va at a TARA LS = tas) Tele We". Sethe ae HARVARD UNIVERSITY IR iS LIBRARY OF THE Museum of Comparative Zoology i : * a Sa 1% ist an As ie a0 ; ee Dale eA! S-\WR -N Lew Haye | Peabody Mascon of Natural History Yale University | MUS. P, ZOOL. Bulletin 21 evn JUN 24 aie NANVArS Menscaen Fossil Mammalia of the” Huertano Formation, Eocene, of Colorado by Peter Robinson vz ar en rail Joe Re ee te FOSSIL MAMMALIA OF THE HUERFANO FORMATION, EOCENE, OF COLORADO Bulletins published by the Peabody Museum of Natural History at Yale University embody research carried out under the auspices of the Museum. ‘The issues are numbered consecutively as independent monographs and appear at irregular intervals. Shorter papers are published at frequent intervals in the Peabody Museum Postilla Series. EpiroriAL Boarp: Elwyn L. Simons, Chairman Willard Hartman A. Lee McAlester N. Philip Ashmole Eprtror: Ellen T. Drake MUS. COMP, ZOOL. LIBRARY JUN 2 4 1966 HAr VAI UNIVERSITy Communications concerning purchase or exchange of publications should be addressed to the Editor, Peabody Museum of Natural History, Yale University, New Haven, Con- necticut 06520, U.S.A. Issued May 10, 1966 PEABODY MUSEUM OF NATURAL HISTORY YALE UNIVERSITY BULLETIN 21 Fossil Mammalia of the Huerfano Formation, Eocene, of Colorado BY PETER ROBINSON University of Colorado Museum Boulder, Colorado NEW HAVEN, CONNECTICUT 1966 Printed in the United States of America CONTENTS List OF ILLUSTRATIONS vi ABSTRACT 1 PART I: GEOLOGY 2 INTRODUCTION 2 ACKNOWLEDGMENTS 3 GEOGRAPHY 4 STRATIGRAPHY 5 Cenozoic System 5 Paleocene Series 5 Eocene Series 6 ?Pliocene Series 16 STRUCTURAL GEOLOGY 17 GEOMORPHOLOGY 19 GEoLocic HisTory 20 PART II: PALEONTOLOGY 22 FosstL LocALitTIEs 22 ‘TAXONOMY | MARSUPIALIA 29 INSECTIVORA 29 PRIMATES 31 TILLODONTIA 43 ‘TAENIODONTA 44 EDENTATA 44 RODENTIA 45 CARNIVORA 46 CoONDYLARTHRA 53 DINOCERATA 60 PANTODONTA 60 PERISSODACTYLA 61 ARTIODACTYLA 68 APPENDIX: STRATIGRAPHIC SECTIONS 72 BIBLIOGRAPHY 79 INDEX 84 Prates I-X 86-95 ILLUSTRATIONS ‘TEXT FIGURES r 2. 3. Chart showing the inferred stratigraphic and age relationships of the Huerfano Index map of Colorado showing the location of Huerfano Park and the appro- priate area (stippled) occupied by the Huerfano and Farisita formations, p. 4. Tectonic map of the Raton Basin, p. 5. and Farisita formations in Huerfano Park and the Spanish Peaks regions, Huer- fano County, Colorado, p. 8. . Sketch map of the Gardner area, Huerfano County, Colorado, showing the location of the major fossil localities and geographic features, p. 17. . Scatter diagram comparison of widths of trigonids and talonids of lower molars of Absarokius from the Lysite, Lost Cabin and Huerfano beds, p. 33. . Scatter diagram comparison of trigonid width and length of M; of Absarokius, pee: . Scatter diagram comparison of widths of trigonid and talonid of M, of Cynodontomys from several localities in the Huerfano formation, p. 38. . Scatter diagram comparison of widths of trigonid and talonid of M, of Cyno- dontomys from the Lysite, Lost Cabin and Huerfano beds, p. 41. 9. Scatter diagram comparison of widths of trigonid and talonid of My, of Micro- syops from the Huerfano and Bridger formations, p. 42. PLATES ee bes ol. Talpavus sp. cf. T. nitidus Fig. 2. Palaeictops bicuspis Fig. 3. Scenopagus priscus Fig, 4,5. rMetacheiromys sp. Fig. 6. Bathyopsis fissidens Big 7. Peratherium sp. Fig. 8. Coryphodon sp. Fig. 9. Palaeosyops fontinalis HW. “Fie. 1. A bsarohius noctivagus nocerai, new subspecies Figs 2. Loveina zephyri Fro. 3. Nyctitherium sp. cf. N. velox ne: Fie. We FHuerfanius rutherfurdi, new genus, new species Kio, 2. Phenacolemur jepseni simpsoni, new subspecies Ris. 3. Shoshonius cooperi Fig. 4,5. Scenopagus edenensis PV 5 Bigs. Cynodontomys scottianus Big.» 2: Cynodontomys knightensis Fig. 3: Hyracotherium craspedotum Fig. 4. Hyracotherium vasacciense ?venticolum Fig. 5,7. Hyracotherium vasacciense vasacciense Bis. 6. Microsyops lundeliusi Vy Kien: Stylinodon sp. Bigs 2: Vulpavus asius Fig. 3. Viverravus gracilis Fig. 4. Uintacyon sp. cf. U. asodes Bis (9: ?Muacis sp. vi VA. Vil. VIIl. IX. Fig, Se are ILLUSTRATIONS Vii Oxyaena sp. cf. O. Lupina Esthonyx acutidens ?Oddectes sp. Muacis parvivorus Oddectes herpestoides Viverravus sicarius Sinopa sp. cf. S. strenua Patriofelis ulta Mesonyx obtusidens Sinopa sp. cf. S. strenua Didymictis ?protenus Didymictis vancleveae, new species Didymictis altidens Didymictis vancleveae, new species Palaeosyops fontinalis Trogosus grangert Notharctus nunienus Palaeosyops fontinalis Eotitanops borealis Lambdotherium popoagicum Helaletes nanus Hyrachyus modestus Eotitanops minimus Helaletes sp. cf. H. nanus Antiacodon pygmaeus huerfanensis, new subspecies Diacodexis sp. cf. D. secans Diacodexts sp. cf. D. chacensis Hyopsodus wortmant Phenacodus vortmani Hyopsodus walcottianus Bunophorus macropternus Hyopsodus paulus Bunophorus sp. cf. B. macropternus — > ae | Me en F ian aly vi My — ey le eee witee . WS” v9e as a4 Atay la tt . iwh rT _ ony Me CLF el sap hy ‘ 58 7 U ni a" may ‘se . >< . a ' adie ita ad P| ; : ae T [4 ‘ wo i rr? Le 8 — i 2) i L. 40" ‘ - ‘ : ¢ : t YALE UNIVERSITY PEABODY MUSEUM OF NATURAL HISTORY BULLETIN No. 21, 95 p., 10 PLs., 9 TEXT-FIGS., 1966 FOSSIL MAMMALIA OF THE HUERFANO FORMATION, EOCENE, OF COLORADO By PETER ROBINSON ABSTRACT Huerfano Basin is a Laramide structure developed between the Sangre de Cristo Mountains (also Laramide) and the Wet Mountains (a rejuvenated Paleozoic structure). Huerfano beds accumulated during a period of relative inactivity in the Laramide Orogeny; fine grained sedimentary rocks derived from mountain areas outside present Huerfano park make up most of the formation. Huerfano sedimentation was brought to a close by a pulse of the Laramide Orogeny within Huerfano Park. Huerfano beds, and their lateral equivalent, the Farisita formation, contain two fossil mammal faunas; one, the Garcia Canyon local fauna, is of Lostcabinian (late early Eocene) age; the younger fauna, the Gardner Butte local fauna, represents a substage older than Blacksforkian (early middle Eocene) and younger than Lostcabinian. The Gardnerbuttean is proposed as a substage of the Wasatchian (early Eocene) stage; characterized by the presence of the following genera: Hyra- cotherium, Coryphodon, Absarokius, Cynodontomys, Didymictis, Eotitanops, Diacodexis and Bunophorus of early Eocene affinity, and Metacheiromys, Mesonyx, Patriofelis, Oddectes, Palaeo- syops, Helaletes and Antiacodon of middle Eocene affinity. FOSSIL MAMMALIA OF THE HUERFANO FORMATION, EOCENE, OF COLORADO* PETER ROBINSON leo eae Ba GEOLOGY INTRODUCTION The study of the Huerfano mammals presented here is the result of a project initiated in 1952 by Dr. George Gaylord Simpson, then Curator of Fossil Mammals and Birds at the American Museum of Natural History and now Alexander Agassiz Professor at Harvard. I started working on the project as one of his field assistants; later Dr. Simpson allowed me to study the fossil mammals. The recent interest of the American Museum of Natural History in Huerfano Park dates from 1952, and the major concern of the museum was collection of fossil mammals from the Huerfano formation. In the same year field parties of the United States Geological Survey (R. B. Johnson and co-workers) started to map Huerfano Park as part of a larger mapping program in the Raton Basin. The U.S.G.S. interests were mainly in coal resources and secondly in oil and gas potentialities. ‘The next year, several parties from the University of Michigan (L. I. Briggs, E. N. Goddard and students) began a long range mapping project of the entire Huerfano Park area. The University of Michigan and the American Museum of Natural History agreed to work on different aspects of the area—the museum collecting fossil mammals from the Huerfano formation and mapping the Eocene rocks and the university concentrating on other geological problems. The occurrence of “Bridger” fossils in Huerfano beds was first noted by Hills (1889). Since the first work of Osborn (1897), faunal levels have been recognized in the Huerfano formation, a lower “Wind River” and upper “Bridger.” In 1916 and 1918 Walter Granger supervised further collection of fossil vertebrates from Huerfano Park and the results of his work were discussed briefly (Granger, 1918). Small segments of the Huerfano formation’s fauna have been described by specialists working on particular groups: Osborn (1919, 1929); Gilmore (1927, 1938); Denison (1938); Wheeler (1960); no one has discussed the entire fauna. Granger recognized the pre-Bridger post-Lost Cabin age of the upper faunal zone of the Huerfano formation (Letter to R. C. Hills dated 14 April 1919) but was prevented from further discussion of the material by the pressure of other work, particularly the Mongolian expeditions. The general consensus has been to equate the upper Huerfano faunal zone with the relatively barren Bridger ‘A’ because the Bridger ‘A’ occupies the appropriate stratigraphic position. Gazin informs me (oral communication) that there may not be a Bridger ‘A’, the particular beds being part of the Green River formation. I see no reason to change the relative position of the upper faunal zone of the Huerfano formation, but there is a semantic question about its temporal position: is it early or middle Eocene? I * Published with the aid of a National Science Foundation Publication Grant No. GN-356. 2 GEOLOGY 3 prefer to consider it the former and discuss my reasons for doing so on page 15; I recognize the essentially arbitrary nature of any placement. However, as Granger noted (letter cited) the upper Huerfano fauna contains elements nicely bridging the gap between the Lost Cabin and Black’s Fork (lower Bridger) faunas. I the so-called Bridger ‘A’ fauna turns out to be equivalent to the upper Huerfano fauna, the Bridger ‘A’ may have to be removed from the early middle Eocene substage (Woodard, others, 1941). Certain problems of the Huerfano fauna are not discussed here; these are: cranial anatomy of the Microsyopidae based on skulls of Microsyops and Cynodontomys collected from locality II and discussion of the quarry sample of Hyracotherium from locality VU. M. C. McKenna is describing the Micro- syopidae and D. B. Kitts will describe the Hyracotherium sample. A. E. Wood described the paramyid rodents in his revision of the Paramyidae and Max Hecht will describe the lizards and snakes. The Huerfano fauna includes at least 58 species of mammals belonging to 42 genera. In addition, perhaps 15-20 species of birds, reptiles, fish and molluscs have been collected, which brings the total faunal sample to approximately 75 species. In many ways Huerfano Park is unique in having a mammal-bearing forma- tion so intimately associated with orogeny. Certain areas of western Wyoming are similarly associated but not as fossiliferous. Because of the association of Laramide deformation and the Huerfano formation, a longer discussion of local and re- gional geology is included than usually is in a faunal study. ACKNOWLEDGMENTS Many people have helped me by criticism, by offering data or both. I am indebted to Drs. S. A. Northrup, V. C. Kelley, J. S. Findley and S. A. Wengerd of the University of New Mexico for criticism of my initial efforts. Drs. L. I. Briggs, E. N. Goddard, David McCulloch, Earle Kauffman and Herbert Tischler of the University of Michigan discussed many of the Huerfano Park problems with me and gave me much data. Drs. Max Hecht and A. E. Wood identified reptiles and rodents, respectively, from the Huerfano formation. Mr. G. T. Benson of Yale University discussed problems of southern Colorado geology with me and showed me the geology of the Stonewall area. Dr. K. M. Waage of Yale has criticized the manuscript and has discussed problems of Mesozoic stratigraphy. The staff of the American Museum of Natural History have been very helpful to me, particularly Drs. E. H. Colbert, Bobb Schaeffer, and M. C. McKenna, Mrs. Rachel H. Nichols and Messrs. Louis Monaco, George O. Whitaker and Joseph A. Nocera. Peter Rona, Ruben Aguirre and Peter Hilgeford helped collect fossils in Huerfano Park. I have discussed problems of Huerfano Park geology or paleontology with Mr. Ross B. Johnson of the United States Geological Survey, Dr. C. Lewis Gazin of the U. S. National Museum, Dr. David Kitts of the University of Oklahoma and Mr. Giles McIntrye of Columbia University. I am extremely indebted to Drs. Joseph T. Gregory formerly of Yale, now at University of California, Berkeley, and George Gaylord Simpson of Harvard for all the help they have given me. The Society of Sigma Xi, the American Museum of Natural History and the John T. Doneghy Jr. Research Fund for Vertebrate Paleontology at Yale University have aided the study with grants for field work. Drs. Mary R. Dawson, Glenn L. Jepsen, John H. Ostrom and Elwyn L. Simons have read the manuscript and offered welcome criticism. 4 FOSSIL MAMMALIA OF THE HUERFANO FORMATION GEOGRAPHY Huerfano Park is a northwest-trending intermountain valley located between the Wet Mountains on the east and the Sangre de Cristo mountains on the west in south-central Colorado. The north boundary of Huerfano Park is the drainage divide between Huerfano River tributaries and streams flowing to the north through the Wet Mountain Valley. The south boundary is a pediment extending from the Spanish Peaks northward to the park. The Huerfano River and its tributaries drain Huerfano Park; the river runs southeastward around the south- east end of the Wet Mountains and then northeastward to the Arkansas River. Muddy Creek, the main tributary of the Huerfano River, joins it at the town of Gardner. Several large hills are found in Huerfano Park. The northernmost is Black Mountain or Promontory Bluff. Little Sheep, Sheep, and Rough Mountains and Mount Mestas are a series of synaxial intrusives of Laramide age standing in the fold belt of the east flank of the Sangre de Cristo Mountains. Dike Mountain, The Black Hills and the Spanish Peaks represent three crossaxial intrusives that lie to the south of Huerfano Park; the Spanish Peaks are noted for the dike networks associated with them (Knopf, 1936, 1956). Settlements in the area are: Gardner, a town of some 200 inhabitants located at the confluence of Muddy Creek and the Huerfano River; Farisita, a hamlet located at the junction of the Huerfano River and Turkey Creek, five miles southeast of Gardner; and Badito, a cluster of houses at the south end of the Wet Mountains where the Dakota sandstone plunges beneath the surface. 108° 1062 104° 40° DENVER Figure 1. An index map of Colorado showing the location of Huerfano Park and the approxi- mate area (stippled) occupied by the Huerfano and Farisita formations. GEOLOGY F Two hamlets are found in the Huerfano Valley above Gardner: Malachite is approximately six miles upstream and Redwing is two miles farther. Eight miles up Muddy Creek from Gardner is the headquarters of the JM Ranch; Bradford, a ghost town, is approximately five miles farther up the Creek. CENOZOIC SYSTEM PALEOCENE SERIES PoIsoN CANYON FORMATION, Hits, 1888: EMEND. 1891. The Poison Canyon formation crops out in a band around the western flank of the Wet Mountains and in front of the foothills belt of the Sangre de Cristo Mountains and southward in the marginal area of the Raton Basin. It rests un- conformably (Black Hills area) or disconformably (Archuleta Draw, Poison Canyon) upon the Pierre shale. The Poison Canyon formation underlies the Huerfano formation conformably at Black Mountain and disconformably in Poison Canyon and Archuleta Draw. The Poison Canyon formation is composed of approximately 1300 feet of yellow to brown, massive, cross-laminated sandstone beds and drab brown or grey silty, often fissile, mudstones. ‘The sandstone beds are resistant and often up\= 4 < WET MOUNTAINS ANTICLINE SANGRE DE CRISTO mHRUST BELT Figure 2. Tectonic map of the Raton Basin modified from John W. Gableman (1956). 6 FOSSIL MAMMALIA OF THE HUERFANO FORMATION form cliffs along valley sides; the mudstone beds are usually covered. Impressions of fossil leaves are common in the shale beds; Briggs and Goddard (1956) report a tropical flora similar to that of the Denver formation from beds 80 feet below the top of the formation at the type locality. The unconformity below the Poison Canyon formation in Huerfano Park in- dicates a depositional break and erosion prior to deposition of the formation. Farther south, in the Raton Mesa region, Johnson and Wood (1956a, b) have shown that the basal beds of the Poison Canyon formation intertongue with the upper beds of the Raton formation in that area. The Trinidad sandstone and the Vermejo formation of late Cretaceous age are present between the Pierre shale and the Raton formation. Because the Poison Canyon formation rests unconformably upon the eroded Trinidad sandstone and Vermejo formation in the Black Hills area (Johnson, Wood and Harbour, 1958), these latter two forma- tions probably extended farther north, perhaps into Huerfano Park. The Raton formation may not have extended far north of the Spanish Peaks area and there- fore not into Huerfano Park. The erosion that preceded the deposition of the Poison Canyon formation is hard to estimate quantitatively. Generally the Cretaceous and younger formations are thinner in Huerfano Park due to the topography of the depositional site as well as to post-depositional erosion. As much as 500 feet of pre-Poison Canyon erosion may have occurred, but the absence of lower Laramide sediments may also be due to non-deposition. If the Trinidad sandstone and the Vermejo formation were eroded from Huerfano Park, the necessary uplift produced little structural discordance. EOCENE SERIES HUERFANO FORMATION, HILLs, 1888; EMEND. 1891 The Huerfano formation crops out in the center of Huerfano Park and south- wards to the southern flank of the Spanish Peaks. As used here, the term Huerfano formation includes the Cuchara formation of current United States Geological Survey usage. In Huerfano Park the Huerfano formation is composed of ap- proximately 2800 feet of pink, red, red-brown and brown silty mudstones and lenses of yellow, tan or green sandstones. The mudstones often contain concretions, which may be encrustations on fossil bone. Farther south, in the Spanish Peaks area, approximately 5000 feet of Huerfano formation is present. The percentage of sandstone increases southward; sandstone forms approximately 10 per cent of the Huerfano formation in Huerfano Park and approximately 50 per cent (John- son, Wood and Harbour, 1958) of the formation in the Spanish Peaks area. The lower part of the Huerfano formation in the Spanish Peaks area is poorly exposed; the upper part is mainly contact metamorphic rocks forming the major portion of the Spanish Peaks. Fossil vertebrates are generally found in the more extensive exposures of the Huerfano formation. Since many of the fossils are small, collecting specimens may require a trained eye. Fossils are more common in mudstones than in sandstones and are more common in Huerfano Park than in the Spanish Peaks area. In Huerfano Park two lower Eocene provincial substages, Lostcabinian and Gardner- buttean, are present. Lower Eocene fossils are present (Robinson, 1960) less than 1000 feet above the base of the formation in the Spanish Peaks area. FARISITA FORMATION, JOHNSON AND Woon, 1956 Johnson and Wood (1956) proposed the term Farisita formation for yellow, tan, buff, white and grey poorly indurated sandstones and silty mudstones. As GEOLOGY a originally described, the Farisita formation was supposed to rest unconformably upon the Huerfano formation. University of Michigan investigators have dis- covered that the Farisita intertongues with the Huerfano formation (Briggs and Goddard, 1956). Faunal evidence indicates that both of the Huerfano faunal zones are present in the Farisita formation; Farisita beds may intertongue with Huerfano beds throughout the entire thickness of the Huerfano formation. Therefore, it seems that the published estimates of 1,000? feet (Briggs and God- dard) and 0-1200 feet (Johnson and Wood) may need revising. Outcrops of the Farisita formation are best exposed between William’s Creek and Turkey Creek, north of Gardner Butte. Similar lithology occurs as inter-beds in the Huerfano formation in the area west of William’s Creek; Farisita beds crop out above Huerfano beds in Black Mountain. erner and Briggs (1958) state that on the basis of lithologic comparisons the source for the Farisita formation was probably the Front Ranges, especially the Wet Mountains. The streams depositing the Farisita sediments are thought to have come from the northwest, in contrast to the streams flowing from the south or southwest that deposited the Huerfano formation. It is difficult to believe that sediments derived from the Wet Mountains were deposited by streams flowing from the northwest, as the Farisita formation lies on the southwest side of the range. It seems more likely that if the Wet Mountains were the source of Farisita sediments, then the sediments came from the northern part of the range and also, perhaps, from the Front Range. The Farisita formation is easy to distinguish from the Huerfano formation by its generally light yellow or tan color, by the lack of induration of the sandstones and by the generally greater percentage of sand in the mudstones. Moreover, fossil wood is common in the Farisita formation and rare in Huerfano beds while fossil vertebrates are scarce in Farisita beds and rather common in the Huerfano facies. Perhaps the evidence that the Farisita and Huerfano formations interfinger through most of their thickness and that they are equivalent in age (except where the Farisita formation overlies Huerfano formation at Black Mountain) should be used as criterion for considering them one unit. The differences between the Huerfano and Farisita are striking and make the two local units mappable. The dissimilarities between the Huerfano and the Cuchara formations are not striking, except at extreme ends of the outcrop belt. ‘The change from sandstone-poor Huerfano beds to sandstone-rich Cuchara beds is gradual; the change from pink, red and brown Huerfano mudstones to yellow and tan Farisita sandstones and mudstones is abrupt. Fossil localities in Farisita beds are listed with those from the Huerfano beds. Both faunal levels present in Huerfano facies are present in Farisita beds; i.e. late early Eocene Lostcabinian and Gardnerbuttean provincial substages. FAciES CHANGES Geologists studying the rocks of the Raton Basin have given formational names to most of the facies of lower Cenozoic rocks displayed in the area. Recent work by the United States Geological Survey has demonstrated facies relationships between the Raton, Poison Canyon and Huerfano (= Cuchara) formations south of the Spanish Peaks (Johnson and Wood, 1956). In Huerfano Park, the Huerfano formation intertongues with the Farisita formation (Briggs and Goddard, 1956) throughout most, if not all, of the thick- ness of the Huerfano formation. At Black Mountain Farisita beds overlie the 8 FOSSIL MAMMALIA OF THE HUERFANO FORMATION Huerfano beds unconformably; whether this unconformity can be traced within the Farisita facies is not known. Fossils collected from the Huerfano facies in Farisita-Huerfano intertonguing zones (Briggs and Goddard, 1956) and Farisita fossils reported on below indicate general agreement in age of the two facies. Unfortunately, no fossils have been collected from Farisita beds above the un- conformity at Black Mountain. Generally the contact between the Poison Canyon formation and the Huerfano formation is disconformable, but it is gradational at Black Mountain. The basal Huerfano beds in the northern part of Huerfano Park vary laterally from massive, cross-laminated white and pink arkosic sandstones and rufous shales into red, yellow, gray and bluish mudstones. The massive sandstones have been called Cuchara formation by Hills (1889-1891), Burbank and Goddard (1937), Johnson, Wood and Harbour (1958) and Briggs and Goddard (1956). ‘These bright pink and white sandstones are unlike the sandstones found in the type Cuchara, which are yellow to brick red in color. Detailed study shows that lateral changes in Huerfano beds are locally prominent. Near the head of Yellowstone Creek in Sec. 7 and 8, T28S, R69W, two PROVINCIAL | HUERFANO PARK SPANISH PEAKS SUBSTAGE BRIDGER 'B' | GARDNER BUTTE Zz Z LOST CABIN HUERFANO FM. Eras oy LYSITE Y GREYBULL ? 2 POISON CANYON FM. iJ = Lid oO Oo Lj ? PALEOCENE Figure 3. A chart showing the inferred stratigraphic and age relationships of the Huerfano and Farisita formations in Huerfano Park and the Spanish Peaks regions, Huerfano County, Colorado. GEOLOGY 9 striking variations are present. One of these is a 50-foot unit of massive, yellow, cross-laminated, conglomeratic sandstone that extends eastward about | mile and then either grades into mudstones or is covered. Individual beds of sandstone may be 6-8 feet thick. On the opposite (south) side of Yellowstone Creek is a red, sandy mudstone containing much detrital mica. Similarly, local changes in lithology are striking in the Poison Canyon and Black Mountain area. In addition to the facies changes mentioned in connection with the contact of the Poison Canyon formation and Huerfano formations, there are beds of conglomerate containing large boulders. The one unifying feature of the many local changes in Huerfano lithology is their presence at the western limit of the formation, where structural complications are common, and where local irregularities in the Eocene drainage would have been most pronounced. Berner and Briggs (1958) state that the source of Huerfano sediment was south or southwest of the present exposures in Huerfano Park. This interpretation is compatible with the increase in percent of sandstone southward, especially in the Spanish Peaks area. The thickness of Laramide sedimentary rocks is greatest in the Spanish Peaks area (Johnson and Wood, 1956, isopach maps) indicating that subsidence was greatest there and perhaps indicating proximity to source of sediments (Culebra Range of Sangre de Cristo Mountains). TERMINOLOGY OF THE LOWER CENOzoIc Rock UNITs IN HUERFANO PARK Which formation name is valid for these lower Eocene rocks, Huerfano or Cuchara? Hills’ first paper (1888) used two terms for the lower Cenozoic rocks of Huerfano Park: Huerfano beds (p. 148) and Poison Canyon series (p. 152) or Poison Canyon conglomerates (p. 151). These terms applied to all three units described in his generalized section on p. 149. His section is (descending order): “Red and yellow sandy clays and marls, sometimes shaly, alternating with yellowish, white, gray and greenish sands ... 2000 feet. Light-red, white, brownish or variegated soft sandstones alternating with red and yellow sandy clays and marls .. . 1500 feet. Yellowish, or light-colored, soft sandstones and conglomerates with occasional beds of yellow clay or marl. . . 3500 to 4500 feet.” Hills noted the similarity of the two upper units, noting the dominance of sandstones in the middle unit. The thickness of Poison Canyon beds reported is probably due to Hills’ failure to notice the anticline in Poison Canyon and there- fore the repetition in strata, or may be due to his not noticing the change in dip, which decreases westward. Hills later (1889) changed his mind regarding the thick- nesses of the individual units, but the total was similar. Hills revised section is (1889, p. 218): ““Marls, clays, soft shales and sands, of red, gray, yellow, green and purple colors, red predominating........ 3300 feet Pink and white massive sandstones................ 300 feet Soft sandstones and fine conglomerates of a yellowish tint, with occasional bands of yellow clay or marl... .3500 feet.” He noted that the last figure might be in excess. Hills stated that the pink and white sandstones of his “middle division” 10 FOSSIL MAMMALIA OF THE HUERFANO FORMATION formed most of the exposures in the Cuchara Valley (1889, p. 219). No specific terms were applied to these beds. In 1891, Hills decided to apply names to his subdivisions of the lower Cenozoic beds which had been called ‘‘Huerfano beds, Huerfano series, and Poison Canyon series” all three terms being synonymous and purposely general (1891, p. 7). His terminology was (1891, p. 9): Huerfano beds “Huerfano series Cuchara beds Poison Canyon beds” Hills’ larger unit would be a group in present usage. Hills stated that the type localities of the units are as follows: Huerfano beds—along Muddy Creek- Huerfano River divide from Mouth of Poison Canyon into the valley of Muddy Creek (1891, p. 7); Guchara beds—Cuchara River Valley especially beds north and south of La Veta; Poison Canyon beds—Poison Canyon, an arroyo entering Muddy Creek from the southwest. However, Hills’ description of the Cuchara beds was taken from the exposures near the mouth of Poison Canyon, 20-25 miles north of the type locality. At Poison Canyon, beds called Cuchara form massive pink and white arkosic sandstones of limited lateral continuity. In the La Veta area, sandstone beds of this pastel colored lithology are absent; instead dark massive, red, micaceous sandstones are present. The term “Huerfano” beds or formation is well entrenched in the literature (works of Hills, Osborn, Burbank and Goddard) while the term Cuchara, almost as old, was not in common usage until more recently (Burbank and Goddard, Johnson and his co-workers). The type locality of the Huerfano beds is good, having approximately 2800 feet of rocks exposed. At the type locality of the Cuchara formation are 5,000 feet of poorly exposed or contact metamorphosed rock (thickness from Johnson, 1958), assuming that the Spanish Peaks can be included in Hills’ “exposures . . . south of La Veta.” Few exposures of Eocene rocks are present between Yellowstone Creek and La Veta. The existing outcrops are in valley walls; the valley bottoms and the drainage divides are covered by alluvium. Robinson (1960) reported the presence of one fossil of early Eocene age in the “Cuchara” formation and noted that two others had been collected from the La Veta area. Locality VIIJa in Huerfano Park is in beds included in Cuchara formation by Johnson, Wood and Harbour (1958). Lambdotherium at this locality indicates a Lostcabinian age. The 4,000 feet of rock above the known fossil from the La Veta area indicates that the Huerfano formation in that area may include beds younger than lower Eocene. Therefore, Huerfano beds in Huerfano Park and in the La Veta area are of general time equivalence. The two areas have lithologic continuity as well. Huerfano beds overlie Poison Canyon beds in Huerfano Park, in the Spanish Peaks area and between (Johnson, Wood and Harbour, 1958). Thick sections, containing the upper beds of the Huerfano formation, are preserved in Huerfano Park and in the Spanish Peaks; only lower beds are found in the exposures from Gardner Butte southwards to La Veta (that is, in beds usually called Cuchara). One might separate the Huerfano and Cuchara formations on age criteria, considering all the beds of Lostcabinian age (Garcia Canyon local fauna) as Cuchara formation and considering the beds of Gardnerbuttean age as Huerfano formation. However, a formation, being a rock unit, is, or should be, defined on lithologic grounds. The recent geologic map of Raton Basin and Huerfano Park by Johnson, Wood and Harbour (1958) shows the inferred contact between the Cuchara forma- GEOLOGY 11 tion and Huerfano formations in the divide between Archuleta Draw and South Oak Creek. The divide is capped by a pediment, therefore critical exposures are confined to valley sides north and south of the divide. Reasons for considering the two formations synonymous are: 1) The two typical lithologies occur only at extreme ends of the outcrop belt; 2) The two formations are general time equivalents; 3) The two formations occupy the same stratigraphic position (overlie the Poison Canyon formation); 4) Exposures in valley walls between La Veta and Huerfano Park indicate that one unit underlies the alluvium throughout most of the intervening distance. Reasons for preferring the term Huerfano formation are: 1) The type section is well exposed, with exposed lower contact; 2) The lithology of the type section of the Cuchara formation is mostly hidden under cover or altered by contact metamorphism; 3) The term “Huerfano” has more distribution in geologic literature and priority (1888 vs. 1891). Environment of Huerfano Park during the deposition of the Huerfano and Farisita formations: The Huerfano and Farisita formations are interdigitating units of water- deposited sandstone and mudstone. The streams that deposited these bodies of sediment came from highlands in the south and southwest (Huerfano facies) or in the north (Farisita facies) (Berner and Briggs, 1958). Huerfano Park, during the Eocene, must have been a flat plain extending from the Sangre de Cristo Moun- tains eastward; perhaps the Wet Mountains formed its eastern limit. Most of the sediment accumulating as the Huerfano facies came from the Culebra unit of the Sangre de Cristo Mountains. At Yellowstone Creek, Poison Canyon and Black Mountain, the Crestone-Blanca unit contributed sediment. The Farisita facies is derived from a “Front Range,” possibly the Wet Mountains, (Berner and Briggs, 1958). The Farisita streams were either more competent or had less fine material to transport than Huerfano streams did, as the Farisita facies generally has a greater percentage of course material. The abundance of coarse material may also reflect a closer source. Huerfano rocks are generally red, red-brown, brown or pink, possibly indicat- ing a source area with an annual rainfall of some 40 inches and a warm climate in well-drained areas (Dunbar and Rodgers, 1957, p. 210). In contrast to the Huerfano rocks, Farisita rocks are tan, buff or yellow. The significance of the color change caused by a change from ferric oxide to ferric hydroxide is not fully known. In the Washakie and Bighorn basins similar color changes occur in Eocene rocks (Van Houten, 1944, p. 188-191, 195-199). In both examples the yellow facies (Green River and Tatman formations) are lake deposits whereas the red facies (Knight and Willwood formations) are flood plain deposits. Perhaps the color change (therefore mineral change) is due to conditions in the provenance, but perhaps the change reflects depositional environment as well. If we assume that the environments of the source areas of Huerfano and Farisita sediments were the same, and were such that red soils were forming, then the difference between present colors of the Farisita and Huerfano formations is due to the effect of depositional or post-depositional environments upon the red sediments. The preservation of the red color in the Huerfano sediments suggests that the conditions of deposition of the sediments were similar to the conditions be FOSSIL MAMMALIA OF THE HUERFANO FORMATION of formation of them. The predominance of ferric hydroxide (“limonite”’) as a coloring agent of the Farisita facies suggests that the area was less well drained than the Huerfano area, since ferric oxide would probably be preserved in a well drained area (it forms in well drained areas, Dunbar and Rodgers, 1957, p. 210). Drainage in the Farisita formation area was poor enough to inhibit the destruction of much of the organic material by soil bacteria, because plant fragments are preserved in the Farisita formation. If we assume that the differences now shown in the color of the Farisita and Huerfano formations are due to differences in the climate in the respective source areas, then depositional conditions must have been good for preserving the dif- ferences. As in most geologic problems, the answer is probably that both differences in the provenance and in the depositional sites control the present color (therefore iron mineral content) of the beds. Certainly the fact that Berner and Briggs (1958) have been able to differentiate source areas for the Huerfano and Farisita beds by studies of the cross-lamination indicates that a different mineralogy in the source areas could affect the resulting sediments. I think that the Farisita formation accumulated in a wetter environment than the Huerfano formation did, perhaps in perennial or annual swamps. The possible presence of swampy or periodically swampy land to the east of a flood plain gives rise to speculation about the nature of the areas east of Huerfano Park, areas in which no supporting evidence to hypotheses about them is likely to be found. Perhaps the Farisita facies is the record of a near-sea deposit. This may seem like an outrageous suggestion, but marine rocks formed the Cannonball of Paleocene age, and the possibility of occurrence of marine beds in other interior Cenozoic deposits should not be overlooked. If the early Eocene is the record of a period of relative quiet in the Laramide orogeny, it might indicate a time when a readvance of the sea could occur. As stated above, the crucial areas are barren of Eocene rocks; whether this is due to non-deposition as well as erosion is a question not likely to be answered. Faunal evidence of environment is useful in that similar or the same species lived in Huerfano Park and in Wyoming during Eocene time and presumably their ecological requirements were similar. Lepidosteus, and crocodiles, indicate warm climates and permanent rivers. The occurrence of presumably arboreal primates such as Notharctus, Anaptomorphidae and Omomyidae indicates the presence of forest areas. Large ungulates do not prove the presence of savannahs but indicate that open country might have existed. The carnivores probably do not indicate any special environment; however, Didymictis is a large dog-like beast and may have been a forest dweller. Other large carnivores, such as Mesonyx, Oxyaena and Patriofelis, were probably specialized for certain habits, such as scavenging, but we cannot deduce their habitats. Detailed study of the lizards and snakes may prove useful for determining habitats. Most of the faunal elements are found at Locality II indicating a definite thanatocoenose. Fish, crocodiles, ungulates and arboreal primates did not occupy the same habitat in life, but they could have been closely associated. The lack of articulated material (only snakes occur in articulation there) at Locality II adds to the evidence in favor of death association. Of all the mammals found at this locality, Microsyops lundeliusi most probably lived nearby. This is indicated by the abundance of this species and the relatively complete preservation of the remains, which include a skull, several pairs of jaws, and jaws with coronoid and GEOLOGY 13 TABLE 1 List of the genera found in upper and lower faunal zones of the Huerfano formation and their distribution in faunal levels above and below the Huerfano faunas. Wasatchian Bridgerian Lost Huerfano Black’s Genera Lysite Cabin Lower Upper Fork Peratherium x i? x Nyctitherium ? iy x x Talpavus x x Scenopagus ? x x Palaeictops x x x x Cynodontomys x x x x Microsyops x x A patemys ? ? x x Phenacolemur x x Loveina x? x x Absarokius x x x Shoshontus x x ? Notharctus x x x x Stylinodon x x x x Esthonyx x x x Trogosus x x x Metacheiromys x x Mesonyx xX me Oxyaena x x x Patriofelis x x Sinopa x x x x x Viverravus x x ore x x Didymictis Ke x X x Vulpavus x x x x Miacis x x x Uintacyon x x Xe x Oédectes iy x x Coryphodon x x x x Bathyopsis x x mi Hyopsodus x x x x x Phenacodus x x bre Hyracotherium x x x x Lambdotherium x x Eotitanops x x x Palaeosyops x x Helaletes x x Heptodon x x x Hyrachyus x x x x Diacodexts x x x x Bunophorus X xe x x Antiacodon x x Paramys x x x x x Thisbemys x x * Leptotomus x x x x x Reithroparamys x x x x Microparamys x ? x x 14 FOSSIL MAMMALIA OF THE HUERFANO FORMATION TABLES? Mammalian species of genera of mammals common to Lostcabinian (or older) and Gardnerbuttean, or Gardnerbuttean and Blacksforkian (or younger) substages Species Lostcabinian— Gardnerbuttean Blacksforkian+ Talpavus nitidus cf x Nyctitherium velox cf x Microsyops lundeliust x x Notharctus nunienus x x Trogosus grangert x Sinopa strenua a cf Viverravus gracilis x x x Viverravus sicarius x x Vulpavus asius x x Miacis parvivorus * x Uintacyon asodes x cf Bathyopsis fissidens x cf Hyopsodus wortmant x x Hyopsodus walcottianus x Hyopsodus paulus x x Hyrachyus modestus cf x Ke angular processes preserved. No other skeletal parts are currently referable to Microsyops. Costillo Pocket (Locality VIII) is the only location where much articulated ma- terial occurs, and this is mainly Hyracotherium, an ubiquitous early Eocene genus. Perhaps more information about the post-cranial skeletons of many of the mammals found in Huerfano Park would allow inferences about their habitats and therefore the habitats present in Huerfano Park. The data on Eocene floras summarized by Dorf (1955) for other western forma- tions can be applied to the Huerfano formation because the same animals occur there and in Wyoming. Huerfano Park was probably a low moderately flat area of tropical or subtropical climate. FAUNAL LEVELS IN THE HUERFANO AND FARISITA FORMATIONS Lostcabinian faunal level—characterized by the presence of Lambdotherium popoagicum, Eotitanops borealis, Loveina zephyri, Hyopsodus walcottianus, Cynodontomys scottianus, Didymictis altidens and the following genera: Phena- colemur, Esthonyx, Oxyaena, ?Heptodon, Phenacodus. This level is the Huerfano A of Osborn and later authors and shows close correspondence with the Lost- cabinian faunal level of the type area in the Wind River Basin (Van Houten, 1945; White, 1952; Osborn, 1897). The Lostcabinian faunal assemblage in the Huerfano formation is here named the Garcia Canyon local fauna. Garcia Canyon has produced more of the lower faunal zone fossils than any other locality. The Lost Cabin beds of the Wind River formation are divisible into two faunal levels, the upper one characterized by the presence of Eotitanops borealis and Lambdotherium popoagicum, and the lower one by the presence of L. popo- agicum and the absence of Eotitanops (Osborn, 1929). The same two levels are probably present in the Lostcabinian beds of the Huerfano formation. The presence of Trogosus sp. cf. T. latidens in the Cathedral Bluffs Tongue (Gazin, 1952, 1962) may be evidence for a Gardnerbuttean or later age for part of the GEOLOGY 15 Cathedral Bluffs tongue. Absarokius witteri (Morris, 1954) of the Cathedral Bluffs tongue is clearly more specialized than A. noctivagus noctivagus from the Wind River or A. n. nocerai from the upper Huerfano beds. The Didymictis from Cathedral Bluffs is closer to D. altidens than to D. vancleveae. The Hexacodus of the Cathedral Bluffs tongue is evidence for a Lostcabinian age of that locality, because its presumed descendant Antiacodon is present in Gardnerbuttean beds. Morris’ (1954, fig. 1) localities are separated by a total of 25 miles, and their relative stratigraphic position is not indicated in his paper. He does not state from which of these localities the Trogosus was collected but does state the locality of the A bsarokius. Gazin (1962, p. 5, 16) indicates that the Dad local fauna at the top of the Hiawatha member of the Wasatch formation is early Lostcabinian. The Tipton tongue of the Green River formation separates the Hiawatha and Cathedral Bluffs members of the Wasatch formation. Gardnerbuttean faunal level—characterized by the presence of the following species of Lostcabinian or earlier affinities: Absarokius noctivagus (but a distinct subspecies A. n. nocerat), Cynodontomys knightensis, Microsyops lundeliust, Shoshonius cooperi, Notharctus nunienus, Bathyosis fissidens, Hyopsodus wortmant, Paramys excavatus gardneri; and of Blacksforkian or later affinities: Viverravus sicarius, Miacis parvivorus, Hyopsodus paulus, Helaletes sp. cf. H. nanus, Leptoto- mus grandis. ‘The following genera are of Lostcabinian or older affinity: Cory- phodon, Hyracotherium, Eotitanops, Didymictis, Bunophorus, Diacodexis. ‘The following genera are of Blacksforkian or younger affinity: Patriofelis, Oddectes, Mesonyx, Palaeosyops, Antiacodon. As far as known, Huerfanius, new genus, is restricted to the Gardnerbuttean substage. The following species or subspecies are restricted to the Gardnerbuttean substage: Absarokius noctivagus nocerat (new subspecies), Trogosus grangeri, Didymictis vancleveae, Eotitanops minimus, Paramys huerfanensis, Paramys excavatus gardneri, Leptotomus huerfanensis, Antiacodon pygmaeus huerfanensis (new subspecies). This faunal level is the “Huerfano B” of Osborn and later authors. The presence of Hyracotherium, Coryphodon, Cynodontomys and Didymictis espe- cially shows its relationship to the Wasatchian stage. The occurrence of Antt- acodon, Scenopagus, Patriofelis, Mesonyx and Oddectes shows a close relationship with the Blacksforkian faunal level. Osborn (1919, 1929) correlated the upper faunal zone with the Bridger A of Matthew (1909). Palaeosyops fontinalis is the only fossil known to have come from the area of Bridger ‘A’ exposures. Recent work by C. L. Gazin (1962, p. 5) indicates that there may not be any Bridger ‘A’, but instead that Bridger ‘B’ rests on the Green River formation. I propose to refer the assemblage of fossils found in the upper faunal zone of the Huerfano formation to the Gardner Butte local fauna and erect a new substage within the early Eocene, the Gardnerbuttean. The name is taken from Gardner Butte, a prominent local landmark located near the southeastern limit of beds containing the Gardner Butte local fauna. The Gardnerbuttean substage is located between the Lostcabinian substage of the Wasatchian stage and the Blacksforkian substage of the Bridgerian (middle Eocene) stage. H. E. Wood (1934) proposed the term Black’s Fork member of the Bridger formation for the Bridger ‘A’ and ‘B’ of Matthew (1909). The faunal assemblage that defines the Blacksforkian substage comes entirely from Bridger ‘B’ beds. Fossils that have been reported to come from Bridger ‘A’ may come from the Green River formation (Gazin, 1962, p. 5). The lithologic change indicates that perhaps a time lapse is present. The presence of Palaeosyops fontinalis in the 16 FOSSIL MAMMALIA OF THE HUERFANO FORMATION so-called Bridger ‘A’ is scant, but pertinent, evidence for a correlation of the Bridger ‘A’ of authors (perhaps the Laney shale member of the Green River formation, Gazin, 1962, p. 5) with the Gardner Butte local fauna of the Huerfano formation. Morris (1954, p. 199) correlated the Cathedral Bluffs tongue of the Wasatch formation in the Washakie Basin with the Bridger ‘A’ of the Bridger Basin. Gazin (1962, p. 18; 1959, p. 135) states that Morris’ Cathedral Bluffs’ fauna is of Lostcabinian age, but may be somewhat younger than the Lost Cabin beds of the Wind River formation. Discussion. Correlation of beds in different depositional basins is always a matter of opinion based on available evidence. Certain criteria are more valuable than others (Dunbar and Rodgers, 1957), but weighing the value of different units of the same criteria (in this case fossil mammals) is hazardous. I believe that large cursorial mammals are more valuable for correlative purposes than small arboreal or fossorial mammals for two reasons: 1, the large cursors are more likely to spread out and more likely to do it quickly; 2, larger animals are more likely to be pre- served, or at least, are found more easily and are more likely to be collected. Rapidity of evolution has always been a criterion for determination of good taxa of correlative value. Certain of the small mammals might evolve more rapidly but not enough is known about them now to be useful in correlation. I am following criteria of standard usage in placing most reliance upon the Perissodactyla and to a slightly lesser degree upon the Pantodonta, Dinocerata, Artiodactyla and Condy- larthra. From the evidence of distribution of various mammals in the Huerfano forma- tion and from data available in recent literature it is obvious that certain species were present in one area at earlier times than they were in other areas. Microsyops lundeliusi is a good example: in the Wind River formation it is associated with Esthonyx acutidens, Hyracotherium vasacciense venticolum, Hyopsodus wortmani, Heptodon brownorum and Eotitanops sp. (White, 1952, p. 203); in the Gardner Butte local fauna Microsyops lundeliusi is associated with Eotitanops minimus, Palaeosyops fontinalis, Trogosus grangeri, Hyopsodus wortmani, Hyracotherium v. vasacciense. In the Wind River formation Lambdotherium is present at the same levels as M. lwndeliusi even though it is absent at the type locality of M. lundeliust (White, 1952). Several species occur earlier in Huerfano beds than they do in known deposits in Wyoming: Hyopsodus paulus, Antiacodon pygmaeus, Scenopagus edenensis are a few. The presence of two titanotheres, E. minimus and P. fontinalis in the Gardner Butte local fauna which could have evolved from the same upper Lostcabinian species, E. borealis, is evidence for post-Lostcabinian age of the Gardner Butte local fauna. P. fontinalis is the probable ancestor of the Blacksforkian P. paludosus providing good evidence for pre-Blacksforkian age of the Gardner Butte local fauna. The presence of Hyracotherium and Cory- phodon and the absence of Orohippus is good support for a pre-Blacksforkian age for the Gardner Butte local fauna. The occurrence of Trogosus in both the Cathe- dral Bluffs member of the Knight formation and the Gardner Butte local fauna in the Huerfano formation does not necessarily imply time equivalence. If the Trogosus locality in the Cathedral Bluffs tongue is in beds containing Hexacodus as well, it may indicate that Trogosus appeared earlier in the Washakie basin. All the elements of a faunal assemblage do not appear simultaneously. ?PLIOCENE SERIES Devit’s HOLE FORMATION, JOHNSON AND Woop, 1956 The Devil’s Hole formation consists of several hundred feet of water-laid tufts. GEOLOGY 17 and a few andesite flows found on the flanks of the Wet and Sangre de Cristo Mountains. Although this was one of the first units to be recognized in the area (Hills, 1888), it was among the last named (Johnson and Wood, 1956a). The age of the Devil’s hole formation is ?Pliocene. This age determination is based upon horse and camel teeth found by R. C. Hills in the area between Muddy and Turkey Creeks (1889, p. 220-221). Hills did not send the fossils to O. C. Marsh for . z ‘ Clx) ' = \A \ we q Te25)S3 S % : rs @ NS MT ( \ As ~ =-.2 ad ia ine i i ey Hea ES i ae I S Bi r= SS by ‘ : Se | Ss Sy Gardner Gardner Butte = Beer rw T™* sw oie =) 1265 v re is eae ian “ot 1X a) Cre SF ‘ Tees: - 0 Hep Villa : | R. 70 W. R. 69 W. Figure 4. A sketch map of the Gardner area, Huerfano County, Colorado showing the location of the major fossil localities and geographic features. identification so the identification is probably his own (letter to O. C. Marsh, dated 26 October, 1889). I do not know where these specimens are now stored. STRUCTURAL GEOLOGY General Statement. Three features determine the basic structure of Huerfano Park: The Greenhorn (Wet Mountains) Anticline, the La Veta Syncline, and the Sangre de Cristo Complex of folds and faults. Minor folds, faults and intrusions are related to the major structure. In east-west profile, the Colorado Rockies are characterized by an eastern anticlinal mountain range (Front Range, Wet Mountains) occupying the site of the Paleozoic Colorado Mountains (King, 1959, p. 104, 105, 118); a structural and depositional basin west of the anticlinal mountains (North-Middle, South Parks; Wet Mountain Valley-Huerfano Park) and a western mountain range or ranges (Northern Sangre de Cristo, Sawatch, Mosquito, Park) which generally has major thrust fault zones and can have a thick section of late Paleozoic rocks making up much of the range (Sangre de Cristo, Mosquito). The syncline and the western 18 FOSSIL MAMMALIA OF THE HUERFANO FORMATION range are predominantly Laramide structures, the eastern mountain range is a rejuvenated late Paleozoic trend (King, 1959; Gableman, 1956, p. 37). The Greenhorn Anticline. The Greenhorn Anticline (Johnson, Wood and Harbour, 1958) is a southwestern bifurcation of the Wet Mountains Anticline which plunges at Badito. The Apishapa Arch merges with the Greenhorn Anti- cline east of Huerfano Park and the combined structure forms the axis of the Wet Mountains. This latter structure, an en echelon extension of the Front Range, occupies the site of a late Paleozoic positive area, part of the Colorado Mountains (King, 1959, p. 118). The Apishapa Arch extends southeastward and connects with the buried Las Animas and Sierra Grande arches (Gableman, 1956, p. 36), all structures forming part of the remnants of late Paleozoic highlands that connected the Colorado Mountains with the Oklahoma ranges (King, 1959, p. 105). Minor faults occur at the southern end of the Wet Mountains, and an extensive thrust fault is present along the northeastern flank (Gableman, 1956); these struc- tures do not greatly alter the anticlinal nature of the range. The La Veta Syncline. The La Veta Syncline (Johnson, Wood and Harbour, 1958) is the structural axis of the Raton Basin, Huerfano Park and, probably, the Wet Mountain Valley (Johnson and Wood, 1956; Gableman, 1956). The Green- horn Anticline, Apishapa Arch and Sierra Grande Arch border the syncline on the east and the Sangre de Cristo Complex borders it on the west. The axes of the eastern arch system, the La Veta Syncline and the Sangre de Cristo Complex diverge as the structures are traced southward. A small eastern prong of the La Veta Syncline, the Del Carbon Syncline, fits into the fork formed by the Greenhorn Anticline and the Apishapa Arch (Gableman, 1956). Similar inter- digitations of fold axes are found in the Canon City Embayment and at the north- ern end of the Front Range (Blackstone, 1953). Many of the minor structures connected with the Front Range and related ranges trend northwestward (Gable- man, 1956; Boos and Boos, 1957; Blackstone, 1953). Wet Mountain Valley is apparently the uneroded northern continuation of the Raton Basin and Huerfano Park. The axis of the Wet Mountain Valley Syncline ( = La Veta Syncline) extends into Pleasant Valley between the Mosquito and Sawatch Ranges (Gableman, 1956). This syncline complex extends from near Salida south to the latitude of Taos, forming one of the major, and more con- tinuous, structural units of the Rocky Mountain region. The Sangre de Cristo Complex. The Sangre de Cristo Complex is a series of parallel folds, thrust faults and intrusions that extend from near Salida, Chaffee County, Colorado to the vicinity of Las Vegas, New Mexico. This complex is here divided into five units: the Salida unit from Salida south to Orient, Colorado; the Crestone-Blanca unit from Orient south to the La Veta Pass area; the Culebra unit from La Veta Pass south to the Colorado-New Mexico boundary; the Mount Wheeler unit to near Taos and a southern unit composed of several small ranges (Truchas, Pecos, Cimarron, Rincon) from Taos south to Las Vegas. The Salida unit is characterized by thrust faults on the western margin (Litsey, 1958; Gableman, 1952) which relate it to the Mosquito Mountains-Coffman Ridge uplift and not to the southern units of the complex. This unit is probably an en echelon extension of the Mosquito Mountains structure. The Crestone-Blanca unit has been described by Burbank and Goddard (1937) and by Briggs and Goddard (1956). This unit is characterized by the flat lobate Bruff Creek Thrust and related thrusts, by a belt of folds on the eastern flank of GEOLOGY 19 the Sangre de Cristo Mountains and by a series of synaxial instrusives occupying the center of the fold belt. These intrusives form the core of Little Sheep, Sheep, and Rough Mountains and Mount Mestas. The thrust faults involve a thick section of upper Paleozoic rocks (Burbank and Goddard, 1937). The Thrust plate is imbricate. Marginal, high angle reverse faults related to the thrust fault cut the Huerfano and Poison Canyon formations of early Eocene and Paleocene age. The Culebra unit has relatively high angle thrusts with little displacement (Asquith, 1958; G. T. Benson, oral communication). This unit differs from the Crestone-Blanca unit in not having an eastern fold belt and in not having stocks; all the synaxial intrusions are dikes, sills or sheets (Benson, oral communication; Johnson, Wood and Harbour, 1958). The Mount Wheeler unit consists of a Precambrian massif with an eastern margin buried under Cenozoic lava flows; this unit may be a structural saddle (G. T. Benson, oral communication). The southern unit consists of several ranges with thrust faults on the eastern flanks. The western margin of the Sangre de Cristo Complex is a series of normal faults that mark the eastern boundary of the Rio Grande Trench. In most areas a Precambrian core is exposed west of the crest of the Sangre de Cristo Mountains. Recent mapping (University of Michigan parties; G. T. Benson) indicates that much of the crest of the Crestone-Blanca and Culebra units is formed by upper Paleozoic rocks. Minor Structures. The Spanish Peaks and Dike Mountain are two intrusive stocks with radiating dike networks that occupy locations on or near to the axis of the La Veta Syncline. The Spanish Peaks are well known for their dike system and the sequence of intrusions displayed there. Petrology of Dike Mountain has not been reported on yet; it may be a single intrusion and dike network or a compound system like the Spanish Peaks. Dikes from these stocks cut the marginal structures of the Sangre de Cristo Mountains and therefore the instrusives are younger than the folding. The cores of the Little Black Hills and the Black Hills are formed by small stocks located at the southern end of the Greenhorn Anticline (Johnson, Wood and Harbour, 1958) and may be late Laramide in age. GEOMORPHOLOGY Huerfano Park is the product of late Cenozoic erosion. The park contains remnants of several surfaces which produce an illusion of a flat intermountain valley. In reality at least 2,000 feet of rock has been removed from the center of the basin and perhaps more from the flanks. Three major surfaces are present in Huerfano Park. The oldest is the present surface of the Wet Mountain Valley which can be traced southward along the flanks of the Sangre de Cristo and Wet Mountains and which caps Black Mountain (L. I. Briggs, oral communication). The second surface slopes to the south indicating that the streams that flowed over it were south flowing, a reversal of conditions found in the old surface. The third (young) surface is a constructional feature formed by the filling in of valleys cut into the second surface. Recent erosion has cut gullies up to 30 feet deep in parts of the young surface. An understanding of the surfaces is necessary since most exposures of the Huerfano formation are found in valley walls exposed between the middle and the young surfaces, and mapping Cenozoic contacts is a case of finding them in adjacent gully sides. Most of the Cenozoic rock beds are hidden by one or another 20 FOSSIL MAMMALIA OF THE HUERFANO FORMATION of these surfaces and their related deposits. Isolated hills such as Rough Mountain, Sheep and Little Sheep Mountains, and the Black Hills produce local irregularities in the pattern of surfaces. Sub- sidiary surfaces are also present indicating less important episodes in the erosional history. The young surface often merges with the middle surface near the heads of the modern arroyos; the middle surface intersects the old surface in the northern part of Huerfano Park. At Gardner the young surface is approximately 7200 feet above sea level and the middle surface about 100 feet higher. The young surface rises to the northwest and is 76001 feet at the JM Ranch (eight miles up Muddy Creek from Gardner). The middle surface is about 50 feet higher than the young surface at the JM. Traced southward the middle surface decreases in elevation to Badito and then remains more or less level (in north-south profile) until the vicinity of La Veta, where it rises again; the rise is due to the presence of the Spanish Peaks. Between Badito and the Spanish Peaks the surface slopes eastward, reflecting the dominant trend in drainage. Locally, outcrops of resistant rock, such as sandstone beds or igneous dikes, produce minor fluctuations in the general trend. GrEoLocic HisToryY The Laramide orogeny produced the present structural pattern, probably in several stages (Burbank and Goddard, 1937). The early Laramide geology of the Wet Mountains indicates that the area was relatively quiet. It is likely that Lara- mide structures in the Front Ranges are really rejuvenated late Paleozoic trends. Sandstone lenses in the upper Pierre beds (Johnson and Wood, 1956a) in the Raton Basin record the initial movements of the Sangre de Cristo Mountains. The Trinidad sandstone is the deposit of a retreating strand (Johnson and Wood, 1956a), and the Vermejo, Raton and Poison Canyon formations record the dominance of terrestrial conditions. Several unconformities and disconformities in the Laramide sequence testify to periods of renewed erosion probably caused by renewed uplift. Coincident with the Laramide uplift of the Sangre de Cristo Mountains was the Laramide depression of the Raton Basin. Isopach maps of the Trinidad and Vermejo formations (Johnson and Wood, 1956a) show that the axis of greatest sedimentation coincides with the axis of the La Veta Syncline. The thickest section of Laramide rocks is preserved in the Spanish Peaks area where at least 8,000 feet of terrestrial rocks occur. Laramide sediments thin out to the east and west of the axis of the La Veta Syncline and less markedly north and south of the Spanish Peaks. Data accumulated by Berner and Briggs (1958) indicate a southwestern (i.e. Culebra Range) source for most of the Huerfano formation in Huerfano Park, and the increase of sandstone in the Huerfano formation in the Spanish Peaks area agrees with the structural and other stratigraphic conclusions. The Culebra range of the Sangre de Cristo Mountains was probably the major source of Laramide sediments, and therefore possibly the most active of the units of the Sangre de Cristo Mountains during the Laramide orogeny. Unfortunately Laramide sediments derived from the Sangre de Cristo Mountains are not found south of the Raton Basin. The southward thinning of the Laramide sediments within the Raton Basin indicates that the basin probably did not extend much farther southward. 1 Altitudes are taken as near the axis of the basin or the nearest stream as possible in order to give some impression of actual slope. GEOLOGY 21 The Huerfano formation (including the Cuchara formation) is comprised of relatively fine-grained sedimentary rocks and indicates a period of relative quiet. However, coarse conglomerates with boulders several feet in diameter are present in the northwestern part of the Huerfano outcrop area (Poison Canyon and Black Mountain) indicating local streams of great competence. ‘The proximity of these conglomerates to the Bruff Creek Thrust Fault indicates the possibility of the thrust being active during Huerfano time. If this were so, Huerfano deposition ceased with the folding and faulting present in the Black Mountain area; renewed quiet is indicated by the Farisita beds that overlie the Huerfano beds uncon- formably at Black Mountain. On the other hand, the Farisita formation records uplift in the Front Ranges, probably the Wet Mountains, during the same interval. When the movement of the Bruff Creek thrust had subsided, Farisita rocks were deposited over the top of the thrust and related faults and folds, showing that active erosion (therefore uplift?) was still in progress in the Front Ranges. The lack of angular unconformi- ties between the Pierre shale and Poison Canyon formation, and between the Poison Canyon formation and the Huerfano formation in the southeastern part of Huerfano Park shows that pre-Huerfano movements in the Wet Mountains were gentle. However, Farisita sediments overlie the Poison Canyon formation unconformably on ‘Turkey Creek, dating the earliest strong movements of the Wet Mountains orogeny as Eocene, probably late early Eocene. Possibly the Wet Mountains uplift was never strong enough to affect beds in the center of the basin so that deposition continued uninterrupted during part or all of the first major pulse of the orogeny. The increase in amount of sediments of the Farisita facies and their spread over the wesiern part of the basin after late Huerfano time indicates either that sedimentation from the Wet Mountains orogeny was sufficient to mask deposition from the Sangre de Cristo Mountains, that the Sangre de Cristo Mountains were not being actively eroded after late Huerfano time, or that depositional conditions had changed to favor the Farisita type lithology. The synaxial intrusives of the Crestone-Blanca unit of the Sangre de Cristo Mountains are probably contemporaneous with the folding or faulting of the sedimentary rocks. This evidence of repeated uplift suggests that the present struc- ture may have developed as a result of several movements. Most of the synaxial intrusives were probably emplaced in the Eocene. The involvement of Poison Canyon and Huerfano beds in the folding that occurred with the Little Sheep and Sheep Mountains intrusions indicates a post-Huerfano age for the movement. The relationship of these intrusives to the folds suggests that they are closely related in time, if not contemporaneous with folding. The Spanish Peaks and Dike Mountain are located east of the fold belt and almost directly upon the axis of the La Veta Syncline. This sequence of intrusions of the dike system of the Spanish Peaks was established by Hills (1901) and Knopf (1936, 1956). All of the Spanish Peaks and Dike Mountain dikes are post-Huerfano and probably post-tectonic intrusion as well, because the dikes from these intru- sives cut the folded belt. Several of the igneous bodies are located well for critical age determination. Dates close to the boundary between the early and middle Eocene could be obtained from Little Sheep Mountain and from a small dike in the overturned anticline east of the JM Ranch. The post-Farisita history of Huerfano Park is not well known, but several inferences can be made. The Devil’s Hole formation of probably Pliocene age rests unconformably upon Farisita beds, leaving a sedimentary gap for most of the 92 FOSSIL MAMMALIA OF THE HUERFANO FORMATION = middle Cenozoic; pyroclastic rocks of the Devil’s Hole formation record the presence of middle or late Cenozoic vulcanism. The Devil’s Hole formation is capped by the present surface of the Wet Mountain Valley. Remnants of this surface occur on the Sangre de Cristo and Wet Mountains, and on top of Black Mountain. Younger surfaces are south-sloping, cut on Huerfano, Farisita and older rocks, showing a change from north-draining to south-draining streams and indicating that at least 2,000 feet of rock has been eroded from the central part of Huerfano Park since the cessation of Devil’s Hole deposition. PART at PALEONTOLOGY HUERFANO AND FARISITA FORMATION FossiL LOCALITIES Eleven major and several minor localities have produced fossil mammals (see locality map). Two definite faunal levels are preserved in the Huerfano facies: a lower faunal level of Lostcabinian age preserved at localities VI, VIII, VIIa, IX; an upper faunal level constituting the Gardner Butte local fauna at localities I, II, II, V. Both faunal levels occur in the Farisita facies: locality IV represents the Lostcabinian, and localities from one mile east to five miles north of Gardner Butte represent the Gardnerbuttean. Locality VII is in the Huerfano facies and probably in the lower faunal zone. Eotitanops borealis from locality VII is a Lostcabinian species but comes from beds higher than the lowest Lambdotherium specimens in both the Huerfano and Wind River formations. Scanty evidence suggests that the fauna of the Farisita facies may differ slightly from the fauna of the Huerfano facies. Except at locality IV, Farisita fossil vertebrates are rare and poorly preserved. Faunal difference is indicated by: the relative abundance (12 specimens) of Hyopsodus walcottianus at locality IV and its scarcity (6 specimens) at locality VI (a relatively rich locality); the lack of Hyopsodus wortmani at locality IV, though it is common at most of the Huerfano facies localities; the possibility that the Hyracotherium at locality IV is H. vasac- ciense venticolum; and the presence of a small Bunophorus at the locality five miles north of Gardner Butte. The lithologies of the Huerfano and Farisita formations are different, record- ing in part different source areas (Berner and Briggs, 1958). Possibly the two lithologies also record different depositional conditions. If this is true then a difference in the fauna should be expected. The localities have the following relative positions. Huerfano facies: Farisita facies: Youngest I ?X Youngest 5 miles N of Gardner Butte DOs Ds Vi PX 1 mile N of Gardner Butte PX 14 mile E of Gardner Butte VII ?X Oldest IV (2 miles east of Gardner Butte) VI, VIII, VIIa Oldest IX PALEONTOLOGY 93 The exact position of most specimens from locality X is not known, but their posi- tion is probably in the upper faunal level. Locality I. Huerfano Muddy Divide, Sec. 14, 15, 16, T26S, R70W. Most specimens are from below the white bed (unit 20 of measured section). Cynodontomys knightensis Microsyops lundeliusi Absarokius noctivagus nocerat, new subspecies Notharctus nunienus Trogosus grangert Paramys excavatus gardnert Viverravus sicarius Didymictis vancleveae, new species Bathyopsis sp. Hyopsodus wortmani Hyracotherium vasacciense vasacciense Palaeosyops fontinalis Helaletes sp. cf. H. nanus Hyrachyus modestus Locality II. “Fossil Creek,” a small tributary arroyo to the arroyo that runs near the Roman Catholic Church in Gardner; in NW), Sec. 12, T26S, R70W. Most of specimens from unit 9 of measured section. ?Peratherium sp. Cynodontomys knightensis Microsyops lundeliust Apatemys, small species Apatemys, large species Absarokius noctivagus nocerat, new subspecies Huerfanius rutherfurdi, new genus and new species Notharctus nunienus Stylinodon sp. Trogosus grangert Leptotomus huerfanensis Leptotomus grandis Paramys excavatus gardneri Thisbemys nini Paramys copet Paramys huerfanensis Sciuravidae—several species Mesonyx obtusidens Sinopa sp. cf. S. strenua Viverravus gracilis Viverravus sicarius Vulpavus astus Miacis parvivorus Uintacyon sp. cf. U. asodes Oddectes herpestoides Coryphodon sp. Hyopsodus wortmani Hyopsodus paulus Hyracotherium vasacciense vasacciense Palaeosyops fontinalis Eotitanops minimus Helaletes sp. cf. H. nanus 24 FOSSIL MAMMALIA OF THE HUERFANO FORMATION Hyrachyus modestus Bunophorus sp. cf. B. macropternus Antiacodon pygmaeus huerfanensis, new subspecies Locality III. East side of arroyo in Elf Sec. 3, T26S, R70W. The arroyo is called either Milligan’s Arroyo or CCC Draw by local residents. A Mr. Milligan had a house near the mouth of this arroyo in the last decade of the Nineteenth Century; some of R. C. Hills’ specimens were collected near the house. Scenopagus priscus, new combination Scenopagus edenensts Nyctitherium sp. cf. N. velox Cynodontomys knightensis Microsyops lundeliusi Absarokius noctivagus nocerat, new subspecies Notharctus nunienus Trogosus grangeri Paramys copei Paramys huerfanensis Paramys excavatus gardneri Thisbemys nini Leptotomus grandis Sciuravidae Mesonyx obtusidens Patriofelis ulta Viverravus gracilis Oddectes herpestoides Bathyopsis fissidens Hyopsodus wortmani Hyopsodus paulus Hyracotherium vasacciense vasacciense Antiacodon pygmaeus huerfanensis, new subspecies Locality V. A small tributary arroyo entering William’s Creek from the west about 2 miles north of the junction of William’s Creek and the Huerfano River. The locality is in the El% Sec. 1, T26S, R70W. The fossils were collected from a red-to-pink mudstone which may be correlative with unit 9 of the locality II section. Locality V is separated from locality II by a small pediment-capped divide. Microsyops lundeliusi Notharctus nunienus Trogosus grangeri Paramys excavatus gardneri Viverravus gracilis Hyopsodus wortmant Hyracotherium vasacciense vasacciense Palaeosyops fontinalis Hyrachyus modestus Locality VII. An arroyo entering the Huerfano River from the south in Sec. 29, T26S, R69W. The fossil-bearing exposures are in the NEl4 Sec. 31 and the NW), Sec. 32 of the same township. Granger called this arroyo “Apodock Gulch,” but the apparent locality of Apodaca Gulch is Sec. 33 and Sec. 28 of the same township. This latter loca- tion is the next arroyo northwest of Garcia Cafion and is part of Locality VI of this report. This locality is probably high in Lower faunal zone. Loveina zephyn Paramys copet Viverravus gracilis Eotitanops borealis PALEONTOLOGY 25 SCATTERED LOCALITIES IN THE UPPER FAUNAL ZONE Locality of University of Michigan. Center of Sec. 31, T25S, R69W, from Huerfano facies in zone of intertonguing between the Huerfano and Farisita formations. Hyopsodus wortmani Eotitanops minimus Locality SEY, Sec. 26, T25S, R70W. Huerfano facies in zone of intertonguing of Huerfano and Farisita formations. Rodent R. C. Hills’ localities. R. C. Hills collected fossil mammals from several localities which I have not been able to relocate. His localities which have been relocated are included in the American Museum locality numbers. Most of Hills’ Huerfano specimens are in Yale Peabody Museum. The notable exception is the type of Trogosus hillst in the U. S. Na- tional Museum. Locality #5 of R. C. Hills. Upper faunal zone. Trogosus grangeri Palaeosyops fontinalis Helaletes sp. cf. H. nanus Near Hausero’s Ranch or | mile south of Hausero’s. Upper faunal zone. Stylinodon sp. Palaeosyops fontinalis “Box Canyon” (see Osborn 1897). Upper faunal zone. Patriofelis ulta Locality—I1 mile ‘‘a little southeast” of Gardner Butte. Upper faunal zone. Palaeosyops fontinalis This important locality is undoubtedly the lowest upper faunal zone locality in Huerfano Park. The locality is probably in Sec. 20, T26S, R69W, or less than 1 mile west of locality IV in the Farisita facies, and near to AMNH locality ‘4 mile east of Gardner Butte.” Locality IV. A small arroyo north of Colorado Highway 69, across the road from the small Roman Catholic cemetery about two miles east of Gardner Butte. SW14 Sec. 21, NW4 Sec. 28, T26S, R69W. The best collecting is from gray-brown mudstone that crops out in the NE part of the arroyo. The bed lacks concretions and has a “soft” texture. The bone from this locality is broken but often has an almost “recent” appearance and color. This locality is in the Farisita facies. Cynodontomys scottianus Viverravus gracilis Coryphodon sp. (14 mi NW of other specimens, in main part of the arroyo). Hyopsodus walcottianus (common) Hyracotherium vasacciense ?venticolum Lambdotherium popoagicum ?Heptodon sp. Bunophorus macropternus SCATTERED LOCALITIES IN THE FARISITA FACIES 1 mile north of Gardner Butte. Sec. 17 or Sec. 18, T26S, R60W. upper faunal zone (?) Metacheiromys sp. 5 miles north of Gardner Butte, east side of William’s Creek. Probably Sec. 29, T25S, R69W. Bunophorus macropternus Y% mile east of Gardner Butte; NW1{ Sec. 20, T26S, R69W. Possibly near to Hills’ locality “1 mile a little southeast of Gardner Butte.” Probably upper faunal zone. 26 FOSSIL MAMMALIA OF THE HUERFANO FORMATION Hyracotherium vasacciense vasacciense Locality VI. Huerfano facies. Exposures in Garcia Cafion and Apodaca Gulch Sec. 33 and Sec. 34, T26S, R69W and in Sec. 3, 4, 5, T27S, R69W. This is the best locality in the lower faunal zone beds. Most of Wortman’s collection of 1897 came from here and probably from unit 13 of measured section; unit 13 has consistently produced good specimens and matches the description of the best bone-bearing layer given by Wortman to Osborn. Granger and Olsen also made good collections here. ?Diacodon or ?Palaeictops sp. Cynodontomys knightensis Cynodontomys scottianus Phenacolemur jepseni simpsoni, new subspecies Loveina zephyri Notharctus nunienus Stylinodon sp. Esthonyx acutidens Paramys excavatus gardneri Oxyaena sp. cf. O. lupina Viverravus gracilis Didymictis altidens Coryphodon sp. Hyopsodus wortmani Hyopsodus walcottianus Phenacodus vortmant Hyracotherium craspedotum Hyracotherium vasacciense vasacciense Lambdotherium popoagicum Bunophorus macropternus Diacodexis chacensis Locality VIII. Exposures on the northeast side of Oak Creek near the road that runs from Farisita westward on the south bank of Huerfano River. Most of the specimens come from the four excavations known as “Costillo Pocket.” NWI, Sec. 2, T27S, R69W. Talpavus sp. cf. T. nitidus Palaeictops bicuspis Cynodontomys knightensis Esthonyx acutidens Leptotomus costillot Didymictis ?protenus Coryphodon sp. Hyopsodus wortmant Hyracotherium craspedotum Hyracotherium vasacciense vasacciense Locality VIIIa. Exposures at the northwest end of the divide between Oak Creek and South Oak Creek, about 14 mile south of the headquarters of the Silverstine Ranch. NWI4 Sec. 20, T27S, R69W. Cynodontomys knightensis Coryphodon sp. Hyopsodus wortmani Lambdotherium popoagicum Locality IX. Archuleta Draw, the arroyo that runs between Oak and South Oak Creeks. Fossil bearing beds are in Sec. 11, T27S, R69W. The fossils came from the lower part of unit 6 of the measured section. PALEONTOLOGY Di Coryphodon sp. Phenacodus vortmani Hyracotherium craspedotum Hyracotherium vasacciense vasacciense Locality X. Black Mountain or Promontory Bluff. Exposures along the southwest face of Black Mountain in Sec. 25, T25S, R71W and Sec. 30, T25S, R70W. The Huerfano forma- tion at locality X is unconformably overlain by the Farisita formation and grades down- ward into the Poison Canyon formation. The section is complicated by faulting and the lower contact is locally hidden beneath the hanging wall of a fault. The lower beds of the Huerfano formation dip steeply or are overturned; the dips decrease higher in the section (northeastward). Complete sections of the Huerfano formation, though probably present, could not be measured due to the faulting and the lack of continuous exposures. The Huerfano formation may be thicker at locality X than at the type locality in the Poison Canyon-Huerfano River-Muddy Creek area. Only one fossil has been collected here in recent years, a tooth of Didymictis vancleveae whose stratigraphic location is noted in Section 7. Granger and Olsen collected a rodent skeleton, Reithroparamys huerfanensis, high in the exposures on Black Mountain and surely from beds equivalent to the upper faunal zone beds a few miles farther southeast- ward. TABLE 3 Non-mammalian fauna of the Huerfano formation Lower faunal Upper faunal Mollusca zone zone Gastropoda, at least three species of snails x x Chordata Osteichthyes Semionotoidea Lepidosteidae Lepidosteus sp. x me Reptilia* Chelonia oS x Squamata Anguidae x Glyptosaurus hillst Gilmore x ? Peltosaurus sp. Se 0S6n45 i536" VOn44e 852i 3. 4125.53 235208 Width trigonid 11 3.5-4.5 4.02 0.26 6.47 3 4.04.7 4.27 Width talonid 11 <420-5.1) 4.33" 0.25 S77 3 4.0-4.9 4.40 M; Length f eSO5=122 0 26. 169 0F61) 929290 2 .631=6.3) 16920 Width trigonid 7 3:0-4.1 3.73 0.37 9.92 1 4.1 Width talonid Soe Siil=su8ee 13555). 20630) W8245 il Sate Famity OMOMYIDAE (sensu Gazin, 1958) Loveina zephyri Simpson, 1940 Plate II, figure 2 Two specimens, AMNH 17554 from locality VI and AMNH 55219 from locality VII, are referred to Loveina zephyri. The M, of No. 17554 resembles the type, AMNH 32517, closely. The paraconid of P, of the type is located more internally than in the Huerfano specimens. This, however, could be normal variation within the species. Measurements (in Millimeters) of Loveina zephyrt are: 17554 55219 Length P, De 1.9 Width P, eth Length M, Width trigonid M, Width talonid M, Length M2 Width trigonid M2 Width talonid M2 mm RS DOR rR bt Con OAK AO PALEONTOLOGY 33 Shoshonius cooperi (Granger, 1910) Plate III, figure 3 One specimen, AMNH 55153, from locality II, is referable to Shoshonius cooperi. The three lower molars are preserved, and nothing can be inferred about the still unknown premolars of this species. The external cingulum on the molars is more pronounced than in the referred specimen AMNH 14665 (Matthew, 1915c, p. 455). Measurements (in Millimeters) of AMNH 55153 are: Width Width Length trigonid talonid Mi 2.4 1.9 1) M: Z.3 1.6 Daal M; 2.8 yf 1.8 Famity ANAPTOMORPHIDAE (sensu Gazin, 1958) Absarokius noctivagus nocerai new subspecies” Plate II, figure 1 Type. AMNH 55215, a fragment of left jaw with two anterior alveoli and P,-Msg. Another specimen, AMNH 55270, is a fragment of the right jaw with P,-M,; 55270 is as- sociated with AMNH 55215 and may be from the same individual. Hypropicm. Type and AMNH 55217, 55218, 55152, 55154, 55155, 55270 and 55292. Horizon AND Locaity. Upper faunal zone, Huerfano formation. Type and nos. 55270, 55292 from locality II; no. 55218 from locality I; nos. 55217, 55152, 55154, 55155 from locality III. O Huerfano Absarokius O Lost Cabin + Lysite M M2 n S OO O ie (oe ®) O = ae © o O Or : 5 2) G ; Wtal O “15 2.0 25 15 20 25mm. 1.0 5 20 Figure 5. A scatter diagram comparison of widths of trigonids (Wtr) and talonids (Wtal) of lower molars of Absarokius from the Lysite, Lost Cabin and Huerfano beds. Diacnosis. M, averages larger size than M, of A. n. noctivagus. M, larger than M, of A. n. noctivagus. A. noctivagus nocerai is smaller than A. witteri. 2 This subspecies is named in honor of Mr. Joseph Nocera of the American Museum of Natural History who collected the type specimen. 34 FOSSIL MAMMALIA OF THE HUERFANO FORMATION Discussion. A. noctivagus nocerai is distinctly larger than A. noctivagus noctivagus from the Lost Cabin member of the Wind River formation. M, of A. noctivagus nocerat is about the same size as the Lost Cabin subspecies. The M, of A. n. nocerat is generally larger but the size ranges of the two subspecies overlap. M, is always larger in the Huerfano subspecies. A gradual increase in size in time is noted for M, and M, of Absarokius (figure 8, 9). A. witteri (Morris, 1954) from the Cathedral Bluffs tongue of the Wasatch forma- tion is a large form that is probably not closely related to this subspecies. The number of teeth in the jaw of Absarokius has been a matter of doubt. Matthew (1915c, p. 465) decided that there were only two teeth in front of P, and gave 1.1.3.3. as the lower dental formula. Morris (1954, p. 200) states that four or five alveoli are present in front of P, in A. abbotii, AMNH 14673; the alveoli are actually in front of P,. AMNH 55215, the type of A. noctivagus nocerat, has two alveoli in front of P,. P, is single-rooted and P, is double rooted. The type specimens of A. n. noctivagus and A. TABLE 6 Measurements (in Millimeters) of lower teeth of Absarokius noctivagus nocerat new subspecies and A. n. noctivagus A. n. nocerat A. n. noctivagus AMNH no. 55215 55270 "So218 155247 55152 15601 15602 type type Variate P, Length NY — — — — -— _ Width i es — — — — _- —— P; Length 1.9 EC — — 1.4 220 —_ Width 2.0 1.8 — — 1.4 2.0 — P, Length 2.6 2.8 2.6 — 255 2 — Width 2.6 220 255 — 2.5 2.6 2.6 M, Length 23 223 2.4 Ia) — 2.4 2a5 Width trigonid 9 1.8 Del DAO) — Zt 2.0 Width talonid jai 250 2.4 DeD Dea 2d 223 M, Length 243 2. 2:6 Diath 2.4 23 2.4 Width trigonid 2 20 2.4 2.4 Deo 2.0 2:0 Width talonid Ze 2.0 2.4 2-3 Dee 22 222 M; Length 205 — 392 2.8 U5) 2.4 — Width trigonid a7 — OMA Det. 1.9 1.6 — Width talonid 1.6 — — 178 1.8 1.5 — Absarokius E = a O © al © Huerfano = © = O © Lost Cabin = ‘Om noctivagus ’ Lysite Length Mz 220 3.0 mm Figure 6. A scatter diagram comparison of trigonid width (Wtr) and length of M, of Absarokius. PALEONTOLOGY 35 wittert have double-rooted P,’s (Matthew, 1915c, p. 464, fig. 36; Morris, 1954, pl. 21, fig. 2). AMNH 55215 shows that there are eight teeth, but the homologies of the two anterior ones are not certain. The alveoli of the two anterior teeth are of equal size; the first might be an incisor or a canine, the second an incisor, canine or a premolar. This question must remain unanswered until the anterior teeth are found. A. noctivagus is the most abundant of the “‘tarsioid’”’ primates in the Huerfano forma- tion. It is not present in the lower faunal zone, but only four ‘“‘tarsioids’’ have been found in the lower zone, so the absence of A. noctivagus may be due to collecting bias. Tooth measurements are listed in Table 6. HUERFANIUS new genus® Type. Huerfanius rutherfurdi new species, only known species of genus. KNOWN DISTRIBUTION. Upper faunal zone, Huerfano formation. D1acnosis. Lower molar teeth (as far as known) similar in size and morphology to Uintanius; lower premolar teeth (as far as known) similar to but smaller than Absarokius, not compressed antero-posteriorly. Huerfanius ruikerfurdi+ new species Plate III, figure 1 Type. AMNH 55216, fragment of left jaw with P,-M,. Hypopicm. Type only. LOCALITY AND DISTRIBUTION. Locality II, 2 miles north of Gardner, Huerfano County, Colorado. Upper faunal zone, Huerfano formation, uppermost lower Eocene. Diacnosis. As for the genus. Discussion. Huerfanius rutherfurdi is similar to Uintanius ameghini in size and in the morphology of the molars. Its premolars are similar to those of Absarokius. The speci- men shows a stage of premolar development intermediate to the condition in Absarokius and Uintanius; the premolars of Huerfanius lack the antero-posterior compression charac- teristic of Uintanius. This specimen shows that Uintanius might have evolved from Absarokius, passing through a stage similar to or represented by Huerfanius. The P, of Uintanius, in addition to being compressed has a paraconid, while the P, of Absarokius and Huerfanius lacks a paraconid. Huerfanius is easily separable from most other anaptomorphids by its Uintanius-like M.. In Absarokius, Anaptomorphus, Tetonius, and Trogolemur the paraconid of M, is either absent or a small cusp close to the metaconid. Huerfanius and Uintanius, however, have prominent paraconids placed on the midline of M,. Measurements (in Millimeters) of the type specimen are: Length P3 Det Width P; ey Length P, 24 Width P, 2VE Length M, 2.0 Width trigonid M, 15 Width talonid M, 1.6 Length M, 1.8 Width trigonid M. 1.4 Width talonid Mz 1.6 3 Named for the Huerfano River. 4The species is named in honor of Mr. Hugo Rutherfurd of Gardner, Colorado, who gen- erously aided the field parties of the American Museum and Yale. 36 FOSSIL MAMMALIA OF THE HUERFANO FORMATION OrpvER PRIMATES (incertae sedis) or INSECTIVORA Famity APATEMYIDAE Apatemys, large species Two specimens of a large Apatemys, AMNH 17454 and YPM 16444, have been collected from locality II. AMNH 17454 has both rami present, but with the exception of an incisor root, lacks teeth, whereas YPM 16444 has the root of the incisor and poorly preserved P,-M,. The jaws are larger than those of A. bellus and smaller than the jaw of the type of A. rodens (see Table 7). The teeth of YPM 16444 are similar in size to those of A. bellus. Unfortunately the teeth are too poorly preserved to permit accurate comparison with Bridger specimens. AMNH 17454 shows that the number of alveoli per tooth is P;, one; P,, one; M,-M, two each. The single rooted P, indicates that this character does occur in larger apatemyids (see Simpson, 1954, p. 3) and that the presence of two roots on P, may not be significant taxonomically. However, the Huerfano Apatemys are sufficiently different to warrant reservations in any reference to Bridger species. Apatemys, small species One specimen of a small Apatemys, YPM 16477, was collected at locality II. It consists of a fragment of the right jaw with incisor root, alveoli of P,_, (one each) and M,; M, and M, are present but poorly preserved. The jaw is comparable to that of A. bellus in size but the teeth are smaller than those of A. bellulus. The specimen is not of good enough quality to be used as a type; the application of a name awaits better material. Measure- ments are given in Table 7. Huerfano Apatemys are unusual in that the teeth are relatively small. In both species the jaw is relatively massive, indicating that the Huerfano forms are relatively more specialized than A. bellus or A. bellulus of the Bridger. The Huerfano specimens may represent undescribed species, but their inclusion within A patemys is reasonable. Although there is a great variability in the size of Apatemys jaws (see table 7) the variation in the size of the teeth is less. The jaw of a referred specimen of A. bellulus, AMNH 12048 is 4.8 mm deep and that of the type of A. rodens is aimost twice as large, 9.3 mm. Yet the difference in the length of the respective P,’s is 1.1 to 1.5 mm. There is a direct correlation between size of incisor and depth of jaw; a correlation also exists between the depth of jaw and molar tooth size, but the relative increase is less. A. bellulus and A. rodens cannot now be shown to belong to the same species, but I would not be surprised if additional data should indicate that they are conspecific. FaAMILy PHENACOLEMURIDAE Phenacolemur jepseni simpsoni new subspecies Plate III, figure 2 Tyre. AMNH 2680, fragment of a right lower jaw with root of incisor, P,-M, and alveoli of M.. Hypopicm. Type only. DISTRIBUTION AND LOCALITY. Lost Cabin equivalent (lower faunal zone), Garcia Canyon local fauna, of Huerfano formation, Huerfano County, Colorado. The specimen was col- lected by Dr. Wortman in 1897 and the exact locality is not known. Most of Wortman’s fossils from the lower faunal zone came from the Garcia Canyon area (locality VI). Diacnosis. P, significantly smaller than that of type of P. j. jepseni from the strati- graphically much lower Almagre facies of the San Jose formation. The lack of expanded talonid heel readily separates this subspecies from the Almagre form. M, similar to P. 7. jepsent. Discussion. P. jepsenit simpsoni is more slender than P. j. jepseni. The reduced size of the P, further exemplifies a trend for size reduction noted by Simpson (1955, p. 428). Phenacolemur has been recorded from Lost Cabin beds (Simpson, 1955), but the where- abouts of the Wyoming specimen are not known. It is interesting that the Huerfano specimen, which has only recently come to light, was probably the first Phenacolemur ever to be collected. 5 Named for Dr. George Gaylord Simpson who recently monographed the Phenacolemuridae. TABLE 7 Measurements of Apatemys (in Millimeters) A, bellulus Huerfano small sp. YPM* AMNH AMNH YPM 13513 12060 12048 16477 Type I width _— — 1.3e 38) depth — — 2.5e 3.0 P; length — — aaa — width = = — — P, length 1.1le 0.9 iL gal — width — 0.7 0.8 — M, length 1.8 1.8 17, IES) width 12 i2 2 1.0 M: length 1.9 — 1.9 1.7e width G8) — 1.4 1.0e Depth of Jaw at M; —_— — 4.8 5.0 A. bellus YPM* USNM* AMNH AMNH* YPM YPM 13512 13276 11425 12047 13520 11766 Type I width 15 1.6e — — — — depth 3.0e oul — — — — P; length — — —_— —_— _— — width — — — — — — P, length — LenS 145 — _- — width — — 0.7 — — — M;, length — — — 2.0 es — width — — — 1.4 1.4 — M:; length 223 2.4 223 Zed 25 225 width 1.6 ah 1.6 1.4 125 17 Depth of Jaw at Mz 5.6 3 S60) — — — Huerfano A. rodens large sp. YPM AMNH 17454 Y¥PM*, USNM* 16444 Right-Left 12973 13277 Type I width 744) Zt Dep) 2-9 2-9 depth 4.1 3.9 3.8 5.4 6.0e P; length 2.2e — — — 5.0e width tla — — — Deal P, length ie — — — iWees) width 1.0 — — — — M; length De, — — — —_ width 1.4 — — — — M; length — — -- — width — — — — — Depth of Jaw at M; 6.7 120 7.0 9.3 — data from Gazin, 1958. estimate. 38 FOSSIL MAMMALIA OF THE HUERFANO FORMATION TABLE 8 Measurements (in Millimeters) of Phenacolemur jepsent specimens P, Mi |B W L_ W-trigonid W-talonid Depth of Jaw at P. j. jepsent type AMNH no. 48005 DOP lead Z.3 1.6 ay 7.1mm P. j. simpsoni type AMNH no. 2680 DAO Seiler Dell 5 16 5.0 Famity MICROSYOPIDAE McKenna (1960) has re-erected Matthew’s family Microsyopidae to include the genera Cynodontomys, Microsyops and Craseops. Three species of microsyopids are found in the Huerfano formation: Cynodontomys scotttianus and C. knightensis in the Garcia Canyon local fauna and C. knightensis and M. lundeliusi in the Gardner Butte local fauna. The distribution of these species is listed in Table 9. TABLE 9 Numbers of individuals of microsyopid species at Huerfano formation localities Species Localities I Pi Se Ve LY Ne VL Vina exe C. scottianus —- —- — — 2 =? — —_- — C. knightensis 1 1i 8 — — 3 1 1 1 M. lundeliusi 2 32 1 1 —- — — —- = u Cynodontomys from the Huerfano formation oO nD O a ° a Site ee ao) Z CO: = ow vs localities O™, wa i © awe o TT O W tal My, “TL aK) eas) 2.0 Zo 3.0 SS) 4.0mm Figure 7. A scatter diagram comparison of widths of trigonid (Wtr) and talonid (Wtal) of My, of Cynodontomys from several localities in the Huerfano formation. PALEONTOLOGY 39 Cynodontomys scotitanus (Cope, 1881) Plate IV, figure 1 Two specimens, AMNH 17543 and YPM 16468, both from locality IV, are referred to this species, and AMNH 17544 from locality VI may also belong to it. C. scottianus is very similar to Microsyops lundeliusi in size (see Tables 10, 12) and may be ancestral to it. The P, of YPM 16468 is not as molariform as the same tooth in C. knightensis. The Huerfano C. scottianus lies near the upper limit of the range of tooth size for the C. scottianus from the Lost Cabin member of the Wind River formation. C. knightensis is at the lower end of this range. This size difference could be explained if the two were actually competitors. However, the known distribution shows that the two species probably inhabited different areas and, perhaps, environments. TABLE 10 Measurements (in Millimeters) of lower teeth of Cynodontomys scottianus from the Huerfano formation M 1 M 2 W- W- W- W- [cocal=5 E> P; IPA thi tale tri- _tal- Specimen hay eee DY, Eo OW OW. L gonid onid L gonid onid AMNH 17543 left Iv — — —_—- — AA 29 632 AS 83 SS 17543 right WV 24°04 3.1 2:0 —- — —- — — 44 3.1 3.5 YPM 16468 Iv — — — a AD Sy 4:3. .73:3,,,,. 13:5 —_- — — Cynodontomys ?scottianus AMNH 17544 VI — — —_- — — — 44 — — 48 — 4.0 Cynodontomys knightensis Gazin, 1952 Plate IV, figure 2 MATERIAL. Locality I: AMNH 17471; Locality Il: AMNH nos. 17473, 17482, 55161, 55162, 55166, 55167, 55168, 55225, 55286, field no. 1952-342; Locality III: AMNH nos. 17019, 17028, 17022, 17470, 55223, 55163, 55164, field no. 1952-538; Locality VI: AMNH nos. 55221, 55222; Locality VIII: AMNH nos. 55287, 55290, 55291; Locality VIIIa: AMNH 55288; Locality X: AMNH 17018. This species is almost ubiquitous in both zones of the Huerfano formation except at locality IV, where it is seemingly replaced by C. scottianus. At least 25 specimens are known from both faunal zones. The absence of C. knightensis from locality IV and the presence of C. scottianus at that locality is probably of ecologic significance, even though the sample is small. Gazin (1952, p. 20) noted that this species was the same size as C. latidens but morpho- logically like C. scottianus. The Huerfano C. knightensis is similar to C. latidens in size, but the mean of the Huerfano sample is slightly greater. The sample of C. knightensis from locality II is composed of specimens whose mean size is slightly smaller than those from locality III. This is shown in the scatter-diagram, figure 7. This may be significant since Microsyops lundeliusi is so common at locality II, where similar sized individuals of the two species might have been competitors. However, the differences shown in figure 7 may be due to depositional agencies or to sampling error. FOSSIL MAMMALIA OF THE HUERFANO FORMATION TABLE 1 Statistics (in Millimeters) of lower teeth of Cynodontomys knightensis Variate Locality N OR M Length Ill 1 2.0 3 xX 1 Sei Width III 1 1.3 Ps; xe 1 1.9 Length II 4 3.2-3.9 Se02 Py III 4 3.4-3.8 3.60 x 1 3.9 Width II 4 2.0-2.7 2.43 Py III 4 2.42.7 295 xX 1 2.6 Length II 6 3.2-3.6 3.45 Mi III 3 3.7-3.8 SE 1/7 VI 1 3.8 VII 2 3.2-3.6 3.40 Width trigonid II 6 2.2-2.5 2.28 Mi III 3 2.3-2.6 DAG VI 1 225 VIII Z D2 Width talonid iD 6 2.42.8 Z.00 Mi ITI 3 2.6-3.0 2.80 VI 1 2.8 VIII 2 2.3-2.4 238 Length I 1 3.6 Me II 5 3.5-3.8 3.62 III 4 3.9-4.1 4.03 VI 2 3.4-3.9 305 Width trigonid I 1 2.4 M, II 5 2.3-2.6 2.46 III 4 2.7-2.9 2.87 VI Z 2.3-2.6 2.45 Width talonid I 1 2.9 M2 II 5 2.6-3.1 2.88 III 4 3.0-3.1 3105 Length I 1 4.2 M; II 3 3.8-4.2 4.10 III 2 4.44.5 4.45 VI 1 4.3 Width trigonid I 1 D2 M; II 3 2.2-2.4 BOX | III 3 2.3-2.6 2.50 Width talonid I 1 2.8 M; II 3 2.3-2.7 253 III 3 2.5-2.8 2.67 VI 1 pie PALEONTOLOGY 41 The greater relative abundance of C. knightensis in the upper faunal zone is probably due to the fact that the upper zone is generally more fossilferous than the lower zone. C. knightensis is very similar to M. elegans from the lower Bridger beds, and probably is ancestral to the latter species. M. elegans has a more molariform P, than C. knightensis does and is slightly larger. Cynodontomys OHuerfanolZ © Huer.L, IIL O Huer WT o Lost Cabin *Lysite “ a5 3,0 55) 4.0mm Figure 8. A scatter diagram comparison of widths of trigonid and talonid of M, of Cynodon- tomys from the Lysite, Lost Cabin and Huerfano beds. Microsyops lundeliusi (White, 1952), New Combination Plate IV, figure 6 MATERIAL. Locality I: AMNH 55214, YPM 16485; Locality II: AMNH nos. 17463, 17464, 17465, 17466, 17469, 17493, 55169, 55170, 55171, 55172, 55173, 55174, 55175, 55176, 55177, 55178, 55179, 55208, 55212, 55213, 55283, 55285; YPM nos. 14614, 16452, 16453, 16454, 16456, 16482, 16484; Locality III: AMNH 17468; Locality V: AMNH 17467. Microsyops lundeliusi is represented by at least 36 specimens from localities I, II, H1 and V. All but four of these specimens are from locality II. The material includes a skull, several partial upper dentitions and many lower jaws. The skull, AMNH 55284, is being described elsewhere by McKenna. White (1952, p. 191) placed this species in Cynodontomys; the type specimen, USNM 18371, lacks the fourth lower premolar. This tooth is preserved in fourteen of the Huer- fano specimens and is more like that of Microsyops elegans than that of Cynodontomys scottianus. In C. scottianus the metaconid of P, is rounded and low. In C. knightensis the metaconid is similar to that of M. elegans, i.e. well developed, but the paraconid is still present. M. lundeliusi P, has a large metaconid, well separated from the protoconid, mak- ing the trigonid as wide as the talonid. The paraconid is lacking. Summations (Table 12) of the widths of P, and P, suggest that sexual dimorphism was present in M. lundeliusi. The separation of the specimens into two groups of equal size is best explained by the presence of males and females. Presumably the males are larger (they are usually so in primates, and Microsyops is here considered a primate). ‘The separa- tion disappears in histograms of widths of the molars. Measurements have been listed for lower teeth only because upper teeth are rare. The relative abundance of M. lundeliusi at locality II is interesting. Locality II is a rich locality; therefore it would be likely to have a plurality of the specimens of any species of upper Huerfano mammals. The relative lack of Cynodontomys (found in approxi- mately equal numbers at localities II and III) and the small size (relative to other C. knightensis) of Cynodontomys at locality II indicate that the two species may have been 42 FOSSIL MAMMALIA OF THE HUERFANO FORMATION Micros yops OBridger CD Ww g al © Bridger B * Huerfano Tibi o) (aN 225 3.0 S5 4.0 4.5 5.0mm Figure 9. A scatter diagram comparison of widths of trigonid (Wtr) and talonid (Wta!) of M, of Microsyops from the Huerfano and Bridger formations. competitors. The possibility exists, but is unsupportable at present, that Microsyops lunde- liusi might have been semiaquatic or a swamp-dweller. The fossil-bearing exposures at locality II are all within 300 yards of a channel sandstone that is at the same level. The abundance of M. lundeliusi may be related to the proximity of a flowing stream. The size of M. lundeliusi (see figure 12) is its most distinguishing character. That the earliest known Microsyops should be so large is startling. M. lwndeliusi is larger than most of the M. annectans from the upper Bridger beds. M. lundeliusi is probably not the TABLE 12 Measurements (in Millimeters) and statistics of lower teeth of Microsyops lundeliust Variate N OR M SD V P; Length 9 3.5-4.6 4.14 0.38 9.18 Width ?male 5 3.3-3.6 3.44 0.13 Seis Width ?female 4 DO Ded 2.63 0.10 3.80 Py Length 13 5.0-6.5 5.67 0.51 8.99 Width ?male 6 4,3-4.5 4.43 0.20 4.51 Width ?female 7 3.6-3.9 Seip! 0.27 7.16 Mi Length 17 4.9-6.2 5053 0.36 6.51 Width trigonid 18 3.4-4.0 Seis 0.21 5.63 Width talonid 16 3.8-4.5 4.19 0225 5.96 M2 Length 18 5.2-6.3 Seal 7 0.27 4.67 Width trigonid 18 3.5-4.2 4.01 0.24 5.99 Width talonid 19 4.1-4.7 4.53 0.17 3315 M; Length 9 5.8-7.9 6.78 0.58 8.63 Width trigonid 9 3.3-4.2 3.76 0742) 10801 Width talonid 9 3.8-4.5 4.19 0.31 7.40 PALEONTOLOGY 43 ancestor of either M. elegans or M. annectans. C. knightensis, especially the Huerfano sample, would seem to be a proper morphological precursor of the Bridger forms. Perhaps M. lundeliusi arose from C. scottianus. Then the genus Microsyops (as here defined) would be polyphyletic in origin. Perhaps M. lundeliusi should be placed in a separate genus, but I cannot recognize any morphological reason for doing this. Placing Cynodontomys in synonymy with Microsyops would solve the semantic problem but not alter the inter- pretation. M. lundeliusi may be the ancestor of Craseops sylvestris, but the latter might have arisen from M. annectans as well. The upper molars of Craseops are more specialized than those of Microsyops (Stock, 1934). OrpvEeR TILLODONTIA Famity ESTHONYCHIDAE SUBFAMILY EsTHONYCHINAE Esthonyx acutidens (Cope, 1881) Plate V, figure 8 Five individuals of Esthonyx acutidens have been collected from the lower faunal zone of the Huerfano formation. Four of the specimens are very poorly preserved but one, AMNH 17531 has the left P,-M, and right P,-M,. No. 17531 came from locality VI: the remaining specimens are not catalogued. SUBFAMILY TROGOSINAE Trogosus grangert Gazin, 1953 Plate VIII, figure 2 PROBABLE SyNonyM: T. hillst Gazin, 1953 MatTERIAL. Locality I: AMNH 17008, type of T. grangert; AMNH 17009; Locality II: AMNH 17495, YPM 16449; No locality data: USNM 17157 (Type of T. hills?). Trogosus grangeri is probably the commonest mammal in the upper faunal zone of the Huerfano formation; it is generally poorly preserved. Fragments of Trogosus incisors and cheek teeth are ubiquitous and are the most useful index fossils for the upper zone because of their large size and distinctive appearance. Gazin’s (1953) revision of the Tillodontia placed two species of Trogosus in the upper Huerfano faunal zone. T. grangeri and T. hillsi are based on skulls and lower jaws with poorly preserved teeth. Gazin (1953, p. 47) noted that the two specimens might be sexual dimorphs of one species. The two individuals might also show age differences. ‘The basicranial differences noted by Gazin might be those of sexual dimorphism. The upper tooth rows of the two types (AMNH 17008, T. grangeri, and USNM 17157, T. hillst) are of similar size: AMNH 17008 P*-M®-85.0 mm, USNM 17517 P®-M*-83.8 mm (Gazin 1953, p. 46, 48). The Huerfano Trogosus show considerable variation in individual size. The length of the lower tooth series (P,-M; at the alveoli) varies from 89.9 mm (AMNH 17008) to 122.2 mm (YPM 16449). The coefficient of variability is large for all variates for which it has been calculated; the variation is from 13.87-17.28. The largest sample is composed of four specimens. In reference to the variation, the difference in size between the same teeth in opposite jaws of the same individual is great. YPM 16449 has as much as a 12 per cent difference in one variate and over five per cent difference in three others. If the larger specimens of the Huerfano Trogosus represent a distinct species, it cannot be either T. grangeri or T. hillsi since these species are described from specimens at the lower end of the size range. The largest specimen collected from the Huerfano formation, YPM 16449, has the peculiar crest between the entoconid and hypoconid of M, mentioned by Gazin (1953, p. 44). This interesting morphological character found in all unworn M,’s from the Huerfano formation, indicates a relationship between all the Huerfano Trogosus possess- ing it and adds to the evidence in favor of a single, variable population. Trogosus grangeri has been collected at localities I, II, III and V. 44 FOSSIL MAMMALIA OF THE HUERFANO FORMATION TABLE. 13 Measurements (in Millimeters) of lower teeth of Trogosus grangert Statistic AMNH nos. 17009 17008 17495 YPM 16449 V left right P; 1b, —= -- —- — 13.4 W —_ — 9.4 P; Ie ilsied/ 14.2 15.9 17.8 18.6 14.16 W 12.9 14.2 14.3 ibe 7 16.4 13.87 M, ILL 16.8 15/50 20.1 — 18.8 W-trigonid 1329 1529 — — 19.5 W-talonid 13.0 — — 18.1 16.7 M: L 20.0 18.2 22.0 22eS W-trigonid 1507 15.8 17.8 20.6 19.3 W-talonid 1552 15.9 eS 19.4 18.9 Monk Sil if 28.9 38.8 38.9 39.1 W-trigonid 1578 15.6 —- Zea 18.5 W-talonid 14.1 14.9 152 20.1 17.6 W-hypoconulid 10.7 12.9 116 12.3 1253 P;-M; Lat alveoli -= 89.9 —- os Ae oD 17.14 M.-M; Lat alveoli 70.0 65.0 88.4 84.4 85.7 14.85 Note: Note the variation between right and left sides of the same individual, YPM no. 16449. OrverR TAENIODONTA Famity STYLINODONTIDAE Stylinodon sp. Plate V, figure 1 MATERIAL. Locality II: AMNH 17451; Locality VI: AMNH 17525; Locality “South of Hausero’s Ranch” (Collected by R. C. Hills): YPM 14616. Three isolated Stylinodon teeth have been collected from the Huerfano formation. The genus being represented in both faunal zones. AMNH 17451 is from a smailer animal than the other specimens. OrpvER EDENTATA SUBORDER Palaeancdonta Famity METACHEIROMYIDAE Metacheiromys sp. Plate I, figures 4, 5 AMNH 18666, an astragalus, from 1 mile north of Gardner Butte is referable to Metacheiromys. The specimen is larger than YPM 13501 referred to M. tatusza by Simpson (1931, p. 305, 356) and may be about the size of M. marsht. A row of small pits is present under the head of the astragalus; the pits follow the rim of the condyle from one side to the other. Comparative measurements are: Bridger Huerfano M. tatusia Metacheiromys sp. YPM 13501 AMNH 18666 Width trochlea 7.0 mm 10.0 Width head Aa] 6.4 PALEONTOLOGY 45 OrDER RODENTIA Famity PARAMYIDAE A. E. Wood has recently (1962) monographed the paramyid rodents including those of the Huerfano formation. As Wood did not indicate which locality most of the specimens came from I do so below. SUBFAMILY PARAMYINAE Paramys copei copei Loomis, 1907 MaTERIAL. Locality II: AMNH nos. 55118, 55124, 55127, 55128, 55134, 55195, 55198; Locality V: AMNH 17455; Locality VI (?): AMNH 2682; Locality VII: AMNH 17546. It is not clear from Wood’s statistical treatment (1962, p. 45, table 8) whether he has included all the specimens from various localities in his chart. If he has, and the sample size indicates so, then the statistical treatment is open to question. In that case samples from widely disparate localities (Wind River, Huerfano and San Jose formations among others) are handled as if they were part of one population. Paramys excavatus gardneri Wood, 1962 MatTEerIAL. Locality I: AMNH nos. 17024, 55146; Locality II: 55120, 55129, 55130, 55133, 55137, 55150, 55193 (type); Locality III: Nos. 17456, 17458, 17459, 55141, 55143, 55196, 55199; Locality V: 17455, 55138, 55139; Locality VI: 17545. Paramys huerfanensis Wood, 1962 MatTERIAL. Locality II: AMNH nos. 55114 (type), 55115; Locality III: 17025. Leptotomus costillot Wood, 1962 MateriAL. Locality VIII: AMNH nos. 55110, 55111 (type), 55113. Leptotomus grandis Wood, 1962 MatERIAL. Locality II: AMNH nos. 17453, 55126, 55192; Locality III: 17452, 17457; Locality 5 of Hills: USNM 20137 (type, misprinted 10137 in Wood, 1962, p. 80); Locality C of Hills: USNM 20135. Leptotomus huerfanensis Wood, 1962 MateRIAL. Locality II: AMNH nos. 55121, 55125 (type), 55149. Leptotomus parvus Wood, 1959 MATERIAL. Locality HI: AMNH 55951. Thisbemys nini Wood, 1962 MatTERIAL. Locality II: AMNH nos. 17026, 55132; Locality III: 17023, 17454, 55937, 55197. SUBFAMILY REITHROPARAMYINAE Reithroparamys huerfanensis Wood, 1962 MATERIAL. Locality X: AMNH 17031 (type). SUBFAMILY MICROPARAMYINAE Microparamys sp. B MATERIAL. Locality Il: AMNH 55200. Wood used the accepted designation of early Bridgerian for the upper faunal zone of the Huerfano formation. However, the paramyid rodents offer some of the best evidence for considering the upper faunal zone late Wasatchian. The presence of Paramys copet copei, Paramys excavatus and Thisbemys nini all indicate Wasatchian age. Wood (1962, p. 80) was uncertain whether Leptotomus grandis was present in the Bridger formation. Only 46 FOSSIL MAMMALIA OF THE HUERFANO FORMATION Leptotomus parvus is definitely present in the Huerfano and Bridger formations (Wood, 1962, p. 85). Famity SCIURAVIDAE Several Sciuravid rodents are present in the Huerfano formation; most of them come from the upper faunal zone. The material is currently being studied by Mary R. Dawson. TABLE 14 Distribution of the rodents in the Huerfano formation Species Locality iy alg ith Mig VItVV, Vio Vale TX: Paramys copet copet ees ee ee ek Me ee e. gardneri D> STL ETA MOBY et SS OH tl petit huerfanensis Actes; AED EE) Pee) Ae Bary Bee) eee es Thisbemys nint S55 Pi Doman We Sete at poke 23 3) ately Leptotomus costillot Soe Se eS huerfanensis = ee grandis ee et parvus Se eee i ee ee Reithroparamys huerfanensis a pee os (See ee eee Microparamys sp. B i ee ss ee ee OrDER CARNIVORA Famity MESONYCHIDAE Mesonyx obtusidens (Cope, 1872) Plate VI, figures 3, 4 Two specimens from the upper faunal zone of the Huerfano formation are referable to Mesonyx obtusidens. AMNH 17423, from locality III, is a jaw fragment with the canine and P.-M, preserved; the cheek teeth are badly chipped. The length, C-M,, of AMNH 17423 is 122 mm, only 5 mm smaller than that given by Matthew (1909, p. 489) for the species: 127 mm. The second Huerfano specimen, AMNH 55272, from locality II, has P*-M? preserved. AMNH 55272 is from a younger individual than the specimen figured by Matthew (1909, p-. 494, figure 94) and has a small metacone on P*. In Matthew’s specimen, AMNH 12643, this cusp has been obliterated by wear. The parastyle and metastyle of M’-M* of AMNH 55272 are less developed than that of AMNH 12643; therefore the upper molars are shorter in the Huerfano specimen. Measurements (in Millimeters) of AMNH 55272 are: Length M! 17.0 Width M1 15.9 Length M2 13.8 Width M2 15.4 FAMILY OXYAENIDAE Oxyaena sp. cf. O. lupina Cope, 1874 Plate V, figures 6, 7 MATERIAL. Locality VI: AMNH nos. 2683, 17557, 55265, 55298. Four fragmentary specimens of Oxyaena have been collected from locality VI. Osborn (1897, p. 255) described O. huerfanensis, basing the species on AMNH 2683. Denison (1938, p. 169) placed O. huerfanensis in synonymy with O. lupina, the type species of the PALEONTOLOGY 47 genus, and referred AMNH 17557 to O. forcipata. As Denison noted, the Huerfano Oxyaena are larger than typical San Jose individuals of O. lupina. The four Huerfano specimens seem to belong in a single population whose individuals were intermediate in size between O. lupina and O. forcipata of the San Jose formation. The Huerfano sample of Oxyaena is younger than the San Jose sample. Although the specimens are either too poorly preserved or too fragmentary for adequate determination, I think the Huerfano sample is closer to O. lupina. Denison’s chart of the progressive development of M, is best when used with a larger sample. However, the statistics trigonid width/trigonid length of M.,, talonid length/length of M.,, and shear angle of M., are nearly the same for O. lupina and O. forcipata. Patriofelis sp. cf. P. ulta Leidy, 1870 Plate VI, figure 2 Synonyms: P. coloradoensis Matthew, 1909, P. compressa Denison, 1938 MATERIAL. Locality III: AMNH 17017 (Type of P. compressa); “Box Canyon,” a locality of R. C. Hills mentioned also by Osborn (1897, p. 251) but not located on any map. YPM 16461; “Upper beds” AMNH 2691 (Type of P. coloradoensis). Three specimens of Patriofelis sp. cf. P. ulta have been collected from the upper faunal zone of the Huerfano formation. All of these specimens are lower jaws; two have been made the types of species: Patriofelis coloradoensis Matthew, 1909 and P. compressa Deni- son, 1938. The type specimen of P. coloradoensis, AMNH 2691, has badly worn teeth but the form of the jaw is Patriofelis-like (Denison, 1938, p. 173). The type specimen of P. compressa, AMNH 17017, is 11 per cent larger than AMNH 2691 and has slightly narrower premolars (Denison, 1938, p. 173). The third specimen, YPM 16461, lacks teeth, but the dentary is almost whole and resembles AMNH 17017 closely. The type of P. ulta is a toothless jaw fragment. Enough of the jaw is preserved to show the distinctive masseteric and temporal fossae that are characteristic of Patriofelis. P. ulta (plastotype YPM 10070) differs from the Huerfano sample mainly in the thickness of the jaw; P. ulta is slightly longer. If the Huerfano population is distinct it is likely to be of only subspecific rank and Matthew’s 1909 name, P. coloradoensis, has priority. ?Patriofelis AMNH 17427, fragments of the left maxilla, both zygomae and the sagittal crest from locality I, is questionably referred to Patriofelis. The infraorbital foramen is unusually close to the alveolus of the canine, Part of the apparent abnormal closeness in this speci- men may be due to fracture of the canine alveolus. AMNH 17427 is the size of P. ulta. Famity HYAENODONTIDAE Sinopa sp. cf. S. strenua (Cope, 1875) Plate VI, figures 1, 5 MATERIAL. Locality II: AMNH nos. 17421, 17422, 17431, 55211, 56501, 56506, 56510. Eight specimens of Sinopa have been collected from locality II. The material is poorly preserved and fragmentary. The sample from locality II resembles Tritemnodon in the closeness of the paracone and metacone of M** but differs from it in the possession of lingual cingula on the upper molars and in the slender proportions of the dentary. Sinopa strenua is recorded from the Grey Bull and Lost Cabin faunas (Matthew, 1915a, p- 76); its possible presence in the upper faunal zone of the Huerfano formation is of little importance but does emphasize the lower Eocene affinities of that zone. Famity MIACIDAE SUBFAMILY VIVERRAVINAE Didymictis sp. cf. D. protenus (Cope, 1874) Plate VII, figures 1, 7 One specimen of Didymictis, YPM 16458 from locality VIII is referable, with doubt, to D. protenus. YPM 16458 has parts of both jaws with right M,., and left M, preserved. It is 48 FOSSIL MAMMALIA OF THE HUERFANO FORMATION smaller than specimens of D. altidens from the Lost Cabin (see Table 15) or Lysite substages (Kelley and Wood, 1954, p. 349) and is close to the size of D. protenus curtidens from the Grey Bull beds (Simpson, 1937, p. 16). Pertinent measurements are listed in Table 15. Didymictis altidens Cope, 1880 Plate VII, figures 3, 5 AMNH 2677, and M,., and other tooth fragments, from locality VI is referable to D. altidens. Several other tooth fragments in the American Museum collection are prob- ably referable to this species. Measurements are listed in Table 15. TABLE 15 Measurements (in Millimeters) of lower teeth of Didymictis from the Huerfano formation and the Lost Cabin beds of the Wind River formation. D. ?protenus D. altidens D. vancleveae YPM AMNH AMNH 16458 4792 4793 4911 4794 2677 17424 56507 P2 Length — type — — — — 10.9 — Width — — — — — — 4.7 — P; Length — —_ — — — — 12.9 — Width — — — — — — 62 — P, Length -— — — — — — 14.4 — Width — — — — —_ — 7:3 — Mi Length aS — — — 15.6 13.0 18.0e 16.9 Width trigonid 6.8 — 10.1 — OFS" © #326 10.7 LS) Width talonid 59 — — — TSP O27 8.0 Tlie M2 Length 7.4 S45 B29 929° = 8.6 10.8 — Width trigonid 4.7 5s On PLONO? "OL ON ei 55.6) 6.8 — Width talonid 4.3 AOE. SOG O85) ae 4.7 6.1 — Depth of jaw at M2 17.4 20.5 — —_— — — 3120 — Didymictis vancleveae new species® Plate VII, figures 2, 4, 6 Typr. AMNH 17424, a fragment of the left jaw P.-M,; M, damaged and lacking paraconid. Hypopicm. Type and AMNH 17030, a skull with the muzzle missing; AMNH 17426, tooth fragments and left M**, AMNH 56507 and M,. Horizon AND Locatity. Upper faunal zone, Huerfano formation, South-central Colo- rado. Specimens from localities I and X. Dracnosis. Larger than D. altidens, one lingual cusp on trigonid of M,, jaw deep and stout. Discussion. Didymictis vancleveae is most easily distinguished from D. altidens by its larger size. The metacone of M’ seems to be less reduced in referred specimens of D. vancleveae, AMNH nos. 17030, 17426, than in a referred specimen of D. alttdens, AMNH 14750, from the Wind River formation. M’ resembles the same tooth in D. protenus. The single lingual cusp on the trigonid of M, may be a variable character. Only the type specimen of D. vancleveae has this tooth preserved. One specimen from the Wind River formation here referred to D. altidens, AMNH 4911, is somewhat larger than most D. altidens material (Table 15) and has the paraconid and metaconid quite close together, and the anterior cingulum of AMNH 4911 is broader than normal, more nearly like 6 Named in honor of Mrs. Eleanor van Cleve of Walsenburg, Colorado, who gave very generous support to the field work related to this study. PALEONTOLOGY 49 D. vancleveae. The single lingual cusp of the trigonid of AMNH 17030 is directly lingual to the protoconid. Perhaps the paraconid fused with the metaconid, AMNH 4911 repre- senting an intermediate structural stage. The location of the lingual trigonid cusp in D. vancleveae is intermediate to the positions of the metaconid and paraconid in D. altidens. Didymictis vancleveae is, perhaps, the youngest known species of Didymictis. Un- fortunately Didymictis is rare in the Huerfano formation, but the meager sample available indicates that the members of this genus increased in size as they evolved. The skull of D. vancleveae must have been at least 190 mm long (estimated length, occipital condyle to premaxilla) compared with about 155 mm for D. altidens (estimate). Giles MacIntyre is revising the Miacidae and the description of this skull is left to him. Viverravus gracilis Marsh, 1872 Plate V, figure 3 ProBABLE SyNonyM: V. dawkinsianus Cope, 1881 MATERIAL. Locality II: AMNH 17429; Locality III: AMNH 56505; Locality IV: YPM 16467; Locality VI: AMNH nos. 2678, 2679, 2681; Locality VII: AMNH 55210. V. gracilis occurs in both upper and lower faunal zones of the Huerfano formation. Wortman (1901, p. 145) included the Lost Cabin species V. dawkinsianus in V. gracilis; Matthew (1915a) regarded the two species as doubtfuily distinct and stated that the Lost Cabin species had smaller size. Tables 16 and 17 show that the Lost Cabin sample exceeded TABLE 16 Measurements (in Millimeters) of lower teeth of type specimens of Viverravus gracilis and V. dawkinsianus Mi, M2 Py W- W- W- W- Horizon L W L_ trigonid talonid L trigonid talonid Bridger B (YPM 11836) type: V. gracilis 525 2.0 B23) Peous 2.8 40 — 2.0 Lost Cabin (AMNH 4788) type: V. dawkinsianus 5.1 1.9 Seow GOS pat Ab Die 2A: 2.0 the Bridger sample in the mean size and also generally the maximum size of the dental dimensions tabulated. Moreover, the type of V. dawkinsianus exceeds the type of V. gracilis (Table 16) in the length of its lower molars. Table 17 shows the complete overlap of the Wind River sample by the more varied Bridger sample for molar widths, despite the larger average size of the older forms. It also shows that the Huerfano sample coincides with the Lost Cabin and Bridger B samples for the same parameters. However, the Huerfano sample has individuals with a shorter P, and may represent a distinct subspecies. No change is noted in the Huerfano specimens in time, most probably due to the few available specimens. Viverravus is variable enough so that erection of a new subspecies should await more material. In the course of determining the species of Viverravus in the Huerfano formation, I inspected all the Viverravus material in both the American and Peabody Museums. The Huerfano specimens show that the hypocone of Mt’ is well developed in V. gracilis, where it forms a broad cingular shelf. The same cusp is missing in V. acutus (Matthew, 1915a, p. 28, fig. 22) and the posterior lingual cingulum is absent. A heavy anterior lingual cingulum occurs on M’ of V. gracilis, AMNH 56505, but the protocone separates it from the posterior cingulum. The protoconal-parastylar distance is relatively greater in V. gracilis than in V. sicarius. 50 FOSSIL MAMMALIA OF THE HUERFANO FORMATION Viverravus sicarius Matthew, 1909 Piate V, figure 12 MATERIAL. Locality I: AMNH 56511; Locality II: AMNH nos. 17432, 56500, 56502, 56509, 56512, 56513, 56514, 56515. Viverravus sicarius occurs at localities I and II. Both the upper and lower teeth are known, and oddly, the upper teeth are more common. The distinguishing feature of V. sicarius, aside from size, is the relative narrowness of the protcconal-parastylar distance of P*, The mean size of the Huerfano sample is smaller than the dimensions of the type specimen, AMNH 11521 from Bridger B. One Huerfano specimen, AMNH 56515, is larger than the type. The hypocone of M?* is not as well developed in the Huerfano individuals as it is in the type. Additional material from both the Bridger and Huerfano formations may indicate a subspecific difference in the two populations; present data are not signifi- cant because of the small size of the Bridger sample. TABLE 17 Measurements (in Millimeters) and statistics of lower teeth of Viverravus gracilis P, Length P, Width Horizon or locality N OR M N OR M Lost Cabin 4 5.1-5.6 5555 4 1.9-2.3 2.10 Huerfano VI 2 4.5-4.7 4.60 2 1.9 1.9 Huerfano IV — — — — — —— Huerfano II 1 —— 4.2 1 —- ib e7 Huerfano VII — =: — -- — — Bridger B 4 5.0-5.6 5e50 4 158=21 1.93 M; Length M, Width trigonid Lost Cabin 5) 5.5-6.2 5.80 4 OES) 3.25 Huerfano VI 2 4.6-5.2 4.90 2 2.9-3.0 2R95 Huerfano IV fl — Sa 1 — oa2 Huerfano II 1 — 52 1 — 20 Huerfano VII 1 a Shi 1 —- 3.3 Bridger B 9 4.9-6.1 Bes) — — — M, Width talonid M: Length Lost Cabin 4 2.42.8 2.60 4 4.1-4.2 4.18 Huerfano VI 2, — 2S 1 — 4.0 Huerfano IV iL oe Do 1 —- 4.0 Huerfano II 1 a 1.9 — — — Huerfano VII 1 — Bd se) il — 4.3 Bridger B 9 2.0-2.6 23 3 3.8-4.4 4.07 M; Width trigonid M: Width talonid Horizon or locality N OR M N OR M Lost Cabin 2 2.4-2.5 2.45 3 1.9-2.0 1.97 Huerfano VI 1 -= Df? = = = Huerfano IV 1 — 2.3 1 — Deal Huerfano II — — — — = = Huerfano VII 1 ae De) 1 -- 2A0) Bridger B 2 1.9-2.3 2:10 3 1.5-2.0 1.83 PALEONTOLOGY 5] TABLE 18 Measurements (in Millimeters) of lower teeth of Viverravus sicarius from the Huerfano and Bridger formations Mi M2 Py W- W- W- W- Locality No. EL W L_trigonid_ talonid L _trigonid talonid Huerfano AMNH II 17432 Vga agers 7.8 4.0 SI a — V 55209 — 2.5 6.8 3.6 Se 4.9 2.6 28 II 56509 G: E275 7.4 Sai) 3h — ~- _ Bridger B AMNH tL AS aaa’ a ina seo: 8.2 4.8 3.4 53 — -— YPM 13116 —_- — (het) 4.5 3.3 — -- “= 12882 —_— — 8.7 4.7 Shel! —_ — — * type V. sicarius. TABLE 19 Measurements (in Millimeters) of upper teeth of Viverravus sicarius from the Huerfano and Bridger formations ps AMNH L-protocone M} Locality nos. L-buccal metastyle W L-buccal L-lingual W Huerfano II 56512 8.0 9.4 a2 — — —- II 56500 8.2 10.3 6.4 Sed 3.6 — II 56502 — — — Dee Seal 8.1 II 56511 8.6 10.0 5.6 Sa 3.4 8.4e I 56511 ors 10.7 5.6 — — 6.5e Bridger B 1211 9.4 10.2 6.0 6rd 3.6 Sou * type V. sicarius. SUBFAMILY MIACINAE Miacis sp. cf. M. parvivorus (Cope, 1872) Plate V, figure 10 AMNH 17435, a left M’ with roots of M* and two alveoli of M°, from locality IT is close to if not conspecific to Miacis parvivorus. The specimen differs from M. latidens of the Lost Cabin in having the posterior margin of the tooth deeply concave. Specimens of M. parvivorus, for example AMNH 11500 (Matthew, 1909, p. 364, fig. 15), have a similar outline. M. latidens and M. parvivorus are of similar size. Pertinent measurements of AMNH 17435 are: Buccal length M’, 3.0 mm; lingual length, 2.3 mm; width, parastyle to protocone M’, 6.5 mm. 52 FOSSIL MAMMALIA OF THE HUERFANO FORMATION Uintacyon sp. cf. U. asodes Gazin, 1952 Plate V, figure 4 A fragment of the right jaw of Uintacyon from locality Il, AMNH 17434, with P,-M, preserved, seems close to U. asodes as far as may be determined from comparison of the photograph (Gazin, 1952, plate 6) of the damaged type specimen, USNM 19351. Measure- ments of the two specimens, listed below, show their similarity in size, the main criteria for association. Width Width Length P, Width P, Length M,; ttrigonid talonid M, USNM 19351 (from Gazin) 6.1mm 823 ES 515 — AMNH 17434 6.4 239 7.3 4.6 oul Oddectes herpestoides Wortman, 1901 Plate V, figure 11 Two specimens referable to Oddectes herpestoides Wortman have been collected in the upper faunal zone: one specimen, AMNH 55289, is from locality I; the other, AMNH 56504, a maxilla fragment with parts of P*-M? is from locality III. Both show great resem- blance to O. proximus Matthew as well as to O. herpestoides Wortman and it is likely that the two species are synonymous. O. herpestoides has priority. O. pugnax (Wortman and Matthew) may also be a synonym, and it has priority over both of the above names. However, the type of O. pugnax is a lower jaw, and direct comparison with the Huerfano specimens is not possible. Giles MacIntyre is currently revising the Miacidae and will resolve the problem of synonymy in Oddectes. The Huerfano specimens are somewhat larger than the type specimen of Oddectes herpestoides but not enough to warrant specific separation. Pertinent measurements are listed in Table 20. TABLE 20 Measurements (in Millimeters) of upper teeth of Oéddectes from the Huerfano and Bridger formations M: No. ps re Lingual Anterior AMNH J W 1G; W Buccal c W 55289 — — — — 3.4 es Si 7/ 56504 4.5 225 5.4 4.3 — —_ Sse! 11495 — — 4.6 Se 7 3-1 23 Sa) type of O. proximus from Bridger B YPM 11861 — — 4.2 4.0 S)eil 2.0 Dez type of O. herpestoides from Bridger B ?Oddectes Plate V, figure 9 AMNH 2681 from the lower faunal zone (locality VI?) is a jaw fragment with P,-M, preserved. Alveoli for Canine, P, (one), P; (two) and the roots of P, are present. Two mental foramina lie below the alveolus of P, and the anterior alveolus of P;. The tips of the protoconids of P, and M, and the talonid rim of M, are lost. The M, is peculiar in that it is smaller than the M, but is not as small as is usual in Miacidae. The metaconid and paraconid of M, are low and of equal size; the protoconid is tall. The paraconid and metaconid are not as close together as in Oddecies, the only miacid that closely resembles this specimen. This specimen may be an insectivore, but it is unlike any known to me. PALEONTOLOGY 53 The miacid-like M, argues against any insectivore relationship. The smaller M, is unlike the hyaenodonts. Measurements (in Millimeters) of AMNH 2681 P, Mi M. Length 3.0 Sal) 3.6 Maximum Width 2A 300 2e3 Vulpavus asius Gazin, 1952 Plate V, figure 2 Two specimens, AMNH nos. 56503 and 56508, referable to Vulpavus asius have been collected at locality II. Both are jaw fragments with M, preserved, in which the paraconid lies close to the metaconid, as noted by Gazin (1952, p. 58, plate 5), for M, and also illustrated in his photograph of the type M,. The lengths of the two teeth are: AMNH 56503, 5.0 mm; 56508, 5.6 mm. The proximity of the paraconid and the metaconid is not usual in Vulpavus. Most American Museum specimens of early Eocene Vulpavus show the paraconid of M, placed more centrally. One specimen, AMNH 16954 from Dorsey Creek in the Bighorn Basin, has the paraconid shifted to the position in V. asius, showing that the condition existed in some pre-Lostcabinian Vulpavus. OrDER CONDYLARTHRA Famity PHENACODONTIDAE Phenacodus vortmant (Cope, 1882) Plate X, figure 6 MATERIAL. Locality VI: AMNH nos. 17547, 55229, 55278, 55279, 55280, 55276. Locality IX: AMNH nos. 55230, 56532. AMNH 17548 from locality VI is questionably referred to P. vortmani. Nine specimens, all fragments of lower jaws, from localities VI and IX are referred to Phenacodus. The specimens possessing a P, agree well with the type of P. vortmani AMNH 4824, also a P,. The measurements of the type are: Length: 8.1mm Width: 5.5 mm In most measurements the coefficient of variability is low, but for the measurements of M, it is unusually high. This is due to the inclusion of one specimen, AMNH 17548. The pertinent statistics are: A: AMNH 17548 included: N OR M SD V Length M; 5 7.8-10.1 8.66 0.94 10.85 Width trigonid 5 Sect a) 6.12 0.85 13.89 Width talonid 5 4.6-6.8 5.36 0.93 1) fs B: AMNH No. 17548 excluded: Length 4 7.8-8.8 8.30 0.58 6.99 Width trigonid 4 Se 4 5.78 0.40 6.92 Width talonid 4 4.6-5.8 5.00 0.54 10.80 Omitting AMNH 17548, the coefficient of variability for the talonid width of M, is still high in set B but is of more reasonable value. The sample is admittedly small, and the high coefficient of variability may be due to this, but the data suggest that the sample of M,’s is heterogeneous. AMNH 17548 may be referable to either P. copei or P. intermedius (Granger, 1915, p. 333) on size characters. Since the specimen has only the third molar, it seems best to consider it P.? vortmani. The lack of other large Phenacodus specimens may be significant. Comparison of the statistics of the Huerfano and Wind River samples of P. vortmani 54 FOSSIL MAMMALIA OF THE HUERFANO FORMATION (see Table 21) shows that in most parameters the Huerfano animals averaged smaller. The Wind River sample used is small (three specimens: AMNH nos. 4824, the type, 17487, and 14801), and the Huerfano sample is not much larger. The evidence is insufficient to establish the Huerfano population a subspecies of P. vortmani until the genus Phenacodus is reviewed and possible synonymy determined. The Wind River specimens used in the above comparison are the smallest Phenacodus from the Lost Cabin fauna at the American Museum of Natural History. P. brachypternus, the only smaller species of Phenacodus, is restricted to the lower part of the early Eocene (Grey Bull and Almagre faunas, Granger, 1915, p. 345). TABLE 21 Measurements (in Millimeters) of lower teeth of Phenacodus vortmani from the Lost Cabin levels of the Huerfano and Wind River formations Huerfano Wind River Variate N OR M SD V N OR M P; Length 1 — 6.9 a = 2° 13h “T2359 Width 1 — 4.1 — — 2 — 4.5 Py Length 3. 7.7-8.0 8.00 — — 3 179-=8-4 5 Sede Width 3 Dil=SOy OLE — — 3. 5.4-5.6 S50 M, Length Saal Hho Olek eS — = 3. ld=lawiaee Width trigonid 4 6.2-6.6 6.35 0.19 3.00 2 — 6.9 Width talonid 2 — 6.4 —_ o 1 _- 6.9 M. Length So Sale MeO) BOnlor ats 2. 7.8=8.0, ) 4296 Width tagonid, 55 9674-698, 6.62)" (0-18 2272 2. TART SY eee Width talonid 3). 8O.1=626; AO9327 ORI9. 3,01 2) = 1 0-1 AS M; Length 4 7.88.8 8.30 0.58 6.99 2. 1.89.52 4598-56 Width trigonid 4 5.3-6.2 5.78 0.40 6.92 2- §.8-655.7 1695 Width talonid 4 4.65.8 5.00 0.54 10.80 2 5-1-Se6e55ec0 Famity HYOPSODONTIDAE Genus HYOPSODUS Leidy, 1870 Hyopsodus, although, or because, it is the commonest lower and middle Eocene mammal, is a systematist’s nightmare. The latest revision of the genus is found in Matthew (1909 and 1915b). Since Hyopsodus is common, statistical treatment of the genus should be more valid than for other less common beasts. Olson and Miller (1958) did not segre- gate their Bridger specimens by individual localities, instead they segregated samples from the two faunal levels (Matthew, 1909). Gazin (oral communication) believes that Matthew’s determination of the stratigraphy of the Bridger formation is open to reevaluation, and for this reason Olson’s and Miller’s conclusions are open to reevaluation. Kelley and Wood (1954) examined the Hyopsodus from the Lysite member of the Wind River formation and were able to show that three populations probably existed in that fauna: H. minor, H. mentalis and H. powellianus. Kelley and Wood measured the maximum width of M, and M.. In these teeth the maximum width may be either on the trigonid or the talonid; a summation of maximum widths therefore is the summation of part of the sample of two variables and is less useful. The maximum width of M, is always across the trigonid. The samples from the Lost Cabin member of the Wind River formation indicate that three populations exist: these are: H. wortmani, the small animals, and H. walcottianus and perhaps H. mentalis, the larger animals (Matthew, 1915b). Two populations seem to be present in each level of the Huerfano formation. In the lower level the equation of these samples with samples from the Lost Cabin member of the Wind River formation is practical. The upper zone samples are harder to allocate. The Garcia Canyon local fauna sample from the Huerfano formation is assigned to H. wortmani and H. walcottianus. PALEONTOLOGY 55 Frequency distributions of measurements of the trigonids of molars of Hyopsodus are from the following levels: Lysite-Bighorn and Wind River basins; Lost Cabin-Wind River and Huerfano basins, Gardner Butte-Huerfano basin; Black’s Fork Bridger Basin. These distributions indicate that the populations of each level in each basin are distinct. The closest similarity between population samples is between the Lost Cabin levels of the Wind River and Huerfano basins. I have not examined the samples from the Largo and Almagre levels of the San Jose formation, or from the Grey Bull level of the Willwood formation. Although the number of named species of Hyopsodus is too great, each popula- tion in each fauna probably deserves subspecific rank. H. mentalis was based on a specimen from the San Jose formation. Referring samples from Wyoming to this species is probably valid, but, until the variation of the San Jose sample is known, it is risky. Determining the species of Hyopsodus on morphological differences is, at the present time, impractical; the animals are variable. Olson and Miller’s treatment is invalidated due to the lack of locality association (see above). My own analysis indicates only one variable (in terms of topography of the tooth) population in the lower Bridger. The sample from Grizzly Buttes (Bridger B) has low coefficients of variability for the parameters used. Hyopsodus wortmani Osborn, 1902 Plate X, figure 5 MATERIAL. Locality I: AMNH nos. 17029, 55253, YPM 16412; Locality IZ: AMNH nos. Wists, 11219, 55256;-56521, 55237, 55233, 55240, 55236, 55258; 55232, 55239) 55234,°55235, 56518, 56519; Locality III: AMNH nos. 55245, 55244, 56520, 17476; Locality V: AMNH nos. 55241, 55243, 55242; Locality VI: AMNH nos. 16907, 17452, 55255; Locality VIII: AMNH 55257; Locality VIIIa: AMNH 55260. Hyopsodus wortmani is present in both faunal levels of the Huerfano formation. The sample from the upper faunal zone is composed of slightly smaller individuals than those of the lower zone sample (Tables 25, 26). This may be related to the presence of H. paulus in the upper zone. H. paulus, though larger than H. wortmani, is much smaller than H. walcottianus from the lower zone. The small individuals of H. wortmani in the upper zone may reflect the competition with a small related species of similar habit. The sample of small Hyopsodus from the Lost Cabin level of the Wind River forma- tion is bimodal (see Tables 22, 23, 24, frequency distributions of trigonid widths). Where measurements are restricted to specimens from one locality in other formations the separation of groups is sharp. Grouping the specimens from the Lost Cabin level of the Wind River formation according to locality might resolve this, but it cannot be done now. The smallest group in the Lost Cabin (Wind River) sample is H. wortmani. The sample with which it intergrades in size may be H. mentalis. H. walcottianus is the large species present. The statistics on the two smaller samples have been calculated together because I have no way of separating them at this time. This explains the high coefficients of variability. Hyopsodus paulus Leidy, 1870 Plate X, figure 9 Two specimens, YPM 16435 from locality III and YPM 16486 from locality II are referable to H. paulus. Both specimens are isolated lower molars and add nothing to our knowledge of the species except its distribution. Hyopsodus walcottianus Matthew, 1915 Plate X, figure 7 MATERIAL. Locality IV: AMNH nos. 17534, 17537, 17540, 17541, 55247, 55249, 55250, 55251, 55254, YPM nos. 16446, 16469, 16470; Locality VI: AMNH nos. 17535, 17536, 17538, 16906, 16905, 55259, YPM 16447. W. walcottianus occurs at localities IV and VI, abundantly at the former. The range of size of the Huerfano specimens overlaps the size range of the H. powellianus popula- 56 FOSSIL MAMMALIA OF THE HUERFANO FORMATION tion from the Lysite fauna of the Willwood formation. The H. walcottianus sample from the Wind River Lost Cabin consists of three specimens whose size range is almost the same as that of the Huerfano sample (see Table 26). The Huerfano, Garcia Canyon, and the Wind River, Lost Cabin, samples are obviously related and both show relationship to the Willwood Lysite sample. Possibly H. walcottianus is a temporal subspecies of H. powellianus, but actual designation of it as such awaits a revision of the genus Hyopsodus. The relative abundance of H. walcottianus at locality IV and the absence of H. wortmani from that locality is further evidence for the unique nature of Locality IV and the Farisita formation in general. TABLE 22 Frequency distribution of widths of trigonids of M, of Hyopsodus from Lysite, Lost Cabin, Gardner Butte and Blacks Fork faunal levels Willwood Wind River Huerfano Gardner Bridger Width Lysite Lysite Lost Cabin Lost Cabin Butte Blacks Fork* 2.0 mm 1 Des Den 1 1 Dye 2 1 2 2.4 2 3 1 259 2 5 2.6 3 1 2 oRSy| 3 4 2.8 2 4 + 6 DAO 3 3 1 13 3.0 2 2 8 26 Saul 1 2 18 S22 1 2 3 Sind 1 2 3 3.4 1 3 1 3.5 1 + 1 3.6 8 1 Sal, 3 3.8 3 1 2 3.9 3 3 1 4.0 4 1 yy 4.1 2 1 1 4.2 3 1 1 4.3 2 1 4.4 2 5 4.5 2 3 * All specimens from Grizzly Buttes. PALEONTOLOGY 57 TABLE, 23 Frequency distribution of widths of trigonids of Mz of Hyopsodus Willwood Wind River Huerfano Gardner Bridger Lysite Lysite Lost Cabin Lost Cabin Butte Blacks Fork* 2.4mm 1 1 2-5 230 3 af 1 2 4 2.8 1 1 9 2.9 1 2 2, 3.0 1 3 3 1 Sol 1 7 3.2 3) 3 1 1 1 323 3 3 4 1 6 3.4 3 3 1 15 3:5 4 3 1 21 3.6 1 3 15 Siar 1 DZ 1 11 328 1 1 3.9 1 2 4.0 6 1 1 4.1 1 7 4.2 6 1 4.3 2 2 4.4 1 5 1 4.5 1 1 1 4.6 2 1 1 4.7 5 1 2 4.8 1 1 3 4.9 3 2. 5.0 1 4 Sea 2 1 a2 1 Z SEG) 1 5.4 Sas) 1 * All specimens from Grizzly Buttes. 58 FOSSIL MAMMALIA OF THE HUERFANO FORMATION TABLE 24 Frequency distributions of widths of trigonids of M; of Hyopsodus Willwood Wind River Huerfano Gardner Bridger Lysite Lysite Lost Cabin Lost Cabin Butte Blacks Fork* 2.2 mm 1 Des 2 2.4 2 2.5 1 4 2.6 2 4 Pal 1 1 3 1 2.8 2 3 1 2.9 3 3 2 2 3.0 1 7 2 1 4 one 4 2 15 g.2 2 2 4 19 303 4 15 3.4 11 Bye) 1 1 4 3.6 7 2 edh 5 1 3.8 4 il 3.9 2 2 4.0 3 3 1 4.1 1 2 4.2 1 2 1 2 4.3 2 4.4 2 1 4.5 2 4.6 2 4.7 1 4.8 4.9 5.0 1 She 1 * All specimens from Grizzly Buttes. Measurements (in Millimeters) of trigonid widths of molars of small species of Hyopsodus, excluding H. minor, from several faunas Fauna Lysite Lysite Lost Cabin Lost Cabin Gardner Butte Blacks Fork Lysite Lysite Lost Cabin Lost Cabin Gardner Butte Blacks Fork Lysite Lysite Lost Cabin Lost Cabin Gardner Butte Blacks Fork PALEONTOLOGY Formation Willwood Wind River Wind River Huerfano Huerfano Bridger AMNH YPM Willwood Wind River Wind River Huerfano Huerfano Bridger AMNH YPM Willwood Wind River Wind River Huerfano Huerfano Bridger AMNH YPM TABLE 25 W w NN DS WW ra oo WW RPnow.s AnuUHN COmMnmN wow Mz bo dS bd bd bd le) nN WwW So paar 59 60 FOSSIL MAMMALIA OF THE HUERFANO FORMATION TABLE 26 Measurements (in Millimeters) and statistics of trigonid widths of molars of large species of Hyopsodus Fauna Formation N OR M S V Mi Lysite Willwood 16 3.9-4.4 4.12 0.14 3.40 Lysite Wind River 27 3.3-4.2 3.66 0.49 13.39 Lost Cabin Wind River 3 4.0-4.5 4.33 — — Lost Cabin Huerfano 7 3.6-4.5 4.19 0.26 6221 M;: Lysite Willwood 19 4.1-5.1 4.68 0.35 7.48 Lysite Wind River Si 3.9-4.7 4.20 0.49 11.69 Lost Cabin Wind River 3 4.8-5.3 SS il7/ — — Lost Cabin Huerfano 19 4.2-5.3 4.85 0.23 4.74 M; Lysite Willwood 9 3.9-4.4 4.10 0.19 4.63 Lysite Wind River 25 3.6-4.2 3.81 0.52 13464 Lost Cabin Wind River 3 4.2-5.1 4.63 — — Lost Cabin Huerfano 11 4.0-5.0 4.42 0.23 520 OrpvER DINOCERATA Famity UINTATHERIIDAE Bathyopsts sp. cf. B. fissidens (Cope, 1881) Plate I, figure 6 Bathyopsts sp. cf. B. fissidens occurs at locality III and probably at locality I. AMNH 17438, a left P, (not ?P,; or P,, Wheeler, 1961) from locality III agrees well with the type AMNH 4820. AMNH 17444, skeletal fragments of a uintathere, from locality I may pertain to B. fissidens; the astragalus is smaller and less specialized than that of Uintatherium. No uintathere remains have been found in the lower faunal zone. Measurements (in Millimeters) of AMNH 17438 are: Length Width trigonid Width talonid 14.8 8.9 8.4 OrpdER PANTODONTA FaMity CORYPHODONTIDAE Coryphodon sp. Plate I, figure 8 Numerous fragmentary specimens of Coryphodon occur in the lower faunal zone, and one specimen, AMNH 56543, a left P,, was collected at locality II. The specimens now prepared are too fragmentary for specific identification. Better material and a revision of the genus Coryphodon will be necessary before accurate determinations can be made. Specimens of this genus occur at localities VI, VIII, VIIa, IX and an isolated specimen, YPM field no. 59-12, was collected in exposures about 14 mile west of locality IV. AMNH nos. 2690 and 17556 are the only catalogued specimens from the lower faunal zone. The Coryphodon specimen from the upper faunal zone may be the youngest record of the genus in North America. PALEONTOLOGY 61 OrDER PERISSODACTYLA FAMILY EQUIDAE Hyracotherium craspedotum (Cope, 1880) Plate IV, figure 3 MATERIAL. (K) = Kitts’ sample: Locality VI: AMNH nos. 17509(K), 17510(K), 17513(K), 17521, 55100(K); Locality IX: AMNH nos. 56530, 56532. The sample of horses used in this study differs from Kitts’ (1956) sample for several reasons: Kitts included material from Locality VIII (Costillo Pocket) in his study, and did not include, in most cases, material from the upper faunal zone. Also, I have the advantage of having material collected since Kitts’ study was completed. Kitts omitted several specimens with numbers of the 2,000 series (collected in 1897) and the 17,000 series (collected in 1916, 1918) from his studies. Differences in data between Kitts’ sample and mine arise from using some different individuals. TABLE 27 Measurements (in Millimeters) of lower teeth of Hyracotherium craspedotum and H. vasacciense from the Huerfano and Farisita formations: Variate Locality N OR M SD V Species M, Length TY SEV, A 2-756 7.40 — — H. vasacciense VI 9 6.48.4 1232 00..68y 9.29 fe IV 2 7.8-9.0 8.40 — = 2 Vi, Xs 4 8.8-9.6 9.20 — — 4H. craspedotum Width trigonid II, III, V. 4 5.2-5.3 5.28 — —— VI 9 4.5-5.9 5.04, 0.50, 5.85 IV 2 5.6-6.1 5.85 — VIE IX 4 6.2-6.5 6.40 — -— Width talonid II, III, V. 4 5.2-5.5 5.335 —— VI 10 4.6-6.2 2 nad) 28. OD IV 2 5.8-6.5 6.15 — Vi; 1x 4 6.5-6.7 6.58 — — M: Length 116 (yd 11) Gee Ve P07 ASI sd 7.61 VI 10 7.1-8.5 7.61 0.54 7.06 IV 3 8.1-9.2 8.65 VI, IX 6 9.7-10.4 10.01 — as Width trigonid II, III, V. 7 4.9-5.8 5.49 — VI 11 5.06.3 S257, 10,48 8.53 IV 3 6.0-6.5 6.17 — Vi, ix 6 6.6-7.3 7.02 — —- Width talonid II, III,V. 7 5.0-5.8 5.44 — VI tly. 457-636 5.65 0.61 10.80 IV 3 5.8-6.8 6.30 VI, IX ind Oo 6.90 — — M; Length II, II, V. 7 10.0-11.8 10.830 — VI 10 9.7-11.1 10.40 0.50 4.80 IV 2? 1213.52" 912620 — Vin ix 3 13.8-14.0 13.93 — — Width trigonid II, III, V. 7 4.8-5.8 5.37 — — VI 9 4.9-6.0 5,52 (0:34 “6.14 IV 2 6.1-6.3 6.200 — — Vir tx 3h 720-73 7.170 — —- Width talonid II, III, V. 7 4.5-5.6 5.07 — a= VI 10 4.4-5.5 5.05 0.45 8.91 IV Zin S9=6.1 6.00 — VinelxX 3 6.6-6.8 6.70 — — 62 FOSSIL MAMMALIA OF THE HUERFANO FORMATION Hyracotherium vasacciense (Cope, 1872) Plate IV, figures 4, 5, 7 MATERIAL. Lower faunal zone: Huerfano facies. (K) = referred by Kitts. H. v. vasac- ciense. Locality VI: AMNH nos. 2586, 17511(K), 17512(K), 17514(K), 17575(K), 17516(K), 17517(K), 17518(K), 17520(K), 55103(K), 55104(K), 55105(K), 55277, YPM 16455; Locality VIII: Kitts referred AMNH 55101, these specimens have not been studied other than to note the presence of this species in the quarry sample. Locality IX: AMNH 56531. Upper faunal zone: Huerfano facies. Locality I: AMNH 56529; Locality II: AMNH nos. 17446, 17447(K), 17449, 55180, 55181, 55183, YPM nos. 16479, 16481; Locality III: AMNH nos. 17021, 17448(K); Locality V: YPM nos. 16440, 16448. Upper faunal zone: Farisita facies, Y mile east of Gardner Butte: AMNH 55185. H. v. (?) venticolum. Lower faunal zone, Farisita facies: Locality IV: AMNH nos. 17522, 55107, 56533, YPM 16436. Hyracotherium vasacciense is present at localities I, II, III, IV, V, VI, VIII and IX and therefore occurs in both faunal levels. Horses are scarcer in the upper faunal zone, but those present closely resemble the lower zone sample. Probably through unfamiliarity with the Huerfano localities, Kitts (1956, p. 46) included two specimens, AMNH nos. 17447 and 17448, from the upper faunal zone in his analysis of the lower zone sample. TABLE 28 Frequency distributions of talonid widths of Ms; of Hyracotherium from the Huerfano formation Width (in Locality Millimeters) VI, IX IV TL Tia 4.4 1 4.5 1 4.6 Aa 2 4.8 2 1 4.9 1 5.0 1 Sat 1 4 1 H. vasacciense vasacciense aS 1 Sac! 1 55) 3 1 556 1 Sees 5.8 5.9 6.0 6.1 1 WH. vasacciense ?venticolum Gn2 6.3 1 1 1 H. craspedotum CONT DN Ue PALEONTOLOGY 63 Kitts placed the smaller Huerfano hyracothere in the subspecies H. vasacciense vasacciense on size criteria and the morphology of P*. He stated that the smaller horses were easily separable from the larger (H. craspedotum) on the basis of size; I found this true for most molar measurements, but I found some overlap in talonid widths of M, (see Table 28). Twinning of the metaconid in molars, noted by Kitts (1956, p. 12) is common in Huerfano specimens of H. vasacciense. ‘Twinned metaconids are also present on DP,., of AMNH 55185. Without premolars for comparison, it is almost impossible to distinguish between the lower jaws of some smaller horses from Bridger, Huerfano and Wind River formations. Kitts (1957, p. 5) cites the shorter heel of M; as a basis for separation of Orohippus and Hyracotherium, but some specimens of M, from these formations are indistinguishable. New data on tooth replacement of Hyracotherium is provided by AMNH 55185. The specimen consists of both jaws, several upper teeth and numerous skeletal fragments. The right jaw has DP,_, and M,-;, M; not fully erupted. M, was in the process of eruption prior to the shedding of DP, , or .. AMNH 55185 is from low in the upper faunal zone, col- lected about 14 mile east of Gardner Butte (Farisita facies). Four specimens of horses from locality IV (Farisita facies) are questionably assigned to H. vasacciense venticolum. They are large for H. v. vasacciense and small for H. craspe- dotum. These are the only horse specimens collected from locality IV and their presence suggests that, if they are correctly assigned to subspecies, H. v. vasacciense and H. v. ventt- colum occupied different habitats in the same basin. AMNH 55185, mentioned above, also comes from the Farisita facies and is one of the smallest horses collected. It is, how- ever, from several hundred feet higher in the upper faunal zone beds. If the locality IV horses are H. v. venticolum, then the known distribution of this subspecies is extended from the Wind River and Debeque formations to include the Farisita formation. The measurements of teeth of specimens of H. v. vasacciense from locality VI have bimodal distribution (see Table 28), best shown in talonid widths of M,. Because the number of specimens in each group is about equal, I suggest that the difference is due to sexual dimorphism. The sample from the upper faunal zone is not sufficient to show this grouping. FAMILY BRONTOTHERIIDAE SUBFAMILY LAMBDOTHERIINAE Lambdotherium popoagicum (Cope, 1880) Plate IX, figure 4 MATERIAL. Locality IV: AMNH nos. 17530, 56534, YPM 16466; Locality VI: AMNH nos. 2688, 17526, 17527, 17528, 56535, 56536, 56537, 56541; Locality VIIIa: AMNH nos. 55264, 55263, 56538. Bonillas (1936, p. 141) placed all other described species of Lambdotherium in synonymy with L. popoagicum. The type specimen of L. magnum (Osborn), AMNH 17527, comes from locality VI of the Huerfano formation. Although that specimen is the largest lambdothere in the American Museum or the Peabody Museum collections, it is within the probable range of variability for the species L. popoagicum. Most of the specimens of Lambdotherium from the Huerfano formation are jaw fragments. Statistics of measurements of lower teeth for both the Huerfano and Wind River specimens in the American Museum are listed in Table 29. The Huerfano sample of Lambdotherium resembles its counterpart in the Wind River formation more than any other pair of mammalian samples from these formations. 64 FOSSIL MAMMALIA OF THE HUERFANO FORMATION TABLE 29 Measurements (in Millimeters) of lower teeth of Lambdotherium popoagicum from the Lost Cabin levels of the Huerfano and Wind River formations Variate Formation N OR M SD V P, I, Huerfano 7 7.3-8.4 7.85 —— — Wind River 4. 7.5-8.0 PE 0.21 Pegi P; W Huerfano 2 4.1-5.0 4.65 — — Wind River 4 4.1-4.7 4.43 0.28 6.33 Ps Ec Huerfano 4 8.4-9.7 8.90 0.59 6.63 Wind River 10 8 .3-9.4 8.99 0.31 3.45 Ps W Huerfano 4 5.46.3 5.85 0.47 8.03 Wind River 10 5.3-6.6 5.82 0.46 7.90 P, 1 Huerfano 6 8.7-10.0 9.37 0.35 Sieh Wind River 9 9.1-10.1 9.43 0.25 2.65 P, W trigonid Huerfano 6 6.3-7.2 6.72 0533 4.91 Wind River 9 6.3-7.3 6.76 O53 4.59 P, W talonid Huerfano 6 6.5-7.5 7.02 0.33 4.70 Wind River 9 6.6-7.7 7.04 0.44 6.25 Mi 16 Huerfano A AO S7—12705 sil 23 0.40 3.50 Wind River 10 10.0-11.8 10.99 0.61 5255 Mi W trigonid Huerfano 5 7.2-8.2 7.84 0.38 4.85 Wind River 10 6.9-8.5 TSH 0.35 4.56 M, W talonid Huerfano 5 7.5-8.7 8.12 0.55 6.77 Wind River 10 7.3-8.5 1.95 0.37 4.65 M: IF Huerfano AS AL 7-120 12ets O. 5 4.19 Wind River OFF TD A0S12 65 alles 0.33 2.68 Me W trigonid Huerfano + 8.6-9.6 OAS 0.44 4.81 Wind River 10 7.8-9.5 8.81 0.38 4.31 M: W talonid Huerfano 6 7.5-9.6 8.85 0.76 8.59 Wind River 9 8.2-9.5 8.89 0.41 4.61 M; 1 Huerfano 8 15.418.5 16.79 1.13 6.73 Wind River 7 15.8-17.8 16.99 0.68 4.00 M; W trigonid Huerfano 9 8.2-10.1 8.94 0.63 1205 Wind River if 8 .3-9.6 9.03 0.43 4.76 M; W talonid Huerfano 7 7.0-9.5 8.18 0.79 9.66 Wind River 8 7.5-9.2 8.23 On55 6.69 Mjes ae Huerfano De AIO -4A4S 42/275 a= —— Wind River 8 37.1-42.2 40.09 1.74 4.34 SUBFAMILY PALAEOSYOPSINAE Palaeosyops fontinalis (Cope, 1882) Plate I, figure 9, Plate VIII, figure 1, Plate IX, figures 1, 3 SyNonyM: Eometarhinus huerfanensts Osborn, 1919 MATERIAL. Locality I: AMNH nos. 17450, 17013, 17416, 56540, 17411, 17412, YPM 16462; Locality II: AMNH nos. 17417, 17425, YPM 16450; Locality V: AMNH nos. 55282, 17415, 17413, 17474; Locality No. 5 of R. C. Hills: YPM 16463; 1 mile south of Hausero’s (Collected by R. C. Hills): YPM 16451; 1 mile a little southeast of Gardner Butte: YPM 16459. The type specimen of P. fontinalis is a partial skull of a young animal with DP*-M*. Another juvenile specimen, YPM 16451, has DP**, M*, and DP,,, and is the only specimen of this species, known to me, that has lower and upper teeth associated. An excellent upper dentition, YPM 16450, consisting of left C-M* and right P’-M?, was col- lected by Peter Hilgeford in 1958. Partial lower dentitions are commoner, but none are PALEONTOLOGY 65 well preserved. No lower dentition with both deciduous and permanent teeth is known to me. Therefore, lower jaws are referred to this species on size criteria and without com- parison to YPM 16451. The inclusion of Eometarhinus huerfanensis in P. fontinalis is based on comparison of the type of E. huerfanensis, AMNH 17412, with skulls of other palaeosyopsines. Osborn’s reconstruction of the skull (1929, fig. 355) is, I think, misleading. The orientation of the nasals is speculative since no true contact remains. Restoration of the nasal posi- tion can be made to agree as well with palaeosyopsines such as AMNH 5104, type of Limnohyops laevidens, or with the type of P. fontinalis, AMNH 5107. The latter has very strongly arched nasals for a young animal. The nasal bones of palaeosyopsines are thin; they might have curved in dessication following death, as some bones do. The teeth of AMNH 17412 are poorly preserved but agree in outline and size with AMNH 17411 and YPM 16450. The maxilla and premaxilla of AMNH 17412 resemble AMNH 5104 closely. The apparently high maxillonasal contact is partly due to crushing. The “horn bosses” mentioned by Osborn (1929, p. 420) are apparently misinterpretations of the nasofrontal suture. No evidence for a hypocone on M, exists as that portion of the tooth is absent. Future restudy of the Palaeosyopsinae will probably reduce the number of species and, perhaps, genera. P. fontinalis may eventually be considered a subspecies of P. paludosus, but it seems best to let the species stand until such time as detailed studies reveal its posi- tion. Measurements of teeth of P. fontinalis from the Huerfano formation are given in Tables 30, 31. TABLE 30 Measurements (in Millimeters) of lower teeth of Palaeosyops fontinalis from the Huerfano formation AMNH nos. YPM nos. Variate 17450 17417 17013 17416 55282 56540 16459 16463 Po, & —— — — — — —- 16.2 -- W — —_ — — — —- 9.1 — Py ©L — ~- a -- 16.4 16.4 1527 = W -— = — 9.6 8.8 9.2 -- ay ey — — -- ISSO ehiay, — 15.8 -- W trigonid — — - 10:2 1046 — ORS) — W talonid — —_— -- 10.7 -- -— a = Mi, -- — 20.5 — 20.6 — — ~- W trigonid — — 1245 -= 13.6 — a Sil, — W talonid —- -- ilejeul — 1303 — 14.3 -- M: L -- DS ie tas iear — ~- —- 28.5 = W trigonid — 16527) 16.0 — — — aye. — W talonid oa Vip S16)..6 — 16.5 — — —_— M, L = a — — — — — 39.7 W trigonid — a -= -- 18.0 — ~- 19.1 W talonid 16.5 -- -= -— 16.4 ~- -- 18.4 Genus EOTITANOPS Osborn, 1907 Eotitanops first occurs a few hundred feet above Lambdotherium in both the Wind River and Huerfano formations. The species first described, E. borealis (Cope), is from the Lost Cabin level of the Wind River formation. Analysis of the Eotitanops material in the American Museum indicates that one population of this genus existed in upper Lost Cabin time in the Wind River basin. This species, for which the name E. borealis has priority, is also found in upper beds of the lower faunal zone of the Huerfano formation. A second species of Eotitanops, E. minimus, is present in the upper beds of the Huer- 66 FOSSIL MAMMALIA OF THE HUERFANO FORMATION TABLE 31 Measurements (in Millimeters) of upper molars of Palaeosyops fontinalis from the Huerfano formation Buccal Lingual Anterior Posterior Specimen length length width width Mi YPM nos. 16451 -= 19.0 23.6 — 16450 23.4 19.5 24.0 23.9 AMNH nos. 17411 —- 19.2 27.3 24.5 5107* 26.6 20.5 29.0 25.9 Type 17413 26.9 21.6 28 — 17425 — 2120 28.3 _— M, YPM no. 16450 29.5 DIRS, 30.5 28.4 AMNH nos. 17411 Si0 24.7 ot 29.7 17425 S33 26.3 8527 — M; YPM no. Width 16450 31.0 21.00 29.0 AMNH nos. 17411 3752 2273 3027 17415 — I — 17413 36.5 Dey 35n3 17425 Sas) DES 33.8 17414 38.2 26.4 34.3 * (From Osborn, 1929, Figure 270) fano formation. Palaeosyops fontinalis occurs alongside E. minimus; the former is larger than E. borealis and the latter is smaller. I think that the two upper zone species are possible descendants of E. borealis. Morphologically, E. borealis resembles P. fontinalis more than it does E. minimus. E. minimus has the hypoconulid of M, relatively reduced, and therefore can be separated from E. borealis on morphological as well as size criteria. As far as I know, E. minimus is found only in the upper faunal zone of the Huerfano formation. Eotitanops borealis (Cope, 1880) Plate IX, figure 2 One specimen, AMNH 17441, an M,, from locality VII, is referable to E. borealis. Osborn (1919) referred this specimen to E. brownianus, but it is very similar to a referred specimen of E. borealis, AMNH 14890, (Osborn, 1929, p. 290). The type of E. borealis is an upper dentition and direct comparison is impossible. Measurements (in Millimeters) of AMNH nos. 17441 and 14890 Ms; Length Width trigonid Width talonid 17441 Doiae 11.8 1.2 14890 23). = 11.3 PALEONTOLOGY 67 Eotitanops minimus (Osborn, 1919) Plate IX, figure 7 MateErIAL. AMNH nos. 17418, 17439 (type), 56539 and YPM 16439. Eotitanops minimus is present in the upper faunal zone of the Huerfano formation, all four of the poorly preserved specimens coming from locality II. Osborn (1919, 1929) referred AMNH 17418 to E. gregoryi, but it is smaller than specimens from the Lost Cabin sample (Table 32). TABLE 32 Measurements (in Millimeters) of second lower molars of Eotitanops from the Huerfano and Wind River formations Width Width Huerfano Length trigonid _ talonid AMNH nos. 17439 14.5 9.1 9.3 Type EL. minimus 17418 iS 10:2 10.9 56539 15.4 9.3 9.8 YPM no. 16439 = 9.8 1022 Wind River AMNH nos. 14890 20.0 12-5 1265 14888 18.6 1176 12.4 14889 17.0 10.6 10.2 Type E. gregoryi 14891 LOR 1275 12.2 Famity HELALETIDAE Helaletes sp. cf. H. nanus Plate IX, figures 5, 8 MATERIAL. Locality I: YPM 16464. Locality I: AMNH nos. 17503, 55262. Locality No. 5 of R. C. Hills: YPM 16457. Several fragmentary specimens of Helaletes, collected from the upper faunal zone, resemble the Bridger species H. nanus. Radinsky (1963, p. 46) considers the upper Huer- fano specimens as possible intermediates between Heptodon and Helaletes. ?Heptodon sp. Three fragmentary specimens from locality IV are referable, with doubt, to Heptodon. These specimens are AMNH nos. 17523, 17524, 17563. They might also be Helaletes. FAMILY HYRACHYIDAE Hyrachyus modestus (Leidy, 1871) Plate IX, figure 6 MATERIAL. Locality I: AMNH nos. 17014, 17436. Locality II: AMNH nos. 17440, 55261, field nos. 1952-277, -342, -407, -476. Locality V: AMNH 17442. Wood referred nos. 17014, 17436, 17440, 17442 to this species. H. E. Wood (1934, p. 192, 196-7) referred the Huerfano specimens of Hyrachyus to H. modestus. As Wood noted, the material is fragmentary, and subsequent collections have not added significantly to knowledge of the species. At present the genus is known only from the upper faunal level. A complete lower molar series is provided by AMNH 17440 and DP,_, by AMNH 55261. AMNH 17436 has fragmentary limb and foot bones. Measurements of H. modestus are listed in Table 33. 68 FOSSIL MAMMALIA OF THE HUERFANO FORMATION TABLE, 33 Measurements (in Millimeters) of lower teeth of Hyrachyus modestus from the Huerfano formation field no. Specimen no. AMNH 17442, 17440, 55261, 1952-407 Locality: V 11 II II Variate Ps. SE 8.4 -- -= —- W 553 -- = -- Ps — 10.0 — — W — 4.2 — -—— Pa ae os 11.9 ~- — W -- 8.8 -— — DE; -- — 13.8 — W trigonid ~- = Gif, — W talonid — — 7.4 — DEA == — 1279 — W trigonid — — (eS) — W talonid — —- S.0 — M, L 14.6 14.2 13.9 1354 W trigonid ORs 9.3 OF? — W talonid 9.9 -— 9.5 — M; = 16.9 — == W trigonid -- 10.3 -— — W talonid —- ftite3 — — M; L 1 L7/e) ibfeadh os — W trigonid 10.8 10.8 — — W talonid 10.3 10.2 —- o- Mi-23 L — 49.9 — — OrpvER ARTIODACTYLA FamMity DICHOBUNIDAE SUBFAMILY DIACODEXINAE (sensu Gazin, 1955) Genus BUNOPHORUS Sinclair, 1914 Bunophorus specimens occur in both the upper faunal zone, localities II and V, and the lower faunal zone, localities IV and VI, of the Huerfano formation. A single specimen was collected from the Farisita formation five miles north of Gardner Butte. Gazin (1952, p. 72-73) states that distinction between Bunophorus and Wasatchia is difficult, and he draws attention to the nature of the wear of the cusps of upper teeth of Bunophorus noting the bunodonty and the circular areas of wear. The protocone, protoconule and metaconule of Wasatchia form a large worn cresent following wear (Sinclair, 1914, p. 262, fig. 6). In older specimens of Bunophorus the worn patches coalesce, the protocone and protoconule patches join first and then the metaconule patch joins. In extremely old specimens it may be impossible to tell the two genera apart. However, Bunophorus molars wear flat, while those of Wasatchia wear in a band, from front to back, across the center of the tooth. This comparison is drawn from one specimen of Wasatchia, AMNH 15673, the type of Wasatchia dorseyana, and four specimens of Bunophorus (given in order of increasing wear), YPM 16472, AMNH nos. 17485, 56524, and 17486. Admittedly, the discussion of Wasatchia is based on insufficient data. The lower teeth of these genera are equally hard to distinguish. Gazin (1952, p. 72) notes that Wasatchia seems to have a paraconid and a hypoconulid, and inflated molars, but these characters can also be found in Bunophorus. Worn lower molars of Wasatchia have a raised trigonid whereas those of Bunophorus are flat. These wear differences cor- PALEONTOLOGY 69 relate with the upper teeth and indicate that the two genera probably had different chewing motions. Bunophorus sp. cf. B. macropternus (Cope, 1882) Plate X, figures 8, 10, 11 MATERIAL. Locality II: AMNH nos. 17485, 17486, 17487, 17488, 17489, 55204, 55252, 56524, 56525, 56527; YPM nos. 16472, 16475. Locality IV: AMNH 17553. Locality V: AMNH 55206. Locality—‘‘East side William’s Creek, five miles north of Gardner Butte.” This set of exposures is in the Farisita formation. AMNH 17561. Two species of Bunophorus have been described, B. macropternus (Cope, 1882, p. 179) and B. etsagicus (Cope, 1882, p. 189). The Huerfano specimens have a small basin on the heel of P, which B. etsagicus lacks. The P, of B. macropternus is not known, therefore designation of the Huerfano Bunophorus as B. macropternus is doubtful. AMNH 17561, from the Farisita facies, is noticeably smaller than Huerfano facies specimens. The sample from the lower faunal zone is not sufficient to indicate any possible change between the populations of the two levels. TABLE 34 Measurements (in Millimeters) and statistics of lower teeth of Bunophorus sp. cf. B. macro- pternus from the upper faunal zone of the Huerfano formation Localities II and V. Variate N OR M SD V Py i Z 7.7-8.1 7.90 - — W 2 4.2-4.6 4.40 —- -- M, L 7 6.5-7.3 6.90 0.26 Seely W trigonid i 4.7-5.4 5.01 0.35 6.99 W talonid 7 5.5-6.2 5.80 0.26 4.48 M: i 6 7.3-8.0 ee dh 0225 3.22 W trigonid 6 6.0-6.7 6.38 0.26 4.08 W talonid 5 6.4-7.2 6.84 0.32 4.68 M; ae 4 8.0-10.3 9.38 0.99 10.55 W trigonid 4 6.0-7.0 6.58 0.34 Sell? W talonid 5 5.3-6.2 5.80 0.32 5.52 Mies 9 4 23.6-25.8 24.70 0.90 3.64 Diacodexis chacensis (Cope, 1875) Plate X, figure 4 One specimen, AMNH 17552, from locality VI is referable to D. chacensis. The speci- men is small for D. chacensis but is within the limits of size stated by Sinclair (1914, p. 291). Measurements are quoted below. Measurements (in Millimeters) of D. chacensis, AMNH 17552 P, Length 4.2 Width 2.2 Mi Length 3.8 Width trigonid 2.6 Width talonid 3.0 M: Length 4.2 Width trigonid 3.0 Width talonid 3.4 70 FOSSIL MAMMALIA OF THE HUERFANO FORMATION Diacodexis sp. cf. D. secans (Cope, 1881) Plate X, figure 3 One specimen of Diacodexis, AMNH 17560, was collected from exposures in the Farisita formation one mile north of Gardner Butte. It resembles Hexacodus Gazin, but the hypoconulid of M,., is not situated as far behind the hypoconid as in Hexacodus. The entoconid of M,., bears accessory cuspules. The dimensions of M, are close to those mentioned by Gazin for a Knight specimen (Gazin, 1952, p. 71). Measurements (in Millimeters) of Diacodexis sp. cf. D. secans, AMNH 17560 P, M, M; Length 5.6 4.3 4.7 Width trigonid a2 a2 3.9 Width talonid Se) 4.2 SUBFAMILY HOMACODONTINAE (sensu Gazin, 1955) Antiacodon pygmaeus huerfanensis new subspecies? Plate x figs- 1,2 Type. AMNH 17490, fragments of both jaws with right P,-M, and Left M, .. Hypopicm. Type and AMNH nos. 17472, 55202, 17491, 55203, 56528. Horizon AND Loca.itry. Upper faunal zone, Huerfano formation, uppermost lower Eocene. Type and AMNH nos. 17472, 55202, 56528 from locality II, AMNH nos. 17491 and 55203 from locality III. D1acnosis. Slightly larger than A. p. pygmaeus from the Black’s Fork member of the Bridger formation. P, similar to that of Hexacodus, with paraconid and metaconid imper- fectly separated from protoconid. Discussion. The similarity of the fourth lower premolar of A. p. huerfanensis to Hexacodus adds support to Gazin’s (1955, p. 11) hypothesis that Hexacodus might be the ancestor of Antiacodon. The size of the individuals in the Huerfano sample is generally larger than that of the Bridger sample measured (collections at the American and Peabody Museums of Natural History), as listed in Table 35. Measurements (in Millimeters) of AMNH 17490, type of Antiacodon pygmaeus huerfanensis new subspecies: P, M; M:, Length 5.0 4.5 4.8 Width trigonid 2.9 Sal Sal Width talonid SES 4.1 7 Subspecific name in reference to Huerfano Park, Colorado and to the Huerfano formation. PALEONTOLOGY 71 TABEE 35 Measurements (in Millimeters) and statistics of Antiacodon pygmaeus from the Huerfano_ . formation and from the Blacks Fork member of the Bridger formation Huerfano Variate N OR M SD V Pao y 5.0-5.4 5.20 — — W 2 2./-2.9 2.80 — -= M: L 1 — 4.5 — W trigonid 1 — Syn! — — W talonid 1 — 3:5 _ = M: L 4 4.8-5.4 5.10 0.26 5.10 W trigonid 4 320-300 3.65 0.10 2.74 W talonid 4 3.8-4.4 4.08 0.21 5.14 M; L 3 5.0-5.5 S27 — —— W trigonid 4 3.0-3.8 3.45 0:38. MeOL W talonid 4 2.9-3.6 S825 ison, tO i, Bridger N OR M SD V Py, 3 2 4.8-4.9 4.85 — —- W if — 226 — — M, L 7 4.24.6 4.41 0.26 5.90 W trigonid 6 2.7-3.0 2.87 0.13 4.53 W talonid 6 2.9-3.5 Sey 0.22 6.94 M: L 7 4,5-4.8 4.64 0.12 2.59 W trigonid 6 3.1-3.8 3.38 0.26 7.69 W talonid 6 3.3-4.0 3.66 0.24 6.54 M: L 5 5.2-5.5 5.32 0.13 2.44 Wtrigonid 4 3.0-3.4 ees) 0.17 Gye) W talonid 5 2.9-3.1 2.98 0.09 3.02 APPENDIX STRATIGRAPHIC SECTIONS Sections listed from south to north SECTION 1. ARCHULETA DRAW Starts in Sec. 1, T27S, R69W, and continues to Sec. 11, T27S, R69W. Pediment: Thickness Huerfano formation (in part): Feet 6. Mudstone, alternating red and brown, beds 5 to 15 feet thick, many concretions, some lenses of sandstone in valley walls, Phenacodus vort- mani, Hyracotherium vasacciense, Hyracotherium craspedotum in TOWER NOU CCE are renee Nasote: si ensye'e a aie, SPccaieyo!s sora ns era SIM crace toenerae eee 385 feet 5s Goveredy 202.07. 5 <6 Geb se oc. 2 eioic wd wie 6 cinta nae cigereeee 20.4 4. Sandstone, medium to fine grained, tan, poorly sorted, very poorly cemented, grains usually subangular; calcareous ............-...02- 28.0 3. Mudstone, sandy, red-brown and standstone, grey-green in alternating beds; outerop weathersitam, COncretIONS o 0 e 26... ces een. eee see 29.6 2. Claystone, dark red, silty, forms dark red layer in outcrops, some green CIAYStONE SS ERIM ETS. 5. vge oper Ne ciclo. hs auenois ie che Si Syste ge SKopabousYaxoy IIS gs esse, > ono S22 1. Mudstone, dark red-brown, sandy partly covered, concretions ........ 53.9 Total measured thickness 530 feet Poison Canyon formation: Sandstone, yellow, arkosic, locally conglomeratic, grains subangular to subrounded, poorly sorted, silt to small pebble size; mudballs; upper surface undulatory. Structurally conformable with unit 1 above. SECTION 2. GARCIA CANYON Starts in Sec. 34, T26S, R69W and continues into Sec. 3, T27S, R69W. Pediment: Thickness Huerfano formation (in part): feet 18. Mudstone, sandy, and sandstone, silty, tan, brown. some red layers... 56.0 7. Mudstone,;sandyered-bro wit .ij2 ste ahs: «ix, tevae os s.0) oie tere o Mea Sere ele 5.6 16. Mudstone, sandy, light brown, color reddens near top, green stringers OVS 5050) BR oreirtadlS sips 4.0. aot ORICON Gt On Oo Oa CIoiocle on oem ee 50 0c 21.4 15. Mudstone, red-brown, sandy, becomes browner and sandier at top, top tan; grades into\sandstone laterally... <. 25. eeoune oc ee eee ome 16.8 14. Mudstone, sandy, brown, weathers tan, large concretions ............ 9.1 13. Mudstone, red-brown, sandy, small concretions; light green siltstone layer 1.5 feet thick at 11.5 feet above base; unit becomes browner near top, Hyracotherium vasacciense, Hyopsodus walcottianus, crocodile, from neat POttont OF MIMIC. 10:50. 2/spas « Scien eee eels eee eree 30.0 12. Sandstone, coarse grained, conglomeratic, some conglomerate lenses, arkosic, cross-laminated, surface of unit 11 channeled; fragments of turtle, small siltstone lenses in standstone. Color grey to green ........ 47.8 Pie) Mudstone: dark-brown, San yi ois.ciciicseresciseo ce ses ei oeietae C earere 5.1 10. Sandstone, yellow to yellow-green, arkosic, conglomeratic, coarse ..... 5.4 9. Sandstone, silty, light brown, some green mudstone lenses ........... 3.6 U4 APPENDIX Garcia Canyon Section (continued) Hy StH SI CO . Sandstone, light brown, silty, becomes siltier and darker near top .... sp AVSCOME, sDTO WM: SATI CY a¥-tchs0toi0\sict svete sets} aoejtalahisy (0 ahs: sss oie) 289) e's) aha l’s eeCOncreron layer ndense iCal CareOuUsy sé ofaysiaps) <1s:ayciatavavain aay o> vm Ce iss dati = Be COVCT CO ots otore bey chan nZ- 0625 ORE Poy h Suess /a1-yoheyeelssaic.e Al'siois aidiehsekchoPauaioim eee ote cd ve SOR WeY0 |) hath Aa HUA ath Se re BAG A i AD St i A UA aa So ee be . Sandstone, silty, brown, medium grained, arkosic jointed, weathers as HALTS TOUNG ISI DOWIGETSH ayers seicis so oi cies ocd od aeetetefe Ee atle eteles Se . Mudstone, brown, sandy; concretions, concretions white, sandy, large, bedi weathersjyellows imap laces va). 0.i geste sae ci- nis eye. sheen eyer leicht ves ays . Sandstone, gray-green, arkosic, calcareous, grains angular and coarse .. Total measured thickness Alluvium SECTION 3. “FOSSIL CREEK” In Sec. 12, T26S, R70W. Pediment: Huerfano formation (in part): 16. 15. 14. es 12. 1 10. 9 Mudstone;dishtoneen, samGiyr jaw.) shi ee eee anteaters erento Mudstone, red-brown, sandy; several concretion beds ............... Mudstone, green-brown, sandy, prominent 0.5 thick concretion layer MLOotraboverbase OL WMI chien ine Saale caps clots hiee ce sateen tale NEWS tone browns SATIGYs « 1 Jala vie) s/el re alate tia 0 wo sr ara otal wae TM Pee MUGSTONe .PLEEM-DLOWM, SATIGY) stirs c)(tta cones ceaie bar tehaiteeae eee aos Minrcstone;: Drow SAN Gy x arses ciet dels vom Nake ins abe ce nee erate Mudstoneeneen- brow .SAMdy: gd, 8 142 cise Na charset ieee tle ai hae Mudstone, red-brown to brown, many concretions, some green clay- stone stringers; grades laterally into white sandstone. This is the bed that produced most of the fossils at locality II. All species reported on faunal lst tom locality ll ypresentimythis bedi i004). tale ae ee eels . Sandstone, gray-green, silty, micaceous, arkosic, coarse to fine grained subangulat grains caleareousy s 5). 5a4.so darn iichs a eebteie ciate etcleee ete e . Sandstone, yellow and white, coarse, silty micaceous, arkosic, yellower bands-are conplomeratic; Harisita faciestseieibaostcaee ile acta . Sandstone, gray-green, silty, arkosic, micaceous; upper part grades into unit 7, thin seams of gypsum cutting across bedding ................ . Mudstone, sandy, red-brown, becomes greenish near top, Notharctus MUNTENUSM eat DasceOlmUnNitas4 4 5c eee Eee . Sandstone, gray-green, silty, micaceous, arkosic; subangular, coarse to finjeserained: (Calcarcouisintsiesisss pacer davasiac Se a ons sivhclc wlaleadhandejer aie eae . Mudstone, dark red-brown, some green stringers, large concretions (NEES UY So es oon dicta BAG ECO CLE OO Cn bOI CIO IGE NOs ede ree eee . Mudstone, brown, micaceous, sandy. Green concretionary layer 1 foot DELOWMEO Pree error teeta acta ctete dia ahah AA ere, jaycis “ievcietsishs, ciniete wiaiow sos . Sandstone, green-gray, micaceous, arkosic, silty, poorly cemented, cal- CALCOUS, CHAINS AME ated. Gel ae exe eo esto ere sie apel aera byes cist whet sts svete Total measured thickness Alluvium 73 5.6 15.8 15.8 308 feet Thickness feet ies 17.8 28.0 9.0 255 feet fics FOSSIL MAMMALIA OF THE HUERFANO FORMATION SECTION 4. HUERFANO RIVER-MUDDY CREEK DIVIDE Bottom of section in Sec. 14, T26S, R70W and continuing into Sec. 16 T26S, R.70 W. This is the area of locality I and the upper part of the type section of the Huerfano formation. The most persistent bed is unit 20 of this section which is probably the same bed as unit 63 of section 5. Unit 16 of section 3 may be the same bed. This is the only marker bed in the Huerfano formation. Most of the fossils from locality I come from below this bed but a few from Sec. 16 come from above it. Pediment Huerfano formation (in part): Thickness feet Sie Mardstone: rem, silty 0). scree ccets eee: cigs eve ee repeat eae eee 10+ 30. Sandstone, yellow to tan, conglomeratic, arkosic, coarse, lenses of red BUNIAOLS LOMO? Mea redogruei rail coe ecm cr ssa Lore eho! aio lave aie a a We ole ooltecarevorerevet certo 28.0 29a Sandstone, coarse tO ane orammeds tail «2s oso 0. cjale al -tore 2 ee alee ee 16.8 28..Mudstone, brown, ted .and green, sandy 2.165... 0402 ac) oe = ae eee 31.2 27. Sandstone, yellow, coarse to fine grained, arkosic, layers peel off in WEAtEKIMO” ChOSS-latMaIM Stes tel. ye eieisier ote) eet tel erel sates et) tee 16.6 26. Mudstone, red-brown and green, sandy, concretionary ............-+-. 8.6 25, EONCretiOn layers its eave tro eho he clea 5) eier ieee eae ee arctan 0.5 2A Mardstone, sOrGwil o)cip ieee cis « «See oral ee so gaie elena te eet ee ee 2.0 23a CONGCKCHOM lay ell mee ecie cou oe leis ecehe sess sie ae eee ain eee 0.7 22. Sandstone, tan, coarse, arkosic, cross-laminated, silty near top ......... 10.1 21.) Mudstone! (red-brown; SaniGy fac). oi sy 20's eves sta sees ee eis are sere ee 2:2 20s Mudstone, Mehtereen, weatiens Wimtle sc. ese ee a> om ire eee 124 19. Mudstone, red-brown, mottled with green .........0.000ccccccceeees 5a 18. Mudstone, red-brown, sandy, some green stringers, concretions ........ 22.4 17. Sandstone, yellow to tan, poorly sorted, arkosic, silty, conglomeratic .... 4.0 16. Mudstone, dark red-brown, color changes at top to dark brown; con- GEECIOMARY, yo isyays a yopouowone ¥ovs) sv oreh =vakoy <¥ spel =jehe//as0.01 shot ed ee heothons) sie eRaete ini tok eee nN 8.4 NSE Madstone, JOTEEN, SATIGY y--ceueryeicr-ass-voncieioie eater hae ete eT 4.6 14> Mudstone, dark red-brown; \concretionaty (.).. p21 bis oats ele 14.7 3-0 Mindstomes stanly SariGy, 0 patay tO) DEOWD: COALS ts ats; crave a1sevaia a hr omiheaicras wis. as ce eae 9.0 AO VEUIGSLONG: TEC DEOWID: (5.3.4 avarevs a/ zie cickars eioials, © badd eat SELES ds ae 7.0 39: Sandstone, tan to brown, coarse conglomeratic .........000.060e-ee00s 6.0 Sear MCIOSEOT CAT ECD LO WIN ra hi 6f oic, o)aheick rahe eriaas Siege iss eS car eames ie 30.3 SU SAMUSLONE «COATSE) CAI AT ROSIC 1% 5 aici craseranc a/c erel otebs: crease oss we 19.0 35; Mudstone, ereen and red, some layers!Sandy) 3.15, j, Mudstone; brows, tsand ya.5.5csaieeess ciety ts, hcl os ae ee eR ere 39.5 Al. sSandsione) tau, CONPIOMELr Atl Cwar ily, a Sf ue Soave oe ee RENE lee ee 59.0 20; Miudstone,; brown; Sand yyy. 2b Based vrei p cals 3 Metacheiromys sp., AMNH 18666, right astragalus from 1 mile north of Gardner Butte, Farisita formation, ventral and dorsal views. x 2. Figure 6. Bathyopsis fissidens, AMNH 17438, a left P, from locality III. x2. Figure 7. Peratherium sp., AMNH field 1952-328, fragment of left jaw with M, from locality If. Figure 8. Coryphodon sp., AMNH 56543, a left P, from locality II. x0.8. Figure 9. Palaeosyops fontinalis, YPM 16463, right M, from locality 5 of R. C. Hills. x2. 86 PEALE It All figures stereopairs Figure 1. Absarokius noctivagus nocerai, new subspecies, Type, AMNH 55215, fragment of left jaw with P.-M, from locality H. x 4.8. Figure 2. Loveina zephyri, AMNH 17554, fragment of left jaw with P,-M, from locality Wie aeee Figure 3. Nyctitherium sp. cf. N. velox. AMNH 55151, fragment of left jaw with M,, from locality III. x 10. 87 PLATE III All figures stereopairs Figure 1. Huerfanius rutherfurdi, new genus, new species, type, AMNH 55216, fragment of left jaw with P,-M, from locality II. x 4.8. Figure 2. Phenacolemur jepseni simpsoni, new subspecies, type, AMNH 2680, fragment of right jaw with P,-M, from the lower faunal zone, probably locality VI. x 4.8. Figure 3. Shoshonius coopert, AMNH 55153, fragment of right jaw with M,, from locality II. x 4.8. Figure 4. Scenopagus edenensis, AMNH 17483, fragment of right jaw with P,-M, from locality III. «4.8. Figure 5. Scenopagus edenensis, type, AMNH 55685, right M,.,, from Morrow Creek Member of Green River formation. 4.8. Unintentionally mounted in reverse. 88 PLATE IV Figure 1. Cynodontomys scottianus, YPM 16468, fragment of right jaw with P,-M, from locality IV. X1.6. Figure 2. Cynodontomys knightensis, AMNH 55225, fragment of right jaw with I, P,-M,; from locality II. x 1.6. Figure 3. Hyracotherium craspedotum, AMNH 17513, fragment of left jaw with M,, from locality VI. x 3.0. Figure 4. Hyracotherium vasacciense ?venticolum, YPM 16436, fragment of right maxilla with DP** from locality IV. x 3.0. Figure 5. Hyracotherium vasacciense vasacciense, YPM 16481, fragment of left maxilla with P*-M? from locality II. x 3.0. Figure 6. Microsyops lundeliusi, YPM 16482, fragment of mandible with right P,-M, from locality II. x 1.6. Figure 7. Hyracotherium vasacciense vasacciense, YPM 16455, fragment of right maxilla with M** from locality VI. x 3. 89 PIEARE, Vi Figure 1. Stylinodon sp, YPM 14616, canine tooth from “South of Hausero’s Ranch.” x 0.8. Figure 2. Vulpavus asius, AMNH 56508, fragment of right jaw with M, from locality II. >Miacis sp. AMNH 56544, fragment of left jaw with M, from locality II. x 1.6. Figures 6, 7. Oxyaena sp. cf. O. lupina, AMNH 55298 from locality VI; fig. 6 right M’, fig. 7 right M.. x0.8. Figure 8. Esthonyx acutidens, AMNH 17531, fragment of left jaw with P,-M, from locality Vix 0:8: Figure 9. ?Oddectes sp.. AMNH 2681, fragment of left jaw with P,-M, from lower faunal zone, probably locality VI. x 1.6. Figure 10. Miacis parvivorus, AMNH 17435, fragment of left maxilla with M? from locality II. x 1.6. Figure 11. Oddectes herpestoides, AMNH 56504, fragment of left maxilla with parts of P®-M? from locality III. x 1.6. Figure 12. Viverravus sicarius, AMNH 56511, fragment of right maxilla with parts of P?-M* from locality II. x 1.6. 90 PAGES VI Figure 1. Sinopa sp. cf. S. strenua, AMNH 17421, right jaw with M,, from locality II. x 0.8. Figure 2. Patriofelis ulta, AMNH 17017 (Type of P. compressa), left jaw with C, P.-M, from locality III. 0.8. Figure 3. Mesonyx obtusidens, AMNH 55272, fragment of right maxilla with P*M? from locality II. x0.8. Figure 4. Mesonyx obtusidens, AMNH 17423, fragment of left jaw with parts of P,-M, from locality III. x0.8. Figure 5. Sinopa sp. cf. S. strenua, AMNH 17421, fragment of left maxilla with M*? from locality II. x0.8. 91 PLATE VII Figures 1, 7. Didymictis ?protenus, YPM 16458, fragment of right jaw with M,. from locality VIII, occlusal and buccal views. x 0.8. Figures 2, 4. Didymictis vancleveae, type, AMNH 17424, fragment of left jaw with P,-M, from locality I, occlusal and buccal views. x 0.8. Figures 3, 5. Didymictis altidens, AMNH 2677, right M,_. from the lower faunal zone, probably locality VI, occlusal and buccal views. x 0.8. Figure 6. Didymictis vancleveae, new species, AMNH 17030, fragment of skull with left P*-M? and right M** from locality I, Palatal view. x 0.8. 92 PLATE VIII Figure 1. Palaeosyops fontinalis, YPM 16450, fragment of left maxilla with C-M?® from locality II. x0.8. Figure 2. Trogosus grangert, YPM 16449, fragment of right jaw with I, P,-M, from locality II. x0.8. Figure 3. Notharctus nunienus, AMNH 55224, fragment of right jaw with P,-M, from locality VI. x 1.6. 93 PEATE 1X Figure 1. Palaeosyops fontinalis, YPM 16451, fragment of left jaw with DP, and part of DP, from “South of Hausero’s Ranch.” 1.5. Figure 2. Eotitanops borealis, AMNH 17441, right M, from locality VII. x 1.5. Figure 3. Palaeosyops fontinalis, YPM 16451, fragment of left maxilla with DP** and part of M’* from “South of Hausero’s Ranch.” 1.5. Figure 4. Lambdotherium popoagicum, YPM 16466, fragment of right maxilla with M** from locality IV. x 1.5. Figure 5. Helaletes nanus, type, YPM 11807, fragment of right jaw with P,-M, from the Bridger formation. 2.0. Figure 6. Hyrachyus modestus, AMNH 55261, fragment of left jaw with DP,.,, M, from locality II. x 2.0. Figure 7. Eotitanops minimus, type, AMNH 17439, left P,-M, from locality II. 2.0. Figure 8. Helaletes sp. cf. H. nanus, YPM 16457, fragment of right jaw with parts of P,-M, from locality 5 of R. C. Hills. 94 READE XxX Figures 1-4 are stereopairs Figure 1. Antiacodon pygmaeus huerfanensis, new subspecies, AMNH 55202, fragment of left jaw with M,_, from locality II. x 1.6. Figure 2. Antiacodon pygmaeus huerfanensis, new subspecies, type, AMNH 17490, frag- ment of right jaw with P,-M, from locality II. x 1.6. Figure 3. Diacodexis sp. cf. D. secans, AMNH 17560, fragment of right jaw with P,-M, from Farisita formation, 1 mile north of Gardner Butte. x 1.6. Figure 4. Diacodexis sp. cf. D. chacensis, AMNH 17552, fragment of left jaw with P,-M, from locality VI. x 1.6. Figure 5. Hyopsodus wortmani, YPM 14612, fragment of left jaw with M,., from locality eS uG: Figure 6. Phenacodus vortmant, AMNH 55230, fragment of right jaw with P,-M, from locality IX. x0.8. Figure 7. Hyopsodus walcottianus, YPM 16446, fragment of left jaw with M,, from locality IV. x 1.6. Figure 8. Bunophorus macropternus, YPM 16475, fragment of right jaw with P,-M, from locality II. 0.8. Figure 9. Hyopsodus paulus, YPM 16435, fragment of right jaw with M, from locality III. LG: Figure 10. Bunophorus macropternus, YPM 16472, fragment of left maxilla with M*™ from locality II. x 0.8. Figure 11. Bunophorus sp. cf. B. macropternus, AMNH 17561, fragment of left jaw with M.., from Farisita formation, east side of Williams Creek, 5 miles north of Gardner Butte. <0.8. 95 a | 2 ¥ 4 th mr | 7 7 7 . 7 ee re iy, - ‘nt ney i i ’ f Wier agile eryiis uae wae oun’ hile maar mo > ai! Pi av hy , : mes mi ig (h tong AbD wey ra i bua! aay. ee ne er uA oo i me | oy.t, M, ip: cae Miah a , 7 a i ee ia 4 i a +i aah ‘ \ ¢ ee a i iy hy ont a , peers i Ve vane it i fi a ; var nT 7 4 1 os 7} ‘ F i 5 ji i - ri 1 i , wa Pe ee bia rr eee ol | ial ing Hie tt ever co - oe + 5 : | hs ve TE WG ss wr; Line 1 yee! iif #1) i fle | Pun oy a vy t We Lae ? oe By Ao. i ay | | ‘Va i yan Ay ttt Laie ms - Hil et, deat ha ir AR othe Ney v nh pte - mt Wikasos ~ wi oan aay Ye he ‘ ‘ 4 F Searinaennes Make panert9 . . tM gah ttt aa) iS . eB eae te