S M I T H S O N I A N C O N T R I B U T I O N S T O P A L E O B I O L O G Y ? N U M B E R 54 The Carnivora of the Edson Local Fauna (Late Hemphillian), Kansas Jessica A. Harrison ISSUED NOV 16 1983 SMITHSONIAN PUBLICATIONS c ^ \ T H S 0 r V 7 ^ ^ NOV 2 11983 1) JJ SMITHSONIAN INSTITUTION PRESS City of Washington 1983 A B S T R A C T Harrison, Jessica A. The Carnivora of the Edson Local Fauna (Late Hem- phillian), Kansas. Smithsonian Contributions to Paleobiology, number 54, 42 pages, 18 figures, 1983.?The late Hemphillian Edson Quarry Local Fauna contains 36 species of amphibians, reptiles, birds, and mammals. The eight species of carnivorans are Cams davisi, a primitive dog; Osteoborus cyonoides, a large borophagine; Agnotherium species, a long-limbed bear; Plesiogulo marshalli, a wolverine; Pliotaxidea nevadensis, a badger; Martinogale alveodens, a skunk; Adel- phatlurus kansensis, a metailurine felid; and Machairodus coloradensis, a machai- rodontine felid. Edson is one of several fossil localities in Sherman County, Kansas, and was deposited in a series of fine sands within the Ogallala Formation. A secondary channel in a braided stream system is proposed as the environment of deposition. The high percentage of juveniles, as well as the vast numbers of the salamander Ambystoma kansensis, indicate accumulation during the spring of the year. The Edson Quarry Local Fauna compares very well with such typically late Hemphillian faunas as Coffee Ranch, Texas, and Optima, Oklahoma. Although only the carnivorans have been treated in depth, a listing of the vertebrate taxa is offered as well. OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. SERIES COVER DESIGN: The trilobite Phacops rana Green. Library of Congress Cataloging in Publication Data Harrison, Jessica A. The Carnivora of the Edson local fauna (late Hemphillian), Kansas. (Smithsonian contributions to paleobiology ; no. 54) Bibliography: p. Supt. of Docs, no.: SI 1.30:54 I. Carnivora, Fossil. 2. Paleontology?Pliocene. 3. Paleontology?Kansas?Sherman County. I. Title. II. Series. QE701.S56 no. 54 [QE882.C15] 560s [569'.74] 83-600029 Contents Page Introduction 1 Previous Work 1 Abbreviations 2 Acknowledgments 3 Location and Stratigraphy 3 Taphonomy 5 Faunal List 7 Biochronology 8 Paleoecology 10 Order CARNIVORA Bowdich, 1821 11 Family CANIDAE Gray, 1821 11 Subfamily CANINAE Gill, 1872 11 Canis davisi Merriam, 1911 11 Subfamily BOROPHAGINAE Simpson, 1945 15 Osteoborus cyonoides (Martin, 1928) 15 Family URSIDAE Gray, 1825 22 Subfamily AGRIOTHERIINAE 22 Agriotherium species 22 Family MUSTELIDAE Swainson, 1835 25 Subfamily GULONINAE Miller, 1912 25 Plesiogulo marshalli (Martin, 1928) 25 Subfamily MELINAE Burmeister, 1850 25 Pliotaxidea nevadensis (Butterworth, 1916) 25 Subfamily MEPHITINAE Gill, 1872 26 Martinogale alveodens Hall, 1930 26 Family FELIDAE Gray, 1821 27 Subfamily MACHAIRODONTINAE Gill, 1872 27 Tribe METAILURINI Beaumont, 1964 27 Adelphailurus kansensis Hibbard, 1934 27 Tribe MACHAIRODONTINI Beaumont, 1964 31 Machairodus coloradensis Cook, 1922 31 Conclusions 37 Literature Cited 39 in The Carnivora of the Edson Local Fauna (Late Hemphillian), Kansas Jessica A. Harrison Introduction Edson Quarry, Kansas, has produced one of the largest and most diverse faunas from the late Hemphillian of North America. As such, it pro? vides an unusually detailed glimpse into this in? terval during which the flora was completing a transition from woodland savanna to grassland, the climate was becoming more xeric, and faunal interchange with Europe, Asia, and South Amer? ica was increasing. Some 36 taxa of vertebrates representing am? phibians, reptiles, birds, and mammals are known from the locality. Eight genera of carnivorans are present in the fauna. Although a complete sys? tematic revision of the Edson Local Fauna is beyond the scope of this paper, a revised faunal list has been compiled. PREVIOUS WORK.?The fossil concentration was discovered in 1924 by H. T. Martin of the Uni? versity of Kansas Museum of Natural History and was subsequently worked by university field parties under the direction of Martin or Claude W. Hibbard into the 1930s. George Sternberg, a professional fossil collector, periodically exca? vated large amounts of material, approximately 80% of the total Edson sample, during the interval Jessica A. Harrison, Department of Paleobiology, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560. from 1933 to 1942. This material was sold to Childs Frick, a Trustee of the American Museum of Natural History, and following his death was bequeathed to that museum and eventually in? corporated along with the rest of the Frick Col? lection into the holdings of the Department of Vertebrate Paleontology. A small number of spec? imens, notably the types of Perognathus dunklei and ?Oryzomys plwcaemcus, were collected in 1935 by David Dunkle for the Museum of Comparative Zoology at Harvard University. Several authors have studied various compo? nents of the Edson fauna. Adams and Martin (1929, 1930) and Taylor (1936, 1941) described the Edson herpetofauna. Tihen (1958, 1962) re? vised much of this early work. The avifauna was dealt with by Wetmore (1937) and Wetmore and Martin (1930). Martin (1928) and Hibbard (1934, 1937, 1939) described several of the smaller mammals and published the first faunal lists. Martinogale, a mephitine mustelid, was named by Hall (1930), and Dunkle (1938) later referred additional material to this genus. Wood (1936) named the rodent, Kansasimys, and recently Wah- lert (1978) has explored its systematic relation? ships. Harrison (1979) named a new lamine camel, Alforjas, and reviewed the wolverine ma? terial (1981, 1982). Prescott (1953) investigated the geology and groundwater resources of Sher? man County. Edson 18.5 Km * Sherman Co. \ 5 5 eo'. r- m' m CC, ' CC f 1 i 1 I f""'' ? S h e r m a n Co. Loc. 4 \ \ 1 \ \ \ \ m E d s o n Q. \ V \ \ \ \ \ \ N I Km \ j Misplaced Q. ? Wallace Co. < ^ West Q. ?\ ? R n i n o H i " Q- Logan Co. FICURE 1.?Geographic location of Edson Quarry and adjacent fossil localities. ABBREVIATIONS.?The following abbreviations are used in this study: F:AM Frick Collection, American Museum of Nat? ural History, New York, New York K U V P University of Kansas Museum of Natural History, Lawrence, Kansas SMM Sternberg Memorial Museum USNM Former United States National Museum, col? lections now in the National Museum of Natural Hislory, Smithsonian Institution, Washington, D.C. N U M B E R 54 ACKNOWLEDGMENTS.?I am grateful to Richard H. Tedford, Department of Vertebrate Paleon? tology, American Museum of Natural History, to Hans-Peter Schultze and Larry Martin, Univer? sity of Kansas Museum of Natural History, and to Richard Zakrzewski, Sternberg Memorial Mu? seum, for the loan of specimens. John Chom unearthed the negative for Figure 2a from the K U V P archives. Debra K. Bennett took the pho? tograph in Figure 3. Earl Manning identified the peccary from Edson Quarry. Henry Galiano was very generous with his knowledge of the metail- urine felids. Richard Tedford offered several use? ful comments pertaining to Canis davisi. Sue Voss was very helpful in translating the work of French paleontologists. This paper represents part of a dissertation submitted to the University of Kan? sas. The project was supported in part by grants from the National Science Foundation and the Geological Society of America. I would especially like to thank Robert J. Emry, Ralph E. Eshelman, Earl E. Manning, and Henry Galiano for the considerable time and thought they expended in the review of this paper. Location and Stratigraphy Edson Quarry is located in the NW 1/4 of the SE 1/4 of Section 25, T 10 S, R 38 W, Sherman County, Kansas (Figure 1). It is one of several fossil localities (Rhino Hill Quarry, West Quarry, Sherman County Locality 4, and Misplaced Quarry) in southwest Sherman County and northeast Wallace County that have produced smaller collections of similar age and composi? tion. Edson Quarry is contained in the Ogallala Formation, which, throughout Sherman County, rests unconformably on the undulatory surface of the Pierre shale. The late Cretaceous Pierre shale, ranging in thickness from 180-275 meters, con? tains numerous thin beds of bentonitic clay rep? resenting altered deposits of volcanic ash (Pres- cott, 1953:25, 64). The Ogallala Formation has been completely eroded in a few small areas of Sherman County, but attains a thickness of up to 90 meters in others. It is composed of a wide variety of fluvial sediments, including gravel, sand, silt, and clay, deposited by streams draining the Rocky Moun? tains. A very hard limestone layer capping the Ogallala sediments in many areas was described by Elias (1931:138, 139) as the "Chlorellopsis lime? stone" or the "Algal limestone." Because the lime? stone contains a high percentage of a fossil alga, Chlorellopsis bradleyi, Elias inferred the presence of a large, shallow lake or system of smaller lakes to provide the lacustrine environment requisite for its deposition. The Ogallala Formation is overlain by a Pleis? tocene loess, the Sanborn Formation, which ranges up to 15 meters (about 50 feet) in thick? ness. Along valley and arroyo floors, Pleistocene alluvial sands and gravels may be found. The nature of the Ogallala deposits in neighboring Decatur, Wallace, Rawlins, and Thomas counties has been described by Elias, 1931 and 1937, Frye, 1945, and Hodson, 1963. Edson Quarry is located in an arroyo draining into the North Fork of the Smoky Hill River. It was last worked by George Sternberg in 1942. The location was reestablished in 1973 through the use of field notes and photographs from the archives of the University of Kansas Museum of Natural History. Legend has it that Sternberg completely excavated the single, large, lenticular bone bed at Edson Quarry (Figure 2a). The approximate extent of the original excavations and the disposition of the backdirt mounds were still discernable in 1978 (Figure 2b); indeed, bone fragments continue to weather from the 40-year- old spoil heaps. Adams and Martin (1929:504) stated that the fossil deposit was "about 20 feet below the level of the prairie." In order to expose a stratigraphic section of the enclosing sediments and to verify, if possible, the vertical position of the fossiliferous layer, two bulldozer trenches were dug, beginning on the arroyo floor that lies some 4 meters (13 feet) below the level of the prairie and continuing down another 4.5 meters. No remains of the original fossil concentration were found, a tribute SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y : ? m U **?< | a l-Jd . - ! & FIGURE 2.?a, Edson Quarry in late 1920's during excavation by H.T Martin and C.J. Hesse; b, Edson Quarry in 1977. to the thoroughness of Sternberg's excavation; however, a few bone fragments were recovered from Trench II at a depth of about 2.2 meters (about 7 feet) or about 6.1 meters (20 feet) below the level of the prairie (Figure 3). The two trenches were about 2.4 meters (8 feet) wide and from 9 to 12 meters (30 to 40 feet) long, with the maximum depth occurring at about midpoint. Trench I was oriented almost N-S, while Trench II ran about N 65? W Stratigraphic sections of both trenches are presented in Figure 4. The sediments are predominantly fine- to me? dium-grained sands that vary in color from light buff to yellow. With the exception of a few rela? tively thin layers, the sands are generally loose and friable. Calcium carbonate impregnated roots and patches of caliche are scattered liberally throughout the section (Figure 5). Primary sedi- NUMBER 54 mentary structures are of a very small scale and subtle nature. Two sequences in the Trench II section show sedimentary structures that enlarge in scale and increase in grain size toward the top. Very small, shallow ripples are overlain by fine laminae and then slightly larger ripples in coarser sand. This sequence is succeeded by another that repeats the progression of small ripples to fine laminae but culminates in festoon cross beds. Miall (1977:1) defines a braided river as "a series of broad, shallow channels and bars, with elevated areas active only during floods, and dry islands." Although in some systems a single, dom? inant channel can be recognized, in others there are several main channels (Rust, 1972:223); moreover, there are several distinct topographic levels. These topographically high channels ex? perience only ephemeral flow at flood stages and are frequently heavily vegetated. They are also characterized by small grain size and repetitive vertical sequences of structures that coarsen up? wards in section. These upward coarsening cycles have been interpreted as due to channel re-occu? pation following avulsion (Eynon and Walker, 1974:67). Bennett (1978) proposed a widespread, braided stream system as part of a depositional model for the strata containing the Rhinoceros Hill Local Fauna (see Figure 1). The sediments at Rhinoc? eros Hill are, with the exception of the diatoma- ceous marl, generally coarser and the primary sedimentary structures are of a much larger scale than those at Edson Quarry. The relatively fine sediments and small-scale primary sedimentary structures at Edson, indicative of a low energy flow regime, could represent a series of deposits in a topographically high channel within the main system of anastomosing channels. The mi? nor amounts of mud renders an overbank or levee interpretation less likely; however, the absence of mud could also be due to scarcity in the source area. Taphonomy "The greatest riddle . . . is when, why, and how did all these assorted creatures, and in such ab- r~T&& ^ ?? '. _ '<,.'. ???>?. ' - . ? " ." '' - * ? ~ ??zSttcfflffflffi ? . ? - ? * ',: FIGURE 3.?View along Trench II looking WNW. solutely countless numbers, get killed . . . and mashed up into this horrific indecency," Sander? son observed (1960:82). The discipline of taphon? omy, through studying the effects of post-mortem and post-depositional events upon skeletal mate? rial, may provide some insight into the problem so melodramatically set forth by Sanderson. Sternberg, who was by far the most active excavator at Edson Quarry, did not record de? tailed field notes or maps. Consequently, no data are available on the orientation of the hundreds of skeletal elements that were recovered. Such data would have been very informative on such subjects as major current direction and flow strength. Very few of the fossil bones were found in articulation with other elements, although size, stage of wear, and ontogenetic age permit several SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y cm r o - 20 40 60 80 -100 - 1 2 0 140 160 -180 - 2 0 0 ?220 - 2 4 0 - 2 6 0 280 L-SOO Trench I &mm-, 3 Slump *2 & Large caliche patch Caliche patches Small cross beds Single, long caliche-impregnated root Coarser grained Small caliche patches Very fine horizontal laminae (=5/cm) Caliche patches j Caliche patches Hard, well-consolidated Bone fragments Very smalt ripples Hard, well-consolidated Small intermittent cross laminae Trench II < & ? Slump Small cross beds Fine laminae Fine laminae Small ripples Small ripples in coarser sand Fine laminae dipping NW Fine laminae dipping SE Caliche patches Very tine laminae ( = 1 2/cm) Small ripples Caliche patches Extremely small ripples FIGURE 4.?Stratigraphic sections of Trenches I and II. Note occurrence of bone fragments at 235 cm in Trench I section. NUMBER 54 , " " > ' . ' ? ? ? V- . ,. ?: FIGURE 5.?One of many calcified roots occurring in Edson Quarry sediments. tentative associations. Two factors that could be responsible for such a high incidence of disarti- culation are a high energy depositional environ? ment or the activity of scavengers. Neither factor is very well supported, since the grain size of the enclosing sediments and the very small scale pri? mary sedimentary structures indicate a low en? ergy environment and very few of the fossil bones exhibit traces of gnawing. A third hypothesis, which accounts for both the low energy depositional environment and the largely dissociated nature of the fossil material, is as follows. A flood, even one of modest propor? tions, would garner the carcasses of animals hav? ing died in the vicinity. Such carcasses, many already partially decayed and disarticulated, could be transported in a stream of fairly low carrying capacity. Transport must have been rel? atively brief since the bones do not appear to be water-worn or tumbled. Narrow, shallow chan? nels would favor the accumulation of skeletal material, whereas the high energy regime of larger channels would result in scattering. Spring is the most common season for flooding in the Great Plains. Ambystomatid salamanders, especially Ambystoma tigrinum mavortium (barred ti? ger salamander), whose range extends over the central and southern Great Plains (Conant, 1975:256), only congregate to breed in ephemeral pools formed by spring rains. The vast numbers of A. kansensis present in the Edson Local Fauna, in addition to the high percentage of mammalian juveniles, as determined by deciduous dentitions and unfused epiphyses, further support the sug? gestion of a spring flood as the concentrating mechanism for the Edson Quarry fossil deposit. Faunal List The following is a list of the 36 vertebrate taxa represented in the Edson Local Fauna. Several discrepancies are apparent between it and the last faunal list to have been published (Hibbard, 1939:460, 461). Within the Amphibia, Tihen (1958:37; 1962:26) synonymized Plioambystoma with Ambystoma and Bufo arenarius with B. hibbardi. Hibbard's faunal lists (1934, 1939) contain the taxon Chelonia, the sea turtle; this has been changed in favor of an undetermined chelydrid. An unidentified snake, represented by two artic? ulated vertebrae, has been added to the Reptilia. Most of the changes have centered around the mammalian taxa. Additions include a talpid, a lagomorph, Agriotherium species, Pliotaxidea neva? densis, a lamine camel, and Texoceras cf. guymonen- sis. The degree of retraction in the nasals of a skull of Aphelops from Edson suggests that A. cf. malachorhinus may be more appropriate than A. cf. mutilus. The equids are represented by Dinohippus interpolatus and a small, very hypsodont, Nannip- pus-like form. Changes involving the Carnivora are discussed in detail within the systematics section of this paper. AMPHIBIA AMBYSTOMIDAE Ambystoma kansensis (Adams, 1929) SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PELOBATIDAE Scaphiopus pliobalrachus Taylor, 1936 BUFONIDAE Bufo hibbardi Taylor, 1936 REPTILIA CHELYDRIDAE Genus and species indet. TESTUDINAE Geochelone species COLUBRIDAE Genus and species indet. AVES COLYMBIDAE Colymbus mgncollis Wetmore, 1937 GRUIDAE Grus nannodes Wetmore and Martin, 1930 SCOLOPACIDAE Genus and species indet. CORVIDAE Genus and species indet. MAMMALIA INSECTIVORA TALPIDAE Genus and species indet. LACOMORPHA Genus and species indet. RODENTIA MYLAGAULIDAE Mylagaulus monodon Cope, 1881 EOMYIDAE Kansasimys dubius Wood, 1936 HETEROMYIDAE Perognalhus dunklei Hibbard, 1939 Prodipodomys kansensis (Hibbard, 1939) CRICETIDAE ? Oryzomys pliocaemcus Hibbard, 1939 Peromyscus martini Hibbard, 1937 CARNIVORA CANIDAE Cams davisi Merriam, 1911 Osleoborus cyonoides (Martin, 1928) URSIDAE Agnothenum species MuSTELIDAE Pleswgulo marshalli (Martin, 1928) Pliolaxidea nevadensis (Butterworth, 1916) Martmogale alveodens Hall, 1930 FELIDAE Adelphailurus kansensis Hibbard, 1934 Machairodus coloradensis Cook, 1922 PERISSODACTYLA EQUIDAE Dinohippus inlerpolalus (Cope, 1893) Neohippanon cf. euryslyle Nanmppus species RHINOCEROTIDAE Aphelops cf. malachorhinus Cope, 1878 Teleoceras fossiger ARTIODACTYLA TAYASSUIDAE Plalygonus rex CAMELIDAE Hemiauchema vera (Matthew, 1924) Atforjas taylon Harrison, 1979 Megatylopus gigas Matthew and Cook, 1909 ANTILOCAPRIDAE Texoceras cf. guymonensis Frick, 1937 Biochronology The age of the Edson Local Fauna has of necessity been determined primarily on the basis of faunal parameters. Throughout the late Neo- gene, the biota of western Kansas was several times disturbed by the deposition of volcanic ash layers. Unfortunately for the paleobiologist, al? though perhaps not so for the then resident biota, no ash falls are recorded in the strata containing Edson Quarry. Hence, its age must rest upon criteria other than radiometric dating. The po? tential contributions of palynology and paleobo- tany are forestalled by the dearth of fossil plant material, although a fossil flora has been de? scribed from nearby Logan County (Chaney and Elias, 1936:23). The Edson Local Fauna has long been held to be a principal correlative of the Coffee Ranch (Hemphill) Local Fauna of Texas (Wood et al., 1941:12). The Coffee Ranch Local Fauna is the principal fauna of the. late Hemphillian North American Provincial Age that extends from late Miocene through early Pliocene, about 10 m.y. to 4 m.y. BP (Berggren and Van Couvering, 1974). Boellstorff (1976, fig. 8) has fission-track dated (glass) a volcanic ash that overlies the Coffee Ranch Local Fauna at 5.3?0.4 m.y. BP. For the same ash Izett (1975:202) has obtained fission- track dates of 4.7?0.8 m.y. BP (glass) and 6.6?0.8 m.y. BP (zircon). Wood et al. (1941:12) recognized the Hem? phillian by "the first appearance of ground sloths, Lutravus, Machairodus, and Taxidea; the last ap? pearance of Aphelops, Blastomeryx, Mylagaulus, Os- NUMBER 54 teoborus, Pliauchenia, Pliohippus, Prosthennops, rhinoc? eroses, Sphenophalos, and Teleoceras." They listed Hypolagus, Megatylopus, Nannippus, and Neohipparion as "characteristic fossils" and Agriotherium, Di- poides, Illingoceros, and Plesiogulo as "index fossils." Late Hemphillian faunas, such as Coffee Ranch, Edson Quarry, and Optima (Guymon), may be distinguished from early Hemphillian faunas, such as the Amebelodon fricki Quarry in Nebraska and the Box T Quarry in Texas, by a reduction in ungulate diversity, by the presence of Pediome- ryx, and by the first appearance of such immigrant taxa as Plesiogulo, Agriotherium, Ochotona, and Prom- imomys (R.H. Tedford, pers. comm., 1976). Al? though the Amebelodon fncki Quarry was originally described as "Kimballian" in age, I herein follow Breyer (1981:1215). In Table 1 the mammalian taxa of the Edson Local Fauna are compared with those of two other late Hemphillian local faunas, the Rhinoc? eros Hill Quarry, which is located in Wallace County about 4 kilometers (2.5 miles) south of Edson Quarry, and the type Coffee Ranch. A high degree of faunal similarity is immediately apparent, especially within the larger mammals. Within the Rodentia Mylagaulus monodon and Kan- sasimys dubius are present at both Edson and Cof? fee Ranch. It is unfortunate that no insectivores, lagomorphs, or rodents have yet been recovered from Rhinoceros Hill Quarry, because it is these smaller animals that are particularly useful in biostratigraphic correlation, especially among neighboring localities. Three, possibly four, of the carnivoran genera are common to all three faunas. Cams davisi was originally described as Leptocyon shermanensis from TABLE 1.?Comparison of mammalian faunas (* = holotype; X = occurrence in fauna) Taxon Soricid gen. and sp. indet. Talpid gen. and sp. indet. Hypolagus sp. Lagomorph gen. and sp. indet. ?Mylodontid gen. and sp. indet. Mylagaulus monodon Kansasimys dubius Spermophilus: sp. Geomyid gen. and sp. indet. Perognathus dunklei Prodipodomys kansensis Heteromyid gen. and sp. indet. Copemys sp. fOryzomys pliocaenicus Peromyscus martini Canis davisi Osleoborus cyonoides Vulpes stenognathus Canid gen. and sp. indet. Agriotherium schneideri Agriotherium sp. Ursid gen. and sp. indet. Edson Quarry X X X * * * * * * * X Coffee Rhino Ranch Hill X X X X X X X X X X X X X X X X X Taxon Plesiogulo marshalli Pliotaxidea nevadensis Marlinogale alveodens Pseudaelurus hibbardi Adelphailurus kansensis Machairodus coloradensis Platybelodon sp. Rhynchothenum sp. Dinohippus inlerpotatus Astrohippus ansae Neohipparion cf. eurystyle Nannippus cf. lentkulare Nannippus sp. Aphelops mutilus Aphelops malachorhmus Teleoceras fossiger Platygonus rex Prosthennops sp. Hemiauchenia vera Alforjas laylori Alforjas sp. Megatylopus gigas Megatylopus matthewi :Tilanotylopus sp. Pediomeryx hemphillensis Texoceras cf. guymonensis Texoceras altidens Texoceras sp. Edson Quarry * X * * X X X X X X X X * X X Coffee Ranch X X * X X X * X X X X X X X * * X Rhino Hill X X X X X X X X X X X X X X X 10 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y Edson. It is listed as Vulpes cf. V. shermanensis at Coffee Ranch (Dalquest, 1969:4, 5) and at Rhi? noceros Hill (Bennett, 1978:15). Osteoborus cyon- oides, in addition to Machairodus coloradensis (listed as M. catocopis by Dalquest (1969:13)), also occur at all three localities. The ursid material from Edson and especially Rhinoceros Hill is fragmen? tary and only marginally diagnostic. Dalquest (1969:9) identified the Coffee Ranch ursid as Indarctos oregonensis, but Schultz and Martin (1975:50) have subsequently referred this mate? rial to Agriotherium species and indicate that it is probably conspecific with the Agriotherium that they described from University of Nebraska State Museum Collecting Locality Sm-101, Sherman County, Nebraska. However, Schultz (1977:75) in a faunal list of Coffee Ranch identifies the ursid as A. schneiden. With regard to the Perissodactyla, Coffee Ranch more closely resembles Rhinoceros Hill Quarry than it does Edson; however, this is prob? ably a reflection of the need for revision in the Edson horses. Dinohippus interpolatus, Neohipparion cf. eurystyle, and Nannippus are common to all three faunas. The two different species of rhinoceros, Aphelops mutilus at Coffee Ranch and Rhinoceros Hill and A. malachorhinus at Edson, may be re? solved at a systematic level. An age assignment of late Hemphillian for the Edson Local Fauna is substantiated by its high degree of faunal similarity to the type Coffee Ranch Local Fauna. The temporal relationship of the Edson Local Fauna to that of Rhinoceros Hill has long been uncertain. Due to the discon? tinuous nature of the enclosing Ogallala sedi? ments and the lack of radiometrically datable strata (i.e., volcanic ashes), the sole basis for determining the relative ages of the two faunas has devolved upon a comparison of the mam? malian components. This problem is com? pounded by the paucity of small mammals in the Rhinoceros Hill fauna. The overall similarity of the larger mammalian taxa supports the hypoth? esis that Edson and Rhinoceros Hill are more or less contemporaneous. The presence of a more derived species of Megatylopus, M. matthewi, and the possible presence of Titanolylopus suggest that Rhinoceros Hill is slightly later Hemphillian than is Edson. However, the occurrence of the Eurasian allochthons, Plesiogulo and Agriotherium, suggests that Edson may be the younger of the two faunas. A more conclusive assessment of the temporal relationship of Edson Quarry and Rhinoceros Hill must await the recovery of small mammals from the latter locality. Paleoecology During the late Clarendonian to late Hem? phillian interval, the flora of the High Plains was evolving from open woodland savanna to pre? dominantly treeless steppe or grassland (Webb, 1977). The Edson Local Fauna is characteristic of the climax of this period. No plant macrofossils or pollen have been recovered from Edson, but Chaney and Elias (1936:23) described a late Mio? cene flora from a diatomaceous marl in nearby Logan County. The Logan County flora is xeric in nature and consists of Celtis kansana, Populus lamottei, Salix coalingensis, Typha lesquereuxi, Ulmus moorei, and Cyperacites species. The elm, U. moorei, most closely resembles U. parvifolia of Asia, which inhabits areas of low rainfall. Chaney and Elias (1936:32) suggest that late Miocene conditions in western Kansas were similar to those of today, and that the mean annual precipitation was ap? proximately 25 inches, some 5 inches more than at present. The fauna of the High Plains was profoundly affected by the transition from savanna to grass? land. A marked decrease in diversity is apparent in Hemphillian faunas as opposed to Barstovian- Clarendonian faunas; most obvious is the paucity of arboreal and browsing species (Gregory, 1971:70). By late Hemphillian time almost all of the large herbivores possessed high-crowned teeth, and a trend for increased hypsodonty may be observed even in the rodents (Webb, 1977; Wilson, 1960). Of the nine ungulate herbivores in the Edson Local Fauna, all but two, Platygonus and Megaty? lopus, are hypsodont. The remaining seven (three horses, one rhinoceros, two camels, and one antil- ocaprid) range from moderately (Alforjas) to ex- NUMBER 54 11 tremely (Nannippus and Texoceras) hypsodont. Of the ungulates, all but Teleoceras are cursorial forms. Among the Carnivora, even the bear (Agriotherium) and the two felids (Adelphailurus and Machairodus) are noted for their exceptionally long limbs. The Edson Local Fauna is typical of an open grassland community dominated by hypsodont, cursorial herbivores, and pursuit-oriented preda? tors. Narrow belts of open forest, consisting mainly of cottonwood, willow, and scrub elm, probably flourished along drainages. These tracts of riparian woodland provided browse and cover for the few brachydont herbivores and those car? nivores commonly associated with forests. The climate, as indicated by the flora and the extensive caliche deposits in the Edson sediments, was xeric although somewhat less so than at present. The only truly mesic elements in the fauna are the crane, Grus nannodes, and the grebe, Colymbus mgricollis. These birds may have been foraging in local pools or stream channels, but, as they are also the most vagile members of the fauna, it is likely that they nested around the fringes of the small lake located at Rhinoceros Hill (Bennett, 1978:46) some three miles (4.8 km) south of Edson Quarry. If the great abundance of the salamander, Ambystoma kansensis, does indeed represent a breeding congregation, it suggests that precipitation, then as now, was probably concen? trated in the spring and secondarily in the early autumn. Order C A R N I V O R A Bowdich, 1821 Family C A N I D A E Gray, 1821 Subfamily CANINAE Gill, 1872 Canis davisi Merriam, 1911 FIGURES 6, 7 Canis davisi Merriam, 1911:242. Leptocyon shermanensis Hibbard, 1937:460 REFERRED SPECIMENS.?KUVP 3608, right ra? mus; K U V P 3280, left P4; K U V P 3281, canine; K U V P 3282, right M1; F:AM 49464, partial palate; F:AM 49470, partial palate, axis, and cervical vertebra; F:AM 49458, left maxilla frag? ment; F:AM 49456, partial skeleton including right and left rami, atlas, axis, cervical vertebrae 3-6, 4 caudal vertebrae, left and right femora, left and right tibiae, left and right fibulae, left and right calcanea, left and right astragali, left and right cuboids, left ectocuneiform, right na? vicular, left and right metatarsi II-V, 6 first phalanges, 3 second phalanges, and 3 third pha? langes; F:AM 49461, right ramus, right calca? neum, partial right scapula, right metatarsi III? V, left metatarsi II?III, partial left metatarsus IV; F:AM 49466, right and left rami; F:AM 49462, left ramus; F:AM 49465, right and left rami; F:AM 49463, partial right ramus; F:AM 49469, partial right ramus; F:AM 49457, right ramal fragment; F:AM 63178, left C1; F:AM 63184, right Ci; F:AM 72826, left radius, distal ulna, right metacarpi II-V, 1 sesamoid, 4 first phalanges, 4 second phalanges, 3 third phalanges; F:AM 72827, left distal humerus; F:AM 63188, left distal radius; F:AM 63185, left metacarpus II; F:AM 72829, left metacarpus V; F:AM 72728, right distal tibia; F:AM 72728A, left distal tibia; F:AM 63177A, left calcaneum; F:AM 72832, right calcaneum; F:AM 72832A, left astragalus; F:AM 63187, right metatarsus II; F:AM 63189, right metatarsus IV; F:AM 72830, right metatar? sus IV; F:AM 63177, right proximal metatarsus IV; F:AM 63186, right metatarsus V; F:AM 72831, right metatarsus V; F:AM 63176, 1 first phalanx, 2 second phalanges, 1 third phalanx; F:AM 63177B, first phalanx; F:AM 63179, first phalanx; F:AM 63181, first phalanx; F:AM 63190, first phalanx. DESCRIPTION.?Hibbard (1937:460) designated a small, slender ramus (KUVP 3608) from Edson Quarry as the type of Leptocyon shermanensis. Ad? ditional material from Edson, a few specimens in the University of Kansas collections, and several more in the Frick Collection (F:AM), consider? ably expand the topotypic sample. Of particular interest is a lower jaw (F:AM 49456) and an associated pair of hindlimbs. The ramus is slightly larger than Hibbard's type, but still slender and FIGURE 6.?Cams davisi: a,b, holotype, K U V P 3608, right ramus, lingual and lateral views; c, F:AM 49456, right ramus, lateral view; d,e, F:AM 49464, partial palate, right lateral and occlusal views;/, F:AM 49456, atlas, dorsal view; g, F:AM 49456, axis, lateral view; h,j, F:AM 27827, dislal left humerus, anterior and posterior views; i, F:AM 72826, right metacarpus II-V, anterior view; k, F:AM 72826, first phalanx, anterior view; /, F:AM 72826, second phalanx, anterior view; m, F:AM 72826, third phalanx, lateral view; n,o, F:AM 72826, left radius, posterior and anterior views (x 1.0). NUMBER 54 FIGURE 7.?Cams davisi: a,b, F:AM 49456, left tibia, anterior and lateral views; c, F:AM 49456, left fibula, lateral view; d,e, F:AM 63177A, left calcaneum, anterior and medial views; flg, F:AM 72832A, left astragalus, anterior and lateral views; h,i, F:AM 49456, right cuboid, anterior and lateral views; j , F:AM 49456, right navicular, proximal view; k, F:AM 49456, right metatarsus I I -V, anterior view; I, F:AM 49456, first phalanx, anterior view; m, F:AM 49456, second phalanx, anterior view; n, F:AM 49456, third phalanx, lateral views (X 1.0). tapering toward the symphysis with a gently curved ventral border. The tooth row is un- crowded, as in the type, with a generous Ci-Pi diastema. The posterior accessory cusps of P3-4 are somewhat stronger in the less worn F:AM 49456. The crown of Mi in the type is badly broken. In other Edson specimens Mi bears a large protoconid separated from the smaller par? aconid by a deep, open notch. The metaconid is well developed. The talonid consists of an ento- 14 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y TABLE 2.?Measurements (cm) of dentition of Cams davisi (O.R. = observed range, X = sample mean, s.d. = standard deviation) TABLE 3.?Measurements (cm) of postcrania of Cams davisi (O.R. = observed range, X = sample mean) Element C1 P1 p2 pa P4 M1 M2 P'-M2 dP 3 dP4 c, Pi P2 Pa P, M, \u M, dP2 dP 3 dP4 length width length width length width length width length width at protocone length width length width length length width at protocone length width length width length width length width length width length width length width length width length width length width length width length width No 2 2 1 1 1 1 2 2 3 3 2 2 3 3 3 3 4 5 4 4 6 6 5 6 O.R. 0.57-0.61 0.37-0.46 0.44 0.26 0.88 0.28 0.95-1.00 0.31-0.33 1.62-1.71 0.70-0.73 1.15 1.36 0.66 1.01 5.93 0.92 0.55 0.71 0.70 0.59-0.64 0.44-0.51 0.37-0.41 0.20-0.25 0.84-0.88 0.29-0.31 0.95-1.03 0.30-0.35 1.03-1.12 0.38-0.49 1.66-1.79 0.60-0.67 0.80-0.90 0.54-0.58 0.35 0.34 0.51 0.14 0.66 0.19 1.08 0.41 X 0.59 0.41 0.97 0.32 1.67 0.71 0.61 0.47 0.39 0.22 0.86 0.30 1.01 0.31 1.07 0.42 1.69 0.65 0.84 0.55 s.d. 0.02 0.02 0.04 0.04 0.04 0.02 0.04 0.01 conid and a slightly larger hypoconid with no connecting crest. A small accessory cusp occurs anterior to the entoconid. The Ma bears a strong anterointernal cingulum. Only one Ma is present in the sample; it is small, single-rooted, and po- Element Atlas Axis Humerus Radius Metacarpus Metacarpus Metacarpus Metacarpus II III IV V First phalanx Second pha anx Third phalanx Tibia Calcaneum Astragalus Cuboid Metatarsus Metatarsus II III Metatarsus IV Metatarsus V First phalanx Second pha anx Third phalanx length width height width at anterior condyles height distal width length distal width length length length length length length length distal width proximodistal proximodistal proximodistal length length length length length length length No 2 2 2 2 2 4 2 1 2 2 2 3 2 1 2 O.R. 2.63 4.80 1.69 2.07 2.53 2.35 11.64 1.78 4.45 4.83 4.75 3.91-3.98 2.02-2.07 1.41-1.43 1.25-1.30 1.66-1.72 3.37-3.68 2.13-2.16 1.40 5.73-6.16 6.32-6.87 6.46-6.51 5.79-6.10 2.41-2.42 1.74 1.07-1.19 X 3.94 2.04 1.42 1.27 1.69 3.52 2.14 5.94 6.59 6.48 5.99 2.41 1.13 sitioned at the base of the ascending ramus. A partial palate (F:AM 49464), bearing right C -M and left P -M , is not directly associated with a lower jaw, but it so obviously complements the lower dentition in size and morphology that there can be little doubt of its belonging to the same species. The canine is long, curved, and sharply pointed. The upper premolars are slender, like the lowers, but without as many accessory cusps. P is small and single-rooted and bears a shallow posterior heel. Diastemata, fore and aft, isolate P , which is larger than P1 and bears a stronger posterior heel. The primary cusp of P3 is slightly higher than that of P2 and its larger posterior heel is in contact with P4. The high paracone and shorter metacone form the slender cutting edge of the upper carnassial. The proto? cone is reduced and anteriorly positioned about NUMBER 54 15 level with the low, sharp parastyle. The paracone of M 1 is larger than the metacone but the dis? crepancy is not so marked as in Vulpes. These cusps are laterally compressed and together form a sharp crest that roughly parallels the external cingulum. Both upper molars bear a marked parastyle, a strong protocone, and a smaller me? taconule. The posterior placement of the broad hypocone results in the curved anterior and pos? terior borders of the crown. M2 is more strongly curved than M1. The limb elements are long, straight, and slen? der. Their morphology is typically canid and suggests an animal of small coyote size. DISCUSSION.?Merriam (1906:5, 6; 1911:242) described Cams davisi from the John Day basin of Oregon. The type specimen, a maxilla fragment bearing right M " , was found as float on the Mascall Formation immediately below the Rat? tlesnake Formation. Thus, the type horizon is in doubt, although Shotwell (1970:73) assigned it definitely to the Rattlesnake. The material of Leptocyon shermanensis from Edson is referrable to Canis davisi (R.H. Tedford and B.E. Taylor, pers. comm., 1977). The upper molars from Edson are very slightly larger than the type of C davisi, but correspond closely in all other characters. The material from the Hemphillian Little Valley fauna of Oregon, which Shotwell (1970:73) re? ferred to C. davisi, is slightly larger than that from Edson. The proportions and curvature of the ramus, as well as the general morphology of the dentition, are similar in the two samples. How? ever, the tooth row is more open, P4 is more slender and blade-like and bears a small para? style, and the metaconid of Mi is not so strong in the Edson specimens. Subfamily BOROPHAGINAE Simpson, 1945 Osteoborus cyonoides (Martin, 1928) FIGURES 8, 9 Hyaenognathus cyonoides Martin, 1928:235. Borophagus cyonoides.?Matthew and Stirton, 1930:173. Osteoborus cyonoides.?Stirton and VanderHoof, 1933; 177. HOLOTYPE.?KUVP 3468, right ramus bearing I1-3, Ci, P2 alveolus, P3-4, M1-2, and M3 alveolus. TYPE-LOCALITY.?Edson Quarry, Sherman County, Kansas. REFERRED SPECIMENS.?KUVP 3470, right partial ramus; F:AM 61640, skull and mandible; F:AM 61641, skull and mandible; F:AM 61642, palate and mandible; F:AM 61643, right maxilla; F:AM 61644, left maxilla; F:AM 61645, partial palate; F:AM 61646, right maxilla; F:AM 61647, left maxilla; F:AM 61648, right partial maxilla; F:AM 61649, partial braincase; F:AM 61650, partial palate; F:AM 61651, right partial ramus; F:AM 61652, right and left rami; F:AM 61653, right and left rami; F:AM 61654, right partial ramus; F:AM 61655, right ramus; F:AM 61656, left Mi; F:AM 61657, left ramus; F:AM 61658, right partial ramus; F:AM 98088, right and left partial rami; F:AM 104719, right M2; F:AM 67646, associated radius, ulna, fibula, partial ti? bia, lumbar vertebra, first phalanx, and third phalanx; F:AM 67647, associated humerus, par? tial ulna, and partial radius; F:AM 67648, hu? merus and partial tibia; F:AM 67649, associated femur head, partial fibula, calcaneum, metapo- dial fragments, 4 first phalanges, and a second phalanx; F:AM 67650, partial humerus; F:AM 67650-A, partial humerus; F:AM 67650-B, par? tial ulna; F:AM 67651, femur; F:AM 67652, metacarpal IV; F:AM 67653, metacarpal IV; F:AM 67654, metatarsal II; F:AM 67656, meta? carpal IV: F:AM 67656-A, metacarpal IV; F:AM 67657, calcaneum; F:AM 67657-A, astragalus; F:AM 67657-B, cuboid; F:AM 67657-C, 2 first phalanges; F:AM 67657-D, first phalanx; F:AM 67658, articulated vertebrae and ribs, including atlas and axis; F:AM 104716, lumbar vertebra; F:AM 104717, partial radius; F:AM 104718, axis. DESCRIPTION.?Although the type of Osteoborus cyonoides, KUVP 3468, is the only member of the hypodigm to have appeared in the literature (Martin, 1928:235), there are also two excellent skulls with associated rami, several unassociated maxillae and rami, and a considerable number of postcranial elements in the American Museum Frick Collections from Edson. Matthew and Stir- 16 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y FIGURE 8?Osteoborus cyonoides, F:AM 61640: a-c, skull, dorsal, lateral, and occlusal views; d.e, rami, lateral and occlusal views (X 0.5). ton (1930:173-178) and Dalquest (1969:5-8) have described in detail the osteology of 0. cyon? oides based upon the large sample of referred material from the Coffee Ranch Local Fauna in Texas. The following description augments the work of Matthew and Stirton, and Dalquest with data on material from the type-locality, in addi? tion to providing a populational framework against which to compare the type ramus. F:AM 61640 is the skull of a fully mature NUMBER 54 17 individual, probably a female. The M 2 is in place and exhibits moderate wear. Both zygomatic arches are incomplete and the sagittal crest, the nasals, and a few of the anterior teeth are broken. With these exceptions, the skull is undamaged and undistorted. The remaining skull, F:AM 61641, is from a very old individual, probably a male. It also is missing portions of the zygomatic arches and the sagittal and lambdoidal crests, and several anterior teeth. The tympanic bullae are broken and the skull has been crushed lat? erally, thereby exaggerating the frontal bulge characteristic of Osteoborus and Borophagus. Both crania agree with the descriptions of Mat? thew and Stirton (1930:174, 175) and Dalquest (1969:5-7), but the Edson Quarry material is slightly smaller than that of Coffee Ranch. The Edson and Coffee Ranch skulls of Osteoborus share the following features: a shortened muzzle; bulg? ing frontals; broad palate; narrow occipital re? gion; well-rounded, prominent tympanic bullae; strong, sharply pointed postorbital processes; and large, wide glenoid fossae. The high degree of individual variation ob? served in the Coffee Ranch Osteoborus is also quite evident in the Edson Quarry material. The upper premolars are particularly variable, not only in size and shape, but also in position. In F:AM 61641 the left P1"4 are all in alignment, but the right P3 is twisted out of line. In F:AM 61640 P2"3 are so twisted that their long axes are almost perpendicular to that of P1 or P4. All of the 11 P4 's retaining a sufficiently unworn or undam? aged antero-external corner bear a distinct par? astyle. The protocone is very reduced, even in unworn teeth, and is rapidly obliterated with little wear. A prominent cingulum is present on the internal side of the carnassial blade, posterior to the protocone. M1 is very large and robust with heavy, blunt cusps. M2 is reduced in size, but still fully functional. The type of 0. cyonoides is from a very young adult exhibiting little wear on P4 . Mi_2 are in place and followed by the M 3 alveolus. It is one of the smaller individuals in the Edson sample. The premolars are well spaced and in alignment, TABLE 4.?Measurements (cm) of skull and upper dentition of Osteoborus cyonoides (O.R. = observed range, X = sample mean, s.d. = standard deviation) Element Occipital condyle to ante- rior incisor alveolus Posterior palatine process M to anterior incisor al? veolus nimum interorbital con? striction Width Width Width I2 I3 c P1 p2 p3 p4 M M P2 P2 P4- P4 across C ' across M across occipital con- dyles length M1 M2 M1 M2 width length width length width length width length width length width length width at protocone length width length width length length length length No. 2 I 2 3 1 3 3 3 2 2 4 4 3 3 7 7 6 6 6 4 7 7 5 5 6 4 7 4 O.R. 18.26-21.14 10.05 4.68-5.21 7.50-7.85 5.67 3.63-4.16 0.63-0.66 0.57-0.61 0.77-0.82 0.78-0.84 1.11-0.132 0.81-0.92 0.56-0.66 0.41-0.46 0.84-1.00 0.42-0.60 1.06-1.27 0.56-0.75 2.19-2.75 1.18-1.43 1.39-1.61 1.76-2.10 0.83-1.02 1.29-1.42 5.08-6.20 5.58-5.97 3.55-4.16 4.42-4.61 X 19.70 4.49 7.62 3.86 0.65 0.59 0.79 0.81 1.23 0.85 0.61 0.43 0.90 0.50 1.21 0.62 2.40 1.26 1.52 1.85 0.91 1.35 5.50 5.85 3.72 4.49 s.d. 0.09 0.05 0.06 0.06 0.07 0.07 0.20 0.11 0.08 0.11 0.07 0.05 0.37 0.18 0.20 0.08 as opposed to some other Edson specimens in which the premolars are severely twisted (F:AM 61640). A ramal fragment (F:AM 61654) bearing Ci (broken), P2, Mi very closely approximates the type for size, stage of wear, and arrangement of cusps. The majority of the Osteoborus mandibles are deeper and heavier than that of the type. All of the lower incisors in the Edson sample have been worn almost to the base of the crown, with the exception of the type and F:AM 61653, in which 18 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y TABLE 5.?Measurements (cm) of ramus and lower dentition of Osteoborus cyonoides (O.R. = observed range, X = sample mean, s.d. = standard deviation) Element Anterior incisor alveolus to dy mandibular con- e Width of ramus below protoconid of Mi Width of mandibular I, I, I, c, p2 Pa W M Ms M Ci- IV IV condyle \U Mi M2 length width length width length width length width length width length width length width length width at metaconid length width length width length length length No. 2 10 3 2 2 3 3 4 4 4 4 6 6 8 8 10 9 10 10 8 8 2 2 5 9 6 O.R. 13.06-14.42 1.33-1.70 2.98-4.06 0.48-0.49 0.31-0.36 0.53-0.60 0.41-0.45 0.56-0.70 0.49-0.55 1.04-1.17 0.85-0.99 0.66-0.77 0.45-0.58 0.74-1.07 0.50-0.75 1.49-1.83 0.97-1.20 2.45-2.76 1.05-1.25 1.05-1.23 0.76-0.92 0.70-0.77 0.60-0.65 8.00-8.57 3.87-4.51 4.92-5.42 X 13.74 1.52 3.40 0.48 0.33 0.56 0.34 0.61 0.52 1.11 0.92 0.70 0.51 0.92 0.62 1.62 1.06 2.67 1.15 1.15 0.83 0.73 0.62 8.26 4.21 5.20 s.d. 0.10 0.57 0.03 0.06 0.03 0.06 0.07 0.04 0.04 0.10 0.07 0.10 0.08 0.09 0.07 0.06 0.05 0.24 0.20 0.18 they exhibit only moderate wear. The lower in? cisors increase in size from I] through I3, and I3 still bears traces of the lateral accessory cuspules present in the type. The lower premolars are variable in size, position, and alignment. Of the 14 P4's present, all have a stepped posterior face. The Mi is a strong tooth with a large, crushing talonid. Evidence of shear is present only in young individuals, such as the type and F:AM 61654. Of 13 Mi's having an unbroken or sufficiently unworn medial face, all bear a well-developed metaconid. M2 was present in 15 out of 18 rami with wear ranging from slight to extreme. Ms was more rarely retained in the socket, being present in only four rami. All M3 's, however, exhibited wear and, as indicated by the manipulation of associated skulls and mandibles, it is M3 that produces a distinct thegosis on the posterior face of M2. This facet is large and concave posteriorly in the right M2 of F:AM 61641. As in the Coffee Ranch sample of Osteoborus, the Edson postcranial material is disproportion? ately low in relation to cranial remains. In the following paragraphs the Edson Osteoborus postcrania are compared with a sample of Pleis? tocene Canis lupus from the Frick Collection. Of the few vertebrae represented, only F:AM 104718, an axis, can be identified with confi? dence. The dorsal spinous process does not project so far anteriorly in Osteoborus as in Cams, but rather it projects further posteriorly in Osteoborus, well beyond the posterior zygopophyses. The width of the atlanto-articular condyles is compa? rable, but the centrum is much shorter in Osteo? borus. The humerus is represented by one almost complete specimen (F:AM 67647) and three dis? tal halves. A humerus of Canis, with a head of comparable size, is longer and slimmer. The greater tuberculum of Osteoborus is not so large as in Canis, but the deltoid crest and the sigmoid curvature of the shaft are more pronounced. The medial epicondyle and, to a lesser extent, the lateral epicondyle are expanded, thus increasing the transverse width greatly over Cams. The ole? cranon fossa is not so deeply pocketed as in Canis, and a well-developed entepicondylar foramen is present. FIGURE 9.?Osteoborus cyonoides: a,b, F:AM 67647, left hu? merus, anterior and posterior views; c,d, F:AM 67646, left ulna, anterolateral and posteromedial views; ef, F:AM 67646, left radius, anterior and posterior views; g, F:AM 104718, axis, lateral view; h, F:AM 67652, metacarpus IV, anterior view; ij, F:AM 67651, right femur, anterior and posterior views; kj, F:AM 67648, left tibia, anterior and posterior views; m, F:AM 67646, fibula, medial view; n, F:AM 67649, right calcaneum, medial view; 0, F:AM 67657A, right astragalus, anterior view; p, F:AM 67649, metatarsus II, anterior view; q, F:AM 67655A, metatarsus III, anterior view; r, F:AM 67656A, metatarsus IV, anterior view (X 0.5). NUMBER 54 20 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y TABLE 6.?Measurements (cm) of postcrania of Osteoborus cyonoides (O.R. = observed range, X = sample mean, s.d. = standard deviation) Element Axis tip of odontoid process to pos? terior surface of centrum antero-posterior of dorsal spi- nous process width at anterior condyles width at posterior zygopo- physes Humerus length proximal end antero? posterior transverse distal end antero? posterior transverse Radius length proximal end antero? posterior transverse distal end antero? posterior transverse Ulna antero-posterior of proximal surface of olecranon minimum antero-posterior of trochlear notch width below trochlear notch antero-posterior of shaft above lateral styloid process width of shaft above lateral styloid process Metacarpus IV length proximal end antero? posterior transverse distal end transverse at condyles No. 1 1 1 1 1 1 1 4 2 1 2 2 1 1 1 1 2 2 2 2 2 1 2 O.R. X s.d. 4.55 5.50 3.50 3.17 17.18 4.61 3.55 1.35-1.60 1.46 0.12 4.72-4.93 4.82 15.92 1.55-1.56 1.55 2.03-2.06 2.04 2.69 1.72 3.06 2.26 0.95-1.03 0.99 1.41-1.50 1.45 0.82-0.96 0.89 5.15-5.77 5.46 1.16-1.17 1.16 0.96 0.91-1.06 0.98 Element Femur head to distal distal end ;ondyles antero? posterior transverse transverse of patellar surface Tibia distal end Fibula distal end Astragalus length transverse at ti antero? posterior transverse antero? posterior transverse bial condyles minimum transverse constric? tion at neck Calcaneum length distal end Metatarsus II length proximal end distal end Metatarsus III length proximal end distal end Metatarsus IV length proximal end distal end antero? posterior transverse antero? posterior transverse transverse at condyles antero? posterior transverse transverse at condyles antero? posterior transverse transverse at condyles No. 1 1 1 1 2 2 1 1 1 1 1 1 2 2 2 2 2 2 2 2 ?J 2 3 1 1 2 O R . 19.48 3.67 3.95 1.84 1.73-1.80 2.61-3.05 1.30 0.87 2.83 1.68 1.11 4.75 1.64-1.70 1.34-1.49 4.70-4.98 1.26-1.28 0.73-0.78 0.88-0.91 5.50-5.65 1.38-1.40 0.99 0.85-0.89 5.78-5.97 1.33 0.84 0.90-0.94 X s.d. 1.76 2.83 1.67 1.41 4.84 1.27 0.75 0.89 5.57 1.39 0.99 0.87 5.86 0.92 The ulna is represented by one proximal half, one distal half, and one specimen complete but for the olecranon. This element is much longer and thinner in Cams. The rugose fossa just distal to the radial notch is not so deep, and the lateral malleolus is much less elongated in Osteoborus. The radius is represented by one proximal half, one distal half, and one specimen complete but for the disto-medial portion. The proximal out? line of the head is more circular in Osteoborus, and NUMBER 54 21 the neck is more constricted. The radial tuberosity is larger, the shaft is shorter and more curved, and the distal articular surface is less transversely elongated than in Canis. The metacarpus is represented by two meta- carpi IV. For elements with distal condyles of comparable size, the shaft is much shorter in Osteoborus, but of approximately equal girth. The proximal articular surface is squarish, not antero? posteriorly elongated as in Canis. The femur is represented by one fairly complete specimen. The neck is more constricted than in Canis, and the greater trochanter is not so high. The distal condyles are not so antero-posteriorly expanded as in Canis, and the patellar surface does not extend so far up the shaft. As in most postcranial elements, the shaft is proportionately shorter and heavier in Osteoborus. The tibia is represented by two specimens that are missing the proximal end. The distal articular surface is transversely elongated in Osteoborus, par? ticularly on the lateral side. The fibula is repre? sented by a single specimen also missing the proximal end. The two small tubercles on the distal end are less distinct and more divergent than in Canis. The astragalus is represented by only one spec? imen. The head is less laterally elongated, more rounded than in Canis, and the neck is shorter. Two calcanea are present, but both lack the sustentaculum. The distal end is elongated an? tero-posteriorly in Osteoborus, rather than trans? versely as in Canis. The metatarsus is represented by one metatar? sal II, two metatarsi III, and two metatarsi IV. The proximal articular surface of metatarsal II is not so antero-posteriorly elongated as in Cams, and the posterior end is not so sharply pointed. The lateral side of the proximal end is more deeply concave for articulation with metatarsal III in Osteoborus. The proximal articular surface of metatarsal IV is also not as antero-posteriorly elongated as in Canis, nor is the lateral side so deeply excavated. Comparison of Osteoborus cyonoides with a late Pleistocene sample of Canis lupus indicates that the limbs of Osteoborus were shorter than those of a wolf, but were proportionately heavier and more robust. The distal limb elements were not as elongated as in Canis lupus or C. latrans. In Canis the olecranon fossa on the humerus is deeper, permitting greater extension of the elbow joint and hence, the entire forelimb. Moreover, in Cams the patellar surface extends further up the femo? ral shaft and the distal condyles are expanded antero-posteriorly, permitting greater flexion and extension at the knee. These modifications for increased length of stride are absent in Osteoborus, suggesting a less cursorial lifestyle for this genus. No articulated vertebral series or associated fore- and hindlimbs are known for Osteoborus; conse? quently, one can only speculate as to the degree of slope along the spine or the relative lengths of fore- and hindlimbs. However tempting it may be to project the hyaenoid parallels observed in the skull, jaws, and dentition onto the body pro? portions, Munthe (1979:33) maintained that the ratio of functional forelimb to hindlimb length in other borophagines was not indicative of a sloping back. DISCUSSION.?Hyaenognathus pachyodon was de? scribed by Merriam (1903:278) from Kern County, California. Martin (1928:235) referred the Edson Quarry borophagine dog to this genus and erected a new species, H. cyonoides. Two years later Matthew and Stirton (1930:173) transferred it to the genus Borophagus and referred to it the material from Coffee Ranch. Later, Stirton and VanderHoof (1933:175) erected a new genus, Os- teoborus, designating 0. cyonoides as the genotypic species. The systematic history of the Borophaginae in general and of Osteoborus in particular has been rendered complex to the point of confusion by numerous synonymies and revisions. Richey (1979:107) offers a most helpful synopsis of pub? lished borophagine species, their previous taxo? nomic history, and holotypic provenience. Many of the species of Osteoborus are based upon single specimens or very small hypodigms. The Edson Quarry material of 0. cyonoides, together with that of Coffee Ranch, forms one of the largest and most complete samples of any of the species of Osteoborus. The postcrania, unknown for many species, are particularly well represented. As Ri? chey (1979) has demonstrated, many of the spe- 22 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y cies of Osteoborus are separated by little more than temporal/geographic differences. Future investi? gation may well indicate that at least 0. secundus, O. direptor, and 0. cyonoides are conspecific. Family URSIDAE Gray, 1825 Subfamily AGRIOTHERIINAE Agriotherium species FIGURES 10, 11 REFERRED SPECIMENS.?KUPV 3603, femur; F:AM 104723, partial C1; F:AM 68229, partial humerus; F:AM 68230, partial radius; F:AM 49477, partial ulna; F:AM 68231, partial ulna; F:AM 68233, partial metacarpus III; F:AM 104720, partial metacarpus IV; F:AM 104721, partial metapodial. DESCRIPTION.?C1 is large and robust. The root is missing as is a fragment from the postero? internal quadrant near the base of the crown. An antero-internal ridge extends from the base of the crown halfway to the tip. In Agriotherium and Indarctos a second ridge is present on the middle of the posterior face of the crown. Ailuropoda also has two ridges, although the anterior one is much more centrally positioned. If the posterior ridge was ever present in the Edson C , it was obliter? ated by wear, but in all else the tooth compares well with canines in the sample of Agriotherium from the late Hemphillian Old Cabin Quarry, Quiburis Formation, Arizona. No ridges are pre? sent in Arctodus or Ursus. The humerus lacks the proximal third and the internal half of the distal articular surface. The posterior surface of the lateral epicondyle is broad in the Edson Quarry specimen, as in Agriotherium and Ailuropoda. Primitively it is narrower, as in Ursus, Hemicyon, and Ursavus. The radius is missing the distal third. As in Agriotherium, Indarctos, Arctodus, Hemicyon, and pos? sibly Ursavus, the proximal surface is not ex? panded to greatly overhang the shaft, as in Ursus and Ailuropoda. The shaft exhibits relatively little curvature. The tuberosity on the posterior face of the proximal end is enlarged, as in Ailuropoda and Agriotherium. Primitively it is small, as in Hemicyon and Ursus, and possibly Ursavus. Even in such a large form as Arctodus, the tuberosity is smaller in relation to the shaft. The radius appears compa? rable to, but slightly smaller than, the radius from Sherman County, Nebraska, which was referred to Agriotherium by Schultz and Martin (1975:50). Two ulnae of greatly disparate dimensions oc? cur in the Edson Quarry sample. The smaller ulna consists of the central portion of the shaft. The distal end is missing entirely and only a small portion of the proximal articular surface remains. In size and proportions this ulna resembles that of Machairodus, also represented in the Edson Quarry Local Fauna. It may be separated from that of Machairodus by the shape of the shaft approximately one quarter of the length from the distal end. In the bear the shaft is rounded, whereas in the cat it is distinctly triangular in cross section. The larger ulna consists of the prox? imal two-thirds and is missing the coronoid proc? ess. The semilunar notch is shallow and wide. Metacarpal III is missing the distal condyle. It is slender and exhibits little curvature. The prox? imal articular surface forms an elongated triangle. The internal and external surfaces are not deeply excavated. Metacarpal IV also lacks the distal end. The shaft is heavy and elongated with many rugose areas for muscle attachment. The external surface is more deeply concave than the internal surface. The proximal articular surface is smoothly convex antero-posteriorly. The femur is very long and slender. Only a small portion of the internal distal condyle is missing. The head does not extend far above the greater trochanter in Agriotherium, Ailuropoda, and Hemicyon, but does in Arctodus and Ursus. The femur of Ursavus is unknown. The lesser trochan? ter is large in Agriotherium, Ailuropoda, and Hemicyon and reduced in Arctodus and Ursus. The shaft is long, slender, and exhibits a slight sigmoid cur? vature, as in Hemicyon, rather than straight, as in Ursus, Ailuropoda, and Arctodus. This may be due to the greater degree of plantigrady present in the latter three genera. The femur, like the smaller of NUMBER 54 FIGURE 10.?Agriotherium species: a, F:AM 104723, left C1, lateral view; b,e, F:AM 49477, right proximal ulna, posteromedial and anterolateral views; c,d, F:AM 104720, proximal metacarpus IV. lateral and medial views; frg, F:AM 68230, right radius, anterior and posterior views (X 0.5). the two ulnae, may be easily confused with the femur oi Machairodus. The primary differences are (1) the head of the femur is approximately level with the greater trochanter in Machairodus, whereas in the bear, the head extends somewhat above it; (2) the patellar surface is wider in the bear and the proximal border is clearly defined and transverse to the long axis, whereas in Ma? chairodus the patellar surface is narrower and has a poorly defined proximal border. The extremely wide range of variation in size in the Edson ursid material may be attributed to 24 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y FIGURE 11.?Agriotherium species, KUVP 3603, left femur: a, anterior view; b, posterior view X 0.33. sexual dimorphism. Such marked disparity in size between sexes is observable not only in most of the extant bears, but in such ursid fossil assem? blages as those of the Redington locality, Qui? buris Formation, and the Wikieup locality, Big Sandy Formation, Arizona. The humerus, the radius, the smaller ulna, the third metacarpus, and the femur form a homogenous size-grouping and are tentatively identified as female. The larger ulna, the fourth metacarpus, and the distal metapodial fragment may be labeled as male. DISCUSSION.?In the late Miocene the hemi- cyonines, the giant bear-dogs of Eurasia and North America, were well on the way to extinc? tion and were being replaced by members of the Ursidae. Agriotherium is one of the first ursid gen? era to appear in North America, and together with its contemporary, Indarctos, achieved a ho- larctic distribution during the Pliocene. Agriotherium and Indarctos belong to a group of ursids characterized by a short, broad skull and long limbs. Arctodus, largest of the short-faced bears, became widespread in the North and South American Pleistocene, and is currently thought to be most closely related to the extant, South American spectacled bear, Tremarctos. If the pre- masseteric fossa is considered to be a derived character, its presence unites Agriotherium, Arcto? dus, and Tremarctos. The presence of a parastyle on P4 is almost certainly apomorphic for ursids and if it has not been developed independently, its presence unites Agriotherium, Indarctos, and Ail? uropoda. It is interesting to speculate upon the possibility of deriving both Ailuropoda and Tre? marctos from a mid-Pliocene Agriotherium/Indarctos stock. This theory is partially supported by kar- yological studies (Wurster and Benirschke, 1968:373; Wurster, 1968, fig. 1; Todd and Press? man, 1968:105) that indicate that Ailuropoda and Tremarctos, two of the most aberrant and geo? graphically disparate ursids, are closely related. Indarctos may be separated from Agriotherium by virtue of the elongation of the talon on M and absence of the premasseteric fossa. The Edson sample does not contain an M , let alone a ramus; therefore, an identification based upon these characters is not practicable. Size alone is a char? acter of little value due to the considerable over? lap between the two genera; however, metapodial proportions appear to be useful in separating them. Agriotherium is more digitigrade, with longer, more slender metapodials; whereas Indarc- NUMBER 54 25 TABLE 7.?Measurements (cm) oi Agriotherium species Element c length width Radius proximal articul Ulna ar surface antero-posterior transverse antero-posterior at level of radial notch antero-posterior below radial notch width below radial notch Metacarpus III proximal end Metacarpus IV proximal end Femur length proximal end distal end antero-posterior transverse antero-posterior transverse transverse transverse width at patellar surface Measurements F:AM 104273 3.25 2.37 F:AM 68230 3.43 4.40 F:AM 49477 5.10 8.33 3.42 F:AM 68233 3.06 1.80 F:AM 104720 3.90 2.58 K U V P 3603 41.24 9.80 8.07 3.98 F:AM 68231 2.43 tos is more plantigrade, with shorter, more mas? sive metapodials. The Box T, Pit 1 locality, Hig- gins area, Lipscomb County, Texas, has pro? duced, in addition to an Indarctos M and assorted postcrania, a fourth metacarpus. The proximal end of this bone is very close in size to that of the Edson Quarry metacarpus IV; however, the Ed? son Quarry specimen is longer by the length of the distal condyle. Based upon the metapodial proportions and the considerable resemblance to the Redington, Arizona, material, I have referred the Edson ursid specimens to Agriotherium. Family MUSTELIDAE Swainson, 1835 Subfamily GULONINAE Miller, 1912 Plesiogulo marshalli (Martin, 1928) Brachypsalis marshalli Martin, 1928:233. Plesiogulo marshalli.?Hibbard, 1934:247. HOLOTYPE.?KUVP 3464, right ramus with P3^, Mi and alveoli of I1-P2 and Mi, and skull fragments bearing the glenoid fossae. TYPE-LOCALITY.?Edson Quarry, Sherman County, Kansas. REFERRED SPECIMENS.?KUVP 3465, right P M1; KUVP 3467, right ramus with dP3-4, Mi_2; KUVP 3606, right ramus with Ci, dP4 ,Mi; F:AM 49479, mandible with right C1-P2, dP4, P4(germ), Mi and left C1-P2, dP3; F:AM 104724, left d ; F:AM 67650-A, right partial humerus. DISCUSSION.?In addition to the holotype oi P. marshalli, Edson Quarry has produced three de? ciduous dentitions representing the only known juveniles of Plesiogulo. The very similar stage of wear in the immature rami suggests that they were possible litter mates. Detailed description, discussion, and illustration of the Edson Quarry wolverine is contained in Harrison (1981) and so is not repeated herein. Subfamily MELINAE Burmeister, 1850 Pliotaxidea nevadensis (Butterworth, 1916) FIGURE 12a Taxidea nevadensis Butterworth, 1916:21. Pliotaxidea nevadensis.?Hall, 1944:11 26 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y FIGURE 12.?Pliotaxidea nevadensis: a, SMM 13979-1, right ramal fragment bearing M2, occlusal view. Martmogale alveo? dens: b, K U V P 3922, left ramal fragment bearing Ci,P2,3, lateral view; c, holotype, K U V P 3473, right ramal fragment bearing P3 (broken), P4-M,, lateral view; d, K U V P 3833, right ramus bearing Ci,P4-Mi, lateral view (X 3.0). REFERRED SPECIMENS.?SMM 13979-1, right M2. DESCRIPTION.?The single M2 is in place in a fragment of the right dentary bearing the poste? rior portion of the alveolus of Mi and a small piece of the coronoid process. The occlusal surface exhibits very little wear. The protoconid and hypoconid are the largest of the five cusps. The metaconid and entoconid are somewhat smaller and weakly separated. A small but distinct acces? sory cusp is present on the labial edge of the crown. Dimensions of the tooth are as follows: length = 4.8 mm; width = 5.2 mm. DISCUSSION.?The M2 in Pliotaxidea is far from diagnostic. However, the Edson Quarry specimen agrees more closely with the illustrations and description off. nevadensis (Butterworth, 1916:21; Hall, 1944:11) than with those of P. garben (Wag? ner, 1976:112). In particular, Wagner describes the M2 of P garberi as bearing four tubercles, the largest being posterior and labial, as opposed to five in the Edson Quarry specimen, with the largest being lingual. Subfamily MEPHITINAE Gill, 1872 Martinogale alveodens Hall, 1930 FIGURE 12A Martmogale alveodens Hall, 1930:147. HOLOTYPE.?KUVP 3473, right partial ramus bearing P3 (broken), P4-M1. TYPE-LOCALITY.?Edson Quarry, Sherman County, Kansas. REFERRED SPECIMENS.?KUVP 3583, right P3; KUVP 3833, right ramus bearing Ci, P4-M1; KUVP 3922, left partial ramus bearing Ci, P2-3. DESCRIPTION.?The type ramus and most of the referred specimens have been described in detail elsewhere (Hall, 1930:148; Dunkle, 1938:181- 184). The sole undescribed specimen from Edson Quarry is a ramal fragment bearing Ci, P2-3 (KUVP 3922). This is the only known P2 of M. alveodens. The tooth is very reduced and closely appressed to Ci. The long axis of P2 diverges from the long axis of the tooth row by approximately 30?. There are no well-defined accessory cusps, although the antero-internal region of the cin? gulum is broadened slightly as in P3 and P4. A derived feature of Mi that was not discussed by either Hall or Dunkle is the presence of two slender accessory rootlets, one directly beneath the protoconid, the other beneath the metaconid. DISCUSSION.?The type-locality of M. alveodens is listed by Hall (1930:147) as "eighteen miles southeast of Goodland, Sherman County, Kan? sas," the location of Edson Quarry. The locality listed by Dunkle (1938:181) is "the SW 1/4 of NUMBER 54 T A B L E 8.?Measurements (cm) of Martinogale alveodens Element Ramus length depth at anterior base of Mi P2 length width P3 length width P4 length width M i length width at metaconid Measu KUVP 3833 20.20 3.20 KUVP 3922 1.96 1.04 KUVP 3922 2.06 1.29 KUVP 3833 2.90 1.67 KUVP 3833 5.27 2.56 rements KUPV 3473 3.50 KUVP 3473 2.90 1.45 KUVP 3473 5.72 2.16 Sec. 26, T 10 S, R 38 W, Sherman County, Kansas," also Edson Quarry. Hall (1930:147) considered Martinogale as a pos? sible ancestor of Mustela. Dunkle (1938:181) in? dicated a greater similarity between Martinogale and Spilogale. I agree with Dunkle as regards the mephitine affinities of Martinogale. Such derived characters as the two accessory rootlets on Mi and the enlarged metaconid support a close re? lationship oi Martinogale to the extant mephitines, Spilogale, Conepatus, and Mephitis. These same two characters serve to differentiate Martinogale from Pliogale, which possesses only an external rootlet and a smaller, more posteriorly placed metaconid. The only other species contained within Mar? tinogale is M. nambiana from New Mexico (Hall, 1930:148). I have examined the type (USNM 1038), which consists of a ramal fragment bearing the posterior alveolus of P2, both alveoli of P3, P4, and the anterior lobe of Mi. The posterior cin? gulum on P4 is expanded into a shelf in M. nambiana, whereas in M. alveodens the posterior cingulum of P4 is narrow. The overall structure of P4 supports the referral of the Edson Quarry material and the New Mexico specimen to sepa? rate species. However, the limited data obtainable from the type of M. nambiana could as readily result in its referral to the genus Pliogale. 27 Family FELIDAE Gray, 1821 Subfamily MACHAIRODONTINAE Gill, 1872 Tribe METAILURINI Beaumont, 1964 Adelphailurus kansensis Hibbard, 1934 FIGURES 13, 14 Adelphailurus kansensis Hibbard, 1934:243. HOLOTYPE.?KUVP 3462, anterior portion of skull with dentition missing only right M1. TYPE-LOCALITY.?Edson Quarry, Sherman County, Kansas. REFERRED SPECIMENS.?F:AM 62224, partial humerus with associated radius and ulna; F:AM 62225, radius; F:AM 62230, partial radius; F:AM 104740, first phalanx; F:AM 104741, first pha? lanx. DESCRIPTION.?The holotype of Adelphailurus kansensis was described in considerable detail by Hibbard (1934:243-246). For the most part I concur with Hibbard; however, I disagree with his description of the upper canine. He main? tained that Adelphailurus possesses only a posterior cutting edge on C1, whereas an anterior cutting edge is indeed present, albeit less pronounced distally than in the proximal half. The referred postcrania were not found in di? rect association with the type dentition. They are distinctly felid in appearance, but fall well below the size range of Machairodus, the only other felid in the Edson Quarry fauna. No postcrania of Adelphailurus have been previously described in the literature. The material is compared to an? other felid of similar size, Puma concolor. Individ? uals whose P ' were of comparable size to those of Adelphailurus were selected for comparison. The humerus is represented by one distal half. In Adelphailurus the distal end is more transversely expanded than in Puma concolor. The medial epi? condyle is more expanded than the lateral epi? condyle. The entepicondylar foramen is longer and wider in P. concolor. Two complete radii and one partial radius are present in the sample. The head is set at a more 28 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y FIGURE 13.?Adelphailurus kansensis, holotype, K U V P 3462, right and left maxillae: a,b, left maxilla, lateral and lingual views; c, articulated maxillae, anterior view; d,e, right maxilla, lateral and lingual views (X 0.5). acute angle in relation to the long axis of the shaft in Adelphailurus. The shaft is straighter in Adelphailurus with less transverse curvature. The distal ulnar articular surface is larger than in P. concolor. The ulna is represented by a single specimen that is missing the distal end and part of the top of the olecranon process. The proximolateral por? tion of the trochlear notch extends further up onto the olecranon process in Adelphailurus. The interosseous crest is quite pronounced in Adel? phailurus. Several phalanges are present. The shaft is more curved in Adelphailurus than in the puma, and the distal condyles are longer antero-poste? riorly and more deeply separated. DISCUSSION.?Hibbard (1934:246) made note of some differences between Metailurus Zdansky from China and Adelphailurus. These two genera are, nevertheless, closely related and belong to a group of predominantly Eurasian felids that have been designated the Metailurini (Beaumont, 1964; Berta and Galiano, 1983) or the Metailu- rinae (Crusafont-Pairo and Aguirre, 1972) (Table 10). Beaumont placed the Metailurini within the Felinae, but Crusafont-Pairo and Aguirre felt that the group occupied a position intermediate between the Felinae and the Machairodontinae and, hence, warranted its own subfamily. Berta and Galiano placed the group within the Ma? chairodontinae. The Metailurini are the least specialized tribe of the Machairodontinae. The upper canine is long and laterally compressed, but not to the extent observed in the Machairodontini or the Smilodontini. The same relative degree of devel? opment applies to such characters as upper incisor enlargement, muzzle procumbency, accessory NUMBER 54 29 FIGURE 14.?Adelphailurus kansensis: a,b, F:AM 10741, first phalanx, anterior and posterior views; c,d, F:AM 62224, right ulna, anterolateral and posteromedial views; e,f F:AM 62224, right distal humerus, anterior and posterior views; g,h, F:AM 62225, left radius, anterior and posterior views (X 0.5). cusps on P3-4, and coronoid process reduction. These primitive expressions of machairodontine characters in the Metailurini were misinterpreted as intermediate by Crusafont-Pairo and Aguirre and resulted in the invalid subfamily Metailuri- nae. Neither Beaumont nor Berta and Galiano offered a diagnosis for the Metailurini; however, two derived characters delineate this group. The upper canine of metailurines bears an antero? internal groove that ranges from fairly deep in Adelphailurus to shallow in Therailurus piveteaui. Contrary to Hibbard (1934:246), this groove is present in both Metailurus major and M. minor. The groove follows the contour of the anterior cutting edge and is quite distinct from the grooves ob? served in a feline C1. The skull of metailurines lacks the massive occipital region of the machai? rodontine skull. The cranial profile of the metail- urine skull is primitively rounded, curving gently through its apex in the frontals just dorsal to the orbits. In a more derived species of Therailurus, the skull assumes the more flattened profile and massive occiput typical of the Machairodontini. Zdansky (1924:123, 131, 137) described three species from China, Metailurus major, M. minor, and Dinofelis abeli. Kretzoi (1929:1298) made Machi- rodus onentalis (Kittl, 1887:329) the type of Pontos- milus, described a new species, P indicus, and referred Felis ogygia (Kaup, 1832:156), Machairodus hungaricus (Kormos, 1911:182), and Machairodus schlosseri (Weithofer, 1888:233) to this genus. Piv- eteau (1948:104) designated Felis diastemata (Astre, 1929:203) as the type of Therailurus. Ewer (1955:588, 599) described T. piveteaui and trans? ferred Megantereon barlowi (Broom, 1937:757) to T barlowi. Stenailurus teilhardi was described by Cru? safont-Pairo and Aguirre (1972:219). The two species of Metailurus that Zdansky 30 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y TABLE 9.?Measurements (cm) of Adelphailurus kansensis Element Humerus distal end Radius length proximal end distal end Ulna antero-posterior a trochlear nc tch minimum antero? posterior transverse maximum transverse of articular surface antero-posterior transverse antero-posterior transverse middle of antero-posterior just distal to trochlear notch transverse just distal to trochlear notch First phalanx length proximal end distal end transverse transverse at condyles Measurements F:AM 62224 1.72 6.05 3.98 F:AM 62224 18.81 1.87 2.62 2.35 3.28 F:AM 62224 1.92 2.69 1.17 F:AM 104740 3.79 1.40 1.10 F:AM 62225 19.15 2.00 2.71 2.26 3.32 F:AM 104741 4.12 1.49 1.10 F:AM 62230 1.85 2.65 described probably represent two different lines within the metailurines, and M. minor may even? tually be removed to a separate genus (H. Gali? ano, pers. comm., 1982). Subsequent workers have largely ignored the existence of the two species, indiscriminately ascribing to M. major characters peculiar to M. minor. Adelphailurus and Stenailurus retain P and have not developed an anterior accessory cusp on P ; however, they share very elongated upper canines that bear minute serrations. Pontosmilus most closely resembles Adel? phailurus and Stenailurus, but has lost P and has slightly larger upper incisors. Viret (1951:89), Ewer (1955:594), and Piveteau (1961:795) were aware of the close relationship of Metailurus and Therailurus. Metailurus major and T. diastemata are very similar in size; both have lost P and have a strong anterior accessory cusp on P'. Their C is not so elongated as in Adelphailurus, Stenailurus, or Pontosmilus, nor is it serrated. Ther- ailurus barlowi from Sterkfontein and T. piveteaui from Kromdraai are among the youngest as well as the most derived of the metailurines; dates of 2.6 m.y. BP and 1.6 m.y. BP have been applied to their respective type-localities (Szalay and Del- son, 1979:10). These species exhibit a trend to? ward a more typically machairodontine mor? phology: the upper incisor complex is increasingly enlarged and procumbent, the upper canines are long and finely serrated, the occipital region is enlarged, and P accessory cusps are more robust. Beaumont (1964, 1978), Crusafont-Pairo and Aguirre (1972), and Berta and Galiano (1983) have included Dinofelis with the metailurines. Di? nofelis has moderately enlarged upper incisors, elongated C , and a low coronoid process remi? niscent of machairodontines in general; however, its relationships within the group are uncertain. Thenius (1967:134) suggested that Dinofelis may be the senior synonym of Therailurus. In the literature Adelphailurus has consisted solely of the type specimen. However, in addition to the referred material from the type locality, the Frick Collection at the American Museum NUMBER 54 31 TABLE 10.?Previous classifications of the metailurines. Beaumont, 1964 Crusafont-Pairo and Aguirre, 1972 Berta and Galiano, 1983 Family FELIDAE Subfamily FELINAE Tribe METAILURINI Metailurus Therailurus Dinofelis? Family FELIDAE Subfamily METAILURINAE Metailurus Therailurus Dinofelis Stenailurus Family FELIDAE Subfamily MACHAIRODONTINAE Tribe METAILURINI Metailurus Therailurus Dinofelis Stenailurus Adelphailurus Pontosmilus also contains undescribed specimens from Op? tima, Oklahoma, Wikieup, Arizona, and Bone Valley, Florida. The material from Wikieup is particularly interesting in that there appear to be two species represented. Tribe MACHAIRODONTINI Beaumont, 1964 Machairodus coloradensis Cook, 1922 FIGURES 15-18 Machairodus coloradensis Cook, 1922:7. REFERRED SPECIMENS.?F:AM 104732, left pre? maxilla bearing I1"3; F:AM 104731, partial left maxilla with P3; F:AM 104728, left ramus bear? ing I2-Ci, P4; F:AM 104729, partial right ramus bearing P4-M1; F:AM 104730, right Q ; F:AM 104725, right radius, ulna, scapholunar, pisiform, magnum, metacarpi II-V, 5 first phalanges, 2 second phalanges, and 3 third phalanges; F:AM 104726, left tibia, fibula, calcaneum, astragalus, navicular, ectocuneiform, mesocuneiform, meta? tarsi II-V, 4 first phalanges, 2 second phalanges, and 2 third phalanges; F:AM 104727, femur; F:AM 104734, partial ulna; F:AM 104735, par? tial pelvis; F:AM 104736, partial pelvis; F:AM 104733, partial first phalange. DESCRIPTION.?The skull and upper dentition of Machairodus is represented in the Edson Local Fauna only by a partial premaxilla bearing I '" and by a left maxilla fragment bearing P . The upper incisors are uncrowded; I is slightly smaller than I2 and both bear strong lingual cingula. The caniniform I3 is larger still, and laterally compressed with serrated anterior and posterior ridges. The emerging P is large with a high central cusp and a single anterior and two posterior accessory cusps. The lower jaw and dentition are represented by an isolated Ci and two partial rami, one bearing I2-3, Ci, P3 (broken), P4 and the other bearing P4-M1. Both individuals are young adults as indicated by the barely emergent P4. The lateral ramal flanges are smaller and more rounded than in Dinobastis (Meade, 1961, pis. 2, 3) or Homothenum (Ballesio, 1963, fig. 13; De Bonis, 1976:169). The lower incisors are slightly crowded and increase in size from L (alveolus) through I3. Only the tip of Ci has erupted, and it is uncertain as to whether the tooth is rounded in cross section as stated by Cook (1922:8, 25) and by Dalquest (1969:16) or oval as illustrated in Burt (1931, PI. 46) and Martin and Schultz (1975:57, fig. 4A). However, the isolated Ci (F:AM 104730) is typically machairodontine, lat? erally compressed, and bearing strongly serrated anterolingual and posterior cutting edges. The double-rooted P3 is broken at the level of the alveolus. In addition to a high central cusp and one anterior and one posterior accessory cusp, P4 bears a very small cuspule on the midline of the strong posterior cingulum. Mi is long and slender with well-developed shearing blades; no meta? conid is present. In the following descriptions of the postcrania of Machairodus, comparisons were made primarily with material of Nimravides from the early Hem? phillian Sebastin Place (Savage Ranch), Decatur 32 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y FIGURE 15.?Machairodus coloradensis: a,b, F:AM 104729, right ramal fragment bearing P4-Mi, lateral and medial views; c, F:AM 104736, left innominate with acetabulum, lateral view; d,e, F:AM 104728, left ramal fragment bearing I2-C1, P4 (emerging), medial and lateral views;/;?, F:AM 104725, right metacarpus II, anterior and posterior views; h,i, F:AM 104725, right metacarpus III, anterior and posterior views; j,k, F:AM 104725, right metacarpus IV, anterior and posterior views; l,m, F:AM 104725, right metacarpus V, anterior and posterior views (X 0.5). County, Kansas, in the Frick Collection of the American Museum of Natural History. Machai? rodus and Nimravides have not been identified from the same fauna, but both are Hemphillian in age. Therefore, it is important to be able to differen? tiate between these two genera when confronted by remains of a large cat in a fauna that may be either early or late Hemphillian. This particular sample of Nimravides was selected because of its many associated cranial and postcranial elements including one almost complete skeleton (F:AM 104044). NUMBER 54 33 TABLE 11.?Measurements (cm) of dentition and forelimb of Machairodus coloradensis TABLE 12.?Measurements (cm) of hindlimb oi Machairodus coloradensis Element I1 length width I2 length width I3 length I2 length width I3 length P4 length Mi length width Ramus thickness below Mi Ulna distal end Radius length proximal end distal end antero-posterior transverse antero-posterior transverse antero-posterior transverse maximum midshaft diameter minimum midshaft diameter Metacarpus II length proximal end distal end Metacarpus III length proximal end distal end Metacarpus IV length proximal end distal end Metacarpus V length proximal end distal end antero-posterior transverse antero-posterior transverse antero-posterior transverse antero-posterior transverse antero-posterior transverse antero-posterior transverse antero-posterior transverse antero-posterior transverse Measurements F:AM 104732 0.89 0.69 1.10 0.89 1.30 F:AM 104728 0.85 0.76 1.08 F:AM 104729 2.70 2.94 1.27 1.74 F:AM 104725 3.45 3.25 F:AM 104725 29.85 3.20 4.12 4.34 6.10 2.97 2.37 F:AM 104725 10.99 2.93 2.47 2.30 2.46 F:AM 104725 12.08 2.68 2.35 2.35 2.55 F:AM 104725 11.80 2.52 2.18 2.29 2.48 F:AM 104725 9.76 2.42 1.86 2.13 2.11 Element Femur length greater trochanter to head antero-posterior of head distal end antero-posterior transverse transverse of intercondylar notch Tibia length proximal end distal end Calcaneum length maximum antero transverse antero-posterior transverse ?posterior maximum transverse Astragalus length maximum transverse minimum diameter of neck maximum diameter of head Navicular maximum antero -posterior maximum transverse Metatarsus II length proximal end distal end Metarsus III length proximal end distal end Metatarsus IV length proximal end distal end Metatarsus V length proximal end distal end antero-posterior transverse antero-posterior transverse antero-posterior transverse antero-posterior transverse antero-posterior transverse antero-posterior transverse antero-posterior transverse antero-posterior transverse Measurements F:AM 104727 33.80 8.45 3.89 7.12 6.77 1.74 F:AM 104726 29.50 7.07 3.31 5.36 F:AM 104726 9.72 4.37 4.00 F:AM 104726 4.76 4.86 2.62 3.13 F:AM 104726 3.73 2.95 F:AM 104726 10.42 2.86 1.54 1.85 1.84 F:AM 104726 11.85 3.10 2.26 1.89 2.10 F:AM 104726 11.73 2.52 1.66 1.87 1.78 F:AM 104726 10.36 1.57 1.71 1.61 1.64 34 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y FIGURE 16.?Machairodus coloradensis: a,b, F:AM 104725, right ulna, anterolateral and postero? medial views; c,d, F:AM 104725, right radius, anterior and posterior views; e,f F:AM 104725, right scapholunar, proximal and distal views; g, F:AM 104725, right pisiform, lateral view; h,i, F:AM 104725, right magnum, proximal and distal views; _;', F:AM 104726, first phalanx, anterior view; k, F:AM 104726, second phalanx, anterior view; /, F:AM 104726, third phalanx, lateral view (X 0.5). NUMBER 54 35 FIGURE 17.?Machairodus coloradensis: a,b, F:AM 104727, left femur, anterior and posterior views; c,f F:AM 104726, left tibia, anterior and posterior views; d,e, F:AM 104726, left fibula, lateral and medial views (X 0.5). 36 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y The Edson Quarry sample contains many as? sociated elements of a distal forelimb (F:AM 104725): a radius and ulna with scapholunar, cuneiform (triquetrum), pisiform, trapezium, magnum, metacarpi II-V, and a few phalanges. The shaft of the radius is more curved and the neck more constricted in Machairodus than in Nim? ravides. The attachment surface of the interosseous membrane is much more rugose and extensive in Machairodus. The proximal surface of the head of the radius is more circular in Machairodus and more oval in Nimravides. The two radii are of comparable length, but that of Machairodus is slightly heavier with larger proximal and distal ends. The Edson Quarry ulna is complete but for a portion of the olecranon. It is about the same length as that of Nimravides, but heavier and more triangular in cross section. The radial notch and the distal radial facet are larger in Machairodus, suggesting greater rotational mobility in the an- tebrachium. The styloid process is longer in Ma? chairodus. The radial facet of the scapholunar extends further laterally, reflecting the larger size of the distal radius of Machairodus. The medial surface of the magnum is more deeply excavated in Machairodus. The remaining carpals and meta? carpals are quite similar. The pelvis is represented in the Edson Quarry sample oi Machairodus by two left acetabula, each bearing portions of the three innominate bones. The acetabulum is hemispherical. The femora of Nimravides from the Sebastin Place were both longer than the femur of Machairodus from Edson Quarry, but other differences were subtle and somewhat equivocal. Several associated elements of a distal hindlimb (F:AM 104726) oi Machairo? dus include tibia, fibula, calcaneum, astragalus, navicular, mesocuneiform, ectocuneiform, meta? tarsi II-V, and some phalanges. The elements of the hindlimb oi Machairodus from Edson are con? sistently smaller than those of Nimravides from Sebastin Place, whereas the elements of the fore? limb were of comparable size. The tibia is com? plete and bears a fragment of the proximal fibula on the lateral side of the proximal end. Even though the Edson tibia is shorter than that of Nimravides, the shafts are about equal in girth. The antero-lateral concavity at the proximal end, occupied largely by the M. tibialis anterior, is deeper in Machairodus. The lateral crest extends further distally in Machairodus and the medial malleolus is heavier. The anterior notch just lat? eral to the malleolus is deep in both genera. The fibula is missing all of the distal end; it is slender and straight in Machairodus as in Nimravides. The tarsals and metatarsals do not differ significantly from those of Nimravides. DISCUSSION.?Cope (1887:1019) described the first North American species attributed to Ma? chairodus, M. catacopis, based upon a partial man? dibular symphysis. Cook (1922:7) later described material from Yuma County, Colorado, as Ma? chairodus (Heterofelis) coloradensis. Specimens from the late Hemphillian Coffee Ranch Local Fauna were referred by Burt (1931:262) to M. catocopis. Dalquest (1969:13-25) made extensive compari? sons of additional material from Coffee Ranch, including an almost complete skeleton, with spec? imens of Smilodon californicus Bovard, as well as illustrations from Merriam and Stock (1932). Matthew (1924:148) synonymized M. coloradensis with M. catacopis, but Martin and Schultz (1975:60) present evidence that the type of M. catocopis is referrable to Nimravides, thus resurrect? ing M. coloradensis as a valid species. They refer the material from Coffee Ranch to M. coloradensis, along with the type material from Colorado. Kurten (1963:97) divided machairodontines into two tribes, the Homotheriini and the Smi- lodontini. Churcher (1968:272, 273) summarized the Smilodontini as having "more massive crania, relatively hypsodont and smooth-edged sabres of an oval cross-section, incisive teeth arranged in a transverse row, protoradix on the upper carnas- sial, and a heavier and more massive build," and the Homotheriini as having "lighter crania, less hypsodont and compressed sabres, incisive teeth separated from each other and arranged in an arc, no protoradix on the upper carnassial, both milk and permanent dentitions generally serrated when unworn, and a lighter and more lion-like skeleton." Berta and Galiano (1983) accepted Kurten's Smilodontini, but enlarged the Machai- rodontini (Beaumont, 1964:840) to include Ma- NUMBER 54 37 FIGURE 18.?Machairodus coloradensis, F:AM 104726, left tarsus: a,b, calcaneum, medial and anterior views; c,d, astragalus, anterior and posterior views; e,f ectocuneiform, proximal and distal views; g,h, navicular, proximal and distal views; i, mesocuneiform, proximal view; j , first phalanx, anterior view; k, second phalanx, anterior view; l,m, metatarsus II, anterior and posterior views; n,o, metatarsus III, anterior and posterior views; p,q, metatarsus IV, anterior and posterior views; r,s, metatarsus V, anterior and posterior views (X 0.5). chairodus, Nimravides, Miomachairodus, Homothenum, and Dinobastis. Based upon the almost complete skeleton from Coffee Ranch, Dalquest (1969:25) suggested that Machairodus would have had a large head and stout neck, a slender trunk, long, slender limbs, and relatively large feet and claws. The material from Edson Quarry does not con? tradict Dalquests's conception. Conclusions The 36 taxa of amphibians, reptiles, birds, and mammals comprising the Edson Quarry Local Fauna represent one of the largest and most diverse vertebrate faunas in the North American Hemphillian. The fauna was deposited in a series of fine sands within the Ogallala Formation of Sherman County, Kansas. Analysis of grain size and primary sedimentary structures suggests a low energy environment of deposition, probably that of a high, secondary channel in a braided stream system. The deposit of fossil material may represent the remains of carcasses accumulated, but not necessarily killed, by a spring flood and concentrated in a small channel. Eight carnivorans are present in the Edson 38 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y Quarry Local Fauna. Canis davisi is a primitive dog of medium size. Osteoborus cyonoides, a large borophagine canid, is the most abundant carni- voran in the fauna. One C1 and a few postcrania are referrable to Agriotherium species. Three genera of mustelids are present. Plesiogulo marshalli, a wolverine, is represented by three deciduous den? titions, in addition to the mature type ramus and two almost complete bacula. The two remaining mustelids are Pliotaxidea nevadensis, the rarest mammal in the fauna, and Martinogale alveodens, a small skunk. The two felids at Edson are Adel? phailurus kansensis, a New World member of the Metailurini, and Machairodus coloradensis. The Ed? son sample contains the only known postcranial elements referrable to Adelphailurus. Due to the discontinuous nature of the enclos? ing sediments and the absence of radiometrically datable strata (i.e., volcanic ash layers) and pa? leobotanical material, the age of the Edson Local Fauna, Late Hemphillian, has been determined on the basis of vertebrate faunal parameters. Edson compares well with the Late Hemphillian Coffee Ranch Local Fauna of Texas and with the Optima Local Fauna, a principal correlative. The fauna contains such typically Hemphillian taxa as Megatylopus, Hemiauchenia, Osteoborus, and Ma? chairodus, in addition to the Eurasian immigrants, Plesiogulo and Agriotherium. The Edson Local Fauna is probably contemporaneous with the nearby Rhinoceros Hill Local Fauna. Literature Cited Adams, L.A., and H.T. Martin 1929. A New Urodele from the Lower Pliocene of Kan? sas. American Journal of Science, 17(102):504-520, figures 1-32. 1930. An Addition to the Urodele Fauna of Kansas from the Lower Pliocene. University of Kansas Science Bulletin, 19(14):289-297, plates 30-32. Astre, G. 1929. Sur un Felin a particularites Ursoides des limons Pliocenes du Rousillon. Bulletin de la Socie'te Ge'olo- gique de France, 29:199-204, figures 1-3. Ballesio, R. 1963. Monographic d'un Machairodus du Gisement Vil- lafranchien de Seneze: Homotherium crenatidens Fa- brini. Travaux du Laboratoire de Geologic de la Faculte des Sciences de Lyon, new series, 9:1-129, figures 1-57, plates 1-4. Beaumont, G. de 1964. Remarques sur la classification des Felidae. Eclogae Ge'ologicae Helvetiae, 57:837-845. 1978. Notes complementaires sur quelques Felides (Car? nivores). Archives des Sciences, 31 (3) :219-222, figures 1-2. Bennett, D.K. 1978. Paleontology, Paleoecology, and Biostratigraphy of the Rhinoceros Hill Fauna (Hemiphillian: Lat? est Miocene) and Sedimentology of the Enclosing Ogallala Formation, Ash Hollow Member, Wal? lace County, Kansas. Masters thesis, Department of Geology, University of Kansas, Lawrence, Kan? sas. Berggren, W.A., and J. VanCouvering 1974. The Late Neogene Biostratigraphy, Geochronol- ogy, and Paleoclimatology of the Last 15 Million Years in Marine and Continental Sequences. Pa- laeogeography, Palaeoclimalology, and Palaeoecology, 16(1, 2):1-216, figures 1-15. Berta, A., and H. Galiano In press. Megantereon hesperus (Felidae) from the Late Hemphillian of Florida. Journal of Paleontology. Boellstorff, J . 1976. The Succession of Late Cenozoic Ashes in the Great Plains: A Progress Report. In Guidebook to the 24th Annual Meeting of the Midwestern Friends of the Pleistocene, 37-71, figures 1-8. Breyer, J.A. 1981. The Kimballian Land-Mammal Age: Mene, Mene, Tekel, Upharsin (Dan. 5:25). Journal of Paleontology, 55(6): 1207-1216, figures 1-5. Broom, R. 1937. On Some New Pleistocene Mammals from Lime? stone Caves of the Transvaal. South African Journal of Science, 33:750-768, figures 1-8. Burt, W.H. 1931. Machaerodus catacopis Cope from the Pliocene of Texas. University of California Publications, Bulletin of the Department of Geological Sciences, 20(7):261-292, plates 43-50. Butterworth, E M . 1916. A New Mustelid from the Thousand Creek Pli? ocene of Nevada. University of California Publications, Bulletin of the Department of Geological Sciences, 10(2) :21-24, 1 figure. Chaney, R.W., and M.K. Elias 1936. Late Tertiary Floras from the High Plains, with a Chapter on the Lower Pliocene Vertebrate Fossils from the Ogallala Formation (Lavern Zone) of Beaver County, Oklahoma. Carnegie Institution of Washington Publication, 476(I):l-72, figures 1-10, plates 1-7. Churcher, C.S. 1968. The Affinities of Dinobaslis serus Cope, 1893. Qua- ternana, 8:263-275, 1 figure. Conant, R. 1975. A Field Guide to Reptiles and Amphibians of Eastern and Central North America. 429 pages. Boston: Houghton Mifflin Company. Cook, HJ . 1922. A Pliocene Fauna from Yuma County, Colorado, with Notes on the Closely Related Snake Creek Beds of Nebraska. Proceedings of the Colorado Museum of Natural History, 4(2):3-15, plates 1-14. Cope, E.D. 1887. A Saber-Tooth Tiger from the Loup Fork. American Naturalist, 21:1019-1020. Crusafont-Pairo, M., and E. Aguirre 1972. Stenailurus, felide nouveau, du Turolien d'Espagne. Annales des Paleontologic (Vertebres), 58(2):211-223, figures 1, 2, plate 1. Dalquest, W.W. 1969. Pliocene Carnivores of the Coffee Ranch (Type Hemphill) Local Fauna. Bulletin of the Texas Me? morial Museum, 15:1-43, figures 1-11, plate 1. De Bonis, L. 1976. Un Felide a longues canines de la colline de Perrier (Puy-De-Dome): Ses Rapport avec les felines Ma- 39 40 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y chairodontes. Annales des Paleontologic (Vertebres), 62(2): 159-198, figures 1-13, plate 1. Dunkle, D.H. 1938. A Lower J aw ot Martinogale alveodens Hall. University of Kansas Science Bulletin, 15(8): 181 -185, plate 21. Elias, M.K. 1931. The Geology of Wallace County, Kansas. Bulletin of the Kansas Geological Survey, 18:1-254, figures 1-7, plates 1-42. 1937. Geology of Rawlins and Decatur Counties, with Special Reference to Water Resources. Mineral Resources Circular of the Kansas Geological Survey, 7:1-25, figures 1-4. Ewer, R.F. 1955. The Fossil Carnivores of the Transvaal Caves: Machairodontinae. Proceedings of the Zoological So? ciety of London, 125(3, 4):587-615, figures 1-11, plates 1-3. Eynon, G , and R.G. Walker 1974. Facies Relationships in Pleistocene Outwash Grav? els, Southern Ontario: A Model for Bar Growth in Braided Rivers. Sedimentology, 21:43-70, figures 1-13. Frye, J .C. 1945. Geology and Ground-Water Resources of Thomas County, Kansas. Bulletin of the Kansas Geological Survey, 59:1-110, figures 1-13, plates 1-6. Gregory, J .T . 1971. Speculations on the Significance of Fossil Verte? brates for the Antiquity of the Great Plains of North America. Abhandlungen Hessisches Landesamles fur Bodenforshungen, 60:64-72, figure 1. Hall, E.R. 1930. Three New Genera of Mustelidae from the Later Tertiary of North America. Journal of Mammalogy, 11:146-155, plates 7-8. 1944. A New Genus of American Pliocene Badger with Remarks on the Relationships of Badgers of the Northern Hemisphere. Carnegie Institution of Wash? ington Publication, 551:9-23, figures 1, 2, plates 1, 2. Harrison, J.A. 1979. Revision of the Camelinae (Artiodactyla, Tylo- poda) and Description of the New Genus Alforjas. University of Kansas Paleontological Contributions, 95:1-28, figures 1-3, plates 1-7. 1981. A Review of the Extinct Wolverine, Plesiogulo (Car? nivora: Mustelidae), from North America. Smith? sonian Contributions to Paleobiology, 46:1-27, figures 1-16. 1982. The Baculum of Plesiogulo (Carnivora: Musteli? dae). Journal of Paleontology, 56:1266-1273, plates 1-3. Hibbard, C.W. 1934. Two New Genera of F'elidae from the Middle Pliocene of Kansas. Transactions of the Kansas Acad? emy of Science, 37:239-255, plates 4-6. 1937. Additional Fauna of Edson Quarry of the Middle Pliocene of Kansas. American Midland Naturalist, 18(3):460-464, figures 1-4. 1939. Notes on Additional Fauna of Edson Quarry of the Middle Pliocene of Kansas. Transactions of the Kansas Academy of Science, 42:457-462, figures 1-6. Hodson, W.G. 1963. Geology and Ground-Water Resources of Wallace County, Kansas. Bulletin of the Kansas Geological Survey, 161:1-108, figures 1-22, plates 1-3. Izett, G A . 1975. Late Cenozoic Sedimentation and Deformation in Northern Colorado and Adjoining Areas. In B.M. Curtis, editor, Cenozoic History of the Southern Rocky Mountains. Bulletin of the Geological Society of America, 144:179-209. Kaup, J.J. 1832. Description d'Ossments fossiles de Mammiferes inconnus jusqu'a present qui se trouvent au Mu? seum grand-ducal de Darmstadt. Archiv fur Miner- alogie, Geognosie, Bergbau,undHuttenkunde, 3:150-158. Kittl, E. 1887. Beitrage zur Kenntniss der Fossilen Saugethiere von Maragha in Persien, I: Carnivoren. Annalen des Nalurhistonschen Hofmuseums, 2:317-338, plates 14-18. Kormos, T. 1911. Der Pliozane Knochenfund bei Polgardi. Fbldtam Kozlony,4\(\, 2):171-189, figures 11-19. Kretzoi, N. 1929. Materialien zur Phylogenetischen Klassifikation der Aeluroideen. In 10th International Congress of Zoology, Budapest, (1927), part 2:1293-1355. Kurten, B. 1963. Notes on Some Pleistocene Mammal Migrations from the Palaearctic to the Nearctic. Eiszeitalter und Gegenwart, 14:96-103, figures 1-2. Martin, H.T. 1928. Two New Carnivores from the Pliocene of Kansas. Journal of Mammalogy, 9:233-236, plates 20-21. Martin, L.D., and C B . Schultz 1975. Scimitar-toothed Cats, Machairodus and Nimravides, from the Pliocene of Kansas and Nebraska. Bulletin of the University of Nebraska State Museum, 10(l):55-63, figures 1-5. Matthew, W.D. 1924. Third Contribution to the Snake Creek Fauna. Bulletin of the American Museum of Natural History, 50(2):59-210, figures 1-63. Matthew, W.D., and R.A. Stirton 1930. Osteology and Affinities of Borophagus. University of California Publications, Bulletin of the Department of Geological Sciences, 19(7):71-126, figures 1, 2, plates 21-34. NUMBER 54 41 Meade, G.E. 1961. The Saber-toothed Cat, Dinobastis serus. Bulletin of the Texas Memorial Museum, 2 (2): 23-60, figure 1, plates 1-9. Merriam, J .C. 1903. The Pliocene and Quaternary Canidae of the Great Valley of California. University of California Publications, Bulletin of the Department of Geological Sciences, 3(14):277-290, plates 28-30. 1906. Carnivora from the Tertiary Formations of the John Day Region. University of California Publica? tions, Bulletin of the Department of Geology, 5(1): 1-64, plates 1-6. 1911. Tertiary Mammal Beds of Virgin Valley and Thousand Creek of Northwestern Nevada. Univer? sity of California Publications, Bulletin of the Department of Geology, 6(11): 199-304, plates 32-33. Merriam, J . C , and C. Stock 1932. The Felidae of Rancho La Brea. Carnegie Institution of Washington Publication, 422:1-231, figures 1-152, plates 1-42. Miall, A.D. 1977. A Review of the Braided-River Depositional En? vironment. Earth-Science Review, 13:1-62, figures 1-16. Munthe, L.K. 1979. The Skeleton of the Borophaginae (Carnivora: Canidae): Morphology and Function. Doctoral Dissertation, Department of Paleontology, Uni? versity of California, Berkeley, California. Piveteau, J . 1948. Un Felide du Pliocene du Roussillon. Annales des Paleontologic (Verte'bre's), 34:99-124, figures 1-14. 1961. In Felidae. In Piveteau, editor, Traite de Paleontolo? gy, 1(6):769-797. Paris: Masson. Prescott, G.C., Jr . 1953. Geology and Ground-Water Resources of Sher? man County, Kansas. Bulletin of the Kansas Geolog? ical Survey, 105:1-130, figures 1-12, plates 1-11. Richey, K.A. 1979. Variation and Evolution in the Premolar Teeth of Osteoborus and Borophagus (Canidae). Transactions of the Nebraska Academy of Sciences, 7:105-123, figures 1-11. Rust, B.R. 1972. Structure and Process in a Braided River. Sedimen? tology, 18:221-246, figures 1-14. Sanderson, I T . 1960. Riddle of the Frozen Giants. Saturday Evening Post, January 16:39-83. Schultz, C.B., and L.D. Mart in 1975. Bears (Ursidae) from the Late Cenozoic of Ne? braska. Bulletin of the University of Nebraska State Museum, 10(l):47-54, figures 1-3. Schultz, G.E. 1977. Guidebook for Field Conference on Late Cenozoic Biostratigraphy of the Texas Panhandle and Ad? jacent Oklahoma. Special Publication of the Kilgore Research Center, West Texas Slate University, 1:1-160, figures 1-35. Shotwell, J.A. 1970. Pliocene Mammals of Southeast Oregon and Ad? jacent Idaho. Bulletin of the University of Oregon Museum of Natural Hisotry, 17:1-103, figures 1-43. Stirton, R.A., and V.L. VanderHoof 1933. Osteoborus, a New Genus of Dogs, and Its Relation to Borophagus Cope. University of California Publica? tions, Bulletin of the Department of Geological Sciences, 23(4):175-182, figure 1-3. Szalay, F.S., and E. Delson 1979. Evolutionary History of the Primates. 580 pages. New York: Academic Press. Taylor, E.H. 1936. Una nueva fauna de Batracios Anuros del Plioceno Medio de Kansas. Anales del Inslilucion de Biologia (Mexico), 7(4):513-529, plates 1, 2. 1941. Extinct Toads and Salamanders from Middle Pli? ocene Beds of Wallace and Sherman Counties, Kansas. Bulletin of the Kansas Geological Survey, 38(6): 177-196, figures 1-7. Thenius, E. 1967. Zur Phylogenie der Feliden (Carnivora, Mam? malia). Zeitschnft fur Zoologische Systematik und Evo- lulionsforschung, 5:129-143, figure 1. Tihen, J.A. 1958. Comments on the Osteology and Phylogeny of Ambystomatid Salamanders. Bulletin of the Florida State Museum, 3:1-49, figures 1-11. 1962. The New World Bufonids. American Midland Natu? ralist, 68(1): 1-50, figures 1-62. Todd, N.B., and S.R. Pressman 1968. The Karyotype of the Lesser Panda (Ailurus fulgens) and General Remarks on the Phylogeny and Af? finities of the Panda. Carnivore Genetics Newsletter, 5:105-108, figures 1, 2. Viret, J . 1951. Catalogue critique de la faune des Mammiferes Miocenes de la Grive Saint-Alban (Isere). Nouvelles Archives du Museum d'Histoire Naturelle de Lyon, 3:1-104, figures 1-23, plates 1-4. Wagner, H.M. 1976. A New Species of Pliotaxidea (Mustelidae: Carni? vora) from California. Journal of Paleontology, 50(1): 107-127, figures 1-19. Wahle r t . J .H . 1978. Cranial Foramina and Relationships of the Eo- myoidea (Rodentia, Geomorpha): Skull and Up? per Teeth of Kansasimys. American Museum of Natural History Novilates, 2645:1-16, figures 1-8. 42 SMITHSONIAN CONTRIBUTIONS TO P A L E O B I O L O G Y Webb, S.D. 1977. A History of Savanna Vertebrates in the New World, Part I: North America. Annual Review of Ecology and Systematics, 8:355-380, figures 1, 2. Weithofer, A. 1888. Beitrage zur Kenntniss der Fauna von Pikermi bei Athen. Beitrage zur Paldonlologie Oesterreich-Ungar- ischen, 6(3):225-292, plates 10-19. Wetmore, A. 1937. The Earred Grebe and Other Birds from the Pli? ocene of Kansas. Condor, 39:40. Wetmore, A., and H.T. Martin 1930. A Fossil Crane from the Pliocene of Kansas. Condor, 32:62-63, figures 23-25. Wilson, R.W. 1960. Early Miocene Rodents and Insectivores from Northeastern Colorado. University of Kansas Paleon? tological Contributions, Vertebrata, 7:1-92, figures 1-131. Wood, A 1936. Wood, H J.B. 1941. Wurster, 1968. Wurster, 1968. Zdansky, 1924. A New Rodent from the Pliocene of Kansas. Jour? nal of Paleontology, 10(5):392-394, figures 1, 2. E., W. Chaney, J . Clark, E.H. Colbert, G L . Jepsen, Reeside, Jr., and C. Stock Nomenclature and Correlation of the North Amer? ican Continental Tertiary. Bulletin of the Geological Society of America, 52:1-48, plate 1. D.H. Cytogenetic and Phylogenetic Studies in Carni? vora. In K. Benirschke, editor, Conference on Com? parative Mammalian Cylogenetics, Dartmouth Medical School, 1968, pages 310-329. New York: Springer- Verlag Press. D.H., and K. Benirschke Comparative Cytogenetic Studies in the Order Carnivora. Chromosoma, 24:336-382, figures 1-34. O Jungertiare Carnivoren Chinas. Paleontologica sm- ica, series C, 2(1): 1-155.