THE PEARCE­SELLARDS Series NUMBER 32 LATE HEMPHILLIAN MAMMALS OF THE OCOTE LOCAL FAUNA, GUANAJUATO, MEXICO Walter W. Dalquest and Oswaldo Mooser February 29, 1980 TexasMemorialMuseum/2400Trinity/ The University ofTexas atAustin Austin, Texas 78705 William G. Reeder, Director The Pearce-Sellards Series is an occasional, miscellaneous series of brief reportsof museum and museum-associated field investigations and other research. All manuscripts are subjected to extramural peer review before being accepted. The seriestitlecommemoratesthefirsttwo directorsoftheTexasMemorialMuseum,bothnowdeceased:Dr. J.E.PearceandDr.E.H.Sellards,professorsofanthro­pology and geology,respectively, atThe University ofTexasat Austin. A complete price list of Pearce-Sellards papers and other bulletins and miscel­laneous publicationsof the museumwillbesentuponrequest. —James Reddell, Editor LATE HEMPHILLIAN MAMMALS OF THE OCOTE LOCAL FAUNA,GUANAJUATO, MEXICO Walter W. Dalquest 1 and Oswaldo Mooser2 ABSTRACT The Ocote local fauna is described from several thousand fossils, mostly isolated teeth, of largemammals,obtainedfromsedimentsnear thevillageofLosRodriguez,DistrictofSanMigueldeAllende,northeast of the city of San Miguel de Allende, Guanajuato, Mexico. Teleoceras ocotensis, Desrnathyusbrachydontus, and Palaeolama guanajuatensis are described as new, the name Paenemarmota mexicana (Wilson) is revived, and descriptions of the dentitions of six species of horses are given. It is suggestedthat the Pliocene evolution of Desrnathyus and Palaeolama took place on the Mexican Plateau. The evolutionary stage of the Ocote mammals suggests that the fauna is of late Hemphillian age. RESUMEN Se describelafaunalocal“Ocote,”basado sobrevariesmilesdepiezasfosilesprincipahnentedientes aislados de mamiferos grandes, obtenido de sedimentos geologicoscerca del Rancho “El Ocote,” Con­gregacion de los Rodriguez, Distrito de San Miguel de Allende, situado al nordeste de la Ciudad de San Miguel de Allende, Estado de Guanajuato, Mexico. Teleoceras ocotensis, Desrnathyus brachydontus, y Palaeolama guanajuatensis se describen como especies nuevas. El nombre Paenemarmota mexicana (Wilson) es revivido y se hace la descripcion de la denticion de seis especies de caballos. Se sugiere quelaevolucion deDesrnathyusyPalaeolama durante elPliocenotuvolugarenlaAltaplanicieMexicana. El estado evolutivo de los mamiferos de Ocote sugiere que la edad de la fauna es Hemphill tardio. INTRODUCTION The Ocote local launa was named and briefly described by Arellano (1951). Specimens were first collected at a small canyon locally called “Arroyo de la Carreta” by EnriqueDiaz Lozano, in the 1940’5. These were examined by Claude Hibbard in 1949 and Chester Stock in 1950. A much larger collection made by Arellano in 1950 was the basis of his faunal list (1951). Between 1954 and 1900, Mooser and persons employed by him ob­tained thousands of bones and teeth at Arroyo de la Carrcta, and these, subsequentlypurchased by Midwestern State University, form the basis of the present report. The name “Ocote" is derived from the name of the ranch where the fossils occur. The siteis near thesmallvillageofLosRodriguez,DistrictofSanMigueldeAllende,Stateof Guanajuato, 20 km north and 5 km east of the city of San Miguel de Allende, at an ele­vation of 2,000 meters. The species listed in Arellano’s (1951) preliminary faunal list are listed also by Silva-Barcenas (1969). The horses of the Ocote local fauna have been described in several papers by Mooser (1958, i960, 1964, 1965, 1968, 1973). We know of no other published, original contributions dealing with the Ocote local fauna. Arellano termed the fossiliferous horizon a “bone bed” about two meters thick, of white to light green or yellowish silty sand with coarser detritus disseminated, or in pockets orlenses.Thesitehas alsobeen describedby Mooser(1958). In theaccounts thatfollow,catalogue numbersareoftheMidwesternStateUniversity’sCollection of Fossil Vertebrates (MWU) and of The University of Texas at Austin’s Texas Memorial Museum Collection (TMM). Measurements are in millimeters and in diameters of teeth, the first measurement is the length of the tooth, the second value is the width. 1Department of Biology, Midwestern State University, Wichita Falls, Texas 76308. 2Francisco I. Madero 309, Aguascalientes, Aguascalientes, Mexico. This study was supported by a grant from Midwestern State University, made availablethrough the efforts of Dr. Travis White. We are indebted to Dr. S. David Webb for helpfulcriticism of the manuscript. ORDER RODENTIA FAMILY SCIURIDAE Paenemarmota mexicana (Wilson) Marmota mexicana Wilson, 1949, Carnegie Inst. Washington Publ., 584:168. Threealmostunwornteeth(11056) arefromagiantgroundsquirrel:anupperMl ,an upper M 3, and a lower Mj. The two Ml’s match teeth ofPaenemarmota barbouri Hib­bard and Schultz, as described by Repenning (1962), but the third molar differs in having a distinct and prominent anteroposterior bisection of the posterior valley (dental terminology is from Bryant, 1945). A high ridge reaches from the posterior cingulum to the metaloph. The labial part of the valley becomes a closed lake that will remain as a deep pit until a late stage of wear. The slightly expanded rnetaconule pinches the lingualpart of the valley to a crescent. The ridge is marked on the external face of the posteriorcingulum by a sharp vertical groove. This condition was recognized by Repenning (1962:547)in theupperM3,solHemphillianPaenemarmotafromtheYepomeralocalfaunaof Chihuahua and the Goleta local fauna of Michoacan. I he separation of the posterior valley is a character readily apparent to the unaided eye, even in specimens with moderately worn teeth. It occurs in the upper M 3 ’s of Paenemarmota from all three known Hemphillian occurrences of the genus in Mexico. A divided valley is absent in described specimens from the United States. The divided posterior valley may be diagnostic of either Hemphillian or Mexican Paenemarmota. The Mexican, Hemphillian, giant ground squirrels are thus set off from the Blancan animals of the United States, and the divided valley is presumed to be a character of spe­cific value. The namc Paenemarmota mexicana (Wilson), type locality Yepomera, Chihua­hua, placed in the synonymy ofP. barbouri by Repenning (1962), is revived. The teeth from Ocote, measured at the base of the enamel as described by Repenning, are smaller than all of the teeth listed by Repenning (1962): upper M l , 6.2 x 6.7; M 3, 7.3 x 6.5; lowerMj, 6.4 x 6.3. They may be near minimum size for the genus. ORDER EDENTATA FAMILY MEGALONYCHIDAE Megalonyx sp. A sloth tooth (TMM 41685-9) is referable to Megalonyx. It is somewhat triangular in shape, and is probably an upper anterior molariform. It is 35 high, of uniform diameter throughoutitscrown,and isthereforefrom anadult.Theleastdiameter(anteroposterior),measured near the top of the crown, is 13.7 and the greatest diameter (transverse) is 19.0 at the same level. The tooth is too large for Pliometanastes Hirshfeld and Webb, but is comparable in size to remains from Hemphillian deposits that have been referred to Megalonyx (Hirsh­feld and Webb, 1968). Fig. I.—A,Megalonyxsp.,tooth,probablyupperanteriormolarifonn,inlateralandocclusalviews. B, Paenemarmota mexicana (Wilson), upper third molar, occlusal view. ORDER CARNIVORA FAMILY CANIDAE Osteoborus cf. 0. cyonoides (Martin) Hyaenognathus cyonoides Martin, J. Mammal., 9:235. The bone-eating dog is represented in the collection by one upper (I 1058) and one lower (1 1059) carnassial. The l>4 is lightly worn, and the protocone is broken away. The tooth is 27.8 long. The Mj is almost unworn and measures 26.7 x 11.3. The measurements of these teeth are within the range of variation for Osteoborus cyonoidesfromtheCoffeeRanchlocalfaunaofTexas(Dalquest, 1969).Thelowermolar is a hit lower-crowned than any of the available Texas teeth. Webb (1969) described Osteoborus dudleyi (White) and 0. galushai Webb, from Hemphillian sites in Florida. The two species are approximately equal in size to the type species 0. cyonoides, and to the specimens from Guanajuato.The Mexican teeth may represent an undescribed species but thetwocamassialsarenotsufficientevidencetoestablish this. FAMILY URSIDAE Indarctos? A huge, almost unworn, upper canine (11062) can only he from a hear. The root and part of the crown have been broken away. The preserved parts of the tooth are every­where covered with enamel. A small but prominent cingulum on one side of the tooth suggests that the base of the specimen was near the true base of the tooth. There is a small ridge along the posterior face. At the base the fossil measures 28.6 x 21.9. FAMILY FELIDAE 1lachairodus sp. A lower carnassial (11063) is from a sabertooth and is w'ell preserved, except that one root is broken away. The blade of the tooth shows moderate wear, at the oblique angleusual in sabertooth cats. Measurements are 28.0 x 1 1.8. The tooth is similar to the lower first molars of Machairodus coloradensis Cook (for use of this name rather thanM. cato­ copis see Martin and Schultz, 1975), but smaller. Carnassials from the Coffee Ranch local fauna of Texas range from 32.0 to 35.8 in length and 12.4 to 13.8 in breadth (Dalquest,1969). ihe measurements of the Mexican tooth are like those of the camassial of a jawfromthelateHemphillianAxtellocalfaunaofTexas(Mawhy, 1965). ORDER PROBOSCIDEA FAMILY MASTODONTIDAE Rhynchotherium sp. All of the mastodon material from Ocole seems referable to a single genus and species,and we referitsomewhathesitantlytoRhynchotherium.Themostinformativespecimenis a fragmentary left lower jaw ramus with Mj and M 2 (11251). The jaw is from a younganimal. Mj is greatly worn, M 2 is slighdy worn, and the ramus is broken just behind the anterior edge of M The anterior crest of the unerupted M 3 is visible in the broken bone. 3. Most of the left side of the symphysis, and a bit of the side of the right symphysis, are present. Ihe anterior end of the symphysis, with the alveolus for a tusk, is missing. The Ocote mastodon is also represented by two Uisk fragments and numerous teeth and tooth fragments. All lower teeth, except P2 , are represented and show the charac­teristic pattern ofRhynchotherium. All lower teeth except M 2 and M 3 have a low ante­rior crest, or cingulum, consisting of a variable number of small, usually beadlike cusps,set off from the anterior main crest by a sharp, narrow valley. The second and third molars lack an anterior crest. Ihe main lophids consist of large, trenchant ridges, each composed of an external cusp that wears to a distinct, simple, trefoil, and an internal cusp in the form of a simple transverse lophid. In the Ocote mastodon the lingual cusps,when worn nearly to the base, exhibit an indistinct trefoil or hourglass pattern. All teetli have a posterior crest of one or two simple, flattened cusps. The posterior crest of M 3usually is of two simple, conelike, somewhatprocumbent cusps. The cusp pattern of the lower teeth of the Ocote mastodon is like the pattern of Rhynchotherium praecursor (Cope), from the Blanco local fauna of Texas (Dalquest,19.5) but is slightly more complicated. In R. praecursor the internal cusps of the main crests are simple and wedge-like, and do not show an hourglass or indistinct trefoil pat­ ternwhenwornto thebase.In otherfeatures,asfarastheycanbecompared, theOcote and Blanco mastodon teeth are similar, and the Mexican form may be ancestral to the Blanco animal. The mental foramen of the jaw is round, 9 in diameter, and is located 68 above the bottom of the ramus and 80 anterior to the anterior edge of Mj. The symphysis is solidlyfused but there is a distinct suture. The width of the tongue groove is about 38, and the groove is semicircular in cross section. The edges of the groove are bordered dorsally bysharp edges or crests. The depth of the ramus under the middle of Mj is about 140;breadth at this point about 81. The depth of the ramus under the middle of is approx­ M2 imately 140:breadth about 95. The largest tusk fragment (8863) is about 60 long and has the complete, rounded,unworn tip. At 45 from the tip it measures 22.3 wide and 20.9 high, and is almost round.1 he enamel band extends halfway around the circumference of the tusk fragment. The other fragment (11253) is only about 30 long and lacks the tip. It is oval in cross section, 20.4 x 16.4; the lesser measurement is presumably the transverse. Enamel,2 thick, torms a ventral band 15.5 wide, but some enamel seems to have been broken away. Upper teeth are represented only by a nearly complete third molar and several damaged and fragmentary third molars. The best specimen measures 197.0 x 89.2. There is no anterior crest. Main crests one to three are worn, especially on the outside, to ex­pose the trefoils. Main crest four and the posterior crest are unworn. The typical patternis exposed. Specific identification of the Ocote mastodon is not practical at this time. The geno­type of Rhynchotherium is R. praecursor Cope, 1893, from the Blanco local fauna of lexas. Ihe Ocote animal (Hemphillian Age) is probably specifically distinct from the younger Blanco (Blancan Age) Rhynchotherium, but better material is required to establish this. 1 hree species of Rhynchotherium were recorded from Mexico by Alvarez (1965).I he type locality of Mastodon shepardi Leidy is Contra Costa County, California. The specimens from the Valley of Mexico referred to this species by Cope (1884) lacked lower tusks and thus are not Rhynchotherium. See also Pichardo del Barrio (1961) and Ochoterena and Silva-Barcenas (1970). Ihe Rhynchotherium tlascalae and R. browni are available for the Ocote names mastodon, but the relationships of these forms to R. praecursor must be determined before identification is possible. ORDER PERISSODACTYLA FAMILY RHINOCEROTIDAE Teleoceras ocotensis new species Holotype.—TMiM 41685-8, upper third molar, crown well preserved but roots mostlymissing. Referred.—Upper incisor (11065); isolated upper premolars and molars (11066­110.8);lowerjawfragmentwith P3-Mj (11079);isolatedlowerpremolarsandmolars (11080-11086); tusk fragment (11087); metacarpal (11090); metatarsal (11091). Type locality.—Arroyo de la Carreta, near the Village of Los Rodriguez, District of San Miguel de Allende, Guanajuato, Mexico; Ocote local fauna, Hemphillian age, Plio­cene epoch. Diagnosis.—A large Teleoceras with large upper lateral incisor, hypsodont cheek teeth, upper first premolar present, median valley of upper cheek teeth closed to form isolated or nearly isolated lakes at early age, cingula strongly developed on upper teeth. Description.—The rhinoceros teeth from Ocote suggest a large animal, equal in size to Hemphillian species from the Great Plains. The upper incisor is shaped like the upperincisors of described species, as far as they are known, but is large. The unworn tooth is bean-shaped in lateral view, almost heart-shaped in cross section, with a trenchant edge.The enamel of the incisor is delicately wrinkled. The first upper premolar is similar in appearance to the upper first premolars of Clarendonian and early Hemphillian Teleoceras from the Great Plains. Middle Hemp­hillian Teleoceras may lack the upper P l . Premolars 2 to 4 are represented by lightly to well worn teeth. The teeth are like Hemphillian Teleoceras from the Great Plains in that the enamel of the lingual borders of the teeth creates lophs that close the median valleys to form lakes. In the teeth from Ocote, however, the closure seems to have occurred at an earlier age than in specimens of Teleoceras from the Great Plains. There is a single moderately worn upper M l , but no upper M 2 ’s. There are 5 lightly to moderately worn upper M3,s. In all of these the posterior crochet, on the melaloph,extends anteriorly to contact the protoloph. The two folds are appressed but not fused. This condition occurs sporadically but to a lesser degree in individual teeth from Hemphil­ lian sites on the Great Plains, but is the norm for upper molars from Ocote. Cingula are strongly developed on upper teeth. On P 3 and P 4 , cingula extend com­pletely across the lingual margins of the teeth, and between the protoloph and metaloph are sharp ridges rising to the height of the crown and leaving deep pits externally that persist as isolated enamel islands to a late stage of tooth wear. On M 1tin*, cingula are similarly developed hut are not as strong. There are no M 2 ’s in the collection. On M 3, there are strong cingula at the anterolingual margin of the protoloph and at the posteriormarginoftheectoloph,butthecinguladonotconnectacrossthemedianvalley(Fig. 2).I he cingula of the premolars of T. ocotensis are stronger than those of HemphillianTeleoceras from the Great Plains of the United States, and a cingulum at the lingual I oto’ border of M and at the posterior end of the ectoloph of M 3 is rarely present, and if present only very weakly developed, on Hemphillian species from the Great Plains. Fig. 2.-Teleocerasocotensisnewspecies,holotype,upperthirdmolarinlabialandocclusalviews. The lower dentition is represented by the jaw fragment with P2-P4 and numerous complete and fragmentary’ premolars and molars. The lower teeth seem to he like lower teeth from the Great Plains, and no distinguishing characters can he noted in them. Etymology.—The species is named from its occurrence in the Ocote local fauna. Discussion. Tanner (1975) has described a stratigraphically successive series of speciesof Teleoceras from the northern Great Plains and his work has been helpful in identifica­tion of the Ocote rhinoceros. Successive species, from Miocene (Clarendonian) throughthe lower Pliocene, are: Teleoceras major Hatcher, T. fossiger (Cope), T. hicksi Cook, and T.schultziTanner. TheKimball localfauna seemstobealmostexactlyequivalenttothe Coffee Ranch local fauna (type of the Hemphillian Land Mammal Age), and T. schultzi is older than the late Hemphillian T. ocotensis. Teleoceras ocotensis exhibits an advanced dentition in having large, strongly hypso­dontteeth. It seems to showtheclimaxforthegenusintheclosureofthemedianvalleyof the molars by the crochet to form isolated lakes, a trend extending from the Miocene through the Pliocene. The Ocote rhinoceros seems to he relatively primitive, however, in the strongly developed cingula on the upper cheek teeth, large lateral incisor, and reten­tion of the upper first premolar. Arellano (1951) listed Aphelops? in his preliminary list of the Ocote local fauna, hut thismustbe anerror.We havefoundonlyTeleoceras. Table 1. Measurements of teeth of Teleoceras ocotensis. Length of upper teeth was measured alongthe ectoloph and breadth at the base of the enamel. Some lower teeth called M2’s may he M3’s. I pper I eeth Lower Teeth Incisor P1 P2 39.4 x 17 9 19.5x18.6 32.2 x 33.3 30.2 x 32.0 31.1x33.8 P2 P3 P4 21.8 x 14.7 35.0 x 22.7 35.0 x 23.4 35.6 x 22.3 46.4 x24.2 P3 P4 ‘ M1? M3 ? x 295 32.0 x 34.0 38.4x44.5 45.6 x 44.5 46.7 x51.6 72.2 x51.7 67-9 x ? 71.2 x 55.0 Ml M2 42.0 x 25.6 48.9x31.9 49.6 x 31.0 61.7x32.4 65.2 x 34.1 60.9 x 36.7 58.8 x 32.9 72.0 x 53.1 74.5 x ? FAMILY EQUIDAE Astrohippus albidens Mooser Astrohippus albidens Mooser, 1968, Southwest. Nat., 13:2. This is the most abundant species of horse in the Ocole local fauna. It is representedby several fragmentary lower jaws and more than two thousand isolated upper and lower teeth. The teethofAstrohippusalbidensarestronglyhypsodont.Theupperteetharealmost straight to moderately curved. The lakes are simple, the styles moderately strong. The protocone is broadly oval in shape and usually deeply grooved on the lingual side. A hypoconal groove is absent in 90 of 176 teeth. The groove is only rarely found in the lower 40 mm of the crown. A distinct groove is present in 45 teeth, and a slender-walled hypoconal fossette was found in 41 teeth. In the lower cheek teeth, protoconid and hypoconid are broad and flattened. Meta­conid and metastylid are well elevated, are stretched and elongated, and the internal valley (linguaflexid) is sharply “V’ shaped. Parastylids are weak or absent, paralophids strong and prominent. The ectollexid (metaconid-metastylid valley) is deeper in the molars than in the premolars but does not enter into the metaconid-metastylid isthmus. Astrohippus albidens is not strongly differentiated from A. stockii, but the followingdifferences are apparent: size slightly larger, protocones longer and more deeply grooved,and linguaflexids of lower molars sharply V-shaped. Nearly all isolated lower cheek teeth of A. albidens have sharply V-shaped linguaflexids, and teeth thought to be molars tend to have the groove most sharply V-shaped. Permanent lower dentitions of A. stockii figured by Lance (1950) show U-shaped or broadly V-shaped linguaflexids. Mooser (1973) listed Astrohippus ansae (Matthew and Slirton) (type from the Coffee Ranch local fauna of Texas) and A. stockii Lance (type from the Yepomera local fauna of Chihuahua) as members of the Ocote local fauna. The teeth illustrated by Mooser show the characters of those species but we feel that the teeth are only atavistic indivi­duals showing the characters of more primitive species, such as might be selected from any large collection ol horse teeth. Teeth can he found in the collection that show various degrees ot intermediacy between the older species and the progressive conditions that are the norm for the Ocote population, Astrohippus albidens. The teeth with more complicated enamel borders ol the lakes, ungrooved protocones, and strongly developedhypoconal grooves, are slightly aberrant individuals of A. albidens. Teeth of A. albidens are distinctly more hypsodont than those ofA. ansae. Protohippus castilli Cope, from Tehuichila, Veracruz (actually in the state ofHidalgoaccording to Silva-Barcenas, 1969), seems referable to the genus Astrohippus. The holo­type tooth is apparently the only representative of the species, and according to Gidley(1907), the tooth was lost. However, it must have been found, for Osborn (1918) gave a newiigureofit. 1hetoothdiffersfromteethofAstrohippusalbidensinhavingastronglydeveloped metastyle, rounded, ungrooved protocone, distinct hypoconal groove, and rather strongly curved crown. Ihe tooth, it typical ot its population, represents a more primitive species than A. albidens. Table2. MeasurementsofteethofAstrohippusalbidens. Occlusallengthsofsomelowerdentitions are: P2-M3,116.8,128.6;P2-P4,61.0,64.7,68.3; M,55.6,57.9,61.4,62.1, 63.9. Lengths of isolated teeth measured 40 above bases N Mean Range p‘ 50 23.4 21.8-25.9 p3 - 100 19.9 18.6-22.6 M 1 -M2 100 18.7 16.7-22.6 M3 50 18.3 16.5-19.9 P2 50 20.4 18.7-22.4 P 3~P 4 100 20.8 19.9-22.1 M.-M, 100 18.9 17.4-21.0 M3 50 20.9 18.8-23.5 Mean occlusal lengths and breadths of teeth in lower jaws p 2 4 21.8 x 10.5 P 3 4 19.8x11.6 p 4 4 19.2x10.9 M, 4 18.8 x 10.8 M2 4 18.0 x 10.0 M3 3 20.5x 9.6 Astrohippus albidens is the most advanced species of its genus yet described. It, or a relatedform,maybeancestraltothegenusAsinus, thedominanthorsesofthePleisto­cene ofNorth America. Dinohippus ocotensis (Mooser) Hippotigris ocotensis Mooser, 1958, Anal. Inst. Biol. Mexico, 28:360. Protohippus muelleri Mooser, 1965. Anal. Inst. Biol. Mexico, 35:157. Dinohippus muelleri. Mooser, 1973, Southwest. Nat., 18:258. Equus (Dolichohippus) mesamexicanus Mooser, 1973, Southwest. Nat., 18:261. AlargehorseisrepresentedintheOcotelocalfaunaby afragmentarylowerjawramus with P2-M3 (holotype of D. ocotensis), a fragmentary ramus with part of P3 and com­ plete P4-M3 (holotype of /). muelleri), about 100 upper cheek teeth and about 300 lower cheek teeth, including the holotype tooth of E. mesamexicanus. 1his ancestral zebra was recognized and described as a zebra by Mooser. The teeth of the holotype of Dinohippus ocotensis are but lightly worn; crown height of Ity is about 11;ofM 2about64.Theteethareslenderandlightlybuilt. Protohippus muelleri was based on a jaw fragment with greatly worn teeth. The bases of the teeth are not exposed but are certainly less than 30 in crown height, and probablyless than 20. The teeth are broad and the enamel relatively thick. In spite of the heavy wear and advanced age of this individual, metaconid and metastylid of P4 are elevated on a distinct isthmus. 1 he apparent differences separating these two specimens are due to difference in ageat the time of death and possibly to sexual dimorphism. The type of D. muelleri was probably a male, the type of D. ocotensis a female. The large collection of isolated lower teeth includes relatively narrow specimens, probably from females, broad teeth, probablyfrom males, and numerous intermediates also. The holotype of D. muelleri is near the maximum breadth for the species, but is exceeded in breadth by some isolated teeth. Some lightly worn teeth are broad and have heavy enamel. With advanced age these would resemble the holotype of D. muelleri. The enamel pattern, however, of these un­worn teeth exhibits the characteristic features of D. ocotensis, including the wrinkled entoflexid characteristic of the genus Dinohippus. D. muelleri is a synonym of D. ocotensis. Equus (Dolichohippus) mesamexicanus was based on an upper P3 and referred upper and lower cheek teeth. Upper teeth referred to E. mesamexicanus had elongated and deeply grooved protocones. Upper teeth referred to D. muelleri had rounded, ungrooved protocones. Lower teeth referred to E. mesamexicanus were relatively slender. Lower teeth referred to D. muelleri were broad, with swollen protoconids. The differencessepa­rating lower teeth referred to D. muelleri and E. mesamexicanus are probably due to sexual dimorphism. Lower teeth of D. interpolator from the Coffee Ranch local fauna of Texas can similarly be separated into slender and broad kinds, probably the result of sexual dimorphism. Teeth with rounded, ungrooved, and posteriorly placed protocones thought to be characteristic of Dinohippus muelleri are probably atypical, atavistic specimens, such as often are found in large collections ofhorse teeth from a single locality. The teeth exhibit features more typical of/), interpolatus of the older Coffee Ranch local fauna of Texas. Specimensshowing theprotocone shapethought tobediagnosticofD.muelleriarevery rare; specimens with the protocones slightly grooved and slightly angular are more com­mon. These intermediate specimens and the “typical" D. muelleri teeth almost all appear to be premolars. Almost all of the upper teeth thought to be molars are of the E. mesa­mexicanus type, with elongated, grooved protocones. Most apparent premolars are of the E. mesamexicanus type also. In the upper teeth referred to Dinohippus muelleri by Mooser (1973), the protoconalisthmus is very short and narrow. This is a very rare condition in the collection of large horse teeth from Ocote, and is not confined to specimens with the D. muelleri pattern. The extreme is reached in a tooth, otherwise typical of/), ocotensis, in which the proto­ cone is completely isolated, as in the hipparions. Lower teeth of the large horse, represented by nearly 300 specimens, do not exhibit any features suggesting that more than one species is represented. Most of these teeth, when only lightly worn, have the floor of the entoflexid wrinkled, often greatly so. The wrinkling sometimes extends to the floor ol the metaflexid as well, and the metaconid­metastylid isthmus of the premolars, less often the molars, has small, semicircular bulges.Not uncommonly the bulges on the isthmus are at the same height on anterior and pos­terior faces ot the column, and form round, bead-like structures. The ectollexids of the premolars are relatively short and do not enter between the metaconid and metasylid;metaconid and metastylid are perched on slender isthmuses. The ectollexids of the molars are deep and blunt-ended, usually almost square-ended, and do penetrate between meta­conids and metastylids. Often there is no true metaconid-metastylid isthmus but instead a short antroisthmus and postisthmus along the edges of the ectoflexid, while the meta­conid and metastylid are widely separated and connected by a metaisthmus (for explana­tion ol terms see Skinner, 1972). Premolars and molars are easily recognized, even in advanced stages of wear. he characters thought to separate upper teeth of the three nominal species of large I horses in the Ocote local fauna grade into one another, and are now thought to be the result of differential wear on the teeth, sexual dimorphism, and individual variation within a species. Ihe lower cheek teeth cannot be separated and, as a group, exhibit a suite of characters unique in the Ocote horse fauna. All of these specimens probably be­ long to one species, for which the oldest name is Ilippotigris ocotensis (Dinohippusocotensis). Dinohippus ocotensis differs from D. mexicanus (Lance), from the Yepomera local fauna of Chihuahua, in having the protocones of the upper cheek teeth more elongate, more angular, more deeply grooved, and in having an anterior extension or spur ex­tendingin advance oftheisthmus. (We referhere to thelarge dentitionsfiguredby Lance, 1950, figs. 4 and 5. The holotype dentition of D. mexicanus is so much smaller than the larger referred specimens that we wonder ifmore than one species might be involved.) Dinohippus ocotensis differs more from D. interpolatus, of the Coffee Ranch local fauna of Texas, than it does from D. mexicanus. D. interpolatus has the upper cheek teeth shorter-crowned, more curved, protocone round or oval, with no or virtually no lingual groove, and almost never is there a trace of an anterior spur extending anterior to the protoconal isthmus. Lower cheek teeth ol D. interpolatus are somewhat shorter-crowned than those of D. mexicanus and D. ocotensis but no other consistent differences between the three species were found. The characters of Dinohippus ocotensis are very close to those of the Blancan zebra,Dolichohippus simplicidens (Cope). Ihe Blancan zebra is larger, tbe anterior spur on the protocone larger, and the wrinkling of the enamel of the floor of the entoflexid does not extend asfardownthecrownasitdoesintheHemphillianspecies.Greaterdifferencesin the skull and post-cranial skeleton probably exist, and the Blancan and Hemphillian spe­cies properly belong in separate genera. However, it seems clear that a species of Dino­hippus, possibly D. ocotensis, was ancestral to Dolichohippus simplicidens. Both upper and lower teeth of Dinohippus ocotensis change in proportions along the height of the crown. Usually the diameter of a tooth is smaller and relatively shorter anteroposteriorly near the base than near the top of the crown. An upper premolar,probably P 4 , with the occlusal surface very slightly worn and roots just formed, crown height 71.2, was measured at 20, 40 and 60 above the base of tbe enamel. Anteropos­terior diameters are: 24.5; 25.8; 26.8. A lower P 3 , crown height 71, at similar increments above tbe base of the enamel, measured: 25.6; 26.0; 28.0. Tapering of teeth is not con­sistent. Some teeth show greater tapering but others retain almost uniform diameter throughout the height of the crown. Comparable measurements can be obtained only at a uniform height above the bases of teeth. Breadth of teeth of Dinohippus can he accu­rately measured only at the occlusal surface; elsewhere the enamel is partly concealed bycementum. Length of cheek teeth can be measured almost anywhere along the length of the crown. In measurements of horse teeth used here, only the length of the tooth is given, unless otherwise specified. Measurements of teeth of Dinohippus ocotensis were taken 35 above termination of the tooth enamel. UpperP2,s(14teeth)rangedfrom30.4to36.0.I teethareprobablymalesand smaller ones females but the series is too small to show a distinctive curve. M 3 s (27) form a bimodal curve. Probable female teetb range from 20.9 to 23.7; prob­able males from 24.3 to 27.2. Upper P 3-M 2,s cannot be separated with certainty on morphological features and measurements are not given here. The 105 teeth in this group include eight possibleclasses:malesandfemalesofP3 P4 Ml ,and M 2. ,, Lower P 2 ’s (28 teeth) ranging Irom 27.1 to 29.5 may be females and from 29.5 to 31.7 may be males. Lower M3 ’s (55 teeth) measuring 26.0 to 28.6 are probably females and from 29.1 to 31.2 are probably males. Lower premolars and molars are readily separated, but P 3 cannot be distinguishedfrom P 4 and Mj cannot be distinguished from M 2. 1heP3-P4(121teeth)measurementsdonotform acurvethatcanbereadilyseparatedinto lour classes. Doubtless there is considerable overlap in measurements of teeth of males and females, and of P 3 and P4 . However 99 teeth (82%) range between 23.6 and 26.7and thisis thetypicallengthforthelowerP3 andP4ofD. ocotensis. Overlapping of measurements also must be responsible for the lack of four distinct classes in measurements of lower MrM2 s (111 teeth). However 107 teeth (96%) rangefrom 22.2 to 25.0. Maximum crown height in upper P2 of D. ocotensis is approximately 64; of upperM about60.MaximumcrownheightoflowerP2is50.7;P3orP4,75.0;ofMjorM2, 3, 71.1; ofM 3, 68.2. Measurements of the dentition of the hololype jaw of D. ocotensis (measured at the occlusal surface) are: P2-M3 , 163.0; P 2-P482.6; Mj-M3 , 80.4; P2 , 30.0 x 13.3; P3 , , 27.0x15.0;P4,25.4x12.5;Mj,25.0x11.5;M 2,27.0x11.2;M 28.4x9.0. 3, Measurements of the dentition of the holotype jaw of Protohippus muelleri are: P4-M3,99.3;M,-M3.75.2;P4,24.5x17.0;M1?22.8x 13.5;M 22.5x 13.5;M 3, 2, 29.5 x 12.5. The greatly worn teeth are relatively shorter than would be the case were they measured 35 above tbe base of the enamel. The holotype upper P 3 of Equus (Dolichohippus) mesamexicanus measures 26.7 x 26.0. Neohipparion otomiiMooser Neohipparion otomii Mooser, 1960, Anal. Inst. Biol. Mexico, 30:376. When Mooser described Neohipparion otomii he designated as the holotype a series of 12 upper molars and premolars, selected from his large collection, not from one indivi­dual horse, and two lower jaw fragments. Silva-Barcenas (1969:10) noted that this was incorrect, and selected as lectotype a lower jaw fragment (MWU 11378) with the com­plete and well-preserved cheek tooth series. The common Neohipparion of the Ocote local fauna is represented by two fragmen­tary jaws and more than two hundred isolated teeth. N. otomii is similar to N. floresi Stirton,ofthe Yepomeralocalfauna,huttheprotoconesoftheupperteethareelongatedand have pointed ends rather than the “fishtail shape typical of N. floresi The pli caballin of the Ocote horse is usually bifurcated, a rare condition in the Yepomera horse. N. otomii seems to he related to, probably is descended from, hut is more advanced than N. floresi. In 18upperP2,s,theprotoconeisisolatedin4teeth,unitedtotheprotoselenebya narrow isthmus in 13, and by a broad isthmus in one. Eight teeth have a small, slender,simple pli caballin and 10 have the pli caballin doubled or even more complicated. In only 7 teeth are the fossettes normal and separated. In 1 1 teeth the labial borders are fused to form a single heavy enamel hand across the tooth. In 7 of these teeth the thin enamel forming the complicated anterior, lingual and labial borders of the fossettes is joined to form a single large fossette. In four teeth the thin enamel, medially, abuts the thick labial enamel to form two fossettes dependent from the labial border. Intermediateteeth (P3-M2)arehypsodont: thehighestcrown measures 74mm,and there are numerous teeth exceeding 68 mm. All teeth are slightly curved, and fossettes are extremelycomplicated. Premolarsare separatedfrom molarsbylargersize,thickerenamel, and bolder enamel pattern. P 3 cannot be distinguished from P 4 and M 1 cannot he sepa­ , ratedfromM 2.Measurementsof25upperP3,sandP4,srangefrom23.0to26.2x 17.0 to 22.7; means 23.7 x 19.2. Measurements of 35 upper M 1’s and M2,s range from 19.1 to 22.2 x 14.4to 21.8; means 20.9 x 19.5. Thirty upper M 3,s have isolated protocones, complicated fossettes, simple pli caballins,andlong,curvedcrowns.Maximumheightis71.4.Measurementsare: 20.7to26.6(mean22.7) x 16.2 to 19.2 (mean 17.5). The lower cheek teeth of Neohipparion otomii are distinctive. The metaconid and metastylid are rounded and prostrate, their axes nearly parallel with the anteroposterioraxis of the tooth. The intervening linguaflexid is broadly and shallowly “U” shaped or, in about half of the teeth, “W” shaped, with a prominent median spur. The metaconid­metastylid isthmus is of complicated shape, usually with slender anterior and posteriorprojections. Paralophids and parastylids are strong and prominent. Premolars 3 and 4 cannot be distinguished from molars 1 and 2. The highest P 2 measures 74.5. Measurements of 36 lower P 2 ’s are: 19.8 to 24.9 (mean22.0) x 9.7 to 12.5 (mean 11.0). The highest M 3 measures 72.5. Measurements of 25 lowerM3sare:22.5to27.0(mean25.2)x 9.8to11.4(mean10.3). ISeohipparion manias Mooser Neohipparion monias Mooser, 1964, Anal. Inst. Biol. Mexico, 34:394. This species was described on the basis of two upper molars (MWU 11130, holotype;MWU 11131, paratype). The teeth differ from the upper molars of Neohipparion otomii as follows: crowns very high, teeth narrow anteroposteriorly, lakes compressed and simple, pli caballin obsolete, and protocones lingually placed. We were tempted to consider this form a synonym of Neohipparion otomii but, if the listed characters are simply extreme variants of that species, we should expect to find teeth showing intermediate conditions, but we did not. Further, just as the characters of N. otomii suggest relationship to but advance over N. floresi Stirton, of the Yepomeralocal fauna of Chihuahua, so do the characters of N. monias show relationship to but advance over N. arellanoi Stirton, of the Yepomera local fauna. Stirton (1955) noted that the collecting sites containing remains of N. floresi and N. arellanoi were different,but were close together and at the same stratigraphic level. Perhaps the ecological de­mands of N. otomii were different from those of N. monias, and the latter rarely visited the Ocote habitat, where N. otomii was common. Stirton (1955) suggested that Neohipparion arellanoi might prove to be a synonym of A. floresi, and N. fnonias may yet prove to he a synonym ofiV. otomii. However, at this time it appears that N. otomii is a species descendent from N. floresi and N. monias is the lineal descendant of N. arellanoi. Ihe holotype tooth of Neohipparion monias measures 20.3 x 21.8; the paratypemeasures 19.2 x 21.0. Nannippus aztecus Mooser Nannippus aztecus Mooser, 1968, Southwest. Nat., 13:7. Nannippus aztecus is known from the holotype maxillary, two mandibularfragments, and about 80 isolated cheek teeth. It is a diminutive species, whose relations are clearlywith N. minor (Sellards) of the Hemphillian of Florida. Lance (1950) referred tiny Nan­nippus remains from the Yepomera local fauna of Chihuahua to N. cf. minor, and this material also is close to V. aztecus. The collection from Ocote is relatively large. Un­fortunately, tiny Nannippus from elsewhere are known from relatively few specimens. The holotype maxillary7 of N. aztecus, carefully described by Mooser (1968), is from an aged individual. The protocones are connected to the protoselenes in all three pre­molars, a condition thought by Mooser to be the result of wear. However other teeth,includingsomethatarecertainlyP3,sand P4,s,areaswornastheteethoftheholotype, or even more so, but the protocones of these teeth are isolated. Teeth of Nannippus aztecus are extremely hypsodont, and the enamel pattern remains almost to the bottom of the teeth; typically, the protocones of all cheek teeth except P 2 are isolated to the base. We now regard the connected protocones ofP3 and P 4 of the holotype as an ab­normal condition, exposed only near the tooth base. Dr. S. David Webb has provided specimens, including a cast of the neotype palate, of Nannippus minor. This small collection, like that from Ocote, includes a seemingly dis­proportionate number of greatly worn teeth. In lightly to moderately worn upper teeth from Ocote, the protocones are slender ovals, and the oval shape, though it becomes broader with wear, remains consistent, even in aged horses. In N. minor the protocones are rounded, even in young individuals, and become almost perfectly round with age. Until greatly worn, upper teeth of N. minor have strongly developed pli caballin folds on both premolars and molars. The premolars ofN. aztecus have strong pli caballin folds but M 1 has only a slight fold, and M 2 and M 3 have none. There is some question as to the consistency of this last character, for identification of isolated teeth, other than P 2 and M is often difficult, and the comparative material of N. minor is limited. 3, In lower cheek teeth, faint parastylid ridges are present in N. aztecus but are absent in N. minor. In both species the ectoflexid penetrates the protoconid-hypoconid isthmus of the molars but not the premolars. Material available suggests that Nannippus minor and N. aztecus are distinct species. More and better material from Florida might show the apparent differences to not be significant. N. cf. minor, from the Yepomera local fauna, may be closer to N. aztecus than toN. minor. None of the upper teeth of Nannippus aztecus appear to be near full crown height.Three lower teeth, probably premolars, are between 34 and 35 high, and appear to be near full crown height. Upper and lower P2’s and M3’s are readily identified, and lower P3 and P4 are easilyseparated from Mj and M 2. Some uncertainty exists as to separating the remaining uppercheek teeth, but it is felt that the teeth are probably correctly placed, and not more than one tooth from their proper position in any case (e.g., some teeth termed M2,M 1’smighthe s, etc.). There is some doubt as to separation oflower P3 from P4 , but most are probably correctly identified. We are unable to separate lower Mj’s from M 2 ’s. The measurements that follow are means; ligures in parentheses are the number of teeth measured. All teeth are lightly to moderately worn; none are of aged individuals. UpperP2(i),18.1x4.4;upperP3(2),16.1x14.8;upperP4(3), 14.8x13.8;upperM 1(3), 14.1 x 13.4; upper M 2 (3), 13.6 x 13.0; upper M 3 (6), 13.3 x 11.2; lower P 2(2), 16.1X7.9;lowerP3(5), 14.3x8.3;lowerP4(4), 14.1x7.9;10werMjorM2(9), 13.5 x 7.3; lowerM 3 (5), 16.1 x 6.5. I he diminutive species of Nannippus seem to be limited to southern Hernphillianlocalities, from Morida to Mexico. Whereverfound they are associated with large species,except at Yepomera, Chihuahua, where no large species has been reported. A few fossils seemto suggestthatthelineoftinyspeciesextendedintotheBlancanage,butspecimensdiscovered to date are poor. Nannippus hesperides Mooser Nannippus hesperides Mooser, 1968, Southwest. Nat., 13:10. I he only material representing a large Nannippus found to date at Ocote is an uppertooth,probablyanMl ,andawornlowerP3orP4 Thespeciesistheraresthorseinthe . fauna. Only the lower tooth was figured in the original description. The holotype (thoughonly lightly worn and well preserved) has the occlusal surface eroded, and the enamel pattern could not be clearly distinguished. The occlusal surface of the tooth has been sanded and polished to reveal the enamel pattern (Fig. 3). The crown height was origi­nally 57.Thetoothmeasures 18.6x 14.3.Thelowerpremolar measures 17.4x 10.3. Nannippus hesperides is very similar to N. lenticularis (Cope), from the Coffee Ranch local fauna of Texas. An upper M 1 from the Coffee Ranch is in almost the identical stageof wear to the holotype tooth of N. hesperides. It measures: crown height, 53.5; dia­ meter, 19.0 X 14.0. Seven lower P 3 s and P4 ’s, in the same stage of wear as the lower tooth from Ocote, have mean diameters of 18.1 x 9.9. Nannippus montezuma (Leidy) [llippotherium montezuma Leidy, 1883, type from “Lacualtipan,” now Zacualtipan, Hidalgo] is apparently known only from the holotype,an upper P 3 . The crown is 47 high, and appears from the figure (Osborn, 1918:198) to be but slightly worn. The diameter is 19.5 x 17. The protoconc is long and slender with pointed ends but would doubtless appear more lens-shaped when worn more deeply. The anteroposterior length of the tooth is within the range observed in N. lenticularis from the Coffee Ranch but the stated breadth is extreme, and may be an error. The enamel pattern, including shape of protocone, can be matched by lightly worn upper teeth from the Coffee Ranch and by the holotype of N. aztecus. Nannippus peninsulatum (Cope) [Hippotherium peninsulatum Cope, 1885, type fromTehuichila, Veracruz, actually from the state of Hidalgo, according to Silva-Barcenas,1969], is known from an upper tooth, probably an M The tooth is high-crowned 2. (51.1) and, to judge from the figure (Osborn, 1918:198), only lightly worn. It measures 17.5 x 15.0. The protocone is elongated and has pointed ends; it will probably show a lens shape on further wear. I his tooth too can be matched by specimens from tin; Coffee Ranch local fauna of Texas, but the crown appears to be more curved than is usual in N. lenticularis. 11ms three species oi large Nannippus have been named from the Hemphillian of Mexico, two from single teeth and one from two teeth. All three appear doubtfully dis­tinct trom A. lenticularis. Gidley (1907) considered Nannippus peninsulatum to be synonymous with N. montezuma. Better material may show that all or some of the nominal Mexican species are valid, or may be synonymous with N. lenticularis. Nannippusmontezuma (Leidy) has priority over the three younger names. Until the relationships of the large Nannippus are determined, we suggest that all four names be retained. Fig. 3.—NannippushesperidesMooser.A,uppertooth,probablyfirstmolar,holotype,inlabialand occlusal views. B, lower tooth, probably third or fourth premolar, in labial and occlusal views. ORDER ARTIODACTYLA FAMILY TAYASSUIDAE Desmathyus brachydontus new species Holotype.—TMM 41685-13, lower third molar, probably of a male. Referred.-19 upper third molars, 27 lower third molars, 11 upper second molars,10 lower second molars, 69 other isolated cheek teeth, and two tusk fragments. Type locality.—Arroyo de la Carreta, near the village of Los Rodriguez, District of San Miguel de Allende, Guanajuato, Mexico; Ocote local fauna, Hemphillian Age, Plio­cene Epoch. Diagnosis.—A very large peccary, larger than Desmathyus pinensis Matthew, the type of the genus, and other described species. Almost exactly the size of the Blancan-agePlatygonus bicalcaratus Cope. Distinguished from described species of Desmathyus bysmall but distinct posterior lobe of upper M 3 and cusps of molars only lightly wrinkled. Readily distinguished from species ofPlatygonus by extremely brachydont dentition. Description.—The holotype tooth is beautifully preserved except that the roots have been damaged. The crown is almost unworn. The 46 referred upper and lower third molars are also well preserved and range from unworn to moderately worn. In all of these teeth the cusps are only moderately wrinkled, the valleys are deep and sharp, and the cingula are strong. The enamel pattern, in occlusal view, is almost identical to that of Platygonus bicalcaratus Cope, from early Blancan deposits of the Great Plains of the United States, except that the cusps lie closer together, more appressed at their lips, than in the Blancan animal. In unworn teeth the tips of the cusps are barely separated while in P. bicalcaratus the cusps are well separated. In lateral view, however, D. brachydontusdiffers from all described Platygonus in its strongly brachydont teeth. In upper and lower third molars in comparable stages of wear, teeth ofD. brachydontus are readily identified in all instances. Dr. David S. Webb has pointed out that the present peccary teeth from Ocote are best referred to Desmathyus rather than to Platygonusand that similar teeth occur in , Hemphillian deposits in Horida. Dr. Webb kindly provided several specimens from Florida,andresemblanceofFloridaandMexicanfossils isclose.MeasurementsofFlorida teeth are within or nearly within the range of variation found in Mexican teeth but the Floridateethhave thecusps ofthethirdmolarsmorewrinkled.Of26lowerM3’sfrom Ocote,onehastheposteriorcusp aswrinkledasdoesthesingleFloridaM andanother 3, tooth has the anterior cusp as wrinkled as the anterior cusp of the single lower M 3 from Florida. The enamel pattern of the Ocote teeth is relatively simple and none have both cusps as wrinkled as does the Florida tooth. An upper M 2 from Florida has the cingulumstrongerandmorewrinkledthananyof 10upperM2,sfromOcote.TwolowerM2’sfrom Florida are available, one measuring 21.4 in length (equal to the longest upper M 2 from Ocote) and one measuring 21.6 (exceeding all Ocote teeth in length). The cusps and cingula of these teeth are also more complicated than those of the Ocote teeth. Differ­ences are slight but, in view of the limited material available for comparisons, appear to be significant. Only upper and lower second and third molars were studied. The 69 other teeth available include premolars and first molars of juvenile, mature and senile peccaries, and it has proven impossible to determine with certainty the correct place in the tooth row of many teeth, or even to separate some upper from lower teeth. Upper third molars, Fig.4.—Desmathyusbrachydontusnew species,holotype,lowerthirdmolar,inlingualandocclusal views. however, have small hut well-developed posterior lobes (variably present in species of Platygonus) that give the tooth a triangular appearance, and upper second molars are large, almost square, and have heavy cingula. Lower third molars are quite elongated,have strong posterior lobes and lack cingula. Lower second molars resemble upper second molars but cingula are weak or lacking. °O Measurements of molars were made at the base of the crown, where the enamel has its greatestdimensions.Measurementsoflengthofupperthirdmolars,andoflength and breadth ol lower third molars, form bimodal curves. Possibly there was sexual dimorph­ism in size of these teeth in Desmathyus hrachydontus, as there is in the domestic pig.However, such size difference in the sexes is not strong in living peccaries, so far as I can discover, nor has it been reported in Platygonus. Possible male and female specimens are grouped together in Table 3. A fragment of upper tusk has a maximum diameter of 22.3 x 13.5, and a fragmentthat appears to be of a lower tusk has a maximum diameter of 17.6 x 15.6. Table 3. Measurements of upper and lower molars of Desmathyus hrachydontus. Length Breadth N Mean Range Mean Range M2 11 20.8 19.7-23.0 17.1 14.2-19.4 M3 19 24.8 22.6-27.6 18.2 16.6-20.8 M 2 10 20.7 19.2-21.4 15.9 14.2-17.0 M3 26 29.4 25.8-34.4 16.6 14.8-19.8 Etymology.—The name is based on the extremely brachydonl teeth characteristic of the species. Discussion.—Desmathyus (including Pediohyus Loomis) is known primarily from earlyMiocene sites of the Great Plains area of the United States (Matthew, 1907; Loomis,1910) but the genus has been little mentioned in recent years. During the Clarendonian and Hemphillian ages, the common peccary of the Great Plains was Prosthenops. It now appears that Desmathyus was displaced southward where it persisted to give rise to Blancan and Pleistocene species ofPlatygonus. Webb hasnoted(personalcommunication)thattheDesmathyusoftheHemphillianof Florida is almost intermediate between Desmathyus and Platygonus, and the same is true of theOcotepeccary.DirectcomparisonofteethofD.hrachydontuswithBlancan speci­mens of Platygonus bicalcaratus shows that Matthew (1924) was correct in consideringDesmathyus to be the ancestor of Platygonus. A hiatus remains between known Hemp­hillian species of Desmathyus and Platygonus bicalcaratus, but specimens demonstratingthe transition may eventually be found in deposits on the Mexican Plateau. FAMILY CAMELIDAE Megatylopus cf. Megatylopus matthewi Webb Megatylopus matthewi Webb, J 965, Bull. Los Angeles Co. Mus. Sci., 1:42. Several isolated camel teeth are very well preserved and are referable to Megatylopus.Direct comparison with topotypes from the Coltee Ranch of Texas shows no significantdifferences. Cheek teeth, especially upper teeth, of Hemphillian camels taper upward swiftly. The upward taper is less iu the anteroposterior diameter, and measurements of length taken at a uniform height above the base of the enamel are usually comparable. The taper in the transverse diameter, however, is often so great that a difference of a millimeter or two inthe height at which measurements are made will result in significant differences.Measure­ments ol transverse diameter of camel teeth, especially upper teeth, must he used with caution. Iwo upper P 3,s (8869, 8870), smm above the base measure 24.9 x 14.5 and 25.4 x 14.1. Iwo teeth that are probably upper M 1’s (8866, 8868) measure, 25 above the base, 44.4 x 29.6 and 43.3 x 28.1. A tooth thought to he the upper M 2 (8864) measures 49.6 x 38.5 at 25 above the base and lower Mj or M 2 (8865) measures 50.9 x 23.2. Pliauchenia sp. A fragment of lower jaw witli part of M 2 and complete M 3 (8871), two isolated lowerM 2 s, or M 3 s with the posterior lobe broken away (8867, 11238), and an astragalus(11240), represent a large camel, much smaller than Megatylopus, but too large and heavy-jawed for a species of Hemiauchenia. Teeth are like those of the large camel from the Coffee Ranch ofTexas thathas been referred to Pliauchenia. The M 3 in the jaw fragment (8871) measures (10 above base) 40.8 x 18.5.The lowerM 2 (or damaged M 3) measures (10 above base) 44.4 x 16.1. The astragalus measures: greatest length, 65.5 plus; greatest breadth, 48.0. Hemiauchenia sp. There are ten complete molars, some tooth fragments, and four astragali, apparentlyreferable to Hemiauchenia. The teeth are much smaller and more lightly built than teeth ol Pliaucheniasp.,butlargerand higher-crowned than teethofPalaeolamaguanajuatensis. They differ additionally from teeth of Palaeolama in that the styles of the molars are distinctly weaker and less prominent. The Hemiauchenia teeth closely resemble teeth referredtoHemiaucheniafrom theCoffeeRanchlocalfaunaofTexas.Ifmorecomplete materialwereavailable, theOcotecamelmightprovetobespecificallydistinctbut,until such specimens are found, itseems best to leave the specific identity undetermined. Measurements of teeth (5 above base of enamel) are: upper M 1or M 2 (2), 25.7 x 18.1, 25.4 x 17.7; lower M 1 or M 2 (5), 24.5 x 13.7, 25.6 x 14.8, 24.3 x 14.0, 26.7 x 13.2, 24.3x 14.2,lowerM3(3),33.5x 14.0,30.8x 13.2,32.2x14.8.Theastragalimeasure: greatestlength,60.6, 56.2,54.2,and51.0;greatestbreadth,41.2,39.8, 39.8,and38.8. Palaeolama guanajuatensis new species Hololype. IMM 41685-11, upper third premolar, well preserved except that the roots are broken away. Referred.-Right maxillary (9010) largely encased in hard matrix hut crowns of P 3 and P exposed. M 1 broken away, M 2 and M 3 present; 75 isolated upper and lower cheek teeth; 4 astragali; 1 medial phalanx. Type locality.—Arroyo de la Carreta, near the village of Los Rodriguez, District of San Miguel de Allende, Guanajuato, Mexico; Ocote local fauna, Hemphillian Age, Pliocene Epoch. Diagnosis.—A tiny camel, approximately the size of the modern domestic llama. Differing from Hemphillian and Pleistocene species of Hemiauchenia in smaller size,shorter-crowned teeth, upper P 3 with three rather than two roots, and styles of lower molars better developed. Similar to Pleistocene species of Palaeolama in size and basic tooth pattern hut styles not as well developed and ectoloph between styles gently convex rather than drawn out into ribs. Description.—l he holotype tooth is well preserved and only lightly worn. It measures 11. i x 10.0, is triangular in shape, has a single open crescent of moderate depth, and shows the remains of three roots. Ihe tooth is essentially similar to upperP 3,s of Palaeo­ lama mirifica (Simpson) (Wehh, 1974). There is a lightly worn low'er P 4 (9005) measuring 14.8 x 7.3, and a moderately worn P 4 (9013), measuring 13.7 x 7.9. Both have the typi­cal,complicatedenamelpatternofPalaeolama(Webb, 1974;181). The maxillary fragment is largely unprepared, hut most of the teeth are exposed.Further preparation might endanger the specimen. The moderately worn P 3 -M 3 has an Fig. 5.—Palaeolama guanajuatensis new species, holotype, upper third premolar, in labial, occlusal and inferior views. alveolar length of about 84.5. The occlusal length of the individual teeth, measured along the ectoloph, are: P3 , 11.1; P 4 12.0; M l , ? ; M 2, 23.5; M 3, 23.0. The premolars are smaller than, and the molars approximately the size of, the few specimens ofPalaeolama available to Webb (1974) from Florida. Upper and lower molars are well represented in the collection. The teeth are small and brachydont. Theycloselyresembleteethofthe livingdomestic llamaofSouthAmerica, and the Pleistocene Palaeolama mirifica. They are readily separated, however, by the , relatively poor development of the ribs (the sharp labial bulges) of the upper and lower molars. P. guanajuatensis is congeneric with and probably ancestral to P. mirifica. Etymology.—The species is named for the Mexican state of Guanajuato. Discussion.—Small camels are known from Hemphillian deposits of the United States,but those we have seen seem referable to the genera Hemiauchenia or Plinuchenia. The presence in the Ocote local fauna of a species of camel probably ancestral to the Pleisto­cene forms of Palaeolama of the southeastern United States and South America suggeststhatthe originof theSouthAmericancamelidstookplaceinMexicoduringHemphilliantimes. Both Palaeolama and Hemiauchenia occur in the Ocote local fauna and are well differentiated. Divergence or radiation of the ancestral llama stock must have occurred in early Hemphiliian or Clarendonian time. Measurements.—All teeth were measured 5 above the base of the enamel. In llamas,the upper M 2 is larger than M l and the posterior lobe of M 3 is relatively narrow. How­ , ever, in the series of isolated upper molars from Ocote, we could not reliably separateM 1 from M 2 , and even have doubts about the identity of some M 3,s. Variation in relative development of the posterior lobe causes extreme variation in the length of lower M 3. The four astragali measure: greatest length, 48.1, 41.3, 41.2, 38.2; greatest breadth,30.1, 28.4, 27.3, 25.0. The greatest length of the calcaneum is 85.9. The ungual phalanxmeasures: greatest length, 78.0; breadth at proximal end, 20.0; midshaft breadth, 11.0;breadth of distal end, 16.1. Table 4. Measurements of teeth of Palaeolama guanajuatensis. Length Breadth N Mean Range Mean Range Upper M 1 -M 2 30 21.2 18.1-23.6 15.7 13.0-18.5 (combined) Upper M3 17 22.6 21.0-24.2 15.0 12.5-16.6 Lower P4 2 14.4 14.1-14.6 8.0 7.7-8.2 Lower Mj 7 18.6 17.4-20.8 10.9 9.2-14.3 Lower M2 9 20.9 20.0-21.9 10.9 10.0-11.7 Lower M3 12 26.5 21.1-29.5 11.2 9.7-12.8 FAMILY ANTILOCAPRIDAE Unidentified Antilocaprid A large antilocaprid was common at Ocote, and the collection contains a maxillaryfragmentwithM 1-M3(8872)andmanyisolatedupperandlowercheekteeth. The teeth are large and hypsodont. They are of the proper size to refer to Hexobelo­meryx fricki Furlong, of the Hemphillian of Florida (see Webb, 1973). In view of the large number of described species of Hemphillian antilocaprids, and in the absence ot a horn core from Ocote, the prongbuck from Ocote is left unidentified. Teeth of antilocaprids are strongly wedge-shaped, broader above than near the base. Measurements are comparable only when taken at a uniform level above the base of the tooth enamel. The Ocote teeth were measured 20 above the base of the enamel. UpperM 1 s are difficult to separate from M 2 ’s. At 20 above the base of the enamel, diameters of M and M2soverlap.However,Mlssarelesswedge-shapedthan M2’s,andhigheron thecrowntheM2sarelongerthanteethhereidentifiedas M1’s. Table 5. Measurements of teeth of unidentified antilocaprid. Length Breadth N Mean Range Mean Range Upper M1 10 12.5 10.8-13.7 9.5 8.1-11.7 Upper M2 8 13.4 12.0-14.0 10.7 9.7-11.7 Upper M3 15 18.6 16.7-20.3 10.0 9.4-11.3 Lower Mj 1 10.2 6.1 Lower M2 11 12.9 12.0-14.5 6.9 6.5-7.8 Lower M3 6 23.8 22.5-24.1 7.2 6.8-7.S AGE AND RELATIONSHIPS OF THE OCOTE LOCAL FAUNA Since there are no sequential strata at the Ocote collecting site that might aid in accu­ rate dating of the Ocote sediments and, so far as we know, no datable volcanic ashes or basalts intimately associated with the sediments, the position of the Ocote local fauna among described Hemphillian local faunas must be determinedby biostratigraphic corre­lation with other local faunas. The Ocote local fauna is clearly of late Hemphillian age. In comparisons with Claren­donian and early Hemphillian local faunas, the evolutionary level of the Ocote mammal species is much advanced over species of the older faunas. On the other hand, the Ocote local fauna is much older than described early Blancan local faunas. By the Blancan age,rhinoceroses (Aphelops, Teleoceras), so characteristic of Hemphillian local faunas, had become extinct. Of tbe four genera of Ocote horses, Neohipparion was extinct by the early Blancan age, Dinohippus had evolved into Equus (Dolichohippus), Astrohippus had evolvedintoAsinusand only Nannippuspersisted into theBlancan. Mammalian fossils of middle and late Hemphillian age are known from many localities in the United States. However, except in Texas and Nebraska, there are few up-to-datefaunal lists available from these sites, making difficult the correlation of the Ocote local fauna with that from elsewhere. The lack of microvertebrate fossils, such as shrews and mice, is another serious obstacle to correlation. Nevertheless the Ocote fossils, and espe­cially the horses, do permit rather close placement of the Ocote local fauna in the late Hemphillian series. Hemphillian local faunas from the Great Basin, the Pacific Coast and e. stward to Nevada and Arizona differ significantly from the Ocote local fauna. For example, Shot-well (1958 and elsewhere) lists Hipparion and Neohipparion, but not Nannippus or anyof the caballine horse genera, from Oregon. MacDonald and Pelletier (1958 or 1961) and MacDonald (1959) list only Pliohippus from the Hemphillian of Nevada, other than the occurrence of Neohipparion leptode Merriam from the Thousand Creeks local fauna, and Lindsay and Tessman (1974) similarly list only Pliohippus from Hemphillian local faunas of Arizona. The significance of the rarity of hipparion horses in the southwest, and the apparent absence of caballine horses in Oregon, during Hemphillian time, is unclear. Both tribes are represented in most Hemphillian local faunas of the Great Plains and in Chihua­hua. On present evidence it would seem best to avoid attempts to correlate the Ocote local fauna with Hemphillian local faunas from the western United States. In the eastern United States, Hemphillian land mammals are known only from Florida (Webb, 1969). Complete faunal lists are not available but horses include a tiny Nannippus much like N. aztecus from Ocote along with a larger Nannippus (N. ingenuus Leidy),apparently similar to N. hesperides Mooser, an advanced species of Osteohorus (0. dud­leyi) the size of the Ocote hyaenoid dog, and other faunal elements resembling those of JO’ O Ocote. Detailed comparison of the Ocote local fauna with the mammals of the upperBoneValleyfaunawouldbeinformative,butisnotpossibleat thistime. A Hemphillian fauna has been listed from Nebraska by Schultz, Schultz and Martin (1970). Tanner (1967) has used this assemblage, the Kimball local fauna, as the typelocal fauna of his ‘'Kimballian Land Mammal Age. This mammalian age was intended for ~ the time interval between Hemphillian and Blancan ages, and thus toOfurnish a technical name for the time period usually termed late Hemphillian. I see little utility in a formal name for the period usually called late Hemphillian. A more serious criticism is that the Kimballian Land Mammal Age is based on a local fauna, tbe Kimball local fauna, that is not intermediate between Hemphillian and Blancan ages but instead is almost exactly the age oi or a trifle older than the Coffee Ranch local fauna ofTexas, the type local fauna of the Hemphillian Land Mammal Age. Several species, including a peccary, Prosthenopsgrciffhami Schultz and Martin, a rhinoceros, Aphelops kimballensis Tanner, a horse, Neo­hipparion eurystyle (Cope), and an antilocaprid, Texoceros altidens (Matthew) (=Texo­cerosguymonensis 1rick),andperhapsotherspeciesalso,appeartobetbesameorvery closely related in the two local faunas. For purposes of correlation the Kimball and Coffee Ranch local faunas may be considered informally as mid-Hemphillian in age.The Ocote local fauna resembles the Coffee Ranch and Kimball local faunas at the generic level. However, when materials are suitable for specific comparisons, the Ocote mammals are seen to be distinctly more advanced than those of the Coffee Ranch and Kimball local faunas. Some of these differences have been detailed in the precedingaccounts of the horses. Teleoceras apparently does not occur in the Coffee Ranch local faunabuttheOcoteTeleoceras isadvanced overT.schultziTanneroftheKimballlocal fauna. Several "latest Hemphillian’' local faunas are listed by Schultz (1977) from Texas but,although the faunal lists certaiidy suggest that the local faunas are younger than the Coffee Ranch local fauna or other described mid-Hemphillian local faunas, uncertain identifications make detailed comparisons with the Ocote local fauna difficult. If the Astrohippus cf. stocki is truly A. stocki, and the Dinohippus cf. mexicanus is actually D. mexicanus, these local faunas correlate with the Yepomera local fauna of Chihua­hua and are slightly older than the Ocote local fauna. At present we can only suggestthat the evolutionary level of the latest Hemphillian local faunas of Texas is approxi­matelythatoftheOcotelocalfaunaor abitolder. Tbe Yepomera local fauna of Chihuahua (Lance, 1950;Stirton, 1955) is chronologi­cally close to the Ocote local fauna. The giant ground squirrel (Paenemarmota mexicana) appears to be the same species in tbe two local faunas. Tbe horses of the Ocote local fauna are similar to but slightly advanced over the Yepomera horses. Thus, the Ocote local fauna appears perhaps slightly younger than the Yepomera local fauna. On the basis of these correlations, the Ocote local fauna appears to be the youngestHemphillian local fauna well represented by fossil materials. There is an obvious gap in the fossil record between the Ocote local fauna and well-known early Blancan faunas such as the Rexroad of Kansas (Skinner and Hibbard, 1972), a gap seemingly not yetrepresented by fossils from the Great Plains area of the United States or on the Mexican Plateau. Broadly speaking the late Hemphillian faunas of the Mexican Plateau and the Great Plains have close formal continuity. Successional species of the horses and Teleocerasand perhaps the giant ground , squirrel and other forms, seem to show steps in an evolutionary sequence, in chronolo­gicalorder, ontheGreatPlainsoftheUnitedStatesandontheMexicanPlateau,ending, 'O’ for Pliocene forms, in the Ocote species. However two species, Desmathyus brachydontusand Palaeolama guanajuatensis, are present in the Ocote local fauna that are not repre­sented, so far as we know, by ancestral forms in the Pliocene of the mid-continental United States. Thus certain lineages seem to have centered uniquely in the Mexican Plateau well south of the classic fossil fields of the High Plains. Additional Hemphillianlocal faunas, both older and younger than the Ocote, may well be discovered on the Mexi­can Plateau, thus shedding further light on the evolution of the Pliocene mammal fauna. LITERATURE CITED Arellano, A. R. V. 1951. Research on the continental Neogene ofMexico. American J. Sci., 249:604-616. Alvarez, 4. 1965. Catalog© Paleomastozoologico Mexican©. Dept. Prehist., Inst. Nac. Anthropol. Hist. Mexico Publ., 17:1-70. Bryant, M. D. 1945. Phylogeny of Nearctic Sciuridae. American Midi. Nat., 33:257-390. Cope, E. D. 1884. The extinct mammalia of the Valley of Mexico. Proc. American Phil.Soc., 22:1-21. Cope, E. D. 1893. A preliminary report on the vertebrate paleontology of the Llano Estacado. 4th Ann. Rept. Geol. Surv. Texas, 1892:1-137. Dalquest, W. W. 1969. Pliocene carnivores of the Coffee Ranch. Texas Mem. Mus. Bull.,15:1-43. Dalquest, W. W. 1975. Vertebrate fossils from the Blanco local fauna of Texas. Occas Papers Mus. Texas Tech Univ., 30:1-52. Gidley, J. W. 1907. Revision of the Miocene and Pliocene Equidae of North America American Mus. Nat. Hist. Bull., 23:865-934. Hirshfield, S. E., and S. I). Webh. 1968. Plio-Pleistocene megalonychid sloths of North America. Florida State Mus. Bull., 12:213-296. Lance, J. F. 1950. Paleontologfa y estratigrafia del Plioceno de Yepomcra, Eslado de Chihuahua, la. Parte: equidos, except© Neohipparion. Univ. Nac. Ant. Mexico, Inst. Geol. Bob, 54:1-81. Lindsay, E. H., and N. 1. lessman. 1974. Cenozoic vertebrate localities and faunas in Arizona. J. Arizona Acad. Sci., 9:3-24. Loomis, F. B. 1910. A new genusof peccaries. American J. Sci., 30:381-384. MacDonald, J. R. 1950. The middle Pliocene mammalian fauna from Smiths Valley, Nevada. 1959. J. PaleontoL, 33:872-887.MacDonald, J. R., and W. J. Pelletier. 1958 (or 1961). Pliocene mammalian faunas of Nevada, U. S. A. Congr. Geol. Internac., Ses. 20, 1956, Sec. 7, pp. 365-388, Mexico D. F. Martin, L. D., and C. B. Schultz. 1975. Scimitar-toothed cats, Machairodus and Nitnra­vides, from the Pliocene of Kansas and Nebraska. Univ. Nebraska State Mus Bull10:55-63. Matthew, W. D. 1907. A lower Miocene fauna from South Dakota. Bull. American Mus. Nat. Hist., 23:169-219. Mawby, J. E. 1965. Machairodonts from the late Cenozoic of the Panhandle of Texas. J. Mammal., 46:573-584. Mooser, 0. 1958. Una cebra fosil de laMesa Central de Mexico. Anal. Inst Biol Mexico 28:359-363. Mooser, 0. 1960. Un equido fosil del genero Neohipparion de la Mesa Central de Mexico.Anal. Inst. Biol., Mexico, 30:375-388. Mooser,0. 1963.Neohipparionmaniasn.sp.equidofosildelPliocenodelaMesaCentral de Mexico. Anal. Inst. Biol., Mexico, 34:393-396.Mooser, 0. 1965. Una nueva especie de equido del generoProtohippus del Plioceno de laMesa Central deMexico. Anal. Inst.Biol., Mexico, 35:157-158.Mooser, 0. 1968. I ossil Equidae from the middle Pliocene of the Central Plateau ofMexico. Southwest. Nat., 13:1-12. Mooser, 0. 1973. Pliocene horses of the Ocote local fauna, Central Plateau of Mexico.Southwest. Nat., 18:259-268. Ochoterena, H., and A. Silva-Barcenas. 1970. Cuvieronius arellanoi sp. n., mastodonte del Pliocene del Estado de Oaxaca. Univ. Nac. Aut. Mexico, Inst. Geol., Paleontol. Mexi­cana, 33:1-22. Osborn,H.F. 1918.EquidaeoftheOligocene,MioceneandPlioceneofNorthAmerica,iconographic revision. Mem. American Mus. Nat. Hist., 2:328. Pichardo del Barrio, M. 1961. Prohoscideos fosiles de Mexico. Dept. Prehist., Inst. Nac. Antropol. Hist. Mexico, Invest., 4:1-38. Repenning, C. A. 1962. Ihe giant ground squirrel Paenemarmota. J. Mammal., 36:540­ 556. Schultz, C. 8., M. R. Schultz, and L. D. Martin. 1970. A new tribe of saber-toothed cats (Barbourofelini) from the Pliocene ofNorth America. Bull. Univ. Nebraska State Mus.,9:1-31. Schultz, G. E. 1977. Guidebook field conference on late Cenozoic biostratigraphy of the TexasPanhandleandadjacentOklahoma.WestTexasStateUniv.,Spec.Puhl., 1:1-160. Shotwell, J. A. 1958. Inter-community7 relationships in Hemphillian (mid-Pliocene)mammals. Ecology, 39:271-282. Silva-Barcenas, A. 1969. Localidades de vertehrados fosiles en la Republica Mexicana. Univ. Nac. Aut. Mexico, Inst. Geol., Paleontol. Mexicana, 28:1-34. Skinner, M. E., and C. W . Hibbard, et al. 1972. Early Pleistocene pre-glacial and glacialrocks andfaunasofnorthcentralNebraska.AmericanMus.Nat.Hist.Bull., 148:1-148. Stirton, R. M. 1955. Two new species of the equid genus Neohipparion from the middle Pliocene, Chihuahua, Mexico. J. Paleontol., 29:886-902. Tanner, L. G. 1967. A new species of rhinoceros, Aphelops kimballensis, from the latest Pliocene of Nebraska. Bull. Univ. Nebraska State Mus., 6:1-16. Tanner, L. G. 1975. Stratigraphical occurrences of Teleoceras, with a new Kimballian species from Nebraska. Univ. Nebraska State Mus. Bull., 10:23-33. Webb, S. D. 1965. The Pliocene Canidae of Florida. Florida State Mus. Bull., 14:273-308. Webb, S. D. 1973. Pliocene pronghorns from Florida. J. Mammal., 54:203-221.