Manteoceratinae. 

6. Dolichorhininae. 

7. Rhadinorhininae. 

8. Diplaoodontinae. 

9. Brontopinae. 

10. Menodontinae. 

11. Megaceropinae. 

12. Brontotheriinae. 

II. Superfamily Chalicotheroidea (chalicotheres) : 

2. Family ChaUcotheriidae : 

Subfamily (phylum) : 

13. Moropinae (America). 

14. Chalicotheriinae (Europe). 
III. Superfamil}^ Hippoidea (paleotheres, horses) : 

3. Family Palaeotheriidae : 

Subfamily (phylum) : 

15. Paloplotheriinae. 

16. Palaeotheriinae. 

4. Family Equidae: 

Subfamily (phylum) : 

17. Hyracotheriinae. 

18. Anchitheriinae. 

19. Protohippinae. 

20. Equinae. 

B. Lophodont suborder or branch: 

IV. Superfamily Tapiroidea (tapirs, helaletids, lophio- 
donts) : 

5. Family Tapiridae: 

Subfamily (phylum) : 

21. Systemodontinae. 

22. Tapirinae. 

6. Family Lophiodontidae: 

Subfamily (phylum) : 

23. Helaletinae. 

24. Lophiodontinae. 



B. Lophodont suborder or branch — Continued. 

V. Superfamily Rhinocerotoidea (amynodonts, hyraco- 
donts, rhinoceroses) : 

7. Familj' Amynodontidae: 

Subfamily (phylum) : 

25. Amynodontinae. 

8. Famiily Hyracodontidae: 

Subfamily (phylum) : 

26. Hyrachyinae. 

27. Triplopodinae. 

28. Hyracodontinae. 

9. Family Rhinocerotidae: 

Subfamily (phjdum) : 

29. Aceratheriinae. 

30. Diceratheriinae. 

31. Teleoceratinae. 

32. Dicerorhinae. 

33. Rhinooerotinae. 

34. Dicerinae. 

35. Elasmotheriinae. 

CONVERGENCE IN HABITAT AND HABIT 

It is vezy important to observe that although the 
nine typical perissodactyl families as a whole diverge 
and radiate from one another, the branches into which 
some of them subdivide under the law of adaptive 
radiation converge, because these seek similar habitats 
and assume similar habits. 

Cursorial habit in the plains habitat. — Cursorial 
habits were independently assumed in one branch of 
the titanotheres, in all branches of the horses except 
one subfamily, in two branches of the lophiodonts, 
in three subfamily branches of the rhinoceroses. 

Forest-living habit. — The very reverse of the cur- 
sorial plains habit was independently assumed by 
the chalicotheres, by the tapirs, possibly by certain 
branches of the titanotheres (Dolichorhininae), also 
by certain branches of the rhinoceroses. 

Aquatic habit and river habitat. — The swamp-living 
and aquatic habit appears to have been assumed 
independently by one branch of the titanotheres 
(Metarhinus) , by the amphibious rhinoceroses (Amy- 
nodontidae), and possibly by certam of the true short- 
footed rhinoceroses (Teleoceratinae). 

Mediportal habit, browsing type. — Mediportal brows- 
ing forms are characteristic of all the middle stages of 
titanothere evolution, of the early stages of the 
lophiodonts and hyracodonts, and of all the early 
stages of the rhinoceroses. 

Graviportal habit, browsing and grazing types. — The 
graviportal habit is illustrated in certain stages of 
the evolution of the titanotheres, in the true lophio- 
donts, in most of the upper Miocene, Pliocene, and 
Recent subfamilies of the rhinoceroses. 

Secondary and independent adoption of similar 
habitats and habits in the different families of peris- 
sodactyls has given rise to the analogous, parallel, 
convergent, and homoplastic adaptations in skull, 
limb, and foot structure described in this chapter and 
in Chapter IX and has produced resemblances so close 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



779 



that it may be difficult to determine the family from 
a single bone, and it therefore becomes necessary to 
determine family affinity from the teeth. 

DESCENT OF THE BUNOSEIENODONT FAMILIES 

/, Family 1 (see table on p. 778). — When the titano- 
theres emerge in North America in Wind River time 
(lower Ypresian) they appear to have already separated 
into two divisions, namely, the cursorial Lambdotherii- 
nae and the subcursorial Eotitanopinae. The subfam- 
ily Eotitanopinae is believed to have been the source of 
the mediportal Limnohyops and the graviportal Palaeo- 
syops of the middle Eocene. From an unknown direct 
source arise the short-horned titanotheres (Man- 
feocems), which are related to the great graviportal 
short-horned titanotheres of the lower Oligocene 
(Brontops). Related to this stem are the forest-, 
swamp-, and river-dwelling Dolichorhininae. Not far 
from this source are the ancestors of the long-horned 
titanotheres {Brontotherium., Megacerops), which ter- 
minate in the graviportal stage and become extinct 
at the end of lower Oligocene time. 

//, Family 2. — The chalicotheres (Ancylopoda of 
Cope) in skull and tooth structure appear to occupy 
a position intermediate between the Titanotheroidea 
and the Hippoidea (paleotheres and horses). At the 
end of middle Eocene time (upper Bridger, Bartonian) 
there occur the Pernatlierium (of France) and the 
Eomoropus (of Wyoming), sharply distinguished in 
foot structure by the possession of powerful claws, 
which were probably used as weapons of defense and 
for grasping tree branches, since the fore limbs are 
not of the fossorial type. The skull and body are 
proportioned like the forest-living okapi, and the great 
rarity of remains of these animals in the plains and 
fluviatile deposits between middle Eocene and lower 
Pliocene time, comparable to the rarity of remains of 
forest-dwelling tapirs, points to their adoption of the 
forest habitat and forest-dwelling habit. 

///, Family 3. — The paleotheres were probably 
closely related in very remote times to the horses, 
although they first emerge geologically in Europe in 
Lutetian (middle Eocene) time, their earlier history 
being unknown. They divide into the browsing medi- 
portal Palaeotheriinae and the slenderly built Paloplo- 
theriinae, which imitate the true horses. 

Ill, Family 4- — The horses emerge as small, progres- 
sive cursorial types in Europe and America in Sparna- 
cian (middle Wasatch) time in the genera Hyraco- 
therium and Eohippus. There is some evidence that 
they subdivide in early Oligocene time into (a) a 
purely cursorial grazing type, (b) subcursorial brows- 
ing types, and (c) subcursorial forest-living types. 
In the types last named tridactylism is persistent in 
a phylum that leads into the forest-frequenting 
Hypohippus, which has crested, short-crowned teeth, 
functionally resembling those of the forest-living 



tapirs, and spreading, three-toed feet. The main line 
of horse evolution was into the extreme plains-living, 
grazing, cursorial types of hipparions, primitive and 
progressive horses, zebras, and asses, the latter retain- 
iug much of the browsing habit. 

DESCENT OF THE lOPHODONT FAMIIIES 

The lophodont families spring from two great 
branches, the Tapiroidea (IV) and the Rhinocerotoidea 
(V). The families of the Tapiroidea are less uniform 
in the structure of their grinding teeth (Osborn, 
1895.105, p. 359, fig. 6), but there is considerable 
evidence from the transitions in the skull, tooth, 
and foot structure that the Tapiridae, Helaletidae, 
and Lophiodontidae sprang from a common stem, 
the lophiodonts being closest in their grinding-tooth 
structure to the rhinoceroses. 

IV, Family 5. — The tapirs emerge in the subcur- 
sorial Systemodon of the lower Eocene of America. 
Systemodon is very abundant only in the Systemodon 
zone, for throughout their entire geologic history 
tapiroid remains are most rare, and we assume that 
these animals early adopted the forest-dwelling habit 
and hence escaped fossilization. The American line 
leads through Isectolophus of the middle and upper 
Eocene, while the Protapirus of the lower Oligocene 
also represents a side branch of the main family, not 
typically tapiroid in its molar-tooth structure and 
therefore classed by Peterson with the Pseudo- 
tapirinae. The tapirs throughout their history have 
preserved the mediportal size and the forest, swamp, 
and river frequenting habitat, although high moun- 
tain dwelling forms evolve in the Andes. 

IV, Family 6. — A true lophodont branch in its 
molar-tooth structure, known as the Helaletinae, 
emerges in the extremely cursorial Heptodon of the 
Heptodon zone of the Wasatch, as a contemporary of 
Systemodon and Eohippus. The skulls of Heptodon 
and Helaletes, we have seen, closely parallel those of 
the tapir in the recession of the nasal bones in adapta- 
tion to the development of a long prehensile upper lip ; 
but the grinding teeth are exactly like those of the 
typical lophiodonts of the upper Eocene of Europe. 
The American stages are Helaletes of Bridger B and 
DesmatotTierium of Washakie B. Unlike that of the 
tapirs the manus in the Oligocene stage (Colodon) be- 
comes excessively elongate, slender, and highly curso- 
rial, although persistently tridactyl. These cursorial 
lophiodonts were probably represented also in Europe. 

The true lophiodonts (Lophiodontidae) are medi- 
portal quadrupeds, emerging in the Sparnacian of 
Europe and breaking up into four independent 
phyletic lines (Deperet). These include the cursorial 
slender-limbed chasmotherines and the heavy-limbed 
true lophiodonts. The latter became extremely gravi- 
portal, suddenly terminating in heavy-bodied, pon- 
derous forms in the lower Oligocene of Europe. 



780 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



T', Family 7. — The amynodonts, a ven" distinct 
aquatic branch of the Rhinocerotoidea, are first 
Icnown at the base of the upper Eocene in America 
and Europe, their distinctive features being very 
large upper and lower canine tusks (which are lost 
in the true rhinoceroses), broad, flat skulls, very promi- 
nent orbits, and broad, spreading, tetradactyl fore feet. 
The pi'eservation of these animals in the river-channel 
•sandstones {Metamynodon zone) of the lower Oligocene, 
their protruding orbit structure, and their spreading 
fore feet suggest amphibious habits. They are more 
.aquatic than any other perissodactyls except possibly 
Metarhinus of the titanothere family. 

T^, Family S. — The hja-acodonts divide into three 
subfamilies, all of cursorial and subcursorial habits 
and of the plains and meadow habitat. The light- 
limbed Triplopodinae of the American upper Eocene 
are subcursorial. The Hyracodontinae of the lower 
and middle Oligocene are the most highly cursorial of 
all the known rhinoceroses; they are, in fact, small- 
headed, tridactyl, swift-running rhinoceroses. The 
Hyrachyinae are more conservative mediportal forms, 
resembling the tapir in proportions, characteristic of 
the middle and upper Eocene of America. They 
stand very close to the origin of the true rhinoceroses, 
if not directly ancestral to them. Apparently none of 
the hyracodonts found their way into Europe, though 
certain fossil remains of middle Europe have been 
referred to Hyracodon. 

V, Family 9. — The true rhinoceroses suddenly 
emerge in two subfamilies of the lower Oligocene, 
namely, the pair-horned diceratheres {Diceratherium) , 
of the Sannoisian and White River, of Europe and 
America respectively, and the hornless Aceratheriinae, 
animals having the same geographic range, which are 
distinguished by the retarded development of the 
median horn. These mediportal animals expand in 
number and variety immediately after the extinction 
of the titanotheres in lower Oligocene time and be- 
come the dominant quadrupeds of middle and upper 
Oligocene time both in Europe and America. The 
diceratheres became extinct in middle Oligocene time, 
but the aceratheres developed a middle horn on the 
top of the skull (Osborn, 1899.166), and it is possible 
that the AceratJierium incisivum of the lower Pliocene 
may have given rise to the gigantic single-horned 
ElasmotJierium of the Pleistocene; but this descent 
is highly conjectural. In lower Miocene time the 
short-footed, graviportal Teleoceratinae emerge in 
Europe and Asia and migrate to America. In body 
and limb structure they even present analogies to 
the hippopotami but exhibit no aquatic adaptations 
in the skull. 

The Dicerorhinae also appear in the lower Miocene 
of Europe and become the dominant European 
rhinoceroses in Pliocene and Pleistocene time, a 
.branch surviving to-day in the existing Sumatran 



rhinoceros, a forest-frequenting, browsing animal, 
relatively primitive in cranial and dental structure, 
mediportal in proportions. 

In the lower Pliocene of Eui'ope also appear the 
Dicerinae, related to the existing African white 
rhinoceros Ceratotherium simum, which is distinctly a 
grazing, hypsodont, extremely graviportal type. 

In the meantime the true Rhinocerotinae appear in 
Asia in the lower Miocene. 

In the Pleistocene of Eurasia occur the steppe-fre- 
quenting, grazing, hypsodont Elasmotheriinae, gigan- 
tic animals of graviportal type, which may represent 
a branch from the original Aceratheriinae stock 
(Osborn, 1900.192). 

STJEVIVAI AND EXTINCTION OF THE PERISSODACTYLA 

The theoretic causes of the extinction of so many 
branches of the Perissodactyla in early Oligocene time 
and the survival to the present time of only five of 
these branches — the Equinae, Tapiridae, Dicerorhinae 
(Sumatran), Dicerinae (African), and Rhinocerotinae 
(Indian) are discussed in Chapter XI, section 2. 

PHYLETIC BRANCHING OF THE TITANOTHERES 

As compared with other perissodactyls the adaptive 
radiation and phyletic evolution of the titanotheres is 
seen to be limited by their conservative grinding tooth 
structure and their closely correlated foot structure. 
The grinding tooth of the titanotheres is mechanically 
incapable of transformation into the grazing type; 
the feet do not evolve in a cursorial direction after 
the first essaying of cursorial structure in Latnhdo- 
therium. Consequently the evolution and specializa- 
tion of the titanotheres took place principally within 
the browsing habitats of meadows, the borders of for- 
ests, and the borders of streams and rivers, the last 
affording amphibious and possibly aquatic habitats. 
In respect both to hypsodonty of the teeth, which is 
invariably a grazing adaptation, and to elongation of 
the feet for the seasonal migrations connected with 
grazing habits, the titanotheres are greatly inferior in 
plasticity to the rhinoceroses, which independently de- 
velop an extreme cursorial type (Hyracodon), an extreme 
aquatic type {Amynodon), also two extreme grazing 
types {Elasmotherium and Ceratotherium simum). 

The phyletic branching of the titanotheres is sum- 
marized in Figure 697. 

SECTION 3. SUMMARY OF THE CRANIAL AND SKEL- 
ETAL EVOLUTION OF THE TITANOTHERES 

GENERAL CONCLUSIONS REACHED 

Epitome of the evolution. — An epitome of the entire 
evolution of the skull, teeth, and skeleton of the titano- 
theres brings out three principal conclusions: 

First, the evolution of the feet "and limbs, as already 
described in detail, from the mediportal to the gravi- 
portal condition, closely parallels similar stages of 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



781 




Figure 707. — Outlines of the body form of the perissodactyls drawn to the same scale 

The largest known member of each family is selected for comparison. The aiumals are grouped according to their natural relation- 
ships, indicated especially by the pattern of the molar teeth, as follows: 

Ehinocerotoid group: A, Melamynoion, family Amynodontidae. graviportal aquatic rhinoceros, lower Oligocene; B, Byracodon, 
family Hyracodontidae, cuisorial rhinoceros, lower Oligocene; C, Ceratolherium simum, family Rhinocerotidae, living white 
rhinoceros, graviportal. Tapiroid group: D, Tapirus terrestris, family Tapiridae, living tapir. 

Hippoid group: B, Palaeotherium, family Palaeotheriidae, mediportal, lower Eocene; F, Equus przewalshi, family Equidae, living 
horse, cursorial. 

Chalicotheroid group: G, Moropus, family Chalicotheriidae, clawed perissodactyl, lower Miocene. 

Titanotheroid group: if, Bwvtotherium plalyceras, family Brontotheriidae, graviportal, lower Oligocene. 



782 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



evolution in all the other perissodactyls. This is 
clearly brought out in Chapter IX, sections 1-2. 

Second, the evolution of the grinding teeth is de- 
cidedly inferior to that of all the other families of 
perissodactyls. The titanothere molar, though capa- 
ble of indefinite enlargement, is not capable of trans- 
formation into a hypsodont tooth. 

Third, the transformation of the skull, on the con- 
trary, is a far more extreme specialization in the titano- 
theres than in the tapirs or even in the horses; it is, 
on the whole, more extreme than in the rhinoceroses. 
Consequently, cranial evolution and transformation 
are among the most distinctive characteristics of the 
titanothere family. In many characters, as pointed 
out in Chapter I (pp. 28-32), the cranial transformation 
in the titanotheres is analogous to that of the rhinocer- 
oses; in other characters it is unique. 

In brief, although the feet and teeth of the titano- 
theres are conservative, retaining more or less their 
ancestral form, the skull is extremely progressive, 
presenting greater variety of form and proportion than 
that of any other perissodactyl. In every subfamily 
the skull is so plastic as to be in a continuous state 
of transformation, radiating into great extremes of 
structure, which show the widest possible differences of 
proportion and in which all resemblance to the ances- 
tral type is completely lost. The horns finally become 
the dominant feature of the skull and appear to condi- 
tion the evolution of all the other parts. 

The transformation of the Eotitanops type of skull 
into the Brontotherium type, which is epitomized in 
Figure 709, as compared with the transformation in 
other perissodactyls, shows the following principal 
features: (1) Loss or reduction of parts is relatively 
infrequent; (2) few rectigradations (new parts) arise, 
the only absolutely new features being the horns and 
additional cusps on the premolar teeth; (3) increase in 
bulk is enormous and with one exception is continuous 
and progressive in every branch; (4) changes of pro- 
portion are great in all parts of the skeleton, especially 
in the skull. 

Loss or reduction of parts. — The known losses and 
numerical reductions of parts may be summarized as 
follows: (1) Incisor teeth (f) retained in some phyla 
(Brontotherium, Brontops) in more or less functional 
condition; incisors entirely lost (%) in other phyla 
(Menodus, Mega cer ops); (2) intermediate conules on 
the superior grinding teeth lost in all phyla, a loss 
that was the final cause of the mechanical imperfection 
of the teeth and of the extinction of the family, in 
contrast with the horses, in which the conules save 
the race; (3) trapezium, inner bone of the second row 
of the carpus, variable or absent. 

Rectigradations. — New cusps and cuspules appear 
on the upper and lower premolar teeth, being inde- 
pendently developed in each of the titanothere phyla 
at more or less rapid rates of evolution, and osseous 



horns appear at the junction of the nasal and frontal 
bones, being independently developed in five distinct 
phyla and more or less rapidly evolving. 

Harmonic increase in size. — During the period of 
time represented by the Wind River, Bridger, Washa- 
kie, true Uinta (C), and Chadron formations there 
was a great increase in bulk or mass of body in the 
titanotheres. In the smallest known true titanothere, 
Eotitanops gregoryi, the skull was smaller than that of 
a wolf, and the bodj' therefore probably weighed less 
than 150 pounds. On the other hand, the largest 
titanothere certainly weighed much more than an 
adult black rhinoceros, whose body weight is estimated 
in Brehm's Tierleben as 1,600 kilograms, or 3,500 
pounds. It would weigh less, however, than a large 
African elephant, whose estimated weight is 4,000 
kilograms, or 8,800 pounds. Therefore, if we assign 
a weight of 2,800 kilograms, or about 6,000 pounds, 
to the largest titanothere its weight would have been 
about 40 times as great as that of its diminutive 
ancestor Eotitanops gregoryi. 

Changes in proportions (allometrons) . — Each phylum 
of the titanotheres has its distinctive rate of increase 
of the grinding area of the teeth in relation to the 
length of the skull, as shown in the accompanying 
table. Of the Oligocene phyla, Menodus has the 
relatively largest grinding area; Megacerops and Bron- 
totherium have the relatively smallest. 

Measurements , in millimeters, showing progressive increase in 
length of true molar series as compared with total length of 
skull 



Eotitanops borealis 

Limnohyops laticeps 

Palaeosyops major, Am. Mus. 12182 

Palaeosyops leidyi, Am. Mus. 1544 

Palaeosyops leidyi, Am. Mus. 1516 

Manteoceras manteoceras. Am. Mus. 

1569 

Manteoceras manteoceras, Am. Mus. 

1545, now in Nat. Mus 

Manteoceras washakiensis. Am. Mus. 

13165 

Telmatherium ultimum, Am. Mus. 2060_ 
Brontotherium leidyi, Carnegie Mus. 93_ 

Brontotlierium gigas 

Menodus giganteus. Field Mus. 5927 

Menodus giganteus, Am. Mus. 505 



Basilar 
length 
of skull 



313 
"410 
389 
415 

414 

492 

523 

-490 
510 
665 
830 
825 
777 



54 

90 

94 

100 

102 

103 

118 

116 
129 
190 
241 
270 
250 



Ratio of 
m'-ms to 
basilar 
length 
(per 
cent) 



17 
21 
21 
24 
24 

20 



23 
25 
28 
29 
32 
32 



Comparison of the percentage increase of the 
body as a whole and of the grinding teeth as a whole 
shows that the increase of the grinding teeth in 
bulk approximately kept pace with the increase in 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



783 



the bulk of the animal as a whole. Such an increase 
might have been expected, for it can not be said that 
the grinding teeth of Brontotherium gigas are greatly 
superior mechanically to those of Eotitanops iorealis. 
If these grinders of Brontotherium gigas had been 
perfected through hypsodonty and plication of the 
enamel, as in the hypsodont horses, rhinoceroses 
(Elasmotheriujn) , and elephants, it is obvious that the 
bulk of the grinding teeth need not have been so 
great. In other words, BrontotJierium provided for 
the greatly increased size of its body by increasing the 
size of its grinding teeth rather than by mechanically 
improving them. 

The above series of comparative measurements 
shows that in the Oligocene titanotheres the true 
molars (m^ to mf) were, relatively to the skull. 



of the disharmonic evolution which is in progress in all 
parts of the skull and teeth. The above comparisons 
include changes which distinguish the two generic 
extremes in the evolution of the family, namely, 
Eotitanops and BrontotJierium; but the following table 
shows that similar changes distinguish each of the 
three principal phyla of the lower Oligocene, namely, 
Brontops, Menodus, and Brontotherium. 

Each genus has its distinctive velocity in the 
evolution of each cranial and dental character. The 
researches made for this monograph have shown that 
the most significant diagnostic feature of a genus is 
the relative rate of evolution of the separate parts of 
which the skull and teeth are composed. For ex- 
ample, we observe that in Brontops the increases in 
different parts of the skull and teeth are more nearly 



MINUS, RETROGRESSIVE EVOL. PLUS, PROGRESSIVE EVOLUTION 


-S.0 .0 ^ 30 .0 ,,0 T ,0 . 30 « so « ,0 » . ,00 ^^"^l"T,0 


fpmj:-co/2d) 


Brontops b7vch?—J3. robusf 
MenodiW Twloc. ^—^.giifaji^ 
Bronto'J^ leidyi --B.curUmi. 




t3l. 


"726 


LENGTH OF Srontops 
TOOTH ROW Menodus 
(p'-TTi^) Brontotheruxm 




♦ 25 


" + 20 


BREADTH OF Broniops 
SKULL iMenodus 
Trans, xy^'' Brontotherium 




+ 39 


♦ 10 


♦ 40 


LENGTH OF ^^^^<^P^ 
HORNS W^riodi^ 

Brontotheruun 




+ H3 


♦ 300 




LENGTH OF Brontops ■ T^ 






B n/a 


+ 33 


jyronzozn, _^ 


BREADTHOF Broniops 
NASALS {Menodus 

Brontot/herium. 






3rontops 

LENGTH .r J7 

^y^j {Menodus 

BrordotheriuTn 




4f 


LENGTH 't""^'" 




♦ 32 


P'-P* 


BrontotheriuTTz, 


jiT(estJ 



Figure 708. — Disharmonic evolution, progressive and retrogressive, shown in eight characters of the skull 

and teeth of Brontops, Menodus, and Brontotherium 

Range of species from the lower Titanotherium zone (Chadron A) through the upper Titanotherium zone (Chadron C) . 



twice as long as in the Eocene Eotitanops. For 
example, in Brontotherium gigas the true molars are 
29 per cent of the skull length, in Eotitanops horealis 
only 17 per cent; yet in Menodus the molars were 
even proportionately larger, namely, 32 per cent of 
the skull length. In the grinders as a whole the 
relative gain in the true molar series is, however, offset 
by the loss in the premolar series, for the ratio of the 
premolars to molars drops from 63 per cent in Eoti- 
tanops to 42 per cent in Brontotherium curtuin. Yet 
the gain in the molar series more than offsets the rela- 
tive loss in the premolar series. 

DISHARMONIC EVOLUTION IN LENGTH AND BREADTH OF 
SKULL 

The above disharmonic evolution of the true molars 
as compared with the skull and of the premolars as 
compared with the true molars is a very simple example 



harmonic, whereas in the related short-horned Meno- 
dus they are most widely disharmonic. This con- 
trast may be summarized as follows: 

Percentages of disharmonic evolution in eight characters in Bron- 
tops, Menodus, and Brontotherium 





Brontops 
braehy- 
cephalus 


Menodus 
giganteus 


Bronto- 
therium 
curtum 




+ 31 

+ 25 
+ 39 
+ 143 
-11 
+ 60 
+ 24 
+ 32 


+ 37 

+ 75 

+ 7 

+ 300 

+ 33 

+ 9 

+ 47 


26 




+ 20 


Breadth of skull 


+ 40 




+ 250 




-56 




2. 9 


Length of true molars (m'-m^) 

Length of premolars (p^-p*) 


+ 22 
+ 12 



784 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



These percentages of increase and decrease are 
quite distinct from relative size. For example, the 
grinding series of Alenodus is relatively much longer 
than that of Brontotherium, yet the percentage of 
increase is less during lower Oligocene time; the horns 
of Menodus are much shorter than those of Bronto- 
therium, yet the percentage of increase in Menodus is 
greater. 



In Menodus the skull gains 37 per cent in length and 
7 per cent in breadth; it lengthens 30 per cent faster 
than it broadens. 

EVOLUTION OF THE SKULL IN CORRELATION (COADAP- 
TATION) WITH THAT OF THE TEETH AND HORNS 

The skull evolution of the titanotheres passes 
through two great phases. The first, in the lower and 
middle Eocene, is in coadaptation to the grinding 




Figure 709. — Evolution of the skull in the titanotheres 

Top and palatal views. One-twelttb natural size. A, Eiiitanops princeps: B, Manteoeeras manteoceras: C, Brontotherium leidyi; 
D, Brontotherium curtum. Note the lengthening of the cranial portion of the skull, the origin of the horn swellings and 
their forward displacement in front of the orbits, the widening of the occipital crests, and the spreading of the buccal process 
of the zygomatic arches. The palatal view reveals the progressive widening of the molar teeth, the partial molarization of the 
premolars, and the relatively conservative (nonprogressive) character of the base of the cranium. 



The most interesting contrast is in the skulls, 
namely, between those which are progressively brachy- 
cephalic, like Brontops, and those which are progres- 
sively dolichocephalic, like Menodus. In Brontops the 
skuU gains 31 per cent in length and 39 per cent in 
breadth; it broadens 8 per cent faster than it lengthens. 



teeth; the second, in the upper Eocene and lower 
Oligocene, is in coadaptation to the grinding teeth and 
to the support of the increasingly powerful horns. 
Both these phases are affected in all their characters 
by progressive brachycephaly or dolichocephaly, pro- 
ceeding from an original mesaticephalic condition. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



785 



Eocene browsing skull -phase. — In the Eocene phase 
we conapare the primitive titanothere skull, as seen 
above in the synoptic plate, with the skulls of other 
Eocene and recent browsing perissodactyls, having 
short-crowned, brachyodont grinding teeth, in which 
there is no special modification of the nares for a pre- 
hensile upper lip. The general resemblance 
in proportions is quite close; the face and 
cranium are of equal length, the orbit is placed 
above the last superior grinder. 

The changes that took place are profoundly 
disharmonic. They show progressive brachy- 
cephaly, mesaticephaly, and dolichocephaly, 
affecting more or less similarly all other trans- 
formations, and are as follows: 

1. The cranium is greatly elongated — 
opisthopic dolichocephaly, dolichocrany. 

2. The face is relatively abbreviated— - 
proopic brachycephaly, brachyopy. 

3. The orbits are shifted forward (dis- 
placed) from above m^ to above p^. 

4. The horns, which originate directly 
above the orbits, are shifted forward to a point 
above the canines, in the long-horned forms 
especially. 

I 5. As the horns shift forward in the long- 
horned forms the nasals are absorbed and 
relatively abbreviated. 

6. The cranium is flattened and elongated 
both in the long-horned and short-horned 
dolichocephalic and mesaticephalic genera. 

7. The occiput is broadened as the cranium 
is flattened. 

8. To fill the spaces of the enlarging skull 
the frontal ethmoid and olfactory sinuses are 
greatly expanded. 

9. The zygomatic arches are broadly ex- 
panded — buccal brachycephaly. 

10. The auditory meatus is closed inferiorly 
by the approximation of the postglenoid and 
post-tympanic processes. 

11. The dental series, expanding in upper 
Eocene time, is relatively constant throughout 
the lower Oligocene, taking up about half the 
length of the skull. 

The above eleven changes, which are begun 
in the upper Eocene phyla of titanotheres, 
culminate in all the Oligocene phyla. In the 
Oligocene they represent the second phase of 
titanothere skull evolution, in which there 
is a double coadaptation to the functions of the 
great grinding teeth and of the horns, respectively. 
In coadaptation to the greater or less dominance 
of the grinding teeth or of the horns we find a large 
number of correlated and compensatory characters. 
For example, in Menodus, in which the grinding teeth 
are superbly developed and which show some tendency 



even to hj'psodonty, the horns remain of moderate 
length, the canine tusks in compensation become 
enlarged and relatively effective weapons, the nasals 
are moderately abbreviated, and the skull as a whole 
is mesaticephalic, the zygomatic arches being moder- 
ately expanded. 




Orbit (ant. border)' 



'Uwfmesostyle) 



Brojiiotheriujiv 



Figure 710. — Relative proportions of the skulls of Manteoceras and 
Brontothei'iutn 

Contrast in the proportions of the facial and cranial parts of the skull in Eocene and OUgocene- 
titanotheres. Middle stage (Manteoceras) in dots, lower Oligocene stage (BrontotJierium) in 
heavy black line. Both figures are reduced to the same absolute length from premasillaries to 
occipital condyles. In A the premaxillaries and the condyles, respectively, of the two skulls 
are superposed, showing the forward shifting of the horns and orbits and the lengthening of 
the cranial region in the Oligocene type. In B the anterior rim of the orbits of the two skulls 
is made to coincide, emphasizing the shortening of the face in the Oligocene type. 

Brontotherium presents an extreme contrast. The 
horns, which are greatly elongated, are the dominant 
character of the skull; the nasals are absorbed and 
abbreviated; the canine teeth are blunt and less effec- 
tive as weapons; the grinding teeth evolve slowly and 
do not become subhypsodont; the buccal brachy- 
cephaly is very extreme. 



786 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



ARE THE PKOPOETIONS OF SKULL AND TEETH 
ADAPTIVE? 

BracTiycepTiahj, mesatkephaly , dolichocephahj. — Be- 
sides the dominance of the development of the grind- 
ing teeth and horns there is repeatedly manifested in 
the Eocene and Oligocene titanothere crania progres- 
sive brachycephaly, persistent mesaticephaly, and 
progressive dolichocephaly. The adaptive nature of 
these profound changes of proportion, which affect 
the skull as a whole, is less obvious than the domi- 
nance of the grinding teeth and horns, because in the 
Eocene titanotheres Palaeosyops and DolicliorMnus 
extreme brachycephaly and dolichocephaly appear to 
be expressed in general changes of proportion which 
affect all the characters of the skull and teeth and 
seem to proceed to inadaptive extremes. Similarly, 




Figure 711. — Proportions of skulls of Palaeosyops, Dolichorhinus, and 
Eotitanops 

Midsection of the crania of Palaeosyops (A) and Dolichorhinus (B), superposed, showing the 
straight (orthocephalic) and the flexed (cyptocephalic) condition of the faciocranial axes 
C, Hypothetic outline of the skull of Eotitanops, drawn to the same scale. 

in the comparison above of Brontops and Menodus we 
observe that the Brontops skulls are increasing in 
buccal brachycephaly; moreover, that this brachyce- 
phalic tendency is manifest in all the measurements 
of the teeth as compared with those of the dolichoce- 
phalic Menodus. Thus we can immediately distin- 
guish a grinding tooth of Menodus because it is longer 
than it is broad, in contrast to a grinder of Brontops, 
which is broader than it is long. 

Conclusion. — It is not demonstrated that all propor- 
tional changes are adaptive; some appear to be 
relatively inadaptive. 

The above is an epitome of the entire evolution of 
the titanothere skull and of the main tendencies of 
titanothere skull evolution, which are altogether dif- 
ferent from those in any other line of perissodactyls. 
Each of the Eocene branches, as described in detail 
in Chapter V, exhibits a distinctive line of cranial 



and dental evolution in which various extremes of 
brachycephaly, dolichocephaly, and cyptocephalj^ 
evolve more or less rapidly. 

RADIATION AND DIVERGENCE IN EOCENE SKULLS 

The divergences of the skull from the primitive type 
in the various Eocene branches are doubtless adapta- 
tions to the browsing and grazing habits in different 
habitats. While in each subfamily branch the 11 main 
tendencies of titanothere skull evolution, enumerated 
above, are more or less clearly manifested, there ap- 
pear certain special generic tendencies distinctive of 
each line. Some of these changes are prophetic of 
those that occur in the Oligocene, although they may 
appear in lines that become extinct in the Eocene and 
do not lead into the Oligocene genera. They are 

therefore parallel or convergent and not truly 

ancestral characters. 

EVOLUTION OF THE SKULL IN EOCENE TIME 
PARTLY PROPHETIC OF THAT IN OLIGOCENE 
TIME 

Ahhreviation of the jace. — The abbreviation 
of the muzzle and premaxillary borders in 
Palaeosyops and Metarhinus, the abbreviation 
of the premolar series and elongation of the 
molar series in all the Eocene phyla is pro- 
phetic of Oligocene evolution. 

Shifting and reduction of orhits. — The for- 
ward shifting and reduction in the size of the 
orbits occur independently in .several of the 
Eocene phyla, such as Palaeosyops, Manteo- 
ceras, and Metarhinus. 

Flattening of the cranium. — The spreading 
of the supratemporal crests, the flattening 
of the top of the cranium, the closure of the 
superior cranial and facial sutures occur inde- 
pendently in Palaeosyops and Dolichorhinus. 
In the latter the flattening of the cranium 
proceeds more rapidly than in Manteoceras, 
which is ancestral to certain of the Oligocene titano- 
theres (Brontops). This feature misled Osborn (1908. 
318) to suppose that Dolichorhinus was an ancestor 
of the Oligocene genera, an error corrected by Hatcher 
in his description of Protitanotherium. 

Massive and slender zygomatic arches. — The massive, 
wide, arching zygomata of Palaeosyops present an 
independent parallel to those of the basal Oligocene 
titanotheres. The zygomata of Telmatotherium are 
deep but not spreading; the zygomata of Manteoceras 
are prophetic of the Oligocene Brontops; the slender 
zygomata of Diplacodon are analogous to those of 
Dolichorhinus. 

INDEPENDENT EVOLUTION OF THE SKULL IN EOCENE 
TIME NONPROPHETIC OF THAT IN OLIGOCENE TIME 

The features of the evolution of the Eocene skull 
that are quite independent of those of the evolution 
of the Oligocene skull are principally the following: 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



787 



Brachycephaly, extreme of, in the Palaeosyopinae. 

Dolichocephaly, extreme of, in the Dolichorhininae. 

Cyptocephaly, deflected facial region, extreme of, 
in Dolichorhininae. 

Cyptocephaly, upturned facial region, beginnings of, 
in Metarhinus, Rhadinorhinus (possibly prophetic of 
Megacerops). 

The pi-ofound changes in the cranial axis produced 
by the cyptocephaly of JDolichorJiinus in comparison 
with the orthocephaly of Palaeosyops are shown in the 
accompanying diagrams of Palaeosyops and Dolicho- 
rhinus. Gregory was the first to observe the up- 
turned facial region of Metarhinus as possibly sugges- 
tive of the greatly abbreviated and upturned muzzle 
of Megacerops of the lower Oligocene. 

The deflected cranial axis of Dolichorhinus, as 
described in Chapter V, section 4, is connected 
functionally with its peculiar habits of feeding. These 
cranial sections are important not only as illustrating 
the flexure in the cranial axes, but also the great 



development of the air sinuses of the facial and cranial 

regions and the relatively small area occupied by the 

brain. 

ZYGOMATIC INDICES 

As explained in Chapter V, section 1, the zygomatic 
index is obtained as follows: Transverse measure- 
ment of the zygomatic arches X 100 h- basilar length 
of the skull. 

The terms brachycephaly and dolichocephaly as 
used in this monograph are based chiefly on the 
zygomatic index in the comparison of the total length 
of the skull with the width across the zygomatic 
arches, rather than in the comparison of the actual 
width and length of the cranium proper. From the 
following table of zygomatic indices it appears that 
the primitive titanotheres, like other primitive peris- 
sodactyls, are dolichocephalic or mesaticephalic. 
Arranged in descending order according to increase of 
zygomatic breadth, irrespective of direct descent, the 
crania are as follows : 



Proportions of length and hreadth of crania of titanotheres 

[Measm-ements in Diillimeters] 



Form of skull 



I. Primitive crania: 

Phenacodus wortmani 

Eohippus venticolus 

Hyrachy us sp 

Lophiodon leptorhynclius 

Eotitanops borealis 

II. Progressive dolichocephaly: 

Metarhinus earlei 

Mesatirhinus petersoni (No. 1556), 
Dolichorhinus (No. 1852, ?) 



III. Progressive brachycephaly: 

Manteoceras inanteoceras — 

No. 12678 

No. 2353 

No. 1569 ■ 

Telmatherium ultimum 

IV. Palaeosyops series: 

Palaeosypos leidyi (No. 1544) 

V. Brontops series, progressive brachycephaly: 

Brontops braohy cephalus 

Brontops dispar 

Brontops robustus 

Diploclonus amplus 

VI. Brontotherium series, progressive brachycephaly: 

Brontotherium leidyi 

Brontotherium hatcheri 

Brontotherium gigas 

Brontotherium platyceras 

VII. Menodus series, dolichocephaly, progressive dolichocephaly: 

AUops serotinus 

Allops crassicornis 

Menodus giganteus 



147 
129 
263 
360 
313 

393 
438 
550 



500 

"465 

492 

500 

415 



58 

53 

122 

176 

163 

240 
207 
240 



°294 
277 

" 310 
300 

310 



73-87 

77-87 

91 

66 

74 

89 

110 

74 

75 

62-70 



Dolichocephalic. 
Do. 
Do. 
Do. 
Do. 

Mesaticephalic . 
Dolichocephalic. 
DoUchocephalic (hy- 
perdolichocephalic) 



Mesaticephalic. 
Do. 
Do. 
Do. 

Brachycephalic. 

Do. 
Do. 
Do. 
Do. 

DolichocephaUc. 
Mesaticephalic. 
Brachycephalic. 
Hyperbrachycephalic. 

Mesaticephalic . 

Do. 
Dolichocephalic . 



' Estimated. 

101959— 29— VOL 2- 



788 



TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



It is yerj important to note that the indices given 
in the table are zygomatic (buccal) (brachycephalic 
and dolichocephalic) as distinguished from true cranial 
(head) indices. The cranium proper is extremely 
elongate. For example, in BrontotJierium curtum it 
extends backward, overhanging the occiput, and for- 
ward, overhanging the orbits — in fact, the elongation 
of the craniofacial summit of the skull is strictly 
dolichocephalic in the original sense of the word as 
defined by Ketzius. 

DIFFEEENTIAI EVOIUTION IN SKUH PROPORTIONS IN 
THE SEVERAL PHYIA OF TITANOTHERES 

During the Eocene the dominant tendency in the 
Manteoceras and Dolichorhinus phyla was increase in 
the length of the middle portion of the cranium. This 
became extreme in DolichorMnus. In Palaeosyops and 
Limnohyops the middle portion of the skull was not 
elongated. During the upper Eocene and lower Oligo- 
cene this increase of the middle portion of the skull was 
surpassed in most phyla by the broadening of the skull, 
and especially of the zygomatic arches. During lower 
Oligocene time (Chadron A to C) the three main phyla 
exhibit approximately equal increments in total 
cranial length, Menodus being only slightly in advance 
of the others in its percentage increase, as shown in 
the accompanying diagram (fig. 708). 

During the same period the zygomatic breadth, 
already very great in male skulls of Brontops hrachy- 
cepJialus, shows a percentage of increase to the stage 
of Brontops robustus relatively less than in the Meno- 
dus phylum. The reason Menodus during the same 
geologic period acquired greater zygomatic increase is 
that the primitive form Menodus heloceras was mark- 
edly mesaticephalic, lacking the zygomatic expansions 
entirely, whereas in the largest male, M. giganteus, 
these expansions caused the skull to attain a width 
index of 55.3. Similarly, the horns appear to increase 
in length very rapidly in Menodus, because of the wide 
range from the extremely short horns of the lower 
Oligocene M. heloceras to those of the great Field 
Museum skull of M. ingens, in which they have an 
outside length of 290 millimeters. 

In the Menodus phylum all the parts except the 
horns enlarge more uniformly and harmonioiisly than 
in the BrontotJierium phylum. It should be noted 
that although the total length of the basilar axis of 
the skull exhibits a similar percentage of increase in 
the three main phyla, this increase is differently 
divided between the different parts of the skull. 
In the Menodus phylum, for example, the facial 
region elongates naore rapidly than in the Bronto- 
therium phylum, in which the midcranial region 
elongates rapidly and the facial region is abbreviated. 



PALATE AND SHIFTING POSTERIOR NARES 
In general, there is a broadening of the palate and 
gradual shifting backward of the posterior nares. In 
the more primitive forms the posterior nares opened 
more anteriorly between the second pair of grinding 
teeth, m-. In Palaeosyops and Limnohyops they 
frequently open opposite the middle of m^ In other 
middle Eocene genera the posterior nares open either 
opposite the posterior part of m^ or between m^ and 
m^ In the Oligocene genera the opening of the 
posterior nares is variable, generally opposite m' and 
sometimes behind m'. 

RUDIMENTS OF HORNS ARISING INDEPENDENTLY IN 
EOCENE PHYLA 

Development of liorns and teeth. — The independent 
origin of the horn rudiments at the junction of the 
frontal and nasal bones in members of four sub- 
families of Eocene titanotheres is one of the most 
significant facts discovered in the course of the 
researches for this monograph. In the accompanying 
table the more or less rapid development of the 
incipient horns is contrasted with the more or less 
rapid development of the incipient cusps on the 
premolar grinding teeth. 

As is shown very clearly in Chapter V (pp. 266-267) 
and Chapter XI (pp. 810-811) the horn rudiments 
appear geologically early and are very prominent in cer- 
tain phyla ( Manteoceras, Mesatirhinus) at a time when 
in other contemporaneous phyla (Palaeosyops, Telma- 
therium) the skulls are still smooth, entirely hornless. 
It is only in the late stages of Palaeosyops and in geo- 
logically late stages of Telmatherium that the horn 
rudiments appear at all. These facts have a double 
significance, which is fully discussed in Chapter XI 
(pp. 883-884). First, it would seem that there is a pre- 
disposition for the evolution of horns in a certain 
region of the skull; second, that this predisposition 
manifests itself not uniformly but in some phyla 
earlier than in others. There are two deductions from 
these phenomena: 

First, the origin of frontonasal horns is a titanothere 
family characteristic. 

Second, this origin at earlier or later geologic stages 
is a subfamily or generic characteristic. 

Comparison with other perissodactyls. — In the Peris- 
sodactyla, as in all other ungulates, there are three 
more or less concomitant steps in the origin of horns, 
namely : 

1 . Psychic predisposition to use the horn as a means 
of offense and defense, correlated with other offensive 
and defensive psychoses, linked also with male and 
female sex glands. 

2. Dermal and epidermal thickening, derm pads, 
and epidermal horny sheaths, protecting layer. 

3. Osseous swellings, exostoses, bony swellings be- 
neath the dermal pads or horny sheaths. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 

Unequal evolution of horns and of premolars in the titanotheres 

Horns 



789 





Lambdotherium 
zone (Wind 
Eiver B). (500 
feet) 


Orohippus zone 
(Bridger A and B). 
(650 feet) 


TJintatherium zone 
(Bridger C and D). 
(725 feet) 


Eobasileus - Dolicho- 
rhinus zone (Washa- 
kie B and Uinta A 
andB). (250 feet) 


Dipiacodon zone 
(Uinta C). (600 feet) 


Titanotherium zone 
(Chadron A to C). 
(180 feet) 


Lambdotheriinae 

Eolitanopinae 


Hornless. 
Hornless. 


Hornless 


"Horns" barely 
visible and 
evolving very 
slowly. 

Horns readily 
visible. 




Horns barely in- 
dicated and 
not develop- 
ing at all. 

Horns expand- 
ing rapidly. 




Telmatheriinae 






Manteoceratinae 










(including Proti- 
tanotherium) . 
Dolichorhininae . 




Horns incipient . 


do 


Horns expand- 
ing gradually. 




Brontopinae . _ -_ 






Menodontinae . 












very small, ex- 
panding rap- 
idly. 
Do 


Brontotheriinae 


























small, expand- 
ing very rap- 
idly. 



Premolars 



Eotitanopinae 


Premolar s 
evolving 
slowlj'. 


Premolar s 

e V 1 V i n g 
slowly . 


Premolars 

evolving 

slowly. 
Premolar s 

evolving very 

slowly. 
Premolar s 

e v 1 v i n g 

slowly . 
Premolar s 

evolving more 

rapidly. 


Premolar s 

evolving 

slowly. 
Premolars 

evolving more 

rapidly. 


Premolar s 
evolving 
slowly. 


















(including Proti- 
tanotherium) . 
Dolichorhininae. _ 




























ing slowly. 














ing more rap- 
idly. 














ing very rap- 
idly. 



790 



TITANOTHERES OF ANCIENT Wl'OMING, DAKOTA, AND NEBRASKA 



Bony exostoses. — All that we see in fossil skulls are 
the bony exostoses, or thickenings of the outer bony 
layer, with expansion of the cancelous bony tissue 
beneath, as in the horn rudiment of PaJaeosyops in the 
accompanying figures. 

In the titanotheres, as in the Amblypoda {Baihy- 
opsis, Coryphodon, Vintaiherium) , the bony exostosis 
expands into a bony horn covered with a thickened 
dermal pad lacking any evidence of a horny epidermal 
sheath and certainly without a distinct epidermal 
horn composed of agglutinated hairs of the rhinoceros 
type. The bony horns of titanotheres and amblypods 
become unique among ungulates as great fighting 
weapons covered with thickened skin pads, rounded, 
triangular, oval, and finally platelike in form. The 
absence of horny sheaths is indicated by the absence 
of the channels for nutrient blood vessels such as 
surround the bony horn cores of ruminants and of 
rhinoceroses. The dermal pad of the titanotheres 
may have been similar originally to the "wart" of 
the wart hog (Phacochoerus), which we may observe 
as partly protecting the ears when two wart hogs are 
butting each other. ^- The rounded bony horns of the 
giraffes are probably degenerate because partly 
covered with hairs and seldom used." 

MODES OF ORIGIN OF THE HORNS OF THE TITANOTHERES 

1. Horns originate independently at different periods 
in four Eocene titanothere phyla. 

2. Horns originate invariably above or slightly in 
front of the orbits, on the line of the nasofrontal suture. 
The frontal element more or less completely overlaps 
the underlying nasal elements. 

3. Horns apparently have a predisposition or pre- 
determination to originate in this particular region of 
the face; no evidence of bony horn rudiments is 
observed in any other part of the skull; the rudiments 
are invariably continuous with the frontals pos- 
teriorly, with the nasals anteriorly. 

4. Rudiments of horns are smooth, rounded, not 
primarily rugose, with no surface indications of 
channels for large blood vessels. They are totally 
unlike the vascular horn rudiments of rhinoceroses, such 
as Aceratherium incisivum. 

5. The evidence indicates the existence of thickened 
epidermis rather than of superficial horny sheaths. 

6. Horns are equally developed (so far as known) 
in both sexes in the original Eocene stages. 

7. Horns are secondarily more prominently devel- 
oped in males and less prominently in females. 

8. Horns secondarily shift in position on the skull 
from a point immediately above the orbits to a point 
more or less above the canines. 

" Memorandum by W. K. Gregory. See also note on the "horn of Phacoclioaus" 
in review of 0. C. Marsh's "The ' Brontotheridae,' a new family of fossil mammals.'' 
Nature. Jan. 22, 1874, p. 277. 

'J Valuable papers on the genesis of horns are those of Hans Gadow, 1902.1; E. 
Ray Lankester, 1902.1; J. ririeh Durst, 1902.1; and especially Max Weber, 1904.1, 
p. 23, all cited in full in the bibliography of this chapter. 



9. Horns are originally elongate and oval in form, 
secondarily of triangular form {Menodus), of rounded 
form {Megacerops), of transversely oval form {Bronto- 
therium). 

10. No evidence is found that the bony horn core 
develops an independent center of ossification. 

From the beginning the titanothere horns are paired 
and supraorbital in position, as in the Cervidae, 
Antilopinae, and Giraffidae, among the artiodactyls. 
and as in Colonoceras among the perissodactyls. 

PHYLETIC DIVERGENCE IN TIME OF ORIGIN OF HORNS 

As is shown in the accompanying diagram (fig. 712), 
the horns are precocious in origin in the Dolichorhin- 
inae — that is, they appear in Mesafirhinus of Bridger 
C as elongate oval swellings overhanging the orbits, 
at a pei'iod when no trace of horn swellings can be 
observed on the face of the contemporary Palaeo- 
syops leidyi. These elongate oval horns of Mesati- 
rJiinus develop into the prominent oval horns of 
DolicJiorJiinus Tiyognathus of Washakie B. 

Contemporary with MesatirJiinus is the phylum 
Manteoceras manteoceras, the "prophet horn" titano- 
there, also found near the base of Bridger C. The 
apparent successor of Manteoceras, namely, Protitano- 
therium emarginatum of Uinta C 1, exhibits a horn 
type which is directly successive to that of Manteo- 
ceras and similar in form except that it is moved 
much farther forward over the narial opening. 

ACCELERATED DIRECT EVOLUTION OF HORNS IN THE 
"PROPHET HORN" PHYLUM 

In Figure 712 are illustrated five direct stages 
of evolution, namely, (1) the rudiment of the horn 
in Manteoceras of Bridger C and its shifting forward 
in Protitanotherium. of Uinta C 1, (2) its ontogenesis in 
Brontops hrachycepJialus of Chadron A, (3) its adult 
form in Brontops hrachyceplialus of Chadron A, (4) its 
adult form in AUops marshi of Chadron B, and (5) its 
ultimate form in Brontops robustus of Chadron C, 
final stage of evolution. 

RETARDED EVOLUTION OF HORNS IN THE PALAEOSYO- 
PINAE AND TELMATHERIINAE 

In contrast to the early geologic appearance of the 
bony horn in the Eometarhinus, Mesatirhinus, Doli- 
chorhinus, and Manteoceras phyla we observe the 
extreme retardation of their appearance in the Palae- 
osyops and Telmatherium phyla. No horn rudiments 
are present on the sides of the face in the species of 
Limnohyops and Palaeosyops of Bridger C, contem- 
porary with Manteoceras and MesatirJiinus. As late 
as Bridger D Palaeosyops roiustus shows a rounded 
bony horn rudiment, which becomes quite conspicu- 
ous in aged individuals. Skulls of the contemporary 
Telmatherium are unknown, so we can not state 
positively that Telmatherium at this stage lacks the 
horn rudiment. This rudiment is first directly 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



791 



observed as a very faint sessile, rounded swelling, on 
the side of the face of a male Telmaiherium ultimum 
of Uinta C 1. It would appear from this evidence, 
which is very strong, if not absolutely conclusive, 
that whereas in Manfeoceras and Dolichorhinus the 
rudiments of horns were doubtless present at the 
time of the deposition of Bridger B and are well 



rliinus) is entirely consistent with the retarded or 
accelerated appearance of other characters, such as the 
secondary cusps on the premolar teeth, as shown in 
the above table. It is also consistent with the fact 
that the Oligocene titanotheres divide into two great 
branches, namely, the short-horned titanotheres, 
including the Menodus and Brontops lines, and the 




Figure 712. — Evolution of the frontonasal horn swelling in members of the titanothere subfamilies 
Manteoceratinae and Brontopinae 

A, Manteoceras manteoceras, middle Eocene; B, Proiiianotherium emarginntum, upper Eocene; C, calf stage of Oligocene titanothere (,fBron- 
tops irachycephalus): D, young titanothere (Brontops hrachycephalus); E, Allops marshi; F, Brontops robustus. The horn swelling 
is at the junction of the nasals and frontals, above the orbits. The frontal swellings grow forward, overlapping the nasals. In the 
final stage the horn is far in front of the orbits and may be much compressed anteroposteriorly. 



developed at the base of Bridger C, they do not appear 
in Palaeosyops until Bridger D, and not certainly in 
TelmatJiermm until Uinta C 1. 

This indirect evidence of the retarded evolution of 
the horns in certain of the titanotheres {Palaeosyops, 
Telmatherium) and of their acceleration in other 
Eocene lines (EometarMnus, Manteoceras, Mesati- 



long-horned titanotheres, including the Megacerops 
and Brontotherium lines. It is, moreover, consistent 
with the evidence repeatedly brought forward in this 
monograph, that the chief difference between phyla 
consists in the distinctive velocities of their character 
evolution rather than in the presence or absence of 
these characters. 



792 



TIT.USrOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



DIVERGENCE IN THE FORM OF HORNS OF THE 
OLIGOCENE TITANOTHERES 

The wide divergence in the form of the Oligocene 
horns is followed in great detail in Chapter VI and 
need not be recapitulated here. In the short-horned 
titanotheres there were doubtless very considerable 
differences in the manner in which the horns were 
used, both in combats between bulls for the possession 
of females and as weapons of offense and defense in 
the combats between the titanotheres and their 
carnivorous enemies. It is probable that the short- 
horned titanotheres used a side thrust of the head, 
while the long-horned types used also a vertical, 
tossing motion. 



ing would explain the absence of sharp points at the 
tips of the horns (as in the ruminants and rhinoceroses) 
as well as the absence of the pointed horny epidermal 
sheath, which is the conspicuous offensive and defen- 
sive weapon in the rhinoceroses. 

CORRELATION (COADAPTATION) OF HORNS WITH 
CRANIA! AND DENTAI CHARACTERS 

There is some reason to believe that the extreme 
development of the horns in Brontotherium platyceras 
is correlated with extreme development of the buccal 
plates on the zygomatic arches, which may have 
served as defensive structures for the side of the skull 
and face. These gigantic buccal plates are most 
strongly developed on the skulls having the largest 




A 

Figure 713. — "Brain easts" (intracranial casts) of titanotheres (A, B, C) compared with the brain of a 

recent rhinoceros (D) 

A, Palaeosyops leidyi, Am. Mus. 1544; B, Mesatirhinus petersoni, Princeton Mus. 10041; C, Menodus giganteus (type of Broniothermm 
ingens), after Marsh; D, Rhinoceros sondaicus, after Beddard and Treves. In the earliest stage (.Palaeosyops) the olfactory lobes 
are relatively large and the cerebral hemispheres are small and do not overlap the cerebellum. The brain cast of Mesatirhinus (B) 
is perhaps somewhat distorted because the cerebra appear to be abnormally wide across the frontal lobes. It is evidently a more 
progressive type, however, than Palaeosyops. In Menodus giganteus (C) the cerebral portion appears to be considerably enlarged 
so as to partly overlap the cerebellum. The anterior lobes, however, are narrower than in the rhinoceros, and the surface was 
probably much less convoluted. The brain as a whole is also much smaller in proportion to the bulk of the animal. 



The connecting crest between the horns is strongly 
developed in both the short-horned and the long- 
horned genera (with the exception of Megacerops) 
and is a very conspicuous feature of the evolution of 
the horns in Bj-ontotherium. This connecting crest 
doubtless served an important mechanical function 
in withstanding the strain of a lateral thrust. In 
Megacerops, on the contrary, there is little or no con- 
necting crest, indicating the absence of lateral thrust 
and the use of the horns principally in the tossing up 
and down motion. In Brontotherium platyceras the 
horns attain supreme development, and in head to 
head combats the connecting crest of one animal 
might engage the malar ridges at the side of the horn 
of the opponent. This general use of the horns for 
lateral striking and butting as well as for vertical toss- 



horns; and that they are not correlated with the 
masseter temporalis and other muscles of mastication 
is proved by the fact that the dental and masticating 
structures in Brontotherium are decidedly inferior to 
the same organs in the short-horned Menodus. These 
buccal plates are also most strongly developed in the 
large males and do not appear in the small-horned 
females. They are thus secondary sexual characters, 
correlated and coadaptive with the powerful develop- 
ment of the horn, as are also the entire osseous struc- 
ture of the anterior part of the skull, the inclosure and 
reduced size of the orbits, the expansion and rugosities 
of the occiput, the powerful ligamentum nuchae at- 
tached to the spines of the anterior dorsal vertebrae — 
in short, the entire bony and muscular structure of 
the anterior part of the body. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



793 



There is some reason to believe that in the combats 
between the males for the possession of females the 
horns attained very high sexual selection value, that 
there was an incessant selection of the large horns at 
the expense of the selection of other characters, and 
that this breeding for large horns may have injuriously 
affected the evolution of the grinding teeth, which is 
inferior in Brontotherium to the molar-tooth evolution 
in the short-horned Menodus. This problem is dis- 
cussed in Chapter XI as among the causes of the 
extinction of the titanotheres. 

CORRELATION OF HORNS WITH SEX 

Sexual disparity in horn development becomes more 
marked as we ascend to higher geologic levels. The 




EVOLUTION OF THE BRAIN 

The brain cavity of MesatirJiinus peter soni and of 
Palaeosyops leidyi, as shown in the intracranial casts 
(fig. 713), is relatively small. Thus it may be stated 
that both the Eocene and Oligocene titanotheres — that 
is, the whole titanothere family — were characterized 
by a small brain. 

We are indebted to Marsh (1884.1) for a discussion 
of the brain characters of certain Eocene titanotheres 
and contemporary Eocene ungulates as compared with 
those of certain related modernized forms; also for 
excellent illustrations (fig. 714) reproduced herewith. 
In the titanotheres (LimnoJiyops laticeps and f Palaeo- 
syops rolustus) we observe again that the brain is 









Relative size of brain and skull in titauutlieres and other Eocene 
perissodactyls, an artiodactyl, and an amblypod 

After Marsh, 1884.1. Scales various. A, Coryphoion hamaius; B, Limnohyops laticeps; C, Palaeosyops robuslus; 
D, Colonoceras agrestis; E, Hyrachyus bairdianus; F, Amynodon advenus; G, Eporeodon socialis 



brontothere females are smaller and wholly different 
in skull structure from the males; there are also some 
differences in dental structure which are difficult to 
comprehend, such as the apparent prominence of the 
cingulum on the grinding teeth of the females of 
Brontotherium and its absence on the grinding teeth 
of the males. Even in the male Brontotherium the 
canine teeth are relatively small and inoffensive as 
weapons, as compared with those of Menodus. In the 
females the canines are greatly reduced in size. This 
would indicate that, as among the existing horses and 
cattle, the males stood guard over the herds of titano- 
theres, protecting the females and the young. 



relatively smaller than in the contemporary Rhino- 
cerotoidea (Colonoceras, Hyrachyus, and Amynodon). 
It is also relatively smaller than in the modern Tapirus, 
Rhinoceros, and Equus cahallus. 

The disparity in the size of the brain of Dolicho- 
rhinus hyognathus and of the modern Equus cahallus is 
very striking. Inasmuch as the skulls of these two 
animals are approximately of the same length, they 
afford an excellent basis of comparison. The ratio of 
weight of brain to weight of body in all the other large 
ungulates has been carefully calculated by Max 
Weber (1897.1) as follows. 



794 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Uatio of weight of brain to weight of body in certain ungulates 



PERISSODACTYLA 



Tapirus indicus L 

Tapirus americanus L. 

Tapirus americanus L_ 
Tapirus americanus L_ 
Equus zebra L 



PHOBOSriDEA 



Katio of 

brain to 

body 



100 

947 
140 
247 




758 0. 13 



Elephas africanus L 


1:375 


.25 


Lived 10 years. 


Elephas indicus L 


1:439 


.23 


Lived about 25 
years. 




1:125 


.8 






1:560 


. 17 




ARTIODACTYLA 










1:305 


. 33 




Camelopardalis giraffa L — 


1:392 


.259 


2 montlis old. 


Camelopardalis giraffa 


1:761 


. 14 


Young. 


Schreb. 








Camelopardalis giraffa 


1:777 


.12 


Lived 22 years. 


Sclireb. 








Boselaphus tragocamelus 


1:585 


. 17 




Sundw. 









[.00 

. 10 
.71 
.40 



1 month 4 days 

old. 
Fatlier of above. 
Young animal. 
Prettv thin. 



The weight of the brains of the early Tertiary 
ungulates has been considered by Lartet (1868.1), 
by Bruce (1883.1), by Marsh (1884.1) by Cope, and 
by Osborn. Cope refers to Lartet's anticipation of 
Marsh's laws of brain weight, which are as follows: 

1. All Tertiary mammals had small brains. 

2. The size of the brain gradually increased during 
the Tertiary period. 

3. This increase was confined mainly to the cerebral 
hemispheres, or higher part of the brain. 

4. In some groups the convolutions of the brain 
have gradually become more complex. 

5. In some groups the cerebellum and the olfactory 
lobes have even diminished in size. 

6. There is some evidence that the same general 
law of brain growth holds good for birds and reptiles 
from the Cretaceous to the present time. 

Marsh enunciated also two additional laws: 

1. The brain of a mammal belonging to a vigorous 
race, fitted for a long survival, is larger than the aver- 
age brain of that period in the same group. 

2. The brain of a mammal of a declining race is 
smaller than the average brain of its contemporaries 
of the same group. 



DENTAL MECHANISM 

The evolution of the upper grinding teeth of the 
titanotheres, as shown in Figure 718, A-E, proceeds 
from a purely brachj^odont, low-crowned browsing 
type to a semihypsodont browsing and grazing type. 
In the earlier stages of evolution {Eotitanops, Lim- 
nohyops, Palaeosyops) the mode of mastication and the 
action of the mandible indicate that the principal 
motion of the jaws was vertical, like that of the 
omnivorous Carnivora. 

In this chopping action of the jaws the pointed in- 
cisors, the relatively long, pointed canines, the rela- 
tively simple premolars, and the low-crowned molars 
were correlated with seven changes as follows: (1) 
The unique feature of the dental evolution is the semi- 
hypsodonty (elongation) of the outer side of the crowns 
of the upper grinding teeth and persistent brachyo- 
donty of the inner side; (2) with this change came the 
inclination of the wearing plane of the tooth row 
toward the midline of the tooth, (3) the complication 
of the premolars by the addition of internal cusps, 
(4) the ratchet and cog like relations of the upper and 
lower molars, (5) the changes in the proportions and 
the arrangement of the masticating muscles, (6) the 
increasing importance of the oblique shearing effect in 
the swing of the mandible, together with (7) its pro- 
nounced fore-and-aft rocking motion (W. K. Gregory). 
These seven coadapted changes resulted in a dental 
mechanism which combined cutting and triturating 
functions in a very complex manner and which may 
have been capable of masticating as wide a range of 
vegetable food as that of the modern Rhinoceros 
(Opsiceros) Mcornis. 

The titanotheres, like other perissodactyls, doubtless 
had a simple stomach and were incapable of ruminat- 
ing. Thus the digestive apparatus as a whole was 
inferior to that of the contemporary camels (Lep- 
taucJienia), of the numerous oreodonts with crescentic 
teeth, and of the ancestors of the modern selenodont 
artiodactyls. Consequently the titanotheres may 
have derived less nutriment from a similar amount of 
food to that consumed by the contemporary artio- 
dactyls or may virtually have been debarred from 
certain kinds of food and certain feeding ranges. The 
wide and increasing prevalence of ruminant artio- 
dactyls after the decadence of the titanotheres seems 
to imply a superior digestive apparatus if not a supe- 
rior dental mechanism, or else a profound, change of 
herbage. 

After the extinction of the titanotheres the only suc- 
cessful competitors of the artiodactyls were certain 
perissodactyls, like the horses and rhinoceroses, espe- 
cially the great Siberian rhinoceroses {Elasmotherium) ■ 
I The Equidae, starting with the same type of molar 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



795 



teeth as the titanotheres, namely, the bunolophoseleno- 
dont, gradually evolved hypsodont molars with highly 
complex enamel foldings and thereby attained an 
efficient grinding mechanism capable of triturating 
the driest and hardest kinds of grasses. (See fig. 718.) 
Consequently the imperfect browsing and grazing 
mechanism of the teeth may have been one 
of the principal causes of the extinction of the 
titanotheres. (See Chap. XI.) As more fully 
discussed in Chapter XI all bunoselenodont 
Eocene perissodactyls and artiodactyls became 
extinct in late Eocene or early Oligocene time 
except only the chalicotheres, which theoret- 
ically retreated to the forests and retained the 
pure browsing habit. 

GEADUAI CHANGE OF DIET 

The evolution of the grinding teeth in the 
titanotheres may indicate that all the phyla at 
first fed upon coarse tubers, roots, and the 
twigs and leaves of low bushes. The cypto- 
cephalic Dolichorhinus may have fared upon 
the harder and smaller varieties of plants and 
thus presented a nearer approach to a true 
grazer. The Oligocene titanotheres probably 
fared upon larger and coarser varieties of plants, 
which were plucked up by the powerful pre- 
hensile lips. The degeneration or complete loss 
of the incisor teeth was supplemented by the 
prehensile lip action. The swollen premolar 
teeth of the ProtitanotJierium stage, ancestral 
to Brontops, seem adapted for browsing and 
crushing food rather than for grinding. The 
pestle and mortar construction of the Oligocene 
titanothere grinding teeth implies that the food 
required to be broken and crushed before being 
ground. The very size of the molars seems to 
imply very coarse food. The selection of food was 
apparently never of the harder siliceous varieties 
sought by the contemporary rhinoceroses and 
horses, otherwise hypsodoiit evolution might have 
been more rapid. The shape of the skull in Oligo- 
cene forms seems to imply that the head was not 
rapidly thrown backward, but that the food was 
sought near the ground, among the low shrubs. 
The browsing African rhinoceros R. (Opsiceros) 
hicornis seems to seek shrubs and browsing food in 
a country affording admirable opportunity for 
grazing. 



diameters of the teeth, whereas in the mesaticephalic 
titanotheres there is an excess of the anteroposterior 
diameters over the transverse. The facts set forth 
below should be noted: 

1. All the species measured show a considerable 
degree of fluctuation in the anteroposterior and trans- 

Cer. 




Figure 715. — Brain of Menodus compared with that of Rhinoceros and 
other quadrupeds 

Proportions of brain in archaic fleft column) and modernized (right column) mammals of 
similar size. Dots show olfactory lobes; black lines show cerebral hemispheres; dashes show 
cerebellum and medulla. All one-fourth natural size. Modified after Osborn, 1910.346. 
A, Araocyon, Eocene flesh eater; Cams (the dog), modern flesh eater. B, Plicnacodus, 
Eocene primitive ungulate; Sus, the domestic pig. C, Coryplwdon, Eocene ungulate; 
Ehinoceros, living rhinoceros of same size. D, Uintalherium, massive Eocene ungulate 
E, Menodus giganteus, Oligocene titanothere. 



HARMONY OF PROPORTIONS OF HEAD AND GRINDING 
TEETH 

A consideration of the proportions of the head and 
the grinding teeth affords a demonstration of the 
principle that the evolution of the 11 principal char- 
acters of the titanothere skull and dentition is corre- 
lated with pervading bjachycephaly or dolichocephaly. 
In the brachycephalic Oligocene titanotheres there is 
an excess of the transverse over the anteroposterior 



verse diameters of all the teeth, a fact which may be 
due in part to geologic crushing of the rock matrix 
and in part to actual variability in proportions. 

2. There seems to be a higher degree of fluctuation 
in the diameters of individual teeth than in the total 
anteroposterior measurements of p'-m^ which may 
indicate compensatory growth. 

3. Apparently each species has a typical mean for- 
mula and curve for the excess of the transverse over 



796 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



the anteroposterior measurements in each of the grind- 
ing teeth, p'-m^. 

4. The correlation of the proportions of the indi- 
vidual tooth respectively with the proportions of the 



molars as compared with the contemporary horses, in 
which the premolars are rapidly molarized; (4) rare 
development of mesostyle (median ridge of outer wall) 
of premolar; (5) feeble development of entoconid of 






Figure 716. — "Brain easts" (intracranial casts) of titanotheres 

A, Palaeosyops leidyi, Am. Mus. 1544, left side view; B, Mesatirhinus pctersoni, Princeton Mus. 
10041, left side view; C, MesatirhiTius -peicrsoni, Am. Mus. 1509, right side view; Di, Menodus 
giganteus i=" BrontotheTtum ingejis" of Marsh), left side view, from a cast made under the direction 
of Professor Marsh; D2, the same, top view. All one-third natural size. In Palaeosyops (A) we see 
the very large olfactory lobes and the small cerebrum. The brain of Mesatirhinus (B, C) appears 
to be shorter anteroposteriorly. 

brachycephalic or dolichocephalic skull is well shown 
in the curves (fig. 719), the contrast between the doli- 
chocephalic and the brachycephalic forms of grinding 
teeth being very clear, as well as the progressive 
broadening of the grinders with the progressive bra- 
chycephaly of the Brontops hrachycephdlus and the 

B. rohustus phylum. 

5. The Brontops and Menodus curves of the short- 
horned titanotheres are of the same general type, con- 
trasting with those of the Brontotheriinae. 

6. In the Brontotheriinae p*-m' are typically broad; 
in Menodus they are typically narrow. 

Although more numerous measurements are needed 
to make these results absolutely precise or decisive, 
the general conclusion is sustained that each tooth in 
each genus and species has its distinctive average and 
typical proportions. 

ARRESTED EVOIUTION IN THE TEETH 

The teeth of the titanotheres show the following 
instances of arrested or retrogressive evolution: (1) 
Ketrogressive evolution of the incisors; (2) retro- 
gressive evolution of the canines in many series; 
(3) imperfect or retarded molarization of the pre- 




FiGUBE 717. — Evolution of the upper molar 
tooth in titanotheres 

Posterior view of a series of left upper molars, showing the gradual 
deepening of the ectoloph, the persistently low, conic form of 
the protocone and hypocone, and the loss of the conules. A, 
Eoiiianops borealis, lower Eocene; B, Limnohyops priscus, 
middle Eocene; C, Manleoceras washakiemis upper middle 
Eocene; T>, TeJmaihenum uUimum , upper Eocene; E, Menodus 
giganteus, lower Oligocene. 

the premolars; (6) hypsodontism of the molars con- 
fined to the ectoloph, or outer wall of the crown, a 
mechanically imperfect evolution; (7) elongation of 
the molar crowns prevented by the position of the 
orbits immediately above the grinding teeth. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



797 



FORAMINA (NERVE, ARTERIAL) OF THE ATLAS VERTE- 
BRA IN THE TITANOTHERES AND OTHER PERISSO- 
DACTYLS 

Primitive. — The primitive condition of ttie atlas 
(fig. 720, A) is as follows: (1) The first spinal nerve 
issues laterally, the inferior branch passing down 



cit., figs. 28, 29) the groove is bridged over. In the 
Oligocene genera the groove is open in certain forms 
(Brontops rohustus; see PL CXCV of this memoir), 
and bridged in others. This bridging is partly an age 
character. In all modern tapirs, horses, and rhinoc- 
eroses the atlantal groove is bridged over. 




-Pattern of an upper molar of five perissodactyls, showing the widely different extent of the 
enamel edges and surfaces 
A, Menodus gigaiiteus, an Oligocene titanothere; B, Palaeosyops Tohustus, a middle Eocene titanothere: C, Teleoceras, a Miocene rhinoceros; 
D, ElasmoilieTium, a very highly specialized Pleistocene rhinoceros; E, Equus, a modern horse. Scale uniform. In A and B the total length 
of the enamel edges is relatively short; in D and E it is excessively long. The more simple molar crowns, with short enamel edges, are adapted 
to crushing coarse vegetation; the more complex crowns, with greatly elongated enamel edges, are adapted to the trituration of hard siliceous 
shrubs and grasses. 



through the "atlantal groove" in front of the pleura- 
pophysis; (2) the same groove also surrounds a 
branch of the occipital artery; (3) the pleurapophysis 
is perforated for a short space on the posterior face 
and under side by the vertebrarterial canal. 

Secondary. — Progressive characters of the atlas 
in the Perissodactyla are: 

1. The first character is the bridging over of the 
atlantal notch or groove by the pleurapophysis, this 
resulting in a condition like that seen in Tapirus (fig. 
720, C). This bridging over of the atlantal groove 
is rare in Eocene titanotheres, although it is repre- 
sented in a supposed Palaeosyops atlas figured by 
Earle (1892.1, pi. 13, fig. 29; see bibliography, p. 698); 
is not present in Oligocene titanotheres. 

2. The loss of the vertebrarterial canal on the lower 
side of the pleurapophysis. It is a striking peculi- 
arity that some of the Oligocene titanotheres lose this 
canal entirely. In some cases this loss seems to be 
preceded by the constriction of the canal, as in the 
atlas of Telmatherium ultimum; in others the vertebral 
artery apparently evaded the canal entirely and passed 
directly from the axis beneath the pleurapophysis to 
its possible junction with the occipital artery (fig. 
720, B). 

Titanotheres. — The titanothere family characters of 
the atlas vertebra are as follows: 

1. Atlantal foramen open or bridged. The first 
cervical nerve and a branch of the occipital artery 
deeply indent or groove the anterior borders of the 
pleurapophysis; in certain forms (Manteoceras, Lim- 
nohyops, fPalaeosyops, DolicJiorhinus, see accompany- 
ing fig. 720) the groove is not bridged over by bone; 
in other forms {Palaeosyops as figured by Earle, op. 



2. The vertebrarterial canal pierces the inferior 
lamella of the pleurapophysis in the Eocene Bron- 
totheriidae as in the modern Tapiridae, whereas in 
the modern Equidae and Rhinocerotidae it enters 



/ Progressive brachycephaly U Contrast of dolichocephaly 1 
and brachycephaly \ 


-f-/0 


-2 














A 

B 
C 














D 

B 
C 

A 






















\ 










\ 












\ 


\ 










\ 






















1/ 


', \ 
























\\ 










j 1 




1 






// 




I'm 














M 








J 1 




1', 










1 / 




\ 












\i 








ji 












/ 






\'. 


^ 












\V 


/ 


/// 






^ 


1^ 




/ 








A 


-A 


/'/ 








'^^ 


\ 


/ 










f-' 














1 
























































\ 


^ 


pt p2 p3 p* ^/ rr^ /7j3 p/ pZ p3 p-f- rn' rrv' m^ 


A.3rOTjTops lracf>ycepha7us average A. Menodus yt^anteus 
S-^rontops dispar averaye .5. A^eyacerops Oucca 
C.JBronTaps roiustus average C. 3ronrotl,eri,.m curtum 
J^^JQronTotherium ^t^as 



Figure 719. — Evolution of distinctive generic and 
specific proportions in the grinding teeth of the Oli- 
gocene titanotheres 

I, Increasing excess of transverse over anteroposterior diameters and 
progressive brachycephalic tendency in the Brontops phylum; also 
illustrating the mean formulae for three species of Brontops. II, Con- 
trast between the extreme brachycephalic tendency in Brontotherium 
and Megacerops with the extreme dolichocephalic tendency in Menodus 
giganteus. In Brontotherium p<-m' are exceptionally broad; in Meno- 
dus p^-mi are exceptionally narrow. 

the superior face of the lamella. A distinctive pro- 
gressive feature in certain of the Brontotheriidae, 
first seen in the upper Uinta forms Telmatherium 
ultimum and Protitanotherium, is that the verte- 
brarterial canal tends to close or become obliterated 



798 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



entirely, causing the vertebral artery to pass beneath 
the pleurapophysis without perforating it. Thus in 
certain Oligocene titanotheres the vertebrarterial 
canal is reduced or wanting. 



By comparison with the tapirs, rhinoceroses, 
paleotheres, horses, and other perissodactyls we have 
found (1) that the locomotor skeleton of the titano- 
theres exhibits certain distinctive titanothere family 




Figure 720. — Atlas and axis of titanotheres and other perissodactyls, showing probable course of 

the arteries through atlas 

Based on comparison with the horse and rhinoceros, Gregory and Christman, 1920. One-fourth natural size. 

A, Manteoceras, Am. Mus. 12204. The transverse process of the atlas was pierced posteriorly by the vertebral artery, which prob- 

ably then united (below the transverse process) with the ascending trunk of the occipital artery, which in turn ran dorsally and 
passed through the atlanta Igroove and foramen, again turning sharply downward and finally turning forward and entering 
the cranial cavity along the floor of the spinal canal. The suboccipital nerve (N. cervicalis I) passed outward and downward 
through the atlantal foramen. These are the conditions in typical placental mammals, and the positions of the foramina and 
grooves of the atlas indicate that they were retained by Eocene titanotheres from Eotitanops onward. 

B, Brontoihenum leidyi, Carnegie Mus. 114. The vertebral artery did not pierce the transverse process of the atlas posteriorly 
but probably passed immediately below it. The transition from the primitive conditions shown in A apparently took place 
in Eothanoiherium osborni, of the upper Eocene, for Mr. Peterson records the fact that in the type atlas the base of the trans- 
verse process was pierced by a small foramen, whereas in the paratype of the same species there is no evidence of the foramen. 

C, Taphus americanus. The atlantal groove is bridged over by the anterior border of the transverse process. 

D, Opskeros bicornis, black rhinoceros. The vertebral artery enters the base of the transverse process on the upper surface, but 
the course of the tuimel is otherwise normal. The atlantal groove is bridged, as in C. 

E, Equus caballus. The vertebral artery pierces the transverse process on its upper surface and then joins the occipital artery. The 

atlantal groove is bridged. 



PROGRESSIVE ADAPTATIONS OF THE PECTORAL AND 
PELVIC ARCHES, THE LIMBS, AND THE FEET OF THE 
TITANOTHERES 

From the chief adaptations to speed and weight in 
the limbs of the larger ungulates (Chap. VIII, sec. 3) 
we may deduce from the bony limb structure of the 
titanotheres some knowledge of their relative size — 
that is, their height and weight combined — as well as 
of the speed and general mode of locomotion of the 
different Eocene and Oligocene forms. 



features; (2) that in its cursorial, mediportal, and 
graviportal adaptations it closely parallels the peris- 
sodactyls of other families, living and extinct; (3) that 
through divergence from each other in modes of 
feeding and locomotion there evolved a great variety 
of titanotheres, adapted to the varied feeding and 
soil conditions of the mountain-basin and plains 
regions. Our inquiry has been directed to three 
determinations — proportions of the feet and limbs; 
proportions of the head and body of the animal as a 
whole; weight of the animal as a whole. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



799 



ANCESTRAL TITANOTHERES SUBCURSORIAX 
The limbs of LamidotheTium are unquestionably of 



cursorial animals, and that the mediportal and 
graviportal forms were secondary, a generalization 



primitive cursorial type, and those of Eotitanops are 1 that is indicated also by the skull and dentition. 



'^ID^ 




D 




Figure 721. — Evolution of the pelvis in the titanotheres 

Four progressive stages of adaptation: A, Ttie subeursorial Eotitanops borealis, Am. Mus. 14887, lower Eocene; B, the mediportal Palaeosyops major. 
Am. Mus. 13116, middle Eocene (lower Bridger); C, the subgiaviportal Manteoceras manteoceras, Am. Mus. 2358, upper Eocene of Washakie 
Basin; D, the graviportal Brontotherium gigas, Am. Mus. 492, lower Oligocene. All one-eighth natural size. In the first stage, Eotitanops, the 
pelvis is long and narrow. 3<: in other primitive perissodactyls. In succeeding stages the gluteal blade of the ilium becomes transversely 
expanded, so that in the final stage the transverse diameter across the ilia greatly e-xceeds the anterosposterior diameter. All this constitutes 
an adaptation to progressively graviportal habits, as in the elephants. 



of subeursorial type, but these genera point alike to 
a more primitive cursorial ancestry. These facts 
appear to warrant the generalization that all the 
earlyjtitanotheres (see pp. 733, 736) were small, light. 



A significant fact, to which W. K. Gregory first 
called attention and which is commented on early in 
this chapter, is that of all the early Eocene peris- 
sodaotyls laiown, the oldest representatives of three 



800 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



families — the Eqiiidae, Helaletinae (Lophiodontidae), 
and Hyracodontidae — are cursorial or subcursorial. 
This prevalence of primitive cursorial types points 
to the hypothesis that the most primitive titanotheres, 
like the other Perissodactyla, the Artiodactyla, and 
the Condylarthra, were largely (if not wholly) cursorial 
or subcursorial animals, from which the larger medi- 
portal and graviportal titanotheres evolved. This hy- 
pothesis, as Gregory (1910.1) points out, is sustained 
by the carpal and tarsal structure in most of the early 
perissodactyls and by what is known of Lamhdothermm. 
In accordance with the principle of survival through 
natural selection, the early perissodactyls, being of 
diminutive size and without defensive weapons of 
any kind, would survive by adopting a relatively 
swift gait to escape their enemies. 

In the remote ancestral perissodactyl types, includ- 
ing the titanotheres, so far as known, the arches and 
limbs were slender, the manus and pes were relatively 
elongate, and the carpals and tarsals were high and 
narrow. 



The details of structure of the Jimbs of the lower Eo- 
cene LamhdotJierium and of Eotifanops are described on 

pages 590-598. 

THE TITANOTHERE PES 

The early titanotheres show the following char- 
acters in common with other early perissodactyls: 
(1) Tridactyl, no trace of D. I or D. V; (2) digits 
relatively elongate, laterally compressed; (3) astrag- 
alus with elongate neck and three entirely separate 
astragalocalcaneal facets (Osborn, 1890.51), the ectal 
(e), the vertically elongated "sustentaculars, " and the 
"inferior" (i), the last usually small and separate; 
(4) astragalocuboidal facet narrowing to a point 
anteriorly, whereas in the middle Eocene titanotheres 
it is progressively broad; (5) calcaneofibular facet, 
or pit, functional in titanotheres and equines. 

The proportions and facets of the pes of titano- 
theres must be interpreted by considering the common 
ancestry of the titanotheres with the other perisso- 
dactyls and the special submediportal divergence of the 
pes. Thus, in the evolution of the feet of the several 
forms of the perissodactyls the contrasts are as follows : 



Forms of the elements of the pes in three types oj perissodactyls 



Element 


Cursorial 


Submediportal 


Graviportal 


Digits 

Tarsal elements 

Sustentaculum astragali 


Elongated or stilted 

Narrow and elevated 

Contracted and elevated 


Relatively short, expanded 

Broader 

Spreading 


Broad and stout. 
Extremely broad. 
Spreading and continuous. 


Expanding 


Massive. 











Judging by these tests the gradations as to weight 
and speed among these lower Eocene titanotheres, as 
compared with contemporary perissodactyls, were as 
follows : 

Ancestral titanothere Eotitanops, larger, least cursorial. 

Cursorial titanothere Lambdotherium, smaller, transi- 
tional. 

Cursorial lophiodont Heptodon, more slender and 
cursorial. 

Cursorial horse Eohippus, most slender and cursorial. 

Thus all the elements of the pes of Eotitanops may 
be distinguished in all their facets and other parts as 
broader and more expanded laterally than those of 
Heptodon and Eohippus, which are compressed laterally 
and expanded anteroposteriorly. 

PROGRESSIVE STAGES OF THE MANUS 

A comparison of the evolution of the manus in the 
ascending series of titanotheres, as illustrated in 
Figures 722-723, shows a complete transition from 
the cursorial through the mediportal to the graviportal 
type, modified throughout by certain peculiar titano- 
there family tendencies, namely, (1) increase in 
size; (2) secondary enlargement of D. V, which 
finally nearly equals D. II in size, a titanothere 
peculiarity; (3) secondary substitution of paraxonic 



formesaxonic condition, a titanothere peculiarity; (4) 
flattening of all the small bones of the carpus; (5) ab- 
breviation of all the phalanges, especially of the ter- 
minal phalanges supporting the horn sheaths; (6) pro- 
gressive expansion of the lunar upon the magnum in 
adaptation to tetradactyly, to the paraxonic condi- 
tion, and to the progressive graviportal condition; 
(7) gradual atrophy of the trapezium, which becomes 
vestigial. 

THE TITANOTHERE MAGNUM 

The progressive stages in the evolution of a single 
bone, the magnum, demonstrate that each bone of the 
carpus passes through a complete succession of stages 
from the cursorial to the graviportal type; that each 
bone reflects what it is correlated and coadapted 
with, namely, the bodily history of the entire or- 
ganism; that each bone if examined with sufficient 
thoroughness and minuteness would reveal generic and 
probably specific characters capable of definition. 

The magnum, as the central bone of the carpus, in 
the course of adaptation from the cursorial condition 
in Eotitanops to the extreme graviportal condition in 
Brontotherium, changes its diameters completely, 
evolving from a high, narrow form (Eotitanops) into 
a broad, low, flattened form (Brontotherium). It has 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



801 



six facets, as follows: Magnum-trapezoid, magnum- 
scaphoid, magnum-lunar, magnum-unciform, mag- 
num-Mtc III, magnum-Mtc II. The most significant 
change is that in the passage from the vertical to the 



EVOIUTION OF THE ASTRAGALUS 

The astragalus of the tarsus displays similarly a 
complete transformation in its facets in passing from 
the cursorial through the mediportal to the^ gravi- 




FiGUHE 722.- 



g -w ^ c 

-Three stages in the evolution of the manus in titanotheres 



A, Subcursorial (Eotiianops); B, mediportal (Limnohyops): C, graviportal (Brontotherium). Not drawn to scale. In all three figures the 
height from the top of the lunar to the bottom of the third metacarpal is arbitrarily made equal. In this way the progressive relative 
widening of the metacarpals, the shortening of the phalanges, and the great widening of the carpals are revealed. 





Figure 723. — Six stages in the evolution of the manus in titanotheres 

Lambdotherium, lower Eocene; B, Eotitanops, lower Eocene: C, Limnohyops mortoconus, middle Eocene; D, Manteoceras manteoceras, middle Eocene; E, "Biplacodon," 
upper Eocene; E, Brontotherium, lower Oligocene. The earlier members of this series retain the more primitive perissodactyl cursorial pattern of the manus. Th& 
carpus is of the interlocking, displaced type, and the third digit is predominant. In the mediportal and graviportal members the whole manu? becomes very 
wide, digits III and IV become subequal, and the lunar and magnum broaden. 







Figure 724. — Evolution of the magnum in titanotheres 



Cursorial, mediportal, and graviportal types. This series shows the progressive widening of the magnum (the increase of the 
horizontal compared with the vertical diameter) in correlation with the widening of the whole foot, a graviportal adaptation. 
The facets for scaphoid, lunar, and Mtc III are greatly widened. A, Eotitanops borealis, Am. Mus. 296, lower Eocene; 
B, Mesatirliinus petersoni, Princeton Mus. 10013, middle Eocene; C, Manieoceras manteoceras, Am. Mus. 12204, middle 
Eocene; T>, Palaeosyops copei (?referred), Am. Mus. 12205, middle Eocene; E, Palacosyops roiusius, Am. Mus. 1581, middle 
Eocene; F, Dolichorhinus longiceps?, Carnegie Mus. 2865, upper Eocene; G, BronMlierium gigas, Am. Mus. M43, lower 
Oligocene. 



lateral expansion, the magnum passes beneath the 
lunar (as described above in this chapter) and the 
lunar also spreads over the upper surface of the 
magnum. Thus, as was first pointed out by Osborn 
(1890.51, sec. 4), the growth of the magnum and 
lunar is reciprocal. 



portal type, in the course of which it parallels other 
perissodactyls in their similar transformation, while 
it also displays certain peculiar titanothere family 
features. In the diagram of Figure 701 we observe 
a series of stages corresponding with those observed 
in the manus and in the magnum, namely, the trans- 



S02 



TITANOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



formation from a small, high, narrow type of astrag- 
alus, adapted to the cm-sorial foot of LamMoiherkim ^ 
through the broadening mediportal stages exhibited 
in Limnoliyops and Manteoceras, into the subgravi- 
portal Protitanotlierium and the massive graviportal 
Brontoiherium. 

The following five detailed changes are evolved, all 
■of which are more or less closely paralleled in other 
lines of ungulates in the passage from the cursorial 
to the graviportal condition: (1) A maximum vertical 
diameter (height and width index) is replaced by a 
maximum horizontal diameter (height and width 
index); (2) the neck, or cervix, is relatively abbrevi- 
ated; (3) the astragalocuboidal facet is greatly broad- 
ened, a process analogous to the extension of the 
lunar upon the magnum in the manus; (4) the ecto- 
sustentacular and inferior astragalocalcaneal facets 
are separated; (5) the astragalocalcaneal facet is 
vertically elongated in Lambdotherium and becomes 
rounded in Brontotherium. 

SECTION 4. BIBLIOGRAPHY FOR CHAPTERS VIII-X 

Allen, Joel Asaph. 

1876.1. The American bisons, living and extinct: Harvard 
Coll. Mus. Comp. Zoology Mem., vol. 4, No. 
10, pp. i-ix, 1-246, pis. 1-12, 1 map. 
Baum, H. 

1898.1 (with Ellenberger, W., and Dittrich, Hermann). 
See Ellenberger, W., 1898.1. 
JBeddard, Frank E. 

1889.1 (and Treves, Frederick). On the anatomy of 
Rhinoceros sumatrensis: Zool. Soc. London 
Proc, 1889, pp. 7-25. 
X)e Blainville, H. M. D. 

1816.1. Prodrome d'une nouvelle distribution syst^mati- 
que du regne animal: Soc. philomath. Paris 
Bull., 1816, pp. 105-124; Jour. phys. chimie, 
hist, nat., arts, vol. 83, pp. 244-267. 
1839.1. Osteographie ou Description ioonographique com- 
paree du squelette et du systeme dentaire des 
mammiferes recentes et fossiles * * * ou- 
vrage accompagne de 323 planches lithographiees 
sous sa direction par M. J. C. Werner, text 4 
vols., plates 4 vols., Paris, J. B. Bailliere et 
Fils, 1839-1864. 
Boas, J. B. V. 

1908.1 (and PauUi, Simop). The elephant's head, studies 
in the comparative anatomy of the organs of 
the head of the Indian elephant and other 
mammals: Pt. 1, The facial muscles and the 
proboscis, pp. 1-79 (two columns to the page, 
each numbered), pis. 1-17, Copenhagen, pub- 
lished at the cost of the Carlsberg Fund. 
BoRELLi, Giovanni Alfonso. 

1680.1. De motu animalium, edited by Carolus Joannes a 
Jesu, 2 parts, 1680, 1681. 
Bruce, Adam Todd. 

1883.1. Observations upon the brain casts of Tertiary 
mammals: Princeton Coll. E. M. Mus. Geol- 
ogy and Archeology Contr., No. 3, pp. 36-45, 
Ipl. 



BURMEISTER, C. H. C. 

1867.1. Fauna argentina, Pt. 1, Mamlferos f(5siles: 
Buenos Aires Mus. publico Anales, vol. 1, pp. 
87-300, pis. 9-14. (Imparidigitata, p. 235.) 
Cope, Edward Drinker. 

1887.1. The Perissodactyla: Am. Naturalist, vol. 21, pp. 

985-1007, 1060-1076. 
1889.3. Tlie mechanical causes of the development of the 
hard parts of the Mammalia: Jour. Morphology, 
vol. 3, pp. 137-277, pis. 9-19. 
Cunningham, D. J. 

1903.1. Text-book of anatomy, xxix, 1309 pp., New York, 
William Wood & Co. 
Cuvier, Baron Georges Leopold Chretien Frederic Dago- 

BERT. 

1850.1. Anatomie compar(5e, recueil de planches de myolo- 
gie dessin^es par G. Cuvier, ou ex6cutfe sous 
ses yeux par M. Laurillard, public * * * 
sous la direction de MM. Laurillard et Mercier, 
2 vols., Paris. 
Dbperet, Charles. 

1903.1. Etudes pal^ontologiques sur les Lop/iiodon du Min- 
ervois, structure du crane, des membres, et 
affinites g^nerales des Lophiodon: Lyon Mus. 
hist, arch., vol. 9, pp. 1-45, pis. 1-4. 
Dittrich, Hermann. 

1898.1 (with Ellenberger, W., and Baum, H.). See 
Ellenberger, W., 1898.1. 
Dubois, Eugen. 

1891.1. Ueber die Abhangigkeit des Hirngewichtes von 
Korpergrosse bei den Saugethieren: Arohiv 
Anthropologic, Band 25 pp. 1-32. 
Durst, J. Ulrich. 

1902.1. Versuoh einer Entwicklungsgeschicte der Horner 

der Cavicornia nach Untersuchungen am 
Hausrinde; Sonderabdruck, Forschungen auf 
dem Gebiete der Landwirtschaft (Festschrift 
zur Feier des 70. Geburtstages von Prof. Dr. 
Ad. Kraemer), pp. 1-47, Frauenfield, Huber 
& Co. 

1902.2. Sur le developpement des cornes chez les cavi- 

cornes: Paris Mus. hist. nat. Bull., vol. 8, No. 

3, pp. 197-203. 
1903.1. Les lois mecaniques dans le developpement du 

crane des cavicornes: Acad. sci. France Compt. 

Rend., vol. 137, pp. 342-344. 
Ellenberger, W. 

1898.1 (with Baum, H., and Dittrich, Hermann). Hand- 

buch der Anatomie der Tiere fiir Ktinstler, Lief. 

1, Rind, 1898; Lief. 2, Pferd, 1898; Lief. 3, 

Rind, 1899; Lief. 4, Pferd, 1900; Lief. 5, Pferd, 

1901; Leipzig, Theodore Weicher. 
Gadow, Hans. 

1902.1. The evolution of horns and antlers: Nature, vol. 

65, p. 549, April 10, 1902. 
Gaudry, Albert. 

1862.1. Animaux fossiles et g^ologie de lAttique d'apres 

les recherches faites en 1855-56 et en 1860 sous 

les auspices de I'Academie des sciences, te.xt and 

atlas, Paris, F. Savay, 1862-1867. 
1908.1. Fossiles de Patagonie — -les attitudes de quelques 

animaux: Annales de paleontologie, vol. 1, pp. 

1-42. 
1909.1. Fossiles de Patagonie — le Pyroiherium: Annales de 

paleontologie, vol. 4, pp. 1-28, pis. 1-7. 



THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



803 



GiDLEY, James Williams. 

1903.1. A new three-toed horse: Am. Mus. Nat. Hist. 
Bui)., vol. 13, pp. 465-476. 
Gregory, William King. 

1910.1. The orders of mammals: Am. Mus. Nat. Hist. 
Bull., vol. 27, pp. 1-524. 

1912.1. Notes on the principles of quadrupedal locomotion 
and on the mechanism of the limbs in hoofed 
animals: New York Acad. Sci. Annals, vol. 22, 
pp. 267-294, pi. 34. 

1920.1. Studies in comparative myology and osteology, 5, 
On the anatomy of the preorbital fossae of 
Equidae and other ungulates: Am. Mus. Nat. 
Hist. Bull., vol. 42, pp. 265-283, 1 pi., 25 figs. 
Haughton, Samuel. 

1864.1. Notes on animal mechanics, 16, On the muscular 
anatomy of the rhinoceros: Roy. Irish Acad. 
Proc, vol. 9, pp. 515-524. 
Hatcroft, J. B. 

1900.1. Animal mechanics: Textbook of physiology, E. A. 
Schafer, editor, vol. 2, pp. 228-273, Edinburgh 
and London, Young J. Pentland. 
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. 
Labtet, Edouard. 

1868.1. De quelques cas de progression organique v6rifi- 
ables dans la succession des temps g^ologiques 
sur des mammiferes de meme famille et de meme 
genre: Acad. sci. France Compt. Rend., vol. 
68, pp. 1119-1122. 

LUCIANI, LXTIGI. 

1905.1. Physiologie des Menschen, ins Deutsche iibertragen 
und bearbeitet von Dr. Silvestro Baglioni und 
Dr. Hans Winterstein, mit einer Einfiihrung 
von Dr. Max Verworn, Jena, Gustav Fischer, 
Band 1, 1905; Band 2, 1906; Band 3, 1907; 
Band 4, 1911. 

Marey, E. J. 

1874.1. Animal mechanism, a treatise on terrestrial and 
aerial locomotion: Internat. Sci. Ser., vol. 11, 
London, Henry S. King & Co. 

Marsh, Othnibl Charles. 

1884.1. Dinocerata, a monograph of an extinct order of 

gigantic mammals: U. S. Geol. Survey Mon. 
10, xviii, 237 pp., 56 pis. 
Matthew, William Diller. 

1897.2. A revision of the Puerco fauna: Am. Mus. Nat. 

Hist. Bull., vol. 9, pp. 259-323. 
1909.1. The Carnivora and Insectivora of the Bridger 

Basin, middle Eocene: Am. Mus. Nat. Hist. 

Mem., vol. 9, pt. 6, pp. 291-565, pis. 43-52. 
Mtjrie, J. W. 

1872.1. On the Malayan tapir, Rhinochoerus sumatranus 

(Gray) : Jour. Anat. Physiol., vol. 5, pp. 

131-169, pis. 8-10. 

MUYBRIDGE, EaDWEABD. 

1907.1. Animals in motion, an electrophotographic in- 
vestigation of consecutive phases of muscular 
actions, 3d impression, London, Chapman & 
Hall. 
101959— 29— VOL 2 8 



OsBORN, Henry Fairfield. 

1890.51. The Mammaha of the Uinta formation, pt. 3, 
The Perissodactyla; pt. 4, The evolution of 
the ungulate foot: Am. Philos. Soc. Trans., 
vol. 16, pt. 3, pp. 541-569, pis. 7-9. 

1892.67 (and Wortman, J. L.). Fossil mammals of the 
Wahsatch and Wind River beds, collection of 
1891: Am. Mus. Nat. Hist. Bull., vol. 4, pp. 81- 
147. 

1893.82. Rise of the Mammalia in North America: Am. 
Jour. Sci., 3d ser., vol. 46, pp. 379-392, 448-466, 
pis. 5-10. 

1894.89. A division of the eutherian mammals into Meso- 
placentalia and Cenoplacentalia: New York 
Acad. Sci. Trans., vol. 13, pp. 234^237. 

1895.105 (and Wortman, J. L.). Perissodaetyls of the lower 
Miocene White River beds: Am. Mus. Nat. 
Hist. Bull., vol. 7, pp. 343-375, pis. 8-11. 

1898.143. The e.xtinct rhinoceroses: Am. Mus. Nat. Hist. 
Mem., vol. 1, pt. 3, pp. 75-164, pis. 12a-20. 

1898.148. Evolution of the Amblypoda, pt. 1, Taligrada 
and Pantodonta: Am. Mus. Nat. Hist. Bull., 
vol. 10, pp. 169-218. 

1899.166. The origin of mammals (abstract of discussion 
before 4th International Congress of Geologists, 
Cambridge, 1898): Am. Jour. Sci., 4th ser., 
vol. 7, pp. 92-96. 

1900.181. The angulation of the limbs of Proboscidea, 
Dinocerata, and other quadrupeds in adapta- 
tion to weight: Am. NaturaUst, vol. 34, pp. 
89-94. 

1900.192. Phylogeny of the rhinoceroses of Europe, 

Rhinoceros contributions 5: Am. Mus. Nat. 
Hist. Bull., vol. 13, pp. 229-267. 

1900.193. Oxyaena and Patriofelis restudied as terrestrial 

creodonts: Am. Mus. Nat. Hist. BuU., vol. 
43, pp. 269-281. 

1908.318. New or little known titanotheres from the 
Eocene and OUgocene: Am. Mus. Nat. Hist. 
Bull., vol. 24, pp. 599-617. 

1913.398. Eomoropus, an American Eocene chaUcothere: 
Am. Mus. Nat. Hist. BuU., vol. 32, pp. 261-274. 
Owen, Richard. 

1848.1. Description of teeth and portions of Jaws of two 
anthracotheroid quadrupeds {Hyopotamus vectia- 
nus and H. bovinus) discovered by the Marchio- 
ness of Hastings in the Eocene deposits of 
the northwest coast of the Isle of Wight, with 
an attempt to develop Cuvier's idea of the 
classification of pachyderms by the number of 
the toes: Geol. Soc. London Quart. Jour., vol. 
4, pt. 1, pp. 103-141, pis. 7-8. (Perissodactyla 
proposed, p. 131.) 
Parsons, F. G. 

1897.1. See Windle, Bertram, C. A., 1897.1. 

1901.1. See Windle, Bertram, C. A., 1901.1. 

1903.1. See Windle, Bertram, C. A., 1903.1. 
Paulli, Simon. 

1908.1. See Boas, J. E. V., 1908.1. 

RfjTIMEYER, L. 

] 865. 1. Beitrage zu einer palaeontologischen Geschichte der 
Wiederkauer zunachst an Linn6's Genus Bos: 
Naturf. Gesell. Basel Verh., Band 4, Theil 4, 
Heft 2, pp. 299-354. 



804 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Rtdkr, John A. 

1S77.1. On the laws of digital reduction : Am. Naturalist, 
vol. 11, pp. 603-607. 
Schmaltz, Reinhold. 

1905.1. Atlas der Anatomie des Pferdes, Teil 1, Das Skelett 
des Rumpfes und der Gliedmassen, 1905; Teil 
2, Topograpliische Myologie, 1909; 2. Auflage, 
Berlin, Schoetz. 
Stillman, J. D. B. 

1882.1. The horse in motion, as shown by instantaneous 
photography, with a study on animal mechanics, 
etc., Boston, Osgood & Co. 
Treves, Sir Frederick. 

1889.1. See Beddard, Frank E., 1889.1. 
Turner, H. N., jr. 

1850.1. Contributions to the anatomy of the tapir: Zool. 
Soc. London Proc, 1850, pp. 102-106. 
Weber, Max. 

J897.1. Vorstudien iiber das Hirngewicht der Saugethiere, 
Festschrift zum 70. Geburtstage von Carl 
Gegenbaur, Band 3, pp. 103-123, Leipzig. 
1904.1. Die Saugetiere, Einfiihring in die Anatomie und 
Systematik der recenten und fossilen Mam- 
malia, 866 pp., Jena, Fischer. 



Weisse, Faneuil D. 

1899.1. Practical human anatomy, a working guide for 
students of medicine and a ready reference for 
surgeons and physicians, 456 pp., 222 pis.. New 
York, William Wood & Co. 
WiNDLE, Bertram C. A. 

1897.1 (and Parsons, F. G.). On the myology of the terres- 
trial Carnivora, pt. 1, Muscles of the head, 
neck, and fore limb: Zool. Soc. London Proc, 
1897, pp. 370-409. 

1901.1 (and Parsons, F. G.). On the muscles of the Ungu- 
lata, pt. 1, Muscles of the head, neck, and fore 
limb: Zool. Soc. London Proc, 1901, vol. 2, pp. 
656-704. 

1903.1 (and Parsons, F. G.). On the muscles of the Ungu- 
lata, pt. 2, Muscles of the hind limb and trunk: 
Zool. Soc. London Proc, 1903, vol. 2, pp. 261- 
298. 
WoRTMAN, Jacob Lawson. 

1892.1. See Osborn, Henry Fairfield, 1892.67. 

1895.1. See Osborn, Henry Fairfield, 1895.105. 



CHAPTER XI 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



In section 1 of this chapter two chief processes of 
evolution are traced in detail in the long history of 
the titanotheres — the evolution of new proportions 
(allometrons) and the rise and evolution of new char- 
acters (rectigradations). Both processes proceed along 
direct, continuous, orthogenetic lines, and both are of 
great biologic and sytematic significance. 

In section 2 the causes of the extinction of the 
titanotheres and other quadrupeds are considered in 
the light of Darwin's theory of natural selection. 



Biocharacters {single characters). — Throughout the 
research made for this monograph far more attention 
has been devoted to a study of the modes of evolution 
of "single characters," which we may term "biochar- 
acters," than to artificial groupings into "mutations," 
"species," and "genera." The titanotheres exhibit 
with extraordinary clearness the origin, evolution, 
survival, and elimination of biocharacters. Detailed 
observations extending over nineteen years are not 
without reward. The long history disclosed by hun- 



H 


■ 






1 




^■K^v^^H 


■ 


^^^^B^^I^^p 


^^^B 


^^^Bp^'*-''^^^^ 


[^Hp^^'" 




W' 




1 


H 


H| 


Ifei; ■■ ''^ 


1^,,; '\^ 


1- 


'"jM^^^^t 




1 


^M 


^m 






■ 


Im^^^M 


^^^m^ 


3 


H 


9 




^^^l^f^s 


1. 


OB* 

s ., ■ 


m^ti^iBM^cik'^^ 


J 



Figure 725. — Restorations of Eoiitanops borealis (right) and Broniotherium plalyceras (left): the earliest and 
the latest stage in the evolution of the titanotheres 



SECTION 1. MODES AND CAUSES OF THE ORIGIN 
AND EVOLUTION OF NEW ADAPTIVE CHARAC- 
TERS (RECTIGRADATIONS) AND NEW PROPOR- 
TIONS (ALLOMETRONS) IN OLD CHARACTERS 

DISTINCTIONS BETWEEN MODES AND CAUSES 
Evolution of the heredity germ. — A host of new char- 
acters and of new proportions in old characters separate 
Brontotherium from Eoiitanops, and all these characters 
are germinal, hereditary. The problem here consid- 
ered is. How and why did the heredity germ of Eoiitanops 
evolve into that of Brontotherium ? "Ho w ' ' is a question 
of the modes of evolution; this question we have in 
large part answered. "Why" is a question of the causes 
of evolution; this question we have answered only in 
small part. Let us examine four points of distinction 
between "modes" and "causes." 



dreds, even thousands, of titanothere biocharacters in 
the teeth and skeleton throws light upon three of the 
still unsolved great problems of the modes of evolution, 
namely : 

Modes of origin of new adaptive biocharacters = 
rectigradations. 

Modes of origin of new proportions in biocharac- 
ters = allometrons. 

Modes of survival and elimination of biocharacters 
= selection. 

Distinctions to he noted. — We must keep constantly 
in mind the distinction between the comparatively 
well understood modes of evolution, which follow cer- 
tain well-established principles that apply to all living 
beings, and the still mysterious causes of evolution, 
which follow principles yet to be discovered. The 

805 



806 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



so-called "laws of evolution" thus far discovered relate 
chiefly to modes and in but small degree to causes. 
Nevertheless the observation of the modes of evolu- 
tion exemplified in the titanotheres and other animals 
bears adversely or favorably upon one or another of 
the various theories that have been advanced as to 
the causes of evolution. 

The study of evolution may be divided into a 
study of facts more or less fully ascertained by com- 
parative studies of the invertebrates and vertebrates 
through the sciences of zoology, comparative anatomy, 
experimental zoology, genetics, invertebrate and verte- 
brate paleontology, and a consideration of hypotheses 
as to the causes of evolution, such as those presented 
by Buffon, Lamarck, Spencer, Darwin, Wallace, Cope, 
Weismann, De Vries, Osborn, Loeb, Morgan, and 
other naturalists, experimentalists, and biochemists. 




£otitanops yre^oryi 



^rontotheriUTrt platyceras 




Lower £x>cene 

Figure 726. 



ZSO.ooo generations 

soo.ooo years of 

^eoloqic 'time 

/.ower Olifocene 

Relations of the heredity germ of Eotitanops gregoryi to that of 
Brontotherium platyceras 
Animals are one-fiftietb natural size. 



SPECTJIATION AS TO THE CAUSES OF EVOLUTION 

Many vertebrate paleontologists have contributed 
to our knowledge of the modes of evolution of the 
extinct vertebrates, notably Cuvier, Owen, Huxley, 
Gaudry, Leidy, Cope, Marsh, Kowalevsky, Williston, 
Scott, Osborn, Deperet, Smith Woodward, Andrews, 
Dollo, Abel, Merriam, Matthew, and Gregory. Only 
a few — among them Buffon, Lamarck, St. Hilaire, 
Spencer, Cope, and Osborn — have attempted even to 
speculate upon the causes of evolution of the verte- 
brates. Lamarck, Spencer, and Cope are the principal 
contributors to the Lamarckian hypothesis of inherit- 
ance by the germ of the mechanical modifications of 
the body, and to the related hypotheses of kineto- 
genesis, of growth force, and of "bathmism, " all 
dependent upon the supposed inheritance by offspring 
of the influences of bodily characters acquired through 



the habits of the parents. Scott has pointed out the 
very important distinction observable in vertebrates 
between the principle underlying the "mutations" of 
Waagen and the "variations" and "variability" of 
Darwin. Osborn has developed a hypothesis for the 
principle of rectigradation as of potential homology, 
and has proposed a theory of tetraplasy in growth 
and development and of tetrakinesis in evolution, the 
latter based upon conceptions of energy. 

DISTINCTION BETWEEN INVISIBLE GERM EVOLUTION 
AND VISIBLE BODILY EVOLUTION IN THE TITANO- 
THERES 

In observing the modes of evolution, and still more 
in speculating upon the causes of evolution, it is 
necessary to keep in mind always the sharp distinction 
between the invisible evolution of the germ and the 
visible evolution of the body. The germ 
contains the heredity units of all the predis- 
positions and potentialities of body form and 
function, units figuratively known as 
"factors," "determiners," or "genes," 
which may control the development into 
normal visible body form only under the 
favorable influence of normal habit and en- 
vironment. The body of the titanothere, on 
the other hand, is the component of the four- 
fold (tetraplastic) influence of the animal's 
germ cells, of its habits, of its physical 
environment, and of its life environment. 

Germ evolution represents incessant 
changes in heredity; bodily development 
represents four factors — heredity, physical 
environment, life environment, and habit 
(ontogeny). With this distinction clearly 
in mind, if we contrast the beginning and 
the end of titanothere bodily evolution we 
observe that the heredity germ of Bron- 
totherium platyceras was very different 
from that of its remote ancestor Eotitanops 
lorealis. Perhaps 500,000 years of geologic time 
separated Eotitanops from BrontotJierium. Allowing 
one generation on the average for every two years, 
two hundred and fifty thousand generations of titano- 
theres separated Eotitanops horealis from Brontotherium 
platyceras. 

Since the adult Brontotherium differs very widely 
from the adult Eotitanops in all its visible biocharac- 
ters it certainly reflects a great number of new or 
transformed germinal predispositions. For example, 
say the germ of Eotitanops contains 300,000 predis- 
positions, that of Brontotherium contains perhaps 
500,000. How did the predispositions of these 
200,000 new or transformed biocharacters enter the 
heredity germ? Whatever the number, certainly 
the heredity germ of Brontotherium was in its predis- 
positions and potentialities vastly different, more 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



807 



complex, and more imposing than that of Eotitanops. 
How was the heredity germ thus enriched; how did all 
these new predispositions of character, new qualities, 
and new potentialities enter the germinal substance? 
We know from comparative embryology and heredity 
that there was a direct continuity between the heredity 
germ of Eotitanops and that of Brontotherium (fig. 
727); also that the germinal predispositions and 
potentialities were handed down continuously, and 
that while the germ flowed outward into the form and 
fimctions of the body there is little or no Lamarckian 
evidence that the modifications of the form and 
functions of the body flowed inward into the germ of 
the succeeding generation — that is, there is little 
evidence in titanothere evolution for the Lamarck- 
Cope-Spencer doctrine of kinetogenesis, or inherit- 
ance of bodily modifications. 



PEINCIPIES OF TETEAPIASY IN INDIVIDUAI DEVELOP- 
MENT; ONTOGENY 

One contrast between the causes of individual 
bodUy development (ontogeny) and those of racial 
(germinal) evolution (phylogeny) is that the former are 
observedly tetraplastic, the latter are hypothetically 
so — that is, the influences of the body and of the 
environment on the germ are as yet hypothetic. 

We may now speak briefly of the fundamental 
tetraplastic principles (Osborn, 1908.308) of indi- 
vidual development, which wifl be more fully explained 
farther on in this chapter (pp. 835-838). 

To distinguish clearly the fact that every visible 
titanothere is the component of four physico-chemico- 
mechanical influences and is continually subject to 
selection in competition with other titanotheres, it is 
necessary to keep in view the following scheme : 



Tetraplasy: Inseparable action, reaction, and interaction of Jour complexes of causes, together with natural 

selection, on development 

1. Titanothere germ evolution 

Genesis of new germinal predispositions and potentialities, giving rise to new 
biocharacters, to new changes of form and proportion, continuous and discontinuous, 
to the hastening or retarding of existing biocharacters in development and evolution, 
to new predispositions to correlation and compensation of biocharacters, which give 
rise to new germinal "mutations," "ascending mutations," " rectigradations, " 
" allometrons," etc. 

2. Ontogeny 



The development of the body (soma) of the individual titanothere; influence 
of habit, use, and disuse upon the germinal predispositions of function and structure; 
genesis of new somatic modifications, of new somatic characters, of somatic chabges 
of form and proportion, of somatic hastening and retarding of biocharacters, of 
somatic correlation and compensation of biocharacters that may give rise to allome- 
trons, "ontogenetic variations," "ontogenetic species," etc. 

3. Physical environment 

Geographic environment (earth, air, water, temperature, etc.). If highly 
favorable, environment intensifies and overdevelops certain germinal predisposi- 
tions, potentialities, tendencies of every biocharacter in the soma; if unfavorable, 
it dwarfs, arrests, inhibits, or actually suppresses them. Environmental infiuence 
on the appearance of new somatic biocharacters, on changing form and proportions 
of somatic biocharacters, on the hastening and retarding of the individual develop- 
ment of somatic biocharacters, on the correlation and compensation of somatic bio- 
characters, concluding in extreme cases in the formation of "environmental species," 
"climatic variations," "geographic species," "geographic variations," "geographic 
varieties," etc. 

4. Life environment 

Plant and animal complex — flora and fauna, biota surrounding the organisms. 
Influence of all the competing organisms on the predispositions and potentialities 
of the titanothere germ, chemico-physical influence of different kinds of food, me- 
chanical influence of food, and complex of the combined influence of food and of 
habit on ontogeny. Competition with titanotheres of the same and of other varieties 
in the search for food. 



5. "Natural selection" of Darwin; "survival 
of the fittest" of Spencer 

Effect of competition with members of 
the same and other species on all the 
functions and structures developed from 
the predispositions and potentialities of the 
^ germ as influenced by environment and 
ontogeny, including the direct action of 
selection on all biocharacters that have 
survival value, resulting in the selection of 
the germinal predispositions from which they 
spring, according to the germinal selection 
theory of Weissman, which includes, in 
part, the "organic selection" or "coinci- 
dent selection" of Osborn, Morgan, and 
Baldwin. 



808 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Thus four physico-chemical and mechanical factors 
interoperate with selection according to Osborn's prin- 
ciple of tetraplasy. Every stage in the development 
of each titanothere, from youth to old age, was thus 
tetraplastic ; every visible biocharacter was the re- 
sultant of these fourfold factors. Natural selection is 
not in the same sense a factor or a force; it is an 
arbiter between the individual titanotheres as molded 
by hereditary predisposition, by life habit, and by the 
physical and the life environment. These are the 
observed modes of titanothere development. The 
hypothetic action of these modes of evolution is con- 
sidered on pages 835-838 in the discussion of tetrakinesis. 



y/ie !/oi/n^ 
TTtanotheres 
of eyery 
stage 

inherit their 
btocharacters 
(rectigradational 
and proportional) 
from the 
germ cells 
of the parent ^™^ 
Titanotheres. 
The body 
of each 
Titanothere 



offshoot 
of the 
qerm cell 
of its 
parent 




Brontot/ierium /eidyi Lower Olk^ocene 

Qenerations o/ Titanotheres 
between Cocene and Oiiyocene 




Cotitanops qreoorui 



IDDLE LOCENE 



-OCENE 



FiGUKE 727. — Continuity of tlie heredity germ from Eotitanops to 
Brontops, according to Weismann's theory of the "continuity of the 
germ plasm" 



PRINCIPLES OF SINGLE CHARACTER (BIOCHARACTER) 
EVOLUTION 

In the titanotheres, as in other animals, the body 
is composed of a large number of correlated single ele- 
ments, characters, or biocharacters, each of which has 
its more or less independent origin, development, 
transformation, rise, or decline. These biocharacters, 
which are separable in individual development and in 
evolution, are separable also in heredity, a fact known 
by comparing them with similar biocharacters in the 
hybridization of living animals like the horse, a 
perissodactyl remotely related to the titanotheres. 



A biocharacter (Osborn, 1917.460) is any single part 
or function of an animal which is known to have sepa- 
rate origin, growth, evolution, ontogenetic or individual 
rate (velocity) of development, phylogenetic or racial 
rate (velocity) of evolution, and a presumable separa- 
bility in the heredity of the germ. A biocharacter 
partly corresponds to the "unit character" of Men- 
delisra. 

Examples of biocharacters of the titanotheres are 
the single cusps of the grinding teeth, the single horns 
on the skull, the old and new proportions of the skull 
and of the limb bones. The term is thus very elastic, 
because it refers to larger and smaller character groups 
which evolve and are heritable together more 
or less as units. The term is both bodily 
(somatic) and germinal in application, for each 
visible biocharacter may have a number of 
genes or determiners in the heredity germ. 

Four chief kinds of biocharacters in the teeth 
and skeleton are common to the titanotheres 
and other mammals of the paleontologic series, 
some of which correspond with the heritable 
variations and fluctuations of Darwin, as 
shown below. 

1. Saltation biocharacters. — Numerical saltations; sud- 
denly appearing complete additional parts; multiplica- 
tion of parts, rarely adaptive (such as supernumerary 
vertebrae, ribs, digits, teeth), duplication of existing 
parts (such as duplication of dental cusps, duplication 
of the tips of the horns) . Similar to certain numerical 
and "meristic" variations in other animals (such as 
hyperdactyl}', hyperphalangy). These suddenly appear- 
ing new characters are complete structures; they are 
all germinal in origin and are usually abnormal. 

2. Quantitative variations, fluctuations. — Variations 
(Darwin), chance variations, minor saltations, minor 
discontinuities (Bateson), minor plus and minus vari- 
ations; the quantitative and intensive fluctuation of 
characters around a mean (Quetelet), in a single curve, 
or in a double curve where two separate evolutionary 
phyla are hybridized. These minute accidental char- 
acters include in part the "mutations of De Vries, " 
which take accidental, chance, or variable directions, 
as distinguished from the rectigradations of Osborn. 
All are germinal in origin and are supposed by Darwin 
to form the chief material out of which selection builds 
up adaptations through gradual accumulation in one 
direction. 

3. Rectigradation biocharacters. — Rectigradations of 
Osborn; new characters developing adaptively, ortho- 

genetically, from the beginning, arising from minute and 
inconspicuous rudiments (such as new units of structure, 
hornlets, dental cusplets, new dental folds and plications). 
These characters are also germinal in origin, but, unlike salta- 
tions, they develop orthogenetically and very graduall}' and 
never appear at first as complete structures. They are invari- 
ably normal. 

4. Proportion biocharacters. — Allometrons of Osborn, quan- 
titative biocharacters, plus and minus proportions, skull 
proportions (such as dolichocephaly and brachycephaly) , limb 
proportions (such as dolichomely and brachymely, adaptive or 
neutral). These are all proportional, quantitative, in existing 
units of structure. Unlike saltations and rectigradations they 
may first appear in ontogeny, in the body, and subsequently in 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHEEES 



809 



the germ. Whatever their cause or origin they become true 
germinal biocharacters, separable in heredity and developing 
orthogeneticaUy. 

All these four kinds of biocharacters are observed in 
many complete series of fossils. In the titanotheres 
the rise of rectigradation and proportion biocharacters 
constitutes dominant modes of evolution. The salta- 
tions seem to be abnormalities or defects in the germ 
and do not appear to play any part in the modes of 
evolution of the titanotheres. Theoretically (Darwin) 
the minor fluctuations of proportion may afford the 
material out of which certain new adaptive propor- 
tional characters are evolving. 

The six problems in biocharacter evolution in the 
body and germ are as follows: 

1. Genesis, initiation of new biocharacters: Do they arise 
suddenly or continuously? 

2. New proportions: Do proportions arise suddenly or grad- 
ually? 

3. Correlation: How are biocharacters correlated? 

4. Velocity: How are the motions of biocharacters regulated? 

5. Heritage: What is the behavior of mingled biocharacters? 

6. Genesis: Do new biocharacters first appear in the body or 
in the potentiality of the germ? 

SEPARABILITY AND COORDINATION OF BIOCHARACTERS 

Although in certain principles of form (morphology) 
and function (physiology) as well as of inheritance 
(heredity) biocharacters are separable, in other 
aspects they are closely coordinated and correlated 
with other biocharacters so as to play their part in the 
organism. Consequently biocharacters in course of 
development display a great number of different and 
invariably correlated modes of evolution. Five prin- 
ciples of the evolution of biocharacters are set forth 
below. 

I. Origin, genesis 

1. Saltations, major. 

2. Saltations, minor; mutations of De Vries. 

3. Continuous genesis of rectigradations. 

4. Sudden genesis of new proportions. 

II. Proportion, genesis 

1. Fluctuation, plus and minus variation. 

2. Continuous evolution of proportion. 

3. Influence of habit. 

4. Influence of internal secretions. 

III. Cooperation, coordination 

1. Compensation in development and degeneration. 

2. Grouping or correlation of function. 

3. Mechanical interrelation of separate parts. 

4. Linkage of biocharacters with male or female sex, respec- 
tively. 

5. Coordination by internal secretions. 

IV. Motion, velocity 

1. Acceleration in ontogeny. 

2. Retardation in ontogeny. 

3. Acceleration in phylogeny. 



4. Retardation in phylogeny. 

5. Arrest of development in phylogeny. 

6. Acceleration or retardation by internal secretions. 

V. Heritage 

1. Biocharacter heritage, separabiUty, segregation. 

2. Biocharacter heritage blending, intermingling forms of 
biocharacters from two separate parents or strains. 

3. Biocharacter heritage, "specificity" in strict repetition 
of parental or racial character. 

4. Separability and blending of proportions. 

INITIATION OF BIOCHARACTERS: IS IT ENVIRONMENTAL, 
ONTOGENETIC, OR GERMINAL? 

Modes of initiation. — The observed time and place 
of the visible initiation or genesis of each new bio- 
character, whether environmental, somatic, or ger- 
minal, is a matter of first importance. There are 
contrasts — for example, the visible genesis of each 
cusp rectigradation is at birth; the visible genesis of 
each horn rectigradation is in the adult. Does this 
mean that the new cusps and new horns have different 
causes? 

In the observation of each biocharacter it is desirable 
to make the following notes as to its initiation. 

Initiation in origin: Is the first appearance germinal 
or somatic? For example, all rectigradations in the 
teeth so far as observed make their first appearance 
in the germ. 

Initiation in motion (acceleration, retardation, 
balance, arrest): Is it germinal or somatic? For 
example, does the initiation in velocity make its first 
appearance in the germ or in the body? In most 
animals the acceleration of characters in ontogeny 
and phylogeny is purely germinal, as in the limb 
proportions of praecoces (birds and quadrupeds that 
are capable of rapid motion at birth). 

Initiation in form evolution (rectigradations, allo- 
metrons): Is it germinal or somatic? For example, 
although all known rectigradations are congenital, 
many allometrons may arise first in the soma. 

Initiation in cooperation, coordination, correla- 
tion, compensation, sex linkage, etc. : Is this germinal 
or somatic? Besides the congenital correlation of 
form and function many kinds of coordination arise 
ontogenetically. 

Germinal initiation. — Close observation of the time 
of the initiation of each biocharacter, whether obvi- 
ously environmental in its first appearance or obviously 
ontogenetic, is very important. Certain proportion 
characters, such as the proportions of the skuU, ap- 
parently arise in the germ; other proportion characters 
may first appear somatically through the use and 
disuse of parts, as for example in the limbs. We are 
absolutely siu-e that the initiation of many biochar- 
acters is germinal, whatever may be their causes. 
This is particularly true of the rectigradation cusps 
of the grinding teeth. Again, the rudiments (recti- 



810 



TIT,\NOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



gradations) of new cusps on the grinding teeth are 
developed in the jaw long before birth; obyiously 
their first appearance is germinal. In apparent but 
not actual contrast, the rudiments (rectigradations) 
of horny outgrowths upon the skull first appear in 
late ontogenetic stages; after a long period of geologic 
evolution these appear at birth or before birth as 
germinal biocharacters. This leaves open the ques- 



jBirth of net^ characters, first appearance 
arly middle or iate stages of deyelopTnen.t 
'.s rudiments , oro^ressiye and prophetic strucTi 
Death of old characters, vestiges 
retrogressive and reversional striutures 





The death plact 



Figure 728.- 



horns first appear as rudiments in adult stages of 
individual development, but after a very long geologic 
interval they are thrust forward into prenatal stages, 
so that the young titanotheres have horns at birth. 
Such characters are said to be accelerated. Other 
characters are ontogenetically retarded: they lose their 
ontogenetic velocity. For example, the lateral digits 
of the horse are so slowed down in rate of develop- 
ment that finally they are confined to the 
germinal stage, in which we find a vestige of 
one terminal phalanx (Ewart, 1894.1), which 
is never present in the young (fig. 730, Bi). 
This gradual acceleration or retardation of 
biocharacters furnishes one of the most strik- 
ing examples of continued evolution m adap- 
tation; one "genus" or "species" slips gradu- 
ally into another (for example, MerycMppus 
into Pliohifpus) not through the sudden or 
abrupt falling out of a character but through 
the adaptation,, acceleration, or retardation of 
certain similarly functional biocharacters. 

PJiylogenetic velocity of biocharacters. — Phy- 
logenetic velocity is a quite different property 
of biocharacters from that described above. 



-Germinal origin and disappearance of characters 

Showing that the germinal cycle of certain biocharacters, such as rectigradations and certain 
allometrons, first appears in the germ and that after a long period of acceleration and visible 
evolution in the soma these biocharacters may undergo retrogressive evolution or retardation 
and may be retained in the germ as latent or potential predispositions and reappear in the 
soma only occasionally as reversional structures. This kind of germinal velocity, or character J^ ^g g^ property discovered and demonstrated 

by Osborn in the researches made for this 
tion whether the initiation of the horns of the titano- 
there is somatic or germinal. Followers of Lamarck, 
Spencer, and Cope would contend that as the horns 
first appear in late stages of somatic development the 
initiation is somatic (ontogenetic) — that horns first 
appear as the results of growth, localized by habit. 
This question is discussed below under the heading 
"Horns arise as typical rectigradations." 

DIFFERING HEREDITARY VELOCITIES; RATES OF MOTION 
OF BIOCHARACTERS 

Definitions. — The term velocity applied to a bio- 
character refers to its separate rate of development 
in the body (ontogeny) as well as to its separate rate 
of appearance in evolution (phylogeny). These two 
kinds of hereditary velocity are quite distiact from 
each other and are both, in different ways, highly 
adaptive. The resemblances and contrasts between 
these two kinds of biocharacter velocity are shown in 
the following principles and examples. 

Ontogenetic velocity of biocharacters. — The movement 
of biocharacters has been demonstrated both in em- 
bryology (Agassiz; Haeckel, biogenetic law) and in 
paleontology (Hyatt, 1866.1, law of acceleration and 
retardation). The principle as developed by Hyatt 
and Cope (1887.1, 1896.1) in paleontology is that 
biocharacters which may originate at or near adult 
development of the individual may be inherited in 
successive generations in earlier and earlier stages of 
individual development until they exist only in the 
extremely young or are actually skipped as stages of 
development. In the titanotheres, for example, the 



monograph, namely, that certain similar biocharacters 
which are commonly regarded in comparative 
anatomy as "homologous,"^* "homogenous," 
"homorphous," really arise independently, although 
they appear at earlier stages of geologic time in 
certain phyla of the titanotheres than in others. 
For example, all the five or six branches of Eocene tita- 
notheres eventually exhibit horn rudiments, rudiments 
that evolve gradually into "homologous" horns. 
These horn rudiments are well developed in the adult 


^or/r stage 


Ontogenetic motion 




PAylogenetic motion | 


w 


7 


W 


M 


Z 


I 


'Phylum — * 

Horns \ 
A 1 

Premolar z 
-S 1 


I 


X 


M 


M 


Y 


r 


























* 






























' 






























* 










* 


Indwidual de\/elopment Vhyletic evoliiXion 

showing ho IV each character showing how each character has 
is accelerated or retarded a different velocity in each 
in the time of its appearance. phylum in which it appears. 


Figure 729.— Contrast between the ontogenetic and phylo- 
1 genetic velocity of a biocharacter 

In its ontogenetic velocity the horn rectigradation appears in all titanotheres in a 
late stage of individual development, which we may call stage VI. Through onto- 
genetic velocity it is thrust forward until it finally appears in stage I, before bhth 
In B the same biocharacters (that is, the horn rectigradations and cusp rectigrada- 
tions) exhibit a different phylogenetic velocity in each one of the sis phyla of 
titanotheres in which they independently arise. 

stages of certain phyla, whereas in other phyla the 
skulls are still hornless; in other words, each phylum 


" Homology 
character from 


in 
the 


the 
ger 


stri 
mo 


3test 
as 


sen 
mik 


se i 
ran 


3 homogeny— that 
tecedent ancestral c 


s, d 
bar 


irec 
cte 


de 


riva 


tion 


of a 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



811 



has its distinctive velocity of the horn biocharacter; 
and the same is true of the new cusps appearing on the 
teeth. 

Therefore when we compare similar biocharacters of 
every kind in two distinct phyla of the titanotheres we 
find that not only the horns but every biocharacter has 




Figure 730. — The limbs of four embryos, showing the 
ontogenetic velocity of the bones of the fore limb in 
horses 

A, 20 millimeters, to D, 350 millimeters. After Ewart, 1894 (PI. 11)^ 
rearranged. Observe the acceleration of proportions in the forearm 
and manus. 

its distinctive phyletic velocity. For example, close 
examination by Gregory of the actual percentage incre- 
ments of proportions in the evolving Palaeosyops skull 
demonstrates three principles of motion, namely, (1) 
that each cranial, dental, and skeletal biocharacter is 
evolving at a different rate, some faster, some slower; 
(2) that a biocharacter which early in geologic time 
evolves rapidly may afterward evolve more slowly, 
while the velocity of other continuous proportion bio- 
characters may be increasing; (3) that the movement 
of a biocharacter may be reversed, so that after a long 
period of phyletic velocity it may take an opposite 
direction. 

The following five principles of biocharacter velocity 
are well established in this monograph : 

1. All biocharacters are moving, each having its 
separate velocity; the movement is either progressive 
or retrogressive. 

2. The phyla of titanotheres are distinguished by 
relative rates of velocity of similar biocharacters in 
geologic time, also by the acceleration or retardation 
of similar biocharacters in ontogeny. 



3. This biocharacter velocity may be cumulative 
and run to an extreme — that is, to overdevelopment of 
certain organs and structures. 

4. Juxtaposed biocharacters may show entirely dis- 
tinct velocities. 

5. The retrogression and subsequent recession of a 
biocharacter may reverse the order of its appearance. 

Having discussed the interpretation of the various 
phenomena which are observed in the evolution of bio- 
characters in titanotheres we are ready to interpret 
the evolution of other series of American quadrupeds, 
such as horses, rhinoceroses, elephants, which are now 
under observation by Osborn. 

MODES OF BIOCHARACTER EVOIUTION ACTUAIIY 
OBSERVED 

MAMMALIAN PHYLA DISTINGUISHED BY DIFFERENT 
ONTOGENETIC AND PHYLOGENETIC VELOCITIES IN 
THE SAME BIOCHARACTERS 

Members of different phyla may possess exactly the 
same biocharacters, but each of these biocharacters 
may exhibit a different rate of phylogenetic progress. 
It is by these contrasts in the rate of evolution of a 
large number of similar biocharacters that the phyla 
of titanotheres and of other mammals may be clearly 
distinguished. In 1902 the fact was established that 
throughout the whole geologic period when the 
Titanotherium-he&Tmg beds (Chadron formation) were 
being deposited as many as eight different phyla of 



'PAenomena of Ontogeny 


Tfetarded 


Characters 


Accelerated \ 


5 


4 


3 


2 


/ 




/ 


2 


3 


■4- 


5 












A 






















iB 






















C 


— 


u>. 












-<- 




— 


— D 






















T 










_^ 












-<_ 


— r 























G 

























, 




— 










-^- 




— 









.7" 


— 




— 








Degenerating 
Petrogressive 


Baianced 
/Arrested 
Persistent 


il>eyeioping 
Progressit^e 



Figure 731. — Relative velocities (acceleration, 
balance, or retardation) of a series of biochar- 
acters (A- J) 

Some are progressive, developing rapidly; others, though 
adjacent, are balanced or are retrogressive and degenerate. 

titanotheres were independently evolving in the same 
general region. These eight phyla correspond more 
or less closely to the several "genera" originally pro- 
posed by Cope and Marsh. 

Study of these separate phyla by the author with 
the cooperation of Dr. W. K. Gregory was directed to 
show what each of these phyla was like at the begin- 
ning of the deposition of the Tiiano^Aermm-bearing 



512 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



beds and what each was like at the end, and thus to 
compare the beginning and the end of each phylum in 
respect to its several biocharacters. Measurements 
were made of the elements shown in the accompany- 
ing table, including (1) length of skull, (2) breadth of 
skull, (3) cephalic index, (4) length of superior grind- 
ing teeth, (5) length of horns, and each titano there 



phylum is seen to have its absolute increase, rep- 
resented by the plus sign ( + ). In a sixth part 
measured, the length of the nasals, there is an absolute 
decrease, represented by the minus sign ( — ). It is 
seen that not only are the absolute increments and 
decrements different in each phylum but that the 
relative increments and decrements are distinctive. 



Maximum increase or decrease {in males) of certain elements in -phyla distinguished hy di^erent rates in the genesis 
and transformation of biocharacters during the deposition of the Titanotherium ieds 

[All figures approximate] 



Menodontine group Cshort-horned) 



Mesaticephalic to 
brachycephalic 



Brontops Diploclonus 



Mesaticephalic to 
dolicbocephalic 



Brontotheriine group (long- 
horned) 



Mesaticephalic to 
brachycephalic 



Megacerops Brontotherium 



Basilar length of skull: 

Percentage 

MilUmeters 

Zygomatic breadth of skull: 

Percentage 

Millimeters 

Primitive and progressive cephalic index (limits) 

Superior grinding teeth (premolar-molar series) : 

Percentage 

Millimeters 

Free length of horns : ' 

Percentage 

Millimeters 

Free length of nasals: 

Percentage 

Millimeters 



+ 31 

+ 185 

+ 39 

+ 187 
72-87 

+ 34 
+ 96 

+ 143 
+ 122 



(?) 
(?) 

(?) 
(?) 
85-91 

(?) 
(?) 

(?) 
(?) 

(?) 
(?) 



+ 17 
+ 110 

+ 67 
+ 230 
67-76 

+ 30 

+ 85 

+ 240 
+ 140 

-31 
-33 



+ 36 
+ 222 

+ 15 

+ 73 

62-70,<' 79 

+ 75 
+ 200 

+ 300 
+ 220 

+ 33 
+ 43 



(?) 
(?) 

(?) 



(?) to 84 

(?) 
(?) 

(?) 
(?) 

(?) 
(?) 



+ 32 
+ 215 

+ 61 

+ 270 

66 to 80-87 

+ 25 

+ 75 

+ 260 
+ 283 



-66 
-76 



DISTINCTION BETWEEN RECTIGEADATION BIOCHARAC- 
TERS AND ALLOMETRON BIOCHARACTERS 

Apparently two kinds of biocharacters arise in the 
hard parts of titanotheres and other mammals. The 
difference is observed in their respective modes of 
origin and evolution. Theoretically we may attribute 
them to different complexes of causes. 

"Eectigradation"55 (Osborn, 1907.301, p. 228) is a 
designation for the earliest discernible stage of an 
absolutely new adaptive character of an orthogenetic 
kind. Such rudiments of new characters were first 
termed (Osborn, 1891.53) "definite variations" to 
distinguish them from the theoretic "indefinite" or 
"chance" variations of Darwin. When the shadowy 
rudiment of a new dental cusp or of a new horn first 
appears it is termed a rectigradation. It marks a 
numerical change, the addition of the rudiment of a 
new biocharacter that was not previously present, 
which when observed in successive generations is 
found to develop into an important adaptive character. 

Quite distinct is an " allometron" ^^ (Osborn, 1912.372, 
p. 250). It is not a numerically new character but a 

" Rectigradation>rec('M, straight; gradus, step. 

M Allometron>dXXoIos, different, changed; liirpoy, the contents or thing meas- 
ured; hence alloiometron or allometron, a changed quantity or measurement. 



' Height above narial opening. 

new proportion ^' in an old character, which may be 
expressed in differences of measurement, in ratios and 
indices. This evolution of new proportions in existing 
characters is a quantitative change; although a highly 
adaptive and important process it adds no new unit 
to the organism. These changes of proportion in the 
titanotheres were investigated very carefully by Osborn 
and Gregory between 1902 and 1916. In the skull 
there arise such contrasting proportions as brachy- 
cephaly and dolichocephaly; in the feet, brachypody 
and dolichopody. We learn from breeding that both 
kinds of change, rectigradations and allometrons, are 
germinal or become so; both become separable in 
heredity. 

As certain differences of opinion and interpreta- 
tion as to the distinctness of rectigradations and 
allometrons have arisen between three observers (H. 
F. Osborn, W. D. Matthew, W. K. Gregory) it is 
important to state very clearly some of the resem- 
blances and differences between these two kinds of 
biocharacters. 

s' Proportion >j3ro, tor; portia, share, part; the relation ot one thing to another 
in respect to size, quantity, magnitude of corresponding parts, capacity, or 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



813 



RESEMBLANCES BETWEEN KECTIGRADATIONS AND 
ALLOMETRONS AS BIOCHARACTERS 

Rectigradations and allometrons have certain points 
of resemblance: (1) Both are orthogenetic, continuous, 
or gradual in origin, rather than discontinuous, 
saltatory, or tychogenetic (tLxv, chance, fortuity; 
ykveai-s, genesis); (2) both appear in increasing inten- 
sity in successive "ascending mutations" of the same 
phyla, a new dental cusp, for example, becoming more 
distinct, the brachycephaly becoming more pronounced; 
(3) both are subject to similar laws of cellular growth, 
the rectigradations more localized, the allometrons 
more diffused; (i) both display differences in phylo- 
genetic velocity; similar rectigradations and allo- 
metrons arise in different geologic stages; (5) both 



display differences in ontogenetic velocity; horn 
rudiments and skull proportions, for example, may be 
accelerated or retarded in development; (6) both are 
subject to the law of correlation, compensation, and 
coordination with other parts; (7) both may or may 
not be correlated with sex; (8) after rectigradations 
such as rudiments of horns have first appeared, they 
are subject to changes of proportion — that is, to 
allometrons. 

CONTRASTS BETWEEN RECTIGRADATIONS AND 
ALLOMETRONS AS BIOCHARACTERS 

The contrasts between the features of rectigrada- 
tions and allometrons are shown in the accompanying 
table. 



Contrasts between rectigradations and allometrons 



Features of rectigradations 



Features of allometrons 



1. Rectigradations are strictly new biocharacters — local out- 

growths — such as dental cuspules or horn rudiments. 

2. Rectigradations give rise to so-called "homologies," more 

strictly "homomorphs"; the mesostyle of Mesohippus and 
that of Palaeosyops are said to be homologous, although 
one animal is not descended from the other, but both are 
descended from a very remote common ancestor that had 
no mesostyle. 

3. Rectigradations are neomorphs, new outgrowths, numeri- 

cally new biocharacters — for example, a tooth having 
three cusps develops five or six cusps. 

4. Rectigradations appear chiefly in near or remote descendants 

of common ancestors; the mammals of the order Peris- 
sodactyla, for example, evolve many similar rectigrada- 
tions in their grinding teeth which do not appear in the 
teeth of the Artiodactyla or the Amblj'poda. 

5. Rectigradations apparently have little or no observable 

adaptive value at their origin, little or no selection sur- 
vival value. 

6. The origin of similar rectigradations in independent phyla of 

mammals and families of mammals always gives rise to 
parallelism, convergence — for example, the molar teeth of 
Eohippus, a horse, converge toward those of Notharctus, a 
lemur, in the independent origin of similar cuspules. 

7. Rectigradations are not known to be produced experimentally 

in ontogeny. So far as observed they are from the begin- 
ning germinal biocharacters. 

8. Rectigradations as neomorphs are relatively rare or infre- 

quent in mammals; see Miller's definitions of the species 
of European mammals (1912.1). 

9. Rectigradations may be measured by indices and ratios only 

after they have developed to a sufficient size — for example, 
changes of proportion in horns and cu^ps (allometrons). 

10. Rectigradations characteristic of a phylum in one geologic 

period may appear in rudimentary form at a later period 
in another phylum as "homomorphs" — such as the 
enamel folds known as crista, crochet, antecrochet in 
horses, titanotheres, rhinoceroses. 



1. Allometrons are generally changes of old proportions not 

localized — for example, dolichocephaly, brachycephaly, 
hypsodonty. 

2. Allometrons give rise to analogies, never to homologies in 

any sense; dolichocephaly, for example, is an analogous 
change of the cranium in Homo and in Menodus. 



3. Allometrons are allomorphs, or heteromorphs, or changes of 

proportion — they are quantitative; a narrow tooth, for 
example, becomes a broad tooth. 

4. Allometrons are totally independent of remote ancestral 

affinity — for example, Perissodactyla, Artiodactyla, Am- 
blypoda may alike become dolichocephalic. They are 
partly independent of near ancestral afiinity; nearly re- 
lated races of men of the same species (Homo sapiens) 
become dolichocephahc or brachycephalic. 

5. Allometrons may have adaptive or mechanical significance 

throughout; they may consist of changes of limb propor- 
tion, which appear to be adaptable or inadaptable through- 
out their evolution. 

6. The origin of allometrons in animals of the same ancestry 

or of nearly allied strain may give rise either to conver- 
gence or to divergence; a dolichocephalic phylum, for 
example, may diverge from a brachycephalic phylum. 

7. Allometrons may be produced experimentally in ontogeny; 

new limb proportions, for example, may arise from new or 
modified habits. All allometrons are not certainly known 
to be germinal at the beginning. 

8. Allometrons are universal phenomena, constantly appearing 

in specific definition; see Miller, op. cit. 

9. Allometrons constantly afford changing indices and ratios in 

all stages of ontogeny and of phylogeny. 

10. Similar or dissimilar allometrons may arise in related phyla 

at different periods of geologic time. 



814 



TITANOTHERES OF ANCIENT "WYOMING, DAKOTA, AND NEBRASKA. 

Contrasts between rectigradations and allometrons — Continued 



Features of rectigradations 



11. Rectigradations appear to indicate a germinal predisposition 

or predetermination to vary in the same direction, these 
predispositions being most apparent in the more closely- 
related phyla and less apparent in remotely related phyla. 

12. The rise of similar rectigradations may confidently be 

predicted in geologic descendants of similar phyla — for 
example, similar and convergent rectigradations succes- 
sively develop in the premolar teeth of the entire order 
Perissodactyla. 



Features of allometrons 



11. Similar or entirely dissimilar allometrons may arise in 

closely related branches of the same stock; dolichocephaly 
and brachycephaly, for example, may arise in closely 
affiliated branches of Homo sapiens. 

12. Occurrence of similar allometrons can never be predicted; 

even within the family Brontotheriidae, even within the 
genus Telmaiherium, dissimilar and divergent allometrons 
may arise during the same geologic stage. 



Theoretic conclusions concerning predisposition to vary (Oshorn) 



Causes of rectigradations 



Causes of allometrons 



13. Rectigradations appear to be due in part to germinal potenti- 

alities as to definite or determinate variation which are 
found in the common ancestors of several phyla. The 
apparent potentiality to definite variation may be analo- 
gous to the similar mutations (De Vries) arising from the 
same stocks. (See section on causes of evolution, p. 834.) 

14. Similar rectigradations arise in connection with changes of 

environment and of habit (or ontogeny), but it is not 
known whether there is any real causal relation between 
changes of environment and of ontogeny and the origin 
of definite rectigradations like dental cusps and horns. 



13. Allometrons exhibit hereditary predispositions and definite 

variations only within descendants of the same phylum — 
for example, a certain brachy cephalic or dolichocephalic 
tendency once displayed within a phylum seems to be 
cumulative in successive generations. 

14. Allometrons are also directly observable as the result of 

changes of environment and of habit, but it is not known 
whether these are merely somatic, ontogenetic, or are 
actually germinal. 



PRINCIPLES OF EECTIGRADATION 

HORNS ARISE AS TYPICAL RECTIGRADATIONS 

ORTHOGENETIC, CONTINUOUS 

Phylogenesis of horns in the titanotheres. — Horns are 
biocharacters having very complex germinal and 
somatic relations. That they are separable in heredity 
is shown by the sudden and complete disappearance 



1. Horns arise from excessively rudimentary begin- 
nings — inconspicuous bony swellings which in their 
initial stages can hardly be detected on the surface 
of the skull. 

2. These horn rudiments arise independently in the 
same region of the skull in different phyla of the 
titanotheres at earlier or later geologic periods, 
respectively. 



fALAeOSYOPS 




OAOH/PPUS 




Figure 732. — Rectigradations in the teeth of Eocene ungulates 

A, Orohippus sp., a primitive equine: B, Falaeosyops palvdosus, a primitive titanothere. From specimens in the 
American Museum of Natural History. Internal view of the lower teeth. The circles mark the cusps, which 
appear independently in the phyla of the horse and the titanothere. They are at first barely visible but increase 
in size in successive geologic levels. 



of horns in certain domestic breeds of cattle and by the 
sudden appearance of rudiments of horns as major 
saltations in many orders of mammals that are 
naturally hornless. Horns are typical rectigradations 
in the titanotheres, as distinguished from allometrons, 
for the following seven reasons: 



3. The horn rudiments evolve continuously and 
very gradually change in form and proportion 
(allometrons) . 

4. In the earliest geologic stages in which horns 
have been observed they are found only in adult 
individuals, but through ontogenetic acceleration 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



815 



they are gradually pushed forward into younger and 
younger ontogenetic stages until finally they appear 
on the skull before birth. 

5. The horn rudiments, through phylogenetic 
acceleration or retardation, appear at earlier or later 
geologic stages in certain phyla than in other phyla 
(fig. 734). 



sh ape of the horn (fig. 733) is like that seen in the 
ontogeny of the horns of cattle. 

Ontogenesis of the horns of cattle. — The ontogenesis 
(development) of the bony horns of domestic cattle (PI. 
XLV) appears to be closely similar to the phylogene- 
sis (evolution) of the horn rectigradations in titano- 
theres. The comparison is useful. The horn biochar- 




FiGURE 733. — -Rectigradations and allometrons in the skulls, teeth, and feet of titanotheres 

Eectigradations are shown on the cuspules of the lower second premolar teeth; allometrons are shown in the proportions of the head (brachy- 
cephalic, mesaticephalic, dolichocephalic); allometrons are shown in the proportions of the median metacarpal (dolichopodal, mesatipodal, 
brachypodal). I, Eoiitanops, an ancestral lower Eocene mesaticephalic, dolichopodal titanothere; n, Palaeosyops, a broad-headed (brachy- 
cephalio), broad-footed (brachypodal) upper middle Eocene titanothere; UI, Telmaiherium, a medium-headed (mesaticephalic) upper Eocene 
titanothere; IV, Manteoceras, a medium-headed (mesaticephalic), medium-footed (mesatipodal) middle Eocene titanothere; V, DolicliorMnus, 
a long-headed (dolichocephalic) short-footed (brachypodal) upper Eocene titanothere. II-V represent four independent phyla of Eocene 
titanotheres which are widely divergent in the allometric evolution of the head and of the feet but are convergent in the independent evolu- 
tion of similar cusp rectigradations on the teeth and similar horn rudiments (H) on the skull. 



6. At the time of their first appearance horns seem 
to be equally developed in both sexes, but gradually 
they become much larger and more formidable in the 
males than in the females. 

7. In the titanotheres the horn swelling rises at the 
junction of the nasals and the frontals (black shading 
in fig. 733). It is borne chiefly on the nasals in 
dolichocephalic skulls, chiefly on the frontals in 
brachycephalic skulls. The original low, rounded 



acter in cattle is known to involve a large number of 
coordinated and correlated biocharacters, all of which 
are germinal in origin, namely: (1) A psychic predis- 
position to use the horn, which is manifested at or soon 
after birth; (2) a thickening of the epidermal cells 
above the bony horn rudiment, which in ontogeny 
appears earlier than the bony rudiment itself; (3) 
early agglutination of the hair to form the rudiments of 
the horny sheath; (4) appearance of the bony swelling, 



816 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



the osseous horn, at the junction of the f rentals and 
parietals (PL XLV, C); (5) shifting of the frontals and 
the osseous horns to the back of the cranium through 
allometry; (6) internal secretions from the interstitial 
cells of the male (testis) or female (ovary) germ 
glands. 



the junction of the frontals and parietals, the shifting 
of the osseous horns forward from above the eyes, and 
the changes in size and proportions, described in Chap- 
ter V, section 2, and Chapter X, section 3. 

The horn rudiments in different phyla, like the cusp 
rectigradations, show marked differences in phylo- 




FiGTJRE 734. — Independent appearance of rudiments 
though in the same part of the skull, 

PHYLOGENESIS AND INITIATION OF HORNS IN 
TITANOTHERES °* 

In contrast to the development of the horns of cattle 
all that we paleontologists observe in the evolution of 
the horns of the titanotheres is the rudimentary ap- 
pearance of the bony swelling of the osseous horns at 

s' See dctafled descriptions in Chapter V. 



of the horn (H) at diflferent stages of geologic time, 
in different phyla of the titanotheres 

genetic and ontogenetic velocity. One of the most sig- 
nificant discoveries made concerning the titanotheres 
is that their horn rudiments arise long after the phyla 
have been separated from one another, earlier in some 
phyla and later in others, but always on substantially 
the same part of the skull. This is probably true in 
general of the origin of horns in mammalian families. 
Each family probably has its predisposition to develop 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



817 



horns in a certain way, but in each this predisposition 
manifests itself at different times. According to the 
Lamarckian principle, the germinal psychic predisposi- 
tion to use the horn should appear first, and this initia- 
tion should be followed after a certain period by the 
ontogenetic bony horn rudiment due to habit, but 
there is no evidence for this sequence. The horn rec- 
tigradations, like the cusp rectigradations, seem to 
be initiated as a germinal phenomenon; it would appear 
that the psychic tendency and the horn rudiment ap- 
pear simultaneously, but as yet there is no positive 
evidence on this important matter of initiation. 

DENTAL RECTIGRADATIONS 

The grinding teeth of the titanotheres are 
sluggish in evolution as compared with those 
in other perissodactyls, especially the horseS) 
in which the grinding teeth are rapidly accel- 
erated in evolution. Nevertheless they ex- 
emplify even more perfectly than the horns 
the modes of rectigradation. Interest in the 
cusp evolution is enhanced by the fact that 
the cusp rectigradations in the lower teeth 
are mechanically coordinated and timed with 
the cusp rectigradations in the upper teeth. 
The causes of this constant and perfect 
mechanical adjustment of the rectigradations 
and allometry of the upper and lower grind- 
ing teeth of titanotheres and other mammals 
presents one of the most difficult problems in 
germinal evolution, the various explanations 
of which wiU be discussed in the third part of 
this section, under the heading "Theoretic 
causes of evolution." Here it may be sufficient to 
point out the modes of rectigradation without making 
reference to the theory of the underlying causes. 

The adaptive cenotelic (Gregory, 1910.1, 1914.1) 
principle in the rectigradations and allometrons of the 
titanotheres, as in all other perissodactyl quadrupeds, 
is to molarize the premolar teeth — that is, to convert 
the premolars into teeth of the molar pattern, adap- 
tively because a herbivorous animal can more effec- 
tually crush and grind its food with sLx complete grind- 
ers (molars and molariform premolars) than with three 
complete grinders (molars) and three incomplete 



grinders (premolars). As shown in detail in Chap- 
ter V this molarization takes the same course in all the- 
Eocene titanotheres, namely: 

1. Molarization generally, but not invariably, is first 
introduced in the fourth premolar (p|), which is usu- 
ally the most progressive tooth of the premolar series. 
Thus, these teeth are the first to acquire a partial molar 
pattern. (For an exception, note Palaeosyops copei.) 

2. The molarization in p| and the other premolar 
teeth of the titanotheres never attains the same degree 
of mechanical perfection that it does in the horses; 
the premolar teeth never become exactly like the 
molars — they are not perfectly molarized. 




Figure 735. — Cusp rectigradations in Telmaiherium 

The cusp rectigradations in the lower teeth are coordinated and timed with the cusp rectigradations- 
in the upper teeth, as shown here in the occlusion ot the jaws of Telmaiherium cuUrideJis, 
Princeton Mus. 10027 (type); one-half natural size. After W. K. Gregory. A, Crown view^ 
lower teeth (heavy lines) superposed on the upper teeth (Ught lines); B, inner side view. 

3. The molarization advances at a uniform rate, 
respectively and successively, in pf, p|, pf, in the 
order named, but affects only slightly p|. This phylo- 
genetic acceleration of the premolar rectigradations is 
more rapid in pf and less rapid in pf . 

4. It follows that the fourth upper and lower pre- 
molars (pf) pass through a series of stages of recti- 
gradation, which are taken up successively and passed 
through by p|, pf , whereas pf , with its simple mechan- 
ical crown, retains its remote ancestral condition. 

These four stages of molarization are illustrated in 
Figures 732, 735, and the accompanying analytical table 



818 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Mechanically correlated rectigradations and allometrons of the upper and lower premolar teeth oj titanotheres 

(Palaeosyops) 



Lower premolars 



Upper premolars 



1. The conical tooth or protoconid pi becomes elongate 
anteroposteriorly; rudimentary anterior and posterior ridges 
run down from summit to base, becoming more sharply marked 
and deflected linguad at the base; concomitantly the single fang 
tends to become divided. Total, one cusp. 

2. In p2 the posterior vertical cingulum ridge acquires a deep 
notch about halfway up the crown of the protoconid, separating 
off a posterior lobe (or hypolophid); the tooth broadens trans- 
versely, and the anterior and posterior fangs become wholly 
separate; a rudimentary paraconid is seen. Total, three cusps. 

.3. In p3 the paraconid becomes stronger; the hypolophid or 
posterior lobe grows higher; the protoconid broadens at the top, 
producing finally a well-marked oblique crest (or metalophid); 
concomitantly the hypoconid lobe becomes broader than the pro- 
toconid lobe,' concavities appear on the internal face of both 
cusps; by the deepening of these concavities the anterior and 
posterior crests or metalophid and hypolophid become sharply 
defined. Total, three cusps, a rudimentary protolophid, a 
metalophid, and a hypolophid. 

4. In pi the paraconid is slightly stronger; the transverse ridge 
at the summit of the protoconid divides, by the insinking of an 
anteroposterior nick, into the protoconid proper and the meta- 
conid, thus completing the protolophid; the posterior face of the 
hypoconid broadens into the hypolophid; the valley separating 
the protolophid from the hypolophid becomes straighter and 
more transverse, and thus these anterior and posterior crests 
become more symmetrical. Total, four cusps and two crests. 

The molarization of p4 is now complete except that an ento- 
conid is not developed until the tetartocone is developed above. 



1. The laterally compressed conical tooth p' exhibits ridges 
on the anterior and posterior faces and a continuous internal 
cingulum; the crown is a protocone. Total, one cusp. 



2. The triangular second premolar p- consists of the high 
protocone; externally devoid of cingulum, internally the cingu- 
lum gives rise to the well-developed deuterocone; while on the 
posterior ridge of the protocone is a very slight swelling which 
indicates the rudiment of the tritocone. Total, three cusps. 

3. The third premolar p' now has a true ectoloph, with the 
rudiment of a parastyle, and an incomplete external cingulum; 
the deuterocone is larger; it is embraced on the posterior as 
well as the anterior side (but not lingually) by the cingulum; 
it exhibits a faint anterior crest or protoloph; the tritocone nearly 
equals the protocone in size. Total, four cusps, an ectoloph, 
and a rudimentary protoloph. 



4. The fourth premolar p* exhibits a still more pronounced 
external cingulum, a prominent parastyle, a subequal protocone 
and tritocone, between which is the faintest rudiment of a meso- 
style. The deuterocone is still larger and not yet surrounded 
basally by the cingulum; it exhibits a slightly stronger anterior 
crest or protoloph, on which is seen a small protooonule. Total, 
six cusps, an ectoloph, and a protoloph. 

The molarization of p^ is now complete except that a tetarto- 
cone is not formed until the entoconid is formed below (as in 
Diplacodon) . 



Summarizing this numerical addition of rectigrada- 
tion cusps we get the following results in the com- 
parison of similar teeth from successive levels. 

Number of reciigradation cusps on lower premolars of EoceriC 
titanotheres 



First premolar-. 
Second premolar 
Third premolar.. 
Fourth premolar 



As stated above, this successive numerical addition 
of cusps as rectigradations affords one means of 
arbitrarily dividing the stages of evolution in each of 
the phyla and of naming them species, in the Lin- 
naean sense, and mutations, in the Waagen sense, in 



a manner simUar to that commonly employed for 
other quadrupeds. 

EECTIGRADATIONS OF OSBORN CONTRASTED WITH 
MUTATIONS OF WAAGEN 

Osborn's rectigradations, which are single char- 
acters, differ from Waagen's mutations, which are 
subunits of classification or taxonomy. One or more 
of Osborn's rectigradations when followed through a 
series of stages would constitute the subspecific stage 
known as a "mutation of Waagen." The same 
rectigradations (homomorphs) appear on the premolar 
teeth of one or more independently evolving phyla, but 
in the rectigradation cusps, as in the rectigradation 
horns, each phylum exhibits its own rate of evolution 
for each rectigradation, as well as its own phylogenetic 
velocity or rate of change. 

The foUowiug table presents crudely the actual 
modes of evolution in monophyletic, diphyletic, and 
triphyletic series of titanotheres, terminating in three 
"genera," A, B, C. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



819 



Rectigradations in raonophyletic, dipTiyletic, and triphyletic series, including genera, species, and ascending 

mutations of Waagen 



Geologic level of the Bridger 



Three-branch or triphyletic series (genus A) 



Mutations of Mutations of Mutations of 
species A^ species A^ species A3 



Two-branch or diphyletic 
series (genus B) 



Mutations c 
species B' 



Mutations of 
species B^ 



Single-branch 

or monophy- 

letic series 

(genus C): 

Mutations of 

species C 



Horizon C le (top) 

Horizon B 2e 

Horizon B 1, 1 

Horizon A 3 

Horizon A 2, 2. 

Horizon A 1, 1 (bottoni) 



Ale 
41 d 

Ale 
Alb 

Ala 

A' 



A2b 

A2. 



A3d 

ASd 

A3c 
A3b 
A3 a 
A3 



Ble 

Bid 
Bu 
Bib 



B2d 
B2C 
B2b 
B2a 

b2 



0= 

Cid 
Ci-^ 

Qlb 

CI" 
CI 



The table begins with three "Linnaean genera," 
A, B, C, each genus corresponding to one to three 
phyla, and each genus persisting throughout a long 
geologic period, namely, from horizon A 1, 1 to horizon 
C le. The table represents thirty-four "ascending 
mutations " of Waagen, distinguished by the successive 
addition of new rectigradations (1, 2, 3, etc.) and by 
changes of proportion (allometrons) (a, b, c, d, etc.). 

Tracing each phylum from the lowest to the highest 
geologic level through ascending mutations we observe 
that 

Mutation A'*^ may exhibit four new rectigradation 
cusps and final allometrons. 

Mutation A^ "^ may exhibit three new rectigradation 
cusps and subfinal allometrons. 

Mutation A'" may exhibit two new rectigradation 
cusps and intermediate allometrons. 

Mutation A'" may exhibit one new rectigradation 
cusp and incipient allometrons. 

Mutation A'" may exhibit no rectigradations and 
no changes of proportion. 

"mutations" and "species" 

Reading horizontally on any given geologic level, 
say horizon A, level 2, 2, we may discover six "ascend- 
ing mutations" of Waagen, namely, A'*, a^, A^*, B'"', 
b^ C", as shown in the table above. 

This- entire geologic section of the Eocene theoreti- 
cally shows, among the 34 ascending mutations of 
Waagen, 6 Linnaean species: A'"'^, A^"^", A^""*, B'"'*, 

Thus, through the contemporaneous evolution at 
different phylogenetic velocities of "homologous" 
(more strictly "homomorphous") rectigradations and 
allometrons new rectigradations and proportions at 
their first appearance correspond with the least or 
minimal systematic divisions of the zoologic system- 
atist. When a number of these biocharacters continue 
101959— 29— VOL 2 9 



to progress they constitute the "ascending mutations" 
of Waagen; when a number of these ascending muta- 
tions are similarly grouped they become the "species" 
of Linnaeus. Therefore the simultaneous evolution of 
a large number of biocharacters at different rates 
of evolution constitutes the divergence of chai'acter 
that underlies both the old zoologic and the new 
paleontologic classification. 

These terms and distinctions, which are very diffi- 
cult to express or to understand in descriptive text, 
are readily comprehended when members of five or 
six independently evolving phyla are compared in the 
museum, character by character, or as shown in the 
plates and synthetic diagrams. 

This mode of dental evolution through rectigrada- 
tion and allometry is precisely what we observe in 
other families of mammals, such as the Equidae, the 
Rhinocerotidae, the Canidae. 

HERITAGE SEPARABILITY OF RECTIGRADATION DENTAL 
CUSPS AND FOLDS 

The different rates of development of the premolar 
cusps indicate that they are true biocharacters, be- 
cause they display germinal (that is heritage) separ- 
ability. Exactly similar homomorphous biocharacters 
appear in the grinding teeth of the horse and of the 
ass, accompanied by a number of secondary recti- 
gradation enamel foldings and plications. In the 
hybrid mule, offspring of the male ass and female 
horse, it is found that these rectigradation folds are 
completely separated as shown in Figure 736. 

RECTIGRADATION CUSPS ON TITANOTHERE TEETH ARE 
UNLIKE SALTATIONS 

In all the hundreds of stages of titanothere dental 
evolution studied not a single instance has been 
observed in which a rectigradation cusp and fold 
appear suddenly, wholly formed; in every instance 



820 



TIT.^NOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



the origin of these new cuspules is so contionnus and 
gradual that they become discernible only after the 
passage of long periods of geologic time. They are 
also invariabty orthogenetic, evolving in a uniform 
direction. This mode of evolution recalls the ob- 
servation of Waagen as to the mode of origin of his 
mutations. In any large series of titanothere or 
other mammalian teeth one occasionally finds also 
major and minor saltations, smaller or larger cuspules, 
which are usually duplications of the rectigradation 
cusps, or reversions of former rectigradation cusps, 
such as the hypocone on the last superior molar. 

RECTIGRADATIONS INFLUENCED BY DEGREE OF 
ZOOLOGIC AFFINITY 

The addition of new dental cusps on the teeth and 
of horn swellinais on the skuU as rectigradations is 



PRINCIPLES OF PROPORTION 
PROPORTION BIOCHARACTERS (ALLOMETRONS) 

Changes in proportion common. — Allometrons play a 
larger part in the evolution of the titanotheres than 
rectigradations. All the bones of the skull and the 
skeleton are continuously undergoing a change of 
proportion — in other words, developing allometrons. 
Continuous change of proportion is a dominant mode 
of titanothere evolution; the addition of new parts 
and the loss of old parts have been second iu impor- 
tance. It is demonstrated throughout this monograph 
and summarized in the comparative measurements 
and various tables that all changes of proportion, as 
a whole, represent the sum of changes of separable 
parts, biocharacters, each of which has its own rate of 
evolution. For example, if the top parts of a titano- 






FiGUKE 736. — Separabilit}- of rectigradation biocharacters in the grinding teeth of the hybrid offspring of a 
male ass {Equus asinus) and female horse {E. caballus) 

Section through the crown of the fourth superior premolar tooth (p') showing that the mule inherits most of the enamel fold rectigradations 
of the ass but few of those of the horse. After Osborn, 1912.3. The following is a summary of the rectigradation biocharacters in the 
grinding tooth of the mule: 

Distinctly asslike.-. - ^jii peculiar to ass. 



Less distinctly asslike 6] 

Common to horse and ass 5=5 common to both. 

Distinctly horselike - 2\ 

Less distinctly horselikc 4J" 



6 peculiar to horse. 



influenced by nearness or remoteness of ancestral 
zoologic afhnity. (See predeterminate rectigradation 
under "Theoretic causes of evolution," p. 834.) De- 
grees of resemblance in the origia of these new bio- 
characters always increase with closeness and intensity 
of kinship. Thus the kinds of rectigradations (whether 
cuspules or hornlets), as well as the rates of change 
(phylogenetic velocities), are strongest in phyletic 
members of the same species; they are somewhat less 
strong in phyletic members of the same genera; they 
become increasingly dissimilar as we compare phyletic 
members of the same subfamilies, orders, etc. Never- 
theless the most striking principle in the mode of 
origin of rectigradations is that similar rectigradations 
sooner or later tend to appear in descendants of 
common ancestral forms. This is in striking contrast 
to the mode of evolution of allometrons; for both dis- 
similar and similar proportions tend to appear in the 
descendants of common ancestors. 



there skull (DolicJiorJiinus, see below) evolve more 
rapidly than the bottom parts, the axis of the skull 
must be bent downward, and the skull as a whole in 
consequence becomes cyptocephalic; if all the com- 
ponent bones broaden more rapidly than they lengthen 
the skull becomes brachycephalic. 

The causes of changes in the proportions of the 
skull are. probably the same in all other mammals as 
in man. The processes and results are distinguished 
as follows: 



Process 


Result 


Uniform lengthening or shortening 


Harmonic facial 


and 


of all parts — that is, at uniform 


cranial structure. 




velocities. 






Differential lengthening, shortening, 


Disliarmonic facial 


and 


or broadening of certain parts — 


cranial structure. 




that is, at different velocities. 







CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



821 



Allometrons differential {disharmonic) in origin. — 
At first glance it would seem that if the bones of the 
skull were slretched longitudinally as a whole the 
stretching would give rise to dolichocephaly, or if 
they were stretched transversely as a whole the 
stretching would give rise to brachycephaly, and 
approximately this mode of evolution is seen in skulls 
where all parts are harmonically lengthened or broad- 
ened in nearly equal percentages; but the actual mode 
of evolution in the titanotheres involves disharmonic 
and differential lengthening or broadening of each 
part. On analyzing closely by comparative measure- 
ments the progressive changes in titanothere skulls 
we find that each of the adjacent bony elements acts 
differently and more or less independently; the skull 
is not broadened or lengthened as a whole, as if com- 
posed of and stretched like India rubber, but each 
part of it acts as a distinct biocharacter and has its 
separate velocity or rate of change. This differential 
principle is illustrated on the accompanying diagram 
(fig. 738) showing some of the contrasts in the parts 
evolving in typical broad-headed {Palaeosyops) and 
long-headed {DolicJiorJiinus) skulls. 

Allometrons continuous in origin. — Do changes of 
proportion arise by saltation or continuously? Observa- 
tions made by the author on titanotheres, rhinoceroses, 
and horses compared with those made by numerous 
anthropologists on man demonstrate beyond doubt that 
all adaptive changes in proportion are continuous in 
origin; also that, like rectigradations, they take certain 
directions (from causes which will be considered in the 
next section) and are cumulative or progressive in 
successive generations. This kind of proportional 
change, as demonstrated below, may be suspended or 
even reversed in direction. This continuity in the 
evolution of proportions is demonstrated whenever a 
series that is nearly complete, or unbroken, is meas- 
ured, as they are in the measurement tables of Menodus, 
Brontotherium, and Megacerops in the sections treating 
of these genera. (See Chap. VI.) 

Exactly similar results have been obtained by Osborn 
in the comparative measurement of closely successive 
series of Oligocene horses. The same is indicated 
among the Oligocene rhinoceroses but is not yet posi- 
tively demonstrated. According to the unanimous tes- 
timony of anthropologists (Ripley, Races of Europe, 
1899.1, p. 624), the form of the human head is the 
result of very gradual change, either by elongation 
(toward dolichocephaly) or by broadening (toward 
brachycephaly) . In the native Indian races of America 
also, which are believed to be of the same remote racial 
origin, considerable diversity in the proportions of the 
head has gradually evolved under geographic isolation. 
Similarly Keith (1911.1) observes that in course of their 
evolution the tendency of one Negro tribe has been 
toward the accentuation of one set of skeletal propor- 
tions, of another tribe toward another set. The 



Dinka acquire high stature and narrow heads, the 
typical Nigerians low stature and narrow heads, the 
Basoko wide, short heads and low stature, the Bunms 
broad heads and high stature. 

Causes of head allometrons unhnown. — It is probable, 
but not yet demonstrated, that whatever the causes 





FiGTJRE 737. — Harmonic and disharmonic natural brachj'cepli- 
aly and dolichocephaly and artificial broadening or lengthen- 
ing of the outline of a skull 

A. Palaeosyops robustus, a natural brachycephalic skull; Ai, outlines of the same 
skull artificially stretched into dolichocephalic form (harmonic); B, Mesatirhinus 
megaThinus, a natural mesaticephalic skull; Bi, outlines of the same skull arti- 
ficially stretched on India rubber into dolichocephalic form (harmonic); C, 
DoUchOTliinus hijognailius, a natural dolichocephalic skull, viewed from above; Ci, 
DolkhorMnus hyognathus, a natural dolichocephalic skull (disharmonic) . The con- 
trast between) Ai artificial) and C (natural) and between Bi (artificial) and Ci 
(natural) is clearly displayed. 

of proportional change may be they are similar in 
the evolution of "bi'oad heads," "long heads," "long 
limbs," "short limbs," etc., among the titanotheres, 
the horses, man, and all other mammals. In man 
also there is considerable evidence that the evolution 
of proportion biocharacters is differential, namely, 
that while there is often harmonic evolution of head 



822 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



form, ill which all parts are more or less imiformly 
elongated or broadened together, there may also be 
disharmonic (that is, differential) evolution of head 
form, in which certain bony parts, such as the cranium, 
are lengthened while others, such as the cheek bones, 
are broadened. 

The continuous proportional evolution of the skull 
of the titanotheres has been demonstrated by the 
long investigations of the writer and his assistant. 
Dr. W. K. Gregory, which involved thousands of 



parative anatomy and in the investigations made for 
the present monograph.^' It is as follows: 

1. Elongation of the skull as a whole = dolichocephaly. 
Elongation of the cranium only = doliohocrany. 

2. Broadening of the skull as a whole = brachycephaly. 
Broadening of the cranium only = brachycrany. 

3. Elongation of the face only (that is, proopic dolichocephaly, 

as in £gM us) =dolichopy. 

4. Abbreviation of the face only = brachyopy. 

5. Elevation of the skull and head as a whole, as in Delphinus 
= acrocephaly, hj-poicephaly. 




~7^roportion 



breadth increments 



T^ercentcufe 



zp /5 /o ^ 



rindinq teeth to foramen ovale \ — JDolickarhL — 



i-enyth of nasals 



T^a laeosyops 
IDolichork in us 



T^espectiVe facial and cranial lengt/i of 7-hlaeosyops 



nespective premolar and molar len^iliofT-'alcLeosyops 



J)olichor}t I n.u.5 



AenctA increments 



S 20 2S 30 35 40 45 SO SS 60 65 70 75 



Figure 738. — Differential and distinctive] increment m every propoitional biocharacter group in eight bio- 
characters of Palaeosyops and Dohchoihmiis 

measurements of specimens belonging to a number of 
successive phyletic series, especially in Palaeosyops, 
Megacerops, Menodus, and BrontotJierium. This con- 
tinuity has been shown in the molar-cephalic index, 
which is the breadth across the cheek arches X 100 -4- the 
basilar length of the skull; also in other indices, such 
as the faciocranial, in which the continuous trend of 
proportional change has been carefully measured. It 
has been demonstrated that elongation of the skull 
as a whole (dolichocephaly) and broadening of the 
skull (brachycephaly) may arise independently in 
every phylum or line of descent. 

As shown in Figure 739, the titanotheres, like man, 
exhibit facial abbreviation (brachyopy) and cranial 
elongation (dolichocrany) in contrast to the facial 
elongation (dolichopy) and cranial abbreviation (bra- 
chycrany) of the horses. The changes of head form 
in Dolichorhinus are shown to be continuous and to 
result in dolichocephaly and cyptocephaly — that is, 
bending down of the face upon the cranium, as in the 
reindeer (Rangifer) and the hartebeest (Buhalis). 
The adaptive mechanical significance or selection 
survival value of these allometrons is sometimes 
apparent, sometimes obscure. 

TERMINOLOGY INDICATING PROPORTIONS OF THE SKULL 
AND SKELETON 



The general terminology of the changes in the pro- 
portion of the skull and postcranial skeleton has been 
gradually developed in the study of human and com- 




Th/aeosyoyps Manteoceras Da/i'cAorAinu^ 

Figure 739. — Continuous origin of allometron biochar- 
acters in the cranium and skull of man (A) and of 
the titanotheres (B) 
In reference to man the words brachycephalic, mesaticephalic, and doli- 
chocephalic denote respectively, brachycranial, mesaticranial, and 
doliehocranial — in other words, they describe the proportions of the 
cranial cavity. In reference to the titanotheres the same words are used 
to describe the relative length and breadth of the entire skull. Note 
especially the relative position of the orbit (o) . 

^8 The author's chief contributions to the craniometry of the titanotheres and 
other ungulates are the following: Dolichocephaly and brachycephaly in the lower 
mammals: Am. Mus. Nat. Hist. Bull., vol. 16, art. 7, pp. 77-89, Feb. 3, 1902; 
Coincident evolution through rectigradation (third paper): Science, new ser., 
vol. 27, No. 697, pp. 749-752, May 8, 1908; Skull measurements in man and 
the hoofed mammals: Science, new ser„ vol. 35, No. 902, p. 596, Apr. 12, 1912; 
Craniometry of the Equidae: Am. Mus. Nat. Hist. Mem., new ser., vol. 1, pt. 3, 
pp. 57-100, flgs. 1-17, June, 1912. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



823 



6. Enlargement of the head as a whole = macrooephaly. 
Relative diminution of the head as a whole = microcephaly 

7. Elongation of the limbs as a whole = dolichomely. 
Elongation of the feet = dolichopody. 

8. Shortening of the limbs as a whole = brachymely. 
Shortening of the feet = brachypody. 

PROPORTIONAL EVOLUTION IN THE TYPICAL 
BRACHYCEPHAL PALAEOSYOPS 

As the course of the evolution of Palaeosyops from 
generation to generation is toward solidity, the skulls 
become progressively larger, the individual parts 
become more massive, the sagittal and occipital crests 
become more robust, the linea aspera of muscular 
origin and insertion becomes more pronounced, the 
zygomata become more widely arched and thicker in 
section posteriorly, and the zygomatic flanges become 
deeper. 



syops is partly ancestral or inherited (paleotely), and 
it partly looks forward (cenotely) to a similar pro- 
portional evolution in a certain direction or a sum of 
proportional changes whose effects are less clearly 
manifest in lower stages and become more and more 
evident in higher stages. The harmonic tendency 
toward progressive weight and brachycephaly is ac- 
companied by the disharmonic tendency in certain 
changes of proportion. 

BRACHYCEPHALIC ALLOMETRONS AFFECT ALL 
BIOCHARACTERS; HARMONIC 

To a certain degree brachycephaly affects every 
contour, relation, bone, and tooth in the skull. Even 
in the earliest known species of Palaeosyops brachy- 
cephaly is the dominant evolutionary trend, although 
we still find reminiscences of less brachycephalic 
ancestors, forms more like the mesaticephalic Lim- 




FiGUHE 740. — Proportions and flexure of skull in horses and titanotheres 

I, Horse proportions: Eguus, face elongated, cranium abbreviated; Eohippus, cranium and face equal. II, Titanothere proportions: 
Brontotherium, face abbreviated, cranium elongated; EotUanops, cranium and face equal. Ill, Skull flexure (cyptocephaly) in ruminants 
{EangiftTj Bubalis) and titanotheres (Palaeost/ops, brachycephalic; Bolichorhinus, dolichocephalic). 



CONTRAST BETWEEN EVOLUTION AND GROWTH 

The first impression from these modes of change is 
that in Palaeosyops skull evolution is like a continu- 
ous growth, as if the skulls found in the higher geo- 
logic stages represented extremely long-lived, mature, 
but not senescent forms of the skulls found in earlier 
geologic stages. This impression is only partly true. 
In evolution new proportions develop through the 
separate rate of change of different biocharacters; 
consequently evolution is more fully differential than 
growth. 

DIFFERENTIAL (DISHARMONIC) ALLOMETRONS IN 
PALAEOSYOPS 

In the evolution of the genus Palaeosyops between 
the lower and upper Bridger beds (a geologic interval 
of 1,500 to 2,000 feet) there is an average increase of 
about 14 per cent in the size of the skull, but this 
ranges in the different bones of the skull from 8 per 
cent minimum to 73 per cent maximum. Thus there 
is a prolonged harmonic evolution toward brachy- 
cephaly and a separate differential evolution of bio- 
characters. Therefore the generic character of Palaeo- 



noJiyops and Telmatherium. The modes of departure 
from an ancestral, more generalized type toward the 
extreme brachycephalic type are apparent in every 
biocharacter, as shown in the eighteen structural 
features listed below:*" 

1. Ancestral, smaller, mesaticephalic type, only partly known, 
passes into the progressively larger, more brachycephalic type. 

2. Skull as a whole increases in massiveness, showing a more 
rapid increase in breadth than in length — that is, as 1.5 to 1. 

3. Premaxillary symphysis becomes short and rounded. 

4. Maxillaries are increasingly abbreviated. 

5. The nasals, originally more elongate and obtuse, become 
more abbreviate and pointed. 

6. A smooth, plane forehead develops a prominent forehead 
convexit}'. 

7. The malar below the orbit deepens with extreme rapidity. 

8. The zygomata become more arching and deeper. 

9. Postglenoid and post-tympanic processes, originally sepa- 
rate and leaving external auditory meatus open below, become 
almost united, closing external auditory meatus below. 

10. Paroccipital process, distinct at base from post-tympanic 
process, becomes confluent at the base with post-tympanic 
process. 

11. Basifacial plane, originally more horizontal, becomes 
displaced downw-ard. 

12. Palatines exhibit an abbreviated exposure on the palate. 

'» other progressive characters are enumerated on p. 257. 



824 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



13. Foramen ovale approaching foramen lacerum medium. 

14. Postcanine diastemata disappearing. 

15. JMandibular rami develop a short, triangular coronoid 
process, condyle elevated, mandible with depressed angle; 
face of coronoid flattened in front, tending to overhang third 
inferior molar; rami increasing in depth and thickness; diastema 
between canines and first premolars disappearing. 



attributed to Palaeosyops are the expressions of its 
progressive brachycephaly. These so-called generic 
characters are based upon the harmonic trend, or 
direction, of evolution in the skull as a whole with 
its correlated changes in a large number of biochar- 
acters. Some of the "specific" characters which the 




Figure 741. — Skulls of three species (probably successive) of Palaeosyops from the middle Eocene, Bridger 

formation 
A, P. major, level B 3; B, P. Icidyi, level probably upper C or D; C, P. robustus, level probably D. 



16. Incisors become more transverse in position; canines 
stouter and rounder. 

17. Molars become broader than long; parastyle robust; 
robust internal and external cingula. 

18. Hypoconulid of ma from a subcrescentic form acquires a 
progressively conic form. 

HARMONIC PROPORTION TREND CONSTITUTES THE 
GENERIC CHARACTER OF PALAEOSYOPS 

The chief "generic" characters which Leidy, Cope, 
Earle, Osborn, and others have from time to time 



same authors have attributed to Palaeosyops are the 
successive stages in its progressive brachycephaly. 
Other specific characters consist of successive rectigra- 
dations which are added to the teeth and the skull. 

FOURFOLD MODES OF ALLOMETRIC BRACHYCEPHALY 

From comparison with the skulls of other mammals 
we draw four important conclusions in the study of the 
allometric brachycephaly of the skull of Palaeosyops 
as a whole, namely: 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



825 



1. No addition oj hiocharacters in Palaeosyops. — In 
Palaeosyops no distinctively new elements or organs 
are added to the skull through progressive brachy- 
cephaly and no characters are lost, but allometrons 
are established by the abbreviation of some parts 
and the expansion of conjoining parts. For example, 
the postglenoid and post-tympanic processes tend to 
close in the auditory meatus. 

2. Correlation oj hiocharacters. — Age and sex both 
exert a positive divergent influence on all the 
quantitative characters of the skull, such as 
massiveness and robustness. Old male skulls 
are more progressive in form, more prophetic 
of higher successive stages than young male 
skulls, and some female skulls are more con- 
servative in form than male skulls. In brachy- 
cephalic phyla the females preserve the more 
mesaticephalic indices of ancestral forms 
Similar conclusions are summarized by Dr- 
C. Hart Merriam (1895.1), from his studies 
on the pocket gophers (Geomyidae), as follows: 

The female generally has the brain case broader and 
flatter, the zygomata narrower and less angular, the 
jugal narrower anteriorl.y, the rostrum and nasals 
shorter, and the skull as a whole smoother. In other 
words, the cranium of the female is much less special- 
ized than that of the male and often points sugges- 
tively to the stock from which the species was derived. 
It thus happens in the case of series of species in which 
the suocsssive forms in the development of a partic- 
ular type are still extant (as in the texensis-bursarius 
series) that the female resembles the male of the species j 
next below in the line of descent more than the male of 
her own species. 

Concerning the influence of age and mus- 
culature the same author observes (op. cit., p. 32): 

Fundamental characters are based on structures and rela- 
tions that enter into the ground plan of the skull and are of 
high morphologic weight; superficial characters are the result 
of special adaptations and particular miuscular strains and are 
of little value except as affording recognition marks for species 
and in some instances for genera also. The fundamental struc- 
tures are mostly hidden, comprising the floor of the brain case, 
the craniofacial axis, and the turbinated bones. 

Thus young jaws of the species of Palaeosyops and 
Limnohyops are more similar to each other than old 
jaws. 



3. Remoteness of zoologic affinity (Jamily, supra- 
generic, and generic). — Such affinity apparently exerts 
little or no controlling influence on the brachycephalic 
evolution of Palaeosyops, because we observe that 
other related generic forms {Limnohyops, Telmathe- 
rium) are evolving differently along their own lines. 

4. Closeness of zoologic affinity {generic, specific, 
varietal). — These relationships exert a positive con- 
trolling influence because of the kinship of (1) a 




Figure 742. — Outlines of skulls of Palaeosyops, Dolichorhinus, and 
Eotitanops, illustrating the evolution of brachycephalic and dolicho- 
cephalic types 
A, Brachyceptialic .skull of P. leidyl (shaded), superposed upon B, dolichocephalic skull of 
D. Injognaihus (heavy outline); C, hypothetic outline of the skull of E. gregoryi, drawn 
to the same scale as A and B. This illustration shows the contrasts produced in the 
evolution of brachycephaly, dolichocephaly, and cyptocephaly from the primitive mesa- 
ticephaly of Eotitanops. 

common progressive brachycephalic tendency, (2) the 
kinship of similar habit and habitat, and (3) the kin- 
ship of a similar incidence of competition and selection 

DIFFERENTIAL ALLOMETRONS OF SPECIFIC VALUE 

Whfle the skull as a whole is increasingly brachy- 
cephalic the differentia] rate of increase in its separate 
parts furnishes an important share of the material 
which paleontologists have seized upon in the sys- 
tematic description of "species." This is shown in 
the careful analysis of the measurements given in the 
following table: 



826 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Differential measurements, in millimeters, in seventeen dental and cranial characters oj shulls transitional between 

Palaeosyops major, P. leidyi, and P. rohustus 



[All from the Bridger formatiou] 





« 


.=■ 


S 


S 


■3 


•3 





^"' 


S 







H D 





?.^ 




n 


53 








s 


.a 




i 


g 


gg 


■^ 


S^ 


g 




































a 


^ 


> 


> 


fS 


S 


. 


'"' 


a 


<^^ 






a 


p-S 




ffl 


^c 


CO 






" 


S 


al 


p^ 




Ph 


epn 


PM 


s*^ 






^i 


!" 


S>D 


3 0) 




3 ?fi 


oP 


"S'^ 


fb 


1 " 


f| 


l§ 


s's 




S 


3^ 


s 


s 


gp 
< 

^:0 




si 




|p^ 


if^ 


is 


Sf^ 


a-s 






c 


1 


< 




1 


1 


s 


1 


1 


'■S» 


." 




1 


s'S 




1 


1 


s 


i 






s 


^ 


1 




i 




S 


II 




Pm 


frn 


P4 


PU 


C^ 




PM 


1^ 


CM 







0-2 


P4 


o3 


1. P^, transverse- _ _ -- 


?!4 




?,5 


95 






»98 


27 


4 


1 


8 


9 


19 


1 




148 


151 


151 


158 








"170 


9 4 


6 


11 


9 7 


14 


1 1 




94 
3Qn 




102 
413 


100 
415 


— - 


94 


»103 
-440 


100 


6.4 

5 8 


1.6 

1 7 


3 

6 5 


.75 
1 6 


9 
11 


75 




1 


5. Cranium, postorbital process of frontals to condyles 
































183 




193 


199 






»906 




5 6 


1 4 


3 1 


7 


19 


1 




















[4.3 


1 










6. Face, postorbital process, frontals to premaxillaries 


?07 




990 


916 






"934 




] or 


or 


[6 3 


1 5 


13 


1 




















le. 2 


1. 5 










7. Breadth across zygomata _ _ __.--_ 


°?S8 




•■?S5 


310 




-395 


°340 




13 


3 ? 


15 


3 7 


18 


1. 5 








181 


195 




994 










16 


4 








»1fifi 




163 


166 




-179 










5 5 


1 4 






10. iSIaximum width, one condyle - - 


5? 




56 


57 




58 


»61 




6 8 


1 7 


9 7 


9 4 


14 


1 1 






53 


53 


69 




60 










17 


4 






12. Depth, malar below postorbital process malar.. . . . _ 


40 


50 


55 


58 


59 


59 


70 


"■70 


38 ■ 


9 


96 


6 5 


73 


6 


13. Free length of nasals 


86 




Q9 


99 


103 




"1 08 




8 


9 


15 


3 7 


95 


9 


14. Postorbital process of frontal to tip of nasals 


180 




1'89 


189 






"908 




5 


1 9 


11 


9 7 


15 5 


1 9 


15. Postorbital process of frontal to nasal canthus 


91 
19fi 


- — 


101 
908 


101 

90S 


108 








12 
5 


3? 
1 9 


7 


1.7 


20 


1 6 


16. Postorbital process of frontal to tip of occiput 










17. Tip of nasals to tip of frontal convexity. . 


240 


I24O? 


240 


240 


- — 


-250 


— - 








4 


1 


4 


33 











On account of the breaks in the phyletic series and 
the effects of distortion the results presented in the 
above Palaeosyops table are approximations. They 
would have been much more strddng had it been 
possible to compare completely the earliest, smallest 
known form, P. paludosus, with one of the latest and 
largest forms, P. grangeri, and had all the skulls been 
free from distortion. In this table are compared only 
the three intermediate stages, P. major, P. leidyi, 
P. rohustus, from which the following principal 
generalizations may be made. 

HARMONIC ALLOMETRY EXCEPTIONAL 

Of the 17 characters mentioned in the above table 
2 only are harmonic, giving exactly the same per- 
centages of increase. Other characters which show 
near approximations to harmonic increase are (4) 
the length of the skull, (2) the length of the grinding 
teeth, (1) the breadth of the fourth premolar, (10) 
the width of the occipital condyle. These nearly 
harmonic increments range from 11 to 14 per cent. 

The skull increases 7 per cent faster in breadth than 
in length, the breadth increment being 1.6 times the 



length increment. From P. major to P. robustus 
the skull increases in length 50 millimeters, or 11 
per cent; it increases in breadth 57 millimeters, or 
18 per cent. 

Cranial and facial increments are subequal. The 
gain in the length of the preorbital region is 12 per 
cent; the gain in the postorbital region is 13 per cent. 
Thus in Palaeosyops the face and the cranium increase 
at approximately equal rates, whereas in most other 
titanotheres the cranium increases more rapidly than 
the face. For example, in DolicJiorJiinus (as compared 
with MesatirJiinus) during a given period the face 
increases 26 per cent and the cranium 38 per cent. 

The greatest gain differentially is in the breadth of 
the masseteric muscle area of the zygomata. On 
comparing a young adult specimen of P. major with 
an aged specimen of P. rohustus we observe a gain of 
73 per cent in the depth of the malar below the orbit. 

When we compare these percentages with those 
based on similar measurements of skulls of specimens 
of other genera we again establish as the distinct 
generic character of Palaeosyops that it is a genus 
characterized by certain definite (that is, brachy- 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



827 



cephalic) and highly differential rates of evolution in 
different parts of its skull and teeth. Although 
progressive brachycephaly is the chief harmonic trend 
of development, this harmonic trend affects every bone 
in different degree. 

DOLICHORHINUS: ADAPTATION OF THE LENGTHENED 
HEAD TO THE SUPPOSED HABIT OF GRAZING 

In the Dolichorhinus phylum (Dolichorhininae) we 
are afforded a unique opportunity of studying the 
allometry that transforms a mesaticephalic skull 
{Mesatirhinus peter soni) into a dolichocephalic skull 
(DolicJiorJiinus hyognathus) with a strong tendency 
to cyptocephaly. The comparison between the older 
and newer forms is not absolutely exact because we 
can not yet demonstrate positively that Dolicho- 
rhinus was directly descended from Mesatirhinus, but 
we are certain that the changes came about through 
a general lengthening of the skull with a relative 
narrowing; moreover, that this is the result of a 
differential evolution in which some bones grow much 
more rapidly than others. At least part of these 
changes are apparently correlated with adaptation 
to the habit of grazing. 

A comparison of Dolichorhinus and Palaeosyops dis- 
closes the following eight features of Dolichorhinus: 
(1) There is a general dolichocephalic increment of 
33 per cent; (2) harmonic increment is rare; (3) incre- 
ment in length greatly exceeds increment in breadth; 
(4) cranial increment (dolichocrany) exceeds facial 
increment (dolichopy); (5) the increment in breadth is 
the minimum; (6) head bending (cyptocephaly) is due 
to the increment in the length of the roof bones of 
the skull in excess of those of the basicranial line; 
(7) the increment of the grinding series and of the 
cranium is harmonic; (8) the increment of different 
members of the grinding series is differential, dis- 
harmonic. 

DOLICHOCEPHALIC INCREMENTS OF MESATIRHINUS AND 
DOLICHORHINUS 

As Mesatirhinus petersoni is itself a dolichocephalic 
form these changes are not so marked as they would 
be were we to compare Dolichorhinus hyognathus with 
its unknown mesaticephalic ancestors, yet we discover 
very marked allometry. Partly owing to the crushed 
condition of the skulls a very high degree of accuracy 
can not be claimed for the percentages in the table, 
although they are based on averages obtained from 
a very large number of measurements. The two ghief 
peculiarities of Dolichorhinus hyognathus as compared 
with Mesatirhinus petersoni are the excessive length 
of the skull as a whole and the convexity (cypto- 
cephaly) of the cranium. 

The cyptocephaly of the skull was brought about 
by more rapid evolution of the top or upper bones 
than of the bottom, the basicranial and palatal bones. 



The total skull measurement (cranium to face) from 
the tip of the nasals to the tip of the occiput along a 
dorsal line increased 44 per cent, whereas the measure- 
ment along the inferior line, premaxillaries to condyles, 
increased only 33 per cent. The top of the cranium 
proper (postoi-bital process, frontals to tip of occiput) 
shows an increase of 40 per cent, whereas the cor- 
responding measurement of the basal line (original 
boundary of posterior nares to posterior face of 
condyles) shows an increase of only 30 per cent. 
The basis of the cranium, the cranio-facial angle, has 
been produced downward with relation to the basis 
of the face from 157° (or 23°) in if. petersoni and 136° 
(or 44°) in D. hyognathus — that is, through 21°. 

The skull as a whole, premaxillaries to condyles, 
and the cheek teeth as a whole, p'-m^, increased 
about equally in length (33 per cent). A similar 
equal increase of the skull and cheek teeth also marks 
the progressive dolichocephaly of the Oligocene genus 
Menodus. The cranium lengthens more rapidly than 
the face. The proportional evolution of the cranium 
and face can best be measured on a horizontal line 
from the tip of the premaxillary to the condyle, 
by projecting on this line the tip of the postorbital 
process of the frontal. The face increment (26 per cent) 
is thus seen to be 12 per cent less than the cranium 
increment (38 per cent) and 7 per cent less than the 
increment of the skull as a whole (33 per cent). The 
region of the greatest elongation of the cranium is 
that between the hinder border of m^ and the foramen 
ovale, which increases 54 per cent. In front of and 
behind this region the increments in length are less, 
as shown in the series of detailed comparative measure- 
ments below. It is in this greatly stretched region 
that many anatomical peculiarities, such as the 
extremely elongate posterior nares and the protrusion 
of the ethmoturbinal, the secondary palate, are noted. 

Other features of differential evolution in length 
are as follows: The premaxillaries and the postcanine 
diastemata have lengthened faster (increment 44 per 
cent) than the skull as a whole (33 per cent). The 
free portion of the nasals has increased 55 per cent, 
and the nasals as a whole have increased 70 per cent 
as compared with the skull as a whole (33 per cent). 
The zygomata have increased 44 per cent, or much 
more than the skull as a whole (33 per cent). There 
is a marked forward displacement of the postglenoid 
process, the distance between the postglenoid and 
post-tympanic processes having increased 55 per cent, 
whereas the corresponding distance between the 
foramen ovale and foramen condylare has increased 
only 34 per cent. As a result of this forward dis- 
placement of the postglenoid process, the external 
auditory meatus has been rotated forward from an 
angle of 78° in M. petersoni to a corresponding angle 
of 53° in D. hyognathus. The anteroposterior diameter 
or opening of the external auditory meatus has 



S2S 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



received the enormous increment of 79 per cent, 
which may be contrasted with the actual closure of 
this same region in the brachvcephalic Palaeosyops. 

SLOW BUT DIFFERENTIAL EVOLUTION IN BREADTH OF 
SKULL OF DOLICHORHINUS 

While the skull of DolichorMnus was rapidly 
elongating (increment 33 per cent) it was much more 
slowly evolving in breadth (increment 9 to 15 per 
cent). Thus the disparity between the length and 
breadth was constantly increasing, as shown in the 
following comparisons: (1) The transverse measure- 
ment across the zygomata increased 15 per cent, or 
less than one-half the increment in length of the skull 
(33 per cent). Similarly, the breadth across the post- 
glenoid and post-tympanic processes increased only 
13 per cent. (2) The space across the infraorbital 
shelves increased only 9 per cent, partly because 
these shelves were already relatively broad in M- 
peter soni. (3) Consistent with the bracing of this 
long and narrow skull we observe that the occipital 
condyles widen much more rapidly than the remainder 
of the skull, the increment being 28 per cent. (4) 
Similarly, the occiput increased in breadth 68 per 
cent in order to provide a large surface of attachment 
for the cervical and cephalohumeral muscles. The 
fact that the occiput increased in height only 20 per 
cent indicates that the horns were used in an obliquely 
lateral rather than in a tossing motion. (5) The 
transverse measurement across the horn region shows 
the relatively high increment of 35 per cent, while the 
transverse measurement across the postorbital proc- 
esses of the frontals increased 23 per cent. (6) A 
noteworthy fact is that the palatal space between the 
anterior premolars receives the large increment of 45 
per cent. This broadening of the palate anteriorly 
would allow greater space for the incoming food. 

DIFFERENTIAL EVOLUTION IN THE GRINDING TEETH 
OF DOLICHORHINUS 

It is a striking fact that the teeth as a whole in- 
creased in length at practically the same rate as the 
skull as a whole, namely, 32 per cent. A second fact 
of great interest is that the dental series lengthens 
much more rapidly than the individual teeth increase 
in width, the transverse increment of the teeth being 
but 17 per cent while the total longitudinal increment 
is 32 per cent. The teeth thus share the general 
elongation of the skull. The grinding areas of the 
cheek teeth as a whole become larger. Measuring the 
grinding areas by the sum of the rectangles circum- 
scribed by each tooth in the series we find that the 
grinders of D. TiyognatTius have received an increment 
of 68 per cent over those of M. petersoni. This 
increase is differential, however, because the grinding 
area of the premolars has increased only 41 per cent 
as against 76 per cent increase in the true molars. 



This is the reverse of what occurs in the horse, in 
which the premolar increment is greater than the 
molar increment. 

The increment of the total grinding area of the 
crowns among the individual teeth is distributed as 
follows: P', a very slight increment in size; p^, an 
-ncrement in grinding area of 48 per cent; p^~^, an 
ncrement in grinding area of 36 per cent; true 
molars, m'"^, an increment in grinding area of 80 per 
cent. Thus the true molars received by far the 
greatest increment. 

CONTRASTS OF DIFFERENTIAL EVOLUTION IN BRACHV- 
CEPHALIC AND DOLICHOCEPHALIC SKULLS 

The net results of these increments in length and 
width may be summarized as follows: 

Percentages of increase in size of skull and teeth of DolichorMnus 
and Palaeosyops 





Dolicho- 
cephaly 

(Doli- 
cho- 

rhinus) 


Brachy- 

cephaly 
(Palaeo- 
syops) 


Increment in length: 

1. Entire skull, premaxillaries to condyles. 

2. Top of skull, tip of nasals to tip of occi- 

put 


33 

44 
38 
26 
44 
70 

15 

13 
9 
28 
68 
35 

32 


11 


3. Cranium proper, postorbital region 

4. Face proper, preorbital region 

5. Premaxillaries, anterior part of face 


13 
12 


Increment in breadth : 


18 


8. Postglenoid and post-tympanic proc- 




















Elongation of grinding teeth: 
13. Total 


14 


14. Premolar series 









9 









SUMMARY OF HARMONIC AND DIFFERENTIAL ALIOME- 
TRONS IN THE SKULLS AND FEET AND AN INTERPRE- 
TATION OF THE PHYLOGENY OF THE TITANOTHERES 

By W. K. Gregory 

HISTORY OF RESEARCH 

The first step toward the discovery of the allometric 
principle was the analysis of dolichocephaly and 
brachycephaly in the titanotheres by Osborn in 1902 
(1902.208). At that time it seems to have been 
thought that the tendencies toward the long-headed 
and short-headed forms were mutually exclusive and 
that their divergent mass effects indicated a very 
ancient separation between the corresponding genera. 
It was also apparently assumed (Osborn, 1902.207) 
that the tendencies acted more or less uniformly on 
all parts of the skull. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



829 




'a.(n.nas.plj 





FiGWJBE 743. — Skull of Dolichorhinus hyognathus 
Am. Mils. 18S1 (type ot " Telmatotherium cormitum"). Top (A), palatal (B), and left side (C) views. 



830 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



In 1904, by a study of the progressive percent- 
ages of increments in the skulls of Palaeosyops and 
Bolicliorliinus for this monograph, Gregory was able 
to establish the principle of difl'erential as opposed 
to harmonic dolichocephaly. In 1907 Osborn, in 
considering the nearness of affinity of the subdoli- 
chocephalic MesaiirMnus to the relatively broad- 
skulled Manteoceras, concluded that broadening and 
lengthening were "quantitative" characters, which 
were largely independent of remote hereditary control. 
In 1908 Gregoiy's studies indicated that the doli- 
chocephalic Menodus and the brachycephalic Brontops, 
together with the intermediate genus Allops, con- 



/ Progressive brachycephaly U Contrast of dolichocephaly I 
and brachycephaly \ 


■f-ZO 
-f-/0 



-2 














A 

B 
C 












n 

B 
C 

A 




























/ 






















1 






















/-' 






















// 










.1 1 




ij 






,'/ 


/ 


1;\ 








J/ A 


■.J 












1' 1 










1 / 












^/ 




y," 








1/ 


/ 




11 












"^ 












\y 


/ 


/'/ 












/ 








A 


■l< 


/'/ 










\ 












f-' 






































































\ 


^ 


P' P' P' P* '"' '"' '"' P' P^ P^ P* '"' "' "-^ 


A -Srontops Irrachycepftalus average A.Menoclu.s tji^anteu-s 
S^&rontops dispar averafe ^. A^e^acerops Ifucco 
C.3rontop5 rodustu.5 average C - £rontotherLu.fn curtum 
2).^ronToTheriu.m fi^as 



Figure 744. — Differences in the proportions of premo- 
lars and molars corresponding to differences in the 
proportions of the skull 

Progressive brachycephaly: Graph I shows numerical excess of trans- 
verse over anteroposterior diameters of the upper premolars and molars 
in successive stages of the progressively brachycephalic Brontops 
phylum. 

Contrast of dolichocephaly and brachycephaly: Graph II shows contrasts 
in the proportions of the upper premolars and molars between 
dolichocephalic (.^) and brachycephalic titanotheres (B, C, D). The 
dolichocephalic form (A) has relatively narrow premolars and posi- 
tively narrow molars, in which the transverse becomes less than the 
anteroposterior diameter. 

stituted a separate group, the menodontine, in con- 
trast to the brontotheriine group, including the rela- 
tively long-skulled Brontotherium and the broad- 
skulled Megacerops. 

In a study of reversal in the proportions of the feet, 
Gregory in 1910 (1910.1) adduced evidence to show 
that the broad-footed genera Palaeosyops, Manteo- 
ceras, Megacerops, Brontotherium had been derived 
from narrow-footed stem forms and that their broad 
magnum had been derived from a narrow magnum 
having a wedge-shaped lower end, much like that of 
Eotitanops. 

By a close study of the Eocene titanotheres in the 
spring of 1914 Osborn and Gregory discovered a 
principle that may afford a key to the phylogeny of 
the titanotheres: Differential lengthening and broad- 
ening of all parts of the skull and feet have acted 
either simultaneously or in combination at different 
rates in the several phyla. 



APPLICATION OF THE PROPORTIONAL REVERSAL 
PRINCIPLE TO THE TITANOTHERES 

Eotitanops. — Eotitanops horealis is dolichocephalic. 
The details of the dentition and infraorbital, malar, 
and other characters of the skull show that it is 
structurally ancestral, or nearly so, to the progres- 
sively brachycephalic LimnoJiyops and Palaeosyops. 
This fact favors the view that the narrow-footed 
Eotitanops ^' gave rise to the broad-footed Palaeosyops. 

Telmatherium. — Telmatherium unquestionably re- 
sembles Manteoceras in some characters, the Palaeo- 
syopinae in others. A close study of the skull and 
dentition of T. cultridens shows no feature inconsistent 
with remote derivation from a mesaticephalic species 
allied generically to Eotitanops horealis. Telmathe- 
rium ultimum conserved the mesaticephalic features 
of the basicranial region, but the middle part of the 
skull, like that in all Uinta Basin forms, had already 
undei'gone considerable elongation. The premolar 
series, p^-p*, as shown by percentage ratios, is rela- 
tively longer in T. cultridens than in Eotitanops; but 
in T. ultimum the molars have lengthened so much 
that the premolar series is again relatively short. In 
this species, however, the molars have also broadened 
decidedly, so that they are very large and somewhat 
broad. Here, then, was a combination of differential 
broadening and lengthening. 

Manteoceras. — As compared with the very primitive 
and ancient titanothere Eotitanops, the hypothetical 
ancestors of Manteoceras in the Wind River were also 
mesaticephalic and probably allied to the forms that 
are ancestral to Telmatherium and to Palaeosyops. 
The Bridger Manteoceras shows a broadening of all 
parts, as well as a lengthening of the middle portion 
of the skull; the grinding series also shows a slight 
broadening and a relative increase in size. But in 
the Uinta C stage (M. uintensis Douglass) a marked 
dolichocephalic tendency has supervened, affecting 
especially the postcanine diastema and the length of 
the premolars and of the first and second molars. In 
the feet Manteoceras manteoceras shows a marked 
secondary broadening of the slender-footed type- and 
is clearly allied to the narrow-footed Mesatirhinus. 

Protitanotherium. — Protitanotherium, to judge from 
the resemblances in the dentition and skull to 
Manteoceras uintensis, was possibly a derivative of 
some earlier species of Manteoceras. In common with 
all other upper Eocene and Oligocene titanotheres 
Protitanotherium had probably suffered an elongation 
of the middle portion of the skull, but in addition to 
and superseding this, a strong brachycephalic tendency 
had set in which, to judge from the front part of the 
skull and dentition, had brought Protitanotherium 
near to the stem of Brontops. 

" The fifth digit of E. iorealis is perhaps not reduced past hope of subsequent 
increase. It is broken off in the type. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



831 



Brontops. — The elongation of the midcranium 
having been effected, Brontops began to develop 
buccal brachycephaly and finally carried it to an 
extreme in B. rolustus and "Diploclonus" amplus. 

Allops. — Allops lengthened the skull, and especially 
the molar series, but also showed some brachycephalic 



gressively increase rather than decrease in size; also 
the close alliance of Menodus and Brontops points 
rather to the possible derivation of Menodus from 
some subdolichocephalic form of Manteoceras, like 
M. uintensis, yet retaining the sharper-edged canines 
of certain species of Bridger Manteoceras. At any 




£btitanops V^^ J E 

A^anteocera-s 
Figure 745. — Skulls of Eocene titanotheres, showing changes in proportion and in development of horns 
Hornless series: A, EotUanops borealis: B, Limnohyops prisons: C, Palaeosyops leiiyi; D, Tdmatherium ultimum. 
Precociously horned series: E, Manteoceras manteoceras: F, Mesatirhinus petersoni: Q, Metarltinus earlei: H, DolichorUnus hyognathus. 



tendency. Its transitional character between Bron- 
tops and Menodus points to the convergence of those 
two lines at a not very distant date, perhaps during 
some stage of the upper Eocene. 

Menodus. — Menodus was probably not derived from 
Telmatherium, for the incisors of that genus pro- 



rate, Menodus carried dolichocephaly to an extreme, 
yet shows a slow progressive broadening across the 
zygomata {M. heloceras to M. giganteus). 

Mesatirhinus. — The skull of Mesatirhinus as com- 
pared with that of Eotifanops had already lengthened 
all the parts. This dolichocephalic tendency became 



832 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



very pronounced in the most advanced members of 
MesatirJiiniis petersoni and culminated not only in 
the long-skulled Dolichorhinus but even more in the 
allied Sphenocoelus. The feet of the Mesatirhinus 
were derived from those of Eotitanops by a slight 
relative broadening of the carpals, tarsals, and 
metapodials. Dolichorhinus, contrary to former views, 
was not dolichopodal but so far as the evidence shows 
was subbrachypodal, the foot proportions being 



markedly hypsodont and had very flat ectolophs. 
Certain specimens of Metarhinus fluviatilis had a lower 
jaw like that of a minute Megacerops. Not all 
metarhines were small; a lower jaw (Am. Mus. 1859) 
shows, in fact, clear indications of a species of Meta- 
rhinus larger than B. diploconus. 

Rhadinorliinus. — As elsewhere fully shown, Rhadino- 
rhinus in many characters foreshadows the stem form 
of the Brontotheriinae; R. diploconus especially is 







Figure 746. — Models of heads of Oligocene titanotheres showing general proportions, especially form of lip and horns 
A, Brontops (brachy cephalic); -B, Menodus (dolichocephalic); C, Megacerops (hyperbrachycephalic); D, Brontotherium (brachycephalic). 



derived by moderate broadening and little or no 
lengthening from those of forms like Mesatirhinus. 

Metarhinus. — Metarhinus is undoubtedly a near 
ally of Mesatirhinus and even parallels Dolichorhinus 
very closely in many characters of the dentition and 
skull; but in Metarhinus the basicranial and mid- 
cranial regions, after becoming subdolichocephalic, 




iyonyh 



Figure 747. — Separability and imperfect blending of allometric biocharac- 
ters in the facial bones of the horse {Equus caballus, 2), ass {E. asinus, 
(?), and mule 
Left column: Bones of the side of the face, preorbital region, c, The bump on the forehead of the 

horse and mule not observed in the ass. 5, The point at which the section of the nasals is taken. 
Middle column: Nasal bones of the horse, ass, and mule viewed from above, showing that the nasal 

bones of the mule closely approximate in proportions those of the horse, s. Point at which the 

transverse section is taken. 
Right column: Transverse section of the nasal bones of the horse, ass, and mule at the point 

indicated by s, showing the shallow nasal bones of the ass, a pure monophyletie type; the vari- 

abihty ( F'- V) in the depth of the nasals in the horse, which is due to the fact that the domestic 

JE. caballus is a type derived by interbreeding of several distinct races, and the intermediate 

depth of the nasals in the mule. 

ceased to elongate, and the elongation of the face was 
retarded at an early period. The molars became 



distinguished by the short, upturned face, the narrow 
preorbital region, the rounded malar-lacrimal bar, the 
saddle-shaped skull top, the narrow, pointed nasals, and 
many other brontotheriine characters. The mingling 
of dolichocephalic and brachycephalic characters in 
R. diploconus foreshadows a similar mingling in Bronto- 
therium, which has an elongate skull and a short face. 
Possible bearings of the facts stated on phy- 
logeny. — Assuming provisionally the correct- 
ness of this view, the history of the bronto- 
theriine skull may have been as follows: 

1. A mesaticephalic member of Eotitanops 
gave rise to the subdolichocephalic Rhadino- 
rhinus, in which, however, the face was not 
lengthened. The middle and basicranial parts 
of the skull continued to lengthen, but per- 
haps in late upper Eocene time the tendency 
toward brachycephaly gathered momentum 
and resulted in the broad zygomata, the broad 
frontals, and the widely spreading horns of 
Brontotherium. The molars in the Rhadino- 
rhinus stage were relatively long anteroposte- 
riorly, but later they widened rapidly with the 
skull. The premolars also widened, and the 
incipient tetartocones were carried to an ex- 
treme in Brontotherium. 

2. In the allied Megacerops the broadening 
tendency early attained predominance, result- 
ing in a broad, short-faced skull. 

3. The changing proportions of the feet in 
the same phylum were hypothetically as fol- 
lows: The narrow foot of Eotitanops gave 
rise, chiefly by an increase in size but also by 

some degree of broadening, to the narrow foot of 
Rhadinorhinus ; the narrow foot of Rhadinorhinus may 



/%/f 



Horse 



CAUSES OF THE EVOLUTION AND EXTINCTION OP THE TITANOTHERES 



833 



have broadened out, especially the magnum and 
astragalus, into the broad foot of Brontotherium. 

4. These examples, for which detailed evidence is 
given elsewhere, thus illustrate the supposed inter- 
mingling and alternate triumph of tendencies toward 
lengthening and broadening of the parts in the various 
phyla of titanotheres. 

IRREVERSIBLE AND REVERSIBLE EVOLUTION OF 
ALLOMETHONS 

The marked trend in certain directions which we 
have termed "progressive brachycephaly, " "progres- 
sive dolichocephaly, " and "cyptocephaly" in some 
cases runs to an extreme, so that the skull reaches a 
limit of broadening, of lengthening, or of bending that 
is apparently an inadaptive extreme — an assertion as 
to adaptation, however, which may be a matter of 
opinion or of theory rather than a statement of 
established fact. The broadening and lengthening 
tendency seems certainly to be cumulative in succes- 
sive generations, affecting the skull as a whole. 

Such a tendency, however, does not apply to parts 
of skulls in which reversible allometry is observed. 
On comparing the skull of Brontotherium leidyi with 
those of all lower Eocene Perissodactyla we see at 
once that there has been a marked lengthening of 
the cranium proper. On comparing the skull of 
this form with that of its late successor B. platyceras 
we see that the progressive lengthening has either 
ceased or has been overshadowed by a marked broad- 
ening. Here, then, is an example of lengthening that 
has been superseded by broadening. A similar proc- 
ess is seen in Brontops. Although Brontops hrachy- 
cepJialus has a I'elatively longer cranium proper than 
the primitive Perissodactyla, the lengthening was later 
overshadowed by the broadening, both of the skull 
top and of the zygomata, that culminates in B. rohustus 
and " Diploclonus" amplus. 

The observation that broadening of the skull or 
certain of its parts sometimes succeeds lengthening is 
paralleled in the evolution of the feet, where gravi- 
portal (weight-bearing) proportions often succeed 
subcursorial (slender) proportions. This reversible 
mode of evolution in proportions is of very wide appli- 
cation in the descent of the perissodactyls and other 
ungulates. Once admitted as a working hypothesis it 
clears up many hitherto confused problems — the 
structurally ancestral position of the light-limbed 
HyracJiyus to the heavy-limbed amynodonts and 
rhinoceroses, of the narrow-skulled, light-footed Eoti- 
tanops to Palaeosyops and perhaps to Teltnatherium, of 
the light-footed MesatirJiinus to the relatively broad- 
footed DolichorJiinus . If the principle is valid that 
in graviportal types broadening succeeds lengthening, 
then there is less difficulty in deriving the broad-footed 
Oligocene titanotheres with their very broad magnum 



from the more narrow-footed Eocene types with a 
narrow (that is, primitive perissodactyl) magnum. 
(See Gregory, 1910.1, pp. 386, 392-394, 450.) 

SEPAEABIIITY AND CORRELATION OF BIOCHARACTERS 

SEPARABILITY OF ALLOMETRONS IN HEREDITY 

The above-described differential modes of evolution 
in the titanotheres furnish presumptive evidence of 
the germinal separability of allometrons in their 
velocities of change as biocharacters. The fact that 
every allometron, like every rectigradation, has a 
more or less distinctive time of appearance and velocity 
of change shows that it is also connected with separable 
germinal genes. This conclusion is confirmed by 
hybridization in recent mammals. 

The central idea of Mendel's discovery is that when 
two contrasting tendencies (such as dolichocephaly 
and brachycephaly) enter a hybrid, one from each 
parent, they separate in the germ cells of the hybrid, 
so that some of the germ cells are purely dolichocephalic 
and others are purely brachycephalic, like those of the 
original parents. Chance meetings of these germ cells 
give rise to ratios characteristic of Mendelian heredity. 

The very important question whether the "broad 
heads" and the "long heads" as allometric biochar- 
acters have a germinal separability like that of the 
"tallness" or the "shortness" observed by Mendel 
in his classic experiments on the pea plant can not 
be determined in the titanotheres but can be inferred 
from crosses between "long heads" and "medium 
heads" in the related horse family, the Equidae. 
Here, as pointed out by Osborn (1912.372, pp. 
177-190), the dolichocephalic skull of Equus caballus, 
when crossed with the relatively mesaticephalic skull 
of E. asinus, gives in the mule a dolichocephalic skull 
and not an intermediate, blended form. 

In crossing the female horse with the male ass 
it is seen that the skull of the resulting mule has the 
dominant dolichocephalic proportions of the horse, 
with slight blending or intermediate proportions of 
certain bones (like the lacrimal) between the forms 
observed in the horse and ass. 

Careful comparison of the proportion indices in the 
hybrid mule proves that it inherits from its mother, 
the horse, very closely the following: 

1 . Its cephalic index ; a long, narrow skull as a whole. 

2. Its diastema index; a long diastema, or space for, 
the bit. 

3. Its craniofacial index; a relatively short cranium 
and a very long face. 

4. Its orbital index; a long, oval orbit. 

5. Its molar index; a relatively long and narrow set 
of grinding teeth. 

6. Its occiput-vertex angle index; a vertically placed 
occiput. The ass has a much more ancient type of 
skull than the horse. 



834 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Thus the proportions of the horse skull acquired 
through a long process of allometric evolution prove 
to be separable biocharacters in the germ. This also 
proves (Osborn, 1912.372) that proportional charac- 
ters, although evolving continuously, may exhibit the 
same germinal separability as the suddenly appearing 
saltations and mutations of De Vries. Out of 28 
characters that were carefully compared in the skull 
and the teeth of the horse, the ass, and the mule 18 
found in the mule are obviously derived either from 
one parent or the other and show very slight tendency 
to blend, whereas 10 characters show a distinct tend- 
ency to blend. 

Yet, so far as our observation shows, rectigradations 
are more decidedly and distinctly separable in the 
germ than allometrons. 

CORRELATION, COORDINATION; COMPENSATION OF 
RECTIGRADATIONS AND ALLOMETRONS 

In the evolution of rectigradations and allometrons 
all the biocharacters are harmoniously adjusted to 




Figure 748. — Contrasts between the hypotheses of Lamarck (A), 
Weismann (B), and Osborn (C) regarding the causes of evolution 



such as the development of the horns at the expense of 
the nasal bones. Such compensation correlation is 
comparatively rare in the titanotheres in contrast with 
other quadrupeds. 

4. Sex correlation or linkage with male or female sex 
of proportion biocharacters and to a less extent of 
rectigradation biocharacters, as in the size of the 
canine teeth, the size of the horns, the proportions of 
the skull, the varying velocities of ontogenetic and 
phylogenetic changes in the dentition and all parts of 
the skeleton, the brachycephaly of males and mesa- 
ticephaly of females of the same species. 

5. Germinal correlation, by which titanotheres of 
common ancestry exhibit similar evolutionary tend- 
encies, as in progressive brachycephaly and doli- 
chocephaly, acceleration and retardation. 

6. Psychic correlation, as of the horns, with psychic 
predisposition to use the horns, as explained above in 
a consideration of ontogenesis of the horns in cattle. 

The accompanying diagram (fig. 748) represents 

crudely the extreme contrasts between the Lamarckian, 

the Darwinian, and the tetrakinetic points of 

f,ypot/,es,s view. 

THEORETIC CAUSES OF THE EVOLUTION OF NEW 
CHARACTERS AND NEW PROPORTIONS 

THEORIES ADVANCED TO EXPLAIN THE ORIGIN 
OF RECTIGRADATIONS AND ALLOMETRONS 

Having compared some of the modes of the 
evolution of the titanotheres with those of the 
evolution of other mammals, we may now con- 
sider theories that have been advanced to explain 



The Lamarckian conception of causes is chiefly external, centripetal; the Darwin- Weismann the tWO kinds of evolution upon which OUr atten- 

conception of causes is chiefly internal, centrifugal; the tetrakinetic hypothesis is chiefly tion is here Centered bv rCCtiffradationS and bv 

internal-external, or both centripetal and centrifugal. The balance of existing evidence is in ,. o • * j • 'Z! ii 

favor of B, the centrifugal hypothesis, since the tetrakinetic theory rests only upon inference ailomctrons. oetting aside aS unsCientlllC all 
from the observed modes of evolution rather than upon actual observation of the external, 
centripetal origin of any single somatic character. 



serve the organism as a whole. The adjustments are 
of six different kinds, as follows: 

1. Mechanical correlation of the rectigradations and 
allometrons in the upper and lower grinding teeth. 
The reason for this continuous and perfect adjustment 
is perhaps the most difficult to comprehend in the 
whole range of titanothere evolution. 

2. Mechanical correlation of the developing horns, 
head, shoulders, and entire locomotor skeleton, with 
the use of the horns as offensive and defensive struc- 
tures, with the use of the lips and grinding teeth in 
the prehension and comminution of food, with the 
development of the limbs, partly for offense and de- 
fense, partly for locomotion. In this general correla- 
tion the evolution of various allometrons and recti- 
gradations is continuously and perfectly adjusted. 

3. Proportion compensation, a mode of correlation 
and coordination, by which the development of one 
character is effected through the sacrifice of another. 



ideas of "vitalism" and "internal perfecting 
tendencies," we may group all modern theories 
of the evolution of form and function under three 
heads, as follows: 

1. External initiation and causation (Lamarckism) ; 
modifications of the environment and of the soma. 
Modifications introduced by change of environment 
and change of habit are in some manner impressed 
upon the germ, so that they reappear more or less 
fully developed in offspring. This idea is inherent in 
all theories of "direct action of environment" and 
"inheritance of acquired characters," as developed by 
Buffon, Lamarck, Spencer, Cope, and others. This 
general interpretation may be refeiTcd to as the 
Lamarckian. 

Evidence: According to this theory germinal varia- 
tions should be found to originate, to follow, and to 
correspond closely with individual somatic modifica- 
tions. They should be orthogenetic and should 
invariably be followed by similar changes inherited 
in the bodies of descendants. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



835 



2. Internal initiation and causation (Darwinism): 
Chance, accidental, fortuitous heritable variations in 
the germ, minor germinal saltations, germinal muta- 
tions (De Vries), predispositions that give rise to 
favorable or unfavorable alterations of character and 
function, of survival value, accumulated under the 
action of natural selection or natural elimination. This 
is the pure Darwinian interpretation. 

Evidence : According to this theory variations should 
originate no adaptive tendency except through the 
cumulative action of selection in a given direction; 
in their first appearance variations should be indefinite, 
indeterminate, and should trend in all directions. 

3. Internal-external initiation and causation (tetra- 
kinesis) : According to this theory evolution is neither 
chiefly external (environmental, ontogenetic) nor inter- 
nal (germinal) but is due to the combined action of 
internal and external causes accumulated by natural 
selection. This theory (Osborn), which is based upon 
the data of paleontology, experimental zoology, and 
physiology, contains certain elements of both the 
Lamarckian and the Darwinian interpretations, 
although it is unlike either. This may be referred 
to as the principle of the inseparable action of four 
factors in development (Osborn, 1908.308), as the 
theory of tetrakinesis in evolution, or as the theory 
of the action, reaction, and interaction of four com- 
plexes of energy under the continuous operation of 
natural selection (Osborn, 1917.463). 

Evidence: According to this interpretation variation 
should move chiefly along definite lines; it should be 
orthogenetic; but, unlike the procedure hypothesized 
by Lamarckian interpretation, variations may or may 
not be preceded by similar bodily modifications. 

ANALYSIS OF THE EVIDENCE ON THE MODES OF ORI- 
GIN OF VARIATION AS CONSIDERED IN DARWINISM, 
LAMARCKISM, AND TETRAKINESIS 

It is obvious that sharp, clear, unbiased observation 
and analysis of the modes of origin of variations may 
have a crucial bearing on the choice between the three 
theories stated above, of the causes of evolution. 
For example: 

1. If in their germinal variations — rectigradations 
and allometrons — mammals are observed invariably 
to follow antecedent bodily (somatic) and environ- 
mental modifications, such modes of origin would 
constitute strong evidence for the Lamarckian theory. 
It will be pointed out, however, that germinal varia- 
tions do not follow antecedent bodily modifications 
(somations) as a rule but only in certain instances. 
These exceptions cast strong doubt on the pure 
Lamarckian interpretation. 

2. If new biocharacters (rectigradations) are her- 
alded in indefinite, indeterminate, accidental varia- 
tions, by the rudiments of new, spontaneously ap- 
pearing characters at various points in the teeth and 

101059— 29— VOL 2 10 



skeleton, if from these unstable origins stability and 
adaptation are gradually attained by the survival of 
certain characters and the elimination of others, such 
modes of origin would constitute strong presumptive 
evidence in favor of the Darwinian interpretation — 
namely, that the adaptive arises by selection out of 
fortuitous variations. If, on the contrary, minor 
organs, saltations, fluctuations, sports, are observed 
to have no relation to the main trend of adaptation, 
this would tend to show that the Darwinian hypothesis 
affords an inadequate explanation of the evolution 
of the titanothere. 

3. If the modes of variation we have been observing 
in the titanotheres — rectigradations and allometrons — 
are in the main definite, determinate, and generally 
adaptive in direction but do not invariably follow the 
direction of individual somatic modifications, there 
must be some cause or complex of causes other 
than those afforded exclusively by either the pure 
Lamarckian of the pure Darwinian hypothesis. 
These rectigradations and allometrons are the modes 
of variation, observed in the titanotheres and other 
mammals, which have led Osborn to propose and to 
develop the theory originally termed (Osborn, 1908. 
308) "the four inseparable factors of evolution," 
more recently termed "the theory of tetrakinesis" 
(Osborn, 1917.462). 

OBSERVED PRINCIPLE OF TETRAPLASTIC DEVELOPMENT 
OF BODY form; THEORETIC PRINCIPLE OF THE 
TETRAKINETIC EVOLUTION OF THE GERM 

Basis of the tetraJcinetic theory. — The tetraldnetic 
theory,- which has been developed gradually by the 
author between the years 1893 and 1917,^' is based 
upon the modern conception that the visible bodily 
evolution of the titanotheres is the result of the in- 
visible germinal evolution of the titanotheres; conse- 
quently that the modes of bodily evolution reflect 
the modes of germinal evolution. If the modes of 
bodily evolution are orderly, determinate, generally 
adaptive in direction (entirely distinct from the trial 
and error modes of pure Darwinism), there must be 
causes of this orderly evolution of the germ. The 
conclusion is reached that these causes are not purely 
internal, germinal, nor purely external, environmental, 
somatic; that they are in some manner internal- 
external. 

The fourfold principle of development. — It is observed 
that in the development of its bodily form and func- 
tion each titanothere is tetraplastic in the sense that 
from the stage of the fertilized germ to adult age it is 
plastically molded by the action, reaction, and inter- 
action of four centers of influence, namely, that con- 
tained in the heredity germ, that found in the 
developing organism, that of the life environment, 

Q2 The special papers and researches in which this principle and theory are treated 
are listed in the bibliography at the end of this chapter. 



836 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



and that of the physical environment. If these four 
influences are all normal (typical) we shall have a 
typical (normal) titano there; but if any of these 
influences is disturbed or altered we shall have an 
at3''pical (abnormal) titanothere. Therefore the true 
interpretation of every titanothere, composed, as it 
is, of hundreds or thousands of biocharacters, is that 
each biocharacter is subject to the normal or abnormal 
condition of these four factors of development. This 
is the principle that was erroneously termed by the 
author "the four inseparable factors of evolution." 
The use of the term "evolution" was an error; the 
principle should be termed "the four inseparable fac- 
tors of somatic development," or the principle of 
tetraplasv. 

The theory of the fourfold principle of evolution. — The 
relation of this principle of tetraplasy in development 
to the subsequently formulated theoiy of tetrakinesis 
in evolution is as follows: The theory of tetrakinesis 
applies to germinal evolution as distinguished from 
somatic development; it is based upon the premise 
that unless the causes of germinal evolution are wholly 
internal and inherent in the germ itself, a supposition 
which is not supported by the modes of evolution 
actually observed, they must be internal-external; they 
must consist of some kind or degree of physico- 
chemical (that is, energetic) action, reaction, and 
interaction between environmental, somatic, and 
germinal changes which remain to be discovered. 
Moreover, as aU physiological and functional rela- 
tions thus far observed in mammals are either physi- 
cal or chemical, and as aU physico-chemical relations 



represent phenomena of energy, we are impelled to 
believe that the relations between the physical envi- 
ronment, the life environment, the developing organ- 
ism, and the germ will prove to be "energetic." 
Hence our theory is termed tetrakinetic (rerpa, four; 
KLvrjTTjs, energy) and may be expressed as follows: 

In each titanothere the phenomena of life represent 
the action, reaction, and interaction of four complexes 
of physico-chemical energy, those of (1) the geographic 
environment; (2) the developing organism (proto- 
plasm, germ chromatin, of the body cells); (3) the 
germ chromatin of the reproductive cells, seat of 
heredity; (4) the life environment. Upon the resultant 
actions, reactions, and interactions of latent potential 
(stored) and kinetic (active) energy in each organism 
natural selection is constantly operating wherever 
there is competition with the corresponding actions, 
reactions, and interactions of other titanotheres and 
other organisms. 

According to this energy conception of evolution 
Darwin's principle of natural selection is not set aside 
but remains extremely important; it is constantly 
operating not only between organisms as a whole but 
on their separate parts, and especially upon their 
separate manifestations of energy, that have or lack 
survival value, as shown in the following summary 
representing the interchange of energy in the develop- 
ing titanothere and between it and rival or competing 
titanotheres and other animals. In each animal all 
the individual changes noted in paragraphs 2, 3, and 4 
take place in competition with similar changes in its 
rivals, under Darwin's law of natural selection. 



Diagrammatic exposition of the development of organisms under the tetralcinetic theory 



Organism A (the developing titanothere) 

affected by the actions, reactions, and 
interactions of 

1. Physical environment: Physico-chem- 

ical energies of space, of the sun, 
earth, air, and water. 

2. Titanothere organism: Physico-chem- 

ical energies of the developing indi- 
vidual in the tissues, cells, proto- 
plasm, and cell chromatin of all 
biocharacters. 

3. Titanothere heredity germ: Physico- 

chemical energies of the heredity 
chromatin, included in the repro- 
ductive cells and tissues, the predis- 
positions of all biocharacters. 

4. Life environment: Physico-chemical 

energies of other organisms. 

In explanation of the above summary it may be said 
that in development the principle of tetraplasy is well 
established; it rests upon a great variety of observa- 
tions on the physiology and anatomy — the function 
and form — of many kinds of plants and animals. 
According to this principle every biocharacter derives 
from the germ the potential impulse or energy to 



A and B-Z mutually affected through the 
operation of 

Darwin's law 

of 

natural selection, 

survival of the fittest: competition, and 
the selection or elimination of energies 
and forms having or lacking survival 
value. 



Organisms B -Z (rival or competing titano- 
theres and other animals) 
affected by the actions, reactions, and 
interactions of 

1. Physical environment: Physico-chem- 

ical energies of space, of the sun, 
earth, air, and water. 

2. Animal organisms: Physico-chemical 

energies of the developing indi- 
vidual in the tissues, cells, proto- 
plasm, and cell chromatin of all 
biocharacters. 

3. Animal heredity germ: Physico-chem- 

ical energies of the heredity chro- 
matin, included in the reproduc- 
tive cells and tissues, the predisposi- 
tions of all biocharacters. 

4. Life environment: Physico-chemical 

energies of other organisms. 



reproduce its typical ancestral form or function in 
the most minute detail. Each biocharacter is tetra- 
plastic {reTpa, four; TrXatro-oj, to mold) in the sense 
that its form is molded by the four influences of the 
germ, the individual development (ontogeny), the 
physical environment, and the life environment. 
These influences are complexes of energy and matter 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



837 



which constantly keep in touch with one another 
through physico-chemical action, reaction, and inter- 
action. An atypical or abnormal condition, such as 
excess, depletion, or defect in any one of these four 
complexes of energy, produces disturbances of some 



jOia^ram skowiny tfiat 
lb attain, normtxc 
cle\^elopment CA ) 
a// characters require t/ie 
normal energy of Nered/ty, 
of friYtronment, 
of life En..-ronn,ent, 
and of Ontogeny 


I 

1 


1 

1 


^ 

1 

^ 


1 
1 

1 


,1 

S 

1 






A 


B 


c 


D 




//- 


H 


/^^ 


0-^ 


/> 


Z^ 








^ 


,1 


1 




Qerminal potency 
J-^rediSpoSitions of charac^/ 






P 


~Rtysico - cAemi-ca./ 
^peograp?nc.XUmat.c Infiueace, 






Tunct{orzaUrTfluenc<^s, ^tc. 
ComLmed supernorTnal 
actions^ reactions, and 
interactions of heredity, 
Etii^ironment, and Ontogeny 
■favoraMe 'to o\>'erdei^elopment 
of a proportionai character - 






1 


1 

3 




'* 


'' 






yktraplasy 


//entaite 

Qerm 

Characters 


Abn-Z/eriTaik 

Somatic 

Characters 



Figure 749. — Cumulative or favorable influence 
of heredity, ontogeny, and environment 

kind, and the visible biocharacter ceases to be typical 
or normal. If such disturbances of normality affect 
many phases in the development of the biocharacter, 
the visible organism will be extremely abnormal. 
Thus to abnormal conditions in the germ, in the on- 



of the fourfold influences may convey its separate 
or its combined effect upon abnormality. It is 
equally true that departures from the typical con- 
dition which may coincide with the adaptive direction 
of the evolution of the biocharacter may again be 
fourfold — that is, they may be attributable to favorable 
predispositions in the germ, favorable modifications 
and accommodations of the soma, favorable influences 
of the environment in the form of physico-chemical 
influence, favorable influences of the life environ- 
ment in the form of food or of competition, which 
may favor the development of certain biocharacters, 
as, for example, the so-called habitudinal (Gulick) 
and organic (Osborn, 1897.125) selection. 

As regards evidence, the theory of tetrakinesis in 
evolution is quite distinct from the principle of tetra- 
plasy in development. It is in the nature of a trial 
hypothesis as to the causes of evolution, because we 
do not yet understand the influence, respectively, of 
ontogeny, environment, life environment, and selec- 
tion upon the evolution of the germ. Therefore the 
theory of tetrakinesis awaits further observation and 
experiment as to the respective influence of physico- 
chemical actions, reactions, and interactions on spe- 
cific biocharacters of the germ. 

The following formulae, suggested by Gregory 
(Osborn, 1912.378, p. 307), present clearly the manner 
in which we may analyze the possible causes of incre- 
ment — for example, quantity, intensity, plus and minus 
variations — in any visible biocharacter wherever we 
observe and compare members of a series. 

QUANTITATIVE INCREMENT OF THE FOUR SEPARABLE 
FACTORS IN DEVELOPMENT AND EVOLUTION 

Let T represent the typical or normal condition of a 
given biocharacter in the ancestral species as to recti- 



Tetraplasy Actions, Tfeactions, and Interactions 

-Diocharacters, 'Predispositions, and T^otentiatities of the Qerm 




Figure 750. — Diagram illustrating the tetraplastic theory 

Conception of a centrifugal [-^] stream of potential genes, determiners, predispositions, heritages, as passing into the life of the organism from 
the heredity germ, both into the new reproductive organs and into every cell of the body (chromatin), and as acting, reacting, and 
interacting with the inorganic environment, the life environment, and the ontogeny of the organism itself. At present there is com- 
paratively little direct evidence of a centripetal [<-] or reversed current of energy such as to eflect specific changes in the heredity 
chromatin. 



togeny, in the physical environment, or in the life 
environment, or in all combined, are attributable 
the visible abnormalities of the biocharacter. Each 



gradation, allometry, area, volume, degree, or inten- 
sity of coloring, strength of muscle, or any biochar- 
acter capable of quantitative determination. Let Tn 



838 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



represent the average condition of the same bio- 
character in a species or mutation descendant from T. 
Then Tn — T is the measure of the evolution of the 
biocharacter, and 100 (Tn — T) T is its percentage in- 
crement. Hence 



^"=t+(4)t 



where Ynr, represents the total percentagii increment. 

Suppose that this total percentage of increment in 
the biocharacter were made up of the following con- 
tributory increments: 

H (heredity), that part of the total percentage 
increment which may be ascribed hypothetically to 
an orthogenetic or germinal tendency to accelerate, 
balance, or retard the typical condition of T. 

(ontogeny), that part of the total percentage incre- 
ment wliich may be ascribed hypothetically to modi- 
fications arising in the soma, evoked either by use and 
disuse or by physico-chemical correlation (interaction) 
during individual development. 

E (environment), that part of the total percentage 
increment which may be ascribed hypothetically to the 
reaction to the geographic and physical environment. 

L.E (life environment), that part of the total per- 
centage increment which may be ascribed hypo- 
thetically to the action of surrounding titanotheres 
and other organisms. 

S (selection), that part of the total percentage 
increment which may be ascribed to the cumulative 
effect of natural selection upon germinal variations, 
fluctuations, etc. 

Substituting the above symbols in the formula 

we obtain the formula 

/ H + + E + L.E + S \ 

^° ^"^V 100 J 

whence 

'Tn-T' 



Tn = T- 



H + O + E + L.E = 100 + S 



{^) 



In paleontologic research on certain proportion 

biocharacters in continuous phyla one may be able 

T 
to determine the numerical value of Tn — m, and 

hence to obtain an exact measm-e of the total per- 
centage of increment due to the combined action of 
H + O + E + L.E + S taken together, but not to measure 
any one of these factors separately. 

In the application of the tetrakinetic theory it is 
the causes of the invisible, germinal increment which 
we have to explain — the increment, for example, 
which separates the germ of Brontotherium from that 
of Eotitanops. 

It is not claimed that the theory of tetrakinesis, in 
which the incessant action of natural selection plays 
a large part, will explain the origin of a single bio- 



character, rectigradation, or allometron. It is still 
in the stage of a working hypothesis, to be tested by 
observation and experiment, in the way, for example, 
that the original Lamarckian and Darwinian theories 
have been and are being tested. The only insight 
we now have into the possible worldng of this theory 
is afforded by the phenomena of organic selection and 
of interaction in the changes in the velocity of certain 
biocharacters which are known to be due in the soma 
to the circulation in the system of "physico-chemical 
messengers" of the kind designated among physiolo- 
gists as enzymes and internal secretions, including 
the hormones (accelerators) and chalones (retarders) 
of the individual growth of certain biocharacters. So 
far as known at present these are purely somatic phe- 
nomena. Experiment may prove that similar inter- 
Tetrajola sy 
Centrifu.aal stream of ener(^y 



envtro^^ 




Figure 751. — Diagram illustrating the 
principle of tetraplasy 

Represencing the conception of a centrifugal stream of 
potential heritages and predispositions passing from 
the heredity germ into the life of the organism (ontog- 
eny), both into the new reproductive organs and into 
every cell of the body (protoplasm and cell chromatin), 
as acting, reacting, and interacting with all the other 
functions and structures within the body as well as 
with the stream of energy interchanged between the 
body and the physical environment and the life en- 
•vironment, showing that every visible biocharacter 
represents the action, reaction, and interaction of 
these four complexes of energy. 

actions originally arising through adaptive modifica- 
tions of the body (soma) may affect the correspond- 
ing predispositions and potentialities of the germ. ' 

ANALYSIS OF THE MODES OF VARIATION; THEORETIC 
IMPORTANCE OF INITIATION 

We have set forth above an analysis of the modes 
of variation; let us apply this analysis to proportions. 

Until the causes, as yet unknown, of predisposition, 
of increment, of acceleration, of retardation, of recti- 
gradation, of allometry in the germ itself are discovered 
experimentally we must confine our observations to 
what we see going on in the body and must seek to 
answer the all-important question whether the initia- 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



839 



tion of any given biocharacter is environmental, 
somatic, or germinal. If, for example, in all the 
instances observed the inititation of a certain allo- 
metron is primarily somatic and secondarily germinal, 
this fact would be presumptive evidence that in this 
particular biocharacter somatic action, reaction, and 
interaction precede germinal predisposition. Even if 
it were proved that a certain allometron is invariably 
somatic in its initiation and secondarily germinal, it 
would not be logical to assume that all allometry is 
first somatic and then germinal; in fact, we shall point 
out that certain allometrons are germinal from the 
first; they are not, so far as we loiow, initiated in 
ontogeny. Moreover, invariable sequence is not 
invariable cause and effect. 

BEARING OF SALTATION VERSUS CONTINUITY ON 

THE LAMARCKIAN, DARWINIAN, AND 

TETRAKINETIC THEORIES 

According to the Darwinian theory variations should 
be observed as chiefly discontinuous, saltatory, fortu- 
itous, adaptive or nonadaptive, furnishing material 
for natural selection. 

According to the Lamarckian and the tetrakinetic 
theories variations shoidd be mainly continuous and 
adaptive, rarely saltatory and fortuitous. What are 
the observed facts of variation? 

DARWIN's hypothesis of FORTUITOUS SALTATION 
AND FLUCTUATION 

After years of observation of domesticated plants 
and animals under artificial selection, Darwin reached 
the conclusion that there were three kinds of new 
variation characters for natural selection to work 
upon — first, hereditary minor saltations ("minute 
heritable variations"); second, hereditary fluctua- 
tions of proportion; third, somatogenic modifications 
of proportion, which he finally believed (Lamarckian 
theory) to be inherited. 

Darwin (1859.1) held that evolution is due chiefly 
to the natural selection of "heritable individual difl'er- 
ences" — that is, variations — his real meaning as to 
these individual differences being found in the hun- 
dreds of examples he cited in the "Origin of species" 
and in "Variations of animals and plants under 
domestication." Poulton (1909.1, pp. 49-50) re- 
marks as to Darwin: "His observation and study 
of nature led him to the conviction that large varia- 
tions [that is, major saltations], although abundant, 
were rarely selected, but that evolution proceeded 
gradually and by small steps." Plate (1909.1) es- 
tablished clearly that the "individual differences" 
of Darwin are practically identical with the "muta- 
tions" of De Vries. Osborn (1912.362, pp. 76-82) 
held that Darwin's conception was that evolution 
develops chiefly through the natural selection of minor 
saltations; but that Darwin's "individual differences" 



are in the nature of both major and minor saltations, 
structural or functional, and always hereditary, as 
is shown in the observations that he assembled in 
commenting on the genesis of the race horse and of the 
greyhound, breeds which he used by way of illustra- 
tion of the genesis of new forms in nature. In con- 
sidering these breeds he pointed to such suddenly 
appearing new characters as horn rudiments, tail- 
lessness, curlmess of the hair, characters which are 
"discontinuous" in Bateson's sense, "mutations" in 
that of De Vries. Intermingled with these minor 
saltations Darwin cited others which are obviously 
reversional. That he believed chiefly in the accumu- 
lation of favorable minor saltations there can be no 
question, but, for the admirable reason that no 
evidence had been adduced in nature of evolution by 
major saltations, he rejected the hypothesis, which 
originated with Geoffroy St. Hflaire, of the appear- 
ance under certain environmental conditions of en- 
tirely new types of animals and plants or of new pro- 
foundly modified organs. 

In brief, Darwin held chiefly that by cumulative 
natural selection of minor saltations a character could 
slowly be shifted in an adaptive direction. These 
minor saltations were in his opinion the fortuitous 
or chance material out of which nature "selects" 
its adaptations, utilizing the adaptive and rejecting 
the inadaptive. He guarded the word "chance," 
however, by stating that it might merely cover our 
ignorance of the unlcnown causes of variation. 

Darwin also believed in the natural selection of 
heritable fluctuations of proportion, as illustrated in 
his classic rebuttal of the Lamarckian explanation of 
the mode of origin of the long neck of the giraffe, 
namely: 

So under nature with the nascent giraffe, the individuals 
which were the highest browsers and were able during dearths 
to reach even an inch or two above the others will often have 
been preserved; for they will have roamed over the whole 
country in search of food. * * * These slight proportional 
differences will favor survival and will be transmitted to off- 
spring. ("Origin of species," ed. of Appleton, 1909, p. 27.) 

Naturally Darwin could not draw such sharp dis- 
tinctions in the definition of variation either in 
language or in example as we may to-day, profiting 
as we do by the 50 years of experiment and analysis 
that have passed since his time. The chief emphasis 
in the above passage is in the words "slight pro- 
portional differences," differences that we now classify 
as fluctuations or fluctuating variabilities. 

In conclusion, critical reexamination of Darwin's 
writings leads us to dissent entirely from the influen- 
tial opinion of De Vries that there was always a doubt 
in Darwin's mind as to whether (1) the selection of 
minor saltations or (2) the selection of fluctuations 
played the larger part in the origin of species. The 
actual examples that Darwin cited and his repeated 
emphasis prove that minor saltations like the De 



840 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Vries mutations were chieflj^ iu his mind. This 
raises the inquiry as to tlie normal or abnormal char- 
acter of major and minor saltations. 

MOST SALTATIONS IN MAMMALS ABNORMAL 

Are saltations natural or artificial products? Are 
they normal or abnormal? Against the repeatedly 
observed continuous origin of normal adaptive recti- 
gradations and allometrons in titanotheres and other 
quadrupeds still stands the pure Darwinian hypothesis 
of the fortuitous and discontinuous origin of similar 
biocharacters as material for natural selection. 



It is true that some characters, such as horns, do 
arise suddenly. It is therefore interesting to compare 
20 distinct types or kinds of major and minor salta- 
tions in 11 different families of mammals — man, 
horses, cattle, sheep, deer, pigs, dogs, cats, rabbits, 
guinea pigs, mice, in which saltations that are 
closely or exactly similar repeatedly occur. Our chief 
authorities are Allen, Azara, Bateson, Brinkerhoff, 
Castle, Darwin, Davenport, Haeckel, Percival, Poul- 
ton, Kidgeway, Koot, Seton, Sutton, Twining. The 
accompanying table presents the very impressive 
results obtained by this comparison. 



Comparison of twenty saltations in eleven Tcinds of mammals 





Man 


Horses 


Cattle 


Sheep 


Deer 


Pigs 


Dogs 


Cats 


Rabbits 


Guinea 
pigs 


Mice 


Neoproportions (allometrons): 




X 
X 








X 
X 














X 










Horn saltations, horn biocharacters: 

3. Sudden development of horns in horn- 


X 


X 








X 






4. Sudden disappearance of horns in horned 


X 


X 
X 
X 


























1 


1 
















1 














X 






1 






Loss of biocharacters: 

8. Earlessness (absence of both external 








































X 






10. Taillessness (absence of caudal verte- 


X 


X 
X 


X 

X 
X 

X 


X 








X 


X 










X 


X 


Saltatlons of form: 

12. Excessivelj' fine or silky hair 


X 


X 






X 


X 


X 


X 










14. Curled hair over the entire body 


X 
X 
X 
X 

X 


X 
X 


X 
X 






X 
X 


1 








X X 


X 




16. Local or general epidermal thickenings. _ 






X 


X 
X 


X 
X 
X 






X 

X 
X 
X 




1 






Foot biocharacters: 

18. Polydactylism (supernumerary digits)... 

19. Syndactylism (consolidation of paired 


X 


X 


X 


X 




X 




Jaw biocharacters: 


X 




X 































Teratology and Mendelism have alike revealed the 
fact that most saltations represent failures or abnor- 
malities in the germinal mechanism. Some occur 
under natural conditions, but most of them under arti- 
ficial conditions of environment. Many of these 
saltations are dominant characters in heredity, and 
in some domestic breeds, through either artificial 
selection or failure of artificial selection, they become 
fairly numerous. We have tailless cats, solid-footed 
(syndactyl) pigs, polydactyl cats, curly-haired horses, 
short-legged sheep (ancon), and hornless cattle of 
several breeds (Angus, niata). It is possible also 
that races of . wild animals have been established 



through germinal saltations, such as the four-horned 
antelope of India {Tetracerus) , which is comparable 
to the four-horned domestic breed of sheep. 

Paleontologic and anatomical research on the 
mammals for over a century has failed to demonstrate 
in a state of nature a single example of the sudden 
origin of such important characters as the horns. 
There is reason to believe that if peculiar or anomalous 
mammals do arise under natural conditions they are 
driven away from the herds and not allowed to breed. 
Thus it may be said that in neither zoology nor 
paleontology has any evidence been adduced thus far 
of the saltatory origin of allometrons or rectigrada- 
tions under natural conditions. 



CAUSES or THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



841 



Bateson (1894.1) revived the saltation hypothesis 
of Darwia as the exclusive mode of the origiri of 
species. Saltationists, now known as mutationists 
(of De Vries), refer to Bateson's work as laying the 
sure foundations for the theory of the origin of species 
through heritable saltations. At the time Bateson's 
work appeared it suffered a searching review from 
Scott (1894.1, pp. 355-374). Matthew (Osborn, 
1912.372) examined it critically in the light of ver- 
tebrate paleontology and reached the following 
conclusions : 

Of the 323 cases of discontinuity cited in mammals the 
greater part are obviously teratological and have no direct 
significance in relation to paleontologie evolution except for 
a very few instances such as the supernumerary or fourth molar 
teeth of Otocyon. While not significant [in evolution] these 
teratological cases are interesting because they show the 
prevalence of homoeosis and indicate that many of the remain- 
ing cases which might [otherwise] be considered normal salta- 
tions or reversions may actually be teratologic, but disguised by 
homoeosis; all of the possibly significant cases (such as the 
supernumerary molars) are thereby placed under suspicion. 
Setting aside this suspicion the minority of the "significant" 
cases in teeth and feet may be said to afford evidence of the 
meristic variability of vestigial and rudimentary structures. 
Bateson's statement that such variability is related not to 
nonfunctionalism but to terminal position in a series appears 
to me directly in conflict with his [Bateson's] own evidence, 
as it certainly is with all my experience. This accords with 
commonly observed data in paleontology, for no paleontologist 
would question that vestigial teeth or bones are apt to [finally] 
disappear by "discontinuous" evolution. As to the appearance 
by salatory evolution of new and primarily functional parts in 
teeth or feet, I know of no adequate paleontologie evidence in 
its favor. It is either demonstrably false or decidedly improb- 
able. In the cases of supernumerary teeth {Otocyon, Myrme- 
cobius, Cetacea, etc.) saltatory evolution may be regarded as 
reasonable in default of any paleontologie evidence to the 
contrary. Meristic or numerical evolution in fully functional 
vertebrae is intrinsically probable as the only method of evolu- 
tionary change. 

The fact that so many cases of supernumerary teeth are 
associated with asymmetry throws doubt on the significance 
of all such cases; asymmetric variations and those occurring 
only in upper or only in lower teeth have no analogy in paleon- 
tology; such cases as occur abnormally are recognized as of a 
different and nonsignificant class than normal evolutionary 
changes. 

Summary of Bateson's 323 cases 





Nonsignifi- 
cant 


Possibly 
significant 


Variations 
vestigial 
or rever- 
sional 


1 . Vertebrae . _ . _ 


17 

83 

110 


27 
10 




2. Teeth 


76 


3. Feet _ _ ._ . ._ 












210 ! 37 


76 



The summary of our conclusions is as follows: 
Of the 323 cases cited by Bateson of discontiuuity 
in the vertebrae, the teeth, and the skulls of modern 
mammals, 286 are abnormal, teratological, reversional, 



and have absolutely no significance in evolution. Ten 
cases of additional fourth molar teeth are possibly of 
significance, because among the mammals there are a 
few genera with fourth molars, which probably have 
arisen by saltation, as in Otocyon and certain African 
races of Homo. Of aU the cases cited by Bateson 
there remaia only 27 which may be ranked as prob- 
ably significant — that is, of the Idnd which may be of 
actual importance in normal evolution. These are 
the addition or the reduction of the number of verte- 
brae ki the spinal column, which is of real significance 
because of the well-lvnown variations in the vertebral 
formulae observed among the different genera and 
even species of mammals, such as Equus caballus 
(Nordic breed), with six lumbar vertebrae, and E. 
caballus (Arab breed), with five lumbar vertebrae. 
The occasional "origin of species" through vertebral 
saltation is also rendered probable by the fact that 
vertebrae can be added to or subtracted from the 
spinal column only discontinuously — that is, through 
germinal saltation. 

This evidence, in its bearing on the principle that 
conspicuous major saltations are not a part of the 
normal evolution of mammals, also throws prima 
facie doubt upon the less conspicuous, minor salta- 
tions. It is these saltations which, as stated above, 
Darwia believed to be among the chief materials out 
of which adaptations were accumulated through the 
action of natural selection. The only laiown differ- 
ence between major and minor saltations is one of 
degree, not of kind. 

The fact that the vast majority of germinal anoma- 
lies cited in Bateson's researches have no significance 
in evolution imder natural conditions, with the excep- 
tion of the sudden addition or disappearance of com- 
pletely adaptive organs, like vertebrae and teeth, also 
throws many minor germinal saltations under sus- 
picion as material for natural selection, important as 
they doubtless are in artificial selection and hybridi- 
zation. This opinion may be contrasted with that of 
Punnett (1911.1, p. 15): 

Speaking generally, species do not grade gradually from one 
to the other, but the differences between them are sharp and 
specific. Whence comes this prevalence of discontinuity if the 
process by which they have arisen is one of accumulation of 
minute and almost imperceptible differences? Why are not 
intermediates of all sorts more abundantly produced in nature 
than is actually known to be the case? 

The fact is that intermediates are found abundantly 
in mammalian zoology (see, for example, Osgood, 
1909.1) as well as in paleontology. 

MENDELIAN DISCONTINUITY IN HEREDITY 

In 1903 the saltation hypothesis was again revived, 
after the rediscovery of the law of Mendel (1865). The 
wide prevalence of the separable characters in hered- 
ity and of the separateness of so-called "imit char- 



842 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



acters" as they appear in the body, and the demon- 
strated separableness of their "factors," "determiners," 
"genes "in the germ, gave rise afresh to the tlieoretic 
assumption of discontinuity of origin of all characters 
in the germ. In 1912 (1912.372) Osborn established 
the fact that this assumption is a non sequitur. 

Mendelian research on the mammals has been con- 
fined mainly to color characters in certain species of 
rodents, chiefly mice and guinea pigs, although it has 
been extended also to horses and cattle, which were 
studied less by experiment than from the records of 
stud books. The first striliing general result is the 
principle of antithesis in the heritage of characters 
that mutually exclude each other, as originally typi- 
fied by the antithesis of Mendel's "tallness" and 
"shortness," "smooth coat" and "wrinkled coat" in 
peas. The second MendeUan principle is that when 
these antithetic characters meet in the germ cells the 
dominant character becomes visible or patent in the 
offspring, whereas the recessive character remains 
latent in the germ, perhaps to reappear in later crosses 
in the Mendehan ratios. 

Similar dominance of head proportion appears in the 
first cross between the ass ( c? ) and horse ( 9 ) as exem- 
plified in the mule, which exhibits the dominant 
dolichocephaly of the horse and at the same time the 
dominant molar-tooth pattern of the ass. Other pro- 
portional characters in this hybrid are known to be 
imperfectly "dominant," or "blending." In such cases 
inheritance is blending and non-Mendelian, for there 
is no evidence of size character segregation in the 
germ other than the increased variability of the second 
hybrid generation. 

TRUTH AND EEEOK IN JOHANNSEN's PURE LINE 
SALTATION PRINCIPLE 

The "pure line" is another biologic principle from 
which both true and false conclusions have been 
drawn. This principle is based upon breeding expe- 
riments on pure strains of garden beans which tend 
to show that certain kinds of characters arise only 
discontinuously by saltation. Johannsen (1911.1, pp. 
129-159), through experiments on successive genera- 
tions of self-fertUizing plants (the garden bean), 
reached the illogical opinion as to all plants and ani- 
mals that the apparent continuity of visible, somatic 
form is delusive, and that in the origin of "mutations" 
and "species" there is invariably a germinal saltation 
or discontinuity. His opinions may be summarized 
and paraphrased as follows: 

The developing organism which the paleontologist or the 
zoologist observes may be called the phenotj-pe, or visible type. 
Vertebrate paleontologists and zoologists comparing huge col- 
lections in museums have erected in phylogenetic speculation 
a science of phenotypes which is not of value in a study of 
germinal evolution (genetics) because the description of pheno- 
types is inadequate as the starting point for genetic investiga- 
tion. The adaptation of phenotypes through the direct influence 



of environment [Buffon's factor] or of use and disuse [Lamarck's 
factor] is not of genetic [hereditary] importance. Ontogeny is 
an expression of the heredity germ, but the heredity germ is 
not affected by ontogenesis. The conception of evolution by 
continuous transitions from one type to another has imposed 
itself upon zoologists, botanists [and paleontologists] who are 
examining chiefly shifting visible forms in very fine gradations. 
There may be such a continuity in visible form but not in the 
hereditj' germ from which they spring. All degrees of con- 
tinuity between phenotypes may be found, but real germinal 
transitions must be distinguished from the transitions which 
we observe in visible museum specimens. 

The delusive nature of phenotype continuity is shown by the 
examination of pure lines of plants. A typical pure line is 
composed of the descendants of one pure strain of an organism 
exclusively propagated by self-fertilization, as in the case of the 
garden bean, which demonstrates the stability of germinal con- 
stitution in successive generations where undisturbed by cross 
breeding with other strains. 

Johannsen proposes the term gene to designate the germinal 
"factors," "determiners" (of authors), of various visible char- 
acters. The sum total of all the genes in the fertilized germ 
cell he terms the genotype. A group of similar genotypes — that 
is, of pure-strain individuals — he terms the biotype. The 
genotype, the sum total of the genes, can be examined only 
by the qualities and reactions of the phenotypes under experi- 
ment. Such examination shows that within pure lines — if no 
mutations (De Vries) or other disturbances have been at work — 
there are no genotypioal (that is, germinal) differences in the 
characters under examination. For example, the mutations of 
De Vries observed in nature have shown themselves as consid- 
erable discontinuous saltations. 

The fallacy of Johannsen's argument lies in the 
dictum "ab uno disce omnes." He may be entirely 
right as to the origin of certain germinal characters 
and entirely wrong as to the origin of others. Never- 
theless, he concludes from his observations on the 
bean that the essential point in all evolution is the 
sudden alteration, loss or gain, of the genes, the 
germinal constituents, of the genotype. In his opin- 
ion all evidence as to "mutations of De Vries" points 
to the discontinuity of the changes in question. In 
the theory of origin of allometrons, the crucial point 
in the application of Johannsen's pure-line hypothesis 
would be the assumed stability of the genes, factors, or 
determiners, as distinguished from their assumed 
fluctuation. 

EXPERIMENTS IN THE ARTIFICIAL SELECTION OF 
VARIATIONS OF PROPORTION 

Against such hypotheses of the stability of all 
germinal determiners of proportional, quantitative, 
intensive characters may be set the evidence adduced 
by many experimentalists as to the heritage of quan- 
titative variation. As opposed to Johannsen, Castle 
(1916.1) observes that in his experiments every 
"unit character" — that is, every germinal deter- 
miner — is subject to quantitative variation — that is, 
its visible expression in the body varies — and it is 
clear that these variations have a germinal basis, 
because they are inherited. By selection, plus or 
minus, through a series of generations we can intensify 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



843 



or diminish, the expression of a character — that is, we 
can modify the germinal character itself. These 
germinal quantitative variations (in color patterns 
and pigmented areas) behave as simple units — that 
is, biocharacters. Blending of size (Castle, 1916.1) is 
demonstrated in rabbits, in which the offspring of a 
large and a small rabbit are of intermediate size — that 
is, neither small size nor large size dominating in the 
cross. "Size," observes Castle, "is an imstable 
character, ever varying." Slow changes in size can 
be effected by artificial selection without any crossing 
whatever. Progressive diminution in size may be 
effected by crossing small breeds. 

More recently Jennings (1917.1, p. 305) observes: 

It appears to me that the experimental work in Mendel- 
ism * * * is supplying a complete foundation for evolution 
through the accumulation by [artificial] selection of minute 
gradation. * * * A visible character may be modified in 
the finest gradations by alterations in diverse parts of the 
germinal material. The objections raised by the mutationists 
to gradual change through [artificial] selection are breaking 
down as a result of the thoroughness of the mutationists' own 
studies. 

Jennings is of the opinion (op. cit., p. 306) that 
Mendelian heredity acts as an accelerator to the 
effectiveness of [artificial] selection. The sum of 
Castle's and Jennings's observations and opinions is 
that certain germinal characters may be altered by 
artificially selecting in each generation the merely 
quantitative variation that goes farthest in the desired 
direction. 

Against this general experimental conclusion are 
earlier (Punnett, 1911.1, p. 138; Davenport, 1910) 
and more recent (Pearl, 1917.1, pp. 72, 73) observations 
that in certain biocharacters fluctuations are not 
heritable, and that it is impossible through cumulative 
artificial selection of fluctuations to establish a new 
quantitative mean. 

To sum up: The years of experiment since the 
rediscovery of Mendelism in 1903 present consid- 
erable but not yet conclusive evidence that certain 
proportional characters as well as certain numerically 
new characters can be accumulated through artificial 
selection, and that certain other characters are not 
cumulative through artificial selection of quantitative 
variations but are dependent upon the saltation, or 
sudden appearance, of new character genes or sudden 
gradations of character in the germ. 

The observations (Castle, Jennings) in experimental 
heredity that certain germinal quantitative varia- 
tions of form and color are heritable and can be 
guided in certain directions by artificial selection do 
not necessarily prove that allometrons have arisen 
by similar processes under natural selection but 
render it probable that heritable fluctuations of 
proportion of survival value are selected and accumu- 
lated under natural selection, as in Darwin's theory of 
the causes of the long neck of the giraffe. 



THEORETIC AND EXPERIMENTAL CAUSES OF THE 
EVOLUTION OF ALLOMETRONS 

The above studies in experimental heredity and 
selection have a very important bearing upon our 
interpretations of heredity in the titanotheres and 
other quadrupeds known from their fossilized remains. 
It should not be assumed that aU allometrons are 
due to the same causes. 

Among titanotheres adaptation through changes of 
proportion has been shown to constitute the larger 
part of the bodily evolution. In all mammals the 
response of the musculature and the skeleton to 
changes of environment and habitat is more conspicu- 
ous in proportions than in any other character groups. 
Allometric adaptation includes a continuous and 
perfect adjustment of all the demands of weight and 
speed in locomotion and of offense and defense in the 
capture or warding off of enemies. It follows, as 
already remarked, that evidence for the Lamarckian, 
the Darwinian, and the tetrakinetic theories of causa- 
tion must be most closely analyzed. The following 12 
modes of allometric change, germinal, ontogenetic, 
environmental, selectional, are observable: 

1. Germinal aUometrons, arising by continuous or gradual 

change. 

2. Germinal aUometrons, arising by sudden changes (saltations) . 

3. Certain germinal allometrons, uninfluenced by the direct 

action of environment. 

4. Germinal aUometrons, apparently influenced by direct 

action of environment. 

5. Germinal aUometrons, fluctuating and nonfluctuating. 

6. Certain germinal aUometrons, of high survival selection 

value. 

7. Other germinal aUometrons, of apparently no survival 

selection value. 

8. Ontogenetic aUometrons, experimentaUy influenced by 

changes of habit. 

9. Ontogenetic allometrons, experimentally influenced by 

changes of environment. 

10. Germinal aUometrons — prenatal, adolescent, adult, male 

sexual, and female sexual development. 

11. Ontogenic allometrons, influenced by glandular internal 

secretions, enzymes, and other organic catalyzers. 

12. Germinal (?) allometrons, influenced by organic catalyzers. 

GERMZNAI AUOMETRONS AEISIWG BY CONTINUOUS OK GKADUAI 
CHANGE 

The continuity of proportional change observed in 
all the phyla of titanotheres described in this mono- 
graph, as well as in several other phyla of imgulates, 
such as the rhinoceroses and the horses, constitutes 
in itself very strong evidence against the adequacy of 
the Darwinian theory of the natural selection of indi- 
vidual fluctuations. 

Proportional continuity is in the nature of the con- 
tinuous "mutations of Waagen" rather than of the 
discontinuous "mutations of De Vries." The general 
contrast is very ably presented by Scott (1894.7, p. 
355) in his article entitled "Mutations and variations." 
Progressive brachycephaly and progressive dolicho- 
cephaly in the titanotheres point to the presence of 



844 



TITANOTHERES OF ^^NCIENT WYOMING, DAKOTA, AND NEBRASKA 



some similarly acting influence affecting generation 
after generation in a similar manner, operating like 
the "Alutationsrichtung" of Neumayr. Within cer- 
tain phyla a tendency or predetermination to evolve 
in breadth or length orthogenetically appears to be 
established, flowing in one direction like a tide, on the 
surface of which occur individual fluctuations and 
variations, like waves and ripples. The overdevelop- 
ment of certain proportions, as seen in extreme brachy- 
cephaly and dolichocephaly, for example, seems to be 
a manifestation of this orthogenetic impulse, which 
appears to cany certain skulls to inadaptive extremes. 
As shown in section 2 of this chapter, on the causes 
of extinction, such extremes of structm-e tend to 
be eliminated in the struggle for existence and to be 
replaced by head forms of intermediate development. 

The phenomenon of overevolution has been likened 
by Loomis to the principle of "momentum" in 
mechanics. As described below it appears as if 
certain progressive proportions were continually in- 
fluenced by certain internal secretions, or organic 
catalyzers, which continue to accelerate these allo- 
metric biocharacters in genei'ation after generation 
and finally to overaccelerate them beyond the point 
of utility. This is manifest in certain cases of the 
harmonic, uniform dolichocephaly or brachycephaly 
of the greater part of the skull, where certain parts 
are sacrificed or overcrowded, in apparent defiance of 
the principle of compensation. Again, as macro- 
cephaly and microcephaly arise as abnormalities due 
to abnormal internal secretions, it would appear proba- 
ble that normal internal secretions are in some way 
connected with the normal evolution of head pro- 
portions. 

Subject to future experimental test it would appear 
probable that continuity of head form evolution is 
not caused through the Darwinian principle of con- 
tinuous selection of "individual variations" and 
"individual fluctuations" but by the intermediate 
influence of some kind of interaction, as implied in the 
tetrakinetic theory. 

GEEMHTAI All OMETEONS ARISING BY SUDDEIT CHANGES (SALTATIONS) ; 
INTEEACTION THEORY 

In considering major saltations above we have 
alluded briefly to the numerous examples of the 
sudden appearance of germinal allometrons in domes- 
ticated animals, such as are seen in the short-limbed 
an con breed of sheep. No evidence has thus far 
been adduced that any wild breeds of animals have 
originated suddenly in this way. The fact that all 
four limbs in quadrupeds are similarly affected by 
harmonic abbreviation in cases of sudden brachy- 
mely and that the opposite hands in bipeds are 
similarly affected in cases of sudden brachydac- 
tyly points again to the possibility of abnormal 
chemical messenger interaction as being the imme- 
diate cause of these kinds of abbreviation. There 



are in man two kinds of brachydactyly — (1) that in 
which the hand is harmonically abbreviated and 
(2) that in which the terminal phalanges are lacking. 
The former, congenital brachydactyly, bears strong 
superficial resemblance to the ontogenetic brachy- 
dactyly caused by the abnormal internal secretions 
of certain parts of the pituitary gland (Gushing, 
1912.1). This again suggests the hypothesis that 
congenital brachymely and brachydactyly are due 
to congenital defects in the hereditary chemical 
messenger interaction system. The value of this 
hypothesis can be tested only by physiological 
experiment. So far as they are understood these 
phenomena of discontinuity of proportion, like the 
phenomena of continuity of proportion, appear to be 
connected with some internal interacting chemical 
messenger system, which may be at once the seat of 
balance, of retardation, and of the sudden acceleration 
manifested in saltations of proportions. 

CERTAIN GERMINAI ALLOMETRONS UNINFLUENCED BY THE DIRECT 
ACTION OF ENVIRONMENT 

Apart from the well-laiown influence of environ- 
ment on harmonic size, dwarfing or gigantism, the 
field, experimental, and anatomical evidence on the 
influence of environment on germinal allometrons is 
conflicting. As to field zoologic observations Gerrit 
S. Miller (letter to author, 1912) observes: 

I can give you plenty of illustrations of "direct action of 
environment" on color characters in mammals; in fact, you 
have only to tell me how many you want, and I will fill your 
order. As to proportions, size, and cranial characters, the 
case is quite different. So far as I have been able to make out 
the action is here indirect, as, for instance, when a cat in a 
wet district takes to a fish diet and [ontogenetically] develops 
the characters of Aelurus, or when a viverrid does the same and 
becomes Cynogale. You can hardly say that the differences 
here observed are due to the direct action of the environment. 
They are rather due to the ontogenetic modifications caused 
by habit. I know of plenty of cases of color changes of definite 
character afl'ecting many different members of a fauna in the 
same way, as you go from a wet to a dry region, or from a high 
to a low region, but I do not know of any parallel set of changes 
in proportions, size, or structure. 

The field zoologist has yet to demonstrate that there 
is direct relation between certain environments and 
certain allometrons distinct from the indirect action 
of environment which may arise as the ontogenic 
result of differences of locomotion, of food, and of 
feeding habits in different environments, as, for ex- 
ample, the influence of the purely carnivorous diet 
of the Eskimo on their head and jaw conformation. 
The ichthyologists (Jordan, Evermann) demonstrate 
that the ontogenic proportions of spines in certain 
fishes are directly related to the chemical content of 
their watery medium. 

Anthropologists differ somewhat as to the direct 
action of environment on the evolution of the form of 
the human head. Elliot Smith (letter to author, 
August 12, 1911) finds no evidence that environment 



CAUSES OP THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



845 



directly produces any modification of head propor- 
tion. He implies that such modification, if natural, 
would show only after thousands of years of residence 
in a single locality, during which the indirect influ- 
ence of change of feeding habits might play an impor- 
tant part. He continues: 

I do not believe for a moment in Boas's observations [of 1911] 
on the direct and rapid influence of environment in modifying 
ontogenic head form. In my opinion the conditions of doli- 
chocephaly and braohycephaly in man must have developed 
very slowly through exceedingly long periods of time and in 
widely separated areas, amidst widelj' different environments. 
Brachycephaly has thus become especially distinctive of the 
Central Asian high plateau populations, doUchocephaly of the 
littoral and plains-dwelling peoples. But these unit characters 
are now so fixed that environment is powerless to modify them 
in a thousand 3'ears or so. 

The observations of Hrdlicka (1910.1, p. 214) on 
the forms of the cranial vault and the face of the 
Eskimo lead us to believe that they are attributable 
rather to feeding habits than to the direct action of 
environment. 

Apart from the difTerences of opinion as to the 
direct influence of environment on head form as dis- 
tinguished from the modifications of habit and the 
influence of human selection, there is absolute una- 
nimity both of evidence and of opinion as to the con- 
tinuity of allometric evolution in the human species 
which establishes different extremes of head form 
under conditions of geographic and social isolation. 

GERMINAL AllOMETEOFS APPARENTLY INFLUENCED BY DIRECT ACTION 
OF ENVIRONMENT 

The dwarfing as well as the gigantism that arise 
continuously or suddenly are usually harmonic — that 
is, all parts are contracted or enlarged harmoniously 
owing to generally unfavorable or highly favorable 
environmental conditions. This is generally true of 
dwarfed breeds that arise through long-continued 
exposure to a dwarfing environment. For example, 
the Celtic pony {E. caballus celticus Ewart) appears 
to be a harmonic dwarf of its remote Arab-like ances- 
tor. Doubtless disharmonic dwarfing and gigantism 
arise under the prolonged direct action of the environ- 
ment, but at present no examples can be cited. 

It would be interesting to ascertain through experi- 
ment whether there is any relation between the har- 
monic dwarfing attributable to a prolonged unfavor- 
able environment on successive generations and that 
induced experimentally in a few months by removal 
of the thyroid or parathyroid glands, as observed in 
dogs and sheep. It does not appear probable that any 
relation will be found, because thyroid and parathy- 
roid dwarfing is apt to be disharmonic, causing the 
acceleration of some parts and the general retardation 
of others. 



GERMINAL ALLOMETEONS, FLUCTUATING AND NONFLUCTUATING 

The Darwinian theory, as we have shown above, 
depends upon the continuous natural selection of 
"fluctuating variations" that are continous in a given 
direction. It appears to be highly probable as to 
fluctuations that have high survival selection value, 
such as fluctuations in the length in the entire series 
of cervical vertebrae in the neck of the giraffe, an 
evolution which is generally harmonic. Harmonic 
allometrons generally, such as length of neck, length 
oi limb, length of feet, may possess high survival 
selection value and favor the Darwinian theory. 

The causation of disharmonic allometrons, which in 
the titanothere skull are far more numerous than the 
harmonic, presents, on the contrary, great theoretic 
difficulties. The abbreviation of the face coordinately 
with the elongation of the cranium is an example of 
disharmony. The separate velocity of ontogenetic 
and phylogenetic movement of a very large number of 
biocharacters is so perfectly coordinated and adjusted 
in every stage of development and evolution as to 
make the Darwinian "fortuitous fluctuation" hypothe- 
sis in a high degree untenable. 

CERTAIN GERMINAL ALLOMETEONS OF HIGH SURVIVAL SELECTION VALUE 

AU parts of the skeleton change in proportion during 
the entire period of growth, prenatal and postnatal. 
These changes of proportion are partly paleotelic — 
that is, they are reversive to ancestral stages — partly 
cenotelic — that is, they are adapted to existing con- 
ditions. 

Certain prenatal germinal proportions exhibit espe- 
cially high survival value; they are examples of 
perfect mechanical adjustments of form that rise 
from the germ before they have been developed by 
individual modification. Apparently fatal to the 
Lamarckian theory that all aUometrons are due to 
the inheritance of modifications acquu-ed postnatally 
are the many well-lvnown instances of prenatal allo- 
metrons, such as those of certain precocious birds — 
the Praecoces, for example — which are adapted to 
run immediately on leaving the shell. Certain pre- 
cocious desert-living cursorial mammals are also 
capable of rapid movement immediately after birth. 
The limbs of such animals have before birth the 
proportions of those of adults. Conspicuous among 
these are the horses. In the feral state the foals 
must be able immediately after birth to keep up 
with the herd, and in adaptation to this necessity 
the members that contribute most to speed are pre- 
natally so much accelerated iu development that 
they attain almost their full adult length before birth. 
Thus in thoroughbred horses the postnatal increase 
in the length of the cannon bones is only 3 centimeters 



846 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



(Ewart, 1909.1). Osbom observed the survival value 
of these limb proportions among the semiferal mus- 
tangs of western Louisiana. While the herd was 
being rounded up one of the mares began to give 
birth to a colt; she lay down for a moment, the foal 
was cast, and in an incredibly short time the mother 
and foal were running with the herd. Stevenson- 
Hamilton (1912.1, p. 539) observes that within a few 
hours after birth all young antelopes and zebras, 
OMTUg to their limb proportions, are able to keep up 
in the most surprising manner with the full-grown 
animals, even when those are going at full speed. 

These are instances where the Darwinian theory of 
the selection of germinal fluctuations of proportion 
appears to be applicable; whereas the Lamarckian 
theory of the inheritance of acquired changes of 
proportion is apparently not applicable. They are 
notable instances of the high survival value of certain 
allometrons which are apparently initiated in the 
germ before they are initiated in ontogeny. 



ONTOGENETIC AllOMETEONS EXPERIMENTALLY INFLUENCED BY 
CHANGES OF HABIT 

The reverse principle is equally true, namely, that 
certain adaptive proportions are initiated in ontogeny. 
Notable instances of this principle are found in the 
experiments of Regnault (1911.1), whereby it is dem- 
onstrated that cursorial limb proportions may be 
transformed into saltatory limb proportions experi- 
mentally. 

In the normal dog the tibiofemoral proportion — that 
is, length of the tibia as compared with that of the 
femur — is 95 per cent. The percentage rises to 104 
in dogs that are forced congenitally or experimentally 
into the cursorial habit. Thus a dog born without 
fore legs sits upright and learns to progress by imper- 
fect or awkward leaps, in response to which the tibia 
gradually elongates. Similar results are recorded by 
Fuld (1901.1) through the amputation of the fore 
limbs in newborn dogs, in which, through the enforced 
adoption of the saltatory gait, the tibiofemoral index 




A Bi B 

Figure 752. — Accelerated elongation of the limbs in the young zebra and guanaco 

A, Zebra and colt; B, guanaco; B'. young guanaco. Note the relative length of limb which enables the animal at birth to keep up with its mother. 

Modified after Loomis. 



OTHER GERMINAL ALLOMETRONS OF APPARENTLY NO SURVIVAL 
SELECTION VALUE; PREDETERMINATION 

No anthropologist has offered any satisfactory 
explanation as to the adaptive significance of the 
dolichocephaly and brachycephaly of the human 
head. Broad heads and long heads have lived in 
northern Europe for 30,000 years under the same 
environment, with the same feeding habits and under 
the same social conditions, and without displaying 
marked differences of intellectual aptitudes. As to 
causation Boas writes (letter to author, April 8, 1911): 

So far, the matter [of the origin and cause of head form] is 
very perple.xing to me. I feel, however, very strongly that 
changes in type [in head proportion] are very liable to be pro- 
gressive in definite directions. * * * To my mind it seems 
no more difficult to assume that this predetermined direction 
should continue from generation to generation than to make 
the much more difficult assumption that notwithstanding all 
internal changes the egg cell of one generation should be 
absolutely indentical with that of the preceding generation. 



rises to 101-104 as compared with 95 in the normal 
dog. At the same time the femur and tibia, taking 
on the entire weight of the body in locomotion, become 
more massive and exhibit many marked modifications. 
Contrary to Regnault's opinion (1911.1), this very 
strildng example of adaptive modification does not 
demonstrate the Lamarckian principle until it is 
shown that such modifications are inherited. 

HARMONIC OR CONFLICTING INFLUENCE OF THE FOUR FACTORS OF 
EVOLUTION 

It is clear from the foregoing and from other 
examples that will be given that heredity may pre- 
dispose allometric adaptations which may be de- 
veloped, intensified, and more or less perfected by 
ontogeny, by physical environment, and by life 
environment, or the very reverse. The internal and 
the external tendencies, if all in the same direction, 
result in the maximum development of a given 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



847 



biocharacter. In the above example of the evolution 
of a saltatory (leaping) type of quadruped such as 
the kangaroo, favorable individual exercise (ontogeny) 
may carry a hereditary predisposition a little beyond 
the normal, say from A to B. Thus the saltatory 
quadruped may be born with a short femur and a 
Jong tibia, the normal length of which is A. By the 
habit of long leaping this tibia may be lengthened to 
JB. In nature this long leaping may have a dele- 
terious effect on other parts of the aninial or on its 
progeny, ia which case natural selection would tend 
to bring the tibia of the jumper back to A. Or long 
leaping may be beneficial both in the avoidance of 
•enemies and the procm'ement of food and not dele- 
terious to offspring, in which case all individuals wall 
rise to B in innumerable generations until finally 
natural selection will bring heredity up to B. This 
illustration affords an example of the relation between 



segments without a corresponding relative increase 
in the breadth of the skull. Neotoma exhibits a 
progressive dolichopy and dolichocephaly in ontogeny. 
The development of the skull of Neotoma exhibits a 
different ratio in each of the eighteen points measured, 
just as the evolution of the titanothere skull exhibits 
a different ratio and index in every bone measured. 
In passing from adolescence to senility proportions 
wholly ontogenetic may readily be mistaken for dif- 
ferences of subspecific or even specific value. In a 
very large series of crania examined, selected from 
localities less than 25 miles apart, subject to similar 
environmental conditions, AUen observed, in addition 
to the ontogenetic allometrons, a large amount of in- 
dividual variation both in the form and in the relative 
size of every element in the adidt skull. He says as 
to size that there are dwarfs and giants. It is still 
difficult to determine how far this apparently sponta- 





FiGURE 753. — Femur of dog, normal and as modified b}- amputation or congenital absence of the fore limb 

A, Dog congenitally devoid of fore limbs. A^ Skeleton of the same animal, showing the proportional enlargement of the hind limb and elon- 
gation of the femur; after Eegnault, 1911. B, Normal femur, tibia, and fibula of dog. C, Femur, tibia, and fibula of a dog whose 
fore limbs were amputated. B and C after Fuld, 1901. All one-nmth natural size. 



ontogeny and heredity that is established in that form 
of survival of the fittest that is known as organic or 
•coincident selection, a theory simultaneously pro- 
posed by Osborn, Baldwin, and L. Morgan. 

ONTOGENETIC ALIOMETEONS EXPEKIMENTAIIY INFLUENCED BY 
CHANGES OF ENVIEONMENT 

Certain featm-es of proportion are highly sensitive 
to envii'onment. The fine, narrow hoofs of the Arab 
horse, for example, are developed only in reaction 
to the hard rocky or sandy soil of semidesert regions. 
If the animal is transferred to the bottom lands or 
grassy meadows of a humid region its hoofs become 
broader and flatter. Doubtless all the proportions 
of the body are subject, either directly or indirectly, 
to similar ontogenetic modifications brought about by 
changes of environment. 

GEEMINAl ALIOMETEONS: PRENATAL, ADOLESCENT, ADULT 

Allen (1894.1, p. 234) observes that in Neotoma 
micropus there is an incessant change in the general 
proportions of the skull coordinated with growth, 
due mainly to the lengthening of the several skull 



' neous variability is actually due to environmental, 
ontogenetic, or germinal mfluence and how far to life- 
environmental influence. 

A Darwinian interpretation of the evolution of the 
skull of Neotoma under the hypothesis of organic 
selection would be that the progressive elongation of 
the face (dolichopy) and the skuU (dolichocephaly) 
is attributable to the combined action of natural 
selection of heritable fluctuations with ontogenetic 
modifications (allometrons due to feeding habits), on 
the theory that all germinal variations favoring length 
of the face and the head are better adapted to the 
feeding habits of Neotoma and thus would favor 
survival. Thus congenitally elongate skulls which in 
course of individual mechanical development most 
rapidly responded to the process of individual adapta- 
tion and to the upbending of the face upon the cranium 
would tend in the long run to be selected. To similar 
hypothetical coincidence of germinal variation, of 
selection, and of ontogeny might be theoretically 
attributed the disharmonic evolution of all the different 
parts of the skull. 



848 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



ONTOGENETIC AlIOMETRONS INFLUENCED BY GIANDUIAR INTERNAL 
SECRETIONS, ENZYMES, AND OTHER ORGANIC CATALYZERS 

The normal hereditary proportions of the skeleton 
as well as of all parts of the animal body are sustained 
through causes of two kinds: first, normal heredity, 
ontogeny, physical environment, and life environ- 





FiGTJRB 754. — Brachj-dactj-ly and dolk-hodactyly 

A, Brachydactyly (short-flngeredness) due to germinal saltation — 
that is, congenital; after Drinkwater. B, Brachydactyly that is 
not congenital but developed secondarily after birth through exces- 
sive internal secretions of the pituitary gland. C, Dolichodactyly 
(long-flngeredness) [developed after birth, attributed to abnor- 
mally insufficient secretions of the pituitary gland. B and C 
after Cashing, 1912.1. 

ment — that is, tetraplasy, the principle 
fully explained above; second, normal 
interactions between all parts of the 
body, probably included in tetraplasy. 
These normal interactions are in part 
sustained by the nervous system, in part 
by the various kinds of physico-chemical 
catalyzers Icnown as enzymes, internal 
secretions, including hormones (accel- 
erators) and chalones (retarders), sum- 
marized by Osborn (1917.462, pp. 250- 
251), from the researches of Simpson, 
Ashner, Gushing, Schafer, Goodale, 
Lillie, as follows: 

These include many changes of proportion 
in mammals, which are not Icnown to have 
a selective survival value. We may instance 
in man, for example, the long-head form 
(doUochocephaly) and the broad-head form 
(brachycephaly) , or the long-fingered form 
(dolichodactyly) and the short-fingered form 
(brachydactyly) , which have been interpreted 
as congenital characters appearing at birth 
and tending to be transmitted to offspring. 
Brachydactyly may be transmitted through 
several generations, but until recently no one 
has suggested what may be its possible cause. 

It has now been found [Gushing, 1912.1, 
pp. 253, 256] that both the short-fingered 
condition and the long-fingered condition 
may be induced during the lifetime of the 
individual in a previously healthy and nor- 
mal pair of hands by a diseased or injured 
condition of the pituitary body at the base of 
the brain. If the secretions of the pituitary are abnormally ac- 
tive (hyperpituitarism) the hand becomes broad and the fin- 
gers stumpy (fig. 118, B) . If the secretions of the pituitary are 
abnormally reduced (hypopituitarism) the fingers become taper- 
ing and slender (fig. 118, C). Thus in a most remarkable man- 
ner the internal secretions of a very ancient ductless gland. 



attached to the brain and originating in the roof of the mouth 
in our most remote fishlike ancestors, affect the proportions 
both of flush and bones in the fingers, as well as the propor- 
tions of many other parts of the body. 

Whether this is a mere coincidence of a heredity chromatin 
congenital character with a mere bodily chemical messenger 
character it would be premature to say. It certainly appears 
that chemical interactions from the pituitary body are con- 
nected with the normal and abnormal development of propor- 
tions in distant parts of the body. 

Allometrons, as shown above, are partly examples 
of biocharacter velocity. Some of the above phe- 
nomena of abnormal interaction in the organism may 
give us an insight into the possible causes of slow or 
rapid movement, of acceleration or retardation, in 
the origm of allometrons. These internal secretions 
may be connected with the fact that one proportion 
is retarded, as if suffering from inertia, while a con- 
joining proportion biocharacter is full of life and 
velocity, accelerated, like the alert soldier in the 
regiment. Whether or not internal secretions prove 
to be among the causes of the evolution of proportion, 





Figure 755. — Examples of dwarfing due to removal or abnormal functioning of 
certain glands 

Harmonic and disharmonic alterations of proportion in mammals due to removal of the thyroid, para- 
thyroid, pituitary, and other glands. A (rightl, Normal sheep of fourteen months; Oeft) sheep at the 
same age from which the thyroid and parathyroids were removed at the age of two months: after Suther- 
land Simpson. B (right). Normal dog of twelve months; (left) a dog of the same age and litter from 
which the pituitary gland was removed at the age of two months; after Aschner. C, Dwarfed ma- 
crocephalic pigmies of the hills compared with tall microcephalic plains men of west-centrfil New 
Guinea; after Rawling. The question arises whether the dwarfing and macrocephaly of the pigmies 
may not be due to abnormal interactions (internal secretions) of certain glands like the thyroid. 

they are certainly among the causes of normal and 
abnormal development of proportion. Theoretically 
it is conceivable, as suggested by Cunningham 
(1908.1), that since hormones and chalones determine 
the rate of development of many organs they may 



I 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



849 



also act upon the physico-chemical messengers to the 
germ cells, the activities of which correspond poten- 
tially with certain activities of the body cells (Osborn, 
1917.462, pp. 77-78). 

GEEMINAI it) ALIOMETEONS INFIDENCED BY 0E6AHIC CATALYZERS 

There is no evidence at present that germinal pre- 
dispositions of proportion are influenced by the hor- 
mones and chalones which from the above observa- 
tions are shown to exert so profound an influence on 
bodily proportions. This can be tested only by very 
careful experimental work. One reason for believing 
that the heredity germ may be influenced in this 
manner is that it is the germinal or chromatin material 
of the body cells which responds to these organic 
catalyzers, as shown in the transplantation experi- 
ments of the placing of the ovary into the male body 
and the testes into the female body, which results in 
the internal secretions from these glands evoking the 
male characteristics in the chromatin of the female 
body cells and female characteristics in the chromatin 
of the male body cells. 

These latent characters can reside only in the germ 
chromatin. Consequently in the germ chromatin of 
the germ cells proper there may be possible influences 
of interaction awaiting discovery through experiment. 
Such discovery would lend support to the tetrakinetic 
theory. 

SUMMARY OF THEORETIC CAUSES OF EVOLUTION 

EVIDENCE AGAINST THE LAMARCKIAN PRINCIPLE 

The balance of evidence in titanothere evolution is 
decidedly against the Lamarckian principle of the in- 
heritance of acquired characters, which was so strongly 
advocated, from paleontologic and mechanical data, 
by Cope. Neither the mode of initiation of allome- 
trons nor that of rectigradations, when carefully ana- 
lyzed, appears to sustain this principle. Though it is 
a convenient explanation of the development and 
degeneration of certain organs it does not account for 
all origins nor for most of the adaptations of pro- 
portion. 

EVIDENCE AGAINST THE THEORY OF SELECTION OF 
MINUTE VARIATIONS 

The evidence is also strongly against Darwin's 
hypothesis of the selection of minute variations as of 
survival value. The chief counter evidence is that 
new characters (rectigradations) and new propor- 
tions (allometrons) do not arise by chance but from 
some undiscovered definite orthogenetic principle that 
affects the germ. 

EVIDENCE FAVORABLE TO THE SELECTION OF CERTAIN 
FLUCTUATIONS 

The Darwinian principle, on the other hand, afi^ords 
the most reasonable hypothesis we have to offer at 
present of the evolution of those larger proportions 



that have a distinct survival value — for example, the 
length of neck of the giraffe, the length of limbs of 
the unborn colt. Acceleration of adult limb propor- 
tions into fetal stages, to protect the young at birth, 
affords the best example of the survival value of 
germinal fluctuating characters and the strongest 
refutation of the Lamarckian hypothesis. 

UNKNOWN CAUSES OF THE ORIGIN OF RECTIGRADATIONS 

We are wholly at a loss to advance any explanation 
of the causes of rectigradations, of the orthogenetic 
appearance of horn rudiments on the head, or of 
cuspules on the teeth. Although these new characters 
arise independently in different phyla, thus pointing 
to some kind of germinal predetermination or pre- 
disposition, they do not indicate an internal perfecting 
tendency, because they are timed with reference to 
external-internal reactions. Consequently there ap- 
pears to be some kind of interaction between environ- 



, ■>. fiAy/a 




Figure 756. — Theoretic selection of fluctua- 
tions of proportion 

Gradual shifting of proportions (d, c, be, b, as of length of 
neck) in successive generations. Showing the shifting of 
the fluctuations of frequency in a series of characters (a-e). 
In a lower geologic stage (A) d is the most frequent char- 
acter; in the next higher geologic stage (B) c is the most 
frequent character; in the next higher stage (C) 6c is the 
most frequent character; in the highest stage (D) b is the 
most frequent character. This shitting is consistent with 
the Lamarckian, the Darwinian, or the tetrakinetic theory. 

ment, habit, and the time of appearance of these new 
organs; but we have no inkling as to what this rela- 
tion is — whether, in fact, it is causal unless it consists, 
of some kind of physico-chemical interaction. 

NECESSITY OF EXPERIMENT ON THE TETRAKINETIC 
PRINCIPLE 

The tetrakinetic principle is a working hypothesis or 
line of suggestion as to a possible relation between the 
external energies of the environment and the body 
and the internal energies of the development and 
evolution of the germ. That some such causal rela- 
tion exists is faintly indicated by the modes of evolu- 
tion both of rectigradations and of certain propor- 
tions. As there are certain energizers in the body, 
catalytic messengers that hasten the development of 
one organ and retard the development of another, we 
may now obtain a glimpse of what may possibly 



850 



TITAN OTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



prove to be one of the real causes of evolution — the 
evolution not of the organs themselves but of the 
energizing chemical or physical messengers that control 
ontogenetic or phylogenetic velocity. 

BIBLIOGRAPHY OF LITERATURE RELATING TO THE 
THEORIES OF EVOLUTION CITED IN SECTION 1 
Allen, Joel Asaph. 

1894.1. Cranial variations in Neotoma micropiis due to 
growth and individual differentiation: Am. 
Mus. Nat. Hist. Bull., vol. 6, pp. 233-248, pi. 4, 
Aug. 3, 1894. 
B,\TEsoN, William. 

1894.1. Materials for the stud}' of variation treated with 
especial regard to discontinuity in the origin 
of species, 598 pp., London, Macmillan & Co., 
1894. 
Boas, Franz. 

1903.1. Heredity in head form: Am. Anthropologist, new 
ser., vol. 5, No. 3, pp. 530-538, July-September, 
1903. 

1911.1. The mind of primitive man, 294 pp., New York, 
Macmillan Co., 1911. 
Castle, W. E. 

1912.1. The inconstancy of unit characters: Am. Nat- 
uralist, vol. 46, No. 546, pp. 352-362, 1912. 

1916.1. Genetics and eugenics, a text-book for students of 
biology and a reference book for animal and 
plant breeders, 353 pp., 135 figs., 37 tables, 
Cambridge [Mass.] and London, 1916. 
Cope, Edward Dkinkek. 

1887.1. The origin of the fittest, essays on evolution, 467 
pp., D. Appleton & Co., 1887. (This work is 
composed of 21 essays, all of which had been 
previously published elsewhere.) 

1889.1. The mechanical causes of the development of the 
hard parts of the Mammalia: Jour. Morphol- 
ogy, vol. 2, pp. 137-277, pis. 9-14, 93 figs., 1889. 

1896. 1. The primary factors of organic evolution, 547 pp., 
120 figs., Chicago, Open Court Publishing Co., 
1896. 

CUMMINGS, E. R. 

1909.1. Paleontology and the recapitulation theory: 
Indiana Acad. Sei. Proc. 25th anniversary 
meeting, 1909, pp. 305-340. 
Cunningham, J. T. 

1908.1. The heredity of secondary sexual characters in 
relation to hormones, a theory of the heredity 
of somatogenic characters: Archiv Entwick- 
lungsmechanik der Organismens, Band 26, pp. 
372-428, Nov. 24, 1908. 
CtJSHiNG, Harvey. 

1912.1. The pituitary body and its disorders, clinical 
states produced by disorders of the hypophysis 
cerebri, Harvey lecture, 1910, amplified, 341 
pp., colored frontispiece, 319 figs., Philadelphia 
and London, J. B. Lippincott Co., 1912. 
Darbishirb, a. D. 

1911.1. Breeding and the Mendelian discovery, 282 pp., 
London, Cassell & Co., 1911. 
Darwin, Charles Robert. 

1859.1. On the origin of species by means of natural 
selection, or the preservation of favoured races 
in the struggle for life, 502 pp., London, John 
Murray, November, 1859. 
1868.1. The variation of animals and plants under domesti- 
cation, 2 vols., vol. 1, viii, 411 pp.; vol. 2, viii, 
486 pp., illus., London, John Murray, 1868. 



EWART, J. COSSAR. 

1894.1. The development of the skeleton of the limbs of the 
horse, with observations on Polydactyly: Jour. 
Anat. and Physiol., Normal and Pathol., 
Human and Comparative, vol. 28 (new ser., 
vol. 8), pt. 2, January, pp. 236-256; pt. 3, April, 
pp. 342-369 (not concluded), pi. 12. (Re- 
printed with slight changes in Jour. Comp. 
Pathology and Therapeutics, vol. 7, 1894, No. 1, 
Mar. 31, pp. 17-31, pi. 2; No. 2, June 30, pp. 
91-110 (not concluded), pi. 7. 

1909.1. The rate of growth in the horse: Livestock 
Jour. Almanac, 1909 (not paged). 
FuLD, Ernst. 

1901.1. Uber Veriinderungen der Hinterbeinknochen von 
Hunden, etc.: Archiv Entwicklungsmechanik 
der Organismens, Band 19, p. 1, 1901. 
Gregory, William King. 

1910.1. The orders of mammals : Am. Mus. Nat. Hist. Bull., 
vol. 28, pp. 3-524, 32 figs., February, 1910. 

1914.1. Evidence regarding evolution of locomotive organs: 
New York Acad. Sci. Annals, vol. 23 (1913-14), 
pp. 267-268, Apr. 30, 1914. 
Hrdlicka, Ales. 

1910.1. Contribution to the anthropology of central and 
Smith Sound Eskimo: Am. Mus. Nat. Hist. 
Anthropol. Papers, vol. 5, pt. 2, pp. 177-285, 
pis. 9-23, 1910. 
Hyatt, Alpheus. 

1866.1. On the parallelism between the different stages 
of life in the individual and those in the entire 
group of the molluscous order Tetrabranchiata: 
Boston Soc. Nat. Hist. Mem., vol. 1, pt. 2, 
pp. 193-209, 1866. (See also minutes of 
meeting of Feb. 21, 1866, Boston Soc. Nat. 
Hist. Proc, vol. 10, pp. 302-303, May, 1866.) 
Jackson, Robert Tracy. 

1913.1. Alpheus Hyatt and his principles of research: Am. 
Naturalist, vol. 47, No. 556, pp. 195-205, 
April, 1913. 
Jennings, H. S. 

1917.1. Modifying factors and multiple allelomorphs in 
relation to the results of selection: Am. Nat- 
uralist, vol. 51, No. 605, pp. 301-306, May, 
1927. 

JOHANNSEN, W. 

1911.1. The genotype conception of heredity: Am. Nat- 
uralist, vol. 45, No. 531, pp. 129-159, March, 
191L 
Keith, Arthur. 

1911.1. On certain physical characters of the negroes of 

the Congo Free State and Nigeria: Roy. 

Anthrop. Inst. Great Britain and Ireland Jour., 

vol. 41, pp. 40-71, pis. 1-4, January-June, 1911. 

LoNGLEY, W. H. 

1917.1. Studies upon the biological significance of animal 
coloration: Am. Naturalist, vol. 51, No. 605, 
pp. 257-285, May, 1917. 
Love, Harry H. 

1910.1. Are fluctuations inherited? Contr. 6, Laboratory 
experimental plant-breeding, Cornell Univ.: 
Am. Naturahst, vol. 44, No. 523, pp. 412-423, 
July, 1910. 
Merriam, C. Hart. 

1895.1. Monographic revision of the pocket gophers, 
family Geomyidae (exclusive of the species 
Thomomys) : U. S. Dept. Agr. Bur. Biol. Survey 
North Am. Fauna, No. 8, 258 pp., 19 pis., 
4 maps, 71 figs., Jan. 31, 1S95. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



851 



Miller, Gerhit Smith. 

1912.1. Catalogue of the mammals of western Europe 

(Europe exclusive of Russia) in the collection 

of the British Museum, 1019 pp., 213 figs., 

Longmans, Green & Co.. 1912. 
Morgan, T. H. 

1914.1. The mechanism of heredity as indicated by the 

inheritance of linked characters: Pop. Sci. 

Monthly, vol. 84, No. 1, pp. 5-16, January, 

1914. 
OsBORN, Henry Fairfield. 

1889.46. The palaeontologieal evidence for the transmission 

of acquired characters: Am. Naturalist, vol. 23, 
No. 271, pp. 561-566, July, 1889; Am. Assoc. 
Adv. Sci. Proc, vol. 38 [meeting, Toronto, 
August, 1889], pp. 273-276, July, 1890; British 
Assoc. Adv. Sci. Kept, [meeting, Newoastle- 
upon-T3'ne, September, 1889], London, 1890. 

1890.47. The palaeontologieal evidence for the transmission 

of acquired characters: Nature, vol. 41, pp. 
227-228, January, 1890. 

1890.48. Evolution and heredity: Marine Biological Labo- 

ratory, Woods Hole, Biol. Lectures, Boston, 
Ginn & Co., 1890. 
1891.53. Are acquired variations inherited? opening a 
discussion upon the Lamarckian principle in 
evolution, American Society of Naturalists, 
Boston, December 31, 1890: Am. Naturalist, 
vol. 25, No, 291, pp. 191-216, March, 1891. 
1892.58. The Cartwright lectures for 1892 before the 
alumni of the College of Physicians and 
Surgeons, New York — Present problems in 
evolution and heredity: 1, The contemporary 
evolution of man; 2, Difficulties in the heredity 
theory; 3, Heredity and the germ cells: New 
York Medical Record, vol. 20, March 5 and 
April 23, 1892. 

1893.86. Alte und Neue Probleme der Phylogenese, Sep. 
Abdr. aus d. Ergebnisse der Anatomic und 
Entwickelungsgeschichte, Band 3, pp. 584r-619, 
Gottingen, Fr. Merkel u. R. Bonnet, 1893. 

1894.92. Certain principles of progressive!}' adaptive varia- 
tion observed in fossil series: British Assoc. 
Adv. Sci. Rept., 1894, p. 693 (title only) ; Nature, 
vol. 50, No. 1296, p. 435, August 30, 1894. 

1895.94. Environment in its influence upon the successive 
stages of development and as a cause of varia- 
tion, opening discussion before the American 
Society of Naturalists, Baltimore, December 27, 
1894: Science, new ser., vol. 1, No. 2, pp. 35-36, 
January 11, 1895. 

1895.97. The hereditary mechanism and the search for the 
unknown factors of evolution: Marine Biological 
Laboratory, Woods Hole, Biol. Lectures, Boston, 
Ginn & Co., 1895; Am. Naturalist, vol. 29, 
No. 341, pp. 418-439, May, 1895. 

1896.108. A mode of evolution requiring neither natural 
selection nor the inheritance of acquired char- 
acters (abstract) : New York Acad. Sci. Trans., 
vol. 15, March 9 and AprU 13, 1896, pp. 141-142, 
148. 

1896.114. Ontogenic and phylogenio variation: Science, 
new ser., vol. 4, No. 100, pp. 786-789, Novem- 
ber 27, 1896. 

1897.125. Organic selection: Science, new ser., vol. 6, 
No. 146, pp. 583-587, October 15, 1897. 

1897.127. The limits of organic selection: Am. Naturalist, 
vol. 31, No. 371, pp. 944-951, November, 1897. 
101959— 29— VOL 2 11 



OsBORN, Henbt Fairfield — Continued. 

1898.153. Modification and variation, and the hmits of 
organic selection, a joint discussion with 
Prof. Edward B. Poulton, of Oxford University 
(abstract) : Am.- Assoc. Adv. Sci. Proc, vol. 46 
p. 239, June, 1898. 

1902.207. Dohchocephaly and brachycephaly in the lower 

mammals: Am. Mus. Nat. Hist. Bull., vol. 16, 
pp. 77-89, Feb. 3, 1902. Abstract in New 
York Acad. Sci. Annals, vol. 14, No. 5, p. 147, 
March 4, 1902. 

1902.208. The four phyla of Oligocene titanotheres: Am. 

Mus. Nat. Hist. BuU., vol. 16, pp. 91-109, 
February 18, 1902. 

1902.212. Homoplasy as a law of latent or potent homology : 
Am. Naturalist, vol. 36, pp. 259-271, April, 
1902. 

1905.271. The ideas and terms of modern philosophical 
anatomy (condensed) : Science, new ser., vol. 
21, No. 547, pp. 959-961, June 23, 1905; Jour. 
Philosophy, Psychology, and Sci. Methods, 
vol. 2, No. 17, pp. 455-458, August 17, 1905. 

1907.303. Evolution as it appears to the paleontologist, 
address before the Seventh International Zoo- 
logical Congress, Section of Paleozoology, 
Boston, August, 1907: Science, new ser., vol. 
26, No. 674, pp. 744r-749, Nov. 29, 1907; 
Seventh Internat. Zool. Cong. Proc. Boston 
meeting, August 19-24, 1907, pp. 733-739, 
Cambridge, Mass., 1912. 

1908.308. The four inseparable factors of evolution, theory 
of their distinct and combined action in the 
transformation of the titanotheres, an extinct 
family of hoofed animals in the order Perisso- 
dactyla: Science, new ser., vol. 27, No. 682, 
pp. 148-150, January 24, 1908. 

1908.314. Coincident evolution through reotigradations 
(tliird paper): Science, new ser., vol. 27, No. 
697, pp. 749-752, May 8, 1908. 

1909.327. Darwin and palaeontology, in Fifty years of 
Darwinism, pp. 209-250, New York, Henry 
Holt & Co., April, 1909. 

1911.353. "Mutations" of Waagen and "mutations" of 
De Vries or [and] "rectigradations" of Osborn: 
Science, new ser., vol. 33, No. 844, p. 328, 
March 3, 1911. 

1911.356. Biological conclusions drawn from the study of 
the titanotheres: Science, new ser., vol. 33, 
No. 856, pp. 825-828, May 26, 1911. 

1912.362. Darwin's theory of evolution by the selection of 
minor saltations: Am. Naturalist, vol. 46, pp. 
76-82, February, 1912. 

1912.366. First use of the word " genotype " : Science, new 

ser., vol. 35, No. 896, pp. 340-341, March 1, 
1912. 

1912.367. Phylogeny and ontogeny of the horns of mam- 

mals: Science, new ser., vol. 35, No. 902, pp. 
595-596, April 12, 1912. 

1912.368. Skull measurements in man and the hoofed 

mammals: Science, new ser., vol. 35, No. 902, 
p. 596, April 12, 1912; New York Acad. Sci. 
Annals, vol. 22, pp. 341-342, April 20, 1913. 
1912.372. The continuous origin of certain unit characters 
as observed by a paleontologist, Harvey lec- 
ture: Am. Naturalist, vol. 46, No. 544, pp. 
185-206, April, 1912, No. 545, pp. 249-278, 
May, 1912; Harvey Soc. Volume, 7th ser., 
November, 1912, pp. 153-204 (with some 
alterations) . 



852 



TIT.'USfOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



OsBORX, Henry Fairfield — Continued. 

1912. 37S. Tetraplasy, the law of the four inseparable 
factors of evolution: Acad. Nat. Sci. Phila- 
delphia Jour., vol. 15, 2d ser., pp. 275-309, 
September 14, 1912. 

1912. 3S2. Craniometry of the Equidae: Am. Mus. Nat. 
Hist. Mem., new ser., vol. 1, pt. 3, pp. 57-100, 
figs. 1-17, June, 1912. 

1914.409. Recent results in the phylogenj' of the titano- 
theres: Geol. Soc. America Bull., vol. 25, No. 
3, pp. 403-405, September 15, 1914. 

1914.412. Rectigradations and allometrons in relation to 
the conception of the "mutations of Waagen," 
of species, genera, and phyla (abstract) : Geol. 
Soc. America Bull., vol. 25, No. 3, pp. 411-416, 
September 15, 1914. 

1915.416. Origin of single characters as observed in fossil 
and living animals and plants: Am. Naturalist, 
vol. 49, No. 580, pp. 193-239, April, 1915. 

1915.421. The origin of new adaptive characters [a replj']: 
Nature, vol. 96, No. 2402, pp. 284-285, No- 
vember 11, 1915. 

1916.426. Mutations of Waagen, Mutationsrichtung of 
Neumayr, mutants of De Vries; relations of 
these phenomena in evolution [title only]: 
Paleont. Soc. Proc, in Geol. Soc. America 
Bull., vol. 27, p. 148, March 31, 1916. 

1916.431. The origin and evolution of life upon the earth 
(part): Sci. Monthly, vol. 3, 1916, July, pp. 
5-22; August, pp. 170-190; September, pp. 
289-307; October, pp. 313-334; November, pp. 
502-513; December, pp. 601-614, iUus. (For 
complete text see below, 1917.462.) 

1917.442. Application of the laws of action, reaction, and 
interaction in life evolution: Nat. Acad. Sci. 
Proc, vol. 3, pp. 7-9, January, 1917. 

1917.462. The origin and evolution of life on the theory of 
action, reaction, and interaction, Hale lectures 
of the National Academy of Sciences, Wash- 
ington, April, 1916, 322 pp., frontispiece, 135 
figs., Charles Scribner's Sons, October 5, 1917. 
Osgood, Wilfred H. 

1909.1. Revision of the mice of the American genus Pero- 
myscus: U. S. Dept. Agr. Bur. Biol. Survey 
North Am. Fauna, No. 28, 285 pp., April 17, 
1909. 
Pearl, Raymoxd. 

1909.1 (and Surface, Frank M.). Is there a cumulative 
effect of selection?: Zeitschr. induktive Verer- 
bungslehre. Band 2, Heft 4, pp. 257-275, 1909. 

1917.1. The selection problem: Am. Naturalist, vol. 51, 
pp. 65-91, February, 1917. 
Plate, L. 

1909.1. Darwinismus und Landwirthschaft, 24 pp., Berlin, 
Paul Parey, 1909. 

POULTON, E. B. 

1909.1. Charles Darwin and the origin of .species, 302 pp., 
London, Longmans, Green & Co., 1909. 
PUNNETT, R. C. 

1911.1. Mendelism, 3d ed., 192 pp., Macmillan Co., 1911. 
Rbgnaltlt, Felix. 

1911.1. Le chien ectromele et les theories de Lamarck: 
Biologica, 1st year. No. 10, pp. 333-337, figs. 
1-5, Oct. 15, 1911. 
Ripley, William Z. 

1899.1. The races of Europe, a sociological study, 624 pp.. 
New York, D. Appleton & Co., 1899". 



Scott, William Berryman. 

1891.2. On the osteology of Mesohippus and Leptomeryx, 

with observations on the modes and factors of 

evolution in the JMammalia: Jour. Morphology, 

vol. 5, No. 13, pp. 301-406, pis. 22-23, figs. 1, 

A-G, December, 1891. 
1894.1. On variations and mutations: Am. Jour. Sci., 3d 

ser., vol. 48, No. 287, pp. 355-374, November, 

1894. 
1896.1. Palaeontology as a morphological discipline, 

Marine Biol. Laboratory Woods Hole, Biol. 

Lectures, pp. 41-61, Boston, Ginn & Co., 1896. 
1898.4. The methods of palaeontological inquiry. Marine 

Biol. Laboratory Woods Hole, Biol. Lectures, 

pp. 59-78, Boston, Ginn & Co., 1898. 
Stevenson-Hamilton, Major J. 

1912.1. Animal life in Africa, 539 pp., illus., London, Wm. 

Heinemann, 1912. 
Surface, Fr-ank M. 

1909.1. See Pearl, Raymond. 

SECTION 2. NATURAL SELECTION IN MAMMALS; 
CAUSES OF THE EXTINCTION OF THE TITANO- 
THERES AND OTHER QUADRUPEDS 

In section 1 of this chapter we have treated the 
origin and evolution of single biocharacters, studied in 
the light of the Lamarckian, the Darwinian (through 
natural selection), and the tetrakinetic theories. In 
section 2 we shall consider the mammals as a whole, 
inquire as to the causes of their survival and extinc- 
tion, and determine how far these processes are 
attributable to variations and fluctuations in single 
biocharacters and how far to the combinations of 
biocharacters that make up organisms. 

EXTINCTION OF FAUNAS IN THE AGE OF MAMMALS 

The sudden extinction of the titanotheres is one of 
the most striking phenomena of the age of mammals; 
it has generally been dismissed in a few words, but it 
is by no means so simple as it at first appears, because 
it seems to call into question ail the laws of natural 
selection and the relation of these laws to various 
degrees of fitness or adaptation. 

Among the more general questions of titanothere 
family evolution are (1) causes favoring increase in 
size, increase in specialization, and multiplication of 
the titanotheres; (2) causes bringing about diminution 
in size, arrest of specialization, diminution of number, 
and elimination of the phyla; (.3) causes terminating 
in the extinction of the family. 

The great law of mammalian adaptation through 
the elimination of the least adaptive becomes less 
sweeping in its effects as geologic time advances and 
the Mammalia become more highly perfected. Thus 
extinction is neither on the same grand zoologic scale 
nor chiefly due to the same causes through the succes- 
sive geologic epochs of the Tertiary period. During 
the middle of the Eocene epoch extinction is chiefly 
that of inadaptive orders of archaic mammals; in the 
late Eocene and through Oligocene time extinction is 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



853 



preeminently that of inadaptive families. Miocene 
time completes the elimination of inadaptive families 
and is broadly characterized by the extinction of in- 
adaptive genera, a process which continues through- 
out Pliocene time. The especial feature of Pleistocene 
extinction is the ruthless and world-wide regional 
extermination (as distinguished from extinction) of 
certain highly adaptive mammals, including both 
genera and species, over great regions of the world. 

Eocene extinction of orders appears to be due chiefly 
to the competition between lower and higher orders of 
mammals — that is, between mammals respectively 
more or less adaptive in intelligence and in skeletal 
and dental structure. In Oligocene time there 



a dying tree, decay in one branch after another and 
finally disappear entirely; or a very flourishing order 
like the Perissodactyla of lower Oligocene time, with 
its nine or more families, may be cut down to three 
families in a comparatively brief geologic period, and 
its world-wide geographic distribution may narrow 
down tb a few favorable regions. (See fig. 757.) 

EXTINCTION OF BOTH THE ADAPTIVE AND THE 
INADAPTIVE 

Some obviously inadaptive zoologic branches, such 
as the titanotheres, may be cut off, and again some 
highly adaptive and resourceful animals, such as the 
horses of the Pleistocene, may yield to new dangers. 




Figure 757. — Extinction of archaic mammals 

Titanothere epoch dotted. Struggle for existence between the archaic (solid black) and modernized (outline) kinds of clawed and hoofed 
mammals. The surviving insectivores and marsupials are the only remnants of the archaic type. 



occurred more distinctly the first fatal effects of the 
inadaptation of cjuadrupeds to environmental changes 
in living conditions, especially to changes in plant 
life. In Miocene and Pliocene time there occurred as 
a cumulative process the survival of the forms best 
adapted to the flora of the open and increasingly arid 
regions of the plains and uplands. In Pleistocene 
time all the direct and incidental effects of cold secular 
changes of climate appear to be the chief causes of 
extinction. 

GRADUAL OR SUDDEN EXTINCTION 

Even in immense periods of geologic time, extinction, 
preceded by regional extermination, may be extremely 
gradual or it may appear to be sudden or even cataclys- 
mal, as it seems to have been with the titanotheres. A 
once flourishing order like the Amblypoda may, like 



Adaptation is constantly varying in kind. Among the 
living races of Africa, for example, a continent which 
to-day gives us the closest parallel with Tertiary con- 
ditions of North America, we observe that elimination 
is constantly standardizmg mammals through destruc- 
tion of the least adaptive individuals and by this 
means is keeping the race up to the highest contem- 
porary plane. Among extinct races, such as the 
titanotheres, we observe that elimination is pruning 
off the most specialized phyla of all grades, especially 
those which exhibit the most extreme specialization. 
Each of these geologic phyla was subjected to com- 
petition as a whole with other phyla, which may have 
evolved contemporaneously; and elimination is the 
death penalty of the relative inadaptation of a phylum 
as a whole, inadaptation either to present or to chang- 
ing environment. 



854 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Changes of environment are almost infinitelj^ com- 
plex. Mammals which from various causes may be 
perfectly, even exquisiteh" adapted in nine hundred and 
ninety-nine out of a thousand biocharacters may be 
suddenly cut off through inadaptation in one remain- 
ing biocharacter; for example, through nonimmunity 
to certain infectious diseases (Smith, T., 1912.1) or 
through deficiency of a single organ. 



with diminution, impoverishment of numbers; teleo- 
logically, from the standpoint of adaptation, it is the 
arbiter of the fitness of whole phyla of organisms and 
their success in reproduction, the crucial test lying 
in the adaptation or inadaptation of correlated parts 
of the individual animals composing each phylum 
or in the collective capacity of the phylum as a whole 
to maintain itself. 




Figure 758. — Affinities and duration of nine families of Perissodactyla 

Extinction of titanotiicres (dotted). The titanotheres play a part in the early history of the order. Their period of extinction is coincident 
with that of the supposed aquatic rhinoceroses known as amynodonts and that of the light-limbed cursorial rhinoceroses known as hyraco- 
donts. These animals possessed the adaptive rhinoceros-like tooth structure, yet they became extinct coincidently with the titano- 
theres, which were provided with cone-and-crescent grinding teeth. The family of chalicotheres, which had cone-and-crescent teeth like 
those of the titanotheres, survived until late Pliocene time. 



PHASES OF EXTINCTION 

The process of extinction usually presents three 
phases, as follows: (1) Numerical diminution, local 
in its effects; (2) regional extermination or dis- 
appearance, local in effect; (3) extinction of certain 
races in all parts of the world. These phases are apt 
to be cumulative — that is, numerical diminution is 
apt to lead to regional extermination through a 
series of causes which may finally lead to world-wide 
extinction. Geographically extinction may be local, 
continental, or world-wide; numerically it may begin 



MULTIPLE CAUSES OF EXTINCTION 

The first generalization that will be drawn from the 
broad comparative study made by the author is that 
there are as many causes and modes of extinction as 
there are vulnerable points in the structure, functions, 
habits, and life development of organisms, and as 
there are changes and vicissitudes of environment; in 
other words, the causes of elimination of individuals 
are almost infinitely varied. Whenever the rate of 
elimination exceeds the rate of reproduction there 
follows numerical diminution, which may lead to 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



855 



extinction. A second broad generalization is that 
unfavorable effects are cumulative — that is, mammals 
are so closely adjusted to their environment as a 
whole that one series of unfavorable conditions seems 
to lead to or to induce another series of unfavorable 
conditions, and the process of numerical diminution 
is intensified. 

Life development proceeds under the law of tetra- 
plasy; consequently we look for the causes of extinc- 
tion in the four tetraldnetic centers of capturing, 
storing, realizing, and reproducing vital energy, 
namely, 

1. Heredity: the germ, germ evolution; reproduction. 

2. Ontogeny: the body, habits and use, modifications, 
plasticity. 

3. Physical environment: geologic, climatic. 

4. Life environment: plants, animals. 

The older historic classification of these causes is as 
follows : 
I. Environmental, external causes: 

1. Geologic and physiographic changes of land masses 

and their connections. 

2. Climatic, periodic changes, secular cold, heat, 

moisture, etc. 

3. Biotic changes in environing plant, insect, bird, and 

mammal life. 
II. Internal causes: 

4. The law of natural selection. Survival value of 

single characters. Degrees of survival value in 
systematic divisions. Survival value of minute 
variations in single characters. Adaptation and 
inadaptation in size, in certain organs, in intelli- 
gence, in reproduction, in plasticity or accommo- 
dation, in immunity to disease, in specialization 
in the trend of de%'elopment, in the potentiality of 
further evolution, etc. 

THE lAW OF WATUEAI SELECTION 

In studying the law of natural selection we have to 
view it in three aspects, as follows: 

1. Selection as it concerns the origin, rise, development, 
competition, decline, and disappearance of each of the phyla of 
the titanotheres, for example, and finally of the family as a 
whole. 

2. Selection as it concerns each of the organs that compose 
the individual titanotheres, their rise, development, decline, 
and disappearance. 

3. Selection as it concerns or is supposed to concern each of 
the "Single characters" (biocharacters) of which these various 
organs are composed. 

The theory of natural selection was originally 
stated independently by Darwin and Wallace.*^ 
Darwin" writes: 

Can it be doubted that * * * ^uy minute variation in 
structure, habits, or instincts adapting that individual better 

^3 Darwin, Charles, and Wallace, A. E., Onthetendency of species to form varie- 
ties and on the perpetuation of varieties and species by natural means of selection: 
Linneau See. (Zoology) Jour. Proc, vol. 3, pp. 45-62, London, August, 1858. 

"^ Darwin, Charles, Extract from an unpublished worlv on species, consisting 
of a portion of a chapter entitled " On the variation of organic beings in a state of 
nature; on the natural means of selection; on the comparison of domestic races and 
true species": Idem, pp. 46-50. Expanded and published as "On 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, November, 1859. 



to the new conditions would tell upon its vigor and health? 
In the struggle it would have a better chance of surviving; and 
those of its offspring which inherited the variation, be it ever so 
slight, would also have a better chance. 

Wallace^' writes: 

All the variations from the typical form of a species must 
have some definite effect, however slight, on the habits or 
capacities of the individuals. * * * if^ on the other hand, 
any species should produce a variety having slightly increased 
powers of preserving existence, that variety must inevitably in 
time acquire a superiority in numbers. * * * Xhe variety 
would now have replaced the species, of which it would be a 
more perfectly developed and more highly organized form. 
It would be in all respects better adapted to secure its safety 
and to prolong its individual existence and that of the race. 

From 1858 until to-day one link in the Darwin- 
Wallace chain of thought has been a contested point 
in the theory of natural selection. It may be para- 
phrased thus: 

Any minute heritable or germinal variation, be it 
ever so slight, which better adapts an individual to 
new conditions will give that individual and those of its 
offspring which inherit the variation a better chance 
of survival. Or, stated more briefly and in more 
modern terms: Germinal variations, be they ever so 
minute or slight, have a chance of being preserved 
and accumulated if they have adaptive or survival 
value. 

As shown in section 1 of this chapter, very careful 
examination of the actual examples cited by Darwin to 
illustrate his meaning demonstrates that under 
"minute variations" he had in mind two kinds of 
variation, which we now distinguish as follows: 

1. Minute individual variations, heritable, corres- 
ponding closely with the "mutations of De Vries.' 
These Darwin believed to be the chief cause of 
evolution. 

2. Variations of proportion, now known as fluctua- 
tions, as in Darwin's explanation of the origin of the 
long neck of the giraffe, where he observes that any 
increase in the length of the neck of a giraffe would 
increase the animal's chance of survival in times of 
drought. The evidence for these hypotheses will now 
be reexamined, and the conclusions regarding natural 
selection will be summed up at the end of this chapter. 
These hypotheses were also examined at the end of 
section 1, especially in relation to the principle of 
tetraplasy and the tetrakinetic theory. 

Does the actual order of evolution of the titanotheres 
favor the original Darwin-Wallace hypothesis? Do 
these animals evolve in the Darwin-Wallace way? 

The independent development of hundreds if not 
thousands of distinct characters in different parts of 
the individual titanotheres has put before us with 
crystalline clearness this crucial test of the Darwin- 
Wallaxje theory of selection — namely, what observable 

" Wallace, A. E., On the tendency of varieties to depart indefinitely from the 
original type: Idem, pp. 53-62, reprinted with slight changes in " Contributions to 
the theory of natural selection," London, Macmillan & Co., 1870. 



856 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



influence has natural selection on the origin and trans- 
formation of biocharacters, on rectigradation and pro- 
portion? What degrees of heritable variation in single 
characters hare a sin-vival value that can be proved by 
observation or experiment? 

To answer this question finally it is necessary, first, 
to consider the laws of survival and elimination acting 
upon quadrupeds as a whole. These laws have been 
made the subject of a long special investigation by 
the author for this monograph, the results of which 
were published in 1906 in a paper entitled "The 
causes of extinction of mammals" (1906.287). Later 
results were published in 1910 in "The age of 
mammals in Europe and North America" (1910.346). 

HISTORY OF OPINION 

Although the mam trend of the present inquiry as 
to the environmental and the internal causes of ex- 
tinction had been suggested by the middle of the nine- 
teenth century, discovery and observation since 
LyeU's and Darwin's time furnish new material for 
induction both as to the environmental and the 
internal causes of extinction. The true method of 
inquiry is to make close and continuous comparison 
of the present with past ages, based on the uniformi- 
tarian doctrines of Lyell and Darwin. Although we 
know few positive cases of natural extinction in our 
times, we must study all possible causes of numerical 
diminution, for we have every reason to believe, with 
Darwm, that numerical diminution is one of the high- 
roads to actual extmction. The conception of the 
similarity of the past and present causes of survival 
and of the numerous internal causes of variation, 
development, and decline only gradually took its 
modern form. 

CATACLYSMAL HYPOTHESKS 

Whewell (1837.1) clearly set forth the history of 
opinion between 1796, the Buffon-Cuvier period, and 
1837, the year of the publication of his "History of 
the inductive sciences." 

Buffon (1749.1), in commenting on the giant extinct 
fauna of northern Asia and Siberia, pointed out that 
parts of the globe now submerged were formerly ele- 
vated; he adumbrated the idea of the separation of 
faunas, such as the mammoths of Siberia and America, 
by continental depression, wliich resulted in the sub- 
mergence of old migration routes. He attributed the 
disappearance of these animals from the north partly 
to refrigeration and partly to migration toward the 
south. 

Cuvier (1825.1) more fully developed the cataclys- 
mal hypotheses of regional extermination through sub- 
mergence and through excessive refrigeration. In his 
"Discours" he observes: 

Let us suppo.se that a great invasion of the sea covers with a 
mass of sand or other deposits the continent of Austraha; it 



would bury the carcasses of tlie kangaroos, wombats, dasyures, 
bandicoots, flying phalangers, as well as of the duckbills 
{Ornithorhynchus) and spiny anteaters {Echidna). It would 
entirely destroy the species of all these animals, because none 
of them exist in any other country. 

After advancing the hj^pothesis that Australia thus 
depopulated might again be populated from Asia, he 
continues: 

To carry the hypothesis still further, after the Asiatic animals 
had migrated into Australia let us imagine that a second revo- 
lution destroyed Asia, the original home of these animals. 

In other words, Cuvier believed in the hypothesis of 
a total depopulation of a continent through a great 
and sudden physical revolution. This was in keeping 
with the cataclysmal geologic theories of the time. 
He explained the repopulation of continents through 
migration from other continents in which life had per- 
sisted. This idea was taken up by Lyell. 

D'Orbigny appeared as a disciple of Cuvier in his 
travels in South America. He writes: 

I was in a position to study the effects of inundations on the 
mammals of the province of Moxos (Bolivia), where these 
inundations are periodical, and I am certain that there the 
animals instinctively fi}' from the fluvial tide and take refuge 
in places further removed from the water and on pieces of 
high ground, where they find themselves momentarily congre- 
gated together. Ruminants sometimes die for want of food 
under such circumstances, and the natives mention certain 
years when this has occurred, but their bodies remain far from 
the rivers on small plateaux, or in the depths of the forests. 

In summing up his results as to the destruction of 
the Pampean Pleistocene fauna D'Orbigny (1835.1) 
says that it is to the sudden rise of the Cordilleras 
he attributes the sudden movement of the sea, which 
invaded the continent all at once, carried off and over- 
whelmed the mastodons which inhabited the eastern 
flanks of the Bolivian Cordillera, the megatheriums, 
megalonyxes, and the multitude of animals daily being 
discovered in the caverns and the fissures of the 
mountains of Brazil — all the species, in fact, which 
are extinct. Again he observes: 

My final conclusion from the geological facts I observed in 
America is that there was a perfect coincidence between the 
upheaval of the Cordilleras and the destruction of the great 
race of animals and the great deposit of Pampas mud. 

The stable continents. North America and Africa, 
underwent moderate fluctuations of land area in the 
Tertiary period as compared with the highly unstable 
continents of Europe, Australia, and the southern half 
of South America. Europe was the scene of alter- 
nating marine and fresh-water conditions, of varying 
coast lines, of insular and archipelagic land masses — 
changes which are all to be more seriously studied in 
connection with extinction than they have been here- 
tofore. 

It must be stated at once that the grand phenomena 
of extinction in unstable Europe from basal Eocene to 
Pliocene time broadly coincide with those observed in 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



857 



the stable continent of America. This very striking 
fact shows, first, how cautious we must be in connecting 
extinction directly with physiographic changes; sec- 
ond, that extinction from internal causes or through 
competition with other species is more manifold than 
extinction through external causes or the struggle with 
environment. 

Wallace (1876.1) discussed the extinction of the large 
Pliocene and early Pleistocene mammals of Australia 
chiefly from the standpoint of a physiographer — that 
is, he attributed Australian extinction chiefly to the 
increased competition, or struggle for existence, caused 
by the progressively contracted land area due to sub- 
sidence; also to possible glacial conditions. 

The most recent advocate of this now abandoned 
cataclysmal hypothesis is Howorth. In his interest- 
ing and encyclopedic work "The mammoth and .the 
flood" (Howorth, 1887.7, pp. xvii-xviu) he marshals 
a large number of facts in support of the hypothesis of 
widely destructive floods both in Asia and South 
America in Pleistocene time. He observes: 

These facts, I claim, prove several conclusions. Tiiey 
prove that a very great catastrophe or cataclysm occurred at 
the close of the Mammoth period, by wMch that animal, with 
its companions, were overwlielmed over a very large part of the 
earth's surface. Secondly, that this catastrophe involved a 
very widespread flood of water, which not only kiUed the ani- 
mals but also buried them under continuous beds of loam or 
gravel. Thirdlj', that the same catastrophe was accompanied 
by a verj' great and sudden change of climate in Siberia, by 
which the animals wliich had previously Hved in fairly tem- 
perate conditions were frozen in their flesh under ground and 
have remained frozen ever since. 

UNIFORMITARIAN THEORIES 
lYEH ON EXTINCTION 

Development of opinion. — Under the influence of the 
geologist Hutton the fu'st note for modern methods of 
explanation and research as to the causes of extinction 
was struck by Lyell (1830.7). 

Lyell's remarkable discussions of "changes of the 
organic world now in progress" and their bearing on 
the phenomena of life in geologic times will be foimd 
in the first edition of his "Principles of geology" 
(1830.1, vol. 2), a volume which was dedicated 
December 8, 1831, and published in January, 1832. 
Darwin departed on his voyage December 27, 1831, 
and took Lyell's work with him; he was profoundly 
influenced by it. 

Lyell himself had been greatly influenced by Buf- 
fon's theories as to the destructive action of physio- 
graphic and climatic changes and by the uniformita- 
rianism of Lamarck; also by Cuvier in respect to the 
repopulation of devastated land areas by migration, 
although in other respects he was an archopponent 
of Cuvier's cataclysmal hypotheses. 

The following citations from Lyell's work, volume 
2, chapter 10, give a clear insight into his opinions at 
that time: 



[Centers of creation.] — For we assume, on grounds before 
stated [chapter 8] that the original stock of each species is intro- 
duced into one spot of the earth only, and, consequently, no 
species can be at once indigenous in the Arctic and Antarc- 
tic circles [p. 170]. * * * The following may, perhaps, 
be reconcilable with known facts: Each species may have had 
its origin in a single pair, or individual, where an individual was 
sufficient, and species may have been created in succession at 
such times and in such places as to enable them to multiply 
and endure for an appointed period, and occupy an appointed 
space on the globe [p. 124]. * * * Now this congregating, 
in a small space, of .many peculiar species would give an 
appearance of centers or foci of creation, as they have been 
termed, as if there were favorite points where the creative 
energy has been in greater action than in others and where 
the numbers of pecuhar organic beings have consequently 
become more considerable [p. 126], 

[Physiographic changes and accommodation.] — Each change in 
the physical geography of large regions must occasion the 
extinction of species [p. 158]. * * * Species we have stated 
are, in general, local, some being confined to extremely small 
spots and depending for their existence on a combination of 
causes which, if they are to be met with elsewhere, occur only 
in some very remote region. Hence it must happen that when 
the nature of these localities is changed the species will perish; 
for it win rarely happen that the cause which alters the character 
of the district will afford new facilities to the species to establish 
itself elsewhere [p. 166]. * * * if^ therefore, we admit 
incessant fluctuations in the physical geography, we must, 
at the same time, concede the successive extinction of terrestrial 
and aquatic species to be part of the economy of our system 
[p. 168]. * * * To pursue this train of reasoning farther is 
unnecessary; the reader has onl}^ to reflect on what we have 
said of the habitations and the stations of organic beings in 
general; * * * he will immediately perceive that, amidst 
the vicissitudes of the earth's surface, species can not be im- 
mortal but must perish one after the other, like the individuals 
wliich compose them. There is no possibiUty of escaping from 
this conclusion, without resorting to some h3-pothesis as violent 
as that of Lamarck, who imagined, as we have before seen, that 
species are each of them endowed with indefinite powers of mod- 
if3dng their organization, in conformity to the endless changes 
of circumstances to which thej' are exposed [p. 169]. * * * 
But the power of accommodation to new circumstances is great 
in certain species and might enable many to pass from one 
zone to another, if the mean annual heat of the atmosphere 
and the ocean were greatlj' altered [p. 171]. * * * This 
argument is applicable not merely to climate, but to any other 
cause of mutation. However slowly a lake may be converted 
into a marsh, or a marsh into a meadow, it is evident that 
before the lacustrine plants can acquire the power of living in 
marshes, or the marsh plants of living in a less humid soil, 
other species, already existing in the region and fitted for these 
several stations, will intrude and keep possession of the ground 
[p. 174]. * * * ^ faint image of the certain doom of a 
species less fitted to struggle with some new condition in a 
region which it previously inhabited and where it has to con- 
tend with a more vigorous species is presented by the extirpa- 
tion of savage tribes of men by the advancing colony of some 
civilized nation. In this case the contest is merely between 
two different races, each gifted with equal capacities of im- 
provement — between two varieties, moreover, of a species 
which exceeds all others in its aptitude to accommodate its 
habits to the most extraordinary variations of circumstances 
[p. 175]. 

[Climate.] — Some of the effects which must attend every 
general alteration of clirhate are sufficiently peculiar to claim a 
separate consideration before concluding the present chapter 



858 



TIT.-VNOTHBRES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



[p. 169]. * * * We have before stated that, during seasons 
of extraordinarj' severity, many northern birds and, in some 
countries, many quadrupeds migrate soutliwards. If tliese 
cold seasons were to become frequent, in consequence of a 
gradual and general refrigeration of the atmosphere, such 
migrations would be more and more regular, until, at length, 
many animals now confined to the Arctic regions would become 
the tenants of the temperate zone; while the inhabitants of the 
latter would approach nearer to the Equator [p. 169]. * * * 
But although some species might thus be preserved, every 
great change of climate must be fatal to many which can find 
no place of retreat, when their original habitations become unfit 
for them. For if the general temperature be on the rise, then 
is there no cooler region whither the polar species can take 
refuge; if it be on the decline, then the animals and plants 
previously established between the Tropics have no resource 
[p. 170]. * * * Let us now consider more particularly the 
effect of vicissitudes of climate in causing one species to give 
way before the increasing numbers of some other [p. 172]. 

* * * That they would be supplanted by other species at 
each variation of climate may be inferred from what we have 
before said of the known local exterminations of species which 
have resulted from the multiplication of others. Some minute 
insect, perhaps, might be the cause of destruction to the huge 
and powerful elephant [p. 174]. 

[Moving sands.] — If we attribute the origin of a great part 
of the desert of Africa to the gradual progress of moving sands, 
driven eastward bj' the westerly winds, we may safely infer 
that a variety of species must have been annihilated by this 
cause alone. The sand flood has been inundating, from time 
immemorial, the rich lands on the west of the Nile, and we 
have only to multiplj' this effect a sufficient number of times 
in order to understand how, in the lapse of ages, a whole group 
of terrestrial animals and plants may become extinct [p. 166]. 

* * * In a small portion of so vast a space, we may infer, 
from analogy, that there were many peculiar species of plants 
and animals which must have been banished by the sand, and 
their habitations invaded by the camel and by birds and insects 
formed for the arid sands [p. 166]. * * * If it be imagined, 
for example, that the aboriginal quadrupeds, birds, and otiier 
animals of Africa emigrated in consequence of the advance 
of drift sand and colonized Arabia, the indigenous Arabian 
species must have given way before them and have been 
reduced in number or destroj'ed [p. 167]. 

[Repopulation through migration.] — So great is the instability 
of the earth's surface that if Nature were not continually 
engaged in the task of sowing seeds and colonizing animals, 
the depopulation of a certain portion of the habitable sea and 
land would in a few years be considerable [p. 158]. * * * jf^ 
therefore, the Author of Nature had not been prodigal of those 
numerous contrivances before aUuded to, for spreading all 
classes of organic beings over the earth — if he had not ordained 
that the fluctuations of the animate and inanimate creation 
should be in perfect harmony with each other, it is evident 
that considerable spaces, now the most habitable on the globe, 
would soon be as devoid of life as are the Alpine snows, or the 
dark abysses of the ocean, or the moving sands of the Sahara 
[p. 1591. 

Summary. — Summing up Ly ell's opinions of 1831 
we find them remarkably modern and Darwinian, 
both in. observation and in reasoning. The chief 
points as to extinction are the following: (1) The 
destructive effect of physiographic changes in Ter- 
tiary time (p. 308) is to be interpreted through the 
extinction of species at the present time; (2) tempera- 
ture barriers are important factors in distribution 



(p. 172); (3) the destructive action of floods (p. 199) 
may be understood through the disappearance of large 
numbers of horses in South America (pp. 249, 312) 
and of buffaloes in India (p. 250); (4) animals may 
have also perished in large numbers in bogs (p. 217); 
(5) changes of climate induced by physiographic 
changes (p. 308) influence both distribution (p. 169) 
and competition between species (p. 172); (6) com- 
petition between related species is illustrated by that 
between the horse and ass, as introduced into South 
America; (7) competition between unrelated animals 
is illustrated by the unchecked increase of goats 
(p. 153), of asses (p. 153), of cattle and horses (pp. 
152-153), and of insects (p. 320); (8) increase, how- 
ever, suffers many checks, as was first noted by Buffon 
(p. 154); (9) there is constant competition between 
established species (p. 142) and newly introduced 
species; (10) the balance of nature (pp. 133, 138, 139) 
is brought to an equilibrium through the relations of 
plants and insects (p. 132), and unlimited increase is 
subject to other checks (pp. 133, 138, 139); (11) many 
species might be destroyed through the devastation of 
plant life caused by locusts (p. 137), or by the advance 
of sands and sandstorms in desert regions (p. 166), as 
illustrated in the burial of camels and other animals; 
(12) plagues of ants would be highly destructive of 
certain types of animals (p. 137). Under the influence 
of Lamarck, Lyell observed that extinction might be 
averted by "accommodation" to a new environment, 
but he argued that under many conditions accommo- 
dation would not be rapid enough, as Lamarck sup- 
posed, to avert extinction. 

Especially interesting is Lyell's calculation (p. 183) 
of the time required for the extinction of one species. 
If the earth were divided into twenty regions of equal 
area one of these would about equal the dimensions 
of Europe and might contain a twentieth part of the 
million species which we would suppose to exist. In 
such a region one species, according to the assumed 
rate of mortality, would perish in 20 years, or 5 
species out of 50,000 in the course of a century. In 
the class Mammalia it would require more than 8,000 
years to lose one group in a region of the dimensions 
of Europe. 

BAIANCE OF NATURE (lYEII, DARWIN, WALLACE) 

Lyell. — Referring to that subtle adjustment of the 
sum of all internal and external causes known as the 
balance of nature, Lyell (1872.1, vol. 2, pp. 455-456) 
observed: 

Every new condition in the state of the organic or inorganic 
creation, a new animal or plant, an additional snow-clad moun- 
tain, any permanent change, however slight in comparison to the 
whole, gives rise to a new order of things and may make a 
material change in regard to some one or more species. Yet a 
swarm of locusts, or a frost of extreme intensity, or an epidemic 
disease may pass away without any great apparent derange- 
ment; no species may be lost, and all may soon recover their 
former relative numbers, because the same scourges may have 



CAUSES OF THE EVOI;XJTION AND EXTINCTION OF THE TITANOTHERES 



859 



visited the region again and again, at preceding periods. Every 
plant that was incapable of resisting such a degree of cold, every 
animal which was exposed to be entirely cut off by an epidemic 
or by famine caused by the consumption of vegetation by the 
locusts, may have perished already, so that the subsequ'int 
recurrence of similar catastrophes is attended only by a 
temporary change. 

As a geologist Lyell believed that the destructive 
influence of geologic and physiographic changes was 
extremely gradual. In 1863 (1863.7, p. 374) he 
observed : 

It is probable that causes more general and powerful than the 
agency of man, alterations in climate, variations in the range 
of many species of animals, vertebrate and invertebrate, and of 
plants, geographical changes in the height, depth, and extent of 
land and sea, some or all of these combined, have given rise, in a 
vast series of years, to the annihilation not only of many large 
Mammalia but to the disappearance of the Cyrena fluminalis, 
once common in the rivers of Europe, and to the different range 
or relative abundance of other shells which we find in the 
European drifts. 

Darwin. — The next great student of this subject was 
Charles Darwin. His voyage around the world 
(1831-1836) on the exploring ship Beagle (1839.1) 
afforded him an extraordinary opportunity of testing 
the theories advanced in Ly ell's "Principles" as first 
published in the previous year (1831). Especially in 
the Pampean and Patagonian regions he contrasts the 
BufTon-Cuvier and the Lyell hypotheses in favor of the 
latter. He dismisses catastrophic causes and in 
general attributes extinction to a cessation of those 
world-wide conditions of life which were favorable to 
the larger quadrupeds in Europe, Asia, Australia, 
North and South America. In South America and 
elsewhere (1) he does not favor the extreme theory of 
the destructive influence of the glacial epoch, and he 
cites the supposed postglacial survival of Macrau- 
chenia and Mastodon. "It could hardly have been a 
change of temperature," he observes (p. 170), "which 
at about the same time destroyed the inhabitants of 
tropical, temperate, and Arctic latitudes on both sides 
of the globe." (2) He dismisses the possibility of 
extinction by man; (3) also of an extended drought 
in South America, calling attention to the Pampean 
horse as an animal which could have survived a 
drought. 

In seeking to establish a general law of extinction 
Darwin makes the following propositions: (1) Ani- 
mals naturally increase in geometrical ratio; (2) the 
food supply, however, remains constant; (3) any 
great increase in numbers is thus impossible and must 
by some means be checked; (4) we are seldom able 
to state the cause of this check beyond saying that it 
is determined by some slight difference in climate, 
food, or the number of enemies; (5) we are therefore 
driven to the conclusion that causes generally quite 
inappreciable by us determine whether a given species 
shall be abundant or scanty in numbers; (6) compara- 
tive rarity is the plainest evidence of less favorable 



conditions of existence; (7) rarity frequently precedes 
extinction, and if the too rapid increase of species, 
even the most favored, is steadily checked, why 
should we feel such great astonishment at the rarity 
being carried a step farther to extinction? 

Darwin's earlier views (1839-1845), developed 
from observations made during his voyage, on droughts, 
floods, insect life, epidemics, were more fully elab- 
orated in connection with the publication of his 
theory of natural selection (1858, 1859). In his 
"Origin of species" (1859) he discusses more fully the 
checks to increase as follows: (1) Climate as directly 
unfavorable; (2) climate as indirectly unfavorable by 
favoring other forms or by increasing the number of 
certain competitors; (3) unchecked increase frequently 
followed by epidemics, possibly in part by facility of 
diffusion of parasites among the crowded animals; 
(4) finally, since a large stock of individuals, relatively 
to the number of enemies, is absolutely necessary for 
the preservation of a species, a diminished number 
would tend to extinction; (5) any form (p. 133) that is 
represented by few individuals will run a good chance 
of utter extinction during great fluctuations in the 
nature of the seasons or from a temporary increase in 
the number of its enemies; (6) diminution in number 
presents less opportunity for producing favorable 
variations, hence rare species will be less quickly 
modified or improved within any given period. 

Wallace. — Wallace in his long series of contributions 
.to the natural selection and Darwinian theory devel- 
oped similar views (1858.1). In "Natural selection" 
(1870.1, p. 14) he observes: 

To discover how the extinct species have from time to time 
been replaced by new ones down to the very latest geological 
period is the most difficult and at the same time the most 
interesting problem in the natural history of the earth. * * * 
Whenever the physical or organic conditions change, to however 
small an extent, some corresponding change will be jDroduced 
in the flora and fauna, since, considering the severe struggle for 
existence and the complex relations of the various organisms, 
it is hardly possible that the change should not be beneficial 
to some species and hurtful to others. 

The majority of the speculations of these great 
naturalists are abundantly confirmed by modern 
paleontology. The lines of thought and investiga- 
tion developed by Lyell and Darwin are precisely 
those which we are pursuing to-day, but our greatly 
expanded knowledge of paleontologic history, of the 
biologic results of physiographic revolutions, of rela- 
tively sudden extinctions at the close of certain 
periods, of the means of the spread of epidemics by 
insect agencies, of the elimination of certain structural 
types opens up entirely new fields of actual observa- 
tion and comparison for the formation of more precise 
and definite conclusions. In short, we are in a posi- 
tion to substitute for the ingenious and profound 
speculations of Lyell and Darwin a number of con- 
crete examples of extinction that can certainly be 
traced to definite and specific causes. 



860 



TITAXOTHERES OF ANCIEiSTT WYOMING, DAKOTA, AND NEBRASKA 



ENVIRONMENTAI CAUSES OF EXTINCTION 

THE PHYSICAL EXVIKONMENT 

PHYSIOGKAPHIC CHANGES 

In our inquiry as to the causes of extinction we may 
first consider those which originate witli clianges in 
the environment brought about by geographic revo- 
lutions — such clianges in land masses, including eleva- 
tion and subsidence, as act directly upon all the physi- 
cal conditions of climate, moisture, or desiccation, 
changing vegetation, etc., and facilitating or restrict- 
ing migration, with resulting new competitions, etc. 

Base-leveling a cause of extinction. — The effects of 
base-leveling, or the erosion of mountainous and other 
land surfaces to a single peneplain, has been especially 
discussed by Woodworth (1894.1) and Adams (1901.7; 
1904.1). Woodworth points out that the develop- 
ment of the peneplain, the wideniug of the lowlands, 
plains, and jungles nearly to tide level in Cretaceous 
time, was highly favorable to the water-loving Reptilia 
and imfavorable to mammalian life. New species are 
brought into competition when a mountain system is 
leveled off, thus throwing the life of its opposite slopes 
into the same field. The degradation of uplands has 
a very direct effect, since organisms suited to steep 
slopes and high altitudes with low temperatures must 
migrate, vary, or live at a disadvantage as the surface 
is lowered by denudation. Other things being equal, 
the endemic lowland forms will have an advantage 
over those organisms which are living under the trial 
of alternating environment with the added stress of 
contest with hitherto unmet species. Periods of base- 
leveling are characterized by relative stability of the 
land with refei'ence to the sea, while periods of glacia- 
tion are characterized by relative instability. 

Insular conditions. — The substitution of insular for 
archipelagic or continental conditions by subsidence 
has undoubtedly been a potent cause both of extermi- 
nation in certain localities and of the survival of 
geologically ancient primitive forms (Wallace), such 
as Monotremata and Marsupialia, in the AustraUan 
region. It may be said at once, with Lyell, that 
most of the causes both of survival and extinction 
which prevail on continents are intensified on islands; 
yet on islands the phenomena are those of local 
extermination and modification rather than the 
general extinction of families and orders, which is our 
real subject of inquiry. 

Dwarjed Pliocene and Pleistocene island, lije. — In the 
islands of Malta, Cyprus, and Crete, as explored by 
Miss Bate (1905.1), we have fine examples of compara- 
tively recent insulation. 

It appears probable that Cyprus became an island 
first, because (1) no submerged bank connects it 
with the mainland, and the 200-fathom line is reached 
within a short distance of the coast line; (2) the ter- 
restrial fauna and avifauna include several distinct 



races peculiar to the island, a fact confirmed by 
Kobelt from his study of the recent Mollusca. The 
reduced existing Cyprus fauna contains a mingling of 
European and north African forms and shows the 
effects of deforestation in historic times. The largest 
animal on the island is the moufflon {Ovis opMon), 25 
inches high at the shoulders; yet this is the smallest 
of all the wild sheep and is related to east Persian 
species. 

The affinity of Malta to Sicily is indicated by the 
occurrence of two species, Hippopotamus pentlandi 
and Elephas mnaidriensis, in the cavern deposits of 
both islands. The early separation of Cyprus is 
indicated by the fact that E. Cypriotes and H. minu- 
tus are both more primitive than the Maltese-Sicilian 
species. Crete also includes antelope and deer in its 
Pleistocene fauna. 

Pleistocene extinct fauna of the Mediterranean islands 



Cyprus- 
Malta __ 

Sicily.-- 
Sardinia 



Proboscidea (pigmy elephants) 



E. Cypriotes ._ 

E. melitensis, E. mnaidri- 
ensis. 

E. mnaidriensis 

E. lamormorae 



Artiodactyla (pigmy 
hippopotami) 



H. minutus. 
H. pentlandi. 



H. pentlandi. 



The occurrence of these specifically different though 
apparently closely related races of small elephants and 
hippopotami in widely separated islands is an instance 
of independent development with some divergence 
from common ancestors. 

Volcanic ash. — The comparatively recent recogni- 
tion that vast areas, both in the Rocky Mountain re- 
gion and in Patagonia, were the seats of deposition of 
volcanic dust arising from volcanic eruptions indicates 
that this may be considered among the means of 
numerical diminution if not of extermination in certain 
regions. There is evidence in Patagonia especially 
that many animals were overtaken and buried entire 
in volcanic dust; but this is less frequently true of the 
volcanic dust deposits in the Eocene of North 
America, where few "entire" burials have been found. 

Floods. — It appears that floods have veiy frequently 
been a cause of marked diminution in number, which 
is a very different result from regional extermination 
or extinction. The instances cited by Howorth, by 
D'Orbigny, and by Azara (as quoted by Darwin) 
indicate that floods may be considered among the 
secondary or contributary causes of extermination. 

In the summer of 1867, it is recorded, over 2,000 
buffalo out of a herd of about 4,000 lost their lives in 
the quicksands of the River Platte while attempting 
to cross (Hornaday, 1889.7, p. 421). ' The floods and 
swollen rivers of Pleistocene time presented a new set 
of conditions. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



861 



Conclusions. — Since there is little parallel between 
the Eocene changes in Europe and those of western 
America, and a very close parallel between the phe- 
nomena of extinction in Europe and those of western 
America, the conclusion is inevitable that physio- 
graphic changes are not directly potent factors of ex- 
tinction but are indirectly potent through the biotic 
changes which they induce, new types of adaptation, 
new forms of competition, etc. 

CHANGES OF CLIMATE 

We have now to consider changes in temperature 
and moisture as brought about by geologic and physio- 
graphic changes and as- effecting in turn biotic 
changes — changes in the fauna and flora. 

Pleistocene secular increase of cold. — The effects of 
secular lowering of temperature must be analyzed with 
some care. At first sight the original theories of 
extinction of Buffon and Cuvier regarding refrigeration 
or direct action of cold are very simple and plausible, 
but on examining all the instances of extinction which 
occurred during the Pleistocene or glacial epoch we 
find that this simple or obvious explanation is not the 
true one. To take the glacial epoch, which affords 
the chief illustration, it is more in general accord with 
the facts to say that this period originated certain new 
conditions of life which diminished numbers and 
hastened extinction. These conditions include such 
phenomena as deforestation, enforced migrations, 
overcrowding, changes of food, unfavorable conditions 
of mating and reproduction, new relations to enemies, 
and other indirect results. The reason refrigeration 
can not be considered alone or as a direct cause lies in 
the remarkable powers of plastic adaptation which 
many mammals, including man, exhibit to secular or 
regional lowering of temperature. The horse of North 
America became extinct during the glacial epoch, 
yet there is every reason to believe that its extermina- 
tion was not directly connected with the Ice Age. 

General phenomena. — Among the observed climatic 
phenomena and biotic effects of imusual cold periods 
are the following: Harsh or unusual conditions of life 
caused by snowstorms, blizzards, ice floods, diminu- 
tion of food supply, limited choice of food caused by 
change of flora and deforestation, enforced choice of 
deleterious food, changes in fertility and reproduction 
rate, dangers to young, arrested growth, diminution 
in number causing diminished herds, enforced migra- 
tions, crowding southward. 

Protective adaptation. — Secular cold is very slowly 
progressive and has generally been accompanied step 
by step by progressive adaptation among the mammals 
to resist cold. Resistance depends upon (a) internal- 
heat producing power, which is a progressive adapta- 
tion of the ascending series of mammals as distinguished 
from the reptiles; (&) acquisition of a warm external cov- 
ering of wool and hair; (c) development of subcutane- 



ous and other fatty layers; (d) power of hibernation. 
The well-known instances of adaptation to extreme 
cold among the Proboscidea (Elephas primigenius, 
woolly mammoth. Mastodon americanus, mastodon), 
rhinoceroses {R. ticliorhinus, woolly rhinoceros), horses 
(E. przewalsTci), northern ruminants (Saiga antelope, 
Bactrian camel, Barren-Ground and Arctic caribou, 
musk ox) indicate again that we must not assume that 
refrigeration was a direct or sole cause of extinction. 

Glacial and postglacial extinction. — Wallace observes 
(1876.1, p. 151): 

We have proof in both Europe and North America that just 
about the time these large animals were disappearing all the 
northern parts of these continents were wrapped in a mantle 
of ice; and we have every reason to believe that the presence 
of this large quantity of ice (known to have been thousands of 
feet if not some miles in thickness) must have acted in various 
ways to have produced alterations of level of the ocean as 
well as vast local floods, which would have combined with the 
e.xcessive cold to destroy animal life. 

And again (1881.1, p. 117): 

We can therefore hardly fail to be right in attributing the 
wonderful changes in animal and vegetable life that have 
occurred in Europe and North America between the Miocene 
period and the present day, in part at least, to the two or more 
cold epochs that have probably intervened. These changes 
consist, first, in the extinction of a whole host of the higher 
animal forms; and, secondly, in a complete change of types 
due to extinction and emigration, leading to a much greater 
difference between the vegetable and animal forms of the 
Eastern and Western hemispheres than before existed. 

Wallace's views regarding glacial extinction are 
modified by the present four-glaciation theory. The 
large Afi-ican-Asiatic fauna of Europe, namely, the 
hippopotami, the rhinoceroses, the elephants, and the 
mammoths, survived the first, second, and third 
glaciations but perished before or during the extremely 
severe climatic conditions of the fourth glaciation. 

Freeh observes (1906.1): 

The chief problem in regard to extinction is in general 
confined to the question whether the great climatic changes of 
former geological periods coincide with the great transforma- 
tions of the plant and animal kingdoms. Corresponding with 
the fall in temperature at the close of the Tertiary in Europe 
is the disappearance of all the tropical and subtropical animals 
(tapirs, mastodons, hipparions, and the last tropical antelopes, 
Protragelaphus and P alaeoryx) . In the Quaternar}^, before the 
advent of the [last] glacial period the last remnants of the sub- 
tropical fauna vanished from central Europe (Hippopotamus 
major, M achaerodus) also the tropical elephants and rhinoc- 
eroses characterized like the recent forms by a practically 
hairless skin. The mammoth and the Arctic rhinoceros had 
in their long and dense woolly coat perhaps the best protection 
against cold that any terrestrial animal has ever possessed- 
The adaptations to cold, equally well defined in large and small 
animals of the circumpolar fauna, presuppose a long period of 
time for their development. The mammoth probably existed 
in Siberia under conditions of temperature similar to those 
prevailing at the present time, before the beginning of the 
glacial period. The general fall in temperature was probably 
first marked in northern Siberia and thus it was here that 
conditions first became favorable for the evolution of an Arctic 
fauna. The general drop of 4° C. in temperature moved the 



862 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



northern limit of trees further to the south. The large rumi- 
nants were thus forced southeast, southwest, and west, since 
the desert plateaus of central Asia shut off a retreat to the 
south. 

By far the most impressive of all the examples of 
extinction laiown to us is this continental extermina- 
tion of nunierous forms of mammalian life, both 
carnivores and herbivores, which was contemporaneous 
with this great secular change of climate. It is no 
exaggeration to say that the faunal aspect of Australia, 
South America, North America, Asia, and Europe 
was profoundly altered as it had never been before. 



This alteration entirely changed the zoologic aspect 
of four-fifths of the earth's surface, and it is unavoid- 
able that we should attribute it to the long series of 
direct and indirect changes connected with the four 
or five waves of advancing and retreating low tem- 
perature and moisture. 

The end of the Pliocene and beginning of the 
Pleistocene found North America populated with the 
kinds of great herbivorous quadrupeds tabulated 
below, all of which disappeared in North America 
during the Ice Age, although some of them survived 
in other forms in South America, Europe, and Asia. 



Herhivorous quadrupeds of North America that became extinct during the Ice Age 



Suborder 


Family 


Kind or genus 


Time of extinction 


Artiodactyla. - ._ .- 


Camelidae. 


Camels, llamas . _ .. 


Before the end of the Kansan stage; not known after the 








Aftonian stage. 


Perissodactj'la 


Equidae 








Aftonian except E. complicatus, which is found in 








Sangamon deposits. 




Tapiridae 

Mastodontinae- 




Before the Wisconsin, but after or during Sangamon. 


Proboscidea 


Mastodons ._ 


After the Wisconsin glacial stage. 




Elephantinae 


Elephants (mammoths) _ _ 


Do. 


Edentata 


Gravigrada 


Giant sloths: 








Megalonyx 


Do. 






Megatherium 


Probably before the end of the Kansan stage. 






Paramylodon 


Mylodon (Paramylodon) not known after the Wiscon- 
sin probably extinct during or after the Sangamon 
stage. 




Glyptodontia 




Probably became extinct in the first third of the Pleisto- 






cene, before the Kansan stage. 



It would be natural to assume that extinction was 
directly brought about by these successive changes of 
temperature and moisture and the changes in the 
fauna and flora consequent upon the great physio- 
graphic changes; but this simple explanation is beset 
with many difficulties and contradictions, and the 
results must be analyzed with some care. In Europe 
the Mediterranean Sea presented a barrier to escape 
or migration southward, but in North America there 
were broad continental areas and high plateaus that 
afforded easy routes of migration southward and every 
means of escape. It is therefore more in accord with 
the facts to say that contemporaneous with the 
glacial epoch in North America there were developed 
certain new conditions of life that directly or indirectly 
resulted in extermination. 

Numerical diminution in hard winters. — These 
phenomena have been discussed by Nehring (1890.1) 
in connection with the Ice Age in noting the effects 
of severe storms on the steppes of Russia, where 
animals are not protected by stretches of forest. 
Garman (1883.1) maintains that snow was the chief 
cause of the extermination of the Pleistocene species 
of bison and horses in North America. We owe to 
Seton (1909.1) a review of the observations of Mun- 



son, Spears, Bunn, Henry, and others of exceptional 
periods of numerical diminution on the Great Plains. 
Henry in 1799 (1897.1, vol. 1, p. 174, and footnote) 
counted "herds" of drowned and mired buffalo along 
the banks of Qu'Appelle River. Treachei'ous ice in 
the spring of 1801 destroyed great numbers of buffalo 
along the shores of Red River (Alberta) ; for two daj^s 
and nights a continuous line of carcasses floated by. 
The blizzard of 1871-72 diminished the last great 
buffalo herd of South Dakota. In 1880 large numbers 
of buffalo perished in a blizzard in the same region, 
50 skeletons being found together in one ravine. 
The hard winter of 1893 (thermometer -61° F.) 
killed off four-fifths of the antelope near Fort Assinni- 
boine, Mont. Carcasses of about 900 animals, which 
had starved to death on account of the deep snow, 
were found in one ravine. 

Dangers of numerical diminution. — While a sharp 
distinction must be drawn between actual extermi- 
nation on a continent and a temporary diminution 
caused, for example, by cold waves, floods, or other 
unfavorable conditions of life, it is very important to 
observe, as first suggested by Darwin, that extreme 
diminution in numbers has as great dangers as- 
extreme multiplication in numbers and may lead to- 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



863 



extermination. For example, a herd of animals may 
be reduced to the danger point in numbers so that 
they can no longer protect their young. Bell, former 
Acting Director of the Geological Survey of Canada, 
observes that the small surviving herd of woodland 
bison (Bison hison athabascae Rhoads) of British 
Columbia, although conserved by the Government, will 
probably be destroyed gradually through the Idlling 
of the calves by wolves, the bulls not being sufficiently 
numerous to protect the calves. 

Diminished herds and inbreeding of hison. — In a 
paper entitled "Das allmahliche Aussterben des 
Wisents (Bison honasus Linnaeus) iin Forste von 
Bjelowjesha," Eugen Biichner (1895.1) gives a detailed 
history of the bison herd in the Bieloviejsha (or 
Bialowitza) forest, Province of Grodno, in Lithuania, 
during the present century. 

A careful stud}' of the breeding habits of the bison in the 
Bieloviejsha forest and elsewhere leaves no room for doubt that 
the present slow rate of reproduction is an abnormal condition, 
and that to it is due the rapid approach of the extinction which 
is the certain fate of the herd under consideration. This dimin- 
ished fertility the author regards as a stigma of degeneration 
caused by inbreeding. * * * Another indication of the 
degenerate condition of the Bieloviejsha herd is seen in the 
great excess of bulls, which probably outnumber the cows two 
to one. This is doubtless a result of inbreeding, for Diising 
(1884.1) has shown that close inbreeding, like a reduced condi- 
tion of nutrition, is favorable to the production of an excess of 
males. * * * Jq conclusion, the author considers that his 
studies of the history of the Bieloviejsha bison leave scarcely 
room for doubt that inbreeding is the cause of the final extinc- 
tion of most large mammals. Inbreeding must begin and lead 
gradually but certainly to the extinction of a species when it, 
through any cause, has become so reduced in numbers as to be 
separated into isolated colonies. 

Numerical diminution of llamas. — The observations 
of Prichard (1902.1, pp. 132, 189, 255) in Patagonia 
afford an interesting instance of numerical diminution 
among the Camelidae. 

Around the lake lay piled the skulls and bones of dead game, 
guanaco {Lama huanachus) and a few huemules (Furcifer 
chilensis). These animals come down to live on the lower 
ground and near unfrozen water during the cold season, and 
there, when the weather is particularly severe, 'they die in 
' crowds. We saw their skeletons, in one or two places literally 
heaped one upon the other [p. 132]. * * * Again we came 
upon a second death place of guanaco, which made a scene 
strange and striking enough. There can not have been less 
than 500 lying there in positions forced and ungainly as the 
most ill-taken snapshot photograph could produce. Their long 
necks were outstretched, the rime of the weather upon their 
decaying hides, and their bone joints ghstening through the 
wounds made by the beaks of carrion birds. They had died 
during the severities of the previous winter and lay literally 
piled one upon another [p. 189]. * * * xhe meaning of 
this I gathered from Mr. Ernest Cattle. He told me that in 
the winter of 1899 enormous numbers of guanaco sought Lake 
Argentine and died of starvation upon its shores. In the 
severities of winter they seek drinking places, where there are 
large masses of water likely to be unfrozen. The few last 
winters in Patagonia have been so severe as to work great havoc 
among the herds of guanaco [p. 255]. 



Numerical diminution of Sirenia. — Interesting in- 
stances of the effects of "cold waves" in a subtropical 
region are those cited by Bangs (1895.1) regarding 
the rivers of Florida, especially in the case of the 
manatee (Manatus manatus). These animals are 
extremely sensitive to a lowering of temperature. 
During the winter of 1894-95 there was an unusually 
cold wave followed by marked niunerical diminution, 
several of these animals being found ashore dead. 

Deforestation and secular cold. — The main natural 
causes of deforestation appear to be (1) extreme heat 
and secular desiccation; (2) periodical fires that destroy 
the young trees; (3) excessive browsing, which destroys 
the young trees; (4) excessive cold resulting in pro- 
longed and deep snow mantling; (5) continuously 
frozen subsoil or tundra condition; (6) plagues of 
insects and other forest destroyers. 

After considering the former abundant mammalian 
life m Alaska, IVTaddren (1904.1, pp. 65-66) sum- 
marizes his conclusions as follows: 

I. That while remnants of the large Pleistocene mammal 
herds may have survived down to the Recent period and in 
some cases their direct descendants, as the musk ox, to the 
present, most of them became extinct in Alaska with the close 
of the Pleistocene. 

II. The most rational way of explaining this extinction of 
animal life is by a gradual changing of the climate from more 
temperate conditions, permitting a forest vegetation much 
farther north than now, to the more severe climate of to-day, 
which, subduing the vegetation and thus reducing the food 
supply besides directly discomforting the animals themselves, 
has left only those forms capable of adapting themselves to the 
Recent conditions surviving in these regions to the present. 

Influence of cold and snow on food supply aiid choice 
of food. — The death of great numbers of animals 
from hunger or starvation through the covering of 
food during the winter under heavy layers of snow is 
commonly observed among the large herds of some 
of the domesticated horses and cattle on the western 
plains. In fact, it is most probable that during the 
glacial epoch the great winter snow blankets covering 
the natural food were the chief cause of extinction 
rather than the actual influence of the cold itself. 

Under these conditions horses are driven to eat 
food which is very deleterious to them, such as the 
branches of willows. Under the influence of hunger 
cattle and sheep will feed eagerly and indiscriminately 
on plants that may be injurious to them or to their 
young, as recorded by Chesnut and others in the 
United States Department of Agriculture. The 
indirect results of hunger may therefore be quite as 
effective as actual starvation. 

Animals vary greatly in adaptability to new condi- 
tions caused by prolonged cold and heavy snowfall. 
Horses remove snow even to depths of 3 or 4 feet and 
find food to carry them through the winter, whereas 
under the same conditions cattle starve. 

Influence of cold during the reproduction period. — 
Exceptional cold waves or unusually prolonged cold 



S64 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



seasons may cause a temporary loss of food supply or 
cause the death of the young, which in northern lati- 
tudes are usually born in spring. The diminution or 
loss of yoiuig from this cause might act as the first of a 
series of destructive effects of a progressive secular 
change. These may be summarized as follows from 
actual zoological observations^" among the Cervidae: 
(1) Disturbed conditions during the conjugation 
(pairing, mating, rutting) period; (2) enfeebled 
(through hunger) condition of females during partu- 
rition period; (3) severe weather conditions, ice storms, 
crusted snow, prolonged wet and sleet at time of birth; 
(4) bulls unable to protect herds; (5) cows unable to 
protect young from Carnivora through starved con- 
dition, or abandoning them when attacked by wolves; 
(6) enfeebled and unprotected condition of quadru- 
peds favorable to increased food supply and conse- 
quent multiplication of cursorial and other Carnivora, 
especially Canidae and Felidae. 

These zoological observations are to a certain extent 
borne out in paleontology by Andrew Leith Adams's 
observation (1879.1, p. 98) of the exceptionally large 
number of milk teeth of elephants found in certain 
Pleistocene deposits, which appears to indicate a high 
mortality of the young. 

Merriam (1892.1), in what he has called the "law of 
temperature control," has directed attention to the 
physiological effects of a lowering of temperature upon 
diminished or increased fertility and the rate of repro- 
duction. It is stated as follows: Temperature, by 
controlling reproduction, predetermines the possibil- 
ities of distribution; it fixes the limits beyond which 
species can not pass; it defines broad transcontinental 
barriers within which certain forms may thrive if other 
conditions permit, but outside of which they can not 
exist, be the other conditions never so favorable, 
because the sexes are not fertile. 

In discussing how species are checked in their efforts 
to overrun the earth Merriam points out that more 
potent than geographic barriers are climatic bai'riers 
(as observed by Humboldt), and of these temper- 
ature is more potent than humidity. First, in 1892, 
Merriam attempted to show (1892.1, pp. 45-46) that 
the distribution of terrestrial animals is governed less 
by the yearly isotherm or mean annual temperature 
than by the total temperature during the period of 
reproductive activity and of growth (adolescence). 
This reproductive period in the Tropics extends over 
many months or nearly the whole year but within the 
Arctic Circle and summits of high mountains is of two 
months or less duration. Later results that Merriam 
(1894.1) obtained from extensive comparison of tem- 
peratures and distribution justified the belief that 
animals and plants (lower austral and tropical types 
coming from the south) are restricted in northward 

B^ Communicated by Mr. Madison Grant, secretary of tlie Zoological Society 
of New Yorli. 



distribution by the total quantity of heat prevailing 
during the season of development and reproduction. 
Conversely, animals and plants (upper austral, transi- 
tion, and boreal types coming from the north) are 
restricted in southward distribution by the mean tem- 
perature of a brief period covering the hottest part 
of the year. Thus in the transition zone boreal and 
austral types mingle in the equable climate of the 
Pacific coast of California, whereas they are sharply 
sepai'ated by the inequable extremes of cold and heat 
of the interior continental plateau. 

Lowering of temperature and diminished fertility. — 
The favorable influence of high temperature on fer- 
tility and reproduction is well illustrated in the early 
age of reproduction in tropical lands and the increase 
in the fertility of the human species toward the 
Equator, and as low temperature is thus a barrier to 
reproduction it is reasonable to suppose that the 
secular lowering of temperature may have been one 
among the causes of the extinction of animals during 
the glacial age. Thus certain mammals, although 
they were otherwise becoming adapted to the effects 
of cold and were discovering new means of feeding, 
may have become extinct through the subtle inhibi- 
tion of fertility and the lowering of the rate of repro- 
duction. 

Increasing moisture in Temperate Zone. — Humidity, 
observes Merriam (1894.1) in the work cited above, 
is a less potent factor than temperature in limiting 
the distribution of the mammals of North America 
[that is, in cold and temperate climates]. Thus 
many genera adapted to restrictions of temperature 
zones range east and west completely across the 
American Continent, inhabiting alike the humid and 
arid subdivisions, but no genus adapted to certain 
restrictions of humidity ranges north and south along 
the temperature zones. Thus, according to Merriam, 
humidity governs the details of distribution of a few 
species of mammals within certain temperature zones; 
temperature establishes the great wall, and humidity 
establishes the lesser barriers. 

This law may be understood as the direct influence 
of temperature and humidity respectively; the indi- 
rect or secondary influences of temperature and 
humidity may be entirely different. Thus Merriam's 
generalization would not apply to central Africa, 
Central America, or other tropical countries where 
insect and disease barriers exist which are generated 
and favored by high temperature and increasing 
moisture. Because of the connection between insect 
distribution and moisture in Africa, humidity is con- 
sidered to be a very potent factor in the distribution 
of animals. 

Moisture in the Equatorial Zone. — We observe that 
increased rainfall may have the following effects, 
especially upon the Herbivora : (1 ) It may diminish the 
supply of harder grasses to which certain quadrupeds 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



865 



had become thoroughly adapted; (2) it may intro- 
duce new poisonous or deleterious plants; (3) it may 
be the means of erecting new forest barriers or new 
forest migration tracts for certain Carnivora; (4) it 
may be the means of introducing new insect and other 
pests, as well as new insect barriers; (5) it may be the 
means of introducing new diseases and new insect 
carriers of disease. 

Besides the changes in plant food that are brought 
about by diminished moistui'e, as indicated below, 
there are the effects of increased moisture which may 
be equally if not more important. Moist conditions 
may cause increase of certain forms of plant life dele- 
terious to all Herbivora. Dry or moderately dry 
conditions, provided they are not too extreme, are 
generally more favorable to quadrupeds than moist 
conditions. The plateau and forest regions most 
densely populated with quadruped life, such as those 
of Africa, are regions of moderate rainfall and even of 
prolonged summer droughts. On the other hand, the 
regions least densely populated with mammals are 
those of heavy rainfall, dense forests and vegetation, 
such as those of the equatorial belt of South America 
and of Africa. 

It follows that periods of secular increasing moisture, 
such as the warm interglacial stages of Pleistocene 
time in the Northern Hemisphere, may have proved 
unfavorable to certain large quadrupeds through in- 
crease of humidity and rise in temperature even 
prior to the advent of extreme cold. 

Insect 'barriers and moisture. — It is now a matter of 
general observation that, especially in tropical, insect- 
infested countries, dry seasons result in the reduction, 
moist seasons in the increase of disease. Dry localities 
are generally favorable, moist localities generally 
unfavorable to quadruped life. For example, Shipley 
(1906.1) observes of the tsetse fly that its 

northern limit corresponds with a line drawn from the Gambia 
its southern limit is about on a level with the northern limit of 
■ Zululand. Most writers agree that the tsetse is not found in 
the open veldt, that it must have cover. Warm, moist, steam}', 
hollows, containing water and clothed with forest growth, are 
the haunts chosen. 

This subject will be treated at length under insect 
life as a factor in extinction. 

General ejfects of decreasing moistwe; secular desic- 
cation. — We observe that deci'easing moisture (1) 
changes the character of the food supply by diminish- 
ing the softer and more succulent vegetation and in- 
creasing the harder and more resistant vegetation; 
(2) increases the length and severity of the dry 
season; (3) removes forest barriers and admits new 
competitors; (4) eliminates animals incapable of 
traveling long distances for food and water, or living 
on a limited or irregular water supply; (5) favors 
grazing quadrupeds and eliminates browsing and 
forest-living quadrupeds; (6) favors hypsodont (long- 



toothed) and is inimical to brachyodont (short- 
toothed) quadrupeds; (7) favors cursorial (rapid- 
moving) and is inimical to mediportal (slow-moving) 
quadrupeds. 

Secular desiccation (Hann, 1903.1, p. 375) of certain 
very extensive regions has occurred in parts of all the 
great continents — North America, Australia, Asia, 
and Africa — and on each continent we observe a 
general concomitant modification and extinction of 
certain kinds of quadrupeds. In general, there ap- 
pears to have been a progressive decrease of moisture 
in the Northern Hemisphere, beginning in Oligocene 
time, which was checked only by the humidity of 
preglacial time. Increasingly prolonged summer 
droughts were characteristic of the late Miocene and 
Pliocene of Europe, and we are beginning to ac- 
cumulate evidence that the same conditions pre- 
vailed in North America. 

The great regions of the world where decreasing 
moisture has introduced a series of changes ending in 
the extinction of a great number of quadrupeds are 
(1) North America (Western Plains region, Interior 
Basin, arid Plateau and Mountain region), beginning 
in Oligocene time; (2) South America (Patagonia and 
Pampean region), beginning in late Pliocene time; 
(3) north-central Africa (the Fayum district), begin- 
ning in Oligocene time; (4) central Australia, begin- 
ning in Pleistocene time. 

American paleontologists, also Stirling (Australia), 
Andrews (Fayum), and Ameghino (Patagonia), de- 
scribe faunas adapted to much moister and more 
hospitable conditions than those which now prevail in 
the regions they consider. 

Sandstorms. — Bravard, who did much to explore 
the Pleistocene fauna of South America, concurring 
with Lyell, argued that the great South American 
mammals were overwhelmed alive by moving sand, 
as the simoom is said to overwhelm creatures in the 
Sahara. He invoked for the purpose an exaggerated 
form of the hurricane, still known in South America as 
the pampero, and urged that, having been thus 
killed, they were afterward covered by the sand. 
Against this hypothesis Burmeister observes that the 
greater number of the remains that are found are 
isolated skeletons, most of which, curiously enough, are 
without heads or tails. Such mutilations do not 
sustain the idea that the creatures were overwhelmed 
alive by moving sand. 

Secular desiccation and vegetation. — The indirect in- 
fluences of dry secular changes of climate on quad- 
rupeds are apparently quite as effective factors in 
extinction as the direct influences, such as changes in 
vegetation due to decrease of moisture, which make 
certain types of quadrupeds that were perfectly 
adapted to one kind of plant food largely or wholly 
inadapted to the new or altered kinds of food. 



866 



TITAN0THERE3 OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Influence of drougJits in central Africa. — The in- 
fluence of the gradual decrease of moisture in a country 
is clearly illustrated in the conditions which prevail in 
the African Continent to-day, as observed by Gregory, 
Foa, and Schillings. Thirst, like hunger, first drives 
quadrupeds to take extreme risks, which they would 
absolutely avoid during natural conditions. The 
drinking places or water pools during long seasons of 
drought become fewer in number and more widely 
separated, and large animals driven to them by thirst 
are more readily attacked and killed by Carnivora. 
The pools become separated by distances of 30 or 40 
miles, thus necessitating long excursions to and from 



that the central Sahara was once by no means the 
desert it is now. North of the Sahara — that is, along 
the Mediterranean coast of Africa — there is evidence 
of profound changes in climate during and since 
Pleistocene time. In Pleistocene time this region was 
still distinctively a part of the African or Ethiopian 
region. After upper Pliocene time this region enjoyed 
a warm temperate climate characterized by abruptly 
alternating dry and rainy seasons; there is evidence of 
periods of excessive rainfall at the beginning of the 
Pleistocene epoch. Various indications point to in- 
creasingly long periods of drought and progressive 
secular desiccation of this great region as the Pleisto- 




FiGURE 759. — Influence of secular desiccation on the archaic and modern orders of mammals 



the feeding places, during which the quadrupeds are 
exposed to attack. Finally some of the pools dry 
up entirely and, as J. W. Gregory observes (1896.1, 
p. 268): "Here and there around a water hole we 
found acres of ground white with the bones of rhinoc- 
•eroses and zebra, gazelle and antelope, jackal and 
hyena, * * * all the bones were there fresh and 
ungnawed." These animals, which had not migrated, 
liad "crowded around the dwindling pools and fought 
for the last drops of water." Such perishing of 
animals in great numbers from thirst would bring 
about diminished herds, spoken of above as the final 
■cause of extinction through inability to protect the 
joung. 

Pleistocene desiccation oj northern Africa. — There are 
archeological and even historical proofs (Herodotus) 



cene advanced, resulting in the partial extinction and 
partial migration of the great equatorial quadruped 
life to central and southern Africa (Pomel, 1895.1). 
Thus the elephants, rhinoceroses, hippopotami, and 
giraffes disappeared. A typical African fauna was 
replaced by an equally typical European fauna, which 
included the bear and the deer. 

Desiccation and extinction in central Australia. — • 
Wallace's opinions as to the causes of the extinction 
of animals in Australia have been cited more as to the 
effect of the conditions during the glacial epoch and to 
continental contraction in general than as to the special 
causes of extinction in Australia. More recent re- 
search, set forth by the geologist Tate (1889.1) and 
the zoologists Hedley (1894.1) and Baldwin Spencer 
(1896.1), shows that in Pliocene time heavy rainfall or 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



867 



pluvial conditions, great inland seas or fresh-water 
lakes (first surmised by Stuart) favored the develop- 
ment of large marsupials. Conversely the rise of an 
eastern coastal range was followed by diminished rain 
supply and progressive desiccation of the interior 
region. 

Spencer observes (1896.1, p. 183): 

The larger forms now extinct, such as species of Diprotodon, 
Nototherium, Phascolonus, Macropus, Protemnodon, etc., reached 
their greatest development in Pliocene time and were character- 
istic of the eastern interior, spreading southward round the 
western end of the Dividing Range into Victoria. They do 
not seem to have reached the eastern coastal district. * * * 
In post-Pliocene time, with the increasing desiccation of the 
whole central area, they became extinct, though this extinction 
can not be attributed wholly to the drying up of the land, 
because in certain parts, such as western Victoria, to which they 
reached, the state of desiccation did not supervene; but at the 
same time it may perhaps be justly argued that the desiccation 
of the vast area of the interior was the largest factor in their 
extinction. 

The discovery (in 1892) of the great Lake Calla- 
bona bone deposit in the interior of South Australia 
abundantly confirms the "desiccation" theory. Dr. 
E. C. Stirling (1899.1, pp. ii-iii) describes this remark- 
able deposit as follows: 

There is, however, compensation for the unpromising physical 
features of Lake Callabona in the fact that its bed proves to be a 
veritable necropolis of gigantic extinct marsupials and birds 
which have apparently died where they lie, literally in hundreds. 
The facts that the bones of individuals are often unbroken, 
close together, and frequently in their proper relative positions 
(vide pi. A, fig. 3), the attitude of many of the bodies, and the 
character of the matrix in which they are embedded negative 
any theory that they have been carried thither by floods. The 
probability is, rather, that they met their death by being entombed 
in the effort to reach food or water, just as even now happens 
in dry seasons to hundreds of cattle which, exhausted by thirst 
and starvation, are unable to extricate themselves from the 
boggy places that they have entered in pursuit either of water 
or of the little green herbage due to its presence. The accu- 
mulation of so many bodies in one locality points to the fact of 
their assemblage around one of the last remaining oases in the 
region of desiccation which succeeded an antecedent condition 
of plenteous rains and abundant waters. An identical explana- 
tion has been suggested by Mr. Daintree (1872.1) in his "Notes 
on the geology of the colony of Queensland." 

Alkali and salt deposits. — One effect of increasing 
desiccation is the increased number of alkali lakes, 
licks, and springs, and other localities of salt deposits. 
Alkali is much sought by certain wild animals as a 
substitute for salt. Western stockraisers disagree as 
to the effects of alkali upon sheep and cattle, some 
believing that it can not take the place of salt. Ches- 
nut (1901.1, p. 20) notes that alkali may possibly pre- 
dispose to the "loco habit," the eating of a narcotic 
weed. When domesticated animals are not salted 
regularly they soon discover localities where large 
quantities of alkali are found in the soil and visit 
such places frequently for the purpose of eating this 
alkali soil. (Op. cit., p. 87.) 

101959— 29— VOL 2 12 



THE LIVING ENVIRONMENT 



We have thus considered the relations of cold, heat, 
moisture, and desiccation to the hunger, the thirst, 
and the feeding and the migrating habits of animals. 
We may now look at the food supply of the Herbivora, 
especially in its relation to unusual conditions of life. 

Forestation, deforestation, and reforestation. — Forests 
furnish a condition necessary to the existence of 
certain quadrupeds, especially the browsing animals, 
such as the Cervidae, which have brachyodont (short- 
crowned) teeth, and the Proboscidea, including 
especially the brachyodont mastodon. Among Artio- 
dactyla the deer, among Perissodactyla the tapirs, 
among Proboscidea the mastodons are typical forest 
animals. Conditions, therefore, which cause defores- 
tation would become a means of extinction; such con- 
ditions are (a) intense cold and heavy snow capping, 
(b) intense dryness, (c) destruction of young trees by 
the smaller browsing animals. 

It is probable that the interior of Australia and the 
Pampean region of South America were in Pliocene 
and early Pleistocene time partly covered with forests. 
It is certain that the Holarctic region (the circumpolar 
belt) was forested in early Pleistocene time. Our 
western arid region was extensively forested at one 
period. Several of the smaller islands of the Mediter- 
ranean have been deforested. Reforestation would 
confine and limit the desert and plains types. Pro- 
gressive moisture and reforestation would be very 
unfavorable to the horse. (Morris, 1895.1, p. 261.) 
Thus both migration barriers and migration tracts are 
formed by forests. 

Woldrich (1882.1), in considering the glacial epoch, 
separates into two subclasses the remains of the 
Quaternary forest fauna of northern Europe, one 
living entirely in the forests and the other in the inter- 
mediate zone between trees and grass. The former 
comprised Alces pahnatus, Cervus elapJius, C. capreolus, 
Rangifer tarandus, Bos, Sus, Castor, Sciurus, Myoxus, 
Arvicola glareolus, Mus sylvaticus, Tetrao urogallus, 
T. tetrix, etc. The other and intermediate class con- 
sisted of Rhinoceros antiquitatis, ElepJias primigenius. 
Hippopotamus (in certain districts), Bison prisons. Bos 
primigenius, Megaceros, Rangifer tarandus, Equus 
fossilis, etc. All these animals frequent woods; in 
fact, their daily need of food compels them to live for 
the most part of the time in forests. The very abun- 
dance of their remains proves how luxuriant the 
contemporary forests must have been, a fact supported 
by the presence of large quantities of the remains of 
the capercailzie, essentially a forest-loving grouse, in 
the caves of Belgium, Yorkshire, etc. 

Torrell (1876.1), a Swedish geologist, has argued that 
it was largely the destruction of the forests in Denmark 
and Scania which during Neolithic (late Pleistocene) 



868 



TIT,\NOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



time extinguished the bison, the elk, the boar, the 
capercailzie, and the tortoise (Emys lutaria). 

Poisonous plants. — Domesticated Herbivora gener- 
ally but not invariably learn to avoid poisonous plants 
or become more or less immune to their deleterious 
effects. Generally they are driven by hunger to eat 
poisonous plants, but there are some exceptions to 
this rule. 

The theory that deleterious or poisonous plants are 
resorted to by animals under unusual conditions of life 
is still to be considered among the possible causes of 
extinction of wild animals. Poisonous plants are 
widely distributed. Under the unnatural conditions 
of extreme cold, drought, enforced migration, starva- 
tion, etc., they may have influenced extinction, 
especially on diminished herds. 

Chesnut's observations on plant poisoning among 
the domesticated animals are summarized in a letter 
written by him to the author (July 9, 1902): 

So far as 1113- observations have extended the chief circum- 
stance leading to death from poisonous plants is an irregularity 
of the food suppl}' caused by more or less unusual conditions. 
It does not seem reasonable to suppose that wild animals are 
frequently poisoned in their native grazing grounds. Sudden 
disasters, however, might drive them from their feeding grounds 
into pastures quite unfamihar to them, where they would 
undoubtedly be more or less at a loss to distinguish between 
poisonous and nonpoisonous plants. 

Hornaday, on the other hand, in a letter written to 
the author in 1906, observes: 

There appears to be no evidence that any wild herbivorous 
species ever has been seriouslj' affected by eating poisonous 
plants. All the cases cited by Chesnut and other authors under 
this head are domestic species onlj-. Tame species have, 
through man's care for their wants, lost the discriminating 
instinct which protects the wild species. For example, the 
antelope, buffalo, elk, and mule deer never eat loco weed on the 
southern plains; the deer of the Appalachian region do not eat 
the leaves of the laurel, which are so fatal to the domestic goat. 

Beebe in a letter to the author dated January 17, 
1912, observes: 

Darwin was in error when he stated " that "The root of the 
Aconitum napellus becomes innocuous in frigid climates." 
Hooker was also wrong when in the "Flora of British India" 
he lumped a number of distinct species under the name Aconi- 
tum napellus, which, as a fact, does not occur in British India. 
The poisonous aconite of Sikhim is A. ferox. In Nepal four 
varieties of A. napellus are found, two of which (A. napellus 
and A. rigidum) are poisonous and two (A. multifidium and 
rotundifolium) are eaten by Bhutian hill men, and also by the 
Impeyan pheasant (Lophophorus) and panda (Aelurus). The 
domestic sheep can not distinguish between the poisonous 
and nonpoisonous forms (sheep are protected by muzzling); 
the wild creatures can distinguish. 

Wild animals are certainly protected, both by in- 
stinct and by experience, from poisoning by plants. 
This fact, however, does not exclude the possibility 
that, imder certain harsh or unusual conditions of 

«" Animals and plants under domestication, Westminster ed., II ,p. 255. 



life, deleterious or poisonous plants may have has- 
tened a diminution of their number. There may also 
have been occasions when such plants constituted a 
new and unfamiliar danger. Some molds and smuts 
that appear on the Gramineae are periodic, not 
constant; under changing geographic conditions cer- 
tain narcotic plants may have been iutroduced and 
spread rather suddenly; ergot, a morbid growth arising 
from a diseased condition of the ovary of various 
grasses, produced by a fungus of the genus Claviceps, 
may cause diseases of the hoof. We must also con- 
sider the effect of the introduction of certain poisonous 
plants that do not injure the parent but may injuri- 
ously affect or kill the suckling young. The distribu- 
tion of most of these plants is related to increase or 
diminution of rainfall. 

There are also marked variations in the degree of 
immunity. Linnaeus, in his "Tour in Scania," as 
cited by Lyell (1872.1, p. 440), tells us that goats 
were turned into an island that abounded with 
Agrostis arwndinacea, which they would not eat, so 
that they perished by famine, but that horses which 
followed them grew fat on the plant. The goat, he 
also says, thrives on meadow sweet and water hem- 
lock, plants that are injurious to cattle. Observa- 
tions in South Africa (Hutcheon, 1906.1) give similar 
results. The " chinkerinchee " plant {Orniihogalum) 
is poisonous to horses, and one of the ragworts (Senecio) 
is an irritant causing cirrhosis of the liver in cattle 
and horses. Tulps (species of Moraea) also give 
trouble. Most of the cattle thus lost are not accus- 
tomed to the country or are very hungry trek cattle. 

Fatal effects of wet seasons. — Chesnut (1901.1, p. 19) 
observes that most of the plants that are especially 
poisonous during the wet season are so much shriveled 
in the dry season as to be absolutely unpalatable. 
Sheep owners have accordingly found that mountain 
ranges whose growths are extremely dangerous for 
sheep during the wet season of early summer are Ciuite 
safe from July to September, inclusive. Similarly, 
during -the wet season and when feeding immediately 
after heavy rainstorms domesticated animals are more 
likely to pull up the roots of plants than when the 
ground is dry, and (idem, p. 26) in many poisonous 
plants it is chiefly the roots that contain the poison. 

Snowstorms as affecting eating of poisonous plants. — 
After heavy snowstorms, when the grass is covered by 
snow, it often happens that only the taller species 
of plants are exposed (idem, p. 27). At such times 
the poisonous larkspurs {Delphinium glaucum) are 
greedily eaten by cattle, which at other times avoid 
these plants. This danger is increased by the fact 
that ruminants do not feel at ease so long as their 
stomachs are not full, and they are inclined to eat 
anything in sight after a snowfall. In seasons of 
drought certain poisonous leguminous plants remain 
green and tempting after the grasses have become 



CAUSES OP THE EVOLUTION AND EXTINCTION OP THE TITANOTHERES 



869 



thoroughly dried. Cattle on the range then take the 
loco and lupine (idem, p. 29). 

Enforced migration and poisonous plants. — Chesnut 
observes (idem, p. 21) that domesticated animals 
when feeding quietly on the range exercise considerable 
choice in selecting forage plants, but that when they 
are being driven 6 or 8 miles a day they may be forced 
by hunger to bite off almost all kinds of plants that 
grow along the course of their travel. Enforced 
migration among wild animals might similarly cause 
them to become less fastidious about food. 

Distribution oj poisonous plants. — The chief poison- 
ous plants of the Montana stock ranges (Chesnut, 
1901.1) are the death camass (Zygadenus), favored 
by moderate moisture and taken by sheep; the "tall 
larkspur" {Delphinium glaucum), favored by moderate 
moisture, taken by cattle; the "purple larkspur" 
(D. iicolor), taken by sheep; the water hemlock 
(Cicuta), found along watercourses, taken by cattle 
and sheep; the white loco (AragaUus), taken by horses, 
sheep, and cattle. Lupines (Lupinus) in certain 
stages of growth are poisonous to sheep. Ergot 
{Claviceps purpurea) occurs in Montana on a variety 
of grasses and is occasionally poisonous to horses and 
cattle, producing a disease of the limbs. On a large 
ranch of Wyoming, according to Walter Granger 
(letter, 1904), ergot appeared as a result of irrigation, 
rendering a large tract fatal to horses and cattle by 
causing a disease of the hoofs. 

A leguminous plant of Egypt (Lotus arabicus), 
recently investigated by Dunstan and Henry, as a 
growing plant is poisonous to horses, sheep, and 
goats. It contains a glucecoid termed "lotusin," 
which is poisonous when taken into the stomach 
(Chesnut, 1902.1, p. 1019). Its seeds when ripe, 
however, are commonly used as fodder. 

Narcotic plants. — Among narcotic plants "loco 
weeds" are the most interesting as one of the possible 
causes of the extinction of wild animals. "Loco," 
a Spanish word meaning mad or crazy, is applied in 
northern Mexico and the southern United States to 
certain plants which so affect the brain of animals as to 
give them all the symptoms of brain disease. As 
described in the important paper of Chesnut (1899.1, 
pp. 403, 404) the weeds called "loco" belong to genera 
of the pea family. He writes (1902.1, pp. 87-90): 

For many years a disease called loco, affecting cattle, horses, 
and sheep, has been generally known to the stockmen of the 
western ranges. This disease has most commonly been attrib- 
uted to the action of certain plants, more rarely to that of 
alkali. Several species of plants have been suspected of pro- 
ducing the loco condition in animals and have been called loco 
plants or loco weeds and also crazy weeds, from the nature of 
the disease. Nearly all of the plants which have been considered 
loco weeds belong to two genera of the pea family. Astragalus 
and Aragallus. These genera are represented by numerous 
species on the western stock ranges. * * * From a general 
description given of the loco disease it is apparent that this 
condition might very justly be termed a perverted appetite. 



* * * 'pjie horse and the sheep are the animals which are 
most frequently affected by loco disease. Cattle occasionally 
acquire the loco habit, but the cases are comparatively rare. 
In certain parts of Montana the habit became so widespread 
among horses that the raising of them was abandoned until 
the locoed animals were disposed of and other horses which 
had not the loco habit had been imported. * * * During the 
progress of field work in Montana in 1900 about 650 locoed 
sheep and 150 locoed horses were seen. 

Mechanically dangerous plants. — Occasional losses 
of stock occur in Montana from plants acting mechan- 
ically. For example, the sharp-barbed awns of the 
porcupine grass (Stipa spartea) and squirrel tad 
(Hordeum juhatum) when the plants are maturing 
separate and, enteruig the mouth, throat, eyes, and 
ears of stock, affect the tissues and give rise to ulcers 
which cause intense suffering and necessitate killing. 
(Chesnut, 1902.1, pp. 50-51.) Similarly the corn- 
stalk disease is sometimes attributed to malnutrition 
or impaction of the alimentary canal. 

In this connection may be cited an observation 
recorded by Thisel ton-Dyer (1902.1), which happens 
to bear upon the life of goats: 

The introduction of the sweetbrier into New South Wales, 
Australia, in many parts of which it is naturalized, affords a 
striking illustration of the mode in which the balance of nature 
may be disturbed in a wholly unforeseen way. * * * The fruit 
of the sweetbrier {Rosa ruhiginosa) consists of a fleshy receptacle 
lined with silky hairs, which contains the seedlike carpels. 

* * * The hairy linings of the fruit caused the death of a 
number of goats by forming hairy calculi, which mechanically 
occluded the lumen of the bowels. These goats were put on 
the land with the idea that they would eat down the briers and 
ultimately eradicate them, but the briers came out best and 
eradicated the goats. The cattle running on the land are also 
very fond of the brier berries, and from time to time one will 
die, and on post-mortem [examination] no pathological changes 
can be found in any of the organs, nor do the hairy calculi 
appear in them, although their various stomachs are one mass 
of the brier seeds. 

INSECTS AND PROTOZOA 

The main features of physical environment, such as 
moisture and desiccation, heat and cold, can not be 
considered by themselves or solely in relation to the 
plant life; they must be studied also in relation to 
the insect life which they condition. Insect and 
parasitic life is now known to be one of the greatest 
factors in the numerical reduction and probably 
therefore in the extermination of mammals. Greater 
progress has been made in this study than in any 
other since Darwin's time, yet Lyell and Darwin both 
adumbrated the modern discoveries. 

Insects and the food supply. — The periodic devasta- 
tions of certain insects, especially those caused by 
locusts as cited by Lyell, in Europe, Arabia, India, 
and northern Africa, are sufficient to cause the 
reduction of certain species. As Lyell concludes 
(1872.1, vol. 2, p. 445): 

The occurrence of such events at certain intervals, in hot 
countries, like the severe winters and damp summers returning 
after a series of years in the Temperate Zone, may effect the 



870 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



proportional numbers of almost all classes of animals and plants 
and probably prove fatal to the existence of many which would 
otherwise thrive there; while, on the contrary, the same occur- 
rences can scarcely fail to be favorable to certain species which, 
if deprived of such aid, might not maintain their ground. 

Insect barriers. — We may first consider the influence 
of the introduction into liabitual feeding grounds of 
various forms of insect life which render tliese grounds 
practically uninhabitable and either kill or drive the 
animals out. Thus Darwin, as quoted by Wallace 
(1889.1, p. 19), observes: 

In several parts of the world insects determine the existence of 
cattle. Perhaps Paraguay offers the most curious instance of 
this; for here neither cattle nor horses nor dogs have ever 
run wild, though they swarm southward and northward in a 
feral state; and Azara and Rengger have shown that this is 
caused by the greater numbers, in Paraguaj', of a certain fly 
which lays its eggs in the navels of these animals when first 
born. The increase of these flies, numerous as they are, must 
be habitually checked by some means, probablj' bj' other 
parasitic insects. Hence, if certain insectivorous birds were 
to decrease in Paraguay, the parasitic insects would probably 
increase; and this would lessen the number of the navel-fre- 
quenting flies; then cattle and horses would become feral, and 
this would greatly alter (as indeed I have observed in parts of 
South America) the vegetation; this again would largely afi'ect 
the insects, and this, as we have just seen in Staffordshire, the 
insectivorous birds, and so onward in ever-increasing circles of 
complexit}'. 

Insects and infection. — As noted above, the most 
striking advance toward a complete theory of natural 
extinction has come from recent discoveries regarding 
the real nature of animal diseases and how they are 
communicated. Only recently have we come thor- 
oughly to realize, first, that insects are the most 
active means of introducing and spreading fatal dis- 
eases over great geographic areas and on a vast scale; 
second, that certain immune mammals become the 
bearers and disseminators of these diseases. 

Aflalo in his paper "The beasts that perish"** 
has discussed many of the various causes of exter- 
mination and gives disease a prominent place. 
Among the Carnivora there are the nonepidemic dis- 
eases, such as distemper, affecting dogs, foxes, wolves, 
cats, and other wild felines. The more rare and 
sporadic epidemics claim victims among the Carni- 
vora wholesale. The prevalence of rabies among 
foxes was observed on the continent from 1830 to 1838 
in Switzerland, also in Wiirttemberg and Baden. 

Tides in Africa. — Roosevelt (1910.1) dwells on the 
extent to which African wild mammals are persecuted 
and infested with ticks, to which, however, they seem 
to -have become so habituated that they dread them 
much less than the biting flies; the ticks, even where 
they do not introduce disease germs, are very weakening 
as bloodsuckers. Many birds devote themselves to 
little else than the picking off and eating of the ticks 
and fleas infesting the mammals; if these birds were 
killed off the mammals would suffer far more. 

68 The original article has not been accessible to the author. 



Johnston notes (1910.1) that certain types of heron 
(egret) are perpetually snapping at tsetse and other 
flies which settle on oxen or game and probably 
destroy a considerable proportion of these disease- 
carrying insects. 

Ticks as rapid spreaders of disease among domestic 
ruminants. — Piroplasma parvum is a protozoan which, 
imlike the trypanosome, invades the blood corpuscle. 
It is malignant with cattle along the greater part of the 
east coast of Africa, causing what is known as east- 
coast fever. The infection is usually transmitted by 
ticks, most frequently by the brown tick, Rhipi- 
cepJialus appendiculatus , also by R. (OerotJierium) 
simus. Migrating or trekking cattle may carry the 
ticks many miles a day and thus spread the disease 
rapidly over a wide area. The larva creeps on an 
infected animal, sucks some of its blood, drops off, lies 
among the roots of the grass, and passes its first molt, 
becoming a nympha, then an imago, in either of which 
stages it may infect a healthy animal by creeping 
from the grass. The tick is very hardy and may 
survive with its infection for a year, but after a year 
or 15 months the infected ticks are all dead, so that 
healthy cattle may reenter the field without risk. It 
takes two years to starve the ticks out of a country by 
removing the cattle. This method was employed in 
the Wichita National Bison Preserve, the bison (Bison 
americanus) being very susceptible to "red-water 
fever." If this tick had been introduced centuries 
ago among the bison it might have exterminated them. 

^Yide geographic distribution. — The geographic dis- 
tribution of the species of Piroplasma is very wide; 
first discovered in North America, it is now epidemic 
throughout most of South Africa. Piroplasma bigemi- 
num similarly causes the "Texas" or "red-water" 
fever of our Southern States. It is conveyed by a 
tick. The germs are latent, and the blood of an animal 
that has recovered from Texas fever remains infec- 
tive. Thus apparently healthy cattle may infect 
imported susceptible cattle. Such latency has a sig- 
nificant bearing upon the theory that natural extinc- 
tion may be caused by similar germs. Although they 
acquire immunity, the domesticated native Bovidae 
act as reservoirs of the disease, in contrast to the 
tsetse-fly disease, in which the wild Bovidae act as 
reservoirs. The further fact that native cattle may 
become immune has an important theoretical bearing 
on the natural origin of immunity to the tsetse-fly 
and other diseases among the wild Bovidae and wild 
Equidae. 

TicTcs among Equidae. — The biliary fever of domes- 
ticated Equidae (horses, mules, donkeys) is conveyed 
by a corpuscle parasite (Piroplasma equi), which is 
spread by the red tick (RhipicepJialus evartsi), the 
infection being taken in the nymphal and transferred 
in the adult stage. The native South African horses, 
like cattle that are exposed to Texas fever, become 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



871 



immune to the disease and are said to be "salted," 
but horses that have recovered from the disease con- 
tinue to act as reservoirs and remain sources of infec- 
tion throughout their lives. 

The same is true of Piroplasma canis among the 
Carnivora spread by the dog tick {Haemophysalis 
leachii). The blood of recovered animals remains 
infective. 



EPmEBncs 



Epidemics in North America. — We are indebted to 
Thompson Seton (1909.1) for bringing together a 
large number of observations recorded by many 
authors on epidemics among the American mammals. 
In the summer of 1873 a fatal epidemic is said to have 
destroyed three-fourths to nine-tenths of the prong- 
horn antelopes (Antilocapra) in the area between Yel- 
lowstone and Missouri Rivers in Montana (Allen). 
An epizootic distemper called black tongue killed 
thousands of white-tailed deer in Texas in 1856 
(Walton). Tuberculosis broke out among the New 
England deer in 1901, destroying chiefly the older 
ones (Murch). A species of distemper killed off vast 
numbers of the beaver of upper Red Deer River, 
Alberta, about the year 1800, and they have never 
been so plentiful in that region and eastward to Hudson 
Bay since that year (Tanner). In western Manitoba 
the wonderful "rabbit year," 1886, in which these 
animals are said to have multiplied to the number of 
5,000 to the square mile, was followed in the ensuing 
winter by a destructive plague due to a staphylococcus 
which may have started in some skin wound or para- 
sitic skin disease (Little); from a condition of extra- 
ordinary abundance the rabbits practically disap- 
peared. According to Hornaday a parasite, the 
bloodsucking stomach worm (Strongylus strigosus), 
may have caused the so-called seven-year plague 
among northern varying hares (Lepus variabilis) and 
the rabbits of the West. Seton observes that the 
great periodic decrease in the lynx, which in every 10 
years or so reduces their number to about one-tenth 
of the maximum, is probably due indirectly to the 
periodic rabbit plagues — that is, to starvation through 
failure of the rabbit supply. Many emaciated bodies 
of the lynx are found in such seasons. 

Epidemics in Europe. — Fleming, in his "Animal 
plagues" (1871.1), enumerates eighty-six epidemics 
that affect wild quadrupeds and birds. In the list 
are diseases that affect nearly every wild species in 
Europe and some in the New World, including the 
red deer {Cervus elapJius), the reindeer (Rangijer 
tarandus), the chamois (Rupicapra tragus), and the 
wild hog; also, among the Carnivora, wolves, foxes, 
and bears; among the Rodentia, hares, rabbits, and 
rats. Various forms of tuberculosis account for a 
large percentage of death among domesticated ani- 
mals. Among animal plagues anthrax was formerly 
the most rapid and deadly, but it is now perhaps the 



least common, owing to Pasteur's discoveries. Amer- 
ican zoologists are familiar with the spread of disease 
from domesticated to nondomesticated animals — of 
the sheep scab, for instance, to the wild sheep {Ovis 
montana). 

The feline and ursine Carnivora are protected by 
their relatively nongregarious habits ; canids are highly 
gregarious and consequently would be more subject 
to the spread of eipdemics. On the contrary, the 
gregarious Herbivora offer favorable conditions for the 
spread of disease. 

Epidemics in Africa. — In his "Great Rift Valley" 
Gregory (1896.1, pp. 265-266) observes that the great 
herds of game which roamed over the steppes of South 
Africa are being rapidly decreased. Man no doubt 
has played the leading part in the annihilation of the 
enormous herds that once thronged Cape Colony. The 
fact that during the last few years the game has re- 
treated from the Somali coast into the interior shows 
how easily it can be driven from a district. In South 
Africa, however, man's work of extermination has 
probably been insignificant as compared with that of 
natural agencies, such as lions and disease. Vast 
herds of the wild buft'alo {Bubalus ca.-ffer) were exter- 
minated between 1890 and 1893 by the cattle disease 
(rinderpest), which also killed off the gnu and giraffe. 
Gumming (1855.1, vol. 1, p. 138) observed as early as 
1855 that 

the goat in many districts is subject to a disease called by the 
Boers "brunt sickta," or burnt sickness, owing to the animals 
afflicted with it exhibiting the appearance of having been burnt. 
It is incurable, and if the animals afflicted are not speedily killed, 
or put out of the way, the contagion rapidly spreads, and it is not 
uncommon for a farmer to lose his entire flock with it. This 
sad distemper also extends itself to the ferae naturae. I have 
shot hartebeests, black wildebeests, blesbucks, and springbucks, 
with their bodies covered with this disease. I have known 
seasons when the three latter animals were so generally affected 
by it that the vast plains throughout which thej' are found were 
covered with hundreds of skulls and skeletons of those that had 
died therefrom. 

Howorth observes (1887.1, p. 174): 

Frequenters of the forest with whom I have conversed, 
whether Europeans or Singhalese, are consistent in their assur- 
ances that thej' have never found the remains of an elephant 
which had died a natural death. One chief, the Wannyah of the 
Trincomalee district, told a friend of mine that once after a 
severe murrain, which had swept the province, he found the 
carcasses of elephants that had died of disease. 

Entozoa in elepJiants. — Several specimens of the 
African elephant (Loxodon africanus Blumenbach) 
were autopsied by MacCallum (1907.1). Little has 
been written about the Entozoa of this species, 
which apparently differ in many respects from the 
better-known forms infecting Elephas indicus. 

It is said that elephants succumb to the infestation 
of parasites more than from any other cause, and the 
worm that most often destroys these animals is a 
species of ParampMstomum. Members of this genus 



872 



TITAN OTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



of trematodes are commonly parasitic among the 
Herbivora in tropical countries. They infest the 
stomach and intestines in great numbers and not only 
deplete the animal's energy by exacting a great 
quantity of blood but disturb the lining membrane of 
the stomach and intestine so as to interfere seriously 
with the secretions and digestion. Several species of 
worms may be found in the same region. 

When it is suffering from the attacks of these 
parasites the elephant is said to make instinctively an 
attempt to get rid of them by eating 10 to 20 pounds of 
a certain red earth, which acts as a purgative, dis- 
lodging and expelling a lot of worms. In India, where 
there are domesticated herds of 400 or 500 of these 
valuable animals, the loss from Entozoa is sometimes 
serious. 

Parasites in air sinuses. — Larvae that invade the 
frontal sinus of the skull are among the possible 
causes of the extermination of animals. An old 
trapper and close observer in British Columbia, Mr. 
Charles Smith, informs the author that both the wild 
sheep of the region {Ovis montana) and the wapiti 
{Cervus canadensis) are seriously affected or killed 
by inflammation caused by these larvae. The over- 
crowded caribou of Labrador and Newfoimdland 
suffer from a fly which lays its eggs in the nasal 
passages. 

Allen (1906.1, p. 201) states that there is a patho- 
logic modification in the skulls of peccaries {Tayassu), 
due to parasites lodging in the orbitosphenoid and 
adjacent parts. He observes: 

In this connection an examination has been made of nearly 
50 skulls of peccaries, in the [American] Museum collection 
from various parts of South Amierica, with the following 
results: In a series of 17 skulls of T. pecari from Santa Marta 
district of Colombia, all were found diseased in the manner 
above described, so that this condition might readily be mis- 
taken for the normal. In 20 skulls of T. torvum about 80 per 
cent show the diseased condition strongly, and others show 
traces. Of 9 skulls of T. iajacu from Chapada, Matto Grosso, 
Brazil, 2 only are normal. The inflation of the bones forming 
the antero-inferior wall of the orbit, through the invasion of 
these parts by some parasite, is so general in the whole group 
of peccaries that the absence of such conditions seems to be 
almost exceptional. 

Moisture favoring the spread of diseases carried by 
insects. — The presence of the blood protozoan para- 
sites known as trypanosomes, combined with certain 
fhes which act as disease carriers, is in many countries 
correlated with wet weather. This is especially true 
of the disease known in India as "surra," the history of 
which was first suspected by Surgeon Major Lewis in 
1888 (1888.1). In South America the mal de caderas 
affects horses, asses, cattle, hogs, and certain other 
animals and is attributed to the protozoan known as 
Trypanosoma equinum. It is distinctively a wet- 
weather disease, almost completely disappearing in 
dry seasons. Asses, swine, and water hogs are said 
to be affected, and horses are never known to recover. 



It is chronic in course, lasting fi-om two to five months 
in horses and from six to twelve in asses and mules. 
See Voges (1902.1) for fuller details. 

Extermination of the Equidae. — The wide geographic 
range of surra and related diseases is significant with 
reference to former periods in the history of the 
Equidae. All authors now agree with Lewis that the 
disease is carried by flies and coincides with wet 
weather, occurring chiefly during or immediately after 
heavy rainfall, though sporadic cases may occur at 
other seasons of the year. In the "Emergency 
report on surra" by D. E. Salmon, C. W. Stiles, and 
A. Hassall (1902.1, p. 18), this is described as chiefly 
a wet-weather disease, invariably fatal to horses and 
mules, occurring in other animals, such as camels and 
elephants, more rarely in ruminants, and transmis- 
sible to goats, sheep, and other mammals. In India 
it is said to affect horses, camels, and elephants. It 
occurs in Burma, Persia, Tonkin, and Chosen. In 
Africa there is the similar nagana, or tsetse-fly disease, 
more accurately described by Bruce (1905.1, p. 333). 
In Algiers, France, and Spain the dourine or maladie 
de coit attacks the horse and ass in particular and 
may be transmitted to certain other animals. It is 
attributed to a trypanosome, T. equiperdum. In the 
Philippines surra caused the death of 2,000 army horses 
in six months. The intermediary is a fly, Stomoxys 
calcitrans. It was also reported (Curry, 1902.1) as 
affecting the carabao. Bos (Buhalus) Tcerabau, but 
according to Lingard ruminants are not particularly 
susceptible. 

An interesting bit of advice given to those in 
charge of horses in the Philippines may have some 
bearing upon the origin of colors in certain quadru- 
peds: "Avoid light-colored animals as much as 
possible; the darker the animal the safer he appears 
to be from the attack of flies." In this connection 
we recall the dark color of the true Bovinae, the wild 
cattle. 

Tsetse-fly disease of domesticated Equidae and 
Bovidae. — The nagana or tsetse-fly disease of Africa 
is caused by Trypanosoma brucei (Plimmer and 
Bradford, 1889-1; 1902.1); the carrier is the tsetse 
fly (Glossina morsitans). Together this trypanosome 
and its host, the fly, render thousands of square miles 
of Africa uninhabitable; no horses, dogs, or cattle can 
venture even for a day into the "fly country." After 
all the nonimmune animals of the country have been 
killed off and thus there no longer exist sources of 
infection, the tsetse fly spreads abroad out of the "fly 
country," still giving rise to the disease. This strange 
fact led to the discovery of the fact noted above — 
that many of the immune wild ruminants carry the 
same trypanosome {T. hrucei) in small numbers in 
their blood and thus act as continuous reservoirs of 
the infection; it is from them that the fly receives 
fresh supplies of the infectious parasite. A similar 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



873 



parasite also lives in the blood of healthy rarts. Anal- 
ogous to this is the "sleeping sickness" that affects 
man, which has spread rapidly from west to east 
Africa, carried by a fly, Glossina palpalis, claiming 
hundreds of thousands of victims. 

Rinderpest in Africa. — The rinderpest or cattle 
disease has been the greatest destroyer of the wild 
African quadrupeds. This disease has been known 
from time immemorial in Europe and central Asia. 
It is believed by some to have entered the Nile 
provinces of Africa in 1880, to have reached the 
Transvaal in 1896, and thus to have traveled the 
whole length of the Dark Continent. It has been 
spread largely through the infection of wild ruminants. 
It is fatal to the wild buffalo, Bos (Bubalus), the kudu 
{Strepsiceros Icudu), the sable antelope {Hippotragus 
niger), the gnu (Connochaetes albojuhafus and 0. 
taurina), also in the Philippines to the carabao, 
Bos (Buhalus) Icerahau. It is fatal to 90 to 100 per 
cent of domesticated cattle. The parasite causing 
rinderpest is undiscovered, no natural immunity is 
known (methods of artificial immunity were dis- 
covered in 1893), and it is distinguished by the ease 
and rapidity with which it spreads in all countries, 
climates, and seasons, being carried even on the clothes 
and person of man. It therefore appears improbable 
(Bruce, 1905.1) that insects have anything to do with 
it. It may be due to a wind-borne bacterial organism. 

By analogy we may imagine that a disease affecting 
the Pleistocene horses of North America may have 
traveled an equal distance, from Texas to Patagonia, 
and destroyed all the South American Equidae. 

Moist rhinarium and other eliminating causes in 
Africa. — The rinderpest of 1889-1899 illustrated the 
eliminating value of a single organ, the external 
nostril. Rinderpest (see above) is fatal to domestic 
cattle, to all the Bovinae and to the antelopes most 
closely approximating to them, including the tragela- 
phines (eland, kudu, bushbuck, reedbuck, dyker), all 
these ruminants having a large, moist rhinarium, 
whereas animals with a small, part hairy rhinarium — - 
the sable antelope, roan antelope, hartebeest, impala — 
suffered much less if at all. 

The wildebeest and hyena, although immune to 
rinderpest, are' hosts of the nagana trypanosomes, 
which are carried by the tsetse fly. In the region of 
the middle Zambezi the buffalo perished in enormous 
numbers, the remains of 200 animals being observed 
within a few acres — piles of bones and skulls, a veri- 
table Golgotha. Geographically the course of the 
rinderpest was most erratic, here diverted by a range 
of mountains, there sweeping down one bank of a 
river and leaving the other untouched. Subsequent 
to the rinderpest epidemic the tsetse fly was elimi- 
nated over certain areas, especially those that had 
been covered by the rinderpest south of the Zambezi. 
Whether the elimination of the fly was in any way 



connected with the rinderpest is a matter of specu- 
lation. 

As a result of the rinderpest epidemic in South 
Africa the last of the elands disappeared and the 
buffalo were reduced to one herd of about 20. In the 
Sabi Bush the kudu were greatly thinned down, 
whereas south of the Zambezi, in British Nyasaland, 
the rinderpest made no difference whatever in the 
status of the fly; possibly climatic conditions or some 
parasitic enemy caused the decrease of the flies. 

Pleistocene extinction of horses. — Among all the 
problems of Pleistocene extinction presented in 
America that of the horses is certainly one of the most 
difficult. These animals are far superior to cattle in 
their adaptability to changing conditions of life and 
in resourcefulness in severe wuiters. They were very 
numerous in North America at the beginning of 
Pleistocene time, whereas at the end of it they were 
apparently extinct. Similar extinction occurred both 
in North and South America at this period. The 
numerous and highly specialized horses of Mexico 
shared in this extinction, although we might have 
regarded the high plateau of Mexico as a ready means 
of escape from the more severe conditions in the north 
in the glacial epoch. It has consequently been 
suggested by Osborn (1906.287) and by others that 
the American horses may have been swept out of 
existence by some epidemic disease or diseases. This 
theory has subsequently received some support 
through the discovery by Cockerell (1907.1; 1909.1) 
of two species of fossil tsetse fly (Glossina) very similar 
to the African types of to-day. These flies are found 
in the Florissant lake beds of Colorado, which are 
regarded as of upper Miocene age, and were contem- 
poraneous with the varied equine fauna. 

Ticlcs and horses. — Ticks, even those that bear no 
infection, form effective barriers to the introduction 
of quadrupeds into certain regions. In some forested 
parts of South and Central America they endanger 
human life. In certain regions of Africa ticks are 
practically fatal to horses. As stated orally by Dr. 
D. G. Elliot, thousands of ticks may gather on a horse 
as a result of a single night's grazing. The mane 
especially serves to collect these pests. Thus the 
falling mane of the northern horse is distinctly dis- 
advantageous as compared with the upright manes of 
the asses and zebras. Ticks are capable of driving 
certain types of animals entirely out of a country and 
of indirectly causing certain modiflcations of the hair 
and epidermis. 

In the forests of southern California prickly seeds 
of certain plants find their way into the nostril sac of 
horses and cause serious if not fatal abscesses unless 
removed. 

Distribution of horse siclcness. — Horse sickness is 
local in distribution, prevailing in low countries and 
during wet seasons. The same climatic relation is 



874 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



true of the heart-water disease of cattle, goats, and 
sheep (Bovidae), which is similar in distribution to the 
horse sickness and is caiTied by the bont tick {Am- 
hlyomma hehraeum), in that it dies out on the high 
veldt. The "catarrhal fever" of sheep has a distribu- 
tion ill South Africa similar to that of horse sickness 
and is probably carried by means of the same night- 
feeding insect. The infection causing horse sickness 
is not carried into the high country nor during the 
dry season; the parasite is unknown and is believed 
to be ultramicroscopic ; it is believed to be carried in 
the blood, because the one-thousandth part of a single 
drop of blood injected under the skin of a healthy 
animal will cause death; some horses require a larger 
dose than others, indicating fluctuations in power of 
resistance or immunity. Unlike the foregoing dis- 
eases it is not endemic or permanent but occurs in 
epidemics at intervals of 10 to 20 years. Its geographic 
distribution in South Africa is very wide, including 
Natal, Zululand, the greater part of Rhodesia, 
Bechuanaland, and Portuguese East Africa. Horses 
placed in fly-proof shelters, even in exceedingly 
unhealthy places, in no case incur the disease. The 
particular fly or insect carrier is stiU unknown. As 
in several of the foregoing diseases the infective power 
of the blood persists for years. 

Summary as to natural extinction iy epidemics. — To 
summarize these remarkable observations, which we 
owe to the labors of Lewis, Koch, Theiler, Kilborne, 
Smith, Watkins-Pitchford, Lounsbury, Salmon, Cur- 
tice, Stiles, Hassall, Taylor, and many others, we 
undoubtedly have an agency that must be considered 
an occasional if not a frequent cause of the extinction 
of quadrupeds in the past. 

It will be noted (1) that in the case of the tsetse-fly 
disease the wild ruminants are the permanent though 
unharmed carriers to domestic animals of the infective 
protozoan; (2) that, on the contrary, in the Texas 
fever, or "red-water fever" the native immune 
Bovidae are the permanent carriers of the disease 
organism to the wild Bovidae (bison); (3) that the 
rinderpest now appears to be in an early stage of its 
history as a disease in which neither domesticated nor 
wild Bovidae have become naturally immune, and all 
the Bovidae act as reservoirs; (4) that in the east- 
coast fever the infective ticks survive for a year and 
the permanent carriers of the infective organism are 
not yet discovered; (5) that in the biliary fever of 
domesticated horses the recovered equines act as 
reservoirs; (6) similarly again, that in the "horse sick- 
ness" of South Africa the infective power of the blood 
in a recovered animal persists for years. 

Thus in these modern cases we have all the theo- 
retical conditions favorable to the wide distribution of 
insect-borne diseases which in past times may have 
attacked various types of quadrupeds and resulted in 
extermination before natural immunity was acquired. 



The relation of birds to the life of insects that attack 
mammals, and thus to the existence of mammals 
themselves, has already been mentioned. The closest 
relation of birds to the protection of mammal life is 
that of destroyers of the insects which infest and 
infect mammals. 

The bird that is most notably destructive to mam- 
mals is said to be the kea, the New Zealand parrot 
(WaUace, 1889.1, p. 75). The sheep-killing habit of 
the kea has recently been declared mythical. W. B. 
Benham (1906.1; 1907.1), however, asserts that the kea 
causes great mortality among the sheep of the high 
mountainous country of the South Island; in other 
parts of New Zealand it is absent. The kea perches 
on the back of the sheep and soon eats a hole into the 
abdomen. It is not known whether this is for the 
purpose of getting the blood or of eating the kidney 
fat, as was formerly supposed. The sheep are often 
found with a hole so large that the entrails exude and 
the animal has to be killed. The kea does a great deal 
of its work at night, and for this reason a great many 
people have not been able to observe it and have 
doubted the truth of the statements. Other birds are 
known to have turned from herbivorous to carnivorous 
habits. 

MAMMALS 

Forms oj competition. — From consideration of the 
struggle of animals with their physical environment 
and with their living plant, insect, and bird environ- 
ment, we now pass to consideration of the more 
intimate struggle with other mammals. It is impor- 
tant to note that the struggle in this intermammalian 
competition is always intensified during periods of 
geographic, climatic, and biotic change. This struggle 
naturally presents a variety of phases, such as com- 
petition of lower and higher types of mammals, 
struggle between Herbivora and Carnivora, struggle 
between Herbivora and Herbivora, competition be- 
tween resident and newly introduced forms, and com- 
petition between less and more adaptive types. 

Even in comparison with the supposed climatic, 
plant, and insect agencies, the competition between 
lower and higher types, or between less and more 
adaptive types, or between slowly producing and rapidly 
producing types has been the chief direct agency of 
extinction, because it has worked more widely and 
over longer periods of time. 

During the introduction of new types in the Ter- 
tiaiy period in North America, we witness more than 
once the apparently unchecked multiplication of cer- 
tain local or native mammals; the repeated intro- 
duction or migration of new mammals, either 
singly or in waves; and the slow or rapid sequent 
extinction of certain local mammals. These cycles 
of change may be due to remote geographic conditions 
but are frequently connected with or sequent upon 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



875 



independent local physiographic changes. Early theo- 
retic observations on this principle are those of Lyell 
and Darwin. Lyell observes (1872.1, vol. 2, p. 451): 

Extension of the range of one species alters that of others. 
In reference to the extinction of species it is important to bear 
in mind that when any region is stocked with as great a variety 
of animals and plants as its productive powers will enable it 
to support, the addition of any new species to the permanent 
numerical increase of one previously established must always 
be attended either by the local extermination or the numerical 
decrease of some others. 

Wallace observes: "There is good reason to believe 
that the most effective agent in the extinction of 
species is the pressure of other species, whether as 
enemies or merely as competitors. " Darwin enter- 
tained similar views. 

The evidence afforded by paleontology does not 
fully support the sweeping statement made by Wal- 
lace, for paleontologic research indicates that compe- 
tition is chiefly between phyla. The conclusion 
drawn from these exceptions is similar to that of 
Darwin — that the keenest competitors are animals of 
most nearly similar feeding habits. 

There are, however, exceptions to Darwin's conclu- 
sion also, as several examples to be cited demonstrate. 
The extinction of the titanotheres and elotheres may 
have been due entirely to changes in vegetation rather 
than to competition with other Herbivora. Again, 
the survival of the opossums (didelphids) in North 
America is a striking instance of successful competition 
of a very generalized type of marsupial with numerous 
other small Carnivora by the adoption of nocturnal 
habits. 

Competition hetween placentals and marsupials. — 
The gradual introduction of placentals into Australia 
has recently been summarized by Lucas and Le Souef 
(Lucas, A. H. S., 1909.1, p. 4). The rodents probably 
became denizens of the continent in later Tertiary 
time. They probably came from Asia, for there are 
no groups related to the peculiar genera of South 
America and the Cape. The rodents all belong to the 
widely distributed family of the Muridae. The most 
specialized genus is Conilurus (Hapalotis), comprising 
graceful little rats, which take the place of the leaping 
jerboas of Africa and Asia. Then came in the dingo, 
a wolf from eastern Asia. Lastly, introduced directly 
from Europe, came a disastrous and prolific popula- 
tion of black and brown rats, the common mouse, the 
rabbit and the hare, and, worst of all, the fox. Dur- 
ing the year 1908 no fewer than 18 million rabbit skins 
passed through the Sydney market, besides a nearly 
equal number through the Melbourne market, and 
vast numbers of rabbits were exported in cold storage, 
but these inroads produced httle effectin exterminating 
the pest. 

Baldwin Spencer observes (1896.1, p. 127) that the 
existing marsupials are severely handicapped when in 



competition with the rodents by having to carry 
their yoimg in the pouch; at the age when a young 
marsupial at sight of danger at once flies to its moth- 
er's pouch, a young rat or rabbit is taking care of itself. 
If a hawk or eagle catches the mother rabbit the yoxmg 
one is left, or vice versa. The marsupial mother 
has to carry the young ones, and not only does the 
extra weight prevent her from gaining shelter, but if 
taught both she and the young ones are sacrificed. 
A very slight difference in speed will save or lose an 
animal's life. When hard pressed a kangaroo will 
throw the young out of the pouch so as to travel 
faster. The same author (idem, pp. 55-57) in con- 
sidering the causes of the extinction of the giant 
Pleistocene kangaroos attributes it to overdevelopment 
in size during a humid period followed by strenuous 
competition of food during an arid period. 

Reduction of food supply by smaller Herhivora. — The 
great changes in the life of the countries encirclirfg the 
Mediterranean are attributable in part to changes of 
climate, in part to the destructive agency of man, 
and in part to the destructive agency of animals 
introduced and protected by man. The change 
both in soil and vegetation has been caused indirectly 
by deforestation of the hills and moimtains, and 
this has been largely due to the tmrestricted browsing 
of large herds of sheep and goats, which has been going 
on since long before the Christian era. Even now 
goats may be seen in certain parts of Palestine and 
Greece destroying the last of the forests by killing 
the seedling trees. The destruction of the forests 
leads to the washing away of the soil, imfitting a 
country to support any of the larger Herbivora. 
Man has played so large a part in this disturbance of 
the natiu-al order that it is hazardous to use these 
illustrations as analogous to natural conditions in the 
distant past. 

The competition of the smaller Herbivora, especially 
on islands, is one which, although by no means demon- 
strated, is a possible cause of the extinction of the 
larger Herbivora in past time. For example, in 
South Dakota and Nebraska the small browsers, 
such as the oreodonts and horses, which swarmed in 
herds in lower Oligocene time, may possibly have 
affected the food supply of the large titanotheres. 
In Oligocene Europe, similai'ly, the great multipli- 
cation of the small browsers, known as cenotheres, 
may have been prejudicial to the larger mammals. 

The introduction of new forms of browsing and 
grazing animals may in certain periods have disturbed 
the balance of nature and altered the character and 
amount of food supply and even the water supply and 
set up new forms of animal competition in certain 
regions. The placental rabbits to-day certainly 
exert a great influence on the natural food supply of 
the marsupial Herbivora of Australia. Cases of 
overmultiplication are rare in nature, yet not un- 



S76 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Iviiown, as seen in periodic multiplication of the hares 
in British Columbia cited above. The United States 
Biological Survey has recorded the startling total of 
10,000 to 12,000 field mice to the acre in certain parts 
of Nevada. 

We may therefore reasonably consider the part 
which the rapid multiplication of the smaller browsing 
animals may have played during the Tertiary period of 
Europe. 

Food competition especially intense on islands. — -The 
influence of the goats on islands is cited by Wallace 
(1881. 1, pp. 280, 283-286) and Palmer. 

Palmer (1899.1, p. 89) observes: 

Sheep and goats when numerous are likely to caus3 widespread 
injury, particularly in forested regions. An instructive example 
of the damage done by goats is that on St. Helena, described by 
Wallace. St. Helena is a mountainous island scarcely 50 square 
miles ia extent, and its highest summits reach an elevation ot 
2,700 feet. At the time of its discovery, about the beginning of 
the sixteenth century, it is said to have been covered by a dense 
forest; to-day it is described as a comparatively barren rocky 
desert. This change has been largely brought about by goats 
first introduced by the Portuguese in 1513, which multiplied so 
fast that in 75 yeai's they existed by thousands. Browsing on 
the young trees and shrubs, they rapidly brought about the 
destruction of the vegetation whicn protected the steep slopes. 
Witn the disappearance of the undergrowth began the washing 
of the soil by tropical rains and the destruction of the forests. 
In 1709 the governor reported that the timber was rapidly 
disappearing and that the goats should be destroyed if the forests 
were to be preserved. This advice was not heeded, and only 
a centur3' later, in 1810, another governor reported the total 
destruction of the forests by the goats. 

The Santa Barbara Islands and Santa Catalina, off the coast 
of southern California, and the island of Guadalupe, off the 
Lower California coast, are utilized as ranges for goats. All 
these islands are dry and more or less covered with brush, but 
arborescent vegetation is comparatively scarce. The goats 
practically run wild and already exist in considerable numbers. 
As yet the goats have not been on the islands long enough 
to cause any serious effects on the vegetation, and they may 
never bring about the ruin which has been wrought on St. 
Helena. But it is scarcely possible for the islands to be grazed 
by goats for an indefinite length of time without suffering serious 
damage. 

Herhivora in relation to the Carnivora. — In all the 
examples of recent conditions of life cited above the 
imrestricted feeding and rapid multiplication of small 
Herbivora have taken place under artificial conditions 
of protection of these animals from Carnivora. 

It is quite possible that m certain regions under 
natural conditions the Carnivora may have become 
diminished through epidemics or other causes, thus 
promoting the multiplication of the smaller browsing 
animals, so fatal to vegetation and to the normal 
distribution of food supply of a country. For a long 
period South America was singularly deficient in 
Carnivora and was overrun with the smaller Herbivora. 

Introduction oj Carnivora. — Among the striking 
examples of the effects of the introduction and com- 
petition of Carnivora in past and recent times are: 



1 . In the Eocene epoch the true or higher Carnivora 
entered into competition with the lower Creodonta of 
Europe and North America; this was followed by the 
final extinction of the last of the Creodonta in the 
lower Oligocene. 

2. In the Pliocene epoch true Carnivora, namely, 
the Canidae and two destructive types of Felidae — the 
saber-tooths (Machaerodontinae) and the true cats 
(Felinae) — suddenly invaded South America. They 
entered a faunal region which, subsequent to the ex- 
tinction of the marsupial carnivores (Thylacinidae) 
in the Oligocene epoch, had been entirely free from 
large Carnivora. 

3. The introduction of the dingo {Canis dingo) in 
the Australian mainland was followed by the extinction 
of the Tasmanian wolf (Thylacinus) and "devil" 
(Sarcopliilus) , animals which survived only in Tas- 
mania. The fox, which was introduced into Australia, 
like the rabbit, has increased so rapidly that it has 
become a veritable menace to native life. 

4. The introduction of the mongoose (Herpestes) in 
various countries has been fatal to the entire small 
endemic fauna. 

In each instance superior mechanical adaptation, 
intelligence, and plasticity in respect to change of 
habitat have played an important part. The Car- 
nivora, therefore, in their relation to the balance of 
nature, to the destruction of competing Carnivora and 
Herbivora, and especially in relation to the young of 
the larger Herbivora, form special topics for examina- 
tion with regard to extinction. 

Direct elimination hy carnivores. — The question as 
to how far the mammals of prey have been a direct 
cause of extinction at various times of various forms 
of quadruped life is considerably disputed. Morris 
(1895.1) observes: "So far as existing evidence goes, 
then, it seems probable that hostile aggression, while 
it may have been occasionally an indirect has rarely 
been the direct cause of the extinction of species." 
A similar opinion has been orally expressed to the 
author by D. G. Elliot — that no wild animal causes 
the extinction of another wild animal directly. Such 
a negative conclusion may hold true of undiminished 
herds and of conditions where carnivorous and her- 
bivorous animals have evolved together and, as in the 
evolution of the modern battleship, modes of defense 
have evolved simultaneously with modes of attack. 

This negative view in our opinion does not hold 
true where newly introduced Carnivora find quad- 
rupeds unprovided with adequate means of defense, 
as in the invasion of South America by carnivores 
from North America in late Pliocene time. Nor does 
it hold true of diminished herds of quadrupeds that 
are unable adequately to defend their young. 

Therefore we must consider the Carnivora as among 
the direct causes of the final extinction of diminished 
groups or reduced herds of animals which are struggling 



CAUSES OP THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



877 



to maintain their numbers against adverse conditions 
of pliysical environment, such as extreme cold, Iieat, 
and drought; of changing food supply; of competition 
with other quadrupeds; and of epidemics. Of Carniv- 
ora as destroyers of mammalian life the best recent 
treatment is that of J. Stevenson-Hamilton (1912.1, 
pp. 160-162, 355, 357). He says: 

It seems to be one of nature's provisions that the females of 
most wild animals are, generally speaking, more alert and 
nervous than the males and therefore, in proportion to their 
numbers, less frequently fall victims to the stealthy cat tribe 
than do the solitary members of the other sex. It is thus 
insured that the destruction of the mothers of the race by 
natural means shall not exceed the due bounds of economy 
nor be such as to hamper or retard proper progressive increase. 

Sudden introduction of smaller Carnivora. — T. S. 
Palmer (1899.1, pp. 93-94) has given a striking sum- 
mary of the influence of the mongoose: 

The common mongoose of India {Herpestes mungo or H. 
griseus), * * * a well-known destro3'er of rats, lizards, and 
snakes, was introduced into Jamaica * * * for the purpose 
of ridding cane fields of rats. * * * Various remedies were 
tried, but apparently with little success, until in February, 1872, 
nine individuals of the mongoose, four males and five females, 
from India, were introduced. These animals increased with 
remarkable rapidity and soon spread to all parts of the island, 
even to the tops of the highest mountains. A decrease in the 
number of rats was soon noticeable. * * * The mongoose 
increased, and as the rats diminished, its omnivorous habits 
became more and more apparent. It destro3-ed young pigs, 
kids, lambs, kittens, puppies, the native "coney" or capromys, 
poultry, game, birds which nested on or near the ground, eggs, 
snakes, ground lizards, frogs, turtles' eggs, and land crabs. It 
was also known to eat ripe bananas, pineapples, young corn, 
avocado pears, sweet potatoes, coconuts, and other fruits. 
Toward the close of the second decade the mongoose, originally 
considered very beneficial, came to be regarded as the greatest 
pest ever introduced into the island. Poultry and domesticated 
animals suffered from its depredations, and the short-tailed 
capromys {Capromys brachyurus) , which was formerly numerous, 
became almost extinct in some of the mountainous districts. 
The ground dove {Columhigallina passerina) and the quail dove 
{Geotrygon montana) became rare, and the introduced bobwhite, 
or quail, was almost exterminated. The pecuhar Jamaica petrel 
{Aestrelata carihhoea) , which nested in the mountains of the 
island, hkewise became almost exterminated. Snakes, repre- 
sented by at least 5 species, all harmless, and lizards, including 
about 20 species, were greatly diminished in numbers. The 
same thing was true of the land and fresh-water tortoises and 
the marine turtle {Chelone viridis), which formerly laid its eggs 
in abundance in the loose sand on the north coast. The destruc- 
tion of insectivorous birds, snakes, and lizards was followed by 
an increase in several injurious insects, particularly ticks, which 
became a serious pest, and a coccid moth, the larvae of which 
bore into the pimento trees. 

CONTRASTS BETWEEN EXTERNAL (ENVIRONMENTAL) 
AND INTERNAL CAUSES OF EXTINCTION 

Summarizing the nature and the effect of the exter- 
nal (environmental) causes of extinction we may note 
some features in which they contrast with those pecu- 
liar to extinction from internal causes, namely: 

1. In large part the external causes of extinction 
originate with cosmic changes — changes in the earth 



itself — such as the elevation or depression of the earth's 
surface and the extension or contraction of areas of land 
and water. These changes cause progressive increase 
of heat or cold imder conditions of either moisture or 
dryness; progressive increase of moisture or desiccation; 
and consequent changes of sod, vegetation, forestation, 
and water supply. From these physiographic and cli- 
matic changes result the introduction of new compet- 
itors for food and new enemies, new insect pests and 
new diseases. 

2. Under changing physiographic conditions, even 
though extremely gradual, many species and genera 
have become extinct. Secular desiccation in successive 
epochs of the Tertiary period, beginning in Oligocene 

. time but intensified chiefly in late Pliocene time, grad- 
ually became fatal to most of the browsing animals — - 
indeed, quite as destructive to them as the cold and 
moisture of the glacial epoch. 

3. Perhaps the most distinctive feature of extinction 
due to external causes is that it may exterminate the 
fit and the unfit alike, the adaptive and the inadaptive; 
it may destroy rather than improve a fauna. The 
extinction of many forms was certainly a consequence 
of the extension of glacial ice in North America and in 
Europe and of the desiccation in Australia. 

4. An equally notable featiire of the external causes 
of the extinction of a fauna is that they have generally 
acted locally or in certain regions of the earth's surface 
only, so that a part of the fauna affected was left to 
survive elsewhere. For example, the extinction of the 
horses and the proboscideans in North and South 
America during Pleistocene time did not prevent their 
survival in the Old World. 

In contrast with extinction thus arising from 
external causes is extinction arising primarily from 
internal causes, such as relative inadaptation, or 
imfitness, overspecialization, and irreversibility of 
evolution. These causes have acted simultaneously 
aU over the world, even imder imiformly favorable 
climatic conditions as, for example, in the extinction 
of the great orders of Creodonta, Amblypoda, and 
Condylarthra during the Eocene epoch. The extinc- 
tion that originates in internal causes — that is, in 
relative internal fitness or unfitness — may improve a 
fauna by eliminating the least-adapted members 
instead of destroying a fauna. 

INTERNAL CAUSES OF PRESERVATION AND EXTINCTION 

IMMUNITY AND ADAPTATION 

Conditions in Africa. — The existing conditions 
among the large quadrupeds of Africa are of especial 
interest because of the increasing evidence that the 
conditions in North America in Oligocene, Miocene, 
and Pliocene time are most closely paralleled by those 
that now prevail in the great upland region of Africa — 
the central Life belt as distinguished from the coast 
belt. 



878 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Recent discoveries (Bruce, 1905.1) indicate that 
immunity from disease has been one of the most potent 
causes of the adaptation of animals to their environ- 
ment, and that, conversely, nonimmunity has probably 
been one of the potent causes of their diminution and 
extinction. T. H. Morgan (1903.1) includes the 
phenomena of immunity among the adaptive processes. 
He states as his personal opinion, however, that 
certain of these phenomena could not be due to selec- 
tive processes. Similarly Leo Loeb (1905.1) believes 
that acquired immunity can not invariably be ex- 
plained as an adaptive phenomenon. 

Variations in immunity. — In Africa certain diseases 
are fatal to both wild and domesticated animals; 
others are fatal to domesticated animals but not to 
wild animals. Some animals succumb to certain 
diseases through successive generations; others; 
especially "natives," acquire immunity in the second 
generation. Still more remarkable is the fact that 
both wild and domesticated immunes may act as 
reservoirs of disease organisms which, through flies or 
ticks, may be carried to nonimmunes. The wild 
ruminants of Africa among the Bovidae especially — 
the buft'alo {Bos {Bubalus) cajfer), the kudu {Strep- 
siceros Tcudu), the wildebeest (ConnocJiaetes) — carry 
about in the fluid portion of their blood, without 
themselves suft'ering any harm, certain protozoan 
trypanosomes which are fatal when borne by flies to 
domesticated horses (Equidae), dogs (Canidae), and 
cattle (Bovidae). 

Thus causes favorable either to the genesis of disease 
organisms or to the acquirement of immunity or to the 
propagation and distribution of flies and ticks become 
matters of prime interest in relation to extinction. 

Origin of immunity as a "unit" character. — For the 
student of .extinction an important point to note in 
connection with "horse sickness" is that although 
means of artificial immunity are thus far undiscovered 
degrees of immunity and of natural immunity some- 
times occur. Such variations in respect to immunity 
would in a state of natm-e lead to the survival or grad- 
ual selection of immune forms and the consequent 
production of immune races. 

BULK NOT INHERENTLY INADAPTIVE 

Bulk supposed to he inadaptive. — There is a wide- 
spread belief or tradition that bulky animals have 
tended to disappear first — that bulk is in itself in- 
adaptive. Thus Owen, although as late as 1877 
(1877.1, pp. ix-x) disposed to attribute the extinction 
of the large mammals of Australia to the agency of 
man, advanced the theory that bulky size may be a 
disadvantage under changed conditions. He writes: 

In proportion to the bulk of a species is the difficulty of the 
contest which, as a living organized whole, the individual of 
such species has to maintain against the surrounding agencies 



that are ever tending to dissolve the vital bond and subjugate 
the living matter to the ordihary chemical and physical forces. 
Any changes, therefore, in such e.xternal conditions as a species 
may have been originally adapted to exist in wiU mihtate 
against that existence in a degree proportionate, perhaps in a 
geometrical ratio, to the bulk of the species. If a dry season 
be gradually prolonged, the large mammal will suffer from the 
drought sooner than the small one; if any alteration of climate 
affect the quantity' of vegetable food, the bulky herbivore will 
first feel the effects of stinted nourishment. * * * The 
actual presence, therefore, of small species of animals in coun- 
tries where larger species of the same natural families formerly 
existed is not the consequence of any gradual diminution of the 
size of such species but is the result of circumstances which 
may be illustrated by the fable of the "oak and the reed"; 
the smaller and feebler animals have bent, as it were, and 
accommodated themselves to changes which have destroyed 
the larger species. 

Morris observes (1895.1, p. 254): 

One tendency, which has particularly manifested itself in 
herbivorous animals, has frequently led directly to their de- 
struction. This is the tendency to increase in size through the 
double influence of abundance of food and little waste of tissue 
through exertion. In the sluggish grass eaters, dwelling on plains 
covered with rich herbage, or leaf and twig eaters in tropical 
forests, the nutritive agencies are in excess of those of waste, 
and these animals seem always to have tended to an increase 
in size, until those of least exertion and greatest powers of 
obtaining food became enormous in dimensions. An example 
of the same kind among the Carnivora is the Greenland whale, 
which, while feeding on minute forms, obtains them in enor- 
mous quantities with little muscular exertion and has in con- 
sequence become of extraordinary dimensions. 

Statistics as to handicap of hulk. — This general 
opinion as to the fatality of bulk in mammals is open 
to question. When we examine the matter statistically 
we find that a far larger number of families of small 
mammals have become extinct than of large mammals. 
The only families of bulky land mammals that have 
become entirely extinct since the beginning of Tertiary 
time are the following : 

Families of bulky land mammals that have become extinct since 
Cretaceous time 



Coryphodonts 

Uintatheres 

Titanotheres 

Py rotheres 

Lophiodonts 

Barytheres 

Arsinoitheres 

Elotheres 

Amynodonts 

Chalicotheres 

Diprotodonts 

Giant sloths (3 

families) . 
Glyptodonts 

Toxodonts 



Lower Eocene 

Upper Eocene 

Lower Oligocene.. 

Eocene 

Upper Eocene 

Upper Eocene 

Lower 01igocene__ 

Lower Miocene 

Lower Oligocene— 

Lower Pliocene 

Lower Pleistocene 
Pleistocene 

Pleistocene 

Pleistocene 



Holarctica. 

North America, Asia. 

Holarctica. 

South America. 

Eurasia. 

North Africa. 

North Africa. 

North America. 

North America. 

Eurasia. 

Australia. 

North and South 

Anaerica. 
North and South 

America. 
South America. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



879 



Number of extinct and surviving families of large and of small 
mammals since Cretaceous lime 





Families 
extinct 


Families 
surviving 


Percentage 
surviving 




16 

48 


9 
90 


36 




65 







The above figures would seem to indicate that the 
chance of survival of small mammals is nearly twice 
that of the survival of bulky mammals, but when 
we consider that many extinct families of small mam- 
mals remain to be discovered among the Marsupialia, 
Insectivora, Cheiroptera, and especially the Rodentia, 
the disparity indicated by the figures given is not so 
great. 

Among the large surviving land mammals are the 
families of elephants, hippopotami, rhinoceroses, 
giraffes, elands, and elks, and among the sea mammals 
are walruses and the various families of Cetacea, 
which, so far as known, are the largest mammals of all 
time. 

Further, we may recall the facts (1) that in lower 
Eocene time the small Condylarthra became extinct 
far earlier than the large Amblypoda; (2) that in 
lower Oligocene time many families of small Artio- 
dactyla and Perissodactyla became extinct at the 
same period as the large titanotheres, lophiodonts, 
and amynodonts; (3) that in Pleistocene time the 
relatively small Mylodon disappeared as early as the 
large Megatherium; (4) that the extinction of the 
mammoth in North America during or after the glacial 
epoch attracts attention because of the animal's large 
size, but, as indicated above, many smaller quadrupeds 
disappeared at the same time — for example, the horses 
and the sloths. 

Bulk and slow breeding not inimical to rapid evolu- 
tion. — The following arguments of Wallace and of 
Andrews in regard to bulk and slow breeding are in 
part fallacious and receive little support from paleon- 
tology. Wallace remarks (1876.1, vol. 1, pp. 158-159) : 

There is, however, another cause for the extinction of large 
rather than small animals whenever an important change of 
conditions occurs, which has been suggested to me by a cor- 
respondent but which has not, I believe, been adduced by Mr. 
Darwin or by any other writer on the subject. It is dependent 
on the fact that large animals as compared with small ones are 
almost invariably slow breeders, and as they also necessarily 
exist in much smaller numbers in a given area, they offer far 
less materials for favorable variations than do smaller animals 
In such an extreme case as that of the rabbit and elephant, the 
young born each year in the world are probably as some mil- 
lions to one; and it is very easily conceivable that in a thousand 
years the former might, under pressure of rapidly changing 
conditions, become modified into a distinct species, while the 
latter, not offering enough favorable variations to effect a 
suitable adaptation, would become extinct. 



The above argument, however, is not in accord with 
the facts ; the slow-breeding elephant evolved with far 
greater rapidity than the swift-breeding mouse. 

C. W. Andrews has recently (1903.1, p. 2) revived 
this argument that the lengthening of the time taken 
to attain sexual maturity may affect the rate of evolu- 
tion and under changed conditions where a rapid rate 
of evolution is essential may cause extinction. He 
writes : 

In many ungulates this increased longevity is indicated by 
various modifications of the teeth, tending to give them a 
longer period of wear: generally this end is attained by the in- 
creasing hypselodonty of the cheek-teeth. A necessary con- 
sequence of the longer individual life will be that in a given 
period fewer generations will succeed one another, and the rate 
of evolution of the stock will therefore be lowered in the same 
proportion. If now the conditions of life undergo change, the 
question whether a given group of animals will survive or be- 
come e.xtinct will depend upon whether it can undergo suf- 
ficiently rapid variation to enable it to avoid getting so far out 
of harmony with its surroundings that further existence becomes 
impossible. It seems to follow, then, that the smaller animals, 
in which the generations succeed one another rapidly, will have a 
better chance of surviving than the larger and more slowly 
breeding forms, which at the same time will be stiU further 
handicapped if, as is usually the case, they are more highly 
specialized than the smaller forms and therefore have a more 
restricted range of possible variation. 

This argument is contradicted by all the facts of 
paleontology: there is no relation between rapid 
breeding and rapid evolution. 

As against these purely hypothetical considerations 
paleontology shows that during Pliocene and Pleisto- 
cene times the slow-breeding Proboscidea evolved 
quite as rapidly if not more rapidly than the rapid- 
breeding Rodentia. 

Relation oj bulJc to nutrition. — Stromer (1905.1, pp. 
97-132) discusses in detail the relation of bulk to 
nutrition. He observes that nourishment or lack of 
it must not be overestimated as a factor in extinction, 
for it has been shown among mammals that the nourish- 
ing surfaces are only squared as the mammals increase 
in size, while the bulk of the body is cubed, and that 
small forms eat much more relatively than large forms. 
The fact remains that the great terrestrial animals need 
as a rule a large amount of vegetable food and also an 
abundance of water. Aquatic mammals, on the other 
hand, embrace more giant carnivores (cetaceans, 
walruses) than herbivores (hippopotami, Rhytina). 
Naturally, defenseless giant forms incapable of fight- 
ing can maintain their existence only in the absence 
of destructive Carnivora — for example, the giant rodent 
{AmUyrliiza) of the Lesser Antilles. Such factors as 
the diminution of food supply would weigh much more 
heavily against ponderous animals like Glyptodon and 
Megatherium than against agile ones such as giraffes, 
elephants, and rhinoceroses. While the glacial epoch 
may have destroyed the northern representatives of 
Hippopotamus, it must be recalled that this genus at 



880 



TIT.\NOTHERES OP ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



the same time became extinct in India, Java, and 
Madagascar, while many of the large European mam- 
mals survived the glacial epoch. Some very large 
animals, such as the elephant, for instance, have been 
shown to be veiy adaptable. At the present time the 
elephant is found in the damp jungle and parklike 
districts, in tree and bush steppes, in low-tying plains, 
on plateaus and mountains. Salensky (1905.1) has 
lately shown that the mammoth was fully adapted to 
life on grassy plains of high latitudes. The mastodon 
also occupied many different life habitats. 

Conclusions as to effect of hulk. — Our conclusion is 
that bulk is not intrinsically or per se fatal. All the 
bulky mammals which have disappeared have pos- 
sessed some single inadaptive organ or set of organs 
which it seems probable would have proved equally 
fatal to animals of small size; they have lacked in 
intelligence, resourcefulness, or brain power; they have 
specialized to an extreme in certain directions; they 
have reached a point where there was only a single 
mode of living left to them, and when this was altered 
or when a particular source of food supply was cut off 
they perished with it. On the other hand, intelligent, 
resourceful, and adaptively constructed large mam- 
mals, lilce the elephant, have survived great physio- 
graphic and bio tic changes. Presence or absence of 
such adaptations is almost equally fatal to small ani- 
mals; bulk in itself is not a cause of extinction; in- 
adaptive bulky animals have disappeared side by side 
with the inadaptive diminutive animals. 

The chief inadaptations of large mammals are the 
following: (1) Disadvantage of the large amount of 
food required by a large animal, which is offset by the 
advantage that many large animals can travel long 
distances; (2) diminished birth rate, which is a charac- 
teristic of large animals, is a point to be noticed; as 
a rule, the larger the animals the fewer the young and 
the less able a species would be quickly to regain 
numerical strength after some widespread diminution 
in number; the diminished birth rate is, however, 
offset by greater longevity and greater power to pro- 
tect young from enemies. Darwin (1859.1, p. 64) 
observes : 

The elephant is reckoned to be the slowest breeder of all 
known animals, and I have taken some pains to estimate its 
probable minimum rate of natural increase; it wiU be under 
the mark to assume that it breeds when 30 years old and goes 
on breeding till 90 years old, bringing forth three pair of young 
in this interval; if this be so, at the end of the fifth century there 
would be alive 15 million elephants, descended from the first 
pair. 

VALUE OF SINGLE ORGANS IN SURVIVAL OR EXTINCTION 

Percentages oj natural increase and decrease. — Recent 
conditions among the large African mammals as ob- 
served by Stevenson-Hamilton (1912.1, pp. 9-95) show 
the effects of numerical reduction and the highly varied 
action of the same causes on different stocks and dif- 



ferent organs. As to numerical increase or decrease 
the author estimates (p. 10) that (1) the actual net 
annual increase of stock of the larger antelopes under 
ordinary favorable natural conditions is not above 5 
per cent; (2) in a region where the carnivorous ani- 
mals have been destroyed or reduced in number the 
net annual increase is about 10 per cent; (3) the gross 
natural increase of the larger antelopes in ordinary 
years is 20 per cent, which is offset by the death an- 
nually of 5 per cent from old age and the destruction 
of 15 per cent by the Carnivora, especially the young 
and worn-out animals; (4) anything that checks the 
gross natural increase and hastens the natural decrease 
means the permanent reduction and perhaps the final 
extinction of the stock. This balance affects the sur- 
vival or elimination value of single organs. 

Extinction oj Artiodactyla. — Woldemar Kovalevsky, 
a Russian paleontologist, was one of the pioneers 
in the consideration of the survival or extinction value 
of single organs. He observes, in his great mono- 
graph (1876.1, p. 152), that the extinction of all 
Artiodactyla having an inadaptive foot structure 
and inadaptive grinding teeth occurred as follows: 
Upper Eocene: XipJiodon, Ano-plotTierium, Biplopus; 
Oligocene: Hyopotamus, AntJiracotTierium, Entelodon. 
He pointed out that the inadaptation of the foot in 
these animals consisted of a mechanical defect in the 
manus, the third metacarpal not spreading above to 
articulate with the trapezium as in the adaptive manus 
of the pig and hippopotamus, and that the inadapta- ■ 
tion in the grinders consisted of the persistent short 
or brachyodont crowns, bunoselenodont and bunodont, 
composed of partially rounded cones. The feet, 
being mechanically weak in the function of the carpals 
and metacarpals, were incapable of the elongation into 
cannon bones, a cursorial or speed adaptation which 
saved the lives of artiodactyls with the adaptively re- 
duced digits. The short teeth were by his theory 
not adapted to a supposed change of vegetation from 
softer herbage to harder Gramineae. His paleozoologic 
supposition that such a change of food occurred was 
independently confirmed by the paleobotanists Saporta 
and Marion. His conclusion as to single organs caus- 
ing extinction (which was original) has since been 
abundantly confirmed by subsequent observations of 
the extinction of all forms of quadrupeds having these 
inadaptive types of short-crowned grinders, both in 
North America and in India. 

Inadaptation of cone and crescent teeth. — The buno- 
selenodont or cone and crescent molar pattern is 
typified in the titanotheres and consists of one or 
two detached cones on the inner side of the upper 
grinding teeth and of two crescents on the outer side. 
It is adapted to browsing on coarse, soft food rather 
than on hard, fine food. 

This dental type presents a cul de sac of evolution, 
because it is incapable of transformation either into 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



881 



the elongate or hypsodont grazing type or into the 
lophodont or crested type. In the absence of trans- 
verse crests between the inner cones and the outer 
crescents elongation would produce a number of 
separate columns and crescents. 

It is interesting to observe, in the evolution of the 
titanotheres, the attempt of nature to create a hypso- 
dont crown out of this cone and crescent type by the 
elongation of the crescents on the outer side of the 
superior grinding teeth. The inner cones do not 
share in this elongation; the result is a tooth half 
hypsodont and half brachyodont, obviously a very 
poor mechanism. 

The bunoselenodont tooth was therefore incapable 
of further evolution in any direction; no mechanical 
progress or perfection was possible. 

Against the theory that the form of the teeth alone 
caused the extinction of these great animals we must 
record the fact that the giant elotheres of lower 
Oligocene time, which had teeth that were still less 
effective (of an all-cone pattern), survived to a later 
geologic period than the titanotheres. If we knew 
all about the life of these elotheres we should probably 
find that they had some compensating advantage in 
local habitat, perhaps recourse to river-border life. 

The hypothesis that the bunoselenodont tooth 
pattern was, if not the sole, at least a potent cause of 
extermination is supported by strong collateral 
evidence, which is presented in the accompanying 
diagram (PI. XLVIII). It appears that all buno- 
selenodont quadrupeds, whether belonging to the 
Artiodactyla or the Perissodactyla, disappeared during 
the Oligocene epoch or early in the Pliocene. The 
only bunoselenodont mammals that survived are the 
chalicotheres, which persisted in the Northern Hemi- 
sphere until the end of Pliocene time, probably in a 
forest habitat. The defective tooth structure in 
these animals was probably compensated for by the 
development of giant claws on the feet, which gave 
these strange quadrupeds certain advantages similar 
to those that were enjoj^ed by the giant sloths. 

Elongation (hypsodonty) of the molar teeth in relation 
to longevity and reproductive power. — The elongation 
of the molar teeth is obviously a direct advantage in 
the promotion of longevity and fertility, an advantage 
that might lead to selection of fluctuations in length. 
The elongate crowns of the teeth of the elephants 
and the horses enable individuals to live many years 
and to produce a large number of young. Elephants, 
which are provided with successive long-crowned 
teeth, live, according to Darwin, between 90 and 
100 years and, although slow-breeding animals, pro- 
duce a large number of young. Horses, which have 
long-crowned teeth, live to the age of 25 years and, 
foaling every year, multiply with great rapidity. 

In contrast, an animal like the titanothere Palaeo- 
syops, with its short-crowned teeth, would live a com- 



paratively short time and produce comparatively 
few young. Through long periods of geologic time 
this relation of longevity to reproduction would 
tend in the same habitat to replace races that had 
short-crowned teeth and that were therefore short 
lived with races that had long-crowned teeth and 
that were therefore long lived. 

As a matter of fact, this theoretical condition is 
modified by change of habitat, because we observe 
that the short-crowned Cervidae of browsing habits 
hold their own in the forests and that the generally 
long-crowned Antilopidae are perfectly adapted to life 
on the plains. 

Inadaptation of small hrain. — The chief advantages 
of brain capacity undoubtedly appear in relation to 
adaptability of habit and resourcefulness in times of 
exposure, to alertness in avoiding new dangers to 
which the young may be exposed, and to enterprise in 
seeking new habitats, qualities that are especially 
valuable at times of climatic change and of severe 
competition. 

Modern quadrupeds differ widely in regard to 
resourcefulness under adverse conditions of environ- 
ment, as illustrated on the western plains during the 
great winter storms. When sheep, cattle, and horses 
meet the im expected conditions incident to a blizzard, 
the sheep disappear first, the cattle second, the horses 
last — that is, the horses, largely owing to instincts 
which they have inherited from northern ancestral 
strains, meet new conditions of life in most extraor- 
dinary ways, whereas cattle of southern wild ancestral 
strains are far less resourceful. 

The paleontologist knows nothing of the psychic 
qualities of an extinct animal; he can judge its 
brain power only by examining its intracranial cast, 
which often reproduces the size and external form 
of the brain with exact fidelity. Lartet, following 
Cuvier, was among the first to allude to the law 
of progressive cerebral development of certain of the 
Tertiary mammals. He says (1868.1, pp. 1120-1122) : 

II rfeulterait en effet d'un certain nombre d'observations 
relevees a divers (Stages de la stratigraphie tertiaire, que, plus 
les mammiferes remontent dans I'anoiennetil des temps geolo- 
giques, plus le volume de leur cerveau se r^duit par rapport 
au volume de leur tete et aux dimensions totales de leur 
corps. * * * On a dit que les plus grands mammiferes 
sont ceus qui vivent le plus longtemps; ce qui serait plus pres 
de la verity, c'est que la longevite normale parait s'aocroitre 
en raison directe du volume absolu du cerveau. 

Marsh in 1884 considered brain size in relation to 
the final extinction of thQ uintatheres. He observes 
(1884.1, p. 190): 

The small brain, highly specialized characters, and huge bulk 
rendered them incapable of adapting themselves to new con- 
ditions, and a change of surroundings brought extinction. 
* * * The Dinocerata, with their very diminutive brain, 
fixed characters, and massive frames, flourished as long as the 
conditions were especially favorable, but with the first geo- 
logical change they perished and left no descendants 



882 



TIT,\NOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



In discussing the brain, especially after referring to 
the general law of brain growth (that is, evolution) 
during the Tertiary period, he states: 

To this general law of brain growth two additions may now 
be made, which, brieflj- stated, are as follows: (I) The brain of 
a mammal belonging to a vigorous race, fitted for a long survival, 
is larger than the average brain, of that period, in the same 
group; (2) the brain of a mammal of a declining race is 
smaller than the average of its contemporaries of the same 
group. 

The ratio of brain weight to body weight in recent 
mammals has been made the subject of an exhaustive 
investigation by Max Weber (1896.1), who publishes 
the ratios between brain weight and body weight in 



The chief measure of the brain capacity of extinct 
mammals is indicated by (1) absolute size and weight 
of the brain, (2) ratio of brain weight to body weight, 
(3) development of the cerebral convolutions, (4) 
proportion between the frontal and parieto-occipital 
lobes of the cerebrum. 

Exceptions: certain small-hrained types survive. — The 

series of diagrams in Plate XLIX shows the enormous 

contrast between the brains of a number of Eocene 

mammals and those of a number of existing mammals 

of equal size and equal bulk. The contrast is drawn 

between pairs of animals of somewhat similar habits 

of life. Thus the Eocene Arctocyon (A) is contrasted 

with the existing dog (Canis), the Eocene Phena- 

codus (B) is contrasted with the 

domestic pig (Sus), both animals 

possessing bunodont teeth and 

somewhat similar omnivorous 

feeding habits. The Eocene 

Coryphodon (C) is contrasted with 

the existing rhinoceros. Finally 

the Eocene Uintatherium (D) is 

contrasted with the Oligocene 

titanothere Menodus giganteus. 

The brains of the archaic mam- 
mals shown on the left in Plate 
XLIX {Arctocyon, Phenacodus, 
Coryphodon, Uintatherium) are not 
only extremely small in proportion 
to the size of the animals but are 
of a low type; they have large 
olfactory lobes, small cerebral 
hemispheres, with small frontal 
lobes; and they are relatively 
smooth — that is, they are without 
convolutions. 

An equally intrinsic inadapta- 
tion seems to have been the inca- 
pacity for progressive increase in 
size of brain in successive geologic 
Figure 760. — Brain proportions in Eocene perissodactyls, an artiodactyl, and an periods; in other words, the brain 
amblypod of similar size ^^^ consequently the psychic 

D, coionoceras agresiis: E, Hyrachyus powers appear to be in a Condition 
of arrested development. The 










A., Coryphodon hamatus; 



Ltmnohyops laiiceps; C, Palaeosyops Tobustu 
bahdianus; F, Amynodon advenus; G, Eporeodonsocialis. 



thirteen existing orders of mammals. In general, (1) 
small mammals have relatively larger brains than 
large mammals; (2) within a natural order brain 
weight does not increase relatively with body weight; 
(3) in growing individuals the relative brain weight 
decreases ; (4) in absolute brain weight man is exceeded 
only by the Proboscidea and Cetacea, in relative brain 
weight man is exceeded only by certain of the small 
mammals; (5) in the very primitive marsupials the 
ratios of brain weight to body weight vary from 
1:110 to 1:711 as compared with the Carnivora, 
in which the ratio varies from 1:100 to 1:546. Brain 
weight, however, does not register the great disparity 
in intelligence as a survival factor between the Mar- 
supialia and the Carnivora fissipedia. 



only exception to this rule is seen in one family of 
archaic creodonts, the Hyaenodontidae, in which the 
brain attained considerable size in the surviving mem- 
bers of lower Oligocene time. 

This limited brain power and arrested brain evolu- 
tion placed these archaic quadrupeds at a great dis- 
advantage in competition with the more advanced 
placental mammals, which suddenly appeared in lower 
Eocene time. The cursorial Phenacodontidae meas- 
ured their psychic powers with the cursorial Equidae; 
the small-brained Creodonta generally competed with 
the incoming true Carnivora, with their progressive 
increase in size of brain. 

The long survival and steady increase in size of 
the clumsy Amblypoda — Pantolambda (basal Eocene), 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



883 



Coryphodon (lower Eocene), and UintatJierium (upper 
Eocene)— despite arrested brain development, is one 
of the most astonishing phenomena of Eocene mam- 
malian life. Their survival may be accounted for by 
the development of very effective defensive weapons, 
the tusks and horns wherewith they stood off their 
enemies and protected their young. The final extinc- 
tion of these mammals may be attributed to two 
causes — low brain power, which may have inhibited 
the proper defense and care of the young, and the 
arrested evolution of the grinding teeth, which were 
actually no larger and little more effective for the 
comminution of food in the giant Dinocerata of upper 
Eocene time than in the smaller Coryphodon of lower 
Eocene time. 

In a race that especially develops tusks and horns, 
both probably favored by sexual selection, the grind- 
ing teeth tend to show arrested evolution. The 
disappearance of the small phenacodonts long prior 
to that of the large Amblypoda is another illustration 
of the fallacy of the widespread belief that bulky 
animals tend to disappear first. 

The competition of the archaic small-brained 
creodont Carnivora with the diminutive large-brained 
pro-Carnivora in Eocene time may be only remotely 
compared with the extinction of the Tasmanian 
wolf (Thylacinus) and Tasmanian "devil" (Sarco- 
pMlus) through the introduction of the true dog 
{Canis dingo) on the Australian mainland. The 
steady increase in size of the creodont carnivores, as 
displayed in Patriqfelis and the enormously powerful 
Harpagolestes, may be placed parallel with the increas- 
ing size of the equally small-brained Amblypoda. 

Exceptions: large-brained types perish. — Against this 
strongly cumulative evidence that the brain is one of 
the single organs whose development has been a deci- 
sive factor in the preservation or extinction of races of 
animals we must place certain exceptions. The extinct 
short-limbed rhinoceros (Teleoceras) of the upper Mio- 
cene evidently had a larger brain than the surviving 
true rhinoceros. Again, it appears from intracranial 
casts that the Pleistocene mastodon of North America 
had a brain fully as large as that of the existing ele- 
phants. In both Teleoceras and the mastodon great 
cerebral development failed to preserve the race. Con- 
versely, certain very small-brained animals, notably 
the North American opossums (Didelphiidae) and the 
Insectivora, have survived, even to the present time. 
The interpretation put upon these exceptions is that a 
large cerebral development may be insufficient to 
compensate for the possession of certain disadvanta- 
geous organs. 

INADAPTATION OF EXTREME SPECIALIZATION 

Extreme specialization may constitute a chain of 
causes that lead to extinction. An animal may 
specialize to an extraordinary degree in a single mode 
101959— 29— VOL 2 13 



of subsistence, and the food upon which it subsists may 
be diminished or destroyed ; or a race of animals may 
expend its energy largely in the development of certain 
single organs, such as horns or tusks, which become 
dominant and interfere with the proper development 
of other organs. If a race undergoes a marked de- 
generation or loss of organs it can not retrace its steps, 
because, to use DoUo's dictum, evolution is irrever- 
sible — that is, lost parts can not be regained. 

SURVIVAL OF THE UNSPECIALIZED 

The extinction of the highly specialized has been so 
frequently observed that it was formulated into a law 
by Cope (1896.1, p. 173): 

Agassiz and Dana pointed out this fact in taxonomy, and 
I expressed it as an evolutionary law under the name of the 
"doctrine of the unspecialized." This describes the fact that 
the highly developed or specialized types of one geologic period 
have not been the parents of the types of succeeding periods, 
but that the descent has been derived from the less specialized 
of preceding ages. No better example of this law can be 
found than man himself, who preserves in his general structure 
the type that was prevalent during the Eocene period, adding 
thereto his superior brain structure. 

The validity of this law is due to the fact that the specialized 
types of all periods have been generally incapable of adaptation 
to the changed conditions which characterized the advent of 
new periods. Changes of climate and food consequent on 
disturbances of the earth's crust have rendered existence 
impossible to many plants and animals and have rendered life 
precarious to others. Such changes have been often especially 
severe in their effects on species of large size, which required 
food in large quantities. The results have been degeneracy or 
extinction. On the other hand, plants and animals of un- 
specialized habits have survived. 

IRREVERSIBLE EVOLUTION 

As shown above in the discussion on bunoselenodont 
teeth, an organ must be capable of development in 
a new direction to supply the new needs of the 
organism; if the organ is lost, degenerate, reduced, 
or mechanically incapable of further development 
it constitutes a bar to the survival of its possessor 
(DoUo, 1893.1). For example, a change of vegetation 
in Oligocene time appeared to favor the animals 
that were capable of cropping their food as well as 
grinding it up into fine particles with the molar teeth. 
This change found the titanotheres possessed of 
roimded, button-like incisor teeth, totally useless for 
cropping. This loss of useful cropping teeth may not 
have been compensated for, as it was with the square- 
lipped grazing rhinoceroses, by the evolution of 
effective upper and lower Hps. It may be said, 
therefore, that the titanotheres had lost all power of 
adaptation to grazing habits through the degeneration, 
simplification, or absence of their incisor teeth. 

INADAPTATION OF DOMINANT ORGANS 

Organs that have reached a stage of development so 
extreme as to require a larger share of the sum total 
of bodily nutrition than their general or apparent 



884 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



utility justifies may be Icnown as dominant organs. 
They violate the St. Hilaire law of economy of growth 
and compensation of parts, or the highest combination 
of adaptable structure through the subservience of 
each organ to all the other organs. The great horns 
of the titanotheres, the giant horns of the Irish deer 
(Ceruus megaceros), the single tooth of the narwhal 
(Monodon)', the giant tusks of the elephants are 
examples of dominant organs. 

The very development of such organs has long been 
considered one of the possible causes of extinction, 
We immediately observe that certain dominant organs, 
like the tusks of the elephant, serve a great variety 
of useful purposes; they are by no means developed 
solely for service in the struggle between the bulls for 
possession of the females. Again, if large horns are 
considered fatal, it is noteworthy that the small- 
horned titanotheres {Menodus, Brontops) disappeared 
simultaneously with the long-horned titanotheres 
(Brontotherium and Megacerops). Were it not for 
this fact we might have attributed the extinction of 
the titanotheres to the possession of such apparently 
useless dominant organs as their horns. 

SELECTION OF SEXUALLY DOMINANT ORGANS 

The general explanation that has been offered for 
certain dominant characters peculiar to the males 
is that they have been developed or perhaps over- 
developed, through the competition of the males for 
the possession of the females. Thus, in respect to 
horns, incisor and canine tusks there has been an 
incidence of selection on organs that were employed 
for combat between males and that were of Httle or 
no use at other times except in standing off carnivorous 
enemies. 

Noteworthy facts derived from our study of the 
titanotheres are that the horns first arise alike in 
both sexes as rudiments, or extremely small horns; 
that they appear to be equally developed in the males 
and the females; and that they gradually become 
distinctively male characters, so that the sexes are 
sharply separated by the size and development of the 
horns. The horns were undoubtedly of advantage 
to the males in their sexual combats for the possession 
of the females, and a constant selection of individuals 
with the largest horns may have been in process. 
This main emphasis of natural selection on characters 
which are useful for competitive and combative 
purposes, it is argued, may have been the cause of the 
arrested evolution of the grinding teeth. 

It is certainly true that the grinding teeth attained 
a higher mechanical perfection in the short-horned 
titanotheres than in the long-horned brontotheres, 
and this would serve to prove that in the menodonts 
the teeth were being favored by selection, while in the 
brontotheres the horns were being favored by selection. 
As noted above, the superiority of the menodonts in 



the matter of tooth structure did not save them from 
the extinction which overtook the brontotheres. 

CAUSES OF OVERDEVELOPMENT 

Naturalists generally have attributed overdevelop- 
ment to the selection of favorable fluctuations in size, 
but Cope (1896.1, p. 480) invoked a certain kind of 
intrinsic hereditary force. He says: "I have therefore 
assumed as a working hypothesis the existence of the 
bathmic energy and have inquired how far the facts 
in our possession sustain it." Similarly, the over- 
development of dominant organs has recently been 
explained by F. B. Loomis (1905.1, p. 843) as the 
result of "momentum." After discussing many 
known instances of organs that appear to have passed 
a stage of utility, he observes: 

The above are selected examples in which a feature once 
useful has been developed beyond its maximum utility. Many 
others equally striking might be cited, the e.xplanation of all of 
which is extremely difficult unless such a factor as momentum is 
called in. In the light of this factor, however, a logical and ap- 
parent cause is found. Momentum also explains why a charac- 
ter that originated in accordance with the environment develops 
so rapidly, and why, when an animal had reached adjustment 
to its surroundings, it still goes on beyond a perfect adjustment. 
It may be laid down as a rule, then, that a variation started along 
any line tends to carry that line of development to its ultimate, 
being driven by momentum. If the feature is detrimental the 
group dies out; if, however, it is merely a minor feature it makes 
a handicap. A line of development may be stopped and its 
momentum overcome, but the tendency is to keep right on. 

RATES OF BREEDING AND EXTINCTION 

It has already been shown that failure or inability 
to protect the young, or possibly the inhibition of 
fertihty through low temperature, must be among the 
causes of extinction, and as extinction rapidly follows 
when the death rate exceeds the birth rate, the rate 
of breeding must be considered in any inquiry as to 
extinction. Arthur Erwin Brown, in a letter written 
October 24, 1907, has called the attention of the 
writer to this subject and has contributed some 
valuable notes. He observes: 

The thought suggests itself that the long breeding period may 
have been a factor in the decline of the Perissodactyla. In aU 
perissodactyls gestation is slow. Hodgson 70 years ago gave 
the Indian rhinoceros 17 to 18 months; it does not appear that 
these figures have been amended. 

Dr. Frank Baker, late superintendent of the 
National Zoological Park at Washington, fixed the 
gestation of Tapirus terrestris at 396 days. The gesta- 
tion period of such an animal as the tapir, adapted to 
the long summers and short winters of temperate 
climates, would naturally be unadapted to the long 
winters and short summers of the glacial climates of 
Pleistocene time; consequently the extinction of the 
Tapiridae in North America during late Pleistocene 
time may have been due to the exposure of the newly 
born young to climatic conditions to which they were 
not adapted. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



885 



Stevenson-Hamilton (1912.1, p. 44) observes that 
the gestation period of the African antelopes is about 
eight months. Most females seek sheltered spots 
before calving. About calving time great destruction 
is wrought by the Carnivora, especially by the hunting 
dogs (Lycaon pictus), which run down the little ante- 
lopes with great ease and destroy large numbers of 
them. Within a very few days of birth all young 
antelopes and zebras are capable of keeping up in the 
most surprising manner with the fully grown ani- 
mals even when the latter are going at full speed. 
■ (See p. 846.) 

SELF-EXTINCTION THROUGH ARRESTED VARIATION 

The Italian geologist Brocchi (1814.1, vol. 1), the 
author of an able work on the fossil shells of the sub- 
Apennine hills, advanced the hypothesis of some 
regular, constant law by which species may disappear 
from the earth gradually and in succession. The 
death of a species, like the death of an individual, 
he suggested, may depend upon certain peculiarities 
of constitution conferred upon the species at its 
origin, and as the longevity of the individual depends 
upon a certain force of vitality, which after a period 
grows weaker and weaker, so the longevity of the 
species may depend upon the quantity of prolific 
power originally bestowed upon it; after a season the 
species may decline in energy and its fecundity may 
be gradually lessened from century to century. 

Lyell opposed this doctrine on the ground that 
there is seldom evidence of physiological deterioration 
in the last representatives of a species. 

Neumayr also opposed this doctrine. He remarks 
(1889.1, vol. 1, pp. 142-143, 147): 

It has been assumed that species, like individuals, go through 
a prescribed cj'cle of life, that they arise, flourish, decline, and 
die, unless they undergo a sort of rejuvenation through gradual 
variation, and thus extinction has been attributed to the ina- 
bility to continue to vary. It is undeniable that countless forms 
have become extinct because they could not adapt themselves 
rapidly enough to changing conditions, even where these 
changes covered thousands of years. However, paleontology 
affords no proof — nor is any evidence to be found elsewhere — 
for the more extreme view that forms maintain their power to 
vary for only a limited time, after which they become rigid and 
unadaptable. That any animal ever ceased to show variations 
is a purely arbitrary assumption [p. 143]. Nor is the analog}' 
between decay and death of the individual and of the species 
or family justified, for senile degeneration and death are by no 
means peculiar to all living organisms. Among the Protozoa 
death is the result of external violence, not the necessary out- 
come and prescribed end of the life cycle, and only among the 
higher organisms with more complicated methods of repro- 
duction does it become so. In everj' instance extinction can 
be explained on the basis of the struggle for existence and 
without recourse to any mysterious inner causes [p. 147]. 

The idea that self-extinction is caused by an inherent 
arrest of variation was expressed in another form by 
Darwin and Wallace, namely, that, as a limitation or 
cessation of variation would cut off material for 



improvement through selection, a fixed or nonadapt- 
able type would arise, and extinction would follow. 
It has been revived or discussed by Doderlein 
(1888.1), Rosa (1903.1), Abel (1904.1), Plate (1904.1), 
and Stromer (1905.1). The rdle assigned to the limi- 
tation of variation (independent of the efJects of envi- 
ronment) as a cause of extinction again depends upon 
direct observation as to the modes of evolution. If 
there is a hereditary progressive trend (such as 
"Mutationsrichtung") in evolution leading in certain 
directions, is there also an arrest of such movement? 

Doderlein (1888.1, p. 394) advanced the opinion, 
especially as a consequence of certain opinions of 
Cope, that in a long-continued evolution in one direc- 
tion there is inherited not so much a definite condition 
as a tendency to continue to develop in that direction. 
In spite of the law of inertia (Tragheitsgesetz) this 
inherited tendency continues even if it is no longer 
useful to the organism unless it is offset by powerful 
counterforces, such as natural selection under condi- 
tions of severe competition. Thus organs may arise 
that are directly harmful to their possessors and may 
contribute to their destruction, as, for instance, the 
excessively large canines of the last machaerodont 
tigers and the gigantic antlers of Cervus eurycerus and 
C. dicranius. 

Rosa (1903.1) discussed the hypothesis of the pro- 
gressive reduction of variability and emphasized the 
fact that highly specialized organisms may show 
variations but that these variations do not lead to 
new phyla, or, if they do, only to a slight extent. 

Plate's discussion and criticism (1904.1, pp. 641- 
655) of Rosa's views is as follows: 

Daniel Rosa in 1899 published a paper [1899.1] attempting to 
prove that there is a progressive reduction of variability in 
species, which leads ultimately to their extinction. In the 
first place, Rosa bases his "law" on the fact that evolution in 
many cases means a constantly increasing specialization. 
Species tend to adapt themselves more and more closely to 
definite environmental conditions and become ever more 
specialized. The smaller the circle of external factors to which 
a species is adapted, the more limited becomes its "phylogenetic 
capacity" (Rosa), the capacity to produce new and distinctly 
different species. As pointed out by Cope in his "law of the 
unspecialized," the great phylogenetic lines originated in 
unspecialized forms with great potential adaptability. But 
the struggle for existence drives organisms into more and more 
extreme specialization, thereby diminishing their chances of 
becoming stem forms of great phylogenetic lines. Therefore 
one may speak of a "law of progressive specialization." Rosa 
calls it the "law of progressively diminished variation" and 
extends its application beyond the cases of very highly special- 
ized forms. * * * 

However, instances like that of the African elephant, which 
shows several geographic varieties, go to prove that highly 
specialized genera do not lose the power to produce new species, 
although they can obviously riot become the originators of 
entirely new and divergent groups. Thus specialization limits 
the "breadth of evolution" (Evolutionsbreite) but does not, 
as maintained by Rosa, check evolution itself. * * * 

Rosa goes on to set up a law of progressively diminished 
variability as the cause of progressively diminished variation. 



886 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



This leads directly to the further assumption that every species 
approaches its constant form (Rosa, p. 58). Thus he supposes 
that through inner causes, which we are at present unable to 
analyze, the power to vary gradually vanishes in the course of 
evolution, and that every phyletic series leads finally to rigid 
forms, which, being unequal to any further changes in their 
environment, are doomed to extinction. * * * 

Rosa's law of progressively diminished variability is based on 
a fallacy, for, except in cases of parasitism, etc., evolution leads 
to greater and greater complication, and variability, which in 
general depends on the number of variable elements, increases 
as the number of tissues and organs and the number of their 
characters is augmented. If, as Rosa maintains, variability 
and the number of variable elements became progressively 
reduced in the course of evolution, the higher organisms could 
not have been derived from the lower. 

In high specialization, as for instance, when the primitive 
mammalian foot is converted into the fossorial extremity of 
the mole, the saltatorial extremity of the kangaroo, or the curso- 
rial extremity of the horse, many characters are no doubt lost 
and variability appears to be diminished, but this is only the 
negative aspect, and in reality these adaptations are due to 
the acquisition of new characters by certain structures. * * * 

Rosa conceives of the splitting up of forms into families, 
genera, and species as a distribution among the descendants of 
possible variations a, b, c, d, e, . . . z, of every organ of 
the stem form. Thus one phyletic series will have the possi- 
bilities a-f, another g-o, a third p-z with respect to the same 
organ — for instance, the foot. The end result in each case will 
be rigid forms, one with the foot a, another with h, a third 
with k, etc. This conception is erroneous. As a matter of 
fact, the stem form is split up into new species by the acquisition 
of new characters, but not by a distribution of variations among 
the various groups of descendants. 

Thus, finally, while Rosa is right in holding that highly spe- 
cialized species readily become extinct under altered environ- 
mental conditions, he is mistaken in believing that variability 
itself is limited by a one-sided organization. Variability has 
nowhere been wholly wanting; it has mierely worked too slowly 
and too incompletely at times to prevent extinction. There- 
fore not only megatheres and ichthyosaurs but countless simple 
Protozoa, coelenterates, and echinoderms as well have become 
extinct. 

Simiiarly, C. B. Crampton (1902.1), as cited by 
C. W. Andrews (1903.1, p. 1), suggests internal causes 
of extinction as follows: 

In a recent paper by Mr. C. B. Crampton a possible inherent 
cause of extinction is suggested. It is impossible to do justice 
to this interesting paper in a short note, but the gist of the 
argument seems to be as follows: In the original unicellular 
organism the possibilities of variation are almost infinite, but 
as soon as evolution along any line begins, these possibilities 
are restricted, and become more and more so the more highly 
specialized the animal is; in short, the potential variation of an 
organism becomes less and less as specialization advances. 
Furthermore, under the influence of natural selection in each 
generation the individuals which tend to vary in the same 
direction will survive, while at the same time, as already 
pointed out, their capacity for variation becomes more and 
more restricted. The consequence of this will be that the more 
highly specialized any stock becomes the more the individuals 
composing it will come to resemble one another, until at length 
the same results as arise from close interbreeding, viz, weaken- 
ing of the stock, and, finally, extinction, may follow. 

Abel, in discussing the whole subject of extinction 
(1904.1, pp. 739-748), favors the hypothesis of the 



reduction of variabihty as one of the causes of extinc- 
tion. He observes: 

Even though it wiU never be possible to unravel the ulti- 
mate causes of the extinction of species, it remains certain 
that not only external factors but the internal organization as 
well are to be considered in seeking the ultimate explanation. 
In most cases probably an excessive, one-sided specialization 
in combination with a reduction of variability has led to extinc- 
tion. * * * Much difiiculty has arisen from the confusion 
of the conception of senility of certain types with that of pre- 
destined duration or of the restriction of vital energy. Because 
certain series of extinct animals show signs of degeneration in 
the last stages of their history it was assumed [for example, by 
Brocchi] that every species, genus, family, etc., has a prescribed 
limit of existence, at the end of which the species decays and 
dies like the individual. The opponents of the hypothesis of 
progressive reduction of variabihty pointed to the existence 
of persistent types from the Cambrian to the present time. 
Whenever we keep in mind the conception of progressive reduc- 
tion of variability such facts can be interpreted in a new light. 
Then it is seen that one-sided specialization and associated 
with it the reduction of variability cause a weakening of the 
entire constitution and invite extinction, while, on the other 
hand, conservative, persistent types may live through long 
periods of time. Furthermore, a review of the past shows that 
rapidly and richly varying groups die out sooner than slowly 
developing series. 

Darwin, Haeckel, and Weismann are upholders of the theory 
of unlimited variability. Haeckel admits that groups that are 
becoming extinct produce no new varieties. As a matter of 
fact, a large number of highly specialized forms have died out 
as the result of their inability to vary sufficiently, and as 
Wallace [1889.1] pointed out, the possibility of successful 
adaptation stands in direct relation to the number of favorable 
variations. However, the number of variations declines as 
specialization progresses, and it is this restriction of the limits 
of variation which makes us better able to determine relation- 
ships of forms near the end of their phylogenetic evolution than 
in the early stages where variations and mutations are more 
numerous and the species, which arise in an explosive manner, 
differ greatly from one another. * * * But not only hyper- 
trophy of the entire body but hypertrophy of certain organs 
to a great extent also appears toward the end of phylogenetic 
series. Very often this may furnish cause for extinction. The 
question of excessive specialization at the end of a phylogenetic 
series leads to a consideration of the problem of degeneration 
or "paracme" [Haeckel, 1906.1, pp. 366, 383]. 

SPECIAL FEATURES OF THE EXTINCTION OF THE 
TITANOTHERES 

In considering the possible causes of the extinction 
of the titanotheres and comparing those operating in 
the case of other mammals we may now sum up the 
observations which have been more or less fully com- 
mented upon in this chapter. 

Increase of individual size. — The general uniform 
increase in individual size in the titanotheres in both 
the Oligocene and the Eocene epochs indicates that 
all conditions of life were then remarkably favorable. 
This is apparently demonstrated by the fact that with 
one exception the animals in every line of ascent, 
both Eocene and Oligocene, steadily increased in size. 
The single exception is the diminutive, supposed 
aquatic species MetarTiinus jluviatilis of the basal upper 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



887 



Eocene. The practically universal increase betokens 
in general the long continuance of favorable condi- 
tions of life, which seem to have affected all lines 
of descent. 

Extinction at the "period of maximum growth and 
size. — Almost without exception extinction in each of 
the lines of Eocene and Oligocene titanotheres was 
not preceded by dwarfing but occurred at the period 
of greatest prosperity and after the completion of a 
long period of unchecked development. Large size 
in itself is not a cause of extinction. 

No numerical increase apparent. — Although our data 
are*far from sufficiently complete to warrant a positive 
statement, we may say that there is at present no evi- 
dence of a marked general increase of numbers among 
the titanotheres such as we observe among other 
mammalian phyla — the oreodonts and the horses in 
certain beds, for example. 

Adequate brain capacity. — Although the brain capac- 
ity of the titanotheres is not so large relatively as that 
of some existing mammals of similar gize, the brain 
steadily increases in size as we pass from the Eocene 
to the Oligocene forms, and there is not sufiicient 
ground to consider lack of intelligence as a cause of the 
sudden extinction of these animals. 

Molar tooth structure in relation to longevity. — The 
direct correlation between the longevity and the fer- 
tility of the quadrupeds with hypsodontism (elonga- 
tion of the crowns of the grinding teeth), enabling 
animals to live a great number of years, has been 
pointed out. Horses, with their long-crowned teeth, 
live for 25 years and, foaling every year, would produce 
22 young; elephants, with their long-crowned teeth, 
live 90 years and produce several pairs of young. In 
contrast, such a titanothere as the Eocene Palaeosyops, 
with its short-crowned teeth, would live for a compar- 
atively short period and would produce comparatively 
few yoimg. Theoretically this principle might be one 
of the means of explaining the early dying out of the 
broad-skulled genus Palaeosyops, with its short-crowned 
molar teeth, were it not for the contradictory fact that 
Manteoceras, also with short-crowned molars, survived, 
whereas Telmatherium, with relatively long-crowned 
teeth, apparently became extinct. 

Inadaptation of the grinding teeth. — We have shown 
that the upper Oligocene titanotheres were apparently 
making an effort to evolve a hypsodont tooth pattern 
by the elongation of the outer side (ectoloph) of the 
superior grinding teeth, but that apparently this effort 
was futile because of the complete separation of the 
protocone on the inner side of the tooth. Such a 
tooth is half hypsodont, half brachyodont, and is 
obviously inadaptive. 

Apparently the titanotheres reached a cul de sac of 
evolution in their grinding teeth, and this, on the 
whole, seems to be the single set of organs in which the 
titanotheres probably failed to meet the new condi- 



tions of food on the Great Plains. We conclude from 
the extinction of the large-toothed Menodus ingens that 
it was not the size of the teeth but the mechanical 
pattern that was unadapted to the new environment; 
no further mechanical progress or perfection was 
possible, hence the cul de sac. 

It is noteworthy that every family of mammals 
that experimented with teeth of this type (Anoplo- 
theriidae, Anthracotheriidae, Chalicotheriidae) became 
extinct sooner or later in Tertiary time. 

There is strong collateral evidence for Kovalevsky's 
theory that the influence of the teeth is a chief factor 
in extinction. The only type of mammal with buno- 
selenodont grinding teeth which survived through 
Miocene time was the chalicothere, probably because 
it became a forest-frequenting animal. 

A combination of other causes does not preclude the 
probability that the inadaptation of the teeth was the 
chief or leading cause of the numerical reduction and 
finally of the extinction of the titanotheres. 

CONCLXISIONS EEGAEDING THE THEORY OF NATTJRAI 
SELECTION OF DARWIN AND WALLACE 

Negative evidence as to minute variation. — The 
foregoing survey of the evolution of the titanotheres 
and of the causes of extinction among mammals in 
past and present time does not appear to support the 
distinctive feature of Darwin's theory that minute 
heritable variations (that is, mutations) have sufficient 
survival value to guide the course of evolution. The 
chief exceptions to this statement are seen in the very 
high survival value of immunity to disease and, con- 
versely, the high eliminating value of nonimmunity, 
which may possibly appear in some individuals as 
sudden variations or saltations. 

Degrees of survival value; utility of organs. — Paleon- 
tology and mammalian zoology now afford positive 
evidence that the structural and functional inadapta- 
tion of certain organs to environment has been a 
primary cause of extinction at all times but chiefly 
during stress of changes in climate, in physical con- 
ditions, and in life environment. We observe that 
certain combinations of mechanical structure and 
function have finally proved fatal to all mammals 
possessing them. Thus we are able to state positively 
that certain single organs of the bony or dental 
mechanism have had distinct survival or elimination 
value. 

Combination of organs and adaptability. — There is 
evidence that in past and present time those animals 
tend to survive which present the highest adaptive 
combination of favorable characters, of fully formed 
organs, the highest adaptability in structure or in 
habit, the highest potentiality of further evolution 
j toward new favorable types of habit and of structure. 

Survival value of single organs. — There is evidence 
I that in phyletic, generic, and family selection, not 



TITAN OTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



only such general adaptive or inadaptive combinations 
occur as those above cited, but single adaptive or 
inadaptive organs, such as the brain, the limbs, the 
feet, the teeth, the horns, have in course of time been 
contributory causes of survival or of elimination, 
partly because of direct adaptation or inadaptation, 
partly because of indirect inadaptation to changes of 
environment and of life environment. 

Positive evidence. — Simultaneously over large areas 
of the world extreme specialization, the development 
of certain dominant characters, has been followed by 
extinction: large-brained have replaced small-brained 
phyla; certain types of limb and foot structure have 
proved fatal to their possessors. This is direct con- 
firmation of the broad original form of the Darwin- 
Wallace hypothesis so far as it applies to individuals 
as a whole and to the groups of individuals known as 
varieties, races, species. 

Negative evidence. — This evidence, however, does 
not touch the queston of the survival value of "minute 
variations in structure, be they ever so slight," to use 
the Darwin- Wallace language. It does not appear 
that paleontology or mammalian zoology can give 
the final answer to this special feature of the Darwin- 
Wallace hypothesis. It is, however, here pointed out 
under the heading "Modes and factors of evolution" 
(sec. 1 of this chapter) that certain fluctuations of 
proportion may have distinct survival value and tend 
to be accumulated through natural selection. 

Minute hiocharacters and survival. — It is impossible 
to determine from paleontology how early in the rise 
of a new character, such as the minute rectigradations 
on the grinding teeth, there may be a sufficient sur- 
vival value, or utihty, to bring the character under the 
action of selection; opinions differ on this point; we 
can neither prove nor disprove the survival value of 
certain minute rectigradations. Our own opinion, 
incapable as yet of demonstration, is that natural 
selection may influence the transformation as soon as 
a number of these biocharacters (either rectigradations 
or aUometrons) reach such a stage of development as 
the "mutation of Waagen." It is not impossible that 
there may be natural selection in these mutation stages- 
Our opinion is more positive against the hypothesis 
that selection is the cause of these rectigradations. 

General conclusions. — Our general conclusions as to 
the part natural selection plays on minute variations, 
mutations, and fluctuations are negative. It is 
reasonably certain that under Darwin's general law of 
selection each strain of individual descent is improved 
as a whole, but it is still uncertain, largely hypothet- 
ical, undemonstrated experimentally, how far special 
selection explains the adaptive evolution of minute 
separate organs or parts, such as the incipient cusps on 
the teeth (rectigradations), the incipient rudiments 
of the horn (rectigradations), the countless minor 
changes of form and proportion (aUometrons). As 



shown in section 1 of this chapter, these biocharacters 
seem to be largely orthogenetic in origin and to arise 
quite independently of natural selection. The con- 
clusion we have reached is that new rectigradational 
and proportional characters may become a cause of 
survival or extinction as soon as they reach a stage of 
survival value — as soon as they are important enough 
to cause the life or the death of the animal in compe- 
tition with other characters. 

SUMMARY OF CONCLUSIONS 

The present aspects of the evidence appear to be as 
follows : 

Intervariation selection: There is no evidence that 
minute individual variations, except immunity to 
disease, have been a cause of the origin or evolution of 
new characters. 

Interfluctuation selection : There is considerable evi- 
dence from both paleontology and zoology that the 
selection of fluctuations of proportion may have been 
a potent factor in evolution. 

Interrectigradation selection: There is no reason to 
believe that minute rectigradations, in their most 
rudimentary stages, have sufficient survival value in 
competition with other biocharacters to affect the life 
of an organism. 

Interorgan selection: It is demonstrated that single 
structures, organs, or functions can certainly be the 
causes of survival or extinction, especially under con- 
ditions of stress of environment — biologic or physical. 

Interindividual selection: The chief action of selec- 
tion between individuals seems to be to standardize — 
to keep every individual as a whole up to or above the 
average of adaptability of the group to which it be- 
longs. Individuals that combine the largest number 
of favorable, adaptive, and adaptable characters are 
constantly being selected. In this process organic or 
coincident selection plays a large part. 

Intergroup selection: By a study of local adaptive 
radiation of the titanotheres we have discovered many 
separate lines of descent in which similar biocharacters 
were evolving at different rates. There has been more 
or less competition between these groups, varieties, 
races, and species. Here we observe especially the 
survival of the unspecialized forms and the extinction 
of the extremely specialized forms. 

Intergeneric, subfamily, family, and superfamily 
selection : The principle just stated applies to selection 
in relation to the survival or elimination of subfamily, 
family, and superfamily divisions. 

To sum up: Natural selection appears to be oper- 
ating incessantly and positively to preserve and accu- 
mulate characters that have survival value and to 
eliminate all other characters. Numerical reduction 
within a group and even extinction may be due to a 
single eliminating cause or to a combination of 
eliminating causes. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHEEES 



BIBLIOGRAPHY OF LITERATURE RELATING TO THE 
EXTINCTION OF FAUNAS CITED IN SECTION 3 

[A'few entries in this bibliographj', marked with an asterislr ('), are Incomplete 
in title and reference because after diligent search they have not been found in 
American libraries.] 

Abel, Othenio. 

1904.1. tjber das Aussterben der Arten: Cong. g^ol. 
internat., 9' sess., Vienna, 1903, Compt. rend., 
pp. 739-748, 1904. 
Adams, Andrew Leith. 

1877.1. Monograph on the British fossil elephants: 
Paleont. Soc, vol. 31, pp. 1-68, pis. 1-5, 
February, 1877; vol. 33, pp. 69-146, pis. 6-15, 
May, 1879; vol. 35, pp. 147-265, pis. 16-28, 
May, 1881. 
Adams, Charles C. 

1901.1. Base-leveling and its faunal significance, with 
illustrations from southeastern United States: 
Am. Naturalist, vol. 35, No. 418, pp. 839-851, 
October, 1901. 

1902.1. Southeastern United States as a center of geo- 
graphic distribution of flora and fauna: Marine 
Biol. Lab. Woods Hole Biol. Bull., vol. 3, 
pp. 115-131, 1902. 

1904.1. On the analogy between the departure from opti- 
mum vital conditions and departure from 
geographic life centers: Science, new ser., 
vol. 19, No. 475, pp. 210-211, Feb. 5, 1904. 

1905.1. The postglacial dispersal of the North American 
biota: Marine Biol. Lab. Woods Hole Biol. 
Bull., vol. 9, pp. 53-71, 1905. 
Aflalo, F. G. 

*1. The beasts that perish: St. James Gazette. 
Allen, Joel Asaph. 

1876.1. The American bisons, living and extinct: Harvard 
Coll. Mus. Comp. Zoology Mem., vol. 4, 
No. 10, pp. v-ix, 1-246, pis. 1-12, 1 map. 

1906.1. Mammals from the States of Sinaloa and Jalisco, 
Mexico, collected by J. H. Batty during 1904 
and 1905: Am. Mus. Nat. Hist. Bull., vol. 22, 
pp. 191-262, July 25, 1906. 
Andrews, C. W. 

1903.1. Some suggestions on extinction: Geol. Mag., new 
ser., dec. 4, vol. 10, No. 463, pp. 1-2, January, 
1903. 
Balfghe, Isaac Bayley (editor). See Schimper, A. F. W. 
Bangs, Outram. 

1895.1. The present standing of the Florida manatee, 
Trichechus latirostris (Harlan), in the Indian 
River waters: Am. Naturalist, vol. 29, No. 345, 
pp. 783-787, September, 1895. 
Bate, Dorothea M. A. 

1905.1. Four and a half months in Crete in search of 
Pleistocene mammalian remains: Geol. Mag., 
new ser., dec. 5, vol. 2, pp. 193-202, pis. 9-10, 
May 5, 1905. 
Bateson, William. 

1894.1. Materials for the study of variation, treated with 
special regard to discontinuity in the origin of 
species, 598 pp., 209 figs., London, Macmillan, 
1894. 
Benham, W. B. 

1906.1. Carnivorous habits of the New Zealand kea 
parrot: Nature, vol. 73, No. 1902, p. 559, 
Apr. 12, 1906. 

1907.1. Notes on the flesh-eating propensity of the kea 
(Nestor notabilis): New Zealand Inst. Trans, 
and Proc, vol. 39 (vol. 22, new ser.), pp. 71-89, 
June, 1907. 



Blanford, W. T. 

1888.1. The fauna of British India, including Ceylon and 
Burma: Mammalia, 617 pp., 196 figs., London, 
Berhn, Calcutta, Bombay, 1888-1891. 
Bosshard, H. (translator). 

1903.1. See Rosa, Daniel. 
Bradford, J. R., and Plimmbr, H. G. 

1899.1. A preliminary note on the morphology and distri- 
bution of the organism found in the tsetse-fly 
disease: Roy. Soc. London Proc, vol. 45, No. 
418, pp. 274-275, Aug. 31, 1899. 

1902.1. The Trypanosoma brucei, the organism found in 
nagana, or tsetse-fly disease: Quart. Jour. 
Micr. Sci., vol. 45, new ser., pp. 449-471, 1902. 
Brooks, Harlow. 

1902.1. Cause of death of five orangs and one chimpanzee 
by invasion of Balantidium coli, a ciliated in- 
fusorian: New York Zool. Soc. Sixth Ann. 
Rept., p. 101, 1902. 
Bruce, Col. D. 

1905.1. The advance in our knowledge of the causation 
and methods of prevention of stock diseases in 
South Africa during the last ten years, address 
before British Association for the Advancement 
of Science: Nature, vol. 72, no. 1872, pp. 496- 
503, Sept. 14, 1905; Science, new ser., vol. 22, 
No. 558, pp. 289-299, Sept. 8, 1905; No. 559, 
pp. 327-333, Sept. 15, 1905. 

BtJCHNER, EUGBN. 

1895.1. Das aUmahliohe Aussterben des Wisents [Bison 
bonasus (Linn.)] im Forste von Bjelowjesha: 
Acad. imp. sci. St.-P6tersbourg M6m., vol. 3, 
No. 2, pp. 1-30, 1895. 
BuFFON, Count Georges Louis Leclerc de, and others. 

1749.1. Histoire naturelle, g^nerale et particulifere, avec la 
description du Cabinet du Roi, 1st ed., 44 vols., 
illus., Paris, 1749-1804. (Edition referred to 
by Flourens in his sketch of Buffon's work.) 
Calvert, Philip P. 

1900.1. The means of defense of animals: 1, The struggle 
for existence: Sci. Am. Suppl., vol. 49, No. 1272, 
pp. 20396-20397, May 19, 1900; 2, The pro- 
tection of the young: Idem, No. 1276, pp. 
20456-20457, June 16, 1900; 3, The protection 
of the food supply: Idem, No. 1277, pp. 20466- 
20467, June 23, 1900; 4, Protection against 
living animals: Idem, vol. 60, No. 1280, pp. 
20516-20517, July 14, 1900; 5, Protection 
against climatic changes: Idem, No. 1281, pp. 
20535-20537, July 21, 1900. 
Cakruthers, W. 

1872.1. See Daintree, R. 
Cattell, J. McK. 

1897.1. The alleged extinction of lines of descent: Science, 
new ser., vol. 6, No. 148, pp. 668-669, 1897. 
Chesnut, Victor I\. 

1898.1. Thirty poisonous plants of the United States: 

U. S. Dept. Agr. Farmers Bull. 86, 32 pp., 1898. 

1899.1. Preliminary catalogue of plants poisonous to 

stock: U. S. Dept. Agr. Bur. Animal Industry 

Fifteenth Ann. Rept., 1898, pp. 387-420, 1899. 

1901.1. Some poisonous plants of the northern stock 

ranges: U. S. Dept. Agr. Yearbook, 1900, pp. 

305-324, pis. 32-34, 1901. 

1901.2 (and Wilcox, E. V.). The stock-poisoning plants of 

Montana, a preliminary report: U. S. Dept. 

Agr. Div. Botany Bull. 26, 150 pp., 1901. 



890 



TIT.iNOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



Chbsnut, Victor K. — Continued. 

1902.1. Problems in the chemistry and toxicology of plant 
substances: Science, new ser., vol. 15, No. 391, 
pp. 1016-102S, June 27, 1902. 
Clements, Fbederic Edward. 

1905.1. Research methods in ecology, 334 pp., 85 figs., 
Lincoln, Nebr., Universitj'- Publishing Co., 1905. 
CocKRELL, Theodore Dbu Alison. 

1907.1. A fossil tsetse fly in Colorado: Nature, vol. 76, 

No. 1973, p. 414, Aug. 22, 1907. 

1907.2. Some Old World types of insects in the Miocene of 

Colorado: Science, new ser., vol. 26, No. 666, 
pp. 446-447, Oct. 4, 1907. 

1908.1. Fossil insects from Florissant, Colo.: Am. Mus. 

Nat. Hist. BuU., vol. 24, pp. 59-69, pi. 5, Feb. 
7, 1908. 

1908.2. Florissant, a Miocene Pompeii: Pop. Sci. Monthly, 

vol. 73, No. 2, pp. 112-126, Aug., 1908. 
1909.1. Another fossil tsetse flj' [GZossiraa osborrei]: Nature, 

vol. 80, No. 2057, p. 128, Apr., 1909. 
1910.1. The Miocene trees of the Rocky Mountains: Am. 
Naturalist, vol. 44, No. 517, pp. 31-47, Jan., 
1910. 
1916.1. The third fossD tsetse fly [Glossina velerna]: Nature, 
vol. 9S, p. 70, Sept. 28, 1916. 
Cope, Edward Drinker. 

1896.1. The primary factors of organic evolution, 547 pp. 
120 figs., Chicago and London, Open Court 
Publishing Co., 1896. 
CouES, Elliott (editor). See Henry, Alexander. 
Crampton, C. B. 

1902.1. A suggestion on extinction: Roy. Phys. Soc. for 
Promotion of Zoology and Other Branches of 
Nat. Hist. Proc, vol. 14 (1900-1901), p. 461, 
1902. 
Crawford, Albert C. 

1908.1. Barium, a cause of the loco-weed disease: U. S. 
Dept. Agr. Bur. Plant Industry Bull. 129, 87 
pp., Aug. 22, 1908. 
Crochard, Lieutenant. 

1885.1. Nouveau silex tallies du Sahara Alg^rien: Mat^ri- 

aux pour I'histoire primitive et naturelle de 

I'homme, 19th year, 3d ser., vol. 2, p. 143, 

March, 1885. 

CuviER, Baron Georges Leopold Chretien Frederic Dago- 

bert. 

1825.1. Discours sur les revolutions de la surface du globe, 
et sur les changemens qu'eUes ont produits 
dans le regne animal, Recherches sur les 
ossemens fossiles, 3d ed., vol. 1, pp. 1-167, 
1825. 
Gumming, Gordon. 

1855.1. The lion hunter in South Africa, five years of a 
hunter's life in the far interior of South Africa, 
with anecdotes of the chase and notices of the 
native tribes, 2 vols., London, John Murray, 
1855. 
CuRRT, Capt. J. J. 

1902.1. Report on a parasitic disease in horses, mules, and 
carabao in the PhiUppine Islands [from a 
report made to the Surgeon-General of the 
Army, Dec. 17, 1901]: Am. Medicine, vol. 3, 
No. 13, pp. 512-513, Mar. 29, 1902. 
Daintree, R. 

1872.1 (with Etheridge, R., and Carruthers, W.). Notes 
on the geology of the colony of Queensland, 
with an appendix containing descriptions of 
the fossils: Geol. Soc. London Quart. Jour., 
vol. 28, pp. 271-360, pis. 10-27, figs. 1-19, 
1872. 



Dall, William Healet. 

1902.1. On the preservation of the marine animals of the 
northwest coast: Smithsonian Inst. Ann. Rept. 
for 1901, pp. 683-088, 1902. 
Dahwin, Charles Robert. 

1839.1. Journal and remarks: Narrative of the surveying 
voyages of His Majesty's Ships Adventure and 
Beagle, between the years 1826 and 1836, 
describing their examination of the southern 
shores of South America and the Beagle's 
circumnavigation of the globe, vol. 3, xiv, 615 
pp., maps, illus., London, Henry Colburn, 1839. 
1858.1 (and Wallace, Alfred Russel). On the tendency of 
species to form varieties and on the perpetua- 
tion of varieties and species by natural means 
of selection: Linnean Soc. London Jour. Proc. 
(Zoology), vol. 3, pp. 45-62, Aug., 1858. 
1859.1. On the origin of species by means of natural 
selection, or the preservation of favored races 
in the struggle for life, 602 pp., London, John 
Murray, 1859. 
1909.1. Journal of researches into the natural history and 
geology of the countries visited during the 
voyage of H. M. S. Beagle round the world, 512 
pp., D. Appleton & Co., 1909. 
Dbshayes, G. p. 

1832.1. See Lyell, Charles, 1830.1. 
Doderlein, L. 

1888.1. Phylogenetische Betrachtungen: Biol. Centralbl., 
vol. 7, pp. 394-402, 1888. 
DoLLo, Louis. 

1893.1. Les lois de revolution: Soc. beige geol. Bull., vol. 
7, pp. 164-166, 1893. 
D'Orbignt, Alcide Dbssalines. 

1835.1. Voyage dans I'Amerique meridionale, Paris and 
Strassbourg, Bertrand & Levrault, 1835-1847. 
(Quoted by Darwin, 1909.1, and Howorth, 
1887.1.) 
DijsiNG, Carl. 

1884.1. Die Regulierung des Geschlectsverhaltnisses bei 
der Vermehrung der Menschen, Tiere und 
Pflanzen: Medicin.-naturwiss.-Gesell. Jena 
Zeitschr. f. Naturwiss., Band 17 (neue Folge, 
Band 10), pp. 593-594, 1884. 
Etheridge, R. 

1872.1. See Daintree, R. 
Fisher, William R. (translator). See Schimper, A. F. W. 
Fleming, George. 

1871.1. Animal plagues, their history, nature, and preven- 
tion, vol. 1, London, Chapman & Hall, 1871; 
vol. 2, London, Bailliere, 1882. 
Flourens, p. 

1844.1. Buffon, Histoire de ses travaux et de ses iddes, 
367 pp., Paris, 1844. (See also Buffon, 1749.1.) 
FoA, Edouard. 

1899.1. After big game in central Africa, records of a sports- 
man from August, 1894, to November, 1897, 
when crossing the Dark Continent from the 
mouth of the Zambesi to the French Congo, 
330 pp., illus., London, Adams & Chas. Black, 
1899. 
Frankland, W. Ashbt. 

1902.1. See Salmon, D. E. 
Frech, Fritz. 

1906.1. Uber die Griinde des Aussterbens der vorzeitlichen 
Tierwelt: Archiv f. Rassen- und Gesell.-Biol. 
einsohl. Rassen- und GeseU. Hygiene, Jahrg. 3, 
Heft 4, pp. 469-498, July-August, 1906. 



CAUSES OP THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



891 



Gasman, S. 

1884.1. Remarks on the extinction of the fossil horse in 
America: Boston Soc. Nat. Hist. Proc, vol. 22, 
pp. 252-253, December, 1884. 
Gibbons, A. St. H. 

1898.1. Exploration and hunting in central Africa, 1895-96, 
400 pp., illus., London, Methuen & Co., 1898. 
Ghatacap, Louis P. 

1896.1. Fossils and fossilization: Am. Naturalist, vol. 30, 
pp. 902-912, 993-1003, pis. 23-27, 1896; vol. 31 , 
pp. 16-33, 191-199, 285-293, 1897. 
Gregory, John Waltkr. 

1896.1. The Great Rift Valley, being the narrative of a 
journey to Mount Kenya and Lake Baringo, 
with some account of the geolog}', natural his- 
tory, anthropology, and future prospects of 
British East Africa, 422 pp., 23 figs., 1 map, 
London, John Murray, 1896. 
Gregory, William King. 

1910.1. The orders of mammals: Am. Mus. Nat. Hist. 
Bull., vol. 27, pp. 3-524. 32 figs., February 1910. 
Groom, Percy (editor). See Schimper, A. F. W. 
Habenicht, Hermann. 

1890.1. Die Todesursache diluvialer Saugethiere: Natur- 
wiss. Wochenschr. Berlin, Band 1, pp. 448-449, 
1890. 
Haeckbl, Ernst. 

1906.1. Prinzipien der generellen Morphologie der Organis- 
men, wortlicher Abdruck eines Teiles der 1866 
erschienenen generellen Morphologie (Allge- 
meine Grundztige der organischen Formen- 
Wissenschaft mechanisch begriindet durch die 
von Charles Darwin reformierte Deszendenz- 
Theorie), 447 pp., Berhn, Georg Reimer, 1906. 
Hann, Julius. 

1903.1. Handbook of climatology (translation by Robt. de 
Courcy Ward), 437 pp.. New York and London, 
Macmillan Co., 1903. 
Harting, J. E. 

1892.1. The fox in Australia: Zoologist, vol. 16, 3d ser , 

No. 185, pp. 189-190, May, 1892. 
1896.1. Present status of the European bison: Zoologist, 
vol. 20, 3d ser., No. 238, p. 377, Oct. 4, 1896. 
Hassall, Albert. 

1902.1 and 1902.2. See Salmon, D. E. 
Hedley, Charles. 

1894.1. The faunal regions of Australia: Australian Assoc. 
Adv. Sci. Rept. Fifth Meeting (Adelaide, 1893), 
vol. 5, pp. 444-451, 1894. 
Heller, Edmund. 

1914.1. See Roosevelt, Theodore. 
Henry, Alexander. 

1897.1. The manuscript journals of Alexander Henry 
* * * and of David Thompson * * * 
1799-1814, edited by Elliott Coues, 3 vols., 
Harper, 1897. 
HoRNADAY, William T. 

1889.1. The extermination of the American bison: U. S. 
Nat. Mus. Rept. for 1887, pp. 367-548, pis. 
1-21, 1 map, 1889. 
1904.1. The American natural history, 449 pp., illus., Chas. 
Scribner's Sons 1904. 
Howorth, Henry H. 

1887.1. The mammoth and the flood, an attempt to con- 
front the theory of uniformity with the facts 
of recent geology, 464 pp., London, Sampson 
Low, Marston, Searle & Rivington, 1887. 
1896.1. The cause of the mammoth's extinction: Nat. Sci., 
vol. 9, pp. 142-143, August, 1896. 



Hudson, W. H. 

1892.1. The naturalist in La Plata, 383 pp., illus., London, 

Chapman & Hall, 1892. 
1893.1. Idle days in Patagonia, 256 pp., illus., New York, 
D. Appleton & Co., January, 1893. 
Hunt, Reid. 

1908.1. The loco-weed disease: Science, new ser., vol. 28, 
No. 721, pp. 570-571, Oct. 23, 1908. 

HUTCHEON, D. 

1906.1. Poisoning of horses by Orniihogalum thysoides or 
"chinkerinchee": Agr. Jour. Cape of Good 
Hope, vol. 28, No. 2, pp. 150, 165-172, Febru- 
ary, 1906. 
Johnston, Sir H. H. 

1910.1. The Roosevelts in Africa (review): Nature, vol. 
85, No. 2142, pp. 77-80, Nov. 17, 1910. 
Keybs, Charles R. 

1898.1. The genetic classification of geological phenomena: 
Jour. Geology, vol. 6, pp. 809-815, 1898. 
KoKEN, Ernst. 

1892.1. Die Geschichte des Saugethierstammes nach den 
Entdeokungen und Arbeiten der letzten Jahre: 
Naturwiss. Rundschau, 7. Jahrg., No. 14, 
pp. 16&-174, Apr. 2, 1892; No. 15, pp. 185-188, 
Apr. 9, 1892; No. 19, pp. 233-240, May 7, 1892. 
1893.1. Die Vorwelt und ihre Entwicklungsgeschichte, 
654 pp., 2 pis., 117 figs., Leipzig, 1893. 
Kovalevsky, Woldemar. 

1873.1. Sur V Anchitherium aurelianense Cuv. et sur 
I'histoire pal6ontologique des chevaux: Acad, 
imp. sci. St.-P6tersbourg M6m., 7° ser., vol. 20, 
Nos. 5 et seq., pp. 1-73, pis. 1-3, St. Petersburg, 
1873. 
1876.1. Monographie der Gattung Anthracotherium Cuv. 
und Versuch einer natiirliehen Classification 
der fossilen Hufthiere: Palaeontographica, 
Band 22, neue Folge 2, Lief. 3-5, pp. i-iv, 
133-347, pis. 7-17, 1876. 
Langkavel, B. 

1896.1. Rattenplagen auf Inseln: Zool. Garten, Jahrg. 37, 
p. 107, 1896. 
Lankester, E. Ray. 

1902.1. On Okapia, a new genus of Giraffidae, from central 
Africa: Zool. Soc. London Trans., vol. 16, pt. 6, 
pp. 279-307, pis. 30-32, August, 1902. 
Lapparent, a. de. 

*1892.1. Les anciens glaciers: Correspondant, Paris, 1892. 
(Referred to by Marcellin Boule in L'Anthro- 
pologie, Paris, 1893, pp. 217-220.) 
Lartet, finOUAHD. 

1868.1. De quelques cas de progression organique v^ri- 
fiables dans la succession des temps g^ologiques 
sur des mammiferes de meme famiUe et de 
meme genre: Compt. Rend., vol. 66, pp. 1119- 
1122. (Separate dated June 1, 1868.) 
Lbndenpeld, R. von. 

*1. Bilder aus dem australischen Urwald. 
Le Souef, Dudley W. H. 

1909.1. See Lucas, A. H. S. 
Lewis, Timothy Richards. 

1888.1. Flagellated organisms in the blood of animals, 
physiological and pathological researches of the 
late Timothy Richards Lewis, London, 1888. 
LoEB, Leo. 

1905.1. Immunity and adaptation: Marine Biol. Lab. 
Woods Hole Biol. Bull., vol. 9, pp. 141-151, 
1905. 



892 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



LooMis, F. B. 

1905.1. Momentum in variation: Am. Naturalist, vol. 39, 
No. 467, pp. 839-843, figs. 1, 2, November, 
1905. 
Losing, J. Alden. 

1902.1. Notes on the destruction of animal life in Alaska: 
New York Zool. Soc. Sixth Ann. Kept., pp. 
141-144, April, 1902. 

LOUNSBTTRT, C. P. 

1906.1. Locust birds and locust poison: Agr. Jour. Cape 

of Good Hope, 1906, pp. 364-366. 

1906.2. The Brazil fruit-fly parasites: Agr. Jour. Cape of 

Good Hope, 1906, pp. 538-540. 

1906.3. Ticks and African coast fever: Agr. Jour. Cape of 

Good Hope, 1906, pp. 634-654. 
Lucas, A. H. S. 

1909.1 (and Le Souef, Dudley W. H.). The animals of 
Australia- — Mammals, reptiles, and amphibi- 
ans, 327 pp., illus., London, Whitcombe & 
Tombs, Ltd., 1909. 
Lucas, Frederic A. 

1901.1. Animals of the past, 258 pp., illus., McClure, 
Phillips & Co., 1901. (Reprinted with one 
new chapter and with prefactory note by 
W. D. Matthew as Am. Mus. Nat. Hist. 
Handbook ser., No. 4, 266 pp., illus., June 5, 
1913.) 
Ltdekker, Richard. 

1903.1. Mostly mammals: Zoological essays, 383 pp., illus. 

London, Hutchinson & Co., 1903. 
1904.1. The supply of valuable furs: Nature, vol. 71, No. 
1831, pp. 115-117, Dec. 1, 1904. 
Ltell, Charles. 

1830.1. Principles of geology, being an attempt to explain 
the former changes of the earth's surface by 
reference to causes now in operation, London, 
John Murray, vol. 1, 511 pp., 1830; vol. 2, 330 
pp., 1832; vol. 3, 398 pp. plus 109 pp. of ap- 
pendices, glossary, and index, 1833. (Vol. 2 
contains "Tables of fossil shells," by G. P. 
Deshayes, as pp. 1-60 of the Appendix.) 
1863.1. The geological evidences of the antiquity of man, 
with remarks on theories of the origin of 
species by variation, 520 pp., 58 figs., pis. 1, 2, 
London, John Murray, 1863. 
1872.1. Principles of geology, 11th ed.. New York, 
Appleton & Co., vol. 1, 1877; vol. 2, 1872. 
MacCullum, G. a. 

1917.1. A new species of trematode {Cladorchis gigas) 
parasitic in elephants: Am. Mus. Nat. Hist. 
BuU., vol. 37, pp. 865-871, Dec. 27, 1917. 
Maddren, Alfred G. 

1905.1. Smithsonian exploration in Alaska in 1904, in 
search of mammoth and other fossil remains: 
Smithsonian Misc. Coll., vol. 49 (Pub. 1584), 
117 pp., 7 pis., 3 figs.., 1905. 
Makrinee, George R. 

1907.1. Notes on the natural history of the kea, with 
special reference to its reputed sheep-kiUing 
propensities: New Zealand Inst. Trans, and 
Proc, vol. 39 (new ser., vol. 22), pp. 271-305, 
June, 1907. 
1908.1. Additional notes on the kea: New Zealand Inst. 
Trans, and Proc, vol. 46, pp. 534^537, June, 
1908. 
Marsh, George P. 

1864.1. Man and nature, Charles Scribner's Sons, 1864. 
(Republished in 1874 under the title " The earth 
as modified by human action." Reprint, 1907, 
629 pp.) 



Marsh, Othniel C. 

1884.1. Dinocerata, a monograph of an extinct order of 
gigantic mammals: U. S. Geol. Survey Mon. 
10, 237 pp., 56 pis., 1884. 
Merriam, C. Hart. 

1892.1. The geographic distribution of life in North 
America with special reference to the Mam- 
malia: Biol. Soc. Washington Proc, vol. 7, pp. 
1-64, 1 map, Apr., 1892. 
1894.1. Laws of temperature control of the geographic 
distribution of terrestrial animals and plants: 
Nat. Geog. Mag., vol. 6, pp. 229-238, pis. 
12-14, maps, Dec. 29, 1894. 
1909.1. Report of the Chief of the Bureau of Biological 
Survey: U. S. Dept. Agr. Ann. Repts. for 
1908, pp. 571-590, 1909. 
Merrill, E. T. 

*1900.1. Sports afield, March, 1900. 
MiLLAis, John Guille. 

1895.1. A breath from the veldt, 236 pp., 25 pis., London, 
1895. 
Miller, Gerrit S., jr. 

1896.1. The fate of a European bison herd: Science, new 
ser., vol. 4, No. 99, pp. 744-745, Nov. 20, 1896. 
Moffat, C. B. 

1907.1. The problems of an island fauna: Irish Naturalist, 
vol. 16, pp. 133-146, April, 1907; p. 207, June, 
1907. 
Morgan, T. H. 

1903.1. Evolution and adaptation, 470 pp., 7 figs., New 
York and London, Macmillan Co., 1903. 
Morris, Charles. 

1895.1. The extinction of species: Acad. Nat. Sci. Phila- 
delphia Proc, vol. 47 (1895), pp. 253-263, 
1896. 
Morse, E. W. 

1887.1. What American zoologists have done for evolu- 
tion: Science, vol. 10, No. 236, pp. 73-76, 1887. 
Murphy, Robert Cushman. 

1917.1. Faunal conditions in South Georgia: Science, new 
ser., vol. 46, No. 1179, pp. 112-113, Aug. 3, 
1917. 
MusY, M. 

1898.1. Essai sur la chasse aux siecles passes et appauvris- 
sement de la faune fribourgeoise : Soc. fribour- 
geoise sci. nat. Bull., vol. 7, pp. 37-82, 1898. 
Neheing, A. 

1890.1. Schneestiirme als Todesursache diluvialer Sauge- 
thiere: Naturwiss. Wochenschr. Berlin, Band 
5, pp. 71-74, 516-519, 1890. 
Neumayr, M. 

1899.1. Die Stamme des Thierreiches, vol. 1, 603 pp., 192 
figs., Vienna and Prague, F. Tempsky, 1889. 
Newton, A. 

1887.1. The theory of evolution and the extinction of mod- 
ern faunas, address of the president of the 
Biological Section: British Assoc. Adv. Sci. 
(Manchester, 1887) Rept., pp. 727-733, 1888. 
Nuttall, George H. F. 

1901.1 (and Shipley, A. E.). Studies in relation to malaria: 
Jour. Hygiene, vol. 1, pp. 45-77, pis. 1, 2, 
1901; vol. 2, pp. 58-84, 1902. 
OsBORN, Henry Fairfield. 

1904.246. Preservation of the wild animals of North Amer- 
ica, address before the Boone and Crockett 
Club, Washington, Jan. 23, 1904, 27 pp.; Forest 
and Stream, Apr. 16, 1904, pp. 312-313. 



CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



893 



OsBORN, Henry Fairfield — Continued. 

1906.287. The causes of extinction of Mammalia; Am. 
Naturalist, vol. 40, No. 479, pp. 769-795, 
November, 1906; No. 480, pp. 829-859, Decem- 
ber, 1906. 

1910.346. The age of mammals in Europe, Asia, and North 
America, 635 pp., iUus., New York, MacmiUan 
Co., 1910. 
Owen, Sir Richard. 

1877.1. Researches on the fossil remains of the extinct 
mammals of Australia, with a notice of the 
extinct marsupials of England, vol. 1, text, xv, 
522 pp., 1 pi., iUus., vol. 2, atlas of 131 plates, 
London, J. Erxleben, 1877. 
Packard, Alpheus S. 

1886.1. Geological extinction and some of its apparent 
causes: Am. Naturalist, vol. 20, No. 1, pp. 29- 
40, January, 1886. 

1898.1. A half century of evolution, with special reference 
to the effects of geological changes on animal 
life: Am. Naturalist, vol. 32, No. 381, pp. 623- 
674, September, 1898; Am. Assoc. Adv. Sci. 
Proc. 47th meeting, 1898, pp. 311-356. 

1903.1. Report of the acting superintendent of the Yellow- 
stone National Park to the Secretary of the 
Interior. 
Palmer, T. S. 

1899.1. The danger of introducing noxious animals and 
birds: U. S. Dept. Agr. Yearbook for 1898, pp. 
87-110, 1899. 
PicHA, Jean. 

1896.1. Note sur les captures d'animaux sauvages: Soc. 
d'^tudes sci. Angers BuU., new ser., vol. 25, 
pp. 497-498, 1896. 
Plate, L. 

1904.1. Gibt es ein Gesetz der progressive Reduktion der 
Variabilitiit? Archiv f. Rassen- u. Gesell.- 
Biol., Jahrg. 1, Heft 5, pp. 641-655, September- 
October, 1904. 
Plimmer, H. G. 

1899.1 See Bradford, J. R. 

1902.1. See Bradford, J. R. 
PoMEL, Nicolas Auguste. 

1895.1. Les 616phants quaternaires, carte gfiologique de 
I'Alg^rie: PalSont. Mon., 67 pp., 16 pis., 
1895; Abstract in Compt. Rend., vol. 123, pp. 
975-976, 1896. 
Prichard, H. Hesketh. 

1902.1. Through the heart of Patagonia, 346 pp., illus., 
New York, D. Appleton & Co., 1902. 
Reid, Clement. 

1895.1. The origin of Megaceros marl: Irish Naturalist, 
vol. 4, pp. 131-132, 1895. 
Rosa, Daniel. 

1899.1. La riduzione progressiva deUa variabilita e suoi 
rapporti coll' estinzione e coll' origine delle 
specie, 135 pp., Torino, Carlo Claussen, 1899. 

1903.1. Die progressive Reduktion der Variabihtat und 
ihre Beziehungen zum Aussterben und zur 
Entstehung der Arten (translation by H. 
Bosshard), 106 pp., Jena, G. Fischer, 1903. 
Roosevelt, Theodore. 

1910.1. African game trails, an account of the African 

wanderings of an American hunter-naturalist, 
583 pp., illus., Charles Scribner's Sons, 1910. 

1910.2. See Johnston, Sir H. H. 
1912.1. See Stevenson-Hamilton, Maj. J. 

1914.1 (and HeUer, Edmund). Life histories of African 
game animals, 2 vols., iUus., 40 faunal maps, 
Charles Scribner's Sons, 1914. 



Salensky, W. 

1905.1. Uber die Hauptresultate der Erforschung des im 
Jahre 1901 am Ufer der Beresowka enteckten 
maenlichen Mammutkadavers: 16th Cong, in- 
ternat. zool. (Berne, 1904), Compt. rend, 
stances, pp. 67-86, 1905. 
Salmon, D. E. 

1901.1 (and Stiles, C. W.). Cattle ticks (Ixodoidea) of 
the United States: U. S. Dept. Agr. Bur. 
Animal Industry Seventeenth Ann. Rept., 
1900, pp. 380-491, pis. 74-98, figs. 47-238, 1902. 
1902.1 (with Stiles, C. W., and Hassall, Albert). Emer- 
gency report on surra [Salmon, Stiles], with a 
bibliography of surra and allied trypanoso- 
matic diseases [Hassall]: U. S. Dept. Agr. Bur. 
Animal Industry Bull. 42, pp. 11-152, 1902. 
1902.1 (with Stiles, C. W., Hassall, Albert, Frankland, 
W. A., and Taylor, Louise). Eleven miscel- 
laneous papers on animal parasites: U. S. Dept. 
Agr. Bur. Animal Industry Bull. 35, 61 pp., 
5 pis. (1 colored), 38 figs., 1902. 
Schillings, C. G. 

1906.1. With flashlight and rifle in equatorial East Africa 
(translated by Fred. Whyte), 814 pp., Hutch- 
inson, 1906. 
1907.1. In wildest Africa (translated by Fred. Whyte), 
2 vols., 724 pp., Hutchinson, 1907. 
SCHIMPER, A. F. W. 

1903.1. Plant geography upon a physiological basis 
(authorized EngUsh translation by Wm. R. 
Fisher, revised and edited by Percy Groom and 
Isaac Ba}'ley Balfour), 839 pp., 502 figs., 4 maps, 
Oxford, Clarendon Press, 1903. 
Seton, Ernest Thompson. 

1909.1. Life histories of northern animals, an account of 

the mammals of Manitoba, 2 vols., 1267 pp., 
illus.. New York, Chas. Scribner's Sons, 1909. 
Shipley, Arthur E. 

1901.2. See Nuttall, George H. F., 1901.1. 

1906.1. Insects as carriers of disease: Nature, vol. 73, 
No. 1888, pp. 235-238, Jan. 4, 1906. 
Smith, Theobald. 

1912.1. Note on infectious abortion in cattle: Science, 
new ser., vol. 36, No. 926, pp. 409-412, Sept. 27, 
1912. 
Spencer, Baldwin. 

1896.1. Through Larapinta land, a narrative of the Horn 
expedition to central Australia: Report on the 
work of the Horn scientific expedition to central 
AustraUa, pt. 1, pp. 1-199, 1896. 
Stephens, J. W. W. 

1904.1. Sleeping sickness: Nature, vol. 69, No. 1879, 
pp. 345-347, Feb. 11, 1904. 
Stevenson-Hamilton, Maj. J. 

1912.1. Animal fife in Africa, with a foreword by Theodore 
Roosevelt, 539 pp., IUus., London, Wm. 
Heinemann, 1912. 
Stiles, Charles Wardell. 
1901.1. See Salmon, D. E. 

1902.1. See Salmon, D. E. 

1902.2. See Salmon, D. E. 
Stirling, E. C. 

1899.1. Fossil remains of Lake Callabona, pt. 1, Descrip- 
tion of the bones of the manus and pes of 
Diprotodon australis Owen: Roy. Soc. South 
Australia Mem., vol. 1, pt. 1, pp. 1-40, pis. 
1-18, 1899. 



89i 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



SrOCKLET, C. W. 

1910.1. Rhinoceros under conditions of scant water: Field , 
Apr. 2, 1910. (Noted in Nature, vol. 83, No . 
2111, p. 19S, Apr. 14, 1910.) 
Stromer, Ernst. 

1905.1. Fossile Wirbeltier-Reste aus dem Uadi Faregh 
und Uadi Natrto in Sgypten: Senckenberg. 
naturforsch. Gesell. Abh., Band 29, Heft 2, 
pp. 14-132, pi. 20, 1905. 
Tate, Ralph. 

1889.1. On the influence of physiographic changes in the 
distribution of life in Australia: Australia Assoc . 
Adv. Sci. Rept. First Meeting (Sydney, 1888) , 
vol. 1, pp. 312-325, 1889. 
Taylor, Louise. 

1902.1. See Salmon, D. E., 1902.2. 
Taylor, Walter P. 

1917.1. The vertebrate zoologist and national efficiency: 
Science, new ser., vol. 46, No. 1180, pp. 123- 
127, Aug. 10, 1917. 
Thiselton-Dyer, W. T. 

1902.1. The sweetbrier as a goat exterminator: Nature, 
vol. 66, No. 1697, p. 31, May 8, 1902. 
Thompson, David. See Henry, Alexander, 1897.1. 
Tobkel, Otto. 

1876.1. Sur les traces les plus anciennes de I'existence de 
I'homme en SuSde: Cong, internat. anthropolo- 
gic et archfiologie pr^historique, 7= sess., 
Stockholm, 1874, Compt. rend., vol. 1, pp. 
16-17; vol. 2, pp. 861-876, 1876. 
Torre Y, Harry Beal. 

1902.1. An unusual occurrence of Dinoflagellaia on the 
California coast: Am. Naturalist, vol. 36, No. 
423, pp. 187-192, March, 1902. 

VOGES, O. 

1902.1. Das Mai de Caderas: Zeitschr. f. Hygiene u. 
Infectionskrankheiten, Band 39, pp. 323-371. 
Wallace, Alfred Russel. 

1858.1. See Darwin, Charles R. 

1870.1. Contributions to the theory of natural selection, a 
series of essays, 371 pp., London, Macmillan 
& Co., 1870. 

1876.1. The geographical distribution of animals, with a 
study of the relations of living and extinct 
faunas as elucidating the past changes of the 
earth's surface, vol. 1, 503 pp., maps, pis. 1-13; 
vol. 2, 607 pp., maps, pis. 14-20, London, 
Macmillan & Co., 1876. 

1881.1. Island life, or the phenomena and causes of insular 
faunas and floras, including a revision and 
attempted solution of the problem of geo- 
logical climates, 522 pp., maps. New York, 
Harper & Bros., 1881. 



Wallace, Alfred Russel — Continued. 

1889.1. Darwinism, an exposition of the theory of natural 
selection with some of its appUcations, 494 pp., 
frontispiece, 37 figs., London and New York, 
Macmillan & Co., 1889. 
Ward, Robert de Courcy (translator). See Hahn, Julius. 
Weber, Max. 

1896.1. Vorstudien uber das Hirngewicht der Saugethiere, 
Festsohr. f. Carl Gegenbaur, pp. 105-123, 
Leipzig, 1896. 
1904.1. Die Saugetiere, 866 pp., Jena, Gustav Fischer, 1904. 
Weisgerber, Doctor. 

1885.1. Les ages de pierre du Sahara central (reported 

by fimile Cartailhac) : Mat6riaux pour I'his- 

toire primitive et naturelle de I'homme, 19th 

year, 3d ser., vol. 2, pp. 124-126, March, 1885. 

Whewell, Rev. William. 

1837.1. History of the inductive sciences from the earliest 
to the present times, vol. 1, 437 pp.; vol. 2, 
534 pp.; vol. 3, 524 pp., London and Cambridge, 
1837. 
Whyte, Fred, (translator). See Schillings, C. G. 
Wilcox, E. V. 

1901.1. See Chesnut, Victor K., 1901.2. 

WOLDRICH, JOH. N. 

1882.1. Die diluvialen Faunen Mitteleuropas und eine 
heutige Sareptaner Steppenfauna in Niederos- 
terreich: Anthr. GeseU. Mitt., Band 11 (neue 
Folge, Band 1), pp. 183-190, 1882. 

WOODWORTH, J. B. 

1894.1. The relation between base-leveUng and organic 
evolution: Am. Geologist, vol. 14, pp. 209-235, 
1894. 
ZiTTEL, Charles. 

1876.1. Sur les silex taillfe trouvfe dans le Desert libyque: 
Cong, internat. anthropologie et arch6ologie, 7* 
sess., Stockholm, 1874, Compt. rend., vol. 1, 
pp. 76-79, 1876. 
ZiTTEL, Karl A. von. 

1887.1. Handbuch der Palaontologie, 900 pp., 719 figs.^ 
Munich and Leipzig, 1887-1890. Revised 
edition, "Grundzuge der Palaontologie," 971 
pp., 2048 figs., Munich and Berlin, 1895. 
Revised and edited by F. Broili, E. Koken, 
M. Schlosser, 2. Abteilung, Vertebrata, 598 
pp., 749 figs., 1911. 



PLATE XLY 



PLATE XLV 

Ontogenesis of the Horns of Domestic Cattle 

Development of the horny sheath and the osseous horn in seven stages. Drawn from preparations made by S. H. Chubb, in the 
collections of the American Museum of Natural History. 

A, Foetal skull, approximatelj- in the fifth month of the intrauterine growth. The future horn area is indicated by a thickening of 

the epidermal cells, from which arise 40 scattered hairs (first stage of the hair tuft) . 

B, Foetal skuU approximateh' of the sixth or seventh month. Epidermal thickening enlarged, covered with the pointed and partly 

agglutinated hair tuft (beginning of the horn sheath). 

C, Foetal skull of the ninth month, in which first appears the osseous horn rudiment covered with a cap of thickened epidermis 

and a rudiment of the horn sheath of the agglutinated hairs. 

D, Skull of calf two weeks after birth, in which the osseous horn and the horn sheath are shifting back toward the occiput, through 

the ontogenetic allometric development of the frontals and recession of the parietals. 

E, Skull of calf two months after birth, showing the allometric shifting of the small osseous horn and the horn sheath toward the 

occiput. 

F, Yearling skull of 18 months, in which the osseous horn and the sheath are fuUy shifted back upon the occiput. 

G, Adult skull of the ninth year, showing the completed osseous. 



n. S. GEOLOGICAL SURVEY 



MONOGRAPH 65 PLATE XLV 




ONTOGENESIS OF THE HORNS OF DOMESTIC CATTLE 



n. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE XLVI 




MODELS OF HEADS OF EOCENE TITANOTHERES, SHOWING BRACHYCEPHALY, MESATICEPHALY, AND 

DOLICHOCEPHALY 



A, Palaeosyops leidyi, brachycephalic. B, Manteoceras manteoceras^ mesaticephalic. C, Telmatherium ultimum, mesatice- 
phalic. D, Dolichorhinus hyognathus, dolichocephalic. H, Horn s-welling 




101959— 29— VOL 2- 



V. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE XLVIII 



Bunodont 



Bunoselenodont and 
Brac/iyselcnodont 



SeLefwdont 




EXTINCTION OR SURVIVAL OF 23 FAMILIES OF EVEN-TOED ARTIODACTYL UNGULATES 

Note the extinilion of 13 families having bunodont and bunoselenodont grinding teeth, the survival of 3 bunodont famili 
and the expansion of 7 selenodont famihes 



V. S. GEOLOGICAL SUUVEY 



MONOGKAPH 55 PLATE XLIX 




BRAIN OF EOCENE MAMMALS COMPARED WITH THAT OF MODERN AND OTHER MAMMALS 



Archaic forms m left column; modernized forms in right column. Olfadlory lobes represented by dots; ce: 
black lines; cerebellum and medulla by dashes. A, Ardlocyon. Eocene flesh eater; Ca-nis, the dog moderr 
nacodus. Eocene primitive ungulate; Sus, the domestic pig. C, Coryphodon, Eocene ungulate; 'Rhinoce 
of same s,ie. D, Uintatherium, massive Eocene ungulate. E, Menodus gigantens, Oligocene titanothere 
natural size 



ebral hemispheres by 
flesh eater. B, Phe- 

OS, living rhinoceros 
All over one-fourth 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE L 





TYPE SKULLS OF PALAEOSYOPS LEIDYI AND MESATIRHINUS PETERSONI 



A, Palaeosyops leidyi, type (Am. Mus. 1544), Bridger Basin, Wyo., level Bridger C or D 
skull (Am. Mus. 12184), CattaU Springs, Bridger Basin, Wyo., level Bridger D 3; lov 
Meadows. Washakie Basin, Wyo., level Washakie A. Both one-third natural size 



Mesatirhinus petersoni, type 
aw (Am. Mus. 1512), Laclede 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 65 PLATE LI 




SKULLS OF TELMATHERIUM ULTIMUM AND MANTEOCERAS MANTEOCERAS 

A, Telmatherium ultimum, type (Am. Mus. 2060), Uinta Basin, Utah, level Uinta C. B, Manteoceras manteoceras (Am. Mus. 2353), south of Hayilack 

Mountain, Washakie Basin, Wyo., level Washakie A(7), canines and incisors partly reilored from Am. Mus. 1566. Both one-third natural sise 




.3 - 
S « 
« 2 



3 



< 

Q 1 ti 

CO 15 .„ 

3 D - 

X .m 

H .S « 

< s s 

Z a -^ 

On-' 

S-3 

>< I > 

as 3^ 






Si 



^ « 

n '^ 
S 5 



5 « c 



V. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LIII 



/H^"#^ 






SKULL FORM IN EOCENE TITANOTHERES 

A, Brachycephalic type, Pdlaeosyops leidyi (Am. Mus. 1516). B, Mesaticephalic type, Manteoceras manteoceras (Am. Mus. 1569). C, Dolichocephalic^ 

type, DolichorliinMi hyognathns (Am. Mus. 1851) 



PLATE LIV 



PLATE LIV 

Upper and Lower Grinding Teeth of the Lower Eocene Titanotheres, Lambdotherium and Eotitanops 

Drawings by E. S. Christman. Natural size. (See pp. 279, 281) 

A, Lambdotherium popoagicum (Am. Mus. 2688), Huerfano Basin, Colo., upper fourth premolar and molars. 

B, L. popoagicum (Am. Mus. 4864 and 4880), m-, upper premolars and molars. 

C, L. primaevum, type (Amherst Mus. 254), Buffalo Basin, Wyo., upper molars. 

D, The same, lower molars. 

E', L. popoagicum (Am. Mus. 2989), Wind River Valley, Wyo., incomplete lower dentition in jaw fragment, external view. Pmj 

in this view is restored from Am. Mus. 4863, a somewhat larger animal. 
&, The same, internal view. 

E, Eotitanops gregoryi, type (Am. Mus. 14889), Alkali Creek, Buck Spring, Wind River Basin, Wyo., upper level of "Big Red 

Pocket," 100 feet above heavj' red stratum, level Wind River B ("Lost Cabin"), second and third left upper molars. 
G, The same, right lower premolar-molar series; drawing reversed. 
H', E. borealis, type (Am. Mus. 4892), Wind River Valley, Wj'o., Wind River formation, level undetermined, right upper cheek 

teeth. 
H^, The same, outer side of p* and m'. 
I', E. borealis (Am. Mus. 14SS7), Dry Muddy Creek 12 miles above mouth, Wind River Basin, Wyo., Wind River formation, level 

undetermined, left upper premolar-molar series. 
P, The same, outer side view of p^p'. 
J, E. princeps, type (Am. Mus. 296), Wind River Basin, Wyo., Wind River formation, level undetermined, left lower grinding 

teeth, crown view; drawing reversed. 



PLATE LV 



PLATE LV 

COMPAEISON OF InCISOES AND CaNINES IN EoCENE TiTANOTHERES FROM THE IJPPER LeVELS OF BrIDGER BaSIN 

AND THE Lower Levels of Washakie Basin 
Drawings by E. S. Christman. Natural size. (See p. 298) 

A, Palaeosyops aff. P. robustus (Am. Mus. 1584), Bridger Basin, Wyo. 

B, Telmatherium cultridens (Princeton Mus. 10027), Sage Creek Spring, Bridger Basin, Wyo. 

C, Manteoceras manteoceras (Am. Mus. 12683), Sage Creelv Spring, Bridger Basin, Wyo. 

D, Telmatherium ultimum, paratype (Am. Mus. 2004), Wiiite River, Utah, premaxillary region crushed laterally. 

E, Dolichorhinus hyognathus (Am. Mus. 13164), Washakie Basin, Wyo., Haystack Mountain, 10 feet above pink stratum, level 

Washakie B. 





u n u 

« I 3" 

1:^1 



. 9 W 

2 fl-^ 



.3 M 



^ o,'i 






>; ^ 



s s a 
o ■§ o. 

■- '-' ^ 



<^4i 



o ;: 



2; 



HI 

> a 
a ■§« 



w".S 



^ III 
a . " -? 



Sh)^ 



^S 






■S . S3 
9 an ^ 




101959— 29--VOL 2- 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 65 PLATE LVIII 







UPPER DENTITION IN FOUR SPECIES OF PALAEOSYOPS 

Illuilrating successive ^ages in the evolution of the premolars. (See p. 314.) Dra\vings by E. S. Chri^man. Natural sise. A, Palaeosyops 
paludosus (Am. Mus. 13032), 5 miles south of Granger, Wyo., level Bridger B 1, pl-m'. B, P. major (Am. Mus. 12182), middle of Ck>tton- 
vi^ood Creek, Bridger Basin, Wyo., level Bridger B 3, p^-m^, drawing reversed. C, P. rohu^us, type (Yale Mus. 11122), part of upper 
Bridger formation, probably level C or D, p^-m^, drawing reversed. D, Palaeosyops sp. (Am. Mus, 2361), south of Hayitack Mountain, 
Washakie Basin, Wyo., level Washakie A, pl-m? 










s a 



^.7 

m 0. 



I 1 

^ .0 









" ■- "3 



H t 

'S « ~ 



ZB 8 



S-s 






•S Art 



PLATE LX 



PLATE LX 

Upper Dentition of Limnohyops and Palaeosyops 
Illustrating three stages in the evolution of the premolars. (See pp. 306-336.) Drawings by E. S. Christman. Natural size 

A, Limnohyops prisons, type (Am. Mus. 11687), Grizzly Buttes West, Bridger Basin, Wyo., level Bridger B 2; pi-m'. In this species 

the second, third, and fourth premolars are in a relatively low stage of evolution. 

B, Palaeosyops leidyi, type (Am. Mus. 1544), Bridger Basin, Wyo., level Bridger C 4(?), i'-m'. In this species the premolars are 

in a more advanced stage of evolution. 

C, P. copei, type (Am. Mus. 1170^,- Henrys Fork, Lone Tree, Bridger Basin, Wyo., level Bridger D 3, pi-m^, drawing reversed. 

In this species the premolars are in a very advanced stage of evolution. 



D. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXI 




TYPE SKULL OF PALAEOSYOPS LEIDYI, PALATAL VIEW 
Am. Mus. 1544, Bridger Basin, Wyo., level Bridger C 4(?). (See p. 327.) One-half natural size 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXII 




UPPER AND LOWER PREMOLARS OF LIMNOHYOPS AND PALAEOSYOPS 

Drawings by E. S. Chriilman. (See pp. 312, 314.) Natural size. A-K, External (lateral) views of the upper premolars, show^ing 
various ^ages in the separation of the cingulum from the edloloph; L-O, internal (medial) views of the low^er premolarSj 
showing various ^agcs in the evolution of p2, ps. A, Limnohyops prisctis, type (Am. Mus. 11687), Gristly Buttes We^, 
Bridger Basin, Wyo., level Bridger B 2, p^-pS drawing reversed. B, L. mo-noconusl (Am. Mus. 5102), Cotton^vood Creek, 
Bridger Basin, Wyo., level Bridger B, p--p*, drawing reversed. C, L. laticeps, type (Yale Mus. 11000), Bridger formation, 
level?, p3, p^, drawing reversed. D, Palaeosyops -major, neotype (Am. Mus. 12182), middle of Cotton\vood Creek, Bridger 
Basin, Wyo., level Bridger B 3, p^-p^ drawing reversed. E, P. paludosus, one of the cotypcs (Nat. Mus. 762), level Bridger 
B(?) 1-2, p<. F, Palaeosyops sp. (Am. Mus. 2361), south of Hayrack Mountain, Washakie Basin, Wyo., level Washakie A, 
p2-p*- G, P. major (Am. Mus. 12182, same as D), p^, drawing reversed. H, P. rohu^us (Am. Mus. 1558), Twin Buttes, 
Bridger Basin, Wyo., level Bridger C or D, p^, drawing reversed. I, P. leidyi (Princeton Mus. 10009), level Bridger C or D, 
pi-p^, dra^ving reversed. J, P. rohu^us (Am. Mus. 1554), Henrys Fork, Bridger Basin, Wyo., level Bridger C(?) or D, p^-pS 
drawing reversed. K, P. grangeri, type (Am. Mus. 12189), Twin Buttes, Bridger Basin, Wyo., 200 feet below red stratum, 
level Bridger C 1, p'-pS draw^ing reversed. L, L. priscus (Am. Mus. 11688), Grizzly Buttes Bail, Bridger Basin, Wyo., level 
Bridger B 2, p2-pj, draw^ing reversed. M, P. paludosus (Am. Mus. 11680), Little Dry Creek, Bridger Basin, Wyo., level 
Bridger B 1, p2-p4, drawing reversed. N, P. major (Am. Mus. 12181), middle of Cottonw^ood Creek, Bridger Basin, Wyo., 
level Bridger B 3, p2-pi, drawing reversed. O, P. leidyi (Am. Mus. 12200), Henrys Fork, Lone Tree, 50 feet below white 
^ratum, level Bridger C 1, p2-p4, drawing i 



TJ. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXVI 






SKULL AND JAW REFERRED:.T0 STHENODECTES INCISIVUS 

A, Referred skull (Field Mus. 12168) and lower jaw (Field Mus. 12166), "Amynodon beds," level Uinta B 2, drav 
\ fourth natural size; B, the same skull, basal view; C, the same jaw, top view 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXVIII 





Drawings by E. S. Chriitman 

k^_^ upper J3iW with crown vi 

inner view of low^er i 



INCISORS AND CANINES OF PROTITANOTHERIUM EMARGINATUM 
atural sise. (See p. 378.) A, Tj^e (Princeton Mus. 11242), level Uinta C 1 (of true Uinta formation), front part of 



e, front part of lower ja-w ^vith 
'■ of mi, ma, dra-wing reversed 



■■ of incisors and 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXIX 





B2 





UPPER AND LOWER TEETH OF PROTITANOTHERIUM 

Progressive Stages. Natural size. (See pp. 379, 380.) A, Protitanotherium superhum, type (Am. Mus. 2501), 
White River, Uinta Basin, Utah, level Uinta C (true Uinta formation), upper molars (m', m') found 
associated with type lower jaw. Bi, P. superhum (Am. Mus. 2501), lower premolars of type; B2, the 
same, inner side view. C, P. entargimatum, type (Princeton Mus. 11242), inner side view of lower premolars 



TT. S. GEOLOGICAL SUEVEY 



MONOGRAPH 65 PLATE LXX 




LOWER DENTITION OF BRACHYDIASTEMATHERIUM TRANSILVANICUM 
(After Bockh and Maty; see p. 382.) One-half natural sise. A, Type jaw, Andrasha^a, Transylvania, Hungary, 
Tapper Eocene; A', outer side vievi' of front part of low^er jaw with canine and pi-mi; A^, inner view of pi-mi; 
A3, back vie-w of mi; AS front view of mi; A*, superior view of fragmentary low^er ja-w and dentition; A^, 
; A', sedlion of left loiver canine; A', fragments of lower incisors 



external vie' 

101959— 29— VOL 2 



■ of left lower 

— 16 



PLATE LXXI 



PLATE LXXI 

Lower Dentition of Mesatirhinus, Metarhinus, and Dolichorhinus 
Drawings by E. S. Christman. Natural size. (See pp. 393, 394, 404, 424, 425) 

A, Mesatirhinus petersoni (Ara. Mus. 1512), Laclede Meadows, Washakie Basin, Wyo., level?, left lower premolar-molar series, 

crown view. 

B, Mesatirhinus sp. (Am. Mus. 2355), lower beds, south of Haystack Mountain, Washakie Basin, Wyo., level Washakie A(7) (Glove 

Spring), right lower premolar-molar series, crown view, drawing reversed. 

C, Dolichorhinus hyognathus (Am. Mus. 1850), White River, Utah, level Uinta B 2, left lower cheek teeth (pi-ms), crown view. 
D', Metarhinus fluviaiilis (Am. Mus. 2059), White River, Utah, level Uinta B 1, external view of lower dentition. 

D^, The same, crown view. 




i 




PLATE LXXIII 



PLATE LXXIII 

Upper Dentition of Dolichorhinus 

Drawings A, B by Sidney Prentiss; C by E. S. Christman. Natural size. (See pp. 405, 406, 416) 

A, Dolichorhinus heterodon, type (Carnegie Mus. 2340), level "Uinta B 2 or C 1, upper cheeli teeth (after Douglass), drawing reversed. 

B, D. longiceps, type (Carnegie Mus. 2347), level Uinta B 1, right upper cheek teeth (after Douglass). 

C, D. intermedius, type (Am. Mus. 1837), White River, Utah, level Uinta B 2, drawing reversed. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXXV 






SKULLS AND JAW OF DOLICHORHINUS 

<After E. S. Riggs; see pp. 405, 406, 417.) One-fourth natural sfce. A, Dolichorhinus superior, type (Field Mus. 
12188), Uinta B 1. B, D. flumi-nalis, type (Field Mus. 1220S), "Amynodon sandftone, Uinta B" (level Uinta 
B 2 of this monograph). C, D. langiceps (Field Mus. 12200), referred jaw 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXXVI 






SKULLS OF DOLICHORHINUS, SIDE VIEW 

(After E. S. Riggs; see pp. 405, 405, 417.) One-fourth natural sije. A, Dolichorhinus superior, type (Field Mus. 
12188). B, D. longiceps, referred epecimen (Field Mus. 12175), drawing reversed. C, D. fluminalis, type (Field 
Mus. 12205) *^Amynodon sand^one," level Uinta B 2, drawing reversed 



IT. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXXVII 




SKULLS OF DOLICHORHINUS, PALATAL VIEW 



One-fourth natural size. A, Dolichorhinus superior^ type (Field Mus. 12188; after 
sandAone, Uinta B" (level Uinta B 2 of this monograph) (after E. S. Riggs). 
palatal view (after Douglass) 



S. Riggs). B, D. fluminalis, type (Field Mus. 12205), "Amynodon 
3, D. longiceps, type (Carnegie Mus. 2347), low in level Uinta B, 



V. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXXVIIl 



m 





SKULLS AND JAWS OF METARHINUS 



: Riggs- see p. 421.) One-fourth natural size. A. Metarhinus criAatus, type (Field ^us. 12194). -Upper Meta.Wm.s beds WWte River 
'•• B M Lr ei referred (Field Mus. 12187), "Upper MetdrHinu. beds," level Uinta B 1. C, M. r^ar^us, type (Field Mus. 12186). D, M . 
ear^.. ;eferredYowe: I^ (F^eld Mus. 12178), "Upper Met...i.u. beds." Uinta Basin, level Uinta B 1^ E, M. ..p..^.. referred lower ,aw (Field 
Mus. 12195). "Upper Metarhinus beds." Uinta Basin, level Uinta B 1. side vievir, drawing reversed; W, the same, top view 



After E, 
Cany 



XJ. S. GEOLOGICAL SUHVET 



MOXOGRAPH 55 PLATE LXXIX 






SKULLS OF METARHINUS 

(After E. S. Riggs; see p. 421.) One-fourth natural site. A, Metarhinus cria.atus, type 
(Field Mus. 12194), drawing reversed. B, M. earlei, referred (Field Mus. 12187), "Upper 
Metarhinus beds" (level Uinta B 1), Uinta Basin, drawing reversed. C, M. riparius, type 
(Field Mus. 12186), "Upper Metarhinus beds" (level Uinta B 1), Uinta Basin, drawing 
reversed 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 65 PLATE LXXX 




SKULLS OF RHADINORHINUS AND METARHINUS 



<After E. S. Riggs; see pp. 426, 430.) One-fourth natural size. A 
beds'* (level Uinta upper B 1), Uinta Basin, top view; A^, the 
Metarhinus beds," Uinta Basin, basal view 



, Rhadinorhinus abhotti, type (Field Mus. 12179), "Upper Metarhinus 
iame, basal view. B, M. earlei, referred (Field Mus. 12187), "Upper 



U. S. GEOLOGICAL SUBVEY 



MONOGRAPH 55 PLATE LXXXI 





UPPER DENTITION OF EOTITANOTHERIUM OSBORNI AND DIPLACODON ELATUS 

(After Peterson; see p. 435.) Drawings by Sidney Prentiss. One-hajf natural sise. A, Eotitanotherium osborni^ type (Carnegie 
Mus. 2859), upper ja-%v and dentition. B, Diplacodon elatus, type (Yale Mus. 11180), upper dentition 

101959— 29— VOL 2 17 



V. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXXXII 







COMPARISON OF UPPER TEETH OF RHADINORHINUS, DIPLACODON. MENODUS, AND BRONTOTHERIUM 
Drawings by E. S. ChriStman. A is natural size; in B, C, and D the premolar-molar scries is reduced to tKe same length as in A. (See pp. 434, 
474.) A, Rhadinorhinus diploconus, type (Am. Mus. 1863), White River, Utah, level Uinta B 1 or 2. B, Diplacodon elatus, type (Yale Mus. 
10320), level Uinta C(?). C, Menodus cf. M. proutii (Carnegie Mus. SS8), Chadron formation. D, Brontotherium gigas elatus (Am. Mus. 
492), Cheyenne River, S. Dak., Chadron formation 



PLATE LXXXIII 



PLATE LXXXIII 

Skull of Brontops brachycephalus 
Plate prepared under the direction of Professor Marsli. One-fourth natural size. (See p. 4S4 and PI. LXXXVII) 

Nat. Mus. 4947, skull B, paratype, female; Big Badlands, S. Dak., "west side of Quinn Draw, near the Cheyenne River" (J. B, 
Hatcher); Tiianotherium zone ( = Chadron formation) of South Dakota, level lower A (Hatcher). Ai, Top view; A2, side view; 
A3, front view. The female sex is inferred from the small horns, slender zygomata, less brachycephalio proportions, and weak 
canines. The horns are smaller and more primitive than in any other known Oligooene titanothere; the skull comes from an- 
extremely low level in the "Titanoiherium beds" (14.4 feet above the Pierre shale and 130.6 feet below the top of the "Tiiano- 
therium beds," according to Hatcher) and seems to represent a slightly more primitive stage than the type of the species. 



n. S. GEOLOGICAL SURVEY 



MONOGRAPH 53 PLATE LXXXIIT 




A2 




SKULL OF BRONTOPS BRACHYCErHALUS 






U. S. GEOLOGICAL SUEVET 



MONOGRAPH 65 PLATE LXXXIV 





SKULL OF BRONTOPS BRACHYCEPHALUS 

Plate prepared under the direftion of Professor Marsh. One-fourth natural size. (See p. 487. also Pis. LXXXV and LXXXVI.) Nat 
Mus. 4258, skull F, female, referred to B. hrachycephalus. (According to the records in the U. S. National Museum this xs skull 
F of the 1886 coUecSion. not skuU F of Hatcher's diary of 1887.) Big Badlands. S. Dak., Chadron formation (Tttanotftermm zone), 
level middle B (Hatcher). This immature skull is vertically crushed, so that the nasals and horns are tilted upward and bacK- 
ward, and the apparent breadth of the skuU is increased. The sutures show the participation of both nasals aond frontals in the 
horns. The lacrLal is represented as entering the outer ridge of the horn. The skull is larger than that of the type and comes 
from a higher level. The dental measurements are much smaller than in any skull referred to B. dispar 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE LXXXVI 





SKULL OF BRONTOPS BRACHYCEPHALUS 
Plate prepared under the diretftion of Professor Marsh. One-fourth natural sue (See p 488, also Pis. LXXXIV and LXXXVni.) 
Nat. Mus. 4258, skuU F, referred, Big Badlands, S. Dak., Chadron formation (Titanotherium sone), level middle B (Hatcher). 
Al, Basal view, a partial reconftru(£lion, the adlual condition being shown in Plate LXXXVIII. The tooth rows as here repre- 
sented are somewhat too redlilinear, and the details of the premolars are mcorrea m several points. The suture between the 
basi-, pre-, and alisphenoids is well shown. A', Front view showing the sutural relations of the frontals, nasals, lacrimals, 
maxillaries, malars, etc., relations which are essentially the same as in Eocene titanotheres 



PLATE LXXXVII 



PLATE LXXXVII 

Skulls of Brontops brachycephalus 
One-fourth natural size. (See pp. 484, 485, also PL LXXXIII) 

A, Paratype (Nat. Mus. 4947), skull B, female, Big Badlands, S. Dak., "west side of Quinn Draw, near the Cheyenne River" 

(J. B. Hatcher); Titanotherium zone (=Chadron formation), lowest levels (14.4 feet above Pierre shale). As compared with 
the type male skuU (B) this aged female is smaller, more elongate, and slenderer, with small canines and unexpanded zygomata. 
The much worn molar teeth appear somewhat more primitive than those of the type, especially m'. 

B, Type (Nat. Mus. 4261), skull c, male, Big Badlands (probably Corral Draw), S. Dak., near base of lower Titanotherium zone. 

Generic affinity with Brontops robustus is indicated by the presence of one or two round-topped incisors on each side (the inner 
pair often drop out in adults) , the stout canines, the retarded postero-internal cusps on p<, the general brachycephalic form of 
the skull, and the character of the horns. 





Su 



£ 2 S a 5 

■ " .tj i7 P 

u - ^ -^ "3 

> B3 2 « a 

■"3 e CO -a 

" a " « js 



" H > b 

■g -o ^ E 

S £ c" ^ 

ft) " -^ 



<= -a ^ i 



? 3 



> **- y « rt -a 

m '' ii >>u a B 

3 fc^" B 1 "> « -S 

_ ^ "^ » S * "2 

Z " -^ 5 ^3 3 2 

5 o ^ "3 (g .g i 

S T3 -^ bo a >i 

m ^ ^ S S "3 

>; i. ^ ^ bh " 



9 S-« 



s S ?= -^ 

° E S «>■ -a § 

W S u 3 .3 u 
." t ;S ^ -S '^ 

« s * 5x e 
o 2 a " a ° 

a r § -s a .2 
«^- §3 3 

> Q B. ^-io g- 

B " M 13 -S 

►J|m*a'H 
^ a „ a '3 S 

S !3 •£ m -^ 2 

■3 " -a i ^ « 
a'm is 2i u 1 
? -"ass 

m" £ a - " o- 
S (fl a « a to 



i 5 .0 



_^ 



>" Tl ^ 

' 5 fi e * a 

. £« S b 

4> m T3 4) 4) *» 

.H rt -a s a 

3 'u J J « •» 

•3 °o? 2 i 

b ft ^ ^ J3 u 

■w ?> a T3 " '^ 

"■ a .a 2 

a ^ B « C » 

t" J5 S ? Si, 3 

3 a ■; h " So 

M S *» 4> a 

"^ K I -S ■£ u 



TJ. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE XC 




SKULL OF BRONTOPS DISPAR 

Lte prepared under the diredlion of Professor Marsh. One-sixth natural si:;e. (See p. 489.) Nat. Mus. 
4245 (erroneously lettered skull P, according to J. B. Hatcher), locality uncertain, Chadron formation. 
Ai, Top view; A-, palatal view; A3, side view. This skull exhibits a certain resemblance to Allops 
marshi, but the measurements are closer to those of B. dispar. The original is badly crushed and as here 
reconstructed the tooth rows may perhaps be too redtilinear 



PLATE XGI 



PLATE XCI 

Skull of Brontops dispar 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See pp. 489, 490) 

Nat. Mus. 4290, skull K, Big Badlands, S. Dak., Titanoiherium zone, level middle B (J. B. Hatcher). Referred male skull, type of 
B. validiis. Vertical crushing has lessened the vertical diameters of the skull top and tilted the horns upward and backward. 
The skull is very robust and braohy cephalic. The stout horns are rounded in section, with a rugose area at the tip for the 
attachment of the thickened epidermal cap. The general conformation of the skull, especially of the nasals and horns, suggests 
possible ancestry to B. robustus. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE XCI 




SKULL OF BUONTOPS DISPAR 





F 


^I^^^^^^HI 








^^^^^^^^H^^^^^^^^^'^&i 


n 




^^^^^^^r^ ■M^^^^^^^^' ^Jti^l 


■ 


^^^^'^<-K^'^^^^^^^^ 1 


^^^^^^^H '' ^&u'*''''' \.^^ ^t9 


1 


^^^B. ''^'' v^V 'm ■ /"'a 


^^^^^^^^H j^^ '^^SpjWIK&i.AAi "J 




^^^^^^^^■>' ^^'K^t^ -«ir 


^^^^B -II^B^p^^^?/ 


' 


^BPB^^^^:-.,. ,'./?^' 




\ 


■■:ymm^.- 










« 


Mii^^fei&!.^...,-^s. 






^^^^^^^^1 s . i'flSHHHH|HH^^HfH^MiH| 




■l^^ppss^^^ ^ 


^^^H.' flSP'^^^^v^ 






^^^^^^9^^' ^ilJM 




^^H ^J^^^^^|_j 


^^^^^^^^^^^^^^^^^^^^^^^^^■ll^^^^^ X ^^9 








1 


^E^ ^ 5^M^^^^^^^| 




I 


^^^^■y|i£4. ^^^^^^^#^ 


^^^^H ^^^^^H^BhIP^^'^'U - 


^ 


^I^^^HHt&^t ^^^m-jatil 


^^^^V -^^^^^^^H^^ ' 




' ^^^■^■■l^v^^^ .^^■P^'^O^. 


■IP ^^^^H'v '' 




.^^^^^^^|s«««^^^^. ;^^P'' 


^^■R«^^ ' 'V^s^^^^^^Hk, 










m- ^^^^^^Bmr~~l 




- ; \ ^ 


■ ■ ■ ^^ "^^P^^^T^^ \ 




1 

f 








^^H^ ' ^^^^ i.m^ ^-v. 


~'->.. 


'^ j^*r- 






|gg|nHHH|^^^^*^^^^^« 


^^^k^^SkTZ^^ik^^l 






^^^^^^^^^^^^^^^^^^^^■£r& > >• i^*i^^^^l 




^Hr y' Jjdj^^^^^^^^^^^l 


^^^^^^^^^^^^H^^. 


•> i.^ 
^i^' 

^ 





ig 
s^ 



■S.s 



K-3 



1° 

i CO 



° is 

§11 

m w g 

►J -4 



S«3 



^^1 

<J J m a 

CM <j 

rt' j5 

- at 

« a K 



5 3 J3 

2 3 ei 

S ■£ " 

ill 

c 





101959— 29— VOL 2- 



PLATE XGIII 



PLATE XCIII 

Skulls of Beontops dispar 
About one-fifth natural size. (See pp. 490, 492) 

A, Nat. Mus. 1217, skuU p {lAllops serotinus), Big Badlands (probably Indian Draw), S. Dak., Titanoiherium zone, level "some- 

what doubtful, probably lower C" (Hatcher). This very progressive skull shows certain of the characters common to Brontops, 
Allops, and Menodus, in contrast with the Brontotherium-Megacerops group. 

B, Nat. Mus. 4941, skuU D, type. Hat Creek, Nebr., Titanoiherium zone, level middle B (Hatcher). This laterally crushed skull, 

the genotype of Brontops, was evidently brachycephalic, with stout rounded horns, thick nasals, short thick canines, and 
heavy zygomata. The horns are well in front of the orbits. It is thus more or less intermediate in form between the highly 
specialized B. robustus and the primitive B. brachycephalus and also exhibits convergent resemblance to the type of Brontothe- 
rium tichoceras. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE XCIII 




SKULLS OF BRONTOPS DISPAR 



U. S. GEOLOGICAL SURVEY MONOGRAPH 55 PLATE XCIV 






SKULLS OF BRONTOPS DISPAR 

Plate prepared vinder the direa:ion of Professor Marsh. One-sixth natural size. 
(See pp. 489, 491, also Pis. XCI, XCII, and XCin.) A, Nat. Mus. 4941, skull 
D, type, reconilrudled from the crushed original; only the front parts of the 
zygomata are shown. B, Nat. Mus. 4290, skuU K, type of B. validus. The 
skull is crushed down-ward, thus lessening the height of the narial opening. 
The horns are rounded in mid-se(5tion, ^vith slender tips, the nasals ^out and 
broad diftaUy, the buccal s^vellings heavy. The inner incisors are small or 
absent. C, Nat. Mus. 4245 (erroneously lettered P). A more slender skull 
Tvhich some-what resembles Allops marshi. (See PI. CXIV) 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE XCV 




SKULL OF BRONTOPS ROBUSTUS? 

One-fourth natural size. (See p. 497.) Princeton Mus. 10061, Chadron formation. The specific reference is somewhat doubtful; the skull 
resembles that of the type of B. rohu^us in the forward pitch of the short horns, in the short nasals, and in the form of the incisors 
and c 



PLATE XGVI 



PLATE XCVI 

Type Skull and Jaw of Beontops robustus 

Plate prepared under the direction of Professor Marsh. One-third natural size. (See p. 494) 

Yale Mus. 12048, Dry Creek, 5 miles northeast of Chadron, Nebr. (J. B. Hatcher), Chadron formation, probably upper levels. 
This great skuU represents the culminating member of the Broniops phylum. It is highly specialized in its massiveness and great 
breadth, in the position of the horns, which are far in front of the orbits, in the great development of the external basal ridge 
of the horns, in their flattened basal section, and in the thickness and abbreviation of the nasals. Relationship with Brontops 
dispar is indicated by the shortness of skull top and base, by the absence of the mid-parietal eminence, and by the form of the 
premolars. Convergent resemblances to Brontotherium are seen in the flatness of the horns, abbreviation of nasals, swollen 
zygomata, short, robust canines, and heavy round-topped incisors. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE XCVI 




TYPE SKULL AND JAW OF BRONTOPS ROBUSTUS 



PLATE XGYII 



PLATE XCVII 

Type Skull and Jaw of Brontops robustus 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 494, also PI. XCVI) 

Yale Mus. 1204S, Dry Creek, 5 miles northeast of Chadron, Nebr. (J. B. Hatcher), Chadron formation, probably upper levels. 
Slight vertical crushing may have somewhat diminished the vertical diameters. The skull is relatively much shorter than in 
Brontoiherium, and the occiput is not produced backward. 



U. S. GEOLOGICAL SUEVEY 



MONOGRAPH 65 PLATE XCVII 




TYPE SKULL AND JAW OF BRONTOPS ROBUSTUS 



PLATE XGVIII 



PLATE XCVIII 

Type Skull of Beontops robustus 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 494, also Pis. XCVI and XCVII) 

Yale Mus. 12048, Dry Creek, 5 miles northeast of Chadron, Nebr. (J. B. Hatcher), Chadron formation, probably upper levels. 
The top view shows well the extreme breadth of the skull as compared with that of B. dispar (PI. XCI). In Brontoiherium 
the skuU top is much longer, the occipital vertex is produced backward, and there is little or no constriction of the vertex 
in front of the occiput. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE XCVllI 




TYPE SKULL OP BRONTOPS ROBUSTUS 



PLATE XGIX 



PLATE XCIX 

Type Skull and Jaw of Brontops robustxjs 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 494, also Pis. XCVI-XCVIII) 

Yale Mus. 12048, Dry Creek, 5 miles northeast of Chadron, Nebr. (J. B. Hatcher), Chadron formation, probably upper levels. 
The front view of the skull, compared with that of Brontops dispar (PL XCIV), shows the increased specialization in its greater 
size and width, expanded zygomata, and thickened horns. 



U. S. GEOLOGICAL SURVEY 



MONOGKAPU 55 PLATE XCIX 




-5 m^^ 








"^ ""^^ 




.cd. 



TYPE SKULL AND JAW OF BRONTOPS ROBUSTUS 



PLATE G 



101959— 29— VOL 2 19 



PLATE C 

Type Skull of Brontops robustus 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 494) 

Yale Mus. 12048, Dry Creek, 5 miles northeast of Chadron, Nebr. (J. B. Hatcher), Chadron formation, probably upper levels. 
The palatal view shows well the extreme breadth, which is characteristic of this species. The dentition is shown also in 
Plate CI. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE C 




TYPE SKULL OF BRONTOPS ROBUSTUS 



PLATE CI 



PLATE CI 

Dentition of Type Skull of Brontops robustus 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 495.) 

Yale Mus. 12048, Dry Creek, 5 miles northeast of Chadron, Nebr. (J. B. Hatcher), Chadron formation, probably upper levels. 
The upper teeth (Bi, B2) may be compared with those of B. dispar (PI. LXXXVIII). In B. robustus the internal cingula of 
the premolars are somewhat reduced, but the postero-internal cusps or tetartocones are better developed. There is also 
a wide postcanine diastema. Although the canines suggest those of Broniotherium (PL CLXIII) the incisors are not broadly 
cingulate posteriorly, as in that genus; the tetartocones, especially in p^, are more retarded; p^ is also less molariform than in 
Broniotherium; and all the cheek teeth are narrower transversely. The lower teeth (Aj, Aj) show reduced external cingula 
on the premolars and molars. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 56 PLATE CI 




B2 





A2 




Al 



DENTITION OF TYPE SKULL OF BRONTOPS ROBUSTUS 



PLATE Gil 



PLATE CII 

Lower Jaw of Bkontops robustus 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 496) 

Yale Mus. 12048, type, Dry Creek, 5 miles northeast of Chadron, Nebr. (J. B. Hatcher), Chadron formation, probably upper levels. 
Ai, Top view; Ao, side view. The very shallow symphyseal region of this jaw by convergent evolution resembles that of 
Brontotherium. 



U. S. GEOLOGICAL SURVEY 



MONOGllAPlI 55 PLATE OH 




LOWER JAW OF BRONTOPS ROBUSTUS 



PLATE cm 



PLATE cm 

Type Skull and Jaw of Brontops eobustus 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 494) 

Yale Mus. 12048, Dry Creek, 5 miles northeast of Chadron, Nebr. (J. B. Hatcher), Chadron formation, probably upper levels. 
Posterior view of skuU and jaw, illustrating the great breadth of the skull, the deep fossae for the attachment of the heavy 
neck muscles, and the transversely expanded condjde of the jaw. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CIII 





TYPE SKULL AND JAW OF BRONTOPS ROBUSTUS 



D. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CIV 



■ 


■■ 






WB 


^^^^H 


I 


f» 


a^*** 


j. ' ^^B^^^l 


fli 


^^^^R^^l^ ■ '■- J^2^'^^*k-.t '^'^^^1 


^^p 




^ 


^ ;^;;^ 






L 


1 ^ 


^W 


f -> 


( 


,:^^PH 


■ 


L;j| 




"^%4 


■i 


BiiiMF!lir?!:f'^. - ^^^ 


■ 


Im 








g^^g 




TYPE SKULL OF DIPLOCLONUS BICORNUTUS 

ic-fifth natural size. (See p. 501.) Am. Mus. 1475, Quinn Draw, Big Badlands, S. Dak., Chadron formation. This skull is 
referred to the genus Diploclonns from the presence of an accessory knob or hornlct on the antero'intemal face of the horn. 
It differs from jDiploclonus amplus in the smaller size of the horns and in the lesser ^vidth of the zygomata (the narrowness 
of the skull is emphasized by lateral crushing). The side view[(AO shows the chara(5leriilics of the menodontine group (deeply 
concave skull top, long ^out nasals, grinding teeth w^ith sharp external cingula). The top view (A^) sugge^s that of Menodus. 
This skull was at firil regarded as that of a female, but the ^out canines and zygomata do not support this view 



PLATE GV 



PLATE CV 

Skull Related to Diploclonus tyleri 

Am. Mus. lOSl, original paratj-pe (see p. 502) of Z). bicornutus Osborn, Cheyenne River, S. Dak., Chadron formation. A', Side 
view; A-, top view. One-fifth natural size. This skull was at first regarded by Osborn as a male of D. bicornutus, but, as 
observed by Lull, its closer resemblances are with the tj'pe of D. tyleri. (See PI. CVII.) It approaches the type of D. amylus 
in breadth, but the horns are not so flat. The accessory hornlets are well marked. The length of the nasals is conjectural. 
The constricted parietal region and very concave skull top are characteristic of the menodontine group. The longitudinal 
slit in the parietal vertex is in the region of the sagittal fossa of Manteoceras and of a parietal opening in certain skulls of 
Brontops brachycephalus; apparently the outer tabula of the parietal bone was thin in this region. Irregular projections of 
the temporal crest above the mid-cranial region are seen, as in Allops, and they may have marked the dorsal edge of the 
insertion area for the temporal muscle. 



O. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CV 














i 


•■^l 










^^^^^^^ 


^^ J&di 


m 


"'*^""" W^" 




m 


--..^ 


■^ 


"^^^^^^ttii 




i. 


w 


-::•,:-#); 


w 




. ..■;iV_:S' 


l^^f^H^v 


\ 




^j^ 5?^;-'" 


iKii^ #^ 


■:-4 


^^ 


1 


kg 


-r 


1 




1^ 










^ 


j 


^' *: : 



SKULL RELATED TO DIPLOCLONUS TYLERI 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 65 PLATE CVI 





TYPE SKULL, JAW, AND TEETH OF DIPLOCLONUS TYLERI 

After original figures by Prof. R. S. Lull. (See p. 503.) One-sixth natural sise. Amheril College Zool. Coll. 327, 
type, "near the head of Bear Creek, a tributary of the Cheyenne River [S. Dak.]. The exadt locality -was on 
the north side of Spring Draw Basin, about 10 miles from the mouth of Bear Creek. Here some 200 feet of titano- 
there beds were found, lying upon Fort Pierre deposits, in which titanothere bones were discovered from a 
point 6 feet above the contadt upward; the specimen under consideration lying 35 feet above the base of the beds, 
hence in the upper part of the lower division."— J. B. Hatcher, Am. Naturaliil, vol. 27, p. 218, 1893, cited by 
R. S. Lull, 1905, p. 443. This large skull recalls Brontopi robuiius or a progressive B. dispar in general appearance 
but is referred to the subgenus Diploclonus partly because it has a large accessory hornlet. The canine is very 
^out; in p* there is pradlically but one internal cusp, the po^ero-internal cusp being barely defined and con- 
tinuous with the cingulum. The general pattern of the premolars is the same as in Allops and Brontops. The 
lower jaw recalls that of Brontops robuflus. (See Pis. XCVn, XCIX, and ClI) 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CVII 




NASOFRONTAL REGION OF SKULLS OF DIPLOCLONUS 

A, D. bicornutus, type (Am. Mus. 1476). B, D. tyleriO), referred (Am. Mus. 1081); 
the nasals are largely reilored. C, D. tyleri, type (Amheril College Zool. CoU. 
327; after LuU). In these skulls we observe a progressive enlargement and 
widening of the horns, which culminates in D. amplus. (See pp. 501, 503, also 
Pis. CIV, CV, and CVI) 



U. S. GEOLOGICAL SURVET 



MONOGRAPH 55 PLATE CVIII 





TYPE SKULLS OF DIPLOCJLONUS AMPLUS AND D. TYLERI 

One-sixth natural siie. (See pp. 503, 504, also Pis. CVI and CIX.) A, Diploclorvus amplus, type (Yale Mus. l^Ol'^)' 
South Dakota, Chadron formation. Figure prepared under the direaion of Professor Marsh. This view exhibits 
well the pronounced brachycephaly, the flattening of the horns, the prominent hornlets, and the abbreviation 
of the nasals. (See PI. CVII.) B, D. tyleri, type (Amherft College Zool. Coll. 327), near the head of Bear Creek, a 
tributary of Cheyenne River, S. Dak., Chadron formation. (After R. S. Lull.) In the top view this skull 
approaches D. amplus 



U, S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CTX 





TYPE SKULL OF DIPLOCLONUS AMPLUS 

Plate prepared under the dire(5tion of Professor Marsh. One-fourth natural si^e. (See p. 504.) Yale Mus. 12015a, type. South Dakota, Chadron forma- 
tion. The lateral view (AO exhibits well the concave skull top, ^out sygomata, thick, pointed horns, and short nasals. The form of the canine 
recalls that of Allops and of Menodus; the upper premolars have external cingula. In front view (A 2) the pointed horns, pronounced hornlets, 
and Tvide zygomata are especially noteworthy. (Compare PI. CVII) 

101959--29— VOL 2 20 



V. S. GEOLOGICAL SURVEY 



MONOGRAPH 65 PLATE CX 






1 


^^S^F^^^^H 


'%^ 






M 


IHfll 


■■ ^^^^Hf^^^l^ 


Igj^l^^ 


^^■HT'' ^ ' ' iifnirif n 


".,. i 


^^m 


^^^^^^IP^ ''lu^f Vl^^ %&^ 


"waw""' 




:|l* 






^ 






^ ^^..■ 


^^:%:.:t.t «*■*-' 





SKULL REFERRED TO DIPLOCLONUS AMPLUS7 

Nat. Mus. 4710, skull e, Indian Draw, Big Badlands, S. Dak., Chadron formation (in 1901 Mr. Hatcher did not recall the exadl level). One-fourth 
natural size. (See p. 505.) This skull is provisionally regarded as the female of D. amplus, although its canines seem hardly small enough to support 
this view. As compared with the type of D. amplus the zygomatic swellings and horns are much less developed 



PLATE CXI 



PLATE CXI 

Type Skull of Allops walcotti 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 609) 

Nat. Mus. 4260, skull Q, type, South Dakota, probably CorralDraw, Tilanotherium zone, lower levels (J. B. Hatcher). This is 
regarded as the most primitive Allops. The specinaen, which is badly crushed laterally, was at first referred by Osborn to 
Menodus heloceras, but it differs from Menodus in the more elongate oval section of the horns and the more tapering form of 
the nasal and in the facts that two distinct incisor alveoli are retained and that the canines are compressed anteroposteriorly. 
It is therefore considered a very primitive member of the Allops phylum. The species shares much in common with Menodus 
heloceras (dolichocephaly, elongate grinders, sharp internal and external cingula, etc.) and according to the "group method" 
of classification would be placed in a primitive genus ancestral to both Menodus and Allops. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXI 




TYPE SKULL OF ALLOPS WALCOTTI 



O. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXIIl 




SKULL OF ALLOPS MARSHI7 

;ld Mus. P6900. Ai, Side vievi^; A2, front viewr. One-fourth natural size. (See p. 508.) This finely preserved skull and lower 
jaw are somewhat smaller than the type of Allops walcotti. The species is more brachycephalic than Brontops hrachycephalus 
and has smaller horns than the typical Allops. The relatively large canines indicate that the specimen is a male and not a female 
of some larger animal. There is a large incisor in the lower jaw. In general this specimen seems neare^ to Allops nuirshi, 
although the horns are smaller 



PLATE GXIV 



PLATE CXIV 

Skulls of Allops marshi 

One-fifth natural size. (See pp. 508, 512, 513) 

A, Am. Mus. 1445, paratype, Cheyenne River, S. Dak., Chadron formation. Crushing from above has lessened the vertical diam- 

eters of this skuU. It is more massive than the type skull and less massive than that of Brontops dispar. 

B, British Mus. 4446M. This well-preserved skull differs from the type chiefly in the greater elevation of the occiput, accenting 

the deeply concave upper profile characteristic of aU Menodontinae. The dental measurements are given on page 508. 

C, Am. Mus. 501, type, South Dakota, probably Cheyenne River Badlands, Chadron formation. This specimen was at first placed 

in the genus Brontops near B. brachycephalus and B. dispar. It is, however, less brachycephalic, its horns have a trihedral 
basal section, the canine is compressed anteroposteriorly, and the premolars and molars are of moderate breadth, with internal 
and external cingula. This combination of characters is seen only in the genus Allops, and the upper premolars of the present 
specimen closely resemble those of Allops serotinus. The upward pitch of the nasals is due to crushing. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXIV 




^M 


H^^H 


^H 


^^E^^I^^^^^^^^^^*^***'''3t^^l 


H 


Ik.^ 


^1 




^tr 


hIa&'^^^^W ^^ 


^^1 




^K Mah^tf 


HHHt ^^vwk -^ % 


^r^^" 


^^^^^^^ ii»i,i ' 'iT^ •' ' i^ !ViPw ^^^H 


KUiHHil 


^^^Htk/^' '^^wi •^l 


life 


- .^tfiw^ ^-^^wkT »^H 


^H 


Wk^^j^ 


iNi 




^^^^^ 


' ^^'ffiP' w^^ 


^5^^ 




^^^^^r 




-^ '■ 


"tjJin^^^^lB, '^^al^il 


K'^ 


^^^9fe ^*'^8?'' 




^^^^^^niliU 




1^ 


^ 


■l^^l 




SKULLS OF ALLOPS MARSHI 



PLATE GXV 



PLATE CXV 

Skulls of Allops marshi 
One-fourth natural size. (See pp. 512, 513) 

A, Am. Mus. 1445, paratype, Cheyenne River, S. Dak., Chadron formation. This skull is more brachycephalic than the type, but 

less so than in Brontops dispar. The width across the horns foreshadows Allops crassicornis. The constricted parietal region 
is characteristic of all the Menodontinae. 

B, Am. Mus. 501, type; South Dakota, probably Cheyenne River Badlands, Chadron formation. The skull is slightly crushed in 

the left side anteriorh-. The canine is compressed anteroposteriorly, the grinding teeth are of moderate width, with marked 
internal and external cingula. The premolars have deep medifossettes, as in Menodus, and closely resemble those of Allops 
serotinus. The well-preserved basicranial region shows well the fossa for the internal extension of the interarticular disk and 
the structures in the region of the periotic. 



PLATE GXVI 



PLATE CXVI 

Skulls of Allops marshi? 
One-fifth natural size. (See p. 514) 

A, Nat. Mus. 1213, skull E, Big Cottonwood Creek, Nebr., Titanotherium zone, level A 2 (J. B. Hatcher). (If this is skull E of 

the 1887 collection, the locality is as given; if it is skull E of the 1886 collection, the locality is uncertain.) This specimen has 
small horns and weak zygomata and is probably a female. The grinding teeth are relatively broader than in the type skull 
and suggest those of B. dispar. 

B, Nat. Mus. 1215, skull A'. This subbrachycephalic skull recalls that of Brontops brachycephalus. Its reference to Allops marshi 

appears somewhat doubtful. 





101959— 29— VOL 2- 



PLATE GXVII 



PLATE CXVII 

Skulls of Allops serotinus 
One-seventh natural size. (See pp. 515, 516, 517, also Pis. CXIX and CXX) 

A, Nat. Mus. 4251, skuU H, type. Big Badlands, S. Dak., Titanolherium zone, level C (J. B. Hatcher).' The horns are widely 

divergent, trihedral in basal section; the connecting crest lies in the posterior plane of the horns; the nasals are abbreviated 
by the forward growth of the horns; the parietal region is narrow, the zygomatic expansion moderate, the skull top long, the 
occiput produced backward. 

B, Am. Mus. 520, Cheyenne River, S. Dak., Chadron formation. A skull referred to A. serotinus in which the above-mentioned 

features are emphasized. 

C, Nat. Mus. 4938, skull j, Big Badlands, probably Indian Draw, S. Dak., Titanolherium zone, level middle C (J. B. Hatcher). A 

skull referred to A. serotinus. The horns and nasals are more massive than in the preceding specimen, and the parietal crest 
is broader. 

D, Nat. Mus. 2151, skuU I, collection of 1886, Big Badlands, S. Dak., Chadron formation, level upper C, 80 feet above Pierre 

shale (J. B. Hatcher). Skull of a supposed female with small horns and slender zygomata, referred to A. serotinus. 

1 Hatcher says, "77 feet above Fort Pierre shale; 34 feet below base of Oreodon beds, allowing liberally for summit of Jitanolherium beds." Darton (in letter) gives 40.7 
feet above Pierre shale for the same skull H. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXVII 




SKULLS OF ALLOPS SEROTINUS 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXVUI 




SKULL OF ALLOPS SEROTINUS 



(See p. 517, also PL 



Am. Mus. 520, Cheyenne River Badlands, S. Dak., Chadron formation. One-fourth natural f 

CXVn.) A skull referred to Allops serotinus. The generic chara(5ters of Allops which are seen in this view are as 
follows: Incisors 2-1, canines compressed anteropo^eriorly, premolars with sharp external and internal cingula, 
grinding teeth and skull of moderate breadth, zygomatic expansion moderate, horns ^videly divergent. The reference 
of this specimen to A. serotinus is based especially on the detailed resemblances in the premolars to those of the referred 
skull Nat. Mus. 4938 (PL CXX, B). The poitero-internal cusps of the premolars are less developed than in A. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXIX 






FRONT VIEWS OF SKULLS OF ALLOPS 

Plate prepared under the direiftion of Professor Marsh. One-sixth natural size. 
(See pp. 515, 517.) A, Allops serotinus, skuU H, coUeiftion of 1886, type (Nat. 
Mus. 4251), South Dakota, Chadron formation, level C. The front view- 
shows well the long, wridely divergent horns, thick nasals, and moderately 
expanded zygomata, which are charadteri^ic of this species. B, Allops crassi' 
cornis, skuU Z', colledlion of 1888, type (Nat. Mus. 4289), South Dakota, Chad- 
dron formation, level lower C (in Hatcher's tabular lift of skuUs; later (1901) 
he was "inclined to place this higher up"). In this species the horns are 
thicker and the zygomata more massive than in A. serotinus. C, Allops sero- 
tinus, skull Ji, referred (Nat. Mus. 4945), Big Badlands, S. Dak., Chadron 
formation, level not lifted. The horns are more pointed at the tip than in 
other specimens of Allops 




f >-. 
— fi u 

m'fi 

> g C R i? 
~ » c ;- '3 

'C .S J= S S 

§ I ^1 

to o 3 -^ '' 



E E 



[0 2 

^ "m "*" » -a 
CO .S «J c a 



^ t< * 5j » 

1 3 i.1 § 

. 3 03 . u 

i ifl i 

ft S S fi 

I -^ -^ ^ ! 
.2 a,,. S " 



< I'i 



a. a 



» S s S 

< I'c !:•£ 

^ -s "3 - ; 

M .0 g ffl B 

>^ Ji « 6 

" ^•al 2 
^ X S fi « 

Cm rt tl * 
o 2 "3 a g 
^-° c u 

" » u -C J3 
tC ^ ^ ^ 13 
J^ - « S 

2. u - c 
2 E H js 

^ -5 ft a J3 
<il j: j3 j: .-, 



PLATE GXXI 



PLATE CXXI 

Type Skull of Allops crassicornis 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 517) 

Nat. Mus. 4289, skull Z', type, Big Badlands, S. Dak., Chadron formation, level lower C. A photograph of the specimen as it 
actually appears in the crushed condition is shown in Plate CXX, A. The aspect of the skull in this plate illustrates the 
mingling of Brontops and Menodus characters seen in this genus. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXXI 



0- 






TYPE SKULL OF ALLOPS CRASSICORNIS 



PLATE GXXII 



PLATE CXXII 

Skull, Jaw, and Teeth of Allops ceassicobnis? seu marshi 

A\ Skull, right side; A^, right upper dentition; B', lower jaw, inner view of left half; B^, lower dentition, crown view. Figures of 
skull one-fifth natural size; figures of teeth one-half natural size. (See p. 508.) 

British Mus. 5743M, female skuU and jaws referred to Allops crassicnrnis? , one individual. The premolar pattern is essentially 
identical with that of the type A. crassicornis, although the resemblance is obscured by the greater wear, which makes the 
internal cusps appear more circular in the present specimen. The small horns, slender zygomata, and rather small canines 
imply female sex. The horns are very different from those of the male Allops, but so also are those of the supposed female 
of A. serotinus. The evidence offered by the premolars seems to justify the reference of this skull to an advanced stage of 
Allops. 

In the side view of the skull the tapering nasals suggest reference to a small species of Bronlotherium, and this idea is at first rein- 
forced by the general appearance of the premolars, which have large circular postero-internal cusps. But reference to Bronto- 
iherium is excluded by the presence of sharp external and internal cingula on the premolars, by the narrowness of the molars^ 
and by the depth of the jaw in the region of the symphysis. 

This specimen is important on account of its association with a lower jaw and with a considerable part of the skeleton. (See p. 680.) 



S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXXII 





SKULL, JAW, AND TEETH OF ALLOPS CRASSICORNIS7 SEU MARSHI 



PLATE GXXIII 



PLATE CXXIII 

Type Skull of Menodus heloceeas 

A', Palatal view; A^, top view; A^, lateral view, left side; A^, posterior view. One-fourth natural size. (See p. 524.) 
Am. Mus. 6360, type. Horsetail Creek, northeastern Colorado, Chadron formation. This very imperfect type skull e.xhibits well 
the most primitive known stage in the generic characters of Menodus, such as the length and slenderness, the unexpanded 
zygomata, and the narrow parietal crest. The horns also, although in an early phylogenetic stage, are trihedral in basal 
section, as in Menodus trigonoceras, from which the nasal region is restored. The palate is unnaturally widened by crushing. 
The teeth so far as preserved present nothing noteworthy. The basicranial region is elongate. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 65 PLATE CXXIII 




TYPE SKULL OF MENODUS HELOCERAS 



PLATE GXXIV 



PLATE CXXIV 

Skulls of Menodus heloceras and Brontotherium hatcheri 

Plate prepared under the direction of Professor Cope. One-fourth natural size. (See p. 524, also PL CXXIII) 

A>, A', A3, Type skull of Menodus heloceras (Am. Mus. 6360), Horsetail Creek, northeastern Colorado, Chadron formation. 
B, Brontotherium hatcheri (Am. Mus. 6353), Colorado, Chadron formation. This palate was originallj' referred by Cope to "Sym- 
borodon acer," but because of the presence of incisors it is referred to Brontotherium bj' the author. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXXIV 




Ai ^'m^p 



SKULLS OF MENODUS HELOCERAS, TYPE, AND BRONTOTHERIUM HATCHERI 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXXV 




JAWS OF MENODUS PROUTII AND MENODUS TORVUS 

, Menodus proutii, type (Nat. Mus. 113) (genotype of Tit another ium Leidy, see pp. 205, 527), "Mauvais Terres * * * 
of White River," Chadron formation. One-third natural size Generic agreement with the genotype of Menodus 
giganteus Pomel (see p. 204) is seen in the elongate molars with sharp external cingula. The jaw fragment and teeth 
also agree in generic characters with the jaws Am. Mus. 1007, 1067 (see PI. CXXXIV, A'), referred to Menodus 
Menodus proutii, as represented by this jaw and by the skull Am. Mus. 9335, has the dimensions of 
as and is much sm,aller than M. giganteus. A-, M. proutii, imperfedl third low^er molar of type (Am. 
evi/. (See A'.) One-half natural si^e. The tooth shows w^ell the heavy external cingulum 
Menodus torvus, type of Symhorodon torvus Cope (Am. Mus. 6365), Horsetail Creek, north- 
ea^em Colorado, Chadron form.ation. One-fourth natural size. This is the true genotype of Symborodon, because 
Cope in his original description of Symborodon torvus (see p. 211) based the generic description primarily upon this 
jaw. It shows generic affinity with Menodus giganteus, and hence Symhorodon becomes a synonym of Menodus. 
The incisors were wanting or ve^igial, the premolars and molars were somewhat elongate v/ith heavy external 
cingula, and the general form, of the ja-w recalls that of Menodus. The front view (B^) shows the edentulous i 
border. In size this type of M. torvus is intermediate between M. heloceras and M. proutii 



trigonoceras 
M. trigonoci 
Mus. 113) 
of Menodi 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXXVI 





JAW REFERRED TO MENODUS PROUTU (TRIGONOCERAS) 

Onc'third natural size. (See p. 524.) British Mus, 4447M. The measurements of this specimen (mi-ms 245 mm.) indicate close 
specific reference to M. proutii, and it might almo^ be seletfled as a neotype of that species. This jaw exhibits clearly the chief 
characters of Menodus, The cheek teeth are elongate anteriorly, w^ith heavy external cingula; the canine is compressed and 
conical with heavy external cingulum; the chin is fuller than in Brontotherium and Brontops. Compare with M. proutii 
(PI. CXXV), M. trigonocerasl (giganteus?) (PI. CXXXIV), and M. giganteus (PI. CXLI) 



PLATE GXXVII 



PLATE CXXVII 

Type Jaw of Menodus torvus 

Plate prepared under the direction of Professor Cope. One-half natural size. (See p. 210, 527, also Pis. CXXV, CLVIII, and 

CLIX) 

Type jaw of Symborodon tortus Cope and genotype of Symborodon Cope (Am. Mus. 6365), Horsetail Creek, northeastern Colorado, 

Chadron formation. 
Cope recognized the relationship of this animal to the species named by him "Symborodon trigonoceras." This jaw is a little smaller 

than in the tj-pical M. trigonoceras. 



0. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXXVII 




PLATE GXXVIII 



PLATE CXXVIII 

Type Skull of Menodus tbigonoceras 
Plate prepared under the direction of Professor Cope. One-third natural size. (See p. 529, also PI. CLVII) 

Am. Mus. 6355, type. Horsetail Creek, northeastern Colorado, Chadron formation. Generic characters are displayed in the 
top view as follows: The skull is elongate, the zygomatic swellings are moderate (in the lithograph these swellings are too 
heavy and broad). The nasals are elongate and distally square. The horns are trihedral at the base, the tips pointed and 
recurved; the horns are directed outward and upward; the connecting crest is developed in the plane of the posterior face of 
the horns; the middle of the occiput is not extended behind the condyles. 

A study of the table of measurements (p. 523) indicates that the lower jaw of this skull would very probably have dimensions 
slightly larger than those of the type jaw of M. proutii; it is quite possible that M. trigonoceras may prove to be specifically 
identical with M. ■proutii and therefore a synonym. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXXVIII 





TYPE SKULL OF MENODUS TRIGONOCERAS 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXXIX 




TYPE SKULL OF MENODUS TRIGONOCERAS 



Am. Mus. 6355, Horsetail Cree 
This viewT shows the i 



:, northea^ei 
scored parts i 



1 Colorado, Chadron formation. One-fourth 

I outline. (Compare Mewodus giganteus, PI. CXL) 



PLATE GXXX 



PLATE CXXX 

Skull of Menodus trigonoceras (Cotype) and a Radius Doubtfully Referred 

Plate prepared under the direction of Professor Cope. One-third natural- size. (See pp. 214, 529) 

Ai, A2, Am. Mus. 6356, cotype. Horsetail Creek, northeastern Colorado, Chadron formation. This skull shows the elongate, 

distaUy expanded nasals and the trihedral, outward-directed horns of Menodus. 
Bi, Bj, Left radius, front and distal views of the lower portion. The reference of this fragment to M. trigonoceras is doubtful. 



MONOGRAPH 55 PLATE CXXX 



U. S. GEOLOGICAL SURVEY 




SKULL OF MENODUS TRIGONOCERAS, COTYPE, AND A RADIUS DOUBTFULLY REFERRED 



PLATE CXXXl 



PLATE CXXXI 

Palate of Menodus trigonoceras 

Plate prepared under the direction of Professor Cope. One-half natural size. (See p. 529) 

Am. Mus. 6356, Horsetail Creek, northeastern Colorado, Chadron formation. Upper teeth of cotype skull. The teeth are more 
accurately figured in Plate CXXXII, B, but this lithograph shows well enough the anteroposteriorlj' elongate molars and the 
progressive premolar cusps of Menodus. The external cingula should have been more clearly defined. The incisor alveoli 
are represented too large, as the specimen shows that the roots of the incisors were small, as in Menodus. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXXXI 





PALATE OF MENODUS TRIGONOCERAS 



PLATE GXXXII 



101959— 29— VOL 2 23 



PLATE CXXXII 

Upper Molar Series of Menodus and Allops 
One-half natural size. (See p. 523) 

A, Molars referred to Menodus tonus?, Carnegie Mus. 558 or 3068. Ttiis specimen represents a stage between the more primitive 

M. heloceras and the more advanced M. proutii {trigonoceras) . It is considerably larger than M. heloceras; it is smaller and 
much more primitive than M. proutii (trigonoceras), in the form both of its premolars, which have small postero-internal 
cusps and rounded internal borders, and of its molars, which are narrower. The dimensions of the teeth approach those which 
we might expect to find in the upper teeth of M. torvus, a form hitherto known only from the lower jaw. 

B, Upper teeth of Menodus trigonoceras, cotype (Am. Mus. 6356), Horsetail Creek, northeastern Colorado, Chadron formation. The 

drawing shows the alveoli for the vestigial incisors, the conical cingulate canine, the sharp cingula of the premolars and molars , 
the weU-developed postero-internal cusps of the premolars, joined by narrow ridges to the antero-internal cusps, and the 
relatively elongate molars — all characteristic of Menodus. 

C, Molars referred to Allops serotinus (Am. Mus. 520), Cheyenne River, S. Dak., Chadron formation. The teeth of Allops differ 

from those of Menodus as follows: One or two upper incisors persist on each side; the canine is flattened anteroposteriorly, 
the premolars and molars are wider transversely, the postero-internal cusps of the premolars in A. serotinus are less distinct 
than in Menodus. In all these characters, except the flattening of the canine, Allops resembles Brontops. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXXXII 






UPPER MOLAR SERIES OF MENODUS AND ALLOPS 



PLATE GXXXIII 



PLATE CXXXIII 

Upper Teeth of Menodus and Bkontotherium 
One-half natural size. (See pp. 523, 554) 

A, Menodus giganteus, neotype (Am. Mus. 505), South Dakota, probably Cheyenne River, Chadron formation. 

B, Brontotherium gigas (Am. Mus. 492), Cheyenne River Badlands, S. Dak., Chadron formation. 
The chief contrasts shown in these figures are as follows: 



Brontotherium (B) 



Upper incisors 

Canines 

Premolars 

Internal border of p^ 

External V's of fourth premolar 

External cingulum of premolars 

Internal cingulum of premolars 

Protostyles of premolars and molars. 
Hypocones of m', m^ 

Proportions of molars 

External cingulum of molars 

Internal cingulum of molars 



Vestigial or none 

Long, conic 

Transversely narrower, less molariform, 
with postero-internal cusps connected 
with antero-internal cusps by a long, 
slender bridge. 

Rounded 

Barely suggested 

Very sharp 

Sharp 

Feeble or absent 

Less protuberant internally 

More elongate anteroposteriorly 

Sharply defined 

Delicate, more or less continuous 



Two incisors, large and typically flat 
topped. 

Short, recurved, swelling, with massive 
posterior cingulum. 

Very broad, more molariform, postero- 
internal cusps very large, subcircular, 
and barely if at all connected with 
antero-internal cusps. 

Subquadrate. 

Well defined. 

Obsolete. 

Massive or partly confluent with base of 
crown. 

Very large. 

More protuberant. 

Wider. 

Obsolete. 

Discontinuous — that is, lacking opposite 
protocones. 



J4 




^■•S S.S 




PLATE GXXXV 



PLATE CXXXV 

Skulls of Menodus trigonoceras? and Menodus giganteus 
One-fourth natural size. (See pp. 523, 528, 531) 

A, Skull of female Menodus trigonoceras? (Am. Mus. 1067), Hat Creek Badlands, Nebr., Chadron formation. Front view of dis- 

torted skuU with associated lower jaw. This specimen is smaller than the supposed female of M. giganteus (B), but it is a 
little larger than the cotype of M. trigonoceras. It may be a female of M. varians, which it somewhat resembles in the right 
horn. The incisive borders were apparently edentulous. The slender canines are cingulate. The thick horns are trihedral 
in basal section, the left horn badly distorted. The nasals are wide distaUy, with two small processes on each side of a median 
notch. These processes with the whole rugose end of the nasals probably served for the attachment of a powerful levator 
labii superioris muscle. The buccal swellings of the zygomata are small. 

B, Skull of female of M. giganteus (Am. Mus. 506), Big Badlands, Cheyenne River, S. Dak., Chadron formation. This skull illus- 

trates well the effect of lateral crushing. Slender proportions are shown by the horns, canines, and zygomata. In dental 
measurements this skull is somewhat inferior to those in the males of M. ingens. 



PLATE CXXXVI 



PLATE CXXXVI 

Skull of Menodus giganteus 

One-fourth natural size. (See pp. 523, 531, also PI. CXXXV) 

Am. Mus. 506, female, Big Badlands, Cheyenne River, S. Dak., Chadron formation. Crushed skull and associated lower jaw. 
The horns are much smaller than in the males of this species. The buccal swellings of the z3'gomata were very slight. The 
dental characters noted in Plate CXXIV are also shown here. The contour of the lower jaw, although somewhat distorted, 
conforms in general with that of Menodus giganteus. (See PI. CXXXIX.) 



^^^^^^^^^^^^^Hr ' . ^'^^^^^^^^•->^>^^«K 


^H 


^^^^^^^^^^^^^V .Jh^^^It ' _J^^^M^ 


i^^'^*" ^^^^^^^^^^1 




B^'~ i^^^^^^H 




Ef^& .'-'' ^^^^1 


^^^^^Pr' ' ^<^l^Hi^^^ ^ / -^'^JRd 


^^> "^ yi^^^^^^^^^^l 


^^^^^^^^^Hr 'S^B^^^^^^^^ 1 ji^-^U. ^ .fjaMlVrSIE^ 


ne^s» I'l^^^^H 


^^^H/^-'wHRi^T t^^^::$^S^ 


1 ^^^ 'l^^^^l 




InHr l^^^^H 


^^^^^^^^^^^^■mH^h^^^HBco- .^^^ v'^^^^^i^Bsv/' >^^^^Kv^^^MHife^'<^ 


Ji^ ' ' (^^H 


^^^^^^^^^^K^^^I^^HkiCsb^j^ 


flVNtJ^i' v^^^H 




^3^ i^^^^H 


^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^H ▼ '^ 'jift ^^^^^^^V^ '^£^HM)'^^l^£^^Hi 


1 \ J fl^^H 


^^^^^^r ''.^S^^ H^l 


T ' ^'^^M 




V \{ - 'l^^l 


^^^^^^^^^^^^^K^^^^'^^ ^^^^^t^^^ . ^^^^HK^^ 


^B 




._,,,i 


^' Sk^' ^ ^ l^ii^^^SMHIite 


^ &. ^^jj^^^^^^^i 




^Hjl^^^l 


■IhI^^I 


^^1 



PLATE GXXXVII 



101959— 29— VOL 2- 



PLATE CXXXVII 

Skull of Menodus giganteus 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. Side view. (See pp. 523, 532) 

Type of "Brontotherhim ingens" Marsh (Yale Mus. 12010), Colorado, Chadron formation. This fine skull, which Marsh erroneously 
referred to Broniotherium, possesses aU the distinctive characters of Menodus, among which may be noted the following: Skull 
dolichocephalic, with slight zj'gomatic expansion and narrow parietal vertex; horns divergent, trihedral in basal section, con- 
necting cr&st at back of horns; nasals elongate, wide distally; cheek teeth with sharp external cingula. 



PLATE CXXXVIII 



PLATE CXXXVIII 

Skull of Menodus giganteus 
Plate prepared under the direction of Professor Marsh. One-fourth natural size. Top view. (See pp. 523, 532 and PI. CXXXVII) 







a "^ 






■S S S 






2 ^ C 
in o -^ 

"■SI 



S J 2 



? ? 



* I s 





U. S. GEOLOGICAL SDRVET 



MONOGRAPH 55 PLATE CXL 




SKULL OF MENODUS GIGANTEUS 

One-fourth natural size. (See pp. 523, 532, also PI. CXXXIX.) Field Mus. P 5927, male, Phinney Springs, S. Dak., Chadron 
formation, uppermost level. The front vie-wr sho^vs the great length and tapering character of the horns, the fundtionally 
edentulous incisive borders, the great v^idth of the nasals, and the moderate expansion of the zygomata 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 65 PLATE CXLI 




SKULL OF MENODUS GIGANTEUS 

One-fourth natural size. (See p. 532.) Am. Mus. 505, male. Big Badlands, S. Dak., probably Cheyenne River, Chadron 
formation. This skull exhibits the dental and cranial charadlers of Menodus giganteus as described above. (See 
Pis. CXXXIII, GXXXVII, CXXXVm, and CXXXIX.) The horns show well the pointed tips and flattened external 
base. The basicranial region sho-ws the alisphenoid canal, the foramen ovale, the paired and median protuberances 
for the redli capitis antici muscles, the periotic bone, the proximal end of the stylohyoid, and the condylar fo 



PLATE GXLII 



PLATE CXLII 

Type Skull of Menodus varians 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 536) 

Menodus (Menops) varians Marsh, type (Yale Mus. 12060), South Dakota, Chadron formation. This skull is referred to Menodus 
on the following grounds: It is dolichocephalic, with little zygomatic expansion; the incisors are reduced or vestigial; the 
canines are conical and cingulate; the cheek teeth are elongate anteroposteriorly, and the premolars and molars have the 
chief characters described above (PI. CXXXIII) ; the nasals are long and distally broad, and the horns are trihedral at the 
base; the dental and other measurements in general agree with those of Menodus giganteus. Noteworthj' features of the den- 
tition are the presence of an accessory cusp behind the postero-internal cusp of p'' (a rare or unique character); the great size 
and depth of the medifossettes of the premolars; the advanced development of the protostyles of the molars; the quadrate 
character of m'', which has the weU-developed hypocone surrounded posteriorly by a broad cingulum. These characters, 
together with the thickened tips of the horns, the greater width of the skull, and the widened base of the nasals, appear to 
indicate a mutation or species distinct from the typical M. giganteus {"ingens"). From Allops crassicornis (see PI. CXX) the 
present form is distinguished by its more pronounced dolichocephalic characters, especially in the dentition. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXLII 




TYPE SKULL OF MENODUS VARIANS 



PLATE GXLIII 



PLATE CXLIII 

Type Skull of Megacerops copei 
Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See pp. 543, 548, also PI. CXLIV) 

Skull V, type (Nat. Mas. 4711), Big Badlands, Indian Draw, S. Dak., Chadron formation, level upper C (Hatcher). Ai, Side 

view; A2, palate. 
The vertical position of the horns in this figure is largely due to crushing. A more accurate representation of the teeth is given in 

Plate CXLV. 
This skuU was referred by Marsh to Menodus { = Diconodon) monlanus but differs generically from the type in the nasals, premolars, 

and molars. 
Characteristic Megacerops features observed in this skuU are the extremely short face, upturned premolar series, small and closely 

approximated canines, horns of subcylindrical basal section without connecting crest. Affinity with Brontotherium is seen in 

the general contour of the skull, midparietal convexity, expanded zygomata, decurved nasals, advanced submolariform 

premolars, and short, bulbous canines. 



D. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE OXLIII 



/-' 



Ci. 



\y"^ 






-po 



Ai 



TYPE SKULL OF MEGACEROPS COPEI 



U. S. GEOLOGICAL SURVEY MONOGRAPH 55 PLATE CXHV 




SKULLS OF MEGACEROPS COPEI AND M. BUCCO 

One-fourth natural size. (See pp. 543, 544, 548, also Pis. CXLIO and CXLV.) A, Mega- 
cerops copei, skull V', type (Nat. Mus. 4711), Indian Draw, Big Badlands, S. Dak., 
Chadron foi-mation, level upper C (Hatcher). In this front view lateral crushing 
has emphasized the eredt position of the horns. The close appression of the zygO' 
mata to the brain case is also due to crushing. This skull differs from M. acer, espe-- 
cially in the horns. Figure prepared under the dirc(5lion of Professor Marsh. B, 
Female skull referred to 'Megacerops bucco, skull O^ (Nat. Mus. 4705), South Dakota, 
Titanotherium zone, upper levels of middle part (Hatcher). Figure prepared under 
the diretftion of the author. Vertical crushing has diminished the vertical and increased 
the transverse diameters of this specimen. Its resemblances to the male of M. hucco 
as well as to the female of M. acer may be noted in Plates CLV and CL, B 



PLATE GXLV 



PLATE CXLV 
Skulls of Megacerops copei and M. bucco 

One-fourth natural size. (See pp. 543, 544, 548, also Pis. CXLIII and CXLIV) 

A, Megacerops copei, skull V, type (Nat. Mus. 4711), Indian Draw, Big Badlands, S. Dak., Chadron formation, level upper C 

(Hatcher). The skull is crushed laterally, thus concealing its brachycephaly. The premolars are in an advanced stage; 
witness the large round postero-internal cusps and submolariform appearance of p*. The molars are very large, with large 
protostyles and jutting hypocones. The canines are short and stumpy. 

B, Female skull referred to Megacerops bucco, skuU 0' (Nat. Mus. 4705), South Dakota, Titanotherium zone, upper levels of middle 

part (Hatcher). A very old supposed female. P* very progressive, with a large mesostyle; premolars with reduced internal 
cingula; canines very small. Skull very brachy cephalic, zygomata widely expanded (unusual in females). 




101959— 29— VOL 2 25 



PLATE GXLVI 



PLATE CXLVI 

Skulls of Megacehops acer 
Side view. One-fourtia natural size. (See p. 545) 

A, Type (Am. Mus. 6348) , Horsetail Creek, northeastern Colorado, Chadron formation. A more perfect skull of this species is 

shown in Figure 452, p. 548. Horns elongate, subcylindrical in basal section, with low connecting crest. Affinity with 
Bro?itotkerium seen in general configuration of nasals, horns, and cranium. 

B, Referred female skull (Am. Mus. 6350; type of Symhorodon altirostris Cope), Horsetail Creek, northeastern Colorado, Chadron 

formation. Regarded by Osborn as a female of M. acer, but the specific reference is uncertain. The basal section of the 
horns and the general configuration of the skull are similar to M. acer, but the nasals are shorter; the horns are also short and 
without connecting crest. The front view (PL CL, B) shows well the high position of the nasals {"altirostris"), the alveoli 
for two small incisors, the large, rounded opening of the infraorbital canal. Affinity with the supposed female of M. bucco 
(figured in PI. CXLV) is indicated, especially in the top view (PI. CXLVIII) and in the palate (PI. CL; compare PI. CLII). 



V. S. GEOLOGICAL SURVEY 



MONOGHAPH 55 PLATE CXLVI 





■ 


^^^^^^K.i)/' 


(''i 9|H^^^^^^^^^^H 


^1 


M 


^^ ''^m 


1 


K 


_^ 


H 


^^E I'^^H 


^P^ " 'flhlHl 


mnffi" 


^^^RRP^^ 


IP"'"" X ^ jsl 




r'^H 


B ' ''^^HP 


1^^^ 


■^^^f K ^T'afe''^ 


11^; 


•'^ ^H^ 


M 


^^^^^ 


I^J 


^mm 


^ 


BB 


y^^ 


HI 


1^1 




SKULLS OF MEGACEROPS ACER 



PLATE GXLVII 



PLATE CXLVII 

Skulls of Megacerops acer 

Plate prepared under the direction of Professor Cope. Side view. One-fourth natural size. (See p. 545, also PI. CXLVI) 

A, Type (Am. Mus. 6348), Horsetail Creek, northeastern Colorado, Chadron formation. 

B, Female skull referred to M. acer? (Am. Mus. 6350), Horsetail Creek, northeastern Colorado, Chadron formation. This is 

Cope's type of Symborodon altirostris. The canine tooth, which has since been lost, was more robust than in the supposed 
female figured in Plate CXLV. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 65 PLATE CXLVII 





SKULLS or MEGACEROPS ACER 



PLATE GXLVIII 



PLATE CXLVIII 
Skulls of Megacerops acer 
Top view. One-fourth natural size. (See p. 545, also PI. CXLVI) 

A, Referred female skull (Am. Mus. 6350; type of Symborodon altirostris Cope), Horsetail Creek, northeastern Colorado, Chadron 

formation. (Compare with supposed M. bucco female, PL CLVI.) 

B, Type (Am. Mus. 6348), Horsetail Creek, northeastern Colorado, Chadron formation. The top of the skull is long and com- 

paratively narrow, as in males of Brontotherium. 



PLATE GXLIX 



PLATE CXLIX 

Skulls of Megacerops acer 

Plate prepared under the direction of Professor Cope. Top view. One-fourth natural size. (See p. 545, also PI. CXLVII) 

A, Female skull referred to M. acer? (Am. Mus. 6350; type of Sijmborodon altirostris Cope), Horsetail Creek, northeastern Colorado, 

Chadron formation. This figure is not as accurate as the photographic figures in Plate CXLVII. Affinity with M. bucco is 
indicated. 

B, Type (Am. Mus. 6348), Horsetail Creek, northeastern Colorado. 



PLATE GLI 



101959— 29— VOL 2 26 



PLATE CLI 

Female Skull Referred to Megacerops acer? (Type of Symborodon altirostris Cope) 

Plate prepared under the direction of Professor Cope. (See p. 545) 

Am. Mus. 6350, Horsetail Creek, northeastern Colorado, Chadron formation. Palatal view. One-third natural size. This figure 
is not as accurate as the photograph reproduced in Plate CLII, the characters of the teeth especially being very poorly por- 
trayed. Affinity with M. bucco is indicated. 



U. S. GEOLOGICAL SURVEY 



"MONOGRAPH 55 PLATE CH 




FEMALE SKULL OF MEGACEROPS ACER?, TYPE OF SYMBORODON ALTIBOSTRIS 



V. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLII 




FEMALE SKULL OF MEGACEROPS ACER? 

One-third natural size. (See p. S4S, also Pis. CXLVI, CXLVTI, and CLI.) Referred female skuU (Am. Mus. 6350; type of 
Symborodon altiro^ris Ckspe), Horsetail Creek, northeai4ern Colorado, Chadron formation. The dentition reveals the 
generic charadters mentioned in the explanation of Plate CXLV. Very clear evidences of close relationship with the 
skuU figured in Plate CXLV, A 



PLATE GLIII 



PLATE CLIII 

Type Skull of Megacerops btjcco and Jaw of Brontops? 
Plate prepared under the direction of Professor Cope. One-third natural size. (See p. 544) 

A., Megacerops {" Symhorodon") bucco, type (Am. Mus. 6345a), Horsetail Creek, northeastern Colorado, Chadron formation. The 
skull is poorly preserved and is vertically crushed, thus rendering its specific affinities somewhat uncertain. The figure is 
rather unsatisfactory, especially in the form of the horns and nasals. Of the two cotype skulls referred to in Cope's original 
description of Symborodon bucco this one is taken as the leototype, because upon it were based nearly all the measurements 
and specific characters Usted by Cope. 

B, Lower jaw of Brontops? (Am. Mus. 6345b), Horsetail Creek, northeastern Colorado, Chadron formation. Placed by Cope 
■nith the type skull of Megacerops {"Symborodon") bucco but probably pertains to the genus Brontops. It differs from the 
supposed true Megacerops figured in Plate CLVIII, A, in general contour and in the possession of well-developed external 
cingula on the cheek teeth. 

G, Undetermined jaw fragment from Colorado, referred by Cope to Megacerops {"Symborodon") bucco. It may pertain to Brontops. 
(See PL CLVIII, B.) 



PLATE GLIV 



PLATE CLIV 

Type Skull op Megacerops bucco 

Plate prepared under the direction of Professor Cope. One-half natural size. (See p. 544) 

Megacerops {" Symhorodon") bucco, type (Am. Mus. 6345a), Horsetail Creek, northeastern Colorado, Chadron formation. Basal 
view. The dentition reveals generic aflBnity mth the specimen figured in Plate CLII. The enormous size of the zygomata 
is alluded to in the name "bucco" ("one with swollen cheeks"). 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 56 PLATE CLIV 




TYPE SKULL OF MEGACEROPS ("SYMBORODON") BUCCO 



PLATE GLV 



PLATE CLV 

Type Skull op Megacerops bucco 

Plate prepared under the direction of Professor Cope. One-third natural size. (See p. 544) 

Megacerops {"Symborodon") bucco, type (Am. Mus. 6345a), Horsetail Creek, northeastern Colorado, Chadron formation. The 
skull is crushed down, this somewhat increasing its apparent breadth. The horns are poorly preserved but may have resembled 
those of the skuU figured in Plate CLVI. The enormous zygomata rival those of Brontotherium. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLV 




TYPE SKULL OF MEGACEROPS ("SYMBORODON") BUCCO 



D. S. GEOLOGICAL SURVEY 



MONOGBAPH 55 PLATE CLVI 




SKULL OF MEGACEROPS BUCC07 

Referred female skull, skull 0> (Nat. Mus. 4705), Big Badlands, S. Dak., Titanotherium zone, upper levels of middle part 
(Hatcher). One-fourth natural size. Vertical crushing has probably emphasized the breadth and flatness. This 
is regarded by the author as a female skull of M. bwcco 



PLATE GLVII 



PLATE CLVII 

Skulls of Menodus, Brontotheeium, and Megacerops and a Fragment of a Humerus 

Plate prepared under the direction of Professor Cope. One-third natural size. (See pp. 529, 546, also Pis. CXXVIII, CXXIX, 

and CLXXXIII) 

A, Menodus trigonoceras, type (Am. Mus. 6355), Horsetail Creek, northeastern Colorado, Chadron formation. Front view. Tliis 

species was referred by Cope to "Symborodon." 

B, Occiput referred to Brontotherium curtum? (Am. Mus. 6346), Horsetail Creek, northeastern Colorado, Chadron formation. 

This specimen was one of the cotypes of Cope's Symborodon bucco (see p. 212) but probably pertains to a large Brontotherium. 

C, Female skull referred to Megacerops acerf (Am. Mus. 6349), Colorado, Chadron formation. Occipital view of a small skull 

regarded b}' Osborn as a female of M. acer. 
Di, D2, Front and distal views of an undetermined humerus from the Chadron formation, Colorado. 



PLATE GLVIII 



PLATE CLVIII 

Jaws of Megaceeops eiggsi and Menodus torvus 
Plate prepared under the direction of Professor Cope. One-third natural size. (See pp. 527, 550) 

A, Megacerops riggsi, type (Am. Mus. 6364), Horsetail Creek, northeastern Colorado, Chadron formation. This jaw was referred 

by Cope to "Symborodon acer" but seems far too small to pertain to that species. It is in fact extremely small for an OUgocene 
titanothere. (See measurements, p. 242.) Reasons for referring it to Megacerops are the extremely short, thick horizontal 
ramus, low ascending ramus and coronoid, very short symphysis and premolars, molars with obsolete external cingula — 
characters which one would expect to find in the jaw of Megacerops. 

B, Undetermined jaw fragment referred by Cope to "S. bucco." (See PI. CLIII.) 

C, Menodus {"Symborodon") tonus, type (Am. Mus. 6365), Horsetail Creek, northeastern Colorado, Chadron formation. This 

jaw is the genotype of Symborodon (see p. 210), Cope's original description of the genus having been based upon this jaw, 
although in his subsequent descriptions he erroneously included other material (jaws and skuUs) belonging to other genera. 
The jaw seems to belong in the genus Menodus, as indicated by its general contour, measurements, absence of lower incisors, 
and robust external cingula on molars. The skulls which Cope referred to "Symborodon" appear not to be congeneric with 
this jaw but to pertain to Leidy's Megacerops. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLVIII 




^Vn\^ 




TYPE JAWS OF MEGACEROPS RIGGSI AND MENODUS TORVUS 



PLATE GLIX 



101959— 29— VOL 2 27 



PLATE CLIX 

Jaws of Megacerops riggsi and of Menodus torvus 
Plate prepared under the direction of Professor Cope. One-third natural size. (See pp. 527, 550, also PI. CLVIII) 

A, Lower jaw of Megacerops riggsi, type (Am. Mus. 6364), Horsetail Creek, northeastern Colorado, Chadron formation. This 

shows the extremely short thick rami and short symphysis. 

B, Unidentified jaw fragment from Colorado, referred by Cope to "Menodus prouiii," which suggests that he intended to include 

under that term titanotheres with two incisors on each side. 

C, Lower jaw of Menodus {"Symborodon") torvus, type (Am. Mus. 6365), Horsetail Creek, northeastern Colorado, Chadron forma- 

tion. This jaw, the genotype of Symborodon, shows the long, slender rami and long symphyseal region characteristic of 
Menodus jaws. 




o 








PLATE GLX 



PLATE CLX 

Jaws of Brontops sp. and of Megacerops riggsi 

Plate prepared under the direction of Professor Cope. One-third natural size. (See p. 550) 

A', A^, Jaw referred to Brontops? sp. (Am. Mus. 6345b), Horsetail Creek, northeastern Colorado, Chadron formation. Incomplete 

lower jaw, seen from below and from the inner side; placed b}' Cope with the skull of "Symborodon bucco" (see PI. CLIII) 

but probably pertaining to a different genus and species. 
B, Megacerops riggsi, type (Am. Mus. 63G4), Horsetail Creek, northeastern Colorado, Chadron formation. (See also Pis. CLVIII 

and CLIX.) This type lower jaw, seen from above, shows a short, thick symphysis, a short premolar series, and molars with 

external cingula obsolete — characters we should expect to find in Megacerops. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLX 




JAWS OF BRONTOPS SP. AND MEGACEROPS RIGGSI 



U, S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXI 




SKULLS OF BRONTOTHERIUM LEIDYI 

One-fourth natural size. (See p. 559.) A, Skull R, type (Nat. Mus. 4249), Big Badlands, S. Dak., Chadron formitiDn, level middle A. Figure prepared 
under the diredlion of Professor Marsh. This olde^ and mo^ primitive member of the Brontotherium phylum foreshadows the later bronto- 
theres in the flattened oval sedlion of the horns, in the tapering decurved nasals, and especially in the progressive chara(5ters of the premolir 
dentition. (See PI. CLXIII.) It is larger than the contemporary primitive members of the Brontops and Menodus phyla. B, Referred skull 
(Carnegie Mus. 93), Hat Creek, Nebr., Chadron formation, level lower A ("15 or 20 feet from bottom of lower beds" — ^J. B. Hatcher). Figure 
prepared under the diretflion of the author. This well-preserved skull has the horns a little longer than in the type. Male sex is indicated by the 
large size of the c 



PLATE CLXII 



PLATE CLXII 

Skulls of Brontotherium leidyi 
One-fourth natural size. (See p. 559) 

A, Skull R, type (Nat. Mus. 4249), Big Badlands, S. Dak., Chadron formation, level middle A. Figure prepared under the direc- 

tion of Professor Marsh. The nasals are long and tapering, the small horns are flattened posteriorly; the skull top is long, 
and the zygomata slender. 

B, Referred skull (Carnegie Mus. 93), Hat Creek, Nebr., Chadron formation, level lower A ("15 or 20 feet from bottom of lower 

beds" — J. B. Hatcher). This and the preceding skull show that the Brontotherium phylum was already well differentiated 
from the other genera at the period when the lower "Titanotherium beds" were laid down. 




"rt 13 u g " 
* u ■« t-i 



■ht 



1 .2 |3 ,; » 
! •- •? -S "3 



c 2 h ° ^ « _c 

d rt S* -w I- •£ 

■s -s s -s i ^ I 

« ^ "" S* ^ 

3 « 3 .b 3 U 
[I, Zi ^ rt (J '^ •£ 
g J3 £ •£ -S 3 « 

- -3 1 '^ I !j ■£ M 

•a BS ■^ 3!io^ 
■2 ^1 ff S'?!* 



,.^ H 



> « .2 






!<•? 



■S S 



! S g 
! ^1 



ft g -g -S 



- «i a c -i, a ^ -g 
« .2 ° a u *: S o 

S^l3g«efe^ 

■ K CL 3 k > "Tl 
SuMmOuOS 

" aB'-i; >ti2 

s « .a . "? g a ?" 

" £ J ^ ^ S *i 
< ^ 8 £ "'^ 

g = 2 -S g ifl < I 
^■S °<sjsh Sit 



8*^ 
# 



•M- S 



<? -S -S J ■ . „ 

mill 



' j3 -- a 




P > * .3 



T3 



•2 



a 2 



H 


S! s-S g 


^ 


i??i 


/. 


•J B U 





x'' s-S 


c! 




n 


b 


3 3 a 










^ 


S S g S 


< 


.1 j: . 


D 


arnegi 

ushed) 
ulum 


< 

D 


d skull (C 
e skull is 5 
he iftout i 
slightly cr 
ternal cing 




}> j: " ~^ X 




fe f^ =3 s :; 




^^itn 




^•Sl-H J 




S ^ !c C J! 










'*«' > " 




S.I > ^ S 








si ■^ 5 - " 




•2 ^ B 1 J 








•rt 1 O >.'^ 




i-'3M 








C'^ O a 




■£ a s J > 








1 i IJ § 





PLATE CLXV 



PLATE CLXV 

Skull of Brontotheritjm leidyi 

One-third natural size. (See p. 559) 

Referred skull (Carnegie Mus. 93), Hat Creek, Nebr., Cbadron formation, level lower A ("15 or 20 feet from bottom of lower 
beds" — J. B. Hatcher). This view exhibits the very progressive Brontotherium characters of the dentition in this oldest 
known member of the phylum, especially the massive swollen canines, the large incisors, the broad submolariform premolars 
with circular internal cusps, the vestigial external cingula, the blunt internal cingula, and the wide molars, with stout proto- 
styles and protuberant hypocones. In measurements the dentition nearly equals that in certain specimens of B. halcheri, 
and the basilar skull length is only 5 centimeters less 



U. S. GEOLOGICAL SUEVBT 



MONOGRAPH 65 PLATE CLXV 





1 


IBPi9l 


■ 


^■■1 




■ 


^^^^HiH^ l^U&^C^j wii'-i i^^S^^^^^^^I 


^1 


^^^^^^^H 




■ 




^1 


^^^^^^^^^^1 




1 


^^m|#H 


1 


^^^^H 




1^ 


^^Bi^^S 


1 


^^^^1 




N 


y^K^^H j^^^l 


1^1 


^^^^^^^^H 






^^i^^Kiq^BSSbEsI ^^Bk\ ^^H 


^^^1 


^^^^^^^^^^^^^^^^^^H 




Q 


wa^^SjSMm \ ^m^rM 


yWH 


^^^^^^^H^H 






yl^n»l«^ 


H 


H^^^^H 


^^^^^^m 

^^^^^^^m 


^ 


^^^Sk ^fv-d 


H 


I^^^^^H 


^mk.-f 


^ 


^^Hp '^^^ 


■ 


M ipPi i^^^^^^^l 




JflM 


I^^^^K»^m. W^ ^ pHfag^^Wk 


^m 


.t-^ «x j^n^^^^^^^^^H 


^^M^^ ':^ M 


jHJ 


^^w^F^ . iS^j'f^ 


^B 


^^^l^^^^^^^^l 




^^^T 


^fcl iw 


8 




^^v J^^^^tk ^ 


1 


wi 


1 




■^■i^^E^^^H 


1 


i^y m| 


1 


^^^^^^^^m '< '''-^'^^1 


1 /-^.'."i^^H 


^1 


■ jf ^PB^ro v^MB^K 


i.1 


^^^^^V '- ^^^H 






^f'l^ 






^^^^^^^^^^^ v 




^^^j'* * hT" * 


j: 






,^' 


^^iaL/ 


*<'^ 


^^^H 




^^Pt^$^* '' "*^ '''^^^It •" -"<|m 


^^^&£f°^ K 


"^^ 


i^'^^^^t v^'K^^^f'^B 


Pli 


M^^^^^^^^^m 


^^^^^^HHjl^ 


i 


^1 


1 





SKULL OF BRONTOTHERIUM LEIDYI 



--4€|-, 



PLATE GLXVI 



101959— 29— VOL 2 28 



PLATE CLXVI 

Type and Referred Specimens of Brontotherixjm hypoceras 
(See p. 562) 

A, Horn, type of Brontolherium hypoceras (Am. Mus. 6361), northeastern Colorado, Chadron formation. Natural size. Of the 
type of this species several skull fragments are all that remain; among these is the tip of the right horn, here shown in the 
front (A'), back (A-), and top (A') views. The horn has a compressed oval section, as in brontotheres, but is very small. 

B', SkuU referred to Brontoiherium hypoceras?, neotj'pe (Nat. Mus. 4273), skuU 1, Indian Draw, Big Badlands, S. Dak., Titanolherium 
zone, probably level upper A (Hatcher). One-third natural size. The horn tip closely resembles that of the type of B. 
hypoceras. B-, The same. Upper front teeth, natural size. The dental measurements are about the same as in B. leidyi, 
and this may represent a more progressive mutation from that species. The extreme forward position of the horns is empha- 
sized by crushing. The massive swollen canines and large incisor are of Brontolherium type. The inner incisors have probably 
dropped out as the animal grew old. 

C, Horn and nasal referred to Brontotherium hypoceras (Nat. Mus. 4940), skull m, Indian Draw, Big Badlands, S. Dak., Titano- 
lherium zone, level upper A (Hatcher). One-third natural size. This specimen is referred to B. hypoceras on account of the 
similarity in the horn and nasals. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXVII 





I 


■ 


Y^ 




1 


1 


1 1 


^^^^- 




^^P 


i^ifl 


^^^^HG^^ -^ 


1 


^ 


~''^^^^^l 




Ai 



TYPE SKULL OF BRONTOTHERIUM HATCHERI 

One-fifth natural size. (Sec p. 563.) Skull a, type (Nat. Mus. 1216), Big Badlands, S. Dak. C?Ck>rral Draw), Titano' 
therium zone, level upper B (Hatcher). Ai, Palatal view. The grinding teeth conform to the Brontotherium 
pattern, ^vhich, apart from size, appears to show little variation in the diflferent species. A3, Front view of 
horns. The horns in this animal are considerably shorter than those of the typical B. gigas (which is recorded 
from a higher level), and they are also less flattened, especially at the base; the connecfting creft is low^ 



n. S. GEOLOGICAL SDRVEY 



MONOGRAPH 55 PLATE CLXVIII 




TYPE SKULL OF BRONTOTHERIUM HATCHERI 
One-fourth natural size. (See p. 563.) Skull a, type (Nat. Mus. 1216), Big Badlands, S. Dak. (?Corral Draw), Titanotherium 
zone, level upper B (Hatcher). Vertical crushing may have emphasized the great breadth of this skull, and the horns have 
been crushed back above the orbits. The thick horns are not much flattened. The supratemporal creils have parallel 
sides, as in many Brontotheriinae, whereas in the Menodontinae they curve inward in the parietal region. The buccal 
swellings are almoS as large as in the moil advanced species of the genus 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXIX 




A, 




SKULL REFERRED TO BRONTOTHERIUM HATCHERI 

:e p. 563.) Am. Mxis. 1070, Hat Creek Badlands, Nebr., Chadron formation. A<, Upper dentition. 
One-third natural size. The generic characters are well shown; incisors large, canines swollen, 
premolars wide and submolariform -w^ith large circular internal cusps, external cingula absent, 
internal cingula thick and -wrinkled, molars wide, with large proto^yles and hypocones, cingula 
veitigial. The molars of the right side have been elongated by crushing. A^, Front view of skull. 
OnC'fourth natural size. . The horns are short and thick, with lowr connetfting cre^. The infra- 
orbital canal, as in other brontotheres, is subcircular. The dimensions of the ja-w belonging with 
this skull (PI. CLXVIII) do not differ greatly from those of the type jaw of Brontoth^ium gigas 
Mirsh 



PLATE GLXX 



PLATE CLXX 

Front View of Skulls of Three Brontotheres 
Plate prepared under the direction of Professor Marsh. One-sixth natural size. (See pp. 562, 563, 668) 

A, Skull referred to Brontolherium hypoceras? (Nat. Mus. 4702), skull K', South Dakota, doubtfully recorded from the lower part 

of the Tiianotherium zone (Hatcher). This skull has small slender horns, but the grinding teeth are as large as in certain skulls 
referred to B. gigas. The width across the zygomata is probably increased by flattening. 

B, Skull referred to Brontolherium hatcheri (Nat. Mus. 4255), skull Qi, South Dakota, Tiianotherium zone, level upper B (Hatcher), 

middle part of Titanotherium zone. The horns are a little longer than in the type of B. hatcheri but considerably shorter and 
less flattened than in the typical B. gigas. The zygomata are not shown. 

C, Skull referred to Brontotherium gigas (Yale Mus. 12061; type skull of Titanops elatus Marsh), South Dakota, exact locality not 

published, Chadron formation. The horns are long and the nasals narrow. This stage is characteristic of the upper part of 
the Tiianotherium zone (Hatcher). 



D. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXX 






FRONT VIEW OF SKULLS OF THREE BRONTOTHERES 



PLATE CLXXI 



PLATE CLXXl 

Type Jaw of Brontotherium gigas 
Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 568) 

Yale Mus. 12009, Colorado, Chadron formation. Ai, Outer side view; A2, front view; A3, crown view of dentition. This important 
genotype, described in June, 1873, together with other jaws and limb bones, formed the material which Marsh used in defining 
the genus Brontotherium, family Brontotheriidae, and it was also the first representative of the flat-horned species which were 
later described under the generic name Titanops. 

The skull subsequently named by Marsh Brontotherium ingens does not belong in the same genus with B. gigas and is now referred 
to Menodus. Generic characters shown in this jaw are the shallow chin, the swollen canines, the lack of external cingula on 
the premolars. Pi has been shed; the alveoli for incisors 1 and 2 of the left side are visible. The jaw belonged to a somewhat 
young adult. The measurements do not differ much from those of jaws called Brontotherium hatcheri by Osborn. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXXI 




TYPE JAW OF BRONTOTHERIUM GIGAS 



a. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXXII 




JAWS OF TWO SPECIES OF BRONTOTHERIUM 

le-fourth natural siie. (See pp. 565, 573.) A, Jaw referred to BrontotHeriutn hatcheri? (Am. Mus. 1070), Hat Creek badlands, Nebr., Chadron forma- 
tion. This ja^v belongs with the skull shown in Plate CLXIX. In the measurements of the teeth this specimen is a little larger than the type of 
B. gigas. The latter, how^ever, is curved upward in front and does not projedt poileriorly. B, Jaw referred to Brontothsrium medium? (Am. 
Mus. 1051), Cheyenne River badlands, S. Dak., Chadron formation. This huge jaw^ represents one of the largest knovi^n titanotheres. Its meas- 
urements are given on page 569. It differs from other known brontotheres in having an angulate chin and in the large size of the teeth 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXXIII 




SKULL OF BRONTOTHERIUM GIGAS ELATUS 

One-fourth natural size. (See p. 568, also PI. CLXXIV.) Skull referred to Brontotherium gigas (Am. Mus. 492), Cheyenne River badlands, 
S. Dak., Chadron formation. The horns are much longer than in B. hatcheri but not so slender as in the type of Titanops elatus. The width 
and flatness of the buccal expansion of the zygomata are somewhat increased by crushing 




•T3 h ^ -J- 



b! S 



2 H 









D 

H 

< 

►J 2 '■ 

w < ^ 



j5 0< 



s -2 



cd SI 

< s !■ - 

►J 
.3 mO 
d -S- 

■£ 9 S! 
;, i; to 

i, C -o _; 

. T) B 

11 is 



a3 j: Tj 












PLATE GLXXV 



101959— 29— VOL 2- 



PLATE CLXXV 

Skull of Brontotheritjm gigas elatus 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 570) 

Skull referred to Bronlotherium gigas (Yale Mus. 12061; type of Titanops elatus Marsh), South Dakota, Chadron formation, probably 
upper levels. The skuU is a little crushed downward in front. Marsh did not recognize the generic relationship of this animal 
■with liis Bronlotherium gigas type jaw, and therefore he established for it the new genus Titanops. In 1902 Osborn referred 
this skuU and jaw to Bronlotherium gigas, the type of which belonged to a somewhat younger animal. The present jaw is a 
little larger than the type B. gigas jaw, but this may possibly be due to greater age rather than to specific differences. 




*«^a!aseSB£j,kiia 



PLATE GLXXVI 



PLATE CLXXVI 

Type Skull of Brontotherium medium 

Plate prepared under the direction of Professor Marsli. One-fourth natural size. (See p. 573) 

Skull w, type (Nat. Mus. 4256), Big Badlands, S. Dak. (?Indian Draw), Chadron formation, level upper C. Ai, Front view; 
A2, palatal view. The specific name medium was given in allusion to the intermediate length of the nasals, which are shorter 
than in the tj-pe of elatus but longer than in type of curium. (See Pis. CLXXV and CLXXVII.) The horns are of the 
same general form as those of B. curium and B. platyceras, though not quite so fiat. The generic characters noted above 
(Pis. CXXXIII and CLXIII) are well shown in the dentition. 



U. S. GEOLOGICAL SURVEY 



^^. 



MONOGRAPH 55 PLATE CLXXVI 






• 1 v 



K" 





Ai 



TYPE SKULL OF BRONTOTHERIUM MEDIUM 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXXVII 




SKULL OF BRONTOTHERIUM CURTUM 

Plate prepared under the dirediion of Professor Marsh. One-fourth natural sise. (See p. 575.) SkuU q (Nat. Mus. 4946), Big Bad- 
lands, S. Dak. (TIndian Dra-w^), Chadron formation, level upper C (Hatcher). This skull is somewhat more progressive than the ■ 
type of B. medium, with broader zygomatic arches, flatter horns, and more abbreviated nasals. The dental series is a little smaller. 
The right horn has been broken off and healed during life (See PI. CLXXXV^^B) 



V. S. GEOLOGICAL STJRVEY 



MONOGRAPH 55 PLATE CLXXVIII 




TYPE SKULL OF BRONTOTHERItJM CURTUM 
Plate prepared under the direiftion of Professor Marsh. One-fourth natural size. (See p. S7S, also PI. Ca^XXIX.) Yale Mus. 12013, type 
of Brmtotherium C'Titanops-) curtum Marsh, Colorado, Chadron formation, upper beds (Hatcher). This iUge is very near the 
culmination of the BrorttotHerium phylum. The nasals are very short, the horns long and much flattened, the buccal expansions 
of the zygomata enormous. The incisor alveoli are not preserved in this specimen. The cheek teeth conform to the Brontotherium 
pattern already described. (See PI. CLXV) 



PLATE CLXXIX 



PLATE CLXXIX 

Type Skull of Beontotherium cuetum 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 575, also PL CLXXVIII) 

Yale Mus. 12013, type of Brontotherium {"Titanops") curium Marsh, Colorado, Chadron formation, upper beds (Hatcher). A], 
Side view. This shows the marked flattening and forward position of the horns, the shortness of the nasals, the high narial 
incisure, and the massive zygomata. A2, Front view of horns. It is possible that this species is the same as the Brontotherium 
platyceras of Osborn; if so the name platyceras would have priority over curtum. (See PI. CLXXXI.) 



S. GEOLOGICAL SURVEY 



MONOGRAPH 65 PLATE CLXXXII 




SKULL OF BRONTOTHERIUM PLATYCaERAS 
Photograph furnished by the Field Museum through the courtesy of Mr. E. S. Riggs. One-fifth natural size. (See p. 581, also PI. CLXXXI.) 
Referred skull (Field Mus. 12161), Phinney Spring, S. Dak., Chadron formation, upper beds. The top view shows well the very elongate 
ftraight-sided skull top and the widely divergent, very flattened horns with their high connediing creft. The zygomatic width and 
expansion are foreshortened in the photograph 



PLATE GLXXXIII 



PLATE CLXXXIII 

Skull of Brontotherium curtum and Horn of Brontotherium hypoceras 
Plate prepared under the direction of Professor Cope. (See pp. 212, 562, also Pis. CLXVI and CLXXXV) 

A, Skull referred to Brontotherium curtum (Am. Mus. 6346), cotype of Symborodon hucco Cope, Horsetail Creek, northeastern Colo- 
rado, Chadron formation. One-third natural size. This is one of the specimens upon which Cope founded his species 
Symborodon bucco, but it was not the leading specimen of the original description, and it appears to be generically distinct 
from the lectotype of that species. (See p. 212 and PI. CLV.) The reference to Brontotherium curtum is based especially 
upon the similarity in the zygomatic arches and cranial vertex. 

Bi, B2, B3, Fragment of horn, part of type of Brontotherium hypoceras (Cope) (Am. Mus. 6361), northeastern Colorado, Chadron 
formation. (Compare PI. CLXVI.) 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXXXIV 




(See pp. 
n gigas? 
Occiput 



SKULLS OF BRONTOTHERIUM GIGAS AND B. CURTUM 

Plate prepared under the diredlion of Professor Marsh. One-sixth natural size 
568, 575, also Pis. CLXXV and CXXXIX.) A, Skull referred to Brontotheri 
(Yale Mus. 12061?), South Dakota, Chadron fonnation, probably upper levels 
of type skull of Titanops elatus Marsh? As compared with B. curtum, the occiput 
narrow and the depression for the ligamentum nuchae is less extensive. B, BrontO' 
therium curtum, type (Yale Mus. 12013), Colorado, Chadron forxaation, upper beds 
(Hatcher). Occiput of type skull of Titanops curtus Marsh. The occiput is vi'ider, with 
a deep insertion area for the ligamentum nuchae and widely flaring creels for the re<5tus 
capitis lateralis muscles 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXXXV 




SKULLS OF BRONTOTHERIUM CURTUM 

One-fifth natural size. (See p. 575^ also Pis. CLXXVH and CLXXXIII.) A, Occiput referred to Brontotherium curtum (Am. 
Mus. 6346), Horsetail Creek, northea^em Colorado, Chadron: ormation. As remarked above (PI. CLXXXIII), this 
specimen is one of the cotyp^s seledled by Cope for his species of "Symhorodon bucco." The insertion areas for the liga- 
mentum nuchae and for the re(5tus capitis lateralis muscles are well shown. The skull is a little crushed downw/ard. 
B, Front view of skull referred to Brontotherium curtum? (Nat. Mus. 4946), skull q. Big Badlands, S. Dak. (TIndian Drawr), 
Chadron formation, level upper C (Hatcher). This view shows well the elevation of the nasal region in uncrushed 
skulls. The right horn has been broken off during life. The infraorbital canals are subcircular, as in other brontotheres 



PLATE GLXXXVI 



101959— 29— VOL 2 30 



PLATE CLXXXVI 

Skulls of Beontotheeium and Megaceeops 
Plate prepared under the direction of Professor Marsh. One-sixth natural size. (See p. 575) 

A, Brontotherium curtum, type (Yale Mus. 12013), Colorado, Chadron formation, upper levels. Front view. Type of Tiianops 

curtus Marsh. The long horns in this highly speciaUzed brontothere are much flattened in cross section and much larger than 
in the type of B. platyceras. The incisors are not preserved. 

B, Skull referred to Megacerops? sp. (Nat. Mus. 1220?). This skull appears to be allied to Megacerops tonus or M. acer. As in 

that genus, the nasals are short and thick, the horns elongate, slender, the zygomata massive and thick vertically. The 
presence of two incisors does not definitely exclude this specimen from Megacerops, because the type of M. acer retains alveoli 
for two small incisors. 

C, Type skuU of Brontotherium dolichoceras (Harvard Mus.), South Dakota, Chadron formation. In this specimen the horns are 

more slender and less widened transversely than in other brontotheres. They are suboval at the base. In Megacerops a 
connecting crest is usually lacking and the horns are thicker at the base. 



TJ. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXXXVI 




SKULLS OF BRONTOTHERIUM AND MEGACEROPS 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CLXXXVIII 




SKULL OF MALE BRONTOTHERIUM PLATYCERAS 

One-fifth natural size. (See p. 579, also PI. CLXXXIX.) Skull referred to Brontotherium platyceras (Am. Mus. 1448), Cheyenne 
River badlands, S. Dak., Chadron formation. This brontothere has the longed recorded horns and the shortest recorded 
nasals. The horns in size resemble those of B. curtum and are far larger than those of the type of B. platyceras. The width 
of the zygomata has been increased by vertical crushing, which has also lessened the height of the anterior narial opening 
and caused the horns to pitch sharply forward. The jaw here show^n (Am. Mus. 1051) may belong to some other species 



PLATE GLXXXIX 



PLATE CLXXXIX 

Skull of Brontotherium platyceras 

One-fifth natural size. (See p. 579, also PI. CLXXXVIII) 

Skull referred to Brontotherium platyceras (Am. Mus. 144S), Cheyenne River Badlands, S. Dak., Chadron formation, probably 
from the upper part of the Titanotherium zone. This side view shows the extreme length of the horns and height of the con- 
necting crest in this brontothere. Vertical crushing has pitched the horns forward, so that the nasals are now far in advance 
of the skuU. The zygomata also are displaced forward. The jaw (Am. Mus. 1061) does not belong with the skull. The 
relationship of this animal to B. platyceras and B. curtum is discussed above. (See PL CLXXXVIII.) 



PLATE GXG 



PLATE CXC 

Skulls of Female Brontotherium 
(See pp. 571, 576, also PI. CXCII) 

A, Female skull referred to Brontotherium? gigas? (Am. Mus. 1006), Big Badlands, S. Dak., Chadron formation. One-fifth natural 
size. This specimen was referred to the genus and species B. gigas on account of the similarity of the basal section of the 
horns, but the characters of the canines and premolars suggest reference to Brontops cf. B. robustus. 

fii, B^, Female skuU referred to Brontotherium curtum (Am. Mus. 1005), Big Badlands, S. Dak., Chadron formation. One-fifth 
natural size. This specimen is undoubtedly a brontothere, as shown especially by its premolars. The short horns are very 
massive at the base, and the connecting crest is high. Somewhat similar horns were described by Cope under the name 
Menodus peltoceras. (See p. 229.) 



V. S. GEOLOGICAL SURVEY 



MONOGRAPH 6S PLATE CXC 




SKXJLLS OF FEMALE BRONTOTHERIUM 



PLATE GXGI 



PLATE CXCI 

Skull of Male Beontotherium ramosum 
One-sixth natural size. (See p. 577) 

Am. Mus. 1447, tj'pe, Quinn Draw, Big Badlands, S. Dali., Chadron formation, upper levels. Ai, Side view; A2, front view. The 
specific name refers to the branching character of the horn tips. The horns are extremely broad, resembling those of the Field 
Museum specimen shown in Plate CLXXXI. Vertical crushing has flattened the skull and zygomata. The lower jaw (Am. 
Mus. 1062; Cheyenne River badlands, S. Dak., Chadron formation) does not belong with this skull. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXCl 




SKULL OF MALE BRONTOTHERIUM RAMOSUM 



101959— 29— VOL 2 31 



U. S. GEOLOGICAL SURVBV 



MONOGRAPH 55 PLATE CXCII 




SKULL OF FEMALE BRONTOTHERIUM CURTUM 

One-fourth natural size. (See p. 576, also PI. CXC.) Referred skull (Am. Mus. lOOS), Big Badlands, S. Dak., Chadron 
formation. The premolars and molars show the characters typical of Brontotherium, save that the poilero-intemal 
cusps of the premolars are not well separated from the antero-internal cusps. The canines are slender. The 
zygomata, though ^tout, are less expanded than in long-homed males 



TJ S. GEOLOGICAL STJRVET 



MONOGRAPH 65 PLATE CXCIII 





SKULL OF FEMALE BRONTOTHERIUM CURTXJM 
One-fifth natural sizie. (See p. 577, also PI. CXCIV.) Referred skuU (British Mus. S629). Ai, Top view; A', front view. This 
specimen so Wrongly resembles Cope's fragmentary type of Menodus peltoceras that it might be seleifted as a neotype. Kin- 
ship with the supposed female of B. curtum (PI. CXCII) is indicated especially in the horns, nasals, zygomata, and dentition; 
but the canines seem too large to belong to a female, so that possibly this and the supposed female together with Cope's type 
may represent a di^indt species of Brontothe- 



C. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXCIV 




SKULL OF FEMALE BRONTOTHERIUM CURTUM 

One-fourth natural size. (See p. 577, also PI. CXCIH.) British Mus. 5629. This palatal view shows well the long 
and the premolar charadters, ^vhich are similar to those of the specimen figured in Plate CXCIl 



PLATE GXGV 



PLATE CXCV 

Atlas of the Type of Brontops eobustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 667) 

Type (Yale Mus. 12048). 1, Lateral view, left side; 2, superior view; 3, anterior view; 4, inferior view; 5, posterior view, a, Facet 
for condyle of skull; 6, facet for axis; n, neural canal; s, neural spine; (, transverse process. 



U. S. GEOLOGICAL SURVEY 



M0N06EAPH 65 PLATE CXCV 








ATLAS OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GXGVI 



PLATE CXCVI 

Axis of the Type of Brontops robusttjs Marsh 

Plate prepared under the direction of Professor Marsh. One fourth natural size. (See p. 667) 

1, Lateral view, left side; 2, superior view; 3, anterior view; 4, inferior view; 5, posterior view, a, Facet for atlas;/, foramen for 
vertebral artery; n, neural canal; o, odontoid process; z', postzygapophysis. The principal muscle and ligament attachments 
of this vertebra are given on page 713. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXCVI 









AXIS OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GXGVII 



PLATE CXCVII 

Fourth Cervical Vertebra of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 667) 

1, Lateral view, left side; 2, superior view; 3, anterior view; 4, inferior view; 5, posterior view, a, Anterior facet of centrum; 
/, foramen for vertebral artery; n, neural canal; s, neural spine; t, transverse process; z, prezygapophysis; z', postzygapophysis. 
The principal muscle and ligament attachments of this vertebra are given on page 713. 



U. S. GEOLOGICAL SURVEY 



MONOGEAPH 55 PLATE CXCVII 








FOURTH CERVICAL VERTEBRA OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GXGVIII 



PLATE CXCVIII 

Second Dorsal Vertebra of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 667) 

1, Lateral view, left side; 2, superior view; 3, anterior view; 4, posterior view, a. Anterior facet of centrum; n, neural canal; p, 
posterior facet of centrum; r, facet for capitulum of second rib; r', facet for tuberculum of second rib; r", facet for capitulum 
of third rib; t, transverse process; z, prezygapophysis; z', postzygapophysis. The principal muscle and ligament attachments- 
of this vertebra are given on page 713. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CXCVIII 




SECOND DORSAL VERTEBRA OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE CXCIX 



PLATE CXCIX 

Tenth Dorsal Vertebra op the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 667) 

1, Lateral view, left side; 2, superior view; 3, anterior view; 4, inferior view; 5, posterior view, a, Anterior facet of centrum; n, 
neural canal; p, posterior facet of centrum; r, facet for capitulum of tenth rib; r', facet for tuberculum of tenth rib; r" , facet 
for capitulum of eleventh rib; s, neural spine; t, transverse process; z, prezygapophysis; z', postzygapophysis. 



U. S. GEOLOGICAL SURVEY 



MONOGKAPH 55 PLATE CXCIX 







TENTH DORSAL VERTEBRA OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GC 



PLATE CC 

Second Lumbar Vertebra of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 667) 

1, Lateral view, left side; 2, superior view; 3, anterior view; 4, inferior view; 5, posterior view, a, Anterior facet of centrum; 
n, neural canal; p, posterior facet of centrum; s, neural spine; t, transverse process; z, prezygapophysis; z', postzygapophysis. 
The chief muscle and ligament attachments of this vertebra are given on page 713. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATR CC 








SECOND LUMBAR VERTEBRA OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE CGI 



101959--29— VOL 2 32 



PLATE CCI 

Caudal Vertebrae of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 667) 

1, First caudal, lateral view, left side; 2, the same, anterior view; 3, the same, posterior view; 4, second caudal, lateral view, left 
side; 5, second caudal, posterior view; 6, third caudal, lateral view, left side; 7, the same, anterior view; 8, the same, posterior 
view; 9, fourth caudal, lateral view, left side; 10, the same, anterior view; 11, fifth(?) caudal, lateral view, left side; 12, the 
same, anterior view; 13, the same, posterior view; 14, seventh(?) caudal, lateral view, left side; 15, the same, anterior view; 
16, the same, posterior view; 17, tenth(?) caudal, lateral view, left side; 18, the same, anterior view; 19, twelfth(?) caudal, lateral 
view, left side; 20, the same, anterior view; 21, thirteenthC?) caudal, lateral view, left side; 22, fifteenth(?) caudal, lateral 
view, left side; 23, sixteenth(?) caudal, lateral view, left side, a, Anterior; c, hemapophysis; n, neural canal; p, posterior; 
s, neural spine; t, transverse process; z, prezj'gapophysis. 



U. S. GEOLOGICAL SUEVEY 
22 



MONOGRAPH 55 PLATE CCI 



^ 




14 


















CAUDAL VERTEBRAE OP THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGII 



PLATE ecu 

Second and Tenth Left Ribs of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 667) 

1, Second rib, outer view; 2, the same, anterior; 2a, the same, proximal end; 3, the same, inner view; 4, the same, posterior view; 
4a, the same, distal view; 5, tenth rib, anterior view; 5a, the same, proximal end, dorsal view; 5b, the same, proximal end, 
antero-internal view; 5c, the same, distal view; 6, the same, posterior view; 6a, the same, internal view; 6b, the same, distal 
end, internal view. 



XI. S. GEOLOGICAL SURVEY 



MONOGEAPH So PLATE CCII 




SECOND AND TENTH LEFT RIBS OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE CCIII 



PLATE CCIII 

Fourth Rib of the Type op Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 667)) 

1, Outer view; 2, anterior view; 3, inner view; 3a, inner view; 4, posterior view; 5, superior view of upper portion, h, Head of rib; 
i, tubercle of rib. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCIII 





FOURTH RIB OP THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGIV 



PLATE CCIV 

Left Scapula of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 668) 

1, Lateral or external view, left side; 2, medial or inner view; 3, posterior view; 4, inferior or distal view, a, Tuberosity for the 
trapezius muscle, running below into a ridge for the acromial portion of the deltoid; c, coracoid process with rugosity for the 
biceps and coracobrachialis muscle; ch, cervical or anterior border of the scapula, ending below in a tuberosity which was 
probably the origin of the caput longum tricipitis and infraspinatus secundus muscles; s, spine of the scapula; ss, suprascapular 
border with rugositj' for insertion of the rhomboideus muscle. The principal muscle attachments of the scapula are given 
on page 714. 




LEFT SCAPULA OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GCV 



PLATE CCV 

Left Humerus of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 668) 

1, Anterior view; 2, medial or inner view; 3, posterior view; 3a, section of shaft at narrowest point; 4, lateral or outer view; 
5, proximal end; 6, distal end. a, Distal articular surface; d, deltoid ridge; e, external condyle; h, head of humerus; 
i, internal condj'le; o, anconeal fossa; s, supinator ridge; t, great tuberosity; t' , small tuberosity. The principal muscle 
attachments of the humerus are given on page 715. 



U. S. GEOLOGICAL SURVEY 



MONOfiliiPH 55 PLATE CCV 







LEFT HUMERUS OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 




PLATE GGVI 



PLATE CCVI 

Left Radius of the Type of Beontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 668) 

1, Anterior view; 2, medial or inner view; 2a, section of shaft at point indicated by dotted line; 3, posterior view; 4, lateral or 
outer view; 5, proximal end; 6, distal end. h, Articular facet for humerus; I, articular facet for lunar; s, articular facet for 
scaphoid; ul, articular facet for ulna. The principal muscle attachments of the radius are given on page 716. 



U. S. GEOLOGICAL SURVEY 



MONOGKAPH 55 PLATE CCVI 
4 




LEFT RADIUS OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGVII 



PLATE CCVII 

Left Ulna of the Type of Brontops robustus Mersh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 668) 

1, Medial or inner view; 2, anterior view; 3, posterior view, c, Anconaeal process; h, great sigmoid cavity; o, olecranon; p, facet 
for cuneiform (pyramidal) ; ps, facet for pisiform. 2a, Distal end. The principal muscle attachments of the ulna are given 
on page 716. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCVII 






2a 




^ -Cj 



:.¥' 



LEFT ULNA OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGVIII 



PLATE CCVIII 

Left Scaphoid of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 668) 

1, Inner or radial view; 2, posterior view; 3, outer or ulnar view; 4, anterior view; 5, proximal surface; 6, distal surface. I, V, 
Facets for articulation with lunar; m, facet for articulation with magnum; r, facet for articulation with radius; td, facet for 
articulation with trapezoid. The ligamentous attachments of the scaphoid are given on page 716. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCVIII 









LEFT SCAPHOID OF THE TYPE OF BRONTOPS ROBUSTUS MAPSH 



PLATE GGIX 



101959— 29— VOL 2—33 



PLATE CCIX 

Left Cuneiform Carpi (Pyramidal) and Eight Pisiform of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 668) 

1, Left pyramidal bone, anterior view; 2, the same, medial or inner view (radial side); 3, the same, posterior view; 4, the same, 
lateral or outer view (ulnar side); 5, the same, proximal end; 6, the same, distal end. l-l', Facets for articidation with lunar, 
?-, facet for articulation with radius; ps, facet for articulation with pisiform; ul, facet for articulation with ulna; un, facet for 
articulation with unciform. 

7, Right pisiform bone, inferior view; 8, the same, lateral or outer view (ulnar side); 9, the same, superior view; 10, the same, 
medial or inner view (radial side); 11, the same, proximal end; 12, the same, distal end. p, Facet for articulation with 
pyramidal; id, facet for articulation with ulna. 

The ligamentous attachments of these elements are given on page 716. 



U. a GEOLOGICAL SURVEY 
1 



MONOGRAPH 55 PLATE CCIX 






Ul I 









12 





10 




LEFT CUNEIFORM CARPI AND RIGHT PISIFORM OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGX 



PLATE CCX 

Left Trapezoid and Left Magnum of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 668) 

1, Left trapezoid, anterior view; 2, the same, medial or inner view (radial side); 3, the same, posterior view; 4, the same, lateral 
or outer view (ulnar side); 5, the same, proximal end; 6, the same, distal end. m, m', Facets for articulation with magnum; 
mcll, facet for articulation with second metacarpal; s, facet for articulation with scaphoid. 

7, Left magnum, anterior view; 8, the same, medial or inner view (radial side) ; 9, the same, posterior view; 10, the same, lateral 
or outer view (ulnar side); 11, the same, proximal view; 12, the same, distal view. I, Facet for articulation with lunar; 
mcII, facet for articulation with second metacarpal; mcIII, facet for articulation with third metacarpal; mcIV, facet for 
articulation with fourth metacarpal; s, facet for articulation with scaphoid; id, td' , facets for articulation with trapezoid; wn, 
facet for articulation with unciform. 

The principal ligamentous and muscular attachments of these elements are given on page 716. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCX 









12 




rncll '"• 








LEFT TRAPEZOID AND LEFT MAGNUM OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXI 



PLATE CCXI 

Left Unciform of the Type of Brontops robustus Marsh 
Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 668) 

1, Anterior view; 2, medial or inner view (radial side); 3, posterior view; 4, lateral or outer view (ulnar side); 5, proximal view; 
6, distal view. I, Facet for articulation with lunar; m, facet for articulation with magnum; mcIII, facet for articulation with 
third metacarpal; tncIV, facet for articulation with fourth metacarpal; mcV, facet for articulation with fifth metacarpal; 
p, facet for articulation with cuneiform (pyramidal). The ligamentous and muscular attachments of these elements are given 
on pages 716, 717. 



U. S. GEOLOGICAL SURVEY 



MONOGEAPH 55 PLATE CCXI 









LEFT UNCIFORM OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXII 



PLATE CCXII 

Second Left Metacarpal of the Type of Brontops robustus Marsh 
Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 668) 

1, Anterior view; la, section of point indicated by dotted line; 2, medial or inner view (radial side) ; 3, posterior view; 4, lateral 
or outer view (ulnar side); 4a, outline of section at point indicated by dotted line; 5, proximal end; 6, distal end. m, Facet 
for articulation with magnum; mcIII, facet for articulation with third metacarpal; ph, facet for articulation with first 
phalanx; s, facet for articulation with sesamoid; td, facet for articulation with trapezoid. The principal ligamentous and 
muscular attachments of this element are given on pages 716, 717. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCXII 




SECOND LEFT METACARPAL OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXIII 



I 



PLATE CCXIII 

Third Left Metacarpal of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 668) 

1, Anterior view; la, section at point indicated by dotted line; 2, medial or inner view (radial side); 2a, outline of section at point 
indicated by dotted line; 3, posterior view; 4, lateral or outer view (ulnar side); 5, proximal end; 6, distal end. m, Facet 
for articulation with magnum; mcll, facet for articulation with second metacarpal; mcIV, facet for articulation with fourth 
metacarpal; ph, facet for articulation with first phalanx; s, facet for articulation with sesamoid; un, facet for articulation 
with unciform. The principal ligamentous and muscular attachments of these elements are given on pages 716, 717. 



U. S. GEOLOGICAL SURVEY 

5 



MONOGRAPH 55 PLATE CCXIII 




2a 




^11} 



THIRD LEFT METACARPAL OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXIV 



PLATE CCXIV 

Fourth Lkft Metacarpal of the Type of Brontops robustus Marsh 
Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 668) 

1, Anterior view; la, section at point marked by dotted line; 2, medial or inner view (radial side); 2a, outline of section at point 
indicated by dotted line; 3, posterior view; 4, lateral or outer view (ulnar side); 5, proximal end; 6, distal end. m, Facet 
for articulation with magnum; mcIII-mcIII' , facets for articulation with third metacarpal; mcV, facet for articulation with 
fifth metacarpal; p/i, facet for articulation with first phalanx; s, facet for articulation with sesamoid; un, facet for articulation 
with unciform. The principal ligamentous and muscular attachments of this element are given on pages 716, 717. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCXIV 




FOURTH LEFT METACARPAL OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GCXV 



PLATE CCXV 

Fifth Left Metacarpal of the Type of Brontops eobustus Marsh 
Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 668) 

1, Anterior view; la, section of shaft; 2, medial or inner view (radial side); 3, posterior view; 4, lateral or outer view (ulnar side); 
5, proximal end; 6, distal end. mcIV, Facet for articulation with fourth metacarpal; ph, facet for articulation with first 
phalanx; s, facet for articulation with sesamoid; un, facet for articulation with unciform. The principal ligamentous and 
muscular attachments of this element are given on pages 716, 717. 




MONOGRAPH 55 PLATE CCXV 
4 





la 







FIFTH LEFT METACARPAL OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXVI 



PLATE CCXVl 

Proximal Phalanges of Left Manus of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 668) 

1, Proximal phalanx of inner or second digit, anterior view; 2, the same, posterior view; 3, the same, lateral or outer view; 4, the 
same, proximal end; 5, the 'same, distal end: 6, proximal phalanx of third digit, anterior view; 7, the same, medial or inner 
view; 8, the same, posterior view; 9, the same, proximal end; 10, the same, distal end; 11, proximal phalanx of fourth digit, 
anterior view; 12, the same, medial or inner view; 13, the same, posterior view; 14, the same, proximal end; 15, the same, 
distal end; 16, proximal phalanx of outer or fifth digit, anterior view; 17, the same, posterior view; 18, the same, lateral or 
outer view; 19, the same, proximal end; 20, the same, distal end. The principal ligamentous and muscular attachments of 
this element are given on page 717. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH SS PLATE CCXVI 




PROXIMAL PHALANGES OF LEFT MANUS OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXVII 



101959— 29— VOL 2 34 



PLATE CCXVII 

Phalanges and Sesamoids of Manus of the Type of Beontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 668) 

Second phalanx of second(?) digit, lateral or outer view; 2, the same, anterior view; 3, the same, medial or inner view; 4, the 
same, posterior view; 5, the same, proximal view; 6, the same, distal view; 7, distal phalanx of second(?) digit, lateral or 
outer view; 8, the same, anterior view; 9, the same, medial or inner view; 10, the same, posterior view; 11, the same, pro.ximal 
or superior view; 12, the same, distal or inferior view; 13, metacarpal sesamoid of second (?) digit, lateral or outer view; 14, the 
same, anterior view; 15, the same, medial or inner view; 16, the same, posterior view; 17, the same, end view; 18, phalangeal 
sesamoid of second(?) digit, medial or inner view; 19, the same, posterior view; 20, the same, anterior view; 21, the same, 
proximal view; 22, the same, distal view. The principal ligamentous and muscular attachments of these elements are given 
on page 717. • 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 00 PLATli CCXVil 




















PHALANGES AND SESAMOIDS OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE CGXVIII 



PLATE CCXVIII 

Pelvis and Sacrum of the Type of Brontops robustus Marsh 

Plate prepared under th? direction of Professor Marsh. Front view. One-fourth natural size. (See pp. 668, 694) 

a, Acetabulum; c, crest of ilium; /, /, obturator fenestra; il, ilium; is, ischium; p, pubis; s, sacrum; z, prezygapophysis of first 
sacral vertebra. 



U. S. Cil'XJLOGlCAL SURVEY 



MONOdltAPlI M PLATE CCXVIIl 




PELVIS AND SACRUM OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXIX 



PLATE CCXIX 

Pelvis and Sacrum of the Type op Beontops eobustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See pp. 668, 694) 

1, Superior view; 2, inferior view, a, Acetabulum; c, crest of ilium; /, /, obturator fenestra; is, ischium; p, pubis; s, sacrum; 
s', caudo-sacral vertebra; z, prezygapophysis of first sacral vertebra. 



n. 8. GEOLOGICAL SURVEY 



MONOGEAPH 86 PLATE CCXIX 




PELVIS AND SACRUM OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXX 



PLATE CCXX 

Left Femur of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsli. One-fourth natural size. (See p. 668) 

1, Anterior view; la, section of shaft at break near middle; 2, medial or inner view; 3, posterior view; 4, proximal end; 5, distal 
end. e, Outer condyle; h, head of femur; i, inner condyle; ic, intercondylar notch; I, pit for ligamentum teres; p, facet for 
articulation with patella; t, great trochanter; t', lesser trochanter; t", third trochanter. The principal ligamentous and 
muscular attachments of the femur are given on page 719. 



MONOGRAPH 55 PLATE CCXX 



U. S. GEOLOGICAL SURVEY 




LEFT FEMUR OF THE TYPE OF BRONTOPS ROliUSTUS MAKSH 



PLATE GGXXI 



PLATE CCXXI 

Left Tibia of the Type op Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 669) 

1, Anterior view; la, section of shaft at break near middle; 2, medial or inner view; 3, posterior view; 4, lateral or outer view, 
showing fibular side; 5, proximal end; 6, distal end. a, Facet for articulation with astragalus; e, facet for articulation with 
outer condyle of femur; /, /', facets for articulation with fibula; i, facet for articulation with inner condyle; p, procnemial 
ridge; s, spine of tibia. The principal ligamentous and muscular attachments of the tibia are given on page 720. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCXXI 




LEFT TIBIA OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXXII 



PLATE CCXXII 

Left Patella and Left Fibula of the Type of Brontops robustus Marsh 
Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 669) 

1, Left patella, anterior view; 2, the same, medial or inner view; 3, the same, posterior view; 4, the same, lateral or outer view. 

/, Facet for articulation with femur. 
5, Left fibula, lateral or outer view; 5a, the same, section of shaft at break near middle; 6, the same, anterior view; 7, the same, 

medial or inner view; 8, the same, posterior view, a, Facet for articulation with astragalus; c, facet for articulation with 

calcaneum; t, t', facets for articulation with tibia. The principal hgamentous and muscular attachments of the fibula are 

given on page 720. 



U. S. GEOLOGICAL SURVEY 

1 





5a. 



A 




MONOGRAPH 65 PLATE CCXXII 
3 




i?,i 



LEFT PATELLA AND LEFT FIBULA OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE CGXXIII 



PLATE CCXXIII 

Left Astragalus of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 669) 

1, Anterior view; 2, medial or inner view (tibial side) ; 3, posterior view; 4, lateral or outer view (fibular side) ; 5, proximal view; 
6, distal view, c, Ectal; c', sustentacular; c", distal facets for calcaneum; ch, cuboid facet; /, fibular facet; n, navicular 
facet; t, tibial facet. The ligamentous attachments of the astragalus are given on page 724. 



U. S. GEOLOGICAL SDRVEY 



MONOGRAPH 55 PLATE CCXXin 









LEFT ASTRAGALUS OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXXIV 



PLATE CCXXIV 

Left Calcaneum of the Type of Brontops eobustus Marsh 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 669) 

1, Anterosuperior view; 2, external view; 3, postero-inferior view; 4, proximal view; 5, distal view, a, Astragalar facet; o', sus- 
tentacular facet; a", distal astragalar facet; c6, cuboid facet; /, fibular facet; t, tibial facet. The principal ligamentous and 
muscular attachments of the calcaneum are given on pages 721, 722. 



U. S. GEOLOGICAL SURVEY 
5 



MONOGRAPH 56 PLATE CCXXIV 




LEFT CALCANEUM OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXXY 



101959— 29— VOL 2 35 



PLATE CCXXV 

Left Navicular and Left Cuboid of the Type of Brontops robustus Marsh 
Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 669) 

1, Left navicular, external view; 2, the same, internal view; 3, the same, posterior view; 4, the same, external view; 5, the same, 
superior view; 6, the same, inferior view; 7, left cuboid, anterior view; 8, the same, internal view; 9, the same, posterior 
view; 10, the same, external view; 11, the same, superior view; 12, the same, inferior view, a, Astragalar facet; c, calcanear 
facet; ec, ectocuneiform facet; m, mesocuneiform facet; mtlll, facet for metacarpal III ; mtIV, facet for metacarpal IV. 
The principal ligamentous and muscular attachments of these elements are given on page 724. 



U. S. GEOLOGICAL SURVEY 

4 








MONOGRAPH 55 PLATE CCXXV 
2 





^'^ mtm 








LEFT NAVICULAR AND LEFT CUBOID OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXXVI 



PLATE CCXXVI 

Left Ectocuneiform and Left Mesocuneiform of the Type of Brontops robustus Marsh 
Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 669) 

1, Left mesocuneiform, anterior view; 2, the same, internal view; 3, the same, posterior view; 4, the same, external view; 5, the 
same, superior view; 6, the same, inferior view; 7, left ectocuneiform, anterior view; 8, the same, internal view; 9, the same, 
posterior view; 10, the same, external view; 11, the same, superior view; 12, the same, inferior view, cb, Cuboid facet; 
ec, ec', facets for ectocuneiform; mtll, mtW , facets for metatarsal II; mtlll, facet for metatarsal III; n, navicular facet. 
The principal Hgamentous and muscular attachments of these elements are given on page 724. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCXXVI 












12 






LEFT ECTOCUNEIFORM AND LEFT MESOCUNEIFORM 
OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXXVII 



PLATE CCXXVII 

Left Fourth Metatarsal of the Type of Brontops robustus Marsh 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 669) 

1, Anterior view; 2, medial or inner view (tibial side); 3, posterior view; 4, lateral or outer view (fibular side); 5, proximal view; 
6, distal view; 7, cross section of shaft, cb, Facet for cuboid; mtlll, facet for third metatarsal; ph, facet for proximal phalanx; 
s, s, facets for sesamoids. The principal ligamentous and muscular attachments of this element are given on page 724. 



U. S. GEOLOGICAL SURVEY 



MONOGKAPH 55 PLATE CCXXVII 
4 








ph *.?;.— 




cb 




LEFT FOURTH METATARSAL OF THE TYPE OF BRONTOPS ROBUSTUS MARSH 



PLATE GGXXVIII 



PLATE CCXXVIIl 

Left Manus and Left Pes of the type of Brontops robustus Marsh 
Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See pp. 668, 669) 

1, Left manus, anterior view; 2, left pes, anterior view, a, Astragalus; c, calcaneum; cb, cuboid; ms, mesocuneiform; n, navicu- 
lar; p, pyramidal (cuneiform); ps, pisiform; fi, radius; s, scaphoid; tr, trapezium; V, ulna; unc, unciform; II, III, IV, V, 
digits. The principal ligamientous and muscular attachments of these elements are given on pages 717, 724. 



J 



PLATE GCXXIX 



PLATE CCXXIX 

Restoration of the Skeleton of Brontops robustus Marsh 
Plate prepared under the direction of Professor Marsli. One-seventli natural size. (See p. 




RESTORATION OF THE SKELETON OF BKONTOPS ROBUSTUS MARSH 



PLATE GCXXX 



PLATE CCXXX 

Left Lunar and Right Trapezoid Referred by Marsh to Brontotherium gigas 

May be associated T\-ith type. Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 692) 

1, Left lunar, anterior view; 2, the same, medial or inner view (radial side) ; 3, the same, posterior view; 4, the same, lateral or 
outer view (ulnar side); 5, the same, superior or proximal view; 6, the same, inferior or distal view, to, m', Facets for mag- 
num; p, p' , facets for pryamidal (cuneiform); r, facet for radius; s, facet for scaphoid; un, facet for unciform. 

7, Right trapezoid, anterior view; 8, the same, lateral or outer view (ulnar side) ; 9, the same, posterior view; 10, the same, medial 
or inner view (radial side); 11, the same, superior or proximal view; 12, the same, inferior or distal view, m, m' , Facets for 
magnum; mcll, facet for second metacarpal; p, p' , facets for cuneiform; r, facet for radius; s, facet for scaphoid; t, facet 
for trapezium; un, facet for unciform. 



V. S. GEOLOGICAL SURVEY 

7 



MONOGRAPH 55 PLATE CCXXX 

9 






12 














LEFT LUNAR AND EIGHT TRAPEZOID REFERRED BY MARSH TO BRONTOTHERIUM GIGAS 



PLATE GCXXXI 



PLATE CCXXXI 

Pelvis and Sacrum Referred by Marsh to Brontotherium gigas 

U. S. National Museum. Plate prepared under the direction of Professor Marsh. One-fourth natural size. (See p. 694) 

1, Dorsal view; 2, posterior view, a, Acetabulum; c, crest of ilium; /, thyroid, or obturator foramen; /',/", vacuities, or unossified 
regions in gluteal surface of ilium; p, pubis; il, ilium; is, ischium; p, pubis; s', caudosacral vertebra. 



U. S. GEOLOGICAL SURVEY 



MONOGBAPH 65 PLATE CCXXXI 




PELVIS AND SACRUM REFERRED BY MARSH TO BRONTOTHERIUM GIGAS 



PLATE GGXXXII 



PLATE CCXXXII 

Left Second Metatarsal Referred by Marsh to Brontotherium gigas 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 694) 

1, Anterior view; 2, medial or inner view (tibial side); 3, posterior view; 4, lateral or outer view (fibular side); 5, proximal view; 
6, distal view; 7, cross section of shaft, ec, Facet for ectocuneiform; ms, facet for mesocuneiform; milll, facet for metatarsal 
III; ph, facet for proximal phalanx; s, s, facets for sesamoids. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCXXXII 












LEFT SECOND METATARSAL REFERRED BY MARSH TO BRONTOTHERIUM GIGAS 



PLATE GGXXXIII 



101959— 29— VOL 2 36 



PLATE CCXXXIII 

Left Third Metatarsal Referred by Marsh to Brontotherium gigas 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 694) 

1, Anterior view; 2, medial or inner view (tibial side); 3, posterior view; 4, lateral or outer view (fibular side); 5, proximal view; 
6, distal view; 7, cross section of shaft, cb, Facet for cuboid; ec, facet for eetocuneiform; milV, facet for fourth metatarsal 
ph, facet for proximal phalanx; s, s, facets for sesamoids. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCXXXIII 













LEFT THIRD METATARSAL REFERRED BY MARSH TO BRONTOTHERHIM GIGAS 



PLATE GGXXXIV 



PLATE CCXXXIV 

Left Fourth Metatarsal Referred by Marsh to Brontotherium gigas 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 694) 

1, Anterior view; 2, medial or inner view (tibial side); 3, posterior view; 4, lateral or outer view (fibular side); 5, proximal view; 
6, distal view; 7, cross section of shaft, cb, Facet for cuboid; mtsIII, facet for median or third metatarsal; ph, facet for 
proximal phalanx; s, s, facets for sesamoids. 



U. S. GEOLOGICAL SURVEY 



MONOGRAPH 55 PLATE CCXXXIV 












LEFT FOURTH METATARSAL REFERRED BY MARSH TO BRONTOTHERIUM GIGAS 



PLATE GGXXXV 



PLATE CCXXXV 

Phalanges of Left Second, Third, and Fourth Digits Referred by Marsh to Brontotherium gigas 

Plate prepared under the direction of Professor Marsh. One-half natural size. (See p. 694) 

1 , Proximal phalanx of second digit, lateral or outer view; 2, the same, anterior view; 3, the same, medial or inner view; 4, the same ; 
posterior view; 5, the same, superior or proximal view; 6, the same, inferior or distal view; 7, second phalanx of second digit, 
lateral or outer view; 8, the same, anterior view; 9, the same, medial or inner view; 10, the same, posterior view; 11, the same, 
superior or proximal view; 12, the same, distal view; 13, distal phalanx of second digit, anterior view; 14, the same, posterior 
view; 15, the same, proximal view; 16, proximal phalanx of third digit, lateral or outer view; 17, the same, anterior view; 
18, the same, inner or medial view; 19, the same, posterior view; 20, the same, proximal view; 21, the same, distal view; 
22, second phalanx of third digit, lateral or outer view; 23, the same, anterior view; 24, the same, inner or medial view; 25, 
the same, posterior view; 26, the same, proximal view; 27, the same, distal end; 28, distal phalanx of third digit, anterior 
view; 29, the same, posterior view; 30, the same, proximal view. 



U. S. GEOLOGICAL SUEVEY 



MONOGRAPH 53 PLATE CCXXXV 












28 










%'^: 










,^^-:'^'^s^. 










PHALANGES OF LEFT SECOND, THIRD, AND FOURTH DIGITS 
REFERRED BY MARSH TO BRONTOTHERIUM GIGAS 



U. S. GEOLOGICAL SI'UVEY 



MONOGRAPH 65 PLATE CCXXXVI 




SKELETONS OF MEGACEROPS ACER, MALE AND FEMALE 

Mounted in the Colorado Museum of Natural History, Denver, by J. D. Figgins, Director. Photograph furnished by courtesy of Mr. Figgins. 
Height of male to top of dorsal spine, 1,759 millimeters. One twenty-fifth natural sije 



APPENDIX 



EOCENE AND OLIGOGENE TITANOTHERES 
OF MONGOLIA 



BY 
HENRY FAIRFIELD OSBORN 



896 



CONTENTS 



History of theory and of exploration 899 

Asiatic centers of origin of mammals 899 

Explorations in Mongolia 899 

Orders of mammals represented in the collections- _ 900 

Families absent 900 

Carnivora 901 

Titanotheres 901 

Older formations 901 

The Tertiary formations of Mongolia and their faunas.- 901 

Names, names of subdivisions, and thickness 901 

Pliocene or Pleistocene: Hung Kureh formation 903 

Miocene (?) : Pang Iviang formation 904 

Miocene: Loh formation 905 

Oligocene: Hsanda Gol formation 905 

Oligocene: Houldjin formation 907 

Lower Oligocene: Ardyn Obo formation 908 

Summit of the Eocene: Shara Murun formation 910 

Upper Eocene: Irdin Manha formation 911 

Middle (?) Eocene: Arshanto formation... 912 



The Tertiary formations of Mongolia — Continued. Fage 

Basal Eocene or upper Cretaceous: Gashato formation. 913 
History of the discovery of remains of titanotheres in 

Mongolia 913 

Generic and specific characters of the Mongolian titano- 
theres 916 

Descriptions of species 918 

Subfamily Dolichorhininae 918 

Subfamily Manteoceratinae (Brontopinae) 924 

Subfamily Telmatheriinae 932 

Subfamily Brontopinae 936 

Subfamily Menodontinae 938 

Mongolian titanotheres of uncertain generic reference 939 

Subfamily Manteoceratinae (Brontopinae) 939 

Subfamily Dolichorhininae 940 

Titanotheres of eastern Europe 941 

Surviving embolotheres of M ongolia 942 

Epilogue of the titanothere monograph 944 

Bibliography 944 

897 



ILLUSTRATIONS 



Figure 761. World chart of 1900 explaining Osborn's theory of the central Asiatic origin of orders of holarctic mammals-- 900 

762. Chief centers of known zoogeograpliic distribution of titanotheres 901 

763. Map of eastern and central Asia, showing area of central Mongolia explored by American Museum Asiatic 

expedition of 1922 and 1923 902 

764. Route map of American Museum expedition of 1922 and 1923 in southeastern and central Mongolia, showing 

the three localities where titanotheres were discovered 902 

765. Map of central Mongolia (Gobi Desert region) traversed by third Asiatic expedition 903 

766. Map and section of eastern Altai region, showing location of Tertiary formations 904 

767. Map and section of east-central Gobi Desert, showing location of chief Cretaceous and Tertiary formations 

discovered in this region 906 

768. Ardyn Obo formation (lower Oligocene) 908 

769. Ardyn Obo formation and American Museum camp of 1923 909 

770. Field sketch of chief fossihf erous beds of the Shara Murun formation. 910 

771. Field sketch map of exposures of Irdin Manha formation 911 

772. Irdin Manha formation, Irdin Manha bluff, looking northward across the Kalgan-Urga trail 912 

773. Type lower jaw of Protitanotherium grangeri from the Irdin Manha formation 914 

774. Comparative views of jaws of upper Eocene titanotheres from the Irdin Manha and Shara Murun formations. 917 

775. Comparative occipital views of chief titanotheres of Mongolia in descending geologic order 917 

776. Crania and jaws of the type and parat^'pe of Dolichorhinus haiseni 920 

777. Comparative views of superior and inferior grinding teeth of Dolichorhinus kaiseni and of the lower jaws of D. 

kaiseni and Protitanotherium grangeri 921 

778. Internal aspect of inferior grinding teeth of five species of titanotheres from Mongolia, Burma, and Utah 922 

779. Comparison of superior grinding teeth of Protitanotherium andrewsi and P. mongoliense 923 

780. Jaws oi Protitanotherium gra7igeri in plsLce, Irdin Manha formation 924 

781. Female skuU and jaws of Protitanotherium grangeri (type), altered by vertical crushing of symphyseal region 

of jaw 925 

782. Type female cranium and jaws of Protitanotherium grangeri 926 

783. Type right ramus, fragment, vnth six grinding teeth, of Protitanotherium mongoliense 927 

784. Palate and superior dentition of Protitanotherium mongoliense (neotype) 928 

785. Complete female skull of Protitanotherium mongoliense (referred) 929 

786. Lower jaw of Protitanotherium andrewsi, a finely preserved specimen, for comparison with the imperfect lower 

jaw of the type of Protitanotherium mongoliense 930 

787. Type cranium of Protitanotherium andrewsi 931 

788. Referred cranium and jaw of Protitanotherium andrewsi 933 

789. Types of Telmatherium and of Dolichorhinus from the Irdin Manha and Shara Murun formations 934 

790. Jaws and maxillae of type and paratype of Telmatherium berkeyi 935 

791. Cranium and superior dentition of type of Brontops gobiensis 937 

792. Referred jaw of Brontops gobiensis and lower grinding tooth of Menodus mongoliensis (type) 938 

793. Type jaw of Manteoceras? irdinensis, Irdin Manha formation 939 

794. Type jaw (fragment) ot Metarhinus? mongoliensis compaved with Protitanotherium grangeri 940 

795. Titanotheres of Rumelia, in the Balkan Peninsula; of Transylvania, in southeastern Hungary; and of Bohemia- 942 

796. Restoration of Embolotherium andrewsi 943 

797. Crania of three species of Embolotherium 944 



EOCENE AND OLIGOGENE TITANOTHERES OF MONGOLIA 



By Henry Fairfield Osborn 



The Eocene and Oligocene titanotheres of Mongolia, 
discovered in the years 1922 and 1923, after the manu- 
script of this monograph had been completed and 
sent to the Geological Survey, are described and 
figured in this appendix. The outstanding features 
of the Tertiary formations in which they occur are 
also described and figured. The appendix closes with 
a special bibliography relating to the Tertiary of 
Mongolia and its fauna, so far as it had been described 
up to the end of the year 1925. 

HISTORY OF THEORY AND OF EXPLORATION 

ASIATIC CENTERS OF ORIGIN OF MAMMALS 

The third Asiatic expedition of the American 
Museum of Natural History started in 1921 to pre- 
pare for its exploration of the Gobi Desert of Mongolia 
and to test the theory advanced in the year 1900 by 
the writer of the present monograph that the high 
plateau region of central Asia would prove to be the 
arena of the evolution and adaptive radiation of 13 of 
the principal orders of mammals. This theory was 
set forth by the writer before the New York Academy 
of Sciences in two presidential addresses in the years 
1899 and 1900 '^ and was graphically illustrated by 
a world chart (fig. 761) on which the names of 21 
orders of placental mammals were printed, together 
with the places of their origin and distribution accord- 
ing to this theory. 

In the text and figures of these addresses the chief 
continental centers of origin of these orders of mam- 
mals appeared as follows: 

Africa: 

Proboscidea (elephants and mastodonts) . 

H3'racoidea (rock conies). 

Sirenia (manatees and dugongs) . 
Madagascar: 

Lemuroidea (lemurs). 
Oceania : 

Cetacea : 

Archaeoceti (primitive whales) . 
Mystacoceti (wha/lebone whales). 
Odontoceti (toothed whales). 
Central Asia: 

Insectivora (insectivores) . 

Cheiroptera (bats). 

Creodonta (primitive carnivores). 

Garni vora (modern carnivores). 

Tillodontia (tillodonts) . 

Rodentia (rodents). 

Taeniodonta (primitive edentates) . 

«• Osborn, H. F., Science, April, 1900, p. 567. 



Central Asia — Continued. 

Primates (Mesodonta). 

Amblj'poda (Coryphodon and Dinoceras). 

Condylarthra (condylarth ungulates). 

Perissodactyla (odd-toed ungulates). 

Ancylopoda (clawed perissodactyls) . 

Artiodactyla (even-toed ungulates) . 
India: 

Anthropoidea (primates) . 
South America: 

Edentata (edentates). 

Litopterna (cursorial ungulates) . 

Toxodontia (toxodont ungulates) . 

Tj'potheria (typotheres) . 

Osborn (1900.187, pp. 55, 56) observed: 

Until the Pleistocene, northern Asia is unknown paleon- 
tologically — here is a region for explorers; we may consider it 
as part of a broad Eurasiatic land area extending from the 
Rocky Mountain region to Great Britain. Every year's dis- 
covery increases the resemblances and diminishes the differ- 
ences between Europe and the Rocky Mountain region. In 
the absence of all knowledge of Asia, we find the pure or autoch- 
thonous fauna of the holarctic region distributed in western 
Europe and in western North America. 

C. W. Andrews (Osborn, 1901-1923) says: 

The above chart of 1900 has been verified by the discovery 
(1901-1904) of numerous ancestors of the orders Sirenia, 
Proboscidea, Hyracoidea, Primates, and Archaeoceti in North 
Africa; more recently (1922, 1923) in central Asia by the dis- 
covery of members of the orders Insectivora, Tillodontia, 
Carnivora, Rodentia, Ambl3'poda, Perissodactyla, Ancylopoda, 
and Artiodactyla. 

Osborn, in 1925, added: "Charles W. Andrews's 
discoveries (1901-1904) demonstrate that the primi- 
tive whales (Archaeoceti) originated in northern 
Africa. Osborn is now inclined to place the origin of 
the Anthropoidea (Primates) in central Asia." 

EXPLORATIONS IN MONGOLIA 

During the seasons of 1922 and 1923 the third 
Asiatic expedition, under the leadership of Roy Chap- 
man Andrews, discovered 10 geologic formations of 
Tertiary age, which contained altogether a very large 
number of fossil mammals. Part of the territory 
Andrews explored in 1922 had been crossed by Raphael 
Pumpelly in 1862-1865; by Obruchev in 1892-1894, 
who applied the name "Gobi series" to the later sedi- 
ments; and by Chernov in 1908. Similar territory 
farther south had been crossed by Von Richthofen in 
1877, and it was he who gave the name Khan-Khai 
beds to the Tertiary sediments he found in the desert 
region. The underlying series as developed in China, 



900 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



from the Jurassic downward, had been described by 
Von Richthofen and later by Bailey Willis in his 
"Researches in China" (1907). 

The geologic personnel of the third Asiatic expedi- 
tion iacluded Charles P. Berkey as chief geologist, 
Frederick K. Morris as geologist and topographer, 
and Walter Granger as vertebrate paleontologist. 
The party under Andrews consisted of 25 persons — 
S Americans, 8 Chinese, and 9 RIongohans. 

A single mammalian tooth found by Obruchev in 
the Iren Dabasu Basm (Houldjin? formation) was 
identified by the Vienna jSIuseum as a rhiuocerid ; it is 
probably BalucMiherium, which occurs in the Houldjin 



ORDERS OF MAMMALS REPRESENTED IN THE COIIECTIONS 

The entire Mongolian collections from the 10 
Tertiary formations are more or less clearly recognized 
as belonging to the following orders of mammals: 

Inseotivora of the primitive familj' Pantolestidae. 

Creodonta, archaic small-brained carnivores, fairly numerous. 
Hyaenodontidae, Oxyaenidae, Mesonychidae. 

Carnivora, modernized large-brained carnivores, less numerous,, 
first appearing in the upper Eocene (Miacidae) and lower 
and middle Oligocene (Canidae, Viverridae). 

Rodentia, modernized types of rodents, extremely numerous in 
the middle Oligocene. 

Amblypoda, archaic small-brained ungulates, very rare, dis- 
appearing in the upper Eocene. 




Figure 761.^World chart of 1900 explaining Osborn's theory of the central Asiatic origin of the tliirteen orders of holarctic 

mammals, which spread westward to Europe and eastward to North America 

After Osborn, 1900.182, p. 567, Chart IV. 



gravel. The fossils collected by the third Asiatic 
expedition (1922-23) from the formations of upper 
Eocene and of lower and middle Oligocene age, the 
Mongolian formations best known as yet, so far as 
catalogued, include 406 specimens, distributed among 
zones and formations as follows: 
Baluchitherium grangeri life zone: 

Hsanda Gol formation, upper part (middle Oligocene). 150 
Houldjin formation, upper(?) part (middle Oligocene). 8 
Brontops gohiensis life zone: 

Ardyn Obo (lower OUgocene) 46 

Protitanotherium mongoliewte life zone : 

Shara Murun (uppermost Eocene) 77 

Protitanotherium grangeri life zone: 

Irdin Manha (upper Eocene) 99 

Arshanto (middle? Eocene) 6 

Gashato (basal Eocene or upper Cretaceous) 20 



Artiodactyla, modernized even-toed ungulates, small and rare- 
in the older formations. Tragulina, Helohyidae, Anthraco- 
theriidae. 

Ancylopoda, highly specialized clawed ungulates, ver}' rare. 
Schizotherium in the Ardyn Obo. 

Perissodact3da, modernized odd-toed ungulates, very numerous,, 
representing families of Lophiodontidae, of Amj'nodontidae, 
of Baluchitheriinae, of Hyracodontidae, and of Bronto- 
theriidae. 

Proboscidea, lower Miocene trilophodonts, lower Pleistocene- 
elephants. 

FAMILIES ABSENT 

Conspicuous by their absence in these Eocene and 
Oligocene formations of Mongolia are the primitive 
three-toed horses of the perissodactyl family of the 
Equidae. The role of these small, swift, cursorial 
quadrupeds seems to have been taken in upper 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



901 



Eocene time by the agile lophiodonts of the family 
Lophiodontidae. The tapirs (Tapuidae) are also 
absent; these forest-living animals are rare in all 
flood-plain formations. 

CARNIVORA 

True Carnivora are first known in the upper Eocene 
(Miacidae). They are found also in the middle and 
lower Oligocene (Canidae, Viverridae). The rhinoc- 



OLDER FORMATIONS 

Below these formations is the little-known Arshanto 
formation, of middle (?) Eocene age, and far below 
is the Gashato formation, of basal Eocene or Upper 
Cretaceous age. Above the Ardyn Obo there have 
been discovered the Houldjin and Hsanda Gol for- 
mations, very rich in fossil mammalian life, probably 
representing lower and middle Oligocene time. 




Former land areas Former migration areas Known fossil areas 

Figure 762. — Chief centers of the known zoogeographio distribution of the titanotheres 

Rocky Mountain region; Balkan region; Burma region; Gobi Desert of Mongolia. 



eroses are represented by three families, members of 
all of which are rather rare — first, the amphibious 
Amynodontidae ; second, the small, swift, cursorial 
Hyracodontidae ; third, the large, tall, and slender- 
limbed Baluchitheriinae. 

TITANOTHERES 

The titanothere family (Brontotheriidae) flourished 
in upper Eocene time and was the dominant element 
in the Mongolian fauna. The titanotheres were 
extremely numerous in Mongolia during the deposi- 
tion of the upper Eocene Irdin Manha and Shara 
Murun formations. They also appear in the lower 
Oligocene Ardyn Obo formation and then apparently 
disappear. The Oligocene formations Hsanda Gol 
and Houldjin contain no trace of titanotheres. The 
dominating hoofed mammals of the middle Oligocene 
are the great baluchitheres, of the species Baluchi- 
therium grangeri. The upper Eocene titanotheres are 
as large as the existing rhinoceroses, fully twice the 
size of their American Eocene relatives. Three titano- 
there life zones have been discovered in the flood- 
plain formations, known as the Irdin Manha, Shara 
Murun, and Ardyn Obo; they correspond, respectively, 
to our Bridger (upper part) and Uinta formations 
(Eocene) of Wyoming and Utah and the White River 
group (Oligocene) of Wyoming and South Dakota. 



THE TERTIARY FORMATIONS OF MONGOLIA AND 
THEIR FAUNAS 

NAMES, NAMES OF SUBDIVISIONS, AND THICKNESS 

The oldest name applied to the continental sedi- 
mentary deposits of the Gobi Basin is the Khan-Khai 
of Von Richthofen, who used it (1877) with the sig- 
nificance of "deposits of an evaporating sea." No 
one of these deposits, however, is caused by evapora- 
tion; all are strictly continental in origin, and for this 
reason Obruchev's term Gobi series (1892-1894) is 
preferable. Gobi means desert basin, and Gobi series 
is a suitable name for these scattered desert-basin 
deposits. The name Gobi series was used by Obru- 
chev to include the whole series of late sedimentary 
beds, without distinction as to age. Obruchev's term 
is therefore chosen, although Von Richthofen's has 
priority. 

Some of these continental formations may over- 
lap in time, so that the sum of all the thicknesses 
may not afford an accurate statement of the total 
column. Making all due allowance, however, for 
possible overlap, no less than 6,000 to 8,000 feet of 
Tertiary sedimentary strata have thus far been 
discovered (1922, 1923) above the post-Jurassic 
unconformity in the Gobi Desert of Mongolia. 



902 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 




Figure 763. — Map of eastern and central Asia 
showing (3, 4) tlie area of central Mongolia 
explored by the American Museum Asiatic 
expedition of 1922 and 1923. After Osborn, 
May 25, 1923, p. 2, fig. 1 

Witbin the four circles shown are the middle Oligocene ex- 
posures of southern and central Asia in which remains of 
the giant hornless rhinoceros Baludiittinium have been 
discovered: 1, Bugti beds of Baluchistan, yielding the type 
of Baluchitherium osborni: 2, Indricotlierium ( = Baluchi- 
therium) zone of Tmgai, northern Turkestan; 3, Baluchi- 
tferiuvi gra-ngcri zone (Houldjin formation), southeastern 
Mongolia; 4, Baluchitherium gravgeri type zone (Hsanda 
Gol formation), central iSIongolia. Upper Eocene and 
lower Oligocene titanotheres occur near area 3. Certain 
upper Eocene species of titanotheres and of amynodonts 
are also found in Burma along Irrawaddy River. 




Figure 764. — Route map of the American Museum expedition of 1922 and 1923 in southeastern and central Mongolia, show- 
ing the three localities where titanotheres were discovered. After Berkey and Granger, May 25, 1923, p. 2, fig. 1 

The three localities are the Iren Dabasu Basin, in which lies the Irdin Manha formation (Protitanotherium grangeri zone) , upper Eocene, on the Kalgan-Urga 
trail; Shara Murun (Ptotitanothenum mongoliense zone), uppermost Eocene, on the Kalgan-Sair TJsu-TJliassutai trail; and Ardyn Obo (Brontops gobiensis zone), 
lower Oligocene, on the Kalgan-Sair Usu-XJliassutai trail. These and other fossil fields are indicated by local Mongol names employed to designate the 
respective formations. 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



903 



In the seasons of 1922 and 1923 the third Asiatic 1 formations are more or less clearly recognizable and 
expedition discovered 10 Tertiary formations within | correlated with those of Europe, Asia, and North 
the Gobi series as designated by Obruchev. These | America by the fossils they contain, as follows • 

Tertiary formations of Mongolia, in descending order 



Estimated 

thickness 

in feet 



Geographic region 



Probable or estimated age 



American correlation equivalent 



Hung Kureh_ 
Pang Kiang.- 
Loh 

Hsanda GoL. 

Houldjin 

Ardyn Obo.. 

SharaMurun. 

Irdin Manha. 



Arshanto. 
Gashato- _ 



2,000 

500 

100-1, 000 

3,000 

30 

500 

500 

40-100 

40-100 
300 



Eastern Altai 

Iren Dabasu Basin. 
Eastern Altai 



Hipparion,Camelus? zone. 



_do_ 



Iren Dabasu Basin_ 
Uliassutai trail 



-do. 



Iren Dabasu Basin. 



do 

Eastern Altai. 



Serridentinus (Trilopho- 

don) mongolieusis zone. 

Baluchitherium grangeri 

zone. 
Baluchitherium grangeri? 
zone. 
Brontops gobiensis 
zone. 

Proti t anotlierium 
mongoliense zone. 

Protit anotherium 
grangeri zone. 



Lophiodont-Schloss e r i a 

zone. 
Palaeostylops iturus zone. 



Upper Pliocene or lower 

Pleistocene. 
Pliocene?; age very 

doubtful. 
Lower Miocene 



7,150 



Lower and middle Oli- 

gocene. 
do 



Lower Oligocene. 
Last? appearance of 
titanotheres. 

Summit of Eocene. Ti- 
tanotheres very abun- 
dant. 

Upper Eocene. Titano- 
theres present. 



Middle? Eocene. 



Basal Eocene or Upper 
Cretaceous. 



Oreodou and Metamy- 
nodon. 



White River group, 
Chadron formation. 

Uinta formation (Dipla- 
codon zone). 

Bridger formation 
(Eobasileus zone) , 
horizon B of Uinta 
Basin. 

Bridger formation (?). 

Possibly Puerco, Torre- 
jon, or Fort Union 
formation of the 
Eocene. 



PIIOCENE OR PLEISTOCENE: HUNG KUEEH 
FORMATION 

Berkey and Granger (May 25, 1923, p. 9) 

Thickness. — The sub-Altai Hung Ku- 
reh formation, named after the hills of 
Hung Kureh, measuring in thickness 
approximately 2,000 feet, rests conform- 
ably upon the Hsanda Gol formation 
and is marked above by Pleistocene 
erosion. It lies in the very bottom of 
the basin of Tsagan Nor, between the 
foot of the mountain of Baga Bogdo 
and the Tsagan Nor itself, and consists 
of a series of gravelly sands and con- 
glomerates of great variety and thick- 
ness, at the base of which lie whitish 
and yellowish sands and clayey sands. 
Although the formation is not so well 
exposed as the Hsanda Gol, it has been 
possible to measure more than 1,000 
feet of it, and a reasonable estimate of 
its total thickness in the type locality — 
the hills of Hung Kureh— is 2,000 feet. 
On its extreme margin next to the moun- 
tain of Baga Bogdo the upper beds 
101959— 29— VOL 2 37 




Figure 765. — Map of central Mongolia (Gobi Desert region), traversed by the 
third Asiatic expedition, Roy Chapman Andrews, leader 

Dotted lines show main routes and location of the chief Cretaceous and Tertiary exposures: Iren Dabasu 
(Upper Cretaceous) , north of Irdin Manha (Tertiary, upper Eocene), Pang Kiang (Tertiary, Miocene?) 
Ardyn Obo (Tertiary, lower Oligocene), and Djadochta (IjOwer Cretaceous, Protoceratops zone). This 
area includes about 475,600 square miles and lies northwest of Kalgan and Peking. After Berkey and 
IM orris, The great bathylith of central Mongolia, p. 2, fig. 1, 1924. 



904 



EOCENE AND OLIGOCENE TITANOTHERES OE MONGOLIA 



become a coarse conglomerate, and the influence of this 
rising mountain is shown* by alluvial conglomeratic 
fans. 

Fauna. — These massive Hung Kureh beds are 
fossiliferous only at their base. Here and there a 
bed does carry fossils, particularly in the lowermost 
part of the formation, within 200 or 300 feet of its 



oid; a rodent [Castor]. This fauna indicates an 

uppermost PUocene or lowermost Pleistocene age. 

PLIOCENE?: PANG KIANG FORMATION 

Granger, Berkey, and Morris (October 7, 1924, p. 119) 

About 60 miles southeast of Irdin Manha are the 

Pang Kiang beds, about 500 feet thick, provisionally 

placed in the Pliocene. In some places at least these 



base, where deposits of yellow iron-stained sand and beds rest directly upon old crystalline rocks. Only 




Figure 766. — Map and section of the eastern Altai region, showing location 
of the Tertiary formations 

Hung Kureh, of Pliocene or Pleistocene age; Loh, of lower Miocene age; Hsanda Gol, of middle and lower 
Oligocene age; Gashato, of basal Eocene or Upper Cretaceous age After Berkey and Morris, Basin 
structures in Mongolia, figs. 9 and 10, 1924. 



beds of white sand prevail. The fossil content of 
the lower Hung Kureh as first identified in the field 
and in the American Museum is as follows: A few 
fragments of a Pliocene horse [Hipparion sp.]; a 
few bones and fragments of shells of the eggs of a very 
large bird, probably Struthiolithus ; a large cervid 
[Cerims sp.?]; a proboscidean; a rhinoceros; a camel- 



one fossil has been found in the Pang Eaang, a fragment 
of the jaw of a rodent which Matthew identifies as an 
ochotonid of more recent age than Ohgocene. The 
rodent family Ochotonidae first appears in the middle 
Oligocene of Europe and much more recently in the 
Pleistocene of North America as the pika, or mountain 
hare. 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



905 



Approximate ■position of ten Mongolian formations in the geologic 
column, as determined to the year 19S4 by the fossiliferous levels 
only 

[Thick formations or series like the Hsanda Gol and Hung Kureh doubtless occupy 
a larger part of the column. A query (?) indicates that the level is determined 
only provisionally] 



(?) Hung Kureh. 
(?) Pang Kiang.. 



Feet 
2,000 



600 



Loh 100-1, 000 

(?) Hsanda Gol 3,000 

(?) Houldjln 30 

Ardyn Obo 500 

Shara Murun 500 

Irdin Manha 40-100 

(?) Arshanto 40-100 



(?) 



300 






MIOCENE: LOH FORMATION 

Granger, Berkey, and Morris (October 7, 1924, p. 115) 

Definition. — The Loh formation was first defined as 
follows : 

The lower Miocene clays of Loh, less than 100 feet thick, 
resting upon the Hsanda Gol clays, without any obvious 
physical disconformity. Going southward along their dip 
(fig. 10), we found that they were succeeded by an undeter- 
mined thickness, probably as much as 1,000 feet, of sandy 
clays and sands, in which as yet no fossils have been found. 

Osborn added (November 11, 1924, p. 1): 
In a thin deposit of olive-colored clays and light-gray 
sandstone resting on the red-banded beds of the Hsanda Gol 
formation at Loh, and believed to be of lower Miocene age, 
were found two highly characteristic fossils: (1) Proboscidean. 
A fragmentary series of lower mastodont teeth (Am. Mus. 
19152) which first reveals the presence of an undoubted Serri- 
dentinus in Mongolia, which we name Serrideniinus mongoli- 
ensis. Serridentinus probably marks the arrival of a 
proboscidean related to the M. [Trilophodon] angustidens 
of the lower Miocene of Europe. * * * (2) Rhinocerotine. 
The facial portion of a skull (Am. Mus. 19185) containing 
three grinding teeth and perfectly preserved nasals, which we 
name Baluchitherium mongoliense. 



Fauna. — The type of Serridentinus (Trilophodon) 
mongoliensis is a smaller and more primitive trilo- 
phodont mastodon than its successor Serridentinus 
productus of the upper Miocene marls (lower part of 
Santa Fe formation) of New Mexico. Its ancestors 
are found in Miocene lignitic deposits of western 
Europe, its descendants in PUocene deposits of Texas 
and Florida. The type of Baluchitherium mongoliense 
indicates a hornless rhinoceros little more than half 
the size of the type of Baluchitherium grangeri of the 
Hsanda Gol, but with premolar teeth of a more 
progressive stage. These two fossils from the Loh 
formation, Serridentinus {Trilophodon) mongoliensis 
and Baluchitherium mongoliense, confirm the judgment 
of Granger, Berkey, and Morris as to the probable 
lower Miocene age of the Loh formation. 

OIIGOCENE: HSANDA GOL FORMATION 

Berkey and Granger (May 25, 1923, p. 8) 

Stratigraphic position and lithologic character. — The 
fine series of Tertiary deposits called the Hsanda Gol 
formation, so far as discovered up to 1923, lies uncon- 
formably upon the unevenly eroded surface of the 
Ondai Sair formation, of Lower Cretaceous (Coman- 
che) age. PrevaHingly yellowish conglomerates and 
pebbly sands, varying greatly in quality, constitute 
at least 800 feet of the lowermost beds; the middle 
beds consist of alternating sands, marls, clays, and 
clayey sands of variegated colors, chiefly red, yellow, 
and white, with no apparent uniformity of succession; 
the uppermost beds [fossiliferous] are prevailingly 
sands and clayey sands, are reddish in color, and 
include some beds that are fairly well indurated. 

This whole series of beds stretches along the Hsanda 
Gol, a dry, sandy stream course leading from Mount 
Uskuk through the Ondai Sair locality southward past 
Loh to the bottom of the basin of Tsagan Nor, at the 
foot of Baga Bogdo, a total distance of about 15 miles. 
Altogether, a thickness of approximately 3,000 feet of 
the formation has been measured and estimated from 
measurements, and a large part of it has been inspected 
in detail. 

The uppermost beds of the formation are fossilifer- 
ous, the middle beds are largely barren, and no fossils 
whatever are found in the lowermost beds. In the 
middle beds a fossil is found here and there, but 
fossils are numerous in the uppermost beds only. 
These uppermost beds are judged to be of Miocene 
[Oligocene] age, but whether the whole thickness of 
3,000 feet to the underlying Cretaceous is also Miocene 
[Oligocene] is not known; yet in the absence of any 
physical break in the whole Hsanda Gol formation, it 
is regarded as a unit of Miocene [Oligocene] age. 
Thousands of fossils, chiefly rodents, were collected 
from the uppermost beds, especially in the vicinity of 
Loh, 10 miles downstream from Ondai Sair, and near 
the so-called Grand Gorge, a corresponding erosion 
exposure 10 miles farther west. 



906 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



Fauna (1923-1926).— OngmaWj regarded by Berkey 
and Granger (1923) "as a unit of Miocene age," 
the fossiliferous uppermost levels [sands and clayey 
sands] of the Hsanda Gol are now known to be of 
middle Oligocene age, but it is believed that the for- 
mation may well extend down at the base to upper 
Eocene. 



Matthew and Granger (January 18, 1924, p. 5) 
listed 22 genera and 26 species of Insectivora, Creo- 
donta, Carnivora, Rodentia, Perissodactyla, and 
Artiodactyla of the Hsanda Gol formation. Both the 
Creodonta and the Carnivora are in an Oligocene stage 
of evolution tending to early Oligocene; the Rodentia 
appear to be of lower and middle Oligocene age; the 



Haiiai srranlh 
\:-^-:~J\ graywacke 




. J ~^^Vit'53'^^^■!,V'->~"'T■- 



/,/ .U"i'r r\lrdin Manha 




•Irdin Manha escarpmenf Houldjm escarpmenf^||,^^ pisAsu- 

Sob; pencplanel undraine_d lowland Sobi peneplane und'ralrK 



N 

lowland 
atos 



^wouuo)iy..,:°i^-; 




Figure 767. — Map and section of the east-central Gobi Desert, showing location 
of the chief Cretaceous and Tertiary formations discovered in this region 

Iren Dabasu (Upper Cretaceous dinosaur beds), Houldjin (contains middle Oligocene BaluchWierium zone), 
Irdin Manha (upper Eocene ProtUanotherium grangeri zone), Arshanto (middle (?) Eocene lophiodont- 
Schlosseria zone). Pang Kiang (Miocene?), Shara Murun (uppermost Eocene ProtUanotheTium mongo- 
liense zone). The section shows the Irdin Manha formation (upper Eocene) overlying the Arshanto 
red beds (middle (?) Eocene), which in turn may overlie the lien Dabasu (Upper Cretaceous dinosaur 
beds); also the Houldjin formation (middle and lower Oligocene) apparently overlying the Arshanto 
red beds (middle (?) Eocene). After Berkey and Morris, Basin structures in Mongolia, figs. 12, 
13, 1924. 



The forms taken from the uppermost beds include 
(op. cit., p. 9) " BalucMtherium, a fine skull nearly 5 
feet long [BalucMtherium grangeri], and other rhino- 
cerids; rodents by the hundreds; artiodactyls; insecti- 
vores; and carnivores." The fossiliferous uppermost 
beds of this formation or series are termed the Balu- 
cMtherium grangeri zone. 



Insectivora also appear to be of OHgocene age; one 
artiodactyl, comparable to certain artiodactyls of the 
French Oligocene Phosphorites, is a true primitive 
cervid, Eumeryx. In the perissodactyl hornless rhinoc- 
eros BalucMtherium grangeri the premolar evolution 
is more modernized or recent than that of the Indri- 
cotherium asiaticum Borissiak of Turgai, Turkestan. 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



907 



Preliminary lists of middle 



and upper Eocene and Oligocene Mongolian fossil faunas as determined to June SO, 
1925, by shorn, Matthew, and Granger 





Middle Eocene; 
Arshanto for- 
mation 


Upper Eocene 


Lower Oligocene: Ardyn 
Obo formation 


Middle Oligocene: Upper 
part of Hsanda Gol and 
Houldjin formations 




Irdin Manila formation 


Sliara Murun formation 


Creodonta: 




Andrewsarchus, 
?Harpagolestes, 
?Synoplotherium, 
?Hapalodectes. 












Pterodon hvaenoides 


Hyaenodon eminus._ 

Ardynictis furuneu- 
lus. 

Cynodictis 

Desmatolagus 


Hyaenodon and 'Z 


Oxyaenidae 




sis. 


genera. 


Carnivora : 

Canidae, Miaci- 




Miacis inviotus.. 




Amphicticeps and 3 


dae, Viverridae. 
Rodentia- -_ ._ 






7Desmatolagus 


genera. 
9 genera, 11 species. 






Eudinoceras mongo- 
liensis. 




Artiodactyla: 
Tragulina. 




Archaeomeryx 


Lophiomeryx, (2 spe- 
cies), Miomeryx. 


Eumeryx culminis. 










Anthraeotheriidae 








?Ancodon. 




Entelodontidae _ 








Entelodon dirus. 


Perissodactyla: 

Lophiodontidae 

Rhinocerotoidea : 

Hyracodontidae_ _ 


Schlosseria__ 
Teilhardia.. 


Teleolophus, Desma- 
totherium mongo- 
liense, D. fissum. 

Caenolophus (1 spe- 
cies), Lophialetes. 
?Amynodon 


Deperetella cristata. . 

Caenolophus (4 spe- 
cies) . 

Amynodon (or new- 
genus) . 

Baluchitherium (or 
new genus) . 

Protitanotlierium 
mongoliense, P. 
andrewsi, Dolicho- 
rhinus kaiseni. 


Colodon inceptus, 
Paracolodon curtus. 

Ardynia praecox 

Cadurcotherium 
ardynense. 








Baluchitherium gran- 


Baluchitheriinae. 
Titanotheroidea: 




Protitanotherium 
grangeri, Telma- 
therium berkeyi, 
Dolichorhinus 
olseni, Manteo- 
ceras? mongoli- 
ensis, Metarhinus? 
mongoliensis. 


Brontops gobiensis, 
Menodus mongoli- 
ensis. 

Schizotherium avitum 


geri. 


Chalicotheriidae 

















OLIGOCENE: HOULDJIN FORMATION 

Granger and Berke}' (August 7, 1922, p. 4) 

The type locality of the Houldjin formation is 5 
miles south of Iren Dabasu. It consists of sand and 
gravel, 30 feet in total thickness; with a coarse sandy 
conglomeratic member at base (5 feet), fossUiferous. 
Originally characterized by the following fossil con- 



tent: (1) A rhino cerid [ICaenopus or Praeaceratherium, 
sp., ? Cadurcotherium sp.]; (2) a large carnivore [Ente- 
lodon dirus, equal in size to Dinohyus hollandi, lower 
Miocene of Nebraska]; (3) an artiodactyl of the size 
of a Virginia deer [1 Entelodon sp.]; (4) an enormous 
mammal, possibly related to or identical with Balu- 
chiiherium of Baluchistan [1 Baluchitherium]; (5) a tor- 
toise of large size [^Testudo], 



908 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



It was from the Houldjin gravel, probably, that 
Obruchey obtained the rhmocerid molar tooth {IBalu- 
cMtherium) The Houldjin was regarded by Granger 
and Berkey (1922, p. 4) as of more recent age than 
the imderlying Irdin Manha formation and as equiva- 
lent in age to the Hsanda Gol formation, because 
both contain BalucMtTierium. 

In 1923 Matthew and Granger (December 18, pp. 
1-6) described the fragmentary Houldjin fossils indi- 

Y- ToSairUsu 



Brontops qobiensis skull \^fe 

Fossiliferous escarpment 

A 



Ardyn Obo formation, CadurcotJierium ardynense 
Osborn. The large Baluchitherium is comparable 
with that from Turgai described by Borissiak. The 
smaller rhinocerid is comparable in size with Caenopus 
occidentalis, from the Oligocene of Nebraska, with the 
Epiaceratherium turgaicum Borissiak of Turgai, and 
with the Aceratherium filholi Osborn of the Phos- 
phorites, Oligocene; its relationships remain to be 
determined. 



\ 



Small Jlrtiodac1y/a,etc, \^ Pnmn 



Cadurcother/um ardynense quarry 




Obo 



M 
W 



ToKalgan 



Ardyn Obo 



CROSS SECTION ATARDYN OBO 
LOOKING WESTWARD 



Cadurcotherium ardynense cjuarr/ 




Brontops gobiensis 
' skull 



FiGX7HE 768. — Ardyn Obo formation (lower Oligocene), after field notes and sketches by 

Granger 

Upper part: Sketch map of Ardyn Obo formation, showing locstion of the principal finds, namely, type skull of Brcm- 
tops gobiensis (Am. Mus. 20354J, Cadurcctkerium ardynense quarry, and small Artiodactyla. 

Lower part: Cross section of the Ardyn Obo formation, showing tho levels on which the remains of Cadurcotherium 
ardynense, Brontops gobiensis, and Menodus mongoliemis were found. 



cated above in brackets, also those given in the 
accompanying table. They concluded in general that 
this fauna is of Oligocene age but can not be exactly 
correlated until better known; it may be homotaxic 
with the Hsanda Gol fauna, but this supposition 
rests only on the doubtfully identified Baluchitherium; 
the Houldjin and Hsanda Gol type localities are more 
than a thousand miles apart, and the geologic character 
of the two formations is quite different. The amyno- 
dont 1 Cadurcotherium is comparable with that of the 



LOWER OIIGOCENE: ARDYN OBO FORMATION 

Berkey and Granger (May 25, 1923, p. 12) 

Type locality and notable features. — The type locality 
of the Ardyn Obo formation on the main Kalgan- 
Uliassutai trail is south of Sair Usu, where a great 
escarpment, surmounted by an obo (the inevitable 
guidepost of the Mongolian desert), stands 300 feet 
above the general level of the plain over which the 
Uliassutai trail passes, and where for many miles the 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



909 



edges of the fiat-lying strata are exposed. Tliese 
strata are formed of ratlier loose and slightly indurated 
sandstone and clayey sand of considerable variety of 
texture, quality, and color. In the upper members 
of the Ardyn Obo formation fossils were obtained, 
including several fine specimens of rhinocerids and 
numerous fragments of turtles. Although it is judged 
to be of mid-Tertiary age, the Ardyn Obo can not be 
correlated with either the Hsanda Gol or the Houldjin 
formation. 

Morris and Granger (notes, 1925) report as follows: 
On top of the Ardyn Obo beds lies a heavy sand- 
stone capping, which has preserved at its present 
level this great mesa. The beds are predominantly 



Fauna. — Matthew and Granger (December 18, 1923, 
pp. 1-5) described the small collection of mammals 
obtained in 1922 at "Promontory Bluff," on the Sair 
Usu-Kalgan trail, about 150 miles from Sair Usu and 
350 from Kalgan. They concluded that this appears 
to be an Oligocene fauna; Cadurcotherium, Schizo- 
therium, and Cynodictis are generically characteristic 
of the Phosphorites of Quercy; a giant tortoise 
(Testudo insolitus) is apparently in a rather primitive 
stage. The nearest zoogeographic affinities of this 
fauna are with western Europe rather than with the 
western United States. 

Osborn (1925) added the important titanothere 
genus and species Bronfops gohiensis, which relates 




Figure 769. — Ardyn Obo formation (lower Oligocene) of Mongolia, where the types of Brontops gohiensis and Cadurcothe- 
rium ardynense were discovered 
American Museum camp of 1923 in the foreground. After Am. Mus. negative No. 251603. 



sandy but include some layers of shale. The fossils 
were collected along the face of the great Ardyn Obo 
bluff and at its base, about 300 feet from the summit 
sandstone capping; the remaining 200 feet lies out in 
the basin and is not well exposed. The Cadurco- 
therium ardynense quarry, opened in 1922 and again 
worked ia 1923, was withia 100 feet of the top of the 
mesa, and most of the teeth and bones of this small 
aquatic rhinoceros came from about this level. The 
numerous artiodactyl jaws, accompanied by an occa- 
sional rodent or carnivore, were found on knolls at 
the base of the great bluff, 150 and 200 feet below its 
top. The skull of the titanothere Brontops gohiensis 
was found at the lowest collecting level, about 300 
feet below the top of the sandstone capping. 



this formation to the lower Oligocene of South 
Dakota. 

Brontops gohiensis life zone. — The name "Brontops 
gohiensis life zone" has been assigned by Osborn 
to the Ardyn Obo formation to emphasize the fact 
that Brontops gohiensis is strikingly close in its evolu- 
tion to Brontops hrachycephalus and to Teleodus avus, 
discovered in the lower Titanofherium life zone 
(Chadron formation) in the great Badlands of South 
Dakota. This robust, broad-headed little titano- 
there was discovered some 300 feet below the top 
of the Ardyn Obo formation, which is altogether 
over 500 feet in thickness as measured by Morris. 
We may confidently assign a lower Oligocene age 
to this formation. A single lower molar represents 



910 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



Menodus mongoliensis, a much larger animal than 
Brontops, characteristic of a higher geologic level. 

Osborn (October 19, 1923) described Cadurcotlierium 
ardynense as resembHng Cadurcotherium cayliixi of the 
Oligoceue Phosphorites of France. The quarry con- 
taining these rhinoceroses {Cadurcotherium ardynense) 
lies 200 feet above the Brontops gobiensis level, or 
100 feet below the top of the mesa, which is protected 
by a heavy sandstone capping that preserves it from 
erosion. 

SUMMIT OF THE EOCENE: SHAEA MURUN FORMATION 
Berkey and Granger (Ma}' 25, 1923, p. 12) 

General features. — A hundred miles south of the 
Ardyn Obo formation is a distinct basin carrying 
fossiliferous strata which have been called the Shara 
Murun formation. The sedimentary area is very 
large and the best exposures occur along the borders 
of a 200-foot escarpment. At this place were found 



To Sair Usu 




Dolichorhinus kaiseni 
Protifanotherium mongo/iei 
andremsKtypi 



20252 

Dolichorhinus kaiseniftypej 



FiGTjRE 770. — Field sketch of the chief fossiliferous beds of the Shara Murun formation 
near Ula Usu, on the Kalgan-Uhassutai trail, showing where the types of Prolitano- 
therium andrewsi, P. mongoliense, and Dolichorhinus kaiseni were discovered 

Unlike the Irdin Manha quarries, the fossihferous locaUties are close together in the Shara Murun formation. 
After sketch by Walter Granger, chief paleontologist of the third Asiatic expedition. 

titano there remains resembling those found in the 
Irdin Manha formation early in the season of 1922. 
These beds are assigned a total thickness of 500 feet. 

The first fossil found in this formation was a titano - 
there, which Osborn (October 17, 1923, p. 3) identified 
as ProtitanotJierium mongoliense, a form closely inter- 
mediate in size and characters between Protitano- 
tJierium emarginatum and the large Protifanotherium 
superhum, both of the Uinta formation (horizon C of 
Uinta Basin) of northern Utah (compare fig. 778). 

According to Granger a total thickness of over 300 
feet is exposed at this type locality. The fossils are 
distributed through a vertical thickness of about 150 
feet; this may account for the presence in the upper 
levels of Protitanotherium andrewsi, a more progres- 
sive species than Protitanotherium mongoliense, which, 
in turn, is much more progressive than the species 
Protitanotherium grangeri of the Irdin Manha forma- 



tion. The Shara Murun beds near Ula Usu have 
somewhat the appearance of those of the Irdin Manha 
formation at the Irdin Manha bench — gray shales 
above and red shales at the base — but they weather 
out in deeper gulches. Calcareous concretions are 
rare in the 150 feet of rich fossiliferous beds. There 
is much local slipping of the clays, whereby many of 
the titanothere crania were badly crushed and dis- 
torted, so that they required careful restoration. 
The titanothere bones of the Shara Murun are readily 
N distinguishable by their light 

cream color and softer texture 
from the rust-brown bones found 
in the Irdin Manha. 

Fauna. — The fauna of the 
Shara Murun formation is more 
recent than that of the Irdin 
Manha. The small lophiodonts 
(Desmatotherium) that swarm in 
the Irdin Manha are rare in the 
Shara Murun, which carries 
many larger lophiodonts (Depere- 
tella) and long-limbed baluchi- 
therine (?) rhinoceroses (ancestral 
Baluchitherium?) , as well as 
numerous Amynodon. The great 
herds of lophiodonts found in the 
Irdin Manha are replaced in the 
Shara Murun by small traguline 
artiodactyls. These artiodactyls 
(Archaeomeryx) were found in 
layers that contained no other 
mammals and may represent 
small herds suddenly overcome. 
The large lophiodonts {Depere- 
tella) were also found by them- 
selves, many individuals com- 
pacted in a small area. 
The rhinoceroses, which are rare in the Irdin Manha, 
are fairly abundant in the Shara Murun and are at 
some places mingled with titanotheres but are gener- 
ally segregated. They include one complete skeleton 
of Amynodon in a stage of evolution similar to that of 
Amynodon antiquus of the upper part of the Bridger 
formation (horizon B of Uinta Basin) of Utah. Skele- 
tal remains and jaws and teeth of five very long-limbed 
baluchitherine (?) rhinoceroses were found, which 
may prove to be ancestral to the Baluchitherium 
grangeri of the middle Oligocene. 

The titanotheres of the three species Dolichorhinus 
Icaiseni, Protitanotherium mongoliense, and Protitano- 
therium andrewsi dominate all other mammals, and 
in one quarry (fig. 770) a large deposit of these bones 
was found and worked by Kaisen during June and 
September, 1923. In the Shara Murun, as in the 
upper Eocene " Titanotherium beds" of the Rocky 



To Kolgan 
Wells 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



911 



Mountains, some of the skulls include jaws, but few 
are found associated with parts of the postcranial 
skeleton. Only one fairly complete skeleton of a 
titanothere was foimd in the Shara Murun, a specimen 
that had been in part destroyed by the weather. 
This skeleton bears the number Am. Mus. 20277. 
Incomplete limb and foot bones of six other skeletons 
were also collected. AU the 17 or more specimens of 
titanotheres collected, except one titanothere skull 
with jaws (Am. Mus. 20252), were found in one rather 
small area just north of the Kalgan-Uliassutai trail. 
This area, which has a radius of about a third of a 
mile, was a veritable bone bed, and the large collection 
of remains of titanotheres obtained from it represents 
only a small part of the fossils actually weathered out, 
a fact that indicates the extreme abundance of these 
titanotheres in central Mongolia in upper Eocene 
time. 

Protitanotherium mongoUense lije zone. — The faunis- 
tic name "Protitanotherium mongoUense life zone" was 
given by Osborn to the Shara Murun formation on 
determining that the species of titanotheres which 
they contain are considerably more recent in age than 
those of the Irdin Manha formation. Altogether 
more than 17 specimens of titanotheres have been 
found in a single fossiliferous exposure, distributed as 
follows : 

Protiianolherium andrewsi, the most progressive species, 
represented by the type and six other specimens, also skeletal 
material, probably belonging in upper levels. 

Protiianotheriuin mongoUense, represented b}' the type and 
five other specimens, .also skeletal material. 

Dolichorhinus kaiseni, represented by the type and three 
other specimens. 

UPPER EOCENE: IRDIU MANHA FORMATION 
Granger and Berkey (August 7, 1922, p. 5) 

Locality and characteristic features. — About 25 miles 
south of the Houldjin formation a conspicuous 
member of Obruchev's Gobi series, called the Irdin 
Manha formation, immediately overlies the Upper 
Cretaceous Iren Dabasu formation. It consists of 
cross-bedded sandstones, limy sand, and pebbly 
sandstones and comprises a barren upper member 
(25 + feet thick) and a lower member, the lophiodont- 
bearing bed (4 feet thick). Originally characterized by 
(1) small Lophiodonta of at least two species in great 
abundance [Desmatotherium]; (2) a perissodactyl 
.[Protitanotheri