sti) et : : att EES 5 sts { : ch i a Spevs- tether Biistats rs ett ee3t 3 eee a Stiet BE Pshesisiigissdaree ses fe ssePs sestetetete cS pie wi ; as fiat sat . shi a i SH LH , et a % 4 Se Peo VERTEBRATES ph] AY IN THE American Museum of Natural History. DEPAKEMENT OF. VERTEBRATE PALAZONTOLOGY. Volume II. ARTICLES COLLECTED FROM THE AMERICAN MUSEUM BULLETINS OF THE YEARS 16958-1903. BY HENRY FAIRFIELD OSBORN, J. W. GIDLEY, iui. WORTMAN, F, W. LOOMIS, W. D. MATTHEW, BARNUM BROWN, Or, P. HAY, Re oh eis, WALTER GRANGER, W. K. GREGORY.- SMITHSON AY APR 2 1 1987 SRARIE? Collected and issued for purposes of sale anc New York, December, 1903. § Ce i \ vi : . PRESS OF = ri THE New ERA PRINTING COMPANY, Pm 7 - - LANCASTER, PA. rin Bey tlt MURBUN JAKOTT AE we 7 dl | we aS EMAC EES. PerLORATIONS, AND: RESEARCHES OF THE Department of Vertebrate Paleontology. PREFACE ‘TO’ VOLUME. I. BY HENRY FAIRFIELD OSBORN, CURATOR. In the preface of Volume I| a brief outline was given of the foundation of the Department in May, 1891, together with the successive appointments of the cur- ator and different members of the staff, of the purposes of the Department, of the twenty-one expeditions sent out between 1891 and 1897, and of the acquisition of the Cope collection of fossil mammals; also a sum- mary of the collections, embracing in 1897 the remains of 3,000 animals ; and of the paleontological, geologi- cal and stratigraphical work accomplished ; and a re- sumé of the chief scientific results contained in the twenty-one Bulletins. It appears appropriate to introduce Volume II with a similar summary of the work accomplished in the six years between 1897 and 1903. Staff.— The preface of this second volume must first refer with regret to the resignation of Dr. J. L. Wortman in March, 1899, to take charge of vertebrate palazontology in the newly established Carnegie Mus- eum. Dr. Wortman’s services to this Department both in the field and in scientific writing were of the highest order. Under his able and energetic direction in the field seventeen of the twenty-one expeditions were conducted, laying the foundations especially for 111 IV Preface. the remarkably representative Eocene and Oliogocene collections of mammals and the Jurassic collection of fossils. Dr. W. D. Matthew, now Associate Curator, was appointed in Dr. Wortman’s place. In January, 1900, Dr. O. P. Hay, now Associate Curator of Chelo- nia, joined the staff in connection with the cataloguing and arrangement of the Cope Collection of fossil rep- tiles, amphibians and fishes. Mr. Walter Granger con- tinued in charge especially of the Jurassic field work until the season of 1903, when he reéntered the mam- malian fossil field in the Bridger basin. Mr. J. W. Gidley joined the staff in 1899, with reterence to strengthening our collection of Miocene, Pliocene and Pleistocene mammals, and in 1900 he took charge of the exploration for fossil horses under the Whitney Fund. Mr. Barnum Brown joined the staff in 1896 in connection with the Jurassic explorations, and in 1900 was put in charge of the exploration of the Creta- ceous. We are indebted to Mr. F. B. Loomis, of Amherst College, for codperation in the expeditions of 1900, 1901 and 1902; and to Professor R. S. Lull, of the Amherst Massachusetts Agricultural College, for codperation in the Cretaceous and Jurassic expeditions of 1899 and 1902. In 1903 Mr. Peter Kaison, who formerly served as field and Museum assistant, took charge of the continued excavation of the Bone Cabin quarry. Mr. Adam Hermann has continued as head prepar- tor, and has been constantly perfecting his methods of mounting. Mr. W.K. Gregory joined the staff in 1900 as an assistant to the curator and in charge of the Department Library. Mr. A. E. Anderson has con- Preface. V tinued in charge of the photographic and art depart- ment. Among the artists who have illustrated our publications should be mentioned especially : Messrs. Weber, Christman, Yoshiwara ; Mrs. Stirling and Miss Box. Total Summary of Collections. —The collec- tions made by our field expeditions during the four- teen years embrace the remains of over 5,000 fossil animals. Our collections have also been strengthened by friendly exchange with American and foreign museums, and by purchases, the most important being that of the second portion of the famous collection brought together by Professor Edward D. Cope, known as the Cope Collection of Fossil Fishes, Amphibians, Reptiles and Birds. This was generously presented to the Museum by President Morris K. Jesup in 1902. - At the same time the Pampean Collection of Ameghino, Larroqgue and Brachet, originally exhibited at the Paris Exposition in 1881, and there purchased by Professor Cope, was secured for this Museum by the gift of Messrs. H. O. Havemeyer, D. Willis James, Adrian Iselyn, Henry F. Osborn and the late James M. Con- stable and William E. Dodge, Trustees of the Museum. Among the other purchases are those from Charles H. Sternberg from his explorations in the Kansas chalk. Altogether the purchases have added 1,800 speci- mens. The chief exchanges have been with the museums of London, Paris, Munich, St. Petersburg, Berlin, Stuttgart, Lyons and Buenos Aires. The collections as a whole now include fourteen thousand four hundred specimens, representing : vi Preface. PossilemarnmalSe.,csc ces. scapes neiesw soe camer sed nen sienoinieinisriems 9,373 Fossil reptiles. ............csssesceseeeceeseseneecates sceeeeeuees 2,694 ThtaAS SH Lobes by se sacisneoceotinekseobbea0 Huodestindobstieceo soon. coudabodn 57 Fossil amphibians,.,.......2..cseeeeesseseeseerceeteneee receeeees 593 TASCrIlN AIS) VES A Aen aaciace erasbononcoade ago, neoprcaseaqcse Sscagcaoge 1,202 ARS TEI sec agnnoonenePeonodsdd: cg to Speadasoucnt rscecaboceen 14,420 Beginning in 1891, the horizons have been explored in the following order : Permian, Texas, 1902. Triassic, North Carolina, 1894, 1895. Upper Jurassic (Como Beds), Wyoming, 1897-1903 ; Montana, 1903. Upper Cretaceous (Niobrara), Kansas, 1897. Upper Cretaceous (Fort Pierre), South Dakota, 1903. Upper Cretaceous (Laramie), Wyoming, 1892, 1900 ; Montana, 1902. Basal Eocene (Puerco), New Mexico, 1892. Basal Eocene (Torrejon), New Mexico, 1892, 1896. Lower Eocene (Wasatch), Wyoming, 1891, 1896 ; New Mexico, 1896. Middle Eocene (Wind River), Wyoming, 1891, 1896; . Colorado, 1897. Middle Eocene (Bridger), Wyoming, 1893, 1895, 1903; Colorado, 1897. Upper Eocene (Uinta), Utah, 1894, 1895. Oligocene (White River), South Dakota, 1892, 1894, 1901, 1902. Oligocene (White River), Colorado, 1898, 1901; Montana, 1902. Middle Miocene, Colorado, 1898, 1901, Igo2. Upper Miocene (Loup Fork), South Dakota, 1894, 1902, 1903. Upper Miocene, Texas, 1899, 1900, 1901 ; Montana, 1902. Pliocene (Blanco), Texas, 1900, 1go!. Pleistocene (Sheridan), Nebraska, 1893, 1897; Texas, 1899, Igo0o, 1gOl. Pleistocene Arkansas, 1903. General Exploration. — Two series of expeditions, for mammals and for reptiles, conducted by from two to four parties, have been sent out each year since 1897 in the systematic search for fossil reptiles and mammals in different formations. These expeditions during the past five years, namely between 1898 and 1903, have had a considerable meas- ure of success. They have added some 2,500 speci- Preface. vii mens to the collections, including some thirty more or less complete skeletons, eight of which have already been mounted and placed on exhibition, as follows : Equus scotte. Merycodus osbornt. Oruztholestes hermann. Glyptotherium texanum. Flypohippus equinus. Dinictis squalidens. Neohtpparion whitneyt. Cynodictis grevarius. So Po) RECENT EXPLORATIONS: Exploration for Dinosaurs. — A large part of the field work of the Department since 1897 has been directed to securing the remains of Dinosaurs, especially from the Upper Jurassic (Como Beds) of Wyoming and Colorado, and from the Cretaceous. A rich and extensive deposit, the ‘Bone Cabin Quarry,” was opened up in Wyoming in 1898, and has been worked with good results for six years. Evolution of the Horse. — The generosity of one of the Trustees of the Museum, Mr. William C. Whitney, enabled the Curator to send out in 1899 and subsequent years a series of expeditions into various Tertiary formations of which the chief object was to obtain materials to illustrate the evolution of the horse. GEOLOGICAL RESUETS. The field parties beginning in 1900 have with one or two exceptions made very exact stratigraphical field records. The result is the accumulation of im- portant data concerning the distribution of faunas and especially concerning the subdivision of horizons which were formerly considered single or incorrectly identi- fied. We may summarize these results since 1900 as follows : vill Preface. 1. Uinta (Upper Eocene) subdivided by Peterson into Telmatotherium and Diplacodon Horizons. 2. Puerco (Basal Eocene) subdivided by Wortman and Matthew into Puerco and Torrejon. 3. Two other Eocene horizons determined and corre- lated in the Huerfano Basin of Colorado by Osborn and Matthew. 4. White River (Oligocene) subdivided by Wort- man and Matthew by the addition of two upper di- visions, Protoceras beds and Leptauchenia beds. 5. The three divisions of the White River, first made by Hayden and Leidy, further defined and correctly -correlated with the stratigraphy, and the existence shown of two contemporary faunal phases in each division (Wortman and Matthew). 6. The various JZzocene formations which have been grouped under the general names of Loup Fork and Deep River, distinguished and correctly correlated with their respective Miocene horizons, and the so-called Palo Duro horizon eliminated (Matthew and Gidley). 7. The ALolian Theory. — Our field researches have resulted in replacing the lake basin theory by the fluviatile and colian theory, especially under the ob- servations of Dr. W. D. Matthew, the arguments de- rived from which are fully summarized in his memoir. This demonstration has been nearly simultaneous with that of Messrs. Hatcher, Davis, and Gidley, who have reached and published similar results. 8. Stratigraphical Succession of Horizons. — The most complete paper on stratigraphical succession is that by Matthew, No. 28, entitled ‘A Provisional Classification of the fresh water Tertiaries of the West,” summarizing results obtained by our field parties by Preface. 1X comparison with those obtained previously by Leidy, Cope, Marsh, and others. Of the same nature are the exact researches by Mr. F. B. Loomis, on the Jurassic stratigraphy of Wyoming (Nos. 37, 48.) PAE AZSONTOLOGICAL -RESULYS. Dinosaurs and other Reptiles. Bulletins. — The first paper, No. 27, is an erroneous identification by Osborn of Camarasaurus Cope with Brontosaurus Marsh, it having since been found that Camarasaurus is more nearly allied to Morosaurus. In this paper the vertebral structure of Sauropoda is discussed. The fore and hind limbs of carnivorous and herbivorous Dinosaurs are discussed by Osborn in later papers. The latest contributions are an account of the skull of Creosaurus by Osborn, No. 61, and of the skull of Triceratops by Lull, No. 60. Attention should be called here to a series of import- ant memoirs on fossil reptiles, namely: A complete Mosasaur Skeleton, by Osborn; A Skeleton of Diplo- docus, by Osborn; Ox the Reptilian Subclasses Diap- sida and Synapsida and the early history of the Diapto- sauria, by Osborn, Evolutionary Series. — One of the principal ob- jects of our expeditions has been the completion of the evolutionary or phyletic series so far as possible, with a view to working out the development of differ- ent orders and families from their first appearance to their extinction. To this subject the following papers have been espe- cially devoted: No. 7, Azcestors of the Tapir, by Wortman and Earle; No. 19, on the Ganxodonta and L:dentata, by Wortman ; No. 24, on the Camelida, by x Ie, refa BE Wortman; No. 26, on the Amblypoda, by Osborn ; No. 29, on the Canide, Viverride and Procyonide, by Wortman and Matthew; No. 33, on the AAznocer- oses of Europe, by Osborn ; No. 35, on the Creodonta, by Matthew; No. 40, on the 77tanotheres, by Osborn ; No. 43, on the Eocene Primates, by Osborn. Systematic Revisions. — We have also taken up one group after another, and have succeeded in giving a more or less complete and final revision of certain of the groups, incidentally determining questions as to priority, synonymy and the location of type specimens. Among the papers of this character have been the fol- lowing: No. 2, Spectes of Coryphodon, Earle; No. 8, Upper Cretaceous Mammals, Osborn; No. 12, Puerco Mammals, Osborn and Earle; No. 10, Species of - Flyracothertum, \Nortman ; No. 21, Puerco Mammals, Matthew; No. 26, Pantolambda and Coryphodon, Os- born ; No. 35, Cveodonta, Matthew ; No. 36, American Species of Equus, Gidley; No. 43, Eocene Primates, Osborn ; No. 46, the yfertragulide, Matthew ; 49, Cretaceous Actinopterous Fishes Hay; No. 53, Me. Lebanon Fishes, Hay. Morphological Series.— The chief morpholog- ical or anatomical papers are the following: No. 3, Protoceras, Osborn and Wortman; No. 5, Acerathe- rium tridactylum, Osborn; No. 9, Patriofelis, Wort- man; No. 14, dgrtocherus, Wortman; No. 18, Psz¢fa- cothertum, \Wortman ; No. 22, Teleoceras, Osborn ; No. 23, Coryphodon, Osborn; No. 25, Phenacodus, Osborn: No. 30, Oxyena, Wortman; No. 32, Lguus scorn: Gidley: No. 34, Oxyena and Patriofelis, Osborn ; No. 41, Dinocyon, Matthew; No. 42, Bunelurus, Mat- thew ; No. 45, Ceratogaulus, Matthew ; No. 46, Hypr- Preface. X1 sodus, Matthew; No. 54, Orzztholestes, Osborn; No. 55, Neohipparion, Gidley ; No. 57, Glyptotherium, Os- born ; No. 58, Paramylodon, Brown ; No. 60, 772cera- tops, Lull; No. 61, Cveosaurus, Osborn. Papers especially relating to phzlosophical anatomy are those on the teeth and feet by Osborn and Wort- man and Gidley and especially No. 39, Doltchocephaly and Lrachycephaly, by Osborn and No. 52, Zhe Fle- phants’ Skull, by Gregory. Faunal Lists of Different Horizons. — We have also attempted to give complete faunal lists of different horizons. The most comprehensive paper of this character is No. 28, by Matthew, including faunal lists of ‘the entire Areshwater Tertiaries of the West. Other papers are No. 12, Puerco Mammals, Osborn and Earle; No. 21, Puerco Mammals, by Matthew ; No. 47, Pleistocene Fauna from Hay Springs, by Mat- thew ; No. 50, Lower Oligocene Fauna of the Pipestone Springs, by Matthew; No. 59, Azocene Fauna of Northwestern Texas, by Gidley. Memoirs on Fossil Mammals. — These include The Extinct Rhinoceroses, Part!, by Osborn, and Fosse Mammals of the Tertiary of Northeastern Colorado, by Matthew. The memoir on the Rhinoceroses will be continued and the memoir on the Fvolution of the fforse is 1n preparation by Osborn with the codépera- tion of Mr. J. W. Gidley. PoAOTOGR AP HY. Field Photographs. —In connection with the var- ious expeditions enumerated on page iii a series of 580 field photographs have been taken, the negatives of which are filed in the department for use in connection xii Preface. with our publications as well as for lecture and illus- trative purposes. Museum Photographs. — These include photo- graphs of mounted skeletons of skulls, of feet and other portions of fossil mammals and reptiles number- ing altogether 590 negatives, also photographs of the restorations of fossil mammals and reptiles by Charles R. Knight. RESTORATIONS, AND MODEES: The work of the department would be incomplete without reference to the very interesting series of res- torations of fossil mammals and reptiles executed by the artist Mr. Charles R. Knight, chiefly under the direction of Professor Osborn, a complete list of which is given at the close of this volume and in a new cata- logue which will be issued in 1904. December, 1903. Preface. Xill PUBLICATIONS FROM: THE DEPARTMENT OF VERTEBRATE PALASONTOLOGY.1 MEMOIRS. 1. The Extinct Rhinoceroses. Part I. By Henry Fairfield Osborn. Vol. I, Part III, pp. 75-164, pll. xiia-xx, and 49 text figures. April 22, 1898. Price, $4.20. . A Complete Mosasaur Skeleton. By Henry Fairfield Osborn. Vol. I, Part 1V, pp. 165-188, pll. xxi-xxili, and 15 text figures. Oc- tober 25, 1899. 3. A Skeleton of Diplodocus. By Henry Fairfield Osborn. Vol. I, Part V, pp. 189-214, pll. xxiv-xxvili, and 15 text figures. Octo- N ber 25, 1899. Price of Parts IV and V, issued under one cover, $2,00. 4. Fossil Mammats of the Tertiary of Northeastern Colorado. By W. D. Matthew. Vol. I, Part VII, pp. 353-446, pll. xxxvil-xxxix, and 34 text figures. November, 1901. Price, $2.00. 5. Lhe Reptilian Subclasses Diapsida and Synapsida and the Early Flistory of the Diaptosauria. By Henry Fairfield Osborn. Vol. I, Part VIII, pp. 449-507, pl. xl, and 28 text figures. Novem- ber, 1903, (Price, 42.60. BULLETINS. Contents of Volume I. 1892 1. Fossil Mammats of the Wahsatch and Wind River Beds. Collec- tion of 1891. By Henry Fairfield Osborn and J. L. Wortman. Vol-1V, No: 1, Article xi, pp. 81-147, pl..1v, and 19 text figures. October 20, 1892. Price, 80 cents. Revision of the Species of Coryphodon. By Charles Earle. Vol. IV, No. 1, Article xii, pp. 149-166, 2 text figures. October 18, Kag2) ‘Pree, 15 cents: 3. Characters of Protoceras (Marsh), the New Artiodactyl from the Lower Miocene. By Henry Fairfield Osborn and J. L. Wort- man, Vol Vj, No. 1, Article xvii; pp. 351-371, 6 text figures. December 30, 1892... Price, 25 cents. Nv 1893 4. Artionyx, a New Genus of Ancylopoda. By Henry Fairfield Osborn and Jacob L. Wortman. Vol. V, Article i, pp. 1-18, 5 text figures. February, 1893. Price, 20 cents. 1 Separates of most of these papers can be obtained of the Librarian of the Museum at the prices indicated, or in exchange. XIV Preface. 135 Aceratherium tridactylum from the Lower Miocene of Dakota. By Henry Fairfield Osborn. Vol. V, Article vii, pp. 85-86. April 29, 1893. Price, 15 cents. On the Divisions of the White River or Lower Miocene of Dakota. By J. L. Wortman, M.D. Vol. V, Article ix, pp, 95-105. June 27, 093. ‘Pace, 15 cents. Ancestors of the Tapir from the Lower Miocene of Dakota. By J. L. Wortman and Charles Earle. Vol. V, Article xi, pp. 159- 180, 7 text figures. August 18, 1893. Price, 25 cents. Fossil Mammals of the Upper Cretaceous Beds. By Henry Fair- field Osborn. Vol. V, Article xvii, pp. 311-330, pll. vil and vill, and 4 text figures. December 15, 1893. Price, 40 cents. 1894 Osteology of Patriofelis, a Middle Eocene Creodont. By J. L. Wortman, M.D. Vol. VI, Article, v. pp. 129-164, pl. 1 and 5 text figures. May 24, 1894. Price, 40 cents. Fossil Mammats of the Lower Miocene White River Beds. Col- lection of 1892. By Henry Fairfield Osborn and J. L. Wort- man. Vol. VI, Article vi, pp. 199-228, pll. 11 and 11, and 8 text figures. July 28, 1894. Price, 50 cents. On the Affinities of Leptarctus primus of Leidy. By J. L. Wort- man. Vol. VI,- Article, vii, pp.. 2290-231. July 30, 1928 HCE; 15, cents: 1895 Fossil Mammats of the Puerco Beds. Collection of 1892. By Henry Fairfield Osborn and Charles Earle. Vol. VII, Article i, pp. I-70, 21 text figures. March 8, 1895. Price, 80 cents. Fossil Mammals of the Uinta Basin. Expedition of 1894. By Henry Fairfield Osborn. Vol. VII, Article ii, pp. 71-105, 17 text figures. May 18, 1895. Price, 50 cents. On the Osteology of Agriocherus. By J. L. Wortman. Vol. VII, Article iv, pp. 145-178, pl. 1, and 24 text figures. June 17, 1895. Price, 60 cents. ertssodactyls of the Lower Miocene White River Beds. By Henry Fairfield Osborn and J. L. Wortman. Vol. VII, Article xii, pp. 343-375, pll. viii to x1, and 12 text figures. December 23, 1895. Price, 80 cents. 1896 Species of Hyracotherium and Allied Perissodactyls from the Wah- satch and Wind River Beds of North America. By J. L. Wort- man. Vol. VIII, Article vi, pp. 81-110, pl. m1, and 18 text figures. May 12, 1896. Price, 50 cents. Preface. XV 17. Zhe Cranial Evolution of Titanothertum. By Henry Fairfield Osborn. Vol. VIII, Article ix, pp. 157-197, pll. 111 and Iv, and 13 text figures. July 31, 1896. Price, 75 cents. 18. Psittacotherium, a Member of a New and Primitive Suborder of the Edentata. By Dr. J. L. Wortman. Vol. VIII, Article xvi, pp. 259-262. November 30, 1896. Price, 15 cents. 1897 19. Zhe Ganodonta and thetr Relationship to the Edentata. By J. L. Wortman. Vol. IX, Article vi, pp. 59-110, and 36 text figures. March 22, 1897. Price, 85 cents. 20. The Huerfano Lake Basin, Southern Colorado, and its Wind kiver and Bridger Fauna. By Henry Fairfield Osborn. Vol. IX, Article xxi, pp. 247-258. October 20, 1897. Price, 15 cents. 21. A Revision of the Puerco Fauna. By W. D. Matthew. Vol. IX, Article xxii, pp. 259-323, and 20 text figures. November 16, 1897... Price, 75, cents. List of Casts, Models, Photographs of Skeletons and of Restorations, issued March 15, 1898. Price, 40 cents. BULLETINS. Contents of Volume II. 1898 . A Complete Skeleton of Teleoceras fossiger. Notes upon the Growth and Sexual Characters of this Species. By Henry Fairfield Osborn. Vol. X, Article iv, pp. 51-59, pll. iv, va. March 18, 1898. Price, 25 cents. 23. A Complete Skeleton of Coryphodon radians. Notes upon the Loco- motion of this Animal. By Henry Fairfield Osborn. Vol. X, Article vi, pp. 81-91, pl. x, and 2 text figures. April 4, 1898. Price, 20 cents. 24. The Extinct Camelide of North America and some Associated Forms. By J. L. Wortman, M.D. Vol. X, Article vii, pp. 93-142, pl. XI, and 23 text figures. April g, 1898. Price, 70 cents. 25. Remounted Skeleton of Phenacodus primevus. Comparison with Euprotogonia. By Henry Fairfield Osborn. Vol. X, Article ix, pp. 159-164, pl. x11, and 4 text figures. May 6, 1898. Price, 20 cents. 26. Evolution of the Amblyfoda. PartI. Taligrada and Pantodonta. By Henry Fairfield Osborn. Vol. X, Article xi, pp. 169-218, and 29 text figures. June 3, 1898. Price, 80 cents. iS) i.) 30. oie 33: 34. 36. Preface. Additional Characters of the Great Herbivorous Dinosaur Camara- saurus. By Henry Fairfield Osborn. Vol. X, Article xii, pp. 219-233, 13 text figures. June 4, 1898. Price, 35 cents. 1899 A Provisional Classification of the Fresh Water Tertiary of the West. By W. D. Matthew. Vol. XII, Article iii, pp. 19-77. March 31, 1899. Price, 50 cents. The Ancestry of Certain Members of the Canidae, the Viverride and Procyonide. By J. L. Wortman and W. D. Matthew. Vol. XII, Article vi, pp. 109-139, pl. v1, and Io text figures. June 21, 1699, Price, 40 cents. Restoration of Oxyena lupina Cope, with Descriptions of Certain New Species of Eocene Creodonts. By J. L. Wortman. Vol. XII, Article vii, pp. 139-148, pl. vir, and 3 text figures. June Zt eisgo. Price, 20 cents, Fore and Hind Limbs of Carnivorous and Ferbivorous Dinosaurs from the Jurassic of Wyoming. Dinosaur Contributions, No. 3. By Henry Fairfield Osborn. Vol. XII, Article x1, pp. 161- 172, 8 text figures. October 30, 1899. Price, 25 cents. 1900 . A New Species of Pleistocene Horse from the Staked Plains of Texas. By J. W. Gidley. Vol. XIII, Article xiii, pp. 111-116, and 5 text figures. August 18, 1900. Price, 15 cents. Phylogeny of the Rhinoceroses of Europe. Rhinoceros Contribu- tions No. 5. By Henry Fairfield Osborn. Vol. XIII, Article xIx, pp. 229-267, and 16 text figures. December 11, 1900. Price, 50 cents. Oxyena and Patriofelis Re-studied as Terrestrial Creodonts. By Henry Fairfield Osborn. Vol. XIII, Article xx, pp. 269-281, pll. xvi and xix, and 4 text figures. December 21, 1900. Price, 35 cents. Igor - Additional Observations on the Creodonta. By W. D. Matthew. Vol. XIV, Article i, pp. 1-38, and 17 text figures. January 31, 1901. Price, 50 cents. Tooth Characters and Revision of the North American Species Oat the Genus Equus. By J. W. Gidley. Vol. XIV, Article ix, pp. 91-142, pll. XvilI-xx1, and 27 text figures. May 31, Igo1 Price, $1.20. 37: 38. Ao. 4I. 43. 44. 45. 46. 47. 48. Preface. XV On Jurassic Stratigraphy in Southeastern Wyoming. By F. B. Loomis. Vol. XIV, Article xii, pp. 189-197, pll. xxv1 and met, JUNE T7, 1o0r.. Price, 25 cents: Fore and Hind Limbs of Sauropoda from the Bone Cabin Quarry. By Henry F. Osborn and Walter Granger. Vol. XIV, Article xiii, pp. 199-209, and 6 text figures. July 9, 1901. Price, 20 cents. 1902 Dolichocephaly and Brachycephaly tn the Lower Mammals. By Henry Fairfield Osborn. Vol. XVI, Article vu, pp. 77-89, and 5 text ieures. bebiuary 3,.1902. © Price, 20' cents: The Four Phyla of Oligocene Titanotheres. By Henry Fairfield Osborn. Vol. XVI, Article viii, pp. gI-109, and 13 text figures. February 18, 1902. Price, 20 cents. A Skull of Dinocyon from the Miocene of Texas. By W. D. Matthew. Vol. XVI, Article xi, pp. 129-136, and 4 text figures. April 7, 1902. Price, 20/cents: On the Skull of Bunelurus, a Musteline from the White River Oligocene. By W. D. Matthew. Vol. XVI, Article xii, pp. 137-140, and 3 text figures. April 7, 1902. Price, 20 cents. American Eocene Primates, and the Supposed Rodent Family Mixodectide. By Henry Fairfield Osborn. Vol. XVI, Article Xvli, pp. 169-214, and 40 text figures. June 28, 1902. Price, 75 cents. New Canide from the Miocene of Colorado. By W. D. Matthew. Vol. XVI, Article xxi, pp. 281-290, and 4 text figures. Sep- tember 18, 1902. Price, 15 cents. A Horned Rodent from the Colorado Miocene. With a Revision of the Mylogauli, Beavers, and Hares of the American Tertiary. By W. D. Matthew. Vol. XVI, Article xxii, pp. 291-310, and i text ncites.. .seplember 25. 91902." ‘Price; 35, cents. The Skull of FHypisodus, the Smallest of the Artiodactyla, with a Revision of the Hypertraguide. By W. D. Matthew. Vol. XVI, Article xxiii, pp. 311-316, and 4 text figures. September 25 L@O2. /Price, 10 cents. List of the Pleistocene Fauna from Hay Springs, Nebraska. By W. D. Matthew. Vol. XVI, Article xxiv, pp. 317-322. Sep- tember 25, 1902. Price, 10 cents. On Jurassic Stratigraphy on the West Side of the Black Fiills. Second Paper on American Jurassic Stratigraphy. By F. B. Loomis. Vol. XVI, Article xxxi, pp. 401-407, pll. LIV-LV. November 3, 1902. Price, 25 cents. XViil 49. 50. 53- 54. 55- 56. 57- 59- 60. 61. Preface. 1903 On Certain Genera and Species of North American Cretaceous Actinopterous Fishes. By O. P. Hay. Vol. XIX, Article i, pp. 1-95, pll. 1-v, and 72 text figures. March 4, 1903. Price, $1.95. The Fauna of the Titanotherium Beds at Pipestone Springs, Mon- tana. By W. D. Matthew. Vol. XIX, Article vi, pp. 197-226, and 1g text figures. May 9, 1903. Price, 40 cents. A Fossil Hedgehog from the American Oligocene. By W. D. Matthew. Vol. XIX, Article vii, pp. 227-229, and I text figure. May 9g, 1903. Price, 10 cents. Adaptive Significance of the Shortening of the Elephants Skull. By W.K. Gregory. Vol. XIX, Article ix, pp. 387-394, pl. Xx11 and 4 text figures. July 8, 1903. Price, 10 cents. On a Collection of Upper Cretaceous Fishes ‘from Mount Lebanon, Syria, with Descriptions of Four New Genera and Nineteen New Species. By O. P. Hay. Vol. XIX, Article x, pp. 395-452, pll, xx1v-xxxvul. July 8, 1903. Price, 95 cents. Ornitholestes hermanni, A New Compsognathoid Dinosaur from the Upper Jurassic. By Henry Fairfield Osborn. Vol. XIX, Article xii, pp. 459-464, and 3 text figures. July 23, 1903. Price, 10 cents. A New Three-Toed Horse. By J. W. Gidley. Vol. XIX, Article xiii, pp. 465-476. July 24, 1903. Price, Io cents. On Two Species of Platygonus from the Pliocene of Texas. By J. W. Gidley. Vol. XIX, Article xiv, pp. 477-481, and 5 text figures. July 24, 1903. Price, 10 cents. Glyptotherium texanum, A New Glyptodont, from the Lower Pleistocene of Texas. By Henry Fairfield Osborn. Vol. XIX. Article xvii, pp. 491-494, pl. xLiml. August 17, 1903. Price, 15 cents. A New Genus of Ground Sloth from the Pleistocene of Nebraska. By Barnum Brown. Vol. XIX, Article xxii, pp. 569-583, pll. Land Li. October 28, 1903. Price, 35 cents. The Fresh-Water Tertiary of Northwestern Texas. American Museum Expeditions of 1899-1901. By J. W. Gidley. Vol. XIX, Article xxvi, pp. 617-635, pll. Li1—Lvimt. November 21, 1903. Price, 85 cents. Skull of Triceratops serratus. By Richard Swan Lull. Vol. XIX, Article xxx, pp. 685-695, pl. LIx, and 1 text figure. Decem- ber 24, 1903. Price, 20 cents. The Skull of Creosaurus. By Henry Fairfield Osborn. Vol. XIX, Article xxxi, pp. 697-701, and 2 text figures. December 24, 1g03. Price, 10 cents. ¢ i t Preface. X1X a. Provistonal Guide to the Hall of Fossil Vertebrates. By W. D. Matthew. Edition of 1900-1901, 4 pp. Price, 5 cents. Super- seded by Guide Leaflet No. 3. 4. Guide Leaflet No. 3. Zhe Hall of Fossil Vertebrates. By W. D. Matthew. Supplement to American Museum Journal. Vol. IJ, No. 1. January, 1902. 1 pl. and 1o text figures. Superseded by Guide Leaflet No. 12. c. Guide Leaflet No. 9. Zhe Evolution of the Horse. By W. D. Matthew. Supplement to American Museum Journal. Vol. Ill, No, 1. January, 1903. 7 pll.and 4 text figures. Price, Io cents. d. Guide Leaflet No. 12. Zhe Collection of Fossil Vertebrates. By W. D: Matthew. Supplement to American Museum Journal. Vol. III, No. 5. October, 1903. 1 pl. and 15 text figures. Price, 10 cents. Revised List of Casts, Models, Photographs of Skeletons and of Re- storations, issued March, 1904. Price, 50 cents. A Complete Skeleton of Teleoceras fosst- ger. Notes upon the Growth and Sexual Characters of this Species. By Henry Fairrietp Oszorn. PLATES’ ITV; anp [Va. AUTHOR’S EDITION, extracted from BULLETIN OF THE American DUusewum of Datural History, Vo. X, ARTICLE IV, pp. 51-59. New York, March 18, 1898. ae =<. 7 - : e x j ~ . - - — , i = -: ig : es a2 7 ; ry) Article IV.—A COMPLETE SKELETON OF TELEO- CERAS FOSSIGER. NOTES UPON THE GROWTH AND SEXUAL CHARACTERS OF. THIS SPECIES. By Henry FAIRFIELD OSBORN. PLATES IV AND IVA. The remarkable series of Rhinoceros skulls in the Cope and American Museum Collections from the Upper Miocene or Loup Fork Beds of Kansas and Nebraska, has finally been prepared for exhibition and research purposes. Associated with them, and of very great value, is a complete skeleton representing an aged female of very large size, mounted from materials belonging to several individuals secured by our excavations in Phillips Co., Kansas, under the direction of Dr. Wortman in the months of September, October and November, 1894. By the comparison of the 16 skulls and 13 jaws, representing both sexes and all stages of growth, we are enabled for the first time to positively define the animal long known as Afphelops fossiger, to distinguish it both from Rhinoceros and Aceratherium, and point out its important sexual and individual variations. The writer’s attention was first drawn to the largely disregarded sexual and age characters of fossil Ungulates in studying the group of Titanotheres ;' the extinct Rhinoceroses conform to the laws which were observed in that group, and which are familiar enough among living types, namely: males, of larger size with more robust and rugose skulls; horns, if present, more prominent ; canines largely developed; incisors and anterior premolars dis- _ appearing in adults. We owe to Hatcher® the valuable demonstration that Aphelops fossiger bore a terminal horn upon the nasals, although he as- signed this character to a type which he supposed represented a PP. 157-107-. i ; : * American Geologist, March, 1804, pp. 149-150. [51] 52 Bulletin American Museum of Natural History. [Vol. X, major proves to be a middle-aged male of A. /ossiger, and _ his distinction of Zeleoceras as a genus supersedes Aphelops Cope,’ because Cope’ originally applied the term Apfhelops to A. megalodus, defining it as an Acerathere with only three premolar teeth in the lower jaw. This is true of the type species (4. megalodus), but this species should, so far as we know at present, be referred to the genus Aceratherium, in which the lower premolars vary from four to three in number according to age and individual variation, as in the living Rhinoceros. Technically, however, Hatcher’s definition did not clearly dis- tinguish Zeleoceras from Rhinoceros, as he himself stated (of. c77., page 245). Our abundant material proves not only that Ze/eoceras is a Rhinoceros with a median horn on the tips of the nasals, but that it is fully distinguished from the genus AAznoceros as follows : Horns. Lower Premolars. Digits. Genus RAznoceros.. Upon anterior portion of nasals. .4 in young, 3 in aged individuals........ 3-3 Genus 7¢e/eoceras...Upon tips of nasals........... 3 in young, 2 in aged individuals. .......3-3 The reduction of the lower grinders to 5 in 7. fossiger (as com- pared with 6 in RAznoceros) is a very important and distinctive character, as it absolutely excludes Zeleoceras fossiger from the ancestry of any of the modern Rhinoceroses, and shows it to have represented a distinct side phylum, as Scott and Osborn had already determined from its skeletal characters. EXCAVATION OF THE SKELETON. The Phillips County Quarry,* near Long Island, Kansas, was discovered in 1883 by Mr. Charles Sternberg, who collected for the University of Kansas and for the Harvard University Museum. From the latter collection Scott and Osborn procured materials 1 American Naturalist, March, 1894, pp. 241-246. 2*On Some New Extinct Mammalia from the Tertiary of the Plains.’ Palason, Bull. No. 14, Proc. Am. Phil. Society, July 25, 1873. 8 See Williston, ‘ Restoration of Acerather*um fossiger Cope,’ Kansas University Quarterly, Vol. II, No. 4, April, 1894, pp. 280-290. 3 1898. | Osborn, Skeleton of Teleoceras fossiger. 53 for the restoration which they published in 1890." Subsequent collections were made by Sternberg and Hatcher for the United States Geological Survey, between 1884 and 1886. Later Professor Cragin collected here, and in 1891 Mr. E. P. West of the Univer- sity of Kansas, aided by Mr. T. R. Overton, began the extensive collections which led to the preparation of the skeleton for the University under the direction of Professor Williston.* This skele- ton, as mounted in the Kansas Museum and described by Williston, gives a much more accurate idea of this animal than the pre- vious restoration by Scott and Osborn, in which the chest is represented far too shallow. Its principal dimensions are as follows: Length, not including fail, go it. > height, 4 ft:; greatest girth, 9 ft., 4 in. The measurements of the American Museum skeleton as mounted are: Length, 1o ft. 2 in. to bend of tail; height at withers, 4 ft. I in.; greatest girth, g ft. 2 in. From the above accounts, and especially from our own observa- tions, it is seen that this quarry represents an old bone-bed, probably the deposit of some stream or small river along which the rhinoce- roses herded in great numbers. The materials collected by the American Museum party are extremely numerous, especially in the skeletal parts, the figures running somewhat as follows: Skulls, Meastapilce-e7r vettewice, £59 ; humeri, 13 > radii, 207; ulnz, 10’; carpals, 90; metacarpals, 38; pelves, 5; femora, 8; tibiz, 20; astragali, 22; calcanea, 18. In this typical bone-bed are mingled individuals of both sexes and of all sizes, and the proximity of one specimen to another is not a certain guide. There are certain spots, however, where considerable portions of individual skeletons have drifted to- gether. We associate the skull and pelvis in our mounted speci- - men, for they are of similar age and were found within about six feet of each other, the skull being that of a fully adult female, and the pelvis indicating a corresponding age, because the ilia are united above the sacrum; with the pelvis moreover was found a part of the jaw belonging to the skull; also with this pelvis _ 1 ‘Preliminary Account of the Fossil Mammals from the White River Formation contained in the Museum of Comparative Zodlogy.’ Bull. Mus. Comp. Zodl., Vol. XIII, No. 5, p. 92, 1890. 2 OP. cit. 54 Bulletin American Museum of Natural History. (Vol. X, belong a femur, tibia and fibula, astragalus, calcaneum and cuboid of one side, several metacarpals and metatarsals and two cervical vertebra. The selection of the other limb and foot-bones was made from these as a guide. Similarly about 300 feet distant were found the principal ribs which have been selected for this mount, characterized by the very rugose appearance and oblique lines for the insertion of the abdominal muscles (sacro-lumbalts, longissimus dorst). Near these ribs were large jaw and limb-bones corresponding in size with those placed in the mounted skeleton. Apart from these proba- ble associations, the main principle of selection adopted through- out has been that of the age and size standard, after a careful comparison of all the elements. In each region the largest and oldest bones were chosen. Upon this principle the ribs are shown to be of very great length; the chest girth exceeds that indi- cated in the Scott-Osborn restoration, and equals that in the mount in the Kansas Museum, which has heretofore appeared extreme. In additional support of this correlation of material belonging to different individuals, this bone-bed gives evidence of the existence of only one species of Rhinoceros, namely, T. fossiger. All the differences observed are due to growth, individual and sexual variations, as set forth below. The following description is supplementary to the very full statements of the skeletal characters of Z. fossiger made by Osborn in 1890. DESCRIPTION OF MOUNTED SKELETON. (Museum Catalogue Number, 2604.) Mounting.—The composite skeleton shown in ihe accompany- ing plate (Plate IV) has been mounted with remarkable skill by Mr, Adam Hermann, preparator. All the bones are traversed by small steel rods, rendering them firm and solid and the limbs self- supporting. As shown in the photograph, the only visible parts of the metal framework are the two uprights for the shoulder and skull, and pelvis, ‘The bones are in a beautiful state of preserva- 1898. | Osborn, Skeleton of Teleoceras fossiger. 55 tion, and except in the case of the artificial elongation of a few of the ribs and completion of the upper border of the scapule (from complete scapulz of smaller size) no plaster was necessary. Skull—Nasals smooth, expanding into a laterally-compressed beak anteriorly, probably characteristic of adult females, with sharp sides and a lateral notch. Temporal ridges forming a sessile sagittal crest ; premaxillaries with triple infraorbital fora- mina ; lachrymals with well-defined, knoblike projection ; zygo- mata very deep, with extensive attachment for masseter muscle inserted on a well-defined ridge on angle of jaw. Occiput broad and low. Jaw with a single mental foramen below the second or third premolar, and a marked median depression between the canines upon front surface of the chin, Dentition. — Formula: I, C$, P3, M3. Lower and median upper incisors vestigial; lower canines worn, enamel measuring 4o mm. (1% inches). Adult condition indicated by very slight wear of crown of third superior molar. Upon outer surfaces of upper grinders parastyle nearly obsolete. First upper molar with enamel crown of same length as that of premolars. Second upper molar with sudden elongation or hypsodontism, with enamel crown measuring 7omm. Thus m3 and m% elongate or hyp- sodont, and of great service as reserve teeth for old age. Molars with secondary folds characteristic of the species. Vertebre.—Atlas much narrower than in A&. zzdicus, with verte- braterial canal directly traversing the transverse process ; axis with a very low spine; cervicals 1-3, with transverse process restored ; characters of inferior lamellz somewhat conjectural. Supposed Vertebral Formula: C.7, D.19, L.3, S.5. The above formula is purely conjectural. It is made to con- form to that of the living 2. wuicornis and R. sumatrensis.’ Sev- enteen of the ribs are provided with both capitular and tubercular facets. The ribs are extremely long and powerful, not very widely arched ; lower line of chest nearly reaching the ground, as in 1 The definitely ascertained formula of the Oligocene A. ¢7/dactylum is D, 19; L, 5; S, 3. This animal gives us no clue to A. /ossiger, because it belongs to the Dicerathere series. 56 Bulletin American Museum of Natural History. \|Vol. X, the Hippopotamus ; girth (9’ 2’’) exceeding that of R. wnicornis (8’ 9’’) ; chest section deep and heavy, rather than rounded as in R. unicornts. Scapula very characteristic, triangular in contour, with pointed upper border ; narrow supraspinatus and very broad, triangular infraspinatus fossa ; acromion placed midway on spine, reflected backwards, so that attachment of deltoid and trapezius muscles is elevated. Coracoid process forming a prominent rugosity for short head of biceps ; bicipital tendons passing through a double osseous groove upon front of humerus (as in A. umicornis, R. bi- cornis and R. simus)"* and inserted distally in a prominent rugosity upon front of radius. Humerus exhibiting prominent rugosity for subscapularis muscle interior to inner tendon of biceps ; greater tuberosity consisting of a large incurved hook for supra- spinatus muscle, and a separate and distinct knob for infraspinatus muscle ; powerful deltoid ridge, everted but not hooked ; distally a large external condyle for extensor muscles, which exhibit rugose insertion areas in the proximal ends of metapodials. Pelvis with ilia arching over and coalescing above the five sacral vertebral spines, affording a stout area of origin for /atssimus dorst, erector spine and gluteus maximus muscles, correlated with support and propulsion of the enormous abdomen; a foramen piercing the posterior superior border of the ilium. Femur with rugose but not very prominent greater trochanter; lesser trochan- ter for insertion of gluteus maximus muscle, much less promi- nent than in 2. wacornis, and not hooked ; third trochanter not prominent. Pubis and ischium more slender than in &. waccornis. Tibia with a characteristic fissure in the cnemial crest dividing it into two distinct tuberosities for patellar tendons (unlike single crest of R. unicornis and A. malacorhinus). Fibula fused with tibia in aged individuals. ? Busk‘ On the Ancient or Quaternary Fauna of Gibraltar,’ Trans. Zo6l. Soc., 1877, p. 97. 2 De Blainville, ‘ Ostéographie,’ Atlas 3, Gen, Rhinoceros, PI. iv. 1898. | Osborn, Skeleton of Teleoceras fossiger. 57 COMPARATIVE MEASUREMENTS. Teleoceras Rhinocerus Sossiger. indicus. Feet. | Meters.| Feet. | Meters. Motaiglenathetoy bend Ofitailie anes ces. =o. o'-- 10. 2 | 3.10 | 10.8 3.05 Height, Stl eA er aceers ante 6 toe PARES Cerne Ax Ty | es WILE GG Hye operat cei eh ayar chek 8) drac.s fohcy 0) se Sool! Spiel som 5.64 | 1.69 Breadth, across pelvis. . Sas Sane Sere 2.10 | 0.87 2.8 0.82 oS SRLS ors Graig wri Ie eae aoe 2.11 | 0.89 Skull, fengthucondiyles toys jeriets ste) stale) eel I. 11} 0.59 2.1% | 0.65 width across arches, dental series....... Ee 3)0538 Vert. column, total excluding Gardals 0s 3st wnat: FeO 2530 8.2 2.48 Cervicals, including intervert. pyacee SA Oot O. Oe I. 64) 0.47 Dorsals, Bs A oa anes 4. 64] 1.39 Lumbars, SS “a RMON See Ncoags sate ee oe “5 5| Ose CANGRAIIS Spee eS AROS Oho ee, eR ECR ee on On] (eg 0.21 Rib (? sth)—length (Groundtthetcunve)e neo s|\02- a 0.80 (Pao) ET dee anasto n so Seep eae eno cee 3. 34\) 1-08 Girth of chest at 11th rib (estimated).......... O), 2 |) BAKO 2.9 2.66 Fore limb, total flexed (ball of hum. vert. to LOU) eect crers ates ce aly AICHE EER Pae Rees 2. 44| 0.72 3.5 1.04 GA Pla tems wins aerstaiss Sichala'at siete laid a a FOUuol| toca dal pOLste! NSAI CLUS eteralieys fel hore clepeaa ray sietete eis stam velave ts Seapevete< Net ets OTL 1.4 0.40 IReeUC ULI Sue peeeee Peer Red estat cate casey ersitnis tov ate “ei /oy et ei|t Tal wie ate 0.24 a3 0,38 WOlitrveampeereetetets co ete ttcae ty cae tepe es hits alae elo leldeaielats $06 406 0.34 IWC yvewere, sre cihe cracks celeste icisre oo avale a S65 ||" 4doah al Chios 8 0.20 Hind limb, total flexed (ball of femur vent. to ground) RS OIA ia SRO e Rea aan LS TEOr 2. 74| 0.80 3.84 | 1.13 ELE TU tgueeerensten weve ed ty oepotet se artont ate etoile, ous ofa: oa" St imiasis Sele nests | LOUAT 1.8 0.52 TUITE by Selenite nice DO Sno ae etna ae error) OF23 1.24 | 0.36 INE S006 Goto Ae Os oe oe ENERO ELI Rete ee a ae om O.LOO4l 77s Bell Ong Lower jaw, total length, condyle to tip of GRINNING omactehd of Hee LH aH EE PRIA te ep rocices OSH! Wentition toltiprolscanine:.5-.-2-.--. .-- betel rouse 0.35 (Cian? SEIntES Ao Cbd ce 6 + coe eee ee! (eee 0.26 Piltenothicaudalsi(alonesctinve) a. +. 22.6522 -6\-- 350 bon 5] (Opts From the above measurements it appears that from head to tail T. fossiger is only six inches shorter than 2. wnicornzs, while the _ back is eighteen inches (.580 mm.) nearer the ground. This remarkable lowering of the trunk is chiefly caused by the great reduction of the fore arm, fore leg and metapodials. The hume- rus and femur are respectively only g0 and 110 mm. shorter than in 2. untcornis, while the radius and tibia (typically shorter ele- ments) are respectively 140 and 130 mm. shorter, and the meta- carpals and metatarsals are respectively go and g50 mm. shorter. This limb reduction is very striking. At thesame time the abdom- 58 Bulletin American Museum of Natural History. \|Vol. X, inal girth exceeds that of 2. unicorns, justifying Cope’s conclusion that this animal had rather the proportions of the Hippopotamus than of the Rhinoceros. It will be recalled’ that R. wnicornis has a lower abdominal line than A. sondaicus or R. sumatrensis, or than either of the African Rhinoceroses. 7. fossiger, therefore, had a totally different external appearance from any existing form. JUVENILE, SENILE AND SEXUAL CHARACTERS. There are conspicuous differences in the dentition of different specimens, all of which may be explained as due to influences of growth or sex. In the young calf jaw (No. 2608) the milk cutting teeth are as follows: diz, dcz, dpy. In the young A. sondaicus (R. javanicus), according to de Blainville, we similarly observe two milk incisors. A somewhat older calf of 7. fossiger shows diz, dc;z, dpz. Even in older jaws there is evidence in one case (No. 8391) of two lower incisors upon one side, the formula being: 13-4, Cq, Pz, mz, as indicated by the incisor alveoli. The outer incisors (iz) in the lower jaw tend to drop out ‘at an early age, leaving only the alveoli ; but the vestigial upper incisors (i2) are remarkably tena- cious, although entirely useless. The canines vary strikingly in the sexes. In the females (Nos. 2604-6, 2610-11, 2623), as shown in the photographs, they are of moderate size. In certain males (Phillips Co., Kansas, No. 2612 ; Republican River, Nebraska, Nos. 8391-2) they exceed in size any that have been recorded in other Rhinoceroses living or extinct, as shown in Pl. IVa. No jaw shows any vestige of pz. Ps is present in young jaws, and invariably absent in very aged jaws. Its dehiscence is cor- related with the coming into use of m3. The upper molars, especially the second and third, are extremely hypsodont, the un- worn enamel of the crown measuring, respectively : m2,=m3=. They are reserved for middle and old age. The size of the skull differs considerably in the two sexes, the female skulls (Phillips Co., Kansas, Nos. 2604, 2607, 2622-3; De- catur Co., Kansas, No. 8388; Republican River, Nebraska, No. _ 1 See Sclater ‘On the Rhinoceroses now or lately living in the Society’s Menagerie,’ Trans, Zo6l. Soc., 1875, pp. 645-651. 1808. | Osborn, Skeleton of Teleoceras fossiger. 59 8393) being smaller and less rugose, with less prominent sagittal crests, and decidedly smaller nasals, as shown in Pl. IVa. In old females the nasals acquire a slightly rugose surface, and probably bore a small horn. In the males (Decatur Co., Kansas, Nos. 8385, 8396; Republican River, Nebraska, No. 8420) the nasals become greatly thickened at the extremities (Pl. IVa), forming a vertically compressed plate, which undoubtedly bore a consider- able horn. Differences in size are observed in skulls from various localities, those from Decatur Co., Kansas, and from Nebraska, being larger than those from the Phillips Co. quarry, which are probably due to differences of geological level, the species run- ning into a larger and more robust type before its extermination. Growth-changes in the limbs are especially observed in the close fusion of the fibula with the tibia, and of the remarkable arching over of the sacrum by the superior borders of the ilium ; this whole area above the sacrum forming a solid plate. CONCLUSIONS. T. fossiger may be briefly characterized as a brachycephalic, extremely short-limbed Rhinoceros, partly aquatic in its habits, with a very large brain and no diploé of the skull.’ It parallels the African Rhinoceroses #. s¢mus and &. bicornis, in the form of the humerus, femur and atlas, and in the terminal position of the nasal horn. The occiput, however, is widely different from that of the African Rhinoceroses, as well as of AR. sumatrensis, resem- bling rather that of &. waicornis, although less pitched forward, The limbs are much shorter than in any living type, and, as pointed out by Pavlow,’ at once recall those of &. érachy- pus and R. aurelianensis. A further comparison of 7. fossiger strengthens the resemblance to the latter form. The proportions of the skull, limbs and metapodials are very similar. In both the cnemial crest of the tibia is double; the secondary folds of the superior molars are similar, as well as the general form of the skull. Further details will be given in the writer’s forthcoming Memoir on the Extinct Rhinoceroses. 1 See Scott and Osborn, of c7t.. 1890, p. 93. 2+ Les Rhinoceridz de la Russie et le développement des Rhinoceridz en général,’ Bull. d. la Soc. d. Nat. d. Moscou, 1802. iy ' t . } Se. a * lide o® q M PS eer i A) } ‘ozis [einjeu Yi9NuaMI-3u(¢) ‘sasissof sv4aI042f AO NOLATANS GALNNOW “AJ 3LV1Id X “10A ‘HN CW CV Nivating (Z19z ON) MV[ GNV (96£g ‘ON) TIONS ATVIN ‘(g0gz “ON) Mv[{ GNv (£6€g ‘ON) TIONS ATVINAY “4a8USSOf SV1LAI0IA ‘VAI ALVId ‘X “I0A ‘HON ‘WY Nizai1ng wz i “nF ‘otes upon the Locomotio ; } ata) -Fairristp Osporn *S EDITION, extracted from BULLETIN P- 74- 1808. | Wortman, Extinct Camelide of North America. 95 of some of the material, and in part to other lines which resemble them in certain points of skull and limb structure. These resemblances are, no doubt, due to the close proximity to the point at which the respective phyla began to diverge. For the purpose of bringing into stronger relief the characters of the Cameloids of this horizon, it is necessary to compare them accurately with the cotemporary Selenodonts, and, since several of them apparently represent new genera, they are herewith described. Leptoreodon marshi,' gen. et sp. nov. This genus and species is represented in the collection by an almost perfect skull in good state of preservation, a number of vertebra, and a few fragments of the limbs (No. 2064), which I use as the type. There are several other specimens of a more frag- mentary character which are probably to be referred to the same Fig. x. Side view of skull of Leptoreodon marshi. genus and species, but they contribute little additional informa- tion to the knowledge of the skeleton. The genus differs from all the Oreodonts hitherto described in the possession of a short diastema in front of, and a longer diastema behind, the first superior premolar, together with a considerable diastema between the first and second premolar in the lower jaw. ‘The incisors are 1 This species is dedicated to Prof. O. C. Marsh, in recognition of his numerous contribu- tions to American paleontology. 96 Bulletin American Museum of Natural History. |Nol. X, present in full in both jaws; the inferior canine is small and incisiform ; the superior canine is large, with the characteristic D-pattern of the Oreodonts on cross section, and the first inferior premolar is enlarged and caniniform as in the Oreodonts. The first superior premolar is two-rooted with a high, compressed cutting crown, the second is similar but somewhat larger, the third has a principal broad, lunate external cusp and a faint internal cingular ledge. The fourth premolar crown is composed of a single external and internal cusp, much as in the Oreodonts. The superior molars closely resemble those of Pvroforeodon (Zomeryx) in the composition of the crown, so far as can be determined in their advanced stage of wear in the type specimen. It is impossible to say whether or not there were anterior inter- mediate cusps present, but judging from certain appearances in this region of the crown, I am inclined to think that less worn teeth would show them. The mesostyle consists of a vertical pillar as in the Oreodonts generally, and not of a wide open loup as in Agriocherus. In the lower jaw the incisors and canines are of the typical oreodont pattern, but they are unusually procumbent in position. The first premolar is enlarged and caniniform, the second simple, the third with a small internal cusp and posterior heel, and the fourth similar in pattern, except that the internal cusp is smaller and the heel more pronounced. ‘The lower molars are almost identical in structure with those of the early Oreodonts. The whole skull differs from that of the Oreodonts in its more slender proportions. ‘This is particularly noticeable in the lower jaws, which are relatively long and shallow, especially in the region of the symphysis, in marked contrast with the deep and abrupt chin of the Oreodonts in general. There does not appear to have been a preorbital pit present, and the orbit was not enclosed by bone posteriorly. ‘The present genus may be distin- guished from its contemporaries in the following dental characters, viz.: from Proloreodon (Lomeryx) in the possession of diastema in both jaws and the full number of incisors in the upper jaw.’ 1 In all of our material [ have not yet seen a specimen among the Oreodonts other than Leptoreodon that has a full set of incisors in the upper jaw. Marsh figures the type of Eomeryx pumilis with but two superior incisors, and if Protoxeodon has the full complement as believed by Scott, then the two genera are certainly distinct. In two specimens in the Musuem collection which correspond closely with Protorcodon parvus, as described by Scott there is but a single incisor on each side above, and the premaxilla are widely separated from each other in the median line, 1808. | Wortman, Extinct Camelide of North America. 97 From AHyomeryx it is readily distinguished by. the full number of superior incisors and by the diastema, although it resembles this latter genus, which is described by Marsh’ as having more slender jaws than Protoreodon (Homeryx). From the cameloid, Leptotragulus, it is easily separated by the numerous oreodont characters which the skull exhibits, although the symphyseal region is strikingly similar in the two genera. Of the hind foot, the cuboid, navicular and the head of the third metatarsal are sufficiently preserved to afford characters for identification. These bones indicate an animal with far more slender limbs and feet than any of the Oreodonts with which I am familiar. The navicular has an inconspicuous posterior hook unlike that of the Oreodonts, and, judging from the much reduced facet on the cuboid, the fifth digit was considerably diminished in size if not entirely rudimental. The limb-bones are not well enough preserved to confirm or negative this conclusion of the slender and delicate proportions of the animal, but, upon the whole, I think it may be safely concluded, from the evidence at hand, that Zepforeodon held the same position with reference to the American Oreodontide that Xvphodon did to the European Anoplotheriide. The second genus to be described in this connection contains species somewhat smaller in size and less perfectly selenodont. Bunomeryx montanus, gen. et sp. nov. There are two specimens in the collection which I classify under this head, viz. : an anterior portion of a cranium somewhat crushed, containing the maxillary dentition complete upon one side, together with the greater part of the left mandibular ramus of the left side having all the true molars and the last premolar in good preservation (No. 2071). ‘The second specimen consists of a portion of a lower jaw with a few teeth, the posterior part of the cranium, a nearly complete fore foot, portions of the hind limbs and other parts of the skeleton (No. 2070). The first of these specimens may be taken as the type, but there can be very 1‘ Descriptions of Tertiary Artiodactyles,’ Amer. Jour. Sci., Vol. XLVIII, Sept., 1894, p. 268. [April, 1898.] vis 98 Bulletin American Museum of Natural History. |Vol. X, little doubt that the second specimen is identical with the first and can be regarded as a collateral type. This genus most nearly resembles HYomacodon Marsh, from the Bridger Beds, although it presents some dental characters similar to Dichobune of the European Eocene. The more important generic characters may be stated as follows : Dentition, 13, C1, Pm4, M3. Superior molars, having well-defined cres- centic outer cusps and a distinct mesostyle and parastyle ; first molar provided with two conic internal cusps (protocone and hypocone), with anterior and poste- rior subcrescentic intermediates; second molar having anterior subcrescentic inter- mediate, subconic protocone, a posterior subcrescentic intermediate and no hypocone ; third molar similar to second. ‘The superior premolars are present in full number ; the two anterior have simple cutting crowns, while the crowns of the third and fourth are made up of single external and internal conic cusps well developed. In the lower jaw the structure of the molars is intermediate between the bunodont and selenodont pattern ; there are only three premolars, the anterior two of which have simple compressed crowns, while the last or fourth of which is provided, in addition to the principal cusp, with anterior and internal cusps, together with a well-defined heel. As compared with Homacodon, Bunomeryx is readily distin- guished (1) by the possession of three premolars in the lower jaw ; (2) by the crescentic character of the external cusps of the superior molars ; (3) by the presence of a well-developed para- style and mesostyle ; (4) by the absence of the hypocone on the second superior molar, and (5) by thesubcrescentic character of the intermediates. (6) The internal cusp of the third superior premolar is, moreover, better developed in Bunomeryx than in Hlomacodon, and (7) the fourth inferior premolar is much more advanced in structure. The structure of the inferior molars is much more selenodont in Bunomeryx than in Homacodon. From Dichobune the present genus is readily distinguished by the absence of all traces of the anterior cusp of the trigon in the lower molars as well as the more crescentic character of the outer cusps of the superior molars, and the possession of well-defined mesostyle and parastyle. The complete adult dentition of Décho- dune is apparently not known, but there can be but little doubt that it had the full number, forty-four teeth, in which case Buno- meryx would be sharply distinguished by the inferior premolar formula. I cannot at present say in what manner Bunomeryx 1898. | Wortman, Extinct Camelide of North America. 99 differs from the European Dezlothertum, Spaniotherium, Metrio- theritum, Moutllacithertum and Oxacron of Filhol, which are placed by Zittel in the subfamily Dichobuninz. On account of the very imperfect knowledge we have of these forms, no com- parisons are at present possible. In specimen No. 2071, the upper and posterior portion of the cranium is sufficiently preserved to indicate a relatively high over- hanging occipital and a strong sagittal crest, the latter dividing into two well-marked lateral postorbital branches. In advance of the point of division of these two branches a strong ridge is continued forward upon the frontals in the median line as in many of the lower forms of the Selenodonts. The postorbital process is well developed, but it does not join the molar, so that the orbit is not enclosed by bone posteriorly. There is no evi- dence of the presence of any long horn-cores. Of the fore limb, the distal ends of the ulna and radius are pre- served, but they are considerably crushed. ‘There is apparently little or no tendency to codssification of the bones, although the shafts are closely applied to each other in the lower third of their extent. The articular end of the radius shows distinct facets for scaphoid and lunar, but does not touch the cuneiform. The distal end of the ulna articulates solely with the cuneiform. The carpus is of the typical artiodactyl pattern, and especially resembles that of the earlier Selenodonts. In the proximal row the cuneiform rests exclusively upon the unciform, the lunar about equally upon magnum and unciform, while the scaphoid is supported below by magnum, trapezoid and trapezium. In the distal row the unciform articulates distally with Mt. III, and to a slight extent with Mt. II. In the modern Suillines, the Cameloids and the later Oreodonts, the second metacarpal has lost all con- nection with the magnum, but in the early Oreodonts Mt. IT still retains a contact between these two bones. In /Pvofoceras of the Oligocene a very minute contact is observable. The trapezoid is free, and shows no tendency to unite with the magnum as in Leptomeryx, the later Tragulines and Pecora. The trapezium is not preserved in the specimen, but judging from the well-marked facets upon the scaphoid, trapezoid and Mt. II, it is quite certain that it was not only present and of good size, but 100 Bulletin American Museum of Natural History. |Vol. X, that it supported a very considerable vestige of the first digit. It would not indeed be a matter of surprise to find this digit com- plete in more perfect specimens very much as in Oreodon. There are four metacarpals preserved of which the median ones, Mt. III and IV, are the largest and subequal in size. Mt. II is slightly larger and longer than Mt. V, and in all of them the distal. keels are confined to the palmar surfaces, as in all primitive Ungulates. The phalanges of the fore feet are not known. Of the hind limb the materials are not so complete as of the fore limb, but enough is preserved to make out its more important characters. The fibula was much reduced, and probably incom- plete in the middle part of the shaft. The distal end of the tibia displays no usual form of the more generalized Selenodonts, as do the tarsal bones. The cuboid and navicular were not coossified, and there is evidence of four complete metapodials, the lateral ones, however, being unusually slender and delicate. The first two phalanges resemble those of the early Cameloids, Protoceros and Leptomeryx in their form, as do likewise the unguals in being relatively high-pointed and _ flattened upon their opposed surfaces. Bunomeryx elegans, sp. nov. A second species of this genus is indicated in the collection by a portion of a cranium containing the last three premolars and the molars, in excellent preservation, together with both mandibu- lar rami bearing all of the teeth with the exception of the incisors and canines. The most important difference be- tween the two species is seen in the presence of a short diastema between premolars two and three of the lower Fig. 2. Crown view of upper and lower teeth of PBunomeryx jaw in B. elegans. In f53, Mmontanus elegans. hy., hypocone; fr., pro- . : : tacone - $2; protoconale, this diastema is absent, and the teeth of the lower jaw were apparently in a continuous series or closed row. &. edegans is smaller and more delicate than B. montanus, a fact that is not only indicated by all the teeth but particularly emphasized by the fourth inferior pre- 1898. | Wortman, Extinct Camelide of North America, Ol molar, which is, considerably narrower and has a less development of the internal cusp. Another important distinction between the two species is found in the greater development of the vestigial hypocone of the second superior molar of B. elegans. In 4. montanus this cusp has almost entirely disappeared, the only evidence of its presence being indicated by a cingulum in this portion of the crown. One fact of great interest in connection with this genus is the probable light which it throws upon the homologies of certain cusps of the molar crown in the higher selenodont Artiodactyla. It is here that we witness the actual passage from the bunodont to the selenodont type of molar in this important group. If we can trust. the evidence before us, Bunomeryx is a direct lineal descendant of the Bridger Homacodon, and it is a matter of the utmost moment to note that in the latter genus there are six fully- developed cusps upon the crowns of the first and second superior true molars ; in the third there are only five cusps present. In Bunomeryx, as already indicated, the full six cusps are found on the first superior molar only, while in the second molar there is ‘but a vestige of the postero-internal cusp or hypocone. ‘The evidence appears to be conclusive, therefore, that the true homo- logical hypocone is in process of retrogressive disappearance, and in proportion as this cusp is reduced, the posterior intermediate is pushed out to take its place. As a further evidence of the truth of this proposition it may be stated that the true hypocone of both the first and second molars of Homacodon, as well as the first molar of Sunomeryx, exhibits no tendency whatever to develop a selenodont structure, while the posterior intermediates especially in Bunomeryx, exhibit very decided advances in this direction. The very position of this cusp, moreover, precludes ‘any possibility of its entering into the formation of the single posterior internal crescent of the more perfectly developed seleno- dont molar of the higher types. I believe therefore that the history of the formation of the four crescents of the superior molar crowns of the Selenodonts has been as follows, tracing it from the five-cusped Pantolestes’ of the 1 From this genus I exclude the type of Paztolestes etsagicus Cope as belonging to a distinct genus ancestral to and leading directly up to the bunodont Artiodactyla. It is very probably synonymous with Zohyus distans of Marsh, who properly placed it among the Bunodonts. 102 Bulletin American Museum of Natural History. |Vol. X, Wasatch. The primitive condition of this oldest type of the Artiodactyla was two buniform external cusps, two buniform intermediates, together with one large more or less lunate internal cusp, flanked by a rudimental postero-internal cusp which is clearly shown to be an outgrowth from the cingulum. There is no evidence that this postero-internal cusp was ever developed on the last molar, because in all the forms from Pavtolestes to Buno- meryx it is persistently absent. The next step consisted in the reduc- tion of the large lunate internal cusp and the full development of a well-marked postero-internal cusp, or hypocone, on the first and second molars. ‘This condition is seen in, and is character- istic of, Homacodon. The third step consisted in the disappear- ance of the true hypocone and the gradual usurpation of its place and function by the posterior intermediate in the crown of the second true molar, a condition seen in Aunomeryx. As a fourth step in this development one can readily imagine this process extended to the first true molar, when it would be complete. This hypothesis may be objected to on the ground that Awno- meryx cannot stand as the direct ancestor of any of the Seleno- donts at present known, on account of its reduced premolar dentition in the lower jaw, but if we are to regard the type of superior molar exhibited by either Homacodon, Dichobune or Helo- hyus as the one which preceded, and from which was derived the tetraselenodont or four-crescented crown, then this hypothesis must be accepted as true. The only case so far known wherein the true hypocone has been preserved and has become crescentic, is in Cenothertum and Plestomeryx,and here we have three well-developed crescents upon the posterior moiety of the crown, of which the inner one repre- sents the hypocone and the middle one the posterior interme- diate. It is possible that the cusps of the two anterior superior molars of X7phodon are to be interpreted in the same way, and that the posterior innercrescent is composed solely of the posterior intermediate, the true hypocone having come to occupy a more anterior and median position. In this case the anterior internal crescent would be made up_of protocone and the anterior inter- mediate. Future discovery will no doubt reveal considerable variety in the formation of the internal crescents in the various 1898. | Wortman, Extinct Camelide of North America. 103 phyla of the Selenodonts, but it appears to me certain that the hypothesis herein advanced is the correct one for the formation of the tetraselenodont superior molars of the Cameloids, Pecora, Tragulines, and probably the Oreodonts and Anthracotheres. Parameryx (Leptotragulus) proavus 5S. & O. This genus was first described by Marsh’ and later by Scott and Osborn,’ who considered that it to belongs to the Traguline division of the order. Later Scott gave a fuller account of it’ and placed it in the Tylopoda’ immediately ancestral to Poébro- thertum of the White River Oligocene. The materials in the Museum Collection do not add very materially to the knowledge of this form ; however, there are some important points to be made out from it. There are four specimens which I refer to this species, the most important of which are a fragmentary skeleton containing a fairly good hind foot, together with the posterior part of the last lower molar (No. 2509). The other specimens pertain exclusively to the lower jaw (Nos. 1803, 1805 and 1808). In the lower jaw there is one diagnostic character by means of which the last lower molar can be recognized, and that is the presence of an extra cusp upon the inner border of the heel near its point of junction with the postero-internal cusp. It is by means of this character alone that I associate the fragmentary skeleton with this species. The lower molars are of the typical selenodont pattern, and the cusps more elongated than in any of the cotemporary Selenodonts. The inferior premolars are three in number, the fourth being pro- vided with a well-developed internal cusp and heel. The second and third are simple and without accessory tubercles. In advance of the second premolar there is a considerable diastema, in front of which is the large procumbent alveolus for the canine. The incisors are not preserved, and this region of the jaw is so much broken as not to reveal their alveoli. 1 “Introduction and Succession of Vertebrate Life in America,’ 1877. _ ; ‘ 2‘ Preliminary Report on the Vertebrate Fossils of the Uinta Formation,’ Proc. Am. Philos. Soc., 1877, pp. 255, 264. , i a : i 3 ‘Mammalia of the Uinta Formation,’ Trans. Am. Philos. Soc., N.S., Vol. XVI, Part 1ii, Aug. 20, 1896, pp. 479-486. f 4 Marsh had, however, clearly recognized the affinities of this genus with the Tylopoda ten years previously, since we find in the address above quoted the following statement : * A most interesting line, that leading to the Camels and Llamas, separates from the primitive seleno- dont branch in the Eocene, probably through the genus Parameryx. 104 Bulletin American Museum of Natural History. [Vol. X, Of the bones of the hind foot, the entire tarsus is preserved with the exception of the cuneiform. ‘These parts of the skele- ton present a most striking resemblance to those of Poébrothe- rium in all the details of their structure, the only difference discoverable being that of size. The third metatarsal is present but unfortunately a small part of the shaft is missing so as not to exhibit its full length ; there is enough, however, to indicate that it was unusually long and slender, much flattened upon the sur- face which it offered to the second metatarsal, and that the form of the shaft, moreover, had that peculiar squarish outline upon cross section, a feature so highly characteristic of the Oligocene Cameloids. Another distinctive cameloid feature is seen in the increased size of the medullary cavity. The lateral or fifth meta- podial was reduced to a mere splint, as is indicated by the much- reduced facet upon the cuboid ; this facet is relatively as small as it is in the cuboid of Poébrotherium. The phalanges have about the same proportions and shape as the corresponding bones of the White River species. That Parameryx (Leptotragulus) was a member of the Tylo- poda, as has already been pointed out by Marsh and Scott, there can be very little doubt, but at the same time the evidence is equally conclusive that it does not stand in direct ancestral line with the succeeding Poébrotheres. The evidence against such a conclusion is to be found in the fact that Parameryx (Leplotraga- /us) has only three premolars in the lower jaw, an enlarged canini- form canine and relatively short, thick inferior premolars, the last of which, or fourth, has a considerable development of the internal cusp. It may therefore be looked upon as a precociously specialized side branch which died out at the close of the Eocene and left no modified descendants. Protylopus petersoni,’ ven. et spec. nov. This genus and species is primarily founded upon the anterior portion of a skull from which the left ramus is missing. The specimen is broken obliquely in such a manner as to show upon the right side all of the facial portion, including the orbit and the 1 This species is named in honor of Mr. O. A. Peterson, whose explorations of the Uinta Beds have been attended with such marked success. u : i ce ct aly 1808. | Wortman, Extinct Camelide of North America. 105 anterior root of the zygomatic arch, while upon the left side the greater part of the orbit is missing. Fortunately the skull contains the dentition nearly complete. In association with it were found the greater part of an ulna and radius of the same individual, A second specimen which I refer to this genus and species includes a large part of both hind legs, together with a large number of vertebrz, ribs and other parts of the skeleton. A Ty ray 4 Mii - ‘ Fig. 3. Side view of skull of Protylopus petersoni. third specimen contains hind limbs and vertebrae, while a fourth includes the greater part of a hind foot. The more important generic characters may be stated as follows : Molars tetraselenodont without intermediate cusps. Teeth of the typical number, forty-four, arranged in a continuous series. Canines of both upper and lower jaws small and incisiform, the first inferior premolar not caniniform. The first three superior premolars elongated from before backwards, secant and without accessory cusps, the fourth with single external and internal crescentic cusps. The inferior premolars elongated and cutting, the fourth without internal cusps. Hind feet provided with but two functional digits, the outer ones, second and fifth, reduced to mere vestiges. | Lumbar vertebral formula 7. Ulna and radius, at least in old individuals, coéssified in the middle part of their shafts but free at their proximal ends. The skull is crushed laterally so as not to reveal the exact form of the face, but it can be safely stated that the muzzle had 106 Bulletin American Museum of Natural History. |Vol. X, moderate length, with slightly overhanging nasals, much as in Poébrotherium wilsont. The premaxillz are relatively broad and extend upwards and backwards to articulate with the nasals. The orbit is not enclosed by bone posteriorly, but exhibits a marked tendency towards that peculiar roofing so highly charac- teristic of Pocbrotherium and the later Tylopoda. In advance of this bony shelf is seen a faint though distinct indication of the supraorbital notch, so constant a feature of the cameloid skull. The lower jaws may be described as long and slender, with a considerably elongated symphysis. The superior incisors are relatively small, of a more or less conical form, and directed downwards. ‘The premaxillae were apparently not in contact in the median line. The superior canine is but little larger than the outer incisor, of a more or less hook-shaped appearance, and provided with a distinct sulcus upon the outer portion of the crown as in Poébrotherium wilsont, The first premolar follows after a very short interval and, like the second, is a simple two-rooted cutting tooth. The third premo- lar has a faint internal cingular ledge, while the fourth, as already mentioned, is provided with single external and internal crescents. The molars are much worn, and do not show clearly whether or not intermediate tubercles were present, but I think it may be safely assumed, from the general appearance of the crown, that they were absent. In thesecond and third molars, between the internal crescents, is to be seen a small styliform cingular cusp which is entirely absent, so far as I can determine, in Poébro- thertum. In the lower jaw the incisors are of a more spatulate form and more procumbent in position. As in the upper jaw, the canine is slightly larger than the outer incisor, but of a very marked incisiform pattern. After a very short interval or diastema, is placed the first premolar, a two-rooted tooth whose crown closely resembles that of the canine, the two teeth being about equal in size. The second and Fig. 4. Crown view of lower teeth of Protylopus petersont. third premolars have elongated secant crowns like the Tragulines. 1898. | Wortman, Extinct Camelide of North America. 107 The fourth has a well-marked heel and anterior basal cusp, but there is apparently no internal cusp present. The molars are so much worn that their structure is not very apparent. There can be very little doubt however that they had the usual structure. In the heel of the last molar a prominent accessory cusp is seen upon the border of the inner side near the point where it joins the lower posterior inter- nal cusp. In Poébrotherium this cusp is clearly present, but it has fused with the postero-internal, producing a prominent angle at this portion of the crown. In per- fectly unworn teeth of Poébrothertum wit- soni, it can be readily demonstrated to be an independent cuspule. Of the vertebra, unfortunately, no cervi- cals are known, consequently it is impossi- ble to say whether they exhibit the peculiar features of the more typical Camelidz or not. The dorsals are well represented in specimen No. 2564, the whole series being present, with the exception of the first three or four, together with all the lumbars locked in position. The vertebra resemble those of the modern Llamas closely in their general proportions. The bodies of the anterior dorsals are but moderately keeled, and towards the posterior end of the series strongly keeled; they increase gradually in size from before backward. ‘The neural spine of the fifth is long and recurved, those of the succeeding dorsals decreasing in length posteriorly. ‘The neural spines of the last two are considerably shorter and broader, having an almost vertical direc- tion. The rib facets in the anterior region have their usual relations and positions, the ribs articulating with the vertebra by two Vertebre of Protylopus petersoni (?) 108 Bulletin American Museum of Natural History. (Vol. X, distinct facets, but in the last two the capitular and tubercular facets appear to be fused together as in these dorsals of the Llama. The lumbars are seven in number, the constant formula for the Tylopoda; they resemble closely the corresponding bones of Poébrotherium and the later Cameloids. The sacrum is composed of only four vertebrze, but it is highly probable that another one or two was added from the caudal region as age advanced, just as in Poébrotherium and the modern Llamas. The three anterior vertebrae of the sacrum have very reduced neural spines, while in the fourth the spine is well developed. ‘The ribs do not display any characters of especial importance. The pelvis is in a very fragmentary condition, but it may be stated that the ilium is well expanded, and, so far as one can judge, the whole bone would correspond closely with that of Poébrotherium. Yhe femur is present in its entire length with both ends in a good state of preservation, although the shaft is somewhat crushed. The proximal end has practically the same relations and arrangement of the different parts as that of Poébro- therium and other members of the group. ‘The distal end thus early gives slight though conclusive evidence of the peculiar and characteristic appearances which this part of the bone assumes in the later Camelidze. This is especially seen in the great extension of the condyles backwards behind the median line of the shaft as well as the forward projection of the borders of the rotular groove, which serve to increase the antero-posterior diameter of this part of the bone. Although not clearly indicated on account of crushing, yet there seem to be distinct traces of the beginning of that peculiar depression at the proximal end of the rotular groove so highly characteristic of the later Tylopoda. In a like manner the patella has begun to assume the distinctively cameloid form by the great elongation of its lower border into a long, pointed process. The tibia, which about equals the femur in length, shows a great resemblance to that of Poébrothertum. ‘The cnemial crest is unusually well developed, and extends quite one-third of the way down the shaft. The fibula is much reduced, and although the specimen does not show whether or not the shaft was com- plete, the probabilities are that it consisted of a distal portion — 1898. | Wortman, Extinct Camelide of North America. 109g only. That part of the shaft which is preserved is very slender and closely applied to the shaft of the tibia. In the hind foot the tarsal bones have nearly the same relations as in Poébrotherium. The tuber of the calcaneum is somewhat shorter proportionately than in the White River genus, but otherwise both the caleaneum and astragalus are strikingly alike in the two genera. The cuboid of Protylopus is slightly narrower in proportion to its height than the corresponding bone in Poébrotherium, and the navi- cular is provided with a much better developed pos- terior hook. As in Poébro- therium, there are two cunei- forms present, the inner of which is a vestigial nodule of bone only. There are but two func- tional metapodials, the third and fourth, the second and fifth being reduced to mere vestiges. Upon one side the vestige of the second meta- podial is preserved in place, and it is seen to articulate . : Fig. 6. Left hind foot, femur, tibia and fibula by a peculiar ledge-like facet, , of Protolypus petersoni (?). cad., calcaneum ; as., a: . _astragalus; cé., cuboid; za., navicular; cu®., ex- upon the principal cunel-_ ternal cunciform; /., head of femur; 7., greater 3 ; trochanter; ¢7?., lesser trochanter; ¢., tibia; /@., form. Upon its posterior fibula. ; surface is a distinct facet by which it articulates with the small cuneiform. The remnant of the fifth is not preserved, but the facet by which it articulates with the cuboid is very small, and there can be no doubt that it was as much reduced as the second. The functional metapodials are relatively much shorter than in Poébrotherium, and of a considerably more primitive form. They are well flattened upon their opposed surfaces in the upper half of the extent of their shafts. Below this the inner surfaces of the two bones are well rounded. Unlike the metapodials of Poébrotherium, they lack that characteristic four-sided appear- TIO. Bulletin American Museum of Natural History. |Vol. X, ance of the later Camels, but on the contrary, are more or less triangular upon cross section, especially in the proximal half of their shafts. The metapodials as well as the long bones show their cameloid affinities in the unusually large size of the medul- lary cavities. The phalanges exhibit comparatively few differ- ences from those of Poébrothertum, the unguals being flattened upon their opposed surfaces. The fore foot is entirely unknown, but itis highly probable that it will be found to possess four complete functional toes. It may transpire that the association of this skeleton with the above-described skull is incorrect, and that these bones belong to separate and distinct species; however, they agree so well in the matter of proportionate sizes of the different parts, and both are so distinctly cameloid, that I am persuaded to believe that they refer to one and the same species. It may be noted here, how- ever, that in one of the specimens referred to above (No. 2067), there is evidence that at least one of the lateral metapodials of the hind foot was complete though very slender, and should prob- ably be referred to another species on this account. The bones are, moreover, somewhat more slender and delicate than the one here described. At all events, whatever form of skull belongs with these skeletal parts it is nevertheless certain that the skull of Protylopus, above described, is just such a type as is required to satisfy all the necessary conditions in order to occupy a position in direct ancestral relation with Poébrotherium. ‘The true Tylo- pod phylum is therefore traceable directly to it. Beyond this, there is at present no satisfactory evidence to establish, with any degree of certainty, the identity of the true Camel pedigree. Poebrotherium /ezdy. With a consideration of this genus we pass from the Eocene to the Oligocene representatives of the group. It was estab- lished by Leidy as early as 1847 upon an imperfect skull presen- ted to the Philadelphia Academy by Mr. Alexander Culbertson of Chambersburg, Pa., who was at the time engaged in the western fur trade. It was among the first of the mammalian fossils from the remarkable Bad Lands of the Cheyenne River 1898. | Wortman, Extinct Camelide of North America. 11 region, whose treasures were destined in later years to play such an important part in the development of American palzontology. Leidy at first’ regarded the skull as pertaining to a genus nearly allied to the Musk Deer, but later pointed out its true position among the Camelidez. The generic differences between Poébrothertum and Protylopus are not great, and indeed it would appear at first sight that they are insignificant. It is more than probable, however, as stated Fig. 7. Side view of skull of Poébrothertum wilson?. above, that Protylopus will be found to have four complete and functional digits in the fore hmb. So far as our knowledge extends at the present, the chief distinctions are as follows: In Poébrothertum the molars are much more selenoid and the crowns more lengthened than in Protylopus ; the third superior incisor is larger than the superior canine; the ulna and radius are firmly coossified, even before the epiphyses of the bones are joined to the shaft and the shaft of the fibula has completely disappeared. Poebrotherium wilsoni Ze/dy. This species, although very abundant in the White River Beds of the Cheyenne River region, has not been very fully described. All of the specimens in the Museum collection have been found in the Lower Oreodon level, and it is doubtful if the vertical range of the species extends much above this point. It differs very markedly from its successor, P. Zabéatum, in the practical 1*Ancient Fauna of Nebraska,’ Dec., 1852, p. 2* Extinct Mammalian Fauna of Dakota Piel Nebrasiad? 1860, p. 141. Ll2 Bulletin American Museum of Natural History. \Vol. X, absence of diastemata in the lower jaw. The canines of this series are, moreover, broad and incisiform, being separated from the first premolars by very short diastemata. In the same man- ner the second premolars follow after a very short interval. In Fig. 8. Fore and hind foot of Poébrothertum wilsont. Fig. 9. Ulna and radius and humerus of Poébrothertum wilsont. Fig. ro. Femur and tibia of PoébrotherZum wilsonz. P. labiatum, on the other hand, the lower canines and outer incisors are almost in contact, the canines are subcaniniform in shape, and there is a short diastema in front of, and a long dias- tema behind, the first inferior premolar. The bones of the limbs and other parts of the skeleton are, as far as can be deter- mined, very much alike in the two species. As in /. /abiatum, there is a considerable range in size in the various specimens referred to this species. 1808. | Wortman, Extinct Camelide of North America. 113 Poebrotherium labiatum Cv/e. The type of this species consists of the larger part of a skele- ton of a single individual from the White River Beds of north- eastern Colorado (No. 6520). Associated with this specimen are two almost complete lower jaws from the same locality (Nos. 6517, 6518) showing the characteristic diastemata of P. /abia/um, but considerably smaller. These specimens were erroneously referred by Cope to P. wz/sont. I have not been able to correlate with certainty the level from which these specimens were taken, with that in which similar remains in the Cheyenne River region occur, but judging from Cope’s unpublished sketch of the section of the bed, there can be httle doubt that it corresponds closely with the upper part of the Oreodon horizon. ‘This surmise is strength- ened by the fact that there is one specimen in the collection (No. 638), from the extreme upper part of the Oreodon Bed, which agrees in every way with the type of P. /abiatum, except that it is a little larger. Another specimen from the Cheyenne River Bad Lands includes a lower jaw and a good part of the skeleton. The lower jaw exhibits the characteristic diastemata of P. labiatum, but is much smaller than the type, and of the same size as the two jaws mentioned above. Unfortunately the exact level of this specimen is not known, but it has every appearance of having come from the upper part of* the Oreodon stratum. Whether or not these smaller specimens are to be referred to a species distinct from P. Zabéatum is a matter which requires a greater amount of material than we at present possess in order to decide correctly. So far as one can determine at present, the only distinction between the two is one of size, and this is not great. I have thought best to regard them as belonging to the same species until other differences are shown to exist. Taken as a whole, P. /aééatum, as exemplified by the larger individuals, was considerably larger than /. qw¢/sonz, and in the possession of diastemata in the lower jaw, as well as the more caniniform shape of the lower canines, makes a distinctive approach to the species from the John Day Beds. In this connection it is proper to observe that no remains of Camels are known from the Proto- ceras level of the White River Beds. When such are found they | April, rS98.\ S 114 Bulletin American Museum of Natural History. [Vols will probably establish a complete transition between /. dabiatum and the John Day species. Gomphotherium Co/e. It is especially to Cope that we are indebted for the discovery of Camels in the John Day Beds. The first remains secured by him from this horizon were referred to Poébrotherium, but later he established the genus Gomphotherium' for their reception, ey. ET) inp | Mh MMM 7 | Wi Ming AY), (\ Fig. rr. Side view of skull of Gomphotherium sternbergi. which he distinguished from Poébrotherium by the more simpli- fied character of the crown, and the one-rooted condition of the first superior premolar. As this distinction was founded largely upon error, | am now able to give the more important and true characters which serve to separate the two genera in a satisfactory manner. In Poébrotherium the inferior canine is either in contact with the outer incisor, or is separated from it by a very short dias- tema, and the form of the canine is either like that of an incisor or very imperfectly caniniform. In Gomphotherium, on the other hand, the inferior canine is either separated from the outer incisor by a very distinct diastema or the diastema is absent, and the shape of the canine is strongly pointed and recurved, as 1° The Phylogeny of the Camelida,’ Amer. Nat., 1886, p. 618. 1898. | Wortman, Extinct Camelide of North America. UTS in many of the later Camelide. In Poébrotherium again, the orbit is not inclosed by bone posteriorly, whereas in Gomp/ho- thertum the posterior boundary of the orbit is complete. Another important distinction is seen in the character of the articular facets of the third and fourth metapodials of the fore foot. In Poétbrotheritum these bones give evidence of having been more widely separated in the living animal, and capable of consider- able independent movement, the facets being relatively large and the opposed surfaces comparatively smooth. In Gompho- thertum these facets are much reduced, the metapodials closely applied to each other and their contiguous surfaces much rough- ened, clearly foreshadowing the codssification of these elements into a cannon bone. Gomphotherium sternbergi Copc. The type of this species consists of the greater part of the skeleton of a single individual in good preservation from the lower beds of the John Day Valley, Oregon. Other specimens from the same horizon include more or less perfect foot-bones, fragments of jaws, and other parts of the skeleton. The form of the skull presents a striking resemblance to that of the modern Camels in its general make up. The nasal bones are, however, proportionately longer, the bony roof of the orbits not so broad, and the muzzle apparently more laterally constricted in front of the infraorbital foramen. As compared with Poébrotherium and the Llama the face is less bent down on the basicranial axis, in this respect resembling more the skull of the Camel. ‘The verti- cal depth of the face immediately in front of the orbit is rela- tively greater than in Poébrotherium, and the opening of the posterior nares has a more forward position. A very interesting transition from the relatively low, much- swollen otic bulle of Poébrotherium, to the high, little-swollen condition of these parts in the living species, is observable. The otic bull of all the Camels are highly characteristic ; they con- sist of an inner, longitudinally-directed swollen part, together with an outer vertical buttress, which joins the inner part at an angle, and at the upper limit of which ts placed the external 116 Bulletin American Museum of Natural History. |Vol. X, auditory meatus. Immediately behind the point of junction of these two parts is seen the deep recess where the hyoid arch is articulated to the skull. In Poébrotherium the inner portion of the bulla is much the larger, and the recess for the tympanohyal is inconsiderable. In Gomphotherium the two parts are about equal in size, and the tympanohyal recess much more pro- nounced. In the living genera, Came/us and Auchenia, the nner Fig. r2. Humerus, ulna and radiaus of Gomphothertum sternbergi. part of the bulla is much reduced and the tympanohyal recess is converted into a deep circular pit surrounded by bone. In the skeleton of the limbs the lower end of the femur is peculiar in the unusual size and development of the areas of attachment of the outer and inner heads of the gastrocnemius. This same peculiarity is seen in the femora of old individuals of both Poébrotherium labiatum and Procamelus occidental’s, although to a somewhat less extent, and is doubtless a result of age. The head of the humerus shows the first distinctive change leading to the development of the double bicipital groove, a feature so char- 1808. | Wortman, Extinct Camelide of North America. 1g) acteristic of the later Camelidz. In ne individual in the collec- tion is there evidence, even in those of the most advanced age, of any traces of bony union of the metapodials. In size G. sternbergt exceeded P. labiatum by at least one-third. Fig. 13 Fig. 14. Fig. 13. Femur, tibia and fibula of Gomphothertum sternbergt. Fig, 14. Hind foot of Gomphotheriun sternbergi. Gompotherium cameloides, sp. nov. This species 1s represented in the collection by an almost com- plete mandibular ramus from the uppermost levels of the John Day deposits (No. 8179). To this same species I also refer an upper-dentition (No. 7915), an almost complete fore limb (No. 7912), as well as several other fragments. The chief distinctions 118 Bulletin American Museum of Natural History. |Vol. X, Fig. 15. Lower jaw of GomphotherZum cameloides. between this species and the older G. sternbergi are seen in the increased size and the absence of a diastema between the lower canine and the outer incisor in G, cameloides. The comparative measurements display these differences in size at a glance ; they are as follows: G. sternbergt. G. cameloides. MM. MM. Length of sup. ms. and three posterior pms. .. 60 83 a “inf. ms. and three posterior pms. .. 65 97 eb MRCHtiLe nme OnmCentt toni —me-lenee 110 170 os PeAntehiOnmmetapocdialSa-wec create. 180 228 Fig. 16. Upper teeth of Gomphotherium cameloides. It will therefore be seen that G. cameloides shows the same increase in size over G. sternbergi as G. sternbergt does over Poébrothertum labiatum. Of the bones of the anterior limb, no differences are observable between them and the corresponding Fig 17, Humer- parts of G. sternbergi, except in the matter -of size us of Gomphothe- - rium PASSE De already noted above. 1898. | Wortman, Extinct Camelide of North America. 11g The exact stratigraphical position of this species is several hun- dred feet above that of G. sternbergz, and there can be no doubt whatever that G. cameloides is not only the direct lineal descend- ant of the older species, but is, at the same time, the progeni- tor of the succeeding Loup Fork species. This conclusion is somewhat at variance with the view expressed by Scott,’ in which he says: “ The Camels of the John Day formation do not present any important modifi- cations of the dentition; in some of them the first upper premolar has but a single fang, and others are decidedly re- duced in size; the former Cope fo has erected into a separate genus, Gomphotherium. It seems probable that these forms are not in the direct line of the ’ cameline descent.’ A careful examination of Cope’s type of Gomphotherium sternbergt, the only species, by the way, with the exception of the one above named, which has so far been described from these beds, re- weals’ *the tact “that . <>< BATH VOPSIDIANE es, let oeychste oe ailleueenees epaae a Waeeedereet lke < Haploconuslineatus 3) eee ee foie uf x MAYS COLMICUIA ISH ee eee eee an all Sx Amisonchus Sectonusse sas se cceie leer Aiea aes ee INET. coos coo So 55\|s000¢ x Hemithleus kowalevskianus .....|..... x INCERTA SEDIS. Eictoconodonta casei aera oe Aia.< SHUTGOMOTSYN Goo9 sadaescasoont ex Rrotolambaarneeemereechin ene OS 1 Matthew, Bull. Am. Mus. Nat. Hist,, 1897, p. 297. 2 Op. cit., p. 265. 1898. | Osborn, Evolution of the Amblypoda. Part 1. 183 Family PANTOLAMBDID® Cofe. Genus Pantolambda Cope. Dentition typical. First upper premolar one-rooted. Second, third and fourth three-rooted, with internal cones. Canines rounded. P. bathmodon. | TV GQULELGLUS. | Type: Mandibular ramus, No. 3956. | Type: Jaw, No. 3961. Smaller size. Both dental series con-| Larger size. First lower premolar tinuous. | close to canine and separated from | second by a wide diastema. Pre- molars reduced in size. P. cavirictus' Cope. The type lower jaw, described and figured by Cope (Am. Nat., Vol. XVIII, p. 1111) is peculiar in the close apposition of the first lower premolar to the canine, and the wide diastema behind it. This is the largest type known and the diastema is probably prophetic of the diastema invariably observed in Corvphodon, (Fig. 2). The skull (Fig. 7) was mistakenly described by Osborn and Earle in 1895 (1895, p. 43) as P. bathmodon. It differs in its much greater size from P. bathmodon, and in the absence of dias- temata from the P. caviricfus jaw. Unlike Corvphodon the upper canines are mainly worn upon the inner posterior surface. Fig. 9. Skeleton of Pantolambda bathmodon. Scapula wholly, pelvis partly restored. Composition from several individuals. One-eighth natural size. 1 Am, Nat., Vol. XVII, 1883, p. 968. 2 Am, Nat., Vol. XVI, 1882, p, 418. 184 Bulletin American Museum of Natural History. [Vol. X P. bathmodon' Cope. The composition skeleton of P. bathmodon measures 2 feet 9 inches (830 mm.) from the premaxillaries to the back of the ischiac symphysis, and 1 foot 134 inches at the withers. It is thus about the size and proportions of a large Wolverene (Gu/o luscus). Excepting in the selenodont teeth, it typifies the Aypothetical Protungulate, being more primitive than either Awprofogonia or Phenacodus. The step is that of the Bear, the feet very broad and spread- ing, the wrist and ankle being slightly raised off the ground, and the phal- anges terminating in hoofs. The vertebrez preserved (Nos. 2549, anca )indicate a short, neck (C,6— tomm.) as in Periptychus, and a back increasing in strength and power as we pass towards the lumbar region, Thus the dorsals are short anteriorly (D. 5 =15 mm, No, 2549) and indicate less separation of the zygapophysial and rib-tubercle facets than in P/ena- codus. The lumbars (L. 4= 25 mm., ene, pie No. 2549) are longer; unlike most Astragalus, metacarpal III and phal- Creodonts they present horizontal anges. Am. Mus., Cope Coll., No. : a 3957: rather than vertical zygapophysial facets. The tail is long and powerful. PRIMITIVE OR PROTUNGULATE CHARACTERS.—Among the persistent primitive or Creodont characters of Pantolambda are the following : Brain small, olfactory lobes large, hemispheres smooth. Skull with a sagit- tal crest ; terminal anterior nares ; nasals very Jong and expanding posteriorly; mastoid (periotic) widely exposed and forming lower posterior border of exter- nal auditory meatus ; tympanic bones rudimentary ; zygomatic arches slender ; no alisphenoid canal; basi-cranial foramina separate. Dentition typical; no diastemata; molars tritubercular, incisors small, cylindrical; canines rounded. Girdles: scapula unknown ; ilium acuminate as in Phenacodus. 1 See Cope, Am. Nat., Vol. XVII, p. 406. 1898. | Osborn, Evolution of the Amblypoda, Part J. 185 Fore-limb strongly bent outwards at el- bow (as in Creodonta and Carnivora), ma- nus everted. Humer- us with powerful del- toid, pronator (ente- picondylar) and supi- nator (ectepicondylar) crests; ulna with a convex posterior bor- der; carpus with an os centrale, an ex- tremely small mag- num and short trape- zoid, causing the me- tacarpal IV to be in- serted proximally be- tween the trapezoid and magnum (Fig. 2). Hind-limb straight, with three trochanters upon the femur (Fig. 11). Tibia with a rudimentary spine, a very long cnemial crest (Fig. 11) and femoral facets approximate. Tibia (Fig. 11) articu- lating with calcaneum. Probably an os-¢zbzace (Fig. 12). Mesocu- neiform short (analo- gous to trapezoid in the carpus), so that metatarsal IV articu- lates between ento— and ectocuneiforms (analogous to meta- carpal IV). Articu- lation between tibia and astragalus slant- ing obliquely inwards, very limited in extent, bounded. posteriorly - i Ni 2 Fig. 11. Pantolambda bathmodon. Anterior view of fore and hind limbs. showing powerful development of crests and tro- chanters as in the Creodonta. Humerus, No 2549. Femur, No. 2523 (2551, 2549). Tibia, No. 2551. Ulna, Was 2550, 2547. Ra- dius, No. 2547 (2546). Am. Mus. Coll. 186 Bulletin American Museum of Natural History. [Vol. X, by astragalar foramen which issues posteriorly between the ectal and sustentacu- lar facets (Fig. 12) (as in Creodonta). Astragalo-navicular head broad convex ; two astragalo-calcaneal facets only. Fifth metatarsal curved with a prominent external tuberosity for the peroneus brevis muscle, as in the Bear. All these osteological characters are shared by Periptychus, so far as the skeleton of the latter genus is known; (the carpus of Periptychus is unknown). The skull, vertebra and proportions of the limb bones in the two types are remarkably similar. ORDINAL OR AMBLYPOD CHARACTERS. — Primitive ordinal characters. It will be observed in the definition of the Ambly- poda, p. 180, that many of these above-described primitive charac- ters persist throughout the evolution of the order, and therefore rank as ordinal. MEASUREMENTS: SKULL, TEETH, SKELETON. P. BATHMODON. P. CAVIRICTUS. Mus. Nos. | Mm. |} Mus. Nos. Mm. = se | Skull, condyles to premaxillaries . . est. 2548 157 963 | est.272 Molar-premolar series, superior....... Wereecscre O55 ry 088 ime Ore en || 2550 060 3961 118 Lower jaw, symphysis to angle..... 5 2550 126 er est. 251 Fore-limb: Humerus, total length...)) 2549 124 oe Radius, ¥ ovolil Oxy 083 a Manus, oo ers = o81 Hind-limb : Femur, Me ete 2551 148 is Tibia, - es oy 11g sey Pes, oe i 10g 3963 184 Ilium, total length, as restored........ 7 IIl Sole “WALCithe sera eee scttectens tt Bee ell renee ee E 031 RESTORED SKELETON. Skull to perpendicular of ischiac SYMP MYSIS 5 save esata eve legekotet Mon svcenel sense etses ter eeeeqe as 830 Heioht -atjshoulder bladegen 50s serrl|l(eerei: OH Coy ails: oe yrekr anes 9 | ‘adAjog ‘sugozuvygaja uopoyftuo) | ZI Tne |} (fA ||P 2 SUS /Z Liev & | ‘add T° -sys03 (uopoydoyejapy) “9 | be oL|:: oe decay Heron cl areacho 5 mires) 4 a aio 4 cae Smane gStr Fidel gtogia Rie eon ar 4 | | €9z Sop|::: oto eee ee ee ee Sieerin, ee gf£er é OIC OOO ORC NCRDACON e024 Ss C6\: ee Lo3izlLs SS aivofele oo © oe eienelin eth) efieiiiic ie; ane) |\'n sie el sells alleMe ha |lat sieht | sh me coor a ‘adA 7° > * * sndAyoed x cz | | t¢é cr = 29.0 oitof ort Aare ics CEor ) SPN I a - + | | gh | Les Cor 998% CM he 0S AS vas see 9 ie | cs Ov leccclee: elgz e Acie cachet oe lola ae a GS | 6¢ os al (RS Colt Ba ees ‘xpun uopoyglso7) | t% | Of || OI | ° (oll pl AO ba eG d | adj, «°° °° *snzugo7 wopoydi.1oD | Li | XS || Crs Po oe ers Br Ce i | addy, ‘snyeursiew uopoyddiog | £z b6\Cg}" * *|Egrj019 lies Lvz avaailerere see eee Cnn ee Dee COM hoy Cec ee se A103] Suen “6 A | 9z 6¢ ee oe ZQI eee ggge 4 Nive (espana: * ie | gz ov ers FAleae eee Zoov HD | SERRE CTE a As OLAS CESS OO NP clo carol ZS oor eee Lz Cr Pr Ee do oS ese) 6zEP fol |} ptio 2100 Die coi i zSa} €€ entiiehe | /eeAgke Pill Rohe PACS nao ie. ‘7 gS|Egz\Sob|- =< *|*> >: eet euial|fail@sei|leticieus - ‘ wie fal tele eee On eee ee 698% é eile) ivikeote! ol/oikeiie) cere he a - Sao |SOLSp LE) is as OW? ETOAC COON TAS Words Weleriey(oxeyre €96z SSS eee eee s Ag piinieris||iie lee) s . -y eer ea Loge ) yimyESy oS 3 oH “ = | ISLigbe |ogt|Seba| bIS |*---| ab | - ve |z6r jrgr | 6292 OO TDS 2 SISo7 an | ia || Oi? || tye |) 3k ype PoP oe 60€1t 9 | ‘ad& °° *-snpuedar uopoyddson | 12 ob | GE t= = "ligene) eave i | ‘adA&y,°* + ‘snqyouro (uopoorjoq) “> | 61 FAS AIS oat |e aa ||ooclo.dos] 2otoclson Nooo fea | Age loo osloa mall § @Yofsi7 & | adh ‘supipn.s (uopomyjog) j | V | | fn e S a | Buoy ‘suey 2 @ oy "VIDA alge ‘suei] “Sul sued} “Suo]| ‘yup ‘dng | roquinyy p 2g 3 =)" 8 ‘SIA[ad = a 5 ‘deos NAS ‘2 ) hea | | (eta || eka. |/010.2 ; gift 9 | ‘ad 7 ‘sxzneusn (uopoyfopmpapy )*D | | LGe | Scie len S oie ol6z SA (MOE te * SAH RONG) 2 lRAD S| OLS Toy a hadi LL6z ‘TINYS* °° t2eewe7-com uopoy ~l4107 | ofa |], 4S POR Io OS Ios le Gy odAy SS isrepnr] BS 61 of oe | > Cz1a ong £9Sz HS roves . OO tno Cl a er | 1S) reo) Of || Om |KSm lopral) Yaa: Pal ail yest hes snuils * Je wee | . . leewe CeCeOn clon — OD Cicr ali yy 6 2° . is aS HRC We yAG EO Joes Jee erleeeeleee . ZIEV ‘TOIS Otc O10 Se it |" |PEz SveNelivelie Ce fot} oe Ooo SED eeieds . LICt OSes soda] on 0g 0619 60€ | ewe . seerles ics . ISP mio ni olkx)/6] « \ateiNs) aah ble: wlYavailfal|e) 0) (Wiss) eeu 6 | | Oe | ls, [Pe Bole omeltiye aE. [oo 2 1zeh Dal SANE [i ekg Sn ee a A | | Sz | o€ | LE | ve |SSr |PE19) oLz Sy I enD[ipeeees re < | €z CAM Oro) Gili Cupil hoe coe hnece ed ‘SNIN JEN é ‘adA J,°: * > ‘snnbijqo ane | Cz gto . ne oa sae vesv é J sesaw wean * $y (sna weN| ¢ | ‘add p++ ‘snzeprdsno uopoydds05 | | aes Ged ates 7 : | OVALE? OPA POC MSI ols 7 Pooky tec Y | ‘y[nyS * sadozunydajza uopoygli07 | eS See Fe pace Blom | | a “qian | P8° | coer -Suoj| ‘sues “Suoy| ‘yay ‘dng | - Si BL Be dee oa) te te gh ars eae |e €1 moO gl VI PInuyuwoy) —_SLNANAAOASVAIN AAILVAVdUNOZ) AO ATAV [ 200 Bulletin American Museum of Natural History. |Vol. X, characteristic last lower molar, m2 in the cotype, is wanting. The species must therefore rest upon the characters of the cotype. No lower teeth were found with the fine palate (No. 275, Am. ai ROM a TD Se Fig. 20. Coryphodon elephantopus, Cotype. Lateral, superior and sectional views of cranium. No. 11, U.S. Nat. Mus. Coll. Mus.) rightly associated with the cotype by Earle. The cranial and dental characters of the cotype (Fig. 20) are those of C. testis and C. dobatus upon a smaller scale ; m® has a well-marked postero-external elbow (Fig. 15), and the paracone is sharply dis- 1898. | Osborn, Evolution of the Ambiypoda. Part 1. 201 tinct from the metacone ; the ectoloph thus does not form a con- tinuous crest as in the supposed type. ‘The parietal protuber- ances or rudimentary horns Pa./7, are less pronounced but equally rugose ; the premaxillary symphysial borders are extensive although without contact ; the incisors are equal in size, 11 and 13 being fully as large as i2 in both jaws. This is therefore a smaller and perhaps more primitive type than either C. /odatus or C. ¢estis, although skull No. 275 is definitely recorded by Wortman from the Buffalo Basin, the highest true Wasatch level. Unfortunately the characteristic last lower molar is missing in the cotype ; the series pm 1-m 2 measure 122 mm. ‘The lower jaws of C. elephantopus are also represented either by Cope’s C. obliquus or by his C. latidens (see below). ‘The former is more probably the case for the following reason. According to the ratio of upper and lower teeth established in the C. /estzs jaws (see Table, p. 199, No. 3829), the lower grinders in C. elephantopus should measure 167 mm. The type lower molar of C. od/iguus ap- proximately agrees with this size (see Table, p. 199) and character. The last lower molar of C. obdiguus agrees closely with that of the supposed type of C. elephantopus. We may therefore consider the greater or less development of the entoconid 2, which these molars present, as variations similar to those which we have ob- served in the other species of this series, namely, C. festis and C. lobatus. No complete jaw is nearer this size than No. 4321 (Am. Mus., Cope Coll.), in which the lower grinders measure 172 mm. ; this specimen is also significant because the last inferior molar on the right side agrees in form with C. cuspidatus (¢.e., entoconid 2, dis- tinct), while the same tooth on the left side agrees with C. od/iqguus (z.e., entoconid 2, obsolete). Another proof of the variability of these cusps. This jaw, however, may belong to a small female of C. testis. INCERT# SEDIS. 21. Coryphodon repandus Co/e. Type, No. 4309, Am. Mus., Cope Coll. Superior and inferior molars m*, m®, Ms, Mz. Symphysis of lower jaws. Size=C. éestis, male. Loc., Big Horn, Wyoming. 202 Bulletin American Museum of Natural History. |Vol. X, This is an indeterminate type. It is distinguished by angula- tion of ectoloph in m# (as in C. elephantopus, cotype) ; perhaps also by the more transverse direction of hypolophid in mg ; second incisors only slightly larger than first and third (as in C. elephantopus, cotype). The nearest resemblance is therefore to C. elephantopus, from which it is distinguished by larger size. Superior molars No. 4366, from New Mexico, furnish a transition in the angular form of the ectoloph of m# to the C-. ¢estis type. Altogether C. repandus is of very doubtful validity. 19. C. cinctus Cope. C. (£ctacodon) cinctus. Type: No. 4341, Am, Mus., Cope Coll. Superior molars complete. A strong cusp appearing at postero-external angle of m®*. Loc., Big Horn, Wyoming. The distinctive feature of this type, viz., the quadrate form and postero-external basal cusp of m® (Fig. 15), and to aless extent on m2, are either individual variations or valid specific characters. They are certainly not generic. Lower teeth which may possibly be correlated (Nos. 4329, 4334, 266) have a triangular heel upon the last lower molar (Fig. 16), with entoconid very distinct and extremely short and oblique hypolophid. 24. C. testis Cope. 24. C. (Metalophodon) testis Cope. Type: No. 4317, Am. Mus., Cope Coll. Superior molar series. Originally distinguished by reduction of posterior cres- cent spur in m®. Definition.—Sup. molars=169¢to 1826. Inf. molars=172¢to 192¢. Third superior molar typically oval, with oblique posterior crest with primitive paracone, mesostyle and meta-crescent more or less distinct. Third inferior molar with oblique hypolophid, entoconid 2 reduced or vestigial. Second incisor the largest. This includes the most completely known Coryphodon. It has been heretofore described by Earle and the writer as C. radians, but is now found to be distinct. ‘The identification with Cope’s type of C. desé’s is made by means of a careful comparison with the superior molars in the female skull No. 2963. ‘The form and measurements are identical. As this skull undoubtedly belongs 1898. | Osborn, Evolution of the Amblypoda. Part J. 203 to the same species as the male skull (Fig. 21) and skeleton, all the characters of this fine type are now available. This is the largest Coryphodon but one, and is very abundant in the Middle Wasatch levels, being represented by a magnificent series of skulls and skeletons in our collection. From these the sexual characters are clearly made out. The large male skull is used in the complete mounted skeleton, Fig. 18 4. The smaller Fig. 21. Coryfhodon testts. Large male, showing rudimentatary parietal horns. Upper canines partly restored. Skull No. 2867, lower jaw, No. 2872. Am. Mus. Coll. female type of this species is represented in the skull No. 2963, and jaws (Nos. 2868, 259) in contrast with the powerful male skulls (Nos. 2829, 2867) and jaws (4322). Variations in the last lower molar are considerable, from an oblique to a bilobed (No. 259) or less oblique condition of the posterior crest, with all the stages in reduction of the entoconid 2. Exactly similar variations are found in the lower molars of the larger and smaller members of Series Il. The development of entoconid 2 also varies in the posterior molars upon opposite sides of the same jaws of several specimens of C. lobatus. The osteological characters have been fully described and fig- ured by the writer (this Bulletin, 1898, pp. 81-91). Full charac- ters of the vertebral column are shown in Fig. 23. Certain specimens (skull, No. 2866) of the still larger C. Zobatus have been found below it, and the much smaller C. edephantopus 204 Bulletin American Museum of Natural History. [Vol. X, occurs in the higher levels of Buffalo Basin, Wyoming. Our scanty evidence therefore appears to indicate a retrogression in size in this series, but this is an inference by no means certainly established. Pao? “apai th pe m.1 ma Fig. 22. Coryphodon testis. Superior molar series, male (Am. Mus Coll. No 274); inferior molar series, female (Am. Mus. Coll. No. 2868). 17. C. lobatus Cope. Type, Nat. Mus. Coll. Sup. molar 3; inf. molar 3; part of sup. and inf. canine, indeterminate. Definition.—Sup. molars=193 4. Inf. molars=1964. Dental characters as in C. ¢estis, excepting elongation of protoloph and degeneration of posterior metacone crescent in m®. Astragalus usually lacking astragalar foramen. Cranium massive, with widened parietal horn rudiments. Synonyms. 20. Coryphodon anax. Type: No. 4327, Am. Mus., Cope Coll. Superior molar 3 ; inferior molars, premolars and incisors. Loc., Big Horn, Wyoming. 25. C. (Bathmodon) pachypus Cope. Type: No. 4335, Am. Mus., Cope Coll. Astragalus, calcaneum, pelvis, femur, &c. Indeterminate type. Loc., Big Horn, Wyoming. This is the largest Coryphodon known ; it surpasses C. fest¢/s in size, the ratio being 50: 45, as indicated by the femora. Unfortunately the name C. /obatus is prior to the more appro- priate C. anax,and must supersede it. The lower molars defined 1808. | Osborn, Evolution of the Ambl vpoda. Part J. 205 ye} TepNoseqny *€z *S1q “SqIit lof sjo wood Jaqunu C sTenptaAtpul [eteAas jo uortsod “uoya]axs pejanouw jo uurnyoo [erAqgei19 A ‘JENPIAIPUL O[BUIs B Wo, pSUTUIaJap S ‘azis [BIN}eU YWYSIO-auC) Gtsoz ds t/'s $ yeuvo petayieaqayiea ‘2 U/-g2 \ uawie1oy snout *s248aq UopoYy gg + syooey ren} ed z= 35) ~@ /> of its skull and molar teeth. as C. lobatus Cope, which may be considered as indeterminate types, present exactly the C. avax measure- ments (see Table, p. 198), and exhibit an oblique hypolophid and depressed entoconid 2 upon ms, as shown in Fig. 16. ‘This may be considered the typical C. Jobatus or C. anax third lower molar. It is well shown in the large jaw, No. 4333, and in the frag- ment, No. 4305. A variation, No. 266,in molars of the C. Jobatus size is paralleled by a variation, No. 4239, in molars of the C. cinctus size, as represented in Fig. 16, in which the three cusps form a posterior triangle, as also in C. cuspr- datus (No. 4324). If these are not variations they represent three dis- tinct species, which is possible but not probable. Synonym.—The skeleton defined as C. pachypus by Cope (No. 4335) un- doubtedly belongs here. The astra- galus of C. pachypus and of another specimen (No. 2870) exhibit no astra- galar foramina, and show a wider in- terval on the front face of the astra- galus between the tibial and navicular facets than we find in C. éestis. The massive male skull in our col- lection (No. 2866) is most interesting in its progressive development of the parietal horn thickenings, parallel with those of Uvdntatheritum. Its geological level, however, according to Wortman, is below that of C. éeszzs, an observation very difficult to recon- cile with the more advanced evolution 206 Bulletin American Museum ae Natural History. er x 14. C. cuspidatus Cope. Type, Nat. Mus. Coll. Fragmentary inferior molars 2 and 3, and a portion of the jaw. Originally distinguished by prominent entoconid 2. Cotype, No. 276, Am. Mus. Coll. Complete lower dentition ; upper pm*—m®. This specimen was referred to C. obfiguus by Earle (1892, p. 162). Definition.—Inferior m.pm.=154. Superior m.pm.=150. Last superior molar oval, antero-posteriorly compressed. Last inferior molar with oblique hypolophid and more or less prominent entoconid 2. Metaconid with rudi- mentary metastylid. Inferior incisors unequal in size. This is a diminutive Coryphodon, of the size and very similar in molar type both to C. eocenus Owen, from the London Clay, and to C. owenit Hébert, from the Suessonian of France. The characters of Cope’s type are very indefinite ; the complete upper and) lower series of teeth, No 270. therefore serve as a cotype to define this species, the most diminutive of the series. ‘lhe last lower molar of the right side agrees in form and measurement with Cope’s type, al- though the entoconid 2 is less promin- ent and isolated ; on the left side the entoconid 2 is nearly obsolete, again eRe att ee cr os demonstrating the variability of this Coryphodon lobatus. Am. Mus. cusp. The very small lower canines Cope Coll., No. 4335. fr a ‘ ( % indicate that the animal is a female. The enlarged second incisors and general form of m3 confirm its reference to Series II. A unique feature is the reduplication of the metaconid in m2 and m3 into a rudimentary metastylid, parallel with the large metastylid of Uintatherium. Another example of this species is No. 4324. 13. C. latidens Cope. Type, Nat. Mus. Coll, 4. Lower jaws and teeth, left premaxillary and in- cisors, superior canine. Loc., New Mexico. Definition.—Inferior m. and pm.=156. Inferior molars short and broad with crests nearly or directly transverse (angle=e 85°); entoconid 2 vestigial or wanting. Inferior incisors equal sized. Superior canines nearly straight, antero-posteriorly compressed, subtriangular, with an external ridge. 1898. | Osborn, Evolution of the Amblypoda. Part J. 207 This imperfectly-known animal appears to represent a rather small and specialized form in Series II. It is distinguished from the type of C. elephantopus by the straighter and more compressed superior canine observed in the type, by the transverse position of the crests of the inferior molars, and by the absence of entoconid 2. As shown in the Table, p. 199, the measurements of Cope’s C. /atidens type are identically the same as those of C. seus, although Cope speaks of the latter as being much smaller than the former. Cope has suggested the possible association of C. Zatédens with the cotype skull of C. elephantopus. It appears to be distinguish- ed, however, by the form and compressed section of the superior canine. It is, however, certainly related to Series II by the sub- triangular form of the canine and the characteristic swelling of the jaw below m3. Unfortunately the types have been tempo- rarily misplaced, and no determination of this question by direct comparison can be made at present. If these jaws should prove to belong to C. elephantopus, the species C. obliguus will have to be revived. It will be noted that both types come from New Mexico. In New Mexico, also true Wasatch, we found in 1897 a lower jaw (No. 2563, Fig. 16) of extremely small size, associated with Mentscothertum, Ambloctonus and Didymictis, which may represent a female of this species. Unfortunately the canines are not pre- served. The total lower grinding series does not exceed 125 mm., so that this is the smallest Coryphodon jaw known; the last lower molar measures only 30x 19 mm.; the posterior crest forms an angle of 85° with the long angle of the jaw; a minute vestige of the entoconid 2 can however be observed. 22. Coryphodon curvicristis Cope. Type, No. 4326, Am. Mus., Cope Coll. Lower jaw fragments containing pm.4 to m3; canine. Definition.—Molar crests transverse. Posterior crest of my directly trans- verse, crenulate, depressed. Superior incisors with sharply angulate anterior faces. Canines as in C. Zestis. The systematic position of this species (Fig. 16) is indetermi- nate. It resembles C. /atdens in the transverse crest angulation 208 Bulletin American Museum of Natural History. [Nol. X, and in the rather broad proportions of the molars, but exceeds this species in size. The complete superior canine determines the position of the animal in Series II]. The canine is powerful, curved and antero-posteriorly compressed, partly as the result of pressure. The inferior premolars are exceptionally short. A fourth member of this series, C. vevtanus, is found in the Wind River Beds. It appears to resemble C. /a/idens in the form of the superior canines. Sertes I[TT.—SMALLER CORYPHODONS. SPECIALIZED. RELA- TIVELY NARROW, FLAT-TOPPED SKULLS, (?) WITHOUT PaARIETAL HorN RUDIMENTS. CANINES COMPRESSED LATERALLY AND GROOVED ANTERIORLY. LOWER MOLARS ELONGATE, CRESTS NEARLY OR QUITE TRANSVERSE; M3 BILOBATE, NO ENTO- CONID 2. In 1872 Cope defined certain teeth as MZetalophodon armatus, mistaking the posterior superior molars, m*, of two individuals for m® and m2 of one individual, as can be proved by a compari- son with his type of C. molestus. ‘The latter type moreover gives us the cranial characters and constitutes a valuable cotype. 6. C. (Metalophodon) armatus. Type, No. 4315, Am. Mus., Cope Coll. Superior m®, m2, premolars, supe- rior canine and incisors ; two individuals, probably mingled, fully adult. Cotype, No. 4316, Am. Mus., Cope Coll. Superior m? and m%, mg, pre- molars, etc., juvenile. Definition.—Upper and lower canines greatly compressed, with a deep antero-internal groove upon the upper canine. M® with powerful anterior crest, my without entoconid 2. Lower molars elongate, crests lunate, nearly trans- verse. (Angle with long axis of jaw, 81°.) Synonyms. g. C. simus Cope. Type, U.S. Nat. Mus. Coll. Inferior m and pm=154. Fragmentary skull, probably female. Superior canine. Mandibular rami and teeth. Loc., New Mexico. 1898. | Osborn, Evolution of the Amblypoda. Part TJ. 209 10. C. molestus Cope. Type, U.S. Nat. Mus. Coll. Skull, dentition and parts of skeleton. Loc., New Mexico. 11. C. domas Cope. Type, U.S. Nat. Mus. Coll. Posterior inferior molar. Loc., New Mexico. The slender crests and the elongate form of the posterior lower molars in this species at once distinguish it as a type from mem- bers of the foregoing series, since they form an angle of 81°, or nearly a right angle, with the long axis of the jaw, and m3 is entirely devoid of the entoconid (Fig. 16). Specimen No. 4315, Fig. 17, gives the most distinctive character, shown again in Cope’s type of C. molestus (Cope, 1877, Pl. LVI, fig. 4), which agrees with C. armatus, namely, the flattened form of the canines. Cope himself referred C. lomas to C. molestus (1877, p. 237). * F : . Fig. 25. Coryphodon armatus supe- The type of C. semus has lower _ rior moiars, left Side (type of C. molestus teeth of the same character, rather i long and narrow. The upper m./ m.2 m3 canines are, however, described by Cope as triangular and grooved ; this raises a doubt as to the reference of this type to C. armatus. The juvenile type specimen of C. molestus demonstrates the flat-topped character of the skull (Coll. U.S. Nat. Mus. No. 1119, Cope, op. ci¢., Pl. LVI); the skull is far less expanded laterally, when seen from above, than any of the skulls in Series IT; but this may be in part due to its juvenile and undeveloped condition. “The inferior canine,” observes Cope, “has a flat interior and convex exterior face, which are separated by anteriorly and pos- teriorly directed cutting edges.’”’ The most distinctive feature of the canines therefore is that the antero-posterior diameter greatly exceeds the transverse, as in Uvntatherium. An aberrant feature is the antero-external groove. ‘he median incisors are as large or larger than the others. [Way, 1898] 14 IsERTA SEDIS. 23. C. marginatus Cope. Type, No. 4374, Am. Mus., Cope Coll. Superior molar 3, canine and pm. Loc., Big Horn, Wyoming. This indeterminate type resembles C. avmatus in the form of m#, but differs from it in the form of the canine, which 1s less compressed and may possibly represent a milk tooth. The canine corresponds with Cope’s description of that of C. semus. WIND RIVER TYPES. Cope’s Wind River material of Coryphodon, all of which is now in the American Museum (Nos. 4811, fragments of skull and teeth ; 4812, lower molar, incisors and fragments; 4813, lower jaw and fragments ; 4814-4817, fragmentary teeth ; 4818), merely sufficed to determine the existence of this genus in these beds. Our Wind River collection and the determination of manus No. 4351 (Am. Mus., Cope Coll.) as belonging to the Wind River Beds, is therefore of very great importance. It demonstrates that Coryphodonts of considerable diversity and size persisted into the Wind River period. Owing to the general scarcity of fossil remains in these beds, the relative abundance of these animals cannot be estimated. Of intermediate size is the jaw of No. 2976, described below as C. ventanus ; of smaller size there is a well-preserved skull (No. 2977), type of the new species C. wortmani. ‘They represent respec- tively the persistence of at least two series, namely of Series I, and of Series IIl now discovered for the first time. SUCCESSORS OF SERIES II. 26. Coryphodon ventanus, sp. nov. Type, No. 2976, Am. Mus. Coll. Jaws and lower teeth. Superior incisors and canine. L. metacarpal IV. Definition.—Size of C. testis. Inferior mand pm series=e172. Supe- rior canines posteriorly compressed, with antero-internal depression and long 1898. | Osborn, Evolution of the Amblypoda. Part J. 211 sharp external ridge. Lower canines with short external ridge near apex. Second incisors enlarged; lateral incisors much reduced. Posterior inferior molars with crests more transverse than in C. festis (angle = 74°) a persistent entoconid 2. ( ? Cuneiform articulating with Mtc. V.) This species is clearly distinguished from C. ¢estis by the form of the canines, which in this animal are comparatively straight and lance-shaped (Fig. 17), the long axis transverse (unlike U¢ntatherium), with an antero-internal groove which is Fig. 26. Foot structure of Coryphodon. D, external view of manus of C. ventanus (No. 4351, Coll. Am. Mus.) ; A, superior view of astragalus and calcaneum found near C. simus, no tibial facet (Bathmodon type); 4, lower surface of astragalus, showing caleaneal and cuboidal facets ; C, external view of calcaneum and astragalus, showing reduction of tibio- calcaneal facet. (Coll U.S. Nat. Mus.) worn away by the lower canine. ‘They resemble those of C. /aé- dens Cope (except in the groove), but are much less compressed than those of C. avmatus Cope, besides having the long axis in a different plane. ‘The posterior crest of mz (Fig. 16) differs from those of C. armatus and C. s¢mus in form and in the retention of an entoconid 2, and from that of C. fests in being slightly less oblique. Another character is the very rapid increase in size of the molar series as we pass backward : mi=28, m3=42. To this species belong Nos. 2982, 4813, 4812, 2774, and 2978 of our collection. The latter contains the complete lower teeth which exhibit the marked disproportion between the second and the first and third incisors embodied in the definition of this spe- cies. The incisor proportions are indicated by the length of MOOS, 1140, 12—=59, 1 3— 28. 212 Bulletin American Museum of Natural History. |Vol. X, The metacarpal IV agrees in length (54 mm.), and lends some probability to an association with that of the complete carpus No. 4351 from the Wind River.’ This associated complete carpus (Fig. 26) agrees with some specimens of U7ntathertum in the very exceptional character that the cuneiform articulates with Meta- carpal V. Series I.—PRIMITIVE, NARROW-CRESTED SKULLS. CANINES ROUNDED. INCISORS SUBEQUAL IN SIZE. The lower teeth are unknown, and the ancestral members of this series have not thus far been determined in the underlying Wasatch formation. 27. Coryphodon wortmani, sp. nov. Type, No. 2977, Am. Mus. Coll. Loc., Wind River, Wyo. Definition.—Superior m and pm=154. Superior canines rounded. Occiput very high and narrow. Supratemporal ridges converging posteriorly to form a comparatively narrow sagittal crest. Fig. 27. Coryphodon wortmanit, type. Lateral view of skull and section of superior canine.? The discovery of this type (Figs. 18 and 27) in the high level of the Wind River Beds is most surprising. It is far more primitive both in its narrow cranium and rounded canines than any of the 1 This carpus was mistakenly described by Cope as coming from the Wasatch. Dr. Wort- man identifies it as found by himself in the Wind River. 2 Dedicated to my colleague Dr. J. L. Wortman. = te ee 1898. | Osborn, Evolution of the Amblypoda. Part J. 213 C. testis series found in the Middle Wasatch. It appears like a direct successor of Pantolambda cavirictus. ‘The median incisors are equal to the others, the lateral incisors being slightly the smallest. The canines have a rounded crown somewhat flattened in front by wear. The superior grinding series present a rudimentary posterior crescent on m2 and an oval m* with short oblique ectoloph. The most unique features are the form of the occiput and the cranium as defined above, which is intermediate between that of Pantolambda and Coryphodon armatus. The metatarsal V is short and robust (= 42 mm.), with the characteristic peroneus tuberosity of the true Coryphodon. Both femora are finely preserved (length, = 340 mm.), being of the smallest size known. INSERT#@ SEDIS. The position of the following types with reference to the Series I-III, which we have been considering, is uncertain. 4. Coryphodon radians Cope. Type, No. 4300, Am. Mus., Cope Coll. Superior molars 1, 2 and 3. Prob- ably associated lower jaw, No. ? 4300. Portions of skeleton. Loc., Evanston, Wyoming. Definition.—Third superior molar with a spur (metacrescent) upon the pos- terior crescent of the ectoloph. Third inferior molar without entoconid 2, hypolophid nearly transverse. Lower canines somewhat incisiform. This classic species, which rests upon somewhat uncertainly associated upper and lower teeth, jaws and skeleton, was the first described in America. The structure of the last upper molar is shown in Fig. 15. The last lower molar has crests nearly as transverse and simple as in C. Zatidens. ‘The most distinctive structure is the lower canine which, although badly broken, exhibits a distinct flare at the base of the inner face, as in the incisors, and is apparently becoming incisiform, an interesting approach to Uintatherium. 214 Bulletin American Museum of Natural History. {|Vol. X, 15. Coryphodon hamatus Warsh. Type, Yale Museum No. 1330. Skull and dentition much worn. Cotype, Yale Museum No. 1334. Female skull with perfect superior and in- ferior dentition. Loc., Evanston, Wyoming. Synonym. 18. Coryphodon(Manteodon) subquadratus Cope. Type, No. 4340, Am. Mus., Cope Coll. Superior molar 2, incisors and fragmentary premolar. Loc., Big Horn, Wyoming. Definition.—Size large. Superior molars with quadrate crowns and well developed hypocones upon mi and m2. Inferior molars with nearly trans- verse crests; mg without entoconid 2. This species was mistakenly associated with C. elephantopus by Earle. In size it equals C. ¢es¢’s, but it 1s well distinguished by the quadrate form of the superior molar teeth in which, according to the figures of Marsh (Dinocerata, Fig. 55, p. 52), a representative of the hypocone is present. This is developed from the ridge extending backwards from the protocone. In the inferior molar teeth the crests are nearly transverse, and there is no trace of the entoconid 2. The unique quadrate tooth with a prominent hypocone, type of Manteodon subquadratus (Fig. 15), was without reason considered by Cope as a third superior molar. It proves, upon comparison with Marsh’s cotype made by Dr. Matthew, to resemble a second superior molar of C. hamatus. It differs, however, from C. hama- tus in the more distinct development of the posterior spur of the metacone crescent, a character which may subsequently prove to give it distinct specific rank. The type skull of C. Aamatus is somewhat fractured. The top of the skull of the cotype, a female, is considerably narrower than that of C. fests, female, presenting a condition intermediate between that of C. ¢estis and C. armatus. ‘The canines in this animal, as in other females, are small. 28. Coryphodon singularis, sp. nov. Zype, A hind limb, tibia, fibula and pes No. 2980. Loc., Wind River, Wyo. A small and unique hind foot and limb from the Wind River Beds, found upon the level of C. wortmant, is of excep- 1898. | Osborn, Evolution of the Amblypoda. Part J. 215 tional interest (Fig. 28). Associated lower tooth fragments, put together by Dr. Matthew, resemble those of a small Coryphodon, and clearly separate this animal from Sathyopsts. The differ- ences from the pes of Coryphodon are very significant, as follows : 1. Navicular laterally reduced, excluded from cuboid by ectocuneiform, a unique condition. 2. Ectocuneiform enlarged, articulating with astragalus (unique). 3. Second or middle phalanges greatly abbreviated upon all digits, I-V, as in Uintatherium manus. 4. Front surface of astragalus widened, separating tibial and navicular facets asin Uintatherium. 5. Tibia long and slender, unlike Coryphodon. The measurements of the metatarsals are as follows: Nits el —> Oe VtS ss elf oe Mts ease Mts ViR—42s VES Vie==34 0 Other measurements in Table on page Igg. Fig. 28. Coryphodon singularis. Superior and lateral views of pes. Am. Mus. Coll. No. 2080. This animal thus shows one progressive character (4), two entirely unique and distinctive characters (1, 2); the latter, together with (5), sharply separate it from Coryphodon ; two char- acters, 3, 4, parallel or approach Usntatherium. The other Wind River species, C. wortmant and C. ventanus, are distinguished from this by their typical metapodials, one of which is known in each type. The associated femur (No. 2970) is proportioned like the tibia, long and slender. 216 Bulletin American Museum of Natural History. |Vol. X, Prophetic of this type, perhaps, is the pes of Pantolambda cavirictus (Fig. 12), in which the navicular is reduced upon the outer side and the ectocuneiform is elongated so as to nearly come in contact with the astragalus. Foot STRUCTURE. Cope (1884, 1, p. 1120) proposed the theoretical groups, P/aty- arthra (with flat astragalus) and Amblypoda hyodonta (astragalus without a neck) from which to derive the Amblypoda. Both groups are superfluous now that it is clear that the ancestral Amblypoda can be derived directly from the Creodonta, all of which possess an astragalar neck. Planes EER et oie es ‘ Re Sarees Caan eeey eee Planes of _ Astragalo- Navicular Us (BL Facets. Fig. 29. Angles formed by tibio-astragalar astragalo-navicular facets, to exhibit widening of front face of astragalus. Ur, Ursus; P. Pantolambda,; C, Coryphodon testis; Ut, Uinta- therium ; E, Elephas. The transition is simple. By shortening of the neck of the astra- galus (Fig. 29 P. and C. and W.) the tibio-astragalar facet is gradually brought almost into confluence anteriorly with the astra- galo-navicular facet, as in C. radians. In C. lobatus and C. sin- gularis this space widens as in Uzntatherium. 1. The variables in these feet are the astragalar foramen and the “biale facet. From our present knowledge both these struc- tures (inherited in Coryphodon from Pantolambda) are useless or vestigial, inconstantly developed and therefore not constant specific characters. In Fig. 26 (identical with Coryphodon III, Cope, 1877, Pl. 60), a small astragalus and calcaneum is shown which lacks both astra- galar foramen and tibiale facet. In C. lobatus (No. 4335, type of C. pachypus) there is a large tibiale facet, while the astragalar 1808. | Osborn, Evolution of the Amblypoda. Part J. 217 foramen is not even grooved. In No. 2870 the tibiale facet is irregular, and a groove represents the astragalar foramen. In C. testis, No. 258, the ibiale facet is irregular, the astragalar foramen is wanting; in No. 2869 it is completely bridged over; in No. 4300 (Cope’s cotype) it is partly bridged over. 2. In therelative constancy of the tibiale facet and of the astra- galar foramen or groove, the pes of Uintatherium mirabile is there- fore more primitive than that of Coryphodon. CONCLUSION OF PART I. The phylogenetic conclusions drawn from this analysis of the Taligrada and Pantodonta will be more fully discussed at the close of Part II of this paper, which will treat of the Dinocerata. The two main results thus far brought out are these: First, the demonstration of a number of separate phyletic lines of Coryph- odons; these lines probably represent the local differentiations of the Coryphodon type in adaptation to different feeding ranges, that is, swamp, plain, and upland. ‘The second result is, that certain Coryphodons approach the Dinocerata in some structures as Closely as they depart widely from them in others; for example, C.armatus resembles Uintatherium in canine type, but differs from it in skull type; C. ¢estis approaches Uintatherium in the upper posterior portion of the skull, but differs from it widely in the anterior portion of the skull, and in the structure of the canine teeth ; C. radians shows the assumption of the incisiform shape by the lower canines, so distinctive of Usntathertum. But no Coryphodon is fully known which fills all the conditions of an ancestor of Uzntathertum. Until the skull of Bathyopszs is known the transition between the above types will remain obscure. BIBLIOGRAPHY—PRINCIPAL REFERENCES. Corr, E. D. 1877.—Report upon the Extinct Vertebrata obtained in New Mexico by parties of the Expedition of 1874. U.S. Geol. Sury. west of tooth Mer. Palzontology, Vol. IV, Part ii. y 1884.—The Vertebrata of the Tertiary Formations of the West. U.S. Geol. Surv., Vol. III, Part i, 1884. oe 1884 (1).—The Amblypoda. American Naturalist, Nov. & Dec., 1884, pp. I110-1121 and pp. 1192-1202. Jan., 1885, pp. 40-55. Adie ts PI Dip ee a OTN A ASOT ath gh SLA Ba”, on te Mee rt . Acne ae ee 218 Bulletin American Museum of Natural History. |Vol. X.] EARLE, CHAS. 1892.—Revision of the Species of Coryphodon. Bull. Am. Mus. Nat. Hist., Vol. 1V, 1892, pp. 149-166. MarsH, O. C. 1876.—On some of the Characters of the genus Coryphodon, Owen. Am. Jour. Sci. (3), Vol. XI, pp. 425-428, 1 pl. 1877 (1).—Brain of Coryphodon. American Naturalist, Vol. a5 (De S775 (2).—Principal Characters of the Coryphodontidz. Amer. Jour. Sci. (3), Vol. XIV, pp. 81-85, pl. iv. (6).—Introduction and Succession of Vertebrate Life in Amer- ica. (Vice-President’s Address before the American Associa- tion for the Advancement of Science, Nashville (Tenn.) meeting, Ate. 30.1577.) roc. Assoc., Vol. XOXVileapp: 211-258, 1 pl. 1878. 1884 (g).—Dinocerata: A Monograph of an Extinct Order of Gigantic Mammals. With 56 plates and 200 woodcuts. Washington, 4to, XVIII, pp. 237. Author’s edition. 1893 (7).—Restoration of Coryphodon. Amer. Jour. Sci. (3), Vol. XLVI, pp. 321-326, plates v and vi. New Haven, Oct 1803: OsporNn, H. F. 1893 (4).—Rise of the Mammalia in North America. (Vice- Pres. Add. Amer. Assoc. Ady. Sci., Sec. Zool., Aug. 16, 1893.) Amer. Jour. Sci,, Nov. and Dec., 1893. (5).—Fossil Mammals of the Upper Cretaceous. Bull. Am. Mus. Nat. Hist., Vol. V, p, 15, Dec. 20, 1893. 1895 (1).—Fossil Mammals of the Puerco Beds, Collection of 1892. Bull. Am. Mus. Nat. Hist., Vol: VII, Art. 1; pp: 1-70, Mch. 8, 1895. (With Charles Earle.) 1898.—A complete Skeleton of Coryphodon radians. Notes upon the locomotion of the animal. Bull. Am. Mus. Nat. Hist., Vol. X, April 4, 1898, pp. 81-91. afte) | Ur - tional ( By ‘Hew1 y Pamerenp Ospony.. = y Article XII.—ADDITIONAL CHARACTERS OF THE GREAT HERBIVOROUS DINOSAUR CAMARASAU- RUS. By Henry FAIRFIELD OSBORN. WITH THIRTEEN FIGURES IN TEXT. This gigantic reptile was found in the famous Como Bluffs of Wyoming by Dr. J. L. Wortman, of the Museum party of 1897, and Prof. Wilbur C. Knight, of the University of Wyoming. The Museum number is R. 222.) The bones include the left ilium, the ischia and pubes of both sides, the right and left femora, the left tibia and astragalus, the right scapula and coracoid, two shattered cervicals, two complete dorsal vertebrae, two incomplete dorsals, three or four incomplete ribs, coalesced spines of three sacral vertebree and one sacral centrum, twenty caudals and twelve chevron bones. With the exceptions stated, the bones are in a~ remarkable state of preservation, having been worked out with exceptional skill by Mr. Granger and others, under the direction of Mr. Hermann. This is a large individual. The identification is provisional. The measurements, in comparison with those taken from the type of B. excelsus Marsh, are as follows: Brontosaurus American Museum excelsus. No. 222. WenothrotitemUts. arc cies ielcy sere ers 11eF= Sy iiés (6) ial, Slit 1) abot VB PREL DIAM Ve isie, «eke Mena nine ai te vans tes G3 titeg | (Oia Binley © teiyat pat take PRULDIS the avo afatticas. s oheltvers: sy oye) Seer Brit. Lovin: Bt. LOM, pomiear ISCIIUMce bs ceil cys ics sa 3} ie fey tal 3 ft. gl in uenamS CAP UIA areastfarsrert siscercis sells ais). tities 2) tole Sey Gabe peru iaee COLACOIC naan niaiteremmie teats) «(0h -Fe 2 ft. 216 in. 2 Ties! Feb ol The new points of greatest importance are : First—The discovery of the hitherto unknown characters of the anterior caudal vertebre. Second.—TYhe apparent resemblances of Camarasaurus Cope to Amphicaelias' Cope, to Brontosaurus Marsh, Atlantosaurus Marsh, and Afpatosaurus Marsh. 1This has been anticipated by Marsh (Am. Jour. Sci., Aug., 1881) in his first classification of the Dinosauria, in which Amphicwlias is bracketed after Camarasaurus, and placed in the Atlantosauride. In the final classification of the Dinosauria, however (* The Dinosaurs of North America,’ p. 241), Caszarasaurus and A mphicelias are removed to the Morosauridz. [219] 220 Bulletin American Museum of Natural History. |Vol. X, Third.—The observation of structural analogy to certain stru- thious birds in the anterior dorsals and posterior cervicals. 1. HABITS AND SIZE OF THE ANIMAL IN RELATION TO ITS STRUCTURE. The estimate given by Marsh of the total length of this animal is nearly or quite 60 feet; the tail is figured at about 24 feet. Since the vertebra believed by Marsh to be the third caudal is probably the roth or rrth, the tail should be increased to over 30 feet in length, by the addition of at least seven large anterior caudals. The total number of caudals is estimated at 4o as against 37 in Diplodocus. Marsh has attributed to Brontosaurus 27 precaudal vertebrze, or 13 cervicals and 14 dorso-lumbars. From reasons given below it is probable that there was a larger number of dorso-lumbars, which would still further increase the length of the animal to considerably over 60 feet. We can only conceive of the Camarasaur as a great wading and swimming quadruped, enjoying a habitat similar to that of the Upper St. John River, Florida, at the present time, namely, a relatively firm bottom gently graded to all depths, supporting a richly luxuriant aquatic vegetation, the river banks bordered by sloping shallows of sand (Colorado, Cafion City Beds) or clays (Wyoming, Como Beds). As imagined by Cope in his picture of Amphicelias (‘Century Magazine,’ November, 1887), the animal could walk along the bottom, raising the anterior portion of its body. We believe also that it could swim rapidly, propelled by its light but long and powerful tail, which would be useless upon land. The abundance of cartilage around all the limb joints and the non-osseous nature of many of the carpals and tarsals afford positive evidence that the limbs were not con- tinuously subjected to the hard impact of the enormous weight of the body by motion on land. Feeding was done in the water and along the shores. Excursions upon shore were there- fore like those of the Alligator, mainly for breeding and egg- laying purposes, and they exposed the animal to attack by the Megalosaurs. By means of powerful mid- and posterior-dorsal spines and opisthoccelous vertebre, the entire anterior part of the 1898.| Osborn, Additional Characters of Camarasaurus. 221 body, whzle in the water, could be raised or lowered with the great acetabulum acting as a fulcrum, thus presenting an analogy to the Hadrosaurs, which exerted a similar movement fon land. The long neck, similar in structure and almost as flexible as that of an Emeu (Dromeus), could thus pass through a prodigious are in the search for food either under or above water. The neck motion apparently involved the anterior non-spine-bearing dorsals Fig 1. Droma@us. Cervicals 13 and 14; dorsals1 and 2, showing absence of median spines. Dorsal 3, showing large blunt median spine, resembling that of the Camarasaurus dorsal, Fig. rr of this Bulletin, Am. Mus. Coll., No. 607. as in Dromeus, behind which the comparatively inflexible large spine-bearing dorsals rose to maximum height in the sacrum for the insertion of the ligamentum nuche and elevator muscles. The importance of such an hypothesis of function will appear in the following description and discussion, and it applies to all the Cetiosauria, namely, to the A7orosaurus and Diplodocus types as well, which so far as known are wniform with the Camarasaur type in the peculiar bird-like structure of the posterior cervicals and anterior dorsals and in the possession of a very powerful swim- ming tail. Ten of the caudals in our specimen afford an interesting illus- tration of the cause of the distribution of these large skeletons 222 Bulletin American Museum of Natural History. |Vol. X, over a considerable surface. The dorsal spines, and in some cases the sides of the centra are found to be deeply gashed with the sharp teeth of a carnivorous Dinosaur. The upper portion of several spines, in fact, is entirely bitten away, the upper surface containing jagged transverse grooves, which prove that the carnivore was of great power, and applied its pointed teeth with strong effect to the gristle and muscles upon the sides of this prodigious tail. TasLE J.—MEASUREMENTS OF VERTEBRE. ] | = ; a a bus Lad 7 o ates 2 “e “2 | ae | oe a 8 oe S) (S) S) QO |o |e H ea} | | | Anterior dorsal. i || BICC) 310 | 1050 470 | e864 Posterior dorsal.; ? | 368 BAOm |e SOmaneecr | e715 First caudal....) 162 383 BS 5 ll ALO Sih tear ste etersncu ees 957 Geel CanCkiIljsseao0\) WOE Noodoor CEH lls saogo|lbau céallqaodoulldoce or <3&4 e 5th caudal.... .| 168 | 360 | - 32z BOD alee. [eae e695 | <6 ercthucaudalie sc.) 9677 || B40) mgoen| meso ulesc: .: eee 615 ai emthecandalyeee-- TOW esss 300 7@ Sullierrsveve|| eer 622 3 emeth caudal eee.) i710 305 274 FiO he lcte veh iieeietene 587 agers Sigil Cell, 565.) uO 283 260 OBB est: Hiatere east 537 edAthcaudall sees 073 282 258 BO Sial ences < x “ MORO EONS 6 00.0 bod 0 Go00e 00 CdODOOODE x A. M. WVICCKOSYOP SENET CLUES EGLO Yrnrslefor eta aheroleter=fole(eitol valet el (ae x m POUAILS (Coens abocobagsboogsogunnoor x Jake IN Cynodontomys latidens Cope...... AGH OO Oren creer x os -RODENTIA. ISCHYROMYID-®, Paramys buccatus Cope......... eONoL selapscoicerer es x x eS delicationlWertyzr ce seek seme eo x sf delicatissininsmceidy sacreimierine vcietersisiore Sui CREODONTA. VIVERRID#. Viverrayus* protenus (Cope)..). sce vk -c ie see tue or < x ae ORE NGEE AGREES EO Y2508 6 G0 eG COU CDE DEO CORO OOCE OOS x BC PALZONICTID, | Palzonictis occidentalis Osborn and Wortman....| A.M. AU TAUSTD TOS. LOST Goi NS6hs dO b Po OO AEM Don UC Oe x Wis: MESONYCHID.®. Pachyena’ ossifraga (Cope)...........0000% aatiawile x HWE Ss: CCR CEI OLE MLCALG NVIOLEMMAM «erie = 00 iis 2 oo cies 2s x< A. M. Pachyena gigantea Osborn and Wortman.......... x | “ LDUSTECOS UGLOTIDT IOS, (Oy ENN Be oceHeCO CORIO PCO Oe Dio. | x ae CONDYLARTHRA. | | PHENACODONTID. | | Phenacodus primeyvus Cope *. 00... a. ie see ns x * || AoM. IDNR OslnS GueUhVOHNS (CGN? nooosdganoooceoo00G | | MURS: Phenacodus nunienus Cope®......- Ao BonorDeDdaoapoa) y ; A. M. Phenacodus wortmani Cope®........... sa0006000l| OX x Phenacodus apternus Cope. ..........- s5H00050|| 3X A.M. Phenacodus hemiconus Cope. ......++- podaoe einc.doll pan ee Phenacodus brachypternus Cope ..............--| X | X | ss Phenacodus macropternus Cope ......+eee eee a6. eral. oS a uy SMALOS COWS 55 900000005 Seeman s59005 Y< |i) Wotse a (Ectocion) osbornianus Cope........- soul 3X A.M. I oloyAUe CEES WCAWE 600000 0apapdsoc2ando0bDSGaD < We as RAMUS WH AVG ado0ccndgup9asodaoomccd0DGc x BG MENISCOTHERIID&, Meniscotherium chamense Cope. ............... : x |) WS. ss terree=Tubree) COpesree).cleerieieleclsel- ee AeaNls Meniscotherium tapiacitis Cope ....- IAA CE Ee < of Hyracops socialis Marsh *...... sdabod0GG0G000 DOO don x We 1 Type from Evanston, Wyoming, 2 Perhaps a small variety of P. przmaevus. 3 See note on this species on p. 36. 4 Doubtfully distinguishable from P. xunxzenus. 5 Eohyus Marsh (nom. nud., 1877) is perhaps a synonym of Phenacodus; E. distans (figured, 1894) might be taken for the very uncharacteristic m$ of that genus, and the description of Z£. robustus (1894) corresponds as far as it goes to the lower jaw of P. prim@evus. Wortman (Bull, Am. Mus. Nat. Hist., 1898, p. 101, foot-note) believes that Z. dzs¢tans is founded on the last upper molar of 7rzgonolestes etsagicus ; but the tooth as figured by Marsh is too large for that species, even on the supposition that m® is unreduced, which, judging from the reduction of the heel of mz, is not the case. ® Hyracops, which has been identified by Osborn with Mezzscotherium, differs considerably in its foot structure if Marsh’s figures are accurate. The large magnum, the entirely serial carpus, and the epicuneiform seen in Marsh’s figures of Hyvacops are not present in Wenzsco- therium, which has a carpal and tarsal structure very like that of Aufrotogonia, with small magnum, lunar supported partly on unciform, and other normal primitive features. The metapodials and phalanges are like those of Hyracops. io>) 1899. | Matthew, Fresh-Weater Tertiary of the West. 3 Ill. Wasatcu.—Continued. < g PAY ede eel | eel g fo) & —Q 4 § & eee ea) Ee] ae a | ele i si se alma| 4 AMBLYPODA. CORYPHODONTID&. Coryphodon (Bathmodon) radians Cope.........- henge AS Ns Coryphodon (Metalophodon) testis (Cope)...... x : Conyphodonirepanditss Caper eisai iia | x ae Corypuodoni lobatis Copesc.c sonnet 33: