From the Thomas Farm Local Fauna (Hemingfordian: Gilchrist County, Florida) + Fxgtwr - W So

Home , Amphicyon, Cynelos, Daphoenus, Euoplocyon, Hemicyon, Hemicyoninae

AMERICAN MUSEUM Novlttates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET NEW YORK, N.Y. 10024 U.S.A. NUMBER 2610 DECEMBER 2, 1976

RICHARD H. TEDFORD AND DAVID FRAILEY Review of Some Carnivora (Mammalia) From the Thomas Farm Local Fauna (Hemingfordian: Gilchrist County, Florida) + fXGTwr - W So

b:: ...... -: 00:" t0'Ty' ''0 ;0 f0\f0X; 0 0' .t

a rt, h $; ;s W0iffi;;21 fit75m;' dS d;

Review of Some Carnivora (Mammalia) From the Thomas Farm Local Fauna (Hemingfordian: Gilchrist County, Florida)

RICHARD H. TEDFORD1 AND DAVID FRAILEY2

ABSTRACT Species of the carnivore genera Cynelos rences of each taxon in North America; all four (Amphicyonidae), Hemicyon (Ursidae), and records constitute the first recognized occurrence Euoplocyon (Canidae) are recognized in the of each in eastern North America. The Thomas medial Hemingfordian Thomas Farm Local Farm Cynelos, Hemicyon and musteline are very Fauna. In addition early Mustelinae are indicated closely related to European representatives of by the ramus referred to Miomustela(?) by Olsen. these taxa from the early Burdigalian, confirming The record of the latter and that of Hemicyon previous correlations of the Burdigalian Age with and Euoplocyon represent the earliest occur- at least part of the Hemingfordian "Age." INTRODUCTION Since its discovery in 1931, the Thomas Farm ACKNOWLEDGMENTS Quarry has remained the most productive Terti- We are grateful to Mr. Henry Galiano, De- ary locality in eastern North America and has partment of Vertebrate Paleontology, the Ameri- yielded the largest sample of Hemingfordian can Museum of Natural History, for discussions vertebrates from this part of the continent. and suggestions during the preparation of the Various taxa of the local fauna have been de- present paper. We further express our gratitude scribed and complete faunal lists have been pub- to Drs. Robert M. Hunt, Jr., Department of Geol- lished (Ray, 1957; Olsen, 1962; Patton, 1967) ogy, University of Nebraska and Thomas H. Pat- that have established the Thomas Farm Local ton, Department of Geology, University of Fauna as a reference point for Miocene faunal Florida for their critical comments on the manu- comparisons between eastern and western North script of this work. Dr. Farish Jenkins lent us America. The recognition of Cynelos (Amphicyo- material from the Thomas Farm Collection of nidae), Hemicyon (Ursidae), and Euoplocyon the Museum of Comparative Zoology, Harvard (Canidae) in the Thomas Farm Local Fauna con- University. Mr. Raymond Gooris skillfully ren- stitutes a refinement in our knowledge of this dered the figures. The manuscript was typed by critical local fauna. Ms. Janice Ebenstein.

'Chairman, Department of Vertebrate Paleontology, the American Museum of Natural History. 2Curatorial Assistant, Department of Vertebrate Paleontology, the American Museum of Natural History.

INSTITUTIONAL ABBREVIATIONS at least three genera of amphicyonids (see Kuss, the type may belong to a species of Cyne- AMNH, Department of Vertebrate Paleontology, 1965), the American Museum of Natural History los Jourdan, 1862, the genus to which we believe F:AM, Frick Collection, the American Museum the Thomas Farm species should be assigned. of Natural History There seems little doubt that Absonodaphoenus MCZ, Museum of Comparative Zoology, Harvard is an amphicyonine rather than an amphicyno- University idontine (as Olsen had concluded) because of its MNHN, Museum National d'Histoire Naturelle simple, transversely elongate, tribosphenic upper molars, presence of an M3, strong hypoconid on M1, and large protoconid on M2. Its small size SYSTEMATICS invites comparison with Cynelos and Pseudarctos FAMILY AMPHICYONIDAE TROUESSART, 1885 Schlosser, 1899. It lacks the diagnostic wide, simple premolars, short trigonid, and broad SUBFAMILY AMPHICYONINAE TROUESSART, 1885 talonid of M1 and M2 of Pseudarctos and agrees GENUS CYNELOS JOURDAN, 1862 with the more primitive Cynelos in having a long, Cynelos caroniavorus (White, 1942) deep jaw with diastemata shorter than half the length of the premolars; elongate slender pre- Daphaenus caroniavorus White, 1942 molars with variable development of a posterior Parictis bathygenus White, 1947 cusp on P4; talonids narrower than trigonids on Absonodaphoenus bathygenus (White, 1942), M1-M2; M2 with short trigonid; M'-M2 relatively Olsen, 1958 short and wide, often with concave anterior and RevisedDiagnosis. About the size of C. schlos- posterior borders, in which the protocone forms seri (Dehm, 1950). Ml triangular, crown con- part of a U-shaped crest with the variable present stricted lingual to paracone and metacone; M2 conules. rectangular approximately 3/4 the width of Ml; In his study of the genus, Kuss (1965) re- M3 oval, approximately half the width of M2, stricted Cynelos to two species groups that range with no distinct cusps. P2-M1 separated by dia- from Aquitanian through Burdigalian time. Later stemata approximately equal to half the length Ginsburg (1966) relegated two Oligocene species of P3 (P4-M, diastema may be absent). Talonid of to the genus, each of which appears to represent M2 very reduced with an indistinct hypoconid a member of the two groups present in the early and no entoconid. Miocene. Cynelos caroniavorus closely resembles Discussion. The species ofD. caroniavorus and the C. rugosidens group and particularly the form P. bathygenus were defined on disparate ele- C. schlosseri (Dehm, 1950), known from Bur- ments (upper molars and a lower jaw fragment digalian deposits in the Loire Basin in France respectively), but Olsen (1958) recognized them (Ginsburg, 1974) and limestone fissure fillings at as belonging to a single taxon for which he in- Wintershof-West, Bavaria (Dehm, 1950). The advertently used the wrong specific name. Both American and European forms resemble one White and Olsen indicated that this taxon was another closely in size, in the degree of reduction distinct generically and Olsen formally recog- of M3, the small size of Ml to Mt, the small size nized this in proposing the generic name Absono- of P4 in relationship to M1, and the lack of de- daphoenus for the Thomas Farm material. In his fined conules on the upper molars (see data in discussion of the relationships of this form, he Kuss, 1965). The Thomas Farm species differs made what appears to us to be a prophetic very little from its European relative; in size it remark in a comparison with the type specimen: falls within the lower range of values for dimen- "The upper molars of A. bathygenus compare sions of the cheek teeth of C. schlosseri, but its well with the general form and structure of those lower premolars are relatively smaller and more found in Amphicyon eseri from the Miocene of widely spaced. Further material is needed before Europe, but are not closely allied to any known geography can be overcome as one of the prime genera from North America." Although the name considerations in separation of the American "Amphicyon eseri" has been used for species of form. 1976 TEDFORD AND FRAILEY: CARNIVORA 3

Hunt (1972) has previously recognized the tory (F:AM 98608, fig. 1). Although unques- presence of Cynelos in North America, specifi- tionably of the same species, minor differences cally the Cynelos lemanensis group, which ranges do exist between this referred specimen and the from late Arikareean through the medial holotype of H. johnhenryi (MCZ 4059), which Hemingfordian. It now appears that species provide an indication of individual variation closely allied to the C. rugosidens group were within this species. Also, P1 is missing in the type also present during part of medial Hemingfordian specimen but present in the referred specimen time in North America. and can be described for the first time, as can the relative placement and sizes of the incisor alveoli. The lower incisor alveoli are crowded between FAMILY URSIDAE GRAY, 1825 the large root of the canine and the very rugose symphyseal suture. They form a triangular con- SUBFAMILY HEMICYONINAE FRICK, 1926 figuration with the 12 root falling behind those of GENUS HEMICYON LARTET, 1851 I1 and I3. The root of I1 indicates that it was smaller than the subequal I2TI3. The P1 of H. Subgenus Phoberocyon (Ginsburg, 1955) johnhenryi is relatively large and bladelike in Hemicyon (Phoberocyon) johnhenryi appearance. The highest point of the crown is in (White, 1947) the anterior part of the enamel crest although no Aelurodon johnhenryi White, 1947 cusp is readily apparent. All the teeth of the referred ramus are smaller and more slender than Revised Diagnosis. Relatively unreduced, nar- those of the holotype (table 1). The alveolus for row premolars with well-developed posterior M3 is also smaller as was presumably M3. The accessory cusps on P2-P4; maximum width of M1 talonid of M1 of the holotype has a small ento- across metaconid equal to or slightly exceeded conid, whereas that of the referred specimen has by width of trigonid of M2; narrow talonids on two equal-sized but small cusps in the position of M1 and M2 with large hypoconids and shallow the entoconid. The protoconid and metaconid of basins; small paraconid present on M2; proto- M2 are lower on the referred specimen than these conid of M2 higher than metaconid; mandible cusps on the holotype, although the protoconid slender. is the larger cusp as in the latter. The referred Description. Hemicyon johnhenryi from ramus appears to be shallower anteriorly than the Thomas Farm was described as a new species of holotype but this may be due only to breakage in Aelurodon (White, 1947) using a ramus with the holotype. In both specimens the anterior lower dentition as the type specimen (MCZ edge of the masseteric fossa is bounded by a low 4059). No upper teeth have been referred to this ridge that passes posteroventrally toward the species. Although the mandible and dentition of edge of the large muscle scar for the masseter H. johnhenryi bear a superficial resemblance to medialis on the angular process. The premasse- that of Aelurodon, numerous features differ- teric fossa is shallow and extends anteriorly to entiate Hemicyon and H. johnhenryi from Aelu- beneath the anterior part of M2, and then for- rodon. These are as follows: (1) The mandible of ward as a shallow groove to the position of the Hemicyon has a premasseteric fossa not found in mental foramina. The referred ramus ofH. john- Aelurodon (or any other nonursid carnivore); (2) henryi has two large mental foramina, one below the lower premolars are much smaller in relation P2 and one below P4. On the holotype ramus to M1 than is seen in most Aelurodon; P4, there are three mental foramina; two smaller especially, is large and robust in Aelurodon in foramina beneath P4 in place of the large single contrast to Hemicyon; (3) M2 is larger, relative to foramen seen on the referred ramus. The condyle M1, than in Aelurodon (in Aelurodon the M2 is of the holotype is much heavier than that of the smaller than P4). referred ramus and the height of the ascending A second specimen of Hemicyon johnhenryi, ramus is less. The overall impression of the holo- a left ramus with P1-M2, is housed in the Frick type is of a relatively heavier, more massive jaw. Collection, the American Museum of Natural His- Discussion. Hemicyon johnhenryi is the 4 AMERICAN MUSEUM NOVITATES NO. 2610

B/I

iiibiiib c %ir"-..

FIG. 1. Hemicyon (Phoberocyon) johnhenryi (White, 1947). Holotype, MCZ 4059, right (reversed, drawn from cast), C, P1 alveolus, P2-P4, M1-M2, M3 alveolus. A. Occlusal view. B. Lateral view. Referred specimen, F:AM 98608, left ramus, C alveolus, P1-P4, M1-M2, M3 alveolus. C. Occlusal view. D. Lateral view. All X1/2. earliest occurring species of Hemicyon in North In Europe where the record of Hemicyon America, and, as might be expected, exhibits extends from the Burdigalian to the Tortonian, numerous features that may be considered primi- species with the suite of morphological char- tive for the genus. These serve well to differ- acters that typify H. johnhenryi are restricted to entiate this species from the other North Ameri- the Burdigalian. These species are the sole repre- can species ofHemicyon whose features are more sentatives of the genus Phoberocyon Ginsburg, characteristic of the genus, i.e., demonstrate a 1955 [P. aurelianensis (Frick, 1926), and P. greater emphasis on a massive, bearlike mandible huerzeleri Ginsburg, 1955] distinguished by its and large, crushing M1-M2. The mandible of author from Hemicyon Lartet, 1851, by the Hemicyon johnhenryi is relatively slender; the following features: P3-P4 with posterior accessory masseteric and premasseteric fossae are the cusps; M1 long and slender, talonid low and shallowest among those known for species of inclined forward; M2 of the same width as MI, Hemicyon. The teeth of H. johnhenryi are among massive, low, with strong trigonid cusps, includ- the largest known for the species of Hemicyon ing a strong paraconid, protoconid more elevated but the mandible is much less deep. The pre- than metaconid, talonid as in M1. As Ginsburg molar series is not so drastically reduced as it is (1955, p. 89) pointed out, the species of in the other species nor are diastemata as Phoberocyon show a curious mixture of archaic apparent between them. The talonids of M1 and and specialized features, however they are most M2 of H. johnhenryi are much less basined than easily distinguished from the species of Hemi- in most of the other species as the large hypo- cyon (and Plithocyon Ginsburg, 1955) by primi- conid occupies most of the talonid. The ento- tive features. Consequently, we have chosen to conid is small and variable in expression. A small regard Phoberocyon as a plesiomorphous sister paraconid is retained on M2. group to Hemicyon and have therefore reduced 1 976 TEDFORD AND FRAILEY: CARNIVORA 5 this taxon to the rank of a subgenus to show this 68254) from the Hemingfordian Blick Local relationship. We are aware of the fact that a Fauna (Gawne, 1975, p. 2) of the Zia Sand group so distinguished may be paraphyletic, but records the presence of a long-footed Hemicyon a far richer record of these rare ursids is needed in New Mexico that is comparable in size with for useful phylogenetic analysis. the limbs expected ofH. (P.) johnhenryi. Hemicyon (P.) johnhenryi is closest in size to H. (P.) aurelianensis, but differs in that the paraconid on M2 is smaller and there is no sharp groove separating it from the metaconid. Un- fortunately, nothing is known of the premolars FAMILY CANIDAE GRAY, 1821 of H. (P.) aurelianensis, but the smaller H. (P.) SUBFAMILY BOROPHAGINAE SIMPSON, 1945 huerzeleri has posterior cusps on P3 and P4 (but not P2). Hemicyon (P.) huerzeleri has an M2 GENUS EUOPLOCYON MATTHEW, 1924 trigonid more nearly resembling H. (P.) john- Euoplocyon spissidens (White, 1947) henryi. Considering the great range in dental measurements shown by individuals of the Aelurodon spissidens White, 1947 sexually dimorphic species of ursids, it is possible Enhydrocyon spissidens (White), Olsen, 1958 that the two European Phoberocyon species lie near the extremes of variation for a single species Revised Diagnosis. P4 more slender than that (compare the ranges with comparable dimensions of E. praedator Matthew, 1924. Talonids of Ml for single taxa of better represented hemi- and M2 are longer and wider than in E. praedator. cyonines in the data given by Hurzeler, 1944, and The protoconid and metaconid of M2 are sepa- Ginsburg, 1955). If this is the case then H. (P.) rate, distinct cusps. The lower tooth row of E. johnhenryi would be weakly distinguished from spissidens is less curved than that of E. praedator. the European forms only by having posterior Discussion. Euoplocyon spissidens is known accessory cusps on P2 as well as P3-P4 and perhaps only from lower dentitions. Previous authors shorter diastemata between the premolars. have primarily used the size and placement of the Hemicyon (Phoberocyon) johnhenryi is pres- premolars, the trenchant talonid, and absence or ently known only from the Thomas Farm near absence of a metaconid on M1 in deter- deposit, but associated metacarpals IV-V (F:AM mining the generic affinities of this species. The

TABLE 1 Comparative Measurements of Hemicyon johnhenryi and Hemicyon aurelianensis (Measurements in Millimeters; parentheses indicate an approximate measurement.) Hemicyon Hemicyon johnhenryi aurelianensis F:AM 98607 F:AM 98608 MNHN 228 (Cast of MCZ 4059, holotype) (Holotype)a Length from canine to condyle 201.0 (198.0) Depth of ramus between P2 and P3 36.6 31.8 Depth of ramus between M2 and M3 (42.0) 42.5 Depth of condyle at center 17.0 11.1 P, (length X width) - 8.1 X 4.2 P2 (length X width) 12.4 x 6.5 10.9 X 5.4 P3 (length X width) 14.0 x 6.9 12.9 x 6.0 P4 (length X width) 18.2 X 9.7 16.8 X 8.0 Ml (length X width) 33.6 X 14.3 33.2 X 13.1 35.0 X 13.5 M2 (length X width) 22.0 x 15.0 19.7 X 13.0 (21.0 X 13.5) M3 (alv.) (length X width) 15.0 x 6.9 (11.0 X 6.0) aMeasurements of M, from Hurzeler, 1944, p. 154; M2 from Ginsburg, 1955, fig. 2. 6 AMERICAN MUSEUM NOVITATES NO. 2610 problems inherent in working with limited mate- Euoplocyon to the Simocyoninae Zittel, 1893. rial are apparent in the taxonomic history of E. These features, however, are widely encountered spissidens. White (1947) diagnosed the species among carnivores and for this and other reasons and referred it to Aelurocyon. Olsen (1958), the Simocyoninae can be shown to be a highly using Enhydrocyon crassidens Matthew, 1907, polyphyletic group whose members can be for comparison, noted the less crowded tooth allocated to the Amphicyonidae, Ursidae, and row and the less robust teeth of Enhydrocyon, Canidae. The unusual presence of anterior and also seen in E. spissidens, and the absence of the posterior cingular cusps and a posterior accessory lingual flexure in the trigonid of M1 of Aeluro- cusp on all four premolars, used by Matthew cyon, in contrast to E. spissidens, and transferred (1924) in his diagnosis of E. praedator, are found "Aelurocyon" spissidens to Enhydrocyon Cope, only in the borophagine canids Tomarctus Cope, 1879. Olsen (1958) also noted that the small 1873, Prohyaena Schlosser, 1887, and Aelurodon ridge which occupies the position of the meta- Leidy, 1858. As these three genera have a num- conid on was present in Enhydrocyon crassi- ber of features in common (such as a pre- Ml I3 dens, E. spissidens, and a referred specimen of maxillary-frontal bone contact, an enlarged Euoplocyon praedator (MCZ 17301) but felt with accessory cusps, a parastyle on P4) and that E. spissidens did not otherwise closely probably form a natural group, Euoplocyon resemble Euoplocyon praedator.' For the follow- could perhaps be better placed with these genera ing reasons, we believe that Euoplocyon is the in the Borophaginae Simpson, 1945. more correct generic assignment for E. spissi- Euoplocyon is known from two species: dens: (1) The teeth of Euoplocyon are less Euoplocyon spissidens from the Thomas Farm robust than those of Enhydrocyon; (2) P4 has an Local Fauna (medial Hemingfordian) and E. anterior accessory cusp which is lacking in praedator from the Lower Snake Creek Fauna Enhydrocyon; (3) the talonids of Ml and M2 are (early Barstovian). Gazin (1932) referred a single larger than those of Enhydrocyon; (4) the hypo- Ml from the Skull Spring Fauna (early Bar- conid of Ml in Euoplocyon is more labially stovian) of Oregon to ?Euoplocyon. This latter placed and less bladelike, whereas that of identification remains uncertain, but the genus is Enhydrocyon occupies most of the talonid; (5) nonetheless widespread over North America. Euoplocyon has a Pl, whereas Enhydrocyon often lacks this tooth; and (6) the mandible of FAMILY MUSTELIDAE SWAINSON, 1835 Euoplocyon increases in depth between P3 and SUBFAMILY MUSTELINAE SWAINSON, 1835 MI; the opposite being true in Enhydrocyon. Euoplocyon spissidens differs from E. Miomustela(?) sp. Olsen, 1956 praedator, the only other described species of The small mustelid referred to Miomus- Euoplocyon, in the reasons stated in the diag- tela(?)sp. by Olsen (1956) appears to represent a nosis. The referred ramus MCZ 7310 is compared third carnivore from Thomas Farm belonging to in figure 2 with the type of E. praedator (here a genus also present in the Old World Miocene. figured for the first time). The most obvious Olsen's queried reference to Miomustela Hall, differences are the distinct cusps on the trigonid 1930, must be rejected on the grounds of the of M2 in E. spissidens. In E. praedator the distinctly different morphology of the carnassial trigonid of M2 is composed of a single crest in and the large size of the crowded premolars as which the protoconid and metaconid cannot be indicated by their alveoli. The carnassial has a differentiated. very high compressed trigonid with a small low The trenchant talonid and the absence of a metaconid and a narrow trenchant talonid. The metaconid on Ml led Matthew (1924) to relate presence of an Ml metaconid differentiates the Thomas Farm species from contemporary Pale- 'The presence of a ridge in the position of the meta- ogale Meyer, 1846. Comparable forms are known conid appears to be variable in Euoplocyon. The ridge is from the Runningwater Formation of Nebraska totally absent in the type specimen of E. praedator and among the various species assigned to the (AMNH 18261) but present as a different ridge in other in and specimens (F:AM 27314, 27315) referred to that genus Plesictis Pomel, 1846, Aquitanian species. Burdigalian faunas of Europe. The latter differ so 1 976 TEDFORD AND FRAILEY: CARNIVORA 7

A C gQ

C CP

FIG. 2. Euoplocyon praedator Matthew, 1924. Holotype, AMNH 18261, left ramus, C broken, P1, P2 broken, P3-P4, M1-M3. A. Occlusal view. B. Lateral view. E. spissidens (White, 1947), referred, MCZ 7310, left ramus, P2-P3 alveoli, P4, M1-M2. C. Occlusal view. D. Lateral view. All natural size.

TABLE 2 Comparative Measurements of Euoplocyon spissidens and Euoplocyon praedator (Measurements are in Millimeters; parentheses indicate an approximate measurement.) Euoplocyon Euoplocyon spissidens praedator MCZ 4246a MCZ 7310 AMNH 18261 Type Type

Depth of ramus at P3 19.0 19.4 21.7 Depth of ramus at M2 21.0 22.5 22.9 P, (length X width) - - 5.9 X 3.7 P2 (length X width) length alveolus (6) P3 (length X width) - Length alveolus (8) 8.4 X 4.8 P4 (length X width) 10.0 X 5.0 10.0 X 5.2 10.0 X 5.7 M1 (length X width) 17.0 X 7.0 17.0 X 7.0 18.2 X 7.7 M2 (length X width) 7.8 x 5.1 6.8 x 5.3 M3 (length X width) 3.4 X 3.2 aMeasurements from White, 1947, p. 498. 8 AMERICAN MUSEUM NOVITATES NO. 2610 radically from the holotype of the genotypic species that are now incorrectly included within the cies (P. genettoides Pomel, 1846) that reference procyonid genus Plesictis. Such musteline species to that genus seems as unlikely as reference to are known from Aquitanian and Burdigalian Miomustela. Hough (1948) and de Beaumont faunas of Europe and Hemingfordian faunas of (1968) have shown that P. genettoides, and the North America. closely related P. lemanensis Pomel, 1853, are procyonids although musteloid in dental adapta- LITERATURE CITED tion. The Thomas Farm form is closest to "P." julieni Viret, 1929, and "P." sicaulensis Viret, de Beaumont, G. in size 1950, and 1968. Note sur la region auditive de quelques 1929, (see Dehm, fig. 78) carnivores. Arch. Sci. Geneve, vol. 21, morphology. "Plesictis" julieni is known from a no. 2, pp. 213-224. skull (the type, Viret, 1929) which shows the Dehm, R, inflated bullae with strongly forward directed 1950. Die Raubtiere aus dem Mittel-Miocan meati, and widely separated paroccipital and (Burdigalium) von Wintershof-West bei mastoid processes typical of mustelids and con- Eichstatt in Bayem. Abh. Bayerischen trasting with contemporary Plesictis. These Akad. Wiss. Math.-Nat. KI., vol. 58, pp. features and the parasagittal crests are matched 1-141. in specimens from the Runningwater Formation. Frick, C. Mustelids such as these have many of the features 1926. The Hemicyoninae and an American of the later Cenozoic and living Mustelinae, and Tertiary bear. Bull. Amer. Mus. Nat. Hist., vol. 56, art. 1, pp. 1-119. possibly represent the earliest occurrences of that Gawne, C. E. subfamily in the New World. 1975. Rodents from the Zia Sand Miocene of New Mexico. Amer. Mus. Novitates, no. SUMMARY 2586, pp. 1-25. Gazin, C. L. The presence of the Holarctic carnivore genera 1932. A Miocene mammalian fauna from Cynelos and Hemicyon in the Thomas Farm southeastern Oregon. Publ. Carnegie Local Fauna and their very close morphological Inst. Washington, no. 418, part III, pp. correspondence with European counterparts 37-86. yields further data on the intercontinental corre- Ginsburg, L. lation of these early Miocene mammal faunas. As 1955. De la subdivision du genre Hemicyon the comparisons detailed above reveal, Cynelos Lartet (carnassier du Miocene). Bull. caroniavorus and Hemicyon (Phoberocyon)john- Soc. Geol. France, 6th series, vol. 5, pp. henryi closely resemble the exclusively Burdi- 85-99. galian C. schlosseri and H. (P.) aurelianensis 1966. Les amphicyons des Phosphorites du Quercy. Ann. Paleont., vol. 52, pp. respectively. These European species occur 23-63. together in the Sables de l'Orleanais of the Loire 1974. Les faunes de mammiferes burdigaliens Basin of France. Cynelos schlosseri (type) and H. et Vindoboniens des bassins de la Loire (P.) huerzeleri [closely related to H. (P.) aureli- et de la Garonne. Mem. Bur. du Rech. anensis] occur in the fissure fills at Wintershof- Geol. et Min., vol. 1, no. 78, pp. West in Bavaria. Both assemblages are regarded 153-167. as early Burdigalian Age following the concept Hough, J. R. of that Age held by European paleomammalogists. 1948. The auditory region in some members Euoplocyon is recognized in the Thomas of the Procyonidae, Canidae and Ursi- dae. Bull. Amer. Mus. Nat. Hist., vol. Farm Local Fauna extending both the temporal 92, art. 2, pp. 67-118. and geographic ranges of this genus from early Hunt, R. M., Jr. Barstovian faunas in Oregon and Nebraska into 1972. Miocene amphicyonids (Mammalia, the Hemingfordian of Florida. Carnivora) from the Agate Springs The identification of Miomustela(?)sp. by Quarries, Sioux County, Nebraska. Olsen (1956) is rejected. The mustelid in ques- Amer. Mus. Novitates, no. 2506, pp. tion is apparently a true musteline related to spe- 1-39. 1976 TEDFORD AND FRAILEY: CARNIVORA 9

Hurzeler, J. Jour. Florida Acad. Sci., vol. 25, no. 2, 1944. Zur Revision der europfiischen Hemi- pp. 142-146. cyoniden. Verhandl. Naturf. Gesell. Patton, T. H. Basel, vol. 55, pp. 131-157. 1967. Oligocene and Miocene vertebrates Kuss, S. E. from central Florida. In Southeastern 1965. Revision der europaischen Amphi- Geological Soc. 13th Field Trip, cyoninae (Canidae, Carnivora, Mamm.) Miocene-Pliocene Problems of Penin- ausschliesslich der voroberstampischen sular Florida. H. K. Brooks and J. R. Formen. Sitzber. Heidelberger Akad. Underwood, Jr. (eds.), 36 pp. Wiss. Math.-Nat. KI., vol. for 1965, pp. Ray, C. E. 5-168. 1957. A list, bibliography, and index of the Lartet, E. fossil vertebrates of Florida. Florida 1851. Notice sur la Colline de Sansan, J.-A. Biol. Surv. Special Publ. no. 3, 175 pp. Portes, Auch., pp. 1-45. Viret, J. Matthew, W. D. 1929. Les faunes de mammiferes de l'Oligo- 1924. Third contribution to the Snake Creek cene superieur de la Limagne Bourbon- fauna. Bull. Amer. Mus. Nat. Hist., vol. naise. Ann. Univ. Lyon, vol. 1 (n.s.), 50, art. 2, pp. 59-2 10. no. 47, pp. 1-328. Olsen, S. J. White, T. E. 1956. A small mustelid from the Thomas 1942. The lower Miocene mammal fauna of Farm Miocene. Breviora, no. 51, 5 pp., Florida. Bull. Mus. Comp. Zool., vol. 1 fig., 1 table. 92, no. 1, pp. 149. 1958. Some problematical carnivores from 1947. Additions to the Miocene fauna of the Florida Miocene. Jour. Paleo., vol. North Florida. Ibid., vol. 99, no. 4, pp. 32, no. 3, pp. 595-602. 497-515. 1962. The Thomas Farm fossil quarry. Quart.

+ fXGTwr - W So

b:: ...... -: 00:" t0'Ty' ''0 ;0 f0\f0X; 0 0' .t

a rt, h $; ;s W0iffi;;21 fit75m;' dS d;

i:i: