Miocene Rhinoceroses from the Texas Gulf Coastal Plain

Home , Cormohipparion, Goliad Formation, Menoceras

JOURNALOF PALEONTOLOGY,V. 61, NO. 2, P. 388-423, 16 FIGS.,MARCH 1987

MIOCENE RHINOCEROSES FROM THE TEXAS GULF COASTAL PLAIN

DONALD R. PROTHEROAND EARL M. MANNING Departmentof Geology, OccidentalCollege, Los Angeles, California90041 and Museum of Geoscience, LouisianaState University, Baton Rouge 70803

ABSTRACT--Fourspecies of rhinocerosoccur togetherin the Barstovian(middle Miocene) faunas of southeast Texas, a unique situation in the Miocene of North America. Two are assigned to normal contemporaryHigh Plains species of Aphelopsand Teleoceras,and two to dwarf species of Peracerasand Teleoceras.The dwarfPeraceras is a new species, P. hessei.The dwarfTeleoceras is assignedto Leidy's (1865) species "Rhinoceros"meridianus, previously referredto Aphelops. "Aphelops"profectus is here reassignedto Peraceras. The late Arikareean(early Miocene) Derrick Farm rhino, erroneouslyreferred to "Caenopus premitis"by Wood and Wood (1937), is herereferred to Menocerasarikarense. Menoceras barbouri is reportedfrom the early Hemingfordian(early Miocene) Garvin Gully local fauna of southeast Texas. The rhinos from the early ClarendonianLapara Creek Fauna are tentativelyreferred to Teleocerascf. major. The three common genera of middle late Miocene rhinocerosesof North America (Aphelops, Peraceras, Teleoceras)are rediagnosed.Aphelops and Peracerasare more closely related to the EurasianAceratherium and Chilotherium(all four togetherforming the Aceratheriinae)than they are to the American Teleoceras.Contrary to Heissig (1973), Teleocerasis more closely relatedto the living rhinocerosesand their kin (togetherforming the Rhinocerotini)than it is to the Acer- atheriinae.

INTRODUCTION of Leidy's species. Most authors since Mat- AMONGthe first mammalian fossils described thew (e.g., Hesse, 1943; Quinn, 1955; Patton, from Texas was a fragmentary upper molar 1969) have referredthe largerrhinoceros ma- of a rhinoceros from the Texas Gulf Coastal terial from the middle Miocene of Texas to Plain. The specimen was given to J. Leidy by "Aphelops"meridianus, if generic-leveliden- B. F. Shumard of St. Louis, Missouri. Leidy tification was attempted at all. (1865) described this tooth, illustrated it Hesse (1943) was the first author to rec- (1869, P1. 23, fig. 10; see Figure 10.1-10.3), ognize a smallerrhinoceros in the Texas Gulf and gave it the name Rhinocerosmeridianus. Coast Barstovian faunas. He described and The only locality informationgiven by Leidy named a new species, Aphelops rileyi, in a was that "it was derived from a tertiary [sic] manuscriptthat was incomplete at his death deposit in Washington county [sic], Texas, in 1944, and never published. This smaller and presents much the general aspect of the rhinoceros material was referredto Dicera- Mauvaises Terres fossils of the White River, theriumby Quinn (1955) and Patton (1969), Dakota, with which it is probably cotempo- apparently due to identifications made by H. rary [sic] in age" (1869, p. 229). E. Wood (J. A. Wilson, personal commun.). In 1875 and 1877, Cope referred some Quinn (1955, p. 75) mentioned a small rhi- specimens from the Santa Fe Group of New noceros "no larger than Diceratherium, but Mexico to the species meridianus,but placed with characters of Teleoceras including the this species in his new genus Aphelops.Mat- reduced number of cheek teeth," suggesting thew (1932) pointed out that Cope's "meri- "a primitive but direct ancestor to the upper dianus" did not correspondwell to Leidy's Miocene and Pliocene species." Quinn gave type specimen,and was not from Texas. Mat- no catalogue number for this specimen. Based thew considered the name meridianusinde- on this brief and nondiagnostic description, terminate until adequate topotypes were it is unlikely that he actually recognized the known. Matthew (1932, fig. 2) also figureda dwarf Teleoceras, since there is no complete specimen (UCMP 31043) from near the type skull of it in the Texas Memorial Museum locality that he believed to be representative collections. All of the "Diceratherium"-sized

skulls (probably in reference to Menoceras of Natural History, New York, N.Y.; F:AM, arikarense,the Agate Springrhino) in the lat- Frick Collection, American Museum of Nat- ter collection are of the dwarf Peraceras.Al- uralHistory, New York, N.Y.; l.f., local fauna though many (but not all) advanced Teleoc- (sensu Tedford, 1970); MCZ, Museum of eras reduce the premolarsand lose P2/2, this Comparative Zoology, Harvard University, is not a diagnosticfeature of the genus. Thus, Cambridge, Mass.; N, sample size; PU, "reduced number of cheek teeth" is not Princeton University Museum, Princeton, diagnostic of Teleoceras.In short, the pub- N.J.; OR, observed range;s, standarddevia- lished literature is in great confusion con- tion; TAM, Texas A&M University, College cerning Texas Gulf Coast Miocene rhinoc- Station, Tex.; TMM, Texas Memorial Mu- eroses. seum, University of Texas at Austin;UCMP, In 1936, the FrickLaboratory of the Amer- University of CaliforniaMuseum of Paleon- ican Museum of Natural History began col- tology, Berkeley, Calif.; UNSM, University lecting in the Texas Gulf Coastal Plain. The of Nebraska State Museum, Lincoln, Nebr.; excavations were supervisedchiefly by N. Z. USNM, United States National Museum, Ward,and continued until 1964. Most of the Washington,D.C.; ., mean. large collection from this area remains un- For ease of word processing,the system of described. Descriptions of some of the ro- dental abbreviations of Jepsen (1966) is dents(Wahlert, 1976), horses (MacFadden and adopted here. Thus, upper premolars and Skinner, 1977, 1981; Skinner and MacFad- molars are indicated by upper case "P" or den, 1977), protoceratidartiodactyls (Frick, "M", lower premolars and molars by lower 1937; Patton and Taylor, 1971, 1973), lep- case "p" or "m". tomerycids (Taylor and Webb, 1976), and A note on endings:the -cerassuffix in rhino oreodonts (Schultz and Falkenbach, 1941) genera (such as Peraceras or Teleoceras) is have been published. While curatingthe rhi- neuter in gender;the suffix -ceros (as in Rhi- noceros material in the Frick Collection, the noceros)is masculine.Since adjectivalspecies junior authorfound that the BarstovianTex- names must agree with the genus in gender as Gulf Coastal Plain faunas contained four (International Code of Zoological Nomen- sympatricspecies of rhinoceros.Two of these clature, Article 11, section g), a number of appeared to be similar to the High Plains traditionally-accepted species names have Aphelops megalodus and Teleoceras medi- been corrected.Thus, Peraceras has species cornutum,but there was also a dwarf Teleoc- superciliosumand profectum;Teleoceras has eras and a dwarf aceratherine rhinoceros species medicornutum,proterum, and meri- present. The latter was first identified as a dianum. Similarly, Brachypotheriumameri- dwarfAphelops(Prothero and Sereno, 1980), canus Yatkola and Tanner, 1979, is here but comparisonwith more complete material amended to B. americanum, since the suffix in the Texas Memorial Museum collection -theriumis also neuter. has shown that the dwarf aceratherineis actually a dwarf Peraceras. The occurrenceof four differentrhinoceroses in the Texas Bar- LOCALITIES AND STRATIGRAPHY stovian faunas probably caused some of the The Miocene of the Texas Gulf Coastal taxonomic confusion outlined above. Plain has been collected by a number of in- The interestingmorphological and ecolog- stitutions over the years, but the major col- ical implications of this occurrenceare dis- lections of fossil vertebratesare at the Texas cussedelsewhere (Prothero and Sereno, 1982). MemorialMuseum of the University of Tex- In this paper,the systematics of the Miocene as at Austin, the Texas A&M University rhinoceroses of Texas are reviewed, and an (presently on loan to the Texas Memorial attempt to clear up some of the confusion Museum), and the Frick Collection of the regarding the early, primitive members of the American Museum of Natural History. Each genera Aphelops, Teleoceras, and Peraceras institution has a distinct set of localities, list- is made. ed in Table 1 and shown in Figure 1. The Abbreviations. -AMNH, Department of detailed locality information is given in Hesse Vertebrate Paleontology, American Museum (1943) and Patton and Taylor (1971, table 2). 390 DONALDR. PROTHEROAND EARL M. MANNING

TABLE 1-Areal distributionof fossil localities (see Figure 1).

Num- ber County Fauna Locality 1 Tyler Burkeville West of Doucette, near Woodville 1 Tyler Cold Spring TMM 40662, Belts Creek 2 Tyler Cold Spring TMM 30009, 31087, 8 mi southwestof Woodville 3 Polk Burkeville TMM 30157, Moscow l.f. 4 Polk Cold Spring Near Swartaut 5 Tyler Cold Spring TMM 40623, Push Creek 6 San Jacinto Cold Spring TMM 31219, Cold Springl.f. (=F:AM Donohoe Farm) 6 San Jacinto Cold Spring McMurrayPits 1 and 2 6 San Jacinto Burkeville Trinity River Pit 1 (=TMM 40196) 6 San Jacinto Burkeville StephenCreek 7 San Jacinto Cold Spring TMM 31191, San Houston l.f. 8 San Jacinto Burkeville TMM 31243, Point Blankl.f. 9 San Jacinto Burkeville TMM 31190, Point Blank l.f. 10 San Jacinto Burkeville TMM 31242, Point Blank l.f. 11 Walker Garvin Gully TMM 30873, Aiken Hill l.f. 12 Grimes Burkeville TMM 40290, near Navasota 13 Washington ?Burkeville Leidy'dlocality, fide Hesse, 1943 14 Washington Cold Spring TMM 31272, southeastof ChapellHill 15 Washington Garvin Gully TMM 40106, Farm Road 244 15 Washington Garvin Gully TMM 31048, Garvin Gully l.f. 15 Washington Garvin Gully TMM 40067, Hidalgo Bluff 16 Washington ?CatahoulaFm. DerrickFarm locality 17 Fayette Burkeville TMM 40127, north side of CedarCreek, La Grange 18 Fayette ?Burkeville TMM 31246, Swiss Alp 19 Lavaca ?Burkeville Near Hallettsville(exact locality unknown) 20 Grimes Cold Spring TAM locality 19, Noble Farm l.f. 21 Bee LaparaCreek TMM 30896, Berclairl.f. (=Farish Ranch l.f.) 22 Bee LaparaCreek TMM 31132, Normannal.f. (=BridgeRanch l.f.) 23 Bee LaparaCreek TMM 31081, Berclairl.f.

Further details are available from the appro- Discussion. -The rhinoceros left maxilla priate institution. with dP1,P2-4 from Derrick Farm, Wash- The mammalian stratigraphy of the Texas ington County, Texas, had previously been Gulf Coastal Plain has been extensively dis- assigned to the Oligocene genus Caenopus. cussed (Hesse, 1943; Quinn, 1955; Wilson, This determination was based on its small 1956; Patton, 1969; Patton and Taylor, 1971, size and supposed Oligocene provenance. 1973), and will not be reviewed here. The Apparently the age determination prevented correlations shown in Figure 2 are based on earlier authors from comparing the Derrick the latest studies of the faunas by a number Farm rhino (USNM 6573) with another small of workers (Tedford et al., 1987), incorpo- rhinoceros, Menoceras arikarense, from the rating much unpublished information from earliest Miocene of Agate Spring Quarry, Ne- the Frick Collection. braska. USNM 6573 is virtually identical in SYSTEMATICPALEONTOLOGY size and morphology with the Agate Spring sample of M. arikarense (Figure 3.6, Table Order PERISSODACTYLA 1848 Owen, 2). Wood and Wood (1937) referred the Der- 1937 Suborder CERATOMORPHAWood, rick Farm rhino to Caenopus cf. premitis, but RHINOCEROTOIDEA 1872 Superfamily Gill, this rhino is merely a variant of Trigonias RHINOCEROTIDAE 1845 Family Owen, osborni. The genus Caenopus is now consid- Genus MENOCERAs 1921 Troxell, ered to be a junior synonym of Subhyraco- MENOCERASARIKARENSE (Barbour, 1906) don, so the Derrick Farm rhino cannot be 3.6 Figure 3.1-3.3, referred to this genus (Lucas et al., 1981). Diceratherium arikarense BARBOUR,1906, p. 316. Morphologically, the Derrick Farm rhino is Diceratheriumcooki PETERSON, 1906, p. 282. a poor match for "Caenopus." It has much 1924 Coenopus[sic] sp. GIDLEY, (in Deussen, 1924, weaker lingual cingula and a more fully bi- p. 98, P1. 28). with the directed Caenopuscf. premitis Gregoryand Cook. WOOD lophodont P4, metaloph and unconnected to the AND WOOD, 1937, p. 134. posteriorly proto- Menoceras arikarense(Barbour). TANNER,1969, loph. p. 395. It is curious that this specimen was so badly MIOCENE RHINOCEROS FROM TEXAS 391

c FALL LEON HOUSTON ANGELINA SABINENEWTON TRINITY BELL ROBERTSONTNTY POL MADISON II :3 MILAM GRIMESWALKER WILLIAMSON~ ~ ii i BRAZOS SANSAN64 BURLESON JACIncNTOJACINTO MO GMRY -HARDIN TRAVIS LEE LIBERTY ORANGE BASTROPASI HARRIS JEFFERSON ASHAYS14O FYTE

coL ADO

BEXAR FORT MED GUADALUPE GONZALES 1111111 BENDAV GALVESTOO WILSON WHARTON BRAZORIA ZAVALAFRIO ATALSCOSA\A JACKSON KARNESVICTORIA ;? MATALGORDA\ EXPLANATION

1AD ASCALE i LABAHIAii FAUNAi LAPARA ii FAUNA OIMMITLAnSALLE McMULLENHcULE LIVELLCALHOUN O I MISSIONCREEK SREFUGIO COLDSPRINGFAUNA BURKEVILLE FAUNA WEBB c SANAPATRICIOARC ARARAAAS GARVINGULLEYFAUNA DUV~ MWELLTE17WELLS FOSSILLOCALITIESL B-ii~il~NUECES 0 to20 SCA\LE40 GQ so Miles

FIGURE1-Index map showingdistribution of fossil localities describedin Table 1. (AfterQuinn, 1955, fig. 5.) misidentified, since both Wood and Gidley Farm specimen is Oligocene in age. Deussen were certainly familiar with Menoceras ari- (1924, p. 98) originally reported the specimen karense. Specimens of M. arikarense are from the Oakville Formation and suggested found in most of the major collections in the that the rhino had affinities with middle or United States, and the rhino was first de- late Miocene forms. Weeks (1933, p. 456- scribed in 1906 and thoroughly mono- 457) and Wood and Wood (1937) were con- graphed by Peterson in 1920. Before 1969, vinced that the specimen came from the Ca- M. arikarensewas known as Diceratherium tahoula Formation, which was then consid- cooki, but Tanner (1969) showed that Trox- ered Oligocene in age. J. A. Wilson (personal ell's (1921) distinction of the common Agate commun.) has examined the locality and sug- Springs Quarry rhinoceros, which he named gests that the specimen could easily be de- Menoceras, from the John Day rhinoceros rived from a channel from the overlying Oak- Diceratherium was correct. Tanner (1969) also ville Formation (Figure 2). If it is truly from pointed out that the species arikarense (Bar- the Catahoula Formation, the Oligocene age bour, 15 June 1906) has two month's priority of this unit is not well established, either. The over Troxell's type species of Menoceras, Di- Catahoula Formation is composed of sub- ceratherium cooki (Peterson, 31 August 1906). aerial braided stream deposits and contains Perhaps the confusion of Wood and Gidley no diagnostic marine fossils, only some plant may be due to the supposed Oligocene age of fossils (Deussen, 1924, p. 95; McBride et al., USNM 6573. Recent workers have found that 1968). The Oligocene age was suggested pri- most of the Arikareean is also of Oligocene marily by its position between the Miocene age (Tedford et al., 1987; Prothero and Rens- Oakville Formation and the Eocene Fayette berger, 1985). The age of the Agate Spring Formation. McBride et al. (1968, p. 10) re- fauna is difficult to determine, but it is prob- port a lead-alpha age on zircons from the ably Aquitanian, or earliest Miocene (Hunt, lower part of the correlative Gueydan For- 1972). mation in Live Oak County, Texas. This date Yet, it is not even certain that the Derrick of 24 ? 1 million years would place the Ca- 392 DONALD R. PROTHERO AND EARL M. MANNING

Ma EPOCH NPZ NALMA HIGH PLAINS TEXAS GULF COASTAL PLAIN

940 O 0o 10 , LL C > Farish Ranch I.f. I MinnechaduzaF E- .<

Ranch I.f. Burge F. iBridge (_ 12 I.f. NIc13Goodrich 12 ISam Houston I.f.

W 0o Cold Spring I.f. jw N9 z N8 > O Trinity Lower Snake River I.f.Moscowf. 15 < Cree k F. a) R I W W Point Blank I.f.

LL 17

18 N6 0 0 N6 Garvin L

201 I M rsland F. Cedar Run Aiken Hill .f 21 z te Springs Q Lu 0 z RAga I WaterRunlocality o u: N4 4 22 - L 0.. Marlon F CearRunI(~ikn illI0 <0 r N3 _ 23 Watr

WLu <

FIGURE2-Stratigraphy of Miocene localities from the Texas Gulf Coastal Plain (after Tedford et al., 1987). Abbreviations: Ma, million years before present; NPZ, Neogene planktonic zones; NALMA, North American land mammal ages; F, Fauna; l.f., local fauna (sensu Tedford, 1970). MIOCENE RHINOCEROS FROM TEXAS 393

! rrg •. - ._

; 1??1 _.o~* '; .:•, f

- t

1 L~2 -

4 5 6

ift

" - " 8 ' 9 i• .A- /I .1 . ,, 7.. ' ... * -•" .-.4... '"i" .. . ," '. ?.., 7 89

FIGURE3--Menoceras from the Texas Gulf Coastal Plain. 1, 2, Menocerasarikarense, USNM 6573, stereopairsof crown view; 3, same specimen, lateralview. 4, 5, Menocerasbarbouri, TMM 31048- 48, stereopairof left dP4. 6, Comparison of M. arikarense;left specimen is AMNH 86118 from Agate SpringQuarry; right specimen is USNM 6573 from Derrick Farm. 7-9, M. barbouri,TMM 41662-1, left ramal fragmentin medial, crown, and lateral views. Scales in cm. 394 DONALD R. PROTHEROAND EARL M. MANNING

TABLE2-Measurements (in mm) of Menocerasarika- to be identified. The most diagnostic speci- rense. men is TMM 41662-1, a partial left ramus Derrick with m3 and part of m2 (Figure 3.7-3.9). In Farm size and morphology it most closely ap- rhino M. referred to "Dicera- USNM AgateSpring arikarensel proaches specimens 6573 N x s OR therium"barbouri (Table 3). It is clearly too P2-4 63.0 7 64.7 3.0 59.6-68.5 largeto be referredto Menocerasarikarense, P2 length 19.1 7 21.5 2.0 19.0-25.3 and too small to be referredto Menoceras P2 width 27.5 7 26.8 1.0 25.3-28.6 marslandense. P3 length 22.1 7 22.1 1.7 20.1-23.8 P3 width 33.0 7 33.6 2.7 31.1-37.4 Wood (1964) placed Diceratherium(Me- P4 length 20.5 7 24.1 1.3 22.4-26.0 noceras) barbouriin the subgenus Menocer- P4 width 35.0 7 36.4 2.3 33.3-39.2 as. Tanner (1969) showed that Menocerasis 'Agate Spring sample based on the following speci- a valid genus, although he did not mention mens: AMNH 86224, 86227, 86116, 86118, 86112, Menocerasbarbouri. Diceratherium 14213, 26892. (Menoc- eras) barbouriis here placed in the genus tahoula Formation in the middle late Ari- Menoceras, and removed from Dicerathe- rium. from the kareean, very near the Oligocene-Miocene Specimens Runningwater the of the Ca- Formation ("B" Quarry, Bridgeport Quar- boundary. Thus, upper part were referredto Menoceras tahoula Formation is probably late Arika- ries) falkenbachi reean with the by Tanner(1972), who never mentioned Me- (early Miocene), agreeing age noceras barbouriin his of M. suggestedby the Derrick Farm rhino. diagnosis falkenbachi,and possibly they cannot be distin- MENOCERAS BARBOURI guished. They are nearlyidentical in size and (Wood, 1964) n. comb. morphology, so far as can be determined. Figure 3.4, 3.5, 3.7-3.9 Menocerasfalkenbachi Tanner, 1972, is not with M. barbouri Diceratherium(Menoceras) barbouri WOOD, 1964, synonymized (Wood, 1964) p. 378. here. Further comparisons, that are beyond Diceratheriumsp. QUINN,1955, p. 72. the scope of the present paper, are necessary Diceratheriumsp. PATTON,1969, p. 211. to establish this synonymy. the status of M. M. Referredmaterial. -TMM 41662-1,jaw fragmentwith Whatever falkenbachi, m2-3, fragmentsof other teeth; TMM 31048-48, left barbouriis certainly a valid taxon. It is re- dP4 (unworn);TMM 40067-124, immatureright femur; stricted to the early Hemingfordian (Run- TMM 40106-4, lumbar vertebra; TMM 40067-190, ningwaterFormation in Nebraska, Thomas fragmentaryskull with poor dentition;TMM 40067-71, tooth fragment;TMM 40067-178, brokenmetacarpals; Farm faunain Florida),and its occurrencein TMM 40067-189, tooth fragments;TMM 31048-54, the Garvin Gully faunais strongevidence for edentulouspartial right lower jaw; TMM 31048-3, right an early Hemingfordianage of TMM locality lower molar (?m2);TMM 31048-39, tooth fragment. 41662. Discussion.-Most of the Garvin Gully The only other diagnostic rhino specimen rhinoceros material is too poorly preserved from the Garvin Gully faunais TMM 31048-

TABLE3--Comparison of early HemingfordianMenoceras. m3 length m3 width Specimen N X s N X s Garvin Gully rhino TMM 41662-1 1 37.2 - 1 21.5 M. barbouri MCZ 7445 (type) 1 38.6 - 1 23.0 M. falkenbachi UNSM 1241 (type) 1 39.9 - 1 22.5 M. marslandense Runningwater Fm. "B" Quarry 2 40.4 5.5 2 22.6 4.1 M. arikarense Agate Spring Quarry 14 34.2 1.7 14 19.5 4.0 (OR = 31.0-37.2) (OR = 17.9-29.0) MIOCENE RHINOCEROS FROM TEXAS 395

48, an isolated right dP4 (Figure3.4, 3.5). Its left M3. Moscow 1.f.,Polk Co., Texas:TMM 31057-160, lophs are barely worn, which is unusual for fragmentaryleft upper deciduous premolar; TMM 31057- a since it is one of the first deciduous 43, fragmentarym2; TMM 31057-11, tusk fragments; dP4, TMM 31057-83,left m2; TMM 31057-35,carpals; TMM teeth to erupt.The strongcrochet and the size 31057-28, phalanx. Point Blank 1.f., San Jacinto Co., of the specimen are comparableto deciduous Texas: TMM 31190-29, tibia; TMM 31190-30, femur. teeth of M. "falkenbachi"from Bridgeport Locality 31242, San Jacinto Co., Texas: TMM 31242- found in the Harold Cook Collec- 30, femur. Trinity River Pit 1, San Jacinto Co., Texas: Quarries F:AM 108311, right M1-2; F:AM 108312, right P3; F: tion now in the collections of the American AM 108907, left m2; F:AM 108314, radius; F:AM Museum of Natural History. TMM 31048- 108315, radius; F:AM 108316, proximal ulna; F:AM 48 seems to corroboratethe Runningwater 108318, metacarpalV; F:AM 108317, left navicular;F: affinitiesof the Garvin fauna. AM 108298, proximalleft metatarsalII; F:AM 108946, Gully proximal left metacarpalII; F:AM 108945, proximal ulna. ACERATHERIINAEDollo, 1885 juvenile Subfamily Cold SpringFauna (late Barstovian):Cold Spring1.f., Genus PERACERASCope, 1880 San Jacinto Co., Texas: TMM 31219-228, skull (type); Type species. -Peraceras superciliosum TMM 31219-227, skull;TMM 31219-229, skull;TMM 31219-225, jaw; TMM 31219-220, jaw; TMM 31219- Cope, 1880. TMM left TMM - 231, left femur; 31219-247, femur; Revised diagnosis. Peraceras is character- 31219-244, left pelvis; TMM 31219-278, left ramus; ized by a numberof derived features,includ- TMM 31219-45, left ramus;TMM 31219-123, left i2; ing a brachycephalicskull with procumbent TMM 31219-226, rightm3; TMM 31219-242, rightp2; lambdoid crest and occiput, shortened na- TMM 31219-1, magnum;TMM 31219-48, left femur; flat dorsal skull an TMM 31219-46, radius;TMM 31219-238, rightfemur; sals, profile, upturned TMM 31219-216, right i2; TMM 31219-193, isolated symphysis in females, a short diastema on lower molars;TMM 31219-49, tibia;TMM 31219-209, the lowerjaw, and lingual cingulaon most of atlas; TMM 31219-50, ulna. A. D. Donohoe Pit, San the lower teeth. It possesses the featuresthat Jacinto Co., Texas: F:AM 108340, right and left rami. characterize the Four miles southeastof ChapellHill, WashingtonCo., Aceratheriinae, including Texas:TMM 31272-24, calcaneum.Near Swartaut,Polk greatlyreduced premaxilla and the loss of I1, Co., Texas:F:AM 108264, rightramus with i2, p2-m3. nasal incision retracted to the level above BeltsCreek, Tyler Co., Texas:TMM 40622-1, rightmax- anterior P4, brachydontteeth with weak an- illa with dP1-2. McMurrayPits 1 and 2, San Jacinto tecrochets, and primitive rhinocerotid pro- Co., Texas:F:AM 108268, left ramuswith ml-3; F:AM More derived of Peraceras 108267, rightramus with p2, m2-3; F:AM 108274, left portions. species tibia;F:AM 108307, skull with left and rightP4-M3; F: develop broad zygomatic arches, flaring AM 108337, 2 metatarsalsIII; F:AM 108336, 2 meta- lamboid crests which secondarily create a tarsals II; F:AM 108335, 2 ectocuneiforms; F:AM concave dorsal skull profile, massive broad 108334, left navicular; F:AM 108333, patella; F:AM postglenoid processes, and size. 108332, astragalus;F:AM 108331, distal tibia; F:AM very large 108330, distal right femur; F:AM 108329, distal left femur; F:AM 108328, left F:AM PERACERAS HESSEI n. sp. tibia; 108327, right femur;F:AM 108326, 2 unciforms;F:AM 108322, prox- Figures 4-8 imalradius; F:AM 108319,right humerus; F:AM 108325, Aphelopsn. sp. (small form) HESSE,1943, p. 171. 2 magnums;F:AM 108324, 2 lunars;F:AM 108323, 5 Diceratheriumsp. QUINN,1955, p. 72-75. scaphoids;F:AM 108321, radius;F:AM 108320, distal Diceratheriumsp. PATTON,1969, p. 129. right humerus;F:AM 108310, damaged left M1-3; F: AM 108309, left M ; F:AM 108308, mandible;F:AM Holotype.--TMM 31219-228, skull from 108944, juvenile mandible with dp2-4, ml erupting. Coldspring,Texas. Dentition worn almost to Noble Farm, five miles east of Navasota, Grimes Co., the alveoli Texas:TAM 2191, rightM3, P2, m2; TAM 2674, sym- (Figure4.1-4.4). physiswith left i2; TAM 2591, palatewith rightPl-M2, Distribution.-Burkeville and Cold Spring left M2-3. faunas (early and late Barstovian, middle Fleming Formation,level unknown:Hallettsville lo- Miocene), Texas; late Barstovian of New cality,Lavaca Co., Texas:TAM 2202, skullwith left and Mexico. right P4, M1-3. Etymology.--In honor of Curtis J. Hesse, Diagnosis.--Primitive, small-sized Perac- who first recognized and described the dwarf eras with relatively robust limb elements, re- rhinoceroses. duced occiput, and lambdoid crest with slight lateral flare. Referredmaterial. - Burkeville Fauna (early Barstovi- an): Swiss Alp locality (TMM 31246), FayetteCo., Tex- Description. -The type specimen of Per- as: TMM 31246-1, badlydamaged left maxillawith P3- aceras hessei, TMM 31219-228, is a skull M2. Woodvillelocality, Tyler Co., Texas:F:AM 108313, that has been heavily restored with plaster 396 DONALD R. PROTHERO AND EARL M. MANNING

2 UIIiLLLLLUlWIII

.. i /" i

4 • " . ?

6r .15 - •~-C16 -,~ .l* •:---•"-r•. . •-" .i .• - ,

FIGURE4-Peraceras hessei n. sp. 1-4, TMM 31219-228, holotype skull, in posterior, right lateral, palatal, and dorsal views; 5, F:AM 108311, right dP3-4; 6, TAM 2591, right dPl-M2; 7, TMM 31219-227, left P2-M3. Scales in cm. MIOCENE RHINOCEROS FROM TEXAS 397

.;t;:wv) V i'? i,.•

-. ?, .' •.. ,•.

r---4

• 'C? .•: -• '" -

/ c~c r lI5i :?I 3 1~•

,8 lllll|?* - " ,

FIGURE 5-Peraceras hessei n. sp. 1-3, TMM 31219-225, mandible, in right lateralview, crown view, and close-up of right lateral view of teeth; 4-7, F:AM 109360, skull in dorsal, left lateral, palatal, and posteriorviews; 8, UCMP 31043, right maxilla with P3-M3, crown view. Scales in cm. 00oo

TABLE 4-Skull and upper tooth measurements (in mm).

oo t 2. CEZ Q) 00 ON c) isC) - l C4 0 q)0kOi k< ?cj00k~ q6) q) c q) z 14T- q6P. 4", 4", q6)Z pQQ p 0 Q) C14 ON 00000 q6)oo2 q6)00Z cn O q6)8 C) ,00 . o C~O '~t tV

Foramen magnum to anterior P2 360.0 361.0 - - - 371.0 431.0 427.0 - 498.0 524.0 Occipital crest to Q) tip of nasals 348.0 - - 379.0 - 375.0 413.0 457.0 - 520.0 - 457.0 Zygomatic width 255.0 214.0 220.0 236.0 - 240.0 251.0 299.0 323.0 329.0 - - Palatal width between M3 49.9 62.1 65.2 65.4 - 52.3 55.0 64.2 82.0 62.0 79.5 64.2 Lambdoid crest width 133.0 107.0 150.0 150.0 - 142.0 163.0 186.0 179.0 252.0 214.0 - P2-M3 length 166.0 168.0 - - - 196.0 227.0 225.0 - 253.0 273.0 - M1-M3 length 92.0 90.0 96.0 98.0 - 104.0 126.0 129.0 138.0 153.0 162.0 116.0 P2 length 23.5 24.5 - - 23.6 21.5 24.0 27.7 - 28.0 34.6 - P2 width 28.2 32.0 - - 27.4 32.2 43.1 36.7 - 35.5 39.0 - P3 length 25.0 27.3 26.5 - 27.8 27.6 33.2 32.8 - 32.6 38.3 32.2 P3 width 31.0 37.8 37.0 - 36.4 44.6 56.9 46.1 - 49.5 53.5 43.6 P4 length 27.6 28.5 29.6 - 31.7 30.4 37.5 35.7 - 40.5 41.5 32.7 P4 width 41.5 37.8 40.8 - 39.5 49.0 60.4 49.4 - 60.0 62.7 50.3 Ml length 26.8 30.0 31.5 30.0 34.5 30.9 37.6 42.8 46.4 44.8 52.5 36.0 M1 width 40.4 35.6 39.7 43.8 37.5 49.6 58.0 57.8 61.2 67.5 73.8 58.7 M2 length 30.0 34.3 33.7 34.7 38.2 35.3 43.2 45.0 49.6 57.4 58.2 40.4 M2 width 41.0 37.6 39.6 41.5 37.5 51.0 61.0 59.0 68.3 68.5 73.1 52.0 M3 length 36.7 31.0 32.0 32.5 34.7 37.4 45.8 41.3 47.3 54.3 49.6 39.7 M3 width 37.0 35.4 34.0 37.4 35.8 46.4 59.0 51.0 51.0 66.8 59.5 47.4 MIOCENE RHINOCEROS FROM TEXAS 399

(Figure 4.1-4.4). Enough bone is present, mensions are largerthan is typical for P. hes- however, to ascertainthat the original shape sei (Table 4), but this is probablydue to ex- has been maintained. The skull of P. hessei treme wear. Lateral expansive deformation has shortened, broad nasals characteristicof of the basal diameter of the teeth is common Peraceras,with a relativelyflat dorsalprofile. in very old individuals. Tooth row lengths, The lateralmargins of the nasals curve under which are less susceptibleto deformationwith abruptly. The nasals are not fused. There is age, and all skulldimensions ofF:AM 109360 no sign of a nasal horn, althoughthe anterior are comparableto other specimens of P. hes- tip has been broken off. There are pro- sei. The largecollection of FrickNew Mexico nounced rugosities over the anterior part of Peraceras was compared to the Texas ma- the orbit. The flat dorsal surfaceof the fron- terial,but could not be includedin this study. tals meets the parietal-occipitalregion with The dentition of the type skull, TMM a sharpinflection. The lambdoid crests show 31219-228, is worn almost to the alveoli and little lateral flaring.There is a single narrow is missing P3 on both sides. The dentitions sagittalcrest. The occiputis procumbent.The of TMM 31219-227 (Figure 4.7) and TAM zygomaticarches are deep and widely flaring, 2591 (Figure 4.6) are much less worn and but this may be an artifactof restoration.The serve as the basis for this description. The premaxillaeand the anteriorpart of the max- upper teeth are brachydont,with unreduced illae are missing from the type and other skull premolars.A dP1 is present on TAM 2591. referred to P. hessei. The nasal incision is It has a very weak protoloph and a strong retractedto the level of anteriorP4. There is metaloph, with a closed medifossette. The a small infraorbitalforamen anterior to the upper premolarshave strong continuous an- orbit at the posteriorend of the nasal incision. terior,lingual, and posteriorcingula; the mo- The basicranium of TMM 31219-228 is lars have no cingula. P2 shows a strong pro- mostly plaster restoration. The confluent tocone-hypocone connection, but all of the mastoid-paroccipitalprocesses have been re- other premolarsare molariform.The prefos- stored incorrectly. Comparison with less sette formed on P2 is elliptical in shape. On damaged skulls, such as TMM 31219-227, the relativelyunworn P2 of TAM 2591, there shows that the mastoid process is quite mas- is a small medifossette due to the contact of sive. A small but distinct parocciptalprocess the cristaand crochet.In the other premolars, is present (contraryto the restorationof the however, the crochets are usually uncon- type). There is a shallow anteriorly-facing nected with the cristae. This feature is vari- pocket between the mastoid and paroccipital able, however; in TMM 31219-227, the left processes. The postglenoid process is also P4 has a medifossette, but the right lacks it. massive, with a distinct ventral knob fre- The postfossettes in the premolars are gen- quently found in Peraceras.The postglenoid erally circular or elliptical, with their long process and the mastoid process are not con- axes oriented parallel to the long axis of the fluent.The basicranialforamina and petrosal skull. In P4 and the molars, the postfossettes region are poorly preserved,although the hy- become more triangularin shape, opening poglossal foramen, posterior lacerate fora- posteriorly. men, medial lacerateforamen, foramen ovale, The molarshave strongparallel-sided lophs and posterior opening of the alisphenoid ca- with weak antecrochets.Crochets are mod- nal can be distinguishedon at least one skull. erately well-developed on most molars, but All of the skulls are too badly damaged to cristae are usually absent. The molars have identifyorbital foramina. The pterygoidflange strong anterior cingula, but the lingual cin- shows a stronglateral flare in F:AM 108307. gula are weak and do not connect the pro- The palateis badly damagedin all specimens. toloph and metaloph. Posterior cingula are An additional skull referableto P. hessei usually well-developed, passing labially to (F:AM 109360, Figure 5.4-5.7) is the best enclose triangular postfossettes. M3 is tri- preserved of any skull referable to this species. angular, with a weak crochet and no antecro- Although it is from the Barstovian of New chet. The anterior cingulum on M3 is strong, Mexico, it is discussed in this paper because but the posterior cingulum is absent except it shows the morphology of P. hessei prior to for a small spur (the remnant of the posterior crushing and restoration. The basal tooth di- metacone). 400 DONALD R. PROTHERO AND EARL M. MANNING

TABLE5--Lower jaw and teeth measurements(in mm).

A. megal- P. hessei' odus T. meridianum2 T. medicornutum3 N ? s (type) N f s N 2 s Mandiblelength 5 379.0 21 461.0 2 443.0 20 1 489.0 - Top of condyle to angularprocess 6 159.0 20 208.0 2 200.0 20 1 247.0 - Width symphysis at tusks 4 68.0 15 87.0 2 85.0 1 1 85.0 - Lengthi3-p2 diastema 4 29.0 12 57.0 2 48.0 11 1 54.0 - Depth of jaw below ml 7 69.0 5 74.0 2 69.0 17 2 90.0 4 p2-m3 length 5 183.0 13 217.0 0 - - 0 - - i3 length 1 92.2 - 80.5 0 - - 1 68.8 - i3 basal diameter 2 26.9 2 32.5 0 - - 1 22.5 - p2 length 3 23.3 2 27.7 0 - - 0 - - p2 width 3 16.8 0.4 36.7 0 - - 0 - - p3 length 6 28.9 3 32.8 2 24.0 2 2 33.5 4 p3 width 6 22.1 2 46.1 2 21.6 4 2 26.5 1 p4 length 6 29.7 2 35.7 2 31.0 1 2 42.5 4 p4 width 6 24.1 6 49.4 2 28.0 6 2 31.1 1 ml length 6 28.5 3 37.0 1 38.8 - 2 49.3 1 ml width 6 25.6 7 29.0 2 31.3 2 2 35.8 1 m2 length 7 31.0 5 42.0 2 44.1 3 2 54.6 0.1 m2 width 7 25.8 7 27.7 2 29.4 3 2 36.6 0.6 m3 length 7 31.8 6 44.6 2 49.7 0.6 2 60.1 4 m3 width 7 26.0 8 27.2 2 29.8 3 2 35.6 4 ' Sample of Peracerashessei includes:TMM 31219-225, TMM 31219-220, TMM 31219-278, F:AM 108308, F: AM 108257, F:AM 108346, F:AM 108264. 2 Sample of Teleocerasmeridianum includes: F:AM 108839, F:AM 108338. 3 Sample of Teleocerasmedicornutum includes: AMNH 9832 (holotype),F:AM 108349.

The best preserved mandible is TMM dpl, and p2 overhangs the diastema. The 31219-225 (Figure 5.1), which is complete lower premolarsand molars all have strong except for a broken left i2 and p2. The sym- labial cingula. Each cingulum extends an- physis has a strong upward inflection and terodorsallyto the surface of the preceding short diastema that is characteristicof Per- tooth. Small cingula are present also on the aceras. Although the type specimen of the lingual side of the tooth, but are usually dis- genus, P. superciliosum, lacks a mandible, tinct only along the base of the protolophid. several workers (Stock and Furlong, 1926; Postcranial skeleton.-Most of the post- Dalquest and Hughes, 1966) have correctly cranial material of P. hessei comes from surmisedthat jaws in their collections (Clar- McMurrayPit Number 2 and has undergone endonian forms, probablyfemales, referable much crushingand plasticdeformation. Thus, to P. superciliosum) with upturned sym- the comparisonsgiven below must be inter- physes and short diastematamight belong to pretedwith caution,since it is likelythat many Peraceras.Associated skull and mandiblesof of the differences observed may be due to P. profectum in the Frick Collection have postmortemeffects. The postcranialelements shown this to be true. The righttusk in TMM describedhere are shown in Figures6-8 and 31219-225 is largeand pointed with a strong Table 5. No vertebrae,scapulae, or ribs are medial ridge.It flaresonly slightlyaway from yet known for P. hessei. the midline. Its elongate shape is character- The humerus (Figure 6.1, 6.2) shows rel- istic of a male tusk. F:AM 108267 has an i2 atively little rugosityon the lateralepicondyle alveolus of the appropriate small size and and medial border of the olecranon fossa. stout shape for a female. The posterior part The radius (Figure6.5) is quite small; on the of the jaw shows the normal aceratherinean- proximal end the bicipital rugosity and the gle, condyle, and coronoid process. ulnar facets are very weak. The distal lateral The lower teeth are brachydont with the facets for the ulna (Figure 6.3, 6.4) are also stereotypedrhinocerotid pattern. There is no weakly developed. The facets for the carpals MIOCENE RHINOCEROS FROM TEXAS 401

- rv

J y / ."t?

?) ]

~ ~ , '1. . ' ? i•.?•. , .. ? I r; / i•:•3"• ' "*"; 1 . $ ' ?4.. 21 1P 345 i :1 " g'"•

6 7 10 1

FIGURE6-Peraceras hessei n. sp. 1, 2, F:AM 108319, humerus; 3, 4, TMM 31219-50, ulna; 5, F:AM 108321, radius; 6, 7, F:AM 108327, femur; 8, 9, F:AM 108333, patella; 10, 11, F:AM 108274, tibia. Scales in cm. 402 DONALD R. PROTHERO AND EARL M. MANNING

y 2s

A ""

12 13 15

FIGURE 7-Peraceras hessei n. sp. Manus elements. 1-3, F:AM 108323a, right scaphoid, in lateral, proximal,and distal views; 4-7, F:AM 108324a, left lunar,in dorsal,proximal, distal, and left lateral views; 8-10, F:AM 108325a, right magnum, in distal, dorsal, and left lateral views; 12-14, F:AM 108326b, right unciform, in dorsal, proximal, and left lateral views; 11, 15, F:AM 108318, right metacarpalV in lateral and posteriorviews. Scale in cm. are very deep. The ulna has a short but slen- sigmoid cavity is broadly arched, with a dis- der shaft that is reduced relative to the size tinct facet for the medial condyle of the hu- of the proximal end. The lateral articulation merus. for the humerus and radius bears only a faint The manus (Figure 7) of P. hessei is known ridge separating the two bones. The greater only from the scaphoid, lunar, magnum, un- MIOCENE RHINOCEROS FROM TEXAS 403 ciform, and metacarpals III and V. The a small aceratherinerhinoceros in the collec- scaphoid (Figure7.1-7.3) has the basic prim- tions from the Texas Gulf Coastal Plain. In itive aceratherine shape, except that the his unpublished manuscript (kindly fur- radial facet is smaller and less convex. The nished by J. A. Wilson), he wrestledwith the anteriordorsal lunar facet is small and poorly problem of the similarity of primitive Bar- developed. There is no posteriorlunar facet. stovian rhinoceroses. The confusion among The lunar (Figure7.4-7.7) is gracile, and the Aphelops, Teleoceras,and Peraceras in the ulnar facet is separatedfrom the body of the literature was compounded by taxonomic lunar by a constriction. The lunar bears a concepts that were based on more derived sulcus separatingthe ulnarfacet from the un- forms. The Frick Collection has greatly im- ciform facet. The magnum (Figure7.8-7.10) proved the sampleof late Hemingfordianand is gracile, with a large unciform facet and a Barstovianrhinoceroses and will make it pos- sulcus between the facets for metacarpal II sible to clearup some of taxonomic confusion and the scaphoid. The metacarpal III facet that presently exists. of the magnumis also reduced.The unciform Primitively,Aphelops and Peracerasare so (Figure 7.12-7.14) is gracile, with a less ro- similar in all but certain skull features that bust posteriorprocess. MetacarpalII is quite we were also confused about the affinitiesof small, and the facet for metacarpal III and the dwarfaceratherine (Prothero and Sereno, the posteriorrugosity are less well developed. 1980). Uncrushed skulls with complete den- A tiny, reduced metacarpalV (Figure 7.11, titions from the Texas Memorial Museum 7.15) that could only belong to P. hessei is collections clearly show that this dwarf is a also present in the collection. Contrary to primitive Peraceras. The reasons for this Wood (1964), both Peracerasand Aphelops confusion are discussed above. Many of the had phalanx-bearingfifth metacarpals,now charactersthat presentlydefine Peracerasare known from large quarry samples of both true only of the later, more derived species rhinos. such as P. superciliosum.The present com- Besides the flattening due to postmortem parisons are made chiefly with undescribed crushing, the femur (Figure 6.6, 6.7) has a Peraceras material from the Barstovian of broadly flared greater trochanterwith a tri- New Mexico, Colorado,and Nebraska,which angularproximal surface. The anterior face is assignedto P. profectum.The diagnosis of of the greatertrochanter bears a strongcrest. the genus Peraceras has been revised to re- The third trochanteris rectangularin shape, flect this. Comparisons of Aphelops, Perac- with sharp proximal and distal borders. The eras, and Teleocerasbased on Frick Collec- tibia (Figure 6.10, 6.11) has a slender shaft tion specimens are given in Tables 4-6. and a narrow,laterally-truncated fibular fac- Peracerashessei is clearly a dwarf species, et. No fibula is known for P. hessei. The pa- ratherthan a primitively small, earliermem- tella (Figure 6.8, 6.9) is relatively thin and ber of the genus, because it is the smallest gracile and lacks a ridge on the distal tuber- aceratherinerhinoceros in North America. osity. Its contemporary sister-species, P. profec- The pes of P. hessei (Figure 8) is known tum, is 20 percent larger. Most other acer- from the astragalus, navicular, ectocunei- atherines, including Aphelops,Aceratherium form, and metatarsalsII and III. The astrag- s.s., and Chilotherium,are even largerthan alus (Figure 8.1-8.3) has a lateral calcaneal P. profectum.Indeed, P. hessei is the smallest facetwhich is broaderand less concave. There post-Hemingfordianrhinoceros of any kind is a distinct fossa lateral to the sustentacular that is known. The small size is not due to facet. The navicular(Figure 8.4-8.6) has dis- sexual dimorphismfor severalreasons: 1) extinct distal facets. The ectocuneiform(Figure treme sexual dimorphism in size is almost 8.7-8.9) is quite small and has sufferedthe unknownamong rhinocerotids (Laurie, 1982); effects of proximodistal crushing. Metatarsal 2) both male- and female-shaped lower in- III (Figure 8.10-8.12) is gracile, and the pha- cisors are present in the collection; 3) there langeal articulation extends further antero- are no P. profectum in the Texas faunas. proximally. Metatarsal II is also very gracile There is, however, an intermediate-sized and bears a reduced facet for metatarsal III. aceratherine, UCMP 31043 (Figure 5.8), from Discussion. -Hesse (1943) first recognized J. Fietsam's Farm, east ofAmmansville, Fay- 404 DONALD R. PROTHERO AND EARL M. MANNING

(i 0,

"i.

": '3

-. 11. ? ""~~~' "" " . .."k ..

i, -tC 1011 .I.I. t 12'(

FIGURE8-Peraceras hessei n. sp. Pes elements. 1-3, F:AM 108332, left astragalus, in dorsal, plantar, and distal views; 4-6, F:AM 108334, left navicular, in dorsal, proximal, and distal views; 7-9, F: MIOCENE RHINOCEROS FROM TEXAS 405 ette County, Texas. This locality (Figure 1) would appear to be part of the Cold Spring Fauna, based on its stratigraphicposition. Matthew (1932, p. 420) referredthis specimen (possibly on the basis of location alone) to Aphelops meridianus, but, as discussed below, Leidy's type of meridianus is not an Aphelops at all, but a Teleoceras. UCMP 1 31043 consists of a right maxilla with P3- rrrrmr M3. It is brachydontwith weak antecrochets. Both features suggest a primitive aceratherine. Since primitive Aphelops and Peraceras cannot be distinguishedon upper dentitions alone, the specimencould be referredto either genus. In size (Figure5.8, Table 4) it is much smaller than A. megalodus or P. profectum, and larger than P. hessei. In tooth area it matches the New Mexico specimen of P. hessei (F:AM 109360). But, as pointed out above, 2 this individual has unusuallylarge tooth area for P. hesseibecause of its extremetooth wear. A species assignmentfor UCMP 31043 is not made until a thorough review of the New Mexico material is done. PERACERASPROFECTUM (Matthew, 1899) n. comb. Figure 9 Aceratheriumprofectum MATTHEw, 1899, p. 71. Aphelopsprofectus (Matthew). MATTHEw, 1901, p. 358. Aphelops profectus (Matthew). OSBORN,1904, p. 317. Peraceras n. sp. TANNER,1976, p. 50. Diceratherium jamberi TANNER,1977, p. 121. FIGURE9-Peraceras profectum. 1, 2, F:AM 108338, referredskull, rightlateral and palatal Holotype. -AMNH 9082, a lower jaw with views; 3, F:AM 108338, right ramus,and m3, fragments of upper teeth, right squa- AMNH 9082, holotype ramus. Scale in cm. mosal region of the skull, and atlas. Discussion. -The type of "Aceratherium" profectum, AMNH 9082, consists of frag- (Figure 9.3) with a skull and mandible (F: mentary right and left rami, with both left AM 108338, Figure 9.1, 9.2) from the Bar- and rightm3. Otherfragments include a skull stovian of New Mexico (specifically,from the fragmentconsisting of the right postglenoid Pojoaque Member of the Tesuque Forma- process, posterior portion of the right zygo- tion). Both F:AM 108338 and AMNH 9082 matic arch, and associated occipital and par- show the characteristic features of Peraceras: occipitalbones, and additionalskull and tooth short, upward-inflected symphysis and short fragments.All are from the Barstovian Paw- diastema, and lingual cingula on the lower nee Creek Formation, Logan County, Colo- molars. The skull shows additionalPeraceras rado. The type rami compare extremely well features, such as the procumbent and laterally

AM 108335a,right ectocuneiform, in dorsal,proximal, and distalviews; 10-12, F:AM 108337a, right metatarsalIII, and F:AM 108336b, right metatarsalII, in dorsal, plantar,and proximalviews. Scale in cm. p o\

TABLE6-Measurements of postcranial skeleton (in mm).

Peracerashessei Aphelopsmegalodus Teleocerasmeridianum Teleocerasmedicornutum N f s N f s N X s N X s Scapula length 0 - - 0 - - 1 239.6 - 0 - - b Scapula width 0 - - 0 - - 1 173.0 - 0 - - O Humerus length 1 262.0 - 4 307.3 17 3 264.0 10 2 365.5 2 2: Humerus, proximal width 1 113.6 - 4 119.1 6 2 86.4 0.1 3 131.6 10 a Humerus, midshaft width 1 45.1 - 5 47.0 6 6 47.8 3 3 55.1 2 Humerus, distal width 1 66.5 - 5 94.7 7 6 87.7 3 3 114.2 17 Radius length 1 217.0 - 5 264.2 18 2 215.3 14 2 299.5 8 Radius, midshaft width 3 40.0 8 5 41.3 2 2 41.9 2 3 47.8 1 c~a Radius, distal width 3 61.7 2 5 74.9 10 3 77.0 6 2 89.4 2 Ulna length 1 263.0 - 1 348.0 - 0 - - 0 - - o Ulna, olecranon to sigmoid "3 notch 1 79.8 - 4 96.6 15 0 - - 2 94.0 7 Ulna, midshaft width 2 30.0 0.2 4 36.5 4 0 - - 2 47.6 5 Scaphoid, lateral width 5 42.9 4 5 51.4 2 2 54.1 0.1 1 64.6 - Scaphoid, proximodistal o length 5 26.7 2 5 34.8 1 2 27.1 0.5 1 35.3 - a Metacarpal II length 1 80.2 - 4 109.6 9 2 92.7 2 5 113.9 2 2: Metacarpal II width 1 21.1 - 6 29.8 6 2 31.2 4 5 43.4 2 b Metacarpal III length 0 - - 4 120.5 4 2 103.7 1 4 125.2 9 Metacarpal III width 0 - - 4 36.5 3 2 43.8 2 4 50.9 2 a: Femur length 6 345.7 10 5 387.8 13 1 398.0 - 2 415.0 7 Femur, proximal width 4 130.7 8 3 150.4 9 1 144.8 - 2 149.8 3 Femur, midshaft width 4 38.8 7 4 57.4 10 1 64.3 - 4 66.5 5 Femur, distal width 6 90.8 9 5 108.8 2 2 110.2 1 2 115.1 7 Tibia length 3 263.7 8 3 290.3 20 4 218.0 7 6 271.4 13 Tibia, proximal width 1 84.5 - 3 93.3 6 4 98.6 7 7 110.3 9 Tibia, midshaft width 3 36.2 7 5 40.4 5 4 40.4 6 8 48.5 3 ~I Tibia, distal width 2 66.0 2 6 72.1 7 4 70.2 5 8 85.3 3 2: Astragalus, lateral width 1 49.6 - 7 61.4 4 0 - - 2 71.6 2 2: Astragalus, proximodistal length 1 49.2 - 7 58.8 2 0 - - 2 68.4 5 cll Navicular width 2 35.3 3 2 41.2 0.5 2 39.3 2 1 55.0 - Navicular, proximodistal length 2 14.6 1 2 17.0 2 2 15.4 1 1 20.4 - Metatarsal III length 2 101.4 5 3 99.1 3 1 87.1 - 4 113.7 5 Metatarsal III width 2 26.8 1 3 26.3 3 1 31.5 - 4 45.9 3 MIOCENE RHINOCEROS FROM TEXAS 407 flaringocciput, short pointed nasals, flat dor- TABLE7--Measurements (in mm) of Peracerasprofec- sal skull profile, and laterally-broadened tum. postglenoid process. The rami of F:AM "D. 108338 are very similar in size to the holo- Holotype jamberi" AMNH 9082 It seems clear AMNH F:AM UNSM type, (Table 7). Character 9082 108338 62048 that "Aceratherium"profectum should be referred to Peraceras. Length,angle to symphysis 400 422 - Tanner (1977) described a skull from the Length,angle to late Barstovian Jamber l.f., lower Valentine condyle 215 195 - Formation, Boyd County, Nebraska.Tanner Length,diastema 34 31 Depth below p3 70 77 - consideredthis fauna to be lower Pliocene in Depth below m2 82 84 - but the late Barstovianis now considered m3 length 43.5 45.5 - age, m3 width 26.0 28.1 - to be middle Miocene (Tedfordet al., 1987). Occipitalcondyle to Tanner (1976, 1977) wrestled with assigning nasals - 450 470 this skull to Peracerasor Occipitalcrest to (UNSM 62048) to nasals - 415 453 Diceratherium,but eventuallynamed it a new Occipitalcondyle to dPI - 440 440 species, Diceratheriumjamberi. The only Occipitalcrest to feature of the that dicer- narial notch - 348 354 specimen suggests Foramenmagnum atherine affinitiesis very slight rugosities on to palatalnotch - 250 237 the nasals, which Tanner considered to be Narial notch to tip of nasals - 107 107 evidence for paired nasal ridges. However, Zygomaticbreadth - 258 277 upon examination of this specimen, the in- Width between M2 - 168 170 on the nasals does not Occipitalheight - 177 162 significantroughening Occipitalwidth - 159 165 bear any close resemblanceto the true ridges Width of occipital of Diceratherium. condyle - 92 104 In other "Di- dPl-M3 length - 238 236 every feature,including size, MI-3 length - 126 123 ceratheriumjamberi" is almost a perfect dPl-P4 length - 120 115 match for Peraceras profectum (Table 7). dPI length - 19 23 dP i width - 18 20 Tanner (1977) may not have realized what P2 length - 24 29 primitive Peraceras looked like, since most P2 width - 43 39 of the are based P3 length - 33 32 publisheddescriptions genus P3 width - 56 48 on very derived, late forms such as P. su- P4 length - 38 33 perciliosum.All of the featurescited by Tan- P4 width - 60 57 ner Ml length - 38 40 as evidence for diceratherineaffinities are MI width - 58 55 features that are also found in primitive ac- M2 length - 43 43 eratherines.Thus, Diceratheriumjamberi is M2 width - 61 60 M3 length - 46 47 considered a junior synonym of Peraceras M3 width - 59 58 profectum. This synonymy eliminates the anomalously late occurrence of Dicerathe- rium suggestedby Tanner.

Genus APHELOPSCope, 1874 weak antecrochets;primitive lambdoid crests Revised diagnosis.--Aphelops shows the (skull triangularin posterior view); narrow derived features of a dorsally-archednaso- zygomaticarches; lower tusk (i2) sub-circular frontal profile (primitively flat) and an un- in cross section; and primitive rhinocerotid usually long diastema between the i2 (lower skeletal proportions. Contrary to Quinn tusk) and the first premolar. In most other (1955) and Patton (1969, p. 119), Aphelops features, Aphelops retains the primitive ac- is not particularly"long-legged," relative to eratherine condition, including: greatly re- more primitive (or the living) rhinos. It only duced premaxilla and loss of II; broad, un- seems so in contrast to the unusually short- fused, hornless nasals with only slightly legged Teleoceras,with which it is frequently downturned lateral edges; nasal incision re- found. Later, more derived species of Aphe- tracted to the level above anterior P4; brachy- lops develop mesodont teeth, further nasal dont teeth without cement; upper molars with incision retraction,and long slender nasals. 408 DONALD R. PROTHERO AND EARL M. MANNING

,1 ,

??i ;* /-~

1P1

7 8 FIGURE10- Teleoceras meridianum n. comb. 1, 2, USNM 3177, stereopair of holotype right ?M2; 3, comparison of USNM 3177 (right) with right M 1-3 of F:AM 108306 (left); 4, F:AM 108269, isolated nasals, dorsal view; 5-8, F:AM 108306, referred skull, posterior, right lateral, dorsal, and palatal views. Scales in cm. MIOCENE RHINOCEROS FROM TEXAS 409

APHELOPSMEGALODUS (Cope, 1873) as. The lowerjaw material,however, has the and Aceratherium megalodum COPE,1873, p. 1. procumbent symphysis, long diastema, Aphelopsmegalodus (Cope). COPE,1874, p. 520. lack of cingulacharacteristic of Aphelopsand Aphelopssp. (large)HESSE, 1943, p. 171. not of Peraceras.The postcranialmaterial is all too large to belong to P. profectumand is Holotype.-AMNH 8292, a complete male in size to A. material skull and from the late Barstovianof the comparable megalodus jaws from the early Barstovian of Colorado and Pawnee Creek area, northeast Colorado. the from the Distribution.-Burkeville and Cold Nebraska, particularly sample Spring Olcott Formation (lower Snake Creek Fauna faunas (early and late Barstovian, middle in of Sioux Nebraska. A in Figure 2) County, Miocene), Texas Gulf Coastal Plain, and full of the Texas material is be- other Barstovian localities outside of description many yond the scope of the present paper, since it Texas. would necessitate a full revision and rede- Referredmaterial. --Burkeville Fauna (early Barsto- scription of the genus Aphelops. vian):Trinity River Pit 1, San JacintoCo., Texas:F:AM 108892,2 metacarpalsV; F:AM 108906,4 fibulae;F:AM Subfamily RHINOCEROTINAE 108905,3 tibiae;F:AM 108904, 3 patellae;F:AM 108903, 1821 (expanded) 4 F:AM 3 F:AM Gray, patellae; 108902, partial femora; Tribe TELEOCERATINI 1902 108901, 2 femora; F:AM 108900, partial ulna; F:AM Hay, 108899, 2 radii;F:AM 108898,4 humeri;F:AM 108927, Genus TELEOCERASHatcher, 1894 3 metacarpalsIV; F:AM 108926, 3 metacarpalsIII; F: Revised diagnosis.- Teleocerasis derived AM 108925, 4 metacarpalsII; F:AM 108924, entocu- in neiform;F:AM 108923,2 ectocuneiforms;F:AM 108922, many features,including hypsodont teeth, 2 naviculars;F:AM 108921, pisiform; F:AM 108920, strong antecrochets,greatly reduced premo- trapezoid;F:AM 108919, 3 unciforms;F:AM 108918, lars with loss of dP 1/1 and occasional loss of 3 unciforms;F:AM 108917, cuneiform;F:AM 108916, P2/2, thick cement on teeth, narrow nasals 2 lunars;F:AM 108915, 4 scaphoids;F:AM 108914, 3 with strong downturned lateral edges, en- calcanea;F:AM 108913, 3 calcanea;F:AM 108912, 3 calcanea;F:AM 108911, 3 astragali;F:AM 108910, 4 larged premaxilla and II, broad zygomatic astragali;F:AM 108909, 2 Ml's; F:AM 108908, left M3; arches, flaring lambdoid crests (skull semi- F:AM 108931, 2 metatarsalsIV; F:AM 108930, 3 meta- circularin posterior view), a small terminal tarsals III; F:AM 108929, 2 metatarsalsII. Woodville nasal horn and fused nasals, lower tusk (i2) localities,Tyler Co., Texas:F:AM 108932,proximal right in cross and char- scapula;F:AM 108943,left metacarpalII; F:AM 108942, teardrop-shaped section, rightmetacarpal III; F:AM 108941, rightmetatarsal III; acteristicbody proportionsof a barrel-shaped F:AM 108940, left metatarsalIV; F:AM 108938, 3 iso- trunk and short, robust limbs. lated worn uppermolars; F:AM 108936, left calcaneum; F:AM 108935, right scaphoid;F:AM 108934, rightcal- TELEOCERASMEDICORNUTUM Osborn, 1904 caneum;F:AM 108933, proximal ulna. Figure 11.4 ColdSpring Fauna (late Barstovian): Two miles northeast of Cold Spring, San Jacinto Co., Texas: F:AM Teleoceras medicornutus OSBORN,1904, p. 319. 108956,juvenile mandiblewith dp2-4 and ml erupting. Threemiles north of Cold Spring,San JacintoCo., Tex- Holotype. -AMNH 9832, a complete skull as: F:AM 108955, rightfemur. Near Swartaut,Polk Co., andjaws from the late Barstovianof the Paw- Texas:F:AM 108939, rightMI; F:AM 108937, righti2. nee Creek area, northeast Colorado (Gal- McMurrayPits, San Jacinto Co., Texas:F:AM 108959, patella;F:AM 108958, 3 tibiae;F:AM 108957, 3 femora; breath, 1953). F:AM 108954, right metatarsalII; F:AM 108953, left Distribution. -Burkeville Fauna (early metacarpalII; F:AM 108952, right metacarpalIII; F: Barstovian, middle Miocene), Texas Gulf AM 108950, 3 calcanea;F:AM 108951, astragalus;F: Coastal Plain, and in many other early and AM 108949, cuboid; F:AM 108948, 3 fibulae; F:AM late Barstovian localities outside Texas. 108947, 2 proximalulnae. than all other Referredmaterial. --Burkeville Fauna (early Barsto- Diagnosis.--Smaller species vian): Localities near Doucette and Woodville, Tyler of Aphelops.Nasal incision not retractedfur- Co., Texas: F:AM 108893, left metacarpalII; F:AM ther than above anterior P4 (to posterior P4 108892, left lunar; F:AM 108891, right and left distal in A. malacorhinus, to M1 in A. mutilus). humeri;F:AM 108890, rightastragalus; F:AM 108889, Discussion. there is a large brokenleft m2; F:AM 108888, brokenright ml; F:AM --Although 108887, righti2; F:AM 108886, left ramus with m2-3; amount of Aphelops material from Texas, no F:AM 108897, right metacarpalIV; F:AM 108896, left complete skulls are yet known to show con- metacarpalII; F:AM 108895, rightmetatarsal IV; F:AM clusively that it is Aphelops and not Peracer- 108894, left metatarsalIII. Trinity River Pit 1, San Ja- 410 DONALD R. PROTHEROAND EARL M. MANNING

TABLE 8--M1-3 lengths(in mm) of selected samples of Teleoceras.

Sample N .c s OR T. medicornutum,late Barstovian,Burge Mbr., ValentineFm., Nebr. 5 156.8 4.1 154-163 T. medicornutum,holotype, late Barstovian,Pawnee CreekFm., Colo. 1 154.3 - - T. major,holotype, early Clarendonian,Cap Rock Mbr., Ash Hollow Fm., Nebr. 1 157.0 - - T. major,early Clarendonian,Cap Rock Mbr., Ash Hollow Fm., Nebr. 6 158.2 7.8 148-170 T. cf. major,early Clarendonian,Lapara Creek Fm., Tex. 3 161.0 5.5 157-167 T. major,late Clarendonian,Merritt Reservoir Member, Ash Hollow Fm., Nebr. 13 154.2 7.8 142-164 T. fossiger, early Hemphillian,Ogallala Fm., Kans. 13 172.4 6.1 161-183 T. fossiger, holotype, early Hemphillian,Ogallala Fm., Kans. 1 187.5 - - T. proterum,early Hemphillian,Mixson's Bone Bed, Fla. 2 148.0 2.8 146-150 cinto Co., Texas:F:AM 108837, left p2; F:AM 108836, M3's from the same quarry,Trinity River Pit m3; F:AM 108835, right p2-3; F:AM 108350, rightra- 1, shows none with this condition. F:AM mus with m2-3; F:AM 108349, left ramus with sym- 108838 shows so little wear that the isolation physis and p3-m3; F:AM 108838, partial skull; F:AM 108840,right m2; F:AM 108841,right p2; F:AM 108844, of the anomalous cusp would disappear in left 12;F:AM 108845, atlas;F:AM 108846, 3 vertebrae; older wear stages.The split between the cusp F:AM 108847, caudalvertebrae; F:AM 108848, back of and the metaloph is very shallow and would skull;F:AM 108849, righthumerus; F:AM 108850, left disappearwith wear. Thus, it is felt that this humerus;F:AM 108851, proximalright humerus; F:AM feature is a dental caused 108852, juvenile right humerus;F:AM 108853, proxi- anomaly, possibly mal righthumerus; F:AM 108854,3 radii;F:AM 108855, by an epigenetic disturbance in the devel- 2 ulnae;F:AM 108856, 3 brokenfemora; F:AM 108857, opment of the teeth. Similardental anomalies 2 partial fibulae; F:AM 108858, distal tibia; F:AM are discussedby Archer(1975), Taylor(1982), 108881,4 metacarpalsIV; F:AM 108880, 3 metacarpals and Rose and Smith (1979). Such a distur- III; F:AM 108879, 5 metacarpalsII; F:AM 108878, 2 ectocuneiforms;F:AM 108877, left navicular; F:AM bance would have to have occurredvery early 108876, 4 cuboids; F:AM 108875, 3 magnums;F:AM in development,since it occurson both sides. 108874, 3 trapezoids;F:AM 108873, 4 pisiforms;F:AM Butler (1967) suggestedthat bilaterallysym- 108872, 3 cuneiforms;F:AM 108871, 4 lunars;F:AM metrical dental anomalies result from dis- 108870, right unciform;F:AM 108869, right scaphoid; turbances in the of F:AM 108868, 3 patellae;F:AM 108867, 4 patellae:F: migration mesenchyme AM 108866, 3 left calcanea;F:AM 108865, 3 rightcal- from the neural crest cells. canea; F:AM 108864 4 astragali;F:AM 108863, 4 as- The remaining material of T. medicornu- tragali;F:AM 108862, 2 partialtibiae; F:AM 108861, 2 tum compares favorably with the type and tibiae; F:AM 108860, 3 tibiae; F:AM 108859, 2 tibial with referredmaterial from the Barstovian fragments; F:AM 108885, 3 metatarsals IV; F:AM 108884, 4 metatarsalsIII; F:AM 108883, 2 metatar- of New Mexico, Colorado, and particularly sals II. the largesamples from the Olcott Formation Snake Creek Fauna) of Nebraska. A -P3/3 less reducedthan all oth- (Lower Diagnosis. full descriptionof this materialis beyond the er species of Teleocerasexcept T. meridia- of the since it would incision scope present paper, num. Teeth less hypsodont. Nasal necessitate a full revision and redescription not retractedfurther than above anteriorP3. of the Teleoceratini. Largerthan T. meridianum. Discussion.-The most remarkableTexas TELEOCERAScf. MAJORHatcher, 1894 referredto T. medicornutumis F: specimen Teleoceras 149. AM 108838, a complete skull. The dentition major HATCHER, 1894, p. Teleoceras proterus (Leidy). QUINN, 1955, p. 72. shows the normal Teleocerascondition with Teleocerasproterus (Leidy). PATTON, 1969, p. 214. strongantecrochets and hypsodonty, and the teeth are relativelyunworn. The strikingfea- Discussion.--A large sample of Lapara ture of this specimen is that both M3's have Creek Teleoceras,including many good skulls a distinct, isolated cusp (Figure 11.4) at the and postcranialbones, is stored at the Texas posteromedialend of the metaloph.No other MemorialMuseum. A full descriptionof this rhinocerosthat we have seen shows this. Such material is postponed until a complete re- a feature might be consideredthe basis for a vision of Teleocerascan be undertaken.These new species under a typological species con- LaparaCreek rhinos were referredto Teleoc- cept. However, comparisonof other isolated eras proterusby Quinn (1955), and his iden- MIOCENE RHINOCEROS FROM TEXAS 411 tificationwas followed by Patton (1969). But 108294,5 metacarpalsII; F:AM108292, distal femur; a preliminary comparison of this material F:AM 108293, 2 metacarpalsIII; F:AM 108291, prox- indicates that it is within the imal femur; F:AM 108290, juvenile femur; F:AM (Table 8) range 108289, tibia;F:AM 108288, tibia;F:AM 108287, tibia; of size variation of T. medicornutum or T. F:AM 108286, 2 i2; F:M 108285, femur;F:AM 108284, major, rather than the smaller Hemphillian humerus;F:AM 108283, distal radius; F:AM 108282, rhino, T. proterum,or the largeHemphillian radius;F:AM 108281, humerus;F:AM 108280, humer- form, T. fossiger. The Lapara Creek rhinos us; F:AM 108279, humerus;F:AM 108278, humerus; than T. F:AM 108277, humerus;F:AM 108276, radius;F:AM are also less hypsodont proterumand 108275, radius;F:AM 108339, mandible with left p2- lack the medifossettes seen on the molars of m3, right p3-m3; F:AM 108348, cuneiform; F:AM this species. Temporarily the Lapara Creek 108347, righti2; F:AM 108346, cervicalvertebra; F:AM rhinos are referredto T. of. major(the typical 108345,thoracic vertebrae; F:AM 108342,juvenile skull; Clarendonian which it most F:AM 108341,juvenile skull with left and rightdP -4, early form, M1; F:AM 108839, mandiblewith left p3-m3, rightp2- closely matches) until a full revision of Tel- m3. eoceras can be attempted. Barstovianlocalities (horizon unknown):McKenney The occurrenceof T. cf. major(known from Place, near Stephen Creek, San Jacinto Co., Texas: F: the early Clarendonian)in the LaparaCreek AM 108344, magnum.One-half mile south of William- Fauna with the late son's Farm,near Stephen Creek, San JacintoCo., Texas: agrees Barstovian-early F:AM 108343, distal right humerus. Clarendonianage of this fauna(Patton, 1969; Tedford et al., 1987). Teleocerasproterum, on the other is known from the Diagnosis. - Teleoceras meridianum is by hand, only far the smallest of known of Teleoc- Hemphillian of Florida. species eras. It also has relatively robust limbs. -The USNM TELEOCERASMERIDIANUM Description. type specimen, 3177, has a strong crochet and antecrochet, (Leidy, 1865) n. comb. and is very hypsodont, with a narrow, slit- Figures 10-14 shaped medifossette. All of these features are Rhinocerosmeridianus LEIDY, 1865, p. 176; 1869, well matched by the right M2 ofF:AM 108306 p. 229, Pl. 23. a skull from the Pit 495 (Figure 10.3), McMurray Aphelopsmeridianus (Leidy). COPE,1875, p. USNM 3177 and the M2 of (type specimen only). locality. F:AM 108306 are also closely comparable in size, Holotype. -USNM 3177, broken ?M2. although this is difficult to measure precisely, Distribution. - Burkeville and Cold Spring since the ectoloph of the type is so badly bro- faunas, early and late Barstovian (middle ken and F:AM 108306 is more worn than the Miocene), Texas Gulf Coastal Plain. type. The only morphological differences between the two are due to differences in wear. Referredmaterial. --Burkeville Fauna (early Barsto- vian): West of Doucette, northwestof Woodville, Tyler The crochet of F:AM 108306 is worn down Co., Texas: F:AM 108261, right mandible with p2-4. to the base, and the postfossette is more tri- TrinityRiver Pit 1, SanJacinto Co., Texas:F:AM 108263, angular and open. Since USNM 3177 is clear- left and rightjuvenile edentulous rami; F:AM 108260, ly a teleoceratine and there are no other Tel- right I1; F:AM 108259, right metatarsal III; F:AM 108258, left metatarsalIV: F:AM 108257, left metatarsal eoceras from Texas that match it in size; F: IV: F:AM 108256, scaphoid;F:AM 108255, unciform; AM 108306 is referred to Teleoceras meri- F:AM 108254, magnum; F:AM 108253, humerus; F: dianum. AM 108842, right M3; F:AM 108843, left M3. F:AM 108306 (Figure 10.5-10.8) has been Cold SpringFauna (late Barstovian):Two miles below crushed so it is difficult Swartaut,Polk Co., Texas: F:AM 108272, left juvenile badly dorsoventrally, ramus with dpl-2; left ramal fragmentwith m2. A. D. to interpretthe shape of the skull. The nasals, Donohoe Pit, San JacintoCo., Texas:F:AM 108265, left premaxillae,and partof the maxillahave been and rightrami with p3-m3. Cold Spring1.f., San Jacinto badly broken. An isolated anterior portion of Co., Texas:TMM 31219-223, scapula.McMurray Pits, the nasal bones from the same San Jacinto Co., Texas: F:AM 108262, proximaltibia; pit (F:AM F:AM 108266, skull and mandible;F:AM 108273, right 108269) shows the characteristic fusion, M2; F:AM 108270, righti2; F:AM 108306, skull;F:AM strongly downturned lateral edges, and ter- 108269, proximal nasal bones; F:AM 108305, 2 I1; F: minal horn rugosity of Teleoceras despite its AM 108304, 2 scaphoids;F:AM 108303, cuneiform;F: unusually small size (Figure 10.4). The fron- AM 108302, pisiform;F:AM 108301, magnum;F:AM 108300, unciform;F:AM 108299, 3 ectocuneiforms;F: tals and parietals have been broadened and AM 108298, cuboid;F:AM 108297, 2 naviculars;F:AM flattened by postmortem crushing. The oc- 108296, 3 metatarsalsIV; F:AM 108295, patella;F:AM ciput, however, is relatively undeformed and 412 DONALD R. PROTHERO AND EARL M. MANNING

m~r~rT 1rIr r

1 "1

2 LW11 LL I 3

4 I II I1I FIGURE11-Teleoceras. 1, 2, Teleocerasmeridianum, F:AM 108839, right lateralview of mandibleand crown view of teeth; 3, F:AM 108341, left dPl-4, crown view. 4, Teleocerasmedicornutum, F:AM 108838, right P2-M3, showing peculiarisolated metaloph cusp on M3. Scales in cm. has a broad, flaring lambdoid crest. The zy- McMurray Pit that show the characteristic gomatic arches are also broad and flaring. bladelike shape of Teleoceras. There are no The nasal incision is retracted to the level upper premolars known for T. meridianum. above anterior P3. M1-3 are present on both sides of F:AM The basicranium is poorly preserved. The 108306. Although very worn, they are clearly postglenoid processes are relatively narrow very hypsodont. All three molars have strong and slender. The mastoid process is a flaring, antecrochets. Ml is so worn that the ante- slender wing fused to the postglenoid process. crochet has enclosed a prefossette; the post- Both paroccipital processes are broken off, fossette is nearly circular. M2 (as in the type but they were apparently narrow and slender. specimen) has a distinct protocone, a strong The rest of the basicranium is too poorly pre- antecrochet, and the base of what was once served to distinguish foramina or other de- a prominent crochet. M3 has a strong crochet tailed structures. The pterygoid flanges are and antecrochet. There are no cingula on the short and stubby. The sulcus anterolateral to upper molars. the pterygoid flange is distinct. The second- The mandible (Figure 11.1, 11.2) has a pro- ary palate terminates at the level medial to cumbent symphysis and long diastema. Iso- anterior M3. lated lower tusks of both male (F:AM 108270) No incisors or premolars are preserved on and female (F:AM 108286A-B) morphology F:AM 108306. However, there are a number are known. The coronoid, condyles, and an- of isolated Il's in the sample from the gle of the jaw are robust. The lower premolars MIOCENE RHINOCEROS FROM TEXAS 413

o g2q 1 2 3 4 C ~*.,N

/ .Z 6 7

5 FIGURE 12- Teleoceras meridianum limb elements. 1, 2, F:AM 108279, humerus; 3, F:AM 108276, radius; 4, TMM 31219-223, scapula; 5, F:AM 108285, femur; 6, 7, F:AM 108288, tibia-fibula; 8, 9, F:AM 108295, patella. Scales in cm. 414 DONALD R. PROTHERO AND EARL M. MANNING

6,,,

6 .

. .

1 4

15 16 MIOCENE RHINOCEROS FROM TEXAS 415 are highly reduced. Lower p2 is missing in The manus (Figure 13) of T. meridianum all specimens, although there is a small al- is known from the scaphoid,cuneiform, mag- veolus for it in some. The lower molars are num, unciform,pisiform, and metacarpalsII very hypsodont and lack cingula. The lower and III. They all show a markeddorsoventral cheek tooth patternis typically rhinocerotid. compression. The scaphoid (Figure 13.1- Two nearly complete juvenile skulls also 13.3) is very flattenedproximodistally. It has occur in the McMurrayPit. The largerone, a distinct posteriorlunar facet. The magnum F:AM 108341, has dPl-4 and the Ml still (Figure13.8-13.10) has a relativelylarge pos- deep in the alveolus. The smaller, F:AM terior (plantar)process, with no indication of 108342, has dP2-4 and the tip of Ml just modification for a magnum-unciformartic- above the gumline. DP1 (Fig. 11.3) consists ulation (Harrisonand Manning, 1983). The of a simple ectoloph with a metaloph at right cuneiform (Figure 13.4-13.7) is very dorso- angles to it. DP2-4 all have high-crowned ventrally flattened, with an indistinct poste- ectolophs, stronganterior cingula, strongan- rior process. The unciform (Figure 13.11- tecrochets,and lobate crochets that meet the 13.13) is transverselyflattened, with an in- crista to form a medifossette. The crochet is distinct cuneiform facet. The pisiform (Fig- longest and most intricately lobate on dP4. ure 13.15, 13.16) has a reduced distal swell- The sutures of these skulls are unfused, the ing relative to the size of the articularfacets. zygoma are narrow, the supraorbitalrugosi- MetacarpalII (Figure 13.14) is stubby and ties are undeveloped, and the occiput is un- anteroposteriorlyflattened. Metacarpal III is flared;all of these featuresmight be expected also stubby and flattened, with a distinct in a juvenile skull. proximal facet for metacarpalIV. Postcranial skeleton. -A few badly broken The femur (Figure 12.5) is robust and has vertebraeand some rib fragmentsare all that been subjected to anteroposteriorcrushing. are known of the axial skeleton of T. meri- The greater trochanter is relatively less ro- dianum. The scapula (Figure 12.4) is known bust, without an anteroproximalridge. The from TMM 31219-223. It was badly broken patellartrochlea of the femur is narrow,and and has been incorrectlyrestored in plaster. the tibial trochlea is relatively reduced. The It appears to lack a posterior flexure of the tibia (Figure 12.6, 12.7) has a less distinct scapular spine and the triangularproximal proximal tuberosity. The fibula is a robust end. The parts that are original bone, how- splint that apparentlywas fused to the tibia ever, show that the scapula was broad and at its proximal end. The pes (Figure 14) of robust, but relatively small. T. meridianum is known from the cuboid, Most of the limb elements of T. meridia- navicular,ectocuneiform, and metatarsalIV. num are from the McMurrayPit and have The cuboid (Figure 14.1-14.3) is dorsoven- undergonecrushing and plastic deformation. trally crushed,with a large facet for metatar- The humerus (Figure 12.1, 12.2) is known sal IV. The navicular (Figure 14.4-14.6) has from a number of good specimens. It has a distinct facets for the ectocuneiformand me- weakly developed deltopectoralcrest and lat- socuneiform. The ectocuneiform (Figure eral epicondyle. The deltopectoralcrest lacks 14.7-14.9) is dorsoventrallyflattened. Meta- a prominentanterolateral process. The radius tarsal IV (Figure 14.10-14.11) is short, ro- (Figure 12.3) is drasticallyshortened relative bust, and shows little lateral curvature. to its diameter. The facets and rugosities on In summary,the postcranialskeleton of T. the proximal and distal ends are relatively meridianumshows many of the same trends indistinct. No ulna is known for T. meri- seen in Peraceras hessei: smaller size, fewer dianum. rugosities, weakly developed articular sur-

FIGURE13--Teleoceras meridianum, manus elements. 1-3, F:AM 108304a, right scaphoid, in left lateral, proximal, and distal views; 4-7, F:AM 108303, right cuneiform, in right lateral, dorsal, proximal, and distal views; 8-10, F:AM 108301, right magnum, in proximal, dorsal, and left lateral views; 11-13, F:AM 108300, right unciform, in right lateral, proximal, and left lateral views; 14, F: AM 108293b, right metacarpal III, and F:AM 108294b, right metacarpal II, in dorsal view; 15, 16, F:AM 108302, right pisiform, in plantar and left lateral views. Scale in cm. 416 DONALD R. PROTHERO AND EARL M. MANNING

1 2.3 .

•

?z3' ..... 4i

.r .

?? ?LCf .,

5 6 - .S??- .5 , 11

9 .? ?. ,''? . . ..

...?. ?_

y "7" "? " l ..... t 1I0

FIGURE14- Teleoceras meridianum, pes elements. 1-3, F:AM 108298, left cuboid, in dorsal, left lateral, and proximal views; 4-6, F:AM 108297b, right navicular, in dorsal, proximal, and distal views; 7-9, F:AM 108229b, right ectocuneiform, in dorsal, proximal, and distal views; 10, 11, F: AM 108296a, right metatarsal IV, and F:AM 108259, right metatarsal III, dorsal and proximal views. Scale in cm. MIOCENE RHINOCEROS FROM TEXAS 417

E U faces,and increasedrobustness relative to size. of these features must be A A cDD O No ADD Many interpreted 120 I0o 140 150 160 170 180 with caution. Most of the elements of both Length MI-3 in mm T. meridianum and P. hessei come from the Pit, where have undergone FIGURE15--Comparison of ml-3 lengths of late McMurray they Barstovian Teleoceras.Solid squares= male T. considerable crushing and plastic deforma- medicornutum, Frick Horse and Mastodon tion. Some of the flatteningdescribed above Quarry, Pawnee Creek Fm., Colorado. Open is certainly due to these postmortem effects. squares = female T. medicornutum, Horse and Discussion.-The of = history Leidy's type Mastodon Quarry. Solid triangle type speci- of "Rhinoceros" meridianus is dis- men of T. Pawnee Creek specimen medicornutum, Fm., cussed in the introduction. Most authors have Colorado. Open triangles = T. meridianum, Texas Gulf Coastal Plain. referred it to Aphelops meridianus or considered it indeterminate. However, it appears that none of these authorsactually examined the type specimen in making these state-

TABLE9-Summary of occurrences. Menoc- Teleoc- Teleoc- eras Menoc- Perac- Aphelops eras eras Teleoc- arika- eras eras mega- meri- medicor- eras Locality rense barbouri hessei lodus dianum nutum cf. major DERRICK FARM (USNM) (?Late Arikareean) X ?Catahoula Formation GARVIN GULLY FAUNA (Early Hemingfordian) Oakville Formation TMM 41662 X TMM 40106 X TMM 40667 X TMM 31048 X BURKEVILLE FAUNA (Early Barstovian) Fleming Formation TMM 31190 Point Blank l.f. X X X TMM 31242 X X X TMM 31243 X TMM 31246 Swiss Alp X X X TMM 40127 X TMM 31057 Moscow l.f. X X X X TMM 40290 X Trinity River Pit 1 (F:AM) X X X X TMM 40196 (Trinity River Pit 1) X Doucette-Woodvillearea (F:AM) X X X X COLD SPRING FAUNA (Late Barstovian) Fleming Formation McMurray Pits 1 and 2 (F:AM) X X X Near Swartaut (F:AM) X X X Noble Farm (TAM) X X X Halletsville(TAM) X TMM 31191 San Houston l.f. X TMM 31219 Cold Spring l.f. X X X TMM 30009 Woodville southwest X TMM 31272 Chapell Hill X TMM 40623 Push Creek TMM 40622 Belts Creek X LAPARA CREEK FAUNA (Early Clarendonian) Goliad Formation X 418 DONALD R. PROTHEROAND EARL M. MANNING

\S` Teleocerasmeridianum is as strikingan ex- - 00 ample of dwarfingas Peracerashessei, since it undergoesa 20 percentsize reduction.It is 51 ACRTER11J RHIN RO clearlya dwarfspecies, since it is smallerthan any known teleoceratine,Eurasian or North 4 CACERATHERINAIE 9 RHINOCE ROT INAE American, including its primitive sister-taxon Brachypotheriumamericanum. Teleocer- as meridianumis also slightly smaller than undescribed teleoceratines from the late Hemingfordian Sheep Creek Formation of Nebraska in the Frick Collection. These are the oldest and most primitive Teleoceras FIGURE16--Hypothesis of relationshipsof North AmericanMiocene rhinos. Derived character known, but descriptionof this materialis be- statesat numberednodes explained in text. yond the scope of the present paper. Sexualdimorphism as a cause for the small size of T. meridianumcan be ruledout, since: ments. The specimen (USNM 3177) is a bad- 1) there is no large Teleoceras in the large ly broken left upper ?M2 (Figure 10.1, 10.2). samples of late Barstovian Texas faunas; 2) It is very hypsodont and has a strong ante- extreme size dimorphism is unknown in the crochet. Both of these features characterize rhinocerotids (Laurie, 1982); and 3) both Teleoceras and not Aphelops or Peraceras. male- and female-shaped lower tusks are "Rhinoceros"meridianus is clearlya species present in the sample of T. meridianum.In of Teleoceras. Figure 15, T. meridianumis compared with This conclusion is strengthenedby com- the large samples of T. medicornutumfrom parison with the material described above, the FrickMastodon and Horse Quarryof the particularlythe skull F:AM 108306, shown BarstovianPawnee CreekFormation of Col- in Figure 10.3-10.8. Matthew (1932, p. 420) orado. Teleocerasmeridianum is clearlymuch considered meridianus indeterminate until smaller than any individual of T. medicor- an adequate topotype is found. Unfortu- nutum.Although late BarstovianTexas Gulf nately, Leidy's locality data are so vague that Coast faunas contain only T. meridianum, it cannot be determined with certainty the early Barstovian faunas contain both T. whetherthe specimenscame from the Burke- meridianumand T. medicornutum(Table 9). ville or Cold Spring level in Washington These two rhinos were clearlysympatric and County. Hesse (1943) was of the opinion that very distinct in size from one another,as can Leidy's localities were from the Burkeville be seen by comparisonof the materialof both level, but this is not certain.Almost no other rhinos from Trinity River Pit 1. It is unclear Barstovianrhinoceros material is knownfrom why T. medicornutumwas absentfrom Texas Washington County. If the concept of type in the late Barstovian,when it was very com- area can be extended 75 miles to the east, mon in High Plains faunasfrom the late Bar- then the materialfrom the Cold Springlevel stovian of Colorado, Nebraska, and New in San JacintoCounty described above serves Mexico. Further study of teleoceratine sys- admirablyas a topotype. Even if the type area tematics and biogeographypromises some is consideredrestricted to WashingtonCoun- interestingresults. ty, the type of meridianusis still not indeterminate, since it matches the San Jacinto RELATIONSHIPS OF THE NORTH Countyspecimens described above so closely. AMERICAN MIOCENE RHINOCEROSES Thus, Teleoceras meridianum is a valid The systematic significanceof the distin- species, presentlythe oldest (1865) described guishingcharacters of the various Gulf Coast species that can be referredto the genus. The rhino genera can only be understood from a genus Teleoceraswas establishedby Hatcher broader perspective of North American Mio- in 1894, based on the species T. major from cene rhinoceroses. A broader review of the the early Clarendonianof Nebraska. Aphe- systematic relationships of the higher groups lops fossiger Cope, 1878, from the early of rhinos has been provided elsewhere Hemphillian of Kansas, was assigned to Tel- (Prothero et al., 1986). The relevant parts of eoceras at a later date. that study are summarized here. MIOCENE RHINOCEROS FROM TEXAS 419

Menocerasis more closely relatedto higher species was placedin the genus Aceratherium rhinos than to primitive ones such as Dicer- originallyby Cope. The only majorconfusion atherium. This is suggestedby the following in the history of the genus was the placement characters:strong upper molar crochets; 12 of the species fossiger in Aphelopsby Cope lost; reduced sagittal crest; reduced premax- in 1877. It was not until much later thatfos- illary; nasal incision over posterior P2; ba- siger was properlyplaced in Hatcher's 1894 sicraniumshortened relative to palate;upper genus Teleoceras.Aphelops starts as a rela- molar lingual cingula weak or absent; and tively primitive aceratherine,with the fol- shallow anteroventralnotch on atlas (node 1 lowing derivedcharacters: nasofrontal profile in Figure 16). The taxon "Diceratheriinae" dorsally arched and the i2/dpl diastema as used to unite these two genera (e.g., Heis- slightlylengthened (node 5). The genusranges sig, 1973) is a rareexample of a strictlypoly- from late Hemingfordianto late Hemphillian phyletic (ratherthan paraphyletic)mamma- in America and may be represented in the lian taxon, since the paired nasal horns are Vindobonianof Asia. It is a very useful genus convergentlyobtained in Menocerasand Di- biostratigraphically,as it appearsto steadily ceratherium.The genus Menocerasis further increase in size from the middle Clarendo- characterizedby the presenceof twinned bul- nian until the late Hemphillian. It is one of bous terminal nasal horns in the males and the few lines which does not get smaller in a vestigial metacarpaland posterodorsalzy- the late Hemphillian. gomatic tuberosities (node 2). Menoceras is A few characterssuggest that the other ac- a very short-lived genus, rangingonly from eratherinesare more closely related to each the latest Arikareean(the Agate fauna and its other than they are to Aphelops-a flattened equivalents)to the early Hemingfordian(the dorsalskull profile and shortenednasals (node fauna of the RunningwaterFormation and 6). This suggests that the one American ac- its equivalents). erathere(Peraceras) is closer to the Eurasian Aceratherines,teleoceratines, and rhinoc- aceratheresthan to the other American one erotinines (see characterizationsfollowing) (Aphelops). can be united on the basis of fully bilopho- The genus Peracerasis the least well under- dont upper premolars and an overall in- stood of the American Miocene rhinos. This creasedsize (reversedin the dwarfforms dis- is true for many reasons: primitively it re- cussed earlier) (node 3). These Eurasian sembles Aphelops;the derived forms mimic groupsunderwent an Aquitanian(early Mio- Teleocerasin severalways; it has a very short cene) diversificationwhich apparentlyled to range in time; and, most importantly, it is the formation of the three major groups of common only in a few faunas. It is distin- Miocene rhinos. guished by the following characters:occiput The first of these major groups, Acera- procumbent; mandibular symphysis up- theriinae, was first used by Matthew (1931) turned in females (with i2/dpl diastema in the sense intended here. They are char- shortened);a lingual cingulum usually pres- acterized by a nasal incision retracted to a ent on the lower cheek teeth; i2 (the lower position over anterior P4 (and, presumably, tusk) very reduced in females; the lambdoid by an expandedupper lip), premaxillaegreat- crest slightly broadened(more so in derived ly reduced (with 11 greatly reduced or lost), forms); and the post-glenoid process slightly and reductionof the medial flangeof the low- broadened laterally (node 7). The primitive er tusk (i2) (node 4). Postcranially,the Amer- species of the genus have not been discussed ican aceratheresremain almost entirelyprim- until the present paper. Peraceras ranges at itive in contrast to Teleoceras. The group least from the earlyBarstovian (and probably appeared in North America in the middle earlier) until the middle Clarendonian (the Hemingfordian(fauna of the Box Butte For- Ricardo fauna). The type specimen of Pera- mation), probably as an immigrant from Eur- ceras superciliosum from the Republican asia. The aceratherines may have supplied River Valley of south-central Nebraska, the competition that caused the extinction of sometimes thought to be Hemphillian (Os- the long-lived American Diceratherium lin- born, 1904), is probably of late Barstovian eage. or early Clarendonian age (R. H. Tedford, The genus Aphelops has been more or less personal commun.). The smaller (P. profec- clearly understood since 1873. The type tum) and larger (P. superciliosum) species are 420 DONALD R. PROTHEROAND EARL M. MANNING

apparentlycommon in a few late Barstovian lambdoid crest expanded laterally;a poste- sites such as the Pojoaque fauna from New rior articulation between the scaphoid and Mexico (F:AM)for the formerand the Hotell the lunar;and reducedbony bar between the Ranch fauna from Nebraska(UNSM) for the foramen ovale and the foramenlacerum me- latter.Peraceras apparently became less com- dium (node 9). Because of these characters, mon by the Clarendonianand was gone be- it is recommended that Rhinocerotinae be fore the late Clarendonian.This may have expanded to include the tribe Teleoceratini been due to competitionwith Teleoceras.The as well as the tribe Rhinocerotini, and that history of Peracerasappears to be that of an Aceratheriinaebe confined to Heissig's Ac- unsuccessfulbid by an aceratherineto com- eratherini. pete with Teleoceras.Many of the featuresof The teleoceratines are distinguished by the large Peraceras appear to be poor imi- hypsodont teeth, strong upper molar ante- tations of Teleocerasfeatures-large propor- crochets,reduced premolars (P2/2 lost in de- tioned, high crowned molars; small propor- rived forms), premaxillaand upper tusk (II) tioned premolars; broad zygomatic arch; enlarged,lateral edges of the nasals strongly flaring lambdoid crest; downturning of the downturned,torso lengthenedand barrel-like, lateral edges of the nasals; and a brachyce- limbs shortened (carpals and tarsals "com- phalic skull. pressed" dorsoventrally), zygomatic arch One might conclude from Peracerasthat it broadened, calcaneal tuber elongate, skull was futile for an aceratherineto try to com- brachycephalic,and nasal incision retracted pete with the grazingteleoceratines. But the to over the anterior part of P3 (node 10). success of the sister lineage of Peraceras in They are one of the most distinctive of all Eurasia(Aceratherium s.s. and Chilotherium) higher rhinocerotidtaxa. contradictsthis. Chilotherium,in particular, The relationshipof the American Teleoc- achieved tremendous success in terms of eras to the Afro-EurasianBrachypotherium numbers and distribution as a grazerin the was recognizedearly (Pavlow, 1892) and was late Miocene (Turolian).The success of Chi- given formal taxonomic recognition by Os- lotheriumprobably had much to do with the born in 1900. Teleoceratinesappear to have competitive defeat of the teleoceratine migratedto North America from Eurasiain Brachypotheriumin Eurasia. The Eurasian the Hemingfordian and persisted in North higheraceratherines are characterizedby their America until extinction in the latest Hemp- slightly parasaggital temporal crests and hillian. During almost its entire range, Tel- slightly laterallyflared lower tusks (i2) (node eocerasis by far the most common American 8). Miocene rhino. It has been suggested (e.g., Heissig (1973) suggested that teleocera- by Matthew, 1932) that Teleocerasdoes not tines were closely relatedto aceratherines(he survive as long as Aphelopsin America, but placed them as tribes of his Aceratheriinae). this is refutedby its presencein several very All charactershe cited were either primitive late Hemphillianfaunas (faunasof the upper for both groups(larger lower tusk, absence of Bone Valley Formation, Florida; upper Bi- nasal horn, short mandibularsymphysis, ar- dahochi Formation, Arizona; Eden 1.f., Cal- ticulation of the radius with the cuneiform, ifornia;Sawrock l.f., Kansas). The confusion unequally curved dorsal surfaces of the car- was due to the peculiar rarity of Teleoceras pals and tarsals,and presenceof manus digit in several late Hemphillian faunas (Coffee V) or derivedfor only the Teleoceratini(pres- Ranch, Texas, and Edson, Kansas) when ence of a small nasal horn, uppermolars with comparedto Aphelops.These are some of the a strong antecrochet,and articulationof the few instances when Teleoceras is not the ulna with the lunar).Heissig cited both pres- dominant rhino in the fauna. ence and absence of the last characterin his It is curious that Teleocerasmaterial was diagnosis of Aceratheriinae. described repeatedly long before Hatcher rec- In contrast to Heissig's view, several char- ognized it as a distinct genus in 1894-"Rhi- acters suggest that teleoceratines are more noceros" meridianus Leidy, 1865; "Aphe- closely related to rhinocerotinines than they lops" fossiger Cope, 1878; "Rhinoceros" are to aceratherines. These features include: proterus Leidy, 1885; and "Eusyodon max- small terminal nasal horn present (expanded imus" Leidy, 1886 ("Eusyodon" was origi- in rhinocerotinines) on fused terminal nasals; nally described as a pig tusk, but is an i2 of MIOCENE RHINOCEROS FROM TEXAS 421

Teleocerasproterum). Why the founders of An hypothesis of relationshipsis proposed Americanvertebrate paleontology never rec- for the common North American Miocene ognized the distinct nature of Teleocerasis a rhinoceros genera, Aphelops,Peraceras, and mystery, since it is distinctive even with the Teleoceras.Aphelops and Peraceras are in- most fragmentarymaterial. cludedin the subfamilyAceratheriinae, along Misunderstandingof Teleocerasevolution with the Europeangenera Chilotheriumand could also be a hindrance to American bio- Aceratheriums.s. Teleoceras (a member of stratigraphy. Unlike Aphelops, Teleoceras the tribe Teleoceratini)is more closely related does not graduallyincrease its size through to the Rhinocerotini (which includes all liv- time. The largest Teleoceras are not the ing species) than it is to the Aceratheriinae. youngest ones. Like many early Hemphillian To reflect this, the usage of the subfamily taxa (Aepycamelus, Yumaceras, Tapirus, Rhinocerotinae is expanded to include the Calippus, Nimravides, Pliohippus, Neohip- tribes Teleoceratiniand Rhinocerotini. parion, Epicyon, Leptarctus,Macrogenis, II- Pros- lingoceras, Barbourofelis, Indarctos, ACKNOWLEDGMENTS thennops), Teleocerasfossiger is unusually large. This fact could be incorrectly inter- We thankJ. A. Wilson for comments, crit- preted to make it appear that early Hemp- icisms, a copy of Hesse's unpublishedmanu- hillian ("Kimballian")Teleoceras is more ad- script, and loan of specimens in the TMM vancedthan the smallerlate Hemphillianone. and TAM collections. Withouthis assistance, Rhinocerotininesare, by far, the most di- this study would not have been a thorough versifiedand widespreadof all higherrhinos. treatment.We thankL. Tannerand M. Voor- This group includes all five of the living hies for reviewingthe manuscript.R. H. Ted- species of rhinoceroses (Javan, Indian, Su- ford graciously made the F:AM specimens matran,black, and white) as well as the long available to us for study. We thank R. Purdy lineages of woolly rhinoceroses and elas- and R. J. Emryfor loan of USNM specimens. motheres. The entire group is characterized Preliminaryidentification of the Frick Texas by an anteroventrallyextended nasal tip (al- rhinos was made by F. Blair, B. Taylor, and lowing space for the expanded nasal horn M. Skinner.Their workgreatly simplified our boss) and by slightly parasaggitaltemporal task. C. R. Prothero helped with the photo- crests (node 11). graphs, and M. Morita produced Figures 7, 8, 13, and 14. We thank M. Zepp and L. SUMMARY Cheungfor help with word processing.Proth- The oldest rhinoceros presently known ero was supported by a Columbia Faculty from the Texas Gulf Coastal Plain is Me- Fellowship during the initial preparationof noceras arikarense from the Derrick Farm this paper. locality (?CatahoulaFormation). It is comparableto late Arikareeanrhinos from Agate REFERENCES Spring Quarry, Nebraska. The rhinos from the Garvin Fauna ARCHER,M. 1975. Abnormal dental develop- Gully (early Hemingfor- ment and its significancein dasyuridsand other dian) are referredto Menoceras barbouri(n. marsupials.Memoirs of the QueenslandMu- comb.), also known from the Runningwater seum, 17:251-265. Formationof Nebraskaand the ThomasFarm BARBOUR,E. H. 1906. Notice of a new fossil local faunaof Florida.Barstovian faunas from rhinoceros from Sioux County, Nebraska:Di- the Texas Gulf Coastal Plain contain four ceratherium arikarense. Publication of the Ne- of rhinoceros: the common braskaGeological Survey, ser. 1, 2:313-318. species High BUTLER, P. M. 1967. Dental merism and tooth Plains forms Teleoceras medicornutumand development. Journal of Dental Research, 46: Aphelops megalodus, and the dwarf species 845-850. Teleoceras meridianum n. comb. and Pera- COPE,E. D. 1873. On some new extinct Mam- ceras hessei n. sp. As far as known, this rep- malia from the Tertiaryof the Plains. Palaeon- resents the instance of a North American tological Bulletin, 14:1-2. only . 1874. on the andPlio- Miocene fauna with four Report stratigraphy sympatric species cene vertebratepaleontology of northern Col- of rhinoceros. The early Clarendonian La- orado. U.S. Geological and Geographic Survey para Creek Fauna contains Teleoceras cf. of the Territories, Bulletin, 14:9-28. major. . 1875. Reporton the geologyof that part 422 DONALD R. PROTHERO AND EARL M. MANNING

of northwestern New Mexico examined during middle to late Eocene hyracodontid (Perisso- the field season of 1874. Annual Report upon dactyla: Rhinocerotoidea) from Asia and west- the Geographical Explorations and Surveys West ern North America. Journal of Paleontology, 55: of the 100th Meridian, Appendix LL:98 1-1017. 826-841. .1877. Report upon the extinct Vertebrata MACFADDEN,B. J. AND M. F. SKINNER. 1977. obtained in New Mexico by parties of the ex- Earliest known Hipparion from Holarctica. Na- pedition of 1874. Report of the U.S. Geological ture, 265:532-533. Survey West of the 100th Meridian, 4(2): 1-370. - AND- . 1981. Earliest Holarctic hippar- . 1880. A new genus of rhinoceros. Amer- ion, Cormohipparion goorisi n. sp. (Mammalia, ican Naturalist, 14:540. Equidae), from the Barstovian (medial Mio- DALQUEST,W. W. AND J. T. HUGHES. 1966. A cene) Texas Gulf Coastal Plain. Journal of Pa- new mammalian local fauna from the lower leontology, 55:619-627. Pliocene of Texas. Transactions of the Kansas MATTHEW, W. D. 1899. A provisional classifi- Academy of Science, 69:79-87. cation of the fresh-water Tertiary of the West. DEUSSEN,A. 1924. Geology of the Coastal Plain Bulletin of the American Museum of Natural of Texas west of the Brazos River. U.S. Geo- History, 12:19-75. logical Survey Professional Paper 126, 145 p. . 1901. Fossil mammals of the Tertiary FRICK,C. 1937. Horned ruminants of North of northeastern Colorado. Memoir of the Amer- America. Bulletin of the American Museum of ican Museum of Natural History, 1(7):355-447. Natural History, 69:1-669. . 1931. Critical observations on the phy- GALBREATH,E. C. 1953. A contribution to the logeny of the rhinoceroses. University of Cali- Tertiary geology and paleontology of north- fornia Publications in Geological Science, 20: eastern Colorado. University of Kansas Paleon- 1-9. tological Contributions, 4:1-120. 1932. A review of the rhinoceroses with a HARRISON,J. A. ANDE. M. MANNING. 1983. Ex- description of Aphelops material from the Plio- treme carpal variability in Teleoceras (Rhinoc- cene of Texas. University of California Publi- erotidae, Mammalia). Journal of Vertebrate Pa- cations in Geological Science, 20:411-480. leontology, 3:58-64. MCBRIDE,E. F., W. L. LINDEMANNAND P. S. FREE- HATCHER,J. B. 1894. A median homed rhinoc- MAN. 1968. Lithology and petrology of the eros from the Loup Fork beds of Nebraska. Gueydan (Catahoula) Formation in south Tex- American Geologist, 13:149-150. as. University of Texas Bureau of Economic Ge- HAY, O. P. 1902. Bibliography and catalog of ology, Report of Investigations, 63:1-122. the fossil Vertebrata of North America. U.S. OSBORN,H. F. 1900. Phylogeny of the rhinoc- Geological Survey Bulletin 179, 868 p. eroses of Europe. Bulletin of the American Mu- HEISSIG,K. 1973. Die Unterfamilien und Tribus seum of Natural History, 13:229-267. der rezenten und fossilien Rhinocerotidae 1904. New Miocene rhinoceroses with re- (Mammalia). Saiugetierkundliche Mitteilungen, vision of known species. Bulletin of the Amer- Miinchen, 21(1):25-30. ican Museum of Natural History, 20:307-326. HESSE,C. J. 1943. A preliminary report on the PATTON,T. H. 1969. Miocene and Pliocene ar- Miocene vertebrate faunas of southeast Texas. tiodactyls, Texas Gulf Coastal Plain. Bulletin of Proceedings of the Texas Academy of Science, the Florida State Museum, Biological Science, 26:157-179. 14:115-226. HUNT, R. M. 1972. Miocene amphicyonids - AND B. E. TAYLOR. 1971. The Syntheto- (Mammalia, Camivora) from the Agate Spring ceratinae (Mammalia, Tylopoda, Protocerati- Quarries, Sioux County, Nebraska. American dae). Bulletin of the American Museum of Nat- Museum Novitates, 2506:1-39. ural History, 145:119-218. JEPSEN,G. L. 1966. Early Eocene bat from Wy- - AND- . 1973. The Protoceratinae (Mam- oming. Science, 154(3754):1333-1339. malia, Tylopoda, Protoceratidae) and the sys- LAURIE,A. 1982. Behavioural ecology of the tematics of the Protoceratidae. Bulletin of the greater one-homed rhinoceros (Rhinoceros American Museum of Natural History, 150:347- unicornis). Journal of Zoology, 196:307-341. 414. LEIDY,J. 1865. On the species of Rhinoceros. PAVLOW,M. 1892. Etudes sur l'histoire paleon- Proceedings of the Academy of Natural Sci- tologique des ongules. VI. Les Rhinoceridae de ences, Philadelphia, 1865, p. 94. la Russie et le developpement des Rhinoceridae 1869. On the extinct Mammalia of Dakota en general. Bulletin de Societ6 des Naturalistes and Nebraska, including an account of some de Moscou, 6:137-221. allied forms from other localities, together with PETERSON, O. A. 1906. Preliminary description a synopsis of the mammalian remains of North of two new species of the genus Diceratherium America. Journal of the Academy of Natural Marsh, from the Agate Spring Fossil Quarry. Sciences, Philadelphia, 7:1-472. Science, new ser. 24(609):281-283. S. R. M. SCHOCHAND E. MANNING. 1920. The American diceratheres. Memoir LUCAs, G., . 1981. The systematics of Forstercooperia, a of the Carnegie Museum, 7:399-476. MIOCENE RHINOCEROS FROM TEXAS 423

PROTHERO,D. R., E. MANNINGAND C. B. HANSON. TAYLOR, B. E. AND S. D. WEBB. 1976. Miocene 1986. The phylogeny of the Rhinocerotoidea Leptomerycidae (Artiodactyla, Ruminantia) and (Mammalia, Perissodactyla). Zoological Journal their relationships. American Museum Novi- of the Linnean Society of London, 87:341-366. tates, 2596:1-22. - ANDJ. M. RENSBERGER.1985. Preliminary TAYLOR, R. M. S. 1982. Aberrant maxillary third magnetostratigraphy of the John Day Forma- molars, morphology and developmental rela- tion, Oregon, and the North American Oligo- tions, p. 64-74. In B. Kurten (ed.), Teeth: Form, cene-Miocene boundary. Newsletter of Stratig- Function, and Evolution. Columbia University raphy, 15:59-70. Press, New York. - AND P. C. SERENO. 1980. Allometry and TEDFORD,R. H. 1970. Principles and practice of paleoecology of medial Miocene dwarf rhinoc- mammalian geochronology in North America. eroses from the Texas Gulf Coastal Plain. Geo- Proceedings of the North American Paleonto- logical Society of America, Abstracts with Pro- logical Convention, Pt. F, p. 666-703. grams, 12(7):504. ET AL. 1987. Faunal succession and bio- - AND - . 1982. Allometry and paleoecol- chronology of the Arikareean through Hemp- ogy of medial Miocene dwarf rhinoceroses from hillian interval (late Oligocene through late Mio- the Texas Gulf Coastal Plain. Paleobiology, 8: cene epochs), North America. In M. O. 16-30. Woodburne (ed.), Cenozoic Mammals: Their QUINN,J. H. 1955. Miocene Equidae of the Tex- Temporal Record, Biostratigraphy, and Bio- as Gulf Coastal Plain. University of Texas Bu- chronology. University of California Press, reau of Economic Geology, Publication 5516: Berkeley, California. 5-90. TROXELL,E. L. 1921. A study of Diceratherium ROSE, K. D. AND B. H. SMITH. 1979. Dental and the diceratheres. American Journal of Sci- anomaly in the early Eocene condylarth Ecto- ence, 202(10): 197-208. cion. Journal of Paleontology, 53:756-760. WAHLERT,J. H. 1976. Jimomys labaughi, a new SCHULTZ, C. B. AND C. H. FALKENBACH. 1941. geomyid rodent from the early Barstovian of Ticholeptinae, a new subfamily of oreodonts. North America. American Museum Novitates, Bulletin of the American Museum of Natural 2591:1-6. History, 79:1-105. WEEKS,A. W. 1933. Lissie, Reynosa, and Up- SKINNER, M. F. AND B. J. MACFADDEN. 1977. land Terrace deposits of Coastal Plain of Texas Cormohipparion n. gen. (Mammalia, Equidae) between Brazos River and Rio Grande. Bulletin from the North American Miocene (Barstovi- of the American Association of Petroleum Ge- an-Clarendonian). Journal of Paleontology, 51: ologists, 17:453-487. 912-926. WILSON,J. A. 1956. Miocene formations and STOCK,C. AND E. L. FURLONG. 1926. New canid vertebrate biostratigraphic units, Texas Coastal and rhinocerotid remains from the Ricardo Plain. Bulletin of the American Association of Pliocene of the Mohave Desert, California. Uni- Petroleum Geologists, 40:2233-2246. versity of California Publications in Geological WOOD, H. E., II. 1964. Rhinoceroses from the Science, 16:43-60. Thomas Farm Miocene of Florida. Bulletin of TANNER, L. G. 1969. A new rhinoceros from the the Museum of Comparative Zoology, Harvard Nebraska Miocene. Bulletin of the University University, 130:361-386. of Nebraska State Museum, 8(6):395-412. ANDA. E. WOOD. 1937. Mid-Tertiary ver- 1972. A new species of Menoceras from tebrates from the Texas Coastal Plain: fact and the Marsland Formation of Nebraska. Bulletin fable. American Midland Naturalist, 18:129- of the University of Nebraska State Museum, 146. 9(8):205-213. YATKOLA, D. AND L. G. TANNER. 1979. Brachy- o 1976. A new species of Peraceras from the potherium from the Tertiary of North America. Lower Pliocene of Boyd County, Nebraska. Pro- Occasional Papers of the Museum of Natural ceedings of the Nebraska Academy of Sciences, History, University of Kansas, 77:1-11. 86:50. MANUSCRIPTRECEIVED 10 JULY 1984 A new of Diceratherium from 1977. species REVISEDMANUSCRIPT RECEIVED 3 APRIL 1986 the Lower Pliocene (Valentinian) of Boyd Coun- ty, Nebraska. Transactions of the Nebraska The senior author provided $250 in support of Academy of Sciences, 4:121-128. this article.