BULLETIN OF THE FLORIDA STATE MUSEUM BIOLOGICAL SCIENCES Volume 14 Number 2 MIOCENE AND PLIOCENE ARTIODACTYLS, TEXAS GULF COSTAL PLAIN Thomas Hudson Patton /853 UNIVERSITY OF FLORIDA Gainesville 1969 Numbers of the BULLETIN OF THE FLORIDA STATE MUSEUM are pub- lished at irregular intervals. Volumes contain about 300 pages and are not necessarily completed in any one calendar year. W,WrER AuFFENBERG, Managing Editor OLIVER L. AUSTIN, Jn., Editor Consultant for this issue: DONALD E. SAVAGE Communications concerning purchase or exchange of the publication and all manuscripts should be addressed to the Managing Editor of the Bulletin, Florida State Museum, Seagle Building, Gainesville, Florida 32601. Published June 17, 1969 Price for this issue $1.50 MIOCENE AND PLIOCENE ARTIODACTYLS, TEXAS GULF COASTAL PLAIN THOMAS HUDSON PATTON1 SYNOPSIS: Describes 27 species of fossil artiodactyls from a series of vertically successive mammalian. assemblages in Miocene and Pliocene deposits of the Texas Gulf Coastal Plain and discusses their systematic positions. Among the new forms represented are two camel genera: Australocametus, the probable Aep!/cameZus ancestor, and Nothott/lopus, a very unusual member of the Proto- labis-Pliauchenia lineage. The Floridagulinae are now seen to have had a trans-Coastal Plain distribution extending in time from the middle Heming- fordian Garvin Gully Fauna through the Barstovian Cold Spring Fauna. The Gulf Coast species of the Synthetoceratinae are discussed and the phylogeny of the subfamily outlined. Evidence from this study indicates that the Gulf Coastal Plain constituted a distinct faunal province throughout most of the Tertiary. Whereas many striking similarities exist between the faunas of the Texas Coastal Plain and those of the Great Plains, several groups are true Gulf Coast autochthons. Others, which are distributed in both regions, contain species endemic to the Gulf Coastal Plain. The ages of the Texas Gulf Coastal Plain faunas are revised and correlated with those from the Great Plains. TABLE OF CONTENTS 116 ACKNOWLEDGMENTS ... 117 ABBREVIATIONS . 119 STRATIGRAPHY . 120 DISTINGUISHING CHARACTERISTICS OF THE FAUNAS 126 SYSTEMATIC DESCRIPTIONS 188 Order ARTIODACTYLA 184 Suborder TYLOPODA 184 Family Camelidae -. .. f84 Subfamily Aepycamelinae -.~~..-~.~~......-...-. 184 Subfamily Camelinae ...._............ 150 Subfamily Floridatragulinae __ 168 Suborder RUMINANTIA . 174 Infraorder TRACULINA - - - 174 Family Protoceratidae . 174 1Thomas Hudson Patton is Assistant Curator of fossil vertebrates at the Florida State Museum and Assistant Professor of Zoology at the UBiversity of Florida, Gainesville. His principal research is on Tertiary mammals of the Gulf Coastal Plain and fossil vertebrates of the West Indies. Manuscript received 12 June 1968. - Ed. Patton, Thomas Hudson. 1969. Miocene and Pliocene artiodactyls, Texas Gulf Coastal Plain. Bull. Florida State Mus., vol. 14, no. 2, pp. 115-226. 116 BULLETIN FLORIDA STATE MUSEUM Vol. 14 Subfamily Synthetoceratinae 174 Phylogeny of the Synthetoceratinae 190 Infraorder PECORA 198 Family Cervidae 193 Subfamily Palaeomerycinae 4... .... 198 Subfamily Dromomerycinae 197 Family Antilocapridae - _.. .. 200 Infraorder OREDONTA . 201 Family Merycoidodontidae . 201 Subfamily Ticholeptinae 201 Subfamily Merychyinae 204 AGE AND CORRELATION 205 SUMMARY AND CONCLUSIONS 217 LITERATURE CITED . 220 INTRODUCnON The Miocene sedimentary rocks along the Texas Coastal Plain provide a record of successive mammalian faunas unsurpassed any- where in the Gulf and Atlantic coastal provinces of North America. This is the only such record outside of western North America, and consequently affords a basis for a detailed comparison of faunas from the Coastal Plain with those from the Miocene and Pliocene deposits of the High Plains. A series of vertically successive mammalian assemblages of rela- tively short temporal span gives the paleontologist an opportunity to study and observe not only geographic changes or shifts in animal communities through time, but also to detect microevolutionary steps within the various groups represented. The combination of good stratigraphic control and increasingly clearer evolutionary line- ages provides the basis for a more complete reconstruction of zoogeographical patterns, faunal interchange, and paleoecology than is usually realized in faunal studies. It also facilitates biostratigraphic correlations between faunas of the Texas Coastal Plain and those else- where especially from the more widely known Great Plains deposits. Whereas the Texas faunas show many similarities to those from the Great Plains region in individual taxa, they differ substantially in rela- tive abundance of taxa and, partially, in their faunal associations. Some groups, for example the synthetocerines, which are only sparing- ly represented in the Great Plains deposits, are abundant in the Gulf Coastal Plain faunas and appear to have undergone most of their evolution in the Gulf region. As would be expected, some taxa appear earlier or persist longer in one area than in the other. Yet 1969 PATTON: TEXAS ARTIODACTYLS 117 on the whole a similarity exists between the faunas of the two regions, which is fortunate for interregional biostratigraphic correla- tion, as most of the geochronologic framework of the North American Tertiary is based on the rocks and fossil vetebrate faunas of the Great Plains. Whereas the Great Plains and the Texas Coastal Plain faunas each evolved under their own particular set of environmental conditions, whatever ecological barriers existed between the two regions must have been slight enough to allow a relatively free interchange of forms between them. Still, enough isolation, geo- graphic at least, existed between the faunas of the two regions to permit specific and greater differences to accumulate. Although vertebrate fossils of Miocene and Pliocene age have been known from the Texas Coastal Plain for nearly a century, the area has only recently been investigated in detail. As fossil remains are for the most part fragmentary and widely dispersed throughout these deposits, the area has been generaliy uninviting from both the collecting and systematic standpoints. Only when the unique nature of the faunal elements and their associations was recognized and its importance understood were any significant studies undertakeh. - Papers by Hay ( 1924) and Hesse ( 1943) included the Brst general faunal descriptions of fossil vertebrates from the Texas Coastal Plain. Quinn ( 1952, 1955) studied in detail the fossil horses from these faunas, and later Wilson ( 1957, 1960) described the entelodonts and the carnivores. Because of the unusually fortunate stratigraphic con- Bguration of the Coastal Plain faunas and the relative abundance of remaining undescribed fossil vertebrate material available, the writer undertook a systematic investigation of various artiodactyl groups. It is hoped that data derived from the study of these taxa may serve as a basis for further systematic and biostratigraphic analysis of the Texas faunas and will provide a foundation for reconstructing a sequence of biological events through time and also perhaps the geological and climatic processes that determined these events. ACKNOWLEDGMENTS I am indebted to John A. Wilson of the Department of Geology, the University of Texas, for his suggestions and criticisms during the writing of this paper. The critical reading of the manuscript by W. Charles Bell, Ernest L. Lundelius, and W. Frank Blair, of the University of Texas, by Donald D. Savage of the University of California ( Berkeley) and by Malcolm C. McKenna and Richard Tedford, American Museum of Natural History. is gratefully acknowledged. 118 BULLETIN FLORIDA STATE MUSEUM Vol. 14 I should also like to thank Malcolm C. McKenna, Morris F. Skinner, andmy colleague S. David Webb of the Florida State Museum, the University ofFlorida, for numerous stimulating and enlightening discussions on the strati-graphy and fossil vertebrate faunas of the Tertiary of the western United State5. I am grateful to the Department of Geology of the University of Texas,the Coates Foundation of the University of Texas, The Society of the SigmaXi and the Scientific Research Society of America, the Houston GeologicalSociety, and to the Humble Oil and Refining Company for financial supportduring the course of this project I should like to express my appreciation to authorities at the AmericanMuseum of Natural History, the United States National Museum, the Museumof Comparative Zoology at Harvard University, Texas A&M University, theShuler Museum of Paleontology at Southern Methodist University, the Universityof California Museum of Paleontology ( Berkeley), the Florida State Museum,and the Frick Laboratory of the American Museum of Natural History, forpermitting me to examine fossil material under their care. Finally, I wish to express my deepest thanks to my wife, Ellen FrederickPatton, for her constant encouragement and indispensable support throughoutthe course of this work. The illustrations were drawn by Mrs. MargaretSkeels Stevens. ABBREVIATIONS 1Materials belonging to various institutions are referred to in the text bytheir catalog numbers and the following abbreviations: AMNH American Museum of Natural History FAM Frick Collection , American Museum of Natural History FSGS Florida State Geological Survey MCZ Museum of Comparative Zoology, Harvard University SMU Southern Methodist University TAMU Texas A&M University UCMP University of California, Museum of Paleontology, Berkeley UF/FSM University of Florida, Florida State Museum USNM United States National Museum UTBEG University of Texas, Bureau of Economic Geology FIGURE 1, Map of Miocene biostratigraphic units in the Texas Coastal Plain( from Wilson, 1962). Numbers refer to fossil localities listed in Table 1. LEON HOUSTON ANGELINA SABINE 1969 PATTO N : TEXAS A R TIO D A C TY LS FALLS NEWTON ; ;-1-1-TTir-TTrT' ROBERTSON TRWITY POL BELL 1::· Z:l;&8.. MADISON .rr:r-r-1-1 -,-. MILAM GRIMES WALKER_' 8 2 WILLIAMSON BRAZOS AN 1 1~! 1110--7-7~*H-+*-,4+on*(th.::~a JACINTO BURLESON MI GOMER 8 4 HARDIN ;( TRAVIS LEE LIBERTY .29 ORANLE 14 JEFFERSON ~ BASTROP FAYETTE AUSTIN ~=BI'ld,1~-9~813 HARRIS I CHAMBERS COMAL CALDWELL 17 AARB 6 ·· ~~ COL ADO.,3 - BEXAR FORT BEND MEDINA GUADALUPE GONZALES ' IS AVAGA GALVESTON 0 WILSON IWO31 WHARTON ATASCOSA JACKSON ZAVALA FRIO D T MATAGORDA EXPLANATION KARNES VICTORIA IAD N 28 LABAHIA MISSION FAUNA LAPARA GREEK FAUNA DIMMIT LA SALLE McMULLEN LIVE °A:BE""*'*7 3 ~ REFUGIO COLD SPRING FAUNA BURKEVILLE FAUNA WEBB GARVIN GULLEY FAUNA •17 \\ ouvil~~ET 9 FOSSIL LOCALITIES f SCALE, 9 '_ESL____28 9-- _40'lf'" IZ CD 120 BULLETIN FLORIDA STATE MUSEUM Vol. 14 Table 1. A~EAL DISTRIBUTION OF FOSSIL FAtINAS IN TEXAsl Map no. UTBEG no. County Location 1 81160 Newton Near Burkville2 31087 Tyler Near Town Bluff8 31057 Polk Near Moscow4 31183 Polk Near Goodrich5 31200 Polk Near Goodrich6 81219 San Jacinto Near Cold Spring 7 31191 San Jacinto Near Cold Spring8 31243 San Jacinto Near Point Blank'9-- 31190 San Jacinto Near Point Blank10 81242 San Jacinto Near Point Blank11 30878 Walker Aiken. Hill12 40071 Grimes Near Navasota18 40070 Grimes Sommers Pit14 81272 Washington Near Chapel Hill15 40067 Washington Hidalgo Bluff16 40068 Washington Near Carmine17 81259 Fayette Near LaGrange18 81278 Fayette Near Abiandsville19 81278 Lavaca Near Shiner20 31262 DeWitt Near Concrete21 80896 Bee Near Berclair22 31132 Bee Near Normanna23 81080 Bee Near Berclair24 81170 Bee Near Normanna25 30936 Live Oak Near George West26 30904 Live 6ak Near George West27 31089 Duval Palangana dome28 30895 Goliad Goliad State Park 29 Hardin Sarat6ga field SO 40589 Austin S. F. Austin State Park81 40198 DeWitt Near Hocheim 82 40224 San Jacinto Near Point Blank88 40262 Gonzales Near Shiner 1More detailed geographic information for each of the above localities is on file in the cata-logue of the vertebrate fossil collection of the Bureau of Economic Geology, The Universityof Texas. 2This locality, on the basis of the fossils it contains, more likely falls within the Cold Springrather than Burkeville Fauna. See also Quinn (1955). STRATIGRAPHY Stratigraphic nomenclature employed in this paper is that of Wilson ( 1956), who summarized the synonymy and revised the Miocene and Pliocene formations of the Texas Gulf Coastal Plain. Wilson divides these deposits into three lithostratigraphic units, the Oakville, Fleming, and Goliad Formations ( in ascending order). Within these formations several vertebrate faunas can be recognized 1969 PATTON: TEXAS ARTIODACTYLS 121 ( Quinn, 1955; Wilson, 1956). From oldest to youngest these are the Garvin Gully, Burkeville, Cold Spring, and Lapara Creek Faunas. According to Wilson, ( 1956): The Garvin Gully Fauna has known stratigraphic limits in Grimes, Washing- ton, and Fayette counties. It coincides with the lower Oakville and the Moulton sandstone Members of Renick ( 1986). The stratigraphic limits of the higher faunas are not so well known, but the Burkeville and Cold Spring faunas are known to be within the outcrop limit faunas of the Fleming and Lagarto formations . It is hoped that future mapping and collecting will clarify the relation of the Goliad formation and its members to the faunas found within it. Lithostratigraphic and biostratigraphic relationships are graphi- cally represented in Figures 2 and 3. The age of these faunas is partially reinterpreted in this paper and is outlined under the section titled Age and Correlation. The Miocene and Pliocene deposits of the Texas Gulf Coastal Plain are exposed in arcuate belts of varying width. They strike roughly parallel to the present shoreline of the Gulf of Mexico. The dip is characteristically small, seldom exceeding 2 to 4 feet per mile. Because the shoreline shifted progressively gulfward through sue- cessive geologic epochs since the Eocene, the younger outcrops are found south and east of the older ones. Topographically the Texas Coastal Plain varies from a nearly flat surface near the coast to a higher hilly region farther inland. The near-coast region consists of unconsolidated clay and sand deposits, whereas the inland belt is formed on differentially eroded alternating sandstone lenses and clay. In the hilly regions infacing cuestas, some with scarps more than 100 feet high, are supported by more resistant sandstone lenses; the intervening lowlands coincide with weaker, more easily eroded clay. In this region the Cenozoic was a time of marine deposition on the fluctuating margins of the Gulf, and of continental deposition inland by meandering, anastomosing streams on the relatively flat, subsiding coastal plain. Because of its history of alternating marine transgressions and regressions, the region is characterized by a com- plicated intergrading of continental and marine sediments. Not only does one type of sediment replace another vertically, but some for- mational units grade laterally, or change abruptly, from one facies to another. This particular geologic configuration is the source of much of the confusi6n still present in the stratigraphic nomenclature of Gulf Coast Cenozoic deposits. The Lower Miocene strata present a continental facies at the outcrop, whereas downdip they change SW 122 B U LLE T IN FLO R ID A S TATE M U S E U M Vol. 14 NE WASHINGTON GRIMES WALKER SAN JACINTO POLKCO. CO CO. CO. CO ---4 1 81 -47&25- . -L./ EFLEII,111 IG-FORMATIOI,1 E=r-m-r,==r- j==El se un ly g .le .[C pu aA I rs so j ur et d Ie }S EO ~) S ex ag J o )p o m a iu sl j or t[d E IS !:1 8# S '3 m In o Id L-6- .-i >- 0 1, 1 r n-'-1-16&5&2*%!ber61 ief55Faii@=rE :*0-dflt~w' 1 1~-46-+41 4· ' -ef,i-- -~--~' ~ -'1 -< 7 -_r- Li-*JA®IALLEIFORMATIOIi- =3 - - -, 15.--14---- 5-- -c==E *, 1~z~E -lil= --Z , 7-0 f---proximate tihy accessory tubercle is visible at the base of the crown between the anterior and posterior crescents of Mi and Mz. This feature is not characteristic of the other advanced camels with which Nothotylopus is contemporary. DiscussION. - In view of the incomplete material available for comparison, derivation of Nothotylopus from any of the currently kilown groups of camels is difficult and precise systematic placement tenuous. Nothotylopus exhibits a strange mixture of advanced and conservative featuies: long diastemata; reduced but complex P) and P4; P: absent but Pi two-rooted; and robust, buttressed molars. Reduction and loss of lower premolars has been one of the most consistently employed and reliable characters in differentiating Late Tertiary camels. With the possible exception of Aepycamelus ( = Alti- camelus), all of the major taxa, i.e.: Protolabis, Procamelus, Pliau- chenia, Megatylopus, and Miolabis exhibit some degree of premolar reduction, including the loss of P2 and/or Pi. Heretofore loss of P2 has been a diagnostic character restricted, at least among the large Tertiary camels, to Pliauchenia (Cope, 1874a) and Megatylopus ( Matthew and Cook, 1909) although individuals in some species of Protolabis, e.g. P. longiceps, may lose P2. Whereas reduction may be a parallel convergent character acquired separately by each of these genera , it is more likely a progressive character peculiar to the Proto- 1969 PATTON: TEXAS ARTIODACTYLS 163 labis lineage. For example, there has been reduction and occasional loss of P2 in Protolabis, reduction in Procamelus, and loss in Pliau- chenia and Megat!/lopus. Nothotylopus possesses this and other, sep- arately acquired, progressive characters, but retains enough conserv- ative features to preclude its derivation from any but the Protolabis stock. This conclusion is based primarily on short, two-rooted Pi and reduced P3 and P4 of Protolabis. These features are also present in the small gracilis group of procamelines and in Miolabis. However, the former is a separate, early offshoot of the procamelines, whereas Miolabis, as it is now known, is taxonomically uncertain, perhaps ultimately to be placed in synonymy with Protolabis. The Oxudac- tylus, Aepycamelus, and Hesperocamelus lineage is far too conserv- ative in premolar reduction, and too advanced in upper incisor reduc- tion and dimensions of skull to provide a possible direct ancestor for Nothotylopt,3, while the Procamehts-Pliauchenia-Megatulopus groups are too specialized, especially in degree of hypsodonty. The probable ancestral stock, therefore, should be found among the early members of the genus Protolabis, or perhaps some form intermediate between Protolabis and that line of oxydactylines which presumably gave rise to it. 2cm. FIGURE 18. Nothotylopus camptognathus n. gen. et sp. C UTBEG 31081-26, Type); left ramus, labial and occlusal views. 164 BULLETIN FLORIDA STATE MUSEUM Vol. 14 Megatylopus Matthew and Cook, 1909 In 1875 Cope proposed the genus Pliauchenia to include those camels lacking the second premolar of the lower jaw, In addition to the type species, P. humpreysiana, he described another species, P. vulcanorum, from an upper dentition in which P2 was present. The dental formula for the new genus was therefore given as "premolars 4". P. oulcanorum has subsequently been assigned by Gregory ( 1942) to Procamelus on the basis of retained P2, but lack of P2 has been used as a diagnostic character for all later additions to the genus, In- cluded in this taxon are various presumably related forms ~ which, while sharing P2 loss and succeeding premolar reduttion, are clearly distinct on other grounds. Therefore, as Matthew ( unpubl. MS) sug- gested, it appears that P~ loss marks a stage in the evolution of the protolabine camels and is not characteristic of a group of nearly related species. In an attempt to rectify part of this confusion, Matthew and Cook ( 1909) restricted the very large species of Pli- auchenia to a separate subgenus, Megatylopus, which has since been elevated to generic rank ( Gregory, 1937). . 4)11 ~ 1 g ~: I \% 1% , Al, 0, ./ r , r 2 cm r. -=---- -1-=ER*=5- i -ij FIGURE 19. Megat!/lopus primaeous n. sp. C UTBEG 31081-460, Type); left ramus, labial and occlusal views Ps-Ma. Matthew and Cook ( 1909) originally described Megatylopus as having the following characteristics: Gigantic camels with dentition It Cf PM? Mt, the second premolar absent in both jaws, the first retarded or absent, the reduction of the posterior premolars intermediate between P,·ocamettis and Camelus. Upper molars with prominent external ribs and styles, as in Auchenia. Limbs and feet moderately elongate and very massive, large in proportion to size of skull. 1969 PATTON: TEXAS ARTIODACTYLS 165 Although additional knowledge of Megatylopus has accumulated since its original description, the genus is still imperfectly understood and, therefore, any comparisons and/or systematic conclusions re- garding it and related forms are necessarily tentative. Species that have been included in Megatylopus are the type spe- cies, M. gigas Matthew and Cook ( 1909), M. spatula ( Cope, 1893), M. merriami ( Frick, 1921), M. major ( Leidy, 1886) and M. cochrani ( Hibbard and Riggs, 1949'). M. spatula and M. merriami are regarded by Webb ( 1965) to be properly placed in Titanotylopus Barbour and Schlutz ( 1934). Webb ( 1965) bases his conclusion in part, on the lack of a cranial flexure, the broad anterior lobe of P4, and the non- reduced Pa, features that these two species share in common with Ti- tanotylopus, but which serve to distinguish them from Megatylopus. To the remaining species of Megatylopus must be added two new forms: M. matthewi from the Hemphillian Coffee Ranch Quarry, a distinct species which Matthew ( unpubl. MS) recognized as Para- camelus arenicola, which is listed but not described by Reed and Longnecker ( 1932 ) and which Webb ( 1965 ) described and a rather primitive form from the Lapara Creek Fauna, M. primaeous, which is described below. Megatylopus primaeous new species Figures 19, 20, Table 15 TypE. - UTBEG 31081-460, a left mandible, with Pz alveolus, PrMB· TYPE LOCALMY. - Farish Ranch, Bee County, Texas. OCCURRENCE. - Lapara Creek Fauna. REFERRED MATERIAL. - UTBEG 31081-18A, a partial right ramus with PB-MB; 31081-188, a metapodial. ETYMOLOGY. - primaeous = L. early, young. DIAGNOSS. - An early species of Megatylopus, similar to M. mai0r, but with a small P2 and a larger and more complex P); man- dible slender and vertically shallow. DESCRIPTION. - The type mandible of M. primaeous is long and slender, the depth of the ramus being consistently shallower than Nothot!/lopus or Protolabis (Tables 11, 14, 15). Characters of the ascending ramus are mostly indeterminate because of poor preserva- tion. A small mental foramen occurs just below the center of the an- terior half of Mi; the anterior mental foramen is situated above the posteroinferior border of the symphysis where it deepens ventrally. The alveolus anterior to this foramen contained the lower canine. 166 BULLETIN FLORIDA STATE MUSEUM Vol 14 ' Above the foramen and just below the surface of the bone is the un- erupted Pi. Two small alveoli for P2 occur anterior to Pa. This tooth was obviously quite small and probably nohfunctional. In view of the fact that the more advanced species of Megatylopus are characterized by loss of Ps, the retention of a small ]?2 in M. prinuleous is regarded as representing the last stage of reduction of that tooth ih the Mega- tylopus lineage. The diastema between /C and Pa is approximately 62 mm long. Pa is two-rooted and has well-developed anterior and posterior cus- pids. The central cuspid ( metastylid) is unworn and attains a height of 16 mm above the base of the crown. The crown flares from the base to reach its greatest anteroposterior length ( 17.8 mm). at about two-thirds its vertical height. The anterior lobe or cuspid is rather deeply grooved but only weakly inflected lingually. The posterior lobe has a large posterolabial cuspid ( hypoconid) that is strongly in- flected lingually, its lower slope extending farther lingually than the weakly expressed posterolingual cuspid, or entoconid. These two cuspids do not join towards the base, thus continued wear would not produce a posterior lake. PI is a much larger and more robust tooth than Pa, and is quite hypsodont. The maximum height of the tooth, only slightly worn, is about 27 mm. The anterior cuspid is lingually infiected and set off from the primary cuspid by a prominent groove. A faint posterolabial groove is present between the principal and two posterior cuspids. The posterior cuspids are equally prominent and join posteriorly. In medium stages of wear they will form a posterior lake. Table 15. MEAsuREMENTs of Megat!/lopus primaeuus UTBEG Lower jaw 31081-460 (Type) Condyle to angular process 87.0 Depth of jaw below M, 51.0 Depth of jaw below M, 83.0 Depth of jaw below Pi-Pa diastema 80.0 PrMa 159.0P.-Ma 142.0 Prp. 39.0 Mi-Ma 120.0 P„ length x width 17.8 X 8.2P.,length X width 22.2 X 12.0 Mi, length x width 84.0 x 18.8 M„ length x width 43.0 X 21.0 M.,length X width 50.8 x 20.0 1969 PATTON: TEXAS ARTIODACTYLS 167 The lower molars are hypsodont and are characterized by strongly developed stylids. Mesostylid and metastylid are espe6ially prominent in the holotype. In the only other ramus referred to M. primdeuus ( UTBEG 31081-18A), these stylids are virtually absent. This jaw differs also in its smaller size and more reduced premolars. Whether or not the differences between the two specimens now assigned to M. primaeous are significant is for the present indeterminate. A large, massive metapodial ( UTBEG 31081-18B) is tentatively - assigned to this species. DIsCUSSION. - Of the species of Megatylopus accepted in this paper (M. gigas, M. major, M. matthewi, M. cochrani, and M. pri- maeous), M. primaeous is the most primitive form yet recognized. In a clearly discernible evolutionary trend towards the condition reached by the genus Camelops ( Webb, 1965), the premolars show a continuous reduction in size and number. P2 having already been lost in all species but M. primaeous, this involves the reduction and eventual loss of Pi and P# The absence of Ps ( and the usual absence of Pl ) represents the taxonomic line of demarcation between Megaty- topus and Camelops. M. cochmni from the Rexroad Fauna of Kansas has been demonstrated by Hibbard and Riggs ( 1949) · to be very similar to Camelops and probably represents the most advanced species of the Megat!/lopus lineage. Unaware of the Lapara Creek specimens, Webb ( 1965) considered M. major from Mixson's Bone Bed in Florida to occupy the earliest level of Megatylopus evolution and placed the species M. matthewi from the Hemphillian of Texas between those stages represented by M. gigas and M. cochrani. This arrangement is based primarily on the condition of P# In all species but M. maior and M. primaeuus, the roots of Ps are fused. These species also exhibit greater reduction in size and complexity of P3 than is attained by M. major and M. primaeous. On these grounds, M. primaeous and M. major are clearly more primitive than M. matthewi and the two succeeding species. M. major and M. primaeous are similar and probably closely related, but because of the retention of a small P2 and the greater complexity and size of PB in M. primae- uus, I consider it to be a more primitive species than M. maior. The differences are not great, but because the characters employed in taxanomic differentiation of these forms also reflect relative advances or stages in their evolution, it is useful to recognize and emphasize such distinctions. This is most apparent when considering these species from the standpoint of their stratigraphic occurrence. The evolutionary series beginning with M. primaeous and continuing 168 BULLETIN FLORIDA STATE MUSEUM VoI. 14 2 cm. FIGURE 20. Megat!/lopus primaecus n. sp. C UTBEG 31081-460, Type); Ieftramus, labial and occlusal views. through M. major, M. matthetri, M. gigas, and M. cochrani cor-responds with their relative stratigraphic position. Complications inthis evolutionary-stratigraphic sequence arising from different rates of evolution in different geographic regions, intermigrations, etc., are not as yet apparent. Subfamily Floridatragulinae Maglio, 1966 Floridatragulus White, 1940 The exact systematic position of these puzzling artiodactyls is uncertain. The fragmentary nature of Horidatraguline material from Texas and Florida deposits has prevented much speculation on phyletic affinities, but it is hoped that study of skull and jaw ma- terial now available from the Thomas Farm will lead to a betterunderstanding of the group. As I have previously commented ( Pat-ton, 1966( 67) : 184) : Floridatragutus bears considerable resemblance to members of two familiesof artiodactyls, the Hypertragulidae and the Camelidae . Floridatragulus resemblesmembers of the Hypertragulidae in the presence of the intercolumnar pillars, adouble enamel loop of the heel of M„, rather promhient cingula, and in theoccurrence of a diastema between P·, and P. In the hypertragulids the detachedP, tends to be unicuspid; in Floridatragrihis this tooth is bicuspid, though weaklyso. The anterjor lower premolars of Floridatragulus are laterally compressed as incamels, and P,, although more foreshortened than in most camels, retains the 1969 PATTON: TEXAS ARTIODACTYLS 169 general features of that group. In the skull and jaws of Floridatragulus from the Thomas Farm, now in the Frick Laboratory, the extremely elongated muzzle has four caniniform teeth ( or alveoli) on each side, probably representing modifica- tions of the incisors, C/, and F. This is a characteristic feature of the Tylopoda, in which the upper incisors and anterior premolars may undergo reducti6n, but in which there is never a complete loss of the upper incisors. In contrast, the hypertragulids are marked by either extreme reduction or complete suppression of the upper incisors, although they may be retained in some of the very primitive types. Because of its overall resemblance to the camels, in the same paper I referred Floridatragulus to the Camelidae as a new sub- familyl. I felt that those features that distinguish it from the Cam- elinae are certainly equal in taxonomic weight to those of the other camel subfamilies. Apparently for similar reasons, White ( 1947) tentatively placed Nothokemas as the type genus of a new family in the Hypertraguloidea. Nothokemas' afEnities with the Camelinae are rather substantially documented, but similar assurance cannot be claimed for Floridatragulus. Clearly while Floridatragulus has several distinctive camelid features such as the "hook" on the angle of the lower jaw, the canini- form condition of the upper anterior dentition, and the relatively swollen bullae, other traits are characteristically ruminant" (fide Simpson, 1945). For example the enclosed orbit, the presence on the molars of intercolumnar tubercles, cingula, strong ribs and styles, and the doub16 lobe on the hypoconulid of Ma are all quite primitive features not fully shared by any of the camels, especially the later ones. That the similarity between the various artiodactyl group increases back through geologic time is indisputable ( Matthew, 1905; Scott, 1940; Colbert, 1941; Simpson, 1945). However, , although the camels and ruminants are supposedly recognizable as distinct groups as early as the late Eocene, Floridatragutus shares relatively few features in common with contemporary representatives of its basal stock. Whether or not the morphological similarity between it and the Camelidae is the result of direct evolution, parallelism, or con- vergence is not immediately determinable. For the present, available evidence suggests that Floridatragulus descends from an early and distinct offshoot of the basal tylopod stock and that it ( and Notho- kemas) subsequently occupied an adaptive zone in the Coastal Plain similar to that occupied by some camels in the Great Plains. With the invasion of the Gulf Coast by several diversified artiodactyl genera 1A similar conclusion was reached by Maglio ( 1966) as the quotdd paper was in press. 170 BULLETIN FLORIDA STATE MUSEUM Vol. 14 ( including other camels) during the late Miocene and early Pliocene,the autochthonous Gulf Coast genera became extinct. The known species of Floridatragulus, so far restricted to the GulfCoastal Plain Miocene, are separable into three size groups that cor-respond with their stratigraphic occurrence. The smallest group isrepresented by F. nanus n. sp. from the Garvin Gully and F, barbouri,known only from Thomas Farm. The medium-sized F. texanus n. sp.and F. dolichanthereus have been recovered from the BurkevilleFauna of Texas and the uppermost strata of the Thomas Farm quarry,respectively. The stratigraphically higher Cold Spring Fauna 6fTexas has yielded a still larger form, F. hesperus n. sp., which is12 per cent larger than F. dolichanthereus and 20 per cent largerthan F. barbouri. I,l Icm FIGURE 21. Floridatragulus nanus n. sp. C UTBEG 40067-194, Type); left M„occlusal and labial views. Floridatragulus nanus new species Figure 21 TypE. - UTBEG 40067-194, a left MB· TYPE LOCALrrY. - Hidalgo Bluff, Washington County, Texas. OCCURRENCE. - Garvin Gully Fauna. ETYMOLOGY. - nanos = Gr. small, dwarf. DIAGNOSIS. - The smallest known species of Floridatragulus; in-tercolumnar tubercle strongly developed between protoconid andhypoconid of M„ weakly developed between hypoconid and hypoco-nulid, talonid of MB divided, but differs from that of other species ofFloridatragulus in having a posterior cingulum and two small tuber-cular cusps developed posteriorly between the two grinding surfaces. 1969 PATTON: TEXAS ARTIODACTYLS 171 DESCRIPTION. - This species is recognized in the Garvin Gully Fauna by a single tooth collected from the · Oakville Formation ex- posed at Hidalgo Bluff, Washington County, Texas. This specimen differs from those representing later occurring species of Florida- tragulus not only in its small size ( 20.0 mm X 10.0 mm vs. 24 mm X 12 mm for F. barbouri), but in the nature of the talonid of M3. Whereas all known species of Floridatragulus have an invaginated talonid on M~, none but F. nanus exhibits any cusp development between the two resultant grinding surfaces. Rising between the di- vision of the talonid at its posterior border are two small cuspules, ap- parently developed from infoldings of the posterior wall of the hypoconulid. Both labial and lingual cu5pules show wear. In addi- tion to the above mentioned minor cusps, a small but distinct cin- gulum appears on the talonid of MB, a feature also unique to this species of Floridatfagulus. Floridatragulus texanus new species Figures 22,24B, 24C, Table 16 TypE. - UTBEG 31190-28, a pair of lower jaws with right P2-Ma and left Ps-Pa and Mi-MB· TypE LocALITY. - Near Point Blank, San Jacinto County, Texas. OCCURRENCE. - Burkeville Fauna. DIAGNOSIS. - A medium-sized artiodactyl with brachyodont molars and a very long mandibular symphysial region; diastema between P~ and Ps equal to two thirds of that between Pi and P2; intercol- umnar tubercles occur between protoconid and hypoconid in Mi and MB; talonid of Ma divided into two separate grinding surfaces. Differs from F. dolichanthereus in having larger premolars, and in the less invaginated talonid of M3· DESCRIPTION. - Although the lower molars of Floridatragulus ~e- semble those of Prosynthetoceras, they are separable from that genus on the basis of lower crown height, a larger and divided hypoconulid on Ma, and the presence of an intercolumnar tubercle between the protoconid and hypoconid on Ms. The lower premolars are generally narrower and less reduced anteroposteriorly than those of Prosyn- thetoceras. When the complete mandible and dentition are preserved, the two genera are easily separated by the presence of diastemata separating the anterior premolars of Floridatragulus. DIscussION. - The Texas specimens of F. texanus are very closely similar to the one White ( 1940) described from the Thomas Farm 172 BULLETIN FLORIDA STATE MUSEUM Vol. 14 l l C , --i- >10(1.- 111 --- - U h L_--1 Ass I cm FIGURE 22. Floridatragulus texanus n. sp. C UTBEG 81190-28, Type); left ramus, labial and occlusal views P-MB, deposits of Florida as F. dolichanthereus. The Burkeville specimens, not known to White, differ from the Florida type specimen ( MCZ 3635) and material subsequently assigned to the same taxon in the slightly larger premolars and in the lesser development of the division of the talonid of MB, a feature White emphasized in the original description. In the Thoma5 Farm specimen the invagination of the posterior border of the talonid of Ma, a reflection of this division, is pronounced and extends from the occlusal surface down to the dorsal edge of the basal cingulum. The Burkeville specimens, which exhibit approximately the same amount of wear, show only a very slight in- folding of the posterior border of the talonid, which extends no farther down than 1.0 mm from the occlusal surface. Although the Texas and Florida specimens are very similar, the above-mentioned differences may be real and may possibly reflect a minor genetic divergence resulting from geographic isolation of indefinite, 'but probably small extent and duration. If these differences prove to be consistent, then they assume not only taxonomic, but also zoogeo- graphic importance. For these reasons, I feel that they warrant specific separation. Floridatragulus barbouri White differs from F. dolichanthereus in being slightly smaller and in the shorter diastema between Ps and Pe The anterior half, metaconid and protoconid, of Mi of the holotype (MCZ 4086) of F. barbouri is extremely foreshortened, but judging from the way it is crowded between M2 and P#, this is probably a pathologic condition. No such foreshortening appears in any other 1969 PATTON: TEXAS ARTIODACTYLS 178 of the known specimens of Floridatragulus. Whether F. dolichan- thereus and F. barbouri lived contemporaneously, or whether they represent an actual ancestor-descendent lineage is uncertain. The nature 6f the Thomas Farm deposit and the history of its excavation ( White, 1942; Bader, 1956) suggest that those forms as they are known so far were separated temporally. The largest species, F. dolichanthereus, was recovered from the uppermost layers of the deposit and has not been found in any of the deeper sediments ( White, 1942: 30). The smaller, less advanced F. barbouri was re- covered from a deeper portion of the quarry. Hence there is a reason to accept the second alternative, that F. barbouri represents an earlier group that evolved ( no evidence suggests replacement) into F. doli- chanthereus. Such an interpretation has obvious chronologic impli- cations ( see Age and Correlation). Table 16. MEASUREMENTS OF Floridatragulus dolichanthereus AND texanus n. sp MC]Z 3685 UTBEG 81190-28 Lower jaw ( Type of F. dot. ) ( Type of F. tex. ) Length P,-P, diastema 33.4 Length P:-Pi diastema 29.5 Ps-Mo 22.7 76.7 P»-P. Mi-M. 55.0 P„ length x width 12.1 X 8.9 Po, length X width 11.6 X 4.9 P., length x width 11.0 x 6.5 M„ length x width 18.0 X 8.5 18.2 X 9.1 M:,length X width 16.8 x 11.0 18.1 x 11.4 M»,length X width 25.6 X 12.6 25.2 x 12.0 Floridatragulus hesperus new species Figures 23, 24a, Table 17 TypE. - UTBEG 31219-266, an incomplete right mandible with MrMa· TYPE LOCALITY. - Donahoe Farm, near Cold Spring, San Jacinto County, Texas. OCCURRENCE. - Cold Spring Fauna. ETYMOLOGY, - hesperos = Gr. west. DIAGNOSIS. - A large species of Floridatragulus, some 12 per cent larger than F. dolichanthereus. It is characterized by a rather large, recurved metastylid on M3, and an intercolumnar tubercle between 174 BULLETIN FLORIDA STATE MUSEUM Vol. 14 LA_/ /m:1 -..$..0- Icnn 4626& #Rn.44/7) =~ r r//ry» C 221 S5 FIGURE 28. Floridatragulus hesperus n. sp. C UTBEG 31219-266, Type); right ramus, labial and occlusal views. Table 17. MEASUREMENTS OF Floridatragulus hesperus UTBEC Lower jaw 81219-266 (Type) Mi-Ma 63.01 Mi, length x width 15.1 x 11.4 M., length X width 19.2 X 12.6 Ma, length x width 29.5 X 12.8 1Approximate the hypoconid and hypoconulid, in addition to the one between the hypoconid and protoconid of M3· DISCUSSION. - This larger form from the Cold Spring Fauna oc- curs highest stratigraphically of all the Floridatragulus species and represents the most advanced stage the genus reached. In addition to its large size, the presence of the intercolumnar tubercle on the talonid of M3, the pronounced metastylid, and the better developed anterior cingulum of Ma serve to distinguish it. Suborder RuMINATA Scopoli, 1777 Infraprder TRAGuLINA Flower, 1883 Family PR~roCERATIDAE Marsh, 1891 Subfamily SYNTHEToCERATINAE Frick, 1937 The peculiarly horned synthetocerines comprise the most abun- dantly represented group of artiodactvls recovered from the Miocene 1969 PATTON: TEXAS ARTIODACTYLS 175 A B i/l 2 cm. --~u......--...0,'".-* C . - '--9 . '»fja - FIGURE 24. A. - Floridatragulus hesperus n, sp, C UTBEG 8121919-266, Type); right ramus, labial and occlusal views. B. - Floridatragulus texanus n. sp. ( UTBEG 31190-28, Type); left ramus, labial and occlusal views. C. - Floridatragulus texanus n. sp. ( UTBEG 31190-28, Type); right ramus, labial and occlusal views. 176 BULLETIN FLORIDA STATE MUSEUM Vol. 14 of the Texas Coastal Plain. They appear earliest in the Garvin Gully Fauna in moderate numbers, become dominant in the Burkeville and Cold Spring faunas, and are reduced in numbers in the Lapara Creek Fauna. Like the camels, they were a rapidly evolving group, and because they occur in a direct succession of stratigraphic units of only slightly differing chronologic age, they are parUcularly amenable to studies of microevolutionary change. Total numbers of specimens, as well as relative numbers of comparable anatomical elements, are unfortunately too few to justify a quantitative treatment of taxonomic features. Material available for study consisted of 10 incomplete jaws, 3 incomplete skulls, and numerous isolated teeth. Several complete restral and postorbital horns were recovered. The paucity of definitely assignable synthetocerine postcranial elements soon made it ap- parent that useful taxonomic criteria would necessarily have to be based on tooth and skull morphology. The degree of expressi6n of anatomical features such as accessory cusps and cingula proved to be most consistently diagnostic, and consequently the greatest re- liance was placed on them. The primary cusps showed very little variability in form, either within any one fauna or from one fauna to the next. Where vertically successive faunas were closely similar in age, taxonomic characters usually were not separable in absolute terms, but differed only in degree of their expression. Thus through- out the systematic description, separation is made on the relative prominence, or stage of development, of particular diagnostic features. .-I ~'EEE A , .< i·8 Ia, Icm , 1 , .1, B Ni , FIGURE 25. Prosynthetoceras texanus ( TAMU unnumb.); right upper dentition, occlusal view P,-M», lingual view P,-M:,. 1969 PATTON: TEXAS ARTIODACTYLS 177 The synonymy of Prosynthetoceras species has been treated previously ( Patton, 1966 ( 67) ; 1967). Since completing this manuscript, I have had the opportunity to begin a new study with Beryl Taylor of the American Museum of Natural History ( Patton and Taylor, MS) on synthetocerines in the Frick Collection which were not available at the time this was written. Prosynthetoceras Frick 1937 Prosynthetoceras texanus (Hay) Figures 25, 26, 27, Tables 18, 19 Dromomeryx texanus Hay, 1924, p. 15, pl. II, figs. 8-12. Dromomeryx angustidens Hay, 1924, p. 16, pl. II, figs. 6,7 Meri/codus grandis Hay, 1924, p. 17, pl. III, fig5. 9-11. Protolabis francist Hay, 1924, p. 14, pl. III, figs. 5-8. cf. Mit>labis sp. indet., Simpson, 1982, p. 37. ?Cranioceras texanus (Hay); Frick, 1987, p. 82,97. ?Synthetoceras rileyi Frick, 1987, p. 608. 605, fig. 66. Blastomeryx texanus (Hay); Wood and Wood, 1937, p, 187, pl. I, figs. 5,6 Synd#oceras australis White, 1941, p. 97, pl. XV, figs. 1, la, 2, 2a. ?Syndyoceras texanus ( Hay); Hesse, 1942, p. 168 ( ?Syndeoceras texanus, p. 167, lapsus) Sgnthetoceras (Prosynthetoceras) douglasi White, 1947, p. 504, fig. 33. Nothoken= grandis White, 1947, p. 508 (in part). cf. Miotabis tennis Ray, 1957, p. 18. TypE. - TAMU 2387, a right Ms. TypE LOCALITY, - Garvin Farm, near Navasota, Grimes County, Texas. OCCURRENCE. - Garvin Gully Fauna, Burkeville Fauna. REFERRED MATERIAL. - UTBEG 31190-61, maxilla with P)-MB; TAMU unnumb., partial right maxilla with alveolus for P2, and Pa-MB; UTBEG 31190-31, partial skull; 31190-75, partial rostral horn core; 31084-81, anterior portion of left ramus with P.-Mi; 31190-95, left P#-M,; 31190-102, left Ml-MS; SMU 60898, left P<-M3; and numerous isolated teeth. DIAGNOSIS. - A relatively small synthetocerine with short-shafted rostral horn and subhypsodont teeth. P2 retained. Upper molars are short, broad, and strongly tapered toward occlusal surface; they are characterized by prominent medial V-shaped pillars and bladelike cingula, anteriorly and posteriorly. There is a pronounced develop- ment of styles, especially the mesostyle. Pi present. Pz is reduced, but two-rooted. 178 BULLETIN FLORIDA STATE MUSEUM Vol. 14 DESCRIPTION. - P. texanus was originally described by Hay ( 1924)as Dromomeryx texanus on the basis of a right Ma and two right Ml's. Frick ( 1937) described Synthetoceras rilegi ( = P. texanus) on the basis of a partial mandible and left Mi-Ms. Specimens described by Patton ( 1967) and in this paper provide the only knowledge to date of the cranial armament of this species. Because no anterior portions of the texanus jaws were recovered, Pi is not available for description. However, because it is present in Syndyoceras cooki and P. francisi, it doubtless occurred in P. texanus. P2 is represented only by two alveoli in the single specimen of P. teranus with that portion of the maxilla preserved ( TAMU unnumb.). Ps is elongated anteroposteriorly; the paracone is high and sharp, and the recurved metastyle forms a small cusp posteriorly. The paracone has no rib. A large cingulum extends from the parastyle lingually to join the metastyle at the posterior border of the tooth. P* has both parastyle and metastyle well -developed. The metastyle is more prominent and more strongly recurved than the parastyle. The paracone is separated from the protocone by a deep valley. A rel- atively large accessory cusp extends up from the base of the posterior limb of the protocone, and a smaller cusp is apparent at the base of the anterior limb. All upper molars display strong stylar cusps on the anterior and posterior crests. The entire anterior half of each molar is set farther lingually than the posterior half. This has resulted in a very prom- inent mesostyle and a rather deep recess anterior to it at the junction of the paracone and metacone crests. A strong rib extends up each crest from the base of the crown to the tip of both paracone and metacone. The shape of the crescents varies from an open V in MB to 'a more closed, slightly recurved V in Mi. Highly developed median pillars occur between the anterior and posterior crescents and extend half the height of the crescents in unworn teeth. This relationship is generally maintained with wear; the pillars are worn at virtually the same rate as the crescents. The pillars form an inverted V as both limbs hug closely their respective crescents. On most of the molars these pillars exhibit some wear. On the anterior margin of each molar is a narrow winglike cingulum which projects from just above the crown near the apex of the protocone and extends labially along the side, becoming fainter as it approaches the anterior crest. This feature is present but not so strongly developed along the posterior side of the tooth. The Ml and M2 are broader Table 18. MEASUREMENTS OF THE LOWER DENTITION OF SPECIES OF Prosynthetoceras 1969 PAT TO N : TE X A S A R TIO D A C TY LS 179 P, P» P, M~ M: M. IC-P, UTBEG no. L W L W L W L W L W L W PrM, P,-Ma Mi-Ma diastema P texanus ( G. Gully) SMU 60898 8.4 8.9 9.8 5.0 10.2 6.7 14.1 9.8 16.1 11.8 27.6 11.6 71.5 62.2 54.9 87.8 P. texanus ( Burkeville) 31190-95 10.5 6.1 14.1 10.1 17.4 11.4 26.1 10.8 56.1 81190-102 14.9 10.2 17.1 11.8 26.8 12.1 56.01 P. francisi ( Cold'Spring ) 31219-218 9.9 5.2 10.5 7.0 16.4 11.1 22.0 12.7 29.2 12.7 85.0 74.6 66.0 81219-214 11.1 7.8 16.2 12.1 21.6 14.0 29.8 18.1 85.61 76.2 66.0 66.1 31219-219 11.8 8.1 12.5 21.5 14.5 88.0 18.0 76.5 68.0 31248-5 10.0 5.1 10.5 6.8 15.6 11.8 20.6 18.7 80.0 18.4 88.6 74.5 65.1 81182-47 17.4 10.71 21.8 12.0 31.2 11.6 69.4 1Approximate Table 19. MEASUREMENTS OF UPPER DENTInON OF SPECIES OF Prosynthetoceras 180 B U LLE T IN FLO R ID A S TATE M U S E U M Vol. 14 P' F P' M~ M2 Ma UTBEG no. LW LW L W L W L W L W F.P' F.P' Mi.M' F.MS PS-M' P. texanus ( G. Gully) TAMU unnumb. 9.7 7.2 8.7 11.5 18.8 16.4 17.5 17.9 17.6 15.4 19.5 46.7 64,7 40067-54 8.9 10.0 P. texanus ( Burkeville) 31190-61 6.4 7.0 9.7 11.0 14.6 15.5 18.2 18.8 19.8 18,7 20.6 51.0 68.5 81190-31 17.2 19.9 20.0 22.0 31190-62 16.7 I7.0 17.5 17.6 31190-55 9.6, 12.0 18.6 19.6 81190-39 17.8 18.2 P. francisi ( Cold Spring) 81219-241 9.5 5.6 11.1 6.8 11.0 18.1 18.5 18.6 22.5 20.2 22.1 16.6 80.7 22.1 68.0 94.6 84.4 81219-257 23.3 20.8 81219-254 23.1 20.1 31219-264 20.81 19,6 21.6 20.4 TAMU unnumb. 9.1 5.5 10.2 7.2 11.0 18.0 16.1 16.8 21.2 19.8 22.8 20.8 27.8 19.5 58.8 84.0 76.8 1Approximate 1969 PATTON: TEXAS ARTIODACTYLS 181 than they are long, while all three molars show strong tapering toward the occlusal surface. The lower premolars are all reduced, especially PB Pz is two- rooted and is Similar to but smaller than PB. The primary differences between 1.2 and Pa are the greater expression of the anterolingual flexid and the greater separation of hypoconid and entoconid in the latter. P3 is elongate and is but slightly broader posteriorly than anteriorly. The hypoconid and protoconid are more distinct from the metaconid than in P4. As in P# swellings at the base of the hypoconid and protoconid are noticeable. P4 is slightly wedge-shaped. Most of the reduction in this tooth has occurred in the region of the protoconid. The metaconid is high and sharp; the entoconid is narrow and bladelike. The hypoconid extends posteriorly and lin- gually to join the entoconid. On worn specimens the metaconid, entoconid, and hypoconid join to form a small enamel islet. The base of the hypoconid is swollen to form a bulge at the posterolabial edge of the tooth. There is a less distinct swelling at the base of the protoconid. The lower molars of P. texanus are subhypsodont. The crescents ( protoconid, hypoconid) are symmetrical and show little tendency towards recurving. Stylids are reduced in comparison with those on the upper molars; only the metastylid is pronounced. The anterior crescent ( metaconid) overlaps the posterior crescent ( entoconid) slightly to form a faint stylid. A small median pillar occurs at the base of the crown between the anterior ( protoconid) and posterior ( hypoconid) crests of Mi and Mo. No accesory pillars appear on Ma· In contrast to the median pillars on the upper molars, the lower ones would not show wear until the teeth were worn almost to the base of the crown. No cingula are present on the lower molars. Perhaps the most significant characteristics of the lower molars is the presence of a double enamel loop on the hypoconid of M3 which, with wear, encloses a fossettid. This feature occurs in all protoceratids (and other ruminants) but never is found in camels. In this regard, Ms of Prosynthetoceras resembles that of Florida- tragulus, which is often found in association, but is distinguishable from it on the basis on the strong anterior cingulum, the accessory pillar between the protoconid and hypoconid, and the lower crown height of the latter. Another contemporary genus, Nothokemas, has a similar M3, but in this instance, the f6rmation of the double enamel loop is accomplished by the posterior extension of the metaconid. 182 BULLETIN FLORIDA STATE MUSEUM Vol. 14 ' A-1 4 I cm. FIGURE 26. Prosynthetoceras texanus ( UTBEG 81190-61); occlusal and oblique occlusal views. DIsCUSSION. - The lengths of the type specimens of P. texanus and the specimens from the Garvin Gully Fauna referred to it are within the range of published measurements of S!/nduoceras cooki ( Barbour, 1905; Frick, 1937). Specimens from the Burkeville Fauna are clearly larger. Unfortunately the teeth of the only available specimen of S. cooki are very badly worn, but as far as can be determined they approach closely those of P. texanus. Equally lament- able is the fact that to date no complete synthetocerine horns are known from the Garvin Gully Fauna, and this appears to be the most useful taxonomic character in the separation of the two genera. Three horn fragments (TAMU 2395) from Garvin Gully described by Hay ( 1924) as the type of Merycodus grandis appear to be fragments of the rostral horn of P. texanus. The largest fragment consists of a part of the main beam and one of the two branches, the other branch having been broken off. As is true of synthetocerines, the horn is convex on one side and concave on the other, with the branch leaning posteriorly. The size of this horn corresponds with that expected for the rostral horn of an early P. texanus. If the horn does belong to P. texanus, it can be used to demonstratate a definite advance over the rostraI armament of Synduoceras, for the point of branching is considerably higher on the beam than in Syndljoce«ras. Rostral horns of P. texanus from the Burkeville Fauna ( UTBEG 1969 PATTON: TEXAS ARTIODACTYLS 188 31190-31, -72) are demonstrably much more advanced than those of either Syndyoceras or those specimens of P. texanus from the Garvin Gully Fauna. In addition the premolars of P. texanus appear to have undergone approximately 10 per cent greater reduction than those of Syndyoceras, indicating that the former had developed further along the lines characterizing the later synthetocerines. Frick ( 1937: 605) originally separated the subgenus Prosynthe- .. toceras from the genus Synthtoceras on the basis of smaller size, the smaller and shorter-crewned dentition and retained P2" in Prosyn- thetoceras. It has been demonstrated above that specimens confi- dently assigned to P. texanus (=(?) Synthetoceras rileyi Frick ) have the alveoli for P2; thus P. texanus now can be included in his subgenus Prosynthetoceras. Because I believe the retention or loss of p2 in this linage to be an important and useful taxonomie character most applicable above the species level, I elevated the subgenus Prosyhthetoceras Frick to the rank of genus ( Patton, 1967). The protoceratids White ( 1941, 1947) described from the Thomas Farm of Florida as Prosynthetoceras douglasi ( MCZ 4065, White, 1947) and Syndyoceras australia ( MCZ 3654 and 3642, White, 1941) are indistinguishable from P. texanus on the basis of the material presently available. For this reason, I have referred these species to synonomy with P. texanus ( Hay) ( Patton, 1966( 67).). The Burkeville specimens ( UTBEG 31190-31, a partial skull with horns, left Mi and right M2-Ma; UTBEG 31190-61, maxilla with left Ml-MS, and right PB-MB) are intermediate in size, crown height, and complexity of molars between P. texanus from the Garvin Gully Fauna and P. francisi from the Cold Spring Fauna. The metastyle of p* of the Burkeville P. texanus is large and strongly recurved; the parastyle is small. A tiny tubercle is present at the base of the anterior limb of the protocone. This tooth is more elongate and appears slightly more wedge-shaped than that of the Garvin Gully P. texanus. The metastyle on P< is also large and recurved. Cingula and intercolumnar pillars are less strongly developed than in the Garvin Gully P. texanus, more so than in P. francisi. The shape of crests and crescents of the molars of the Burkeville form are similar to those of the Garvin Gully. The posterior cingulum, when present, is tight against the side of the tooth, is relatively pronounped lingually, and becomes faint as it approaches the anterior edge of the metastyle; these cingula are only faintly visible on Mi-M2 of UTBEG 31190-61. The anterior cingula are strongly pronounced lingually and rise 184 BULLETIN FLORIDA STATE MUSEUM Vol. 14 over half the length of the anterior limb of the protocone forapproximately one-third the width of the tooth, then dip down closeto the base of the crown, becoming increasingly faint towards theparastyle. The median pillars are strong and folded into limbsthat abut closely on their respective crescents. Wear on the pillarsis proportional to the degree of wear on the primary cusps. Thestyles are very well-developed, especially the mesostyle. The post-orbital horns are laterally directed. The shaft of the rostral horns ofthe Burkeville P. texanus is longer than that of the Garvin Gully Iform, attaining a length of approximately 135 mm from the base tothe point of bifurcation. P. texanus specimens from the Burkeville Fauna mark an inter-mediate stage in the evolution of this group of synthetocerinesbetween that of earlier P. texanus and P. francisi of the Cold SpringFauna. An analysis of all the features in the teeth, skulls, and jaws,excepting those whose comparison was prevented by lack of material,indicates a remarkably close relationship between the Garvin Gullyand Burkeville populations. This affinity is witnessed in the inter-gradation of every comparable diagnostic character available in the two series of specimens. Although P. texanus from the BurkevilleFauna more closely approximates P. texanus from the Garvin GullyFauna than P. francisi in size as well as in evolutionary development,separation of these two forms nevertheless involves in part an arbi-trary taxonomic decision. For example, Figure 30 illustrates thedistribution of lengths of left Ma of synthetocerines from the fourbiostratigraphic units included in this study. The separation of P.francisi from the Burkeville P. texanus is apparently consistent and Ireliable. However, there is a moderate overlap in the observedranges of MB lengths of the Garvin Gully P. texanus and the Burke- ville P. texanus, with the Burkeville form showing a definite sizeincrease over the one from Garvin Gully. But although the meanlengths ( Garvin Gully 22.2 mm, Burkeville 23.9 mm) of M3 in the two forms are distinct, the degree of distinction is less than that between the succeeding forms. This same pattern is observed in othersynthetocerine characters. Though characters are not necessarily of equivalent value in different species or subspecies, in a relatively uninterrupted lineagesuch as this one they are almost equally diagnostic. A single changinglineage of successive species, "chronocline" of Simpson ( 1961), offers few nonarbitrary ways to arrive at taxonomic subdivision. Yet becauseof the general biostratigraphic, zoogeographic, and evolutionary util- 1 1969 PATTON: TEXAS ARTIODACTYLS 185 ity derived from segmenting a continuous linage, it is often desirable to do so. It has been possible to recognize certain shifts in the patterns of synthetocerine characters from the Garvin Gully Fauna to the Burkeville Fauna, and because these shifts may be important in interpreting the evolution of this group of artiodactyls, they may eventually warrant separate taxonomic status for the two forms. The Burkeville specimens may represent a temporal or successional il (di: i 1 ; C - ./.~1215<,/.rizy,- ) L_-J 2 cm FIGURE 27. Prosynthetoceras texanus C UTBEG 81190-31); partial skull and postorbital horns. 186 BULLETIN FLORIDA STATE MUSEUM Vol. 14 subspecies of P. ten:anus. This is based on the range of variation of morphological characters and because the degree of their differentia- tion is significantly smaller than the sequential differences observed in other synthetocerine species. However the limited sample of specimens now available for study does not justify separate nomen- clatorial status for the Burkeville specimens at this time. Prosynthetoceras francisi Frick, 1937 Figure 28, Tables 18, 19. TypE. - TAMU unnumb., a skull with canine alveolus, Pz-Ms,rostral horn and bases of paired horns. TYPE LocALITY. - near Cold Spring, San Jacinto County, Texas. REFERRED MATERIAL. - UTBEG 31219-214, partial left mandible with /C and P4-M3; 31219-213, partial left mandible with PrMa;31243-5, partial right mandible with PB-MS, 31183-47, partial left mandible with Mi-M3; 31219-236, rostral horn; 31219-219, partial right mandible with P4-Ms. OCCURRENCE. - Cold Spring Fauna. DIAGNOSIS. - A relatively large synthetocerine with hypsodont molars. P2 highly reduced. The rostral horns are heavy and long; shaft longer than that of P. texanus,· the paired postorbital horns are directed laterally and to the rear ( posterolaterad). Other characters are described by Frick ( 1937: 605) as follows: The type cranium ( Figures 2, ZA, 60-68), as noted above differs from ( S!/n-thetoceras tricornatus) in its considerably smaller size and notably smaller andshorter crowned dentition and retained F. What evidently is the alveolus of anenlarged canine occurs at the anterior base of the rostral horn. The cranialcharacters are better shown than in several of the ( Synthetoceras tricornatus)skulls. Lacrimal' vacuities, as in the case of Protoceras, are absent. Theinverted U-shaped supra-occipital area, tlle peculiar paraoccipital processes and peculiarly constructed bullae are well shown. The paired postorbital horns arebroken; the rostral horn is proportionately heavy and long-forked. DESCRIPTION. - Other than size, the most striking differences in the skull and horns between texanus and francisi is the backward sweep of the frontal horns and the proportionately longer shaft. of the rostral horn in the latter. Other skull comparisons are precluded by lack of diagnostic specimens of texanus. The upper molars of P. francisi are approximately 20 per cent larger than those of the earlier Burkeville P. texanus. Although the teeth of both the type and the most complete specimen ( UTBEG 31219-213) in the Univer- sity of Texas collection are worn, they appear to have been more 4 1969 PATTON: TEXAS ARTIODACTYLS 187 hypsodont than forms lower in the section. The greatest changes have affected the accessory cusps; the posteri6r cingula are no longer visable and the anterior ones are considerably reduced. The median pillars, which are especially prominent in the molars of P. texanus, have undergone reduction. The second and fourth premolars of P. francisi are proportionately smaller than those of the Garvin Gully and Burkeville forms. This is equally true of the third premolar of the type of P. francisi. However, in other specimens confidently assigned to P. francisi and also in the large Synthetoceras tricornatus, the PB is anomalously robust. D '1 ..~ 4542:30*R+4~~,,, *dS.%4 R..*r--LE-«Al-~ 2 cm FIGURE 28. Pfosynthetocems francisi C UTBEG 81219-218); left ramus, occlusal view P»-Ms, labial view of ramus and PrMs. Five rather well preserved lower jaws ( UTBEG 31219-213, 31219-214, 31183-47, 31243-5, 31219-219) known from the Cold Spring Fauna provide an adequate basis for specific identification. In four of the five specimens the complete cheek dentition is present - or indicated, and consists of three premolars and three molars. The caniniform Pi and the base of the incisiform canine are preserved on UTBEG 31219-214, a left mandible, Pi is seldom present. On UTBEG 31219-213, also a left mandible, most of the diastema and symphysis remains intact. The mandibular diastema of P. francisi is considerably longer than that of P. texanus, corresponding to a similar increase in diastema length of the upper jaw. Because the type specimen of P. texanus is broken anterior to P~, the length of the diastema cannot be determined precisely, but judging from sym- metry alone, it appears to be relatively shorter than the later forms. The diastemata of Protoceras and Syndyoceras are much shorter than any of the succeeding forms. Each ramus of P. francisi, except the type described by Frick ( 1937), has a P2 or an alveolus for P2, 188 BULLETIN FLORIDA STATE MUSEUM Vol. 14 but this tooth, when present, is highly reduced. In UTBEG 31219-213 are two small alveoli for P2 whereas UTBEG 31219-214 has only one alveolus. In all other respects the specimens are virtually identical and afford no reason for taxonomic separation. In view of the gradual reduction of the P2 in earlier protoceratids and its complete loss in Synth€toceras tricornatus C based on FAM 32468 from the Clarendon beds), variation in occurrence of this tooth ( and Pi) among individuals within any one population is not surprising, indeed is to be expected. Pa of P. francisi is virtually indistinguishable from the earlier forms. With the exception of the greater recurving of the protoconid, which forms a distinct recess lingually between it and the metaconid, no significant difference in morphology is noted in P+ M2 has a small accessory tubercle at the base of the posterior limb of the protoconid. The MB of P. francisi is high-crowned and lacks any trace of accessory cusps. ~Irgm ' WITA MER51 Al[ZIX ZIP' Upy#f".il'Br L-A-/ L-4-J IcnnA /, A ,18 B '+A H• 111111- 11~11!111 1~Z31//ffWW 1% 111~, 5es¢, \W13: iB:z 21 f J#Y hiss ,-*r#~ '2-3/~13 FIGURE 29. Synthetoceras tricornatus C UTBEG 31182-568); right M., occlusal and lingual views. 1969 PATTON: TEXAS ARTIODACTYLS 189 Synthetoceras Stirton, 1932 Synthetoceras tricornatus Stirton, 1932 Figure 11B, 11C, Table 20 TypE. - Skull, UCMP 31520. OCCURRENCE. - Lapara Creek Fauna. REFERRED SPECIMENS. - UTBEG 30936-214, 31152-563. DISCUSSION. - Stirton ( 1932 ) described a new genus and species of artiodactyl from the Clarendon beds of the Texas Panhandle that subsequently became the type genus of a new subfamily of Proto- ceratidae, the Synthetoceratinae ( Frick, 1937). This nearly complete skull represents the most advanced stage of synthetocerine evolution now known. This species is recognized on the basis of a right Ma ( UTBEG 30936-214) from Live Oak County, and a right M2 ( UTBEG 31132- 563) from Bee County. To date no other synthetocerine remains are known from the Pliocene of the Texas Coastal Plain. Younger Hemphillian MeGehee deposits of Florida have so far yielded only two teeth, both of which are inseparable from Clarendonian speci- mens. The molars from the Lapara Creek Fauna are also inseparable from those of the Clarendon beds. Unfortunately, no better conn- parisons of the forms from these two areas can be made without additional material, especially that involving skull and cranial arma- ment. The Lapara Creek specimens are virtually identical in size and morphology with the previously mentioned two molars ( UF 10044, 10045) recovered from the McGehee deposit ( Hemphillian) of north central Florida. Until more material is forthcoming from the Mc- Gehee quarry, it is probably safe to include that form also in S. tricornatus. Table 20. MEASUREMENTS oF UPPER MOLARS OF Synthetoceras tricornatus Specimen no. Length Width UTBEG 80936-214, RM' 29.3 26.4 UTBEG 81132-568, RM' 29.2 25.2 UCMP 81520, RM' 29.8 25.6 UCMP 81520, LMs 29.6 25.5 UF 10044, RM' 27.0 25.6 UF 10045, RM' 80.0 25.5 190 BULLETIN FLORIDA STATE MUSEUM Vol. 14 S. tricornatus 00 (Lopora Creek) P.francisi 8 8(Cold Spring) - P .texanus •1 -:••(Bufkeville) P . texanus X XX X X (Garvin Gully) ..... 20 25 30 3~Length of Ma in mm FIGURE 30. Distribution of measurements of M, lengths in Gulf Coast syntheto- cerines. PHYLOGENY OF THE SYNTHETOCERATINAE The series of protoceratids collected from successive biostrati- graphic units in the Texas Coastal Plain Miocene exhibit characters that occur in a steplike progression thought to represent their actual phylogeny. Remarkable uniformity in this group exists within each fauna; although sample size is small, the ,degree of variability found among specimens from widely separated localities is comparable to that expected in a single modern pollulation of ruminant artiodactyls. No extreme variants have been recovered to indicate the presence of divergent geographically -separated populations. Thus it appears that these animals at any one time were either members of one large interbreeding population, or perhaps were distributed in very closely related allopatric populations with occasional opportunity for contact, preventing the accumulation of isolating mechanisms. All evidence suggests-that the sequence of synthetocerines from the Garvin Gully, Burkeville, Cold Spring, and Lapara Creek Faunas represents a continuously evolving lineage of huccessive ancestral-descendant pop- ulations, with evolution proceeding in situ by direct, sustained inherit- ance from one generation to the next, involving a slow shift of the adaptive optimum. Dispersal, with the obvious exception of the initial immigration into the Coastal Plain, apparently entailed only emigration from this 1969 PATTON: TEXAS ARTIODACTYLS 191 region, as no discontinuities ascribable to replacement by adaptively superior allochthons are evident. The earliest protoceratids are known from the Great Plains of North America. The genus Protoceras is generally thought to be the ancestral group from which the later forms were derived. Among these, Protoceras nasutus of the White River appears to be the most likely progenitor of the early Miocene Syndyoceras. The skull of the genus Protoceras is characterized by the development of bony protuberances rising from the temporal crests of the parietals, from the supraorbital borders of the frontals, and from the upper margins of the maxillaries. These processes are represented in each Protoceras species, although suppressed to a greater extent in what are believed to be females. In P. celer the maxillary protuberances are broad and platelike, curve outward, and are widely separated at the base. In P. nasutus they are much narrower, more rodlike, and unite suturally over the nasal canal ( Scott, 1940: 577). It is in this character especially that P. nasutus appears to anticipate Syndyoceras. The genus Synd!/oceras has four prominent horns: the rostral or maxillary pair join over the nasal canal to form a common trunk, then bifurcate just above the juncture into two outwardly curving branches; the two frontal horns rise from a point above and behind the orbits and : curve gently towards the midline. The kind of rostral horn seen in Syndyoceras appears to have evolved from the union of the two maxillary protuberances of the type seen in Protoceras nasutus. The frontal horns, on the other hand, were probably derived from the union of the bony processes on the supraorbital borders of the frontals and those on the temporal crests of the parietals, forming a single beam. P. nasutus satisBes those requirements necessary for being placed in a morphologically ancestral position to Syndyoceras and the later synthetocerines. The lack of a significant number of frontal ( postorbital) and nasal ( rostral) horns in any of the availible collections makes difficult the detection of any evolutionary trends in these parts. It is difEcult to determine the constancy of anatomical details observed in contem- poraneous specimens because of the very small samples available from any one fauna, it is especially difficult when comparing speci- mens collected from each of the successive faunas. It becomes easy to confuse variability with phylogeny. Notwithstanding this com- plication, the ancestral horn condition for this lineage seems dis- cernible in Syndyoceras. The postorbital horns of Syndyoceras extend laterally at nearly right angles to the long axis of the skull and curve 192 BULLETIN FLORIDA STATE MUSEUM Vol 14 up and towards the midline in a wide semicircle. At the base, these horns tend to be slightly flared and flattened, but become more rounded toward the middle. No bony protuberances occur along their edges. The only recovered postorbital horns of P. texanus share several characters of Sundyoceras. The horns are directed laterally from the skull and recurve rather sharply toward each other. They differ from those of Synd!/oceras in the slightly greater flare near the base and in the increase in irregularity or roughness of the anterior and posterior edges at the flare. P. francisi shows considerable change from the preceding forms; the horns sweep out fr6m the skull at a low angle and in a posterior direction, have greater Hattening towards the base, and the anterior and posterior edges are marked by more prominent irregularity. This progression in characters of the horn _co-Btinues into Synthetoceras. The evolution of the rostral horn of the synthetocerines has in- volved primarily the lengthening of the main shaft, resulting in an increase of the distance from the base of the rostral horn to the point of bifurcation. There also seems to have been a gradual increase in backward growth of the distal, or forked, end of the rostral horn. This is perhaps a counterbalancing adaption serving partially to overcome the disadvantage of having such a heavy weight towards the tip of the snout. Thus from the preceding discussion and from the data included in the systematic description of these forms, a morphologic, and presumably phylogenetic, series representing the steps in protoceratid evolution from late Oligocene through early Pliocene can be estab- lished. On the generic level this series is as follows: Protoceras-Syndyoceras-Prosynthetoceras-Synthetoceras.1 Several morphologic trends from Synduoceras of the early Miocene to Synthetoceras of the early Pliocene can be discerned. ( 1) increase in size (2) reduction of premolars ( 3) reduction of intercolumnar tubercles and cingula in molars (4) increase in crown height ( 5) increase in diastema length iAs this paper was going into press, Stirton's ( 1967) description of a new genus, Lambdoceras, was published. This genus from the middle Hemingfordian Flint Hill fauna of South Dakota, is in iny opinion very difficult to separate from Prosynthetocems. I have not had the opportunity to study Stirton's specimens; however, the new taxon is discussed in Patton and Taylor ( MS). 1969 PATTON: TEXAS ARTIODACTYLS 198 ( 6) more posteriorly directed postorbital horns ( 7) increase in irregularity of lateral edge of postorbital horns towards the base ( 8) increase in distance from base of rostral horn to point of bifurcation ( 9) gradual develepment of bony covering of canal for nerves and blood vessels supplying rostral horns Infraorder PECORA Linnaeus, 1758 Family CERvIDAE Gray, 1821 Subfamily PALAEOMERYCINAE Matthew, 1904 Blastomeryx Cope, 1877 Blastomeryx elegans Matthew and Cook, 1909 Table 21 TypE. - AMNH 1401, a partial left ramus, with PS-MB· OCCURRENCE. - Lapara Creek Fauna. REFERRED SPECIMENS. - UTBEG 31132-331, partial left ramus with MrMa; 31132-4, dPL 30896-326, partial right ramus with P3-M2 and broken Ms; 30896-496, partial right ramus with PrMB. DEsCRIPTION. - The four referred specimens from Bee County, Texas, are indistinguishable from a series ( AMNH 17345, 17341, 17347, 17348, 17349) collected from the Snake Creek beds of Sioux County, Nebraska, and designed by Matthew and Cook ( 1909) as Blastomeryx elegans. Frick ( 1937) separates the genera Blastomeryx and Longirostromeryx on the basis of the latter's elongate diastema and greatly reduced PG-PI No symphysis is preserved in any of the Lapara Creek specimens, but the premolars are not markedly reduced as in species of Longirostrome,yx. P, is represented by two alveoli. Table 21. MEASUREMENTS OF Blastomeryx elegans AMNH 14101 UTBEG UTBEG UTBEG Measurennent ( Type) 80896-496 86896-326 31132-321 Length Pa-P, 14.8' 14.5 13.8 Length PrMi 45.4 48.4 Length Mi-M. 32.0 80.7 P., length X width 6.2 X 8.0 6.5 x 8.2 6.8 X 80 P#,length X width 7.8 X 4.0 8.2 x 3.5 8.0 X 2.8 M„ length X width 9.0 X 5.2 88 X 4.7 8.4 x 4.8 M:,length X width 10.3 X 5.9 9.0 X 5.9 8.6 x 5.3 9.5 X 5.7 M., length X width 13.9 X 5.6 14.0 x 5.5 12.42 x 5.5 1From illustration (Matthew and Cook, 1909 : 410, fig. 23) .Approximate 194 BULLETIN FLORIDA STATE MUSEUM Vol. 14 Longirostromeryx Frick, 1937 Longirostromerux blicki Frick, 1937 Figure 31, Table 22. TypE. - FAM 31734, a right ramus with symphysis and P2-Ma· OCCURRENCE. - Cold Spring Fauna. REFERRED MXCERIAL. - UTBEG 31183-35, a partial skull with PB-Ma a right ramus with P#-M:{, and two upper canines. DIsCUSSION. - The Cold Spring specimens compare closely to specimens Frick ( 1937 ) assigned to (?) Longirostromerux blicki from North Santa Clara River, New Mexico. In addition to the elongate diastema of Longirostromervx, the genus is also characterized by reduction of P2-P~. ( ?)L. Nicki is known to have an elongate dia- stema, but apparently because it has comparatively large premolars, Frick ( 1937 ) assigned it questionably to Longirostromeryx. f j u 7~.6.4.~f~r --5< ~fr/~.f~2 cm FIGURE 81. Longirostromeryx blicki ( UTBEG 81188-35); occlusal view of partial skull with upper dentition, lateral view of left C/, labial view of right ramus. , 1969 PATTON: TEXAS ARTIODACTYLS 195 The genus Longirostromeryx as defined by Frick ( 1937) includes seven species ( two questionably assigned) : L. wellsi ( Matthew ) ; L. merriami Frick, the type species of Longirostromeryx; L. serpentis Frick; L. clarendonensis Frick; L. nooomexicanus Frick; ( ?) L. vigor- atus ( Hay); and ( ?) L. blicki Frick. I can see no substantial differ- ence between L. weUsi and L. merriami in size or morphology and believe the two species should be synonomized under L. wellsi. The teeth of L. serpentis are'Very worn so that litle cusp morphology remains visible, but the shape and dimensions of the teeth and jaws are comparable to L. we.Zlsi and also should probably be synonymized with that species. L. nouomexicanus from the Santa Cruz beds of New Mexico is based on a partial left ramus with P2-Pa alveoli and P#-MB· Because neither of the two diagnostic characters of Longirostromeryx, reduc- tion of ]?2-Pa and elongate symphysis are present on the type and only specimen, no clear reason exists for including it in this genus. ( ?) L. vigoratus was originally described as Blastomeryx vigoratus by Hay ( 1924) on the basis of a left M2 and left Ma. Again no con- elusive evidence exists for assigning it to either genus. Faced with a similar dilemma in constructing a tenative arrange- ment of New World testudinines and supposed testudinines, Williams 2 ( 1950) states: Dr. Paulo Vanzolini of the Department of Zoology, State of San Paulo, Brazil, has suggested to the author that specific names of this nature to be designated species inquirendae. This nomenclatorial category seems to be unusual or unknown in North American taxonomy. It would, the author believes, serve - a useful purpose. Certainly names of possible but presently indemonstrable validity cannot properly be treated as of equivalent value with adequately based, well-known taxonomi6 categories, nor is it at all reasonable to place such names in synonymy, since this could be done only arbitrarily and doubtfully. There is evident need for a purgatory in which such names may reside until proper ' disposition can be made for them. The International Code of Zoological Nomenclature ( 1961) authorized such a procedure, and under the category species inquiren- dae I prefer to include the taxa L. nouomexicanus and (?)L. vigoratus. L. clarendonensis represents the most advanced stage of Longiro- stromeryx development in premolar reduction . P2 tends to be lost and Pa-1?4 are greatly reduced. (?)L. blicki, on the other hand, was only questionably assigned to Longirostromeryx by Frick ( 1937) apparently because of its comparatively unreduced premolars. The 196 BULLETIN FLORIDA STATE MUSEUM Vol. 14 Table 22. MEASUREMENTS OF Longirostromeryx blicki Measurement UTBEG 81183-85 P3-M, 36.4 31.9Mi-Ma 26.2 F, length x width 4.6' X 5.71P*, length x width 6.0 X 6.6Mi, length X width 8.7 X 9.1M', length x width 10.0 X 10.2M; length X width 9.2 X 9.9P.-M» , 89.0 Mi-M, 31.6P., length X width 7.8 X 4.0Mi, length X width 8.7 X 5.6M:, length x width 9.6 X 6.2Ms, length X width 14.0 X 6.1 1Approximate elongate diastema in this species is well established. Because pre- molar reduction appears to be a progressive character in this genus, it is not surprising to find Barstovian specimens with less premolar reduction than those from Clarendonian deposits. Nor would it be unexpected to find forms conservative in premolar reduction con- temporaneous with the more progressive species in this regard. Longirostromeryx is probably better defined on the basis of the great length of the mandibular symphysis although, as Gregory ( 1942) points out, the extent of variation of this character is as yet undetermined. For this reason I prefer to consider the Cold Spring specimens and those Frick ( 1937 ) referred to (?)L. blicki as belonging to Longirostroineryx and remove them from their questionable status within that genus. ( ?) Longirostromeryx vigoratus ( Hay) Blastomeryx vigoratus Hay, 1924, p. 16, pl. II, figs. 13, 14. TypE. - TAMU 2378, left M2-M3· OCCURRENCE. - Garvin Gully Fauna. DISCUSSION. - Hay ( 1924) described Blastomer[/x uigoratus on the basis of six teeth collected from the Garvin Farm quarry. He characterized the species as being of nearly the same size as B. wellsi Matthew ( Now Longirostromerilx wellsi) from the late Miocene of South Dakota, although Mi of wellsi is no longer and wider ( 11 X 8 mm) than that of vigoratus, which is 10 mm long and 6.3 mm wide. 1969 PATTON: TEXAS ARTIODACTYLS 197 L. wellsi differs from vigoratus also in having more hypsodont teeth. Frick ( 1937 ) has since listed this species as (?) Longirostromeryx oigoratus (.Hay), although the length of its symphysis and propor- tions of its premolars are still unknown. Perhaps this decision was based on the presumed amnity of (?)L. vigoratus to L. wellsi, whose assignment to Longirostromeryx is more confidently founded. In any event, I believe (?)L. vigoraius should be considered a species inquirenda for the same reasons as outlined in the discussion of L. blicki. Subfamily DROMOMERYCINAE Frick, 1937 Barbouromeryx Frick, 1937 Barbouromerux (Bouromeryx) submiUeri Frick, 1937 Figure 32C. TypE. - FAM 33729, a right ramus. OCCURRENCE. - Cold Spring Fauna. REFERRED MATERIAL. - UTBEG 40622-26, a partial right ramus with M2 and M3· DESCRIPTION. - This specimen is from Belt's Creek; Tyler County, Texas. The molars are comparable to and about the size of those of Bouromeryx submilleri from Aphelops Draw, Sioux County, Nebraska, except for a less prominent metastylid on the anterior lobe of M:1. M2 measures 12.7 mm X 10.9 mm; M~, 18.0 X 11.1 mm. DIsCUSSION. - The systematic position of this subgenus is rather unclear. Frick ( 1937) includes it in his subfamily Barbouromerycinae, apparently in the type genus Barbouromeryx. As with other groups, he does not discuss the type subgenus; presumably it would assume the characteristics of Barbouromeryx, but in Frick's treatment of these forms, subgeneric characters seem to bear equal taxonomic weight with generic characters. Under this arrangement the utility of the subgeneric grade is lost by the lack of hierarchical ranking, and as a result systematic relationships within and among the closely allied taxa become obscured. That these taxonomic assignments are not regarded by Frick to reflect presumed genetic relationships is further evidenced by his phylogenetic considerations of the Barbouromery- cinae, in which subgenera are sometimes treated separately from the genera in which they are included ( Flick, 1937: 49, 127). The validity of the subgeneric grade in taxonomy is not denied ( for pertinent re- marks on Frick's usage of subgenera see Simpson, 1945: 267), but its utility in a paleontological series such as this one is questionable. 198 BULLETIN FLORIDA STATE MUSEUM Vol. 14 I do not suggest that Bouromeryx or any other subgenus of the Barbouromerycinae be elevated or reduced in rank at this time. The group is still poorly understood, and a rearrangement of any one of its parts without detailed study and consideration of its relationship to the others is obviously premature. Clearly this genus and the rest of the Dromomerycinae comprise a closely related group of artiodactyls that differ substantially (and, I think fundamentally) from the rest of the Cervoidea with which they are usually classified ( Simpson, 1945 ). I agree with the observations of some others ( Matthew, 1915, 1918; Schlosser, 1924; Pilgrim, 194la, 194lb; Crusa- font Pairo, 1952) that the dromomerycines share many features of the giraffoids and may actually share common ancestry as well. In a later paper Crusafont Pairo ( 1953) concluded that this resemblance results rather from convergence. The issue obviously is not settled. Cranioceras Matthew, 1918 Frick ( 1937: 47) recognizes two groups within the genus Crani- oceras: a subgenus Procranioceras, distinguished by distinctly brachy- odont molars and large-proportioned premolars, and Cranioceras "proper" ( sensu Frick, 1937) which is characterized by a larger size, slightly more hypsodont molars, proportionately smaller premolars, and an occipital horn which in the male is rounder, more posteriorly directed, and longer. The skull of Procranioceras is roughly suggestive of Dromomeryx, but is separated from that genus "in the more anterior position of the orbits, the erectness of the paired horn cores, the absence of any trace of a postorbital flange, and in the characters of the sagittal area, with its remarkable, backwardly directed, forwardly curved median- occipital horn" ( Frick, 1937: 75) . The lower incisors and canines of Craniocerds clatendonensis also closely resemble those of Dromo- meryx. The premolars of C. clarendonensis are "notably brachyodont. Compared to [Droinomeryx], the premolars tend to be proportion- ately larger, the P4 anterior fossette is unformed, and... there is no dPi (Pi)" ( Frick, 1937: 76). Two specimens from the Lapara Creek Fauna appear referable to Cranioceras "proper": a right immature dentition ( UTBEG 31081- 1480), and a left postorbital horn ( UTBEG 31132-455) . Although material is admittedly meager, these specimens are assigned to Cranioceras clarendonensis Frick. Specimens from the Garvin Gully Fauna Hay ( 1924 ) described 1969 PATTON: TEXAS ARTIODACTYLS 199 as Dromomeryx texanus and Frick ( 1937) lists as ?Cranioceras texanus ( Hay) are referred to Prosynthetoceras Frick. Cranioceras clarendonesis Frick, 1937 Figure 32@ TypE. - FAM 32454, a right ramus with symphysis and Ps-Mi REFERRED MATERIAL. - UTBEG 31081-1480, an immature den- tition; and UTBEG 31132-455, partial horn. OCCURRENCE. - Lapara Creek Fauna. DESCRIPTION. - An immature dentition ( UTBEG 31081-1480) from Bee County, Texas, agrees very closely in size and cusp mor- phology with a similar dentition ( FAM 32216A) from the MacAdams Quarry, Clarendon, Texas. Other immature dentitions of similar size ( FAM 32491, Subdromomeryx scotti, and FAM 31199, Bouromer!/X nebrascensis) differ from the Lapara and Clarendon specimens pri- marily in size of tubercles of dIN and in the presence of a "Palaeo- meryx fold" on Mi. The left postorbital horn, UTBEG 31132-455, from this fauna is broken distally, but on the basis of the proximal cross-sectional shape and the direction and taper of the preserved part of the shaft, it more closely approaches the postorbital horns of Cranioceras than Dromomerux, Bouromeryx, or any other similar form. For this reason, 00-2 r A B -*----==----=I-*./.-I.-~ --=K- .#pmmi A" ~11'rl'km ==~ tls4 9/ 40) /,1 FIGURE 82. A. - Cranioceras darendonensis C UTBEG 81182-455); partial horn. B. - Ramoceros ramosus ( UTBEG . 811-73); partial horn. C. - Barbouromeryx (Bouromeryx) submitteri C UTBEG 40622-26); labial and occlusal views of right M:-M». 200 BULLETIN FLORIDA STATE MUSEUM Vol. 14 I tentatively refer this specimen to C. clarendonensis, the only Crani- oceras species presently recognized in the Lapara Creek Fauna . Family Antilocapridae Gray, 1866 Ramoceros Frick, 1937 Ramoceros ramosus ( Cope), 1874c Figure 32B TypE. - USNM 1144, a three-pronged "antler." OCCURRENCE. - Lapara Creek Fauna. REFERRED MATERIAL. - This species is recognized on the basis of a partial four-pointed "antler," or horn, recovered from the Santa Fe Marls of New Mexico. The Lapara Creek specimen is inseparable from horn material assigned by Frick to Ramoceros ramosus from New Mexico, Colorado and Nebraska. DISCUSSION. - Frick ( 1937: 290) defines the genus Ramoceros as: "characterized by the depressed postorbital position of the three-to- four-pointed 'antlers,' which are of remarkably Cervine appearance. the basal 'pedicle' is depressed and directed outwardly and posteri- orly; the tips ( not always symmetrical) Hare widely; the shafts are slender with circular-tending cross section and are deeply curved anteriorly." This genus, included in Frick's Division Merycodontini ( Mery- codontinae of Simpson, 1945), is divided into two subgenera, Para- moceros and Merriamoceros, from which Ramoceros "Proper" ( sensu Frick) is excluded. These subgenera, presently monotypic, are sep- arated from each other by the shape and length of the shaft, and the shape, number, and distribution of terminal prongs of the post- orbital horns. Paramoceros is based on R. (P.) breuicornis from Bar- stow, California, whereas the type species of Merriamoceros is R. (M.) coronatus ( Merriam), also from Barstow. If subgenera are retained, the type subgenus, Ramoceros, must be recognized for the type species of the genus, R. ramosus (Cope). The three groups are de- cidedly different in appearance. Ramoceros "Proper" ( should read Ramoceros (Ramoceros) ) is characterized by Frick ( 1937: 290 ) thusly: "The shaft is long, with a forwardly directed prong ('brow tine') at two-thirds distance above the base, and distally is bi- or tri-forked." Frick (1937: 291) recognizes the subgenus Faramoceros on the following grounds: "The main shaft is short ( in the California sub- genotype) to long ( in Rocky Mountain variation) and the secondary 1969 PATTON: TEXAS ARTIODACTYLS 201 shaft is reduced, the lower and two terminal prongs simulating a tripointed crotch." The third subgenus, Mel'riamoceros, Frick ( 1937: 291) describes as having the following features: "The 'horns' are remarkable for their small size, tendency to palmation and mutitubercular pointing. They recall, in minor measure, miniature moose antlers." Merycodont indet. OCCURRENCE. - Lapara Creek Fauna. REFERRED MATERIAL. - UTBEG 31170-28, a right MS; and UTBEG 31170-68, left M3. DESCRIPTION. - Two small hypsodont M~s from Bee County, Texas, are referred to the Merycodontinae. Reliable generic designa- tion of these isolated molars is virtually impossible and is not at- tempted. Whether they are referable to recognized merycodonts in the Lapara Creek Fauna, e.g., Ramoceros, or represent an otherwise unrecognized group is equally uncertain. Infraorder OREODONTA Osborn, 1910 Although the presence of oreodont material in the Texas Coastal Plain fauna was previously reported by Quinn { 1955) and by Schultz and Falkenbach ( 194la), the former distribution of these animals in the Gulf Coast region has remained largely unknown. In addition to the three genera from Texas described in this paper, oreodonts now are known to occur in Florida deposits as well. Maglio ( 1966) describes Meri/chyus from Thomas Farm, while Patton ( 1967, 1966 [67] ) reports oreodonts from several different Florida localities ranging in age from Late Oligocene to Middle Miocene. Aside from ¥ their local importance, these occurrences demonstrate a greater ac- cessibility of the Gulf Coastal Plain to immigration from the Great Plains than previously had been considered probable. Family MERYCOIDODONTIDAE Thorpe, 1923 Subfamily T~CHOLEPTINAE Schultz and Falkenbach, 194la Ticholeptus Cope, 1878 Ticholeptus rilegi Schultz and Falkenbach, 194la Table 23 TypE. - FAM 42329, a left mandible with Ii (alv.)-M3 and partial femur. 202 BULLETIN FLORIDA STATE MUSEUM Vol. 14 Table 28. MEASUREMENTS OF Ticholeptus rilegii Ramus FAM 42329 Length ( max., including incisors) 165.5Length, /C to condyle incl. 150.0Depth of jaw below coronoid 77.0Depth of jaw below anterior edge of M: 26.5Length, /C-M» incl. 103.5Length, Pi-M. incl. 96.5Length, Pi-P, ind. 39.5Length, M,-M, ind. 57.0 1After Schultz and Falkenbach (19418) TypE LocALrry. - near Cold Spring, San Jacinto County, Texas. OCCURRENCE. - Cold Spring Fauna. DECUSSION. - This species is currently recognized in the Cold Spring Fauna only on the basis of material in the Frick Collection. No material other than the type was reported. Schultz and Falkenbach ( 194la: 83-84) describe this species as being very close to T. h!/psodus Loomis ( 1924) from Sioux County, Nebraska. Their specific, description follows: SKULL - Unknown. MANDIBLE . - Approximate size of that of T lulpsodus; ramus shallow ( even shallower than in T. obliquidens and T. tooheyi); ascending ram zis shallow andnarrow anteroposteriorly; condyle light. DENTITION. - Superior series unknown; in ferior series approximate size ofthose of T. z!/gomaticus and T. hypsodus. LIMBS. - Unknown only from a partial femur; lighter than examples of T.hgpsodus. Ustatochoerus Schultz and Falkenbach 194la Ustatochoerus profectus (Matthew and Cook), 1909 Figure 33, Table 24 TypE. - AMNH 14055, incomplete left ramus with I,-M3 ( /C broken and P4 represented by alveolus). OCCURRENCE. - Lapara Creek Fauna. REFERRED SPECIMENS. - UTBEG 31081-685, a right maxillary with /C-Mi UTBEG 31081-1157, a right maxillary with M2-M3 and alveoli for Pl-Ml. DIsCUSSION. - The Lapara Creek specimens are morphologically inseparable from specimens from the Lower Ash Hollow Formation of Nebraska referred to U. profectus ( Matthew and Cook) by Schultz and Falkenbach ( 194la: 36). The Cmstal Plain specimens fall within 1969 PAT TO N : TE X A S A R TIO D A C TY LS 208 Table 24. MEASUREMEIVTS OF CERTAIN SPECIES OF Ustatochoeruf U. profectus U. medius U. medius Lapara Creek U . profectus studeri U . medius mohavensis nouomexicanus 81081-685 FAM 38621 FAM 48081 FAM 480808 FAM 84464 FAM 48258A Length, C/-M» 184.0 142.5 127.0 122.0 180.0 186.0 Length, Pi-M' 118.0 128.5 113.0 109.0 118.0 121.0 Length, Pi-P' 47.6 55.5 50.0 45.5 46.5 52.5 Length, Ma-M' 75.8 74.0 70.0 67.0 70.5 71.5 1Mostly after Schultz and Falkenbach (194la. Table 1) 204 BULLETIN FLORIDA STATE MUSEUM Vol. 14 the lower limits of the range of variation of U. profectus, represented by U. profectus studeri from the Texas Panhandle, and the upper limits of the range of U. medius ( Leidy), represented by two sub- species, U. medius mohavensis and U. medius nomomexicanus ( Table 24). At present there seems to be no way to separate the jaws of large "varieties" ( sensu Schultz and Falkenbach, 194la) of U. medius from those of the smaller varieties of U. profectus. The Lapara Creek specimens more closely approach U profectus profectus ( not so designated by Schultz and Falkenbach C 194la) but provided for under Article 61(a) of the International Code of Zoological Nomen- clature ( 1961: 59) ) than they do U. medius. Subfamily MERYCHYINAE Simpson, 1945 Merychgus Leidy, 1858 Meiychgus sp, Figure 33 OCCURRENCE. - Garvin Gully Fauna. REFERRED MATERIAL. - UTBEG 40105-5, a fragment of left man- dible lacking dentition. DEsCRIPTION. - The nature and number of alveoli present' in the fragmentary mandible indicate that it represents a juvenile individual. The specimen measure 60 mm from the anterior of Pl alveolus to the posterior of M3 alveolus, falling just below the minimum range of A 6.6..1 B FIGURE 88. A. - Mer!/chyus sp. C UTBEG 40105-5); labial view of edentulous left ramus. B. - Ustatochoerus profectits C UTBEG 81081-685); occlusal view of right maxillary with C /-M». 1969 PATTON: TEXAS ARTIODACTYLS 205 Merychyus minimus Peterson from the Lower Marsland ( Schultz and Falkenbach, 1947) 1, AGE AND CORRELATION This paper, as one of a series of taxonomic and stratigraphic studies of the fossil vertebrate faunas of Tertiary age in the Texas Gulf Coastal Plain ( Hay, 1924; Hesse, 1943; Quinn, 1952, 1955; Wil- son, 1956, 1957, 1959, 1960), provides additional information useful in making biostratigraphic correlations between the Texas faunas and those elsewhere. As each major taxonomic study is completed, and the systematic relationships of the various groups become better known, increasing information on which to base studies of biostrati- graphy, zoogeography, and paleoecology is made available. Because the faunas ( sensu Wilson, 1959) in question correspond to what many biostratigraphers call assemblage zones, new information on previously poorly known components serves to broaden the founda- tion upon which the assemblage zone rests and alleviate overde- pendence on a few taxa for correlation purposes. Evidence presented in this study indicates that the ages assigned to some of the Texas Coastal Plain faunas can be modified ( Fig. 34). GARVIN GULLY FAUNA Studies by Quinn ( 1955) and Wilson ( 1956, 1957, 1960) estab- lished the first clear outline of the general stratigraphic framework within which these faunas occur. Quinn ( 1955) and Wilson ( 1960) considered the Garvin Gully Fauna to be of late Arikareean age.. This view was based primarily on the close similarity of the Garvin Gully taxa to those from the Florida Thomas Farm deposit. The Thomas Farm at that time was included in the Hawthorne Formation, of middle Miocene age. Apparently because Simpson ( 1932) believed material from the Hawthorne Formation at Quincy and Midway, Florida to be somewhat younger than that from the Thomas Farm Quarry, Quinn suggested that the Thomas Farm fossils were prob- 'The Marsland Formation of Schultz and Falkenbach ( 1947) includes more than Peterson's ( 1906) "Upper Harrison," as that unit is interpreted by Cook ( 1960, 1965), McKenna ( 1965), and Skinner ( pers. comm. ). The "Lower Marsland" of Schultz and FaIkenbach is equivalent to Peterson's Upper Harrison and the restricted Marsland of Cook ( 1965) and McKenna ( 1965). The "Upper Mars- land" includes Cook's Runningwater Formation ( See McKenna, 1965, for review of this problem). 206 BULLETIN FLORIDA STATE MUSEUM Vol. 14 ably of "Tampa", or early Miocene age. The ages of both the Haw- thorne and Tampa Formations were determined on the basis of marine invertebrates fossils ( Cooke and Mossom, 1929; Puri and Vernon, 1964), which in the Gulf Coast region have not yet been tied into the terrestial vertebrate faunas. Thus, there is no wide agreement on the equivalence of these sections, nor much informa- tion available on the magnitude of the geochronologic, "offset" between the two. In his section on Age and Correlation, Quinn ( 1955: 72-76) states that "there is nothing in the Garvin Gully fauna indicative of an age later than lower Miocene." The Thomas Farm deposit definitely is not correlative with the Tampa Formation, nor can it be placed with any certainty even in the Hawthorne Formation, a more com- monly accepted correlation. ( Vernon ( 1951) traces the Hawthorne only to within 25 miles of Thomas Farm.) The Florida Geological Survey includes the Thomas Farm sediments in the "Alachua For- mation," which they regard as the continental equivalent of the Haw- thorne. According to Puri and Vernon ( 1964): The placement of the bone-bearing sediments at the Thomas Farm excavation presents a very perplexing geologic problem. Beds of at least two' ages are represented: Upper Eocene, Crystal River Formation; and middle to upper Miocene, Alachua Formation. Limestone of the Crystal River Formation occurs as a pinnacle on the north side of the quarry... and around it, and also over- lapping it are clays, boulder bars composed of Suwannee Limestone fragments ... and cross-bedded lime and quart sands of the Alachua Formation ... Sediments of the "Alachua Formation" occur in fissures, depres- sions, and sinkholes in the underlying Ocala Limestone and are related only in that they are derived primarily from the same source ( the Hawthorne Formation) and in that they accumulated in the same kind of karst-controlled catchment basin. Because this process has been operative since late Oligocene time,1 the deposits are usually totally disjunct and quite varied geochronologically, and in no sense can they be considered to comprise a formation in the generally accepted use of the term ( Code Strat. Nomen., 1961). In any event, because of the peculiar geologic configuration of the Thomas Farm deposit, it is fruitless to attempt to achieve accurate age determina- tion for it using evidence based on lithostratigraphic data. The age of the Thomas Farm fauna generally has been con- iBased on the discovery of a late Oligocene land vertebrate fauna in central Florida ( Patton, 1967, 1969) 1969 PATTON: TEXAS ARTIODACTYLS 207 sidered to be late Arikareean. White ( 1942) believed it to be some- what older, but based his thesis on paleogeographical arguments ( concerned with the extent, depth, and length of existence of the so-called Okefenokee Trough) that many workers dispute. Bader (1956),in a study of the Thomas Farm horses, regarded the fauna as being late Arikareean, but based this conclusion .in part on Quinn s ( 1955 ) estimation of the age of the Thomas Farm in relation to the Garvin Gully Fauna. In a more recent paper, H. E. Wood ( 1964) states that a Thomas Farm rhinoceros, Diceratherium barbouri, is progressive over the classic early Mioeene index fossil from the Harrison beds, D. cooki, and indicates a late Arikareean or early Hemingfordian age for the deposit. On the basis of their closer proximity to and greater number of taxa shared with the geochrono- logically more accurately placed Great Plains faunas, seemingly a more logical approach is to place greater reliance on the Texas Coastal Plain faunas in dating the Thomas Farm, rather than the other way around. Irrespective of the larger argument, there is general agreement that the Thomas Farm fauna is slightly younger than the Garvin Gully. Wilson ( 1957, 1959, 1960) assigned the Garvin Gully Fauna to « the Arikareean on the basis of Daphaenodon cf. superbus, Dinohyus hollandi, and the abundance of Parahippus. He 'pointed out that the association of Daphaenodon, Dinoh!/us, and Parahippus is character- istically Arikareean, as designated by the Wood Committee ( Wood et al., 1941). Wilson ( 1959) further stated that of the horse material of Parahippus size and proportions from the Garvin Gully Fauna, 10 per cent possess Meri/chippus characters, which is similar to the Table 25. A~E RANGE OF GARVIN GULLY GENERA Taxon Arikareean Hemingfordian Barstovian Clarendonian Palaeolagus x Daphaenodon x Amphic!/on xxx Cynodesmus x Dinoh!/us x Merychyus xxx Ox!/dact!#us x Diceratherium x Hypohippus x x Mer!/chippus X X Anchitherium x X X X X X X X Farahippus x x Archaeohippus,· x · x H*podon x 208 BULLETIN FLORIDA STATE MUSEUM Vol. 14 finding of Bader ( 1956) for the Thomas Farm horses. Among in- dividual skulls of these horses from the Thomas Farm, several have dentitions exhibiting both Farahippus and Merychippus features ( White, 1942; Bader, 1956). Wilson ( 1959: 773), following White ( 1942) and Bader ( 1956), concluded that: Since Parahippus is absent in the Burkeville -Fauna it seems obvious that it became extinct by evolution into Merychippus. Once Mer~chippus had evolved it migrated northward and lived alongside surviving parahippines in Nebraska. The appearance of Merychippus then at the type section in Nebraska is one of migration from the Gulf Coast northward. The appearance of Mer!/chippus on the Gulf Coast is by evolution and must be earlier. Continuing, since the appearance of unquestioned Meri/chippus, or that stage of horse evolution represented by "Meri/chippus," i.e. presence of cement, connected crotchet, etc., marks the base of the Hemingfordian, the aforementioned evolutionary transition leading to it must logically precede it also in time, and therefore should be regarded as an Arikareean event. Thus the presence of Daphaenodon and Dinohyus, neither of which is known above the Arikareean, and the Parahippus-Meruchippus transition with its attendant evolutionary- zoogeographical implications, led Wilson to place the Garvin Gully Fauna in the late Arikareean. Although the first appearance of Merychippus has been used by some to recognize the base of the Hemingfordian in North American Tertiary deposits, no evidence shows this single. criterion to be valid. In the generally accepted guide to the nomenclature and correlation of the North American continental Tertiary, Wood et al. ( 1941) make no claim for Meri/chippus as marking the base of the Hemingfordian, but rather simply list its first occurrence as being in the Heming- fordian. Further, Mer!/chippus does not occur at the base of the Hemingfordian in the type area, but actually replaces Parahippus well up in the Hemingford section. In an early work Osborn ( 1918: 98-99), in his section on the generic diagnosis of Meri/chippus, lists the ~rst occurrence of that genus in the Great Plains as being in the Sheep Creek. Morris Skinner ( pers. comm.), in his study of the Tertiary horses of the Great Plains, also contends that Meri/chippus is not found in beds older than Sheep Creek. Cook ( 1960) however recognizes Merychippus in the underlying Runningwater Formation. ( M.C. McKenna ( pers. comm. ) claims that Cook misidentified a younger channel cut into the Runningwater as the Runningwater itself.) In a preliminary paper describing the ecologic factors affecting horse evolution in the Great Plains, Cook ( 1960: 203) states: 1969 PATTON: TEXAS ARTIODACTYLS 209 In the beds representing the earlier stages of what is currently being called the Marsland formation, all known horses had simple patterned, brachydont molars. After Marsland times, and before Sheep Creek times of late middle Miocene, events occurred including another massive regional elevation to the west, with marked erosional effects, and a new series of deposits was laid down across western Nebraska. In these new deposits, now included in the upper half of the Marsland formation, the most astonishing degree of evoluti6n occurred in the horses. The crowns of the molars rapidly became more hypsodont, and more complicated in pattern, and as they grew longer, they deposited, for the frst time cement in the enamel valleys in the teeth, reinforcing and strengthening them. Now, with three degrees of hardness in the crowns, exposed, as they wore down, as being enamel, dentine, and cement, the molars maintained an efficient millstone grinding surface. Correspondingly, through the time represented by these few hundred feet of new deposits, the molars rapidly increased in length, so that by Sheep Creek time the dominant, complicated, hypsodont, Mer!/chippus pattern had replaced the brachydont, simple, Parahippus present at the beginning of Marsland sedimentation. Space here prevents a discussion of other related faunal changes also present. For this series of deposits above the Iower stages of the Marsland ( which were originally named Upper Harrison Beds) and are below the Sheep Creek beds of Matthew and Cook, we are proposing the name Runningwater Formation, after an old, local name of the river which they occur. More evoluti6n is recorded in the teeth of fossil horses in the Runningwater Formation than in all the preceding millions of years in which their fossil record is known. Another paper, going into essential details on the occurrence of this formation and its fossils, is under preparation. Whereas the Parahippus-Merychippus transition as described above is not as yet substantiated, Cook's paper does serve to point out that Merychippus does not appear iii the Great Plains sequence below the Runningwater Formation. Schultz ( 1938), Lugn ( 1938), Elias ( 1942), Cady ( 1940), Schultz and Stout ( 1941), Schultz and Falkenbach ( 194lb), and McKenna ( 1965) all consider the Mars- land Formation to be the basal formation of the Hemingford Group in western Nebraska. Thus, if the concept of the Hemingfordian Stage/ Age is based on the rocks and fossils comprising the Hemingford Group, its lower limit should generally be recognized out of the type area on the basis of fossils that appear in the lowest beds of the Mars- land Formation. Mergehippus on the other hand does not appear in the section until at least as high in the section as the Runningwater Formation, which is younger than the Marsland.1 If the criteria Osborn ( 1918) and Cook ( 1960) use for recognizing Mer!/chippus are valid, then evidence from the stratigraphic occurrence of these 1(Jook and Cook ( 1933) list the fauna from the "Upper Harrison" ( =Marsland) of Nebraska. Their "Upper Rosebud" may be in part equivalent to the Running- water. 210 BULLETIN FLORIDA STATE MUSEUM Vol. 14 fossils indicates that the practice of employing the first occurrence of Merychippus to mark the base of the Hemingfordian, insofar as it is based on the rocks and fossils of the Hemingford Group, ( and this is the stated intention of Wood et al., 1941: 12) is of negative value. Although the Meri/chippus, if it actually is Meri/chippus, from the Garvin Gully Fauna may represent a less advanced stage of evolu- tion than Merychippus occurring in the late Hemingfordian of the Great Plains, it is demonstrably more advanced than the parahippines known from late Arikareean deposits in that region. In point of fact it compares most closely with forms occurring in deposits of much younger age, i.e. the Runningwater and later. Until the faunas of the Marsland and Runningwater Formations ( and those from the even lesser known Red Fill and Box Butte de- posits) become better known, the most def,nite statement I can make about the placement of the Garvin Gully Fauna is that, in terms of I -their respective faunas, it is post- Lower Harrison and pre-"Sheep Creek." On the basis of correspondence of taxa and relative stage of evolution of its faunal components, I believe that of the two the Garvin Gully Fauna is considerably younger than the Lower Harrison and is probably best correlated with the fauna from Runningwater. The same may be said of the Florida Thomas Farm fauna, except that it appears to be slightly younger than the Garvin Gully. BURKEVILLE FAUNA Quinn ( 1955) suggested a close correlation between the Burke- ville Fauna and the presumably advanced components of the Thomas ' Farm, and an even closer correlation with the Quincy and Midway Faunas of Florida . This appears to be valid. The Burkeville Prosyn- thetoceras texanus, which is more advanced than the Garvin Gully form, closely resembles the Thomas Farm P. texanus ( =Prosyntheto- ceras douglasi White and Synd!/oceras australis White). Floridatra- gulus specimens from the Burkeville Fauna are similar to F. dolich- anthereus from the Thomas Farm, but are assigned to a new species, F. texanus. Tomarctus canauus occurs in both faunas, as does Mery- chippus gunteri ( =Hippodon gunteri of Quinn, 1955 ). Its stratigraphic occurrence above the Hemingfordian Garvin Gully- Fauna and the association and relative stage of evolution of its fauna ( Table 26) indicates that the Burkeville Fauna also should be assigned to the middle Hemingfordian. The small steps observed in the phyletic sequences of both Prosynthetoceras and Floridatra- ..1969 PATTON: TEXAS ARTIODACTYLS 211 gulus between the Garvin Gully and Burkeville Faunas indicates that the time separating the two faunas is of relatively short duration. Of the four biostratigraphic units considered in this paper, only the Burkeville Fauna lacks camel material. This seemingly anomalous situation may be in part a reflection of the comparatively poorer state of knowledge of this fauna, suggesting that specimens may yet be found; or it may be attributed to differential environmental sampling, whereby only certain habitats, apparently unfavorable to camels, were near enough to the environment of deposition to contribute samples of their fauna. In any event, it is difficult to conceive of entire camel populations emigrating from the Texas Coast during the time the \ Burkeville deposits were accumulating only to return shortly there- after. Possibly local differential extinction was followed by a new spread, although we have no evidence for this. The absence in the Burkeville Fauna of other grasslands or savan- nah dwelling forms such as Amphicyon and the oreodonts, and the comparatively diminished horse fauna is signiRcant. It is noteworthy that these forms are present in relatively plentiful numbers in the faunas that bracket the Burkeville Fauna in time. If the lack of camels is a result of limited habitat sampling, it is a most unfortunate cir- cumstance with regard to comprehending camel evolution in the Coastal Plain faunas, for without such information it is virtually im- possible to speculate on the nature of the relationship between the camels of the Garvin Gully Fauna and those of the Cold Spring. This Table 26. AGE RANGE OF COLD SPRING GENERA Taxon Arikareean Hemingfordian Barstovian Clarendonian Mylagatdus x Amphicyon xxx Aeturodon ? x x Comphotherium x x Ticholeptus x x Hesperh!/3 x x P Aep!/camelus x x Bouromeryx x x Longirostromeryx x x Diceratherium x x Teleoceras x x Peraceras x x Mergchippus xxx Pliohippus x Nannippus X Neohipportion x Catippus X X 212 BULLETIN FLORIDA STATE MUSEUM Vol. 14 becomes especially critical, for example, in the attempt to establish the possible phylogenetic relationship of AustraZocam€Zus to Aep!/- camelus. COLD SPRING FAUNA The Cold Spring Fauna is considered by Quinn ( 1955) to be ap- proximately of late Hemingfordian age: "The horses of the Cold Spring seem unrelated to any forms other than those from the Calvert formation. The age of the Cold Spring fauna must therefore be evaluated by the relative evolutionary position of the fauna with regard to the Burkeville fauna, which is certainly older, and with regard to the Lapara Creek fauna, which is certainly younger. In this respect the Cold Spring fauna is clearly much closer to the Burkeville fauna than to the Lapara Creek fauna and seems to be ap- proximately late middle Miocene age. Quinn ( 1955: 74) further states that "the indigenous components ( of the Cold Spring Fauna) appear to have no counterparts else- where except for the Calvert material ( Gazin and Collins, 1950), which corresponds with the Cold Spring insofar as the proboscideans are concerned. Several Cold Spring forms, some of which were not fully known to Quinn, are virtually inseparable from forms occurring in western faunas. Ticholeptus filet/i is reported by Schultz and Falk- enbach ( 194la) to resemble Lower Snake Creek material. Measure- ments of the radius-ulna and metatarsus of Cold Spring Aepgcamelus correspond to those of A. alexandrae from the Barstow fauna of Cali- fornia ( Davidson, 1923). Cold Spring specimens assigned to Longi- rostromeryx blicki correspond to the type specimens of that species from the "uppermost" Santa Fe beds ( Mio-Pliocene) of New Mexico ( Frick, 1937; Simpson, 1950 ). None of the allocthonous Cold Spring genera except Dicerath- erium are restricted to the Hemingfordian or below but range either into the Barstovian or the Clarendonian. The camels and the deer of the Cold Spring are advanced over related late Hemingfordian forms in the Great Plains. Gomphotherium is not confidently known to appear in North American faunas earlier than the Barstovian. Mastodon remains have not been recovered from the lower Snake Creek deposits, but in the Great Plains appear for the first time in the Pawnee Creek beds, which are generally regarded as late Bar- stovian. Gomphotherium from the Calvert Formation of Maryland is discussed by Gazin and Collins ( 1950) as being significant in either of two ways: "If the presence of mastodon is to be regarded as placing 1969 PATTON: TEXAS ARTIODACTYLS 213 an early limit on the pos5ible age assignment, then a lower Barstovian age would likely be indicated. On the other hand, there is always the possibility that the Calvert represents the first appearance of mastodons in North America, a conclusion not out of keeping with its geographic remoteness from the western occurrences." In a preceding paragraph, Gazin and Collins ( 1950) conclude that the age of the Calvert is either latest Hemingfordian or early Bar- stovian. Thus here again exists a situation where the fauna in question lies relatively near an age boundary, as Quinn ( 1955) also recog- nized. However, on the basis of evidence deriving from this study that was not available to Quinn, the Cold Spring Fauna assumes a geologically younger aspect than was previously recognized and warrants its inclusion in the Barstovian, probably representing the early to middle part of this age. LAPARA CREEK FAUNA As the most extensive of the Texas Tertiary Gulf Coastal Plain faunas, the Lapara Creek Fauna provides a broad and reliable foun- dation for interregional correlation. The fauna is systematically well- balanced and the material fortunately has been amenable to greater precision in taxonomic determination, as the greater number of spe- cific identifications ( Table 27) witnesses. This relative profusion of taxa is thought not to be merely a result of accidents of preservation, but rather to reflect the great amount of taxonomic differentiation or radiation that occurred in this region in the late Miocene and early Pliocene. Quinn ( 1955: 75) correlates the Lapara Creek Fauna with the "Alachua ( Bone Valley) fauna" of Florida. He subscribes to the age Kellogg ( 1924) assigns to the Bone Valley fauna from his study of its marine mammals. However, his "Alachua ( Bone Valley) fauna" is not a unitary one and attempts to correlate one aspect of the fauna without regard to the lack of isochroneity of the whole are destined to lead to confusion. In a summary of the Alachua problem, Webb ( 1964) states: --Much of the confusion as to the age of the "Alachua Fauna" stemmed from the failure on the part of these early workers to consider the possibility that the Alachua Formation was heterochronic. Leidy unknowingly included fossils from the Pliocene and Pleistocene localities in his studies. Hay recognized that the "Alachua Fauna" included species of Pliocene and Miocene aspect along with typical Pleistocene forms, but he assumed that this indicated the survival of archaic types in the Pleistocene of Florida. This situation was largely remedied by Simpson's ( 1980 ) revision of the Pliocene part of the "Alachua Fauna." He 214 BULLETIN FLORIDA STATE MUSEUM Vol. 14 Table 27. AGE IUNGE OF L~PARA CREEK SPECIES Taxon Barstovian Clarendonian Hemphillian Aeturodon taxoides x xLeptocyon uefer X XUstatochoerus profectus xProcametus occidentatis x Procamelus grandis xSynthetoceras tricornatus xBlastomeryx elegans Cranioceras darendonensis xRamoceros ramosus x xTeleoceras proterus xProtohippus perditus x Ptiohippus supremus xNannippus ingenuus xNannippus tehonensis xNeohipparion coloradense xCalippus placidus x demonstrated, using exact locality data, that certain sites produced only Miocene,others only Pliocene, and still others only Pleistocene terrestrial fossil vertebrates. Thus, whereas Kellogg ( 1924), on evidence derived from marinemammals, considers the "Alachua fauna" to be not older than lateMiocene, Simpson ( 1930) shows that most of Leidy's original "Ala-chua fauna" came from Mixson's Bone Bed near Williston, Florida,which is clearly Hemphillian ( Webb, 1964). Quinn ( 1955: 75) regards the Lapara Creek Fauna to be olderthan the Burge Fauna of Nebraska as indicated by . . . direct comparison of specimens of Protohippus, Hippotigris, and CalippusThe amount of time separation is difficult to determine because the availableBurge sample is inadequate for statistical comparison. In general, the NorthTexas-Clarendon and Nebraska-Lower Ash Hollbw horses are readily distinguish-able from their Lapara Creek relatives by virtue ofthe much greater degree of -hypsodonty in the northern forms, but the Lapara Creek horse, Catippus ahatinus,is barely distinguishable from C. regulus of the Clarendon fauna. At the sametime tl~e Burge Hippotigris is closer to H. clarendonensis than to H. settardsiof the Lapara Creek. . Quinn (1955: 75) observes that species can exist sufficiently longto transcend provincial age boundaries and ·concludes: In consideration of this factor ( that species may transcend time boundaries),and because exact comparisons of. Burge and Lapara Creek forms could not gen-erally be made, several of the Lapara Creek · forms are, in this paper, referred tothe Burge species. In spite of this, the conclusion that the Burge and LaparaCreek are exactly synchronous is not warranted. Some of the forms referred to 1969 PATTON: TEXAS ARTIODACTYLS 215 Burge species will, no doubt, with better comparative materials, be separable; others undoubtably will not. Of the 8 specifically identified artiodactyls with the same or related species occurring in High Plains faunas, 6 are conspecific with forms from the Clarendonian or later, 1 occurs in both the Barstovian and Clarendonian, and only 1 is previously undescribed above the Bar- stovian ( Table 27 ). A direct comparison of the Lapara Creek species with those from the Great Plains and elsewhere also strongly sug- gests a Clarendonian assignment for the fauna. The Lapara Creek oreodont, Ustatochoerus profectus, is very similar to specimens repre- senting that species from the Burge Fauna and from the overlying Minnechaduza Fauna ( both Clarendonian). Synthetoceras tricornatus and Cmniocetas clarendonensis are indistinguishable from their types from the Clarendon beds of the Texas Panhandle, while Blastomeryx elegans is previously described only from the Hemphillian Upper Snake Creek beds. Ramoceros ramosus is found in both Barstovian and Clarendonian deposits in New Mexico ( Erick, 1937) and Cali- fornia ( James, 1963). The Lapara Creek specimens of Procamelus occidentalis, a species heretofore restricted to the Barstovian, are higher crowned and probably more advanced than the Barstovian species. Procamelus grandis from the Lapara Creek is advanced over its Burge equivalent, while Protalabis notiochorinos, a new species from the Lapara Creek Fauna, is advanced over similar forms from both the Burge and the Minnechaduza Faunas. Megatylopus primde- uus, also a new species, is very similar to, but slightly more primitive than, M. maior from the Burge. It seems clear, then, that the age assignment Quinn ( 1955) gives to the Lapara Creek Fauna should be reinterpreted in the light of evidence and information unavailable to him at the time of his study. As Quinn recognized, some forms in the Lapara Creek Fauna are commonly found in Barstovian deposits, and some formerly even re- stricted to beds of that age. Still others were previously thought to occur in beds no older than Hemphillian. The majority of Lapara Creek artiodactyls, however, are characteristic Clarendonian forms, and indicate that those disparities existing between the Lapara Creek Fauna and the Pliocene faunas of the High Plains have not resulted simply from temporal isolation of the faunas in the two regions, but rather from differing rates and directions of dispersal of some of the taxa in question. Thus consideration of the correspondence of fossil taxa between the Lapara Creek Fauna and late Tertiary faunas of the High Plains, as well as evidence derived from the evaluation of the 21 6 B U LL E T IN F LO R ID A S TA TE M U S E U M Vo 14 re lat ive s ta ~ m t~ ito na t° 15 35 :r Y ef ~ ~ ~ % 2 2 f a* 4+08 EPOCH N. A. PROVINCIAL AGES GREAT PLAINS TEXAS COASTAL PLAIN FLORIDA 548-S UPPER UPPER SNAKE BONE VALLEY CREEKass 1190 HEMPHILLIAN LLI vv Z Mc GEHEEZ 4 0 UPPER SNAKE CREEK LABAHIA MISSION wn ee C re ek b ed s of C ol or ad o bu t ol de r th an t C in ne ch ad uz a fa un as o f Te xa s an d Ne br as ka . It is m os t 0 CLARENDONIAN MINNECHADUZA BURGE LAPARA CREEK LOWER VALENTINE PAWNEE BUTTES seunej U Feld le }se o j H n ~ se xa l jo u o ile la u o o otqdEISR EASO ISE ops sninoI,~ BARSTOVIAN COLD SPRING ASHVILLE LOWER SNAKE CREEK SHEEP CREEK BURKEVILLE QUINCY- MIDWAYHEMINGFORDIAN BOX BUTTE THOMAS FARM MI OC EN E GARVIN GULLY na s fro m t he L ow er V al en n e Fo rm at io n o RUNNINGWATER BUDAMARSLAND HARRISON BROOKSVILLE ARIKAREEAN MONROE CREEK ot he r fa un a GERING 1969 PATTON: TEXAS ARTIODACTYLS 217 related with the early Clarendonian Burge Local Fauna of Nebraska ( Fig. 34). SUMMARY AND CONCLUSIONS From a series of vertically successive mammalian assemblage5 in Miocene and Pliocene deposits of the Texas Gulf Coastal Plain 27 species of fossil artiodactyls are described and their systematic posi- tions discussed. These include 10 species ( 5 new) of Camelidae, 3 new species of Floridatragulinae, 3 species of Synthetoceratinae, 5 species of Cervidae, 2 antilocaprids, and 3 oreodonts. Among the new forms recognized are two camel genera: Australocamelus, the probable Aep!/camelus ancestor, and Nothotylopus, a member of the Protolabis- Pliauchenia lineage. The camels are the most numerous group of fossil artiodaftyls from the Texas Coastal Plain. Of the 9 genera represented, 5 (Ox!/- dactylus, Aepycatnelus, Protolabis, Procamelus, and Megatylopus) are well-known from faunas of the Great Plains and Pacific Coast regions, while 4 (Nothokemas, Floridatragulus, Australocamelus, and Notho- tylopus) are so far restricted to the Gulf Coastal Plain. The former genera are geographically widespread and are apparently allocthonous elements contributed by the Great Plains faunas. Oxydactylus bene- dentatus is closely similar to species of Oxydactylus well established in the Great Plains faunas, and the Texas Gulf Coast representative of Procamelus are cospecific with Great Plains forms. Procam€?us species from the Florida Plioeene are based on astragali and are indeterminate, but like Megat!/lopus are probably derived from Texas forms. Protolabis notiochorinos appears to represent an advanced branch from an early P. heterodontus stock, whereas Megat!/lopus primaeous is closely related to, but is more primitive than, M. major from the Nebraska Burge and Minnechaduza Faunas. Apparently M. primaeous gave rise to the M. maior from the Florida MeGehee deposit ( Hemphillian), but whether it is the actual ancestor of M. major from the Burge and Minnechaduza faunas or owes its origin to a Great Plains form ancestral to both the Texas and Nebraska forms, is not known. In any event, it is morphologically the most primitive of the Megatylopus species. Specimens from the Cold Spring and Lapara Creek Faunas assigned to Aep!/camelus sp. are comparable to Great Plains forms, but Australocamelus from the Garvin Gully Fauna qualifies as a structural ancestor to Aepycamelus 218 BULLETIN FLORIDA STATE MUSEUM Vol. 14 and Hesperocamelus and may represent the earliest member of aphylogenetic series that evolved in the Gulf Coastal Plain andprovided the basal stock for the Great Plains and Pacific Coastmembers of the lineage. The lamentable lack of camel material fromthe Burkeville Fauna unfortunately precludes verification of this possi-bility. The three late Tertiary camel genera endemic to the Gulf CoastalPlain are Nothokemak, Australocamelus, and Nothotylopus. OnlyNothokemas is found in Florida ( Thomas Farm) Nothokemgs is mor- 1phologically similar to early Ox!/dactulus, but does not seem derivablefrom any presently known species of that genus. As it occurs in theearliest Miocene Coastal Plain vertebrate faunas ( Garvin Gully andThomas Farm), its origins must be sought elsewhere. The Florida-tragulinae are known only from Miocene deposits in the Gulf CoastalPlain. They occur in the Burkeville and Cold Spring Faunas of Texasand in the Thomas Farm Fauna of Florida. The exact systematicplacement of this group is presently uncertain, but for reasons out-lined on page 000, I have included them in the Camelidae. They haveapparently evolved entirely in the Gulf Coastal Plain, but theirphyletic origin is poorly understood. Austratocamelus and its probable relationships have been mentioned above. The origins of Nothotylo-pus are unclear. This strangely conservative yet in some ways progressive camel seems too specialized to be derived from theProcamelus, Pliauchenia, and Megatulopus groups, but probablyevolved from an early member of the genus Protolabis or some formintermediate between Protolabis and the line of oxydactylines that presumably gave rise to it. Whether or not Nothotylopus is a true jGulf Coast autochthon is presently unanswerable. The Synthetoceratinae are represented in the Gulf Coastal Plain by the genera Prosynthetoceras and Synthetoceras. From P. texanus of the Garvin Gully and Burkeville Faunas to S. tricornatus of the Lapara Creek Fauna, these forms exhibit characters that occur in a step-like progression thought to represent their actual phylogeny. Although ultimately derived. from Great Plains protocerati(is, most of their later evolution occurred in the Gulf Coastal Plain. The Florida Thomas Farm and MaGehee synthetocerines are conspecific with their Texas correlatives, suggesting that few, if any, lasting barriers existed for this species between Texas and Florida. The Claredon Synthetoceras tricornalus from the Texas High Plains prob- ably represents an emigrant population from the Coastal Plain. The cervids and antilocaprids from Texas show their greatest 1969 PATTON: TEXAS ARTIODACTYLS 219 amnities to the Great Plains forms, as witnessed in the identity of taxa shared by the two regions. Only one species, Blastomeryx vigor- atus from the Garvin Gully Fauna, is restricted to the Gulf Coastal Plain; another, Cranioceras clarendonensis, is known from the Lapara Creek Fauna and from the Clarendon beds of the Texas Panhandle. The correspondence of Great Plains and Texas Coastal Plain taxa may be only illusory but the insumcient preserved material from the Texas deposits prevents more accurately founded comparisons. The identification of Thomas Farm cervids suffers from similar defects; a revision by the writer of these and other Thomas Farm artiodactyls has recently been completed ( Patton, 1966 [67] ). Three oreodonts, Merychyus sp., Ticholeptus rile!/i, and Ustato- choerus profectus, have been identified in the Texas Coastal Plain faunas, in the Garvin Gully, Cold Spring, and Lapara Creek Faunas, respectively. No evolutionary sequences are apparent in the Texas Coastal Plain oreodonts. They all seem to be derivable from Great Plains forms, apparently having been introduced to the Coastal Plain as occasional migrants. Until the Oligocene and early Miocene oreodonts from the Serra Vieja and Big Bend regions 6f west Texas are known ( now being studied by J. A. Wilson, University of Texas), this will remain conjectural. The fossil artiodactyls from Miocene and Pliocene deposits in the Texas Coastal Plain fill an important gap in completing the under- standing of the terrestrial mammalian faunas existing at that time in the Gulf Coastal Plain of North America. As Bader ( 1956) and Quinn ( 1955) recognized, the Gulf Coastal Plain was essentially a distinct faunal province. The elements of this province, while greatly influ- enced by immigrants from the Great Plains of the western United States, underwent much of their evolution in situ under the influences of their own peculiar geography and ecology. This is evident not only in the presence of taxonomically distinct forms at the species level and above, but also in the detection of microevolutionary sequences in some of the groups represented. As established by Quinn ( 1955) and Wilson ( 1960), the Miocene and Pliocene deposits on the Texas Coastal Plain contain four fossil land mammal assemblages. The earliest of these is the Garvin Gully Fauna. It is here considered to be of approximately middle Heming- fordian age and correlative with the older portion of the Thomas Farm fauna of Florida. Characteristic taxa are Daphaenodon, Dino- hyus, Parahippus, Oxydactylus benedentatus, Australocamelus orarius, 220 BULLETIN FLORIDA STATE MUSEUM Vol. 14 Nothokemas hiddlgensis, Floridatragulus nanus, and Pros!/nthetoceras texanus. The Burkeville Fauna is of middle Hemingfordian age and cor- responds to the Quincy-Midway faunas of Florida and to the younger portions of the Thomas Farm fauna. Its distinguishing taxa are Tomarctus canavas, Merychippus gunteri, Floridatragulus texanus, and Aphelops meridianus. No camels have yet been found in the Burkeville Fauna. The age of the Cold Spring Fauna is reinterpreted in this paper and is considered to be of middle Barstovian age. The camels and the deer of the Cold Spring are advanced over related late Heming- fordian forms, while the presence of Mylagaulus, Gomphotherium, Teleoceras, Peraceras, Pliohippus, Nannippus, Neohipparion and Calippus warrants an age assignment for the Cold Spring Fauna later than previously designated. The Cold Spring Fauna is probably best correlated with the interval between the Lower Snake Creek of Nebraska and the Pawnee Buttes of northeastern Colorado, the late Miocene portion of the Barstow "fauna" from California, and the Calvert fauna of Maryland. The Lapara Creek Fauna is the most extensive of the Texas Tertiary Coastal Plain faunas and provides a broader basis fpr inter- regional correlation than the preceding faunas. Comparison of the Lapara Creek species with those from the Great Plains ( outlined in section on Age and Correlation strongly suggests a Clarendonian assignment for the fauna. A comparison of these taxa and a consider- ation of their relative stage of evolution indicates that the Lapara Creek Fauna is younger than faunas from the Lower Valentine Formation of Nebraska and the Pawnee Creek beds of Colorado but older than the Clarendon and Minnechaduza faunas of Texas and Nebraska. It appears to be most closely correlative with the Burge Local Fauna of Nebraska. LITERATURE CITED Bader, R. S. 1956. A quantitative study of the Equidae of the Thomas Farm Miocene. Bull. Mus. Comp. Zool. 115( 2): 49-78. Barbour, E. H. 1905. A new Miocene artiodactyl Science. 22 ( 572) : 797-798. ---, and C. B. Schultz. 1984. A new giant camel, Titanotylopus nebraskensis, gen. et sp. nov. Bull. Nebraska State Mus. 1 ( 36): 291-294. Cady, R. C. 1940. The Box Butte member of the Sheep Creek formation, Nebraska. Amer. J. Sci. 288 (9) : 668-667. 1969 PATTON: TEXAS ARTIODACTYLS 221 Cassell, D. E. 1958. Geology of the Cold Spring area and petrology of the Fleming Formation, San Jacinto County, Texas. Unpubl. M.A. thesis, Austin, Univ. Texas. Code of Stratigraphic N6menclature. 1961. Amer. Assoc. Petrol. Geol. Bull. 45( 5) : 545-600. Colbert, E. H. 1941. The osteology and relationships of Archaeomeryx. an ancestral runiinant. Amer, Mus. Novitates. ( 1185) : 1-24. Cook, H. J. 1909. Notice of a new camel from the lower Miocene of Nebraska. Amen Naturalist. 48: 188-189. ---. 1960. New concepts of late Tertiary major crustal deformation in the Rocky Mountain region of North America. Intern. Geol. Congr., 11th Session, Norden, Copenhagen, 1960, Rept. ( 12): 198-212. -. 1965. Runningwater Formation, middle Miocene of Nebraska. Amer. Mus. Novitates. ( 2227) : 1-8. --, and M. C. Cook. 1983. Faunal lists of the Tertiary Vertebrata of Nebraska and adjacent areas. Nebraska Geol. Surv. (5): 1-58. Cooke, C. W., and S. Mossom. 1929. Geology of Florida. 20th Ann. Rept., Florida Geol. Surv. pp 29-227. Cope, E. D. 1874a. Report on the stratigraphy and Pliocene vertebrate pale- ontology of northern Colorado. Bull. U. S. Geol. Geogr. Surv. Tem (1): 9-28. ---. 1874b. Report on the vertebrate paleontology of Colorado. Ann. Rept. Geol. Geogr. Surv. Terr. for 1878. pp. 427-583. ---. 1874c. Notes on the Santa Fe marls, and some of the contained vertebrate fossils. Proc. Acad. Nat. Sci. Philadelphia. 26: 147-152. ---. 1876. On a new genus of Camelidae. Proc. Acad. Nat. Sci. Philadelphia. 28: 144-147. ---. 1877. Report upon the extinct Vertebrata obtained in New Mexico by parties of the expedition of 1874. Geogr, Surv. West 100th Meridian, v. 4, - Paleontology, Washington. pp. 1-870. ---. 1878. A new genus of Oreodontidae. Amer. Naturalist. 12: 129. -. 1.898. A preliminary report on the vertebrate paleontology of the Llano Estacado. 4th Ann. Rept. Geol. Surv. Texas. pp. 1-137. Crusafont Pairo, M, 1952. Los Jirafidos fosiles de Espana. Diput. Provinc. de Barcelona. Mem. y Com. Inst. Geol. 18: 1-289. -- 1953. Sobre el origen de Triceromer!/x C ?Emigrante o autoctono?). Estudios geol., Inst. Investigaciones geologicos, Lucas Mallada. ( 20): 509-517. Davidson, P. 1923. Alticam€lus alexandrae, a new camel from the Barstow upper Miocene of the Mohave Desert. Univ. California Publ., Bull. Dept. Geol. Sci. 14 ( 12): 897-408. Douglass, E. 1900. The Neocene lake beds of western Montana and descriptions of some new vertebrates from the Loup Fork. Unpubl. M.S. thesis, Missoula, Univ. Montana. Elias, M. K. 1942. Tertiary prairie grasses and other herbs from the high plains. Geol. Soc. Amen, Spec. Pap. (41): 1-176. Flower, W. H. 1888. On the arrangement of the orders and families of existing Mammalia. Proc. Zool. Soc. London. pp. 178-186. Frick, C. 1921. Extinct vertebrate faunas of the badlands of Bautista Creek and San Timoteo Canon, southern California. Univ. California Publ., Bull. Dept. Geol. Sci. 12(5): 277-424. 222 BULLETIN FLORIDA STATE MUSEUM Vol. 14 --. 1987. Horned ruminants of North America. Bull. Amer. Mus. Nat. Hist. 69: 1-669. Gazin, C. L., and R. L. Collins. 1950. Remains of land mammals from the Miocene of the Chesapeake Bay region. Smithsonian Misc. Coll. 116(2): 1-21. Gray, J. E. 1821. On the natural arrangements of veftebrose animals. London Med. Reposit. 15 ( 1) : 296-310. --. 1866. Notes on the pronghorn buck ( Antilocapra) and its position in the system. Ann. Mag. Nat. Hist. (Ser. 3), 18: 828-826, 468-469. Gregory, J. T. 1987. Association of skull and limb bones in a new species of camel, with a review of the genus Pliauchenia. Proc. Geol. Soc. Amer. 1986, p. 888. ( Abstr.). ---. 1942. Pliocene vertebrates from Big Spring Canyon, South Dakota. Univ. California Publ., Bull, Dept, Geol. Sci. 26(4): 807-445. Hay, 0. P. 1924. Description of some fossil vertebrates from the upper Miocene of Texas. Proc. Biol. Soc. Washington. 87: 1-19. Henshaw, P. C. 1942. A Tertiary mammalian fauna from the San Antoni6 Mountains near Tonopah, Nevada. Carnegie Inst. Washington, Publ. ( 580):78-168. Hesse, C. J. 1942. Vertebrate paleontology in Texas. Texas Arch. Paleontol. Soc. Bull. 14: 97-119. ---. 1948. A preliminary report on the Miocene vertebrate faunas of southeast Texas. Proc. Trans. Texas Acad. Sci. 26: 157-179. Hibbard, C. W., and E. S. Riggs. 1949. Upper Pliocene vertebrates from Keefe Canyon, Meade County, Kansas. Bull. Geol. Soc. Amer. 60 (5) : 829-860. Illiger, C. 1811. Prodromus sustematis mammalium et avium additis terminis zoographicis utriudque classis. Berlin: C. Salfeld. Pp. 1-301. International Code of zoological Nomenclature. 1961. London: Intern. Trust Zool. Nomenclature. Pp. 1-176. James, G. 1963. Paleontology and nonmarine stratigraphy of the Cuyama Valley badlands, California, P. 1. Geology, faunal interpretations, and systematic descriptions of Chiroptera, Insectivora, and Rodentia. Univ. California Publ., Bull. Dept. Geol. Sci. 45: 1-171. Kellogg, R. 1924. Tertiary pelagic mammals of eastern North America. Bull. Geol. Soc. Amer. 85( 4): 755-766. Leidy, J. 1858. Notice of remains of extinct Vertebrata, from the valley of the Niobrara River. Proc. Acad. Nat. Sci. Philadelphia. March, 1858, pp. 20-29. ---. 1886. Mastodon and 11ama from Florida. Proc. Acad. Nat. Sci. Phila- delphia. pp. 11-12. Linnaeus, C. 1758. Systema naturae per regna tria naturae, secundum classes, ordines, genera, species cum characteribus, differentiis, synonymis, locis. Stock- holm: Laurentii Salvii. 10th ed., pp. 1-824. Loomis, F. B. 1924. Miocene oreodonts in the American Museum. Bull. Amer. Mus. Nat. Hist. 51: 1-37. ---. 1925. Dentition of artiodactyls. Bull. Geol. Soc. Amer. 36: 588-604. Lugn, A. L. 1938. The Nebraska State Geological Survey and the "Valentine Problem." Amer. J. Sci. (36) : 220-227. MacDonald, J. R. 1949. A new Clarendonian fauna from northeastern Nevada. Univ. California Publ., Bull. Dept. Geol. Sci. 28(7): 173-194. ---. 1956. A new Clarendon mammalian fauna from the Truckee formation of 1969 PATTON: TEXAS ARTIODACT¥LS 223 western Navada. j. Paleontol. 80(1): 186-202. -. 1964. Barstovian mammal fauna from Camp Creek, Nevada. Geol. Soc. Amer., 60th Ann. Meeting. p. 41. ( Abstr.). Maglio, V. J. 1966. A revision of fossil selenodont artiodactyls from the Middle Miocene Thomas Farm, Gilchrist County, Florida. Breviora. ( 255) : 1-27. Marsh, 0. C. 1891. A horned artiodactyl ( Protoceras celer) from the Miocene. Amer. J. Sci. ( Ser. 8), 41 : 81-82. Matthew, W. D. 1901. Fossil mammals of the Tertiary of northeasiern Colorado. Mem. Amer. Mus. Nat. Hist. 1(7): 855-447. ---. 1904. A complete skeleton of Merycodus. Bull. Amer. Mus. Nat. Hist. 20: 101-129. --. 1905. Notice of two new genera of mammals from the Oligocene of South Dakota. Bull. Amer. Mus. Nat. Hist. 21: 21-26. -- 1909. Faunal lists of the Tertiary Mammalia of the West. Bull. U. S. Geol. Sum (861): 91-120. ---. 1915. Climate and evolution. Ann. New York Acad. Sci. ( 24): 171-814. ---. 1918. Contributions to the Snake Creek Fauna; with notes-upon the Pleistocene of western Nebraska; American Museum Expedition of 1916. Bull. Amer. Mus. Nat. Hist. 88: 183-229. ---. Unpubl. MS. A revision of the extinct Camelidae with a discussion of their affinities and phylogeny. Osborn Library, Amer. Mus. Nat. Hist. Matthew, W. D., and H. J. Cook. 1909. A Pliocene fauna from western Nebraska. Bull. Amer. Mus. Nat. Hist. 26: 361-414. MeKenna, M. C. 1965. Stratigraphic nomenclature of the Miocene Hemingford Group, Nebraska. Amer. Mus. Novitates. ( 2228) : 1-21. Osborn, H. F. 1907. Evolution of mammalian molar teeth. New York: The ,~ Macmillian Co. Pp. 1-250. ---. 1910. The age of mammals in Europe, Asia, and North America. New York: The Macmillian Co. pp. 1-635. ---. 1918. Equidae of the Oligocene, Mioeene, and Pliocene of North America, iconographic type revision. Mem. Amer. Mus. Nat. Hist., ( new series ). 2(1) : 1-217. Owen, R. 1848. Description of teeth and portions of jaws of two extinct F anthracotheroid quadrupeds... discovered ,..on the N.W. coast of the Isle of Wight, with an attempt to develop Cuvier's idea of the classification of pachyderms by the number of their toes. Quart. J. Geol. Soc. 4( 1): 103-141. Patton, T. H. 1964. The Thomas Farm f6ssil vertebrate locality. In Guidebook, 1964 Field Trip, Soc. of Vert. Paleontol., Univ. of Florida, pp. 12-20. ---. 1966 ( 67). Revision of the selenodont artiodactyls from Thomas Farm. Quart. J. Florida Acad. Sci. 29 (3): 179-190. ---. 1967. Reevaluation of Hay's artiodactyl types from the Miocene of the Texas Coastal Plain. Texas J. Sci. 10( 1): 35-40. -- 1969. An Oligocene land vertebrate fauna from Florida. J. Paleontol. 48(2): 548-546. Peterson, O. A. 1904. Osteology of Oxydactylus, a new genus of camels from the Loup Fork of Nebraska, with descriptions of two new species. Ann. Carnegie Mus. 2: 434-476. --. 1906. The Miocene beds of western Nebraska and eastern Wyoming and their vertebrate faunae. Ann. Carnegie Mus. 4(8): 21-72. 224 BULLETIN FLORIDA STATE MUSEUM Vol. 14 ---. 1911. A new camel from the Miocene of Nebraska. Ann. Carnegie Mus. 7(2): 260-266. Pilgrim, H. G. E. 194la. The relationship of certain variant fossil types of"horn" to those of the living Pecora. Ann. Mag. Nat. Hist. (Ser. 11), 7.172-184. ---. 19414.The dispersal of the Artiodactyla. Biol. Rev. 16: 184-168. Plummer, F. B. 1988. Cenozoic systems in Texas. In the geology of Texas, v. 1, stratigraphy. Univ. Texas Bull. ( 8282 ) : 519-818. Puri, H., and R. 0. Vernon. 1964. Summary -of the geology of Florida and a guidebook to the classic exposures. Florida Geol. Surv, Spec. Publ.(5): 1-812. Quinn, J. H. 1952. Recognition of Hipparions and other horses in middle Miocene mammalian faunas of the Texas Gulf region. Univ. Texas, Bur. Econ. Geol. Rept. Invest. ( 14) : 5-6. --. 1955. Miocene Equidae of the Texas Gulf Coastal Plain. Univ. Texas Publ. ( 5516): 5-90. Ragsdale, J. A. 1960. Petrology of Miocene Oakville Formation, Texas Coastal Plain. Unpubl. M.S. thesis, Austin, Univ. Texas. Ray, C. E. 1957. A list, bibliography, and index of the fossil vertebrates of Florida. Florida Geol. Surv. (3): 1-175. Reed, L. C., and O. M. Longnecker. 1932. The geology of Hemphill County Texas. Univ. Texas Bull. ( 3231): 1-98. Renick, B. C. 1986. The Jackson Group and the Catahoula and Oakville formations in a part of the Texas Gulf Coastal Plain. Univ. Texas Bull. ( 8619): 1-104. Schlosser, M. 1924. Uber die systematische Stellung jungtertiarer Cerviden. Centralbl. Min. Geol. Paleont. Abhandl., ( new series).11(3): 155-264. Schultz, C. B. 1938. The Miocene of western Nebraska. Amer. J. Sci. (Ser. 5), 85(210): 441-444. Schultz, C. B., and C. H. Falkenbach. 194la. Ticholeptinae, a new subfamily of oreodonts. Bull. Amer. Mus. Nat. Hist. 79: 1-105. ---. 194lb. Marsland formation. Bull. Geol. Soc. Amer. 52 ( 10) : 1990-1991. ---. 1947. Merychyinae, a subfamily of oreodonts. Bull. Amer. Mus. Nat. Hist. 88 (4) : 157-286. ---., and T. M. Stout. 1941. Guide for a field conference on the Tertiary and Pleistocene of Nebraska. Univ. Nebraska State Mus., Special Publ. PP. 1-51. ---, L. G. Tanner, and C. H. Harvey 2nd. 1958. Buried soils of the Tertiary of the great plains. Bull. Geol. Soc. Amen 64(12): 1471-1472. Scopoli, G. A. 1777. Introductio ad historiam naturalem sistens genera lapidum, plantarum et animalium hactenus detecta, caracteribus essentialibus donata, in tribus divisa, subinde ad leges naturae. Prague: Gerle. Pp. 1-506. Scott, W. B. 1940. The mammalian fauna of the White River Oligocene. Pt. 4, Artiodactyla. Trans Amer. Phil. Soc., ( new series ). 28( 4): 863-746. Simpson, G. G. 1980. Tertiary land mammals of Florida. Bull. Amen Mus. Nat. Hist. 59(11): 149-211. ---. 1982. Miocene land mammals from Florida. Florida Geol. Surv. Bull. ( 10): 7-21. -- 1945. The principles of classification and a classification of mammals. 1969 PATTON: TEXAS ARTIODACTYLS 225 Bull. Amer. Mus. Nat. Hist. 85: 1-850. ---. 1950. Cenozoic formations and vertebrate faunas. In Guidebook for the 4th Field Conference of the Soc. Vert. Paleontol. in Northwestern New Mexico. Pp. 74-85. ---. 1961. Principles of animal taxonomy. New York: Columbia Univ. Press. Pp. 1-247. Stirton, R. A. 1932. A new genus of Artiodactyla from the Clarendon Lower Pliocene of Texas. Univ. California Publ., Bull. Dept. Geol. Sci. 21(6): 147-168. --. 1967. Relationships of the protoceratid artiodactyls, and description of a new genus. Univ. California Publ., Bull. Dept. Geol. Sci. 72: 1-44. Thorpe, M. R. 1928. The primitive and carnivore-like characters of the Merycoidodontidae. Amer. J. Sci. (Ser. 5), 6: 289-246. Vernon, R. 0. 1951. Geology of Citrus and Levy Counties, Florida. Florida Geol. Surv. ( 83) : 1-256. Webb, S. D. 1964. The Alachua Formation; in Guidebook for the Annual Meeting of the Soc. Vert. Paleontol., Univ. of Florida. pp. 22-29. ---. 1965. The osteology of Camelops. Bull. Los Angeles County Mus. Sci. (1): 1-54. ---. 1969. The Burge and Minnechaduza Clarendonian mammalian faunas. Univ. California Publ., Bull. Dept. Geol. Sci. 78: 1-191. Weeks, A. W. 1945. Oakville, Cuero, and Gohad formations of Texas Coastal Plain betwen Brazos River and Rio Grande. Amen Assoc. Petrol. Geol. Bull. 29(12): 1721-1782. White, T. E. 1940. New Miocene vertebrates from Florida. Proc. New England Zool. Club. 18: 81-88. ---. 1941. Additions to the Miocene fauna of Florida. Proc. New England Zool. Club. 18: 91-98. ---. 1942. The lower Miocene mammal fauna of Florida. Bull. Mus. Comp. ZooI. 92( 1) : 1-49. ---. 1947. Additions to the Miocene fauna of north Florida. Bull. Mus. Comp. Zool. 99(4): 497-515. Williams, E. E. 1950. Testudo cubensis and the evolution of western hemisphere tortoises. Bull. Amer. Mus. Nat. Hist. 95(1): 1-86. Wilson, J. A. 1956. Miocene formations and vertebrate biostratigraphic units, Texas Coastal Plain. Amer. Assoc. Petrol. Geol. Bull. 40(9) 2288-2246. ---. 1957. Early Miocene entelodonts, Texas Coastal Plain. Amer. J. Sci. 255: 641-649. ---. 1959. Stratigraphic concepts in vertebrate paleontology. Amer. J. Sci. 257: 770-778. ---. 1960. Miocene carnivores, Texas Coastal Plain. J. Paleontol. 84 (5): 988-1000. ---. 1962. Tertiary formations between Austin and Houston, with special emphasis of the Miocene and Pliocene. In Geology Gulf Coast and central Texas, Houston Geol. Soc. Ann. Meeting. Geol. Soc. Amer., pp. 842-858. Wood, H. E. 1964. Rhinoceroses from the Thomas Farm Miocene of Florida. Bull. Mus. Comp. Zool. 130(5): 361-886. Wood, H. E., End, and A. E, Wood. 1987. Mid-Tertiary vertebrates from the Texas Coastal Plain; fact and fable. Amer. Midl. Naturalist. 18 (1): 129-146. 226 BULLETIN FLORIDA STATE MUSEUM Vol. 14 Wood, H. E., 2nd, et al. 1941. Nomenclature and correlation of the North American continental Tertiary. Bull. Geol. Soc. Amen 52(1): 1-48. Zittel, K. A. von. 1893. Handbuch der Paleontologie. Abteilungl Palaeo- zoologie. Band IV, Vertebrata ( Mammalia). Munich: R. Oldenbourg. Pp. 1-799. Contributions to the BULLETIN OF THE FLORIDA STATE MUSEUM may be in any field of biology. Manuscripts dealing with natural history or systematic problems involving the southeastern United States or the Caribbean area are solicited especially. Manuscripts should be of medium length-50 to 200 pages. Examination for suitability is made by an Editorial Board. The BULLETIN is distributed worldwide through institutional subscriptions and exchanges only. It is considered the responsibility of the author to distribute his paper to all interested individuals. To aid in this, fifty copies are furnished the author without cost. PREPARATION OF MANUSCRIPT Highly recommended as a guide is the volume: Conference of Biological Editors, Committee on Form and Style. 1960. Style manual, for biological journals. Anier. Inst. Biol. Sci., Washington. 92 p. Manuscripts should be typewritten with double spacing throughout, with ample margins, and on only one side of. the paper. The author should keep a copy; the original and a carbon must be submitted. Tables and legends of figures should be typed on sheets separate from the text. Several legends or tables may be placed on a single sheet. Illustrations, including maps and photographs, should be referred to as "figures." All illustrations are reduced to a maximum of 4-1/4 by 7-1/8 inches. Size scales, wherever they are necessary, should be incorporated into the figure. References to literature should conform with the bibliographic style used in recent numbers of the BULLETIN. Spell out in full the titles of non-English serials and places of publication. Footnote material should be kept to a minimum. However, provide copy for a footnote detailing the title, affiliations, and address of the author (see recent numbers of the BULLETIN). Manuscripts must be accompanied by a synopsis-a brief and factual summary (not a mere description) of the contents and conclusions, which points out the presence of any new information and indicates its relevance. In it list all new organisms described and give their ranges; indicate all taxonomic changes pro- posed. The synopsis, written in full sentences, should be concise, but completely intelligible in itself without references to the paper, thereby enabling the busy reader to decide more surely than he can from the title alone whether the paper merits his reading. The synopsis will be published with the paper. It does not replace the usual conclusions or summary sections. It may also serve as copy for the abstracting services. Manuscripts and all editorial matters should be addressed to: Managing Editor of the BULLrrrN Florida State Museum Seagle Building Gainesville, Florida