special Collections ТИ a HARVARD UNIVERSITY. LIBRARY OF THE MUSEUM OF anc ah, оиа ZOOLOGY Munum V sew hor 23, юа “А 26,1435; JUL 9 6 1935 BULLETIN MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE, IN CAMBRIDGE VOL. LXXVI CAMBRIDGE, MASS., U. S. A. 1935 27 BE a STS ER TEE Tue Cosmos Press, Inc. CAMBRIDGE, Mass., U. 8. A. 1 | | CONTENTS . ]l.—GQoNTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY ‚ or THE Сознич Ноге AREA, Wyoming. І. A Detailed Study of the Strueture and Relationships of a New Zalambdodont Insectivore from the Middle Oligocene. By Erich M. Schlaikjer. (1 plate). November, 1933 ; : . 2.— CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY or THE Совнем Ноге Area, \Ухомтма. II. The Torrington Member of the Lance Formation and a Study of à New Triceratops. By Erich M. Schlaikjer. (6 plates). January, 1935 қ i : > à à 5 \ . 3.—-CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY or THE GosHEN Ноге AREA, Wyoming. ПІ. А New Basal Oligocene Formation. By Erich M. Schlaikjer. (8 plates). January, 1935 5 à У қ А - . 4.—CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY or THE GOSHEN Ноге AnEA, Wyoming. IV. New Verte- brates and the Stratigraphy of the Oligocene and Early Miocene. By Erich M. Schlaikjer. (41 plates). Мау, 1935 . . 5.—8oME ORDOVICIAM Сүвтірв FROM Russia. By Fred B. Phleger, Jr. (1 plate). July, 1935. . 6.—LEANCHOILIA AND OTHER Mip-CAMBRIAN ARTHROPODA. Ву Percy E. Raymond. July, 1935 PAGE 29 69 95 191 208 з] Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vor. LXXVI, No. 1 CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY OF THE GOSHEN HOLE AREA, WYOMING I A DETAILED STUDY OF THE STRUCTURE AND RELATIONSHIPS OF A NEW ZALAMBDODONT INSECTIVORE FROM THE MIDDLE OLIGOCENE. By Евісн М. ScHLAIKJER Wits One PLATE CAMBRIDGE, MASS., U.S. A. PRINTED FOR THE MUSEUM NOVEMBER, 1933 PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOÓLOGY AT HARVARD COLLEGE There have been published of the Вотљеттм Vols. I to LXV, LXVII- LXXIV, of the Memorrs Vols. I to LI, LIV. The BunLETIN and МЕмотвв are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collec- tions and Exploration. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs 18 sold separately. А price list of the publications of the Museum will be sent on application to the Director of the Museum of Comparative Zoólogy, Cambridge, Massachusetts. Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE VOD I XXVI NO. 1 CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY OF THE GOSHEN HOLE AREA, WYOMING I A DETAILED STUDY OF THE STRUCTURE AND RELATIONSHIPS OF A NEW ZALAMBDODONT INSECTIVORE FROM THE MIDDLE OLIGOCENE By Erich М. SCHLAIKJER Wits Ong PLATE CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM NOVEMBER, 1933 zem ms паа Наа No. 1.— Contributions to the Stratigraphy and Palaeontology of the Goshen Hole Area, Wyoming ii A. Detailed Study of the Structure and. Relationships of a New Zalambdodont Insectivore from the Middle Oligocene. By Евсн M. ScHLAIKJER INTRODUCTION This article is the first of a series of contributions on the stratigraphy and palaeontology of the Goshen Hole area. Goshen Hole is the name given to a region including about fifteen hundred square miles in southeastern Wyoming, which is more or less continuously surrounded by an escarpment of from four hundred to a thousand feet in height. The North Platte river bounds the area on the northeast. The escarp- ment begins a few miles west of the town of Torrington, and extends in semi-circular fashion around to the west, southwest, and south where it is interrupted at Bear Creek Mountain, thirty-five miles south of Torrington. To the southeast is Sixty-Six Mountain, and the eastern boundary is defined by two rather distantly separated prominences known as Eagle Nest and 'Table Mountain, which are situated along the Nebraska-Wyoming state line. For the past three summers the author has been in charge of fossil-collecting expeditions sent into this. area by the Museum of Comparative Zoólogy at Harvard College. During this time he has made a detailed study of the stratigraphy of the region and has prepared a geological map. 'The area presents splendid and abundantly fossiliferous outcrops of the Lance, Chadron and Brule formations and a lower Miocene formation. А new pre-Chadron post- Uinta formation is also present. Because of the large number of new fossil forms that have been discovered, and because of the several stratigraphic problems that have arisen, it is deemed advisable to publish a series of contributions, so that this information can be put into print prior to the preparation of a monograph on tbe entire area. Following this first contribution will appear a series of five or six papers dealing with the several formations and their faunas, and a final paper containing a summary of the problems of the whole area, emphasizing especially geological history and correlation of the forma- tions present with those of other areas. By way of acknowledgment I wish to express my deep gratitude to Dr. Thomas Barbour of the Museum of Comparative Zoólogy, who has made it possible for me to carry on this research both in the field and in the laboratory. I owe my thanks to President Henry Fairfield 4 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Osborn and to Curators Walter Granger, George Gaylord Simpson, Barnum Brown, and Charles Craig Mook of the American Museum of Natural History for the kindness and coöperation they have shown in placing at my disposal for study the very excellent vertebrate fossils in that museum. I am indeed profoundly grateful to Professor William K. Gregory, under whose inspiring guidance it has been a privilege to carry on this research. This article is confined to a detailed study of a new Zalambdodont insectivore from the Brule formation. The specimen is a beautifully preserved skull and jaws, and was found approximately two miles south and eight and one-half miles west of Torrington in the 5. E. 24. Sec. 19. T24n. R62 west of the sixth principal meridian. In strati- graphie position, the specimen was collected approximately sixty feet above the Chadron formation and two hundred and eighty feet below the highest outerop of Brule in this locality (see Fig. 1). Fiver channel deposit, corse boulder conglomerate, ES we Fine grained ЗВ pink tyor Brule суду: = М Chonge grodotionol 0 fine pinkish sandy clay. # Fine , ате) ctoss-bedded greensh channel 50106. S ч M BA x a RE kg 204% Tanvishpink cay with green streaks. Hhternodus gregory, elevation #3207 ~ о = 2-3 Lord greemsh-groy sandstone. S ж Reddish-green clay. Leptomeryx. N га Grem sands; somewhat consolidded. LS =: Green clays and sands. Fra. 1. Geological section of a canyon in S. Е.1 Sec. 19. T24n. R62w. Goshen Co., Wyoming. E. M. Schlaikjer 1932. | | SCHLAIKJER: А NEW ZALAMBDODONT INSECTIVORE 5 Inasmuch as the Brule formation is completely represented in the Goshen Hole area and is approximately four hundred feet in thickness, it is evident that the specimen comes from what may be called lower Brule, or the equivalent of what is known as the lower Oreodon zone in South Dakota and Nebraska. The matrix is of the usual very fine- grained pinkish-gray clay so characteristic of the Brule formation throughout the Great Plains area. The fossil record of Zalambdodont insectivores is very incomplete. Only six genera and as many species have been recorded from the entire world. The new species herein described adds a seventh. The genera and species of the fossil Zalambdodonts, the geological forma- tion from which each comes, and the geographical distribution is as follows: DELTATHERIDIIDAE Deltatheridium pretrituberculare Gregory € Simpson. Djadokhta formation. Mongoha. Deltatheroides eretacieus Gregory & Simpson. Djadokhta formation. Mongolia. Hyotheridium dobsoni Gregory & Simpson. Djadokhta formation. Mongolia. PALAEORYCTIDAE Palaeoryctes puercensis Matthew. Torrejon formation. New Mexico, North America. SOLENODONTIDAE Subfamily SOLENODONTINAE NEW. Mieropternodus borealis Matthew. Lower Oligocene formation. Montana, North America. Subfamily APTERNODONTINAE Apternodus mediaevus Matthew. Lower Oligocene formation. Montana, North America. Apternodus gregoryi Schlaikjer. Middle Oligocene formation. Wyoming, North America. It will be noticed at once that the three American Oligocene forms are placed in the family Solenodontidae. This problem of taxonomy will be considered later in these pages. The drawings in this article are by Mrs. Helen Ziska, and the speci- men was most skillfully prepared by Mr. Albert Thomson. | | 6 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY APTERNODUS GREGORYI! sp. nov. Type. M. C. Z. No. 17685. Nearly complete skull and jaws. Col- lected by E. M. Schlaikjer, 1932. Horizon and locality. Middle Oligocene. Lower Brule formation, Goshen County, Wyoming. Specific Characters. This genus includes a single species, the type (described by Matthew 1903). Inasmuch as the type is only that of а partially complete posterior portion of a lower jaw with the last two molar teeth and two paratype jaw fragments, it seems most advisable to list here only those characters of Apternodus gregoryi which differ from those seen in A. mediaevus. Those differences are as follows: 1. Jaw larger and more massive. 2. Coronoid heavier, more antero- posteriorly expanded, and more external to the tooth row. 3. Angle stronger and more inflected. 4. М» з series longer. 5. М» shorter, wider and heavier; talonid more reduced; and paraconid more lin- gually situated with respect to metaconid. 6. M; longer and slightly higher; and talonid larger. In addition, А. gregory differs from para- type А. М. №. Н. No. 9608 in having a very much heavier mandibular condyle; and from paratype A. М. №. Н. No. 9607? in that the central cusp of Рз is less emphasized and the posterior cingulum extends more on the external side of the tooth. DESCRIPTION Upper Dentition. Dental formula ?'7——. Only the roots of two incisors are present on the right side. These are not preserved on the left. There is some doubt about the number of incisor teeth, and since the anterior portion of the premaxillary is broken, it cannot be defi- nitely determined if another tooth were present in front of the two pre- served. It is probable, however, that there were only two. The roots of the two incisors are round and show that the second from the canine is the larger. "There is a diastema of ca. 1 mm. between the last incisor and the canine. The canine is larger, recurved, and has two roots. The crown is ovate in outline. It is made up of a single, large, anteriorly placed cusp and a small, low, conical heel. P? is double rooted and has the same general form as the canine, but it is ! Named in honor of Professor William K. Gregory. ? This specimen is the front part of a left lower jaw with only one tooth preserved. In his description of A. mediaevus Matthew figured this specimen and considered the tooth to be Ра. I examined this specimen under the microscope and found the tooth to be P; and that the crown had been glued on the roots of Py. It is now restored to the roots of Рз on which it fits perfectly. SCHLAIKJER: A NEW ZALAMBDODONT INSECTIVORE 7 only half as large and the cusp is proportionately much lower. The tooth is separated from the canine, as well as from P*, by a 1.5 mm. diastema. P*-M? are three-rooted. P? is almost symmetrically tri- angular with a large, relatively high, centrally placed conical cusp. The anterior, posterior and internal corners are developed into cusp- like einguli. The one at the posterior corner is more developed, higher, and more continuous with the central cone than are the others, thus foreshadowing the condition of Pt. The width of P is slightly greater than its length. It is molariform, and affords an excellent example of Fig. 2. Apternodus gregoryi. Left upper tooth series, Р? to M3. Five times natural size. the transition from the premolar to the molar tooth pattern. Because of the development of the cusp at the posterior angle of the tooth, P* has become obliquely triangular in outline. This cusp is now almost as high as the main central cusp and is in complete union with the latter. The external anterior extension of this posterior cusp is now connected with the сизр at the anterior corner of the tooth and is developed into a complete external cingulum, the central part of which is developed into a prominence which is the homologue of the parastyle of M!. 'The internal angle of the tooth is lower than any other part of the crown and presents a cusp-like cingulum with a single minute conical projection. All the molars have a much greater transverse than an antero-posterior diameter. M! is the largest. The parastyle, metastyle and para-metacone cusp are of equal height. The cusp at the anterior angle of the tooth is not quite of the same height and is confluent with a complete, low, very small anterior cingulum. The internal angle of the tooth is rounded. It presents a prominent cingulum which is cusp- like, and shows a slight differentiation into two areas. Of these two areas, the anterior is the larger and may be called the protocone, while the posterior represents the hypocone. M? is smaller than М! and differs from the latter in the reduction of the metastyle and in that the anterior external eusp is conical, closely approximated to and almost аз high as the parastyle. M? is expanded transversely and shortened antero-posteriorly. This is brought about by the approximation of the 8 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY metastyle and para-metacone cusp; and by the development of the antero-external cusp so that it is larger and higher than the parastyle. The Skull. The right side of the skull is perfectly preserved with the exception of the loss of the anterior tip of the premaxillary. On the left side part of the cranium is chipped away and the cast of part of the brain is exposed. The preorbital.region of the skull is relatively short and uniformly cylindrical. The postorbital region is about twice the length of the preorbital. The skull is that of an adult and most of the sutures are completely fused. The basioccipital is short and is below the level of the palate. The whole skull is bent upward from the basioccipital axis. The occipital condyles are low, broad and widely separated. The foramen magnum faces backward and downward, and there is no evidence of a notch either above or below. It is large, and oval in outline; being broader than high. The whole posterior part of the skull is greatly expanded as а result of the development of the squamosal, mastoid portion of the periotie, and exoccipital bones into a curious plate-like development at the side of the skull. This plate-like structure is rectangular in outline and is oriented obliquely to the long axis of the skull, and is consider- ably below the level of the sagittal crest. The dorsal edge is heavy and curves somewhat downward and outward from behind forward. 'The antero-dorsal corner presents a blunt zygomatic process. This slightly overhangs the large open and very transversely expanded glenoid cavity. The glenoid cavity is elevated; the dorsal edge lying on a level with the inferior part of the orbit. It faces forward and slightly out- ward. The post-glenoid process is small. It projects downward and forward about half way under and along the entire width of the glenoid cavity. The infero-anterior angle of this plate-like expansion is formed by the post-tympanic process. Between this and the post-glenoid process is a large and fairly deep notch for the external auditory meatus. When the jaws are articulated with the skull and are opened, the angle of the jaw fits into this notch. Opening of the mouth, there- fore, must have interfered with the ability of the animal to hear. The post-tympanic process is stout, rounded and marks the most ventral point of the entire skull. Extending backward and upward from this process, the ventral portion of the skull expansion ends in the postero- inferior angle, which is stout and bluntly developed. The posterior margin of the expansion is almost vertical in position and projects only slightly behind the condyles. There seems to be a well defined suture extending dorso-ventrally just anterior to the posterior margin of this plate-like expansion. I have considered this to be the suture between 9 SCHLAIKJER: A NEW ZALAMBDODONT INSECTIVORE the squamosal in front and the mastoid portion of the periotic behind. The sutures between the basioccipital and basisphenoid; basisphenoid (ES = (III Е (Y А == nn, зә т б У ANUS асса асу IS Е LUN == 4 ~ Аист MS, NR ~ > a ШІ өң NN М vn 3 N “ER Й Fig. 3. Apternodus gregoryi. Dorsal, Іа ега] and palatal views of the skull. Enlarged two times natural size. and presphenoid; and presphenoid and vomers cannot be determined. The pterygoid plates are low. The suture between the basioccipital and 10 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY petrosal seems to be well defined. 'The basisphenoid is flat, as in Solenodon and Microgale, and bears no pit as does this bone in the skulls of Centetes, Ericulus, Echrinops and Erinaceus. It possesses по descending lateral wing embracing the tympanie cavity on the mesial side, unlike the condition in all the recent zalambdodonts with the exception of Solenodon. There seems to be a distinct suture between the petrosal and the exoccipital extending from the basioceipital to the foramen lacerum posterius. The petrosal is in close contact with the basioceipital. The anterlor part is concave and forms the dorso-posterior wall of the deep, rounded tympanie cavity. Posteriorly the petrosal is rounded, and postero-externally it presents a knob-like projection. The tympanie is not preserved but it was probably ring-like and was probably braced between the petrosal projection and a rather low swelling which is present on the posterior surface of the post-glenoid process; a condition especially similar to that in Solenodon. Viewed from behind, the skull is bracket-shaped in outline. Tt is broad and relatively low. Complete sutures between the exoccipital and the mastoid portion of the periotie, between the latter and the supraoccipital, and between the supraoccipital and the exoccipital are not distinct. There is some evidence, however, that their arrangement 13 as.shown in Figure 4. Foramina are absent, and there is а shght Ета. 4. Apternodus gregoryi. Posterior view of the skull. Enlarged two times natural size. depression on the margin of the foramen magnum immediately above the condyle. The squamosal is of peculiar shape in that it is the major element in the plate-like development at the side of the skull. The presence of interparietals cannot be determined. The lambdoid and sagittal crests are prominent and massive. Anteriorly the sagittal crest divides into two low but sharply defined ridges leading to the rudimentary postorbital processes. There is a slight postorbital con- SCHLAIKJER: A NEW ZALAMBDODONT INSECTIVORE igi striction. The sutures of the cranium are completely fused. The interorbital region is flattened and broad. 'The antorbital region is cylindrical, and somewhat broadened anteriorly. The nasals are completely fused. The premaxillary is large and is heaviest in the area of the enlarged incisor. There is a distinct suture on the side of the face between the maxillary and premaxillary. The facial portion of the maxillary presents a relatively large depressed area dorso-anteriorly to the orbit. This depression was undoubtedly occupied by the levator labii superioris proprius muscle, which must have been large. The large size of this muscle may indicate that there was present an extended naso-cartilage similiar to that in Solenodon. There is an abrupt trans- verse constriction of the face across the maxillaries from M! to the canine; a condition almost identical with that seen in Solenodon. The boundary of the lachrymal cannot be determined. The orbit is dimin- ished, and the antorbital border is bar-like and set out from the plane of the front of the face. The zygomatic arch is completely lacking. The infraorbital portion of the maxillary shows only the slightest indication of a malar projection, and that projection is above the front of M!. The region of the palate anterior to P? is flattened and narrow. Posteriorly the palate is broad, and is concave transversely and antero- posteriorly. It extends behind M?. The palatines are large and extend forward to M!. Each has a triangular postero-external wing extending back to meet the pterygoid plates. The exact boundaries of the orbito- sphenoid, alisphenoid and vomers are indeterminable. Cranial Foramina. The infraorbital foramen is situated above the center of P?. It is large, round and faces forward due to the outward and forward projection of the antorbital border. The infraorbital canal is short. The lachrymal foramen is round, proportionally large and is mar- ginal in position; facing outward and slightly backward. The anterior palatine foramina are small and set closely together. 'The posterior palatine foramina are minute, and occupy a rather postero-mesial position to the last molars. The spheno-palatine foramen is about equal in size to the optic foramen. The latter is situated immediately and centrally in front of the very large opening formed by the apparent confluence, externally at least, of the foramen lacerum anterius and the foramen rotundum. The alisphenoid canal is smaller than the foramen ovale and is situated antero-internally to the latter. The foramen ovale is large, round and is situated near the internal border of the glenoid cavity. 12 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY The postglenoid foramen is situated inside and at the antero- external corner of the tympanic pit. The foramen lacerum posterius is large, round and occupies a posi- tion slightly internal to and immediately behind the knob-like projec- tion of the petrosal. The condylar foramen is small, oval in shape, situated more mesially than the foramen lacerum posterius and is located at the margin of the articular surface of the condyle. There seem to be only two squamoso-parietal foramina. They are small, oval and are situated antero-internally to the postero-dorsal corner of the plate-like structure at the side of the skull. Measurements of the Skull mm. Exoccipital condyle to anterior of сапіпе.................... 38.5 Width across post-tympanic processes. . ІН Әнет. НОБ Width across postero-inferior corners of auiditosy. plates. nirai 20.0 Greatest length она ову plate йв sein ne ae бота 15.8 Greatest height of andivory, равен cen inde f et ES Sed Width of occipital condyles. . 2 по ЊЕ Height of occiput, basioccipital tó top of saggital de па pullo Height of skull above alveoli of M+.. ай. doa: NE ond ША: Front of antorbital ridge to occipital coxidyle кдм. И.О 82.1 Front of antorbital ridge to front of enlarged incisor.......... 1147 Greatest width across maxillaries above posterior of canines... 7.6 Width across maxillaries on postero-external alveoli of Ms!.... 13.2 Length of tooth row anterior of canine to ar of M? on alveoli......... - 5120220 у Anterior of canine to Р», inclusive; on CE Жаралы 010.8 Eeupth of eine ИСП, ыыр» р 8.0 ООО Width of canine 10525 vix audies ил. мит. постати. Gr (еше ue ol P. coños bus anen adsa ii 28 Width of P* at center of crown on alveoli... os. ao tiroa ai 80 Greatest length ов ode al 7”... 2.5 Width of M! at center of crown on alveoli................... 3.6 Height of МІ, external. . iade әде івано nod del trad Greatest length of M?.. LT 2220 750 12158 Width of M? at center ote crown on ӨҢ ЕМЕН ао Жөні 3.4 О debes ka Бе 1.0 Greatest ОО rata sed ud alie dentia. 3.0 SCHLAIKJER: А NEW ZALAMBDODONT INSECTIVORE Is Lower Dentition. Dental formula s». 3.3. 15 is the largest. The crown is broken off, but the stub of the root is oval in outline; being compressed laterally. Is, the canine and Р» are missing, the alveoli however, indicate that they were single rooted, round, about the same size, and that they were set close together. Рз-Мз are two rooted. P; is oval in outline and consists of a single, centrally-placed, high, Ета. 5. Apternodus gregoryi. A, Superior view of jaws. B, Lateral view of right lower jaw. Enlarged two times natural size. conical cusp, and an anterior and a posterior cingulum. The latter is higher and slightly less developed than the former. P4 is very molari- form and affords a beautiful example, as did the fourth premolar of the upper dentition, of the transition from the premolar to the molar tooth pattern. 'The trigonid is made up of a high external angular сизр, the probable homologue of the molar protoconid, and an antero- and a postero-internal eusp which correspond in position to the para- and metaconids of the molars. The postero-internal cusp is closely approxi- mated to and almost as high as the external cusp. The antero-internal 14 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY cusp is smaller and lower than the other two cusps of the trigonid. There is a small antero-external cingulum. The tooth has a low, minute, posterior heel. M; is the largest tooth of the premolar-molar series. It, as well as М» and з, are of essentially the same pattern as P4; the only major difference being the further separation of the protoconid from the metaconid. This results in a widening of the teeth. Mı_ are wider than they are long and M; is longer than it is wide. The propor- tional increase of the length of M; is brought about by the relative enlargement of the posterior heel. See Figure 5. The Mandible. Тһе mandible is relatively short and heavy. The symphysis is long — about one-third the length of the entire jaw,— and the rami are not fused. This strengthening of the symphysis is in direct relation to the enlargement of the second lower incisor. The ramus Is deep and the posterior mental foramen lies beneath the center of Ра. Just posterior to the tooth row there is a dorso-ventral con- striction of the ramus and the posterior part of the jaw is curved up- ward. The coronoid is very large and stout. И is expanded antero- posteriorly. It is unusually high, is directed upward and forward, and occupies a position external to the tooth row. The angle is small, protrudes slightly behind the condyle, is rounded externally, and is sharply and considerably inflected. The condyle is heavy and broadly expanded transversely. It is horizontal in position and is somewhat elevated above the tooth row. 'The foramen mandibulare is situated just below and in front of the condyle, and is on the level with the mesial base of the tooth row. Measurements of the Jaw mm Posterior od (on alveolus) to сопдуе....;................. 26.5 Rostenor or ТО Wis on Ве ори. Le 15.5 Why On VCO ane А АМ T) No шой аео 2722222 S nU та Depth of jaw undér center of Ms (inte mal): TEOR по Jib de 6.0 Posterior of М» to foramen mandibulare. ................11. 6.5 Height: of @отопоїй from bottom ot алуы мо ом), о род bou 1855 Müdthtoticondyles 250007 8) 200 20081 Pod Ии Е ог о ро 191810 Heichtor М он и Td Anteroposterior length of Mission 2.0 Күте нова een уг 2.6 ето и о Моно м war. & AOL, UE EIL 2) Antero-posterior length of М» SCHLAIKJER: A NEW ZALAMBDODONT INSECTIVORE 15 Widthoi Mob? acest metum Me ~ dido ea Height of М... Boios droht Та esty а 0 Antero-posterio length of M... so cto Usi tme dao 128. Width of Ма. Aad ideo tenis barsa wir ceres ш е Height of Р... ii jasal i De. warren seh ned ahead Antero-posterior length of Py. ио Қылған ta domi egt піні. Width of Parad мл йо niis lo bie: A Den d Jost amour la. Нејоћњо Pine. ciet Hr CD A Donee El aca, 2.6 Antero-postenionilength of: Paakca рки orae астына На 2.0. азо Eir: абын. na ale edel AFFINITIES It is very unfortunate that the type material of the genus Apternodus is so fragmentary. There can be little question, however, that the species deseribed above belongs to that genus, and the differences pointed out between А. mediaevus and А. gregoryi, should, for the present at least, be considered as of specific rather than of generic value. 'The difference between the two species is mainly one of degree. With the information at hand, it seems that А. gregory? has carried to greater emphasis the specializations already established in A. mediaevus, and it is probable that the latter is its direct ancestor. In 1910 Matthew gave a brief description (pp. 33-36) of an un- usually complete insectivore skull and jaw collected from the lower Oligocene, Titanotherium beds “in the neighborhood of Bate's Hole, north of the Laramie Plains" Wyoming (p. 33). He referred this speci- men to A. mediaevus. The specimen is deposited in the museum at the University of Wyoming, and unfortunately I have not had the oppor- tunity of studying it.! Excellent enlarged photographs of it, however, are on file at the American Museum of Natural History. Referring to these and to Matthew's description Т have been able to draw the following conclusions. If the explanation of the plate in Matthew's publication is correct,— that the published photographs are three times the natural size of the specimen,— then the University of Wyo- ming specimen is about two-thirds the sizeof A. gregoryi. If, however, the record with the photograph negatives in the American Museum of Natural History is correct, then the two specimens are approximately of equal size. From an inspection of the photographs, some of the ! The specimen was reported as lost. It has been found since this manuscript went to press, and the results of a study of it will appear in а subsequent paper. 16 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY characters in which the University of Wyoming specimen differs from А. gregoryi are: postorbital region proportionately much longer; face in front of antorbital bar proportionately shorter; proportionately broader across the auditory region; skull higher above M?; palate proportionately broader across M!; length from posterior mental fora- men to anterior of base of coronoid process much longer; no diastema between last I and C, and only the slightest diastema between C and P? and between Р? and P3; M? with greater antero-posterior length and with well developed proto-hypocone сизр; and, M! with well developed anterior and posterior cinguli. In other respects, such as tooth formula and development of auditory plates at the sides of the cranium, the two specimens are very similar. The problem of the relationship of Apternodus to other zalambdo- donts is a most interesting one. There has, as yet, been no fossil form recorded, which may be spoken of as being in the direct ancestral line of this genus. Micropternodus borealis from the lower Oligocene (a lower jaw described by Matthew, 1903) is the most nearly related. The two genera are similar in the following characters: long symphysis of the lower jaw; high trigonid and low talonid; outline of the trigonid; protoconid the tallest cusp of the trigonid; and, metaconid more em- phasized than the paraconid. Apternodus is different from Mieropter- nodus, however, in the greater massiveness of the mandible, the loss of L, the single rooted Р», the talonid more reduced, and the extreme ex- ternal displacement of the coronoid process from the tooth row. These two genera do not seem to be very distantly related, and it is probable, Т think, that their common ancestor need not have lived earlier than middle or late Eocene times. The only other known fossil zalambdodont from America is Palae- oryctes puercenis (described by Matthew, 1913) from the Paleocene (Torrejon formation) of New Mexico. Dr. Matthew (1913, p. 310-314) considered Palaeoryctes, as known, to be a generalized zalambdodont that could have given rise to the diverse modern types. When further discovery reveals more material, however, Palaeoreytes may prove to be somewhat specialized (especially in the anterior region of the skull) and off the line of Apternodus ancestry, although Apternodus must have originated from a form very close to Palacoryctes. As shown by Gregory (1922 p. 156, pl. 5) Palaeoryctes, (or its European equivalent, not yet discovered) could well have been the direct ancestor of Potomogale. Dr. G. G. Simpson in his very excellent paper on " Affinities of the Mongolian Cretaceous Insectivores” (1928) pointed out that the Del- tatheridiidae are near the starting point of the Zalambdodonts. SCHLAIKJER: A NEW ZALAMBDODONT INSECTIVORE фі Deltatheridium, however, was already on the way to give rise to the Creodont type of dentition. Considering the question of the relationship of Apternodus to the recent forms, it seems that Solenodon, the Central American zalambdo- dont, is its nearest living relative. This similarity is established by the following characters which are common to both genera. 1. Slight postorbital constriction. 2. Constrietion of the antorbital part of the face; especially the abrupt constriction across the maxillaries from M!-C. З. Tendency for the mesial basal cusp on M? and especially М! to divide into a protocone and a hypocone. 4. Basisphenoid with no pit. сл . Basisphenoid with no lateral descending wing embracing the tympanie cavity on the mesial side (this character is distinctive of Solenodon and Apternodus among the zalambdodonts). 6. c Ring-like tympanic braced between the petrosal projection and the prominence on the postglenoid process. In fact, the entire basicranial region, with the exception of the descending proc- esses of the auditory plates, is almost identical with that of Solendon. Ч . Long mandibular symphysis, and general shape and massiveness of the mandible. 8. Large, heavy, high and erect coronoid process. 9. Broadly expanded mandibular condyle. 10. Enlarged second lower incisor. In tooth formula, however, Apternodus gregoryi is nearer Echinops and Ericulus, because of the apparent loss of the first upper and lower incisors. In molar pattern the two latter are quite different from А. gregoryi in that the proto- and hypocones are diminutive and are widely separated; being situated on the anterior and posterior sides, respectively, of the para-metacone cusp. In some respects, such as the lack of development of the internal cusp of M? and the lack of well defined proto- and hypocones on M!2, Microgale seems to present a molar pattern more like А. gregory? than does Solenodon. The premolar- molar teeth of Solenodon, however, present no characteristics which could not be derived from the A. gregoryi pattern. All in all, it may be said that the similarity of tooth pattern between Microgale and 18 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY Apternodus, if as great, is not greater than the similarity between that of Microgale and Solenodon (see Fig. 6). гу 5.265 А Ето. 6. Comparative drawings of left upper cheek teeth. А, Deltatheridium pretrituberculare. В, Palaeoryctes puercensis. С, Apternodus gregoryi. D, Microgale dobsoni, A.M.N.H. No. 31261. Е, Solenodon paradoxus, A.M.N.H. No. 28270. (A, & B, after Simpson.) Not to scale. In spite of all the above listed similarities between the two genera, А. gregory? cannot hold claim to a place in the direct line of ancestry | | | | | SCHLAIKJER: А NEW ZALAMBDODONT INSECTIVORE 19 of Solenodon. It presents a number of specialized characters which rule it out of that rank. "Those characters are as follows: development of auditory plates on sides of skull, extreme reduction of zygomatic root on maxillary, probable loss of first or third upper incisor, loss of first lower incisor, second lower premolar single rooted, extreme ex- ternal displacement of coronoid process from tooth row, and extreme Widening of the mandibular condyle (both the latter are probably re- sultant of the broadening of the occipital region of the skull and the development of auditory plates). In 1910, Matthew pointed out, quite correctly I think, that the specializations of Apternodus were of sub- family rank and suggested the name Apternodontinae. From the evidence presented above it seems evident that this subfamily should be included in the family Solenodontidae rather than in the Tenrecidae (Centetidae) as suggested by Matthew. Dobson (1882) defined the family Solenodontidae. Along with a number of distinctive anatomical characters listed, he gave the follow- ing skull characters: dentition en, I, deeply grooved internally, Pt and molars with flat V-shaped crowns and notched external mar- gins, skull with interorbital (postorbital) constriction, and mandibular condyle transversely extended. He also pointed out a number of characters in common with Myogale and Scalops. Leche (1907, pp. 44-46) felt that Dobson had overemphasized the likenesses of Soleno- don with the Talpine genera, and listed a number of characters to show the relationship of that genus to the Centetidae. He did not, however, question the fact that Solenodon should be placed in a separate family. In 1910, Dr. Gregory, as a result of his careful and detailed study of Solenodon paradoxus, listed (pp. 225) several additional skull characters of that genus. Among these are: 13 much enlarged; pseudo- Protocone (protocone) and hypocone forming small distinct cusps; по tympanic wing on the basisphenoid; and petrosal with a small tympanie process. It seems justifiable now to redefine, on the basis of Skull characters at least, the family Solenodontidae. Such a defini- Чоп, together with a definition of the subfamily Apternodontinae and the subfamily Solenodontinae (a new name which I propose) is as follows: Family Solenodontidae. "Tooth formula MIX I, enlarged. Protocone and hypocone forming small distinct сизрз. No tympanie Wing on the basisphenoid. Petrosal with small tympanic process. Slight postorbital constriction. Abrupt constriction across the maxil- aries from M! to C. Mandibular condyle transversely extended. “Arge, heavy, high and erect coronoid process. Bubfamily Apternodontinae. Tooth formula Aaram ade 20] gest nlarged. 20 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY Development of auditory plates at sides of skull. Р» single rooted. Extreme external displacement of the coronoid process from the tooth row. Subfamily Solenodontinae. Tooth formula 2173. Г enlarged. P; two-rooted. Antero-posterior ridge on the talonid confluent with the base of the trigonid. That Mieropternodus is a zalambdodont has not been questioned. Tts exact position in that group, however, has caused considerable speculation. In 1909 (p. 103) Matthew listed it under the insectivores as а member of an indeterminable family. In 1912 (p. 520) Osborn classified it with the Solenodontidae. In 1910 (p. 259) Gregory stated that “Micropternodus is also very likely a member of the Centetidae.” More recently, Matthew (1919, p. 174) mentioned that it “may have been a distant relative of Solenodon." In his description of Micropter- nodus Matthew wrote (1903, p. 205), “Pz is small and one-rooted, canine small, incisors small, sub-equal. No diastema except a slight one behind Рз.” I have examined the specimen, and it is very obvious that the slight diastema behind P; of which Matthew spoke, presents a small but very distinct alveolus. Р» therefore is two-rooted and was probably as large as Рз. The size of the canine alveolus indicates that this tooth was larger than any one of the incisors. Also, 15 is larger than I, or Ла. In comparing Micropternodus with Solenodon, it is similar to the latter in the following characters: dental formula 3-1-3-35 I; larger than either Із ог Iı; Pa two-rooted; antero-posterior ridge on talonid confluent with the base of the trigonid (this especially true of Mi; in Solenodon); general shape of the mandible; and the position of the anterior base of the coronoid process (the posterior part of the jaw is missing). These similarities, it seems to me, entitle Micropternodus to be placed in the subfamily Solenodontinae along with Solenodon. It cannot be said with certainty, however, that M icropternodus is or is not directly ancestral to Solenodon until more material is discovered. IE Micropternodus is not ancestral to Solenodon, it certainly is not far from a form that is, and if this be true, I can see no reason for placing the common ancestor of that form and Micropternodus farther back geologically than latest Eocene. In this consideration of the relationships of Apternodus to other zalambdodonts, mention should be made of two additional fossil forms which have been assigned to the zalambdodont group. They are Neerolestes patagonensis from the Miocene of South America, and Arctoryctes terrenus from the lower Miocene of South Dakota. Necrolestes was described by Ameghino in 1891. At that time he ЕЕ e | SCHLAIKJER: A NEW ZALAMBDODONT INSECTIVORE 2t pointed out its resemblances to Chrysochloris, the Cape golden mole of South Africa. In 1905, Professor W. B. Scott of Princeton restudied the genus, and his conclusions confirmed the suggestion made by Ameghino. He considered Necrolestes as а distant relative of Chryso- chloris (pp. 379-381). In recent years the material at the Princeton Museum was inspected by Dr. G. G. Simpson, and he came to the conclusion that “N ecrolestes resembles Chrysochloris chiefly in adaptive characters in skull and fore-limb, but very significant differences are also seen (the detailed structure of the less plastic parts of the skull is Moreover inadequately known), and the pelvis and hind-limb are quite unlike those of the African form. . . . It seems possible that it is à convergent form and not really a member of the Chrysochloris group, although apparently an insectivore and perhaps even a zalambdodont” (1927. pp. 294-295). Since there seems to be no conclusive evidence that Necrolestes is a zalambdodont, I refrain, for the present at least, from including it in that group. A left humerus from the lower Miocene of South Dakota was briefly described by Matthew in 1906. The specimen was not figured, and was not given a name until a year later when Matthew, without addi- Чопа! discussion, gave the name Arctoryctes terrenus (p. 172). Matthew Considered it as positive evidence of the former existence of Chrys- ochlorids in this country. He further stated that, “The humerus of Chrysochloris is, however, so peculiar and characteristic in form, as described by Dobson (Monograph of the Insectivora) and shown in the figures and specimens with which comparison has been made, that there can be no doubt that the fossil specimen belongs to the family, although somewhat less specialized than the modern genus" (p. 787). Recently I have had the opportunity of studying the specimen in de- tail and I can find hardly the slightest resemblance between it and the humerus of Chrysochloris. Dobson's figure of the anterior view of a left Chrysochloris humerus to which Matthew referred is shown in Fig. 7A. The posterior view is from a specimen in the American Museum of Natural History. Arctoryctes presents only one major Chrysochlorid similarity,—the large size of the bicipital groove (see Fig. 7B), and this character is certainly not distinctive of the Chrysochlorids. It seems to be Talpine in almost every other respect. Comparing it, for example, with the humerus of Scalops the similarities are indeed strik- Ing (see Fig. 7C). It is true that the Arctoryctes humerus is more ex- panded distally, and that the deltoid process is more pronounced and Occupies a more internal position. Also, the surface of the external 22 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY tuberosity is more anteriorly placed, hence the insertion for the brachialis anticus is more on the front of the humerus. These differ- ences, however, are only differences of degree. If the external tuber- Uf. pect ridge / ext cond Jor : 2” radius HAS uina froot. sur ulna 7700/7 sur Sur ofext tub тивну nl tah í Fata Жару» Ni dat YU A... pectoralis may delt ЫЛ ext cond Jor QUEM и ie 2 TEE uina troch sur ulna Doo SUL i плив surofextlub en | ig tat tub bicip ges ext Ib P PE - Decr delt ( C E Mage {с brachanitius) aL. enten Sor i oo ext cona for fi | A Eos y ау LN... | ulnatroch.sur ula Troc. sus | Ета. 7. Anterior and posterior views of left humeri. A, Chrysochloris trevelyani. Anterior view after Dobson, posterior view A.M.N.H. No. 41500. B, Arctoryctes terrenus A.M.N.H. No. 12864. С, Scalops sp. A.M.N.H. No. 36946. Enlarged one and one-half times natural size. osity were rotated externo-posteriorly, and if the bicipital groove were shortened, a form of humerus would be developed almost identical with the Scalops type, although I do not wish to infer by this that Arctoryctes is ancestral to the Talpid humerus. The true moles were | | | | | | | | SCHLAIKJER: А NEW ZALAMBDODONT INSECTIVORE 23 already established in Oligocene times, and the humerus of Talpa from the Oligocene of Europe; though slightly more primitive in some respects, is essentially the same as that of the recent form. . Proscalops secundus, a primitive though a somewhat highly special- ized pre-talpid occurs in the same geological formation in which Arctoryctes was found. Only the skull of this form is known. It is not improbable that Arctoryctes, because of its fossorial specializations, is the humerus of that form or one closely related. "This suggestion was made by Matthew (1928, p. 71) who stated, however, that “if this be 80, Proscalops, P. secundus at least, can hardly belong to the Talpidae, but must represent а distinct family.” It would seem, that if Arcto- ryctes is the humerus of Proscalops, that would be additional evidence for placing Proscalops with the Talpids rather than creating a new family for it. The problem of the affinities of Proscalops, however, is foreign to the subject of this paper. It was suggested by Matthew (1928, p. 71) that on the basis of fos- sorial specializations Arctoryctes might be the humerus of Apternodus, though the latter had not been found after the lower Oligocene. From what has been said above concerning the affinities of Apternodus, it is highly improbable that this form had the Arctoryctes type of humerus. The evidence indicates conclusively, I think, that Arctoryctesis nota Chrysochlorid. It seems to be related most nearly to the Talpids, and ought to be included, on the basis of the known material, with that group. Since Epoicotherium (Xenotherium), the previously supposed Chrysochlorid from the lower Oligocene, has been shown to be an edentate (Simpson 1927), and since Necrolestes and now Arctoryctes have also been eliminated from the Chrysochlorid group, there re- mains no recorded evidence of true Chrysochlorids anywhere in the World with the exception of South Africa where the recent forms occur. 24 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY CONCLUSIONS 1. Apternodus mediaevus was probably in the direct ancestral line of A. gregoryi. 2. Of the known fossil forms, Micropternodus borealis is the nearest relative of Apternodus, and they probably had a common ancestor in middle or late Eocene times. 3. The Apternodus-Micropternodus line was probably derived from a Paleocene form not far removed from Palaeoryctes. Palaeoryctes, as pointed out by Simpson, is a probable descendant of the Deltatheridi- dae. Either it or its European equivalent could well have given rise to Potomogale. 4. Of the recent zalambdodonts, Solenodon is the nearest relative of Apternodus, and the latter should be included in the Solenodontidae under the subfamily Apternodontinae. 5. Micropternodus, if not the direct ancestor of Solenodon is very close to a form that is, and should be included in the same subfamily. 6. Arctoryctes is not à Chrysochlorid. It seems, moreover, to be most nearly related to the Talpids. Necrolestes, as shown by Simpson, is probably not a Chrysochlorid. Though probably an insectivore, it is not included with the zalambdodonts because of the lack of evidence. 7. There is no recorded evidence of the pre-existence of true Chrys- ochlorids. | | SCHLAIKJER: A NEW ZALAMBDODONT INSECTIVORE 25 Recent Solenodon Pleistocene Pliocene Miocene Oligocene A. gregoryi Apternodus mediaevus Micropternodus Eocene ? ш Paleocene Palaeoryctes Deltatheridiidae Cretaceous . Кіа. 8. Diagram illustrating the affinities of the fossil zalambdodont Insectivores. BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY BIBLIOGRAPHY Doßson, С. E. 1882. А monograph of the Insectivora. London, pp. 1-172, pls. 1-28. GREGORY, У. К. 1910. The Orders of mammals. Bull. Amer. Mus. Nat. Hist., 27, pp. 1—524, 32 text figures. 1922. Origin and evolution of the human dentition. Williams and Wilkins Co., Baltimore. pp. 1-548, pls. 15, 353 text figures. Gregory, У. К. and Simpson, G. С. 1926. Cretaceous mammal skulls from Mongolia. Amer. Mus. Novitates, No. 225, pp. 1-20, 19 text figures. LzecHE, У. 1907. Zur Entwicklungsgeschichte des Zahnsystems der Saugethiere, zugleich ein Beitrag zur Stammesgeschichte dieser Thiergruppe. II Theil: Phylogenie. II Heft: Die Familien der Centetidae, Solenodon- tidae und Chrysochloridae. Bibl. Zoöl., Heft 49, pp. 1-158, pl. 1-4, and 108 text figures. Маттнеу, У. D. 1903. 'Тһе fauna of the Titanotherium beds at Pipestone Springs, Montana. Bull. Amer. Mus. Nat. Hist., 19, pp. 197-226, 19 text figures. 1906. Fossil Chrysochloridae in North America. Science, n.s., 24, pp. 786-788. 1907. A lower Miocene fauna from South Dakota. Bull Amer. Mus. Nat. Hist., 23, pp. 169-219, 26 text figures. 1909. Faunal lists of the Tertiary mammals of the West. Bull. U. 8. Geol. Surv., No. 361, pp. 91-138. 2 1910. On the skull of Apternodus and the skeleton of a new artiodactyl. Bull. Amer. Mus. Nat. Hist., 28, pp. 33-42, pl. 6, 5 text figures. 1913. А zalambdodont insectivore from the Basal Eocene. Bull. Amer. Mus. Nat. Hist., 33, pp. 307—314, pl. 60, 6 text figures. 1919. Recent discoveries of fossil vertebrates in the West Indies and their bearing on the origin of the Antillean fauna. Proc. Amer. Phil. Soc., 58, рр. 161-181. 1928. Xenotherium, an edentate. Jour. Mam., 9, No. 1, pp. 70-71. OsBORN, Н. Е. 1907. Evolution of mammalian molar teeth to and from the triangular type. Maemillan Co., New York, pp. 1-250, 215 text figures. Edited by W. K. Gregory. 1910. The age of mammals in Europe, Asia and North America. The Maemillan Co., New York, pp. 1-635, 220 text figures. SCHLAIKJER: A NEW ZALAMBDODONT INSECTIVORE 27 Scorr, W. B. 1905. Reports of the Princeton University Expedition to Patagonia, 1896-99. Princeton and Stuttgart, b, Paleontology, etc., Pt. 2, Insec- tivora, pp. 365-388. Sımpson, С. С. 1927. A North American Oligocene edentate. Ann. Carn. Mus., 17, No. 2, pp. 283-296, pl. 24. 1928. Affinities of the Mongolian Cretaceous insectivores. Amer. Mus. Novitates, No. 330, pp. 1-11, 1 text figure. ES “> ава AS алмас = —— EXPLANATION OF PLATE EXPLANATION OF PLATE APTERNODUS GREGORYI Schlaikjer Dorsal, lateral, and palatal view of skull. Two and one-quarter times natural size. Photographs by Albert Thomson. All figures based on type, M. C. Z. No. 17685. BULL. MUS. COMP. ZOOL. SCHLAIKJER. New ZALAMBDODONT INSECTIVORE. te, ВИЗ ЕВЕ ps _ ^ — | x js 4 Я у oe | я М 3 E | | = 1 _ A аад JAN 23 1935 5171 Bulletin ої the Museum ої Comparative Zoology AT HARVARD COLLEGE Vor. LX XVI, No. 2 CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY OF THE GOSHEN HOLE AREA, WYOMING II THE TORRINGTON MEMBER OF THE LANCE FORMATION AND A STUDY OF A NEW TRICERATOPS Ву Erich M. SCHLAIKJER Уттн Six PLATES | CAMBRIDGE, MASS., U. S. А. PRINTED FOR THE MUSEUM JANUARY, 1935 M | PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOÓLOGY AT HARVARD COLLEGE There have been published of the BuruETIN Vols. I to LXV, LXVII-LXXV, LXXVII, of the Мемотвв Vols. 1 to LIT, LIV. The BULLETIN and МЕмотвв are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collections and Exploration. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately. A price list of the publications of the Museum will be sent on application to the Director of the Museum of Comparative Zoology, Cambridge, Massachusetts. Bulletin of the Museum of Comparative Zoölogy AT HARVARD COLLEGE Vor. LXXVI, No. 2 CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY OF THE GOSHEN HOLE AREA, WYOMING II THE TORRINGTON MEMBER OF THE LANCE FORMATION AND A STUDY OF А NEW TRICERATOPS Ву Евісн М. SCHLAIKJER WirH Sıx PLATES CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM JANUARY, 1935 ———— No. 2.— Contributions to the Stratigraphy and Palaeontology of the Goshen Hole Area, Wyoming H The Torrington Member of the Lance Formation and a Study of a New Triceratops By Евісн М. SCHLAIKJER INTRODUCTION ‚ The present paper constitutes the second part of a series of contribu- tons by the author on thestratigraphy and palaeontology of the Goshen Hole area. This article includes, primarily, a study of the geology and fauna of the Cretaceous formations, with especial attention given to the description of a new continental member of the Lance formation and a study of a new species of T'riceratops. Goshen Hole presents approximately two hundred square miles of Cretaceous exposures which have been mapped by the author as part of his field research for the Museum of Comparative Zoólogy at Har- vard College during the field seasons of 1930-1933 (Fig. 1). Pre- Моця geological work in this area is limited to the reconnaissance by G. I. Adams (published in 1902) on the Patrick and Goshen Quad- Tangles which includes an eastern strip of the Goshen Hole area. dam's geological work, however, was not intended to be detailed and while his map of the geology gives the general distribution of the formations in the region concerned, it is not entirely accurate. I wish to acknowledge my indebtedness to the department of Vertebrate palaeontology at the American Museum of Natural His- tory, and particularly to Curator Barnum Brown through whose Ind offices I was granted permission to study the very splendid “eratopsian collection in that museum. To Dr. С. W. Gilmore I am Indebted not only for his valuable suggestions in the field, but also or the privilege of inspecting the Triceratops material in the National Museum, I desire to express my gratitude to Dr. J. B. Reeside, Jr., 9f the United States Geological Survey for his identification of the vertebrates herein described. Practically every member of the &cologica] faculty at Columbia University has offered valuable Criticisms of this paper, and I am especially grateful to Professors | | 82 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY Charles P. Berkey and William К. Gregory under whose immediate guidance this research has been executed. AA сет езі eae lr pori anne diee м | | ы | | | | ee й | | M | qe 9 je: > р ps a ( | | ps erts | | | E га ЈЕ P | | бок к t DETUR CIMA | nd N. ve ME | | | | eT ae | [чул жо К | | di P | | | | | L—-—-4 pues | | | | à т анаан |: ee Ета 1. Sketch map of Wyoming showing location of Goshen Hole area. The ceratopsian remains were most skillfully prepared and mounted for exhibition by Messrs. Charles Lang and Carl Sorensen. The drawings of the Triceratops were made by Mrs. Helen Ziska. DESCRIPTIVE GEOLOGY General Features The Cretaceous beds occupy a considerable portion of the interior lowlands of Goshen Hole. As indicated on the accompanying map (Plate 1), two main areas of exposure are present, one in the Corn Creek vicinity to the west and the other principally along Horse Creek. These two areas are separated by a strip of lower Oligocene. Several isolated outcrops of Cretaceous surrounded by Tertiary beds lie immediately to the north and to the east of the main exposures. SCHLAIKJER: A NEW TRICERATOPS 33 Two upper Cretaceous formations are represented in the area. On Corn Creek south of Red-Bill Point is a small exposure of dark gray to black carbonaceous, somewhat sandy shales. The age of these shales will be discussed at length in the succeeding pages. Overlying these shales with an unconformity are the typical yellow, fairly un- consolidated sands and shales, which are Lance in age. Because of the low rounded-hill topography of the Lance beds and because of the abundant alluvium no thick sections are measurable. The general lithology of the formation is that of yellow to greenish sands, yellow, red, green carbonaceous shales and lenticular masses of rather in- tricately eross-bedded yellow to brown channel sandstones. Occasion- ally large rounded sandstone concretions, with a diameter as great as two feet occur. On Horse Creek (Sec. 10, T. 22N., В, 61W.) the beds contain a two-foot lenticular layer of lignite coal (see Plate 2) which as been quarried by local people. The coal bed is overlaid by several leet of black shales containing many brackish water invertebrates. Many invertebrate fossils have been found along the banks of Horse Creek from one to three miles southeast of the lignite bed. In places Oyster beds assume a thickness as great as twenty feet. "These are lentieular, however, and grade laterally into sandstones and thin limestone layers. Near the town of Huntley occurs the same sequence of beds containing many brackish water invertebrates. Throughout the Lance, especially in the Horse Creek vicinity Halymenites! occurs abundantly. Reptilian remains are abundant in many localities and & very fine skull of a specialized ceratopsian was collected from the reddish shales on the east side of Horse Creek in the NEM, Sec. 4, - 22N., R. 61W. 'The presence of this specialized ceratopsian to- gether with the stratigraphie sequence of the beds and their relation to the remainder of the Lance has made it possible to determine an "pper Continental member of this formation which has heretofore been unrecorded. The Geological Map Most of the mapping of the Cretaceous beds in the Goshen Hole Area was done during the field season of 1931. Detailed mapping of © western-most exposures in the vicinity of Box Elder Creek and - B. O. Pass was deferred, however, until the summer of 1933 at Which time а topographie survey was made by the author and his ш. p Specimens collected are being studied by Dr. R. W. Brown of the United States Geologi- "rvey who is making an extensive study of this genus. 34 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY party of the western portion of the area. The geologic work was carried on along with this survey, hence enabling more accurate work to be accomplished. 'The author feels that the accompanying map (Plate 1) presents as accurate a record of the formations as can be made without the use, on a large scale, of unusual methods such as extensive excavating. 'The aecuracy with which the formations could be mapped was, of course, limited. First, while exposures are abundant, they are usually neither thick nor very extensive. Second, alluvium abounds and commonly is of eonsiderable thickness. Third, in many places the contact between the Cretaceous and the overlying Tertiary could not be determined exactly because of the presence of debris which has been transported down from the strata above. In most instances, however, close approximation was possible. In order to cope with these difficulties it was necessary to employ every possible method for determining the character of the subsurface rocks. Particular attention was paid to the type of soils, especially in newly plowed fields. Also, while the field work was carried out, many new county roads were graded, thus affording fresh exposures. Materials along irrigation canals proved very helpful, and frequently the walls of the canals would show splendid exposures. Wherever necessary, diggings from water wells and even cellars were inspected. No attempt was made to show on the map the separate members of the Lance formation or the alluvial deposits. To do this accurately would have been an endless procedure and perhaps impossible. Only the principal roads, the main irrigation canals and the major topo- graphic features are indicated on the map. A topographic map of the entire Goshen Hole area will appear later. STRATIGRAPHY All of the Cretaceous exposures in the area, with the exception of the limited outcrop of dark carbonaceous sandy shales in the N. W.14 Sec. 28, T. 22N., R. 63W. along Corn Creek, belong to the Lance forma- tion. The dark shales belong either to the Pierre formation, or repre- sent a level which is transitional between the Pierre and the Fox Hills. Pierre-Fox Hills. Aside from their lithologic character, the dark shales give no evidence of their affinity. The outcrop is very limited and no fossils were found. Additional information! has been obtained, 1 Through the courtesy of Mr. Dean A. McGee, Geologist for the Phillips Petroleum Com- pany. SCHLAIKJER: A NEW TRICERATOPS 85 however, by the drilling of three core holes in the Red-Bill Point Corn Creek vicinity. Only the driller's logs and not the cores have been studied by the author. He has, however, had the opportunity of inspecting the cuttings from five core holes drilled through the same Strata about 15 miles to the north. АП of this drilling was carried out for the Phillips Petroleum Company by the same driller with Whose terminology the author is perfectly familiar. The first hole in the Corn Creek vicinity was drilled about two hundred yards north- west of the west end of the outerop and the surface elevation is ap- proximately 4440 feet. The total depth of the hole was 1031 feet. The driller’s log is as follows: NE corner Sec. 29, T. 22N., Б. 6377. EMEN un РЕЧЕ Ла у а-ы = 17, co duc A AS ТА БАРАЛАР, 17— 38 ее do» ОШ ОООО ше ЕА Ба Во 52— 59 КОР eb л л о ort БУ "ОУ EE uM uus Ls eer c cr 62 65 EEUU enis ovn RI Hee err 64 0 ОО bale мо ННЯ о О ЫТ” р ПН ас жузу 72— 81 ин NA ӨШ ND EN Sor a uv eror а 88— 84 Gray РАДО НА био une en ОО 6844-2792 EEUU A eos ser b ree 92 — 102 Haese tias gonah ae А и, 102— 110 ERE eos een uet {10 11% ШО Bundy знае А О ет А О ner er АНА 112— 114 ШИ И а НА ПА 2. А 0150 ое 120 — 123 Бе 123 — 195 EN n e on se 125— 126 не 120 — 190 о, 130 — 133 ENDE ooo S rers este 19959 180 BE au eor DCN E CN EAE 136 — 138 Hard ШАШ doute ee ME E teers, Meee 138 — 138.1 ee ни ee 138.1 — 140 оные: 140 — 149 "ay sandy shale with black streak.............. 149 — 150 NEU oou уу. ен" 150 — 158 EUM iuuenes ат 54725. 153-- 159 36 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Gray: anndyiaheló. en nee eyed ed tota no 159 — OO RANI TS ен an es 160 — ОСУ darie ВРАЊЕ Залы c ii cord assis 168 — ОКУ ate SAO In. anis Rode Terri rie rusa CL Ig Lg = GIO BIO. A ea en ES EN 16 ~ ӨТ SHOIC 2 А ЛОВ AW ae el ОЕ 150 ИУ ШЕ зд и ан. las MST BO А РАМОК O 1807 - крону dere а САРА РА У УВА RED EINEN 189 — Gray sandy Saale А А dor dede x Do ВИ У in ИОА 190 аи к 205 — Guay М ZN красу datis В «a su C vsu qu ace) vv vp Vac red i глу 226 — Cray sandy "Па л de v. HF er dp EE en. 230 — Sticky. dark Bhale ^. Ud ар Y уы 231— СОНУ ТОШ АНЫЛУ е see WEE HOLEN UR UU 234 — OEL BIO BUT esu ecd ik en be Uni 7” 236 — SUL DIAS сы 243 — Сашойдо shale a... A а 247 — я v CLOS S TT IA Er VER 248 — Gaubonaccoisshslev na TRI US KC HW One FRAIS v CE QE 248.6 — о v л, л TR e e 251— (rav Bands O 22” 253 — СТО У Aae Ва а, iv TUE Cirta evade das ved 256 — ВОЉУ dano. у e у, м v е wile a MAI oe oit nn 260 — Cay SPEM ЛІ x o RES. 262 — ICV Gene RUC ie de M а ГАКА 264 — SUE рі Сто iius des er ce e rers 210 — СНУ A О er A ECCE ESQ etes drea 271— Sticky dark вое, Ми... a cu EDI PAIR 279 — (Om Siue sa c M ПАР DRESS. РО ИРА 280 — Istis dales e p IURE are 287 — ЭОК КОДАШ ем ah D DIO а асан 290 — o dea e 300 — ӨРДҮ ШАША, УУ ee a E od e va ti TR 304 — У ан АША ur. en 307.6 — GAS ен ro RUD I Dci 314 — SUC, о, Кы ле dr TS 319 — Guay ЕУ Te PEARL ил I TOUS LT Tg 321 — GUAY Bal аю лу; ЕРО О ОР ата S ака 324 — ОСИ ОЕШ Sqalote caw ta er ds PUN EVA 244 7 330 — Sticky dark shale and hard tan buttons................... 343 — Е тени НО 7 344 — IUOS. dana АБС LC. су secto пик Ra 348 — (баулы ы алы Ручни о 350 — 209.6 226 230 231 234 286 248 247 248 248.6 251 258 256 260 262 264 270 271 279 280 287 290 300 304 307.6 314 319 321 324 330 SCHLAIKJER: A NEW TRICERATOPS 97 Sticky dark shale—hard tan buttons..................... 858 — 354 | Sticky dark shale—hard tan buttons..................... 854-- 858 М НЕ ОО Bucs 358 — 359 о 359 — 361 | A NX 361— 362 | Sticky dark shale—hard tan buttons... а... а. 936—865 | nube dau tulo o o МІНА 365 — 368 | ET A ТУТ ОЕ Е. 368 — B72 1 sus АТЕНЕ asim ioi EE ato a hdd 88 | EI BH ES s E о ТЕ кул ‚.. 88. 384 | I SUE но a Л ie e 384— 885.6 1 Boobs dare shale у... eee cea 9856. 388 | ШО Cites АУ АШ Ө ууу АРА РЈ OEC 388 — 894 1 ТОУ Еу йау Sale Ое 394— 895 БА ае ыл Dive eu ER ES 395 — 399 Bando ауа АӨ Lok ro a Gass eoe ch OO SIS css 40516 ћ НЛО ВАН л... uU o aC пиона 405.6 — 407 nd les P ML о 407 — 414 Ed nde hes ел OS Т ICH eo d We 414 — 492 Fairly Sticky gray shale........ ЫН ELE IN TS RM 499 — 424 ав cec ЊЕ A | Gray shale....... Mb d LON Sap кй бу. у al le m | Sandy ту shale. N ae a ARD MR. р Bub А АРВ ИВА МАРИА ИН ДРИНИ РАНИ me esed dl | Sandy gray shale. . й з XN Wei. E ARI f | Sticky black shale- Jittle БОШ, ро Mod: | Sticky black shale. . E A л een ale, Sandy gray shale....... torneo nut Квас t obolis св AR | Gray sand with black доні a ВОЮ сомы CU К Stieky dark shale. (hil IAN пе a eA О ted шом О C DRE ODIT RUN брге] ВУКА dao d x MI I Л үу) О К EPIS S NE DD ESSO EE ааа 479 — 484 EN ra ees 484 — 494 аа наз е Й ER Solid gray sand. . A ow As MA Hard bluish gray lime. . а роо Bolid gray sand. PPE AUSTEN ra MIC cm 528 | "Ll еее е есе ее еке се есе 523 — 534 р С У ои | Sandy gray alte. Vb ANS a В | Е Т КЕ ПТ у г | END sides. sie м ese cask hoe ese alls тео: 800 | КО UA. A OS сене EE | По а ум oon 564 — 574 | | d | | | 38 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY BORD DPA. SEU do ESL О erc o ad) Hard анау шата ката e neueren a ra PIE A OOM RA СНО [un rwy cond Р ea MEME T ИОВ САВА е a ne (ШИ ЗАШО Ве и SUNG mbi php А еуен ку кий PH EU ОШУ ЛАШУ Во о, ри ен DE й САША РЕ a uer cs РАИ e ае Ute Л шу уйкулук go а е парка О Sale opis O o t RS SS Spray ИЕ S a ooo A БОШ БУ SEIL edis ру aget mcer E GR T CHOR E BUCO COS СЕН АРАС Ib бақсын e denied БИК ahale and қар tu i era БИШ ИД ШШ en Е ДОВШЕ БЕЛЕ ode iiec eI one Оа ЄВ о ep ны ca a Re ae I Dial ve deu но a Con Ud SH ОТАСЫ атала өшін но SAY UAW bis. Ve Suo datis sug ae E re e eo nator tn messa ud osi easi dui луч: у, Ил ЛДЫ o Pes ЫНЫ МАНА ШШ en: Su Є Зла E ЦАНА d ur dr ics wa rr ЕР Sandy ео: OTIC MN ай Eine lesen ceat СА EN RO NT MT. Sues не Ја Е-е ТКТ an КОТУ CUE ENG, op е ааа CN C EACUS pud опат аршаны. рыл т. Soft gray sand. Sticky dark hale tan button: aaa a З видань ы O ÓN БАП аи PRAL АИ o ede o IER UN E EI ne hue MN PETI. ьо БШ dak Hala. иное ои So cdi аа е АҒАМ оо ea dr ы er AA tio acude С а кл O RD GES ӨН Ө rut ye neck cad en UN Suc ПАШ rip nr nie ns) ac л у г; SCHLAIKJER: A NEW TRICERATOPS Sticky dark shale—tan Әийбопв.......................... DU УМОВ СБ ct s а АЙША s rh orien DR Tan button... . БН оа а RR OC CT SEMIS NEAN SIDES dale NULL Е Red E ре. Gray sand. 5 Black shale and oM de Gray sand. f o RU faba p Te ur WO ISI БЕШ ьо шд 0 Jug ees SV TP а Ре des о Suede o Me code EDS dar Bhdloest QUI. De ile qs Sticky dark shale. . Sticky dark зльоти eod. EOD UU арнады ш зу a Мика Ы ЛЫ SUR C - БЕТТЕ тј рај на a T NEN oo oobis шш Extra hard tan streak . Sticky dark shale. ten Patton. un re Pe нн Gray sand. Hard broken: gray y sandy lime. . Gray sand DUI OUS лы шо АН НОЮ Sandy gray shale. a ЗЫ TUO LSU RR S Sticky dark shale. Sticky dark ТЕМЕН tan Brilon. сайтын aub. qur dedi Sticky dark shale- a ен њи Sandy g отау shale. .... Sticky dark gh: ale and Boel. Sticky dark shale—tan Buttons. PUN ш». Bic PN Le Sticky dark shale. Sandy gray shale. ard gray lime. Sandy gray shale. a Hard gray lime. . a gray shale. . ticky dark shale—hard bu ре 99 0101 — 198 2198—1502 802 — 802.2 802.2 — 808 808 — 809 809 3809.2 800:2 = Sit 811 — 811.4 811.4 — 813 813 — 815 815 — 820 820— 821 821— 831 831— 835 835 — 854 854 — 855 855 — 861 SOL 66019 861.8 — 872 9/27 872.1 872.1— 882.4 882.4 — 883 883 — .890 890 — 896.6 896.6 — 897 897 OTI ОЮ ОД 914 — 925 925 — 932.4 932.4 — 933 933 — 951 951— 958 955 — 901 961 — 968 958 =: 979 919: 985 988 — 1001 1001 — 1002.6 1299100216 — 10038 ..1003.8 — 1006 1006 — 1018 1018 — 1031 сы analysis of this section shows that the first twenty-one feet ег the surface material are unquestionably Lance and can be ——ÁÁ—á— | 40 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY traced as a continuation of the same beds in which were found ceratop- sian bones about a mile to the southeast. Approximately one-fifth of the entire section is sand. Sandy shales and sticky black shales abound and some coal is scattered throughout the entire section. There seems to be little evidence of distinct lithologic breaks other than the pres- ence of marked sand-bearing zones. The most pronounced of these zones is the one at a depth of 407 to 585 feet. Of this 178 feet the last 100 feet is almost entirely sands and sandstones. From this depth downward the shales become less and less sandy, and in the last 120 feet there are no sands and sandstones present. Another core hole was drilled about four miles northeast of the first. The surface elevation is approximately 4320 feet and the total depth of the hole is 602 feet. The following is the driller’s log: Center of So. Line, Sec. 1, T. 22N., В. 68W. sec M Cc cuu 224400001 0—2 KE па ДАР Ju. ал o LU О РОА лк 2— 19 Pe a a АТ A лије 19 — 28 Нага reddisnevellow broken die: оо iS 988 a o ню lo RI AG ООУ вро вро поро ИО BO. О, міти не HC 459 54 80 SEO Rand T ША ME у л зс 59 — 71 а Mucky mo yellow sandy аны. 215) 76 — 79 iue T urs A то 79 — 82 Dark shale—some hard tan thin streaks.................... 82 — 90 and danh ан ek wae М, 90 — 91 Бер ВАО (s cie iiit M Li e 91 — 118 BU or US у ne 113 — 145 Malte AS RDAS лы LAUR ae Sticky gray shale—few thin hard tan streaks................ 150 — 178 Samar or Mee a aa НО ТТ oe Buela T E АРЕАЛ ое 182 — 190 de у... 22 000 190 — 195 Go А uot ЛАЛА А S M ua. M 195 — 195.1 0 и 200 — 206 Jic ВАМ а ул... 206 — 215 Sandy gray shale—much coal in thin РЖ анима ut adt: а PS LAE in 227 — 236 Bane sandy gras shales ob s dde ні а CAES 286 — 243 SCHLAIKJER: A NEW TRICERATOPS 4 Sticky gray shale—some hard tan streaks................... 243 — 268 DD MM NELLE oe cu I у... B NE A e eo a NS EO NL ED С 272 — 280 Sticky gray shale—little ооы...2..................2..2...:. 980— 290 Billy отау shalo—little coal. у... cessere s 290 == 300 SOS Cea DIEA LAC ИЕ, RE AMI рее ads A Р БИ АУ finis o eus cones m rli omo Cn з OL S anna Бр Las OD ТР nce o coke PO ти NN ESO UM Bu useless ibd Sandy gray shale—some hard tan streaks..............:.... 9842 — 347 sidor ae ee ован ери abs 347 — 358.9 (Cog Му ТҮН a nn ГОР O uc ООН 222.22 09810 Pulls Sale M mus оку зы LI IM S HA Extra hard dark chert-like Ште............................4197 — 420.9 Sticky gray shale—some hard tan streaks...................420.9 — 458 Vo. s. d rc ME ERUNT erm т" vista c M NM EU MM CUR ар р EMEN 5р Sticky gray shale—some thin hard tan streaks............... 506 — 540 И Ае, уу... у: EN URS 540 — 548 DEDOS ose К у у ллу Pec RA ES UD QUE. у умы ш реу 2-46 Sticky gray shale—many hard tan streaks............... 550 — 582 р. о оз Sticky gray Shale-—hard tan straks: i: rarau. ans en. 582.8 S 502 Е А PC о одна р 009 OE Sticky gray shale—hard tan streaks. ........................600---600 The first 79 feet of this hole is typical Lanee. The rest of the log Shows remarkably few sand and sandstone layers. The first layer of sand occurs at a depth of 206-215 feet. The only satisfactory corre- lation which can be made with the first core hole 13 that this sand layer is the equivalent of the 18 feet of soft gray sand at a depth of 92-110 feet. Such a correlatión is in keeping with the general litholo- gie similarity of the remainder of the hole with that of the first. The total depth of the hole was only 602 feet, hence the thick sand and sandstone layer present in the first core hole, on the basis of this Correlation, was not reached. The following is the log of the third core hole which is located just to the south of Red-Bill Point. Drilling was started at the surface elevation of 4380 feet. NW, Sec. 2, T. 22N., В. 63W. Surface. Sticky ріні and. yellow Tae. ie PELE etapa Bioken me Qus Lo y CLOW е ee we oir aar r др EE ЛЫГ ТЛ Yellow sand. Sticky yellow shale. Sticky yellow shale—gr preenish- blue streaks. Sticky yellow shale. . Yellow sand. Sticky yellow ad gray itle. NER MN I e ur C. io Ea E Ec ЖОШ рик у оо le OU M RIA A M ea. ПОКЕР Cine duet cu НИ Я lu a RU ok ПН ее eani Р ОИ УТУ ш: е A X d ан не ОИДИ bino Sand “BPAY BHAI a. nn МЕСТЕ ИМЫЛ ОР АША sone well a Mo ce р БРА Re е LAU (Js о T Less ұғ па о edere 6 Ай о Lung УОШУ ВИЦ о TUO РИМОМ РУМЕ Sandy. gray Slo I.e мона PATET a К ЗАПОР e ia чег. улуш ed PEU vecta n Е . 148.8 — 150 Sandy gray shale..... Reddish yellow coarse bond. pne MEE КЕШЕ yellow. Coarse sande. as co Are I LORS UAE cup И e Cake) co e due S р Е ВВ gray ЗОНЫ 6 coal «veo e ONE eg ЕСЕ brown sandy BOALO min en о Uc ОНУ Sale os «a coco boa р но ATO E a Pe Пош Gray lime. Medium coarse gray m white ging: ЖЕУ АВИИ ИРИНА SUM Sticky gray shale—thin coal streaks........................ Gray. sind vsus (preda ae сырға eo Cos D vb о OR rs Fond Кере (банака БЕДЕ A на d p ae ЛИ ARE EST Ie Ук А Soft gray sand. Gray shale. : Gray shale- АРЕН, ires i. Fine light gray нон Gray sand. Fine sandy gray вањ: = one . на сс нш ОЕ етти 22112 Шаш ал БОНУ «ШО A ин BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY 012 12 == 10 ор еі 2172-5486 36 — 48 48 — 60 60 — 64 64 — 67 Oron (dc үз, vO 10 TO= 79-- 79:6 79.6 — 80 90891 ASS бок 91110 110 — 140 140 — 143 143 — 148.8 150-1607 167175 vor б 170—151 ISI IS 182 — 185 180 РА МО ‚ОИ дана 201 231 — 246 246 — 250 250 — 251 251 — 252 252 — 253 253 — 255 255 — 262 262 — 265 265 — 285 285 — 288 288 — 295 295 — 296 SCHLAIKJER: A NEW TRICERATOPS 43 Bine gandy алы ы сала се ата пене 299015 208 он voir лыы do 298 — 302 Mostly coal. . dic qus Ну ПРИС КРИВОНОСА Р +809 Fine sandy gray АЕ Не ОО а ieu кы м o dA 1009-52 Hard white quartz-like lime. x PLI UI ud 321 — 321.5 Fine sandy gray shale—some duh рае d ‘coal. S et e а зді. 5-- 390 Fine sandy gray shale. RON UM узу Зо DU S Bite лар ату Вы! 2o, obe. ei да lee, поша 400 — 418 Gray sand. а Fine sandy gray shale. . RE задо ФА ИН МИ Ss ИО 48816; Cherty bluish-gray шде: ..488.6 — 488.9 Extra hard chert-like Dluish-gray w vith black et eas of lime.. . 488.9 — 490.4 Hard tan streak. . AO шо е Boft gray sand with many black Speck... ына N гүр Gray sand. A Ages eor CH HH eod vello o RS Л 522 — 545.6 Extra hard rough tan збгеяКв............................. 545.6 — 546 Sandy gray shale—little соа].............................. 646 — 550 Sandy gray shale—tittle’coal. ET o E 21200" 8850602 pandy gray shale—dfittle alii cere. dors ne 1, 602 45612 со my dll PUO TOC Cb p CHER ST den a Olea, 6195. ee iuit Мб O is gaia (ais c do дб Gray sand. . Doe pde to cvs УРА ОБ Fairly sticky | gray У als e tan buttons, eta, ТЕҚ Macs VER ou 615 — 622 Sticky gray shale. . Жагы eR Аса або O. Sticky gri ay saxo broken Tán En енин 0299 —= 080 ку gray shale, eae sae see tate 630 — 640 Gray shale... бе Hard chert-like bluish є gray і Lcid tee. Mn Оло 009 Sticky gray shale—some hard tan buttons..:............... 643.4 — 653 ОИ did: «oC we роду, merces. han aee. A868 12965 REN A etos. o eodd. solide ендігі». 2655 O87 ES uni d our hoo ers eod жини. бл боз E o os у... улул... OBB ORB Тап buttons, E ОО Du EUN E eius 60863000 ПИ US I BET з o Tan buttons. Eu dE e си IS е 66816630 ушкш»... са caus an (d hr dr d LOAD 000 Soft gray sand. я o hs sone i dad Noll ep оо Б. ыы. У уз м а а те а р 678 — 686 44 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Since the remains of Triceratops and Trachodon were found a mile to the east in beds which can be traced as the equivalent of the upper thirty feet of this section and since all of the first 180 feet seem to be a lithologie unit, it seems logical to refer to this as unquestionably Lance. At depths 212 — 231 feet occurs а medium coarse white sand. This sand was not encountered in either of the other core holes. The only satisfactory correlation which ean be made of the remainder of the hole with that of the first on Corn Creek is that the 32-foot layer of gray sand at depth 490 feet is the possible equivalent of the first sand layer at the 92-foot depth in hole number one. The data presented in the three core holes gives evidence of a con- siderable anticlinal structure in the Corn Creek vicinity. In hole number two the strata are structurally 234 feet lower than in one, indicating a gentle dip to the northeast. This is in keeping with sur- face conditions which show an anticlinal structure plunging to the northeast. Under a special topie, the structural conditions of the region will be considered more fully. In hole number three, the strata are structurally at least 458 feet lower than they are in one, which is a little less than four miles away. This, of course, does not represent the maximum dip since, in relation to hole number one, number three was not located at right angles to the strike of the fold. Lance. Unconformably overlying the dark carbonaceous sandy shales, оп Corn Creek are from: seven to ten feet of darkish yellow sandy shales. Immediately above this are twenty to thirty feet of reddish yellow coarse sands more or less consolidated. In these sands, fragmentary remains of large ceratopsians were found. The succeed- ing one hundred and fifty feet is composed primarily of alternating beds of yellow sands and sandstones, more or less intricately cross- bedded, and variegated shales. Bones of Triceratops and Trachodon are scattered through this series and are especially abundant near the top. SCHLAIKJER: A NEW TRICERATOPS 45 Along Horse Creek good exposures are numerous although no thick Sections are measurable. Local correlation is possible, however, and а fairly thick series, 150 to 200 feet, is determinable. In the northwest corner of the SW, бес. 14, T. 22N., R. 61W., on the south side of Horse Creek the following section is characteristic: Feet above Feet | Horse Creek Alluvium Gray finely laminated вапабвіопе..................... 1 73% Concretionary yellow gray limestone................... М 721% o ый АСТАН a pc ee ve beo 72 и д 69 Iron-stained limestone, many brackish water fossils...... 1% 67 Yellowish gray limestone, large Реесуройв............. М 6615 КИРА ОО и 66 ау Massive SAM debole: е co Sato 64 ote ce 64 Gray concretionary sandstone.... : 1 62 Gray massive sandstone Oaren ....................... 2 61 Gray finely-bedded sandstone Озігба................... 1 59 Yellow, purple atu Тед сва sales о о S va duod tt 2 58 Gray Mae БОЕ eos Т 56 Gray, purple, red and yellow shale. .................... 4 55 Finely bedded АО, seele 51 Follow «id purple shales. ала e a 4 50 Finely bedded gray sandstone ЫЫ ы ушн аЛ» 2 46 Massive gray.sandstone. и бз је. ый ee 15 44 Finely bedded dark ATA BANGS LOMO, sacas dad Жалы nte 43% ИТ parole and vellow ана а... 41 QEON Das АОН, ен. c 34 Finely laminated ТАУ ВАО ОПО Т ов TIE D й 26 TS Еу напамово c. sse SA 2 25 Yellow, DULDIG NOU RAMOS? ETAT Cia ces si MO Concealed by water of Horse Creek One of the most outstanding features of this section is the presence of the brackish water fossils, Corbicula, Ostrea, ete. These fossil-bearing Strata are found about а mile to the east in WY, Sec. 13, T. 22N., В. 61W. On the south side of the Creek the Ostrea beds are fifteen to twenty feet in thickness. "They are lenticular and thin out into gray Sands and thin yellowish limestone, which contain Halymenites. Four feet of finely bedded reddish yellow and gray sandy shales overlie the 46 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY oyster beds. About a mile and a half northeast of the town of Huntley, NW, Sec. 19, Т. 23N., К. 60W., the oyster bed is about three feet in thickness and contains considerable gray and yellow shale and sandy material. Above the oyster layer are seven feet of finely bedded reddish yellow and gray sandy shales which are succeeded by a two- foot layer of finely bedded yellow sandstone. This sandstone contains at the bottom a four-inch layer of brackish water fossils. One mile northwest, SW14, Sec. 10, Т. 22N., В. 63W., of the lo- cality of the first section given above, the following section was measured: Feet above Feet | Horse Creek * Alluvium Finely bedded red, yellow and gray sand shale with thin sandstone layers. . e Ecke e КТО ЛМЕ НА Т ELO 41- Yellow sandstone. . zs 2- 5 26- Greenish-yellow тп wil ЖЕКЕН leña and. Барон ing shaly at bottom. . a 15 21 Dark carbonaceous ahale e many оре + У n о о пд УЕ 2 6 Lignite.. ЕТСЕ UM 2 4 Darks sbalok: з ; DIEN AND. i 2+ Concealed by water об Hotte ніс. The Corbicula zone was not found in this locality. A slab ої sandstone containing a dozen or more specimens belonging to this genus was found on the surface, however, indicating perhaps that it had come down from the overlying beds which are covered with alluvium. If such is the case, the lower twenty-five feet of this section, or perhaps all of it, is stratigraphically below the beds exposed a mile to the southeast. The beds here dip 4° northwestward for about a hundred yards where they are interrupted by a minor normal fault, which has a southeast dip. There is arching of the strata to the southeast so that within a mile they are dipping gradually to the southeast. North- westward there is a gradual dip away from the fault, so that the beds exposed here are beneath the surface a mile and a half up Horse Creek, NEM, Sec. 4, Т. 22N., К. 61W., where the following section was studied. SCHLAIKJER: A NEW TRICERATOPS 47 Feet above Feet | Horse Creek Alluvium Red, yellow and green carbonaceous shales with lenses of ВИДОВА Ма ВОН ye ® о ее ов 19 56 Yellowsandstone grading laterally into variegated shales MICH ООО A Л Т ЫЛЕ Иә Green, yellow and gray sands grading laterally into varie- gated shales containing sandstone lenses and many iron concretions. Abundant remains of large Tricera- орша Паво ОТПАД uta Mic dci dado 36 Red, green and yellow finely bedded shales and sand- Stones. T'rachodon, Triceratops, Crocodilian, and turtle ТЫШЫН 2 кы а dr ин а ладан ОЇ 24 O праве ы nd 14 Greenish-yellow and gray валав.................... 5 12 аери STONE, А name oa digas y Greenish-yellow and gray sands.................... 5+ Concealed by water of Horse Creek The dinosaur remains in these strata are of primary importance. In the lower part of the uppermost twenty feet of this section the skull and some skeletal parts of a new species of Triceratops were collected. This individual, to be described in detail under a special topic in this paper, represents the last recorded survivor of that great group of dinosaurs. The presence of this specialized Triceratops in addition to the presence of the brackish water fauna which probably 1$ stratigraphically above the Triceratops beds of the Lance in other areas, makes it necessary to give a new name to the continental de- posits above the Corbicula-Ostrea zone. I suggest that it be known as the Torrington member of the Lance Formation,— so named because of its proximity to the established and prosperous town of Torring- ton, Wyoming, which is located on the North Platte river north of the typical exposures on Horse Creek. About a mile northeast of Fox Creek Сар, SW14, Sec. 1, T. 20N., R. 63W., the following section was measured. Feet Yellow and Ову Cross-bedded Sandstone. ana sta. НД ed, purple and yellow carbonaceous shales Trachodon and Tricera- а Bellow eross-bedded sandstone. д чу, АЛ у... Yellow sands.......... [ur слона нь мо не нь 00 Town sandstone with cannon-ball concretions 2’ in diameter ...... Yellow and purple sands... . A > Yellow and gray platey sandstone. Pow Ua а А ур M Yellow sands. ..... Concealed 48 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY The eighteen feet of red, purple and yellow carbonaceous shales con- tain fragmentary remains of Triceratops and Trachodon. The close similarity both in lithology and in the fossils between these shales and those on Horse Creek seem to indicate that the Torrington mem- ber is represented in the Fox Creek Gap vicinity. This conclusion is expressed tentatively in Fig. 2. A mile northwest of Table Mountain, Sec. 17, T. 22N., В. 60W., there is an exposure of yellow and brown sands with large elongate and cannonball concretions of sandstone very similar to those near Fox Creek Gap. Above are five feet of purple shale with many thin limonitie layers, one foot of gray sand and sandstone and finally, five feet of yellow-gray lenticular channel sandstone. Above this sandstone are the typical lower Oligocene clays of the Chadron forma- tion containing titanothere remains (See fig. 2 and plate 3). Correlation. While long range correlations, based on lithology alone, are inadvisable, it is the only method which can be used to de- termine the age of the dark carbonaceous sandy shale in Goshen Hole. In this particular instance, however, this method is unusually re- liable. Considering first the possibility of the entire series being Lance some serious objections are immediately encountered. First, the Lance is typically made up of yellow, brown, tan and reddish sands, sandstones and shales, and while dark members are frequent they are always relatively thin. These dark sands and shales are over 1000 feet thick as recorded in one of the core holes. Nowhere in the Dakotas, Montana, Colorado, or Wyoming has a single dark member of this magnitude been recorded as belonging to the Lance. Furthermore, if these dark shales and sands belong to the Lance, they ought to be underlain by the typical buff and brown sandstones and shales or Triceratops member. Beneath this member should be the typical white and brown members of the Fox Hills formation. On Rawhide Creek about eighteen miles north of Corn Creek a deep test well was drilled, having a total depth of 5628 feet, in Sec. 24, T. 26N., R. 62W. and no such sequence was encountered. In this well, of which only the driller's log is available, the Lance is typically repre- sented by about 440 feet, beneath which is at least 200 feet of typical Fox Hills which in turn grades into the Pierre below. c + SCHLAIKJER: A NEW TRICERATOPS Е 'Suore[erroo рив AZojoyg Zurmoys Águia 9910, SUL 2242402 Оо) |, Alsoo) |-- 902 eno, б 7177209170! MOSYNZZLLI ЖС MORMAMEZIGI 229 MIIAHNEZLEI PS ос отмо?) | X М 6861 TOETS 9SIOH 99} ur әәпет 993 jo Suorjoos лечит) 7 “DIA шен IED == «иг руд юр | ES / ==; ии СУД | e С) toy porua | LOY ION І / / MESINZILO PE ри | e70207224/ ELO! SIYZ ЗАРИ СОМ = | ИРИМ с MEPS NOZL 25} 50 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY 'That the.dark shales and sands in Goshen Hole are Fox Hills in age is very questionable. Certainly they are not lithologically the same as the upper Colgate white sandstone and the lower brown members, which characterize the Fox Hills in practically every lo- cality, wherever seen in the field by the author, and as set forth by Meek and Hayden 1861, Cross 1889, Hatcher 1893, Knowlton 1909, Stanton 1909, 1910, Lloyd 1914, Lloyd and Hares 1915, Winchester, etc. 1916, Thom and Dobbins 1924, and Dobbins and Reeside 1929. It is generally agreed, however, that the Fox Hills and Pierre are transitional (Stanton 1910, Ward 1924, Dobbins and Reeside 1929, and others). That the series in Goshen Hole represents a Fox Hills- Pierre gradational sequence, similar to that which occurs in the Salt Creek oil field and in the Poison Spider district west of Casper, is not improbable. Although most of the sequence, especially the lower part, seems to favor closer affinity with the upper Pierre. Sandy and carbonaceous phases of the upper Pierre are common in western and central Montana, central Wyoming and in eastern Wyoming in the Lance Creek area. A similar lithology is also recorded for the upper Pierre in the log of the deep test well, referred to above, on Rawhide Creek, eighteen miles north of Corn Creek in Goshen Hole. With the data at hand it seems that the most logical correlation, therefore, is to refer the dark carbonaceous sandy shales along Corn Creek to the upper Pierre, possibly to include part, at least, of the Pierre-Fox Hills transitional phase. The yellow, brown, red and greenish-gray sands, sandstones and shales in Goshen Hole may be unequivocally referred to the Lance. Ceratopsian remains occur from a few feet above the Pierre through- out the series in the Corn Creek vicinity. The exact relationship be- tween this series and that in the Horse Creek vicinity cannot be ac- curately determined. The Lance continues under the coal bearing beds on Horse Creek as is indicated in shallow water wells. Since the brackish water beds are not present in the Corn Creek vicinity per- haps they are stratigraphically higher. In this event, the total thick- ness of the Lance would be about 375 feet or more. On the other hand, the outcrops on Horse Creek may represent the westernmost extent of the brackish water deposits which may really be synchronous in disposition with some part of the series at Corn Creek. This relation- ship can be estimated only by subsurface work on Horse Creek to determine the underground thickness of the Lance. The normal thickness for the Lance in other areas is 700 feet or more. Its thinness in Goshen Hole may be explained by the fact that SCHLAIKJER: A NEW TRICERATOPS oL local uplift after Fox Hills times, as is indicated by the absence at Corn Creek of the typical members of that formation, developed the агеа into a structural high so that no thick accumulation of Lance could follow. Since the uppermost Lance is represented in the area by the Torrington member, the thinness of the formation cannot have been caused by surface erosion. The general sequence, then, of the Lance is a lower series of continental deposits 100 to 200 feet in thickness which contain ceratopsian and other dinosaur remains; 80 to 125 feet of brackish-water deposits (see palaeontology section of this paper for complete list of invertebrate fossils); and an uppermost continental series 60 to 100 feet in thickness which the author has named the Torrington member and which contains an advanced form of Triceratops. The lower part of this sequence certainly may be correlated with the ceratopsian beds of the Lance formation in other areas such as Lance Creek, Wyoming, the Dakotas, eastern Montana, etcetera. While the overlying brackish-water beds were being deposited, marine con- ditions must have existed to the east or northeast. The only known record of such conditions is the presence in western North and South Dakota of the Cannonball Marine member of the Lance (Lloyd 1914, Lloyd and Hares 1915). This marine member is above the Triceratops beds: and thins rapidly toward the west grading into the Ludlow lignitic member (Lloyd and Hares 1915, p. 539), and near Yule in Billings County, North Dakota, just east of the Montana State line, oyster beds occur which are considered by Stanton (1920, р. 9) to represent a deposit in “a brackish-water estuary or inlet at the western margin of the Cannonball sea.” On the basis of the avail- able information, a tentative correlation of the brackish-water de- Posits in Goshen Hole with some phase of the Cannonball seems allowable, ] No ceratopsians were found above the Cannonball member. Cal- vert (1912, p. 471), however, states that in eastern Montana ceratop- Stan bones were found in beds which are supposed to be stratigraphi- cally higher than the Cannonball. Further detailed investigation of the upper Lance in this area, as well as in others, may bring to light evidence of the occurrence of a series of beds comparable to the Tor- проп, with advanced forms of Triceratops, showing that that member of the Lance is of considerable geographie extent. That the Torrington member is equivalent to at least part of the Arapahoe-Denver beds in the Denver Basin is indicated by the close affinity of the ceratopsians found in both. This is additional evidence ог considering the Torrington as an upper-most member of the Lance. D2 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY STRUCTURE Since the structural geology of Goshen Hole will be considered at length in the last of this series of contributions, the present paper needs only to include a treatment of the subject insofar as it directly concerns the development of the Cretaceous. The structural conditions illustrated in Goshen Hole are doming, arching, folding, fracturing and minor faulting with major faulting to the northwest. On the Laramie River, a few miles west of the town of old Fort Laramie there is present a major normal fault with a strike E. 20°N. and dipping 6195., accompanied by much minor faulting (found by the author and later to be described in detail in this series). Paralleling the strike of this major zone of fracture is а vast system of minor faults and fissures throughout the Goshen Hole area. The major structural feature is a broad up-warping or doming of the whole region. This condition is made particularly evident by the attitude of the strata exposed along the escarpment which almost completely encircles the region. Everywhere along this prominent topographie feature the beds dip, with a magnitude of from one to three degrees, away from the interior lowlands. The most important structural feature in Goshen Hole is an anti- cline which lies in the vicinity of Corn Creek between Y.B.O. Pass and the town of Yoder. I wish to designate this structure as the Corn Creek Anticline. It has a northeast-southwest strike and the length of its exposed surface is approximately ten miles. At Y.B.O. Pass it plunges under the escarpment with an inclination of from one to three degrees. Rising to the northeast, it appears to assume its maximum height and its maximum northwest-southeast arching just to the south of Red-Bill Point оп Corn Creek, and then plunges to the northeast under the Tertiary about two miles west of the town of Yoder. It may be estimated from the discussion of the stratigraphy above that there is a dip to the northwest of at least four hundred and fifty feet in less than four miles from Corn Creek to the Red-Dill Point vicinity. This dip continues fairly uniformly from Red-Bill Point northwestward for several miles and then becomes increasingly greater, nearer the zone of the large fault west of old Fort Laramie where a seven to eight degree dip is very apparent in the Tertiary strata. Southeastward from Corn Creek the dip is more gentle,— about one degree. If the Trachodon beds near Fox Creek Gap can be correlated with those along Horse Creek, then the Torrington mem- SCHLAIKJER: A NEW TRICERATOPS 58 ber is present in that locality, which may mean the presence there of the full thickness of the Lance as represented in Goshen Hole. There is the possibility, however, that these beds, in spite of the similarity to the Torrington member both in lithology and in the presence of Trachodon fragments, may represent a level in the typical ceratopsian member of the Lance in which, in areas outside the one under con- sideration, Trachodon remains occurs abundantly. In such an event the Lance would be thin, and since the surface elevation is two hundred and eighty feet above the elevation of the Pierre-Fox Hills on Corn Creek this would indicate an arching in the Fox Creek Gap vicinity. Moreover, if the Trachodon beds belong to the Torrington there is also evidence of structure, for subtracting the full thickness of the Lance (375) from the surface elevation the top of the Pierre-Fox Hills would be 150 to 200 feet higher than at the Red-Bill locality. The structure would not, however, be as pronounced as that at Corn Creek. Whether or not the typical Fox Hills is present can be de- termined only by subsurface work. In addition, the escarpment be- tween Big Lone Tree Creek and Fox Creek Gap presents a southwest dip of at least one degree, and Tertiary sediments are present at a much lower elevation between this vicinity and that of Corn Creek. This arching includes a surface area southeastward as far as the Hawk Springs Reservoir and northeastward almost to the town of Huntley. Because of the lack of good exposures and because of the unconsoli- dated nature of the sediments, accurate measurement of dips is for the most part impossible. Local faulting, mentioned above, with measurable dips as great as four degrees are very apparent on Horse Creek several hundred feet north of the coal mine. PALAEONTOLOGY Invertebrates I am indebted to Dr. John B. Reeside, Jr., of the United States Geological Survey, for his identification of the invertebrate collections from the Lance. His list of species and the localities from which they were collected 1s as follows: N. W.14, Sec. 19, T. 23N., В. 61W. Ostrea glabra Meek and Hayden Anomia micronema Meek Corbicula cardiniaeformis White Corbicula planumbona Meek Corbicula sp. 54 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY SW М, Sec. 13, T. 22N., В. 61W. Ostrea glabra Meek and Hayden Melania wyomingensis (Meek) SW М, Sec. 10, T. 22N., R. 61W. Membranipora sp. unnamed Anomia ? sp. Corbula subtrigonalis Meek and Hayden Small Planorbis-like gastropods, undetermined Oyprid ostracods, undetermined NEM, Sec. 15, T. 22N., В. 61W. Ostrea glabra Meek and Hayden Anomia micronema Meek Corbicula planumbona Meek Corbicula augheyi White Corbicula sp. undetermined, internal molds only In commenting on this fauna Dr. Reeside! says, “ АП of the lots represent a brackish-water fauna. There are no strictly fresh-water shells and no typically marine shells. All of the species occur in the Laramie formation of the Denver Basin, and most of them have been recorded from the lower Lance formation. By themselves they do not indicate an extension of the Cannonball sea into southeastern Wyo- ming, though they do indicate the presence somewhere near by of truly marine waters.” To recapitulate what I have stated previously in this paper, the brackish-water deposit which indicates near by marine conditions is above beds containing Triceratops and Trachodon re- mains. The only other recorded marine-brackish water deposits of similar stratigraphic position in the Lance is the Cannonball marine member. Perhaps the brackish-water deposit in Goshen Hole repre- sents a different phase of the Cannonball, or a tongue of that sea which is somewhat earlier than the one represented in North and South Dakota. However, that the two are nearly equivalent in time as stated previously, is strongly suggested by the high stratigraphic position of the conformably overlying continental beds containing the very advanced form of Triceratops. Only from additional evidence procured by extensive field work in the Lance throughout the Great Plains area with particular attention given to its upper members can this problem of correlation be solved. 1 Communication by letter. SCHLAIKJER: A NEW TRICERATOPS 55 VERTEBRATES Vertebrate remains are abundant. Isolated vertebra and scutes of a small crocodilian were collected, which, according to Dr. C. C. Mook!, probably belong to the genus Leidyosuchus. Fragments of large Chelonians are not of unusual occurrence, and Trachodon bones are numerous. None of the discovered remains, however, were suffi- ciently complete to permit specific determinations. Ceratopsians abound. The skull and jaws and some skeletal parts of a new Tricera- tops were collected. The following is a study of this rare and important specimen. Class Reptilia Subclass Diapsida Superorder Archosauria Order Orinthischia Suborder Ceratopsia Family Ceratopsidae TRICERATOPS EURYCEPHALUS ? sp. nov. Type. M. C. Z. No. 1102. Nearly complete skull and jaws and parts of skeleton. Discovered 1930 by Е. М. Schlaikjer and collected 1n 1931. Horizon and locality. Upper Cretaceous. Torrington member of the Lance formation, Goshen County, Wyoming. Specific Characters. 1. Crest greatly expanded in proportion to the length of the skull. 2. Facial region abbreviated and broadly triangu- lar in outline when seen from above. 3. Orbit elevated so that almost two-thirds of its area is above the posterior of the nasal. 4. Antero- inferior corner of the squamosal extended forward, giving a straight anterior border to the squamosal. 5. Dentary short and proportion- ately deep with high coronoid which distally is little antero-posteriorly expanded. 6. Brow-horns proportionately very long and relatively slender. 7. Nasal horn greatly diminished. 8. Olfactory nerves separate and diverge laterally immediately in front of the cerebrum. (See plates IV and V.) Description. None of the rostrum is preserved. The presence of a complete nasal and jaws has permitted, however, a fairly accurate restoration. As is indicated by the facial proportions and by the width ! Verbal communication. ? From the Greek, euros, wide, and cephalos, head. 56 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY of the predentary, it must have been wide and massive. The right nasal is missing but the left is completely preserved. It is slightly vascular, especially along the posterior of the narial margin. The dorsal surface is gently concave antero-posteriorly. Posteriorly the Fie. 3. Triceratops eurycephalus. Lateral view of the skull. One-four- teenth natural size. nasal is broad and anteriorly it converges rather rapidly. The nasal horn base is represented by a small, low, vascular, oval-shaped, rugose prominence which faces obliquely upward. It is more probable that in life this prominence bore a pad-like protuberance, rather than a true horn. Only a fragment of the left maxillary was found. It pre- sents deep alveoli, and externally a rather prominent facial ridge is present. 'lhe orbit is elliptical in outline and is inclined with the in- ferior border slightly forward. It occupies an elevated position in the skull so that almost two-thirds of its area is above the level of the posterior of the nasal. The jugal (see Plate 5) is proportionately fairly short and broad. Its position is nearly vertical as seen from the side view, but distally SCHLAIKJER: А NEW TRICERATOPS of. it is sharply deflected out away from the skull. A low keel is present externally on the lower one-third and is located about two-thirds of the way back. The sides are nearly parallel,— tapering only graduall y and about three-fourths of the way down they begin to converge rapidly and come to a blunt point distally. No epijugal is preserved but one probably was present. The distal ends of the postorbital horn corés are missing but the left is more complete than the right. Proximally they are oval in cross-section, — being much broader behind than in front and becom- Ing more circular distally. They extend obliquely upward from the skull, curving outward arid forward, and near the tips slightly inward. Their posterior surfaces are more deeply vascular than are their lateral or anterior surfaces. They are proportionately longer and more slender than in any other species. In general form they are close to the horns of T. alticornis, from the Denver formation. Another horn (U. S. N. M. No. 7086) in the National Museum collection from the same formation is almost identical with that of T. eurycephalus, al- though it must be remembered that accurate specific determinations and affinities cannot be based on brow horns alone. Perhaps the most unusual single feature of the entire skull is the great width of the crest in proportion to the length (see Plate 5). T his exceptional widening appears to be in no way the result of crush- ing for the symmetry of the parietals and squamosals seem to be little disturbed. The squamosal is proportionately wide and is plow-share- like in general form. The antero-inferior edge is extended more an- teriorly than in any other species of Triceratops. The anterior edge 15, therefore, practically straight and as a result, the infratemporal fossa is restricted. The lateral edge is smooth and comes to a thin margin with no indication of the presence of epioccipitals. The distal end of the squamosal is blunt, hówever, and an epioccipital may have ееп present there. Vascular impressions are well developed in the dorso-marginal anterior area, with few elsewhere on the dorsal and inferior surfaces. The parietals are completely fused, and are pro- portionately very broad. They are gently arched transversely and posteriorly they are flattened and curve upward rather sharply. Not all of the antero-lateral portions are preserved, but they are sufficiently complete to indicate that no parietal fenestra were present. The median ridge is only slightly undulatory and the posterior margin is developed into a smooth thin edge with no indication of epioccipitals. ascular impressions are neither abundant nor well developed and there are no special zones of these impressions. 58 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY In the basi-cranial area only the occipital condyle with parts of the ex-occipitals, the left orbitosphenoid (alisphenoid) and the left ptery- goid are preserved. The occipital condyle is nearly circular in outline and is proportionately unusually small. It is only 6.6 centimeters in diameter. The pterygoid is short and unusually massive. The basioccipital wing and the ridge immediately in front of it are nearly parallel and they are low and especially heavy. The left orbitosphenoid (alisphenoid) is splendidly preserved. It is in no way distorted and it presents excellent articulatory surfaces. The ascending wing is fairly long and narrow, and distally its blunt oval termination fits into a socket in the postorbital under the center of the brow-horn base. The orbitosphenoid was not coossified with any of the surrounding bones. Antero-posteriorly it is proportion- ately long. Posteriorly it terminates just at the anterior of the fifth nerve foramen, above which is the proótie surface, and below which 13 the surface for the basisphenoid. The orbitosphenoid contains all of the brain cavity in front of the fifth nerve, and from this cavity it was possible to get a splendid cast of that region of the brain (see Fig. 5). In size, the brain is a little more than two-thirds that of Triceratops serratus (Hatcher 1907, p. 39, Hay 1909, pl. 9) or Triceratops sp. U. S. N. М. 5740 (Gilmore 1919, р. 103). The post-cerebral constriction seems more pronounced in T. eurycephalus and cranial nerve III is not immediately behind II and inferior to IV, as it is in T. serratus, but is more behind IV. The most outstanding feature of the brain, as preserved, is, however, the separation and rapid lateral divergence of the olfactory nerves im- mediately in front of the cerebrum. This condition probably is directly coincident with the shortened facial region of the skull. A supero-mesial ridge in front of the cerebrum indicates perhaps that a well developed partition was present between the olfactory stalks. In general outline, the jaws are bluntly triangular — being propor- tionately far apart posteriorly. The articulars, angulars, and sphenials are missing. А small part of the anterior end of the right dentary is broken away but this is preserved on the left and there can be no ques- tion about the length of the jaws. The dentary is unusually short and is proportionately deep (see plate 4). The coronoid is high and the distal end faces upward and forward and is only slightly expanded antero-posteriorly. Viewed from in front, it is set considerably ex- ternal to the tooth row. No teeth are preserved, but in the right dentary twenty-two alveoli are present. As stated above, the front of the right dentary is broken away but, as indicated by the left, not SCHLAIKJER: A NEW TRICERATOPS 59 more than one alveolus is missing. The posterior end of alveolar por- tion of the dentary is also missing but certainly not more than two or three alveoli have been lost. The total number of alveoli, therefore, ‚postorbital T Surface = parietal surface | uf У Ms Jrontat surface N Г ‚prootie Surface \ G | М | | AN 7; piluilary Jossa Ета. 4. Triceratops eurycephalus. Antero-internal view of the left orbito- Sphenoid (alisphenoid) bone. Two-thirds natural size. may be conservatively estimated as twenty-five or certainly not more than twenty-six. The predentary is relatively long, flat and massive. A number of skeletal parts are preserved but the skeleton of Tricera- tops is not as yet sufficiently known to enable one to make specific { і { | | | ж 60 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY determinations. It is of interest to note, however, that the right femur of Т. eurycephalus is practically as long as the femur of the large T. elatus (?) skeleton in the American Museum, but it is by far less Win |) ^ ШЕ Кт ЛДІ, туђу ИД 2 Fig. 5. Triceratops eurycephalus. Left view of the cerebral region of the brain. Two-thirds actual size. massive. The scapula, coracoid and other parts of the skeleton such as vertebra, ribs and fragments of the distal limb elements indicate the same condition. From this it ean be concluded that in stature T. eurycephalus was almost as large as any of its Lance relatives but was indeed a more slender-bodied and perhaps a more agile form. Measurements Jagal diia ahi or arpit СО dista а ке. и Width of jugal just above distal tapermg.................. Expanse of frontal region at anterior border of orbits........ Greatest diameter ОЁ ОТО a iy а ., Least diameter of Orbit, v н Ou у un Antero-posterior diameter of horn core 6” above orbit....... Transverse diameter of horn core 6” above orbit . Distance from mid-frontal region M apex of supra- orbital hörn (аз pronen ed iir Жуа A O ero iiri Hd Antero-posterior diameter of nasal horn prominence.......... Greatest width ot ЗОНА Ва у. a DUCIT Of Occipital cond Ae МОМ nn... Тер OF predentaty ee ей eens Combined length of dentary and predentary............... Length of anterior of predentary to posterior of surangular.... Ten ОТОО cow ОВ, Height of coronoid above bottom of jaw................... Distal antero-posterior expansion of coronoid.............. Depth of jaw immediately in front of coronoid............. SCHLAIKJER: A NEW TRICERATOPS 61 Discussion and phylogeny. When it is remembered that not more than twenty-five skulls of Triceratops, which are specifically determin- able, have been collected from a formation, the Lance, representing а duration of time of about five million years and which is represented by abundant out-crops geographically distributed over a vast area from Saskatchewan and Alberta to Mexico, it is not difficult to under- Stand why the determination of the relationships among the various species of Triceratops is such a difficult problem. It is fortunate, however, that a single locality (Niobrara County, Wyoming) has provided all of the type material of the nine valid species thus far described. It is also fortunate that all of the types with the exception of T. serratus were collected by that great naturalist and collector J. B. Hatcher whose accurate field work included a record of the strategraphic position in the Lance of every specimen obtained by him, This information has been very valuable in the study of phy- logeny of the various species. The question of the relation of T. eurycephalus to the other species of Triceratops is of unusual importance not only because this species represents the culminative stages of definite evolutionary trends in Triceratops during Lance times, but because it is intermediate be- tween certain other species and thereby contributes to a better under- Standing of the relation of the different lines of development in this genus, Perhaps the most outstanding single feature of T. eurycephalus is the great breadth of the crest in proportion to the length of the skull. Anthropologists have long used the terms brachycephalie (short- headed), mesaticephalic (middle-headed) and dolichocephalie (long- headed) to define the cranial proportions of the human skull. These Proportions are obtained by dividing the width of the cranium by the length and the result is termed a cephalie “index.” This terminology Was first introduced in the study of fossil mammals by Professor Osborn in 1900 and in this connection has been most elaborately used in his monumental work on the titanotheres (1929). Widening of the mammal skull is mostly the result of lateral growth of the 2ygomata while the cranial region itself may remain proportionately olichocephalic. In Triceratops the widening of the crest is not ex- actly homologous to brachycephaly in certain mamma] skulls. It is, rather, ап actual enhanced transverse growth of two important ае elements, the parietal and squamosal; a transverse growth } ы of antorbital-postorbital skull length proportions. This 5 admirably illustrated by the skulls of Protoceratops in the splendid 62 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY collection at the American Museum. For example, by comparing two skulls, one a fairly young individual (No. 6408) and a second which is an adult (6414), it is shown that there is an actual lengthen- ing of 106 mm. in the antorbital region and 103 mm. in the postorbital region while there is an actual increase of width in the crest of 260 mm., or more than the actual increase of the skull's length. Among the species of Triceratops the same condition is present. For example, the antorbital region of T. hatcheri is proportionately longer than it is in T. brevicornus, yet the crest of T. hatcheri is propor- tionately much wider. Moreover, the proportion of the antorbital region in T. calicornis is the same as that of T. brevicornus, but its crest is proportionately wider than is that of T. hatchert. In other words, as stated above, the crest of Triceratops in proportion to the length of the skull becomes wider irrespective of the antorbital-postor- bital skull length proportions. Of the known forms this widening of the crest is most emphasized in T. eurycephalus. I suggest that this condition be referred to as eurycephaly (wide-headedness) in order to avoid confusion with a somewhat different condition among certain mammalian skulls known аз brachycephaly (short-headedness). For the forms with narrow crests the term stenocephalic! (narrow-headed) seems more fitting than dolicocephalie (long-headed). For the inter- mediate forms mesaticephalic (middle-headed) may be used. In figure six it is shown that the cephalic indices of the Triceratops species cover a wide range. T. flabellatus presents the lowest index. The low figure for the crest width in this species, however, is probably the result of high arching and of crushing. Its normal index would probably be nearer that of T. serratus. On the basis of the cephalic indices shown in figure six, I would recommend the following index range for the terms suggested above: stenocephalic 40 — 65 mesaticephalie 65.1 — 70 eurycephalic 70.1 — 94+ T. prorsus and T. flabellatus are stenocephalic. The former, how- ever, is approaching the mesaticephalic stage which includes T. brevi- cornus and T. serratus. The eurycephalie species are T. hatcheri, T. calicornis, and T. eurycephalus. The latter has a crest which is pro- portionately so much wider than any of the others that it really 18 l'The terms stenocephalic and eurycephalic have been used by early anthropologists to designate narrow and wide skulls in relation to height. The more accepted terms now in use by anthropologists are akrocephalic (narrow in relation to height) and tapeinocephalic (wide in relation to height). Stenocephalic and eurycephalic are used in this paper to designate narrow or wide skulls irrespective of height. | | l | SCHLAIKJER: А NEW TRICERATOPS 63 “hypereurycephalic.” In this connection it is of interest to note that those skulls of Protoceratops, which can be unequivocally referred to the same species, show that from youth to old age the crest becomes proportionately wider. Since Triceratops is not a very distantly re- lated form, it is entirely probable that it, too, possessed this character- istic. If this is true, then 7. eurycephalus, which according to its open Skull Crest Species Age Length Width Index T eurycephalus.......... | immature 138.6 129.7 93.5 + T. calicornis............ | not fully adult 210.0 155:2 73.9+ Ti elatus. ОРАО АИ immature 193.4 ? ? T. hatcheri............. | old 185.4 132.8 71.64- Т. brevicornus.......... old 165.2 112.0 67.84- T. serratus............. | immature О 115.0 67.2+ Ulo pM "| мой 152.3 94.4 61.9+ T. Лаһ Шаша........... very young 187.9 86.4 45.94 а V IC Со: fully grown Й ? Т d obus ЕЕ ld ? ? ? р Fra. 6. Table showing the eurycephalic, mesaticephalic and stenocephalic indices of the Triceratops species. Measurements are in centimeters. Sutures is an immature individual, would have been even more eury- cephalic had it attained old age. In eurycephaly 7. calicornis is nearest to T. eurycephalus, but only slightly more so than is T. hatcheri. In the general form, position and size of the brow horns, however, 1. eurycephalus seems to be intermediate between these two species. It 18 definitely nearer to T. hatcheri than to any other single species Ш the following characters. . Reduction of the nasal horn. - Slight antero-posterior concavity of the nasal. - Structure and form of the jugal. - Small number of teeth (T. hatcheri has only 24 in the maxillary. Ти eurycephalus 25-26 in the mandible. No other known form has fewer than 28 and the average number is 30-31). - Shape and position of the orbit. - Forward extension of the antero-inferior angle of the squamosal. нь со № н 2 к { 4 } { || 64. BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY Because of these affinities, I am inclined to place it phylogeneti- cally (see plate 6) somewhat intermediate between 7. calicornis and T. hatcheri, though closer to the latter. In all of the above mentioned characters Т. brevicornus ranks third in close relationship to T. eury- cephalus. This species, however, is much closer to T. prorsus which seems to represent a line of Triceratops evolution quite apart from the T. elatus-calicornis-eurycephalus-hatcheri group throughout Lance times. Т. brevicornus may be derived directly from T. prorsus as Pro- fessor Lull has shown in his recent and masterful memoir, “ А revision of the Ceratopsia or horned dinosaurs" (1933, p. 28). Professor Lull derives, structurally at least, T. horridus from T. brewicornus. The former, however, occurs in the lowermost Lance already a large and massive form, removed from the 7. prorsus-brevicornus line, and which is known to have persisted late in the Lance. In his tentative phylogenetic chart Professor Lull places, with reservation, Г. obtusus as the ancestor of T. hatcheri. Since T. obtusus possesses a number of specializations in the region of the face, the jugal and squamosal, which I believe eliminates it from the ancestry of T. hatchert and in addition, since 7. eurycephalus by its somewhat intermediate position seems to show a close relationship between T. hatcheri and the T. calicornis-elatus line, I prefer a phylogenetic arrangement as shown in plate 6. Professor Lull does not include T. flabellatus or T. serratus in his chart. He rightfully concludes, however, that they are rather far removed from any of the other species, and that their closest affinity 1s to each other. Nevertheless, in the characters of the nasal, jugal, squamosal and brow horns they seem to be nearest the T. elatus-calicornis-eurycephalus-hatcheri group and, tentatively, I suggest their derivation from some early form in that phylogenetie line. А study of the known Triceratops material reveals that several de- cided evolutionary trends developed throughout each of the diverse specific lines from the beginning of Lance to the closing stage of the Cretaceous. They are as follows: 1. Progressive eurycephaly. 2. Proportionate shortening of antorbital region. 3. Proportionate lengthening of postorbital region. 4. Regressive nasal horn. 5. Progressive development of brow horns. It is in T. eurycephalus, the last known survivor (that is, the last specifically identifiable form; Agathaumas and T. alticornis may be SCHLAIKJER: A NEW TRICERATOPS 65 somewhat later in time) of the ceratopsians, that these concurrent evolutionary trends attain a highly advanced culminative expression. The ancestry of Triceratops is dimmed by the lack of sufficient evidence and is, therefore, a very debatable question. Brachyceratops, however, from the Two Medicine formation of the upper Montanan, affords, in nearly all of its skull and skeletal structures, an admirable ancestor. C. W. Gilmore (1917, p. 38) states, however, that “it is hardly conceivable that an animal like Brachyceratops, having a паза] horn split longitudinally by suture and an outgrowth from the nasal bones, could be the progenitor of later ceratopsians having this horn developed from a center of ossification distinct from the nasal bones.” This seems to be insufficient cause for its elimination, especially when it is remembered that Brachyceratops possesses a separate ossicle above the nasal horn which Gilmore himself suggests (p. 9) “may represent the incipient horn of later ceratopsians." A more elaborate discussion of this subject is foreign to the present study. CONCLUSIONS l. The dark, carbonaceous, sandy shales exposed along Corn Creek are over a thousand feet in thickness as is shown by core- drilling. These beds are Pierre in age and the upper portion may repre- sent a Pierre-Fox Hills transition. 2. The Lance in Goshen Hole can be separated into three divisions. The lowermost is a continental deposit 100-200 feet in thickness Which contains ceratopsian and other reptilian remains. It is therefore equivalent to the Triceratops-bearing Lance in other areas. The middle division із a brackish-water deposit 807-125’ thick which is in a correct Stratigraphic position to be correlated with some phase of the Can- nonball. Above the brackish-water beds is a 60'–100 +" continental deposit which contains, in its upper portion, a new and very advanced form of Triceratops. For this deposit the name, “ Torrington member of the Lance" is suggested. 3. Goshen Hole is a topographie basin developed on a structural dome. Within this basin is a pronounced anticline along Corn Creek ànd a broad low arch in the Horse Creek vicinity. 4. Triceratops eurycephalus is a very advanced species and it pre- Sents a culmination of the main evolutionary trends in Triceratops development during Lance times. Its phylogenetic position is some- What intermediate between T. hatcheri and T. calicornis though it is пеатег to the former. It aids greatly in determining the phylogeny of all of the Triceratops species. 5. Brachyceratops of the Two Medicine formation, appears to be the most probable ancestor of Triceratops. 4 | ( | | 66 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY BIBLIOGRAPHY ApAMS, С. І. 1901. Geology and water resources of the Patrick and Goshen Hole Quadrangles in Eastern Wyoming and Western Nebraska. U. В. Geol. Surv. Water Supply Paper, No. 70, pp. 1-47, pl. 1-11, and 4 text figures. CALvERT, W. R. 1912. Geology of certain lignite fields in Eastern Montana. Bull. U. S. Geol. Surv., No. 471, pp. 187-201 and 1 text figure. Cnoss, W. 1889. The Denver Tertiary formation. Amer. Jour. Sei. 3rd series, 37, pp. 261-282. Dossin, C. E. anp Вввзтрв, J. B. ЈЕ. 1929. The contact of the Fox Hills and Lance formations. U. В. Geol. Surv. P. P. No. 158-B, pp. 9-25, pl. 4 and 5, and 4 text figures. Опмове, С. №. 1917. Brachyceratops. A Ceratopsian dinosaur from the Two Medicine formation of Montana with notes on associated fossil reptiles. U s. Geol. Surv. P. P. 103, рр. 1-44, pls. 1-4 and 57 text figures. 1919. A new restoration of Triceratops with notes on the osteology of the genus. Proc. U. S. Nat. Mus., 55, pp. 97-112, pls. 1-9 and 6 text figures. 1922. The smallest known horned dinosaur, Brachyceratops. Proc. U. S. Nat. Mus., 61, pp. 1-4, pls. 1-4. Снкоову, W. K., anb Mook, С. С. 1925. On Protoceratops, а primitive ceratopsian dinosaur from the lower Cretaceous of Mongolia. Amer. Mus. Novitates, No. 156, pp. 1-9, 3 text figures. HATCHER, J. B. 1893. The Ceratops beds of Converse County, Wyoming. Am. Jour. Sa 3rd ser., 45, pp. 135-144. Harcuer, J. B., Marsa, O. С. AND LULL, TIUS 1907. 'The Ceratopsia. Monograph 49, U. S. Geol. Surv., pp. 1-198, pls. 1-49, 125 text figures. Нах, ©. P. 1909. On the skull and the brain of Trieeratops, with notes on the brain cases of Iguanodon and Megalosaurus. Proc. U. S. Nat. Mus., 36, pp. 95-108, pl. 1-3. SCHLAIKJER: A NEW TRICERATOPS 67 Knowrron, Е. Н. 1909. The stratigraphie relations апа paleontology of the “Hell Creek Beds", “Ceratops beds" and equivalents, and the irreference to the Fort Union Formation. Proc. Wash. Acad. Sci., 11, No. 3, pp. 179-238. Lrovp, E. В. 1914. The Cannonball River lignite field, Morton, Adams and Hettinger Counties, North Dakota. Bull. U. S. Geol. Surv., No. 541, pp. 243-291, pls. 1-14. Тлюур, Е. В. Амр Hares, C. J. 1915. The Cannonball marine member of the Lance formation of North and South Dakota and its bearing on the Lance-Laramie problem. Jour. Geol., 23, pp. 523-547. Lunt, В. 8. 1933. А revision of the Ceratopsia or horned dinosaurs. Memoirs of Peabody Mus., New Haven, Conn., 3, pt. 3, pp. 1-175, pls. 1-17 and 42 text pictures. Mexx, Е. B. AND Наурех, Е. V. 1861. Description of new lower Silurian, Jurassie, Cretaceous and Ter- tiary fossils, collected in Nebraska, . . . with some remarks on the rocks from which they were obtained. Proc. Acad. Nat. Sci. Phila- delphia, 13 pp. Ossory, Н. Е. 1929. The Titanotheres of ancient Wyoming, Dakota and Nebraska. U. S. Geol. Surv. Monograph 55, pp. 1-953, pls. 1-236 and 797 text figures. SCHLAIKJER, E. M. 1933. Contributions to the stratigraphy and paleontology of the Goshen Hole area, Wyoming. Bull. Mus. Comp. Zoöl., 76, No. 1, pp. 1-27 pl. 1 and 7 text figures. STANTON, T. W. 1909. The age and stratigraphy of the “Ceratops beds" of Wyoming and Montana. Proc. Wash. Acad. Sci., 9, No. 3, pp. 239-293. 1910. Fox Hills sandstone and Lance formation (“Ceratops Beds") in South Dakota, North Dakota and eastern Wyoming. Am. Jour. Se., ser. 4, 30, pp. 172-188. 1920. The fauna of the Cannonball marine member of the Lance forma- tion. U. S. Geol. Surv. P. P. 128-A, рр. 1-64, pl. 1-10 and 3 text figures, 68 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Тновм, W. T., ЈЕ. лмо Боввіхв, С. Е. 1924. Stratigraphy of Cretaceous-Eocene transition beds in eastern Montana and the Dakotas. Bull. Geol. Soc. Am., 35, рр. 481-506, pls. 23-25, 3 text figures and 2 tables. Warp, Е. 1924. The Lance problem in South Dakota. Am. Jour. Sc., series 9; f, pp. 65-68. WINcHESTER, D. E., Hares, C. J., Ілоүр, E. R., AND Parks, E. М. 1916. The lignite field of northwest South Dakota. Bull. U. $. Geol. Surv., No. 627, pp. 1-169, pl. 1-11 and 3 text figures. EXPLANATION OF PLATES қ Ё— | PLATE 1 SCHLAIKJER — А NEW TRICERATOPS у | | | | PLATE 1 Map of the Cretaceous formations in Goshen Hole, Wyoming. | і BULL. MUS. COMP. ZOOL. SCHLAIKJER. A New Triceratops. PLATE 1 | | mz | МАР ОҒ ТНЕ CRETACEOUS FORMATIONS IN GOSHEN НОГЕ, WYOMING E | Geology b Scale Explanation | Erich M. Bener 1955 : 5 < j Lane Бет 2 Vertebrates+1++| Invertebrates || ] | | | | pue p^ Sud | | | A и P я | | | м Д амал NA FOSS RESERVE с Busch | | due | | e | ЧАД | SP | | | | Ese radi | | т | A | Veteran / 1 | | 5 | ПАРА T AMD, | 3 (2 | > ca ) | 1/2 ZA | ) | | | | | РА | uy, il Ж E 122 ІШ ШТ WS Вох вы КАДА al TE | IM £M М; МА ht ко pido | 9% OSHEN У SERVOIR SS od "AS > Zany Mp gm „ш Temm - 247%, AU 3 E A PASS = 9 27277 5 4 w, ит, ШУ! “йш, x ^ E Int it N m: CHEYENNE —T ORRINGTON HIGHWAY SSNS a rawn Бу Erich M Schlager Ré4w ме | Í а | 2l y j ; | N | | E mé 2 Нық | ў | | un, - aw] rings і ols | E 2 4 | | | = | ZA | 5 5 - ~ й T | > 7 ; e^ 21 M R LA й 2 % X . > 8 Жее т 3 YY 4 ре Ее ES 7 slowa Cente L DARE 77 2 A | = 4 | SS IND | | = КІ? N I 22001 2, | [A / E M" Sa Te 7, y ; BOR | A и Su, A ій, У | | С ++ E ^n : | | бы ж ЖЕТ? % i E c 5 | | SA ME ЛК а ЗЕЕ % E S28 "n OA DP 2 | | Fox Creek CARY Uis n "иу ун HAWK | a АВИИ lus S qo um л лу [ҮЛ ШҮ ЫШ УУ "m, ИУ “ҮМ | ü % = < 66 s SE zi pi rui ZZ 8 в | | а а 2 SPRINGS Ў = | AW ing саму NIE UN + \ | ДА Be su | fox 5, «b "A A Эт Yo y E T КО E ВЕЕ Y “ay x KIS А < = Ж фи i 2) & ^ Ка e N еңді ЫЗ Fr еШ, une” 44 "mz ТЕЗ ©, %% SS E M DU MS ©. => Vin ТЕДІ SS e AN анти у м, 3. | ez = p М - S 2 | = 2- KR 7, Ni | = СА. ZT | | Тар 2и le SR 2 ЛЕР” ще т! = | | | WE SS жүре, | те | | it 2 uy e СЛОЯ p | 60 AD „7 7 à ai 77 2 7 v = * | | + 2 | mul % 4775 N кестім Cra B P» A m * у 22 r Zi Ш) 2 EU, SS Bet EN LEN ESTER Rew нем В 60w В 58W PLATE 2 SCHLAIKIER— А New Triceratops PLATE 2 1. Typical exposures of the Lance formation along Horse Creek showing local dipping of the beds. 2. Oyster beds along Horse Creek having a measurable thickness of over twenty feet. BULL. MUS. COMP. ZOOL. ScHLAIKJER. A New Triceratops. PLATE 2 | | PLATE 3 PLATE 3 1. Cannonball concretions in the Lance one mile northeast of Fox Creek Gap. 2. Channel sandstone in the Lance on which rest the lower Oligocene clays One mile northwest of Table Mountain. Photographs, plates 2 and 3 by Erich M. Schlaikjer BULL. MUS. COMP. ZOOL. ScHLAIkJER. A New Triceratops. PLATE 3 PLATE 4 SCHLAIKJER— А New TRICERATOPS PLATE 4 Triceratops eurycephalus Schlaikjer. Skull, right view. Approximately one- twelfth actual size. A New Triceratops. PLATE 4 ScHLAIKJER. a о O N в. = Q o [7] => = і = = с ко E E < = 2. SCHLAIKJER — А NEW TRICERATOPS PLATE 5 Triceratops eurycephalus Schlaikjer. Skull, front view. Approximately one- fourteenth actual size. BULL. MUS. COMP. ZOOL. SCHLAIKJER. A New Triceratops. PLATE 5 | AT 55 i = = - = zx ae; 5 си -- те -" -— u — == I E s G = T TER zd EEE PLATE 6 ScHLAIKJER — A New TRICERATOPS PLATE 6 Phylogeny of Triceratops. Uppermost Cretaceous, North America. BULL. MUS. COMP. ZOOL. SchLalkJeR. A New Triceratops. PLATE 6 PHYLOGENY OF TRICERATOPS UPPERMOST CRETACEOUS, NORTH AMERICA ы © ZA | T.CALICORNIS Ж | | T.PRORSUS | | « | | | | | | | | | 4 Ss | | x M | > ~ | | | | T.HORBIDUS ~ — Lo (©) = Ж m (©) ps A © > а zl Ж = LEVEL OF ТҮРЕ == KNOWN VERTICAL RANGE ERICH M.SCHLAIKJER 1933 Bulletin of the Museum of Comparative Zoólogy AT HARVARD COLLEGE Vor. LXXVI, No. 3 3131 CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY OF THE GOSHEN HOLE AREA, WYOMING III A NEW BASAL OLIGOCENE FORMATION By Енісн М. ScHLAIKJER МУ Wits Етент PLATES CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM JANUARY, 1935 ИІ A} H PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOÓLOGY AT HARVARD COLLEGE There have been published of the Burzerin Vols. I to LXV, LXVII-LXXV, LXXVII, of the Memorrs Vols. 1 to LIT, LIV. The BurLETIN and Мемогвз are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collections and Exploration. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately. A price list of the publications of the Museum will be sent on application to the Director of the Museum of Comparative Zoölogy, Cambridge, Massachusetts. Bulletin of the Museum of Comparative Zoülogy AT HARVARD COLLEGE Vor. LXXVI, No. 3 CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY OF THE GOSHEN HOLE AREA, WYOMING ШП A NEW BASAL OLIGOCENE FORMATION By Евісн M. SCHLAIKJER Wits Егснт PLATES CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM JANUARY, 1935 No. 3. — Contributions to the Stratigraphy and Palacontology of the Goshen Hole Area, Wyoming III A New Basal Oligocene Formation. By Erich М. ScHLAIKJER INTRODUCTION Previous to the publication of a paper by O. A. Peterson and L. J. Kay in 1931 on the Duchesne formation of northeastern Utah, the boundary between the Eocene and Oligocene continental sediments Was generally accepted as between the Uinta and the Chadron. The boundary is well defined and was regarded as representing a consider- able break, faunistically at least, between these two formations. This marked separation between the Eocene and Oligocene is now less Pronounced since the discovery of the Duchesne with its inter- mediate fauna. The Oligocene affinities of this fauna led Peterson and Kay to consider it as belonging to the basal part of that geological Period, while Simpson (1933, pp. 84, 85), because of the occurrence of Epihippus? and on the basis of physiographie and stratigraphic evidence places it very tentatively in the Eocene. In certain parts of the Goshen Hole area, Wyoming, there is a series of fluviatile deposits which lie with unconformity on the Lance formation of the upper Cretaceous and which grade upward into the hadron formation of the lower Oligocene. This series of beds was first discovered by the author in July 1930 when he began the field research for the Museum of Comparative Zoölogy in that area. Dur- Ing that and subsequent field seasons he has carried out а stratigraphie study of these beds and has obtained from them a considerable col- lection of vertebrates. А number of the forms represented are new. А description of these together with the results of the stratigraphic Investigation is the subject of this paper. From all the available data It seems justifiable to regard these deposits as representing a new forma- ton constituting, in part at least, a transitional stage between the "ocene and Oligocene periods. N Wish to acknowledge the kindness of Mr. Harold J. Cook of Agate, ven who has most generously turned over to me for study and . "CIpton three specimens which he collected from this formation in ‚the fall of 1932. а mings of the specimens in this paper were made by Mrs. ` 41SKa,, 72 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY STRATIGRAPHY The geographic distribution of this series of sediments is very limited. The combined extent of the exposures in Goshen Hole amounts to not more than five square miles (see Fig. 1). Since the locality is one of low relief no good sections can be measured except along the Fort Laramie irrigation canal, especially one to three miles northwest of the town of Yoder, where deep excavation was neces- sary to get the canal waters across the plunging northeast limb of the Corn Creek anticline. The walls of this canal present excellent ex- posures of the beds and a fairly thick section is measurable. The maximum thickness is indeterminable, however, since nowhere is the deepest contact with the Cretaceous observable. The lowermost part of the canal walls.is composed of fifteen feet of rather finely bedded fairly coarse red sandstone and dark to red sandy clays. The next eighteen feet is red to greenish sandy clays, and these are over-lain by twenty feet of reddish-green clays with channels of intricately cross- bedded, wine colored coarse gravels to fine sandstone which grade laterally into the clays. Resting on this sequence, in places with no apparent stratigraphic break, are the typical greenish-gray Chadron clays (see Fig. 2). In other places as in Sec. 19, T. 23n. R6lw. the series їз capped by a greenish-maroon coarse channel sandstone several feet in thickness with the Chadron clays above. Fossils are abundant and although they frequently occur as frag- ments they are little if at all water worn. The forms thus far dis- covered indicate a fauna which is distinctly transitional between the uppermost Еосепе, or Duchesne, and the lowermost Oligocene, or lower Chadron. The Eocene representative in this fauna is the creodont, Міасіз, but it is an end-member of the Eocene forms and may be considered as a living fossil existing along with its more ad- vanced relatives Hemipsalodon and Hyaenodon at the dawn of Oligo- cene times. The rhinoceroses and titanotheres are definitely early or pre-Chadron members and the horse, so far as is known is decidedly an intermediate form transitional between Epihippus intermedius of the Duchesne and the primitive Mesohippus representatives of the early Chadron. The Hypertragulid present is Герютегух and though it possesses characters more primitive than those found in other species, it is already a typical example of this Oligocene genus. The fauna, therefore, considered in its entirety, seems t00 progressive to be assigned to the Eocene, yet its primitiveness will not allow it to be regarded as of the Chadron. Because of these facts and because 78 SCHLAIKJER: А NEW BASAL OLIGOCENE FORMATION Dong мәз HEH “WN 4243 Ха имел y ло! рән Es РЫ заредфоган, Y R ДЕ ES A СЕБЕ >, Ж XE | м) 5) d Tf 5212 IS чела) p "s й Е | TS e^ J N He Se // E ) E t Be ita а» t £ A ә 7 gore pe ; Є >] ava = \ == == | . JRPS W цомі3 9226 Ха 4801099 NOOZLWIAWO нлаох AHL JO ЧУИ к> 74 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY the beds represent a stratigraphic level, quite apart from the upper Eocene and are easily distinguishable from the Chadron, a new for- mational name is necessary. The name proposed for these deposits is the Yoder formation, so named after the town of Yoder which is situated one and one-half miles southeast of the type locality. After an evaluation of the faunal relationships and after its stratigraphic position is taken into account, the Yoder formation should be re- garded as belonging to the Oligocene rather than to the Eocene system. It is a significant fact that the Yoder formation represents the earliest Oligocene deposition, thus far recorded, in the Great Plains region. Nowhere in the Goshen Hole area are Paleocene or Eocene deposits to be found. It is probable however that the former, at least, were present, since Fort Union beds, hundreds of feet in thick- ness, occur less than eighty miles to the north. Subsequent to Port Union times and perhaps during most if not all of the Eocene, degrada- tion continued and by the close of the Eocene these sediments as well as some of the Cretaceous had been removed. The beginning of the Oligocene was marked by renewed uplift in the Rocky Mountains which resulted in the rejuvenation of all those eastward flowing streams that then began to carry from the uplands out on to the plains the sediments of the early Oligocene. One of those streams flowed across what is now the Goshen Hole area, entrenched itself in the Lance formation of the Cretaceous and deposited the Yoder sediments. Hence for the first time perhaps since the Paleocene this area became one of deposition. The trend of the Yoder river was from west to east as is admirably shown by the channel deposition. The restriction of the formation to the northern portion of Goshen Hole may be the result of local uplift in the Corn Creek vicinity immediately to the south from which perhaps some of the Yoder sediments were de- rived. Most of the sediments, however, originated in the nearby Laramie mountain region to the west. The angular condition of the more elastic materials shows that they have not undergone prolonged transportation and the freshness of the feldspars indicates that dis- integration rather than decomposition was the major process of erosion at the source. PALAEONTOLOGY In the collection of vertebrates from the Yoder there are a few reptilian fragments. These are, isolated fragments of a turtle cara- | | | | R.62y. G SCHLAIKJER: A NEW BASAL OLIGOCENE FORMATION YO nr 15! Ета. 9. ALLUVIUM TYPICAL GRAYISH-GREEN CLAYS, CHANNELS OF WINE-COLORED INTRICATELY CROSS-BEDDED COARSE TO FINE SANDSTONE. SUBHYRACODON WOODI REDDISH-GREEN CLAYS. ATELEODON OSBORNI BRONTOPS BRACHYCEPHALUS HYRACODON ISCHYROLOPHUS RED TO GREENISH SANDY CLAYS, MOSTLY BEDDED. HEMIPSALODON COOKI HYAENODON HORRIDUS CAENOPUS YODERENSIS RED SANDSTONES AND DARK TO LIGHT RED RATHER FINELY BEDDED SANDY CLAYS. MIACIS MATTHEWI LEPTOMERYX YODERI 2MESOHIPPUS SP. CONCEALED. Type section of the Yoder formation. S.E. М. Sec. 29. Т. 23m. oshen Co., Wyoming, E. M. Schlaikjer 1933. 76 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY pace and several crocodilian vertebra. None of this material is generically identifiable. The remainder of the material is mammalian of which the following is a faunal list. Order CARNIVORA Suborder CREODONTA Superfamily MIACOIDEA Family MIACIDAE Miacis matthewi sp. nov. Superfamily OXYAENOIDEA Family HYAENODONTIDAE Hemipsalodon cook? sp. nov. Hyaenodon horridus Leidy Order PERISSODACTYLA у Superfamily EQUOIDEA Family EQUIDAE ? Mesohippus sp. Superfamily BRONTOTHERIOIDEA Family BRONTOTHERIIDAE Ateleodon osborni gen. et. sp. nov. Brontops ? brachycephalus (Osborn) Superfamily RHINOCEROTOIDEA Family HYRACODONTIDAE Hyracodon ischyrolophus sp. nov. Family RHINOCEROTIDAE Caenopus yoderensis sp. nov. Subhyracodon woodi sp. nov. Order ARTIODACTYLA Suborder PECORA Superfamily TRAGULOIDEA Family HYPERTRAGULIDAE Leptomeryx yodert sp. nov. | | | | SCHLAIKJER: A NEW BASAL OLIGOCENE FORMATION {7 DESCRIPTION AND DISCUSSION OF NEW MATERIALS Order CARNIVORA Suborder CREODONTA Superfamily MIACOIDEA Family MIACIDAE MIACIS MATTHEWI! sp. nov. Ў Type. Left lower jaw fragment with Ms апа part of My, М.С. Z. No. 2,094 and associated right lower jaw fragment with Р; and the Toots of P; and Mi, М. С. 7. №. 2,094a. Collected by E. M. Schlaik- Jer, 1932. Horizon and locality. Basal Oligocene, Yoder formation. Goshen County, Wyoming. Specific characters. A large form. Trigonid of M5 low and talonid fairly large and broad. Talonid of Mi broad and rounded posteriorly. E 4 With pronounced posterior accessory tubercle half way up on the principal сизр. Principal сизр erect. Internal cingulum incomplete at the base of principal eusp. External cingulum faintly developed. (See Fig. eU Fig. 3. М. tacis matthewi. Above, lingual and superior views of M» and part of M.. Below, external and superior views of P4. Natural size. Discussion. M. matthewi is one of the largest known members of this genus, In general it is about the size of M. longipes (Peterson) Tom the upper Uinta. It is different from that species, however, in the following characters: 1, Deeper ramus. 2, Principal cusp of P4 more act, Posterior accessory cusp more pronounced and internal cingu- um "complete. 3, Talonid of M; broader. М, larger. In the pro- Portionately larger size of Ms, M. matthewi is closer to M . medius 1 Р OF ач in honor of the late Dr. W. D. Matthew, author of the “Carnivora and Insectivora © Bridger Basin, Middle Eocene,” etc. 78 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Matthew of the lower Washakie but of course is different from that species in size and in the reduced anterior basal cusp of Ра. The relationship of M. matthewi to the other species of this genus will remain obscure until more satisfactory material is discovered. For the present, however, there seems to be no good reason for not considering this species an end member of а M. medius-longipes line of Miacis evolution. It is the last known survivor of that genus. Measurements mm. Eo Vi е (estimated a een Lena. И ыла LR ык их 12170 Width, Or Mr across бшш ren 6.5 RSM SOL OL Wig Sn То Superfamily OXYAENOIDEA Family HYAENODONTIDAE HEMIPSALODON COOKI! sp. nov. Type. Anterior portion of left mandible with Рҙ-Мі and the alveoli of C and Рі-. No. Н. C. 530 Cook Museum, Agate, Nebraska. Found by H. J. Cook, 1932. Associated is the distal end of the left humerus, M. C. Z. No. 2,093. Found by E. M. Schlaikjer, 1931. Horizon and locality. Basal Oligocene. Yoder formation. Goshen County, Wyoming. Specific characters. Smaller in size than H. grandis Cope, but much larger than Pterodon africanus Andrews. Diastema between P» and Рз. Canine proportionately smaller than H. grandis. Width of M; about one-half its length. Humerus very massive and very broad distally. (See Figs. 4 & 5.) Discussion. In 1885 Professor Cope described a right lower creo- dont jaw containing the crown of Мз and the roots of all the other teeth from the lower Oligocene of the Cypress Hills, Canada. To this jaw he applied the name Hemipsalodon grandis. In a later publication (1891 pp. 6-7, pl. II) a fuller description was given and the specimen was figured. In 1908 Lambe (p. 61, pl. VII, Figs. 7-8) described and figured a single canine tooth from the same beds which he rightfully referred to Cope's genus and species. Until 1931 no other material ! Named in honor of Mr. Harold J. Cook of Agate, Nebraska. SCHLAIKJER: A NEW BASAL OLIGOCENE FORMATION 79 of this largest known North American creodont (equalled in size perhaps by Н arpagolestes of the Eocene) had been recorded. Matthew EN Не externa] view nipsalodon cooki. Above, superior view of PM.. Below, of left ramus fragment. Natural size. 1‹ 3 5 nu 20) and others have considered this genus às synonymous 1 ; Í ; ka Pterodon of Europe. A study of the newly discovered material a reconsideration of all the available information, however, causes 80 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY the author to regard Cope's genus as valid. That Hemipsalodon grandis and Н. cooki belong to the same genus is established by the following similarities: 1. Large, deep and compressed mandible. 2. Position and proportions of the cheek teeth. 3. Development of the symphysial region. 4. Position of the mental foramina. The differences between the two species have been cited above. It is unfortunate that none of the upper dentition is known. On the basis of the available data, however, the following characters of Hemip- salodon may be regarded as distinguishing it from Pterodon: 1. Much larger in size. 2. M; proportionately smaller. 3. Mi proportionately less reduced and with large talonid. 4. Р, proportionately heavier and with a much better developed postero-internal basin. That the two genera are closely allied is, nevertheless, beyond question and it is probable that Н emipsalodon was derived from some primitive form of Pterodon such as that repre- sented by P. leptognathus Osborn from the Fayúm of Egypt. The appearance of Hemipsalodon and Hyaenodon in North America is significant in the history of faunal mixing which took place between the Old and New Worlds in earliest Oligocene times. These two specialized creodonts at that time arrived in the homeland from whence their ancestors had migrated early in the Eocene. The distal end of the left humerus shows that that bone was of ex- traordinarily massive dimensions. It is considerably larger than the humerus of Harpagolestes. These facts indicate that the skeleton of H. cooki was large and very massive. Measurements mm. РИ ЛЕН ар ou. set Da a е л 81.0 еп о Ps (on ао 20.8 О 23.0 ео ОЕ Lee oou iu RO ҒА ӨТУ DI HM eps 27.0 Length of Mi pee p EL ME 25.0 Width ol Mar. 2 1 AR oett near ovs 11.0 Width of distal end obohutietuss м 0 cv A REUS BOUT Y 86.0 Нулеморох (NEOHYAENODON Thorpe) HORRIDUS Leidy Material. Fragment of the upper left maxillary with РМ». Н. C. No. 531 Cook Museum, Agate, Nebraska. Found by Н. 7. Cook, 1932. | i | | SCHLAIKJER: A NEW BASAL OLIGOCENE FORMATION 81 Horizon and locality. Basal Oligocene. Yoder formation. Goshen County, Wyoming. Discussion. In size this specimen is slightly smaller than H. horri- dus. 'The evolutionary stage of M? is about the same. The only im- portant difference between the two is a better developed deuterocone t HZISKA Е On IG. 5. Hemipsalodon cooki. Anterior view of distal end of left humerus. © half natural size. 6 Ета. 6. Hyaenodon horridus. Above, crown view of P}M?. Below, ex- ernal view of the same. ‘Two-thirds natural size. d P* of the Yoder specimen. When more complete material is found this form may prove to be specifically distinct from the later species. Until then, it is tentatively referred to H. horridus. Я Measurements mm. E с Hg oe tase) o oC 62.0 BER ш o ss oos шлш . 185 E I n RUNDEN з Ms 17.0 END i ooi LN de 16.0 Length of мр ей i ББ BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Order PERISSODACTYLA Superfamily EQUOIDEA Family EQUIDAE ? MESOHIPPUS sp. Horizon and locality. Basal Oligocene Yoder formation, Goshen County, Wyoming. Collected by E. М. Schlaikjer, 1931. Description and discussion. The specimen is а Jaw fragment con- taining a single tooth, left М», М. C. Z. No. 2,085 (Fig. 7). The tooth appears to represent a form intermediate between Epihippus апа Mesohippus. It is somewhat nearer to the former in size and in that Fra. 7. 2 Mesohippus sp. Lingual and superior views of left Ms. Natural size. the paraconid retains a position internal to the fore and aft alignment of the metaconid-metastylid and entoconid cusps. Of the known species of this genus this specimen seems nearest to E. intermedius Peterson (1931, pp. 66-68). Е. intermedius is, of course, much closer to | Mesohippus than any of the other known Epihippus species and may, | | as suggested by Peterson, represent a distinct genus. The affinities | | of the Yoder specimen seem rather closer to Mesohippus than to | Epihippus especially in the greater breadth of the tooth in proportion to its length and in a better developed endostylid. For these reasons it has been assigned very tentatively to that genus. Of the described species it is very close to Mesohippus westont Cope from the lower Oligocene of the Cypress Hills. Measurements mm Aero DOSte ОПО аон 9.5 Greatest widins re лы ан 8.0 УТО OL сос е 4.5 SCHLAIKJER: A NEW BASAL OLIGOCENE FORMATION 83 Superfamily BRONTOTHERIOIDEA Family BRONTOTHERIIDAE ATELEODON OSBORNT gen. et sp. nov. Type. Symphysial region of a lower jaw and associated left М». Adult M. C. Z. No. 2,100a and No. 2,100b. Collected by E. M Schlaikjer, 1931. Paratype. Anterior part of right ramus containing Mı, Young. М. С. 7. No. 2,099. Collected by E. M. Schlaikjer, 1931. Horizon and locality. Basal Oligocene, Yoder formation. Goshen County, Wyoming. Generic and Specific Characters. Canine and incisors greatly re- duced or absent. Р, diminutive. Pz much reduced. Symphysis short and shallow, Chin convex. Cingulum incomplete over external sur- faces of proto- and hypoconids. Crowns of M; and М» proportionately tall. M; proportionately small. (See plates 2, 3, 4 and 5.) Discussion. А character of this specimen not to be found in any other recorded species and one which seems incredible for any titan- Othere to possess is the apparent great reduction or complete loss of the lower canines and incisors. No teeth are present in the type jaw fragment, although the roots of two and the alveoli of two others are preserved, These are the roots and alveoli of the premolar series. The most anterior tooth v 'as single rooted and the central vertical ridge Оп the posterior wall of the alveolus precludes the possibility of the Toot representing a much reduced and antero-posteriorly compressed "nne. [t is undoubtedly the root of Py. Immediately behind this root are the roots of P; which show that that tooth was much reduced m size. The succeeding alveoli represent Pz-4. The whole premolar Series was very much shortened,— a character also shown in the paratype jaw. For a distance ої 66 mm. in front of P; the bone tapers to a thin edge. The most anterior portion of the jaw, perhaps not More than a half to three-quarters of an inch, is broken away. The Preserved portion, however, is sufficient to indicate that the tapering continued anteriorly. The edge present does not show the slightest evidence of a normal deep canine alveolus which ought to be there if that tooth had been present. This character of the jaw cannot be the Tesult of erosion for the original surface and texture of the bone is lp > eon the Greek а without, re\eos distant, обох teeth. Named in honor of Professor Wald: airfield Osborn, author of “The Titanotheres of Ancient Wyoming, Dakota and Е а, ete. | || f | | 84 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY clearly shown, and there is no indication of any old age or pathological condition. Тһе paratype jaw presents the roots of Рз апа D.P., and the first molar. Although the symphysis is missing, characters such as the proportionately small premolars and shallow ramus as well as the form and dimensions of M; show that it is probably Ateleodon osborni. Both were collected from exactly the same geological level and only a few feet apart. In size, Ateleodon osborni is intermediate between Teleodus uintensis Peterson of the Duchesne and 7. primitus Lambe of the lower Chadron. It differs so far as is known from these, however, aside from the great reduction or complete loss of the lower canines and incisors, in the following characters: 1. A more shallow ramus. 2. Chin more gently convex and much more shallow. 3. Much greater reduction of Ps. Мі is of the same width as in T. wintensis although it is longer and taller erowned. М» is about the same size. 'The great reduction or complete loss of the lower canines and in- cisors is a character certainly of more than specific value in the titan- othere group. Ateleodon osborni represents an aberrant form which probably became extinct at the end of Yoder times or early in the Chadron. It is indicative of one more significant stage in the great adaptive radiation of the titanotheres. Measurements mm. енсін ОДА оша астана а ded mern dee DH n 14.5 (Greatest епо ӨМ s. busca otv La олт ut es 43.5 Greatest. width О Моон ен ав а o Y edo 24.5 (тезтез етот Of Vlos u inn cdo dB BUR S TO Eoo AD 49.5 (Greatest ОГ MBs аи ito disp OLD о ин 30.5 BRONTOPS ?BRACHYCEPHALUS (Osborn) Material. Anterior fragment of left lower jaw with Р, and Мі, and associated right М» and left Т, М. C. Z. No. 2,098 a, b, € c. (See plates 5 and 6.). Collected by E. M. Schlaikjer, 1931. Horizon and locality. Basal Oligocene. Yoder formation. Goshen County, Wyoming. SCHLAIKJER: А NEW BASAL OLIGOCENE FORMATION 85 Discussion. In size, the specimens compare most favorably with B. brachycephalus although somewhat smaller. The known portion of the jaw shows that the ramus was rather shallow, especially anteriorly. P, represents about the same evolutionary stage as in В. brachycepha- lus. It has, however, а very much developed external cingulum, a character also present in М 1». М» also differs from this species in that the external surfaces of the protoconid and hypoconid display greater obliquity; the postero-external corner is much more de- veloped; and, the anterior portion of the tooth is much narrower than the posterior. The upper dentition is represented only by the left I? Which exhibits a perfectly smooth rounded crown. When more material is known, these specimens may be found to represent a new species,— a stage perhaps directly ancestral to Brontops. Measurements mm. Greatest teneti DL сала уы уу л EO ET. 40.0 Greatest width Of Baro cda а UT о 32.0 Greatest length of Ma? ox LAE CURA с OEC AW Gs ТІ 70.0 а. 44.0 Length ӨК ООУ s ІТ e ET. 8.0 йшй Жөнө EU ood ОВА ci^ dia conr n 719 Superfamily RHINOCEROTOIDEA Family HYRACODONTIDAE HyRACODON ISCHYROLOPHUS! sp. nov. Type. Left M? No. H. C. 532 Cook Museum, Agate, Nebraska. Found by H. J. Cook, November, 1932. Horizon and locality. Ваза] Oligocene. Yoder formation. Goshen County, Wyoming. Specific characters. Extension of ectoloph, posterior to metaloph Well developed. Postfossette pronounced. Parastyle internal to para- Cone. Constriction between protocone and protoloph. Posterior of the tooth broad and straight-margined. А From the greek ıaxvpos strong, Aopos ridge, in allusion to the well developed post-metaloph Portion of the ectoloph. 86 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Discussion. The specimen here described is the earliest known Hyracodon and even though the material consists only of a single tooth it is worthy of considerable attention. To be sure, there is some individual variation in M? in each of the several described species but these variations are never interspecific. There are, therefore, cer- tain M? characters in, H. ischyrolophus (listed above) which are specifically distinctive, and which indicate a morphologic stage inter- mediate between Prothyracodon obliquidens (Scott and Osborn) of the Ета. 8. Comparative drawings of the left M*of several hyracodont species. Arranged stratigraphically with the oldest at the left. A, Prothyracodon obliquidens. В, Н. ischyrolophus type. С, Н. petersoni. D, H. nebrascensis. E, H. apertus. A (reversed) and C after Wood, D after Leidy and E after Sinclair. Drawn to scale, x 28. Uinta and H. petersoni Wood of the Chadron, but somewhat nearer the latter (see Fig. 8). Dr. H. E. Wood in his paper on Hyracodon petersoni (1926, p. 317) states that it is possible to derive all known species of Hyracodon from H. petersoni “or perhaps, from something very close toit, with a larger posterior buttress on Ма." H.ischyrolophus, so far as is known, seems to fulfill Dr. Woods’ prophecy. There is, however, still a considerable gap between Prothyracodon obliquidens and H. ischyrolophus. Measurements mm. Antero-posterior diameter of М3........................... 18.0 (Greatest width OL ME teuer 22.0 Family RHINOCEROTIDAE CAENOPUS YODERENSIS Sp. NOV. Type. Left mandible fragment with Муз МІ. Є. 7, No. 2,09% Collected by E. М. Schlaikjer, 1931. Horizon and locality. Basal Oligocene, Yoder formation. Goshen County, Wyoming. SCHLAIKJER: A NEW BASAL OLIGOCENE FORMATION 87 Specific characters. Intermediate in size between Kotrigonias rhinocerinus Wood from the upper Uinta and Caenopus dakotensis from the upper Oligocene. M; with minute internally placed posterior cingulum. Мі proportionately small (See plate 7). Discussion. Caenopus yoderensis forms an admirable morphologi- cal and stratigraphical stage intermediate between Kotrigonias rhinocerinus and C. dakotensis. It is different from the latter, in so far as is known, in its smaller size, in the proportionately small Mı and in the presence of a diminutive posterior cingulum on Mz. In all of these characters, however, it is definitely in advance over E. rhinoceri- nus. C. mitis is a much larger form than any of the other species but it too could easily have been derived from C. yoderensis. It probably represents a line of more rapid development in the adaptive radiation of this genus than does C. dakotensis. The phylogenetie position of ? C. premitis is an open question (See Gregory and Cook, 1928, p. 19 and Wood, 1931, p. 426). Its affinities are tentatively suggested in Fig. 9. Comparative measurements (in mm.) Е. rhinocerinus C. yoderensis C. dakotensis C. mitis М; length 15.4 16.0 21.5 26.0 width 10.0 13.0 15.5 17.5 M: length 17.2 24.0 27.0 ?28.0 width 10.3 14.6 16.5 ?19.0 М; length D. 24.1 26.3 31.0 width 10.5 15.4 16.5 ?19.0 SUBHYRACODON WOODI! sp. nov. Type. Right mandible fragment with DP, and Муз. M. C. Z. No. 2,096. Collected by E. M. Schlaikjer, 1931. Horizon and locality. Basal Oligocene. Yoder formation. Goshen County, Wyoming. Specific characters. Smallest known Subhyracodon. Ра molariform, although the endoconid cingulum is stronger and extends farther for- ward, it does not completely block off the inner end of the talonid valley, however. The antero-internal cingulum is slightly stronger on P, than on the molars but its union with the anterior base of the metaconid is as feebly developed. Anterior and posterior cingula ! Named in honor of Dr. Horace Elmer Wood, 2nd. BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY C. dakotensis = d e E ж То fa C. mitis other species ?C. premitis og З e$ Q (6) еі H © 5 2 Caenopus yoderensis Subhyracodon woodi o Я Ф mI 5 2 | A == о A ү © Hom | ; Eotrigonias rhinocerinus ms Б rhinoceroses. Fra. 9. Diagram showing the affinities of some of the early Oligocene Schlaikjer 1933. SCHLAIKJER: A NEW BASAL OLIGOCENE FORMATION 89 strong. External cingulum weak on P, and absent (?) on the molars. Мі and М» subequal in size. Premolar series short with P; about one- half (measurements on alveoli) the length of Р». Symphysis shallow (See plate 8). Discussion. The inferior tooth formula, as known, is П--, CO, P4, МЗ. The symphysis contains the inferior portion of a fairly large alveolus which is probably that of Is and its position in the jaw indi- cates that that tooth was somewhat procumbent. Of the known species of Subhyracodon, S. woodi is nearest to S. copei and it possesses no characters which would eliminate it as ancestor to all the known species. Its stratigraphic occurrence, of course, is also correct for such а position. This suggested phylogeny is expressed in Fig. 9. Measurements mm ПРОДА or on УБОП 9.0 Шеп о a ООСО e s р Nc IU ee 16.0 еше or Ps on. alveoli(estimated).. 2.22... 19.5 о о. 24.0 Кен О 28.0 ане рк о etu d cr ts 19.5 күни a и e C и с? 30.0 vicc n о M E т 20.0 Length of М; FE E Е AR CD HC МОИМ ось 31.0 Order ARTIODACTYLA Suborder PECORA Family HYPERTRAGULIDAE LEPTOMERYX YODERI SP. NOV. Type. Left M? and associated left lower ramus with Р, М» М. C. Z. No. 2,095. Collected by E. M. Schlaikjer, 1931. Horizon and locality. Basal Oligocene. Yoder formation. Goshen County, Wyoming. Specific Characters. Paracone- metacone crest straight. Mesostyle rather feeble. External rib on metacone completely lacking. Median internal cingule on upper molar small and shared by the bases of both the protocone and metaconule. A weak anterior cingulum. Мз with 90 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY interior and external сизрз of the heel almost completely confluent. М» considerably larger than М). (See Fig. 10). Discussion. Seven species of Герютегух have been previously de- scribed: four from the lower Oligocene, two from the upper Oligocene or lower Miocene and one, L. evans?, has been distinguished from the array of specimens collected from the middle Oligocene or Oreodon beds. Three of these species, L. esulcatus — mammifer — semicinctus were established by Cope on one, two and three isolated teeth re- spectively. L. speciosus Lambe was also described from isolated teeth жае Ета. 10. Leptomeryx yoderi sp. nov. Crown and anterior views of left M? and left P; to Мз. Natural size. and Matthew (1926, p. 4) casts doubt on the validity of L. transmon- tanus Douglass, the type material of which consists of upper cheek teeth and part of the skull. It seems, therefore, that whether or not these species are valid, can be determined only by the discovery of more complete specimens, especially in the type localities and by a more thorough study of the known Leptomeryx material. The species described above is the earliest Leptomeryx recorded and in several respects, such as the proportionately small size of M4 and the less complicated lower premolars, it is less specialized than the other species. P3-4 are less complicated especially in that the postero- internal branch from the main cusp of P, is a simple straight crest, and that the branching of the posterior crest of Рз is short and open. In size this specimen is considerably larger than L. evansi Leidy, L. esulcatus Cope and L. transmontanus Douglass. It is slightly larger than L. speciosus Lambe and slightly smaller than L. mammifer Cope. SCHLAIKJER: A NEW BASAL OLIGOCENE FORMATION 91 It is somewhat smaller than L. obliquidens Lull and much smaller than L. semicinctus Cope which is the largest of all the described species. Several skeletal parts are preserved. They are, the distal end of the left radius, the proximal end of the fused metatarsals III and IV and the right caleaneum and astragulus. None of these skeletal ele- ments possess characters which may be regarded as distinctive of this species, unless their proportionately small size may be considered аз such. Measurements mm. Сеше елеш. ОМ. о ee GS 8.6 Greatest ан РИМ нон oculis A к OE. ш 10.4 Distance from М-М; on alveoli (center)................... 27:9 Greatest Тате ONNE us Мата ed 6.3 a Ob 222.00.227. 5.0 Greatest [erecto Nery. ла за sie Mel инн aaa 7.6 ТАСА Оо nr жий нек e 5.5 Greatest length oi Mes аео лоне 115 ШЕН ИНТ е 5.0 Greatest ЈЕО ОВИ уыш уузу л Л" 7.0 О Or асылда ee ae LS 3.5 Greatest ento m o e ЖИЛЫ E 6.4 CONCLUSIONS l. To a series of fluviatile deposits in Goshen Hole resting with unconformity on the Cretaceous and overlain with conformity or disconformity by the lower Oligocene Chadron formation is applied the new formational name Yoder. 2. The Yoder formation represents the earliest Oligocene deposition known in the Great Plains region and because of stratigraphic and faunal evidence its age is regarded as post-Duchesne and pre-Chadron. 3. The species which are described as new from the Yoder are as follows: Miacis matthewi, Hemipsalodon cooki, Ateleodon osborni (a new genus), Hyracodon ischyrolophus, Caenopus yoderensis, Subhyra- codon woodi and Leptomeryx yoderi. á BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY BIBLIOGRAPH Y Corr, E. D. 1885. 'The White River beds of Swift Current River, Northwest Terri- tory. Amer. Nat., 19, pp. 163; and Geol. and Nat. Hist. Surv. of Can- ada, 1, n.s., part C and appendix I, pp. 80. 1891. On vertebrata from the Tertiary and Cretaceous rocks of the Northwest Territory. 1 — The species from the Oligocene or lower Miocene beds of the Cypress Hills. Сео]. Surv. of Canada, contribu- tions to Canadian paleont., 3 (quarto), pp. 1-25, pls. 1-14. Grecory, W. К. and Соок, Н. J. 1928. New material for the study of evolution. A series of primitive rhinoceros skulls (Trigonias) from the lower Oligocene of Colorado. Proc. Colo. Mus. Nat. Hist., 8, No. 1, pp. 1-32, pls. 6, 5 text figs. LAMBE, L. M. 1908. The vertebrata of the Oligocene of the Cypress Hills, Saskatche- wan. Geol. Surv. of Canada, contributions to Canadian paleont., 3 (quarto), part 4, pp. 1-64, pls. 1-8, 13 text figures. MarrHEW, W. D. 1901. Additional observations on the Creodonta. Bull, Amer. Mus. Nat. Hist., 14, pp. 1-38, 17 text figures. 1909. The Carnivora and Insectivora of the Bridger Basin, Middle Eocene. Mem. Amer. Mus. Nat. Hist., 9, part 4, pp. 291-567, pls. 11, 118 text figures. 1926. On а new primitive deer and two Traguloid genera from the lower Miocene of Nebraska. Amer. Mus. Novitates, No. 215, pp. 1-8, 3 text figures. OsBORN, H. F. 1918. Equidae of the Oligocene, Miocene and Pliocene of North America. Iconographic revision. Mem. Amer. Mus. Nat. Hist., n.s., 2, pp. 1-217, pls. 1-54, 173 text figures. 1929. The titanotheres of ancient Wyoming, Dakota and Nebraska. U. S. Geol. Surv. monograph 55, pts. 1 and 2, pp. 1-953, pls. 1-236 and 797 text figures. PETERSON, O. A. 1919. Report upon the material discovered in the upper Eocene of the Uinta Basin by Earl Douglas in the years 1908-1909, and by O. A. Peterson in 1912. Ann. Carn. Mus., 12, pp. 141-168, pls. 34-47, 19 text figures. 1931. The Upper Uinta formation of northeastern Utah. Ann. Carn. Mus., 20, pp. 293-806, pls. 9-11. SCHLAIKJER: A NEW BASAL OLIGOCENE FORMATION 93 1931. New species from the Oligocene of the Uinta. Ann. Carn. Mus., 21, pp. 61-78, pl. 1, 12 text figures. 1931. New species of the genus Teleodus from the Upper Uinta of northeastern Utah. Ann. Carn. Mus., 20, pp. 307-312, pls. 12-15, 2 text figures. Sımpson, С. С. 1933. Glossary and correlation charts of North American Tertiary mammal-bearing formations. Bull. Amer. Mus. Nat. Hist., 67, article 3, pp. 79-121, 8 text figures. SINCLAIR, W. J. 1922. Hyracodons from the Big Badlands of South Dakota. Proc. Amer. Phil. Soc., 61, pp. 65-79, 8 text figures. ТНОЕРЕ, М. В. 1922. A new genus ої Oligocene Hyaenodontidae. Amer. Jour. Sci., 8, pp. 277-287, 2 text figures. Woop, Н. E. 1926. Hyracodon petersoni, a new cursorial rhinoceros from the lower Oligocene. Ann. Carn. Mus., 16, pp. 315-318, pl. 26, 1 text figure. 1927. Some early Tertiary Rhinoceroses and Hyracodonts. Bull. Amer. Paleont., 13, No. 50, pp. 1-89, pls. 17. EXPLANATION OF PLATES | | | PLATE 1 4 | | | | SCHLAIKJER—Oligocene Formation PLATE 1 Fig. 1. Type locality of the Yoder formation. East bank of the Fort Laramie Canal. Fig. 2. Block of coarse sandstone containing type jaw of Subhyracodon | woodi. Photographs by Erich M. Schlaikjer BULL. MUS. COMP. ZOOL. SCHLAIKJER. OLIGOCENE FORMATION. PLATE 1 PLATE 2 ScHLAIKJER—Oligocene Formation PLATE 2 ATELEODON OSBORNI Schlaikjer Type. Anterior part of right mandible. Natural size. BULL. MUS. COMP. ZOÖL. SCHLAIKJER. OLIGOCENE FORMATION. PLATE 2 me —À— PLATE 3 SCHLAIKJER—Oligocene Formation PLATE 3 ATELEODON OSBORNI Schlaikjer D Туре. Superior view of the symphysial region, and left M». Natural size. | | | BULL. MUS. COMP. ZOÖL. SCHLAIKJER. OLIGOCENE FORMATION. PLATE 3 PLATE 4 ScHLAIKSER—Oligocene Formation PLATE 4 ATELEODON OSBORNI Schlaikjer Paratype, Anterior part of right mandible containing Mi. Young. Natural size. | | BULL. MUS. COMP. ZOOL. SCHLAIKJER. OLIGOCENE FORMATION. PLATE 4 | | | a PLATE 5 SCHLAIKJER—Oligocene Formation PLATE 5 Fig. 1. Ateleodon osborni. Paratype. Superior view of anterior part of right mandible containing Mi. Natural size. Fig. 2. Brontops ? brachycephalus (Osborn). Superior view of anterior part of left mandible containing Ра and Mi, and right М». Two-thirds natural size. BULL. MUS. COMP. 7001. SCHLAIKJER. OLIGOCENE FORMATION. PLATE 5 PLATE 6 SCHLAIKSER—Oligocene Formation PLATE 6 BRONTOPS ? BRACHYCEPHALUS Osborn External view of anterior part of left mandible containing P4 and Mi. Natural size. SCHLAIKJER. OLIGOCENE FORMATION. PLATE 6 BULL. MUS. COMP. ZOÖL. E РЯ POET nd а PLATE 7 ScnrArkjER— Oligocene Formation PLATE 7 CAENOPUS YODERENSIS Schlaikjer Type. Superior and internal views of left mandible containing Mis. Natural size. BULL. MUS. СОМР. ZOÖL. ScHLAIKJER. OLIGOCENE FORMATION. PLATE 7 PLATE 8 ScHLAIKJER—Oligocene Formation PLATE 8 SUBHYRACODON woopt Schlaikjer Fig. 1. Superior view of right mandible containing Ра and Mis. Fig. 2. Internal view of the same, showing P4 not yet erupted. Both figures two-thirds natural size Plates 2-8, photographs by George Nelson SCHLAIKJER. OLIGOCENE FORMATION. PLATE 8 BULL. MUS. COMP. 2001. А. RAAV 1 n 102A MAY 1 0 1938 3129 Bulletin of the Museum of Comparative Zoólogy AT HARVARD COLLEGE Vor. LXXVI, No. 4 CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY OF THE GOSHEN HOLE AREA, WYOMING IV NEW VERTEBRATES AND THE STRATIGRAPHY OF THE OLIGOCENE AND EARLY MIOCENE By Erich M. ScHLAIKJER Wits Forty-one PLATES CAMBRIDGE, MASS., U.S. A. PRINTED FOR THE MUSEUM May, 1935 е PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE There have been published of the BurLzTIN Vols. I to LXV, LXVII-LXXV, LXXVII, of the Memoıss Vols. 1 to LII, LIV. The BurLETIN and Мкмогиз are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collections and Exploration. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately. A price list of the publications of the Museum will be sent on application to the Director of the Museum of Comparative Zoology: Cambridge, Massachusetts. Bulletin of the Museum of Comparative Zoölogy AT HARVARD COLLEGE Vor. LXXVI, No. 4 CONTRIBUTIONS TO THE STRATIGRAPHY AND PALAEONTOLOGY OF THE GOSHEN HOLE AREA, WYOMING IV NEW VERTEBRATES AND THE STRATIGRAPHY OF THE OLIGOCENE AND EARLY MIOCENE By Енісн M. SCHLAIKJER Уйтн Forty-one PLATES CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM Max, 1935 | | Хо. 4. — Contributions to the Stratigraphy and Palaeontology of the Goshen Hole Area, Wyoming IV New Vertebrates and the Stratigraphy of the Oligocene and early Miocene By Евсн М. SCHLAIKJER INTRODUCTION During the field seasons from 1930 to 1933 the writer carried out a geological and palaeontological investigation of the Goshen Hole Area in southeastern Wyoming for the Museum of Comparative Zoólogy at Harvard College. The partial results of this research have appeared in a series of articles dealing with: (1) the description of а new species of Apternodus and a study of the zalambdodont insec- tivores; (2) the Cretaceous formations, giving especial reference to a new member of the Lance, the description of a new Triceratops and the discussion of the phylogeny of the genus; and, (3) a new basal Oligocene formation,— the Yoder and its fauna, most of which was new. The present paper is the fourth and final part of these contribu- tions and comprises a consideration of the Chadron, Brule and lower Miocene formations with a discussion of the faunae and descriptions of many new forms. The writer wishes to acknowledge his sincere appreciation to Pro- fessor William К. Gregory for assistance and advice throughout this research. Acknowledgment is here made of the aid received from Dr. Thomas Barbour and the many courtesies extended by the Mu- seum of Comparative Zoölogy. The Columbia University geology faculty, particularly Professors Charles P. Berkey and Douglass W. Johnson and Doctors H. N. Coryell and G. M. Kay have offered many valuable suggestions and criticisms. The encouragement of Pro- fessor Charles C. Mook of Brooklyn College and his many suggestions In the field during the summer of 1931 have been most welcome. To rofessor Henry Fairfield Osborn and to the other members of the department of vertebrate palaeontology at the American Museum of atural History the writer is indebted for the privilege of studying the Specimens in that institution. The friendly attitude shown the Writer and his associates by the citizens of Goshen County, Wyoming, аз made the field work especially pleasant. 98 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY DESCRIPTIVE GEOLOGY General Features Exposures of the Pierre, Lance and Yoder formations of the Goshen Hole Area have been fully described previously. The remaining out- crops in the interior lowlands are mostly of the Chadron formation which is composed of green, yellow and reddish clays and sandstones. Except for an occasional promontory such as Red Bill Point and a few other more or less sinuous mesa-like expressions, which result from the impeded erosion of the resistant upper Chadron channel sandstones that cap the softer clays, the relief of Goshen Hole is low and gently rolling. Surrounding this vast topographic basin is the almost con- tinuous escarpment which is from four hundred to a thousand feet in height. Exposed along the face of this escarpment are the soft pinkish Brule clays that form the less steep slopes near the base on which is developed a miniature bad-land topography, and the resistant gray sands and sandstones of the lower Harrison formation on which the steep upper limits are developed. The Harrison sediments are rather pervious and the Brule clays are relatively impervious, hence the con- tact of these two formations determines the most important water table of the uplands. This contact is disconformable in some places and unconformable in others. It is, therefore, quite irregular and fre- quently is inclined towards the face of the escarpment which results in the occurrence of many springs. This seepage results in the rapid disintegration of the clays that are soon washed or blown away. In this way the overlying more resistant sands and sandstones are under- mined and the retreat of the escarpment is greatly enhanced. The three hundred and seventy-five square miles of high upland table land is composed of the Harrison formation. Much of this formation is covered with alluvium but where exposures occur erosion has carved it into a bizarre relief mainly because of the variability of resistant and soft layers. A characteristic feature of these beds is the presence of a millenium of well-like excavations of all sizes from two to ten feet in diameter and three to fifteen feet deep. Apparently these are caused by wind action. The drainage system of the area consists of the North Platte River with its main tributary the Laramie River and two tributaries of lesser importance,— Cherry Creek and Bear Creek. Along the North Platte River extensive terraces occur which are frequently from fifty to one hundred feet above the present level of the river. On these terraces | i | | | SCHLAIKJER: NEW FOSSIL VERTEBRATES 99 are rather thick accumulations of sands and gravel which, as is shown by the presence of Archidiskodon imperator! and Equus, are Pleistocene m age. The Laramie River flows across the northwestern corner of the area. Where it crosses the Whalen Fault it has carved a canyon, the south wall of which is four hundred feet high. Deer and Six Mile Creeks are its tributaries which drain the“Military Flats” country just to the south of Old Fort Laramie. Their southwest-northeast trend has been determined, as has also the head of Cherry Creek, by structural conditions in the vicinity of the Whalen Fault. Horse Creek flows in from the southwest, turns abruptly and flows northward between Sixty Six and Bear Creek Mountains. G. I. Adams (1902, p. 25) has sug- gested that the head waters of Horse Creek used to flow eastward into Pumpkin Creek of Banner County, Nebraska before they were captured and diverted to the north by a stream in Goshen Hole that worked its way headward through the Bear Creek-Sixty Six Mountain gap. The Goshen Hole Area presents a number of interesting physiographic problems which are intricately related to adjacent areas and an inter- pretation of which will involve a consideration of most of the region traversed by the North Platte River. This is beyond the scope of the present study. The Topographic Map The topographic map presented as Plate 1 is mainly a compilation from the United States Geological Survey topographic sheets of the Hartville and Goshen Quadrangles. Two hundred and fifty-two square miles in the southwest portion of the area, including townships 20N- 23N of Range 65W and all but a mile and a half strip along the east Side of townships 20N-23N of Range 64W, were mapped by the writer and his accomplished field assistant Mr. David Bradley Cheek during the summer of 1933. The twenty foot contour lines of the section taken from the Hartville Quadrangle were interpolated from the fifty foot contour interval of that sheet. The Geologie Map A geologie map including an eastern strip of the Goshen Hole Area Was published by Mr. G. I. Adams in his “Geology and water resources 1 An excellent skull and some skeletal parts of this species were collected near the town of ngle on the north side of the river, and are now in the University of Nebraska Museum. Li 100 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY of the Patrick and Goshen Hole Quadrangles in eastern Wyoming and western Nebraska" (1902). 'This map was the result of recon- naissance work and illustrates only the general distribution of the formations. The northwestern corner of the area was mapped by Darton in the Hartville Folio (1903). At that early date a portion of the Laramie River vicinity was quite inaccessible and Darton mapped it as lower Miocene and missed the Whalen Fault and approximately two townships of Brule exposures immediately to the west of it. About fifteen square miles of this Втше is included in the area under con- , sideration. Most of the geologie mapping was done by the writer during the field seasons of 1930 and 1931. The southwestern portion was mapped during the summer of 1933 at which time a topographical survey was made of that vicinity. Included in the accompanying map (see Plate 2) are the Cretaceous and Yoder formations, maps of which have pre- viously appeared in parts two and three of these Contributions. For completeness these formations have been included in the present map. Because of lack of sufficient evidence no attempt has been made to differentiate on the map the upper Harrison at the heads of Deer and Cherry Creeks from the lower Harrison. Only along the major streams; the North Platte and Laramie Rivers, has the alluvium been indicated. 'The major topographie features have not been shown since a topo- graphie map (see Plate 1), drawn to the same scale, is published here- with. The towns, railroads, canals, principal roads and a few ranches are shown. Stratigraphy The formations considered in the present discussion, are Oligocene and Miocene in age. The oldest of these! is the Chadron which is the most limited in extent. It has a maximum thickness of eighty feet and is absent in many places, which results in the next succeeding forma- tion, the Brule, resting on the Lance of the Cretaceous. The Brule, typically a flesh-colored clay, is fully represented in the area and 18 normally about four hundred feet in thickness. The earliest Miocene formation is the lower Harrison which rests with disconformity and in places with unconformity on the Brule. It is typically a light to gray rather unconsolidated sand with numerous sandstone concretions and some channel deposits, and has a maximum thickness of about four hundred feet. In the Deer Creek and Cherry section of the area 1 See part III of these “Contributions, etc.” for a description of the Yoder formation,—the oldest Oligocene deposit known in the Great Plains. | | ово —e——C"€———M——— —— ORO O TORO TO тыда тесе оны і SCHLAIKJER: NEW FOSSIL VERTEBRATES 101 there are approximately two hundred feet of somewhat lighter and more consolidated sands which contain an abundant amount of plant remains, and which lie above the typical lower Harrison beds. Strati- graphically the change from the lower Harrison to these strata is gradational although on the basis of the fauna they are correlative with the upper Harrison of northwestern Nebraska. The following 18 а detailed discussion of each of these formations. Chadron. "Three main exposures of the Chadron formation, having а northeast southwest trend, are present in Goshen Hole. The south- easternmost of these extends from the vicinity of Lyman, Nebraska to the northwest corner of Sixty Six Mountain, immediately west of which, across Horse Creek, is an isolated outerop of about two square miles. The second of these exposures occurs around Lyman and Huntley, extends westward to the town of Yoder from where it con- tinues southwestward in belt-like fashion between the Corn Creek Anticline on the northwest.and the Hawk Springs Arch on the south- east. A narrow strip from Yoder to Veteran connects this belt with the northwestern and most extensive exposure, which is of continuous extent from the North Platte River up Cherry Creek southwestward, and which occupies the greater portion of the interior lowlands of the western part of Goshen Hole. The Chadron deposits are composed of grayish, reddish and brownish green clays, which are quite sandy in places; and fine to coarse fairly unconsolidated sands and hard sandstones. The latter are most usual 1n the upper limits of the beds and, capping the softer clays, frequently form mesa-like prominences which are most conspicuously represented four miles southwest of the Harvard Fossil Reserve, at Red Bill Point (see Plate 3), northwest of Reichstadt's Ranch and in the Cherry Creek vicinity southeast of Doty's Ranch. Many of these, some as great as a' mile in length, show the original meandering southwest- northeast course of the Chadron river with an occasional tributary. About two miles northeast of Veteran the main stream channel bends eastward and can be traced by intermittent outcrops to the south bank of the Platte, two miles southwest of Henry, Nebraska, where the upper channel sands and sandstone are more than twenty feet in thickness. Another, somewhat thimner upper channel sandstone out- crops from under the Brule clay just east of Lingle on the north side of the river. This probably is the deposit of a tributary that entered the main Chadron stream from the northwest, for it can be traced along the south side of the Platte to west of Henry where it merges With the more extensive channel sandstone. 102 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY 'The Chadron has a maximum thickness of eighty feet. It is com- monly thinner; and in places, particularly in the southern part of the area, has been entirely removed from the underlying Lance formation before Brule deposition. Usually it rests with unconformity or dis- conformity upon the Lance and where the Yoder formation intervenes it is either conformable or disconformable with those beds. The Chad- ron deposits are easily recognizable not only by their variegated clays and green sandstones but also by the almost constant presence of titanothere remains. In general they are clearly demarked from the overlying pinkish Brule clays but in some localities the contact is not so distinct. Darton (1903, p. 3) recognized this in his comments on the Chadron southwest of Doty's ranch, and the uppermost twenty- foot sandstone in his section 1, on the basis of faunal and stratigraphic evidence, is unquestionably the same channel deposit as the one at the 4240-4265 foot level in the Harvard Fossil Reserve (see beyond under Brule). 'The contact seems least discernable, however, where the upper- most Chadron channel sandstone is absent, for there was little differ- ence in flood plain deposition at the end of Chadron and the beginning of Brule times. Where such deposition took place the formations seem almost conformable except for the slightly darker greenish tinge of the Chadron clay which also, because of its somewhat more uniform texture and composition, weathers into a rounded hillock topography rather than the characteristic "pinnacle" topography of the Brule. A typical example of this may be seen in the S.W.14. Sec. 36. T. 24N. R. 62W, two miles southwest of the Harvard Fossil Reserve. In this locality titanothere remains are abundant in the Chadron clays (two skulls were collected here by the Colorado Museum party). Perhaps the best localities for studying the Chadron-Brule relation- ship are to be found along the south side of the North Platte River. Three miles northwest of Torrington in the S.E. М. Sec. 12. T. 24N. В. 62W. the following section was measured: F'eet above Feet | Irrigation Ditch Typical pinkish clay, somewhat sandy........... 22 53 "Topical PINKIE а sen eds ES 16 37 Very sandy clay which becomes less sandy laterally 10 27 Hard gray cross-bedded sandstone with some un- Consolidated sanai о aan 17 10 Brownman preci! Clavie иеа ta Al es 10 0 Concealed under irrigation ditch................. SCHLAIKJER: NEW FOSSIL VERTEBRATES 108 Titanothere bones occur in the seventeen-foot sandstone and this Stratum is unquestionably the top of the Chadron. The succeeding ten feet of very sandy clays grade laterally into less sandy pinkish typical Brule clay which rest on greenish clays that grade into the channel Sandstone. Four miles south-southeast of this locality in the N.W.14. Sec. 32. T. 24N. В. 6IW. the following sequence occurs at the bottom of the canyon extending northwest from the Harvard quarry: Feet below Feet | Upper Sandstone Dark gray consolidated “ribbon” sandstone. .... 20-25 Pisos se 5 0 Hard sandy pinkish сјау..................... 15 5 Dark gray consolidated channel sandstone. . .... 8-10 20 Green Оду en ne Neu E US 1% 80 Sandy green Gleis t cei о EN 1 31% наты ДЕНО By mere UM p C моз a el 2% 32% Dark clay with purplish tinge................. 5 35 Cream colored clay which erodes into low rounded ШОР у ee у E NM E 26 40 Pinkish clay with green streaks................ 3 66 Dark green and yellowish ОБУ. а > 2% 69 Dark green channel sandstone................. 7 71% во windy діду уу л хс... с. 5 78% Dark green CLAY fa, c РИТМ ME NCC 41% 83% Yellowish dark green яһау clay............... 10% 88 Concealed by alluvinm а In this section the 8-10 foot dark gray consolidated channel sandstone marks the upper limit of the Chadron. Titanothere fragments were found at this level and the overlying pinkish clay is distinctly Brule. A mile to the northeast this channel sandstone outcrops along the Cheyenne-Torrington highway about three hundred yards south of the Union Pacific Railroad tracks. From there the channel takes a Southeastward course and is concealed under the Brule for about Seven miles. Immediately to the north of this tributary a lake must have existed on the flood plain during Chadron times, for in some localities a thick cherty fresh water limestone caps the Chadron clays. Such is to be found two miles south and a mile east of Torrington 104 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY (№. W.14. Sec. 27. Т. 24N. В. 61W.) where the following fairly thick section is measurable: Feet below Feet Upper Sandstone Gray finely bedded coarse to fine consolidated sand- BODE cu d RU c REY л D АСЫҒЫН 15 OUAIS DINK ОД ҮН lo УО E НУ 22 0 Freshwater cherty limestone.................... 2 22 Greenek sindy ОТАУВИ en 40 24 Cherty freshwater limestone: :.......... 00.05... 2 64 Greenish sandy Ау о. 12 66 Е Le у. cus s ОИ Since the lower clays are markedly distinct in their greenish Chadron color, the upper two-foot limestone probably indicates the top of that formation. Three miles to the east this limestone is six feet in thick- ness. Up in the Brule, twenty-two feet above this limestone is a four- foot sandstone which is stratigraphically and lithologically the same as the twenty to twenty-five foot sandstone above the Chadron at the Harvard Fossil Reserve (see Figure 1). Two miles southwest of Henry, Nebraska, the upper channel is more distinct. A section in the S. W.14. Sec. 4. T.23N. R.60W. is as follows: Feet above Feet N. Platte River Light to dark greenish gray intricately cross-bedded pande And ANISTONE, I Los Лу ЫРА ТӨ 21 43 ear alta O Al 13 30 Brown hard clay with greenish streak............ 30 0 Concealed Буте ee ИЛГЕР ИО REOR SA The lower thirty feet of brown clay weathers to a cream color with а purplish tinge and it erodes, as well as the succeeding thirteen feet, into smooth rounded knolls. In the upper channel deposit occasional clay conglomerates are present and greenish clay lenses are frequent. Titanothere remains abound. In Figure 1 the above described sections are graphically shown. At the locality west of Torrington the elevation of the Chadron-Brule contact is 4160 feet. Four miles southeast, at the Harvard quarry it is 4220. This slight dip to the northwest is the result of minor S£4Sec/ZTEANREW NWESESZTZANREN NWESee2ETZINREIN = SWESecSUE MARGO SW4 Sec ПОЗИК 5 4 Gravel Finely bedded sandstone |95592) Pinkish clay “| Channel sandstone Titanolheres Pinkish sandy clay, Pinkish су Greenish clay 4/60 Greenish sandy сву Channel sandstone Green channel sandshre / Titanotheres ie Scale Limestone E Greenish clay T4 3 Dah ы Light уау limy clay 0 $ EMS 1933 ; Ета. 1. Columnar sections of the Chadron in the North Platte River vieinity showing lithology, correlation and Brule relationship. Schlaikjer 1933. SALYYHIHLYYA TISSOA MAN (ЧУГСМІУТНОЗ з. 4 9 106 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY arching in this vicinity which is part of the Corn Creek anticlinal structure. East of the Harvard quarry at the localities given above, the elevations of the contacts are 4180, 4148 and 4120. In other words, there is a hundred-foot drop in a little more than nine miles, or a dip of approximately ten feet per mile. This is the normal inclination eastward from the Rocky Mountains of the sediments in this section of the Great Plains. Brule. Exposures of the Brule have a greater distribution than those of any other formation in the area. The conspicuous badland topog- raphy everywhere to be seen at the base of the Goshen Hole Escarp- ment is developed on this formation and it constitutes the occasional table lands, such as those along the Wyoming-Nebraska state line as well as the uplands in the vicinity of Sixty Six Mountain, and im- mediately west and to the south of Torrington. It consists of: pinkish, flesh-colored relatively hard clays which are somewhat sandy at the base and at the top of the formation; occasional rather extensive cherty limestone lake-deposits; numerous beds of almost pure volcanic ash; and, thick channel sands, sandstones and conglomerates. It rests with disconformity (probably with conformity in places) on the Chadron, and is overlain by the lower Harrison with disconformity and unconformity. The whole of the Brule is represented in the area and has a maximum thickness of four hundred and thirty feet. Not infrequently it is much thinner for, as the result of local uplift, some- times more than half of the beds (as at Y. B. O. Pass) have been removed before deposition of the lower Harrison began. А generalized sequence of the most important phases of the Brule in the area is as follows: Feet Pinkish, flesh colored rather hard clay becoming sandy in its upper ОООО 90 River channel deposit, coarse boulder conglomerate................ 15 Pale pinkish clay with minor channel deposits, limestone lenses and volcan c о здо IS a hice, a A, 285 СТАУ channel ята ОПЕ, ал QUEDA IRL TUR ATE 20 Pinkish clay, sandy towards bottom at Chadron contact............ 20 Total aces о Da NU C palette беу д ead 430 The relationship of the Brule to the Chadron has been discussed above. The Brule clays which immediately overlie the Chadron are almost always sandy at the bottom. The lower channel sandstone is not everywhere present for it thins laterally and grades into clays. It is SCHLAIKJER: NEW FOSSIL VERTEBRATES 107 exposed in its maximum thickness at the Harvard Fossil Reserve and its differentiation from the upper Chadron channel is admirably shown along the road four miles north of Yoder in Sections 9 and 16. Fossils are rare in this channel deposit. Although scraps of bone and teeth could not be generically identified with certainty, they indicate the presence of a perissodactyl about the size of Metamynodon. For this reason and because of their stratigraphic position and relationship to the overlying strata of which the exact level is known, they are ten- tatively correlated with the Metamynodon sandstone of the South Dakota Badlands. The general trend of the Brule stream which this channel represents was from west to east and it occupied the same geographic position as did the main Chadron stream before it. It can be traced across Goshen Hole from four miles south of Doty's ranch, where it outcrops from under higher beds, eastward to north of Yoder, to the Harvard Fossil Reserve, and southeast to Rattlesnake Hill two miles west of Lyman where, in the М.М.М. Sec. 33. T.23N. R.60W. near the bottom of a canyon, it is composed of approximately twenty- five feet of intricately eross-bedded, unconsolidated gray sands with some clay pebbles. The west-east trend of this former stream is further Substantiated by the fact that in most of the eross-bedding the foreset beds are to the east. The lithology of the next succeeding two hundred and eighty-five feet of pale, pinkish, fairly resistant clay is relatively constant through- out the area. A characteristic feature of this phase is the presence of at least three distinct beds of rather pure volcanic ash. The lower two of these beds, although lenticular, are widely distributed in the eastern part of the area and are best shown on the east side of Table Moun- tain. The lowest bed occurs about fifty feet above the Metamynodon sandstone and attains a thickness of six feet (see Plate 3, Figure 2). Twenty-seven feet higher (or seventy-seven feet above the Metamy- nodon sandstone) is the second layer which is a half-foot thick. This ed can be correlated with the two-foot layer which occurs thirty miles to the west in the western part of Goshen Hole. In this vicinity it is approximately eighty-four feet above the Metamynodon sandstone. Eighty feet higher (or two hundred and four feet above the Chadron) 18 another layer of ash only one foot thick. This is probably equivalent to the layer that is at the top of the Brule at Y.B.O. Pass where about half of that formation has been eroded away. Another feature of this middle phase of the Brule is the frequency of lenticular fresh- Water cherty limestone deposits. These are most significantly promi- nent along the low escarpment just south of Torrington,on Rattlesnake | | | | | | 108 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Hill and Table Mountain, and six miles south of Red-Bill Point. In most instances the lime has been almost completely replaced by silica. Frequently these lenses are in part composed of ball-like accumulations which are probably algal in origin. More rarely, they are entirely made up of small fresh water gastropods (now being studied by Dr. William Clinch). This phase of the Brule is richly fossiliferous at certain levels (see below for description of fossil beds). Fragments of Merycoidodon culbertsoni and an abundance of the turtle Stylemys nebrascensis together with the array of forms which occur in the Harvard Fossil Reserve, show that these deposits may be unques- tionably correlated with the oreodont beds of Nebraska and South Dakota. The upper channel conglomerate is best developed eight miles south- west of Torrington in the 5.Е 24. Sec. 19. T.24N. R.62W. (see Figure 1, p. 4 in Part One of these “Contributions, ete."). It is about fifteen feet thick and contains boulders of granite, pegmatite and gneiss as large as a foot in diameter. This undoubtedly marks the largest stream in the area towards the close of the Oligocene. Overlying this channel deposit to the northwest are ninety feet of pinkish, flesh-colored rather. hard clay, the top twenty-five feet of which are quite sandy. Teeth and skeletal fragments of Leptauchenia nitida occur in these clays which demonstrate that they are equivalent to the Leptauchenia beds in other areas. Two rich fossil deposits occur in the lower Brule formation. The stratigraphie position of these is shown in the following section measured in a canyon in the N.W.14. Sec.32. T.24N. R.61W. Feet below Feet Limestone Alluvium (4300 foot level at top)................. 5 Cherty limestone with many gastropods........... 2-5 Bolt Clay ШО ропе a. dl лу, на по Rei ca 2-5 0 Layer. OL DONC UE Сіз 20V. у, у Qu ao dp 1-3 2—5 Ligat РОЗ DA ОУ 10 5-6 Pinkish clay pebbles with fossils................... 1 15-16 АБ лантан ады а DL T LL cO EE алы 14 16-17 Dark gray consolidated “ribbon” sandstone........ 20-25 30-31 Pinkish elay ns О qur cR А НО 5 51—55 Hard.sandy pinkisli/.glay УВ: 15 66—70 Dark gray consolidated channel sandstone. Chadron. 8-10 | | | | SCHLAIKJER: NEW FOSSIL VERTEBRATES 109 The bottom part of this section has been discussed above. The twenty to twenty-five foot Metamynodon channel sandstone makes a sharp bend to the north about three hundred yards west of the Field Labo- ratory (see Figure 2). Its lateral equivalent is about twenty feet of un- raphical survey of a canyon in Section 32 of N and R 61 W of 6% Principal Meridian showing PALEONTOLOGICAL EXCAVATIONS - 1930 Topog Tas conduci MR. ERICH. M. SCHLAIKJER ano PARTY fer МИЗЕЦМ or COMPARATIVE ZOOLOGY farvard University Engineer a surveyor Phil Rouse, Torrington, Wyoming m Fig. 2. Мар of the Harvard Fossil Reserve. Dotted line shows out- Crop of fossil bed. consolidated sands and clays. Above these is a one-foot fossiliferous ауег which contains the rather water worn and broken remains mainly of Subhyracodon occidentale and М esohippus bairdi. Five miles to the 110 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY east in the N.W.14. Sec. 6. T.23N. R.60W. this fossil layer is again present exactly fourteen feet above the Metamynodon sandstone (the latter outcrops in the northeast corner of section 6). Неге the fossils, mostly the remains of Subhyracodon, are more abundant. The Colorado Museum party spent two field seasons collecting in this deposit. Succeeding this lower fossil level in the above section are ten feet of light pinkish hard clay overlain by an almost solid layer of fossils from one to three feet in thickness. This is the richest and most extensive single deposit of Oligocene fossils yet recorded. It outcrops continuously for approximately a half mile (see Figure 2 and Plate 4) and contains in extraordinary abundance the remains of Mesohippus, Pediohippus, Subhyracodon, Cynodictis, Archaeotherium, Poébrotherium and Hypisodus. Of these, the horses and rhinoceroses occur in greatest numbers. In addition, the deposit has yielded the largest collection of pre-Pleistocene fossil birds yet discovered in North America (see Wetmore, 1933). An occasional carapace of the turtle Stylemys neb- rascensis is also present. In thirty-two cubic feet of this deposit the complete or partial remains of 288 birds, 96 horses, 18 rhinoceroses, 8 dogs, 1 camel, 1 hypertragulid, 2 entelodonts and 2 turtles were ac- counted for. On the basis of these figures а conservative estimate shows that there are the remains of probably between twenty and forty million individuals in this deposit! The matrix in which the bones occur is a fine pinkish clay that contains a high percentage of calcium carbonate and volcanic ash. Throughout a large part of the deposit the clay is comparatively hard and is somewhat difficult to work. At the south excavation, however, (see Figure 2) the bones occur in soft clay. This excavation is located in about the center of the deposit. Above the bone-bearing stratum is a two to five-foot layer of soft clay which contains occasional bird and Cynodictis re- mains. Above this is a two to five-foot bed of white, cherty lime- stone, the upper limits of which are rich in the shells of small fresh- water gastropods. This limestone is thickest, as is also the underlying fossil deposit, where the south excavation is located. From there it this to the south and to the north. It continues eastward uninter- ruptedly for about three miles and to the west for an equal distance,— thinning, of course, in those directions. This limestone undoubtedly represents an accumulation in a fresh-water lake which must have been at least three miles wide and six to eight miles long. In its thinner portions plant remains occur in remarkable abundance in the form of roots and branches that vary from a quarter of an inch to two inches in diameter. | | | | | | | | | | SCHLAIKJER: NEW FOSSIL VERTEBRATES ЫШ In this vicinity marsh and lake conditions prevailed throughout the early part of the Brule. Here animals lived in prolific numbers. The amassment of their remains is, however, somewhat puzzling. All of the material is disarticulated, and many of the bones are broken, most of which probably occurred before fossilization. Since hardly a single water-worn bone has been found, it is not likely that this concentration is due to severe stream action. Probably the bones of those animals that died along the shores, as well as on the actual mud flats of this lake when its waters were more restricted during the drier seasons, were washed into the deeper portions during times of freshets and floods and thereby accumulated to form this rich deposit. The main part of this fossil deposit occurs in the south half of the N.W.14. Sec. 32. T.24N. R.61W. which tract of land has been pur- chased by Harvard College and is designated as the Harvard. Fossil Reserve. Lower Harrison. This formation constitutes the high upland table lands of the area and is exposed principally on Sixty Six Mountain (see Plate 7), Bear Creek Mountain, in the southwestern corner of the area and in the vicinity of Deer and Six Mile Creeks south of Old Fort Laramie. Its maximum thickness is approximately four hundred feet and it is mainly composed of light to dark gray unconsolidated sand and sandstones in the form of channels and concretions. (See Plate 6, Figure 2). This lithologie character is particularly true for the lower three hundred feet which are rather abundantly fossiliferous. The Upper one hundred feet is usually lighter in color, is less concretionary and is almost barren of fossils. The lower Harrison formation rests with disconformity and uncon- formity on the Brule and its lowest limits are generally marked by Channel deposits. The Brule-lower Harrison contact is widely shown and its interpretation presents one of the most interesting stratigraphic Problems in the area. Considering the relationship between these two formations, an outstanding feature is that the lower channel phase of the Harrison becomes more and more dominant from Sixty Six and Bear Creek Mountains northward to Y.B.O. Pass, Box Elder Creek, Wilson's Ranch, Cherry Creek and along the northern part of the Goshen Hole Escarpment. A series of sections covering these localities are given below. The following sequence was measured at the east end of Sixty Six ountain at the Nebraska-Wyoming State line in the S.E.14. Sec. 3. T.20N. R.60W. (See Figure 3). 112 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY Feet DEY ВАШ ВАЗ ОР А СРЕ АРВ DR PC 15 Gray Sands with Pipy сор отейопв, ыны de edis ЗА лу The dE 32 ОТОН О нен Eod karina qe IUE exa (Oe eee Pe Loses 114 Gray sands, slightly clayey with pipy concretions................-. 74 VO Cana Sal peut Cu a WAP а whe ape M 1% Finely bedded grayish brown sand with occasional pipy concretions.. 1026 ОТОН Bulls. nias gd aot En E edat a reo ал ае ани 4 Finely редеа erus нала” жмыс. каулы. rhe a o rada iat ночі 3 Gray to pinkish sandy clay (this marks the top of the Brule; elevation ОЛ o MS ҰЛЫН 1614 PR КАШЧУ ШАЙ ЗЫ ҒАРЫ Ы ТАНЫ С sq E S Mi UAE 10+ The sixteen and a half feet of gray to pinkish sandy clay probably belongs to the Brule. It is gradational into the typical pink sandy clay below and is sharply differentiated by a disconformity from the overlying gray sands of the Harrison. Four miles north at Eagle's Nest a duplicate sequence occurs except that the volcanie ash layers are thinner and the pipy concretions are less dominant. Five miles to the northeast at Signal Butte in Nebraska, the lowermost Harrison is composed of rather well stratified sands with only very thin sandstone layers. At the northwest end of Sixty Six Mountain in the Х.Е.М. Sec. 7. T.20N. R.60W. the typical Brule clay is overlain by a three- foot clay conglomerate above which are nineteen feet of clay sands that grade upward into the gray sands with pipy concretions. In the 5.1.14. Sec. 36. T.20N. R.62W. at the east end of Bear Creek Moun- tain the following section was measured. (See Figure 3). Feet Light gray sands few sandstone lenses. No fossils found............ 100+ Gray sands WUD ОУ бое му s dove на nds ы Ка не 128 Rather compact gray sands with many fossils such as Promerycochoerus, ano ra guias, Мевовубть ВО салтынан e ао 6 Gray sands with many pipy concretions and Daemoneliz in places. ... 77 Gray sands finely bedded and in places with aeolian cross-bedding.... 48 Gray sands with pipy concretions. Stenomylus hitchcocki at top...... 16 MOLA НА ил ee 1 Gray ваша with pipy CONGIULIONS а-ы Яп Р. DUI C S 7 Sandy clay with 6-10 foot layer of clay conglomerate at top......... 42 Typical fine grained Brule clay, somewhat валду................... 50+ There is an irregular contact at places between the forty-two-foot layer of sandy clay and the typical Brule clay beneath. At other places the two seem conformable. The six to ten-foot conglomerate | SCHLAIKJER: NEW FOSSIL VERTEBRATES 113 material is reworked Brule clay and is clearly demarked from the overlying gray Harrison sands. This is the only locality where Daemo- nelix specimens were observed. The occurrence of these together with the little camel Stenomylus hitchcocki proves conclusively that these lower Miocene beds are lower Harrison in age. The upper part of the one hundred and twenty-eight-foot bed is sparsely fossiliferous and the top one hundred feet are decidedly lighter in color with only a few sandstone lenses and apparently no fossils. These upper light colored beds have a rather uniform distribution throughout the area and are present wherever a thick section of the Miocene occurs, as at Cherry Creek and near Old Fort Laramie where the Miocene has been down faulted just to the east of the Whalen Fault and the higher levels have not been removed by erosion. Also, they always seem to be gradational into the gray concretionary beds below. For these reasons I am in- clined to include them, provisionally at least, in the lower Harrison. At the west end of Bear Creek Mountain the above section is du- plicated, except that resting on the Brule clay is a foot and a half of сіау conglomerate succeeded by thirteen feet of sandy clay above Which is a two-foot channel sandstone that is overlain by another foot and a half of clay conglomerates. Above the latter are sixteen feet of somewhat clayish sand that grades upward into gray concretionary sands (see Figure 3). These thirty-four feet of material probably represent the forty-two-foot lower bed in the section at the east end of the mountain. Immediately north of Y.B.O. Pass in the N.W.14. Sec. 5. T.21N. R.64W. the Brule-Miocene sequence is as follows: Feet Gray sands, numerous pipy concretions, some aeolian eross-bedding... 45+ Rather consolidated СТАНОВИ НОВ зорова Н ECT URGE ak 5 Gr ay sands with numerous pipy сопсгейопя....................... 36 Light to dark gray sands with many pipy concretions............... 69 И BER yaconsoliddtedisands a... ee 10 Coarse channel sands and boulders. ны муу 5 cahy оаа Sante. ш i САА, 10 Pinkish Brule Clay (elevation of contact 4980)... . 5... esos. 20 Volcanie mu M Tr quM ре c NI 2 О Biule clay лл лш лл a лл ОРН ~ 604- Here the Brule is only about two hundred feet thick,— approximately alf of its normal thickness. The contact with the Miocene is clearly a the bottom of the first ten-foot layer of rather coarse sands. Above ese the five-foot coarse channel sands, with boulders of gneiss, 114 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY granite, quartzite, etc., become less clastic at the top and grade upward into the overlying rather consolidated sands. The remainder of the section presents no distinctive features. Eight miles to the northwest in the N.E.14. Sec. 6. T.22N. R.65W. (three-quarters of a mile north- east of Wilson's Ranch), the channel phase immediately above the Brule is much thicker. In this locality the Brule is represented by practically its maximum thickness (elevation of contact with the Miocene is 5090 feet) and is rather sandy in its upper limits. Following the Brule is а coarse channel conglomerate twenty feet in thickness which is separated by a ten-foot bed of sandy clay from the overlying twenty-six feet of coarse channel conglomerates and sands. (See Plate 5, Figure 2). These channel beds evidently represent the course of the major stream in this area in late Oligocene and early Miocene times. They can be traced continually southwestward to the escarp- ments along Chugwater Creek where they sometimes attain a thick- ness of one hundred feet or more. Five miles due north of Wilson's Ranch and from there northeastward to Cherry Creek the lowermost Miocene sediments are fairly well bedded fine sands with thin sand- stone layers and small globular sandstone concretions as well as occasional beds of somewhat more clastie materials,— typical flood plain and alluvial flat deposits. Two miles north ої Doty's Ranch the channel beds again appear and are from thirty to fifty feet thick. They thin westward and eastward. In the N.W.14. Sec. 36. T.25N. R.63W. the deposit is thinner and less clastic. Its unconformable position on the Brule is shown in Plate 6, Figure 1. In this locality the full width,— almost four miles, of this Miocene river bed can be seen. Another channel sandstone and conglomerate, although of more limited extent, occurs in the northwestern corner of the area where it has been faulted to the surface and is exposed along the Laramie River and just south of Whalen Dam on the North Platte (see Plates 8 and 9). It too has a southwest-northeast trend and can be traced for a number of miles eastward on the north side of the North Platte River. Ап analysis of the above sections, several of which are illustrated in figure 3, reveals the following facts concerning the late Oligocene and early Miocene in the area: 1. Where the Brule is fully represented it becomes more and more sandy in its upper limits. This is probably the result of gradual uplift in the Rocky Mountains towards the close of the Oligocene which resulted in the transportation of more clastic sediments farther out onto the plains. SEHSA3TZONREOWN || SE Sc S6 ZA. NLZSEITTZONREEN MWESecSTEINREHW SWESETTZENKESW NWESESCTZIREIHN | ЗЯТУЧЯЧІЧОУЛ TISSOI MAN :WWfMIVIHOS EMS /935 STI Ета. 3. Columnar sections along the Goshen Hole Escarpment showing Brule-lower Harrison relationship, lithology and correlation. Schlaikjer 1933. BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY 116 2. The end of the Oligocene and the beginning of the Miocene were marked by rejuvenation of the drainage system caused by major uplift of the mountains to the west. This is shown by erosion and rede- position of Brule clay, the presence of channel conglomerates, the numerous volcanic ash layers and the clastic nature of the Miocene sediments. 3. The thick channel deposits in the northwestern portion of the area indicate the position of the main early Miocene river which had a southwest-northeast trend. 4. The thin layer of coarse sands with some conglomerate material at У.В.О. Pass are the deposits of a minor tributary to this major stream. 5. The clay conglomerates at the Bear Creek Mountain and Sixty Six Mountain localities represent relatively rapid accumulations formed by small streams. The material may have been derived, in part at least, from an erosion of the Brule in the Y.B.O. Pass vicinity where there was local uplift at that time. It can be concluded therefore, that the channel deposits at the base of the Miocene are synchronous. Upper Harrison. At the heads of Deer and Cherry Creeks there is a series of light to dark gray sands and sandstones which contain some lenticular beds of clay and which, at certain levels are made up of ' more than fifty percent of plant remains. The total thickness of these beds is between one and three hundred feet. There is no extensive clear demarcation between them and the upper levels of the lower Harrison. About four hundred feet above the Brule-lower Harrison contact, however, there is a fairly abrupt change to coarser sands and sandstones and to numerous but thin clay beds. This zone probably marks the change from lower to upper Harrison. A hundred feet higher good specimens (see beyond under Palaeontology) of Parahippus wyomingensis, Merychyus minimus, Oxydactylus and Palaeocastor were collected. The nearest affinities of these forms occur in the upper Harrison of northwestern Nebraska. For these stratigraphic and pa- laeontological reasons I am considering, tentatively at least, the uppe? three hundred of the seven hundred feet of Miocene in this locality as upper Harrison in age. Correlation Chadron. On the basis of our present knowledge of the fauna from this formation in the area and taking into consideration its stratigraphy, SCHLAIKJER: NEW FOSSIL VERTEBRATES 117 it seems identical with the Chadron in other areas, such as Nebraska and South Dakota. Brule. The reasons for correlating the lower channel phase of the Brule with the Metamynodon sandstone of South Dakota have been cited above. The large assemblage of species collected from the fossil layer, thirty feet higher, at the Harvard Fossil Reserve show without question that that level is equivalent to the upper nodular layer of the middle oreodont beds in the Big Badlands. Species which are distinc- tive of this level are Mesohippus barbouri and М. obliquidens; and species most commonly abundant at this level are Cynodictis gregarius, Subhyracodon occidentale and Archaeotherium mortoni. In addition, the only other remains of Bathornis celeripes found outside of the Harvard quarry were collected from the middle oreodont beds north of Harrison, Nebraska (see Wetmore, 1933, p. 308). In the upper limits of the Brule, Leptauchenia mitida and occasionally Palaeolagus agapetillus are found. These forms definitely indicate the presence of the Leptau- chenia beds. The upper channel deposit of the Brule, although no fossils have been found in it, may be the correlate of the Protoceras sandstone level of the Big Badlands. It certainly occurs at the correct stratigraphie position. It seems, therefore, that the entire Brule for- mation is represented in the Goshen Hole Area. Lower Harrison. The name Arikaree (spelled Arickaree) was first applied by F. W. Cragin (1896, p. 52) to a series of “Light-colored, olive, yellowish and brownish-gray shales” exposed on the Arickaree River in Colorado and Kansas which he referred to “the lower part of the Fox Hills division” of the Cretaceous. In 1898 G. L Adams (p. 27) referred these Arikaree shales to the upper part of the Pierre. Darton in 1899 (p. 735) used the name Arikaree,— apparently not knowing that it was preoccupied, for a part of the lower Miocene, including everything below Ogallala of the Pliocene down to the Gering beds of southeastern Wyoming and western Nebraska, which is com- posed of “a series of sands of gray color everywhere characterized: by layers of dark gray concretions which often have tabular form." J. B. Hatcher (1902, рр. 116-117) divided Darton's Arikaree into the Monroe Creek, the Harrison and the Nebraska beds of Scott. (The term Nebraska Beds is no longer in use. It is a general term that in- cluded what are now known as the upper Harrison and the Valentine). In 1904 Matthew and Gidley gave the name Rosebud beds to the lower Miocene of southwestern South Dakota and northern Nebraska. In 1907 Dr. Matthew recognized two divisions to the Rosebud and rightfully correlated it with the Harrison. Since that time a number 118 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY of the same species, as found in the lower Rosebud, have been recog- nized in the lower Miocene,— unquestionably lower Harrison, of other areas (Spanish Diggings — Lusk localities north of Goshen Hole). The name Gering formation was given by Darton (1899 pp. 735-736) to a series of channel deposits at the base of the Arikaree in the Scotts Bluff vicinity south of the North Platte River. In 1899 Darton stated that, "the character of the deposits and the abrupt change from the typical Arikaree formation have afforded a basis for the discrimination of this formation." In the Scotts Bluff Folio (1902) Darton presented a further study of the Gering and wrote (рр. 2-3), “It is separated from the Brule clay by a distinct erosional unconformity, but appears to merge upward into the Artkaree formation through a few feet of passage beds." I have inspected the Gering formation at the type locality south of Gering, Nebraska and have made a preliminary study of it from that locality westward to Wyoming. It appears to be nothing more than an exaggerated eastward extension of the same conditions that prevail at the base of the lower Harrison in the Goshen Hole Area. In other words, the Gering represents only a local phase of the lower Harrison and ought not to be regarded as a separate formation any more than the Metamynodon or Protoceras channel phases of the Brule should be regarded as distinct formations of the Oligocene. The fauna of the “Gering” is certainly not well known. In the northern part of Goshen Hole approximately fifty feet below the “Gering” channel beds or fifteen feet above the Brule contact the small canid Nothocyon leptodus sp. nov. was collected. While this individual represents a new species it is far more closely related to Nothocyon annectens of the lower Harrison than to any other known species. On the west end of Bear Creek Mountain Cyclopidius densa, a lower Harrison species, was found twelve feet above and another specimen forty feet above the Brule contact. On the east end of Bear Creek Mountain the little camel Stenomylus hitehcocki was collected twenty-four feet above the Brule contact. Previously this species has been recorded only from the lower Harrison formation four and a half miles east of Agate, Nebraska. When а fairly complete fauna of the “Gering” is known, however, that it will be somewhat distinct,— as is every "river fauna" (as the Metamynodon and Protoceras faunas of the Oligocene) is to be expected. А complete list of species from the lowermost Miocene of the Goshen Hole Area is as follows: SCHLAIKJER: NEW FOSSIL VERTEBRATES 119 Palaeolagus hypsodus sp. nov. Cyclopidius densa (Loomis) Hypolagus primitivus sp. nov. Cyclopidius simus Cope Archaeolagus ennisianus (Cope) Cyclopidius lullianus? 'Thorpe Palaeocastor sp. Leptauchenia minora sp. nov. Sciurus sp. Merychus harrisonensis Peterson Nothocyon leptodus sp. nov. Eporeodon cheeki Schlaikjer Mesocyon hortulirosae sp. nov. Mesoreodon chelonyx Scott Plionictis sanguinarius! Loomis Mesoreodon megalodon Peterson Miohippus sp. Promerycochoerus loomisi Schlaikjer Parahippus ?pristinus Osborn Promerycochoerus hatcheri Douglass Hypohippus avus sp. nov. Miotylopus bathygnathus gen. et sp. nov. Menoceras cooki (Peterson) Stenomylus hitchcocki Loomis Dinohyus minimus sp. nov. Ргоютегух sp. Cyclopidius heterodon Cope Nanotragulus intermedius sp. nov. Of the forms that are not new, and excluding Plionictis sanguinarius, Eporeodon cheeki and Promerycochoerus loomisi which have been re- corded only from the Goshen Hole Area, three species,— in addition to Daemonelix, are distinctive of the Harrison. They are Menoceras cooki, Merychyus harrisonensis and Stenomylus hitchcocki. One of these, enoceras cooki together with Cyclopidius densa (also recorded from the lower Harrison southwest of Chadron, Nebraska), Cyclopidius lullianus and Mesoreodon megalodon occur in the lowermost Miocene beds in the “Spanish Diggings"— Lusk, Wyoming area immediately to the north of Goshen Hole. Mesoreodon megalodon was first recorded from the Monroe Creek beds. Four lower Deep River (Fort Logan) Species, all oreodonts, are present,— Cyclopidius heterodon, C. simus, esoreodon chelonyx amd Promerycochoerus hatcheri. One Rosebud Species, Parahippus ?pristinus, and one John Day species, Archaeolagus enmisianus are also in the list. Of the new species, having close affinity With lower Miocene forms from other areas, M esocyon hortulirosae is Nearest M. josephi of the John Day while Nothocyon leptodus, Dinohyus minimus and Nanotragulus intermedius have their closest relationships with typical lower Harrison members. It is obvious, therefore, that 9n the basis of stratigraphic and faunal evidence the lowermost Mi- ocene beds in the Goshen Hole Area are lower Harrison in age. It is of importance that four lower Deep River species are present in the fauna. This is additional evidence for correlating those beds with the lower Harrison. That a number of the Goshen Hole species also occur 1 pa by Dr. Loomis on Sixty Six Mountain, lowermost Miocene. sion, referred to this species, was collected from the lowermost Miocene on Sixty Ountain, by Dr. Loomis. Six 120 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY in the “Spanish Diggings"— Lusk area is not surprising for those beds can be traced eastward into the lower Harrison formation of western Nebraska. The occurrence of Parahippus ?pristinus and Me- soreodon megalodon in the lower Harrison of Goshen Hole is significant. То find a lower Rosebud species and a Monroe Creek species in the same beds is not unusual. While the fauna of the Monroe Creek is very scanty it is hardly separable from that of the lower Rosebud. This together with the stratigraphie occurrence, leaves little doubt that the two are contemporaneous deposits. Likewise, the occurrence of Rosebud species in the lower Harrison is not surprising. As early as 1907, Dr. Matthew because of faunal similarities correlated the Rosebud with the Harrison. I have traced the lower Miocene along an almost continuous series of exposures from the Rosebud-Pineridge district of South Dakota southwestward across Nebraska, via Chadron and Agate, into the Scotts Bluff-Goshen Hole areas and I have been unable to discern where the lower Rosebud ends and the lower Harrison begins. Г, therefore, can see no reason for separating into two distinct formationsasynchronous deposit which is faunally andstratigraphically variable only in so far as was governed by local environmental and depositional conditions. The name Arikaree should not be resurrected for this formation since it is preoccupied, since it is too inclusive аз originally defined and since it is obsolete. Н arrison seems most appro- priate not only because it is perhaps more widely used than Rosebud but also because it has priority. Structure The general structural features of the Goshen Hole Area together with a description of the most important structures within Goshen Hole itself have been considered previously (part IL, pp. 52-53). There is evidence that the area has probably been subjected to at least five diastrophie movements since late Cretaceous times. Two О these, one at the close of the Fox Hills and the other between the close of the Cretaceous and the beginning of the Oligocene have been dis- cussed (see part II, pp. 50-53 and part III, p. 74). As was stated above, the Chadron formation is absent in a large part of the southern portion of Goshen Hole and in those localities the Brule rests directly on the Lance. This is probably the result of uplift and erosion in that vicinity because of the following reasons: 1. The Chadron formation has a relatively uniform thickness and a wide geographic distribution. 2. The Brule-Chadron contact locally presents a fairly pronounce SCHLAIKJER: NEW FOSSIL VERTEBRATES 121 disconformity. 3. The fine grained Chadron clays do not include an appreciable amount of the coarse unconsolidated sands of the Lance, on which they rest, which would have been the case had they been deposited in erosional lows. Local uplift at the end of Brule times is significantly shown in the area. At Ү.В.О. Pass approximately two hundred feet of that forma- tion were removed by erosion before the overl ying Miocene strata were deposited. То the northwest and to the southeast the Brule again at- tains its full thickness. Moreover, additional evidence of disturbance previous to the Miocene is shown by the presence of strong dips, due in part perhaps to slumping, as great as 43? (see Plate 5, Figure 1) at Fox Creek Gap. This tilting does not occur in the Miocene above. The occurrence of an occasional fissure in the Brule, down into which Miocene sands and boulders of Brule clay have been injected, seems to show that severe earthquakes took place concomitant with this up- lift. The largest of these fissures having a maximum width of three feet, crosses the northwest corner of the Harvard Fossil Reserve. It is of interest to note that this fissure has a strike parallel with the Strike of the Corn Creek Anticline and occupies a position which would be on the crest of that structure were its surface expression continued northeastward. "These fissures cannot be regarded as shrinkage cracks resulting from the drying out of the clay because they are not numerous (only three have been observed), they are too large and they have the same orientation,— southwest-northeast. The major diastrophism recorded in the area occurred after the Miocene when the whole area was broadly upwarped or domed. This 1$ shown by the fact that everywhere along the Goshen Hole escarp- ment the strata have a one to three degree dip away from the interior lowlands. Also, throughout the area there is an intricate system of mmor faults and fissures which parallels a major fault zone to the northwest. On the south side of the North Platte River about three miles southeast of the Whalen Dam there is a large normal fault which at this place has a strike N. 70? E. and a dip 61? S.SE. (See Plates 8 and 9). Behind this fault the upper Brule clay, capped by the Brule- „locene conglomerate, has been brought to the surface. On the east side, the uppermost beds of the lower Harrison occur on the Laramie Iver three miles south at a level four hundred feet below the top of the Brule which means a throw of seven hundred feet. This fault, to Which the name Whalen Fault is given, extends northeastward some ten to fifteen miles and it was traced southwest to where it again crosses the Laramie River but in which locality it is more of a crushed zone 122 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY than a single fault. Many minor faults and fissures are present in this faulted zone. For example in Sec. 31 T. 26N. E. 65W. there is a vertical fault with a displacement of not more than fifteen feet. In the im- mediate vicinity of this is a network of chaleedony veins which pro- bably represents deposition of silica in fissures from cold solution since the silica penetrates the walls of these fissures only a short distance. To the southeast of the Whalen Fault the strata are also considerably fractured. This minor faulting, along with the reasonably steep dips (in Sec. 36. T. 26N. R. 65W. the beds dip as much as 7? northwest) has probably had much to do with determining the southwest-northeast trend of Six Mile and Deer Creeks and the head portion of Cherry Creek. On the Laramie River in front of the Whalen Fault the elevation of the top of the lower Harrison is 4250 feet. Assuming that theformations in this locality have thicknesses which are normal for them in the area, there are 1580 feet of sediment above the Pierre shales. In other words the elevation of these shales is 2670 feet or at least 1730 feet lower than they are along Corn Creek over in Goshen Hole. This is in contrast to the very gentle inclination (about ten feet per mile) of the beds southeastward from the vicinity of the Corn Creek Anticline (see Plate 7). PALAEONTOLOGY As stated previously, the turtle Stylemys nebrascensis occurs abun- dantly throughout the Brule. Another form, which has been identified by Dr. Theodore White, as Testudo peragrans occurs with equal abundance in the lower Harrison. More than a dozen specimens of this species were collected. A large individual (over two feet in length), having a pathological abnormality in the carapace, is shown in Plate 10. Dr. Alexander Wetmore of the National Museum has studied the large collection of fossil bird bones collected from the Harvard Fossil Reserve and has described several new forms of which the following is a list: Order FALCONIFORMES Family ACCIPITRIDAE Subfamily BUTEONINAE Buteo antecursor SCHLAIKJER: NEW FOSSIL VERTEBRATES . 126 Order GRUIFORMES Suborder CARIAMAE Family BATHORNITHIDAE Bathornis celeripes Bathornis cursor In the very large collection of fossil mammals from the Chadron, Brule and Harrison formations, forty-two species are represented. Of these, thirteen species and two genera are new and several unquestion- ably new forms are present in the material which has not yet been prepared for study. The following is а complete faunal list of the mammals: Order RODENTIA Suborder DUPLICIDENTATA Family LEPORIDAE Palaeolagus hypsodus sp. nov. Palaeolagus agapetillus Cope Hypolagus primitivus sp. nov. Archaeolagus ennisianus (Cope) Suborder SIMPLICIDENTATA Family CASTORIDAE Palaeocastor sp. Family SCIURIDAE Sciurus sp. Order CARNIVORA Suborder FISSIPEDIA Family CANIDAE Cynodictis gregarius (Cope) Nothocyon leptodus sp. nov. Mesocyon hortulirosae sp. nov. ! Not including Apternodus gregoryi from the middle Oligocene which has been fully described in Part I of these “Contributions, etc.” 124 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY Order PERISSODACTYLA Superfamily EQUOIDEA Family EQUIDAE Pediohippus antiquus gen. et sp. nov. Mesohippus barbouri Schlaikjer Mesohippus obliquidens Osborn Miohippus sp. Parahippus ?pristinus Osborn Parahippus wyomingensis sp. nov. Hypohippus avus sp. nov. Superfamily TAPIROIDEA Family TAPIRIDAE Genus and spectes undescribed Superfamily BRONTOTHERIOIDEA Family BRONTOTHERIIDAE Allops cf. marshi (Osborn) Superfamily RHINOCEROTOIDEA Kost y RHINOCEROTIDAE Subhyracodon occidentale Leidy Menoceras cooki (Peterson) Order ARTIODACTYLA Suborder BUNODONTA Superfamily DICHOBUNOIDEA Family ENTELODONTIDAE Archaeotherium palustris sp. nov. Archaeotherium mortoni Leidy Dinohyus minimus sp. nov. I і | SCHLAIKJER: NEW FOSSIL VERTEBRATES 125 Suborder ANCODONTA Superfamily AGRIOCHOEROIDEA Family OREODONTIDAE Cyclopidius heterodon Cope Cyclopidius densa (Loomis) Leptauchenia minora sp. nov. Merychyus harrisonensis Peterson Merychyus minimus Peterson Eporeodon cheeki Schlaikjer Mesoreodon chelonyx Scott Mesoreodon scotti Schlaikjer Mesoreodon megalodon Peterson Merycoidodon culbertsoni Leidy Promerycochoerus loomisi Schlaikjer Promerycochoerus hatcheri Douglass Suborder TYLOPODA Superfamily CAMELOIDEA Family CAMELIDAE Miotylopus bathygnathus gen. et sp. nov. Stenomylus hitchcocki Loomis Protomery& sp. Oxydactylus sp. Suborder PECORA Superfamily TRAGULOIDEA Family HYPERTRAGULIDAE Nanotragulus intermedius sp. nov. Hypisodus alacer Troxell Leptomeryx evansi Leidy Leptotragulus ultimus sp. nov. ата. PH — он 126 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Order RODENTIA Suborder DUPLICIDENTATA Family LEPORIDAE PALAEOLAGUS HYPSODUS! 8р. nov. Type. A right maxilla with P*-M?, М. C. Z. No. 2,889. Collected by Erich M. Schlaikjer, 1930. Horizon and locality. Lower Miocene. Lower Harrison formation, approximately 150 feet above the Brule-lower Harrison contact. М. У. М. Sec. 21. T. 20N. В. 62W. Goshen County, Wyoming. Specific characters. About the size of Р. agapetillus. Teeth very hypsodont. P? almost quadrate in cross-section. P‘-M? oval in out- line. Re-entrant angles more compressed and proportionately more extensive than in the other species. Teeth almost entirely covered with cement. (See Fig. 4). Ета. 4. Palaeolagus hypsodus. Type. Superior view of P? —M?, Twice naturalsize. Drawn by Helen Ziska. Discussion. Р. hypsodus is the last known member of this typically Oligocene genus. In all of the characters listed above it has progressed somewhat over the “typical” Palaeolagus of the Oligocene, but is not nearly as advanced as Hypolagus which was existing at the same time. This is another striking example of a primitive form, representative of an ancestral stage existing, as a living fossil, along with a more ad- vanced type. Palaeolagus probably became extinct at the close of the lower Miocene. Measurements ro Глегайсе otr те ER ИЛ cU serie 7.8 Бозала РАН А ние dr cda iD ra tere 2.1 Nauta ODE ro e uL еее 2.5 Length Oi... сам ані a a 1.8 МОНО, 8.0 ЈОНА РОГ M i она VE 1.8 Wah or ME пали а а о ee 2.9 LEAM A AA 1.8 O ны oc i Po a. 2.6 From the Greek vy, tall; and oóovs, tooth. SCHLAIKJER: NEW FOSSIL VERTEBRATES 197 PALAEOLAGUS AGAPETILLUS Cope Material. A left maxillary with P2-M? and the root of Мз, M. С. Z. No. 2,890.. Collected by Erich M. Schlaikjer, 1930. Horizon and locality. Upper Oligocene. Brule formation, Leptau- chenia beds. Lone Tree Canyon. Goshen County, Wyoming. Discussion. This specimen compares very closely with the neotype Skull of this species described by Dr. Matthew (1902, pp. 307-308) from the Leptauchenia beds of Logan County, Colorado, except that P4 їз of slightly greater ancero-posterior dimension. P? and M? are greatly reduced and the internal grooves of the premolars and molars are covered with cement in both specimens. . The species P. agapetillus was based by Professor Cope on a lower Jaw. When Dr. Matthew referred the above mentioned skull, as well às à number of jaw fragments, to this species he recognized that the material represented a species distinct from Р. hydeni but stated (p. 308) “Whether the type of P. agapetillus is properly referred to it, I am unable to decide; but leave it provisionally." The status of P. agape- tillus is of по particular concern in this connection and at present must remain problematical. In dental evolution this species seems the most advanced of the P. temnodon-hydeni-agapetillus group which, as sug- gested by Dr. Matthew (1903, p. 218), represents a phylum of Palaeo- lagus with the internal median fold of enamel on the superior premolar- molar series more deeply incised than in the P. brachyodon-turgidus- untermedius phylum. HYPOLAGUS PRIMITIVUS sp. nov. Type. The orbital region of a skull with the right Р?-М3, М. С. Z. No. 2,887. Collected by Erich M. Schlaikjer, 1933. Horizon and locality. Lower Miocene. Lower Harrison formation, Approximately 150 feet above the Brule-lower Harrison contact. S.E.. Sec. 15. T.21N. R. 64W. Goshen County, Wyoming. Specific characters. Slightly smaller than H. vetus (Kellogg). Crenu- ations of re-entrant angle less developed. P? proportionately long (antero-posteriorly) and with crescent present. М? relatively very much reduced. Skull relatively deep with the anterior of orbit gently curved and the orbit proportionately large. (See Figure 5). 1cussion. Р? аз rather small and possesses a shallow but very com- Pressed anterior groove. P? has the greatest antero-posterior dimension ОЁ any tooth in the series. In the center of the crown is the remnant | | | | кескин те ОЕ 128 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY of a rather large crescent. There is a shallow internal groove which becomes less pronounced farther down the crown. At the postero- external corner of the tooth there is a small rootlet. А similar remnant is present on the antero-external corner of P^ and of M! on which tooth it amounts to only a tiny short crest. No crescent is present on P! and all that remains of the re-entrant angle is a faint internal groove. Even this groove is absent on М! although near the center of the occlusial surface is a narrow portion of the crenulated re-entrant Ето. 5. Hyolagus primitivus. Type. Superior view of P2-M3, Twice natural size. Drawn by Helen Ziska. angle. The re-entrant angle is best preserved on M?. It extends at least half way across the crown and crenulations are fairly well de- veloped. Through wear the internal groove has become separated from the more external part of the re-entrant angle but not as much so as in M!. M? is relatively much reduced and the crown is simple in pattern. There is no cement on any of the teeth. The palate is broad and the palatine fissure is developed about аз in Lepus and occupies the same position. The internal nares are broadly open and extend forward to opposite the center of МІ, The skull is deep, the orbit is relatively large and the postorbital process of the frontal is developed about as in Archaeolagus. As shown by Dr. Dice (1917, p. 181) the genus Hypolagus is distinc- tive, not only in the crenulated re-entrant angles of the superior molar teeth, but also in the presence of two exterior re-entrant angles in Ps. Unfortunately the lower jaw of H. primitivus is not known. The oc- currence, however, of a shallow groove extending from the antero- external corner of P? to the center of the crown probably indicates the emphasis of the anterior fold on Ра which would be brought about by the development of two external re-entrant angles on that tooth. H. primitivus affords the only information we have of the upper den- tition of an early Miocene member of this genus. In so far as is known the upper dentition, particularly in the primitive state of P®, in the less erenulated internal re-entrant angles and in the absence of cement, is more primitive than any of the later species. И. primitivus forms an SCHLAIKJER: NEW FOSSIL VERTEBRATES 129 excellent stage which, for the time being, may be considered directly ancestra] to the later members. Too little is known about the later Oligocene and early Miocene lagomorphs to say with certainty what was the ancestry of Hypolagus. Dr. Dice has suggested (1917, р. 182), that Archaeolagus might have been the progenitor of Hypolagus. Occasionally the upper teeth of that genus display a slight folding in the enamel of the internal re-entrant angle. It should be called to mind, however, that this feature is also present in Palacolagus agapetillus and it is probable that Hypolagus may have come directly from that genus. Measurements mm Distance from РМЗ on угол ое... 14.7 Distance from Pi-P* on ‘alveoli. оо иу ^os үү Greatest tengthior P5 КА NS ADULT M LI TET ер 2.4 Grcatest width of Peo у зу IT AT 4.6 Greate t length oi Pe re aa 2.4 Createt wath of P o esL US WALL o 5.2 ео 2.5 Edu o ME UU sU у у ан 5.0 Greatest larh ot М осо eaa MN 1.8 о 2.2 Vertical diameter of orbit. cn. S OE. у о. 14.8 ARCHAEOLAGUS ENNISIANUS (Cope) Material. A left maxilla with P?-M?, M. C. Z. No. 2,888. Collected by Erich M. Schlaikjer, 1930. Horizon and locality. Lower Miocene. Lower Harrison formation, Approximately 200 feet above the Brule-lower Harrison contact. М.М. Sec. 36. T. 20N. В. 62W. Goshen County, Wyoming. wscussion. This specimen seems near to A. ennisianus in size, in the Spacing of the teeth and in the shallow invasion of the internal re-entrant angles into the teeth crowns. There are also faint remnants of exterior enamel crescents which seem to connect with the outer Margins of the teeth. In so far as is known, this is an Archaeolagus Character. It differs from the above species, however, in having а Weaker anterior groove on P? and, as is indicated by the alveolus, in a Proportionately larger М5. This specimen was collected from beds 130 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY which are probably stratigraphically higher than the John Day in which А. ennisianus occurs, and when more completely known, may represent a, distinct species. Suborder SIMPLICIDENTATA Family CASTORIDAE PALAEOCASTOR sp. Two excellent skulls, M. C. Z. Nos. 2,927 and 2,886, and a partial skull and skeleton, M.C.Z. No. 2,916, referable to the genus Palaeo- castor, were collected from the lower Harrison beds in the Goshen Hole Area. A left mandible, M. C. Z. No. 2865, was collected from the Parahippus level of the lower part of the upper Harrison. None of these specimens are assignable to any of the described species of this genus. Dr. R. A. Stirton has in press a monograph on the beavers. This extensive piece of research includes a revision of the group and descriptions of new forms. To avoid possible synonymy and additional complication of the literature it is deemed advisable to postpone an account of the material at hand. Family SCIURIDAE SCIURUS sp. A fine skull, jaws and greater portion of the skeleton, M. C. Z. No. 2,928, which is probably referable to the genus Sciurus was found in the lower Harrison beds on Bear Creek Mountain. А study and de- scription of this specimen will appear later when it has been prepared in the laboratory. Order CARNIVORA Suborder FISSIPEDIA Family CANIDAE CYNODICTIS GREGARIUS (Cope) Material. Right lower jaw with С-Мз, M. С. Z. No. 2,884a, and the left facial portion of a skull with 18-Р? and three upper first molars, M. C. Z. Nos. 2,884b-f. Collected by Erich M. Schlaikjer, 1932. | SCHLAIKJER: NEW FOSSIL VERTEBRATES ОШ Horizon and locality. Middle Oligocene. Brule formation, upper part of the middle oreodont zone. Harvard Fossil Reserve. Goshen County, Wyoming. Discussion. The known stratigraphie range of this species is throughout the oreodont zone or lower and middle Brule of the White River beds. In South Dakota and Nebraska it seems to occur most abundantly in the middle level of this zone. Its occurrence in the Harvard quarry is of significance, therefore, since it aids, as well as a number of other distinctive forms, in correlating this deposit with the middle oreodont level of the Big Badlands. | Dr. Thorpe has stated (1924, pp. 425-427) that C. gregarius (Cope) is identical with C. angustidens (Marsh) and since the latter was described at an earlier date it has precedence over C. gregarius. After а comparison of Professor Cope's type with Dr. Thorpe’s figure of C. angustidens, a jaw fragment with three premolars, it seems that the latter is distinct in at least two important characters. First, the pos- terior basal cusp on Ра is much better developed and second, the ramus, especially anterior to P4, is deeper and less tapering. It is, therefore, considered advisable, for the present at least, to regard C. gregartus as valid. NoTHOCYON LEPTODUS! sp. nov. Type. А fragmentary left lower jaw with Ра-Мо and the alveoli a M; and С-Р;, M. C. Z. No. 2,878. Collected by Erich M. Schlaikjer, Horizon and locality. Lower Miocene. Lower Harrison formation, approximately 15 feet above the Brule-lower Harrison contact. МАЈА. Sec. 7. T. 25N. В. 63W. Goshen County, Wyoming. Specific characters. About the size of N. annectens from the lower larrison of Nebraska. Mi relatively narrow and the endoconid crest- like. M, relatively long. P4 narrow and with the posterior cusp not exterior to the main cusp. Mandible proportionately deep under M; (See Figure 6). Discussion. Of the specimens assigned to Nothocyon, N. leptodus seems closer to the N. annectens-latidens-lemur group of smaller forms rather than to the larger N. geismarianus-gregorü-vulvinus species. Its closest affinities are to N. latidens (Cope) of the upper Oligocene, John ay beds and N. annectens Peterson of the lower Harrison. In size IF rom the Greek, Хептос, narrow; обо05, tooth; in allusion to the narrow Mi. oe een Y $ 1 DINNER 182 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY and in the general tooth proportions it is closer to the latter while in the less crowded condition of the premolars and in the relatively deep mandible it appears to be more closely related to N. latidens. From the early "typical" Nothocyon, N. leptodus, in so far as is known, has di- gressed mainly in the relatively large size and narrow proportions of the inferior premolar-molar teeth. In these characters, №. ammectens is even more primitive. These two species are small primitive forms that Ета. 6. Nothocyon leptodus. Type. Lateral and superior views of the left mandible. Natural size. Drawn by Helen Ziska. existed on into early Miocene. They represent a phylogenetic line less progressive than that of N. gregorii and N. vulvinus, the larger and more specialized members of the genus. From Dr. Matthew's 1924 phylogeny of the Canidae (p. 122) one would infer that he considered all of the American representatives of Cynodictis, at least C. temnodon and C. gregarius, as belonging to Nothocyon. In other parts of the same paper (pp. 119, 120, 128, etc.), however, he repeatedly refers to them as members of “Pseudocyno- dictis,"— а generic name proposed by Schlosser for the American representatives. In his later paper “The phylogeny of the dogs" (1930), Dr. Matthew seemingly regards Cynodictis as including the American species. Whether or not these species should be included in а genus distinct from Cynodictis is of no particular concern in the present consideration. It is of importance, however, that they be regarded as generically distinct from Nothocyon, the species of which in almost all of their dental and cranial characters are decidedly less viverrine and more dog-like. | | | 9 SCHLAIKJER: NEW FOSSIL VERTEBRATES 133 Measurements mm Iren sch ОР PIE ME бана уво ee 34.3 ee ee 5.9 Dd ob» pie ar que dH M rer ЭЭ р неа а 9.3 ОМ аа nM E tou ез 3.9 с 5.3 ИХ а he я 9:0 Depth of mandible under center of М,..................... 114 MESOCYON HORTULIROSAE Sp. NOV. Type. A complete and excellent preserved skull and jaws, M. C. Z. No. 2,102. Collected by Mrs. C. W. Culley!, 1928. Horizon and locality. Lower Miocene. Lower Harrison formation. МУ. Sec. 36. Т. 20N. В. 62W. Goshen County, Wyoming. Paratype. An incomplete pair of lower jaws with the premolar and molar teeth, M. C. Z. No. 2,882. Collected by Erich M. Schlaikjer, 1933. Horizon and locality. Lower Miocene. Lower Harrison forma- tion, approximately 120 feet above the Brule-lower Miocene contact. S.E.14. Sec. 22. Т. 203. К. 60W. Goshen County, Wyoming. Specific characters. In size, about the same as M. josephi and M. brachyops. Face elongate, low and narrow. P* with relatively long heel. P? and P» proportionately small and the latter crowded close to РА М» proportionately wide. Palate narrow. Basicranial area relatively short. (See Figures 7-10, and Plate 11). Description and discussion. The upper tooth series is relatively Ong. ІЗ is proportionately small. The canines are not preserved although the roots and alveoli indicate that they were proportionately smaller than those of M. coryphaeus (Cope) and proportionately larger than those of M. josephi (Cope). There is a short space between the C and P!, between P! and P? and between the latter and P? which is closely set against P4. P3, as in M. josephi, does not possess a posterior cusp. The protocone (deuterocone) of P* is more reduced and, mainly аз à result of this, the shear of that tooth is in a more antero-posterior 1 ; : РЕА 36 А small block of sandstone containing the specimen was used for some tim e os a door ae at the home of Mrs. Culley near Huntley, Wyoming. Later, it acted as one of the ornaments п her rose garden. It was presented to the author in 1931. | | | | | | | | | || | \ | 34 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY e W. Germann. === Fic. 7. Mesocyon hortulirosae. Type. Lateral view of the skull and jaws. Natural size. Drawn by Louis | SCHLAIKJER: NEW FOSSIL VERTEBRATES 155 Mesocyon hortulirosae. 'Type. Palatal view of the skull. Natural size. Drawn by Louise W. Germann. Fra. 8. "UUBULIOH) "AA әвтот Aq UMBIC “Әле үелпзем IMAS oy} Jo MOTA тело] 'edÁT, ‘21050491340 цобооза Й `6 “DIA BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY ——————————" SCHLAIKJER: NEW FOSSIL VERTEBRATES 137 position than in any of the other species. Also, M! is relatively wide and the hypocone cusp does not extend as far anteriorly. M? 15 large, a Cynodesmus feature. As a matter of fact, in nearly all of its dental characters M. hortulirosae is very close to that genus. Of the lower teeth, P» and P; are relatively small. The left Р; has an accessory postero-internal cusp which is lacking in the right. The metaconid of M, is very well developed and the protoconid and metaconid of М» are pronounced and quite separated. _ The basioccipital is broad and displays a heavy median ridge on either side of which is a shallow depression immediately in front of the foramen hypoglossi. The distal end of the paroccipital process is lost but the base indicates that it was short, heavy and firmly braced against the bulla. The bulla is developed about as in Canis except that the endotympanic is proportionately somewhat larger. The ex- ternal auditory meatus is annuloid and on the anterior margin is : slight projection in front of which is the temporal canal. The occipital condyles are narrow, are widely separated above and below and occupy an oblique supra-anterior ventro-inferior position. The supraoccipital 15 high and presents a faint median ridge. The interparietal is small, and the lambdoid and sagittal crests are quite prominent. The cranium is proportionately very expanded posteriorly and the parietal and squamosal bones are about as inclusive as in Canis. "Гһе postor- bital constriction is very marked and the frontal crest is weakly de- veloped and extends out to the suppressed postorbital process. The lacrimal is fairly large and its external margin is restricted to the anterior border of the orbit. The lacrimal foramen is large and round. he malar is extensive and posteriorly it presents a small postorbital Process. The orbit is rather small and is oval in outline. The nasals are narrow throughout most of their length and widen anteriorly. P osteriorly they are wedged between the two long anteriorly projecting Processes of the frontals. The face is somewhat abruptly constricted above P? and widens again across the canines. The infraorbital fora- men is large, circular and is located immediately above the posterior Part of Ps, The dorsal posterior projection of the premaxillary is con- siderably separated from the frontal. A minute foramen incisivum is Present Just anterior to the fairly large oblong palatine fenestrae. The к 18 relatively narrow and flat and the palatine extends for rard X oe the middle of P*. The palatine grooves are very distinct palatine foramina are small, round and are situated opposite Js fTY . " M E а . > The internal narial opening is extended only slightly behind the *5* molar. The dorsal wing of the palatine appears more elongated 138 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY than in Canis because it is not so ventrally extended. The posterior palatine and the sphenopalatine foramina are close together and the latter is slightly larger. The alar canal is large and is separated from the foramen orbitale which is a trifle larger. Slightly above and in front of the latter is the optic foramen which is much smaller. There is no orbitosphenoid ridge anterior to these foramina. The distal ends of the pterygoids are broken away. Dorsally the pterygoids barely unite over the suture between the basisphenoid and vomer. The vomer is narrow and has a low median ridge. Fig. 10. Mesocyon hortulirosae sp. nov. Superior view of the lower jaws- Natural size. Drawn by Louise W. Germann. The mandible is relatively long and slender. The condyle is narrow and is extended transversely. The coronoid is high, compressed and of large antero-posterior dimension. There is a large hook-like process protruding posteriorly and internally at the angle. Immediately an- terior to this is the fairly small foramen mandibulare. The mandibular notch is widely open. | Compared with the other species of Mesocyon, M. hortulirosae 15 nearest to M. josephi (Cope) of the upper John Day. From this species it is different, however, in the following characters: SCHLAIKJER: NEW FOSSIL VERTEBRATES 13 . A relatively smaller 15. . P! closer to the canine. . P? more closely crowded against P4. . Protocone (deuterocone) more reduced. M! more antero-posteriorly constricted internally. - M? proportionately much larger. ‚ Posterior portion of P4 more elongated. - Facial region more shallow. 9 гї сн со юн Considering its general assemblage of characters, M. josephi is struc- turally, as well as stratigraphically, an ideal ancestor for М. hortuli- тозае. One other species, M. robustus Matthew, has been described from the Great Plains region. This species was based on a pair of lower jaws collected from the lower Harrison (lower Rosebud) of South Dakota. In one character, the incipient development of the metaconid on М», M. robustus is distinct from all other species and is close to Temnocyon. In addition, М. hortulirosae is different from М. robustus m its smaller size, in the relatively smaller Р»-М;, and in the relatively larger М» and Мз. Out on the Great Plains in the early Miocene, M. robustus and М. hortulirosae respectively occupied the same habitat as did M. coryphacus and М. josephi in the late Oligocene, John Day times. After a study of M. hortulirosae, as well as the other species of M esocyon 1 am unable to find a single important character in this genus which is not transitional between Daphaenus of the Oligocene and Cynodesmus of the Miocene. Measurements mm. КОК of skull, аа НА ОЇ 130.8 Condylo-basal length . Е ens M M 19/22 Teatest zygomatic breadth on squamosals..........-..-.- 80.5 Greatest breadth across squamosals on cranium............ 48.5 Breadth of upper rim of postglenoid notch................- 48.5 Breadth ON postorbital созсо 18.0 Distance from anterior rim of orbit to pmx. (anterior of I ESSE e A i uu ee 48.5 Istance from anterior rim of orbit to supraoccipital crest ... 94.5 1 io I regard M. drummondanus Douglass from the upper Oligocene? of Montana as belonging Daphaenus rather than to Mesocyon to which genus it was tentatively referred by Mr. Douglass, 140 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY Distance from mandibular condyle to pmx.....:........... Greatest width of condyles . Distance from ventral portion of бонна ЗОНУ ibi supra- OTA на Жал ше M GENI LAE Distance from anterior of I! alveolus to internal nares. ...... Breadth of muzzle at openings of infraorbital foramina . . .... Width of premaxillaries on I? alveoli (external). ............ Distance between Landes. o e у, Pa Л а S Breadth across Povon а Сон у o ДОРА КЕ ее КОШЕ ОГ Жат у жое З мау аи dirum d улл; Greatest таш В аыл л лу у лан шегу: Т ш, Уулу COSO Md ee АР у" Greatest Width oe л л eth оно. Length of МЕН ttr r o dec To S кю PR aot basti а Greatest айн OMA jae а Measurements of referred skeleton Width of distal end of humerus . EUM IU a Length of femur, great trochanter to кауы Ri I ол ҰТА ОЗІ ОТАУ ЕН ld а Smallest transverse diameter of femur shaft................ (Greatest еШ ОЮ a Width OF proximal end от узу уллин НО Width of distal end о а. (риса ЕУ Се ЛО ШШШ o у s E E ND Greatest length or caleaneum. oe vo o ШОШО СТОНА o a a Ma а ГЛ у ге Багш О баран letrero а ea a и: Longsti or eratuta o. ро, ено of metatarsal 4 a ne ee Ingo О ЕСА ВА a e и 102.0 29.5 37.0 66.0 28.0 13.5 14.5 20.7 43.0 54.5 38.5 139.0 32.5 SCHLAIKJER: NEW FOSSIL VERTEBRATES 141 Order PERISSODACTYLA Superfamily EQUOIDEA Family EQUIDAE Pediohippus antiquus! gen. et sp. nov. Type. A left maxilla with P?-M?, M. C. Z. No. 2,790. Collected by Erich M. Schlaikjer, 1930. Paratype. А nearly complete skull with complete cheek dentition, М. C. Z. No. 2942. The following specimens are also referred to this Species: a right maxilla with P?-M?, M. С. Z. No. 2,789; a palate with complete cheek teeth, М. C. Z. No. 2,791; a right maxilla with P!-M5, M. С. 7. No. 2,792; a left maxilla with РАМ, М. С. Z. No. 2,793; and, a large number of jaws and isolated teeth. Horizon and locality. Middle Oligocene, Brule formation, upper part of the middle oreodont zone. Harvard Fossil Reserve. Goshen County, Wyoming. Generic characters. Metaloph with crochet but not united to ectoloph except in M? of the later forms. Upper incisors deeply cupped and Cingule of lower incisors larger than in Mesohippus. A metatarsal HI- Cuboid facet. Specific characters. Somewhat larger than Mesohippus barbourt. Teeth as hypsodont as in Parahippus pristinus. Crochet developed on M!-M3. Р and М! subequal in size. Styles on teeth rather weakly developed. (See Plate 12). Discussion. Аз stated previously horses are the most abundant of the fossils which occur in the Harvard quarry. The occurrence of so many individuals whose remains accumulated, as is designated by Stratigraphic evidence, in a comparatively short time is particularly Welcome and of striking significance. For the first time it is now possible to say with reasonable certainty how much variation is to be expected, m the dentitions at least, in the earlier of the Oligocene horses. Like- Wise, it should be called to mind that this period is probably the most Critical in the evolution of the horse, for the direct ancestral forms of at least five important late Oligocene and early Miocene genera were becoming established. Perhaps never before and not since that time has there been such an adaptive radiation of the Equidae. А study of the dentitions and skeletal remains of over fifty speci- Mens together with a survey of the other Oligocene and early Miocene 1 From the Greek, areduov plains; ит тоў horse. 142 BULLETIN; MUSEUM OF COMPARATIVE ZOÖLOGY species reveals the following outstanding facts concerning evolutionary changes in the direct line of ancestry of the horse from Pediohippus antiquus of the middle Oligocene to Parahippus pristinus of the lower Rosebud: . The cheek teeth do not become more hypsodont. . There is only a slight increase in tooth dimensions. . The proportionate lengthening of the premolars is remarkably small. . The incisor teeth do not become more deeply cupped. . There is practically no additional strengthening of the metatarsal III — cuboid facet. 6. There was a small amount of increase of size and slight changes in pro- portions,— especially in the limbs. 7. À union of the metalophs with the ectolophs. 8. A development of a crochet on P2-M?. бл њ фо № м АП of the above changes except the latter were also accomplished in such collateral forms as Anchitherium (including Kalobatippus perhaps); Archaeohippus and Hypohippus, which survived to late Miocene or early Pliocene times. It was only in Pediohippus, however, that the crochet was developed,— primitively on the last molar, later on all three molars, and eventually on all of the cheek teeth. (See Plate 12). In the Parahippus pristinus stage, in addition to the development of a crochet on the cheek teeth, the metalophs unite with the ectolophs. From this stage onward, concomitant with an increase of hypsodonty, the crochet increases, forms isolated valleys in which cement is de- posited, and later on unites with the protoconule coincident with which the protocone begins to become isolated from the protoloph,— a Merychippus stage in which the essential tooth pattern of Equus 18 established. In Hypohippus of the early Miocene the metaloph has united with the ectoloph but from then on progression of tooth evolu- tion 18 arrested and increase of size is the only dominant change. Hypohippus matthewi of the lower Pliocene, for example, has teeth which are gigantic in size. Each superior molar has a grinding surface area of approximately three square inches, yet structurally the tooth amounts to nothing more than that of an inflated Miohippus from the upper Oligocene. These same conditions prevail in Anchithertum although with size increase more inhibited in this genus,— a feature even more pronounced in the rather rare little Archacohippus. The crochet, so characteristic of Pediohippus, Parahippus and all of the later stages in the direct line of Equus ancestry, is never present on the teeth of these genera. The development of the crochet, therefore, 18 SCHLAIKJER: NEW FOSSIL VERTEBRATES 143 the most important single character in dental change during the evolution of the direct progenitors of the horse, particularly in the formative period from Oligocene to Miocene times. Type M.C.Z.| M.C.Z.| M.C.Z.| M.C.Z.| M.C.Z.| M.C.Z.| Varia- 2789 | 2790 | 2791 | 2792 | 2042 | 2793 | tion P2-Ms 76.0 | 760 | 73.7 | 76.6 | 80.0 ? 6.3 P2-pa 382 | 38.8 | 367 | 36.6 | 392 ? 2.6 Mims 40.9 | 40.7 | 39.9 | 409 | 422 | 367 | 55 P? length 128 | 132 | 121 | 126 | 136 ? 1.5 P? breadth 1872490: 1173 о аде 485 ? 2.0 P5 length ? 140 | 180.| 181 | 134 ? 1.0 P? breadth 15.6: | 155 185. 158. | Lod. ? 23 Р! length 145 | 151 | 133 | 133 | 149 | 186 | 18 Р! breadth i^t 3841 | 1653 У 198 lege о] Daea M! length ы mo las 140. 11454214 оон XE y M! breadth 155.1 4758.1 156 188 | 120. 156 2.1 M? length 147 1.184 4 197 | 142.1 162. | 347.|. 17 M? breadth 154 М Т 170 | i 1 169.4 216 MS length 154.149 | 149.) 188 1.106 | 129 15У en s onis М breadth 17.4 | 166 | 156 | 154 | 152 | 154 | 22 M Tarro 1. Measurements of six specimens of Pediohippus antiquus showing ariation in dentition. 5 The crochet in Pediohippus antiquus shows a certain amount of ariation and some forms correspond very closely to Mesohippus bar- 144 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY bouri which also occurs so abundantly in the Harvard Quarry. In a large assemblage of teeth M. barbouri seems transitional to P. antiquus. One specimen of the latter, M. C. Z. No. 2,791, is almost indistinguish- able from М. barbouri except for the development of an incipient crochet on the last molar. Then too, there is an occasional specimen with a crochet on M? and M? but none on M!. This variability in the number of teeth with a crochet is not surprising, for the existing in the same oecologie niche of hundreds of individuals, representing at least three species, is bound to result in considerable hybridiza- tion, which would certainly account for this variation. In Table 1, measurements are listed for six specimens of P. antiquus. The figures given show a fair amount of variation which is due mainly to a dis- crepancy in size rather than to a difference in proportions. Likewise structural differences in dentition are peculiarly few. The hypostyle, for example, is remarkably constant in its form and position. It is comprised of an enlargement of the central area of the posterior cin- gulum which is connected with the postero-internal face of the hypo- cone by a small conule. Occasionally М? shows a digression from this usual pattern in the absence of the conule and is, therefore, more primitive in that respect. In the genus Pediohippus several previously described species of Mesohippus and Miohippus may be included. They are as follows: 1. Mesohippus portentus. 'This species was described by Mr. Douglass in 1908 (рр. 268-269) as from the “Lower White River (“Titanotherium’) beds, pipestone Creek, near Whitehall, Montana." The {уре is а second right superior molar on the metaloph of which is a well developed crochet. As suggested by Professor Osborn (1918, p. 45), the tall crown, the well developed crochet and the close approximation of the metaloph certainly indieate a form too advanced for the lower Oligocene. This form is very close to “Miohippus” gidleyi (Osborn) of the upper Oligo- cene and when more completely known may be found to represent the same species. ~ - Mesohippus trigonostylus. Professor Osborn described this species in 1918 (pp. 47-48). Recent excavation has revealed a distinct crochet on M. This, of course, was not mentioned in Professor Osborn’s description; neither was it shown on the original figure. Stratigraphieally this speci- men occurs in the Metamynodon sandstone which is slightly older than the beds of the Harvard quarry. ‚ Miohippus brachystylus (Osborn). 'T'his species was found in the Leptau- chenia beds of the upper Oligocene. A crochet is present on P3-M?. № forms an excellent intermediate stage between Pediohippus antiquus and Parahippus pristinus (see Plate 12). eo SCHLAIKJER: NEW FOSSIL VERTEBRATES 145 4. Miohippus intermedius (Osborn and Wortman). 'This is a species from the upper Oligocene of South Dakota. It cannot be regarded as directly ancestral to Parahippus pristinus, however, because: а. The superior incisor teeth are not as advanced as in Mesohippus barbouri and Pediohip- pus antiquus of the middle Oligocene. b. The teeth are considerably larger and the premolars are proportionately too broad. c. A crochet is developed only on M3. (It is of interest to note that M3 not only has a crochet but the metaloph is united with the ectoloph. This tooth which is frequently the most primitive of the series is the most advanced in Pediohippus, for in this genus it is the first tooth to possess а crochet and the first to have the metaloph united with the ectoloph). 5. Miohippus gidleyi (Osborn). This species is somewhat larger than М. intermedius. The crochet is present on P3-M?. It has progressed over M. brachystylus only in size,— being considerably larger than Parahippus pristinus and probably not ancestral to the latter. і In Plate 12 the upper cheek teeth of three Pediohippus species, - І : trigonostylus, P. antiquus and Р. brachystylus, are shown together With the premolar-molar series of Parahippus pristinus. All are types except Pediohippus trigonostylus. Another specimen, Amer. Mus. 674, Was selected instead of the type only because the stage of wear was More nearly like that of P. antiquus. In every respect it is identical With the type, and was collected in the same locality and from the same beds which are slightly older than the deposit at the Harvard Quarry. This series of dentitions illustrates the evolutionary changes Which took place in the premolar-molar teeth of the direct progenitors of the modern horse during early middle Oligocene to lower Miocene times. The more salient of these changes are the following: L The crochet is developed on all of the teeth. It gradually increases in Size and becomes more closely approximated to the protoconule. 2. The union of the metalophs with the ectolophs is completely established. 3. There is an extension of the protolophs so that they unite more firmly with the parastyles which become less angulate. 4. There is a proportionate enlargement of the protocone. 5. The crowns of the teeth become more quadrate. 6. Ра-ра gradually become longer than M!-M$. As mentioned above, except for the presence of a crochet, it is some- times difficult to distinguish between Pediohippus antiquus and Meso- hippus barbouri. Even in this character there is every gradation from well developed crochet on all three molars, to the merest welt on the Mtr 146 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY metaloph of M3, to the absence of a crochet. This is indeed not sur- prising. The more complete the palaeontological record, the greater will be the difficulty of distinguishing the more progressive members of an early genus from the primitive members of a later genus. This is particularly true of the horses of which the evolutionary sequence is especially complete. Certainly the presence of a crochet on the metaloph, together with other skeletal differences such as the meta- tarsal III-cuboid facet, is of greater significance in separating Pedio- hippus antiquus from Mesohippus barbouri than is any character seen in Merychippus gunteri which separates it from Parahippus leonensis or Merychippus primus from Parahippus brevidens. MESOHIPPUS BARBOURI Schlaikjer Material. A rather complete skull, M. C. Z. No. 2,778 (see Plate 13), and a great number of palates, maxilla with P!-M and lower jaws. Also an abundance of fragmentary skulls, isolated teeth and jaw fragments. Collected by the Museum of Comparative Zoólogy expeditions 1930-1934. Horizon and locality. Middle Oligocene. Brule formation, upper part of the middle oreodont zone. Harvard Fossil Reserve, Goshen County, Wyoming. Discussion. Previously this species was known only by a single specimen, the type М. С. 7. No. 17,641, which is an unusually com- plete skeleton from the upper concretionary layer of the Brule for- mation in the Big Badlands, South Dakota. The occurrence of this species in such abundance in the Harvard quarry is of significance. "Together with the assemblage of other genera and species that are found in this rich deposit, it shows that the geological level of the beds at the quarry can be correlated with the upper concretionary zone in the South Dakota Badlands. The skull figured in Plate 13, as far as the characters are preserved is identical with the type. In spite of the close similarity to Pediohippus antiquus, this species is typically smaller, has proportionately narrower premolars, has a more primitive hypostyle and, of course, does not have the metatarsal III-cuboid facet. In addition, the lower teeth are relatively narrower and the symphysis is proportionately shorter and heavier. By comparative measurements of teeth of ten individuals, given in Table 2, it is shown that this species presents little variability in den- tition. SCHLAIKJER: NEW FOSSIL VERTEBRATES 147 MESOHIPPUS OBLIQUIDENS Osborn Material. A left maxilla with P?-M?, М. С. Z. No. 2,785, and а large number of isolated teeth. Collected by the Museum of Com- parative Zoólogy expeditions 1930-1933. M.CZ. | M.CZ. | M.C.Z. | M.C.Z. | M.C.Z. | М.С.2. | M.C.Z. | М.С.2. | MCZ.| MCZ.| Size 2778 | 2779 | 2780 | 2781 | 2782 | 2783 | 2784 | 2786 | 2787 | 2788 | Variation 4 | | Ы. En | 73.9 | 760 ? 76.0 ? ? ? 5.0 34.9 37.8 380. | 397 ? ? 39.8 4.9 41.1 40.6 ? 3.1 length 11.8 | 13.1 | 12.0 118 12.8 12.8 1.3 P? breadth ця 4122 | 13.1 12.7 12.5 113 18 Р: TRES парео co SEN RP er Sa TAE MER a Тү. ength 133 | 129 | 125 | 130 18.9 12.9 1.9 P* breadth | 1.6 P length 2.0 P! breadth 1.3 M! length 5. 18 M! breadth 15.8 16.8 16.7 1.5 M? length 144 135 | 148 13 М? breadth 163 | 168 | 165 | 175 1.6 М length | 13.9 14.5 ? ? ? 1.8 M? breadth | 15 | 16.9 15.9 | 151 ? ? ? 1.9 и | ixi р TABLE 2, Measurements (in millimeters) of ten specimens of Mesohippus arbouri showing variation in dentition. Horizon and locality. Middle Oligocene. Brule formation, upper Part of the middle oreodont zone. Harvard Fossil Reserve. Goshen County, Wyoming. Discussion. This species is based on a fragmentary skull and jaws of an immature individual in the American Museum collection. It was Collected from the same geological level as M. barbouri in the South akota Badlands and is only occasionally found in the Harvard quarry. The teeth of this species are approximately fifteen percent larger than those of M. barbouri. The molars are more hypsodont and each tooth is more rhomboidal in outline. 148 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY MIOHIPPUS SP. Material. А pair of lower jaws, М. C. Z. No. 2,946 of a very young individual. Collected by Erich М. Schlaikjer, 1933. Horizon and locality. Lower Miocene. Lower Harrison formation, approximately 100 feet above the Brule-lower Harrison contact. 5.Е.М. Sec. 3. Т. 20N. В. 60W. Goshen County, Wyoming. Discussion. The incisors and canines were preserved and resemble very much the permanent teeth of Mesohippus. АП of the deciduous premolars were lost but a partially erupted Mi indicates an animal somewhat smaller than Pediohippus intermedius. The incompleteness of this specimen prevents specific determination and it is tentatively assigned to the genus Miohippus. PARAHIPPUS ?PRISTINUS Osborn Material. A left maxilla with DP2-1, М. С. Z. No. 2,867 (see Plate 41), and a right lower jaw with P®-M? of an extremely old individual. Collected by Erich М. Schlaikjer, 1933. Horizon and locality. Lower Miocene. Lower Harrison formation, approximately 200 feet above the Brule-lower Harrison contact. 5.Е.М. Sec. 22. Т. 20N. К. 60W. Goshen County, Wyoming. Discussion. The specimen with DP? compares very closely with one in the American Museum, No. 12,915, from the lower Rosebud formation of South Dakota, which is undoubtedly referable to IA. pristinus. The teeth are somewhat narrower, however, and there is no crochet on DP?. More complete material is necessary for accurate identification. : Specimen M. C. Z. No. 2,867b is the jaw of an extremely old indi- vidual. The crown of М! is so completely worn that the roots of that tooth are no longer joined together. Around the roots of several teeth and down in the jaw there is 'evidence of pathological abnormality. In life this must have been a region of acute infection, a condition not uncommon in very old horses today. Although the dental characters are destroyed by wear, the specimen is of about the same size ап proportions as P. pristinus. PARAHIPPUS WYOMINGENSIS! sp. nov. Type. An unusually complete skeleton, M. C. Z. No. 6,390. Col- lected by Erich M. Schlaikjer, 1933. nsis 1 The writer has in press a rather extensive paper on, “A study of Parahippus wyominge and a discussion of the phylogeny of the genus.” | | I SCHLAIKJER: NEW FOSSIL VERTEBRATES 149 Horizon and locality. Lower Miocene. Lowermost upper Harrison formation, approximately 500 feet above the Brule-lower Harrison contact. S.W.lz. Sec. 10. Т. 24N. R. 65W. Goshen County, Wyoming. Specific characters. Skull length somewhat less than that of P. tyleri. Skull proportionately narrow. I? smaller and 1° considerably smaller than Г. Proportionate shortening of molars and development of crochet intermediate between P. tyleri and P. nebrascensis. Hypostyle Weak and mesially placed so that it is close to the posterior of the hypocone. Hypocone connected with metacone by a thin, narrow, Straight ridge which gives a straight posterior margin to the tooth. Cheek teeth, especially Ра and M+, proportionately broad. Palate very harrow. Lower canines absent. Limbs very long. Lateral digits short. Metatarsal ITI proportionately short. (See Plates 14, 15 and 16). Maximum | Maximum Length Femuro- length length of metatarsal ПТ of femur of tibia |metatarsal ПІ ratio P. wyomingensis M.C.Z. No. 6390 320.0 344.5 223.4 .698 ҮТ алал em ТТ P. sp. Amer, Миз. 13807 285.0 303.0 203.5 714 р, tyleri ү Amer. Миз. 18769 276.0 286.0 198.0 417 Т нарынын кка ЫННА р, nebrascensis Carn. Mus. 1440 325.0 325.0 ? ? ir 3. Comparative limb measurements and the femuro-metatarsal III 108 of four Parahippus specimens. Nanas Perhaps the most outstanding single feature of Р. ioi 7 is the extraordinary length of the limbs in proportion Mii 12% of the skull and of the skeleton. The skull is some- Uf ihe E and is much narrower than that of P. tyleri Loomis, one D pe skulled forms, yet the limbs are longer than ш any is arahippus. In Table 3 comparative measurements of the dw m dn and metatarsal III are given for the available specimens accurate measurements could be had. The nearest approach 150 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY to P. wyomingensis in limb dimensions is P. nebrascensis Peterson. The femora of these species are of approximately the same length yet the tibia of P. wyomingensis is much longer. The length of metatarsal ПІ in the type of P. nebrascensis is not known. Excellently preserved hind limbs and other skeletal parts of a specimen, No. 13,807, in the Ameri- can Museum collection from the same geological level is undoubtedly referable to this species. Both the metatarsal III and the femur of this specimen are considerably shorter than those of Р. wyomingensis. While this species possesses limbs which are longer than in any other known Parahippus, its femuro-metatarsal ratio is lower than in the other species (see Table 3). It is obvious, therefore, that during this development of longer limbs the enhanced growth of the proximal ele- ments was proportionately greater than that of the metatarsals. In every essential character of the cheek teeth as well as in other dental and skull characters, P.wyomingensis forms an admirable inter- mediate stage between P. tyleri and P.nebrascensis. P.australis (Leidy) is known only from the second right upper premolar. The position and isolation of the internal cusps (protocone and hypocone) indicate an evolutionary stage about the same as that of P. nebrascensis, al- though a smaller form, and with the evidence at hand there is nothing to rule P. wyomingensis out of its direct ancestry. Parahippus tyleri-wyomingensis-australis-nebrascensis represent а primitive group of large, brachiodont forms, the later members of which became longer limbed. Measurements mm. Cond oba «2.2. ance ВАННУ 312.7 Zygomatie breadth (Just posterior t0MY...............% 114.7 Greatest breadth across squamosals on malar arches. ....... 1114 Greatest breadth across squamosals on cranium............ 64.5 Interorbital breadth across ітопігів....................... 728 Distance from anterior rim of orbit to pmx................. 171.2 Distance from anterior rim of orbit to supraoccipital crest... 173.1 Breadth of muzzle at anterior root of Mi... лл д... 87.2 Greatest Vidin ot СОШО ыл. ынк TS Du қы а i edd 51.5 Width of premaxillaries at posterior of I? alveoli............ 35.7 Cheek teeth on alveoli (P! to M?) external................. 1168 Distance from Р? to P^ on alveoli (external) ................ 59.0 Distance from М? to M? on alveoli (external).............. 52.0 Bosteriot OL Lu to УМ ОП SV ROLL. s cem rero на mcr 26 128.4 SCHLAIKJER: NEW FOSSIL VERTEBRATES 151 Depthror талу Underantemor ob Mus co ОТО 9005 ОЕ О РО. 55.7 Distance ТОА О ue uc а уг e 58.1 О СО scapula СЗАО te, оо REA 230.0 Length of humerus, head to posterior trochlea.............. 198.8 асат leno th Or a Mop 22. 320.0 Length of femur from head to internal trochlear ridge. ....... 291.6 Length of tibia from medial condyle to internal malleolus.... 344.5 ПСР OL Hohl О ыллыы к ed e у куше 312.0 Greatest length of Сай Mo amuni length ol metatarsal HT. 1. 2.001 A Maximum length of metatarsal ІУ........................ 209.0 HYPOHIPPUS AVUS sp. nov. Type. Almost complete skull of a colt with complete right deciduous premolars and M!-2, also left P4-M?, M. С. Z. №. 2,811. Collected by Erich M. Schlaikjer, 1933. Horizon and locality. Lower Miocene. Lower Harrison formation, approximately 100 feet above the Brule-lower Harrison contact. 5.W.14. Sec. 16. T. 20N. R. 63W. Goshen County, Wyoming. Specific characters. А small form. Metaloph firmly united with the ectoloph on P* and less strongly on M2. Metaloph straight with metaconule very faintly marked. Protoloph not connected with the ectoloph on the crest. Protoconule conical. Anterior cingulum strong, internal cingulum absent. Hypostyle developed as a loop on the inside of the strong posterior cingulum. No hypostyle-hypocone crest. Cheek teeth quadrate in outline, median width of each as great as length on ectoloph. М! smaller than Pt or Ма. Antorbital fossa large and deep. (See Plates 17, 18 and 19). Discussion. Hypohippus avus is the earliest and smallest known Species of this genus. А dental character which is alone sufficient to eliminate it from Miohippus is the union of the metaloph with the ectoloph. Compared with the species of that genus, it is somewhat smaller than М. validus, of the upper Oligocene, a species which in its large size, its broad relatively hypsodont teeth, its well developed hypostyle-hypocone crest and in a number of other features is already specialized towards the Anchitherium praestans line for which it forms ап ideal progenitor. Anchitherium praestans Cope (Kalobatippus praes- tans Osborn) occurs at approximately the same geological level as Hypohippus avus. Miohippus brachylophus (Osborn) from the upper 152 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY Oligocene, John Day is somewhat smaller than И. avus and is more primitive,— being a “typical” Miohippus. Of all the known species of Miohippus, it seems nearest the ancestry of the Hypohippus line. The more striking dental characters which reveal this affinity are: the form of the hypostyle and its isolation from the hypocone; the quadrate form of the cheek teeth; the reduction of Mt; and, the shelf-like de- velopment of the postero- and antero-internal cingula. Dr. A. S. Romer (1926) has proposed that Professor Osborn's genus Kalobatippus be included in the genus Anchitherium. In a more recent paper Dr. Simpson (1932, p.33) has suggested that if this be done, “the more primitive Hypohippus group, H. equinus, H. pertinax and Н. osborni should be placed there. H. affinis (the genotype), H. neva- densis and Н. matthewi are more advanced and distinctive." After а study of the species referred to Hypohippus I am unable to find characters in the known material which would permit the separation of the species into two generically distinct groups. In size, reduction of the protoconule and in tooth proportion H. osborni seems much nearer to И. matthewi, the largest and most specialized of all the known species, than to Anchitherium. After a consideration of these facts, and since the discovery of H. avus it seems preferable, at least until more complete material is found, to regard Hypohippus and Anchithervum as distinct genera derived respectively from Miohippus brachylophus and М. validus of the upper Oligocene. H. avus could indeed have been the ancestor of all the later species. Aside from its tooth characters, and even though it is а young individual, it is strikingly similar to IT. osborni in skull features such as the following: 1. The elongated and expanded cranium. 2. The low facial crests. 3. The proportionately great diameter across the malar portions of the zygomatic arches. 4. The small postorbital bar. 5. The extraordinarily deep and extensive antorbital fossae. Measurements ШШ; пати lange FW п УДО ia ПРУСИ 224.0 Length of P!-M? rwn La o У m 86.0 Length of Р\-М?.. о Greatest length of ps uL qi cR MN UE an ША Greatest отела от Piven u жаа Lr ce dap lores 18.3 Yen test ODD. Oh MICRO аи 17.0 (ба носе breadth, OF ME ма ya 17.8 SCHLAIKJER: NEW FOSSIL VERTEBRATES 100 o length ot МВ d a 17.0 оос bres dih of MET, ee т тещ 19.0 Superfamily TAPIROIDEA Family TAPIRIDAE Genus and species undescribed During the past field season a fragmentary skull and jaws with complete dentition, and the greater. portion of the limbs and other skeletal parts of a tapir were collected from the lower part of the upper Harrison formation in Goshen Hole. This specimen has not as yet been prepared for study. A cursory inspection of the material, however, reveals the fact that this form is a splendid intermediate stage between Protapirus of the Oligocene and the recent 7 apirus. Stratigraphically it is from exactly the same level as Parahippus wyo- Mingensis and was found only a short distance from that specimen. A new species of Oxydactylus and the jaw of a large Palaeocastor were also collected from this same level. The lithology of the beds is a rather fine, unconsolidated gray sandstone which contains a remark- able abundance of plant remains that sometimes constitute over fifty bercent of the deposit. This biotic environment would seem to indicate à tropical, densely forested habitat. The occurrence of Parahippus wyomingensis in this deposit is of oecological significance. The speciali- Zations for forest-living of this species with its relatively long limbs (principally the result of а lengthening of the proximal elements) and rachyodont teeth are in marked contrast to a species such as Para- hippus cognatus with а hypsodont and more complicated dentition and limbs with the distal elements proportionately more elongated, all of Which characters are plains-living and grass-eating specializations. The discovery of this tapir 1s of singular importance since not more than three teeth have been previously recorded from the Miocene (excepting perhaps the uppermost portion of the John Day) of North merica. The meagre fossil record of tapirs should not be misleading. robably they have always been forest dwellers (since Oligocene times at least) and the preservation of their remains was perhaps as occasional 25 was accumulation of sediments in such areas. The family Tapiridae 48 persisted in the New World during the past sixty million years and there is every reason to believe that during the Miocene the mem- bers of this family occurred in numbers commensurate with the living orms. 154 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Superfamily BRONTOTHERIOIDEA Family BRONTOTHERIIDAE ALLOPS CF. MARSHI (Osborn) А right superior P? and a right superior M2, М. С. Z. No. 2948a and b, were collected from about the middle of the Chadron as exposed in Goshen Hole. These teeth were found in the N.W.14. Sec. 19. Т. 23N. В. 61W. In size, in general proportions and in most of the characters they are close to Allops marshi and when more completely known this form may be definitely referred to that species. In one important character,— the rather pronounced confluence of the hypocone with the protocone of Р?, this specimen is closer to Brontops. Although, this character is somewhat variable in the Menodontine titanotheres. Fragments of teeth, skulls, jaws and skeletal parts occur abundantly throughout the Chadron beds in Goshen Hole and are very valuable for purposes of correlation. The Colorado Museum expedition in 1931 collected one fragmentary and two fairly complete skulls. These were found in the uppermost level of the Chadron and when studied may prove to be of considerable morphologic interest. Superfamily RHINOCEROTOIDEA Family RHINOCEROTIDAE SUBHYRACODON OCCIDENTALE (Leidy) Material. Ап adult skull without the premaxillae, M. C. Z. No. 2,919. (See Plate 41). Collected by Erich М. Schlaikjer, 1932. Horizon and locality. Middle Oligocene. Brule formation, upper part of the middle oreodont.zone. Harvard Fossil Reserve. Goshen County, Wyoming. Discussion. This specimen agrees closely with Leidy's figures of the type as well as with specimens in the Harvard Museum collection from the middle Oligocene of South Dakota and Nebraska which are also unquestionably assignable to S. occidentale. More than fifteen good skulls were collected from one hundred square feet in the south excavation at the Harvard quarry. They occur, therefore, in most unusual abundance. When this material is prepared for study, it may be found that more than one species is represented. | f SCHLAIKJER: NEW FOSSIL VERTEBRATES 195 MENOCERAS COOKI (Peterson) Material. | А pair of lower jaws of an old individual, М. C. 7. No. 2,874. Also, a left lower jaw of а young individual with deciduous meisors, DP, and Mi, M. С. Z. No. 2,938. Collected by Erich М. Schlaikjer, 1933. Horizon and locality. Lower Miocene. Lower Harrison formation, approximately 120 feet above the Brule-lower Harrison contact. 5.Е.М. Sec. 22. T. 20N. В. 60W. Goshen County, Wyoming. Discussion. The pair of jaws is of an old individual hence the tooth Characters have been obliterated. In size, proportions and general form, however, the specimen seems indistinguishable from M. cooki Which has a marked range of variation. Normally in this species the first pair of inferior incisors is lost with age. The presence of these teeth in this specimen is, therefore, unusual. Measurements Mass Greatest length ої mandible. Е ter DOE seb un BORG) Istance from mandibular condyle to anterior of jaw....... 388.0 E сотойо к сон аа ғ e бш а 210.0 Depth of mandible under center of Ра..................... 58.5 ВРЕТ mandible under center of Мз.:....22...5.2222..2 069.0 1dth of symphysis across caniniform incisors on alveoli .... 47.0 Length of зур ув нь азарт al 98.0 О КЛИПКЕ 164.5 ВА Қы Pech een. 67.0 ШО йош Mi Ма ве Нун. coe. сы EA. 101.0 Order ARTIODACTYLA Suborder BUNODONTA Superfamily DICHOBUNOIDEA Family ENTELODONTIDAE ARCHAEOTHERIUM PALUSTRIS Sp. nov. Туре. Complete lower jaws without the teeth, М. C. Z. №. 2,980. Collected by Erich M. Schlaikjer, 1930. Horizon, and locality. Middle Oligocene. Brule formation, upper Part of the middle oreodont zone. Harvard Fossil Reserve. Goshen County, Wyoming. ES H 156 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Specific characters. "The largest known Archaeotherium. Symphysial region long, deep and very broad. Ramus proportionately slender. Posterior mental process long and nearly circular in cross section. Pı very large and set close to P». No diastema between Р» and Ps. М, reduced. (See Plate 20). Description and discussion. А. palustris is an exceedingly large en- telodont. The ramus is long and relatively slender. One of the more outstanding features is the very broad and extraordinarily massive symphysial region. The anterior mental process is missing but the base indicates that it was large, oval in cross section and projected downward and somewhat outward. Immediately above its anterior margin is a large mental foramen. None of the teeth are present but the alveoli and some of the roots are preserved. The incisors were very large and the canines enormous. P; was very large and there is no evidence that it was two-rooted. It was set close to Ро and was sepa- rated from the canine alveolus only by a short diastema. Pz and Pa were large and subequal in size. The alveoli of M, are small which indicates that the tooth was considerably reduced. The coronoid proc- ess is very low and heavy. The condyle is large and faces backward. The masseteric fossa is oval-shaped, deep and extensive. The angle is heavy and very expanded. The vascular impression is well marked. Of the described species of Archaeotherium, А. palustris is nearest to A. scotti Sinclair from the Chadron formation. Certainly the latter is the closest approach in size. А. palustris is different from this species, however, in the following characters: larger size; wider and heavier symphysis; premolar teeth more crowded and P; larger; ramus more slender; anterior mental process not flattened; and, the posterior mental process very much longer. Structurally, as well as strati- graphically, A. scotti is an ideal ancestor for A. palustris. Moreover, there is no evident reason why the latter could not be directly ancestral to Вобећатиз, the very large upper Oligocene form which can be con- sidered as directly ancestral to Dinohyus of the lower Miocene. Boócherus is known only by the limbs. Mr. Peterson (1909, pp. 62-63) has shown that “Although B. humerosus is heavier than D. hollandi, the feet of the former are considerably shorter and broader than in the latter.” He further states that, “The absence of an articular facet for metatarsal III on the lower tibial angle of the cuboid in the John Day form (Boöchoerus) is also a noticeable feature." It would seem, therefore, that from the Oligocene to the Miocene there was a change to a longer limbed form with a more firmly arranged tarsus. This development is in keeping with the changing environmental conditions | | | | | | | SCHLAIKJER: NEW FOSSIL VERTEBRATES 157 of early Miocene times when the Great Plains region (the habitat of Dinohyus) was becoming more and more arid, and coincident with this the entelodonts, in the beginning primarily a rarian and palustrian group, were now compelled to adopt a cursorial habit. Their failure to fully achieve this was probably the very cause of their extinction later in the Miocene when environmental changes were even more Severe. Measurements mm. red test length of the Jw. AT EN E ? Width across jaws immediately posterior to anterior mental PROCES A ы ud УАН NIC а оа 105.0 Width across jaws on exterior borders of canine alveoli...... 125+ BE of АИ а a WI HUP 1754- Width across posterior mental tubercles................... 2700 Шер af jaw under Му Л a есе оао S PULS LI esq PR о 178.0 Length of М.-М; не 105.0 Antero-posterior diameter of P; alveolus.................... 35.0 Antero-posterior diameter of M; а1уеоЇйз.................. 29.5 Antero-posterior diameter of Mg alvelous 323... 11313. 88:0 ARCHAEOTHERIUM MORTONI Leidy Material. Ап almost complete pair of lower jaws, M. C. Z. No. 2,914. Collected by Erich M. Schlaikjer, 1930. Horizon and locality. Middle Oligocene. Brule formation, upper Part of the middle oreodont zone. Harvard Fossil Reserve. Goshen County, Wyoming. Discussion. This specimen is almost identical in size and characters With А. mortoni as figured and described by Mr. Peterson (1908), Mr. Troxell (1920) and Professor Sinclair (1921 and 1924). Its oc- currence in the Harvard Fossil Reserve is of stratigraphic importance as 1s also the occurrence of a number of other forms) since it is typi- cally of the upper concretionary beds (middle oreodont zone of the Brule formation) in the Oligocene deposits of South Dakota. DINOHYUS MINIMUS sp. nov. T ype. The symphysial region of the lower jaws with deciduous entition, and the permanent teeth about to erupt, M. C. Z. No. 2,894. Collected by Erich M. Schlaikjer, 1932. 158 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Paratype. Рз and Pz, М. C. Z. Nos. 2,894a and b. Collected by Erich М. Schlaikjer, 1932. Horizon and locality. Lower Miocene. Lower Harrison formation, Promerycochoerus loomisi-Mesocyon hortulirosae zone, approximately 200 feet above the Brule-lower Harrison contact. N.W.14. Sec. 7. T. 20N. R. 60W. Goshen County, Wyoming. Specific characters. About one-fourth smaller than D. hollandi. In- cisors very large. Symphysis relatively expanded. (See Plate 21). Discussion. This specimen possesses an excellent series of deciduous incisors. In their small size, narrow proportions and weakly developed cinguli these teeth resemble the permanent incisors of Archaeothervwm. The lingual symphysial surface is broken away and the permanent incisors are seen ready to erupt. The first of these is approximately twice as large as DI;. It has a heavy internal cingulum and is in a position as if it were going to replace both thefirstandsecond deciduous incisors. The second unerupted tooth is more caniniform than the first. It probably-had an internal cingulum although this remains problemati- cal since that portion of the tooth has been lost. The third tooth is the largest of the three. It is internal to and slightly behind the canine. DP, is small and is set close to the canine. Considerably behind it is the elongated and narrow DP. Р» of the paratype differs from ЮР» in that it is shorter and heavier. P; of the paratype is about one third larger than Р». Dinohyus minimus is the smallest known member of this genus and its affinities are, of course, nearest Dinohyus hollandi the gigantic and. only other known species from the lower Miocene. Measurements mm Length of symphysis......... Nu AL Lo cad E Prod Vased d 150.0 bane verse: diameter ої Die ои ED S BR COEM лиле. 15.5 Transverse diordatet Ob beu No oot nee а Ap. 20.0 талау ете тесе Ор Wen А ы ue аты da e dm Лу. 28.0 Transverse Сесе Ооа ica 20.5 Transverse diameter ot Die 90. у. ш и 2208 Transverse diameter dile e оне 20+ Antero-posterior diameter of C (at base of erown).......... 29.0 "Tirarsversedarmeteti ot o o0 Ав 21.0 Anterospostetior diameter on DD 2 a a Pg ОН 27.0 Aranaverse diameter ot DE sn DS oa o Ия i SCHLAIKJER: NEW FOSSIL VERTEBRATES 159 Antero-posterior diameter of DPa......................... 34.0 паша ставе тата е Сету ОЈ ЮР вав Pao SUE Jodi ume уы es TO idth across canines (external). ......................... 1815 Antero-posterior diameter of Pa (paratype)................ 30.5 Transverse diameter of P; (рата тре кант c в 150 Antero-posterior diameter of P; (paratype meta лаа О Transverse diameter of P; (paratype) ок тыы о 200 Suborder ANCODONTA Superfamily AGRIOCHOEROIDEA Family OREODONTIDAE CYCLOPIDIUS HETERODON Cope Neotype. A fairly complete skull and jaws, M. C. Z. No. 2,849. So referred to this species is an excellent skull and right lower jaw of an immature individual, M. C. Z. No. 2,826. Neotype collected in 1932 and the referred specimen in 1933 by Erich M. Schlaikjer. Horizon and locality. Lower Miocene. Lower Harrison formation. 9th specimens found approximately 110 feet above the Brule-lower arrison contact. М. С. Z. No. 2,849 from S.E.14. Sec. 17. T. 20N. В. 62W. and M. C. Z. No. 2,826 from S.E.. Sec. 22. T. 20N. R. 60W. Goshen County, Wyoming. Characters. One of the smallest known species of Cyclopidius. esostyle, especially on МІ, very weakly developed. Molarsverymuch terally compressed and M! shortened. Premolars and molars greatly Ypsodont. Incisors 2. Lower incisors minute. Skull narrow for ‘Yclopidius. Facial area narrow and rather elongate. Orbits exces- Sively large. Malar portion of the zygomatic arch not developed in- Criorly. Antorbital fossae elongated, narrow and almost separated. (See Plates 29 and 23). escription and discussion. The basicranial area is shortened and axis is rather steep. The bulla is large and oval and has the hyoid and Paroccipital processes deeply embedded in it. On the anterior Surface of the bulla there is a barb-like projection which is in contact ds the pterygoid. When viewed from behind the skull is quadrate 1 outline. The lambdoid crest is low and the sagittal crest is only еа posteriorly. The postorbital bar (incomplete in a young » Widual) is weakly developed and the orbits are proportionately ery large. The upper and lower incisors of the neotype are not pre- la its сія 160 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY served. There are two incisor alveoli preserved, however, in the lower jaw and it is probable that the third, which is more shallow, has been broken away. There is no reason for assuming that the incisor formula of the neotype is different from that (13) of specimen М. C. Z. No. 2,826 which is so exactly similar in all of its other important characters. It must also be borne in mind that the incisor alveoli of Cyclopidius are shallow and when even a small amount of the superior symphysial area is broken away all evidences of incisors may be completely obliterated. This is particularly true of old individuals and is admir- ably shown in specimen M. C. Z. No. 2,842 (See Plate 22) which is unquestionably C. heterodon. The outstanding features of the cheek teeth are: their very hypso- dont development; the transverse compression of the molars; the shortened M!; and, the weak development of the styles on the molars. In 1878 Professor E. D. Cope (P. 222) described the species Сусіо- pidius heterodon which he based on a fragmentary maxilla containing P^ and Мі. In 1884 (P. 559) he referred this species to the genus Pithecistes. Later on Matthew (1899, р. 73) restudied the Cope specimens and rightfully concluded that Pithecistes was a synonym for Cyclopidius. He stated, however, that in C. (P.) brevifacies “there are no distinctions whatsoever from Cyclopidius simus except those due to age of the individual.” After a carefuly study of the material in the American Museum I am convinced that C. (P.) brevifacies is à very short and deep-jawed form which is specifically distinct from C. simus Cope. Matthew further stated that, “Pithecistes decedens is the permanent and P. heterodon probably the milk dentition of a smaller species of Cyelopidius.” Close inspection of the type of C. decedens (consisting of a maxilla with three teeth) reveals that the teeth present are DP? and M! and that this species can, without question, be re- ferred to C. simus as suggested by Loomis (1925, p. 248). C. heterodon, 'as is now shown by the Goshen Hole material, is a valid species and the two teeth of the type are the permanent P^ and МЇ. This rare little Cyclopidius is a species which seems to represent а dwarf line that evolved from some species of Leptauchenia, perhaps from L. nitida itself, in later Oligocene times. Dr. Chester Stock (1930, p. 38-39) has described, from the lower Miocene Sespe deposits O California, the smallest known form of the Leptauchenia-Cyclopidius group which he regarded as “Leptauchenia? (Sespia) californica тї. subgen. and sp.” The characters listed for this new subgenus are 25 follows (p. 38): “Skull with large antorbital vacuities; nasals long а? narrow; ramus of mandible relatively deep. External styles on upper SCHLAIKJER: NEW FOSSIL VERTEBRATES 161 molars less developed than in Leptauchenia. Styles on inner walls of lower molars feebly developed or absent. M; with inner wall of pos- terior lobe projecting but slightly past the inner wall of the middle lobe. No overlapping of inner enamel wall of middle lobe by inner wall of anterior lobe in Мз.” All of these characters are prevalent in C. heterodon and the main distinguishing feature between these two Species is one of size. Мо incisors were preserved in the California specimen and Dr. Stock very tentatively referred it to Leptauchenia. Since its close affinities to C. heterodon are now established it can be referred to the genus Cyclopidius, and on the basis of the known ma- terial, affords an ideal ancestral stage, structurally at least, for C. heterodon. (See Fig. 11). The discovery of C. californica is of particular Interest since it is the first record of occurrence of the Leptauchenia- Cyclopidius group west of the Rocky Mountains. Measurements mm Skull length, lambdoid crest to premaxillary............ 89.0 Length of dental Senes (MUST обо А e al 44.0-48.0 Length of Süpcno premolätes и 18.0-19.0 Length of URENO ШОН Ыл D. а E 23.0-25.5 Length of inferior premolar-molar series................ 44.5 Greatest length of mandible...... заочною Cu у и 74.0 CYCLOPIDIUS DENSA (Loomis) Material. Two excellent adult skulls and jaws, M. C. Z. Nos. 2,850 and 2,866, two skulls of young individuals, M. C. Z. Nos. 2,876 and 2,877, two palates M. C. Z. Nos. 2,067 and 2,851 and a number of isolated jaw fragments and teeth. (See Plates 24 and 25). Collected by Erich M. Schlaikjer, 1931-1933. orizon and locality. Lower Miocene. Lower Harrison formation. oshen County, Wyoming. iscussion. Dr. Loomis described (1925) a new species of Leptau- Chenia from the lower Harrison (lower Rosebud) at Muddy Creek, eastern Wyoming which he named L. densa and to which species he referred a large number of skulls from the same beds and from the сч Harrison of the Goshen Hole Area. He recognized the many Yelopidius-like characters in this species, however, but referred it to "Plauchenia principally because it possessed three lower incisors for 162 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY later he stated,! “In my material the question of 2 or 3 incisors in the upper jaw is not clear. I assumed that if there were 3 below — there were 3 above.” The question of generic distinctions between Leptau- chenia and Cyclopidius has not been definitely settled. Dr. Thorpe (1921, p. 412-413) reviewed the subject and concluded that, “И we assume for the present at least that Cyclopidius has but two incisors in each tooth row, then we have the most marked generic distinction between this genus and Leptauchenia, which has the full complement of forty-four teeth." A study of the Cyclopidius types and a large number of other specimens shows that wherever the incisors are pre- served the formula is I2. One possible exception to this is seen in C. brevifacies which is founded on a lower jaw. In this species only two incisor alveoli are present. Since, however, the jaw is that of a very old individual and since the alveoli present are minute and very shallow, it is entirely possible that the loss of one of the incisors is an old age character,— a condition not infrequent among other groups of mam- mals. Nevertheless, it would not be surprising to find a species of Cyclopidius with less than three or, as suggested by Thorpe, even no inferior incisors. The fact of importance is that Cyclopidius may, and usually does have three lower incisors. АП of the Goshen Hole speci- mens at the Museum of Comparative Zoólogy which can be referred unquestionably to Leptauchenia densa have three lower and two upper incisors. This condition together with the many other "typical" Cy- clopidius characters such as the proportionately shortened skull, the massiveness of the malar portion of the zygomatie arch and the deep and heavy mandible,show that L. densa should be placed in this genus. It is curious that of the lower incisors [о is smaller than I, or ls. № would seem, therefore, that the superior incisor which has been lost is P. Also, if C. brevifacies has but two lower incisors, it is probable that the second rather than either the first or the third has been lost. This is a most unusual character and one which is to be found only in a very few groups of mammals. In size, C. densa is about the same as that of C. simus although its molar series is proportionately much longer and in this, as well as in several other characters, it is intermediate between this species and C. lulliamus, the largest known member of this genus and one which is found in the same beds. There seems to be а marked amount of variation in the species of C. densa. For example, two specimens M. С. Z. Nos. 2,067 and 2,851 (see Plate 25) were found at exactly the same level and only a few feet 1 Communication by letter. SCHLAIKJER: NEW FOSSIL VERTEBRATES 163 apart. The former is unmistakably C. densa. No. 2,851 differs from this specimen principally in that it is somewhat larger and more robust and in the greater development of the internal crest on P?. Ап evalua- tion of these features as specific characters seems unwarranted. They may be accounted for merely as sex differences. Moreover, that No. 2,851 is a hybrid of C. densa and the large and more specialized form C. lullianus which occurs in the same beds, is not improbable. _ Another significant feature, shown in the material at hand, which is present in this species and which is likely to hold true, when more complete material is known, for all the other species and perhaps for Leptauchenia as well, is the change т skull proportions in young and old individuals. Three specimens in the Harvard collection, one, No. 2,876, а very young individual with only the first of the molar teeth in permanent position, another, No. 2,877, with the first two molars In permanent position, and a third, No. 2,866, which is а rather old individual, show that from youth to old age the skull becomes more brachycephalic (see Table 4). M.C.Z. No. 2876 М.С.2. No. 2877|M.C.Z. No. 2866 very young young rather old Greatest; length of skull 108 133 140 I i | Width across malar arches 70 94 102 C у. d Cephalic index .648 707 728 | Tasty 4. Comparative skull measurements and cephalic indices of, three Specimens of Cyclopidius densa. Not only in skull proportions is there a change from youth to old age, but the skull in general seems to become heavier in its construction. his is especially true of the malar arches and of the sagittal andlamb- doid crests. Likewise, in the dental series marked changes take place With age. For example, each molar tooth is sphenoidal in outline, — that is, toward the base of the crown the antero-posterior dimension diminishes while the transverse dimension increases. Therefore, as the tooth is worn it becomes broader and shorter. Also, in some species (ex. C. heterodon) the metaconule is smaller and the protocone is larger at the base of the crown, thus the former becomes smaller and the latter larger with wear. It would seem probable that a prevalence of these features will also be found in Leptauchenia. nee ET ыы == 164 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY LEPTAUCHENIA MINORA!, Sp. nov. Туре. The palate of a young individual with 1°, Рі, DP2-* and Ма, М. С. Z. No. 2,841. Collected by Erich М. Schlaikjer, 1932. Horizon and. locality. Lower Miocene. Lower Harrison formation, approximately 200 feet above the Brule-lower Harrison contact. N.W.14. Sec. 36. T. 20N. R. 62W. Goshen County, Wyoming. Specific characters. The smallest known Leptauchenia,— about one fifth smaller than Г. nitida of the upper Oligocene. Molars propor- tionately long and transversely constricted. Styles not greatly de- veloped. Mesostyle reduced and does not overlap the external wall of the paracone. Palate relatively narrow. Alveolar portion of the premaxillae more reduced and the incisors smaller than in L. nitida. (See Plate 41). Discussion. Since it has been shown that Leptauchenia? (Sespia) californica Stock and Leptauchenia densa Loomis belong to the genus Cyclopidius, there are only three species which can be referred to Leptauchenia. They are L. nitida Leidy, L. decora Leidy and L. major Leidy. The first two are unquestionably from the upper Oligocene and the latter is either upper Oligocene or lowermost Miocene in age. The new species named above adds a fourth to the list and cannot be confused with any of the others because of its size, the compressed molars, the reduced mesostyle and the reduced incisors. The deciduous premolars three and four differ from those of Cyclo- pidius principally in being less molariform. The posterior portion of ЮР? is not nearly so transversely expanded and the internal cusps of DP are more rounded. In Cyclopidius DP* is completely molariform. (Dr. Loomis, 1925, p. 246, Figure 4, in his drawing of the deciduous premolars of C. densa designates, through a lapsus pennae, DP? аз ОР). On first appearance it would seem that the small mesostyle and the compressed molars are primitive characters. It is more probable however, that these are specializations since they are expressed even to a greater degree in C. heterodon. The reduction of the alveolar portion of the premaxillae with the small and very crowded incisors 1$ a further specialization in the Cyclopidius direction. Of the known forms, the nearest relative of L. minora is L. nitida, from which species it could have been derived,— or possibly from an earlier and more primitive species. It represents the last known form of this upper Oligocene genus existing on into the early Miocene. (See Figure 11). ! From the Latin meaning "little опе.” SCHLAIKJER: NEW FOSSIL VERTEBRATES 165 Measurements mm. ТОО of dental senes PM 22 1 ЖА 1 1S ЖЕ ao 462 Tederi at Pi to Ре аьаа оиву а ANDAR ba M AME C Doce TERI OMS 20.0 Greatest lengthook Mi ы Ee РО ee. 9.9 ОТА length ОМ rt 11.6 idih Across the cantes. 10000 л л RUM des 14.1 Width across the first superior їшо]агз..............1.../. 32.0 LEPTAUCHENIA and CYCLOPIDIUS Аз stated above, four species are now assigned to Leidy's genus Leptauchenia. Of these L. nitida is the most primitive, L. major the most specialized and Г. decora is an admirable intermediate form between these two species. Although all three forms occur in the same beds such a phylogenetic arrangement is probable. L. minora is a small rather specialized species which lived in early Miocene times and which probably had L. nitida as its direct ancestor. These re- lationships are graphically illustrated in Figure 11. There has been considerable discussion concerning the validity of the various species referred to the genus Cyclopidius. Dr. Matthew (1899, p. 73) considered C. brevifacies (Cope) as synonomous with C. simus Cope. He also regarded C. decedens (Cope) as the permanent and C. heterdon (Cope) as the milk dentition of a smaller species. In 1925 Dr. Loomis (p. 248) held that all of these species belong to C. Svmus. All of this material was collected in the lower Miocene, lower Deep River beds of Montana. The Goshen Hole specimens aid greatly m clarifying this problem. As shown previously, C. heterodon is un- Questionably a valid species. Of the three upper teeth preserved in the type of C. decedens the second is the very molariform ОР“ and not M! as suggested by Professor Cope and Dr. Matthew. This type is therefore, only a young individual of C. simus. C. brevifacies is a short, broad and deep-jawed form which certainly cannot be referred to any of the other species. It is not improbable that when the skull is nown, C. shucherti Thorpe,— the short and very broad-skulled form, may be referable to this species. In 1884 Professor Cope described C. emydinus, (Amer. Mus. No. 8,115) a splendid skull of an old individual, from the same beds as - simus but considered it as distinet from this species in the following 106 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY characters which he listed (p. 553): “First, the external vertical ridges or crests of the true molars are directed obliquely forwards so as to overlap the external wall of the anterior crescent much more exten- sively than in C. simus. (2) The crown of the true molars have а relatively greater transverse diameter. (3) There is a peculiar process at the external base of the otic bulla, between the paroccipital and postglenoid processes, which may be called the subtympanic process. (4) There is no median occipital keel. (5) The maxillary bone is pro- longed posterior to the last superior molar, which is not in C. simus. (6) The oblique orbitosphenoid ridge is wanting. (7) The ойс bullae are shorter and wider in their form. This character will require con- firmation by examination of many individuals." Both of these speci- mens are in the American Museum and recently they were further prepared,— thus allowing a closer comparison to be made. Dr. Walter Granger, through the courtesy of the Department of Vertebrate Palaeontology has permitted me to reexamine these types and has supplied me with photographs so that C. simus, the genoholotype could be illustrated. (See Plate 26). After a careful inspection of both types, Г am unable to observe any character or characters in C. emy- dinus which distinguish it specifically from C. simus. A reconsideration of the seven characters listed by Professor Cope may Ъе summarized as follows: 1. The anterior projection of the mesostyle becomes greater near the base of the tooth crown. The teeth of C. emydinus are more worn than those of C. simus, hence the mesostyles seem more developed. 2. As stated previously, with wear the molar teeth become shorter (antero- posteriorly) and broader. This, together with a small amount of varia- tion in the species, accounts for the “relatively greater transverse diam- eter" of the molars. 3. The “subtympanic process” is the hyoid process which is partially em- bedded in the bulla. It is present in C. simus as well as in all species of this genus. 4. The absence of a median occipital keel seems to be an old age character coincident with a rugose and heavy development of the lambdoid crest. 5. А prolongation of the maxilla behind M? might also be accounted for a8 an old age character, or an individual variation. 6. The “orbitosphenoid ridge" is faintly present in both, a seeming emphasis of it in C. simus was due to crushing. 7. There is little, if any, difference in the dimensions of the bullae in the two species. In C. simus the bulla is somewhat transversely crushed. SCHLAIKJER: NEW FOSSIL VERTEBRATES 167 Dr. Loomis (1925, p. 248) has rightfully regarded C. incisivus Scott, also from the lower Deep River beds, as belonging to C. emydinus, Which of course, in the light of the above interpretation, is equivalent to C. simus. Geological formation Skull Length | Length length of Pms. | of Ms. L. nitida Leidy Upper Brule 88.9-100.0| 15.8-22.0 | 20.1-25.0 А ж й 2 ds L. decora Leidy Upper Brule 101.6-110.0) 14.8-20.0 | 29.6-31.7 L. major Leidy Upper Brule? 1280-1460 25.4—85.0 | 40.2-42.3 ыыы L. minora sp. nov. Lower Harrison ? | ? % C. lullianus Thorpe Lower Harrison 144.0 30.5 44.2 а Lais ai i EA | олке — —— | ———— C. densa (Loomis) Lower Harrison 123.0-134.0| 25.0-27.0 45.0 — шылар С. simus Cope Lower Harrison (Lower Deep River beds)|126.8-129.8| 24.0-25.0 | 33.5-35.5 о о C. schucherti Lower Miocene Thorpe (Lower Harrison?) % 20.0 28.5 A. —_ — MM (Eh brevifacies Lower Harrison (Cope) |(Lower Deep River beds) ? % ү C. heterodon Cop Lower Harrison 89.0 18.0-19.0 | 23.0-25.5 ое ie e californica Lower Miocene (Stock) (Lower Harrison?) % 15.6 Y umen ary 2. TanLE 5. А list of the species of Leptauchenia and Cyclopidius with the 8eological level and three salient measurements of each. The species of Leptauchenia and Cyclopidius which, for the present, may be regarded as valid are listed in Table 5. n Figure 11 is given a tentative phylogeny of all the known species This is primarily intended to show Little is known of the ancestry of these two of Leptauchenia and Cyclopidius. structural relationships. —o—HÉ— 168 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY genera. Of the scanty remains described from the early Oligocene de- posits, Limnentes Douglass seems, as regarded by Loomis (1924а, p. 15), to be the most likely ancestor, although, when more completely known, it may be found to be nearer the Eporeodon line. C. lullianus te) Ё BE 31.2 BARRE C.densa | C.helerodon Qis E | .|І.тіпога C.brevifacies C.schucherti | C. simus С. californica A |A ~ НЕ ES Ss MAE N: A X NET L. major rj E Z met fa | © зу = L. decora, CIE „| OQ у (do; БК L. nitida, БЕ е | о ? Limnentes (Lower Oligocene) Ета. 11. Phylogenetic chart of the species of Leptauchenia and Cyclopidius. Schlaikjer 1934. A study of the skulls in particular of the Leptauchenia-Cyclopidvus group of oreodonts reveals a number of important facts of which the following are the outstanding: 1. The molar teeth are sphenoidal in general form,— that is, the antero- posterior dimension is less and the transverse dimension is greater near the base of the crown. 2. With wear, therefore, the molar teeth become proportionately shorter and broader. 3. There is considerable variation in the premolar teeth, especially in Р?. 4. The incisor roots are frequently short, particularly those of П-, and when only a small amount of the superior portion of the symphysis is broken away some or all of the alveoli may be destroyed. 5. It appears that with age the skull becomes more brachycephalic. 6. There is à considerable variability in size which may be accounted for primarily as a sex difference. The somewhat larger and more heavily built skulls are probably those of males. A careful consideration of these facts, before new species are described, would certainly seem advisable. 2 SCHLAIKJER: NEW FOSSIL VERTEBRATES 169 MERYCHYUS HARRISONENSIS Peterson Material. Left maxilla with P2-M3 and the root of РЕИС 7, No. 2,869. (See Plate 41). Collected by Erich M. Schlaikjer, 1932. Horizon and locality. Lower Miocene. Lower Harrison formation, Approximately 125 feet above the Brule-lower Harrison contact. N.E.14. Sec. 29. T. 25N. R. 64W. Goshen County, Wyoming. Discussion. The teeth of this specimen, in size and even in the Stage of wear, are identical with those of Peterson's type (1906, p. 37- 40). One of the more important features of the dental series is the rather elongated P* with a strong anterior intermediate crest which divides the front part of the tooth into two basin-like areas. There is also present a weakly developed median crest. This dental character 18 one which seems more typical of the genus Ticholepus rather than of Merychyus. Dr. Loomis (1923, p. 227) has referred this species to the former genus mainly on the basis of size and skull proportions. - harrisonensis is, however, smaller than the “typical” Ticholepus. "hen too, the dental series is shorter than that of such species as T. arenarum Cope and M. siouxensis Loomis. For the present, there- оте, I choose to retain this species in the genus Merychyus. The occur- тепсе of this species in the lower Miocene Beds of the Goshen Hole Area is of significance since it adds to the other corroborative evidence Which substantiates the correlation of these beds with the lower Harri- 80n of northwestern Nebraska. Measurements " mm. а 37.0 оша No tou Ae RE. 44.5 EE ue ue E RU LIS s 10.0 БН onis viu cuo uo fuu. 9.5 BEES Ne es у ло cnt RI. 8.0 КОО ДН у у л ме 8.6 КОО O ааа а“ 11.5 О 12.0 О С dI о 13.5 ength of М2. 15.0 Bene ne. a 15.0 У ОКО a у у A ge | 20.0 COD RO ЫНЫ о ыл cane Mae 16.1 очнь 170 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY MERYCHYUS MINIMUS Peterson Material. Right mandible with P:-M3, М. С. Z. No. 2,868. Col- lected by Erich М. Schlaikjer, 1934. Horizon and locality. Lower Miocene. Lowermost upper Harrison formation, approximately 500 feet above the Brule-lower Harrison contact. 5.Е 14. Sec. 3. Т. 24N. К. 65W. Goshen County, Wyoming. Discussion. This specimen is that of a mature yet fairly young individual and it corresponds in almost every respect with the type of M. minimus Peterson from the upper Harrison formation of Sioux County, Nebraska and is, therefore, of stratigraphie importance. It was found at exactly the same level as Parahippus wyomingensis which is also an upper Harrison form. Measurements mm. Height of mandible at coronoid process..................... 78.0 Depth of mandible at anterior о Mao. 7180070 26.0 Волин ot lower Тео оо. M length oe lower molars: бен еқ аты dea taam 44.0 EPOREODON CHEEKI Schlaikjer Material. А nearly complete skeleton, the type, M. С. Z. №. 17,765. (See Plates 27, 28 and 29). The skeleton of a young individual M. С. 7. No. 2,807, and an immature skull and jaws, М. С. Z. №. 2,870. Collected by Erich M. Schlaikjer, 1932-1933. Horizon and locality. Lower Miocene. Lower Harrison formation. Goshen County, Wyoming. Discussion. The type was fully described by Schlaikjer (1984, р. 220-223) and is now figured in Plates 27, 28 and 29. As pointed out previously, Е. cheeki is the last known survivor of this genus existing as a living fossil in the early Miocene along with such genera as Mesoreodon and Promerycochoerus that are descendants of the upper Oligocene members of this genus. MESOREODON CHELONYX Scott Material. Ап excellent skull, М. С. Z. No. 2,814 (see Plate 30), and two jaws М. C. Z. Nos. 2,819 and 2,824. Collected by Erich M. Schlaikjer, 1932. | | | SCHLAIKJER: NEW FOSSIL VERTEBRATES A Horizon and locality. Lower Miocene. Lower Harrison formation. M. C. Z. No. 2,814 from Bear Creek Mountain and M. C. Z. Nos. 2,819 and 2,824 from 66 Mountain, Goshen County, Wyoming. Discussion. This material is practically identical with the type as described and figured by Professor Scott (1893). The occurrence of this species, as well as such forms as Cyclopidius simus and C. heterodon, m the lower Harrison formation of the Goshen Hole Area is significant since it shows that these deposits (as stated earlier in this paper) can be considered as of the same age as the lower Deep River beds of ontana. MESOREODON MEGALODON Peterson. Material. A complete skull and jaws, adult, M. C. Z. No. 17,751. (See Plate 31). A pair of jaws M. C. Z. No. 2,822 and the jaws of a young individual M. C. Z. No. 2,817. Collected by Erich M. Schlaikjer, 1932-1933. | | Horizon and locality. Lower Miocene. Lower Harrison formation. - С. Z. No. 17,751 from approximately 200 feet above the Brule- lower Harrison contact. 5.Е.М. Sec. 21. T..20N. R.-60W. M. C. Z. No. 2,822 from the same level at Fox Creek Gap. M. C. Z. No. 2,817 ound at the same level and a few feet away from the type of Hypo- (Фрив avus, approximately 100 feet above the Brule-lower Harrison Contact. 5.У/ 14. Sec. 16. T. 20N. В. 63W. Goshen County, Wyoming. Discussion. The type of Mesoreodon megalodon Peterson consists of the front of a skull and jaws collected from the “Monroe Creek beds",— lower Harrison of Sioux County, Nebraska. Dr. Loomis (1933, p. 727) has referred a skeleton to this species from the lower locene (lower Harrison) beds at Muddy Creek in eastern Wyoming, аз well as several skulls and considerable limb material from 66 ountai in Goshen Hole. The material in the Harvard Museum Corresponds identically with Peterson's type. This material occurs at the same level as Merychus harrisonensis a typical lower Harrison Species and is additional evidence, therefore, as suggested in the Stratigraphic section of this paper, that the Monroe Creek and lower arrison beds are synchronous deposits. Мезовеором зсоттт Schlaikjer ғ This Species was based on a nearly complete skeleton, M. C. Z. 9. 17,480 (1934, p. 223-225) and is figured in Plate 32 of this paper. © Specimen was collected by Dr. Е. B. Loomis from the lower N О | | А 172 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Harrison formation at Muddy Creek, Wyoming. No specimens refer- able to this species have as yet been found in the Goshen Hole Area. Nevertheless, at Muddy Creek M. megalodon occurs in the same beds, and it, as well as M. chelonyz, is also found in the lower Harrison of Goshen Hole. АП three species, therefore occur at the same geological level yet they form an excellent evolutionary sequence, structurally, in this genus with M. chelonyx the most primitive, M. scotti the inter- mediate and M. megalodon the most advanced species. PROMERYCOCHOERUS LOOMISI Schlaikjer The type of this species is an excellent skeleton; M. C. Z. No. 2,820, Which was collected from the lower Harrison formation, Bear Creek Mountain, Goshen County, Wyoming. It is figured in Plates 33-35 of this paper. As stated previously (1934, p. 225-229) P. loomisi is a large-skulled, short bodied and long-limbed form. The more outstand- ing features of this species are: the very large and narrow skull, the large but short body and the relatively long limbs. It is an excellent example of gigantism among the oreodonts. P. grandis may be con- sidered its direct ancestor and since the smaller and more primitive species P. hatcheri occurs in the same beds it would seem that Р. loomisi represents а very progressive evolutionary line of Promery- cochoerus. PROMERYCOCHOERUS HATCHERI Douglass Material. An excellent skull, M. C. Z. No. 2,815 (see Plates 36, and 37). А skull and jaws with atlas, axis, two cervicals and an astrag- alus, M. C. Z. No. 17,760 (see Plate 38). A skull and jaw, M. C. Z. No. 2,816, two pairs of jaws, M. C. Z. Nos. 2,821 and 2,825 and the skull and jaws of an old invididual, M. C. Z. No. 2,816. Collected by Erich M. Schlaikjer, 1932-1933 Horizon and locality. Lower Miocene. Lower Harrison formation. M.C.Z. Nos. 2,815, 17,760, 2,816, 2,821, and 2,825, from the Eporeodon cheeki-Mesoreodon megalodon level 66 Mountain, M. C. Z. No. 2,816 from the same level on Bear Creek Mountain, Goshen County, Wy- oming. Discussion. The type of this species was collected from the lower Deep River Beds, Montana, and was briefly described by Douglass, 1907. The dental series was not figured by Douglass but so far as can | | | | SCHLAIKJER: NEW FOSSIL VERTEBRATES 178 be judged from his description and figure of the skull the Harvard Specimens are almost identical with the type except for the smaller size and the slightly more posterior position of the infraorbital foramen, Which characters I regard as individual variations. A typical Promery- cochoerus dental character is the separation of the internal crescent from the median crest in Ра. The pattern of this tooth is primitive, however, in that the posterior crest is not united posteriorly with the posterior crescent (see Plate 37). 'The dental and cranial feature of this Species are structurally intermediate between Eporeodon cheeki and Promerycochoerus loomisi all of which occur at exactly the same geo- logical level. There appears to be a considerable amount of variation in this Species. M. C. Z. No. 17,760 (see Plate 38), for example differs from MC. Zi Ne. 2,815 (see Plates 36 and 37) only in its somewhat smaller size and lighter construction. These differences are probably sexual, — No. 17,760 being the skull of a female. Measurements МОЛ MS. 2,815 17,760 57 9 Greatest length of the skull...................... 280.0 263.0 аав length лл 75.5.7, веб 9400 Greatest breadth across squamosals on malar arches. 156.0 % Length of dental series (C-M inclusive) ........... 133.0 129.0 Length of dental series (PI-M3 inclusive) .......... 1180 1140 istance mom Peto Ba external ЛЕТ ear 550 52; Istance from Mi to M? external................. 69.0 ? MERYCOIDODON CULBERTSONI (Leidy) Of all the fossil mammals which occur in the middle Oligocene deposits of South Dakota and Nebraska, Merycoidodon culbertsoni is Y far the most numerous. Yet, in the Goshen Hole Area with its extensive and abundantly fossiliferous middle Oligocene only a single fragmentary skull and a few isolated teeth were found. This may be accounted for by the fact that the oreodonts were gregarious and ex- Temely localized in their distribution. t | | | | | | | | | piden e — en 174 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Suborder TYLOPODA Superfamily CAMELOIDEA Family CAMELIDAE MIOTYLOPUS BATHYGNATHUS gen. et sp. nov. Type. А partial skull and jaws and an associated pelvis with most of the sacrum, M. C. Z. No. 2,924. Collected by Erich M. Schlaikjer, 1932. Horizon. and locality. Lower Miocene. Lower Harrison formation, approximately 150 feet above the Brule-lower Harrison contact. Х.Е.М. Sec. 20. Т. 20N. R. 60W. Goshen County, Wyoming. Generic characters. Orbit incompletely closed. Tympanic bullae relatively small. Molars brachyodont, relatively small and without mesostyles, except a faint indication of опе on МЗ. Postcranial skeleton, as indicated by the pelvis, probably proportionately small and of light construction. Specific characters. Inferior premolars simple. P; separated from the canine by a long diastema which is subequal to that which separates P; from Py. Styles on inner walls of inferior molars absent or feebly developed. Ramus narrow and relatively deep. Infraorbital foramen above the paracone of M!. Interorbital breadth across frontals very great. (See Plate 39). Description and discussion. In the basicranial area there are a number of rather unusual features. "The basicranial axis has a very gentle slope. The basioccipital is wide, has a short median crest on either side of which is a large and shallow depression which is bounded laterally by a low sharp ridge. Where the basioccipital and basi- sphenoid unite there are two elongated, prominent and rugose protu- berances developed. The left and the posterior part of the right pala- tine and most of the pterygoids are broken away. Posteriorly the latter are widely separated. The tympanic bulla is relatively small, — actually not any larger than that of Poöbrotherium yet the skull is almost twice as large as that of the latter. It possesses only a shallow posterior vertical groove. This is in marked contrast to the deep ravine-like invagination of the bulla of Poébrotherium, Paratylopus and Oxydactylus. The paroceipital processes are narrow and short. They project backward and considerably overhang the very deep and rounded condyloid fossae. There are two hypoglossal foramina in the SCHLAIKJER: NEW FOSSIL VERTEBRATES V left fossa and one in the right. This difference in number seems to be а variable character among the Artiodactyla. The occipital condyles are low and are relatively very broad. The articular surface of the ventral portion is approximately one third greater than the dorsal surface. Ventrally the condyles are separated by a shallow groove. The supraoccipital is low. Ventrally it is trans- Versely convex and dorso-medially it presents a strong crest. The lambdoid erest and the posterior part of the sagittal crest are strongly developed. The postorbital region is proportionately long and the cranium is broad. The frontals are fairl y short, but very wide especially between the orbits. The postorbital process is incompletely preserved although the position and unusual thinness of the considerable portion which is Preserved shows, together with the gentle curvature of the malar border, that the orbit was probably not completely closed. Although the facial region is missing, the jaws indicate that it was relatively Short, narrow and probably not expanded across the alveolar portion of the premaxillaries. The antero-external margin shows that the ant- orbital vacuity is of moderate size. The lacrimal forms the upper half of the anterior margin of the orbit and extends far forward on the face, The facial crest is low and knob-like. The infraorbital foramen 15 above the center of the М! paracone. The palate is very narrow as 15 shown by the preserved portion of the right maxillary and palatine. The jaws are narrow and the symphysis 1s short and steep. Even though the individual was full y adult the rami are not firmly united. he ramus is long, very narrow and exceptionally deep especially Under P, Its inferior margin is almost straight. The infraorbital foramen is large and is situated under the anterior of P. The rather complete pelvis and sacrum associated with the type, and probably from the same individual, shows that in comparison With the skull, the postcranial skeleton was proportionately small. The small size of the acetabulum would indicate a rather slender-limbed orm. Compared with the other genera of camels from the lower Miocene, M totylopus bathygnathus shows no close affinity with any of the known orms. The open orbit, the relatively small bullae and the low-crowned almost styleless molars are Poébrotherium-like, while the relatively short face and large cranium, the deep ramus and the simplified inferior Premolars are habitus characters. In the absence or feeble development of styles on the inner walls of the inferior molars, this form is close to aratylopus sternbergi of the John Day; and in the simplified inferior 176 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY premolars it is close to Protomeryx, also of the upper Oligocene. These seem to be characters of convergence, however, since in the large assemblage of other characters, Miotylopus bathygnathus is so strikingly different from either of these two genera. Recently the writer had the privilege of inspecting a fine camel skull which was collected by Dr. Chester Stock from the lower John Day of the Sespe. This specimen is genericly distinct from Miotylopus and will soon be described by Dr. Stock as a new genus. It possesses а number of characters which seem intermediate between Poébrothervwm and Miotylopus and may be the upper Oligocene progenitor of the latter. Measurements mm. Condylo-basal length of skull (estimated)................... 220.0 ИЗМЕНА ОЈ МЕРУ ND n I c eL хул м ш 41.8 Length of inferior dental series Сз. з. 115.0 Lente of infenor premolars. ма crux cae UM ЕСЕ Ghinterion ШОН ЫТ kre eG LATI ea 48.5 ШШ, in dora bie bois eek на уэ ллу MEC лу А 18.2 IDEM Gono, Udo Р nt 18.5 ШЕРІН Отау under Es a о an een ео Interorbital breadth across frontals............4........... 716 Breadth across squamosals on cranium..................... 568 Tau ot OCGIDILAL Со 36.0 Antero-posterior diameter of acetabulum................... 240 ШОПОВ О pela ұлан алына њи na 80.0 STENOMYLUS HITCHCOCKI Loomis Material Ап excellently preserved posterior portion of a skull and jaws with M3, M. C. Z. No. 2,925. Collected by Erich M. Schlaikjer, 1981. Horizon and locality. Lower Miocene. Lower Harrison formation; approximately 24 feet above the Brule-lower Harrison contact. S.W.14. Sec. 36. Т. 20N. К. 62W. Goshen County, Wyoming. Discussion. This specimen is almost identical with the type of S. hitcheock? which was collected from the famous Stenomylus quarry four and one half miles east of Agate, Nebraska. It differs from the type specimen only in a somewhat longer saggital crest and in а SCHLAIKJER: NEW FOSSIL VERTEBRATES 177 slightly wider and shorter M?. Also, the mandible is somewhat more shallow under Мз. These differences are of little significance, however, Since this species displays a considerable amount of variability, as is Shown in the many skeletons collected from the same quarry as the type. For example, one of these skeletons in the American Museum, No. 14,227, shows the same variations from the type as does the Goshen Hole specimen. Stenomylus, in so far as is known, is a lower Harrison genus and the occurrence of S. hitchcocki in the Goshen Hole Area is of particular importance stratigraphically. PROTOMERYX sp. The writer has been informed by Dr. Frederick Brewster Loomis that in 1927 some local people discovered a nearly complete skeleton of Protomerya in the lower Harrison beds on 66 Mountain. The specimen Was collected and presented to the University of Wyoming Museum. Recently Dr. Loomis inspected the specimen and found it to represent а new species. It has been turned over to him for study and a descrip- tion of it is soon to appear. It is probable that two specimens, М. С. Z. Nos. 17,763 and 2,068, both of which are the facial portions of skulls of young individuals, belong to the same species as the above. They were found by the Writer in the same beds and in the same locality. They have been turned over to Dr. Loomis for study. OXYDACTYLUS sp. Collected in the same block of sandstone containing the skeleton of P arahippus wyomingensis (see above) was an almost complete left front foot of Oxydactylus, M. С. Z. No. 17,762. The metacarpals are Approximately thirty millimeters shorter than those of O. brachyodontus eterson. During the past field season the writer collected an unusually Complete skull and jaws and partial skeleton from the same beds. his individual has an especially short skull and a relatively long pre- Molar-series. It is undoubtedly a new species and when it is prepared or study it will be fully described in a subsequent paper. It is probable that the above mentioned foot is referable to the same species. 178 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Suborder PECORA Superfamily TRAGULOIDEA Family HYPERTRAGULIDAE NANOTRAGULUS INTERMEDIUS Sp. nov. Type. The skull and jaws of an immature individual, M. С. Z. No. 2,103. Collected by Erich M. Schlaikjer, 1931. Horizon and locality. Lower Miocene. Lower Harrison formation, approximately 200 feet above the Brule-lower Harrison contact. N.W.14. Sec. 36. Т. 20N. К. 62W. Goshen County, Wyoming. Paratype. Ап adult skull with the right maxilla and teeth missing, М. С. 7. No. 2,812. Collected by Erich M. Schlaikjer, 1933. Horizon and locality. Lower Miocene. Lower Harrison formation, approximately 120 feet above the Brule-lower Harrison contact. 5.Е.М. Sec. 22. T. 20N. К. 60W. Goshen County, Wyoming. Specific characters. The smallest known Nanotragulus. Molars, especially M!, quadrangular in outline. P with internal cusp present but very small. Parastyle of M? does not overlap M? as much as in М. loomisi. Also, the bulla is smaller and more compressed than in that species. Facial crest reduced. Orbit almost closed behind. ls larger than other incisors. Р» and P; closely appressed. (See Figures 12-13 and Plates 40-41). Description and discussion. The basicranial region is proportion- ately long. The basioccipital is broad and the tympanic bulla is smaller and more transversely compressed than in N. loomisi. The paroccipital process is short and compressed. At its base it is firmly braced against the bulla immediately behind which the hyoid process is deeply em- bedded. The whole basicranial region is very similar to that of Tragulus except that in the latter there is a pterygoid process on the anterior of the bulla and the petrosal bone does not appear on the surface between the bulla and the basioccipital. The external auditory meatus is in- complete and projects upward and backward. A short distance above there is a large squamosal foramen. The occipital condyles are low, broad and rather distantly separated,— more so above than below. The supraoccipital is high, narrow and transversely convex. The mas- toid portion of the periotic extends far up on the side of the skull and is amost in contact with the parietal. The lambdoid and sagittal crests are weakly developed. The cranium is arched and is somewhat ex- panded. There is a slight postorbital constriction. The postorbital bar SCHLAIKJER: NEW FOSSIL VERTEBRATES 179 | та 19.7 Nanotragulus intermedius. Туре. Dorsal, lateral and palatal Views of the skull. Natural size. Drawn by Helen Ziska. 180 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY is almost complete. 'l'he malar is large and elongated. On its external surface there is a prominent ridge which extends forward and grades into the short and low facial crest of the maxilla. The lacrimal is large and occupies a considerable portion of the face immediately in front of the orbit. The facial region is short, deep and very narrow. The infraorbital foramen is above the second molar. It occupies a slightly more posterior position in a young individual. The palate is narrow and the internal nares are rather far forward (opposite the posterior of M2). There is an internal cingule on each of the upper molars of the type specimen. The cingule is absent on the teeth of the adult paratype skull. Since in almost all of their other characters these two specimens are so much alike the presence or absence of a cingule may be regarded as an individual variation. The canine is small and slightly compressed transversely. P! is minute and is situated about midway between the canine and Р». P? is set close to P?. It is two-rooted and has a single- cusped crown which is oval in outline. P? has the general form of E^ although with the internal cusp much reduced. In premolar develop- ment, therefore, №. intermedius forms an ideal intermediate stage be- tween N. loomisi and Hypertragulus. DP? has three external and one postero-internal cusp. РР“ is completely molariform. The mandible is long and slender. Posteriorly it is deep and ex- panded. The mandibular condyle is large, flattened and considerably elevated above the tooth-row. The coronoid is high, slender and re- curved. The symphysis is fairly short and narrow. There is а short diastema between the canine and Ру. The canine is completely incisi- form and is of the same size аз Із. I, and Із are larger. Р. is caniniform. There is no diastema between Р» and Рз. DP, is of the typical artio- dactyl three-erested pattern. DP; has the same general pattern except that the anterior crest is less developed. "There is a minute external cingule on each of the lower molars. Previously two species have been referred to the genus Nanotragulus and the above description adds a third. The type N. loomisi from the lower Harrison of eastern Wyoming was described by Professor Lull (1922, pp. 116-119) who immediately recognized its hypertrag- ulid characters and placed it in the family Hypertragulidae but stated that, “itis not clearly derivable from any known Oligocene form except possibly Hypisodus” (р. 119). Dr. Matthew (1926, p. 3) assigned а fragmentary skull and jaws in the “lower Rosebud” collection at the American Museum to Professor Lull's genus and species and pointed out its close affinity to Hypertragulus calearatus from the middle oli- SCHLAIKJER: NEW FOSSIL VERTEBRATES 181 gocene of South Dakota. He also referred H. ordinatus Matthew (only the jaws of this species are known) to Nanotragulus. Dr. Loomis (1933) has described a fragmentary skeleton, also from the “lower Rosebud,” which he regards аз N. loomisi. In his discussion of the Hypertragulus- Nanotragulus affinity he considered the former “to be the immediate predecessor of Nanotragulus.” . Fig. 13. Nanotragulus intermedius. Type. Lateral view of the left lower Jaw. Natural size. Drawn by Helen Ziska. ‚ On the basis of its rather small and compressed bullae, its incipient Internal cusp on P? and its small size, N. intermedius is the most primi- tive of the described Nanotragulus species. Structurally it forms an ideal predecessor for N. loomisi and in the above characters it is inter- Mediate between this species and Hypertragulus. N. ordinatus (Mat- thew) is the largest species and could be the lineal descendant of N. loomisi. Of the known species of Hypertragulus, H. minutus Lull from the upper Oligocene (upper John Day) of Oregon, is nearest to Nano- tragulus and may, tentatively, be considered directly ancestral to that genus, Measurements ; тато E o baal length... o rogi о 72.0 оо skull qbo dia р Ld 80.0 Greatest breadth across squamosals on eranium............. 23.5 Interorbital bresdth Across montala r ices am is са са а 2155 Istance from anterior rim of orbit to front of ршх.......... 34.5 Istance from anterior rim of orbit to supraoccipital......... 49.0 Greatest width of condyles -o;o о a ОО 14.0 Istance from ventral portion of exoce. to top of supraoce. ... 25.5 182 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY Breadth of muzzle at entrance of infraorbital foramina. ...... 9.5 Distance from anterior of canine to anterior of P?............ 14.0 ЕИО SIS Ба оба ОО о л, De 25.5 Пій А ЕМ Me ME E UA 16.0 Distance from posterior of I; alveolus to mandibular condyle 58.0 Greatest length ої mandible a... ован рат Qr i e Bensth ое зуру и 12.0 Бер OL jaw under. сотопо о о 29.5 ДАДА ООП ТЕ Ме pa Den пок лик Dear ua 6.5 Length of diastema between Р; and Рә.................... 70 Distance rom. to Ре 20.8 ace ШОШ Mite Mass cio uo a ж ie cQ 28.0 HYPISODUS ALACER Troxell Material. A left inferior third molar with a fragment of the man- dible, M. C. Z. No. 2,873. Collected by Erich M. Schlaikjer, 1932. Horizon and locality. Middle Oligocene. Brule formation, upper part of the middle oreodont zone. Harvard Fossil Reserve. Goshen County, Wyoming. Discussion. This specimen is slightly larger than the M; of H. alacer and differs from that species principally in the more reduced heel and in the more transversely compressed character of the meta- cone-metaconule crest. The discovery of more complete material may prove this form specifically distinct from H. alacer. For the present, however, it may be referred to that species. LEPTOMERYX EVANSI Leidy Material. A fragmentary skull and jaws, М. С. Z. No. 2,926. Col- lected by Erich М. Schlaikjer, 1932. Horizon and locality. Lower Oligocene. Chadron formation. N.W.14. Бес. 86. T. 24N. В. 63W. Goshen County, Wyoming. Discussion. The occurrence of this species in the upper level of the Chadron in Goshen Hole is of no particular significance. It has 8 considerable vertical range for it occurs up through the oreodont beds of the Brule formation, and it is found in these deposits throughout the Great Plains Region. SCHLAIKJER: NEW FOSSIL VERTEBRATES 188 LEPTOTRAGULUS ULTIMUS Sp. nov. Type. А superior left third molar, М. С. Z. No. 2,872. Collected by Erich M. Schlaikjer, 1932. Horizon ата locality. Lower Oligocene. Chadron formation. Х.Е.М. Sec. 19. T. 22N. В. 61W. Goshen County, Wyoming. Specific characters. The largest known Leptotragulus. Very heavy internal cingulum. Anterior cingulum moderately developed and pos- tero-internal cingulum absent. (See Plate 41). Discussion. Leptotragulus proavus, the type of the genus, was de- Scribed by Professors Scott and Osborn (1887 and 1890) from the Uinta formation of the Eocene. It was based on a fragmentary lower jaw With several teeth. In 1919 Mr. O. A. Peterson described (pp. 93-97) and figured, from the same formation, several upper and lower den- titions which he rightfully referred to this species. In the same publi- cation he described a new species, also from the Uinta, to which he gave the name L. medius and which he distinguished from L. proavus principally on the basis of the larger styles and absence of cingula on the superior molars. A third species “L.” profectus was described by Dr. Matthew (1903, p. 224) from the Pipestone Springs bed (Chadron) of Montana. The type is a fragmentary lower jaw with Р;-Мі. Аз Shown by Dr. Matthew, this species is a large and very cameloid form. Its inclusion in Leptotragulus was questioned by Dr. Matthew himself (1909, p. 104) and, as suggested by Mr. Peterson (1919, p. 97), perhaps should not be included in that genus. The molar tooth of L. ultimus possesses every advanced character Which would be expected in an Oligocene representative of this genus. These characters showing advanced changes over the upper Eocene forms are as follows: the large size, the proportionately broad paracone- Protocone dimension and the heavy internal cingulum. Г. proavus may be considered directly ancestral to this species. Measurements mm. Greatest antero-posterior dimension of М3,.................. 8.0 Width of M? across PALACOME-PROTOCONE 22/8 22222 ШО Width of M? across metacone-metaconule................... 8.5 | | 4 | ! f 184 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY CONCLUSIONS 1. The Chadron formation rests with conformity and disconformity on the Yoder formation and is overlain with disconformity by the Brule. Lithologically, and in so far as the fauna is known, is it indistinguish- able from the deposits of this formation in Nebraska and South Dakota. 2. The Brule is fully represented in the area. The lower channel sandstone member is probably comparable to the Metamynodon sand- stone in other areas. The rich fossil deposit at the Harvard Fossil Reserve is unquestionably the correlate of the upper nodular layer in the South Dakota Big Bandlands. 3. Most of the lower Miocene in the area is lower Harrison in age and is typically four hundred feet in thickness. The upper three hun- dred feet of the seven hundred feet of the Miocene in Six Mile, Deer and Cherry Creek vicinity is probably upper Harrison. 4. The “Gering formation" represents only a lower channel phase of the lower Harrison and should not be regarded as a distinct for- mation. 5. The Monroe Creek beds, the lower Rosebud and the lower Harri- son are synchronous deposits. The fauna and lithology are variable only in so far as they were governed by local environmental and de- positional conditions. 6. The Goshen Hole Area has been subjected to at least five dias- trophic movements since the late Cretaceous. These took place at the close of the Fox Hills, between the Cretaceous and early Oligocene, at the end of the Chadron, at the end of the Brule and since the end of the Miocene. 7. In the large collection of Oligocene and early Miocene verte- brates, two genera and thirteen species herein described are new. They are as follows: Palacolagus hypsodus, Hypolagus primitivus, Nothoeyon leptodus, Mesocyon hortulirosae, Pediohippus antiquus (a new genus), Parahippus wyomingensis, Hypohippus avus, Archaeotherium palustris, Dinohyus minimus, Leptauchenia minora, Miotylopus bathyganthus (a new genus), Nanotragulus intermedius, and Leptotragulus ultimus. SCHLAIKJER: NEW FOSSIL VERTEBRATES 185 BIBLIOGRAPH Y Арама, G. I 1898. The upper Cretaceous of Kansas. Univ. Geol. Surv. Kans., 4, pp. 13-27. 1901. Geology and water resources of the Patrick and Goshen Hole Quadrangles in Eastern Wyoming and Western Nebraska. U. 8. Geol. Surv. Water Supply Paper, No. 70, pp. 1-47, pl. 1-11, 4 text figures. Corz, E. D. 1878. Description of new Vertebrata from the upper Tertiary formations of the West. Proc. Amer. Phil. Soc., 17, рр. 219-231. 1884. Synopsis of the species of Oreodontidae. Proc. Amer. Phil. Soc., 21, pp. 508—572, 1 text figure. Скдатм, Е. №. 1896. On the stratigraphy of the Platte Series, or upper Cretaceous of the Plains. Colo. College Studies, 6, pp. 47—52. Darton, №. Н. 1899. Preliminary report on the geology and water resources of Nebraska west of one hundred and third meridian. U. S. Geol. Surv. 19th Annual Report, Pt. IV, pp. 719-814, pls. 74-118, text figures 208-230. 1903. Scotts Bluff, Nebraska, quadrangle. U. S. Geol. Surv., Geol. Atlas, Folio No. 82, pp. 1—5, 23 text figures. 1905. Preliminary report on the geology and underground water re- sources of the Central Great Plains. U. S. Geol. Surv., P. P. No. 32, рр. 1-433, pls. 1-72, 18 text figures. 1909. Geology and water resources of the Northern portion of the Black Hills and adjoining regions. U. S. Geol. Surv., P. P. No. 65, pp. 1-105, pls. 1-24, 15 text figures. Dick, В. 1917. Systematic position of several American Tertiary lagomorphs. Bull. Dept. Geol. Univ. Calif., 10, pp. 179-183, 6 text figures. 1929. The phylogeny of the їс idae, with ph La of a new genus. Jour. Mam., 10, pp. 340-344. Dovarass, E. 1908. New vertebrates from the Montana Tertiary. Ann. Carn. Mus., 2, pp. 145-199, pl. 2, 37 text figures. 1907. Merycochoerus and a new genus of Merycoidodonts with some notes on other Agriochoeridae. Ann. Carn. Mus., 4, pp. 84—98, pl. 21. 1907. Some new Merycoidodonts. Ann. Carn. Mus., 4, pp. 99-109, pls. 22-30. 1908. Fossil horses from North Dakota and Montana. Ann. Carn. Mus., 4, pp. 267-277, pls. 65-68. | 186 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY HATCHER, J. B. 1902. Origin of the Oligocene and Miocene deposits of the Great Plains. Proc. Amer. Phil. Soc., 41, pp. 113-181. ШАУ, О: Р, 1924. Description of some fossil vertebrates from the Upper Miocene of Texas. Proc. Biol. Soc. Wash., 37, pp. 1—20, pls. 1—6, 1 text figure. Kxzrroag, L. 1910. Rodent fauna of the late Tertiary beds at Virgin Valley and Thousand Creek, Nevada. Bull. Dept. Geol. Univ. Calif. Б, pp. 421- 437, 20 text figures. Loomis, Е. В. 1910. Osteology and affinities of the genus Stenomylus. Amer. Jour. Sci., 29, pp. 297—323, 30 text figures. 1911. The camels of the Harrison beds, with three new species. Amer. Jour. Sci., 31, pp. 65-70, 3 text figures. 1920. On Ticholeptus rusticus and the genera of Oreodontidae. Amer. Jour. Sci., 50, pp. 281-292, 4 text figures. 1923. Oreodons of the Lower Harrison beds. Amer. Jour. Sci., 6, рр. 222—228, 5 text figures. 1924. Miocene Oreodonts in the American Museum. Bull. Amer. Mus. Nat. Hist. 51, pp. 1—87, 26 text figures. 1924. Тһе Oreodonts of the lower Oligocene. Ann. Carn. Mus., 15, рр. 369-378, pls. 52-54, 3 text figures. 1925. Leptauchenia and Cyclopidius. Amer. Jour. Sci., 9, pp. 241-249, 4 text figures. 1932. The small carnivores of the Miocene. Amer. Jour. Sci, 24, рр. 316—829, 10 text figures. 1933. Тһе ske eton of Nanotragulus. Amer. Jour. Sei., 25, pp. 390-398, 6 text figures. 1933. Three Oreodont skeletons from the Lower Miocene of the Great Plains. Bull. Geol. Soc. Amer., 44, pp. 723-734, 9 text figures. Іллл, В. 8. 1922. Primitive Pecora in the Yale Museum. Amer. Jour. Sci., 4, рр. 111-119, 1 text figure. Marrmew, У. D. 1899. A provisional classification of the fresh-water Tertiary of the West. Bull. Amer. Mus. Nat. Hist., 12, pp. 19-75. 1902. А horned rodent from the Colorado Miocene. With a revision of the Mylagauli, beavers and hare of the American Tertiary. Bull. Amer. Mus., 16, pp. 291—810, 17 text figures. 1903. The fauna of the titanotherium beds at Pipestone Springs, Mon- tana. Bull. Amer. Mus., 19, pp. 197—226, 19 text figures. SCHLAIKJER: NEW FOSSIL VERTEBRATES 187 1907. A lower Miocene fauna from South Dakota. Bull. Amer. Mus., 23, pp. 169—220, 26 text figures. 1924. Third Contribution to the Snake Creek Fauna. Bull. Amer. Mus. Nat. Hist., 50, pp. 59-210, 63 text figures. 1926. On а new primitive deer and two traguloid genera from the lower Miocene of Nebraska. Amer. Mus. Novitates, No. 215, pp. 1-8, 3 text figures. 1930. The phylogeny of the dogs. Jour. Mam., 11, рр. 117-138, 3 text figures. Marruzw, W. D. and Стрьву, J. №. 1904. New or little known mammals from the Miocene of South Dakota. Bull. Amer. Mus. Nat. Hist., 3, pp. 241-268, 15 text figures. Merriam, J. C. 1906. Carnivora from the Tertiary formations of the John Day region. Bull. Dept. Geol. Univ. Calif., 5, pp. 1-64, pls. 1-6, 18 text figures. 1919. Tertiary mammalian fauna of the Mohave Desert. Bull. Dept. Geol. Univ. Calif., No. 5, 11, pp. 437a-e, 438-585, 253 text figures. Ossory, Н. Е. 1918. Equidae of the Oligocene, Miocene and РПосепе of North America. Iconographic type revision. Mem. Amer. Mus. Nat. Hist., n.s., 2, pp. 1-217, pls. 1—54, 173 text figures. Ossory, Н. Е. and WORTMAN, J. Г. 1894. Fossil mammals of the lower Miocene White River beds. Collec- tion of 1892. Bull. Amer. Mus. Nat. Hist., 6, pp. 199-228, pls. 2-3, 8 text figures. 1895. Perissodactyls of the lower Miocene White River beds. Bull. Amer. Mus. Nat. Hist., 7; pp. 343-375, pls. 8-11, 12 text figures. PETERSON, О. А. 1906. The Miocene beds of western Nebraska and eastern Wyoming and their vertebrate faunae. Ann. Carn. Mus., 4, pp. 21-72, pls. 9-19, 20 text figures. 1909. A revision of the Entelodontidae. Mem. Carn. Mus., 4, pp. 41- 156, pls. 54—62, 80 text figures. 1914. The osteology of Promerycochoerus. Ann. Carn. Mus., 9, рр. 149- 219, pls. 33-42, 41 text figures. 1919. Report upon the material discovered in the upper Eocene of the Uinta Basin by Earl Douglass in the years 1908-1909, and by O. A. Peterson in 1912. Ann. Carn. Mus., 12, pp. 141-168, pls. 34—37, 19 text figures. Rome, А, 5. 1926. A Lower Miocene horse, Anchitherium agatense (Osborn). Amer. Jour. Sci., fifth series, 12, No. 70, pp. 325-335, 4 text figures. | f | | 188 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY SCHLAIKJER, E. M. 1932. The osteology of Mesohippus barbouri. Bull. Mus. Comp. Zoöl., 72, pp. 391-410, 5 plates. 1933. А detailed study of the structure and relationships of а new zalambdodont insectivore from the middle Oligocene. Bull. Mus. Comp. Zoól., 76, pp. 1-28, 1 plate, 7 text figures. 1934. Three new oreodonts. Proc. Boston Soc. Nat. Hist., 40, pp. 219- 231. 1935. Torrington member of the Lance formation and a study of a new Trieeratops. Bull. Mus. Comp. Zoöl., 76, pp. 29-68, 7 plates, 5 text figures. 1935. A new basal Oligocene formation. Bull. Mus. Comp. Zoól., T6, pp. 69—94, 8 plates, 10 text figures. Scorr, W. B. 1893. The Mammalia of the Deep River Beds. Trans. Amer. Phil. Soc., 18, m.s., pp. 55-185, pls. 1-6. Scorr, W. B. AND OsBORN, Н. Е. 1887. Preliminary report on the vertebrate fossils of the Uinta forma- tion, collected by the Princeton expedition of 1886. Proc. Amer. Phil. Soc., 24, pp. 255-264. 1890. The mammalia of the Uinta formation. Trans. Amer. Phil. Soc.; 16, pp. 461—572, pls. 7-11, 13 text figures. SELLARDS, E. Н. 1916. Fossil vertebrates from Florida; a new Miocene fauna; new Plio- cene species; the Pleistocene fauna. Bull. Fla. Geol. Surv., 8, pp. 77- 119, pls. 10-14. Sımpson, G. G. 1932. Miocene land mammals from Florida. Bull. Fla. Geol. Surv., 10, pp. 11-41, 23 text figures. SINCLAIR, У. J. 1921. Entelodonts from the Big Badlands of South Dakota in the Geo- logical Museum of Princeton University. Proc. Amer. Phil. Soc., 60, pp. 467-495, 22 text figures. Өмітн, W. S. Т. { 1903. Hartville, Wyoming, quadrangle. U. 8. Geol. Surv., Geol. Atlas, Folio No. 91, pp. 1-6, 3 maps, 14 sections, 6 text figures. Stock, C. 1930. Oreodonts from the Sespe deposits of South Mountain, Ventura County, California. Publ. Carn. Inst. Wash., 404, pp. 27-41, pls. 1-2, 2 text figures. 1933. Carnivora from the Sespe of the Las Posas Hills, California. Publ. Carn. Inst. Wash., no. 440, pp. 30-41, pls. 1-3. SCHLAIKJER: NEW FOSSIL VERTEBRATES 189 Тнонрв, M. В. 1921. John Day Promerycochoeri, with descriptions of five new species and one new subgenus. Amer. Jour. Sci., 1, pp. 215-244, 6 text figures. 1921. John Day Eporeodons with descriptions of new genera and species. Amer. Jour. Sci., 3, pp. 93-111, 16 text figures. 1921. Two new forms of Agriochoerus. Amer. Jour. Sci., 3, pp. 111-119, 4 text figures. 1921. Leptauchenia Leidy and Cyclopidius (Pithecistes) Cope, descrip- tions of new and little known forms in the Marsh collection. Amer. Jour. Sci., 1, pp. 405-419, 6 text figures. 1922. Some Tertiary Carnivora in the Marsh collection, with descrip- tions of new forms. Amer. Jour. Sei., 3, рр. 423-455, 13 text figures. 1924. Тһе generic distinctions between Merycochoerus Leidy and Pro- merycochoerus Douglass. Amer. Jour. Sci., 7, pp. 121-131. 1924. White River (Oligocene) Eporeodons. Amer. Jour. Sci, 7, pp. 219-226. 1931. The osteology of Eporeodon socialis Marsh. Bull. Peabody Mus. Nat. Hist., 2, pp. 1—43, 23 text figures. Твохвы, E. L. 1920. A tiny Oligocene artiodactyl, Hypisodus alacer, sp. nov. Amer. Jour. Sci., 49, pp. 391-398, 4 text figures. 1920. Entelodonts in the Marsh collection. Part I. Morphology and taxonomy. Amer. Jour. Sci., 50, pp. 243-255, pl. 3. 1920. Entelodonts in the Marsh collection. Part II. The genus Archaeo- terium. Amer. Jour. Sci., 50, pp. 361-386, 12 text figures. 1920. Entelodonts in the Marsh collection. Part III. The larger genera and species. Amer. Jour. Sci., 50, pp. 431-445, 20 text figures. EXPLANATION OF PLATES PLATE 1 Sarg neigen ге —— — HM Te TE бснілікхея. New FossiL VERTEBRATES. PLATE 1 A TOPOGRAPHIC MAP OF THE WYOMING Ф GOSHEN HOLE AREA 1955 Scale of miles interval 20 feet ntour RAN гу» ЛЯ | N gS AS Д5; PR G АС 92 Torrington He - 20 д. LUFF)C KSCOTTS B 74 y YY VI R вом m о т т. + Ф S) ( V \ N N N “5 S ( У N INN ч ) WE ~ N N E > 7 У ~ S => UR МА A NN АА \ >) ~ N 4 АУА У = А NN VAVAI VA WS = W AWN SS | ) WRI > WEN У NN hee С ҚЫЗ - ` 4 Sy N WW A D N \ 4 SN N \ #7) SS S N) -- б < чу y^ : EOS ү IN M \\\ ||| ~ IN Y NS Y mw > » N N \ = = DR MIR y BULL. MUS. СОМР. ZOÖL. R sèw R 62 R 63м 64 W Drawn by Schlaikjer and Young PLATE 2 BULL. MUS. COMP. ZOÖL. бсніліклен, New Fossit VenrEBRATES, PLATE 2 A GEOLOGIC MAP OF THE GOSHEN HOLE AREA, WYOMING ‘ GEOLOGY BY ERICH M. SCHLAIKJER 1955 Scale of miles | 2 3 4 M. Explanation edi Alluvium Miocene — Harrison Brule Chadron Yoder У пе Lance Pierre Oligocene Cretáceous EIN 55 NS \ SS SS RS D. “а. эв. етн. чө, э. III 7 YAP SS АХ ААА 2 2 2 zta4 25-1 NI Ж. Г АРА 2 Л UN.PLATTE СО: Л 77 ИСО. 2; о OSHEN EI Ж ПВ Ж 7 Ж OS 5 Ж DIA NY \ TORRINGTON |HIGHWwAY SS ПР ГА 22 Жж CHEYEI N m SN" > > Жи, WE Ии СЛ Z 7 2 В 58У ` Resw | PLATE 3 SCHLAIKJER— New Fossil Vertebrates PLATE 3 Fra. 1. Red-Bill Point showing upper Chadron channel sandstone. Ета. 2. Voleanie ash in the Brule formation at the east side of Table Moun- tain. BULL. MUS. СОМР. ZOÖL. ScHLAIKJER. New Fossit VERTEBRATES. PLATE 3 PLATE 4 SCHLAIKJER—New Fossil Vertebrates PLATE 4 Ета. 1. Looking eastward. Harvard Fossil Reserve. Bone layer immediately below laboratory building. Ета. 2. Collecting a large block of fossils in the Harvard Fossil Reserve. BULL. MUS. COMP. ZOÓL. SCHLAIKJER. New Fossit VERTEBRATES. PLATE 4 PLATE 5 SCHLAIKJER—New Fossil Vertebrates PLATE 5 Ета. 1. Dip in the Brule formation at Fox Creek Pass. Ета. 2. Brule-lower Harrison channel deposit resting on Brule clay. Northeast of Wilson's Ranch. BULL. MUS. СОМР. ZOÓL. ScHLAIKJER. New Fossit VERTEBRATES. PLATE 5 PLATE 6 SCHLAIKJER— New Fossil Vertebrates PLATE 6 Ета.1. Brule-lower Harrison channel deposit resting on Brule clay. Sec. 36. T. 25N. R. 63W. Ета. 2. Eagle Rock at Fox Creek Pass showing the lower Harrison formation, BULL. MUS. COMP. ZOÓL. SSHLAIKJER. New Fossit VERTEBRATES. PLATE 6 ~ а e < = ~ SCHLAIKJER—New Fossil Vertebrates PLATE 7 Looking southeast at Sixty Six Mountain showing the Brule-lower Harri- son contact and the gentle dip southeastward. BULL. MUS. COMP. ZOÓL. SCHLAIKJER. New Fossit VERTEBRATES. PLATE 7 PLATE 8 * SCHLAIKJER—New Fossil Vertebrates PLATE 8 Ета. 1. East view down the North Platte River valley overlooking the upper lower Harrison beds which have been faulted down below the upper Brule (immediate foreground). Ета. 2. Whalen Fault showing the Brule clay above the lower Harrison. BULL. MUS. COMP. ZOÓL. SCHLAIKJER. New Fossit VERTEBRATES. PLATE 8 PLATE 9 ScHLAIKJER—New Fossil Vertebrates PLATE 9 Ета. 1. The Whalen Fault, same locality as in Plate 8. Ета. 2. Close view of the fault-plane. Photographs, Plates 3-9, by Erich M. Schlaikjer BULL, MUS. COMP. ZOÓL. ScHLAIKJER. New FossiL VERTEBRATES. PLATE 9 ScBLAIKJER—New Fossil Vertebrates PLATE 10 Testudo peragrans Hay Referred specimen. Skull and carapace. One-fourth natural size. BULL. MUS. COMP. ZOÖL. SCHLAIKJER. New FossıL VERTEBRATES. РЕАТЕ 10 4 PLATE 11 SCHLAIKJER—New Fossil Vertebrates PLATE 11 Мевосуом HORTULIROSAE Schlaikjer Referred specimen. Right hind limb and right humerus. Approximately three-fourths natural size. Photographs, Plates 10—11, by W. L. Hallowell. ge mo u ро” | BULL, MUS. COMP. ZOÓL. SCHLAIKJER. New Fossit VERTEBRATES. РідтЕ 11 ТІ 2 | 3 | | 4 | PLATE 12 SCHLAIKJER—New Fossil Vertebrates PLATE 12 Ела. 1. Parahippus pristinus Osborn. Type. Palatal view of left PI-M3; Ета. 2. Pediohippus brachystylus (Osborn). Туре. Palatal view of left P!-M3. Ета. 3. Pediohippus antiquus Schlaikjer. Genoholotype. Palatal view of left P2-M3. Ета. 4. Pediohippus trigonostylus (Osborn). Referred specimen. Palatal view of left P2-M?. All Figures natural size. Drawings by Gladys De Grout Schlaikjer. BULL. MUS. COMP. ZOÖL. SCHLAIKJER. New Fossit VERTEBRATES, PLATE 12 PLATE 13 SCHLAIKIER—New Fossil Vertebrates PLATE 13 MESOHIPPUS BARBOURI Schlaikjer Referred specimen. Palatal view of the skull. Natural size. Photograph by W. L. Hallowell. BULL, MUS. COMP. ZOÓL. SCHLAIKJER. New FossıL VERTEBRATES. PLATE 13 PLATE 14 SCHLAIKIER— New Fossil Vertebrates PLATE 14 PARAHIPPUS WYOMINGENSIS Schlaikjer Type. Superior and lateral view of the skull. One-third natural size. + = ul E < zi a Ф ul E < x m ul E T Ш > ЕТ Ф З о u. z Ш Е > m 5 x < El == o an BULL. MUS. COMP. ZOOL. PLATE 15 ScmLAIKJER—New Fossil Vertebrates PLATE 15 PARAHIPPUS WYOMINGENSIS Schlaikjer Type. Palatal views of skull and jaws. One-third natural size. ScHLAIKJER. New Fossit VERTEBRATES. РгАТЕ 15 BULL. MUS. COMP. ZOÖL. PLATE 16 SCHLAIKJER—New Fossil Vertebrates PLATE 16 PARAHIPPUS WYOMINGENSIS Schlaikjer Type. External and anterior views of right hind foot. One-third natural size. Photographs, Plates 14-16, by George Nelson. SCHLAIKJER. New FossiL VERTEBRATES. PLATE 16 BULL. MUS. COMP. ZOÓL. SCHLAIKJER—New Fossil Vertebrates PLATE 17 Hypouirrus avus Schlaikjer Type. Lateral view of skull. Two-thirds natural size. BULL. MUS. COMP. 7001. SCHLAIKJER. New Fossit VERTEBRATES. PLATE 17 PLATE 18 ScHLAIKJER—New Fossil Vertebrates PLATE 18 HYPOHIPPUS Avus Schlaikjer Type. Palatal view of skull. Two-thirds natural size. | і | | | BULL. MUS. COMP. ZOÓL. ScHLAIKJER. New Fossit VERTEBRATES. PLATE 18 PLATE 19 ScHLAIKJER—New Fossil Vertebrates | PLATE 19 НУРОНТРРОВ Avus Schlaikjer Туре. Dorsal view of skull. Two-thirds natural size. | | 1 ScHLAIKJER. New FossiL VERTEBRATES. PLATE 19 BULL. MUS. COMP. ZOOL. PLATE 20 SCHLAIKJER—New Fossil Vertebrates PLATE 20 ARCHAEOTHERIUM PALUSTRIS Schlaikjer Type. Superior view of jaws. One-fourth natural size. BULL. MUS. СОМР. ZOÓL. ScHLAIKJER. New Fossit VERTEBRATES. PLATE 20 PLATE 21 SCHLAIKJER—New Fossil Vertebrates PLATE 21 DINOHYUS minimus Schlaikjer Type. Superior view of symphysial region of the jaws. Three-fourths natural size. BULL. MUS. COMP. ZOÖL. SCHLAIKJER. New Fossit VERTEBRATES. PLATE 21 PLATE 22 SCHLAIKJER— New Fossil Vertebrates PLATE 22 Ета. 1. Cyclopidius heterodon Cope. Neotype. Lateral view of skull. Ета. 2. Cyclopidius heterodon Cope. Referred specimen. Palatal view of pi-M?, Ета. 3. Cyclopidius heterodon Cope. Referred specimen. Right maxilla. Ета. 4. Cyclopidius heterodon Cope. Referred specimen. Right jaw. Ета. 5. Cyclopidius heterodon Cope. Referred specimen. Metacarpal and first phalanx. All figures natural size. N 10 SCHLAIKJER. New Fossit VERTEBRATES. PLATE 22 BULL. MUS. COMP. ZOÓL. ————— À—————— rM УРН >" FE TM PLATE 23 SCHLAIKIJER— New Fossil Vertebrates PLATE 23 CYCLOPIDIUS HETERODON Cope Referred specimen. Dorsal, lateral and palatal views of the skull of a young individual. Natural size. uum En ы стас И а ағы ie = = 3, e N w E < et n a ul E < с a ш E с uj > d Ф Ф о ш. z ш = с ш 5 52 < d En S o зі о о N а: = © © % => = =! ET ES m Å x e E я < — Ay SCHLAIKJER— New Fossil Vertebrates PLATE 24 CYCLOPIDIUS DENSA (Loomis) Referred specimens. Two skulls. Three-fourths natural size. ScHLAIKJER. New FossiL VERTEBRATES. PLATE 24 BULL. MUS. COMP. ZOÖL. PLATE 25 SCHLAIKJER—New Fossil Vertebrates PLATE 25 CYCLOPIDIUS DENSA (Loomis) Referred specimens. Palatal views of dentitions of two individuals. Natural size. Photographs, Plates 17-25, by W. L. Hallowell. BULL. MUS. COMP. ZOÓL. SCHLAIKJER. New Fossit VERTEBRATES. PLATE 25 PLATE 26 SCHLAIKJER—New Fossil Vertebrates e PLATE 26 Ета. 1. Cyclopidius emydinus Cope. Type. Palatal view of the skull. Ета. 2. Cyclopidius heterodon Cope. Type. Palatal view of P4-Mt, Ета. 3. Cyclopidius simus Cope. Type: Palatal view of the skull. AII Figures three-fourths natural size. Photographs from the American Museum collection. BULL. MUS. COMP. ZOOL. SCHLAIKJER. New FossiL VERTEBRATES. PLATE 26 PLATE 27 ScHLAIKJER—New Fossil Vertebrates PLATE 28 EPOREODON CHEEKI Schlaikjer Туре. Palatal view of the skull. Five-ninths natural size. BULL. MUS. COMP. ZOÓL. SCHLAIKJER. New Fossit VERTEBRATES. PLATE 28 PLA'TE 29 ScnrAIKJER—New Fossil Vertebrates PLATE 29 EPOREODON CHEEKI Schlaikjer Type. Left view of the skeleton. Approximately one-ninth natural size. Photographs, Plates 27-29, by George Nelson; BULL. MUS. COMP. ZOÓL. SCHLAIKJER. New Fossit VERTEBRATES. Рідте 29 PLATE 30 SCcHLAIKIER— New Fossil Vertebrates PLATE 30 MESOREODON CHELONYX Scott Referred specimen. Lateral and palatal views of the skull. Two-thirds natural size. ті en E Es < a SCHLAIKJER—New Fossil Vertebrates PLATE 31 MESOREODON MEGALODON Peterson Referred specimen: Lateral view of the skull. Two-thirds natural size. Photographs, Plates 30-31, by W. L. Hallowell. BULL. MUS. COMP. ZOÓL. SCHLAIKJER. New Fossil VERTEBRATES. PLATE 31 PLATE 32 SCHLAIKJER— New Fossil Vertebrates PLATE 32 MESOREODON 8сотті Schlaikjer Type. Lateral view of the skeleton. One-ninth natural size. Photograph by George Nelson. BULL. MUS. COMP. ZOOL. ScHLAIKJER. New FossiL VERTEBRATES. PLATE 32 PLATE 33 ScHLAIKJER—New Fossil Vertebrates PLATE 33 PROMERYCOCHOERUS LOOMISI Schlaikjer Туре. Dorsal view of the skull, Three-eighths natural size. SCHLAIKJER. New FossıL VERTEBRATES. PLATE 33 BULL. MUS. COMP. ZOOL. PLATE 34 SCHLAIKJER—New Fossil Vertebrates PLATE 34 PROMERYCOCHOERUS LOOMISI Schlaikjer Type. Lateral view of the skull. Three-eighths natural size. BULL. MUS. COMP. ZOOL. SCHLAIKJER. New Fossil VERTEBRATES. PLATE 34 PLATE 35 SCHLAIKJER—New Fossil Vertebrates PLATE 35 PROMERYCOCHOERUS LOOMISI Schlaikjer | Type. Lateral view of the skeleton. One-twelfth natural size. Photographs, Plates 33-35, by F. P. Orchard. N | | BULL, MUS. СОМР. ZOÖL. SCHLAIKJER. New Fossit VERTEBRATES. Puate 35 | PLATE 36 SCHLAIKJER—New Fossil Vertebrates PLATE 36 PROMERYCOCHOERUS HATCHERI Douglass Referred specimen. Dorsal view of the skull of male. One-half natural size. BULL. MUS. COMP. ZOOL. SCHLAIKJER. New Fossil VERTEBRATES. PLATE 36 H j | | H | | $ А | E | || 4 | 4 PLATE 37 ge SCHLAIKJER—New Fossil Vertebrates PLATE 37 PROMERYCOCHOERUS HATCHERI Douglass Referred specimen. Palatal view of the skull of male. One-half natural size; BULL. MUS. COMP. ZOOL. SCHLAIKJER. New Fossil VERTEBRATES. Puate 37 PLATE 38 SCHLAIKJER— New Fossil Vertebrates PLATE 38 PROMERYCOCHOERUS HATCHERI Douglass Referred specimen. Lateral view of the skull of female. BULL. MUS. COMP. ZOÓL. Зснілікуєв. New Fossit VERTEBRATES. PLATE 38 PLATE 39 SCHLAIKJER—New Fossil Vertebrates PLATE 39 MIOTYLOPUS BATHYGNATHUS Schlaikjer Genoholotype. Dorsal, lateral and palatal views of the skull and jaws. One-half natural size. Photographs, Plates 36-39, by W. L. Hallowell. BULL. MUS. COMP. ZOOL. SCHLAIKJER. New FossıL VERTEBRATES. PLATE 39 PLATE 40 ScHLAIKJER—New Fossil Vertebrates PLATE 40 NANOTRAGULUS INTERMEDIUS Schlaikjer Type. Dorsal, lateral and palatal views of the skull and jaws. Natural size. Photographs by Albert Thompson. BULL. MUS. COMP. 7001. ScHLAIKJER. New Fossit VERTEBRATES. PLATE 40 ті < E E < а) D SCHLAIKIJER— New Fossil Vertebrates PLATE 41 Ете. 1. Subhyracodon occidentale (Leidy). Referred specimen. Palatal view of PI-M?. One-half natural size. Ета. 2. Parahippus ?pristinus Osborn. Referred specimen. Palatal view of DP24, Natural size. Ета. 3. Merychyus harrisonensis Peterson. Referred specimen. Palatal view of P2-M3. Natural size. Ета. 4. Nanotragulus intermedius Schlaikjer. Paratype. Palatal view of the Skull. Natural size. Ета. 5. Leptauchenia minora Schlaikjer. Type. Palatal view. Natural size. Ета. 6. Leptotragulus ultimus Schlaikjer. Type. Palatal view of МЗ. Nat- ural size. BULL. MUS. COMP. ZOOL. SCHLAIKJER. New Fossil VERTEBRATES. PLATE 41 | d | | | | | | | | а С ыла пер ава 3137 Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Мог. LXXVI Мод SHALER MEMORIAL SERIES SOME ORDOVICIAN CYSTIDS FROM RUSSIA Bv Евер В. PHLEGER, JR. v WrrH ONE PLATE CAMBRIDGE, MASS., U.S. A. PRINTED FOR THE MUSEUM Jury, 1935 A т PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE There have been published of the BuLLeriw Vols. I to LXV, LXVII-LXXV, LXXVII, of the Мемотвз Vols. 1 to LII, LIV. The BuLLETIN and МЕмотвв are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collections and Exploration. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately. A price list of the publications of the Museum will be sent on application to the Director of the Museum of Comparative Zoölogy, Cambridge, Massachusetts. Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vor. ХУТ №. 9 SHALER MEMORIAL SERIES SOME ORDOVICIAN CYSTIDS FROM. RUSSIA Ву Frep B. PHLEGER, JR. Wira ONE PLATE CAMBRIDGE, MASS., 1). 8. A. PRINTED FOR THE MUSEUM JuLY, 1985 No. 5.— Some Ordovician Cystids from Russia Bx FreD B. Рнікакв, JR. IN TRODUCTI ON In 1914 Professor Percy E. Raymond, on a Shaler Memorial Expedition, made large collections from several localities of the Ordovician strata of Russia. The following paper describes the new cystids represented in the collection and redescribes some of the established forms. The writer is grateful to Professor Raymond for the use of the specimens and for his generous aid in the preparation of this paper. Most of the material was obtained from quarries and all the good specimens are from weathered debris not in place. It is believed that most of the cystids are from Lamansky’s! zone Вита but some are probably from his zone Вила. Lamansky’s Brita is the equivalent of the upper part of Raymond’s? Walchow formation and Вита is the lower part of the Kunda formation. It is doubtful if any of the material came from any lower in the section than the zone Вила. Phylum Echinodermata Stachelhauter Class CYSTOIDEA Beutelstrahler Order AMPHOROIDEA Haeckel Family DENDROCYSTIDAE Bather Genus Rhipidocystis Jaekel RHYPIDOCYSTIS GIGAS Jaekel Plate 1, fig. 9 Rhipidocystis gigas Jaekel, Zeitschrift der Deutschen geol. Gesell., 1900, p. 672. Bather, F. A., Caradocian Cistidea from Girvan, Trans. Royal Soc. Edinburgh, 1913, pp. 369-370. Bather lists the published specimens which have been referred to this species, and also the holotype of Rhipidocystis hallica which he refers to R. gigas. None of the specimens so far figured shows the anus. One specimen in the present collection exhibits the anal side 1 Lamansky, von W., Die Aeltesten Silurischen Schichten Russlands, 1905, p. 171. ‘Raymond, P. E., The Correlation of the Ordovician Strata of the Baltic basin with those of Eastern North America, Bull. Mus. Comp. Zool., 1916, 41, no. 3, pp. 188-195. 194 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY of the theca. Jaekel states that the forms included in his family Rhipidocystidae have a simple anal cavity between three plates. Moreover, in the definition of the only genus of that family, Rhipido- cystis, he describes the anus as small and round. But in the definition of В. gigas, the only species of the family, no mention is made of the anus. Evidently none of the specimens studied by Jaekel exhibited the anal area, and his description of the anus in the family and genus was by inference only. The anal pyramid is large, circular, convex, and enclosed by five thecal plates. It has a diameter of about 10 mm. It is composed of 18 to 25 wedge-shaped plates, about one-third of which reach the center of the area; small plates are wedged in between the larger ones. No idea of the shape of the theca is given by the fragmentary speci- men at hand. The plates are thin and vary somewhat in size, but most are about 20 10.25 mm. in width. They are smooth with the exception of very faint axial ridges on a few plates. The author is following Bather's suggestion that this genus is prob- ably to be included in the family Dendrocystidae. Formation and locality: Walchow and Kunda formations, Iswos, Uli, The plesiotype is Mus. Comp. Zoól. no. 4002. Order RHOMBIFERA Bather Superfamily GLYPTOCYSTIDAE Bather Family ECHINOENCRINIDAE Bather Genus ECHINOENCRINITES von Meyer Echinoencrinites von Meyer, Karstens Archiv für Naturlehre, 1826, Bd. VII, p. 185. Jaekel, Otto, Stammesgeschichte der Pelmatozoen, Berlin, 1899, p. 242. Echinoencrinus Bather, F. A., A Treatise on Zoólogy, pt. III, The Echinoder- mata, London, 1900, p. 60, figs. 22, 23. Gonocrinus Eichwald, E., Zeitschr. für Natur u. Heilkunde der medicinischen Acadamie zu St. Petersburg, 1840. Sococystis von Buch, L., Abhand. d. kgl. preuss. Acad. d. Wiss., Berlin, 1845. The following is Bather’s description of the genus Mchinoencrinites: “Differs from the imagined archetype in the restriction of pectini- rhombs to 1 and 5,1 and 6, and 14 and 15,and in the apparent bisection of plate 23. The main grooves may support five brachioles, or only PHLEGER: ORDOVICIAN CYSTIDS 195 two, or may branch yet more; in any case the facets are always close around the mouth. Anal region often prominent. The plates usually have strong radiating folds, often crossed by finer concentric ridges." ECHINOENCRINITES SENCKENBERGII von Meyer Echinoencrinites senckenbergii von Meyer, Karstens Archiv für Naturlehre, 1826, Bd. VII. p. 185. Jaekel, Otto, Stammesgeschichte der Pelmatozoen, Berlin, 1899, p. 248, pl. 13, figs. 1, 2. The writer considers published descriptions of this species inade- quate and submits the following description based on the specimens in the Museum of Comparative Zoólogy : Theca pear-shaped, expanded below to become, in some cases, more or less flat on the aboral side. 'The anus is between plates 7, 8, and 13 (after Bather's analysis) and is rounded in outline; in most cases the opening is projected downward at a definite angle from a triangular eminence on the three enclosing plates (the vertex of the triangle 18 directed upward). The oral region is oval in outline and supports about ten arms; the lips are extended outward and upward or slightly overlap the theca. The radiating folds are rounded to subangular and are continuous over the suture lines. The number of folds on a single plate varies from about 15 to 38, being less in the small specimens than in the large ones. Between the radial folds is а series of cross-ridges, varying in concentration, but in some cases there are as many as six folds to 2 mm. length. In a few cases these concentric folds are continuous over the radial folds; this phenomenon is present in a few specimens only. Most of the plate beaks are moderately projected; in some individuals they are more or less flattened. One partially preserved stem shows about 25 columnals. Each proximal segment has а well-defined engirdling ridge about Imm. in width. On the distal segments, comprising about one-third of the specimen at hand, engirdling ridges are present but they are flattened and widened and show well-developed transverse striae. Measurements Small Average Large Height 9 mm. 14 mm. 22 mm. Width 7 mm. 11 mm. 18 mm. Some fragmentary specimens indicate that a few individuals were larger than any of the above. The number of radial folds varies directly 196 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY with size and there are all gradations. Variation in abundance of concentric folds appears to be quite independent of size. The diagrams (plate 1, figs. 4-8) show the development of the radial folds in a series of five specimens, ranging from the smallest to the largest in the collection. It must be kept in mind that a slight addition to the size of each individual plate can cause a great increase in the size of the theca; thus, the range in size of the plates is not great. Although there is an expected variation, in general it is observed that the folds increase in number in the larger specimens; moreover, the addition of new folds takes place at the suture angles. The resulting apparent variation in ornamentation, caused by the increase in total number of folds, is a direct function of size, and although apparently marked between large and small specimens, in the opinion of the writer does not indicate specific differences. Formation and locality: Lamansky's zone Bia or zone Bing, opposite Iswos; at the mouth of the Sjas river; the mouth of the Lyna river; Iswos; Wassilkowa; and a quarry 3 miles east of Putilowa, 05.5. Hh. The plesiotypes are Mus. Comp. Zoöl. no. 3997. ECHINOENCRINITES INTERLAEVIGATUS Jaekel Echinoencrinites senckenbergü var. interlaevigata Jaekel, Stammesgeschichte der Pelmatozoen, Berlin, 1899, p. 249. Jaekel separated the variety interlaevigata from Eehinoenerinites senckenbergii von Meyer on the basis of the ornamentation being con- fined more or less to the plate centers, resulting in smooth plate borders. Several of the specimens of the Museum of Comparative Zoólogy show this characteristic; but gradations оссиг between the specimens show- ing complete ornamentation and those in which only the plate centers are ornamented. Although a complete series showing the develop- ment of this characteristic is not at hand, the indications are that such could be found. It is important to note that the smooth plate borders are well- developed in the largest and are not indicated in the smallest speci- mens. One medium-sized specimen shows smooth plate borders on the plates of one side of the theca and fully ornamented ones on the other side. In other respects the specimens are like E. senckenbergit. The development of such a character from the typical Е. sencken- bergii could, in the opinion of the writer, be due to three things: (1) PHLEGER: ORDOVICIAN CYSTIDS 197 differential solution—by ground water, perhaps; (2) resorption of plate borders; or (3) old age. There is no indication that solution has taken place, since the plate umbos are plainly not affected; it is difficult to imagine such selective erosion as to produce the appearance shown by these specimens. Furthermore, the characteristic unevenness to be expected on the plate borders does not exist. The smooth rims of the plates must, then, be due to organic causes. Shimer! has observed that flattening out of plications is an old age characteristic of brachiopods. Their former presence on the smooth senescent individual is indicated by growth lines. In E. interlaevigatus fine growth striations are abundant on most of the plate borders. Loss of skeletal features has been ascribed to a progressive loss of the power of growth which is characterized by Minot’, in his work on guinea pigs, as “beginning almost immediately after birth." He raises the question “whether in all living things there is not a certain impulse given at the time of impregnation, and whether this impulse does not fade out so that from the very beginning of the new growth there oc- curs а dimunition in the rate of growth.” Loss of plications has also been observed by R. T. Jackson? in Litor- ina littorea, a gastropod which is at present abundant on the Atlantic coast: “While thus the increase in size becomes less and less for each succeeding growth period, а time is reached, varying with each in- dividual, when another factor enters and actual shrinkage begins. The tendency of the soft parts of animals in old age to contract is familiar to us.” This can explain alteration in hard parts, particularly in а form like a cystid. Shimer has shown that old age smoothness of the shell is caused directly by the contraction of the brachiopod mantle. Jackson‘ has further shown that resorption is an important process in sea urchins. He points out that with an internal skeleton "there is a constant resorption within the plate concurrently with increase in external dimensions." Jackson also points out the role of variation in producing apparent differences: “А regressive variant is one that has attained full characters and then in later life has reverted to the youthful or primitive characters. . . . it is felt that we should be extremely cautious in assuming hybridity when the same results might be attained by variation." 1 Shimer, Н. W., Old Age in Brachiopods, Amer. Nat., 1906, 40, pp. 95-122. ? Minot, C. S., Senescence and Rejuvenation, Jour. of Physiol., 1891, 12, p. 151. ¿Shimer, op. cit., р. 119. 4Jackson, В. T., Memoirs of the Boston Soc. of Nat. Hist., 1912, T pe las 198 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY It is the opinion of the writer that the forms of Echinoenerinites here considered are regressive variants of E. senckenbergW, and that the variation is an old age character brought about possibly by resorption of the plate borders ; the fact that the smooth borders are in nearly every case confined to the large individuals is presented in support of this theory. Many of the plates of the specimens of this group have been distinctly displaced at the sutures, probably indi- cating a weakening of the body tissues in that area. All the forms having well-developed smooth plate borders are placed in E. interlaevigatus for convenience. ; Formation and locality: Lamansky's zone Виша or Виа, Iswos; opposite Iswos, and at the mouth of the Lyna river, U.S.S.R. The plesiotypes are Mus. Comp. Zoól, no. 3996. ECHINOENCRINITES LAHUSENI Jaekel Echinoencrinites lahuseni Jaekel, Stammesgeschichte der Pelmatozoen, Berlin, 1899, pp. 247-248, pl. 13, fig. 10. In the specimens of this species there are observed similar gradations in number of radiating folds as in the writer’s examples of Echinoenerin- ites senckenbergit, although the range in size is not as great. The larg- est is 18mm. high and 12 mm. wide, whereas an average specimen is 14 by 12 mm. This form differs from E. senckenbergii mainly in the location and appearance of the anus, which is between plates 7, 8, 12, and 13. The prominence surrounding the anal opening is roughly square, is less distended than in E. senckenbergit, and is not deflected downward but faces directly to the side. In some specimens there is a depressed area, with ornamentation less prominent at the plate borders than in the region of the umbo, but not actually smooth. The lips of the orifice do not lap over the theca quite to the extent shown in Jaekel's illustration. Formation and locality. Lamansky’s zone Виа or Bing, opposite Iswos; at Iswos, and Wassilkowa, U.S.S.R. The plesiotypes are Mus. Comp. Zoöl. no. 3999. ECHINOENCRINITES SPHEROIDALIS Sp. nov. Plate 1, fig. 2 Theca spherical, flattened below and above, expanded in the median circumference to become as wide as high. The anus is between plates PHLEGER: ORDOVICIAN CYSTIDS 199 7, 8, and 13 and is rounded in outline. It projects downward at a slight angle from a roughly triangular eminence, and is only slightly distended from the theca. А comparatively narrow eminence extends from the anal opening to the beak of each of the enclosing plates. 'The radiating folds are subrounded and extend over the suture lines. There are about 40 to 50 folds on each plate. Cross ridges are abundant between the folds, averaging three to 1mm. of length. The plate beaks are very slightly distended. Measurements: height, 18 mm; width, 18 mm. This species is most closely related to Echinoencrinites senckenbergit von Meyer, but it differs in being as wide as high, in having more radial folds per unit size, in having the anus less projected from the theca, in having narrow, smooth eminences extending from the anus to the beaks of the enclosing plates, and in the comparative flatness of the plate beaks. Formation and locality. Lamansky’s zone Виша or Bing, Iswos, 9.5.5.8. 'The holotype is Mus. Comp. Zoól. по. 3998. ECHINOENCRINITES SIMPLICATUS Sp. nov. Plate 1, fig. 3 This species is like Echinoencrinites lahuseni in all respects except in plate ornamentation, which is more simplified. Аз in that species, there is one radial fold extending from the plate apex to the approxi- mate center of each suture segment. These folds are distinct in their broadly rounded character and are much wider than in E. lahusena, approaching Imm. in width in the larger specimens. Very few auxiliary folds are developed, characteristically only one extending to each su- ture angle; at one of the suture angles in some of the plates (commonly plates 10, 11, 12, 13, and 14) there is a flat raised area. Transverse fila are present but are not as obvious as in Æ. lahusent. On one specimen there is a quadrangular area about 14 mm. in width on all plate borders, slightly insunken, which gives the effect of a geometrie border. Most of the radiating folds are noticeably depressed in crossing this border. The size of the specimens at hand varies as follows: height 9 to 17 mm., width 7 to 12 mm. Formation and locality. Lamansky's zone Виша or Виа, Iswos, U.S.S.R. The holotype is Mus. Comp. Zoöl. no. 3995. BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY EuTRETOCYSTIS gen. nov. There are three cystids in the collection which are like Echinoen- crinites in all respects except in the development of an extra pectini- rhomb on plates 10 and 15. It appears that the importance of the pec- tinirhombs has not been fully recognized by some workers. Bather, in his diagnosis of Echinoencrinites states definitely that pectinirhombs are restricted to plates 1 and 5, 1 and 6, and 14 and 15. Jaekel places in E. lahuseni forms having pectinirhombs on plates 10 and 15. Since the rhombs are, according to Bather, "specialized outgrowths of the ambulacral system" and associated with respira- tion, it is the opinion of the writer that they are of generic importance. Some of Jaekel’s specimens referred to E. lahusent should undoubted- ly be referred to Eutretocystis. Only those specimens with three rhombs have been referred to Echinoencrinites lahusen? by the writer. Genotype, Eutretocystis similis sp. nov. EUTRETOCYSTIS SIMILIS sp. nov. Plate 1, fig: 1 The theca is pear-shaped, expanded below. The anus is between plates 7, 8, and 13 and is rounded in outline; the opening is projected slightly downward. It is in a comparatively smooth triangular eminence formed by the junction of the three enclosing plates; the vertex of the triangle is directed upward. The oral region is oval in outline, apparently supporting ten or more arms. It does not overlap the theca. Radiating folds on the plates are rounded to subrounded. There are 22 folds on the plate of a small specimen 11mm. high and 9 mm. wide, and 32 folds on the corresponding plate of a specimen 21mm. high and 14 mm. wide. Concentric cross-ridges are abundant between the folds, 2 to 3 occurring in each mm. Plate beaks are moderately projected. Formation and locality. Lamansky’s Виша or Bing, Iswos on the Walchow river, U.S.S.R. The holotype is Mus. Comp. Zoól. no. 4001. PHLEGER: ORDOVICIAN CYSTIDS 201 EUTRETOCYSTIS ACUTIROSTRIS SP. nov. Plate 1, fig. 10 The plate beaks rise sharply to a point which extends well away from the rest of the theca. The species is thus characterized by its angularity and in this way is distinguished from Eutretoeystis similis. Formation and locality. Lamansky’s zone Bi, Dr, Iswos, U.S.S.R. The holotype is Mus. Comp. Zoöl. no. 4000. Order DIPLOPORITA Zittel Family PROTOCRINIDAE Bather Genus PROTOCRINUS Eichwald PROTOCRINUS FRAGUM Eichwald Protocrinus fragum Eichwald, Lethaea Rossica, ou Paleontologie de la Russie, Band. I, Stuttgart, 1860, p. 621. Jaekel, Stammesgeschichte der Pelma- tozoen, Berlin, 1899, p. 433. Four specimens of Protoerinus fragum in the collection are rather poorly preserved, but one specimen shows the ambulacral grooves very well. They extend a little farther aborally than Jaekel's descrip- tion indicates, to the upper edge of the lower radials. Most of the diplo- pores have been nearly or completely obliterated. At least two speci- mens are young animals, since they show a feeble stem development. Formation and locality. Zone Віта or Bing of Lamansky, mouth of the Lyna river; Hamantown village; and quarry west of Putilowa, U.S.S.R. The plesiotypes are Mus. Comp. Zoöl. no. 4003. EXPLANATION OF PLATE PurEGER — Ordovician Cystids PLATE Fig. 1. Eutretocystis similis sp. nov. Lateral view showing pectinirhombs on plates 10 and 15, and 14 and 15. х2. Fig. 2. Echinoencrinites spheroidalis sp. nov. Lateral view showing the anal region. x2. Fig. 3. Echinoencrinites simplicatus sp. nov. x 2. Figs. 4 to 8. Showing the development of plate ornamentation in Echinoen- crinites senckenbergii von Meyer. Figs. 4, 5, and 7 are diagrams of plate 12, 6 is plate 11, and 8 is plate 9 of successively larger specimens. x2. Fig. 9. Rhipidocystis gigas Jaekel. The anal pyramid. x2. Fig. 10. Hutretocystis acutirostris sp. nov. Lateral view showing pectini- rhombs on plates 10 and 15, and 14 and 15. x 2. Drawings by Mrs. Cecil Lalicker. unl X 5 = PHLEGER. ORDOVICIAN CYSTIDS | ———————— M BULL. MUS. COMP. ZOOL. JUL 2 6 1935 STE Bulletin of the Museum of Comparative Zoölogy AT HARVARD COLLEGE- Vor. LXXVI, No. 6 LEANCHOILIA AND OTHER MID-CAMBRIAN ARTHROPODA Ву Percy E. Raymond CAMBRIDGE, MASS., U. 5. A. PRINTED FOR THE MUSEUM Jury, 1935 ML Vib T PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE There have been published of the BuLLerin Vols. I to LXV, LXVII-LXXV, LXXVII, of the Мемотвз Vols. 1 to LII, LIV. The BurLETIN and МЕмотвв are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collections and Exploration. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs is sold separately. A price list of the publications of the Museum will be sent on application tothe Director of the Museum of Comparative Zoólogy, Cambridge, Massachusetts. * Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vor. LXXVI, No. 6 LEANCHOILIA AND OTHER MID-CAMBRIAN ARTHROPODA Ву Percy E. RAYMOND CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE- MUSEUM JULY, 1985 No. 6. — Leanchoilia and other Mid-Cambrian Arthropoda By Percy E. RAYMOND Leanchoilia superlata was described by Dr. Walcott (1912, p. 170, pl. 31, fig. 6) at a time when he had only one incomplete specimen. Although he collected others later, he published no further description of this species, but (1931, p. 8, pl. 12, pl. 13, fig. 2, pl. 14, fig. 4, 5.) figured some excellent specimens. Several more or less complete individuals were collected by Mr. W. E. Schevill in the summer of 1930 while a party from the Museum of Comparative Zoólogy! was working in the Middle Cambrian on Burgess Pass. Rather to our surprise, we did not find any specimens in Dr. Walcott's old quarry, all which we obtained being derived from a layer some seventy feet further up the mountain side. In addition to our own specimens, I have been able to study those in the U. S. National Museum, through the kindness of Dr. C. E. Resser, and one in the Peter Redpath Museum, thanks to Dr. T. H. Clark. Leanchoilia was referred by Walcott to the order Anostraca under the subclass Branchiopoda, and included with Opabinia in the family Opabinidae. Fedotov (1925, p. 385) doubted if Leanchoilia were really a bran- chiopod, stating that its general habitus and appendages suggested relationship to the amphipods. He further questioned if there was any close relationship to Opabinia. Henriksen (1928, p. 7) was able to study two complete individuals of Leanchoilia, and found that there were ten segments in the trunk, behind which was a long telson bearing a fringe of setae. He considered the evidence to indicate that the animal is a merostrome, and that the species is identical with Emeraldella micrura Walcott. Hutchinson (1930, p. 2) did not discuss Leanchoilia, merely noting with approval that Henriksen had assigned it to the Merostomata. Description of the test 'The general form of the skeleton is best shown by specimen M.C.Z. 1842, an individual which is flattened dorso-ventrally instead of laterally, as are most of the others. As preserved on the rock, the test is seen from the inside. The head, body, anterior part of the telson, and some appendages are preserved. 1 The material upon which this paper is based was collected with the aid of the income of the William and Adelaide Barbour Fund. 206 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY The head as flattened is roughly triangular, about as broad as long. At the anterior end it is extended into a short rostrum. There is no visible opening of any kind through the test of the head, and no notch in the margin for a pedunculate eye. The body consists of ten segments of approximately equal length. As flattened, the width of the body increases up to the eighth segment; the lateral portions of the ninth and tenth turn backward and inward toward the telson. Throughout the length there is a narrow axial lobe which does not exceed one fifth the total width. It is limited by a pair of very narrow longitudinal furrows on the inside of the test, which must have found expression externally as corresponding linear ridges. Specimen M.C.Z. 1846 B represents an individual in the more com- mon state of preservation, flattened from the side. The anterior end of the head is a pointed rostrum, turned upward in this as in other specimens similarly compressed. The body has ten segments, behind the last of which is a telson. The back of the head and of each segment projects high above the front of the adjacent segment, giving the im- pression that there was a sort of crest along the median line. Closer study shows, however, that several of these angular peaks are double, the shell having broken along the back when it was flattened. A faint longitudinal line indicates the lateral border of the axial lobe. The pleural portions of the segments extend downward and slightly back- ward, each ending in a recurved point. The Appendages The outstanding characteristic of Leanchoilia is the presence of a pair of huge tripartite appendages on the head. These are well pre- served in specimen M.C.Z. 1843, which retains an almost complete one belonging to the right side, and a less perfect one on the left. The innermost segment is not completely exposed on either, but is short, abruptly curved, forming a sort of peduncle by which the appendage is suspended beneath the head. Because of the curvature, the proximal end is nearly parallel to the axis of the appendage, sug- gesting that it was the locus of attachment, but this has not been definitely verified. The second segment is stout, much wider at the outer than at the inner end. The profile of the lower margin is strongly convex, con- tinuing the curvature of the proximal segment. At its distal end it supports two branches, an outer double one, and an inner, stouter, single ramus. The latter consists of two segments, the first short and RAYMOND: MID-CAMBRIAN ARTHROPODA 207 stout, the second elongate, tapering outward nearly to the end, where it is somewhat enlarged. There are three terminal spines, apparently movable, since they are not in the same position in all specimens. One of these spines is nearly straight, very long, and finger-like. The other two are shorter, somewhat stouter, and gently curved. The appendage being described is that of the right side, and is flattened in such a way as to show its inner face. In this position, the first of the segments of the uniramous branch has so broad an articula- tion with the second segment of the appendage that it overlaps a large part of the articulatory end of the double branch. It is therefore on the inner side of the appendage, so is an endopodite. Fra. 1. The right “great appendage” of specimen M.C.Z. 1843, showing the segmented protopodite, the bifurcated exopodite and the spine-bearing endopodite. x2. The exopodite consists of two long slender branches which appear to be entirely independent of one another. Each consists of a very long slender proximal shaft, at the end of which is a lash made up of short segments, each about half as long as wide. The flexible portion appears to be about half as long as the proximal segment. One of the branches is dorsad in its position in regard to the other. 'The upper one articu- lates with the upper part of the distal end of the large second segment of the appendage. The lower one articulates independently with the end of the same segment, its proximal end being below that of the upper branch and outside the first segment of the endopodite. This is best shown on the left large appendage of specimen M.C.Z. 1843. If we use the terminology commonly employed in describing Recent crustaceans, the segment bearing the endopodite is the basipodite, and the short curved one inside it, the coxopodite. Together they make up the protopodite. Measurements: The following measurements of the parts of the appendages on the right side of specimen M.C.Z. 1843 will suggest the 208 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY relative proportions. The coxopodite is 4 mm. long on the median line, and 2.5 mm. wide at the distal end. The basipodite is 7 mm. long, 5 mm. wide at the distal end. Total length of endopodite without spines, 17.25 mm.; length of inner segment, 4.25 mm., width 3 mm.; length of outer segment, 13 mm., width at base, 2 mm., at narrowest part, 1.25 mm. Length of first segment of dorsal branch of exopodite, 16 mm., width at proximal end, 1.5 mm., at distal end, 0.75 mm. Length - of first segment of ventral branch of exopodite, 17 mm.; length of segmented portion, 6 mm. Other appendages of the head All other appendages of the head are small, and specimens so far found do not permit their proper description. It is as yet impossible to determine how many there are, or even their exact nature. Specimen M.C.Z. 1843 shows faint traces of appendages which are pressed upon the proximal segments of the large ones just described. One pair of these may be uniramous tactile organs, but they are so far from their natural position that one can not be sure. The better preserved of the two is 8.5 mm. long. It lies, with the apex turned backward, across the basipodites of both the large appendages. It is nearly straight, but shows traces of segmentation. Its mate is curved. Only a small portion is visible, as it disappears forward beneath other structures. Although specimen M.C.Z. 1844 has appendages which may be like these, they are too poorly preserved to be satisfactory. The other cephalic appendages of specimen M.C.Z. 1843 are too fragmentary to be of any value. Specimen M.C.Z. 1842 shows a pair of appendages in front of the large tactile organs, but whether they belong in front of them is doubtful. They are obviously displaced, turned end for end. One lies just at the margin of the head, which is flattened dorso-ventrally, the specimen revealing the lower side. This appendage belongs to the right side. The anterior end is at the base of the rostrum, the posterior one just in front of the protopodite of the large tactile organ. The append- age is about two and a half times as long as wide, approximately parallel-sided. The posterior end appears to have three tooth-like processes, the inner one reinforced by a wide lateral addition to the shaft. The corresponding limb of the left side is pressed against the inside of the head, so that it is difficult to distinguish its outline. That it can be seen at all is due to the fact that these appendages have weathered to a reddish rust, whereas the greater part of the test is RAYMOND: MID-CAMBRIAN ARTHROPODA 209 yellow. It seems to be somewhat longer and more slender than its companion, but the posterior end shows a similar arrangement of the tooth-like processes. The general form of these two appendages suggests that they are mandibles. It must, however, be admitted that they are not satisfactorily preserved. No trace of them has yet been observed on other specimens. The only other appendages yet seen on the head are fimbriated exopodites of the same general type as those described below аз occur- ring on the thoracic segments. Specimen M.C.Z. 1842 has one extend- ing out from about the middle of the left side of the head, another, but faintly visible, pressed upon the inside of the test in advance of the proximal end of the right "great" appendage, and another pair flattened against the inside of the posterior part of the head shield. Specimen M.C.Z. 1846 B shows parts of two exopodites underneath the head, one of them very clearly preserved. Specimen M.C.Z. 1845 has a row of eleven exopodites on one side. Since there are only ten segments in the trunk, at least one of the appendages must belong to the head, which is not preserved on this individual. The arrangement of the exopodites indicates that at least ten were thoracic appendages. Specimen M.C.Z. 1844 also shows an exopodite belonging to the head. All these cephalic exopodites are broad, nearly circular in outline, differing in this respect from the narrower ones on the segments of the trunk. Appendages on the trunk Specimen M.C.Z. 1846 B is complete, and crushed from the side. It retains the “great appendages" and the telson, so there can be по question as to the identification. The exopodites extend far below the test on the left side, their outer ends being beautifully preserved. Ten are clearly visible, somewhat overlapping one another, and not all in exactly the same plane. Since the number corresponds to the number of thoracic segments, it is probable that they all belong to the same side. This deduction is the more reasonable because the animal is flattened in such а way that it shows somewhat more than half the shell, hence the appendages of the right side must be entirely beneath the test. Each exopodite has a long, broad, blade-like shaft, with close-set finger-like "filaments" projecting from the distal and posterior mar- gins. Except that it is not made up of two segments, and that filaments are present on the end, the exopodite is very like that of Neolenus. 210 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY There seems to be but little difference between those of the anterior and posterior segments of the trunk. To give some idea of the proportions, 1t may be stated that the ex- opodite of the third segment projects 15 mm. beyond the test, is 5 mm. wide 6 mm. from the outer end, and the longest filaments on the pos- terior margin are 4 mm. long. In a linear distance of 4 mm. along the posterior margin there are nine filaments. The distal portion of the Fra.2. Outline of specimen M.C.Z. 1846 B, showing the tips of the thoracic exopodites. The irregular object beneath the dorsal spines represents the filling of the alimentary canal. xll4. exopodite of the first segment is especially well preserved. The outer end is nearly semicircular in outline and carries eleven filaments, the longest 3 mm. long. Specimen M.C.Z. 1845 is flattened dorso-ventrally, preserving the telson and parts of all the segments of the body. Alongside it are eleven exopodites, rather evenly spread, apparently remaining opposite the segments to which they belong. These exopodites expose their inner faces, which are somewhat different from the outer ones shown so well by M.C.Z. 1846. Although no individual appendage is abso- lutely complete, two or three of the anterior ones are almost so. Two | | “ | RAYMOND: MID-CAMBRIAN ARTHROPODA 211 of them, fortunately, show the articulatory ends. The blade has a broad articulation with a proximal segment, which is short, nearly oval in outline as flattened, drawn out mesially into a short, narrow projection. There is no indication of how this protopodite was attached to the ventral membrane. The spinelike inner end was probably not connected with the body, but was free to act as an endobase or gnatho- base. The blade-like portion appears narrower when seen from the inside than from the outside, and its posterior margin is less distinct. This seems to be due to the attachment of the filaments to the inner face some distance in front of the posterior margin of the blade. There is no distinct line along which they are attached, but they may be seen pressed against the inside of about the posterior half of the blade. One exopodite looks as though it were made up of two parts, a short inner and a longer outer one. Since this condition is not seen in any other, it is probably due to an accidental displacement of the filaments. 'The posterior exopodites of this specimen are distinctly shorter than the anterior ones, as would be expected. Endopodites were certainly present on the trunk of Leanchoilia, but the specimens now at hand give little information as to their form, and no suggestion as to their number. Specimen M.C.Z. 1843 shows the distal end of one behind the left *great appendage." It has a long, slender terminal spine with two subordinate ones. Behind it are tips of two or three more appendages which are probably endopodites, for they end in spines which are longer and fewer in number than the filaments on the exopodites. Specimen M.C.Z. 1846 B shows the distal end of a slender endopodite just in front of the first exopodite of the trunk. It is, however, too poorly preserved to be of any value. Alimentary tract Several specimens show in some part the filling of the alimentary canal. It is most conspicuous in specimens M.C.Z. 1846 A and В, which are part and counterpart. This animal died with, perhaps be- cause of, a great mass of food in that part of the canal between the posterior margin of the head and the ninth segment. The alimentary tube was segmented, constructed at the boundaries between the segments of the trunk. The material within the tube consists of a black carbonaceous substance in oval masses of considerable thickness. One of the well preserved pieces, filling the enlargement in a single segment, is 4 mm. long, 3 mm. wide, and about 1 mm. in thickness. 212 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY The surface is transversely laminated, indicating that it had been subjected to the pressure of annular muscles in the walls of the ali- mentary canal. Telson The telson is elongate, tapering gradually to an acute point. In the Peter Redpath Museum at McGill University there is a very good specimen, flattened from above, about 8.5 mm. long and 3 mm. wide at the anterior end. The surface is featureless, without ornamentation or median ridge. Tapering spines line both the sides, appearing to emerge from beneath the telson. They are directed backward, their axes parallel to one another, at a low angle, about 30^, with the axis of the telson. Another isolated telson of about the same size in the Museum of Comparative Zoólogy, appears to show the lower surface. The spines are directed backward at a still lower angle with the axis than those of the McGill specimen. They appear to be inserted in sockets, in a fashion similar to the movable spines on the margins of the metasoma of Limulus. A darker band about the base of each spine indicates the thickening of the chitin bordering the socket. The telson of still another much distorted and imperfect individual supports the contention that the lateral spines are inserted in sockets, for, as flattened, one side is scalloped, the proximal ends of the spines being in the recesses. The telsons of the individuals which have been laterally compressed are too poorly preserved to be of any service. АП of the specimens mentioned above show a small ventral under- turning of the margins of the shell, a sort of doublure; only at the posterior end is the ventral side completely covered. The telson was not therefore a mere spine-like process, as is that of Limulus. Reconstruction Leanchoilia appears to have been a short, laterally compressed, round-backed animal with a sharp rostrum and narrow elongate telson. The head was covered by a fold of shell, the sides of which were progressively deeper toward the back, whereas the anterior end was an uptilted spine. The ten segments of the trunk were provided with deep lateral extensions, protecting the body and appendages. Since there is no trace of the presence of sternites, the under surface must have been a thin membrane, as in the trilobites. The appendages RAYMOND: MID-CAMBRIAN ARTHROPODA 213 were attached to this membrane, and not, with the possible exception of the “great appendages” of the head, to the test. Walcott states that “The large opening on the side of the head indi- cates a large pedunculated eye comparable to that of Opabinia regalis." As a matter of fact, there is no opening on the side of the head, even in the holotype. This specimen shows a reniform depression which occupies a large area on the side of the head. The position and shape are correctly shown in Walcott's figure (1912, pl. 31, fig. 6). Similar areas are to be seen on two other specimens (one numbered 239 N on Fra. 3. А restoration of Leanchoilia superlata Walcott, about one half larger than the average specimen. А green label) in Walcott's collection. Although no lenses are visible on any of these, it seems likely that the reniform areas are the remains of very large sessile compound eyes. If so, Leanchoilia was probably an inhabitant of rather deep water, in the zone of feeble light. The probability that the animal had sessile eyes is increased by the fact that no indications of stalked eyes exist. Not only is none pre- served on any specimen, but there are no notches in the margin of the head-shield, or other evidences of their presence. 'The only satisfactorily known organs of the head are the "great appendages,” already described. These were held beneath the head in a horizontal position by the curved protopodites. Although no proof remains, they were probably the second appendages, the antennae. There is some evidence that there was a pair of uniramous antennules in front of them. The mandibles are known only from the inner por- tions, the gnathobases, which are armed with denticulations at the inner ends. That each had an exopodite is probable, for impressions of such are found ahead of the bases of the “great appendages” on specimen M.C.Z. 1842. These must have belonged to appendages far forward, and hence probably to the mandibles, which are the first 214 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY appendages which would be expected to show the flattened setiferous type of exopodite. Slender endopodites were probabl y present behind the mandibles, for indications of such are seen on specimen M.C.Z. 1843. I believe the head had five pairs of appendages, although there is no satisfactory indication of more than three. That at least two of the pairs had large, somewhat fan-like, setiferous exopodites is, how- ever, clear. Each segment of the trunk bears an appendage with a short coxo- podite, probably attached along its upper surface to the ventral membrane, in such a way as to leave the inner end free to serve as a gnathobase. The long blade-like exopodite is attached to the outer end of the coxopodite. There is no very definite evidence of endopodites, although such organs were presumably present. If so, they were slender, segmented branches such as are suggested by specimen M.C.Z. 1843. The telson was spine-shaped, but with a narrow doublure instead of a complete ventral covering. It bore no appendages, but several pairs of slender spines were articulated in sockets on its margins. Relationships to Other Arthropods Although the structure of Leanchoilia is not fully known, still it obviously, like all the other arthropods from the Burgess shale, has the trilobitan type of appendages, hence can not be referred di- rectly to any of the modern orders of Arthropoda. The antennae are curiously like those of certain Decapoda, a fact of no significance as an indication of relationship. Suggestions as to possible connections with modern animals can be obtained only by reviewing again the whole arthropod fauna of the Burgess shale. Waleott recognized, among the fossils retaining appendages, the following:- Merostomata. Limulava.—Sidneyia. Aglaspina.—Molaria, Habelia, Emeraldella. Trilobita.—Marrella, Nathorstia, Neolenus. Branchiopoda. Anostraca.—Opabinia, Leanchoilia, Yohoia, Bidentia. Notostraca.—Naraoia, Burgessia, Anomalocaris, Waptia. Malacostraca. Hymenocarina.—Hymenocaris. Contrary to the usual custom, Walcott included the Merostomata RAYMOND: MID-CAMBRIAN ARTHROPODA 215 among the Crustacea, hence, although the Limulava resemble euryp- terids and the Aglaspina suggest relationship with the xiphosurans, no member of the fauna was described as an arachnid. Reviewing these arthropods as best I could from the published descriptions and figures, I (1920, p. 106 et seq.) agreed with Walcott that all were crustaceans, and, to avoid the implications of the term Merostomata, removed the Limulava and Aglaspina to new subclasses. I pointed out that, so far as could be determined, all of these animals had tactile antennules, and all had at least some exopodites and en- dopodites of the trilobitan type. My scheme was, then, only slightly different from that of Walcott. I recognized: Trilobita. Naraoia, Nathorstia (= Neolenus). Branchiopoda. Anostraca.—Opabinia. Notostraca.—Burgessia, Waptia, Yohoia. Haplopoda (a new subclass). Marrellina.—Marrella. Aglaspina.—Molaria, Habelia. Xenopoda (а new subclass). Limulava.—Sidneyia, Emeraldella. Malacostraca. Hymenocarina.—H ymenocaris Fedotov (1924, p. 384) reviewed the papers of both Walcott and myself, reaching conclusions which in some respects differ from both. His views may be summarized as follows: Branchiopoda. Anostraca.—Opabinia, Yohoia. Notostraca.—Burgessia. Conchostraca or Cladocera.—Marrella. Branchiopoda or Ostracoda—Hymenocaris. Malacostraca. Leptostraca.—Waptia. Possibly related to Isopoda.—Molaria, Habelia, Emeraldella. Possibly related to Amphipoda.—Leanchoilia. Merostomata (Arachnida). Sidneyia. Henriksen (1928, pp. 1-20) discussed Walcott’s papers, with the advantage of having specimens for comparison. He reached the fol- lowing conclusions: Branchiopoda. Anostraca.—Opabinia, Bidentia. 216 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY Near Notostraca.—Marrella, Burgessia, Yohoia. Malacostraca. Leptostraca.—Waptia, Hymenocaris, Carnarvonia, Anomalocaris. Merostomata. Leanchoilia. Aglaspina.—Molaria, Habelia, Emeraldella. Xiphosura.—Naraoia. Sidneyia was placed in a group sharing the characteristics of the Trilobita, Branchiopoda and Merostomata, but not assigned to any subclass. Walcott (1931) in a posthumous paper, edited by C. E. Resser, has supplied new figures and restorations of Naraoia, Marrella, Burgessia, Leanchoilia, and Waptia. All of these corroborate my opinion of 1920 that the appendages are essentially like those of trilobites. Störmer (1933) has recently done a signal service in pointing out that the trilobite exopodite is a very distinctive thing, having a structure un- like that of any crustacean and paralleled only in Limulus. The exopodites of trilobites are fringed with long, blade-like filaments, which differ entirely from the setae of crustaceans. The exopodites of the other Mid-Cambrian arthropods, so far as known, are of this same type, hence Stórmer has questioned whether they are not, after all, arachnids. It seems to me that the outstanding characteristies of Crustacea are the presence of tactile organs as the first pair of appendages, and the rather universal evidence of the actual or former presence of biramous limbs. That filamentiferous exopodites should be found only in the trilobites, other Mid-Cambrian arthropods, and Limulus, seems merely to strengthen the belief that the latter animal is descended from some group of the former, retaining the primitive type of gill. The trilobitan exopodite was not a very efficientsortof gill. Itschitinous covering was apparently relatively thick, reducing the effectiveness of its respiratory function. It was too clumsy to aid in swimming; in fact it is probable that the better the trilobite swam, the more closely it kept its exopodites folded across the ventral membrane. Hence as better methods of respiration evolved in the Crustacea, exopodites degenerated till only a part of the primitive shaft remained. This has been lost, or has become variously modified in different groups. The xiphosurans, on the other hand, have retained the general shape and some habits of the majority of the trilobites. Bottom dwellers, and relatively poor swimmers, they have retained, although in a very considerably modified form, the ancient type of gill. RAYMOND: MID-CAMBRIAN ARTHROPODA 21 The Mid-Cambrian Branchiopoda In my discussion in 1920 of the fossils described by Walcott from Burgess Pass, I took the position that, for purposes of classification, the general habitus was more important than the nature of the ap- pendages. I therefore followed Walcott in assigning Burgessia, Waptia, Yohoia, and Opabinia to the Branchiopoda. The general form of some of these animals is so similar to that of certain modern branchiopods as to suggest that they were really ancestral. Since I have been able to study actual specimens, however, I have come to doubt the correct- ness of this view. On the other hand, the general form should not be entirely ignored. It seems improbable that animals whose structure so resembles that of branchiopods should not have had some sort of an- cestral relationship. As will be noted in the summary of views cited above, all have agreed that Opabinia is an anostracan branchiopod. Hutchinson (1930, pp. 1-13) has recently restudied the specimens, and found that, although the structure was in most respects comparable to that of modern representatives of this order, still there were certain differences, which caused him to erect for it, and a few other extinct species, a new sub- order, the Palaeanostraca. The anostracan characteristics cited by Hutchinson are: absence of carapace, general form and size, presence of pedunculate eyes, position of the antennae and probable reduction of the antennules, a nearly uniform series of foliaceous trunk limbs, and the presence of a frontal process in the male. As may be seen, these are chiefly characteristics of the body, not of the appendages. By his reference to “foliaceous” limbs, Hutchinson apparently did not mean to imply that the limbs were phyllopodan. As figured and described by him, the appendages of the trunk are long, blade-like, smooth externally, with “уегу strong thick setae" on the inner edge. Г am inclined to think that Hutchinson, un- accustomed to dealing with fossils, was misled by the extraordinary state of preservation of the specimens. Through the kindness of Dr. Resser, I have been able to study Walcott's types (1912, pl. 27, fig. 6, and pl. 28, fig. 1). Both are ex- cellently preserved, but the structure is clearest in the specimen figured by Wolcott on pl. 28 (U.S. Nat. Mus. no. 57,684). This is exposed from the dorsal side with the greater part of the test broken off, so that most of the appendages are revealed. Enough of the shell remains to show that the axial lobe was narrow, and that the pleural lobes were sufficiently wide to cover the appendages when in their 218 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY natural position. No endopodites are seen, but the exopodites much resemble those of Neolenus. I was unable to see any traces of seg- mentation in any exopodite, but since the rounded distal end lacks filaments, that part is probably homologous with the outer segment of Neolenus. The anterior margin appears to be thickened, and the posterior edge of the greater part of the shaft supports long, flattened filaments. The adjacent members of this fringe overlap so much as to give the appearance of a continuous plate. Тыз doubtless accounts for Hutchinson's description of the appendage as carrying a series of “very strong thick setae" on the “internal edge." Walcott's figure 1, pl. 28 has been considerably retouched, but gives a fair idea of the filaments of the exopodites. That the filaments are really discrete is clearly shown by a specimen with the green label numbered 404 (U.S. National Museum). Aecording to this interpretation, the chief non-branchiopodan characteristics are the presence of pleural lobes, a small number of abdominal segments, and the retention of the trilobitan exopodite. Walcott (1931, p. 15) has given much new information about the appendages of Burgessia, showing that they are typically trilobitan, those of the trunk having both exopodites and endopodites. Both antennules and antennae are simple. Waptia has, according to the most recent restoration (Walcott, 1931, p. 20), endopodites only on the thorax, and exopodites only on the anterior segments of the abdomen, а most extraordinary combination. We probably have much to learn about this creature. In any event, the appendages are trilobitan. Burgessia and Waptia are decidedly notostracan in habit. Both Fedotov and Henriksen placed Burgessia in or near the Notostraca, but both assigned Waptia to the Malacostraca. If any great weight is given to its trilobite-like appendages, the latter assignment is obviously impossible. Although, with present knowledge, there are no real connecting links between these Mid-Cambrian “‘branchiopods” and modern ones, future discoveries may yield them. Such forms as Lepidocaris rhyniensis (Scourfield, 1926, pp. 153-187), from even so young a deposit as the Mid-Devonian, will probably fit into the sequence, although rather far along in the series. Scourfield points out that the first three pairs of appendages on the trunk are practically phyllopodan, whereas those on the fourth to eleventh are biramous, similar to those of young copepods. Apparently the condition is what Jackson has charac- terized as a “localized stage in development." The first three pairs show a modified, the others a primitive condition. The anterior limbs RAYMOND: MID-CAMBRIAN ARTHROPODA 219 have a well developed corm, with a gnathobase, five endites, a terminal lobe, and an exite or flabellum appearing as a process on the outer side. The remaining trunk appendages are biramous, the endopodite and exopodite each consisting of a single segment, as does the protopodite. The latter shows rudimentary endites, and a well developed gnathobase. As Scourfield points out, there is a clear transition from one type of limb into the other, showing that the phyllopodan limb is a modifica- tion of the biramous type. Students of the Mid-Cambrian fossils will see, in the copepodoid limbs on the 4th to 11th segments, vestiges of a degenerate trilobitan appendage. It is especially interesting to note that the protopodite is unsegmented throughout the series. The corm and endites are obvi- ously modifications of the trilobitan coxopodite. Scourfield homologizes the terminal lobe with the endopodite, and the flabellum with the exopodite. There seems, however, some reason to interpret the series as indicating that the endopodite was lost, hence the exopodite formed the palmate terminal lobe, and the flabellum was a new outgrowth on the outer side of the corm. Whatever the interpretation, Lepido- caris is an extremely important link in the series conriecting the Cam- brian branchiopod-like animals with the real branchiopods. It is particularly interesting in that it indicates that the phyllopodan type of limb was just beginning to be evolved in Mid-Devonian times, perhaps in response to the invasion of the fresh waters by this type of crustacean. As a result of this review of the evidence accessible up to the present time, 1t seems that the best procedure is to remove the Mid-Cambrian forms from the subclass Branchiopoda, although still considering them as belonging to the stock from which the modern group was derived. The Mid-Cambrian Malacostraca Turning now to another group, it will be noted that Fedotov doubted the propriety of referring Hymenocaris to the Malacostraca as а member of the Leptostraca. I now believe he was right in this, although Henriksen supported the view of Walcott, which I had previously followed. Here again we are confronted with the question of the relative value of form and appendages. In this case, however, we are dealing with an almost extinct group, the Leptostraca being represented at the present time only by four genera of the family Nebaliidae. Although great numbers of bivalved erustaceans have been described 220 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY from Paleozoic strata, very few have been found which show any appendages. The present opinion as to their distribution in various orders is shown by Ulrich and Bassler in their recent monograph on the Cambrian Conchostraca (1931, pp. 4, 5). The small bivalved forms found in the Cambrian are referred to the Conchostraca because the adductor muscle scar of most of them is near the anterior margin, just below the ocular tubercle. The highly problematical Ribeiria, Euchasma, Eopteria, etc., are referred to the Notostraca, as has been the custom for some time. Anomalocaris, Hymenocaris, and Walcott's genera from the Mid-Cambrian are placed in the Hymenocarina, under the Phyllocarida, and recognized as Malacostraca. Other somewhat similar fossils of the Paleozoic are referred to the suborders Ceratio- carina, Discinocarina, and Rhinocarina of Clarke, all supposed to be malacostracans. Ulrich and Bassler consider the present classification as a tentative and unsatisfactory one. They state: “The correct classification of the small Cambrian bivalved Crustacea is a difficult question. If the super- order Branchiopoda were to include the Phyllocarida as an order or were considered as closely allied, we would not hesitate to place them in the same general association, as we are impressed with the view that these Cambrian forms are an early phase of offshoot of the stock from which the Branchiopoda and Phyllocarida on one hand, and the Ostracoda on the other were derived." Looking over the illustrations, one is inclined to agree with them that the Ostracoda may well have been derived from the Bradoriidae and Beyrichonidae. In fact, with the exception of Fordilla troyensis | and Modioloides prisca, all the specimens which they have figured lack the growth-lines of the modern Conchostraca. It is true that these shells differ from those of ostracods in the position of the adductor muscles and some other features, but they do not greatly resemble modern conchostracans. Would it not be better to give a new name to this group than to force them into either the Branchiopoda or Ostra- coda? The important point is that they are probably ancestral to the latter group. We do not know their appendages, but from our experi- ence with the other Cambrian arthropods we can safely say that they were neither phyllopodan nor ostracodan. Among the Cambrian crustacea with large compressed carapace, Hymenocaris perfecta Walcott is the only real guide, although some- thing is known of the appendages of Anomalocaris. The strongly developed endopodites and exopodites of the anterior part of the trunk of Hymenocaris perfecta are distinctly trilobitan, not at all malacostra- RAYMOND: MID-CAMBRIAN ARTHROPODA 221 сап. Should the general resemblance of the body to that of Nebalia outweigh the evidence derived from the appendages 7 In Clarke's classification (Eastman-Zittel, 1900, p. 653 et seq.), Hymenocaris 1s given a very isolated position, having been made the type and only member of a suborder of Phyllocarida, the Hymeno- arina. If this view is correct, it seems that there is little reason to consider that there is a continuous series connecting it with the Nebaliidae. The basis for the separation from the Ceratiocarina, to which suborder most of the Paleozoic fossils of this general type were assigned, was that the carapace was univalved, without rostrum in the former, and bivalved, with a free rostrum, in the latter. The carapace of the nebalioids is of course univalve. Several specimens in the Museum of Comparative Zoólogy show that Hymenocaris perfecta, although univalve, has a deep, narrow furrow along the middle of the back, which may indicate the beginning of a hinge. Specimens flattened from above show a deep emargination in the anterior outline, and two individuals have a rostrum inserted in this cleft. This indicates that the Mid-Cambrian specimens referred to Hymenocaris are, in form at least, not very widely different from Ceratiocaris. It is of course possible that the Hymenocaris of the Mid-Cambrian does not, after all, belong to that genus, but that question may be held in abeyance for the present. The only one of the Paleozoic “bivalved” crustaceans whose appendages are at all well known is Nahecaris (Broili, 1928, pp. 1-18; 1929, pp. 253-63). Some beautifully prepared specimens from the well known Lower Devonian slates of Gemiinden and Bundenbach enabled Broili to determine the nature of many of the appendages. The antennules are biramous, a shaft of two long segments bearing a pair of multisegmented tactile organs. The antennae are also biramous, the exopodite composed of numerous ring-like segments, whereas the less flexible endopodite has the proximal half made up of at least two long segments, the distal of short rings. The other cephalic appendages have not yet been seen. The trunk, according to Broili, consists of eight thoracic and eight abdominal segments. In this count he considers the last apparent segment to be a combination of the last abdominal with the telson. This may very well be the case, as the lateral spines unquestionably are attached to it, and not to the segment ahead of it. The first eight pairs of appendages are slender, the first the most elongate. These are biramous, the exopodite weaker than the endo- ) 222 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY podite. 'The others seem to be uniramous but this is probably a matter of imperfect preservation. Broili has restored the endopodite as consisting of a single blade, but this likewise is probably due to the imperfect preservation of the specimen. In another individual, which he described in his second paper (1929, p. 262), he detected indications of joints. It seems very probable that the endopodites were segmented. Five segments of the abdomen bear biramous pleopods, the endopo- dites and exopodites about equal in strength, each apparently consist- ing of one setiferous segment. Broili was impressed by the differences to be noted on comparison of Hymenocaris, Nahecaris, and the Nebaliidae. He saw more similarity between the appendages of Nahecaris and those of Sergestes, a modern pelagie decapod. He concluded that Nehecaris was really malacos- tracan, but not a phyllocarid, erecting for it the group of Nahecarida, to have the same rank in the classification of the Crustacea as the Phyllocarida. I have been fortunate enough, through the kindness of Mr. T. H. Withers, to be able to see some of the appendage-bearing specimens in the British Museum. The appendages of the trunk are not abso- lutely clear, and, as has been said, it is possible that the endopodites are segmented. The restoration of the tactile organs in the paper of 1929 is much better than the earlier one. In some respects Nahecaris is more like a decapod than like the Nebaliidae. If the endopodites of the thoracic limbs are really unseg- mented, it cannot be ancestral to either. If as seems probable, they are segmented, it is more primitive than either, though probably not directly ancestral, since at least two segments lack appendages. Although considerable modification of the limbs would be necessary to change them so as to conform to the type found in the Nebaliidae, none of these changes is of an improbable nature. Let us look at the chief structural features. In the first place, the carapace of Nahecaris is not bivalved. An excellent specimen in the Museum of Comparative Zoólogy, flattened from above, shows conclusively that the shell was all in one piece. It cannot even be interpreted as two shells in symphysis. One wonders how many of the so-called bivalved phyllocarids really had any separation along the dorsal side. The assumption that such separation existed is based chiefly on the fact that the shells as found, show, in many cases, a perfect outline of one valve only, with no trace of the other. It is therefore inferred that the two valves separated auto- RAYMOND: MID-CAMBRIAN ARTHROPODA 220 matically on the decay of the soft tissues. In many of these cases one valve may be buried beneath its fellow; in others, separation may have taken place because of the weakness of the material along the “hinge.” In still other cases, the carapace appears to have been actually in two parts, as in the Rhinocarina. Curiously enough, Hennig (1922, p. 140) placed Nahecaris in the family Rhinocaridae, a group whose chief characteristic is that the valves are articulated by “interlocking at the single point where they come in contact." Broili points out that Nahecaris has five segments behind the carapace (he counts the an- terior half of the telson as an abdominal segment), whereas the Rhinocaridae show only three at the most. Hence he makes a new family, the Nahecaridae. Since the carapace is all in one piece, and the rostrum is a projection from it, not articulated with it, and since there are four segments (if the whole of the last one be considered the telson), behind the cara- pace, then the general form of Nahecaris corresponds exactly with that of Nebalia. If we compare the appendages in detail, we find a certain amount of agreement. The antennules of Nebalia have a peduncle of four seg- ments, a flagellum of short segments, and a scale representing the other branch. Only two segments of the peduncle of Nahecaris are known, but the full length may not have been seen. The exopodite is lost from the antenna of Nebalia. The endopodite has a long peduncle as in Nahecaris. Lacking the other cephalic appendages of Nahecaris, we may turn to the thoracic appendages. Eight pairs of similar limbs are present in both Nahecaris and Nebalia. In the latter both exopodite and endo- podite are present. Both are flattened, the exopodite unsegmented, the endopodite only obscurely segmented, not separated from the basi- podite. A large epipodite is attached to the outer side of the basipodite. This type of limb closely approximates the phyllopodan, a condition more nearly reached in Nebaliopsis. Paranebalia and Nebaliella, however, show distinctly segmented endopodites (Calman, 1909, pp. 154, 155). If Nahecaris and these latter are properly interpreted, it would not be possible to derive the nebalioid thoracic limb from that of Nahecaris. It is, however, very possible that the appendages of the latter are really segmented, and that the segmentation has been obscured by the preservation or by the process of cleaning. If they were segmented, the segments must have been short, as in Nebaliella. The pleopods of the two groups are very similar, except that in the Nebaliidae the third and fourth of the four pairs are uniramous. 224 BULLETIN: MUSEUM OF COMPARATIVE ZOÓLOGY From the present evidence it appears that both the decapods and the Nebaliidae may have been derived from some Nahecaris-like ancestor, but that more modification has taken place in the latter than in the former line of evolution. Once again the phyllopodan type of limb seems to be a secondary modification. We may, then, look upon the Nebaliidae, not as intermediate in their characteristics between Branchiopoda and Malacostraca, but as specialized descendants of the really primitive group which gave rise to the higher Crustacea. As Broili has pointed out, Hymenocaris appears to be more special- ized than Nahecaris, because it seems to lack appendages on the abdominal segments. I am inclined to believe that further study of Hymenocaris perfecta will show that this animal had appendages on all segments of the trunk. A specimen in the Museum of Comparative Zoology shows fragments of exopodites pressed against the test of several of the segments behind the carapace, the last trace of them being on the segment in front of the telson. Although the specimen is unsatisfactory, for the fragments are obviously displaced, if affords a suggestion of a primitive condition. To recapitulate briefly: the form of Hymenocaris seems to be con- tinued in the Ceratiocarina, to which group the Rhinocarina are allied; appendages are practically unknown in either of these groups. Hymenocaris, with its free rostrum, was probably not ancestral to Nahecaris, nor to the modern Nebaliidae. There is, however, in the Cambrian, а genus which has a carapace somewhat similar to that of Hymenocaris but which has a fixed instead of a free rostrum. This is Tuzoia, Walcott (Resser, 1929), a crustacean with a univalved but much compressed carapace. The body and appendages are, unfortunately, unknown. It seems probable, however, that it belongs to the same group as Hymenocaris, although certainly not to the same family. It may have been some Tuzoia-like animal which was ancestral to Nahecaris. The general form of the carapace, the abdominal segments, and the peculiar telson, are so much alike in the Ceratocarina, Rhinocarina, and Nahecarida аз to make it almost certain that they are all of the same lineage. Lacking any definite knowledge of the appendages of the Ceratio- carina, Rhinocarina, and Discinocarina, it is probably advantageous to group them, as Broili has done, under von Stromer's name Archae- ostraca. If so, we must return to Claus's term Leptostraca for the Nebaliidae, adopt Broili’s Nahecarida, group Hymenocaris with the other arthropods with trilobitan appendages, and discard Packard's Phyllocarida. RAYMOND: MID-CAMBRIAN ARTHROPODA 220 The general result of this entire discussion is to indicate that the Mid-Cambrian crustaceans are related to animals of later date with similar forms, but that the relationships are not by any means so close as some have supposed. Аз yet no fossils have been found which show the steps in the change from the trilobitan to the typical crustacean limb, hence no one of the Mid-Cambrian creatures should be placed in any order whose definition is based upon characteristics of modern animals. Personally I do not think they should be included in the same subclasses. The Position of Leanchoilia As has already been stated, Walcott considered Leanchoilia to be an anostracan branchiopod, Fedotov suggested that it might be an amphi- pod, whereas Henriksen removed it entirely from the Crustacea, think- ing its relationships were with the Merostomata. Му study of the appendages reveals nothing to support Henriksen's position. A spine-like telson is common enough among the Crustacea; at any rate, this single characteristic would not warrant referring the animal to the Arachnida. On the other hand, the “great appendages” of the head have a typically crustacean structure; they, and others of the limbs, are typically biramous. Under the current definition of the class, Leanchoilia must be referred to the Crustacea. It is equally obvious that the genus can not be included either among the branchiopods or the amphipods. As in the branchiopods, the appendages are serially similar, but the limbs in no way resemble the leaf-like organs of the phyllopodan type. On the contrary they are essentially of the trilobitan type, closely comparable to the exopodites of Neolenus. Leanchoilia appears to have been a highly specialized animal, not, so far as I can see, very closely related to anything else, living or fossil. Suggestions for a Classification of Some Paleozoie Crustacea In 1920 (p. 148) I proposed the name Haplopoda for a subclass of Crustacea, to include those members of the larger group with two pairs of uniramous tactile organs, but with the other appendages of the trilobitan type. It happens, unfortunately, that the same name was already in use for a tribe of the Branchiopoda. Itis therefore necessary to replace it, and I now propose the term Homopoda, in allusion to the likeness of the various pairs of limbs. Since it is thus necessary to change the name of the subclass, it is 226 BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY an opportune time to modify the definition. The limits of the group, as first defined, were too narrow for a subclass, and the wording did not actually express what I had in mind at the time. The following scheme will perhaps serve to express the relationships of the various groups so far as our present knowledge will permit: Class CRUSTACEA: Subclass TRILOBITA Walch Subclass HOMOPODA (nomen nov.) Crustacea with two pairs of tactile organs, the other appendages trilobitan. Biramous appendages on some or all the trunk segments. Carapace present or absent. No facial sutures. Order MARRELLINA Raymond Homopoda with trilobite-like form, pleural lobes reduced. An- tennae uniramous. All segments behind the antennal bear biramous appendages. Genus MARRELLA Walcott Order PSEUDANOSTRACA nov. (Leading to Anostraca) Homopoda without carapace. Appendages on all segments, or absent from one or more of the posterior ones. Not all appendages are biramous. Family OPABINIDAE Walcott “Fifteen pedigerous segments followed by a postpedigerous portion divided into two parts, of which at least the anterior one is presumably segmental. Frontal process bifurcated only at the apex." (Hutchin- son, 1930, p. 12). Genus OPABINIA Walcott Family LEANCHOILIIDAE nov. All segments except the telson appendagiferous. Antennae biram- ous, specialized. RAYMOND: MID-CAMBRIAN ARTHROPODA zum Genus LEANCHOILIA Walcott ? Genus ВІРЕХТІА Walcott Family YOHOIIDAE Henriksen Not well enough known for characterization. Genus УонотА Walcott Order PSEUDONOTOSTRACA nov. (Leading to Notostraca) Homopoda with a depressed carapace. Some of the posterior seg- ments without appendages. Family BURGESSIDAE Waleott Carapace large. All segments behind it narrow, without appendages- Genus Burcessia Walcott Family WAPTIDAE Walcott Appendages on all but six of the segments behind the carapace. Telson provided with a pair of swimmerets. Genus МУАРТТА Walcott Family PROTOCARIDAE nov. Carapace distinctly depressed, bilaterally symmetrical with respect to a median "hinge." Trunk with numerous short segments, the anterior ones appendagiferous. Genus Prorocarıs Walcott (emend. lesser) (See Resser, 1929, p. 12) Order HYMENOCARINA Clarke (emended) Homopoda with a compressed carapace. Rostrum free or attached. Biramous appendages on anterior portion of trunk. BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY family HYMENOCARIDAE Salter (emended) Carapace a single shield arched over the anterior part of the body. Rostrum free. Seven to nine segments behind the carapace. Genus Hymenocarıs Salter (emend. Walcott) The genera Hurdia Fieldia and Carnarvonia may belong to this 5 7 7 “ © family. Family ANOMALOCARIDAE nov. Carapace as in Hymenocaris. Trunk elongate, large, with appen- dages on all segments except the last. Genus ANoMALOCARIS Whiteaves Family TUZOIIDAE nov. Carapace with a median keel, prolonged into a rostrum. Surface reticulate. Genus Tuzora Walcott (emend. Resser). Order AGLASPINA Walcott (Leading to Xiphosura?) (See Raymond, 1920, p. 149). Subelass XENOPODA. Raymond. (Leading to Merostomata?) (See Raymond, 1920, p. 149). Subclass ARCHAEOSTRACA von Stromer Crustacea with a compressed or, more rarely, a depressed carapace. Appendages unknown. Order BRADORINA nov. (Leading to Ostracoda) Small, ostracod-like crustaceans with muscle scar high on the anterior end. RAYMOND: MID-CAMBRIAN ARTHROPODA 229 Families BRADORIIDAE Matthew, BEYRICHONIDAE and INDIANIDAE Ulrich and Bassler. (See Ulrich and Bassler, 1931) Order CERATOCARINA Clarke (See Clarke, 1900, 1913) Order RHINOCARINA Clarke (See Clarke, 1900, 1913) Order DISCINOCARINA Clarke (See Clarke, 1900, 1913) Archaeostraca incertae sedis. Euchasma, Eopteria, Ischyrina, Billings; Technophorus, Miller; Ribeiria, Sharpe; Ribeirella, Shubert and Waagen. Subclass MALACOSTRACA Latreille | Order NAHECARIDA Вгош Family NAHECARIDAE Broili Genus NAHECARIS Jaekel " Order LEPTOSTRACA Claus Family NEBALIIDAE Baird ‘ Subclass Branchiopoda Latreille Order LIPOSTRACA Scourfield Carapace absent. Anterior trunk-limbs foliaceous; remainder biramous. Family LEPIDOCARIDIDAE Scourfield Genus LEPIDOCARIS Scourfield BULLETIN: MUSEUM OF COMPARATIVE ZOÖLOGY BIBLIOGRAPH Y Ввопл, Е. Beobachtungen an Nahecaris. Sitzungsberichte der math. naturw. Abteilung der Bayerischen Akad. der Wissenschaften zu München, 1928, pp. 1-18, рі. 1. Ввопл, Е. Beobachtungen an neuen Arthropodenfunden aus den Hunsrück- schiefern. Ibid. 1929, pp. 253-280, pls. 1—5. Carman, W. T. Crustacea, in Lankester, A Treatise on Zoólogy, 1909, pt. 7. CLARKE, J. M. Phyllopoda, Phyllocarida, in Zittel's Text-Book of Palaeon- tology, English edition, 1900, 2d ed. 1913, vol. 1. Керотох, D. On the Relations among the Crustacea, Trilobita, Merostomata and Arachnida. Bull. de l'Académie des Sci. de Russie, 1924 (1925), ser. 6, nos. 12-18, pp. 383-408. Hennig, E. Arthropoden-Funde aus den Bundenbacher Schiefern (Unter- Devon). Palaeontographica, 1922, 64, рр. 131-145, pls. 32-34. HENRIKSEN, К. L. Critical Notes upon some Cambrian Arthropods described by Charles D. Waleott. Vidensk. Meddel. fra Dansk naturh. Forenigen i Köbenhavn, 1928, 86, pp. 1-20. Нотсніхѕом, С. E. Restudy of some Burgess Shale Fossils. Proc. U. S. Nat. Mus., 1930, 78, pp. 1—24, pl. 1. Влумомр, В. E. The Appendages, Anatomy, and Relationships of Trilobites. Mem. Conn. Acad. Arts and Sciences, 1920, 7, pp. 1-169, pls. 1-11. * Виззвв, C. E. New Lower and Middle Cambrian Crustacea. Proc. U. S. Nat. Mus., 1929, 76, pp. 1-18, pls. 1-7. STÓRMER, Г. Are the Trilobites Related to the Arachnids? Am. Jour. Sci., 1933 (5) 26, pp. 147-157. Тлнтсн, E. D. and Bassler, R. S. Cambrian Bivalved Crustacea of the Order Conchostraca. Proc. U. S. Nat. Mus., 1931, 78, pp. 1-130, pls. 1-10. Ульсотт, C. D. Middle Cambrian Branchiopoda, Malacostraca, Trilobita and Merostomata. Smithson. Misel. Colls., 1912, 57, pp. 145-228, pls. 24—84. Warcorr, C. D. Addenda to Descriptions of Burgess Shale Fossils. Edited by C. E. Resser. Smithson. Misel. Colls., 1931, 85, pp. 1-46, pls. 1-23. ІШ Harvard MCZ Librar