THE VERTEBRATE FAUNAS OF THE PLIOCENE RINGOLD FORMATION, SOUTH-CENTRAL WASHINGTON by Eric Paul Gustafson BULLETIN No. 23 o/ the MUSEUM OF NATURAL HISTORY University o/ Oregon Eugene, Oregon March, 1978 The Bulletin of the Museum of Natural History of the University of Oregon is published to increase the knowledge and understanding of the natural history of Oregon. Original articles in the fields of Archaeology, Botany, Ethnology, Geology, Paleontology, and Zoology appear irregularly in consecutively numbered issues. Contributions arise primarily from the research programs and collections of the Museum of Natural History and the Oregon State Museum of Anthropology. However, in keeping with the basic purpose of the publication, contributions are not restricted to these sources and are both technical and popular in character. Communications concerning manuscripts or purchase of copies of the Bulletin should be addressed to the Museum of Natural History, University of Oregon, Eugene, Oregon 97403. A list of previous issues of the Bulletin and prices is printed inside the back cover. THE VERTEBRATE FAUNAS OF THE PLIOCENE RINGOLD FORMATION, SOUTH-CENTRAL WASHINGTON THE VERTEBRATE FAUNAS OF THE PLIOCENE RINGOLD FORMATION, SOUTH-CENTRAL WASHINGTON by ERIC PAUL GUSTAFSON Bulletin No. 23 Museum of Natural History University of Oregon Eugene, Oregon March 1978 j i Dedicated to the late Willis E. Fry, Dr. Claude W. Hibbard, and to my father, Dr. Paul V. Gustafson, all of whom should have lived to see the publication of this report. 1 The Vertebrate Faunas of the Pliocene Ringold Formation, South-Central Washington by ERIC PAUL GUSTAFSON ABSTRACT The vertebrate fauna of the upper Ringold Formation at the White Bluffs, south-central Washington, has been the subject of several short papers since its discovery in the late 19th century. Additional information from more recent collections, which include remains of many small mammals, expands the knowledge of this White Bluffs local fauna. A rhinoceros mandible from the lowest exposures provides evidence of a second, distinctly earlier fauna, the River Road local fauna. Fossiliferous localities can be correlated by reference to two key beds, the White Bluffs tuff and the Tay- lor Flat conglomerate, both of which are widely exposed. Vertebrate fossils are most commonly preserved in stream channel conglomerates and less frequently in silt deposits. The White Bluffs local fauna includes three genera of fresh-water snails, two genera of fish (]ctalurus and Archoplites), unidentified anuran amphibians and small reptiles, two or three genera of turtles ( Clemmys, Chrysemys?, and possibly Testudo), and 25 genera of mammals. Among the mammalian genera are a mole (Scapanus), two leporids (Hypolagus and Nekrolagus), nine rodents (Paenemarmota or Marmota, Sper- mophilus?, Ammospermophilus, Thomomys, Castor, Dipoides, Peromyscus, Neotoma, and Ophiomys), an edentate (Megalonyx), six carnivores ( Canis, Borophagus, Ursus, Trigonictis, F elis, and a machairodont), a proboscidean (Mammut), an equid (Equus), and four artiodactyls (Platygonus , Megatylopus, Hemiauche- nia, and Bretzia). New species are: Hypolagus ringoldensis, a leporid probably derived from H. oregonen- sis Shotwell; Spermophilus? russelli, a large ground squirrel of uncertain affinities; Ammospermophilus hanfordi, a large antelope ground squirrel; Peromyscus nosher, a deer mouse; Ophiomys mcknighti, a mi- crotine closely related to 0. magilli Hibbard from the Sand Draw local fauna of Nebraska but more primi- tive than the latter; and Megalonyx rohrmanni, a small ground sloth similar in form to specimens from Ha- german, Idaho. The White Bluffs local fauna is early Blancan (Pliocene) in age. The faunal assemblage is most similar to that of the Hagerman local fauna of Idaho but is probably slightly older. The predominance of browsing forms among the large mammals (particularly Bretzia, Megalonyx, and Platygonus) indicates that the Ringold flood plain supported considerable riparian forest and open wood- land, environments extremely restricted in eastern Washington today. Savanna or open grassland, suggested by the presence of Equus and possibly by Megatylopus, may have been important away from the streams, but the absence so far of antilocaprids suggests that these habitats were not important near areas of stream depo- sition. The River Road local fauna, containing T eleoceras and ? M egatylopus, is probably late Hemphillian in age. INTRODUCTION The vertebrate fauna of the Ringold Formation has been neglected since publication of its discovery in 1893. Enough well-documented material has now come to light to allow the beginnings of a thorough study. Because of the desirability of detailed strati- graphic analysis of carefully dccumented specimens, this study was restricted primarily to the collections from the Ringold in the Thomas Burke Memorial Washington State Museum. These specimens were collected by V. S. Mallory and others in 1963 and by myself and others commencing in 1967, especially during the summers of 1969 and 1970. The collection has been further augmented by donations of speci- mens collected by Willis E. Fry over a period of sev- 2 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 eral years. Undescribed specimens in the Burke Museum from sediments in other areas of the Columbia Basin which may be correlative with the type Ringold Formation at the White Bluffs are not included in this study. Collections from the White Bluffs exist in other insti- tutions, including the University of California Mu- seum of Paleontology, the Los Angeles County Muse- um of Natural History, the National Museum of Natural History, and the Conner Museum at Wash- ington State University. The White Bluffs fauna is similar in its general composition to the classic early Blancan fauna from Hagerman, Idaho; however, there are a number of important differences. A somewhat different envi- ronment has been sampled, so that some animals which left few remains at Hagerman are common in the White Bluffs. The additions to the described early Blancan fauna of North America, the geographic location in the far northwest of the contiguous U. S., and the unusual abundance of forms, such as the cervid and ground sloth, which are considered to be browsers or indicators of forest habitat, contribute to the importance of this locality and fauna. LOCATION, TOPOGRAPHY, AND CLIMATE The type exposures of the Ringold Formation are located in the south central part of the State of Wash- ington, near the center of the Columbia Basin and within a broad local downwarp known as the Pasco Basin. These fossiliferous sediments form a series of cliffs ( the White Bluffs) averaging about 500 feet high along the east bank of the Columbia River (map, Fig. 1) . The bluffs start several miles north of the city of Richland and extend northward for about 25 miles. The type exposures range in altitude from 340 feet ( about the level of the Columbia River) to about 960 feet above mean sea level. Although the Ringold Formation is mainly ex- posed east of the Columbia River, well logs show that Ringold sediments are present beneath the surface of most of the Pasco Basin. They are consistently cov- ered by glaciofluvial sediments in the basin area be- tween the south end of the bluffs and Wallula Gap and in the area of the Atomic Energy Commission's Han- ford Reservation, which occupies much of the Pasco Basin west of the Columbia. The Pasco Basin is bor- dered on all sides by basalt of the Miocene-early Plio- cene Columbia River Group. The climate over the Ringold area and most of the Columbia Basin is semiarid. The yearly rainfall of less than ten inches supports sagebrush and bunch- grass. The ·6luffs themselves are sparsely vegetated, and steep slopes are often bare of vegetation. Few permanent streams other than the throughflowing Columbia, Yakima, and Snake Rivers are present. A few minor streams, whose flow has recently been augmented by irrigation, have cut narrow canyons into the White Bluffs, but much of the rainfall evapo- rates or escapes into the permeable sediments of the Ringold Formation and overlying deposits. HISTORY OF INVESTIGATIONS Published information on the White Bluffs sedi- ments began with a report by I. C. Russell ( 1893) , who referred the sediments to the "John Day Sys- tem" of Oregon, published a stratigraphic section, and reported the presence of fossil bones. Merriam and Buwalda (1917) separated the Ringold Forma- tion from other central Washington sediments on the basis of lithology and a fauna which at the White Bluffs contained an "advanced or specialized species of Equus" which they concluded could represent late Pliocene or Pleistocene time. They chose the White Bluffs sediments as their type section. Further faunal descriptions by McKnight (1923 ), Stock (1925) , Beck (1936, 1937, 1940, 1949), Strand and Hough (1952), and Taylor (1966), provided minor addi- tions to the fauna described by Merriam and Buwal- da. Fry and Gustafson (1974) described the cervid Bretzia pseudakes from specimens in the Burke Museum White Bluffs collection, and listed a fauna which includes most of the forms described in this paper. Gustafson (1977) noted the presence of Teleo- ceras in the Taylor Flat conglomerate. Recent work on the geology of the Ringold Forma- tion includes papers by Newcomb (1958), which re- defined the formation and described the sediments in some detail, by Brown and McConiga ( 1960) , which described deformation and plotted the position of several key beds, and by Newcomb, Strand, and Frank (1972) which includes a geological map of the central Pasco Basin area, including most of the White Bluffs. Culver (1937) proposed an extension of the use of the name Ringold Formation to include sediments over wide areas of the Columbia Basin, particularly in the Othello and Quincy Basins to the north and in the vicinity of Delight (McChesney Springs) to the east. Fossils from the Othello and Quincy Basins seem to be Blancan in age, thus agreeing with Culver's interpretations. However, the fauna from Delight, whic.h has been described by Cope ( 1889), Matthew (1902), and Hay (1927), may be lrvingtonian (mid- dle Pleistocene) in age, judging from the presence of a Cervus-like deer, Paramylodon, and Mammuthus tooth fragments, and the absence so far of Bison (Fry and Gustafson, 1974). 1978 46° 10 MILES GUSTAFSON: RINGOLD FAUNAS N r n9° j I ___ __J ,,-:-.-.-,._ Outcrop area of type ''-~~:) Ringold Formation County boundaries ~ Major anticlinal axes 3 Figure 1. Map of south-central Washington, showing major structures and outcrop area of type Ringold Formation. Contour line shown is 1000 feet above mean sea level. 4 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 METHODS Effort was made to sample the entire stratigraphic sequence (figs. 3, 4). Most specimens were found in situ. Small specimens were recovered largely by dry- screening, particularly at locality UW A6503, and by washing following the methods described by Hib- bard (1949), particularly at UW A5927. Thicknesses of stratigraphic units were determined from Locke hand level measurements. Measured sections were mostly limited to intervals determining the relative position of fossilbearing layers to nearby key beds. However, three of the sections ( fig. 4, A, C, and E) cover the entire thickness of the known fossiliferous strata in the Ringold Formation. Only those sites worked in 1967 through 1971 are plotted on the detailed sections. Statistical procedures and symbols used here fol- low Simpson, Roe, and Lewontin (1960). ABBREVIATIONS The following abbreviations are used: Institutions: KU-University of Kansas Museum of Natural His- tory DA-University of Arizona UCMP-University of California Museum of Pale- ontology UMMP-University of Michigan Museum of Pale- ontology DO-University of Oregon Museum of Natural His- tory USNM~United States National Museum UW-University of Washington, Thomas Burke Memorial Washington State Museum UW A-Paleontologic locality number prefix, Thom- as Burke Memorial Washington State Museum UWVZ-Vertebrate Zoology collections, Thomas Burke Memorial Washington State Museum Measurements: AP-greatest anteroposterior diameter T-greatest transverse diameter H-greatest crown height e--estimated measurement PER-depth of posteroexternal reentrant on P 3 of leporids Prox.-proximal Dist.-distal Max.-maximum ACKNOWLEDGEMENTS This report benefited from the efforts of many peo- ple, of whom I can name only a few here. Willis E. Fry aided my work through assistance in the field, as well as by donating many excellent specimens. I am indebted to Mr. and Mrs. Paige Burns and John Oram for help with the screen-washing operation. Max Kinne generously allowed me to use his river- bank for washing and for equipment storage. C. W. Hibbard (University of Michigan), Craig Black (University of Kansas), and J. A. Shotwell (Uni- versity of Oregon) provided access to or loans of specimens under their care. The manuscript was read and commented upon by Randall E. Brown, C. W. Hibbard, V. Standish Mal- lory, Stephen C. Porter, John M. Rensberger, Charles A. Repenning, and S. David Webb. I am par- ticularly indebted to Dr. Rensberger, who supervised much of this work, to Dr. Hibbard, whose comments on the fauna were invaluable, and to Dr. Brown for much information on the geology of the Pasco Basin. My wife Jessie deserves my intense gratitude for her encouragement and assistance during all stages of this project. STRATIGRAPHY PHYSICAL UNITS AND RELATIONSHIPS Merriam and Buwalda (1917) proposed the name Ringold Formation to encompass the thick continen- tal deposits exposed in the White Bluffs. They in- cluded a type section of 503 feet of sediments meas- ured at a site "three to four miles below Hanford," apparently in the area of the north end of Savage Is- land. They did not describe areal or stratigraphic limits to the formation . Subsequently Culver (1937) informally proposed the lateral extension of the Rin-- gold Formation to include sediments outside the Pasco Basin, and Newcomb (1958) formally ex- tended the formation in the type area downward to the top of the underlying basalt flows of the Columbia River Group. The position of the upper basalt surface in the Pasco Basin ( or generally the bottom of the Ringold Formation) is known from numerous well logs, and has been mapped by R. E. Brown and D. J. Brown 1957). R. E. Brown (1969) suggests that the 1978 GUSTAFSON: RINGOLD FAUNAS 5 TUN ~.4--~~~--l-~~~--4--1 TlON R28E R29E Figure 2. Map of southern White Bluffs, showing locations of UW fossil localities. Ringold Coulee (see fig. 1) reaches the Columbia River at section 24, Tl2N, R28E. 6 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 Ringold sediments rest mainly on the Elephant Mountain basalt flow*, which in its type locality ( 45 miles to the west of the White Bluffs) occurs within a thick section of the Ellensburg Formation (Waters, 1955) . I have collected a typical Ellensburg fauna from the upper part of the latter section, including Hipparion condoni Merriam, large Geochelone, cam- elid, a small antilocaprid with teeth similar to those of Merycodus, a mastodont (not Manunut), and other fragments, all from above the Elephant Mountain flow. This fauna is probably early Clarendonian in age (Smiley, 1963). The Ellensburg sediments, large- ly deposited by streams flowing east, become thinner towards the Pasco Basin, and no actual contacts have been reported between the Ellensburg and Ringold formations. R. E. Brown and D. J. Brown (1961) suggested that some lower Ringold sediments may be time-equivalents to upper Ellensburg rocks; however, neither convincing lithological evidence nor faunal evidence favoring this view has been presented. The eastward thinning and disappearance of the upper Ellensburg, the lithological continuity of the Ringold Formation, and the considerable time gap between ages of the known faunas suggest that the Ringold Formation may be entirely post-Ellensburg in age. However, the Ringold sediments are not all Blancan in age, as is proven by the presence of rhinoceros in the Taylor Flat conglomerate. The latter is Hemphil- lian or older. Post-Ringold Formation sediments in the Pasco Basin lie mainly in channels carved into the Ringold Formation by the Columbia River and by late Pleis- tocene floods. The primary exceptions are some ba- saltic gravels perched at high levels on the top of the Ringold, ascribed by Bretz, Smith, and Neff (1956) to early floods coming from the northeast across the undissected Ringold Formation surface, and a thin layer of late- or post-Pleistocene loess. A consider- able deposit of basaltic sand and gravel, with some thin silt layers, occurs at the mouth of Ringold Coulee. This was probably deposited under deltaic conditions, as is demonstrated by strong foreset bed- ding exposed in a gravel pit wall. The vast majority of surficial deposits in the Pasco Basin west of the Columbia and below the south end of the bluffs are fluvial and glaciofluvial silts, sands, and gravels of later Pleistocene age. These latter deposits may ex- ceed 200 feet in thickness in places; the upper parts are often rhythmically bedded, pebbly silts known as the Touchet beds (Flint, 1938). * This may actually be the Ward Gap flow (Schminke, 1967), which is closely associated with the Elephant Moun- tain flow over much of the central plateau, or the Ice Harbor flow, which is the uppermost major flow (Brown and Ledger· wood, 1973) . Ringold Formation sediments can be distinguished from later deposits by several lithologic criteria, par- ticularly by the high percentage of non-basaltic clasts, prominent limonite staining, and common oc- currence of well-cemented sandstone layers, none of which are seen in most stratigraphically higher de- posits. Contacts ' between Ringold strata and later deposits are irregular because of the extent to which the Pleistocene floods and migrating Columbia River . have channeled into the Ringold Formation. Detailed information about the post-Ringold de- posits of the central Columbia Basin can be found in papers by Bretz et al., ( 1956) , Brown ( 1968) , Flint (1938), Lupher (1944), Newcomb et al., (1972), Richmond, Fryxell, Neff,, and Weis (1965) , and in many other reports. Descriptions of the Ringold Formation by New- comb (1958, 1972) and Brown and McConiga (1960) provide a good general picture of the extent and structure of these beds in the type area. New- comb described three informal members. The lowest, the. "blue clay" section, consists of a maximum of 300 feet of clay, silt, sand, and gravel. These sedi- ments are not exposed. A distinctive conglomerate, about 165 feet thick, overlies the "blue clay" in the southern part of the type area ; it will be called here the Taylor Flat conglomerate because of the excellent exposures flanking Taylor Flat on the north and south. A majority of the pebbles and cobbles are metamorphic, granitic, or volcanic rocks derived from upriver sources; the rest is mostly basaltic rocks similar to the Yakima Basalt. The top of the Taylor Flat conglomerate is sharply defined and is easily traced south of Ringold Flat, though some in- tertongueing with sand lenses does occur in the more northern exposures. The upper 500 feet of the Ringold Formation is quite heterogeneous. Parts of the section directly overlying the Taylor Flat conglomerate are poorly indurated sandstone and siltstone, with a few thin conglomerate layers and one extensive tuff. Some sandstone beds, particularly those containing or overlying the thin conglomerates, are well-sorted, clean, and white, composed mainly of angular frag- ments of quartz, with lesser amounts of mica and other minerals. These are presumably stream-chan- nel deposits, showing laminar or cross bedding. A common and often repetitive vertical sequence of rock types includes, (bottom to top) successive beds of silt, conglomerate, cross-bedded clean white sand- stone, silty sandstone, and silt. The uppermost por- tions of the bluffs are largely a soft, massive to platy or shaly impure siltstone containing some interbed- ded sandstone. A thick, extensive tuff bed, here named the White 1978 ., c:: ., .., .g ci: Generalized section, central Pasco Basin 000000 OC>o O O O O o o O O ••• 0 000000 000000 O o O 00000 000 00000 0 o o 00 0 00000 000 0 0 .. ·. :, ~ ,: =. =.=. = .. · .. ·. :_.·,.·,.·:·: ::.-:. ::.-·.: :.: _:_: .. · .:.:_:,.·_.::_::·.:_.· .. : :_:_:_:_:_::: :_:::::::: :. .. ; : _:: -.. :· :-: : :- ': ·: ": :· _: :- _: .· ... ~-:-~ ·: ·:·: ·: :_. :··:::: ..... :- _: :· :· :·: : :· :· :· ... ~-: ................ • '0 0 0 0 O • • • • •' • • • 00 O•O •, • • • •, • • • • • , • • · , ·., , , Touchet silts, and loess, lacesc Pleistocene and Recent lace Pleistocene glacial flood deposits upper Ringold Formation * · · · :.::· ·::::· =· ·=· =· ·= ..=· ·=· ·=· ·=·~· ·= .. =· ·==:::t:=Whice Bluffs cuff ... . ............ ·· o •o o ·oo,o·•· · · · ': : .'. -o oo- .. · · ·,.,, . , known r ... :-: : .. :-:-: -: :'·:-::_.: ;_.:: :: . . . . -:-:-:: :: :-:-:-::-:-:-:-: ::.. .·.·.·.·.·.·:.·.·:.·.·. 0000 000 0 00000 00 0 Oo ooo 000000 0000 000000 000000 000000 0 00 0000 00 oo O O o O O 0 0 O 0 0 0 0 0 000 ooo ooo ooo oooooooX 000 00000 00 0000000 000000 0 000000 . . : ; .·. ·:.·. ·. ~·::::. . ..... . · .. .. . . ..... ....... .... .. .... ... .. . .. .. ....... ..... a o o o 0 0 00 O O ge of vercebrace fossils Taylor Flat conglomerate lower Ringold Formation, "blue clay" section ! i silcscone and claystone J::: :: :: ::::i sandstone I o o o o o conglomerate Columbia River Basalt, lace flows and sedimentary incerbeds. • Thicknesses of Ringold Formation scraca Brown (1968), and measured sections. shown are maxima. Information from Newcomb (1958), 7 Figure 3. Generalized section, central Pasco Basin. A Teleoceras mandible found in 1974 was located 80 feet below the top of the Taylor Flat conglomerate (marked with a cross). 8 COLUMNAR SECTIONS OF THE RJNGOW FORMATION. SOUTHERN WA SHI NG TON. F A8813 A9577 E EXPLANATION WHITE BLUFFS. D silt EI] gravel A9328 111111 sand and silt ii White Bluffs tuft II sand D covered No.23 A 700 A9327 A8811 600 ?' A5927 A6525 500 1,00' ., ,, ' ·, ·. , 1978 GUSTAFSON: RINGOLD FAUNAS 9 Bluffs tuff (figs. 3, 4, 5, and 6) occurs within the most fossiliferous part of the section. The tuff is easily recognized by its fine grained texture, white color, and blocky fracture. Microscopically the tuff is al- most entirely fine particles of glass, with the largest fragments being less than 0.2mm in diameter. The stratigraphic position and lateral extent of this key bed was traced by Brown and McConiga (1960). Where the White Bluffs tuff is not preserved, for in- stance in the area of Taylor Flat (fig. 4), large amounts of stream-channel sediments occur in that part of the section where the ash would be expected. Where it is preserved, however, it is remarkably uni- form in thickness and suggests a depositional surfac~ of very low relief. In some places it rests on a definite buried soil. Several of the aforementioned thin conglomerate layers are noteworthy because of their significance concerning depositional environments (see below) and because they contain concentrations of fossil bone. A number of productive localities in the area of the south end of Ringold Flat are all in two later- ally extensive ( though discontinuous) beds of con- glomerate. Locality A6503 is a small quarry in the upper fossiliferous bed 45 feet above the White Bluffs tuff; the lower, less extensive conglomerate is 40 feet below the tuff ( fig. 4) . At A6503, the fossiliferous conglomerate is bounded above and below by clean white sand. The conglomerate at this point is seldom over 20 cm thick. Its level may vary by 15 or 20 cm over the 10 meter long exposure. Other conglomerate lenses at the same level, extending at least half a mile north and south of the quarry, reach a thickness of over a meter in places. Within the quarry, another horizon- tal, very thin conglomerate lens truncates the cross- bedding in the meter-thick sand overlying the fossil- iferous bed. Pebbles and small cobbles (up to 10 cm in diame- ter) in the fossiliferous conglomerate include gran- itic and volcanic rocks, quartzite, agate, and nodules of clay in varying colors, some being grey, light brown, or brick red. The conglomerate at A6503 is dominated by irregular limonite nodules which oc- cur among and cement other particles. The nodules, mostly 2 or 3 cm in diameter, have a brownish rind and a yellow, softer, often partly hollow interior. Sand grains and larger particles occur in the rind but not in the interior. Numerous pieces of fibrous, very poorly preserved wood impregnated with limon- ite were found. The end of deposition of Ringold sediments is marked by a thick layer of caliche found in the up- permost levels of the bluffs. This caliche is not present in areas which have been eroded (Newcomb, 1958) ; it presumably records a dry interval after final flood- plain deposition, but prior to the large scale erosion of Ringold sediments which occurred during the Pleistocene. A second caliche layer, developed in Ringold sediments beneath the Pleistocene sediments in the AEC Hanford reservation area, presumably records a second dry interval after the first stripping of the Ringold Formation, and before Pleistocene sedimentation (R. E. Brown, personal communica- tion, 1973). OCCURRENCE OF FOSSILS Fossils from the Ringold Formation at the White Bluffs in the Burke Museum collection came from more than fifty localities. Some of these, particularly the older sites, may be duplications because of col- lection done by several persons at different times from the same bed or quarry. Two distinct faunas seem to be represented. The best known assemblages, from strata above the Tay- lor Flat conglomerate, may be termed the White Bluffs local fauna; the less well known, from within the Taylor Flat conglomerate, may be termed the River Road local fauna. Biostratigraphic terminol- ogy as used by vertebrate paleontologists is current- ly in a state of flux. The term "local fauna," used to denote an assemblage of fossils from a narrow strati- graphic and geographic range and with essentially uniform taxonomic composition, has been standard for many years. A rationale for the continued use of "local fauna" names has been provided by Tedford (1970), but some authors (notably Fisher and Rens- berger, 1972, and Downs and White, 1968) have used zone terminology (local range zones or concur- rent-range zones) commonly used by marine bio- stratigraphers. The local faunas named above should be considered biostratigraphic in nature; the boun- dary between them can be conveniently placed at the Figure 4. Detailed columnar sections. A measured up gully between localities A6520 and A5927 visible in fig. 5, near north end of boundary between sections 1 and 2, TlON, R28E. B measured at A8810, about 300 feet north of A. C measured up gully beside A6525 in central section 25, TUN, R28E. D measured up dirt road just north of Ringold Wasteway chute, about one-fourth mile north of boundary between section 1, TUN, R28E, and section 36, Tl2N, R28E. E measured at A6503, near southern boundary of section 25, Tl2N, R28E. F measured up north side of canyon in south central section 25, Tl2N, R28E. Scale at right indicates feet above sea level. 10 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 Figure 5. Bluffs at locality A6520 (near center of photograph); resistant stratum just above center is White Bluffs tuff, about 8 feet thick at this site (2.4 meters). top of the Taylor Flat conglomerate, but this is purely a matter of convenience at present. The White Bluffs geographic area may be consid- ered a collecting field by the definition of Macintyre (1966) . Five major areas exist in which fossiliferous beds are most common ( fig. 2) . These are at the south end of the bluffs, the south end of Taylor Flat, Baxter Canyon, the Ringold Flat area, and the Sav- age Island area. The precise stratigraphic position of the sites in the Savage Island area with respect to those further south is not certain because of changes in lithology, lack of northern exposures of the White Bluffs tuff, and major landsliding in the intervening area. Two kinds of sediments in · the upper Ringold Formation, the stream channel gravels and a fine light brown silt, have produced fossils. Most of the sites are l9cal lenses of conglomerate. Since the two types of sediment contain substantially different faunas, field work was concentrated especially on a locality in each lithology. These are UW A5927, a silt deposit, and UW A6503, a stream channel con- glomerate (see above and fig. 4) . The bones and teeth at UW A6503 were usually embedded in a mixed matrix of limonite nodules, clay nodules, pebbles, and sand. Larger bones often lie upon the upper surface of the conglomerate. Most of the bones lay within 15° of horizontal. No articulated bones were found in the conglomer- ate. All bones and teeth show some sign of abrasion, often so extreme that identification is difficult. The abrasion must have occurred while the bones were fairly fresh. The preserved bone material is very brittle, and the many thin or elongated pieces of bone found would not have survived reworking. Concentration of secondary minerals, particularly iron oxides, in the conglomerate lenses has produced a wide variety of colors in the enclosed bones. ·1n contrast to most of the fossiliferous sites, specimens from UW A5927 and UW A6520, near the south end of the bluffs, are found in fine clayey silt- stone. The two sites are at the same stratigraphic level, about a hundred meters apart (fig. 4, A). Sur- face exploration at UW A5927 revealed a group of associated rodent foot bones and a partial skeleton of Ammospermophilus, a small ground squirrel. Screenwashing of sediments from this deposit pro- duced a large number of isolated rodent teeth and 1978 GUSTAFSON: RINGOLD FAUNAS 11 bones, two unerupted peccary teeth, snake vertebrae, and miscellaneous other bones. Most specimens are less than 2 cm in greatest dimension. At least two bone fragments from this site show signs of having been gnawed by rodents or carni- vores. There are no indications of surficial wear from stream transport, and very delicate structures (for example, single fish scales) are preserved undam- aged. Most of the bones are a pale tan color, appar- ently because of the uniformity of the fine-grained sediments. MODE OF DEPOSITION OF THE RINGOLD FORMATION Deposition of the Ringold Formation was almost certainly initiated by a combination of settling in the area of the Pasco Basin and uplift of the Horse Heav· en Hills anticline across the downstream course of the ancestral Columbia River. Sediment accumulation occurred earliest towards the basin center (Brown, 1969, p. 9), and Ringold sediments may eventually have covered most of the central Columbia Plateau (Culver, 1937, and Brown, 1969). Present evidence suggests that the Ringold Forma- tion is composed largely of stream-channel and flood plain deposits. Several early investigators, including Russell (1893), Smith (1903), and Calkins (1905) suggested a lacustrine environment for all or a ma- jor part of the White Bluffs sediments. More recently, Newcomb (1958) has returned to this hypothesis as the major explanation for Ringold Formation depo- sition. Newcomb states that "the successions of silt, fine sand, clay, and volcanic ash exhibit some rhyth- mic changes in a vertical direction, as though they resulted from deposition by the gradually shifting of currents in a large lake." The rhythmic changes observed by Newcomb do exist, but an alternate ex- planation ( see below) is available and seems more probable. Most recent authors, including Waters ( 1955) and Brown ( 1969) , have accepted the hy · pothesis propounded by Merriam and Buwalda (1917), that the Ringold Formation is by and large a complex flood plain deposit. Merriam and Buwalda cited as evidence the poor sorting of most sands, silts, and clays, the rarity of freshwater molluscs usually common in lacustrine environments, and the presence of mammalian fossils at several localities. Although fish bones occur in association with the mammal re- mains, the bones are always disassociated and usually show wear from transportation in an abrasive me- dium. No complete fish skeletons are known, even from the finer shaly beds in the upper part of the secton. Many sections which I examined in the lower bluffs south of Ringold Coulee contain one or more se- quences of sediment types closely matching that se- Figure 6. Close view of White Bluffs tuff just north of Ringold W asteway; note thin layer of tuff at bot- tom separated from the main body of tuff by a darker layer; thickness of tuff in photo is about four and one half feet ( 1.4 meters) . quence shown by Allen (1971, fig. 3) to be typical of the deposits of meandering streams. For example, the Ringold Wasteway North section (fig. 4 D) con- tains the following sequence: 1. an underlying sandy siltstone truncated by a smooth erosion surface; 2. a thin conglomerate with many clay fragments; 3. clean, cross-bedded sandstone, grading upward into 4. overlying thick sandy siltstone. The extensive erosion surface on the top of bed 1 could have been produced by the sideways and down- stream migration of the meandering channel. The coarser fraction of the channel bed sediments, includ- ing gravel, chunks of clay and silt, and occasional bones and teeth, was deposited first, and was im- mediately buried by downstream migration of point- bar deposits, which are mainly fine clean sand, com- monly cross-bedded. As the stream migrated out of the area, deposits became less well sorted, and finally only overbank sediments, mainly impure silt, were deposited by floods. Repetition of this sequence or parts of it, with local variations, is probably the mechanism responsible for Newcomb's "rhythmic changes." The above is not intended to deny the presence of lacustrine lenses in the Ringold Formation. As Waters (1955) states, " ... although the Ringold Forma- tion contains thinly stratified clays, silts, and diato- mite layers that were doubtless deposited in lakes, these lakes appear to have been shallow and to have shifted position." Typical modern floodplains have numerous oxbow lakes and other depressions which fill with water during floods. Fossil mollusks have 12 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 been found at three sites (McKnight, 1923, and Tay- lor, 1966). Taylor, referring to the three species of snails he reports, states: "This assemblage probably represents a shallow-water environment, such as a flood-plain or the edges of a stream or lake, rather than an open lake. The water body might have been subject to seasonal fluctuation, but probably did not dry up entirely." SUMMARY OF PLIOCENE AND PLEISTOCENE EVENTS In earliest Pliocene time the area that is now the Pasco Basin was near the center of a broad, flat plain formed by the upper surface of the last flood basalt flows of the Columbia River Group. The basalts and interbedded sedimeqts beneath the Pasco Basin area are over 10,000 feet thick (Raymond and Tillson, 1968); the earliest known flows may be as old as late Eocene or early Oligocene (Newman, 1969). As Brown (1969) has emphasized, deposition of this enormous thickness of basalt required fairly continu- ous basining at a rate sufficient to offset the thicken- ing of the basalt pile. Deformation of the basalts, other than the continu- ing basining, began during latest Miocene to earliest Pliocene time, on the order of 12 to 15 million years before present (Holmgren, 1969). Tectonic activity beginning at this time included the first significant rise of the Cascade Mountains (Smiley, 1963) and the beginning of the warping of the basalt along east- west or northwest-southeast axes. Considerable in- crease in relief had occurred during or before the inception of deposition of the Ellensburg Formation on the east flank of the Cascades, and before the out- pouring of the last basalt flows which are interbedded with the Ellensburg Formation (Holmgren, 1967). The present Columbia River drainage basin was evidently formed during this initial period of Cascade uplift and basalt deformation. Quartzitic conglom- erates within the Ellensburg Formation indicate the presence of a large river flowing from north to south which could have drained much of the area now drained by the Columbia. The position of the main drainage channel apparently shifted widely due to displacement by late basalt flows and subsequent re- establishment of drainage toward the Columbia Ba- sin center, near Pasco. Cessation of basalt eruptions allowed the center of drainage to become permanent- ly established near its present position, setting the stage for the deposition cycle which produced the Ringold Formation. The first Ringold Formation sediments were de- posited sometime after the main phase of deposition of the Ellensburg Formation, but before the end of the Hemphillian mammalian faunal stage. Most of the visible White Bluffs sediments were deposited in early Blancan time (late Pliocene as usually used by vertebrate paleontologists) according to evidence presented in this paper, except for the Taylor Flat conglomerate which is pre-Blancan at least in part. Regional topography suggests that a combination of continued settling near the basin center and uplift of the Horse Heaven Hills anticline produced sufficient river grade changes to initiate aggradation in the Pasco Basin. Other factors, such as altered precipi- tation patterns or increased sediment load, may well have been involved. Contrary to statements by Newcomb (1958) an.cl earlier authors, the Columbia River was probably never completely impounded. Sedimentation oc- curred on a broad flood plain. The present gradient of the Columbia River is quite low, less than 1.5 feet per mile along the White Bluffs. Late Pliocene gradients must have been considerably less, perhaps near the 0.5 feet per mile suggested by Brown (1969). Sev- eral streams flowing into the Pasco Basin from the east, north, and west may have contributed to sedi- mentation. Sediments in the Othello and Quincy ba- sins further north, referred to the Ringold Formation by some authors, may have been deposited by such tributary streams. The cause of the end of Ringold deposition is not certain. A number of factors may have been involved, for example, a slowing or stoppage of anticlinal up- lift at W allula Gap, increased precipitation, regional uplift, decreased sediment load, or any combination of these and other factors. Newcomb (1958 ) thought that the end of Ringold deposition occurred rather suddenly, and that the ancestral Columbia became quickly entrenched in the Ringold sediments in the area to the west of the White Bluffs. Presence of the high-level caliche suggests some delay in incisement; however, once downcut- ting was underway it may have been rapid. During most of the Pleistocene the sediments of the Pasco Basin underwent degradation primarily by stream and aeolian erosion. The relatively undissected na- ture of the upper surface of much of the remaining Ringold Formation suggests that rainfall ( and there- fore flow in tributary streams in the area) remained low throughout the Pleistocene. Aeolian erosion of the Ringold sediments seems to have provided much of the material now composing the Palouse Forma- tion further to the east (Bryan, 1927; Lewis, 1960) . The present topography of the Pasco Basin is largely the result of glacial meltwater floods, vast in scale, originating in Lake Missoula, Montana, during the Wisconsin glaciation ( Bretz et al., 1956). Early floods, coming from the north and east, deposited basalt gravels over much of the upper Ringold sur- face. Later floods carved Ringold Coulee and other 1978 GUSTAFSON: RINGOLD FAUNAS LIST OF VERTEBRATES FROM THE WHITE BLUFFS TABLE 1 Previously reported Fish Carcharius? sp. (b) catfish (b) percoid? ( b) Reptilia T estudo sp. (a) Clemmys marmorata (d) Aves Marila? sp. (b) Mammalia leporid ( a) micro tine (primitive) ( b) Megalonyx sp. (a,c) M ammut americanum ( c) Equus or Pliohippus sp. (a) Equussp. (a) Platygonus sp. (c) Camelops sp. ( c) camelid (large) (a) camelid near Pliauchenia (a) cervid (a) Rangif er sp. ( c) (a) Merriam and Buwalda, 1917 (b) McKnight, 1923 (c) Strand and Hough, 1952 Recognized in this report Fish I ctalurus vespertinus Archoplites taylori Salmonidae indet. Amphibia Anura indet. Reptilia ?Testudo sp. Clemmys marmorata Chrysemys? sp. Squamata indet. Mammalia Scapanus sp. Hypolagus ringoldensis n. sp. Nekrolagus sp. cf. N. progressus Paenemarmota or Marmota sp. Spermophilus? russelli n . sp. Ammospermophilus hanfordi n. sp. Thomomys sp. cf. T. gidleyi Castor californu:us Dipoides rexroadensis Peromyscus nosher n. sp. Neotoma sp. cf. N. quadriplu:atus Ophiomys mcknighti n. sp. M egalonyx rohrmanni n. sp. Canis davisi Borophagus sp. Ursus sp. cf. U. abstrusus Trigonictis cooki Felis sp. Machairodontinae Mammut americanum Equus (Dolichohippus) sp. cf. E. simplu:idens T eleoceras sp. * Platygonus pearcei Megatylopus cf. M. cochrani ? M egatylopus sp. * Hemiauchenia sp. Bretzia pseudakes ( cl) Brattstrom and Sturn, 1959 * Member of River Road local fauna; all others from White Bluffs local fauna. 13 14 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 channels, and still later floods, travelling down the Columbia valley, formed the huge gravel bars of the basin area and carved the Bluffs back to their present position. The Touchet beds overlie all but a few aeolian deposits. They consist of rhythmically bedded silts and sands, apparently deposited during frequent fluctuations of a large lake during the waning stages of the last glaciation. Post-Pleistocene changes have apparently been slight. SYSTEMATIC PALEONTOLOGY WHITE BLUFFS LOCAL FAUNA Phylum CHORDATA Subphylum VERTEBRATA Class OSTEICHTHYES Identifiable fish bones have been recovered from ten sites. Pectoral fin spines of the catfish (Family lctaluridae) Ictalurus vespertinus Miller and Smith (1967) are the most common identifiable bones. Re- mains of the centrarchid sunfish Archoplites taylori Miller and Smith are also fairly common; both of these species were first reported from the Glenns Fer- ry Formation of Idaho. Specimens of a large salmon- id have also been found. The above identifications were made by R. R.Miller (personal communication, June 1973), who is currently studying the Ringold fishes. Miller (1965) also reported Ictalurus and Ar- choplites from sediments of possible Ringold Forma- tion age near Moses Lake, Washington. Class AMPHIBIA OrderANURA Frog or toad bones, including humeri, ilia, and others, were found at UW A5927 and UW A6520. Class REPTILIA Order TESTUDINATA Family Testudinidae Subfamily Emydinae Genus Clemmys Ritgen Clemmys marmorata (Baird and Girard), 1852 Discussion-Brattstrom and Sturn (1959) re- ported this living species from a site opposite the old town of White Bluffs. Their 17 fragments of bone, mostly marginals, are in the old California Institute of Technology collection, now in the Los Angeles County Museum of Natural History. Genus Chrysemys Gray Chrysemys? sp. (Figure 7) Material-UW A6503: UW 35143, hyoplastron; UW 35203, epiplastron. UW A9328: UW 42455, en- toplastron. The Burke Museum collection also in- eludes many other specimens from several localities whch may belong in this taxon. Discussion-The predominant turtle in the White Bluffs is an emydine with plastron elements verv similar in morphology to those of Recent Chrysemy; , and probably referable to that genus at least in the broad sense used by McDowell (1964). The primary character differentiating these specimens from Clem- mys is the posterior position of the humeral-pectoral sulc.us, which does not cross the entoplastron in any specimen in the Burke Museum collection. Differ- ences in size and form of several marginals suggests that more than one taxon is present in this collection. Figure 7. Plastron of Chrysemys? sp., ventral view. Hyoplastron, UW 35143; epiplastron, UW 35203, both from A6503. Entoplastron, UW 42455, from A9328. Scale line indicates one centimeter. · Anterior towards top of page. 1978 GUSTAFSON: RINGOLD FAUNAS 15 Order SQUAMATA A number of small reptile bones were found, pri- marily at UW A5927. Class MAMMALIA Order INSECTIVORA Family Talpidae Genus Scapanus Pomel, 1848 Scapanus sp. (Figure 8) Material-UW A6503: UW 35108, left humerus. Stratigraphic position-Ringold Formation, 45 feet above White Bluffs tuff, early Blancan. Description-A single humerus, lacking portions of the proximal end, wide in proportion to length ( though not as wide as in Scalopus), with well devel- oped processes as in advanced fossorial talpids. Size is very close to that of Scapanus orarius True. The length of the unbroken humerus was probably about 12.5 mm.; the preserved length is 9.6 mm. The distal maximum width is 8.3 mm. Minimum shaft width ( taken in the same plane as distal width) is 4.3mm. The White Bluffs specimen is closer in morphology to Recent species of Scapanus than to S. (Xeroscaph- eus) proceridens Hutchison ( 1968) from the Hem- phillian Westend Blowout site (U02349) in north- ern Oregon. Judging from comparison with seven specimens of the latter, the White Bluffs specimen is slightly shorter, and is wider in proportion to length. Figure 8. Scapanus sp. , left humerus, UW 35108, anterior side. Scale line indicates one millimeter. Order LAGOMORPHA Family Leporidae Subfamily Archeolaginae Genus Hypolagus Dice, 1917 Hypolagus ringoulensis new species (Figure 9a-d) Type-UW 45015, partial mandible with Pa-M2 and broken M3 , collected by William Shawver in 1965. Type locality-UW A8814, White Bluffs south of Ringold Coulee, Franklin County, Washington, from conglomerate about 45 feet above White Bluffs tuff. Diagnosis-A species of Hypolagus similar in size and morphology to H. oregonensis Shotwell; anterior reentrant of P3 much more consistently present and usually deeper; P 3 is slightly longer antero-posterior- ly; posteroexternal reentrant deeper and more often with distinct anterad inflection at its tip. Referred specimens-The following are UW lo- calities and specimen numbers: A3029: 41548, P 3• A3037: 41975, P 3• A5927: 34977, P3• A6503: 35118- A, 35118B, 35129, 35145A, 35145B, 40396, 40428, 45020, all P3• A6520: 40313, P 3• A6525: 35237, mandible with I, P 3+ A8810: 40259, Pa; 40286, mandible with Pa-M2 ; 40293, mandible with P 3-M1• A8812: 42220, P3• A9305: 42148, Pa. A9851: 46169 4613, both P 3. Stratigraphic range-Ringold Formation, from 40 feet below to 45 feet above White Bluffs tuff, early Blancan. Description-The most distinctive tooth is Pa, of which 21 are known from 11 localities. The Pa is ar- cheolagine in character, having a posteroexternal re- entrant extending from half to two-thirds of the width of the tooth. H. ringoulensis is larger than most spe- cies of Hypolagus but is within the size range of H. oregonensis Shotwell (1956) from the Hemphillian McKay Reservoir fauna of northern Oregon. How- ever, the mean anteroposterior diameter of Pa in H. ringoulensis is significantly larger than in H. oregon- ensis ( Student's t probability of derivation from identical populations is less than .01). Mean trans- verse diameters of Pa for the two species are not sig- nificantly different. H. ringoulensis can be distinguished from all other species of Hypolagus except H. oregonensis by the usual presence of a strong anterior reentrant on P3• This structure is strong (deep or V-shaped) on 72 % of the Pa's of Hypolagus from the White Bluffs but is strong on only 38% of the P 3's from McKay Reser- voir; on the other specimens this reentrant is weak (shallow, broadly U-shaped) or absent. A second, usually weaker anterior reentrant is present on the anterointernal face of Pa of some individuals; when 16 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 a groove is not present this face is usually flattened and finely striated. The anteroexternal reentrant is shallow and is usually not complicated by minor crenulations, though such crenulations do occur. The posteroexternal reentrant is deeper than in H. ore- gonensis in most cases; the depth in specimens from the Ringold Formation includes larger measurements F B E c D :Figure 9. Rabbits and marmot. A. Hypowgus rin- goldensis, type left P 3-M3, UW 45015, occlusal view. B. H. ringoldensis, left P 3 of type specimen, occlusal view. C. H. ringoldensis, right P 3, UW 42148. D. H. ringoldensis, right P3 , UW 40313. E. Nekrolagus cf. N. progressus, left P 3 , UW 351188. F. Paenemarmota or Marmota sp., right M3 , UW 35248. Scale lines indicate one millimeter. Anterior towards top of page in A-E, to right in F. than found in the H. oregonensis sample. The smaller values of depth in H. oregonensis are below the range in H. ringoldensis, and the difference in the means is highly significant (Student's t probability of deriva- tion from identical populations is less than .001 ). In H. ringoldensis the tip of the posteroexternal reen- trant bends slightly anterad in a majority ( 57% ) of the speciments ( 12 of 21). This anterad bend was found in only 12 % of the specimens ( 4 of 34) of H. oregonensis. The other cheek teeth, P4-M3 , are quite typical for Hypolagus, and show no unusual features. The talon- ids of these teeth ( except M3 ) are of lesser transverse width and greater length than in the Recent Lepus califomicus Gray. The enamel borders show no mi- nor crenulations. The lower incisor of H. ringoldensis is similar to that of H. oregonensis; in both the posterior end of the incisor is positioned opposite the posterior part of the trigonid of P 3, and the incisor is considerably longer and less curved than in Recent Lepus califor- nicus Gray. Fenestration of the mandible is less developed than in L. califomicus. The lateral surface of the one an- terior mandible available (UW 35237) is practically smooth and solid in the areas below P 3 and the in- cisor where fenestrae are present in the Recent form. The dorsal outline of the diastema is nearly straight, as in H. oregonensis. The horizontal ramus is gener- ally similar in depth and form to that of L. calif omi- cus, which is of nearly the same size. Discussion-The statistical differentiation of H. ringoldensis from H. oregonensis depends largely on the assumption that each sample has a normal fre- quency distribution. The probability that this as- sumption is true is high in the sample of H. oregonen- sis which is all from one locality (UO loc. 2222) ; a complete intergradation of characters is seen, and the statistical parameters (i.e. the near coincidence of the means and medians) are as would be expected if only one species were represented. The sample from the White Bluffs came from a number of localities through a considerable stratigraphic range. In the two measurements of P3 shown in Fig. 10, the mean is well above the median, so much so in the AP di- mension that one standard deviation above the mean exceeds the observed range. The P3's falling in the lower part of the range tend to be more primitive in enamel pattern, particularly in more frequent lack of the anterior reentrant. There is insufficient reason to separate out two species of Hypolagus from the White Bluffs sample, but the possibility remains that more than one is present. Hypolagus oregonensis was probably ancestral to H. ringoldensis. The type localities are separated by 1978 GUSTAFSON: RINGOLD FAUNAS 17. 2.0 N=27 R N = 79 I PER 1.5 0 N=34 2.5 AP 3.5 Figure 10. Comparison of anteroposterior length, and depth of posteroexternal reentrant in Hypolagus ringoldensis n. sp. from the upper Ringold Formation and H. oregonensis Shotwell from McKay Reservoir, Oregon. Narrow lines indicate ranges; vertical and horizontal lines intersect at means of both. Broad bars extend one standard deviation from means. N = number of measurements used in calculation. only 80 miles. The differences are mainly in char- acter frequencies and means, rather than in discon- tinuities in morphological type. H. ringoldensis tended towards a more complex enamel pattern in PH, away from the simple pattern characteristic of H. vetus (Kellogg ) and other earlier and contempor- ary populations of Hypolagus . There is some evi- dence, in the form of a single P 3 reported by Downey ( 1968, fi gure 2d ) from the Benson local fauna of Arizona, that this line of rabbits was not restricted to the Pacific Northwest. The latter tooth, called Nekro- lagus? by Downey, but lacking the enamel lake char- acteristic of Nekrolagus, is apparently identical in all characters to H. ringoldensis and to some individuals of H. oregonensis. Subfamily Leporinae Genus Nekrolagus Hibbard, 1939 Nekrolagus cf. N. progressus (Hibbard) (Figure 9e) Material--UW A6503; UW 35188, Pa. Stratigraphic position-Ringold Formation, 45 feet above White Bluffs tuff, early Blancan. Description-A single Pa possesses the diagnostic characters of the genus Nekrolagus. An oval enamel lake is separated from both the posterolingual enamel border and the posteroexternal reentrant. The anteri- or face has a small but distinct and cement-filled an- terior reentrant. The tooth is similar in size and pro- portions to the Pa of Nekrolagus progressus (Hib- bard ) from the Rexroad Formation of Kansas. The White Bluffs tooth differs from N. progressus holo- type (KU 4570) in the pinching of the enamel lake, lack of several very smb.ll enamel crenulations, and slightly larger size. Measurements of UW 35188 are: AP=3.7 mm, T=3.6 mm. whereas those of N. progres- sus are: AP=3.5 mm, T=3.7 mm. Discussion-Nekrolagus has been known primar- ily from specimens from the Rexroad Formation of Kansas, but it has also been mentioned from sites in Arizona and Southern California. As noted above, the Arizona specimen may not be Nekrolagus. The Southern California species (Downs and White, TABLE 2 AP T PER AP T PER P a MEASUREMENTS OF H ypolagus FROM THE WHITE BLUFFS, WASHINGTON, AND McKAY RESERVOIR, OREGON H. ringoldensis, White Bluffs, Washington no. of specimens mean range standard deviation 19 21 21 3.43 3.04 1.79 2.8-3.7 2.5-3.5 1.4-2.0 H. oregonensis, McKay Reservoir, Oregon 0.24 0.27 0.19 no. of specimens mean range standard deviation 34 33 34 3.23 2.98 1.51 2.6-3.7 2.5-3.5 1.2-1.9 0.23 0.28 0.15 18 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 1968) has not been described. The White Bluffs oc- currence extends the range of the genus into the Pa- cific Northwest. The genus Nekrolagus is apaprently ancestral to many of the modern leporids, including Lepus and Sylvilagus (Hibbard, 1963 ) . Order RODENTIA Family Sciuridae Genus Paenemarmota Hibbard and Schultz 1948, or Marmota Frisch, 1775. Paenemarmota or Marmota sp. indet. (Figure 9f) Material-UW A3037: UW 41974, lower I. UW A6503: UW 35248, Ma; UW 42264, upper I. Stratigraphic position- Ringold Formation, 45 feet above White Bluffs tuff, early Blancan. Description-Terminology used here for sciurid tooth structures generally follows Bryant ( 1945 ) . UW 35248 is a large, moderately worn M3 • The tooth is slightly smaller (AP=8.0 vs. 8.5 mm) than the M3 of Marmota sawrockensis Hibbard 1964, (late Hemphillian, XI member of Rexroad Fm., Kansas ) , but is larger tlian M3 of Recent species of Marmota. The occlusal outline is similar to that of M3 of Pae- nemarmota barbouri Hibbard and Schultz ( early Blancan, Rexroad Fm., Kansas) but size is consid- erably smaller (Table 3) . The talonid basin is bor- dered on the anterior and anterolateral sides by a shallow but distinct groove (Repennings's [1962] "basin trench" ). A similar groove is seen in Pae- nemarmota, M. sawrockensis, M. nevadensis (Kel- logg) , M. oregonensis Shotwell, and in Cynomys and some species of Spermophilus but is not found in Recent species of M armota. The trigonid basin is small, and is closed medially by the metalophid as in Paenemarmota and M. sawrockensis. The parameta- conid is relatively low. Two incisors, both with stri- ated enamel, may belong to this taxon. .... Discussion-The several large sciurids from tlie Hemphillian and Blancan of western United States which possess an anterolateral groove on tlie talonid have been placed in two genera largely because of the extreme size and proportionately large P 4 4 of one form, Paenemarmota. A generic identification would be premature in the absence of a P 4 • TABLE 3 Paenemarmota OR Marmota SP., MEASUREMENTS OF TEETH Paenemarmota or Marmota sp., Paenemormota Ringold Formation barbouri, from 1/UW 42264 Repenning, 1962, Marmota / I 41974 Tables 1-2 caligata, Ma 35248 max. and min. * uwvz 18837 I/ AP Depth 6.7 ?7.5-9.9 (5 )2 7.6 T Width 4.4 ?4.8-6.8 (5) 5.1 curvaturea 23e1 ?28--45 (5) 19e / I AP Depth 7.2 7.0-8.7 (10) 7.1 T Width 5.0 4.7-5.8 (10) 4.4 curvature3 36e 41--49 (10) 30e Ma T Width4 7.3 7.8-9.3 (7) 6.3e AP Length4 8.0 9.2-11.0 (8) 7.4 * From individuals of several widespread localities not necessarily of a single time period, or therefore of a single population. 1 e indicates estimated measurement. 2 Numbers in parentheses are the number of specimens listed by Repenning (1%2 ) . a Measured for the anterodorsal or anteroventral surface. 4 Measurements or Ma made parallel or at right angle to the labial edge of the tooth. 1978 GUSTAFSON: RINGOLD FAUNAS 19 Genus Spermophilus Cuvier, 1825 Spermophilus? rus;elli new species (Fig. 11) Type-UW 33038, left mandible with broken I and P4-M3• Type locality-UW A5927, (near south end of White Bluffs), Franklin County, Washington. Etymology-Named for I. C. Russell, who first mentioned in print the presence of fossils in the White Bluffs. Diagnosis-A moderately large sciurid, larger than most species of Spermophilus ( except S. patter- soni Wilson) but smaller than most species of Mar- mota; mandible exceptionally deep (10.5-10.8 mm below P 4 ); incomplete protolophid on P4 not at- tached to base of metaconid; protoconulid frequent- ly present; p a about one-third the size of P4 ; meso- style present on P 4 and Ms; metaloph not attached to protocone on P4 ; metaloph incomplete on Ms; meta- conule present on M3, M3 longer than wide. Referred specimens-The following are UW lo- calities and specimen numbers: A5927: 33040, p s_ P4; 33041, Ms; 33052, upper incisors (2); 33058, P4; 33062, mandible with I, P4-M3 ; 33069, lower I; 33070, upper I; 33076, lower I, 33077, upper I, 33079, M2 ; 40599, upper I. c A Stratigraphic position-Ringold Formation, from a siltstone 30 feet below White Bluffs tuff, early Blancan. Discussion-Spermophilus? russelli bears some resemblance, particularly in size, to M armota vetus (Marsh, 1871, p. 121), a form of uncertain age from "Loup Fork Beds" of northern Nebraska, which is known only from the type mandible. S.? russelli dif- fers from M. vetus by lacking any grooving of the lower incisor, having greater depth of the mandible, and by the presence of prominent mesostylids. It differs further from the other marmots in a number of characters, particularly in relatively small size, small P 4 relative to M1, small size of the protolophid of P 4, lack of grooving of the incisors, moderate size of the parastyle of P4, lack of connection between metaloph and protocone on P4 , and incompleteness of the metaloph on Ms. MS of S.? russelli is very similar in size and shape to two teeth reported by Zakrzewski ( 1969, p. 5) from Hagerman, Idaho, as Citellus sp. (large). The Hagerman specimens differ by the lesser prominence of the protoconule and metaconule. The large ground squirrel from Hagerman may prove to be S.? rus- selli when more material is recovered. The only known species of Spermophilus which is larger than S. ?russelli (P4 AP==3.4 mm) is S. patter- .,.; ; ,'• Figure 11. Spermophilus? russelli n. sp .. A. right P3 -4, UW 33040, occlusal view. B. right Ma, UW 33041, occlusal view. C. Type left mandible with P 4-M 3, UW 33038, lateral and occlusal views. Anterior to right in A-B, to left in C. Drawing UC by Mark Orsen. Scale lines indicate one millimeter. 20 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No. 23 soni (P4 AP=4.l mm, Wilson, 1949, p. 170) from the Hemphillian Yepomera fauna of Mexico. The latter resembles Recent species of S. ( Otospermo- philus) except for size and presence of a strong meta- loph on M3 , and it differs from S.? russelli in the lat- ter character. In its general features S.? russelli is similar to the subgenus Otospermophilus which Black (1963) em- phasizes as the most primitive structural group with- in the genus Spermophilus. S.? russelli differs mark- edly from the primitive marmotlike sciurid Pakzearc- tomys in the relatively larger cheek-tooth series and smaller incisors, and from Protospermophilus in hav- ing relatively more distinct lophs, higher tooth crowns, and a p a with distinct anterior and posterior cingula. The teeth of Protospermophilus ( compari- son here is primarily with P. angusticeps specimens from the Deep River Formation, late Hemingfordian of Montana) are more rounded, with robust, almost bulbous cusps. The ancestry of S .? russelli remains in doubt. It seems possible to derive the observed morphology from either a primitive Spermophilus pattern (in which case S.? russelli is rather special- ized, particularly for large size) or from the primi- tive Marmota pattern (in which case it is quite primi- tive, lacking specializations such as increased size of P/, which are seen in Hemphillian marmots). Given the present information, questioned placement into the broadly defined genus Spermophilus seems to be most convenient. Genus Ammospermophilus Merriam, 1892 Ammospermophilus hanf ordi new species (Fig. 12) Type-UW 33078: fragmentary skull including partial frontals, parietal, right I, p a.M3, left I and P4, complete mandibular dentition, most of right and part of left dentaries; several vertebrae and ribs, par- tial scapula, both humeri, part of radius and ulna, metacarpal. Type locality-UW A5927, Franklin County, Washington. Diagnosis-Largest species of Ammospermophil- us, skull apparently about size of skull of Spermo- philus kzteralis (Say) . Ammospermophilus generic TABLE 4 MEASUREMENTS OF Spermophilus? russelli N. SP. Specimen Number Tooth 33038 33062 33058 33079 33069 33076 40599 33040 33041 33052 33070 33077 P4 M2 / I / I pa p a p4 Ma I/ I/ I/ I/ AP 3.8 3.4 3.6 3.7e 4.8 14.6 3.7 2.8 3.1 3.5 3.4 3.8 3.7 3.8 1.8 1.8 3.0 3.9 3.8 3.9 4.1 4.1 T 2.4 3.4 3.8 3.9e 4.0 2.3 3.3 3.4 3.7 3.3 3.5 2.3 2.2 1.7 2.0 3.5 3.8 2.8 2.6 2.5 2.7 Mnd. depth (below P4) 10.8 10.5 1978 GUSTAFSON: RINGOLD FAUNAS 21 characters include: cheek teeth relatively small, low crowned; P 3 a simple single-cusped peglike tooth; metaloph of P 4-M2 separated from protocone by deep notch; protolophid absent and protoconid and para- metaconid appressed on P 4 ; interorbital region rela- tively wide; interorbital width narrower than post- orbital constriction; zygomatic notch opposite an- terior M1. Specific characters include relatively large size; mesostyles present on M1·2 ; metaconule very distinct; mandible relatively deep; small protoconu- lid present or absent, M1_2 with little anteroposterior compression. Referred specimens-The following are UW lo- calities and specimen numbers: A5927: 33048, P4- M3; 34953, M1 or 2 ; 34954, M1 or 2 ; 34956, M1 or 2 ; 34957, M2; 35229, P 4-M2 ; 35230, P4 ; 35232, P4 ; 40371, M1 or 2 ; 40584, M3 ; 40597, M3 ; 40601, Mn; 40608, M2; 40612, P4; 45291, M3 ; 45303, P4, 45311, M1 or 2 ; 45313, M1 or 2 ; 45315, M1 or 2 • A B Stratigraphic position-Ringold Formation, from siltstone 30 feet below White Bluffs tuff, early Blan- can. Description-The skull, as is apparent from the available fragments, is larger than in the Recent spe- cies of Ammospermophilus. Portions of the premax- illary, nasal, maxillary, frontal, parietal, and squa- mosal bones and the right bulla, are present and seem to indicate a skull the size of that of Recent Sperm- ophilus lateralis, or about 42 to 44 mm in maximum length. The frontals are a little wider at the interor- bital notch than in 5. lateralis. The bulla is very sim- ilar in size and form to that of S. lateralis, and does not have transverse indentations at the intrabullar septa as in Eutamias. The zygomatic notch is not as far posterior as in most Recent Spermophilus or as far anterior as in Eutamias or A. fossilis (Clarendon- ian, Cuyama Valley, California; James, 1963, p. 75), but is similar in position to that of Recent Ammo- Figure 12. Ammospermophilus hanfordi n. sp .. A, C,and 1) from holotype UW 33078. A. right P3-M3, occlu- sal view, anterior to right. B. right M10r 2, UW 34956,occlusal view, anterior to right. C. left P4-M3, occlusal view, anterior to left. D. left mandible, lateral view,anterior to left. Scale lines indicate one millimeter. 22 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 spermophilus, opposite the anterior portion of M1. pa is a simple peglike tooth, about 1/2 of the trans- verse width of P4. It is proportionately slightly larg- er than in Eutamias townsendi Bachman. P4 is highly variable in occlusal shape, from ovate to triangular, depending on the prominence of the parastyle and anterior cingulum. In specimen UW 45308 the anterior cingulum is quite small and 'there is no parastyle; in the type the anterior cingulum is larger and a very small parastyle is present; and in UW 40612 the parastyle is a prominent cusp nearly as large as the paracone. The lingual attachment of the cingulum is at the anterior middle of the tooth. The protocone is large, placed slightly forward of the middle of the lingual side of the tooth. The para- cone is slightly smaller than the metacone. The pro- toloph is lower than the metaloph and includes no protoconule. The metaconule is large and distinct. The metaloph does not connect to the protocone until late wear. The posterior cingulum is strong, and con- nects labially to the posterior side of the metacone. There is no mesostyle. M1 and M2 are not separable as isolated teeth. In the type, M1 is equal to M2 in anteroposterior length but slightly narrower transversely. The anterior cin- gulum is straight, nearly parallel to the protoloph, and joins the protocone with a sharp bend about one- fourth the distance from the lingual side of the tooth. The parastyle is distinct. The protocone is large, with the tip just anterior to the middle of the lingual side. The protoloph is a high straight ridge, with no proto- conule. A mesostyle is present, though small. The metacone and paracone are equal in size; the meta- conule is distinct and only slightly smaller than the TABLE 5 MEASUREMENTS OF Ammospermophilus hanfordi N. SP. Specimen 33078 33078 33048 Least lnterorbital breadth Depth of mandible below M1 Length of humerus Proximal max. diameter, humerus Distal max. breadth, humerus Tooth AP T LI/ 2.5 1.5 L p4 1.5 2.0 RI/ 2.5 1.5 pa 0.8 0.9 p4 1.5 2.0 M1 1.6 2.3 M2 1.6 2.4 Ma 1.9 2.2 p4.Ma 6.4 L / I 2.1 1.2 p4 1.6 1.6 M1 1.6 2.1 M2 1.9 2.3 Ma 2.4 2.1 P4-Ma 7.1 R / I 2.2 1.2 p4 1.6 1.4 M1 1.7 1.9 M2 2.0 1.9 Ma 2.2 2.1 P4 1.6 1.6 M1 1.7 2.0 M2 1.7 2.3 Ma 2.2 2.1 P4-Ma 7.2 11.0 5.7 27.3 5.2 7.0 1978 GUSTAFSON: RINGOLD FAUNAS 23 metacone. The metacone and metaconule are con- nected by a low ridge, and the metaconule is distinct- ly separated from the protocone until late wear. The posterior cingulum is rather weak and low. M3 is triangular, with a smaller parastyle and weaker anterior cingulum than the other molars. The anterior cingulum connects to the protocone with a sharp bend as in the other molars. The paracone is strong, and the protoloph is straight but lacks a pro- toconule. There is no distinct metacone, and the metaconule is very small or absent. The posterior cingulum is separated from the protocone by a dis- tinct notch; it is strong and curved strongly and evenly. A small mesostyle is present on 2 of the 4 examples. The posterior cingulum is separated from the paracone by a deep notch. The upper incisors are not grooved, unlike those of Eutamias. They are about the size and form of those of S. /,ateralis, with a more rounded anterior face than in Eutamias. The mandible is similar in size to that of S. later- alis, although the diastema is slightly shorter. The mental foramen is in the same position as in S. lat- eralis. The masseteric fossa is like that of Eutamias townsendi in form, but with stronger dorsal and ven- tral ridges. The cheek tooth row is relatively shorter than in Spermophilus, and the low- crowned teeth are similar to those of E. townsendi in general aspect. The lower incisors are nearly identical in size and form to those of S. lateralis; they lack the grooving seen in Eutamias. The incisors of E. townsendi are much less procumbent and more strongly curved. P4 is narrow anteriorly, with the protoconid and the metaconid closely appressed. A small anterior cusp (protoconulid of Bryant, 1945, fig. 43) is pres- ent on two specimens. The protoconid is larger at the base but lower than the metaconid. The protoconid and hypoconid are equal in size though the hypoco- nid is lower. The ectolophid is low; it is distinct on less-worn teeth but disappears with wear. The pos- terolophid is low but distinct, nearly semicircular, with no visible entoconid. The talonid basin is smooth. DP4 is similar to P4 but smaller, with sharper cusps, a relatively larger talonid basin, the ectolophid indistinct, and two anteroposteriorly flared roots. TABLE 5 (Continued) Specimen Tooth AP T 35229 p4 1.5 1.2 M1 1.8 2.1 M2 2.0 2.2 35230 P4 1.7 1.6 35232 DP4 1.3 1.2 45303 P4 1.5 1.3 40601 M1 1.8 2.1 M2 2.1 2.2 Ms 2.3 2.1 34957 M2 2.1 2.5 40608 M2 2.0 2.3 40612 p4 1.6 1.6 45308 p4 1.5 2.0 34953 Mlor2 1.8 34954 Mlor2 1.8 2.2 34956 Mlor2 1.6 2.0 40371 Mlor2 1.9 2.1 45311 Mlor2 1.9 2.3 45313 Mlor2 1.9 2.4 45315 M1or2 1.7 2.1 40584 Ms 2.1 2.0 40597 M s 1.9 2.2 45291 Ms 2.2 2.1 34981 Length of humerus 27.6 Prox. max. diameter, humerus 5.5 Dist. max. breadth, humerus 7.3 24 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 M1 is a suhquadrate tooth with a low protoconid, weak proh:>lophid, and a short, weak metalophid. The trigonid basin is small and is placed just anterom~- diad from the protoconid. The protoconid and meta- conid are placed slightly less far apart than the hypo- conid and entoconid, with the cusps set near the corners of the tooth. The metaconid is the highest cusp, though it is still rather low. The talonid basin is large and smooth. The ectolophid is quickly worn away. The posterolophid is low and only slightly curved. The ectoconid is distinct on less-worn teeth. A distinct notch is present between the entoconid and the posterior side of the metaconid. M1 is a little smaller in length and width than M2 but larger than P4. M2 is similar to M1 except that the protoconid and metaconid are placed further apart than the hypoco- nid and entoconid. The entoconid is less distinct than onM1. M3 is the largest of the lower cheek teeth. The tri- gonid is similar to that of M2, with a small basin an- teromediad from the protoconid. The hypoconid is large, the entoconid and ectolophid are small and dis- appear quickly with wear. The tooth shape tends to be more triangular than quadrate. A small mesostylid is present on less-worn specimens. The notch between the entoconid and the mesostylid is less deep than on M1_2• The posterolophid is strongly curved. The postcranial bones of A. hanfordi have not been compared to those of Recent Ammospermophil- us; thus a detailed description will not be attempted at present. An exception is made for the humerus, which is especially well preserved. Two complete humeri were found, one of which was associated ( with another partial humerus) with the type, the other found during screenwashing. They are nearly identical in size. The humerus is slender and is proportioned (as in other Ammospermophil- us) more as in Sciurus than in Spermophilus. The lesser trochanter is identical in form to that of Sciurus. The head of the humerus is ovate in dorsal view, narrower than that of Spermophilus, and very similar to that of Sciurus. The deltoid crest is similar to that of A. f ossilis James ( 1963 ) , higher than in Sciurus. The lateral epicondylar ridge is about as prominent as in Sciurus and more prominent than in Spermophilus; however the proximal end of the ridge connects to the shaft farther distad than in Sciurus, so that the ridge flares more abruptly from the shaft. An entepicondylar foramen is present. The humerus resembles that of Sciurus in characters which are found also in other species of Ammospermophilus, and in species of Eutamias, particularly the slender proportions and ovate head. The ratio between the length of the humerus and the (estimated) length of the skull (James, 1963, fig. 32) is close to the regres- sion line drawn by James for the antelope ground squirrels ( A mmospermophilus) , chipmunks (Tami- as ), and ground squirrels ( Spermophilus) ; the hu- merus is relatively shorter than in A. fossilis, flying squirrels ( Glaucomys ) , and tree squirrels ( Sciurus) . Discussion-A. hanfordi, though slightly larger, was evidently rather similar to living Ammospermo- philus in structure. This genus is not now found north of the Great Basin; Recent species are adapted to hot, dry climates. A. hanfordi was (as is indicated by the morphology of the humerus ) probably a ground- dwelling form without marked adaptations for either arboreal or fossorial life, and which appears to have been rather abundant in at least some areas of the Ringold flood plain. Family Geomyidae Genus Thomomys Wied-Neuwied, 1839 Thomomys cf. T. gidleyi Wilson, 1933 (Fig. 13) Type locality-Hagerman, Idaho. Material-UW A5927: UW 40596, M1 or 2, UW A6503 : uw 35110, M1 or 2 ; uw 35111, P 4 ; uw 35139, M1 or 2 ; UW 35209, M1 or 2 ; UW 40427, ?M1 or 2 ; UW 45026, M1 or 2 • Stratigraphic position-Ringold Formation, from 30 feet below to 45 feet above White Bluffs tuff. Description-Several isolated cheek teeth repre- sent a small pocket gopher referable to the genus Thomomys. The teeth are similar to but slightly smaller than those of Recent Thomomys talpoides (Richardson, 1828) and T. gidleyi (early Blancan, Hagerman, Idaho ) and are nearly as hypsodont as T. talpoides. Dentine tracts extend the full height of all Figure 13. Thomomys cf. T. gidleyi. A. left M10r 2, UW 45026, occlusal and labial views, anterior to left. B. left P4, UW 35111, occlusal and labial views, an- terior to right. Line indicates one millimeter. 1978 GUSTAFSON: RINGOLD FAUNAS 25 TABLE 6 MEASUREMENTS OF Thomomys CF. T. gidleyi* Specimen Tooth AP UW 35111 p4 1.70 uw 35139 Mlor2 1.02 uw 35209 Mlor2 0.80 uw 40596 Mlor2 0.80 UW40427 ?Mlor2 0.89 uw 35110 M1or2 1.16 uw 45026 M1or2 l.ll * Measurements taken along occlusal plane. teeth. The shape of the occlusal surfaces, including lingual constriction of the lower molars and labial constriction of the upper molars ( a generic character of Thomomys), is variable, as in T. gidleyi. One tooth (UW 35209), evidently in a late stage of wear, has closed off the lower end and formed a vestigal root. The latter tooth shows no evidence of constriction, and could represent another geomyid. The P 4 , while approximating the occlusal form of P 4 of T. talpoides, differs in several respects. The an- terior surface as well as the sides of the reentrant angles, are less strongly curved than in T. t,alpoides. The tips of the reentrant angles are further apart, leaving more room for the dentine connecting the anterior and posterior loops. A P4 from a young in- dividual of T. gidleyi is similar to but slightly smaller than the White Bluffs specimen. Discussion-Except for size ( which is not greatly different) these molars compare well with those of T. gidleyi Wilson (Zakrzewski, 1969, Text-fig 3). The Ringold specimens may or may not be specific- ally separable from T. gidleyi, but they certainly represent a species at very nearly the same stage of evolution. The presence of a root on the tooth of an old individual may be an indication of a slightly more primitive condition than in T. gidleyi. Family Castoridae Genus Castor Linnaeus, 1758 Castor californicus Kellogg, 1911 (Fig. 14B-E) Type locality-North end of Kettleman Hills, Fresno County, California; "upper Etchegoin," Pec- len coalingensis zone, late Pliocene. Material-UW A6571: 50389, right M1 or 2 . UW A8819: 50387, right M 1 or 2, and 50388, right M1 or 2· UW A9305: 51749, M3 , 51750, mandible fragment, T 1.56 1.51 1.78 1.47 1.78 1.51 1.47 T. gulleyi (range) Hagerman, Idaho AP T 1.60 1.45 1.16-1.36 1.81-2.05 and 51751, I fragment, all probably from one indi- vidual. Stratigraphic position-Ringold Formation, all specimens probably from above White Bluffs tuff, early Blancan. Description-The specimens listed for this taxon are generally slightly larger than in available Recent adult specimens of Castor canadensis. The incisor is identical in size to that listed for C. californicus by Stirton (1935, p. 445, T=8.8 mm). The enamel pat- tern of the molars (Fig. 14 c, d, e) is generally simi- lar to that seen in both C. canadensis and C. califor- nicus; the sample is too small to allow differentiation. The stria and striids are shorter than seen in C. cana- densis and similar to those of C. californicus. The jaw fragment (UW 51750) has two distinctive features: size (larger than in C. canadensis), and the shape of the mental process. In C. canadensis this process is concave on the posterior side, while the anterior side blends smoothly with the anterior bor- der of the jaw. In contrast, the mental process in the Ringold specimen is larger, forms a distinct obtuse angle with the anterior jaw, and has a straight pos- terior border; thus it seems to jut forward much more strongly. A similar mental process is seen in a specimen from Grand View, Idaho (Shotwell, 1970, fig. 15) which was assigned to C. accessor Hay. The mental foramina on UW 51750 are adjacent to the dorsal border of the incisor and are farther dorsal than in most specimens of C. canadensis. Discussion-Two similar-sized Blancan species of Castor have been described from North America. Castor californicus Kellogg (1911) from the "upper Etchegoin" beds of the Kettleman Hills, California, and C. accessor Hay (1927) from Froman Ferry, Idaho, each had single cheek teeth as holotypes. Stir- ton ( 1935) described additional material ( appar- ently topotypic) of C. californicus. No topotypic 26 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 TABLE 7 MEASUREMENTS OiF Castor californocus FROM THE RWGOLD FORMATION Spec,imen Tooth AP s~ M1or2 8.le occlusal ,'517i~ 503&7 50389 Mg Mlor2 M1or2 >3!9-) occlusal '9..'a-e • occlusal "7. [© occlusal 7..'S maximum 10:0e maximum near base 7. 9 occlusal 9.6e occlusal 7.1 occlusal 11.0 maximum near base specimens of C. accessor have been described, al- though Shotwell (1970) assigned a num'her of-speci- mens from Grand View, Idaho, to this !Wecies. This assignment was based on geographic pnoximity and length of stria and striids ( terminology of Stirton, 1935), which in the Grand View specimens are inter- mediate between those of C. californi.cus and the liv- ing C. canadensis. Both of the Blancan species have stria and striids shorter than C. canadensis (in this respect like the White Bluffs specimens) , hut no adequate quantita- tive data is available for lengths of these structures. Because C. californicus appears to he the better founded as well as the earlier name, I prefer to as- sign the White Bluffs specimens to this species. Genus Dipoides Schlosser, 1902 Dipoid,,es ,rexroadensis Hibbard and Riggs, 1949 (Fig. 14A) Type locality-Keefe Canyon, Kamsas, 1University of Kansas lloeality 22, Rexroad Formattion, 1la1e Plio- cene. Materud----OW A9849: 45673, right lo;wer jaw with broken I, complete P 4-M3, found by Marvin Gross, April 1967. Stratigraphi.c positio-Ringold Formation., texact level not known, hut probably above lev.el of White Bluffs tuff, early Blancan. Description-Dental terminology after Wood- hurne, 1961. The mandible is broken behind Mg. The remaining portion is close in size to the mandible of Castor canadensiis, and its form is similar to that of the man- dible of Procastoroides sweeti (Hibbard, 1972, fig. 36D ) . A well defined fossa for attachment of the an- terior ma-sseter muscles is present on the lateral side between miterior P 4 and the middle of M2 • There are four ment:al foramina near the base of P4 , one (an- terior) lall'_.ge, the others smaller and 1j ust posterior to the large one. A mental process ·smaller than that present in Castor canadensis prqjects below the ven- tral hordffl" mt .the posteroventr.al end ,of the sy.mphy- sis. The amtrurodorsal border between 1P 4 and the in_ cisor is SmlllJP!Y angled, more so '1lhan in P. sweeti or P. idahoerosiils 'Shotwell from Grand Viiew, Idaho. The anterior ro@t rcff the coronoid prucess is situated op- posite the J>9Btenior portion of M 1 . The lower iiindisor is rounded (!'1D its wentrolateral , side, with smoath enamel like oilier ,SJJecies of Di- rpoides and P. JS.WJJ.eti, and unlike C~r..Osterior end Palate length, maximum, from anterior maxillaries to posterior end Diastema length, post. C. to ant. M 94 119 35 Mandible UW 32704 Width, maximum preserved Width of dentary at M2 Lel:\gth of all alveoli, M1-a Length of diastema 92 R33 R55 R22 R54 /::;,. ++ ® ® ++ EB EB L34 L57 L20 D.llJ>th of jaw at M1 Length of symphysis, taken along anterior edge 68 x x The upper caniniform of M. rohrmanni is antero- posteriorly elongated, as in all species of Megalonyx except M. mathi.si Hirschfeld and Webb (1968, upper Mehrten Fm., Calif., Hemphillian), and a median lingual bulge is present, though variable. On the type specimen, the lingual bulge is very pronounced, giv- ing the tooth an almost triangular section, while on the juvenile (UW 40471) and on another adult (UW 42146) the lingual bulge is relatively low. The lin- gual bulge is bordered by wide, very shallow grooves, not nearly so deep and prominent as seen in M. jef- fersonii Harlan ( 1825, Pleistocene of W. Virginia; see Leidy, 1855, plate 16). A wide, shallow groove is present on the labial side of the adult caniniforms. EB The upper molariform teeth are very similar to those seen in other species of Megalonyx; the first three are straight, while the last is somewhat curved, concave anteriorly. M1 is wider and more squared on the labial side than the M1 figured by Leidy ( 1855, plate 3) in a skull of M. jefjersonii, resembling more closely the tooth in his plate 16, fig. 9d. In the type of M. rohrmanni, the anterior side of aU the upper molariforms is less convex than shown by Leidy (plate 6, fig. 1, and plate 16, figures 9 and 10). These teeth are variable in shape, however, and no great significance can he placed on the observed differences in the shape of the cheek teeth. The upper molari- forms are not as proportionately wide as in M. lep- tostomus Cope ( 1893) from the Blanco local fauna of Texas. @ @ X 11. lep±ostom1As @ j .... venile /::;,. ,M. sp. (Ha~er man) + M~9a.lon~X rohrmanni EBjuv. l'I +-~~~~----=;:===:::::;====:::::;:::====::;:::====:::;-~ Figure 20. Comparison of measurements of molari- form teeth in Megalonyx rohrmanni n. sp., M. lepto- stomus Cope, and a specimen of Mega/onyx from Hagerman, Idaho. 10 fl. l'I 16 AP 36 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 The shape of the anterior portion of the zygomatic arch of M. rohrmanni differs from that seen in other described species of Megalonyx. In M. rohrmanni the anterior surface of the zygomatic arch merges smoothly with the lateral surface of the maxilliary near the base of C1• This anterior surface descends posterad and slightly laterad, initially at an angle of about 45° (the palate between the cheek teeth is considered horizontal), then curves posterad to an angle of 25° to 35° . In M. jefjersonii, the anterior surface of the zygomatic arch merges with the maxil- lary, at an angle of 80° to 90° from the horizontal, then curves slightly to descend at an angle of about 45° . In M. leptostomus the anterior zygomatic arch flares abruptly from the maxillary surface and pro- jects slightly anterad (Hirschfeld and Webb, 1968 fig. 2). The anterior zygomatic arch in M. mhr- manni is slightly more prominent laterally than in M. jefjersonii (Leidy, 1855, plate 6, fig. 2). Only two of the lower teeth of M. rohrmanni are available, a lower caniniform and a cheek tooth (probably M2 or M3 ), both from adult individuals. The mandible (UW 32704) includes alveoli of all lower teeth. The lower caniniform is an anteroposte- riorly elongated ·oval in shape, very flat on the labial surface, with a slight bulge on the lingual surface. The tooth shows little, if any curvature along its length. The cheek tooth alveoli increase slightly in anteroposterior diameter from first to third. The M2 or M3 (UW 41523) is oval in outline, slightly nar- rower on one side. It is very slightly curved, though evidently not so much as in M. curvidens Matthew (1924, Upper Snake Creek Beds, Nebraska). The anterior border of the symphysis of the man- dible (UW 32704) is nearly straight for the lower three-quarters of its length, and projects forward more than in Leidy's specimens of M. jefjersonii (1855, plate 1). It resembles more the mandible of M. leptostomus figured by Meade (1945, plate 55), and even more the specimen from Hagerman, Idaho, (USNM 13477) which Gazin (1935) assigned to M. leptonyx? The anterior point of the symphysis of M. rohrmanni is not spread laterally as in the man- dible of M. curvidens figured by Hirschfeld and Webb (1968, fig. 6). The dorsal surface of the man- dible on the labial side of the cheek tooth alveoli is narrow and rounded in M. rohrmanni, not shelflike as in the mandible of M. leptostomus (Meade, 1945, plate 55), resembling, again, Gazin's (1935) man- dible from Hagerman, Idaho. The diastema of the Ringold specimen is relatively short and wide, with a distinct sulcus on the labial surface. Discussion-There is no species described in the literature whose characters match those of the speci- mens from the Ringold Formation. The sloth speci- mens from the Glenns Ferry Formation near Hager- man, Idaho (Gazin, 1935) come closest to M. rohr- manni in size ( fig. 15) as well as in characters of the mandible. Gazin referred the Idaho specimens to M. leptonyx? (Marsh, 1874). Hirschfeld and Webb (1968) considered M. leptonyx a nomen dubium because the type, an ungual phalanx, is probably not specifically determinable, its age and locality are uncertain, and the type specimen has been misplaced since before Gazin's paper of 1935. It may be that the Hagerman and Ringold sloths are conspecific; skull material of the Idaho sloth is needed for com- parison. There is no way of ascertaining specific identity or lack of identity between the Ringold speci- mens and Marsh's type of M. leptonyx, or, for that matter, between the Hagerman sloth and M. leptonyx. It seems altogether preferable to describe the excel- lent material from the Ringold Formation as a new species, and, for the present, to apply the name M. leptonyx only to the (missing) type specimen of that species. Order CARNIVORA Family Canidae Genus Canis Linnaeus, 1758 Canis davisi Merriam, 1911 (Figure 2la) Material-UW A3036: UW 41945, M1 fragment. UW A3037: UW 41969, M1 • UW A5134: UW 28747, lower C fragment. UW A 5927: UW 33060, P 2, UW 33061, P 4• UW A6503: UW 35114, P4, UW 40393, M1• UW A8810: UW 40288, P4, UW 40289, lower C fragment. UW A9293: UW 42135, M1 fragment. Stratigraphic position-Ringold Formation, from 35 feet below to 45 feet above White Bluffs tuff, early Blan can. Description-P2 (UW 33060) does not show the small posterior cusp often present in C. latrans, al- though there is a slight widening of the posterior ridge. The slight cingular protocone seen in C. latrans is less distinct on UW 33060. One P4 (UW 35114) is slightly concave on the anterolabial surface, but the other Ringold teeth are flat to slightly convex on this surface as is usual in C. latrans. The Ringold specimens are almost iden- tical to a P 4 of C. davisi (UW 40067) from Westend Blowout, Oregon (late Hemphillian) and to speci- mens described by Shotwell ( 1970, p. 73). M1 shows little other than smaller size and less robust character to distinguish it from the M1 of C. latrans. The talonid is slightly more rounded poste- riorly, and on UW 41969 the entoconid is small and appressed against the hypoconid. The hypoconid is not quite as prominent as in C. latrans and does not I . ., 1978 GUSTAFSON: RINGOLD FAUNAS 37 show a labial basal swelling as in the modern species. The Ringold specimens lack any sign of a cingulum such as that seen on both labial and lingual sides of M1 of a specimen (UMMP V53910) from the Hager- man fauna , which Bjork (1970, p. 13) referred to C. lepophagus Johnston (1938, type specimen from Cita Canyon, Texas). The Hagerman Canis Mi, though near the size of the larger Ringold M1 (UW 41969), is wider and more robust. Discussion-The material from the Ringold Form· ation referable to Canis is indistinguishable from specimens referred to Canis davisi from the Hem- phillian of northern Oregon. All of the Ringold speci- mens are smaller than any specimen of the living Canis latrans or the extinct late Blancan C. lepopha- gus which were available to me. Merriam (1911) described Canis davisi from a maxillary fragment containing M1 and M2 , which was found downhill from outcrops of the Rattlesnake Formation (Hemphillian ) in north-central Oregon. Shotwell (1970) added a number of specimens from other Oregon Hemphillian localities to the hypodigm of the species, including a partial skull with complete dentition. In addition , the Burke Museum collections from W estend Blowout ( UW A8803) and Krebs Ranch include specimens referable to C. davisi. In total, this material is quite adequate for identifica- tion of the White Bluffs Canis materials despite the absence of any upper molars in our collection. Genus Borophagus Cope 1892 Borophagus sp. (Figure2lb) M aterial-UW A6503: UW 35115, left P 4• Stratigraphic position-Ringold Formation, 40 feet above White Bluffs tuff. Description-This unworn tooth is dominated by the massive central cusp. A distinct ridge connects the tip of the central cusp to a low swelling at the anterolingual corner. This swelling continues post- erad along the lingual enamel border, rising at the posterolingual corner to form the rim of the heel or posterior cingulum. The posterior cingulum is wide transversely and is curved convexly upwards. A short ridge is present on the posterior side of the central cusp. Squeezed in between the central cusp and the posterior cingulum is a much smaller posterior cusp. The labial surface of this tooth is very convex, while the lingual surface is slightly concave in profile. Discussion-This specimen is evidently from a borophagine dog, and agrees in general with various species of the Hemphillian genus Osteoborus in the possession of a small cusp on the posterior side of . the P 4 • However, it also agrees in this character with a specimen from the late Blancan Grand View local fauna of Idaho, identified by Shotwell (1970) as Borophagus sp., and differs from other specimens of Borophagus and Osteoborus progressus Hibbard (1944), which lack a posterior cusp on P4 • The size and relatively great width (Table 11) of the Ringold tooth support the reference to Borophagus; presence or absence of the small posterior cusp is probably less significant. The single P 4 is insufficient evidence for specific identification. TABLE 12 MEASUREMENTS OF Canis davisi Specimen number Tooth uw 28747 IC uw 40289 / C UW40393 M1 uw 41945 M1 uw 41969 M1 uw 42135 M1 uw 33060 p2 uw 33061 p4 uw 35114 p4 uw 40288 p4 AP 17.0 18.5 11.9 18.5 17.1 18.4e T H 4.7e 15.0 4.7e 16.8 6.3 6.3e 7.4 6.3e 4.2 8.7 8.0 7.9e 38 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 TABLE 13 MEASUREMENTS OF P4 IN Borophagus H (crown Height height of AP (Length- in plane at right angle to flat posterior surface) T (max. width of posterior along post. lingual post. cusp above tooth) surface) cingulum Borophagus sp. uw 35115 Ringold Fm. Borophagus sp. UO-Fl6343 Glenns Ferry Fm. Grand View, Idaho Borophagus diversidens Midwestern U. 8034 Blanco Fm. Osteoborus cyonides KU6791 Hemphill Fm. 18.5 19.7 20.0 14.85 Family Ursidae Gray, 1825 Genus Ursus Linnaeus, 1758 Ursus sp. cf. U. abstrusus Bjork, 1970 (Figure 21 c-d) Bjork, 1970, P. 16, figure 96. Material,-UW A3036: UW 41944, partial M1. UW A9292: UW 42328, partial M1. Stratigraphic position-Ringold Formation, level uncertain , early Blancan. Description-The first lower molar, though only an anterior fragment, is comparable to the type of Ursus abstrusus from Hagerman, Idaho, which it matches closely in size and morphology. The trigonid is simple, and is proportioned as in U. abstrusus. The paraconid is lower than the protoconid. A small metaconid, appressed to the posterolingual border of the protoconid, is slightly larger than that on U. abstrusus. Unlike U. americanus Pallas, no accessory cuspules are present anterior to the metaconid. The transverse measurement of the trigonid is 7.0 mm. The upper first molar is missing fragments from the posterolingual corner and the anterolabial side, but enough remains to show that the tooth was pro· 15.8 17.2 2.6 15.3 20.0 2.6 14.3 10.5 portioned as in U. americanus and is not as propor- tionately wide as in the tremarctine bears. The para- cone and metacone are set close to the labial edge of the tooth, not further lingually as in the tremarc- tines. The metacone is divided into two cuspules by a transverse groove. The protocone is elongate, and is split, as in the metacone, into two cuspules by a transverse groove. The anterior end of the protocone merges with a vague cingulum, which connects to a small parastyle on the anterior side of the paracone. Unlike U. americanus, there is no labial cingulum and only a suggestion of a lingual cingulum. The transverse measurement is 13.1 mm, and the pre- served anteroposterior diameter is 18.5 mm. The M1 is quite distinct from that of I ndarctos oregonen- sis Merriam, Stock, and Moody (Hemphillian of Oregon) in smaller size, narrower proportions, and and numerous details of enamel morphology (Shot- well, 1970, Fig. 37A). Discussion-The two available specimens are very incomplete, making it difficult to precisely assess affinities. The similarity of the M1 fragment to the type of U. abstrusus and the ursine proportions of the M1 suggest that a true ursine is represented. 1978 GUSTAFSON: RINGOLD FAUNAS 39 A D Figure 21. Carnivora. A. Canis davisi right M1, UW 40393, occlusal and labial views. B. Borophagus sp., P 4 , UW 35115, occlusal and labial views. C-D. Ursus sp. cf. U. abstrusus. C. Left M1, UW 41944, occlusal view. D. Left Mi, anterior fragment, UW 42328, occlusal and lingual views. Lines indicate one centimeter. Anterior to right in A, D; to left in B, C. Family Mustelidae Genus Trigonictis Hibbard, 1941 Trigonictis cooki ( Gazin), 1934 (Figure 22) Gazin, C. L., 1934, p. 142, fig. 2. Material-UW A3027: UW 41527, partial left mandible with P 4-Mi, and alveoli of P 3 and M2• Stratigraphic position-Ringold Formation, the single specimen was float, found 150 to 200 feet above level of road, probably from above White Bluffs tuff, early Blancan. Description-The single available specimen, a par- tial mandible, is almost identical in morphology to specimens of Trigonictis idahoensis and T. cooki from Hagerman, Idaho. The mandible appears to have been chewed upon by some predator or scavenger. Two distinct round punctures ( about 10 mm apart) and a crushed area slightly dorsal to the punctures are present on the medial side of the mandible. Additional grooves, apparently tooth marks, are seen near the posterior mental foramen. Figure 22. Trigonictis cooki, mandible with P 4-M1, UW 41527, occlusal and labial views. Line indicates one centimeter. Anterior to left. Discussion-Zakrzewski (1967a) andBjork(l970) found no consistent morphological characters other than size to separate the two species of Trigonictis described by Gazin (1934), T. cooki and T. i,daho- ensis. The Ringold specimen, with an M1 length of 12.0 mm falls into the gap (M1 AP length from 11.7 to 12.6 mm) used by Bjork to separate these species. The primary statistic favoring division of the Hagerman Trigonictis sample into two species is the high variability in size. The coefficient of variation (Simpson, et al., 1960) for the entire sample (AP length of 23 lower first molars plotted by Bjork) is about 10.1, higher than normally seen in dental char- acters in a concurrent population of one species. Zakrzewski presented statistics suggesting correctly that more variation in size is present in the Hager- man sample than is present in Recent sexually dimor- phic mustelids. In addition to the statistical evidence, there is stratigraphic evidence that at least two dis- tinct populations of Trigonictis are present in the Hagerman sample. Figure 23 plots altitude ( approximately=strati- graphic position) versus size for three parameters, . 40 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No. 23 P4 LENGTH M1 LENGTH MANDIBLE DEPTH x x x BELOW Ml 3200 - - - x x x x x 3100 - - - x x x % )()( ~ 3000 x x - - - x x x 2900 - - - x x x 6 .0 7 .0 8.0 12 13 14 12 IL! 16 (MM) Figure 23. Measurements of Trigonictis teeth and mandibles from the Glenns Ferry Formation at Hager- man, Idaho, plotted against altitude (in feet above sea level) . See text for discussion. AP length of P 4 and M1 and depth of mandible below anterior end of Mi, measured in 8 specimens of Tri gonictis from the UMMP Hagerman collection. The small size of this sample, despite the rather large number of Trigonictis specimens known from the , Glenns Ferry Formation, is due to lack of strati- graphic information about many specimens, includ- ing the types in the U.S. National Museum. An appar- ent increase in size with increasing altitude is evident in all three measurements, with the strongest increase above an altitude of 3050 feet. The significance of this size increase was determin.ed with the chi-square test (Simpson et al., 1960, p. 186). Subsamples were formed above and below the 3050 foot level and above and belowe 7.0 mm for P 4, 13.0 mm for M1, and 14.0 mm for mandible depth. The chi-square test for difference in M1 length is not significant at the 5% confidence level (null probability=.05-.10) ; however, the differences in P 4 length (null probabil- ity=.025-.05) and mandible depth (null probability= less than .005) are statistically significant. It is worthy of note that Galbreath (1972), in a review of Bjork's 1970 paper, also suggested that T. cooki and T. idahoensis were part of a single evolving population. He suggested including Bjork's genus Sminthosinis, a smaller Trigonictis-like ani- mal, in this evolving population, but since the Smin- thosinis specimens are from the upper part of the Hagerman section, the direction of size change would be reversed from enlargement to diminution through time. The samples of Trigonictis above and below 3050 elevation are distinct; they may represent simply two populations or the differences may be the result of a chronocline. Additional stratigraphically con- trolled material may show the latter to be the case. At present we can continue to recognize two arbitrar- ily partitioned species; smaller specimens (T. cooki) in the lower Hagerman section; large specimens (T. idahoensis) in the upper part of the section. The Ringold specimen of Trigonictis falls into the size range of specimens found below the 3050 foot level at Hagerman, and therefore can be placed in T. cooki. Family Felidae Gray, 1821 Genus F elis Linnaeus, 1758 Felis sp. (Figure 24) Material-UW A6571: UW 35258, anterior right mandible with broken canine, complete P 3_4_ Stratigraphic position-Ringold Formation, 45 feet above White Bluffs tuff, early Blancan. Description-This mandible fragment of a cougar- sized cat presents several unusual features. The P 3 I 1978 GUSTAFSON: RINGOLD FAUNAS TABLE 14 MEASUREMENTS, Trigonictis cooki Trigonu:tis cooki Ringold Fm. T. cooki* T. idahoensis* uw 41527 Glenns Ferry Fm. Glenns Ferry Fm. Anterior- Idaho posterior Transverse AP AP AP Length 6.5 12.0 Mandible 12. 7 Depth below M1 * from Bjork, 1970 Diastema, C-Pa Depth of jaw between Pa & P4 c AP T Pa AP T p4 AP T width 3.7 5.3 mean 11.4 range mean 10.7-11.7 13.7 TABLE 15 MEASUREMENTS OF Felis SP. Glenns Ferry Fm. Ringold Fm. F elis /,a,custris Felis sp. USNM 13743R uw 35258 from Bjork, 1970 23.6 11.4 27.7 23.8 14.9 12.2 11.0 9.0 13.5 11.6 7.0 6.4 16.9 15.6 7.7 7.3 Idaho AP range 12.6-14.9 Recent F. concolor uw 27392 12.6 24.2 15.2 9.9 14.2 7.1 17.0 8.1 41 and P 4 are within the size range of F elis concolor in anteroposterior and transverse dimensions, but the mandible is deeper and more elongate than in F. concolor. The diastema is relatively longer than in any species of Felis or Pseudaelurus of which I have seen a specimen or description; however, there is no flange along the ventral margin of the mandible, and this together with the form of premolars and the probably small incisors suggest a lack of any close relationship to any of the sabertooth cats. The sym- physis is rounded anteroventrally as in Lynx, F elis /,a,custris Gazin ( 1933, Hagerman, Idaho), and spe- cies of Pseudaelurus, and not angular as in F. con- color. The symphysis does not extend as far postero- ventrad as in F. concolor. One large mental foramen is located below the center of the diastema. Two very small additional foramina are present, one lower in Figure 24. Felis sp., mandible fragment with Pn, UW 35258, labial view. Line indicates one centimeter. Anterior to right. i 42 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 the mandible and below the middle of P 3, and the other slightly anterior to and slightly higher than the large foramen. Most cats have at least two large mental foramina; F. concolor usually has four nearly equal-sized foramina (Savage, 1960); F. lacustris, Panthera onca, Lynx, and Pseudaelurus generally have two subequal mental foramina. The type of Nimravides thinobates (Macdonald 1948, fig. 13, Clarendonian, Black Hawk Ranch, Calif.), has a single large mental foramen in a position similar to that of the Ringold specimen, as well as having an unusually long diastema; N. thinobates is, however, much larger. The anteroventral portion of the Rin- gold mandible is somewhat swollen, in a manner suggestive of some species of Pseudaelurus. A sharp dorsal ridge along the diastema is most prominent opposite the small anterodorsal mental foramen. The incisors are missing. The alveolus of h is set behind the other two alveoli, as is seen in Lynx and F. lacustris. The alveolus of Ia is largest of the three. The canine is broken off at the base of the enamel. It is relatively wider than the canine of F. concolor, and is similar in proportions to the canine of F. la- custris though larger. Neither P 1 nor P 2 were present. Both Pa and P 4 are large, high-crowned teeth. Their height is considerably greater than in F. concolor or other Recent F elis, and they approach the propor- tions found in the much smaller Lynx. Pa is narn;>wer anteriorly than in F. concolor. P3 has a strong poste- rior cusp, hut a swelling on the anterior end of the tooth is not cuspate. P 4 is much larger than P 3, with a strong anterior and posterior cusps, and a tiny cuspule on the posterior cingulum. There is consider- able vertical relief in the lower enamel borders of the Ringold specimen; the enamel descends much lower over the roots and is higher between the roots than in F. lacustris or concolor, being more like the condition seen in Lynx. Discussion-This specimen shares characters of several taxa of cats, but is probably more closely allied to F. lacustris and the Lynx-like cats than to F. concolor. There are several characters shared with Pseudaelurus, but with the lack of anterior pre- molars and the late age of the deposits, it is better placed in F elis. Although the species is probably new, I prefer to defer formally recognizing it until it can be determined whether the unusual mandibular diastema is a normal feature. Subfamily MACHAIRODONTINAE Gill 1872 genus and species indeterminate Material-UW A3028: UW 41532, upper canine fragment. UW A3037: UW 41970, upper canine fragment. Stratigraphic position-Ringold Formation, early Blancan. Description-Fragments of two large, transversely flattened canines demonstrate the presence of a ma- chairodont in the White Bluffs local fauna. At the base of the enamel, the better preserved fragment (UW 41970) has an anteroposterior diameter of 26.3 mm and a transverse diameter of 10.5 mm. The other (UW 41532) has a transverse diameter of 12.1 mm. No serrations are visible on the damaged anterior and posterior margins of the fragments. They are slightly smaller than upper canines of Homotherium serum (Texas Memorial Museum no. 933-3582, left canine AP=31.3, T=l3.2, from Friesenhahn Cave, Texas, late Pleistocene) and appear to have tapered to form a similar rather short blade. H. serum canines have serrations almost to the base of the anterior enamel. This is a consider- ably smaller animal than ? I schyrosmilus idahoensis Merriam (1918) of the Grand View local fauna, which has canine measurements of 38.0 and 18.2 mm respectively (Shotwell, 1970). The Ringold speci- mens are approximately the size of the canines of Barbourofelis morrisi Schultz, Schultz, and Martin (1970, Hemphillian, Ash Hollow Fm. Nebraska), but they lack the lateral longitudinal grooves of Bar- bourof elis. Order PROBOSCIDEA Family Mammutidae Genus Mammut Blumenbach, 1799 M ammut americanum (Kerr) Material-UW A3031: UW 42668, thoracic ver- tebra centrum. UW A5927: UW 34939, tooth frag- ment. UW A8812: UW 42171, ?tooth fragment. Stratigraphic position-Ringold Formation, from 35 feet below to 45 feet above White Bluffs tuff, early Blancan. Discussion-The mastodon material from the Rin- gold Formation available in the Burke Museum col- lections alone is not diagnostic. M. americanum was reported by Strand and Hough (1952). The identi- fication was based on an excellent specimen (USNM 19101), which includes p11rts of two tusks, right M1-2 , left M1·3 , and M2·3 from both sides, along with parts of both mandibular rami. I could find no major differences between USNM 19101 and undoubted specimens of M. americanum from the late Pleisto- cene, and therefore concur with this identification. The early Blancan date given to these specimens is unexpectedly early for this species. Webb (1974) records M. americanum from the early Pleistocene (Nebraskan-Aftonian) Santa Fe River IB fauna of Florida, a later Blancan locality. Other reports of 1978 GUSTAFSON: RINGOLD FAUNAS 43 Blancan Mammut (Blanco local fauna, Texas; Broad- water quarries, Nebraska; and Hagerman local fauna, Idaho) give no species name. The Blanco specimens are probably not Mammut. Order PERISSODACTYLA Family Equidae Genus Equus Linnaeus, 1758 Equus (Dolichohippus) sp. cf. E. simplicidens Cope, 1893 (Figure 26) Material-UW A3029: UW 41553, fragmentary mandible with dP2 , anterior half of dP3, dP4, M1 , and M2• UW A3030 : UW 41922, astragalus. UW A3036: UW 41943, lower cheek tooth fragment; UW 42331, P3 or 4. UW A3340: UW 41991, P3 or 4• UW A6571: UW 45033, maxillary fragment with M1-3 • UW A- 9292: UW 42321, mandible fragment with P3-M1 : uw 42323, M3 : uw 42324, M1 or 2 • uw A9293: UW 42136, M1 or 2 • Stratigraphic position-Ringold Formation, up to 45 feet above White Bluffs tuff, early Blancan. Description- All specimens seem to pertain to a single large species of horse with very hypsodont teeth. The measurements (Table 14) fall into the range of E. simplicidens from the Broadwater quar- ries of Nebraska (Howe, 1970). The deciduous lower premolars, heavily worn in the specimen from the Ringold Formation (UW 41553), are similar in most details to the equivalent teeth of E. simplicidens from Hagerman in a similar state of wear (Gazin, 1936, fig. 22). The dP2 in material from the Ringold Formation appears less pointed anteriorly because of a relatively wider pro- toconid. The median valleys are slightly deeper and straighter than in E. simplicidens at a similar wear stage. The flexure between the metaconid and meta- stylid is U-shaped in dP4 but V-shaped in dP2• The permanent lower premolars (UW 42321) are little worn, so that the enamel pattern can not be entirely determined from the occlusal surfaces. The broken lower ends, however, show some details. A number of minor crenulations visible on the occlusal surfaces disappear further down the teeth. The lin- gual flexure between metaconid and metastylid is V-shaped, and the median valley does not extend between the metaflexid and entoflexid. A small para- stylid is visible in both P 3 and P 4 near the bases of the teeth. The metaconid is slightly longer antero- posteriorly than the metastylid. There is no sign of a pli caballinid on P 4, but one is present on P 3 in an early stage of wear. Four lower molars are available, two little worn first molars, an unworn M2, and a moderately worn M1 or 2. The molars are · longest near the occlusal surface, due to the anteroposterior elongation of the hypoconulid. Further down the teeth, the hypo- conulid becomes flattened posteriorly, and the teeth become anteroposteriorly shorter. The teeth definitely identified as M1 have a rela- tively simple pattern, lacking a parastylid or pli caballinid; it is not unusual for the pli caballinid to be absent on M1 but present in premolars and M2_3 in E. simplicidens (Skinner and Hibbard, 1972, fig. A Figure 26. E quus ( Dolichohippus) sp. cf. E. sim- plicidens. A-B. M10r 2 , UW 42136, occlusal and an- terior views. C. p 3or4, UW 41991. D. p 3or4, UW 42331. E. Unworn M10r 2, UW 42324, occlusal view. F. M1or2, UW 45103, occlusal view. Lines indicate one centimeter. Anterior to left in A, F; to right in C. D, and E. I I 44 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No. 23 57B). The lingual flexure between the metaconid and metastylid is V-shaped, as in the premolars. On UW 41553 the median valley extends between the metaflexid and entoflexid, but on UW 42321 it does not extend as far linguad. The M1 or 2 (UW 45103) features a small pli caballinid and more rounded metaconid and metastylid. The median valley extends well between the flexids. The protoconid and hypoconid of the Ringold specimens seem to be narrower than in Hagerman E. simplicidens, but this feature varies with wear, and the enamel patterns of the lower molars of the Ringold and Hagerman specimens seem otherwise similar. The characters of the upper cheek dentition show considerable variation. The premolars tend to be slightly larger and less curved than the molars. Two of the molars are strongly curved. The mesostyles vary in height, being weaker in the Ms and some badly worn teeth, but quite strong in most, with the tip rounded, pointed, or in one premolar, expanded as in most species of Equus. The parastyle is strong and either rounded or flattened at the tip. The pro- tocone is anteroposteriorly shorter in the premolars than in the molars ; however, it is distinctly grooved in only one tooth of the eight with the protocone preserved. The protocones are consistently much shorter anteroposteriorly than in Recent Equus, and than in E. simplicidens from Hagerman. The form of the protocones resembles most closely those of the type specimen of E. (D.) francescana (Frick, 1921, p. 322, Blancan, San Timoteo beds, Califor- nia). The pli caballini is absent on one worn molar but is present on the other teeth; however, it tends to be small. The hypoconal groove is highly variable and may be absent in some individual teeth. The pre- fossette consistently has one or two pli protoconules, and a pli prefossette is present on five of nine teeth. Postfossette structure tends to be simpler; five of nine have no accessory crenulations, one has a pli hypostyle, and three have a pli postfossette. Discussion-The usage of the name Dolichohippus for those North American horses previously placed in the subgenus Plesippus follows Skinner (in Skin- ner and Hibbard, 1972, p. 118) . The horse from the Ringold Formation is con- sidered to belong in the Dolichohippus group because of the following characters: large size, great hypso- TABLE 16 MEASUREMENTS1 OF TEETH OF Equus FROM THE RINGOLD FORMATION Specimen number Tooth AP2 TS H4 41553 dP2 33.9 13.7 11 dPs 14.0 14 dP4 30.1 14.2 16 M1 32.2 13.5 78 M2 32.0 11.8 67+ 42321 Pa 31.3 17.0 63+ p4 28.6 15.6 64+ M1 27.1 14.7 69 45103 M1or2 23.7 15.8 41991 p sor4 31.0 24e 27 42331 p sor4 28.1 26.1 43 42136 Mlor2 26.5 25.8 77 42324 Mlor2 30.6 21.3 69+ 45033 M1 21.4 25.0 28e M2 24.1 25.3 26e Ms 27.4 23.3 23e 42323 Ms 30.8 26.3 14 1 In millimeters. 2 AP measured through middle of tooth parallel to occlusal plane and parallel to tooth row. S T measured across mesostyle and protocone. 4 H measured between lowest enamel and highest cusp on lingual side of lower teeth, labial side of upper teeth; necessarily approximate. Plus sign indicates a tooth on which the roots had not formed. 1978 GUSTAFSON: RINGOLD FAUNAS 45 donty (maximum height of cheek teeth over 80 mm), presence ( though not dominance) of elongated, grooved protocones, presence of pli caballinid, sharp metaconid-metastylid groove, and extension of me- dian valley between flexids of lower molars. The upper cheek teeth from the Ringold Forma- tion match the illustrations of E. (D.) francescana (Frick 1921, figs. 1, 2, 3, and 4 of Folder 3) and UW specimens assignable to E. (D.) proversus (Merriam 1916) from the Tehama Formation of California (UW A3198) more closely than tthey match either specimens or illustrations ( e.g. Gazin, 1936, fig. 21) of E. (D.) simplici-dens from Hagerman, Idaho, the Broadwater quarries of Nebraska (Howe, 1970 p. 961, Text fig. 1), and other localities. Merriam and Buwalda ( 1917) briefly described some horse teeth found in the Ringold Formation. Stratigraphically, their specimens were from just above the White Bluffs tuff; the site can be located from a published photograph, their Plate 13, figure 1. Their description suggests a more advanced horse than is indicated by the Burke Museum specimens. 20 AP(mm) 30 WB +I 8 152 H 7 WB +I P.q 8 151 WB 2 8 152 WB +1 B I "I .q H 7 WB --t- 2 8 132 H 6 WB "' 8 153 H 6 WB --+-2 8 70 Figure 25. Ranges and means of anteroposterior tooth measurements of Equus simplicidens from the Broadwater Quarries (B), Nebraska (Howe, 1970), Hagerman (H), Idaho (Gazin, 1936) , and of E. cf. simplicidens from the White Bluffs (WB), Washing- ton. Horizontal lines indicates ranges, vertical lines indicate means; numbers at right of horizontal lines are numbers of specimens measured. Merriam (1918) compared the White Bluffs teeth to E. idalwensis Merriam, from Idaho. Although the Idaho form is advanced (see descriptions by Shot- well, 1970) with respect to the Burke Museum speci- mens, Merriam's Ringold specimens seem to be more so. This record suggests that either more than one species of Equus is present in the Ringold Formation, or that the range of variation in the species is not adequately covered in the materials available to me. Savage (1951, p. 217) believed that given a rea- sonable amount of variation within a species, most of the plesippine ( =dolichohippine) species could be referred to the earliest-described species, E. (D.) simplicidens Cope (1893). This latter opinion has been accepted here, with the reservation that more complete statistical studies of the sort done on Mery- chippus by Downs (1961) may make a subdivision of the group possible. Order ARTIODACTYLA Family Tayassuidae Genus Pwtygonus Le Conte, 1848 Platygonus pearcei Gazin, 1938 (Figure 27) Material-UW A3024: UW 42000, mandible frag- ment with P4-M2• UW A3036: UW41942, / C frag- ment. UW A3037: UW 41972, molar fragment. UW A5136: UW 28734, M1. UW A5137: UW 28733, molar fragment. UW A5927: UW 34966, M1 : UW 35231, P/ ? fragments. UW A6503: UW 35099, C/ : UW 35115, / P: UW 35216, maxillary with P2-M3 : UW 35246, / C. UW A6520: UW 40301, M/ frag- ment: UW 40302, M1 • UW A6571: UW 35257, / M fragment. UW A8807: UW 42161, M? fragment. UW A8809: UW 41516, M2 fragment. UW A9292: UW 42327, M3• Stratigraphic position-Ringold Formation, 40 feet below to 45 feet above White Bluffs tuff, early Blancan. Description-Gazin ( 1938) did not describe the dentition of P. pearcei from Hagerman in detail. The following description is intended to cover the White Bluffs material but will hold for the Hagerman sample except where noted. Three canines, one upper and two lower, have been damaged or badly eroded, but appear to have been similar in size to those of P/,aty- gonus pearcei Gazin ( Hagerman, Idaho) . The upper premolars show progressive molar- iformity from front to rear, manifested in in- creasing rectangularity and size of the posterior cusps (metacone and hypocone) from P2 to P4. The upper teetth differ only slightly from specimens of P. pearcei in the collections of the University of 46 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 TABLE 17 MEASUREMENTS OF TEETH OF Platygonus pearcei AND P. bicalcaratus ( MAXIMA AND MINIMA ONLY) A= P. pearcei, White Bluffs, Ringold Formation, Washington. B= P. pearcei, Hagerman, Idaho, UMMP collection and Gazin, 1938. C= P. bicalcaratus, Keefe Canyon, Rexroad Formation, Hibbard and Riggs, 1949. D= P. bicalcaratus, Blanco Formation, Texas, from Gidley, 1903. Tooth locality AP T no. of specimens p2 A 10.8 9.7 1 B c 9.5.11.8 7.5-11.0 3 D 11 -11.5 10.5-12 2 pa A 11.4 11.2 1 B 10.le 10.8e 1 c 10.8-11.8 9.8-12.4 3 D 12 -13.5 12 -13 2 p4 A 11.7 12.0 1 B 10.2-11.2 12.8-13.2 2 c 11.6-14.0 11.8-13.7 3 D 12 -14 14 2 M1 A 13.5-16.3 12.5-14.7 3 B 14.8 12.6 1 c 14.1-16.1 12.5-12.6 2 D 16.5-17.5 15 -15.5 2 M2 A 18.7 16.4 1 B c 18.0-18.4 15.0-16.5 2 D 19.5-20 17.5-20 2 Ma A 23.1-30.5 17.0-20.5 2 B 21.8-23.0 17.0 2 c 21.0-21.5 16.5-17.4 2 D 20 -25.5 16 -21.5 2 p4 A 14.3 10.7 1 B 11.5-15.4 12.0-14.1 6 c 12.2-13.0 10.5-12.5 3 M1 A 15.1 11.3 1 B 15.1-16.2 11.4-14.3 6 c 14.8-16.0 11.5-12.2 3 M2 A 18.7 14.6 1 B 17.3-21.5 13.3-17.5 4 c 18.2-19.5 14.0-15.5 3 Michigan Museum of Paleontology, and are within the range of variation of that species. paracone, with a very small divided cusp in the posi- tion of a metacone. The paracone is slightly more anterior and higher. A heavy cingulum is present on all sides except the anterolabial. P2 is a rounded triangular tooth, widest posteri- orly. There are two distinct cusps, a protocone and - - ----- 1978 GUSTAFSON: RINGOLD FAUNAS 47 pa is more quadrate than P2 , and is widest an- teriorly. The protocone and paracone are large and subequal; the metacone an.cl hypocone are distinct but much smaller, and the hypocone is set well in from the lingual side. The cingulum is strong, es- pecially on the posterior border, but is absent on the lingual side of the protocone and the labial side of the paracone. P4 is distinctly rectangular, widest anteriorly, and is larger than p a_ The metacone is proportionately larger than in P3, considerably larger than the hypo- cone. The cingulum is strong anteriorly and on the labial side of the metacone, but is weaker posteriorly than in pa_ The posterior cusps on P3 ·4 are apparently larger B and more distinct than on specimens of P. bicalcara- tus Cope from the Blanco Formation of Texas illus- trated by Gidley ( 1903 fig. 1, 2 ) . M1 is badly worn on the complete series. It shows that a cingulum was present on the labial side but not the lingual. Two isolated MPs are available. UW 28734 is worn, though not as badly as in UW 35216. The cusps are subequal in size as in the other MPs, and a cingulum is present on anterior, labial, and posterior sides. UW 40302 is unworn; on it the cingulum is strongest on the anterior side, present on labial and posterior sides, and is present but weak on the lingual side of the protocone. The cusps are well separated and do not form distinct cross-lophs when unworn. The enamel is lightly crenulated. D Figure 27. Platygonus pearcei. A. P2-Ma, UW 35216, occlusal view. B. P4-M2 UW 42000, occlusal view. C. UW 42000, lateral view. D. Ma, UW 42327, occlusal view. Lines indicate one centimeter. Anterior to left in A, B, and D ; to right in C. 48 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 M2 is larger than M1, with the cusps subequal in size but spaced further apart anteriorly than posteri- orly. It is more lophodont than the unworn M1 . A cingulum is present around the labial half of the tooth and on the anterior side of the protocone. M3 is the largest molar. It tapers strongly towards the posterior end; the protocone and paracone are larger than metacone and hypocone. The posterior cusp or heel is part of a broad ridge or posterior cingulum which extends labiad and posterad from the posterior hypocone and connects with the labial cingulum. The labial and anterior cingula are simi- lar to those of M2• The enamel is crenulated where unworn . A second M3 (UW 42327) is questionably referred to this species. It is larger than any reported M3 of Platygonus, but could conceivably be a large variant of P. pearcei. The cusps are large and well separated, the cingulum is stronger than but in the same positions as in the smaller M3, and the enamel is heavily crenulated. Several smaller cusps are pres- ent on the heel, three on the labial side and one (relatively large) on the lingual side. P 4 is molariform in outline but with considerably higher anterior cusps than posterior. The protoconid and metaconid are subequal in size, as are the hypo- conid and entoconid. A small ridgelike cuspid is present near the center of the tooth. The labial cusps are well separated from the lingual cusps by medial grooves. A cingulum is present only on the anterior side of the tooth. This tooth is quite variable in the sample of P. pearcei from Hagerman; it may have a prominent posterolabial cingulum or none, and the central cuspule connecting the anterior and posterior lophs may be double or absent. M1 is a rectangular tooth with four equal-sized, evenly spaced cusps forming anterior and posterior lophs, anterior and posterior cingula, and a small cingulum at lingual and labial ends of the transverse central valley. In the Hagerman sample this tooth often has slightly more prominent cingula, and the lophs are often a little closer together. M2 is larger than M1 but very similar otherwise. A prominent posteromedial cuspule is developed on the posterior cingulum. Discussion-Two species of Platygonus, P. pear- cei Gazin and P. bicalcaratus Cope, are currently recognized from the Blancan of western North America. These two species are difficult to distin- guish; they are almost the same size, and distinguish- ing characters which have been mentioned ( Gazin, 1938) such as development of cingula, height and distinctness of lophs, and development of Mi heels, are all variable. Woodburne (1968), using Gazin's illustrations, suggested that P. pearcei was a longer snouted species with more flaring zygomatic arches than P. bicalcaratus; unfortunately these characters are not determinable in the Ringold material. I have referred the Ringold specimens to P. pearcei because I could not adequately distinguish them from the latter, and because there are apparent differences between the Ringold specimens and published figures of topotypes of P. bicalcaratus. The great variability found by Simpson (1949) in a sample of P. com- pressus LeConte suggests that better criteria may be needed if it is to be shown that P. pearcei and P. bicalcaratus are distinct. Family Camelidae Gray, 1821 Genus M egatylopus Matthew and Cook, 1909 Megatylopus cf. M. cochrani (Hibbard and Riggs), 1949 (Figure 28) Hibbard, C. W., and Riggs, E. S., 1949, p. 854, Fig- ures lOA,B,C, Material-UW A3030: UW 41920, M/ fragment. UW A3339: UW 41557, P 4 • UWA6525: UW 40475, M1• UW A8812: UW 42221, M/ . UW A9292: UW 42320, partial mandible with alveolus for P 3 and badly worn P 4-M3• Questionably referred: UW A30- 27: UW 42003, / 1. UW A3037: UW 41979 / 1. Stratigraphic position-Ringold Formation, up to 45 feet above White Bluffs tuff, early Blancan. Description-The upper cheek teeth of M. cochra- ni are not known from the type locality in the Rex- road Formation, Kansas; the following upper teeth are assigned to this species on the basis of size and similarity to other species of Megatylopus. P 4 is a three-rooted, fairly symmetrical, shield- shaped tooth with distinct styles and rib. A wear facet on the anterior side indicates the presence of P3• The upper molars are hypsodont as indicated by a _____ ...,.:_..::_-_. ___ - - - - -.::._-....:.::.:=..:..-.: .:. Figure 28. Megatylopus sp. cf. M. cochrani, man- dible with alveolus of P 3 and badly worn P 4-M3, UW 42320, lateral view. Line indicates one centimeter. Anterior to right. fragmentary specimen. A strong mesostyle is pres- 1978 GUSTAFSON: RINGOLD FAUNAS 49 ent, but ribs are weak. A worn upper molar (UW 42221) closely resembles the M1 of Megatylopus gigas (Matthew and Cook, 1909), except for some- what smaller size. The lower dentition is represented by a single littleworn molar, probably Mi, and by a partial man- dible with the single root of Pa and badly worn and chipped P4-Ma. The root of Pa is separated from P4 by a 12 mm diastema. The Pa of the type specimen of v r:ochrani is two-rooted. The measurements of the lower dentition of the lower dentition of the Ringold specimens are close to measurements of M egatylopus cochrani (Hibbard and Riggs, 1949) from the Rexroad For- mation of Kansas. The Ringold specimens are at the lower range of measurements of Camelops hesternus Leidy from Rancho la Brea (Webb, 1965 p. 50) which lacks Pa entirely, and are apparently outside the range of variation of M. gigas (Mathew and Cook, 1909) and species of Titanotylopus (Meade, 1945). Although only one root can be seen for Pa in the Ringold mandible, the preserved root is well separated from the anterior root of P 4. Since this was an old individual, Pa could have been broken off and the posterior root resorbed before the animal died. Discussion-Differentiation of the species of Mega- tylopus depends largely on the degree of reduction of the third premolars (Webb, 1965, p. 44), which are not, except for alveoli, preserved in the Ringold Formation sample. The close similarity to M. coch- rani in size suggests a close affinity. Family Camelidae Genus Hemiauchenia Gervais and Ameghino, 1880 H emiauchenia sp. Material-UW A3027: UW 42004, Ma. UW A3036: UW 41939, Ma fragment. UW A3037: UW 41965, C or P. UW A3339: UW 41555, mandible with M1 and partial DP4, M2-a: UW 41556, / M frag- ment. UW A3339?: UW 41554, mandible fragment with DP 4, M1. UW A6503: UW 35154, DP 4; UW 35147, Ma: UW 40392, M1. UW A6571: UW 42985, mandible fragment with DP 4 and partial M1. UW A8812: UW 40468, / 1. UW A9292: UW 42325, M1-2. UW A9305: UW 42147, Ma. Stratigraphic position-Ringold Formation, up to 45 feet above White Bluffs tuff, early Blancan. Description-M1 and M2 are very similar in size, and rib and style development to a specimen of Hemi- auchenia blancoensis (Meade), (KU 7503) from the Rexroad Formation of Kansas, Keefe Canyon locality (Hibbard and Riggs, 1949) : styles are strong, and ribs, though low and rounded, are distinct. The Rin- gold specimen is chipped on the labial side of both upper teeth, so that further details are not available. The mandible (UW 45101) is long and slender. TABLE 18 MEASUREMENTS OF THE TEETH OF Megatylopus FROM THE RINGOLD FORMATION Tooth Specimen number p4 uw 41557 M/ uw 42221 M1 uw 40475 P4 uw 42320 M1 " M2 " Ma " Pa-Ma " Depth of mandible (inside) below anterior P 4 Depth of mandible (inside) below anterior Ma AP 20.9 33.4 36.0 19.4 33.0 38.6 53.2 163e uw 42320 60.7 74.3 * KU 7644, after Hibbard and Riggs, 1949 M egatylopus cochrani, holotype* T AP T 14.5 30.7 16.4 14.0 21.8 15.0 22.2 32.6 24.0 24.8 39.8 26.3 23.3 52.0 21.7 166.2 M. cochrani, holotype 62.2 80.7 50 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 TABLE 19 MEASUREMENTS OF TEETH OF H emiauchenia FROM THE RINGOLD FORMATION Specimen number Tooth AP T 41555 dP4 13.7 M1 27.5 12.3 M2 10.2 41554 dP4 29.5 10.9 M1 30.6 11.8 42985 dP4 35.9 9.6 35154 dP4 9.7 45101 M1 18.5 13.3 M2 27.2 16.1 Ma 40.9 14.6 40392 M1 13.8 42004 Ma 36.6 14.7 35147 Ma 38.0 15.0 42147 Ma 36.4 13.7 Figure 29. Hemiauchenia sp .. A. Mandible with dP4 and M1, UW 41554, lateral view, and occlusal view of M1. B. Mandible with partial P1, alveoli for Pa, badly worn P4-Ma, UW 45101, lateral view. Lines indicate one centimeter. Anterior to right in A, to left in B. 1978 GUSTAFSON: RINGOLD FAUNAS 51 Figure 30. Bretzia pseudalces Fry and Gustafson, restoration. The anterior part of the mandible is considerably longer than seen in H. blancoensis ( distance from anterior root of P4 to posterior border of symphysis= 71 mm versus 46 to 56 mm in Keefe Canyon H. buzn- coensis). It is not as elongate as in H. mirifica Simp- son ( distance from anterior border of P 4 to posterior alveolus of P 1=81 mm versus 99 mm in H. mirifica [Pleistocene, Florida] reported by Hibbard and Dalquest, 1962) . The teeth are close in size and form to H. blancoensis, and are larger than in H. macro- cephala (Cope) from the Pleistocene of Texas. DP4 is a three-lobed tooth with only very low, roun.ded ribs and small styles on the posterolingual corners of the anterior and middle lobes. It is similar in form to DP 4 of H. blancoensis. The permanent premolar crowns are not preserved. Pa was two-rooted and probably rather large; P 4 was somewhat larger but smaller than M1• Both labial and lingual anterior styles are present in Mi, though both styles would be greatly reduced when the tooth was more worn. A strong steplike style occupies the posterolingual corner of the anterior lobe, so that the anterior lobe is placed further lin- guad than the posterior lobe. Development of all styles is similar to that seen in H. blancoensis. The anterior hypolophid does not connect to the metalo- phid at the stage of wear represented by UW 41554. Ribs are low but distinct. The only M2 available is badly worn. A remnant of an anterolabial style can be seen. The four M3's are all badly worn. The anterolabial style is present. The posterior lobe is set labiad from the line of the paralophid and metalophid, but the angle of offset is variable; also the development of a steplike posterolingual style on the metalophid is variable. Characters of Ma are like those of H. blan- coensis at a similar stage of wear. Discussion-The Ringold H emiauchenia would be easily placed in H. blancoensis except for the longer anterior mandible. Additional specimens will be needed to demonstrate whether this character is con- stant or not in the Ringold Formation. The synonomy of Tanupolama with Hemiauchenia follows Webb (1974). . ' . 52 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 Family Cervidae Genus Bretzia Fry and Gustafson, 1974 Bretzia pseudalces Fry and Gustafson, 1974 (Figure 30) Holotype-UW 42971, nearly complete left antler. Type locality~UW A9577, 44 feet below White Bluffs tufI. Stratigraphic range-Ringold Formation, from 44 feet below to 65 feet above White Bluffs tuff, early Blancan. Description-A medium-sized deer, about the size of living Odocoileus hemionus. Antlers broadly pal- mated, with a single anteriorly directed tine between and palm and the burr. The name pseudalces is in reference to the mooselike palmation of the antlers; in other features ( which will be described in a future publication) including the dentition, B. pseudalces is very much like living Odocoileus. Discussion-Bretzia is the most commonly pre- served large mammal in the Ringold Formation This represents a unique situation in Blancan faunas, as deer are otherwise among the rarest of forms. The Burke Museum collection contains several nearly complete antlers, a number of teeth and jaw frag- ments, a skull fragment with pedicles, and other assorted bones. The metapodial identified by Strand and Hough (1952) as Rangifer probably should be referred to Bretzia . RIVER ROAD LOCAL FAUNA Only two identifiable specimens have been found in the Taylor Flat conglomerate of the Ringold For- mation, and of these only one has been positively identified at the genus level. However, because of the combination of low stratigraphic level and ob- vious difference in faunal composition (rhinos are conspicuous by their absence, despite the availa- bility of many hundreds of specimens, in the upper Ringold Formation) ; I consider a separate name to be justified for this fauna. Class MAMMALIA Order PERISSODACTYLA Family Rhinocerotidae Genus Teleoceras Hatcher, 1894 T eleoceras sp. (Figure 31) Material-Left mandibular ramus with symphysis, incisors, Pa-Ma, collected December 27, 1974 near south end of Taylor Flat. UW C71: UW 52685. Stratigraphic position-Ringold Formation, Tay- lor Flat conglomerate, about 70 feet below top of conglomerate, near line of section C of figure 4, late (?) Hemphillian. Discussion-The late discovery of this specimen and the press of other duties have allowed only a pre- liminary identification to be made. Neverthless some mention must be made here because of its biostrati- graphic importance. T eleoceras is a characteristic pre-Blancan genus, ranging from the Barstovian (middle Miocene) through the Hemphillian mam- malian ages, but becoming extinct at the end of the Hemphillian. In fact the Hemphillian was in part defined by the last appearance of Teleoceras (Wood et al., 1941) . The Ringold T eleoceras specimen is about the same size as specimens referred to T. fossiger (Cope), but the genus is badly in need of revision and a specific designation would be premature. The Burke Museum collection includes a partial skull and man- dible of T eleoceras from the W estend Blowout local fauna and a Ma from the McKay Reservoir local fauna, both Hemphillian of Oregon. The latter speci- mens are almost identical in form to the equivalent parts of the mandible from the Taylor Flat con- glomerate. The diastema is short as is typical of Teleoceras, and there is no sign of Pi-2 • The animal was an adult, with all molars worn, though none of the teeth are worn to the roots. The anterior part of P 4 is raised above the molar occlusal plane, and P 3 seems to have been tilted forward; these suggest that the upper premolars were damaged or otherwise wearing abnormally. The coronoid process has the anterad lean commonly seen in Teleoceras. There are no published analyses of evolutionary trends in Teleoceras. Some Clarendonian individuals possess P 2, but older individuals from sediments of the same age have lost this tooth (Webb, 1969), and may have no sign of the alveolus. The stratigraphic superposition of a fauna of early Blancan age and the similarity to specimens of known Hemphillian age are evidence of a Hemphillian age for the River Road local fauna. I tentatively consider the Taylor Flat conglomerate to be late Hemphillian. ,. 1978 GUSTAFSON: RINGOLD FAUNAS 53 Figure 31. Teleoceras sp., mandible from Taylor Flat conglomerate, UW 52685. Ocdusal and left lateral views. Scale line equals 5 centimeters. Order ARTIODACTYLA Family Camelidae ? M egatylopus sp. M aterial-UW A8817: proximal portion of radio- ulna. Stratigraphic position-Ringold Formation, Tay- lor Flat conglomerate, 20 feet below top of con- glomerate, late (?) Hemphillian. Discussion-This bone is similar in morphology to the radioulna of Camelops and its size is reasonable for members of the genus Megatylopus which is found in the ·overlying White Bluffs local fauna and which was probably ancestral to Camelops (Webb, 1965). 54 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 AGE AND CORRELATION Previous age assignments for the Ringold Forma- tion were based on stratigraphic correlations with other formations of the Columbia Basin and upon inadequate information on the faunas from the White Bluffs. The Ringold sediments have been considered early Miocene (Russell, 1893, p. 23), late Miocene "probably equivalent to the Mascall Formation" ( Calkins, 1905) , latest Pliocene or early Pleistocene (Merriam and Buwalda, 1917, p. 260), upper Plio- cene (Beck, 1940), and middle to late Pleistocene (Strand and Hough, 1952, p. 154). Russell and Cal- kins based their assignments on physical correlations with rocks in central Oregon. The assignments by Merriam and Buwalda and by Beck were based pri- marily on horses. The young assignment by Strand and Hough was based primarily on a mastodon. I propose a late Hemphillian age for the Taylor Flat conglomerate of the Ringold Formation at the White Bluffs. This assignment is based on a Teleo- ceras mandible, for reasons already discussed. The Hemphillian-Blancan boundary may not, and prob- ably does not exactly coincide with the top of the conglomerate; however for the present this is a con- venient stratigraphic boundary. The fossil mammalian aggregate in the upper Rin- gold Formation at the White Bluffs represents a North American Mammal Age of early Blancan (Pli- ocene) . The Blancan age is based on the presence of Equus (Dolichohippus), Ophiomys, Borophagus, Trigonictis, and N ekrolagus; on the association of these with other genera, particularly Platygonus, Megmylopus, and large Dipoules; and on the absence of genera characteristic of Hemphillian age (rhino- ceroses, Neohipparion, Prosthenaps, Plesiogulo, Microtoscoptes) or of Irvingtonian age (Equus (Equus), Lepus, Microtus, Mammuthus). Correlation within the Blancan is currently under- going revision on the basis of combined radiometric and paleomagnetic data (Lindsay, Johnson, and Op- dyke, 1975) which when properly combined with biostratigraphic information will eventually provide a much finer-scale correlation. Biostratigraphic correlations with Great Plains and southern faunas suffer because of regional and latitudinal effects. The White Bluffs local fauna lacks such common southern and plains forms as N annip- pus, glyptodonts, Geomys, Ogmodontomys, and Sig- modon. Thus I cannot use the major faunal datum planes set up by Johnson, Opdyke, and Lindsay (1975) in Blancan deposits of the San Pedro Valley, Arizona, which were based on Nannippus and Sig- modon. On the other hand, Scapanus, Ammospermo- philus, Thomomys, and Castor are all western forms in the Blancan, and microtine rodents are scarce or absent in southern faunas. These combine to make the northwestern faunas, especi,ally Grand View, Hagerman, and White Bluffs, appear similar despite probable differences in age. Correlations suggested by individual taxa of the White Bluffs local fauna are as follows; Omalodiscus pattersoni-This snail was reported by Taylor (1966) who states that "the distribution in western North America is limited to upper Plio- cene rocks .... " He reports it from the Glenns Ferry Fm., Idaho, Hagerman and Sand Point local faunas. Fishes--Miller ( personal communication, 1973) considers the identified White Bluffs fishes to be specifically identical to those of the Glenns Ferry Fm., Idaho. Reptiles-These provide no significant informa- tion. Scapanus sp.-The humeri are hard to distinguish from living Scapanus, but provide no other infor- mation. Hypolagus ringoldensis-Similar to but slightly advanced from the McKay Reservoir Hemphillian H. oregonensis. Nekrolagus cf. N. progressus-Similar to speci- mens from the Rexroad Formation, Kansas. Paenemarmota or Marmota sp.-Most similar to M. sawrockensis from the late Hemphillian Sawrock Canyon local fauna; smaller than ( and perhaps therefore earlier than?) specimens from Hagerman and Rexroad local faunas. Spermophilus? russelli-"primitive" in appear- ance, and possibly matched by specimens from Ha- german. Ammospermophilus hanfordi-Provides no sig- nificant information. Thomomys cf. T. gidleyi-Similar to the Hager- man species. Castor calif omicus-Specimens of very similar beavers are known from California, Hagerman, and ( though assigned a different specific name by Shot- well, 1970) Grand View local fauna of Idaho. Dipoides rexroadensis-Similar-size Dipoides are known from Hagerman and Rexroad local faunas. A slightly larger upper molar from the late Blancan Sand Draw local fauna of Nebraska was assigned to this species by Hibbard (1972). Peromyscus nosher-A modem-looking species, similar to P. hagermanensis but not very informative. Neotoma cf. N. quadriplicatus-Most similar to specimens from Fox Canyon fauna of the lower part of the Rexroad Formation, Kansas. This species has -~------ -- ------~~--------- 1978 GUSTAFSON: RINGOLD FAUNAS 55 not been reported from later Blancan localities. Ophiomys mcknighti-This species is more primi- tive than 0 . taylori from Hagerman, and is not greatly advanced beyond Promimomys from the Hemphillian of Oregon. Megalonyx rohrmanni-This species may appear at both Hagerman and Grand View, Idaho. Canis davisi-This canid is common in the late Hemphillian faunas of Oregon, but has not been reported from other Blancan faunas. Borophagus sp.-This genus is found generally in both early and late Blancan faunas. Ursus cf. U. abstrusus-Otherwise known only from the Hagerman local fauna. Trigonictis cooki-Known from both Hagerman and Grand View local faunas. F elis sp.-Provides no significant information. Machairodontinae-Provides no significant in- formation. Mammut americanum-1£ correctly identified, this species provides the strongest evidence for later Blancan (or Pleistocene) age; however, the mam- mutids of the Blancan are poorly known. Equus (Dolichohippus) cf. E. simplicidens-Evi- dence from the horses is contradictory. Merriam and Buwalda ( 1917) found "advanced" specimens of Equus in the White Bluffs; however, specimens in the Burke Museum collection are uniformly primitive in appearance. Platygonus pearcei-All other identified speci- mens of this species are from Hagerman, but P. pear- cei is questionably distinct from P. bicalcaratus, known from several other Blancan localities. Megatylopus cf. M. cochrani-Species reported only from the Rexroad Formation, Kansas, though Megatylopus is reported from Cita Canyon, Texas, and Hagerman. The genus is more widespread in the Hemphillian. Hemiauchenia sp.-Not well enough known to pro- vide significant information. Bretzia pseudalces-This genus is reported only from the White Bluffs. Deer from other Blancan lo- calities have not hen sufficiently described to allow positive identification. From the above listing it is evident that the affini- ties of this fauna are, first , to the Hagerman local fauna; second, to other western Blancan. faunas, in- cluding Grand View; and third, to the Rexroad For- mation faunas (Rexroad, Fox Canyon, and Sawrock Canyon local faunas) of Kansas. Several of the taxa have close relatives in Hemphillian faunas, especially those of northern Oregon. It would be difficult to justify a late Blancan age assignment for the White Bluffs local fauna. Keeping in mind the very small samples available for most of the taxa from the Rin- gold Formation, I would tentatively correlate the White Bluffs local fauna with the faunas of he mid- dle and lower part of the Rexroad Formation, Kan- sas, very early Blancan, and place it earlier than the Hagerman local fauna of Idaho. Author's note.-Specimens of Ophiomys mcknighti from Alturas, California, underlie a basalt dated at 4.9± 0.5 mil· lion years ( C. A. Repenning, personal communication, 1977). If the date is correct and the Ringold Formation specimens of 0 . mcknighti are of about the same age, the White Bluffs local fauna would be about 1.5 million years older than th.e Hagerman local fauna. PALEO ENVIRONMENT Fauna-The most striking feature of the White Bluffs fauna is the high relative abundance of larger mammals with dentitions adapted for browsing rather than grazing. This preponderance is illustrated by figure 32 which graphs the number of localities at which a particular form was found. The number of localities is probably nearly the same as the mini- mum number of individuals because of the large number of localities from which specimens were re- corded and the small number of specimens from most localities. The most common large mammal is Bretzia pseu- dalces, a deer about the size of the Recent mule deer ( Odocoileus hemionus) but with wide palmated ant- lers. It was presumably a browser; modern deer pre- fer areas where brushy cover is available. The adaptive features of Platygonus, the second most common form, have recently been studied by Guilday, Hamilton, and McGrady (1971). A variety of evidence, including the elongated antorbital por- tion of the skull and elongation of the legs, was con- sidered to be indicative of adaptation to an open forest habitat; this conclusion is supported by the tooth structure, which (p. 307) "would seem to be adapted to browsing, much more so than in the Old World Sus, although it is still essentially that of a modified omnivore." Blancan faunas vary considerably in relative abun- dances of large herbivorous mammals. Three ex- amples are listed in Table 20; the fauna from the Keefe Canyon locality of the Rexroad Formation, Kansas (Hibbard and Riggs, 1949), the Grand View 56 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 70 20 / turtle I------------~--' I pla tygonus 1-- ------~ --~ I llemiauc he nia 1-- ------,--' Mega tylo pus fl s h i------~~I Me ga lonyx I Canis J--------1 o------~j Eguus I Mammut >----~~ D marmo t sna ke ~ Ammos pe r mo ph i l us f r og or t oad - s a ber tooth cat - S permo ph1lus ? Peromyscus Nekrolagus Tr1gon 1c t 1s Borophagus Figure 32. Relative abundance of vertebrates in the White Bluffs local fauna, Ringold Formation, Wash- ington. Length of bar indicates number of localities at which a taxon was found. Census of Burke Museum collection made in June, 1972; it includes identified teeth and postcranial material, some of which was not listed in the systematics section. fauna from several localities in the upper Glenns Ferry Formation, Idaho (Shotwell, 1970), and the White Bluffs fauna. The White Bluffs local fauna lacks Stegomastodon and Titanotylopus (sensu Webb, 1965), both mem- bers of Hibbard's (1972, p. 134) "Savannah Valley Communities" and present in both of the other fau- nas: it also lacks antilocaprids, which are present in the Grand View local fauna. Both the Idaho and Kansas faunas are typical of Blancan faunas in gen- eral in their high abundance of Equus and low abun- dance or absence of cervids and Megalonyx. The White Bluffs fauna in contrast includes abun- dant cervids and Megalonyx; Equus is relatively scarce. Hemiauchenia is abundant in all three faunas, as is Platygonus; however, Platygonus is relatively more abundant in the White Bluffs and Keefe Canyon samples than in the Grand View sample. M egatylopus is also present in the Keefe Canyon and White Bluffs faunas, but is not seen at Grand View. Differences in these faunas of large mammals probably reflect local abundances of vegetation types rather than geographical differences. The most sensi- tive "indicators" of vegetation types are presumably the more specialized browsers and grazers, primar- ily cervids, Megalonyx, Mammut, Equus, and antilo- caprids; the first three of these are probably indica- tive of dense woodland, the latter two of open grass- land. Other forms, such as the camels and Platygonus probably ranged through the open woodland-savan- nah habitats. With these "indicators" in mind, it appears that woodland habitats were most prevalent on the Ringold floodplain, and less prevalent but present near the Keefe Canyon and Grand View de- positional sites. Open grassland was most prevalent near the Grand View deposits and least prevalent on the Ringold floodplain. The mammals may be allocated to the following communities (modified after Shotwell, 1958, fig. 13, and Hibbard, 1972,p. 134): Stream-Bank and Lake Bank Communities: Di- poides rexroadensis, Castor californicus. Marsh and Semiaquatic Communities: Ophiomys mcknighti. Woodland Communities: Hypolagus ringolden- sis, Scapanus sp. , (?) Peromyscus nosher, Megalo- nyx rohrmanni, Mammut americanum, Ursus cf. U. abstrusus , (?) Trigonictis cooki, Bretzia pseudalces. Open Woodland-Savannah Communities: Nekro- lagus cf. progressus, Paenemarmota or Marmota sp., Spermophilus? russelli, Thomomys cf. T. gidleyi, Neotoma, Felis sp., Platygonus pearcei, Megatylopus cf. M. cochrani. Open grassland Communities: Equus cf. E. simp- plicidens. Not allocated: Ammospermophilus hanfordi, Canis davisi, Machairodontinae, Hemiauchena sp. These are very preliminary allocations; many of these animals certainly ranged through several habi- tats, and some may have lived in rather different habitats from those occupied by their Recent coun, terparts. The lower vertebrates, including pond turtles, frogs or toads, and snakes, as well as catfish and sun- fish, are suggestive of the presence of quiet water ( oxbow lakes? ) and marshes. There are no areas along the present Columbia River that could support the White Bluffs faunas. This absence of suitable habitat seems to be largely 1978 GUSTAFSON: RINGOLD FAUNAS TABLE 20 RELATIVE ABUNDANCE 1 OF LARGE MAMMALS Keefe Canyon2 Grand View3 White Bluffs Rexroad Glenns Ferry Ringold Formation Formation Formation 1. Titanotylopus 21 Equus 30+ Bretzia 22 2. H emiauchenia 14 H emiauchenia 7 Pwtygonus 14 3. Equus 10+ Stegomastodon 6 H emiauchenia 11 4. P/,atygonus s+ P/,atygonus 3 M egatylopus 10 5. Mammut 2 Titanotylopus 3 Megalonyx 8 6. M egatylopus 1 Cervid 3 Equus 7 7. Megalonyx 1 Megalonyx 2 Mammut 5 8. Stegomastodon 1 Mammut 2 9. Antilocaprid 2 1 Numbers given are approximate minimum numbers of individuals for Keefe Canyon and Grand View, and are numbers of localities for White Bluffs (see fig. 23). 2 Hibbard and Riggs, 1949 3 Shotwell, 1970 57 due to the absence of broad flood plains as well as to the lack of rainfall. The nearest areas which could conceivably support such a fauna today are along the Yakima River near Yakima and for about thirty miles downstream, where some floodplain woodland is developed. The large mammal fauna is depauperate at present ( except for deer and domestic mammals), probably due to late Pleistocene extinctions and heavy use by humans. This area receives little more rainfall than the Pasco Basin (less than 10 inches yearly) yet supports even some drought-sensitive animals such as the mole Scapanus. Columbia Basin areas with slightly higher rainfall ( about 15 inches yearly) support very productive grasses and in his- toric times were famous as excellent grazing land for horses. The faunas of the Ringold Formation near the streams may have been relatively independent of rainfall, as long as permanent streams were present and the floodplain water table remained relatively high. Somewhat greater rainfall may have been pres- ent in the central Columbia Basin during the Plio- cene, perhaps because the Cascade Mountains rain- shadow was not as effective. However, the Pliocene rainfall could have been only about 5 to 7 inches greater yearly and yet have allowed the vegetational changes necessary to support a large and varied fauna on the Columbia Plateau. 58 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 LITERATURE CITED Allen, J. R. L., 1971. "Rivers and their deposits." Sci. Prog., Oxf. 59: pp. 109·122. Baird, S. F., and C. Girard, 1852. "Characteristics of some new reptiles in the museum of the Smith- sonian Institution." Proc. Acad. Nat. Sci., Phila- delphia 6:pp. 125·129. Beck, G. F., 1936. "Lime-rock zone of Columbia Plateau." The Mineralogist 4,8:pp. 12-14. _____ 1937. "Camels of the Columbia Pla- teau." The Mineralogist 5, 3 :pp. 5-6. -----, 1940. "Late Tertiary stratigraphy and paleontology of south-central Washington and adjacent Oregon." Abstracts of section E, Geol. Soc. Am. Bull. 51 :pp. 2018. -----, 1949. "Modern and extinct horses of the Columbia Basin." Geol. Soc. of the Oregon Country Geol. News Letter 15, 7 :pp. 60·62. Bjork, P. R., 1970. "The Carnivora of the Hager- man local fauna (late Pliocene) of southwestern Idaho." Trans. Amer. Phil. Soc., new ser. 60, 7:pp. 1-54. Black, C. C., 1963. "A review of the North American Tertiary Sciuridae.'' Bull. Mus. Comp. Zool. 130, 3 :pp. 109-248. Brattstrom, B. H., and J. Sturn., 1959. "A new species of fossil turtle from the Pliocene of Oregon, with notes on other fossil Clemmys from western North America." Southern Calif. Acad. Sci. Bull. 58,__ 2: pp. 65-71. Bretz, J. Harlan, H. T. U. Smith, and G. E. Neff. 1956. "Channeled scab land of Washington: new data and interpretations." Bull. Geol. Soc. Amer. 67, 8:pp. 957-1050. Brown, R. E., 1968. "A study of reported faulting in the Pasco Basin." AEC Research and Develop- ment Report, BNWL-662. -----, 1969. "Some suggested rates of defor· mation of the basalts in the Pasco Basin, and their implications." Proc. Columbia River Basalt Sym- posium, Cheney, Washington, pp. 179,187. Brown, R. E., and D. J. Brown., 1957. "The sur· face of the basalt series in the Pasco Basin." un- published report, General Electric Company, Rich- land, Washington. -----, 1961. "The Ringold Formation and its relationships to other formations." Northwest Sci- ence 35, 4:154-155. Brown, R. E., and M. W. McConiga., 1960. "Some contributions to the stratigraphy and indicated de- formation of the Ringold formation. Northwest Science 34, 2 :pp. 43-54. Bryan, K., 1927. "The Palouse soil problem" U.S. Geol. Survey Bull. 790-B. Bryant, M. D., 1945. "Phylogeny of the Nearctic Sciuridae." Ameri. Midland Naturalist 33, 2 :pp. 257-390. Calkins, F. C., 1905. "Geology and water resources of a portion of east-central Washington." U.S. Geol. Surv. Water-supply and irrigation paper no. 118, 96 pp. Cope, E. D., 1889. "The vertebrate fauna of the Equus beds." Amer. Nat. 23:pp. 162-165. -----, 1892. "A hyaena and other Carnivora from Texas." Amer. Nat. 26:pp. 1028-1029. -----, 1893. "A preliminary report on the ver- tebrate paleontology of the Llano Estacada." 4th Ann. Rep. Texas Geol. Surv., 136 pp. Culver, H. E., 1937. "Extensions of the Ringold Formation." Northwest Science 11, 3:pp. 57-60 ... Cuvier, F., 1825. "'Des dents des Mammiferes, con- sideres comme caracteres zoologiques." in Diction- aire des Sci. Nat., Paris. Dice, L. R., 1917. "Systematic position of several American Tertiary lagomorphs." U. Cal. Puhl., Bull. Dept. Geology 10, 12 :pp. 179-183. Downey, J. S., 1968. "Late Pliocene lagomorphs of the San Pedro Valley, Arizona." U.S. Geol. Surv. Prof. Pap. 600-D, pp. Dl69-Dl 73. Downs, T., 1961. "A study of variation and evolu- tion in Miocene Merychippus." Los Angeles County Museum Contributions to Science 45 :pp. 1-75. Downs, T., and J. A. White., 1968. "A vertebrate faunal succession in superposed sediments from late Pliocene to middle Pleistocene in California." Inter. Geol. Cong. Report XXIII Session, Proc. Sec. 10:pp. 41-47. Fisher, R.V., and J.M. Rensberger., 1972. "Phy- sical stratigraphy of the John Day Formation, central Oregon." U. Cal. Pub. Geol. Sci. 101 :pp. 1-45. Flint, R. F., 1938. "Origin of the Cheney-Palouse scabland tract." Geol. Soc. America Bull. 49:pp. 461-524. Frick, C., 1921. "Extinct vertebrate faunas of the badlands of Bautista Creek and San Timoteo Can- yon, southern California." U. Cal. Publi., Bull. Dept. Geol. 12:pp. 277-424. Fry, W. E., and E. P. Gustafson., 1974. "Cervids from the Pliocene and Pleistocene of central Wash- ington." lour. Paleo. 48, 2:pp. 375-386. Galbreath, E. C., 1972. Review of Bjork, 1970, '"The Carnivora of the Hagerman local fauna ." Jour. Paleo. 46, 5 :pp. 785-786. Gazin, C. L., 1933. "New felids from the upper Plio- cene of Idaho." }our. Mammalogy 14:pp. 251-256. 1978 GUSTAFSON: RINGOLD FAUNAS 59 , 1934. "Upper Pliocene mustelids from the Snake River Basin of Idaho." Jour. Mam- malogy 15:pp. 137-150. - ----, 1935. · "Gravigrade sloth remains from the late Pliocene and Pleistocene of Idaho." Jour. Mammalogy 16:pp. 52-60. - - --- , 1936. "A study of the fossil horse re· mains from the upper Pliocene of Idaho." Proc. U. S. Nat. Mus. 83, 2985: pp. 281-320. -----, 1938. "Fossil peccary remains from the upper Pliocene of Idaho.'' Jour. Wash. Acad. Sci. 28:pp. 41-49. Gidley, J. W., 1903. "On two species of Platygonus from the Pliocene ·of Texas." Amer. Mus. Nat. Hist. Bull. 19, 14:pp. 477-481. Gill, T., 1872. "Arrangement of the families of mam· mals with analytical tables." Smithsonian Misc. Coll. 11, l:pp. 1-98. Gray, J. E., 1821. "On the natural arrangement of vertebrose animals." London Med. Reposit. 15, 1 :pp. 296-310. - ----, 1825. "Outline of an attempt at the dis- position of the Mammalia into tribes and families." Ann. Philos., N. S., 10:pp. 337-344. Guilday, J.E., H. W. Hamilton, and A. D. Mc- crady., 1971. "The Welsh Cave peccaries (Platy· gonus) and associated fauna, Kentucky Pleisto· cene." Ann Carnegie Mus. Pittsburgh 43, 9 :pp. 249-320. Gustafson, E. P., 1977. First record of Tdeoceras ( Rhinocerotidae) from the Ringold Formation, Pliocene of Washington. Paleo Bios, No. 27, 4pp. Harlan, R. 1825. "Description of two new species of Agama." Jour. Acad. Nat. Sci., Philadelphia, 4:pp. 296-304. Hatcher, J.B., 1894. "A median horned rhinoceros from the Loup Fork beds of Nebraska." Amer. Geologist 13, 3 :pp. 149-150. Hershkowitz, P. 1955. "South American marsh rats, genus Holochilus, with a summary of sigmo· dont rodents.'' Fieldiana, Zool. 37 :pp. 639-688. Hibbard, C. W., 1939. "Four new rabbits from the upper Pliocene of Kansas." Amer. Midi. Nat. 21, 2:pp. 506-513. -----, 1941. "Mammals of the Rexroad Fauna from the upper Pliocene of southwestern Kansas." Trans. Kansas Acad. Sci. 44:pp. 265-313. _____ , 1944. "Two new mammals from the middle Pliocene of Seward County, Kansas." Univ. Kansas Sci. Bull. 30, 1 :pp. 107-115. -----, 1949. ''Techniques of collecting micro· vertebrate fossils." Contrib. Mus. Paleontol., Univ. Mich. 8, 2:pp. 7-19. -----, 1956. "Vertebrate fossils from the Meade Formation of southwestern Kansas." Pap. Mich. Acad. Sci., Arts, Letters 41 :pp. 145-203. -----, 1957. "Two new Cenozoic microtine rodents." Jour. Mammalogy 38:pp. 39-44. --- - - ·, 1959. "Late Cenozoic microtine rodents from Wyoming and Idaho." Pap. Mich. Acad. Sci., Arts, Letters 44:pp. 3-40. - - - - -, 1962. "Two new rodents from the early Pleistocene of Idaho." Jour. Mammalogy 43, 4:pp. 482-485. -----, 1963. "The origin of the P3 pattern of Sylvilagus , Caprolagus, Oryctolagus, and Lepus." Jour. Mammalogy 44, 1 :pp. 1-15. ·- ---, 1964. "A contribution to the Saw Rock Canyon local fauna of Kansas." Pap. Mich, Acad. Sci. , Arts, Letters 49:pp. 115-127. -----, 1967. "New rodents from the late Ceno· zoic of Kansas." Pap. Mich. Acad. Sci., Arts, Letters 52:pp. 115-131. ---- -, 1972. "Class Mammalia." in Skinner et al., 1972, Bull. Amer. Mus. Nat. Hist. 148, 1 :pp. 77-116. Hibbard, C. W., and Dalquest, W., 1962. "Artio· dactyls from the Seymour Formation, Knox Coun· ty, Texas." Pap. Mich. Acad. Sci., Arts, Letters 47:pp. 83-99. Hibbard, C. W., and Riggs, E. S., 1949. "Upper Pliocene vertebrates from Keefe Canyon, Meade County, Kansas." Bull. Geol. Soc. Amer. 60:pp. 829-860. Hibbard, C. W., and C. B. Schultz., 1948. "A new sciurid of Blancan age from Kansas and Nebras- ka.'' Univ. Nebraska State Mus. Bull. 3, 3 :pp. 19-29. Hibbard, C. W., and R. J. Zakrzewski. 1967. " Phyletic trends in the late Cenozoic microtine Ophiomys, gen. nov., from Idaho." Contrib. Mus. Paleont., Univ. Mich. 21, 12 :pp. 255-271. Hirschfeld, S. E., and S. D. Webb., 1968. "Plio· pleistocene megalonychid sloths of North Amer· ica." Bull. Florida State Mus. Biol. Sci. 12, 5 :pp. 213-296. Holmgren, D. A., 1967. "The Yakima-Ellensburg unconformity, central Washington." unpublished MS thesis, Univ. of Wash., pp. 1-69. -----, 1969. "K/ Ar dates and paleomagnetics of the type Yakima Basalt." Proc. 2nd Columbia River Basalt Symposium, Cheney, Wash. pp. 189- 199. Hooper, E. T., 1957. "Dental patterns in mice of the genus Peromyscus." Misc. Pub. Zool., Univ. Mich. 99:pp. 7-59. Howe, J. A. 1970. "The range of variation in Equus (Plesippus) simplicidens Cope from the Broad- water quarries of Nebraska." Jour. Paleontology 44, 5 :pp. 958-968. I : ·. 60 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 Hutchison, J. H., 1968. "Fossil Talpidae (Insecti- vora, Mammalia) from the later Tertiary of Ore- gon." Bull. Mus. Nat. Hist., Univ. Oregon 11 :pp. 1-117. James, G. T., 1963. "Paleontology and non-marine stratigraphy of the Cuyama Valley badlands, Cali- fornia." Univ. Cal. Pub. Geol. Sci. 45-pp. 1-154. Johnson, N.M., N. D. Opdyke, and E. H. Lind- say., 1975. "Magnetic polarity stratigraphy of Pliocene-Pleistocene terrestrial deposits and ver- tebrate faunas, San Pedro Valley, Arizona." Geol. Soc. Amer. Bull. 86, 1 :pp. 5-12. Johnson, C. S., 1938. "Preliminary report on the vertebrate type locality of Cita Canyon, and the description of an ancestral coyote." Amer. J our. Sci. ser. 5, 35:pp. 383-390. Kellogg, Louise., 1911. "A fossil beaver from the Kettleman Hills, California." U. Cal. Pub., Bull. Dept. Geol. 6, 17 :pp. 401-402. Le Conte, J. L., 1848. "On Platygonus Compressuss a new fossil pachyderm." Mem. Amer. Acad. Arts Sci. 3, 8:pp. 257-274. Leidy, J., 1855. "A memoir of the extinct sloth tribe of North America." Smithson. Contrib. Knowl. 7:pp. 1-68. Lewis, P. F., 1960. "Linear topography in the south- western Palouse, Washington-Oregon." Assoc. Amer. Geogr. Ann. 50:pp. 98-111. Lindsay, E. H., N. M. Johnson, and N. D. Op- dyke, 1975. "Preliminary correlation of North American Land Mammal Ages and geomagnetic chronology." C. W. Hibbard Mem. Vol. 3, Univ. Mich. Mus. Paleo. Papers on Paleontology 12 :pp. 111-119. Linnaeus, C. 1758. Systema naturae per regna tria naturae, secundum classes, ordines, genera, spe- cies cum characteribus, difjerentiis, synnymis, lo- cis. 10th Edition, revised. Stockholm, Laurentii Salvii, vol. 1, pt. 1, 532 pp. Lupher, R. L., 1944. "Clastic dikes of the Columbia Basin region, Washington and Idaho.'' Geol. Soc. Amer. Bull. 55:pp. 1431-1462. Marsh, 0. C., 1871. "Notice of some new fossil mam- mals from the Tertiary Formation." Amer. Jour. Sci., ser. 3, 2:pp. 120-127. -----, 1874. "Notice of new Tertiary mam- mals III." Amer. Jour. Sci., ser. 3, 7:pp. 531-534. Matthew, W. D., 1902. "List of the Pleistocene fauna from Hay Springs, Nebraska." Bull. Amer. Mus. Nat. Hist. 16, 24:pp. 317-322. -----, 1924. "Third contribution to the Snake Creek fauna." Bull. Amer. Mus. Nat. Hist. 50, 149: pp. 59-210. Matthew, W. D., and H.J. Cook., 1909. "A Plio- cene fauna from western Nebraska." Bull. Amer. Mus. Nat. Hist. 26:pp. 361-414. Meade, G. E., 1945. "The Blanco fauna." Univ. Texas Puhl. 4401 :pp. 509-556. Merriam, C. H., 1892. "Descriptions of nine new mammals collected by E. W. Nelson in the states of Colima and Jalisco, Mexico." Proc. Biol. Soc. Washington 7:pp. 164-174. Merriam, J. C., 1911. "Tertiary mammal beds of Virgin Valley and Thousand Creek in' northwest- ern Nevada." Univ. Cal. Puhl., Bull. Dept. Geol. 6, 11 :pp. 199-304. -----, 1916. "Relationship of Equus to Plio- hippus suggested by characters of a new species from the Pliocene of California." Univ. Cal. Publi., Bull. Dept. Geol. 9, 18:pp. 525-534. -----, 1918. "New Mammalia from the Idaho Formation." Univ. Cal. Publi., Bull. Dept. Geol. 10, 26 :pp. 523-530. Merriam, J. C., and J. P. Buwalda. 1917. "Age of strata referred to the Ellensburg Formation in the White Bluffs of the Columbia River." Univ. Cal. Puhl. , Bull. Dept. Geol. 10, 15 :pp. 255-266. Miller, R. R., 1965. "Quaternary freshwater fishes of North America." in The Quaternary of the United States. Wright, H. E., jr., and D. G. Frey, eds. Princeton Univ. Press, pp. 569-581. Miller, R. R., and G. R. Smith, 1967. "New fossil fishes from Pio-Pleistocene Lake Idaho." Occ. Pap. Mus., Univ. Mich. 654:pp. 1-24. MacDonald, J. R., 1948. "The Pliocene carnivores of the Black Hawk Ranch fauna." Univ. Cal. Puhl. , Bull. Dept. Geol. 28, 3 :pp. 53-80. McDowell, S. B., 1964. "Partition of the genus Clemmys and related problems in the taxonomy of the aquatic Testudinidae." Proc. Zool. Soc. London 143:pp. 239-279. Macintyre, G. T., 1966. "The Miacidae (Mammalia, Carnivora). Part 1. The systematics of Ictidopap- pus and Protictis." Bull. Amer. Mus. Nat. Hist. 131 :pp. 115-210. McKnight, E. T., 1923. "The White Bluffs forma- tion of the Columbia." Univ. Washington thesis, unpublished. Newcomb, R. C., 1958. "Ringold Formation of Pleistocene age in type locality, the White Bluffs, Washington." Amer. Jour. Sci. 256 :pp. 328-340. Newcomb, R. C., J. R. Strand, and F. J. Frank, 1972. "Geology and ground-water characteristics of the Hanford Reservation of the U. S. Atomic Energy Commission, Washington." U.S. Geol. Survey Prof. Paper 717. 1978 GUSTAFSON: RINGOLD FAUNAS 61 Newman, K. R., 1969. "Palynology of interflow sediments from the Standard Oill Company of California Rattlesnake Hills No. 1 well, Benton County, Washington." Proc. 2nd Columbia River Basalt Symposium, Cheney, Wash., pp. 201-207. Raymond, J. R., and D. D. Tillson. 1968. "Evalua- tion of a thick basalt sequence in south central Washington." AEC Research and Development Report BNWL-776. Repenning, C. A., 1962. "The giant ground squirrel Paenemarmota." lour. Paleontology 36 :pp. 540- 556. -----, 1968. "Mandibular musculature and the orogin of the subfamily Arvicolinae (Roden- tia) ." Acta Zool. Crac. 13, 3 :pp. 29-72. Richmond, G. M., R. Fryxell, G. E. Neff, and P. L. Weis, 1965 "The Cordilleran ice sheet of the northern Rocky Mountains, and related Quater- nary history of the Columbia Plateau." in The Quaternary of the United States. Wright, H. E. Jr., and D. G. Frey, eds. Princeton Univ. Press. pp. 231-242. Russell, I. C., 1893. '"A geological reconnaissance in central Washington." U.S. Geol. Surv. Bull. 108. Savage, D. E., 1951. "Late Cenozoic vertebrates of the San Francisco Bay region.'' Univ. Cal. Puhl. Geol. Sci. 28, 10:pp. 215-314. -----, 1960. "A survey of various late Ceno- zoic vertebrate faunas of the Panhandle of Texas, pt. 3, Felidae." Univ. Cal. Pub. Geol. Sci. 38:pp. 317-343. Say, T., and G. Ord, 1825. "A new genus (Neo- toma) proposed, and a description of the species (N. fioridana) upon which it is founded." lour. Acad. Nat. Sci., Philadelphia 4:pp. 345-349. Schlosser, M., 1902. "Beitrage zur Kenntnis des Saugetierrests aus den suddeutschen Bohnerzen." Geol. Paleontol. Abhandl., new ser., 5:pp. 117- 258. Schmincke, H. U., 1967. Stratigraphy and petro- graphy of four upper Yakima Basalt flows in south- central Washington." Geol. Soc. Amer. Bull. 78: pp. 1385-1422. Schultz, C. B., M. R. Schultz, and L. D. Martin, 1970. "A new tribe of saber-toothed cats (Bar- bourofelini) from the Pliocene of North America." Bull. Univ. Nebraska State Mus. 9, 1 :pp. 1-31. Shotwell, J. A., 1955. "A review of the Pliocene beaver Dipoides." lour. Paleontology 29, 1 :pp. 129-144. -----, 1956. "Hemphillian mammalian assem- blage from northeastern Oregon." Geol. Soc. Amer. Bull. 67 :pp. 717-738. -----, 1958. "Inter-community relationships in Hemphillian (mid-Pliocene) mammals.'' Ecol- ogy 39, 2 :pp. 271-282. -----, 1961. " Late Tertiary biogeography of horses in the northern Great Basin." lour. Pale- ontology 35, 1 :pp. 203-217. -----, 1967. "Peromyscus of the late Tertiary in Oregon." Bull. Mus. Nat. Hist. Oregon 5:pp. 1-35. -----, 1970. " Pliocene mammals of southeast Oregon and adjacent Idaho." Bull. Mus. Nat. Hist. Univ. Oregon 17 :pp. 1-103. Shotwell, J. A., R. G. Bowen, W. L. Gray, D. C. Gregory, D. E. Russell, and D. Taylor, 1963. "The J untura Basin: Studies in earth history and paleoecology." Trans. Amer. Phil. Soc., N. S. 53, l:pp.1-77. Simpson, G. G. 1949. "A fossil deposit in a cave in St. Louis." Amer. Mus. Novit. 1408:pp. 1-46. Simpson, G. G., Anne Roe, and R. C. Lewontin. 1960. Quantitative Zoology. Harcourt Brace, New York. Skinner, M. F., and Hibbard, C. W., with E. D. Gutentag, G. R. Smith, J. G. Lundberg, J. A. Hol- man, and P. V. Rich, 1972. "Early Pleistocene pre- glacial rocks and faunas of northcentral Nebras- ka." Bull. Amer. Mus. Nat. Hist. 148, 1 :pp. 1-148. Smiley, C. J., 1963. "The Ellensburg Flora of Wash- ington." Univ. Cal. Pub. Geol. Sci. 35, 3 :pp. 159- 276. Smith, G. 0., 1903. "Description of the Ellensburg Quadrangle (Washington)." U.S. Geol. Surv. Geol. Atlas, Folio 86. Stirton, R. A., 1935. "A review of the Tertiary beavers." Univ. Cal. Puhl., Bull. Dept. Geol. 23, 13:pp. 391-458. -----, 1940. "Phylogeny of the North Ameri- can Equidae." Univ. Cal. Publi., Bull. Dept. Geol. 25, 4:pp. 165-198. Stock, C., 1928. "Tanupolama, a new genus of llama from the Pleistocene of California." Carnegie Inst. Washington Puhl. 393 :pp. 30-37. -----, 1925. "Cenozoic gravigrade edentates of Western North America." Puhl. Carnegie Inst. Wash. 331:pp. 1-206. Strand, J. R., and J. Hough. 1952. "Age of the Ringold Formation." Northwest Science 26 :pp. 152-154. Taylor, D. W., 1966. "Summary of North American Blancan nonmarine mollusks." Malacologia 4, 1 :pp. 1-172. Tedford, R. H., 1970. "Principles and practices of mammalian geochronology in North America." North Am. Paleont. Convention, Chicago, 1969, Proc., F.pp. 666-703. 62 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.23 Waters, A. C., 1955. "Geomorphology of south-cen- tral Washington, illustrated by the Yakima East quadrangle." Geol. Soc. Amer. Bull. 66:pp. 663- 684. Webb, S. D., 1965. "The osteology of Camelops." Bull. Los Angeles Co. Mus., Sci. 1 :pp. 1-54. -----, 1969. "The Burge and Minnechaduza Clarendonian mammalian faunas of north-central Nebraska." Univ. of Calif. Pub. Geol. Sci. 78:pp. 1-191. _____ ,, 1974. Pleistocene Mammals of Florida. Univ. Florida Press, Gainesville. Wilson, R. W., 1932. "Cosomys, a new genus of vole from the Pliocene of California." Jour. Mammal- ogy 13, 2:pp. 150-154. -----, 1933. "A rodent fauna from later Ceno- zoic beds of southwestern Idaho." Contrib. Pale- ontology, Carnegie Inst. Wash. Puhl. 440 :pp. 117- 135. -----, 1949. "Rodents of the Rincon fauna, western Chihuahua, Mexico." Contrib. Paleontol- ogy, Carnegie Inst. Wash. Puhl. 584:pp. 165-176. Wood, H. E., et. al., 1941. "Nomenclature and cor- relation of the North American continental Ter- tiary." Geol. Soc. Amer. Bull. 52:pp. 1-48. Woodburne, M. 0., 1961. "Upper Pliocene geology and vertebrate paleontology of part of the Meade Basin, Kansas." Pap. Mich. Acad. Sci., Arts, Let- ters 46:pp. 61-101. _____ ,, 1968. "The cranial myology and osteol- ogy of Dicotyles tajacu, the collared peccary, and its bearing on classification." Mem. South. Calif. Acad. Sci. 7:pp. 1-48. Zakrzewski, R. I., 1967a. "The systematic position of Canimartes? from the upper Pliocene of Idaho." Jour. Mammalogy 48:pp. 293-297. -----, 1967b. "The primitive vole, Ogmodon- tomys, from the late Cenozoic of Kansas and Ne- braska." Pap. Mich. Acad. Sci., Arts, Letters 52 :pp. 133-150. -----, "The rodents from the Hagerman Local Fauna, upper Pliocene of Idaho." Contrib. Mus. Paleontology Univ. Mich. 23, 1 :pp. 1-36. -- -------- Bulletins No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8 No. 9 PUBLICATIONS Museum of Natural History, University of Oregon Eugene, Oregon Titles Cenozoic Stratigraphy of the Owyhee Region, Southeastern Oregon, by L. R. Littleman and others; 45 pages,9 plates, 11 figures, December 1965 Notes on Some Upper Miocene Shrews from Oregon, by J. H. Hutchison; 23 pages, 17 figures, March 1966 A New Archaic Cetacean from the Oligocene of Northwest Oregon, by Douglas Emlong; 51 pages, 15 figures, October 1966 The Archaeology of a Late Prehistoric Village in Northwestern California, by Frank C. Leonhardy; 41 pages, 17 figures, March 1967 Peromyscus of the Late Tertiary in Oregon, by J. Arnold Shotwell; 35 pages, 11 figures, June 1967 Ethnomalacology and Paleoecology of the Round Butte Archaeological Sites, Deschutes River Basin, Oregon, by Ernest J. Roscoe; 20 pages, 4 figures, July 1967 Price $1.50 $1.25 $1.50 $1.00 $1.25 $ .75 Its Own Story: The Museum of Natural History; 20 pages no charge Geologic Map of the Owyhee Region, Malheur County, Oregon, by L. R. Kittleman and others; scale, 1:125,000 (1h inch equals 1 mile), September 1967 Late Tertiary Geomyoid Rodents of Oregon, by J. Arnold Shotwell; 51 pages, 28 figures, November 1967 $2.00 $1.25 No. 10 Refinements in Computerized Item Seriation, by W. B. Crayton and LeRoy Johnson, Jr.; 22 pages, 6 figures, March 1968 $ .75 No. 11 Fossil Talpidae (lnsectivora, Mammalia) from the Tertiary of Oregon, by J. H. Hutchi- son; 117 pages, 98 figures, July 1968 $1.25 No. 12 Plants of the Three Sisters Region, Oregon Cascade Range, by Orlin L. Ireland; 130 pages, 34 figures, April 1968 $3.75 No. 13 Historical Background of the Flora of the Pacific Northwest, by LeRoy Detling; 57 pages, 6 figures, July 1968 $1.50 No. 14 Miocene Mammals of Southeast Oregon, by J. Arnold Shotwell; 67 pages, 33 figures, August 1968 $1.25 No. 15 Item Seriation as an Aid for Elementary Scale and Cluster Analysis, by LeRoy Johnson, Jr. ; 46 pages, 19 figures, September 1968 No. 16 The Oligocene Marine Molluscan Fauna of the Eugene Formation in Oregon, by Carole Jean Stentz Hickman; 112 pages, 14 plates, 4 figures, August 1969 No.17 Pliocene Mammals of Southeast Oregon and Adjacent Idaho, by J. Arnold Shotwell; 103 pages, 42 figures, August 1970 No. 18 Smilodonichthys rastrosus, a New Pliocene Salmonid Fish from Western United States, by Ted M. Cavender and Robert Rush Miller; 44 pages, 14 figures, March 1972 No. 19 Journal of First Trip of University of California to John Day Beds of Eastern Oregon, by $1.50 $2.50 $2.00 $1.50 Loye Miller, edited by J. Arnold Shotwell; 21 pages, 7 figures, 1 map, December 1972 $1.00 No. 20 An Approach to the Study of Far Western North American Prehistory: Early Man, by Luther S. Cressman ; 11 pages, August 1973 $ .75 No. 21 Guide to the Geology of the Owyhee Region of Oregon, by Laurence R. Kittleman; 61 pages, 4 plates, 36 figures, 3 maps, September 1973 $2.50 No. 22 Guide to the Geology and Lore of the Wild Reach of the Rogue River, Oregon, by William B. Purdom; 67 pages, 4 plates, 31 figures, 7 maps, May 1977 $3.25 No. 23 The Vertebrate Faunas of the Pliocene Ringold Formation, South-Central Washington, by Eric Paul Gustafson; 62 pages, 31 figures, March 1978 $3.00