Pa l e o Bio s Contributions from tbe University of California Museum of Paleontology, Berkeley No. 15 February 28, 1974 THE JHEMINGFORDIAN MAMMAL FAUNA OF THE VEDDER LOCALITY, BRANCH CANYON FORMATION, SANTA BARBARA COUNTY, CALIFORNIA. PART I: INSECTIVORA, CHIROPTERA, LAGOMORPHA, AND RODENTIA (SC1URIDAE). by J. Howard Hutchison and Everett H. Lindsay no. 15 THE HEMINGFORDIAN MAMMM, FAUNA OF THE VEDDER LOCALITY, BRANCH CANYON FORMATION, SANTA BARBARA COUNTY, CALIFORNIA PART I: INSECTIVORA, CHIROPTERA, LAGOMORPHA, AND RODENTIA (SCIUTUDAE) BY J. HOWARD HUTCHISON1 AND EVERETT H. LINDSAY2 ABSTRACT The Vedder mammal locality of Hemingfordian Land Mammal Age Is located in a non-marine tongue of the predominantly marine Branch Canyon Formation. The Branch Canyon Formation has "Temblor" Stage megainvertebrate fossils and Saucesian Stage foraminiferal fossils. Taxa from the locality include a shrew (cf. Limnoecus sp.), mole (Scalopoides sp.), bat (?Phyllostomatidae), rabbit (Hypolagus cf. H. apachensis), pika (Cuyamalagus dawsoni n. gen., n. sp.), ground squirrel (Miospermophilus sp.), and flying squir- rel (Blackia sp.). Rodents and horses to be discussed in subse- quent parts of the faunal study form the basis for the age deter- mination of the fauna. INTRODUCTION Hemingfordian mammal faunas are rare in the Pacific Coast region of North America. Most of those previously described are large-mammal assemblages. The Vedder locality provides the most diverse small-mammal Hemingfordian fauna yet known from this large area. Description of the fauna is divided into three parts; this part includes the Insectivora, Chiroptera, Lagomorpha, and sciurid Rodentia. Part II will include the eomyid and heteromyid Rodentia; Part III will include the larger mammals. Fossils described below were collected from a single locality discovered by J. G. Vedder in 1967 while mapping the marine Branch Canyon Formation west of the Cuyama Badlands. Most of the fossil vertebrates previously collected in Cuyama Valley were found in sediments of the Caliente Formation on the east and north sides of Cuyama Valley in parts of Ventura and Santa Barbara counties, California. Fossil mammals from the Cuyama Badlands were first reported by C. L. Gazin (1930). Later re- ports on the fossil fauna were made by A. E. Wood (1937), V. L. Museum of Paleontology, University of California, Berkeley, 94720. Department of Geosciences, University of Arizona, Tucson, 85721. 1. no. 15 VanderHoof (1939), J. F. Dougherty (1940), C. Stock (1947), D. E. Savage (1957), and C. A. Repen-ning and J. G. Vedder (1961). G. T. James (1963) published a comprehensive study of the Cuyama Badlands fossil vertebrate fauna based on extensive reconnais- sance and collecting, including underwater screen washing, of previously known and new localities. As noted by earlier authors and summarized by James, the Caliente Formation of Cuyama Valley contains a diverse and long-ranging fauna. James recognized mammal faunas assignable to four North American Land Mammal Ages (Hemingfordian, Barstovian, Clarendonian, and Hemphillian) in the Caliente Formation. ,. The Vedder locality is west of the Cuyama River in a tributary of Tennison Canyon, 34° 511 05" N, 119° 33" 13" W, at an eleva- tion of 3120 feet, Santa Barbara County, California. The beds containing the site are tilted and locally faulted. Fossils were recovered from a mottled, brown and green mudstone overlying a localized six-inch-thick white limestone and underlying oyster- bearing beds. The present sample from the site includes only mammals. Collections from this locality are preserved in the University of California Museum of Paleontology and United States Geological Survey at Menlo Park where they are catalogued under locality numbers V6761 and M-1090 respectively. The latter col- lection consists of only larger vertebrates and the initial dis- covery material. Small vertebrates were recovered by underwater screening of about 750 to 1000 pounds of matrix collected by University parties. One millimeter pore brass screen (24 mesh) was used. The Branch Canyon Formation is a lateral equivalent of the lower part of the Caliente Formation (Repenning and Vedder, 1961). Vedder (in letter, 1972) states, "the red bed zone from which the vertebrate material was collected contains marine invertebrate assemblages of middle Miocene age ("Temblor" Stage) west of M-1090 so that there is no apparent correlation problem...More than 1,500 feet of strata beneath the vertebrate locality con- tain middle Miocene mollusks, and the only Saucesian forams from the marine sequence north of Fox Mountain were collected about 1,400 feet below it." The Hemingfordian age of the Vedder local fauna is based pri- marily upon specific identity of several rodents with those at Hemingfordian sites in the Great Plains (especially Martin Canyon Quarry A of Wilson, 1960; see McKenna, 1965:17, regarding the mammal age assignment of Quarry A) and identity of the equids with those at other Californian localities generally considered Hemingford ian in age. Taxonomic parts of this paper were compiled separately by Lindsay (rodents) and Hutchison (other mammals). Augusta F. Lucas, staff artist of the Museum of Paleontology, drew figures 5 and 7. The authors drew the remaining figures. Terminology 2. no. 15 and abbreviations for dental characters of the Sciuridae are il- lustrated in figure 6. Terminology for other groups follows that in general usage. All measurements are given in millimeters. Abbreviation UCMP refers to University of California Museum of Paleontology. SYSTEMATICS Order INSECTIVORA Illiger, 1811 by J. H. Hutchison Family TALPIDAE Fischer von Waldheim, 1817 Subfamily TALPINAE Fischer von Waldheim, 1817 Tribe SCALOPINI Trouessart, 1879 Genus Scalopoides Wilson, 1960 Stratigraphic and geographic occurrence: Pawnee Creek Formation, Logan County, Colorado; Colter Formation, Teton County, Wyoming; Monroe Creek Formation, Shannon County, South Dakota; Deer Butte Formation, Butte Creek Volcanic Sandstone, and Juntura Formation, Malheur County, Oregon; Beatty Butte, Harney County, Oregon; Guano Ranch, Lake County, Oregon; Branch Canyon Formation, Santa Barbara County, California. Age: Arikareean, Hemingfordian, Barstovian and Clarendonian. Scalopoides sp. Material: UCMP loc. V6761 - UCMP 96281, damaged left M ; UCMP 96282, damaged right humerus lacking distal end, head, and parts of tuberosities and pectoral crest; UCMP 96283, left metacarpal V; UCMP 96284, damaged metacarpal IV; UCMP 96285-96286, two prox- imal phalanges of manus; UCMP 96287, unciform; UCMP 96388, dam- aged ungual phalanx of manus. Description: A damaged M (Fig. la) resembles the Barstovian Scalopoides ripafodiator Hutchison (1968) from Oregon in propor- tions and relatively small metaconule. M of £. isodens Wilson (1960) from the Hemingfordian Martin Canyon Quarry A, Colorado, is not known. The humerus (Fig. lb) resembles Scalopoides in major features and S_. ripafodiator more closely than ^. isodens. It is slightly smaller and proportionately more slender and delicate than that of ^. ripafodiator. 3. no. 15 Fig. 1. Scalopoides sp. a. UCMP 96281, damaged left M , occlusal view. b. UCMP 96282, damaged left humerus, anterior (right) and posterior views, c. UCMP 96283, left metacarpal V, lateral and ventral views, d. UCMP 96285, proximal digit of manus, dorsal and lateral views. Scale lines = 1 mm. 4. no. 15 Metacarpal V (Fig. lc) and metacarpal IV are not directly com- parable to the named species of Scalopoides. In comparison with metacarpal III of j3. ripafodiator (Hutchison, 1968, fig. 58), the Vedder specimens are perhaps more slender, but are not as long as in Neurotrichus gibbsii (Baird). Two proximal phalanges of the manus (Fig. Id), probably from digits II, III or IV, are about as squat as in _S. isodens and thus shorter than some post-Barstovian specimens referred to Scalopoides (Hutchison, 1968;78, Table 17). UCMP 96287 is the first unciform assignable to Scalopoides but at present sheds little light on the relationships of this species. Discussion: General agreement in size, functional grade, and comparative morphology of the Vedder talpid specimens indicates they represent only one species. The overall slenderness of ele- ments indicates a mole less specialized postcranially than the Barstovian Scalopoides ripafodiator from Oregon, which in turn is less specialized than the early Hemingfordian _S. isodens from Colorado. Robustness of the proximal phalanges is probably not significant considering the wide range in proportions in the small samples available for comparison and lack of knowledge of which digit they represent. The Vedder Scalopoides as presently known would provide a temporally suitable ancestor for _S. ripa- fodiator but not for J3. isodens. It indicates the presence of at least two lineages of Scalopoides in the Hemingfordian. Fig. 2. Cf. Limnoecus sp. a. UCMP 82147, right M^ labial, lingual, and occlusal views, b. UCMP 82145, left I , lingual and posterior views. Scale line = 1 mm. 5. no. 15 Family SORICIDAE Fischer von Waldheim, 1817 Subfamily LMNOECINAE Repenning, 1967 Genus Limnoecus Stirton, 1930 Stratigraphic and geographic occurrence: Pawnee Creek Formation, Logan County, Colorado; Split Rock Formation, Fremont County, Wyoming; Valentine Formation, Cerry County, Nebraska; Ogallala Formation, Trego County, Kansas; Butte Creek Volcanic Sandstone and Rome beds, Malheur County, Oregon; Barstow Formation, San Bernardino County, California; Caliente Formation, Ventura County, California; Branch Canyon Formation, Santa Barbara County, Cali- fornia. Age: Hemingfordian, Barstovian, Clarendonian, and Hemphillian. Cf. Limnoecus sp. Material: UCMP loc. V6761 - UCMP 82145, heavily worn left I1; UCMP 82147, nearly complete but damaged and heavily worn right M,; UCMP 82146, left dentary fragment with broken Mp UCMP 96078, 96080, broken tips of two lower anteromost incisors; UCMP 96079, Mi fragment. Description: A single relatively complete Mj (Fig. 2a) is dam- aged and heavily worn but compares favorably with the M-^ of Lim- noecus tricuspis Stirton in preserved details and size. The re- entrant valley between protoconid and hypoconid emerges labially well above the cingulum. Anterior and labial cingula are well developed as in Limnoecus, unlike Antesorex Repenning. The den- tary fragment includes the mental foramen below and anterior to the tip of the Ml protoconid as in Limnoecus, Angustidens Repen- ning, and Antesorex. I (Fig. 2b) has a nonbifid principal cusp, narrow basilabial cingulum, and bicuspid (?) heel with labial cusp predominating in wear. The unworn heel may have had a single, transversely elong- ate cusp, but if so, wear tended to bifurcate this cusp. Fragments of two lower incisors show that at least one promin- ent cusp lies posterior to the tip. Discussion: In combination, position of the mental foramen, structure and delicacy of il and M^, and small size exclude the Vedder shrew from all Miocene genera except Limnoecus (see Repen- ning, 1967). Although in agreement with Limnoecus in the above characters, poor preservation and lack of more diagnostic parts preclude definite assignment of the Vedder shrew to Limnoecus. Only one species seems indicated in the sample. 6. no. 15 Order CHIROPTERA Blumenbach, 1779 by J. H. Hutchison Suborder MICROCHIROPTERA Dobson, 1875 Family PHYLLOSTOMATIDAE Coues and Yarrow, 1875 Stratigraphic and geographic occurrence: Honda Formation, Colom- bia; Caliente Formation, Ventura County, California; ?Branch Can- yon Formation, Santa Barbara County, California; Pleistocene and Recent of North and South America. Age: Hemingfordian to Recent in North America. ?Phyllostomatidae incertae sedis Material: UCMP loc. V6761 - UCMP 82144, left lower canine; UCMP 80324, edentulous and incomplete right dentary. Description: Canine crown (Fig. 3a) consists of a single hemi- conical cusp, flattened posteriorly, and centered over the anter- ior moiety of the root. Labial and lingual cingula rise anterior- ly. Two small cuspids, central and lingual, surmount the poster- ior cingulum; the posterolabial corner of the cingulum is either excavated or removed by shear. The root is robust, straight, and ovate to circular in cross-section. The dentary (Fig. 3b) consists of two fragments found together in the same sieving screen and assumed to belong to one individ- ual, although no contact is preserved. Fragments of roots are preserved in several alveoli. The anterior segment is bounded by breaks at the level of the canine and M2. Four alveoli between Mj^ and canine alveoli are interpreted to have held a double-root- ed P> and either a single-rooted P, and reduced single-rooted P3 or a double-rooted P2 or P~; the former condition seems more like- ly considering the lack of alveolar coalescence and typical phyl- lostomatid anterior root reduction during antemolar shortening. Dental homologies will thus be considered as I?, CI, P2, P3, P4, M1-M3 (following Miller, 1907). The dentary is curved both ver- tically and horizontally and has a large oblique symphyseal scar. A mental foramen exists below P„ and a small foramen is present just posterior to the symphysis. Discussion: Dentary and canine may represent more than one taxon; however, the two specimens indicate comparable-sized individuals and will be treated as one species. Among Microchiroptera available for comparison in the Univer- sity of California Museum of Vertebrate Zoology at Berkeley, members of the Phyllostomatidae most closely approach the charac- 7. no. 15 Fig. 3. ?Phyllostpmatidae incertae sedis. a. UCMP 82144, left lower canine, labial, posterior, lingual, and occlusal views. b. UCMP 80324, right dentary fragments, lingual, dor- sal, and labial views. Scale line = 1 mm. ters of the fossils. Of the recent genera resembling the fossils (Chilonycteris Gray, Fteronotus Gray, Lonchorhina Tomes, Macro- phyllum Gray), no genus is identical to the fossil material, but Macrophyllum macrophyllum (Wied) is similar in general shape of the dentary, position and expanse of symphyseal. scar, dental for- mula, and alveolar positions. In contrast to the fossils, the canine of M. macrophyllum is longer than wide in occlusal dimen- sions and possesses only one cusp on the posterior cingulum. While the Vedder bat certainly represents a new species, the fragmentary nature of these specimens and possibility of future work at the site advise deferring assignment of a new name. 8. no. 15 Order LA.GOMORPHA Brandt, 1855 by J. H. Hutchison Family LEPORIDAE Gray, 1821 Genus Hypolagus Dice, 1917 Hypolagus apachensis Gazin, 1930 Stratigraphic and geographic occurrence: Known only from Cali- fornia - Caliente Formation, Ventura County; Mint Canyon Forma- tion, Los Angeles County; ?Branch Canyon Formation, Santa Barbara County. Age: Hemingfordian, Barstovian, and Clarendonian. Hypolagus cf. H. apachensis Material: UCMP loc. V6761 - UCMP 82482, fragment of right dent- ary with damaged Po-P, . Description: Labial side of dentary (Fig. 4a) below and anterior to Po is fenestrated. Lingually the lower incisor root termin- ates above the level of fenestration and extends posterad to the. anterior moiety of P3. P3 (Fig. 4b, c) is elongate (this is ex- aggerated due to oblique breakage of the crown). Deep cement- filled labial and lingual hypostriae separate talonid from tri- gonid and persist to base of exposed crown. A broad shallow an- terolabial hypostria appears to lack cement although this may have been lost after death. Discussion: Fragmentary nature of the Vedder leporid prevents definitive taxonomic assignment. Combination of elongate P3 and shallow anterolabial hypostria are characteristic of Dawson s (1958:38) Archaeolagus group (A. ennisianus (Cope) and A. macro- cephalus (Matthew), but a similar degree of development is also seen in some species of Hypolagus¦ Persistent internal hypostria on P, is present in some specimens of H. apachensis (e.g., Fig. 25c and discussion, p. 17, of Dawson, 1958). Dawson (1958:63) considered high posterior termination of the lower incisor as typical of Hypolagus, and, in combination with tooth morphology assignment of the Vedder leporid to Hypolagus is more reasonable. The Vedder Hypolagus, as known, is in suitable chronologic and geographic position for ancestry to other known H. apachensis, to which it is tentatively referred. 9. no. 15 Fig. 4. Hypolagus cf. H. apachensis, UCMP 82482, right den- tary fragment with ^-Pa- a* Labial view with arrows indica- ting angle of occlusal views. b. Occlusal view of P.,-P, . c. Occlusal view of P, perpendicular to long axis of tooth. Scale lines = 1 mm. Family OCHOTONIDAE Thomas, 1897 Cuyamalagus dawsoni Hutchison, n. gen., n. sp. Holotype: UCMP 82151, three lower right molariform teeth. Hypodigm: Type and UCMP 82149, 82150, 80327, three incomplete upper molariform teeth. Type locality and age: UCMP locality V6761, Vedder locality, Branch Canyon Formation, Santa Barbara County, California. Hem- ingfordian. Known only from type locality. Diagnosis: Cheek teeth hypsodont but rooted. Upper molariform teeth with concave labial and convex lingual walls indicating unilateral hypsodonty, distinct crescentic valleys present, hy- postriae cross about one third or less of occlusal width in early wear but fade out before reaching lingual base of crowns, antero- lophs wider than posterolophs. Lower molars with talonids and trigonids separate to near base of crowns, talonids nearly as transversely broad as trigonids, posterior wall of trigonids slightly protruded, talonids with central anterior protrusion, posterolophids persistent 20% to 40% of depth, talonids becoming equal in length to trigonids with loss of posterolophids during wear. From Cuyama - Cuyama Valley plus Greek lagos, hare. Specific name after Mary R. Dawson in recognition of her work on fossil lagomorphs. 10. no. 15 Description: Molariform upper teeth show marked transverse cur- vature (Fig. 5a, b). UCMP 80327 shows two small labial roots and one massive lingual root. All specimens have anteroloph wider than posteroloph, distinct hypostriae, and crescentic valleys. Transverse depth of lingual hypostriae varies from fairly deep and close to crescentic valley in little-worn condition (Fig. 5a) to shallow (Fig. 5b, c) in more worn specimens. Crescentic val- leys and hypostriae are cement-filled. The arrow in Figure 5b indicates terminus of the hypostria. Fig. 5. Cuyamalagus dawsoni Hutchison, n. gen., n. sp. a. UCMP 82149, right upper molariform tooth (M2?), occlusal, lingual, and posterior views. b. UCMP 80327, right upper molariform tooth, occlusal and anterior views. c. UCMP 82150, left upper molariform tooth fragment, occlusal view. d-f. UCMP 82151, holotype. d. Right P4 or M-, , occlusal, lingual, and labial views. e. Right M2, occlusal, lingual and labial views. f. Lower portion of right P4 or Mj^, occlusal cross-section and lin- gual view. The small dots along the side of the lower teeth rep- resent levels at which measurements were taken. Scale line = 1 mm. 11. no. 15 Three lower right molariform teeth appear to represent a single individual and unlike the upper teeth, were previously catalogued under, a single number (probably indicating a single sieving screen or bag sample) and thus chosen for the type speci- men. Two specimens (Fig. 5d, e) are complete enough to preserve their total heights. Both appear to be in early wear with hypo- conulids still present and roots (in third specimen also) incom- plete but closing. Trigonids and talonids on all lower teeth are separate to about the basal third of the tooth (75% to 90% of tooth height) but are amply joined by cement (mostly chipped away by breakage). Folds setting off hypoconulids from talonids are also cement-filled. Trigonids and talonids are subequal in width but talonids are longer in early wear, tending to become nearly equal to trigonids after loss of hypoconulids (postero- lophids) with wear (Table 1). Labial margins of trigonids and talonids are persistently more angular than the lingual margins. Anterior protrusion of the talonid is well developed in early wear stages. A small protrusion labial to the center of the pos- terior wall of the trigonid is indicated in all three teeth. In- dications of enamel thinning or loss are discernible on the lab- ial sides of two specimens (Fig. 5d, f). Strong curvature and lesser root development on one specimen (Fig. 5e) suggests that this tooth is an M~. Discussion: Of the twenty named genera of ochotonids, only six (Amphilagus Pomel, Titanomys von Meyer, Desmatolagus Matthew and Granger, Hesperolagomys Clark, Dawson and Wood, Russellagus Storer, and Gripholagomys Green) retain roots and well developed molar crescents, but all these also have distinctly narrower tal- onids than trigonids (a primitive feature in ochotonids). A wide talonid is characteristic of more advanced ochotonids with ever- growing teeth. While it is possible to speculate on the phyletic position of Cuyamalagus, lack of anterior premolars and jaw material leave too much latitude for profitable comparisons. Of the North Amer- ican genera (see Green, 1972), only ?Desmatolagus (Hemingfordian) and Gripholagomys (Arikareean and Hemingfordian) are early enough and structurally primitive enough to be possible ancestral forms or close relatives of ancestral forms of Cuyamalagus. Hespero- lagomys (Barstbvian and Clarendonian) and Russellagus (Barstovian) retain a primitively narrow talonid and are thus unlikely descen- dents of Cuyamalagus. Oreolagus McGrew occurs in the early Hem- ingf ordian (Green, 1972) and later but is too advanced in hypso- donty for close relationship with Cuyamalagus. While relation- ships of all North American Miocene ochotonid genera are largely unknown, the known diversity of these genera (six) is rapidly approaching that of Eurasia (eight in Europe and two in Asia - see Green, 1972, and Dawson, 1967). 12. no. 15 TABLE 1 Measurements of Cuyamalagus dawsoni, n. gen., n. sp. Upper Molariform Teeth Length Width of anteroloph Width of posteroloph UCMP 82149 1.54 2.57 2.20 Lower Molariform Teeth - UCMP 82151 (Type) Length trigonid (top) Length talonid (top) Length trigonid (middle) Length talonid (middle) Width trigonid (top) Width talonid (top) Width trigonid (middle) Width talonid (middle) Height (base of enamel on labial side to top of trigonid) Length of crown incomplete specimen 2.17 UCMP 80327 1.55 P4 or Mx M2 P4 or Mj* 0.99 0.94 - - 1.04 1.32 - - 1.10 1.13 1.04 0.96 1.17 1.03 1.95 1.93 - - 1.77 1.90 - - 2.13 2.10 2.18 1.72 1.84 2.16 6.72 5.85 - - 2.33 2.05 13. no. 15 RODENTIA Bowdich, 1821 by E. H. Lindsay Family SCIURIDAE Gray, 1821 Subfamily SCIURINAE Baird, 1857 Genus Miospermophilus Black, 1963 Stratigraphic and geographic occurrence: Pawnee Creek Formation, Logan County, Colorado; Split Rock Formation, Fremont County, Wyoming; Branch Canyon Formation, Santa Barbara County, and Bar- stow Formation, San Bernardino County, California. Age: Hemingfordian and Barstovian. Miospermophilus sp. indet. Material: UCMP loc. V6761 - UCMP 80326, right dP^; UCMP 82138, left M,; UCMP 82139, right P4; UCMP 82141, left M1 or M2; UCMP 82142, right P4, and UCMP 82143, P3.- Description: P - A small, round, peglike premolar is assigned to this taxon. The premolar is about 0.3 mm wide. dP - UCMP 80326 (Fig. 7b) has the anterocone and part of the anterior cingulum removed by breakage. Paracone and metacone are subequal, smaller than the prominent protocone. Protoconule and metaconule are. absent from the high, narrow protoloph and meta- loph uniting paracone and metacone (respectively) with the anter- ior and posterior sides of the protocone. Posterior cingulum is high and narrow, connecting the basal metacone and posterior pro- tocone. An incipient hypocone is marked by an expansion on the internal posterior cingulum. The anterior cingulum is low where it joins the anterior protocone. A small mesostyle is placed between and slightly lateral to the paracone and metacone; a low lophule joins the mesostyle and posterior basal paracone. The central basin is wide, slightly rugose, and open laterally pos- terior to the mesostyle. The posterior basin is narrow, bounded by the metaloph and posterior cingulum. P - A badly corroded cheek tooth is probably F*. Its occlus- al outline is elongate oval, width greater than length; the an- terolabial margin of the tooth is removed by breakage. The pro- tocone is the most prominent cusp; paracone and metacone are sub- equal. A small metaconule is present on the metaloph; a proto- conule is absent. The protocone is directly lingual to the para- cone, and these cusps are connected by the narrow protoloph. Metaloph connects metacone and metaconule, it terminates short of the protocone. Anterior and posterior cingula were apparently present, but are largely obliterated by corrosion. Parastyle 14. no. 15 —anterior—? upper lower Fig. 6. Dental terminology for sciurid teeth. 1 - Paracone. 2 - Metacone (-id). 3 - Protocone (-id). 4 - Hypoconid. 5 - Entoconid. 6 - Protoconule. 7 - Metaconule. 8 - Mesoconid. 9 - Parastyle. 12 - Protolophule. 13 - Metalophule. 14 - Meta- lophid. 15 - Ectolophid. 17 - Mesostyle (-id). 19 - Anterior cingulum. 20 - Posterior cingulum. 22 - Anterior inner valley. 23 - Central valley. 24 - Posterior inner valley. 25 - Talonid basin. Fig. 7. Sciurid rodents, a-d. Miospermophilus sp. a. UCMP 82141, left M1 or M2, occlusal view. b. UCMP 80326, right dP4, occlusal view. c. UCMP 82139, right P^, occlusal view, d. UCMP 82139, right P4, labial view. e-f. Blackia sp. e. UCMP 82140, left P*, occlusal view. f. UCMP 82137, left M3, occlusal view. Scale line = 1 mm. 15. no. 15 and mesostyle were small or absent but breakage in that area pro- hibits a precise determination. The central basin is relatively narrow, but wider than anterior and posterior basins. M or tr- A well preserved and slightly worn tooth (Fig. 7a) is W- or M . Precise identification of this isolated tooth is impossible without associated teeth. Occlusal outline is sub- quadrate, width greater than length. Protocone, the most promi- nent cusp, is elongated anteroposteriorly. Metacone is lower and smaller than the paracone. An incipient hypocone is indicated by expansion of the lingual part of the posterior cingulum. A small metaconule is indicated by expansion of the metaloph; protoconule is absent. A minute parastyle is present on the labial margin of the anterior cingulum. A small mesostyle is present between and labial to the paracone and metacone. The protoloph is narrow, directed transversely, connecting paracone and anterior basal margin of the protocone. Metaloph is narrow, and connects the metacone and metaconule. A low and faint ridge continues lingual from the metaconule to the posterior base of the protocone. The anterior cingulum is narrow and relatively low, connecting the parastyle and anterior base of the protocone. Parastyle does not join the paracone. The posterior cingulum is low and narrow pos- terior to the metacone and high lingually where it joins the pos- terior side of the protocone. The central basin is relatively wide, slightly rugose, and closed by the mesostyle. The anterior basin is narrow, slightly rugose, and constricted labially. The posterior basin is very narrow, slightly rugose, and constricted both labially and lingually. A faint accessory loph is directed labially from the protocone. P4 - A slightly corroded but relatively unworn tooth (Fig. 7d) is identified as a P.. Occlusal outline of this tooth is obovate, length greater than width; anterior width is less than posterior width. The posterolingual corner of the tooth is subangular. Metaconid and protoconid are prominent, with protoconid lower and placed slightly more posteriorly than the metaconid. Hypoconid is large; entoconid is small but distinct. Mesoconid is absent from the low and narrow ectolophid. Posterolophid is low and narrow, slightly arcuate between the hypoconid and entoconid. A low, rounded mesostylid is present on the lingual border between the metaconid and entoconid. Talonid basin is shallow and slightly rugose. Valley between the metaconid and protoconid is steep and narrow, with no indication of a loph connecting the cusps. M3 - The posterior lower molar is represented by part of a broken tooth. Its occlusal outline is apparently an elongate oval with a rounded posterior margin. Metaconid is prominent and connected to the hypoconid by a high, continuous, wide, arcuate posterior cingulum. Talonid basin is shallow, slightly rugose, and closed posterolingually by the posterior cingulum. Measure- ments are given in Table 2. 16. no.. 15 TABLE 2 Measurements of sciur id rodents Miospermo] ph i lus sp. Length UCMP 82143 P3 0.3 ~~ UCMP 80326 dP4 - - UCMP 82142 p4 1.56 UCMP 82141 M1 or M2 1.95 UCMP 82139 P4 1.88 UCMP 82138 M3 (2.80) Blacfcia sp. UCMP 82140 P4 1.04 UCMP 82137 M3 (1.50) Width 0.3 1.77 1.95 2.34 1.62 1.25 (1.50) Measurements in parentheses ( ) are approximate. Discussion: The teeth of Miospermophilus sp. of the Vedder fauna are larger than those of Miospermophilus bryanti (Wilson) from Martin Canyon Quarry A, Logan County, Colorado, and Miospermophi- lus sp. from the Barstow Formation; they are about the size of Miospermophilus wyomingensis Black from the Split Rock Formation in Fremont County, Wyoming (Black, 1963). Miospermophilus of the Vedder fauna differs from M. wyomingensis in having a broader protocone, less prominent transverse lophs, and a more distinct entoconid. Observed differences between the Vedder Miospermo- philus sp. and M. wyomingensis are slight. Subfamily PETAURISTINAE Simpson, 1945 Genus Blackia Mein, 1970 Stratigraphic and geographic occurrence: Branch Canyon Formation, Santa Barbara County, California; Sansan, La Grive, and Soblay in France; Anwil, Switzerland; Wolfersheim, Germany. Age: Hemingfordian, North America; middle Miocene to Pliocene, Europe. Description: P - A small, transversely elongate, relatively un- worn tooth (Fig. 7e) is considered P . Its prominent protocone 17. is high and narrow. Metacone is lower than the paracone and is placed slightly labial to it. A large parastyle is anterior and slightly lingual to the paracone. Protoconule, metaconule, hypo- cone, and mesostyle are absent. Protoloph and metaloph are nar- row and high, joining anterior and posterior sides of the proto- cone. Anterior and posterior cingula are low, terminating lin^ gually at the base of the protocone, terminating labially at the parastyle (anterior cingulum) and at the posterior side of the metacone (posterior cingulum). The central basin is distinctly rugose, wide, slightly cup-shaped, and closed labially by a low ridge (mesostylar ridge of Mein, 1970) connecting paracone and metacone. Anterior and posterior valleys are narrow, and open where cingula are very low. A distinct but small accessory loph is directed labially from the central area of the protocone. M3 - An isolated M3 (Fig. 7f) is relatively unworn but its an- terolingual corner is missing. The occlusal outline is elongate oval with its long axis directed obliquely. Protoconid is set off from the anterior cingulum and the ectolophid. Hypoconid is embedded in the high arcuate posterior cingulum near the rounded posterolingual corner. Incipient entoconid and mesostylid are indicated by slight swelling on the posterior cingulum along the labial border. Mesoconid and metalophid are absent. Talonid basin is wide, distinctly rugose, slightly cup-shaped, and closed except for the groove separating the posterior cingulum from pos- terior side of the protocone. Measurements are given in Table 2. Discussion: These two isolated cheek teeth are referable to the Petauristinae because of their rugose occlusal surface, reduced anterior cingulum, and distinctive groove on M_ between the an- terior cingulum and protoconid. They are placed in the genus Blackia because of their small size, and simple molar structure (lacking the hypocone and conules in upper cheek teeth and lack- ing the mesoconid in lower cheek teeth). Two species of Blackia have been named. Blackia miocaenica is known from la Grive, Sansan, and Anwil in western Europe. Blackia miocaenica and Blackia sp. from the Branch Canyon Formation are smaller than B_. wolfersheimensis. B_. miocaenica was characterized by Mein (1970) as having a triangular F4 with a high an8 narrow protocone and a strong parastyle. Blackia sp. of the Vedder fauna is very close to B. miocaenica, differing from that species in its slightly wider and more robust P . The North American record of Blackia probably represents a new species but specific assignment is withheld until more diagnostic characters can be defined. 18. no. 15 LITERATURE CITED Black, C. C, 1963. A review of the North American Tertiary Sciuridae. Mus. Comp. Zool. (Harvard) Bull. 130(3): 109- 248. Dawson, M. R., 1958. Later Tertiary Leporidae of North America. Kansas Paleontol. Contrib. Vertebrata 6: 1-75. _____________, 1967. Lagomorph history and the stratigraphic record. JCn: Essays in Paleontology and Stratigraphy. Raymond C. Moore Commemorative Volume. Univ. Kansas Dept. Geol. Special Publ. 2: 287-316. Dougherty, J. F., 1940. A new Miocene mammalian fauna from Caliente Mountain, California. Carnegie Inst. Washington, Publ. 514: 109-143. Gazin, C. L., 1930. A Tertiary vertebrate fauna from the upper Cuyama drainage basin, California. Carnegie Inst. Washing- ton, Publ. 404: 55-76. Green, M., 1972. Lagomorpha from the Rosebud Formation, South Dakota. J. Paleontol. 46(3): 377-385. Hutchison, J. H., 1968. Fossil Talpidae (Insectivora, Mammalia) from the Later Tertiary of Oregon. Univ. Oregon Mus. Nat. Hist. Bull. 11: 1-117. James, G. T., 1963. Paleontology and nonmarine stratigraphy of the Cuyama Valley Badlands, California. Univ. Calif. Publ. Geol. Sci. 45: 1-154. McKenna, M. C, 1965. Stratigraphic nomenclature of the Miocene Hemingford Group, Nebraska. Amer. Mus. Novitates 2228: 1- 21. Mein, P., 1970. Les sciuropteres (Mammalia, Rodentia) Neogenes d'Europe Occidentale. Geobios 3(3): 7-77. Miller, G. S., Jr., 1907. The families and genera of bats. U.S. Nat. Mus. Bull. 57: i-xyii, 1-282, 14 pi. Repenning, C. A., 1967. Subfamilies and genera of the Soricidae. U. S. Geol. Surv. Prof. Paper 565: 1-73. _______________, and J. G. Vedder, 1961. ' Continental vertebrates and their stratigraphic correlation with marine mollusks, eastern Caliente Range, California. U. S. Geol. Surv. Prof. Paper 424C: C235-C239. Savage, D. E., 1957. Age of the Caliente Formation, Caliente Range, California. Geol. Soc. Amer. Bull. 68(12): 1845. Stock, C, 1947. A peculiar carnivore from the Cuyama Miocene, California. Southern Calif. Acad. Sci. Bull. 46: 84-89. VanderHoof, V. L., 1939. New evidence as to the age of the Cuyama beds, California. Geol. Soc. Amer. Bull. 50: 1974. Wilson, R. W. , 1960. Early Miocene rodents and insectivores from northeastern Colorado. Univ. Kansas Paleontol. Contrib. Vertebrata 7: 1-92. Wood, A. E., 1937. Additional material from the Tertiary of the Cuyama Basin of California. Amer. J. Sci. 33: 29-43. 19.