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 ir- regularly 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. J. ARNOLD SHOTWELL, Director Museum of Natnral History University of Oregon Communications concerning manuscripts and purchase of copies of the Bulletin should be addressed in care of the Museum of Natural History, University of Oregon, Eugene, Oregon 97403. A list of previous issues of the Bulletin will be found inside the back cover. I l I ,. j FOSSIL TALPIDAE (lnsectivora, Mammalia) FROM THE LATER TERTIARY OF OREGON FOSSIL TALPIDAE (Insectivora, Mammalia) FROM THE LATER TERTIARY OF OREGON by J. HOWARD HUTCHISON Bulletin No. 11 Museum of Natural History University of Oregon Eugene, Oregon July 1968 CONTENTS Abstract______________________________________ __ ____ _____ ___ ______________________ ______ __________ ____________ 1 Introduction_____________________________ ______________________________________________ ____________________ 1 Classification of recent and North American fossil moles_____________________ _____ 8 Acknowledgments__ _________________________ __ _________________________ __ _____________________ __________ 9 Materials and Methods________________________ _________________ _______ __ ____________________________ 9 Osteology of the living Talpidae___________________________ __ ___________________________________ 13 Systematics___ __ _________________________________________________ __________________ ______ ___________________ 36 Ta 1 pi da e ____ ________ ------------------------------------------------------------------_______ ---------------- 36 Subfamily U ropsilinae___________________________ _________________________________________________ __ 36 Mys ti pter us------------------------------------------------------------------------------------------------ 36 Summary of M ystipterus relationships ______________________________________________________ 46 Suh family G ai Ilardi nae ______________________________________ --------------------------_ . _____ _______ 46 Gaillardia _____________________________________________________ --------------------------------------------- 4 7 Suh fa mi 1 y T alp ina e _________________________________________________ ---------------------------------- 54 Tribe Sea p ton ychini ___________________________________________________ -------------------------------- 55 Tribe T alp i ni _________________ ----------------------------------------------------------------------------- 5 5 Tribe U rotrichini_______________________________________________________________________________________ 55 N eurotrichus__________________________________________________________________________ _______________ __ ___ 55 Tribe Con d y 1 urini ___________________________________________ ------------------------------------------- 58 Tribe S ca 1 op i ni ______________________________________________________ ------------------------------------ 58 Suh tribe Par a seal op i na _______________________ __ _____________ ---------------------------------------- 58 S calo po ides ________________________________ ----------------------------____________________________________ 58 Summary of relationships of the Scalopoides-like moles ____________________________ 80 Domninoides______________________________ __ ______ _________________________________________________________ 80 Subtribe Scalopina_____________________________________________________________________________________ 82 S capano s capte r -------------------------------------------------------------· --------------------------- 83 Scapanus_____________________________________________________________________________________________________ 86 X er o s cap he us ____________________ ---------------------------------------------------------------------· ___ 8 7 Summary of relationships of the Scapanus-like moles ______________________________ 93 A chl yo s capte r ________________ ----------------____________________________ -------------------------------- 96 T al pi dae and Talp inae incertae sed is ______ __ ________ __ _______________ _______________________ IO 1 Summary ____________________________ ------------------------·------------------------------------___________ l 08 Locality d escri p ti o ns _______________________________ -----------------------------------------------____ l 13 References Cited ________ ------------------------------------------------_________________________________ l 15 Add end um ____ __ ______________________________________________________________ ______________________________ l l 7 Funds for publication of Bulletin No. 11 provided by the Natural Science Foundation, grant GB3971. 1 FOSSIL TALPIDAE (Insectivora, Mammalia) FROM THE LATER TERTIARY OF OREGON By J. HOWARD HUTCHISON1 Museum of Paleontology University of California Berkeley ABSTRACT The Talpidae of the late Teritary of Oregon are described and their relationships discussed. Two new genera, Scapanoscapter and Achlyoscapter, a new subgenus, Xeroscapheus, and six new species, Mystipterus pacificus, ?Neurotrichus columbianus, Scalopoides ripafodwtor, Scapanoscapter simplicidens, Scapanus ( Xeroscapheus) proceridens, and Achlyoscapter longirostris, are recognized. INTRODUCTION The present paper is a continuation of the study of the fossil insectivores of Oregon. The fossil Talpidae of North America were, until recently, poorly known and with few exceptions the same may be said for the Old World mem- bers of the family. As compared with the his- tory of other middle and later North American Teritary insectivores, however, the talpid fossil record fares somewhat better. Unquestioned New World moles range in age from Oligocene2 to Recent and represent four subfamilies, the Uropsilinae, Gaillardinae, Talpinae, and Pro- scalopinae. Our knowledge of this long record is partly the result of the highly specialized and easily recognized humerus of talpids. Un- fortunately, the lack of humeri associated with 1 Contribution of the Museum of P aleontology, Univer· of California, Berkeley, and the Paleoecology Laboratory, University of Oregon, Eugene. 2 Paleocene (Torrejonian) to Recent if the badly broken humerus from Montana described by C. A. Reed (1954) is talpid. teeth has led to a confusion as to species, genera and even family assignment of the various ele- ments. The Oregon record is incomplete and does not wholly alleviate this problem but a study of the living genera and well known fossil forms goes a long way toward narrowing the choices available and offers evidence for prob- able associations. At the beginning of this study, the necessity of re-examining the postcranial skeletons of recent moles, particularly of the more obscure genera, was recognized and pursued. Although a superficial description and comparison of se- lected postcranial elements of available living moles is detailed below, a prodigious amount of work remains to be done on the comparative osteology and myology, and the functional morphology of the Recent Talpidae. The majority of the fossil material described in this paper was recovered by underwater sieving techniques under the direction of J. A. Shotwell ( see R. Wilson, 1965). No articulated material was recovered. 2 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 Morphological criteria were the main means of determining association of various elements. However, other criteria may also be useful. Relative abundance of various parts in sieved samples requires the tabulation of each kind of element and the assignment of high abund- ance elements to one form and low ones to another or others. This method is reasonably reliable in large samples. Another method of determining association relies on the proba- bility that when the same combinations of ele- ments recur in various localities they repre- sent the same species or genus. This method be- comes more reliable as the number of samples increases. The latter two methods of deter- mining association necessarily require some degree of taxonomic determination. Recurrent association proved of little use in this study owing to the sparsity of localities. In practice all three means in various combinations were used here to arrive at the assignment of isolated elements to taxa. The previous literature of New World Ter- tiary talpids is dispersed and varied. Reed and Turnbull (1956) have surveyed the literature dealing with the Proscalopinae. I shall restrict my review to the other talpids. Douglass ( 1903) described and figured the first valid North American Tertiary non-pro- scalopine talpid under the name ?Talpa platy- brachys on the basis of four humeri from the Late Miocene Flint Creek beds of Montana. After the work of Douglass, no new taxa of tal- pids were described until 1930, although Mer- riam ( 1911) figured and briefly described the humeri of a mole which he referred question- ably to Scapanus from the middle Pliocene Thousand Creek beds of Nevada, and Matthew ( 1924) mentioned but did not figure a humerus that he referred to Scalops cf. aquaticus from his Hipparion affine zone of the upper Snake Creek beds of Nebraska. In the same paper Matthew briefly described but did not figure a lower jaw fragment with an M3 which he named Talpa incerta. Hall (1930) described and fig- ured an M3 and associated dentary fragment from the early Pliocene of Nevada that he considered to be a bat and named it My- stipterus vespertilio. Patterson and McGrew ( 1939) transferred it to the Soricidae and Clark, Dawson, and Wood (1964) suggested other possibilities except Talpidae. Van Valen (1967) aligns Mystipterus question- ably with the Talpidae. Matthew (1932), in describing new material from the upper Snake Creek beds, figured and briefly de- scribed a partly endentulous dentary that he named Gaillardia thomsoni. Matthew gave no specific assignment for Gaillardia; however, he compared it most closely to members of the Talpidae, excluding it from the Desmaninae on the basis of a difference in dental formula. Wilson (1937) listed but did not describe a Scapanus species from the middle Pliocen~ Rome fauna of Oregon. Hibbard ( 1939) described and figured an incomplete humerus from the middle Pliocene Edson Quarry of Kansas. The humerus is that of the advanced fossorial type, such as Sca- panus and Scalopus, to which he compared it. He did not refer to it either of these genera but suggested that it might represent a new genus. Later in a series of papers on the late Pliocene Rexroad fauna of Kansas (1941A, 1941B, 1949 with Riggs, 1953, 1954) he re- ported a new genus and species of advanced burrowing mole under the name of Hespero- scalops rexroadi. Some of the postcranial ma- terial assigned to Scalopus in the earlier papers was later assigned to Hesperoscalops. In 1953, he stated that Hesperoscalops was probably the ancestral stock from which Scalopus was derived. During this time Gregory (1942) described but did not figure several humeri from the early Pliocene Big Spring Canyon, Lessert Spring and Minnechaduza localities of South Dakota and Nebraska. He noted the similarity of his fossils with ?Talpa platybrachys Doug- lass and the Fort Niobrara material (Domni- noides valentinensis K. M. Reed, 1962). Galbreath (1953) in his study of the verte- brate paleontology of northeastern Colorado, mentions several talpid remains from four dif- 1968 HUTCHISON: FOSSIL T ALPIDAE FROM OREGON 3 ferent faunas. He briefly described but did not figure a "Scalopine sp." humerus from the Cedar Creek member of the White River Oligo- cene which is" ... approximately three-fourths the size of the humerus of Scalopus aquaticus machrinoides, but otherwise does not differ in any respect from that of the Recent form." From the Miocene Martin Canyon local fauna, he described a lower molar and suggested that it " ... could well be the structural ancestor of Hesperoscalops Hibbard." From the late Mio- cene Kennesaw and Viv-Peetz faunas, he re- corded four incomplete humeri which he desig- nated tentatively cf. Condylura. Shotwell in 1956 and 1963, recorded Sca- panus sp. from two different Pliocene locali- ties in Oregon. In the earlier paper he also described a new genus and species of desman mole, H ydroscapheus americanus, based on a number of postcranial elements. In this same year Green (1956) described and figured the lower jaw of a new genus and species of insecti- vore from the early Pliocene of South Dakota that he believed to be a shrew and called Domninoides riparensis. Littleton and Crosth- waite (1957) listed Desmana moschata from the Idaho Formation on teeth identified by J. Hough. Wilson (1960), Tedford (1961), and K .Reed (1962), all pointed out that Domninoides is a mole. Wilson's (1960) paper dealing with the early Miocene rodents and insectivores of northeastern Colorado comprises the most im- portant study of Tertiary North American non- proscalopine moles to date. Two new genera . and species of talpid were described. He believed his M ydecodon martini compared fa- vorably with the living Urotrichus (Dymeco- don) pilirostris (True). His Scalopoides iso- dens did not seem to him to be directly an- cestral to any of the Recent New World moles but embodied a combination of characters of both Neurotrichus and Condylura in the hu- merus. In discussing the teeth of Proscalops secundus Matthew, he noted the similarities of the lower molars with those of Parascalops but discounted the likelihood that Proscalops was ancestral to either Parascalops or Scalopus stating that "needed transformation although not impossible suggests that P. secundus is aberrant." Throughout his discussions Wilson made thorough comparison ( where such is pos- sible) with pertinent Old World fossils, and suggests that Scalopoides and M ydecodon are represented in the Sansan and La Grive faunas of France. Contributions to the history and phylogeny of New World talpids have been made in every year since Wilson's paper. Tedford (1961) described and figured a new species of Sca- panus from the lower Pliocene of the Ricardo Formation of California under the name Sca- panus shultzi. He suggested that his species is " ... most clearly related to the hypsobra- chyodont living species, S. orarius and S. town- sendii, but could conceivably be ancestral to the hypsodont S. latimanus and even Scalopus ( via the Blancan H esperoscalops) as well." In the following year K. Reed described two new species from the great Plains, Domninoides valentinensis from the late Miocene of Ne- braska and H esperoscalops sewardensis from the late Pliocene Saw Rock Canyon fauna of Kansas, which was previously referred to H. rexroadi by Hibbard ( 1963). In addition to the lower jaw and P4, the humerus of D. valen- tinensis is figured. No extensive comparison of the humerus with other fossil moles was made and the other limb bones were not described. Macdonald ( 1963) described a new species under the name Domninoides evelynae based on a couple of fragmentary jaws containing the M2-M3, from the lower Miocene Sharps Formation of South Dakota. On the following page he reported Arctoryctes terrenus Matthew humeri from the same locality and suggested that Domninoides [ evelynae J may be asso- ciated with the humeri. Apparently Macdonald did not see Wilson's (1960) or K. M. Reed's ( 1961) papers which described the lower teeth of Proscalops for there is no mention of these in his paper nor any comparison of his D. evelynae teeth with those of Proscalops. In the same year James (1963) reported an 4 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 undescribed species of Scalopus on the basis of three incomplete humeri from the early Pliocene beds of the Cuyama Valley, Califor- nia. Clark, Dawson, and Wood (1964) in de- scribing some fossil mammals from the early Pliocene of Nevada reported a specimen of Domninoides cf. riparensis and a toothless jaw fragment that they compared to M ystiperus ves- pertilio Hall. Lewis in the same year listed a Talpa sp. from the Barstow Formation of Cali- fornia. Lastly, Firby ( 1966) in a list prepared by S. D. Webb, reported "?lnsectivora, genus undetermined-Arctoryctes species." This rec- ord from the late Miocene of Nevada is appar- ently based on the humerus of unknown affini- ties described by Reed and Downs ( 1959). Reed and Turnbull ( 1965) pointed out that this humerus is neither Arctoryctes nor a pro- scalopine mole. Most of the preceding records need com- ment in order to orient the reader as to the scope and synonymy of the various genera mentioned in the following discussion and de- scriptions. I have studied all specimens except those described by Hibbard, Littleton and Crosthwaite, Clark et al., Green, Lewis and the unreferred talpine of Reed and Turnbull. When Douglass ( 1903) described ?T alpa platybrachys he noted that the shaft is narrow and that "The ridge for the insertion of the pectoralis major muscle is well developed on the inner part of the humerus, but does not extend anywhere near so far outward as in Scalopus ." These features, as well as the gen- eral proportions of the humerus, agree very closely with those referred to Domninoides val- entinensis. Although the species must remain in doubt at present, the assignment of ?T alpa platybrachys to the genus Domninoides seems much more reasonable. Merriam's ( 1911) , Shotwell' s ( 1956, 1963), and Wilson's ( 193 7) records of Scapanus from Oregon and Nevada are valid and are discussed further in the sys- tematic section. The humerus described by Matthew (1924) and referred to Scalops cf. aquaticus certainly agrees closely with the modern Scalopus but the species is indeter- minate. Mathew considered the reference of his T alpa incerta to the genus T alpa as pro- visional. Re-examination and removal of the excess shellac from the Ma revealed that T alpa incerta is neither T alpa nor a talpid but an erinaceid. M ystipterus vespertilio Hall ( 1930) appears to be a uropsiline mole. Additional material of this form is described below in the systematic section. Matthew's (1932) Gaillardia thomsoni and Shotwell' s ( 1956) H ydroscapheus americanus appear to represent the same species and leave little doubt that this form is an aquatic mole which is best assigned to a new subfamily. I have not seen the humerus which Hibbard ( 1939) described from the Edson Quarry. It could represent either Domninoides or Scalo- pus (Hesperoscalops) although I favor the lat- ter judging from its rather stocky appearance ( see Addendum) . There seems little doubt that H esperoscalops rexroadi Hibbard and Hesper- oscalops sewardensis K. Reed ( 1962) are close- ly aligned with the living Scalopus and, as noted by Hibbard ( 1953), probably form the ancestral stock of the modern Scalopus. Indeed the morphological similarities between the specimens of Hesperoscalops and Scalopus are so great that there is little reason for generic distinction. The characters of the species as- signed to H esperoscalops seem consistent enough; however, to retain H esperoscalops as a subgenus and I shall treat it thus in sub- sequent discussions. The majority, if not all of the specimens described by Gregory (1942) from South Dakota and Nebraska, is assign- able to the genus Domninoides. Galbreath's (1953) Scalopine sp. humerus is, as he noted referable to the Recent Scalo- pus aquaticus, although he suggested that this specimen represented the humerus of Pro- scalops and it was so labeled in the University of Kansas collections. Reed and Turnbull's (1965) excellent study of the forearm of Arc- toryctes allows little doubt that Arctoryctes humeri are those of proscalopine moles, and this has been verified by additional associated 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 5 skeletons of proscalopine moles from both Col- orado and Oregon. As far as I can determine from Galbreath's paper, no other specimens were recovered from the same locality as the Scalopus-like humerus. On theoretical grounds, I find this occurrence anachronistic and believe it to be incorrectly located or a contaminant from a higher ( and younger) stratum. I shall not regard it as a valid age or locality record in further discussions. The humeri of cf. Con- dylura also mentioned by Galbreath are in- separable from those referable to Domninoides and are so considered hereafter. The lower tooth of a talpid listed from the Martin Canyon local fauna was not sef;n but, from Galbreath's description, appears to be either a proscalopine mole or Domninoides. The former is more likely in light of Wilson's ( 1960) discussion. Quinn ( 1955) listed T alpa in a faunal list from the Late Miocene Lapara Member of the Goliad Formation of Texas. To my knowledge this specimen ( not seen) has never been figured or described. Assignment to Talpa seems highly unlikely. It is probably a Domninoides but is best referred to as a talpine for the present. Attempts to locate the specimens of Desmana moschata (Linnaeus) reported by Littleton and Crosthwaite (1957) from the Idaho Forma- tion were unsuccessful. The identification is highly questionable on geographic grounds and even more dubious in consideration of the known fossil history of talpids in North Amer- ica. I shall ignore this record in subsequent dis- cussions (see Addendum). Although Domninoides riparensis was soon recognized as a talpid, Tedford ( 1961) was the only author to note its similarity to Para- scalops and Scapanulus. Unfortunately, these two living genera have been rather neglected in the studies of fossil North American talpids. As will be more fully elaborated upon below, they are important for our understanding of North American talpid phylogeny. Scapanus shultzi while still retained within the genus Scapanus, is not given the same phyletic po- sition in the phylogeny of the living species as was proposed by Tedford, but is aligned with the new subgenus Xeroscapheus. K. M. Reed's ( 1962) two new species remain unchanged. M ydecodon martini Wilson ( 1960) appears to be related to Mystipterus vespertilio Hall. I tentatively place Mydecodon as a subgenus of Mystipterus (see discussion under M. vesper- tilio). Although Wilson aligned his Scalo- poides isodens most closely with the Urotri- chus-like moles and Condylura, I believe it has closer affinities with Scapanulus ( which Wilson apparently did not see) and Parascalops. A re-survey of the unidentified bone scrap from Martin Canyon Quarry A yielded additional material of Scalopoides isodens which is de- scribed below. Wilson (p. 49) also described an upper molar under the heading of Soricoid? sp. indet. Re-examination of this specimen reveals that it is an upper molar of a didel- phid marsupial. It agrees quite closely with Peratherium. I agree with Macdonald ( 1963) that Domni- noides evelynae is a likely candidate for asso- ciation with Arctoryctes humeri from the same locality. In so doing I submit that D. evelynae is a proscalopine mole inasmuch as Arctoryctes humeri are now known to belong to proscalo- pine moles. The lower teeth of Domninoides are grossly similar in morphology to those of Proscalops ( compare fig. 4, K. M. Reed 1961, with fig. 2, K. M. Reed 1962). Proscalops evelynae (Macdonald) lacks the distinct meta- stylid of D; valentinensis and D. riparensis, as pointed out by K. M. Reed ( 1961) in a foot- note, and resembles Proscalops in this and other features. That Macdonald confused the lower teeth of Domninoides with those of Proscalops is evident from K. M. Reed's (1962) mention of a personal communication from Macdonald in which he "suggested that the jaws of his new species may represent the lower dentition of Proscalops secundus." Macdonald only dis- cussed some typological problems of the skull of Proscalops secundus under the heading of Domninoides discussion. The removal of this record restricts the known record of Domni- noides to the later Miocene and early Pliocene. The three humeri from the Caliente Forma- 6 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 tion of California assigned by James ( 1963) to Scalopus are quite unlike Scalopus in pro- portions. James noted the resemblances to humeri now referred to Domninoides platy- brachys and D. valentinensis. The configura- tion of the pectoral muscle scars is also like that in those species. Apparently much weight is given to the presence of a small foramen in the proximal end of the humerus. He noted its variability in Scapanus but indicates that its presence in all his specimens is an indication of relationship to Scalopus. I believe that strong similarities with D. platybrachys and D. valen- tinensis align it more closely with Domninoides than with any other living or fossil talpid. Lewis ( 1964) gave no indication where the material ascribed to T alpa sp. from the Bar- stow Formation now resides but the only teeth of a talpid that have been recovered from washing of the Barstow sediments by Tedford are those of Domninoides to which Lewis' T alpa is probably referable. No other talpid elements are thus far known to me from this formation. The humerus reported by Reed and Downs (1959) and listed by Webb in Firby (1966) as Arctoryctes is apparently the distal end of the humerus of Meterix latidens. A list of described moles and the usage as outlined above is sum- marized in Table l. For a group of animals represented by only about two score living species and by about an equal number of fossil forms, the Talpidae has a checkered taxonomic history. There have been almost as many generic names applied as there are living species. C. A. Reed (1951) pre- sented a concise history of the various usages and composition of the proposed subfamilies so that only a brief comment on subsequent classifications is necessary here. The Proscalo- pine proposed by K. M. Reed (1951) form a close-knit endemic group of North American talpids. The remaining moles have been grouped into two or more subfamilies of vary- ing composition by various authors. Borrowing from Gill (1875) and Wehe~ (1928), C. A. Reed ( 1951) arrived at a two-fold system of classification comprised of the Desmaninae and the Talpinae. As useful as this may have been for his purposes, his division of Cabrera's ( 1925) and Simpson's ( 1945) Scalopinae into desman and non-desman groups seems to be an unwarranted simplification of the relationships among these talpids. The inclusion of Condy- lura in the Desmaninae can only be justified on a few ( convergent) features owing to its slight aquatic modifications. The classification used here resembles that of Campbell ( 1939) who in turn used a combi- nation of Thomas' (1921) andWinge's (1917) classifications. He used five subfamilies, the Desmaninae, Urotrichinae, Condylurinae, Scalopinae, and Talpinae. Although he did not use the Uropsilinae, this is implied in his work. I have reduced most of his subfamilies to tribal or subtribal level in order to emphasize the more profound differences between the major adaptive and phylogenetic groups of moles. Van Valen (1967) has arrived at an identical classification of subfamilies and a general simi- larity of the tribes with the notable exception of the composition of the groups of moles includ- ed in the Scalopinae of Simpson ( 1945). Van Valen divides this group into two distinct tribes, the Urotrichini and the Scalopini. His Urotri- chini are all of the living Scalopinae of Simp- son except Scalopus and Scapanus, which com- prise all of his Scalopini. The characters on which he defines these groups revolve around the presence or absence of an enclosed bulla and features of the molar dentition. The less developed humerus is also used but the dis- tinction beween the humerus of Scapanus and that of Domninoides or Parascalops is certain- ly much less striking than that between Neuro- trichus and Scalopoides or Scapanulus. My im- pression of this complex of genera is also that there are two major divisions, the Urotrichini and the Scalopini. However, my Urotrichini includes only Urotrichus and Neurotrichus with the rest falling into the Scalopini. The Scalopini is in turn divided into two subtribes; the Parascalopina, including most of the other genera in Van Valen's Urotrichini, and the Scalopina, which equals Van Valen' s Scalo- 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 7 TABLE 1 SUMMARY OF ORIGINAL REFERENCES AND PRESENT USAGE OF VARIOUS DESCRIBED NEW WORLD TALPIDS OF THE LATER TERTIARY. Original Reference Present Usage ?Talpa platybrachys (Douglass 1903) ........ ... ......... ......... ... ....... Domninoides* platybrachys Scapanus (?) sp. (Merriam 1911) ... .... ... .... .. .................... .......... Scapanus sp. Scalops cf. aquaticus (Matthew 1924) .......................... ..... ... .... Scalopus sp. Talpa incerta (Matthew 1924) ....... ....... ...... ... ... .................. ... ..... Erinaceidae* Mystipterus vespertilio (Hall 1930) ....................... ... ........ .... ...... unchanged Gaillardia thomsoni (Matthew 1932) ......... ........ .. ...... .................. unchanged Scapanus (?) sp. (Wilson 1937) ....... .... ................. ....... ... .......... Scapanus sp. Talpidae (Hibbard 1939) ....... ........ ... ............. ....................... ....... cf. Scalopus* sp. Hesperoscalops rexroadi (Hibbard 1941A) ........... ........... ... ..... Scalopus* (Hesperoscalops *) rexroadi Scalopus sp. (Hibbard 1941B) .................................................... Scalopus (Hesperoscalops*) rexroadi Talpidae (Gregory 1942) ....... ... ...... .... ........ ............ .... ................ Domninoides* sp. Hesperoscalops rexroadi (Hibbard and Riggs 1949) .......... ...... Scalopus* (Hesperoscalops * ) rexroadi H esperoscalops rexroadi ( Hibbard 1950) .. ......... .. ... ... .... ........ ... Scalopus * (H esperoscalops *) rexroadi H esperoscalops rexroadi ( Hibbard 1953) .................... ..... ... .... .. Scalopus" ( H esperoscalops *) sewardensis Scalopine sp. ( Galbreath 1953) ........... ........ ... ... ... ... .... ... ..... ... .. .. Scalopus* sp. Talpid sp. ( Galbreath 1953) .................................. ....... ............... Proscalopinae * Cf. Condylura (Galbreath 1953) ......................... ... ... ....... ..... ..... Domninoides '' sp. Talpa sp. (Quinn 1955) ............... .......... ......... ............................... Talpini* Scapanus sp. (Shotwell 1956) ...................................................... unchanged Hydroscapheus americanus (Shotwell 1956) .... .. ..... .......... .. ....... Gaillardia* thomsoni* Domninoides riparensis ( Green 1956) ......................... .. ............. unchanged Desmana moschata (Littleton and Grosthwaite 1957) ........ .... .... . ?Talpidae"? Mydecodon martini (Wilson 1960) .... ..... .. ...... ...... ........ ........ ..... Mystipterus* (Mydecodon*) martini Scalopoides isodens (Wilson 1960) ....... ............. ........ ..... .... ....... unchanged Scapanus shultzi (Tedford 1961) .. .................. ............ ............... . Scapanus (Xeroscapheus*) shultzi Domninoides valentinensis (K. Reed 1962) .................... ......... ... unchanged Hesperoscalops sewardensis (K. Reed 1962) .... .......... .. ...... ...... ... Scalopus* ( Hesperoscalops) sewardensis Domninoides evelynae ( Macdonald 1963) ....... ... ......... ....... ........ Proscalopinae * Scalopus sp. (James 1963) ..... ........ ... .... ............ .............. ..... ....... Domninoides* sp. Domninoides cf. riparensis ( Clark et al. 1964) .............. ... .. .. ..... .. Scala pus* ( H esperoscalops) sewardensis Talpa sp. (Lewis 1964) ... ... .......... .. ..... ........... ............ ........ ....... ..... Domninoides* sp. '' Asterisk denotes taxonomic name change in this paper. pini. The osteological survey of the living gen- era below and the systematic section indicate the reasons for these groupings. The Scapto- nychini of Van Valen is accepted here for its utility as a waste basket group for the poorly known European middle Tertiary talpids and reception of the enigmatic Scaptonyx.3 I have deferred from assigning any of the fossil Eurasian talpids to my classification 3 Scaptonyx jaegeri Seeman (1938), as has been noted by others, is not Scaptonyx. Most or all of the European form s referred to Scaptonyx are not referable to this genus and probably belong to other tribes. scheme. The European literature is poor in the osteological descriptions or figures which are significant for my scheme of classification. To assign these forms without re-examination of this material would only tend to obscure the groups as defined upon available material. Furthermore, Van Valen (1967) presents a comprehensive classification of fossil and Re- cent talpids. Comments on first or second hand knowledge of some of the fossil Eurasian gen- era are footnoted in appropriate places. The European Talpidae are now under study. 8 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON CLASSIFICATION OF RECENT AND NORTH AMERICAN FOSSIL MOLES Family Talpidae4 Subfamily Uropsilinae Dobson, 1883. Uropsilus A. Milne Edwards, 1871. ( =Rhynchonax Thomas, 1911; =Nasillus Thomas, 1911) Recent; Asia. tMystipterus Hall, 1930. (Including Mydecodon Wilson, 1960) Middle Miocene-Early Pliocene; N. Am. tSubfamily Gaillardinae new tGaillardia Matthew, 1932. (=tHydroscapheus Shotwell, 1956) Middle Plioc.; N. Am. tSubfamily Proscalopinae K. M. Reed, 1961. (=Arctoryctinae, C. A. Reed and Turnbull, 1965) . tCryptoryctes C. A. Reed, 1954. Early Olig.; N. Am. tOligoscalops K. M. Reed, 1961. Middle Olig.; N. Am. tProscalops Matthew, 1901. (Including t Arctoryctes Matthew, 1907) Late Olig.-Early Mioc.; N. Am. tMesoscalops K. M. Reed, 1960. Middle Mioc.; N. Am. Subfamily Desmaninae (Thomas, 1912) Mivart, 1871. (=Myogalina Mivart, 1871) Desmana Gueldenstaedt, 1777. ( tMygale Cuvier, l800;=My- ogale Brandt, 1836) Early Plioc.-Recent; Europe. Recent; Asia. Galemys Kaup, 1829, Plioc.-Recent; Europe. Subfamily Talpinae Fischer von Valdheim, 1817. Tribe Scaptonychini Van Valen, 1967. Scaptonyx Milne Edwards 1872. ?Middle Mioc.; Europe. Recent, Asia Tribe Talpini Fischer von Valdheim, 1817. TalpaLinnaeus, 1758 (=including Mogera Pomel, 1848; Para- scaptor Gill, 1875; Euroscaptor Miller, 1940; Eoscalops Stro- ganov, 1941; Asioscalops [ not Asioscaptor] Stroganov, 1941) Late Miocene-Recent; Europe. Pleist.-Recent; Asia. Scaptochirus A. Milne Edwards. 1867. ( =Chiroscaptor Heude, 1898) Pleist.-Recent; Asia. Tribe Urotrichini Dobson, 1883. Urotrichus Temminick, 1841. (including Dymecodon True, 1886) Recent; Japan. Neurotrichus Gunther, 1880. ?Middle Plioc., Recent; N. Am. Tribe Condylurini Dobson, 1883. Condylura Illinger, 1811. Pleist. Recent; N. Am. Tribe Scalopini Dobson, 1883. Subtribe Parascalopina new tScalopoides Wilson, 1960. Middle Mioc.-?Middle Plioc.; N. Am. No.11 4 Only some of the newer and more common synonyms are listed-see Van Valen (1967) for more extensive syno- nymies. Authorship of the groups follows Van Valen except within the Talpinae where I believe some other authors were the first to adequately define the groups with refer- ence to my concept of them. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON tDomninoides Green, 1956. Late Mioc.-Early Plioc.; N. Am. Scapanulus Thomas, 1912. ?Middle Plioc.-Recent; Asia. Parascalops True, 1894. Pleist.-Recent; N. Am. Subtribe Scalopina (Van Valen, 1967) Winge, 1917. tScapanoscapter new. Late Mioc.; N. Am. Scalopus E. Geoffrey Saint Hilaire, 1803 (=Scalops Illiger, 1811; includes tHesperoscalops Hibbard, 1941) ?Middle Plioc.-Recent; N. Am. Scapanus Pomel, 1848. ( includes t Xeroscapheus new) Early Plioc.-Recent; N. Am. Talpinae incertae sedis Achlyoscapter new. Late Miocene; N. Am. 9 ACKNOWLEDGMENTS Mr. Elliott Adams, University of California, This work would not have been feasible with- Davis. out the great financial, critical, and moral sup- Dr. David Johnson, U.S. National Museum, port of Dr. J. A. Shotwell, Museum of Natural Washington, D. C. History, University of Oregon, Eugene. Mr. Dr. C. 0. Handley, Jr., U. S. National Mu- J. A. Suthard of Salt Lake City, Utah, kindly seum, Washington, D. C. allowed me to study and redescribe unpub- Dr. C. L. Gazin, U. S. National Museum, lished material of Mystipterus vespertilio Washington, D. C. which he collected while completing a Masters Dr. J. R. Macdonald, Los Angeles County degree at the University of California at River- Museum, California. side. C. A. Repenning of the U. S. Geological Dr. S. B. Benson, Museum of Vertebrate Zo- Survey at Menlo Park, California, generously ology, University of California, Berkeley. turned over for study undescribed talpid ma- Dr. Judith King, British Museum (Natural terial which he had collected. Dr. George Stas- History), London. ney, DDS, of Eugene, Oregon, freely provided Dr. C. C. Black, Carnegie Museum, Pitts- radiographs of several specimens. Drs. Lee burg, Pennsylvania. Van Valen and A. Zeigler stimulated discus- Dr. W. H. Burt, University of Michigan Mu- sions and allowed me to view unpublished seum of Zoology, Ann Arbor. manuscripts dealing with fossil and Recent Dr. R. H. Tedford, University of Califor- talpids. Drs. J. A. Shotwell, D. E. Savage, and nia, Riverside. W. A. Clemens read the manuscript and pro- Dr. M. 0. Woodburne, University of Cali- vided many useful criticisms. The following fornia, Riverside. people have greatly aided this endeavor by This research was supported under the co- lending specimens from the collections in their sponsorship of the University of Oregon Mu- keeping or providing specimens for dissection: seum of Natural History and University of Cal- Dr. M. C. McKenna, American Museum of ifornia Museum of Paleontology. Natural History, New York. Dr. C. E. Ray, U. S. National Museum, Washington, D. C. Dr. M. Hildebrand, University of Califor- nia, Davis. Dr. W. A. Clemens, Jr., then at the Museum of Natural History, University of Kansas, Law- rence, Kansas. MATERIALS AND METHODS The rank of subgenus is extensively used throughout this paper. Many of the subgenera used or proposed could be elevated to generic rank based on precedent in the taxonomy of this group. The various genera used here are not ;·.' ~ 10 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 i~ 2t c,j' i '"3 0... Q) "' "' (.) ... ... (/} ,.Q ~ Q) ::s Q) .... s :"Q t Q) > > '+< ('' ~ "'O "' "'O (/} = = (/} ~ (.) = '"3 "' Q) ~ "' 0... "' § Ef "' Number of ·~ "'O .... "' "'O "'O (/} = "' '"3 = :. = ...... ::s ::s "' = ,.Q "' ... = ...... ~ best specimen "' ,.Q ... ... (/} s <..:::: (/} ~ ::s Q) Q) ::s 0 ::s = .... s ~ 0 :..s = or institution ..II: "' (/} ::s "' = "' No.·j· . ~ ::c:: = ~ ~en p:: ...... . USNM 256119 x x x x x x x 1 Uropsilus soricipes I MH 1027 x x x x x x x 1 Urotrichus pilirostris MVZ ! x 10+ Urotrichus talpoides I UO and MVZ x x x x x x x 10+ Neurotrichus gibbsii USNM 172374 x x X" x x X* 1 Galemys pyrenaicus YPM 1189 x x x x x x x 2 Desmana moschata MH 73, MH 115 x x x x x x x 2 Condylura cristata ' AMNH x 2 Scaptonyx fusicaudus USNM 233838 x x x x x x x 4 Talpa europaea MVZ 62754 x 1 Talpa occidentalis USNM 253319 x x x x x x x 2 Talpa micrura USNM 201318 x x 1 Scaptochirus moschatus ;: USNM 240757 x x x x··· 1 Scapanulus oweni MHMH 233 x x x x x x x 1 Parascalops breweri UO and MVZ x x x x x x x 40+ Scapanus ( all species) UF and JHH x x x x x x x 10+ Scalopus aquaticus I ; t This number indicates the minimum number of specimens in all ca tegories. * Fragmentary pelvis or proximal end of a long bone only. e so highly diversified at the species level as to jaws and teeth. No taxa are named on isolated ! warrant the elevation of various subgenera to teeth, whether upper or lower. At the present genus rank. There is, on the other hand, a great rate of collecting, it should not be long before utility to the retention of a rather broad gen- these unnamed forms are known from either ' eric usage within this family. Thus, limb ele- more or better material. ments which are unquestionably referable to Wherever possible, I have attempted to as- the Scapanus complex may be referred to that sign isolated upper teeth to species or genera genus where further refinement is impossible. known from lower teeth from the same locality Rather than run the risk of duplicating taxa when consistency of size and lack of conflict based on different elements, I have followed with more reliable assignments from other the policy of attaching names only to lower samples were met. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 11 Measurements of the larger limb bones were taken with a Vernier caliper. All other meas- ures were taken with either a Gaertner measur- ing microscope or ocular micrometer. The scale is not uniform and scale lines in all figures are equivalent to one millimeter. All the figures except those of the major limb bones in the osteology section were drawn with the aid of an ocular grid in a binocular microscope. The methods of measurement of the various ele- ments are either illustrated in the osteology section, or explained in the text or in Hutchison (1966). Abbreviations used throughout this paper are: AMNH=American Museum of Natural History CIT=California Institute of Technology (Material now the possession of Los Angeles County Museum) JHH-Author's collection M=United States Geological Survey ver- tebrate locality MH=Milton Hildebrand Collection, Da- vis, California MVZ=University of California Museum of Vertebrate Zoology, Berkeley UCR=University of California at River- side UF=University of Florida, Cainsville UK=Museum of Natural History, Uni- versity of Kansas, Lawrence UO= University of Oregon Museum of Natural History, Eugene USNM=United States National Museum YPM=Yale Peabody Museum, New Ha- ven, Connecticut The recent specimens were assembled from diverse sources and ranged in completeness from skulls to entire skeletons. Since the osteo- logical concept of the various proposed group- ings was influenced or dependent upon mater- ial available to me, a detailed listing of ma- terial is presented. Oregon localities from which material is discussed in this report range in age from Bar- stovian to Blancan. The geographic positions are indicated on the map of Figure 1. Other in- formation pertaining to the localities is given in the locality descriptions at the end of the text. Their ages are as follows: BARSTOVIAN Quartz Basin UO 2465 Red Basin I UO 2493 Red Basin II UO 2495 M 1040 Beatty Butte M 1041 Guano Ranch M 1042 Snyder Creek M 1043 CLARENDONIAN Black Butte I UO 2337 West of Riverside UO 2489 Black Butte II UO 2500 HEMPHILLIAN McKay Reservoir UO 2222 Krebs Ranch I UO 2322 Krebs Ranch II UO 2323 Otis Basin UO 234 7 Westend Blowout UO 2349 Bartlett Mountain II UO 2357 Little Valley I UO 2380 Little Valley II UO 2516 Bartlett Mountain UO 2517 Rome CIT 62 Arlington CIT 375 BLANCAN? Enrico Ranch UO 2223 12 4 60 4 5° 4 4° 430 420 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 120° i --- ~ ! L·- ·- ·- ·- ·1 '---, 120° I i i i •Mil042 119° 0 5 JO 20 30 MI LES 1190 118° 117° I I l \ i j j i \ ~ r,.~! f ' j i .-·- ·- i 4 60 4 50 440 43° 420 Figure 1. Location of sites from which talpid material is discussed. See text for additional info rmation. - ------- - -- - --- 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 13 OSTEOLOGY OF LIVING TALPIDAE Comparative studies on the postcranial skel- etons of talpids are rare. The most thorough and useful are those of Campbell (1939) on the comparative shoulder anatomy of the moles and C. A. Reed ( 1951) on the locomotor and appendicular anatomy in soricoid insecti- vores. The osteology of the living moles dis- cussed below is not intended to supplant these studies but rather supplement their discussions with brief descriptions of material that was not discussed or not available to them. The following descriptions should be used in con- junction with these papers. No attempt is made to be complete in the following survey but rather to briefly review or cover the general description of those elements which are most commonly recovered as fossils. The odontological and osteological termin- ology is derived primarily from C. A. Reed ( 1951), but also from a number of other sources. A few terms are coined. In order to avoid changes in terminology resulting from the different orientation of the limbs of vari- ous talpids in life, I have assumed for purposes of discussion of the individual bones that all the elements are oriented as in a generalized plantigrade mammal. Thus the lateral side of the humerus of Scapanus which in life is me- dial with respect to the axis of the body is still referred to as the lateral side. The plantar side of the foot is considered ventral. UROPSILINAE DOBSON, 1883 There is only one generally recognized liv- ing species of this subfamily, Uropsilus sori- cipes. Since no general osteology of this unique form has been set forth in one place, I shall attempt to mention some of the more salient features. The skull is generally shaped like that of N eurotrichus and the bullae are open both ventrally and dorsally. The lacrimal foramen 1ms Figure 2. Zygomatic region of the maxilla, left lateral and ventral views. A; Uropsilus soricipes, USNM 256119, B; Neurotrichus gibbsii, JHH 516. if-infraorbital foramen; ims-insertion of M. masseter superficialis ; lf-lacrimal foramen; sp-subzygomatic process; z-zygomatic arch. 14 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 is unusually large, being as large as the an- terior opening of the infraorbital foramen which is situated immediately below it ( see Fig. 2A). The zygomatic arches are strongly dorsad thus descending on the maxilla from above rather than from behind as in other talpids. Just ventral to the zygomatic process of the maxilla is a laterally projecting tubercle, here referred to as the sub-zygomatic process, which apparently serves for the insertion of the M. masseter superficialis. Apparently, this inser- tion has migrated onto the arch proper in other talpids as the zygomatic arch came to lie more ventrally. The dentition is reduced and the anterior incisors above and below are enlarged. The postfossid of the lower molars opens out lin- gually. The upper molars are peculiar in that there is a very well developed shelf extending posteriorly from the metaconule (Fig. 2A) which results in the noticeably concave poster- ior margin of the teeth, a condition unique to Uropsilus among the living talpids. The specimen in hand (USNM 256119) is somewhat damaged in the thoracic region. There are six lumbar vertebrae and presumably 13 thoracics. Six coossified vertebrae comprise the sacrum, and there are at least 16 and prob- ably a few more caudal vertebrate. The manubrium of Uropsilus is essential- ly like that of Sorex. The shaft of the manu- brium (Fig. 3) is rod-like with an incipient ventral midline ridge. The alae extend later- ally from the anteriormost end of the manu- brium and are not expanded anteroposteriorly as in Neurotrichus. The mesosternum consists of three elements. The clavicle ( Fig. 3) is relatively simple, long, slender, and shows a sigmoid flexure comparable to that in man. Campbell ( 1939) noted that "in the relations of the distal end of the clavicle Ursopsilus stands intermediate be- tween the mole type with the articulation lim- ited to the humerus and the shrew type where it is confined to the acromion. Uropsilus exhibits an articulation both with the high bifed shrew- Figure 3. Uropsilus soricipes, USNM 256119, ven· tral view of manubrium and left clavicle. a-ala; C---clavicle; m-manubrium ; r-costal cartilage. like acromion and with the humerus." The scapula also shows a combination of shrew and mole characters. There is a "large bifid acro- mion, with its distally directed process, on which the clavicle articulates, and the posterior process, the metacromion .... ". "The supra- spinatus fossa is scarcely represented, the in- fraspinatus fossa is a minute groove, and the fossae for the subscapularis and for the teres major are unmarked, ... " (Campbell, 1939, Fig. 21). The glenoid fossa is oblique to the long dimension of the scapula. "In general configuration the humerus of Uropsilus resembles that of Blarina. The distal end is nearly identical. The head of the hu- merus is rounded and not elliptical as in the true moles" ( Campbell, 1939). The distal end is markedly broader than the proximal end (Fig. 40). The bicipital groove is open through- out its length. No ridge connects the lesser tuberosity with the teres tubercle. The clavicu- lar articular facet is ill defined on the greater tuberosity. There is no fossa for the origin of the M. flexor digitorum profundus ligament. The ulna of Uropsilus shows few if any mod- ifications for burrowing. It is long, slender, and broadly sinusoid with the proximal portion more flexed (Fig. 4). The olecranon process occupies proportionately a very small extent of the length of the ulna. The terminus of the olecranon process is not expanded laterally as in the shrews or other moles. The area of in- sertion of the triceps muscle is reflected medi- ally. The triceps scar is not well defined and a 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 15 ridge or crest for the extensor muscles is not developed. The abductor fossa is very narrow and the posterior crest thin. There is no exter- nal olecranon crest and the internal crest is simply the continuation of the shaft which forms the olecranon. The processus anconeus is poorly delineated and the coronoid process is not developed. The semilunar notch is corre- spondingly broadly open anteriorly. The shaft is long and slender, and the relatively narrow distal articular surface forms an oblique angle with the manus. Like the ulna the radius of Uropsilus is also shrew-like. It is long, slender and gently arched (Fig. 4). There is no capitular process. Extend- ing obliquely and distally from the ulp.ar artic- ulation is a narrow ridge which may be the structural predecessor to the strong crest in this region of the Urotrichini. The distal end of the radius is relatively broader than that of the shrews, even Blarina. The manus of Uropsilus shows no specializa- tion for digging. As noted by C. A. Reed ( 1951) the metacarpals of U ropsilus are long, slender, and generally unspecialized as in most shrews. The longest metacarpal is that of digit three ( about six times as long as wide). The carpus of Uropsilus was not completely pre- pared but the scaphoid and lunar appear to be fused to form a scapholunar as in soricids. The phalanges are also long and slender. The un- gual phalanges are laterally compressed and not bifid. The distal articular surfaces of the metacarpals are flatly cylindrical or slightly convex. There is a median keel as in shrews on the ventral side of the trochlea. Although a radial sesamoid was not noted by previous authors, Reed ( 1951) has pointed out that this bone is easily lost in preparation. A long, slen- der radial sesamoid is present in USNM 256119. In the pelvic girdle there is no pubic sym- physis or pseudosymphysis. The ilium is not fused to the sacrum ( in the USNM specimen) and only two sacral vertebrae contact the in- nominate. B A Figure 4. Uropsilus soncipes, USNM 256119,A; right radius and ulna, right lateral view, B; righL ulna, anterior view of proximal end. The femur in shrews may be relatively longer (Sorex) or shorter (Blarina) than that of the moles. In the shrews the distal margin of the third trochanter is distal to the lesser trochanter thus imparting a rather "long look" to the proximal end of the femur when com- pared to that of the true moles. In most talpids the third trochanter tends to become aligned transversely with the lesser trochanter thus clustering the insertions of the pelvic muscles near the head and imparting the broad aspect to the proximal end of the femur. The femur of Uropsilus is relatively as long as that of Blarina and the relationships of the trochanters are shrew-like. Uropsilus also differs from the Talpinae in that the proximal articular sur- face only extends along the dorsal side of the neck toward the greater trochanter, the inter- trochanteric crest is evenly curved and well de- fined, and the third trochanter is below the level of the lesser trochanter. 16 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 The tibiofibula of Uropsilus is characterized by the general shrew-like aspect, a long and slender shaft, simple proximal end of the fib- ula with only an incipient development of the lateral and posterior processes which are rep- resented by small ridges, and the falciform process of the tibia expanded into a wide flange. The pes is larger than the manus and the metatarsals are unspecialized. PROSCALOPINAE K. M. Reed, 1961 Although this subfamily is entirely extinct it seems appropriate to include a brief sum- mary of characteristics for completeness. K. M. Reed ( 1961) gave the following diag- nosis for the Proscalopinae: "Skull chrysochlorid-like; dentition of tal- pid type. Skull wide and deep in posterior part; flexure in maxillary, between P3 and P4, stronger than in N eurotrichus; lateral projec- tions on premaxillaries; slight sagittal and oc- cipital crests; palate long relative to skull length; antorbital rim confined to maxillary, not joining main body of zygoma. Wide lingual shelf on upper molars with hypocone [meta- conule] varyingly developed; dental formula It q P~ M!." Other features of Proscalopinae include the ossified bulla, infraorbital foramen much larger than lacrimal foramen, straight zygo- matic arch, anterior incisor above and below (?) enlarged, postfossids open lingually; humerus very broad, distal and broader than proximal end, bicipital groove long and open, head laterally compressed, fossa for ligament of M. flexor digitorum profundus present: ulna with coronoid process well developed, trans- verse proximal crest; radius short with well developed capitular process; manus broad, scaphoid and lunar separate; ungual phalanges depressed but not bifid; pes probably smaller than manus; rest of skeleton undescribed. See Turnbull and Reed ( 1965) for more detailed descriptions. DESMANINAE (Thomas 1912) Mivart, 1871 The Desmaninae form a very compact group at present. All the known members are semi- aquatic or amphibious and show corresponding osteological and myological modifications. The desmans show marked modifications of the limbs, being longer proportionately in the dis- tal elements of the limbs than are the other moles. The hind feet are webbed and the meta- carpals and phalanges are greatly elongated for support and effectiveness. They also show the characteristic great development of the third trochanter and expansion of the distal end of the femur neither of which are as prominently developed in the other Talpids. The tibia is proportionately very long and deeply grooved for the tendons of the extensor muscles which are characteristic for mammals that use the hind limbs for aquatic locomotion. Dobson's ( 1882) monograph remains the most complete treatment of the myology and oste- ology of the Desmaninae (Desmana) but Schreuder' s ( 1940) revision of fossil des mans is the best source for separating the various fossil and recent genera from each other on os- teological criteria. There has been little confu- sion in the assignment of desman postcranial elements to the correct subfamily. The subfamily can be diagnosed as follows: bulla ossified but small; infraorbital foramen much larger than lacrimal foramen; hypocone- like metaconule moderately developed, I1 and b enlarged; postfossids open lingually; manu- brium with expanded anterior alae, ventral keel weak; mesosternum with three elements; clavicle simple and stout, ventral spine and scapular articulation absent; scapula with metacromion, glenoid fossa nearly perpendic- ular to long dimension, subscapular and teres major fossa present, suprascapular foramen present; humerus (Fig. 40) relatively slender, distal end broader than proximal end, bicipi- tal groove short and partly closed over by fusion of the pectoral process and lesser tuber- osity, head laterally compressed, fossa for M. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 17 flexor digitorum profundus ligament present; ulna without coronoid process, transverse proximal crest present; radius with capitular process absent or poorly developed; manus moderately expanded, ungual phalanges bifid, scaphoid and lunar separate; pubic symphysis present, pseudosymphysis absent, ilium fused to sacrum; five caudal vertebrae in sacrum, three vertebrae contacting innominate; tail very long; femur expanded proximally and distally, shaft compressed anteroposteriorly; tibiofibula long and deeply grooved distally, fibula complex proximally; pes larger than manus, metatarsals laterally compressed and at least some twisted. TALPINAE Fisher Von Valdheim, 1817 SKULL: Complete skulls of fossil moles are seldom found and the discussion of the skull will only be treated lightly here. The general proportions and morphology of the living tal- pid skulls have been described in numerous other sources. Some of the general trends to be noted within the Talpinae are the general broadening of the cranial cavity, increase in coossification of the roofing bones, and devel- opment of the ossified bullae. All these trends are seemingly associated with an increasing degree of fossorial specialization. In such high- ly fossorial moles as Scalopus, Scapanus, and T alpa, the ventral floor of the bulla is formed by a ventrolateral flange of the medial arm o'f the tympanic ( ectotympanic) bone. This flange is only slightly developed in Scaptonyx, the Urotrichini, Condylura, and the Parascal- opina, so that the tympanic membrane is visible in ventral view. The tympanic bone is not fused with the surrounding bones in the later genera but becomes completely fused in the Scalopina, Talpa (Talpa), and Scaptochirus. In Talpa (Mogera) the tympanic is overlapped margin· ally by the bullar wings of the surrounding bones which are not fused with the tympanic. The stapedial artery and the veins still with- in the tympanic cavity are enclosed by bony tubes in the highly fossorial moles-Talpini, and Scalopina. The stapedial artery is only partly enclosed or thinly so in Parascalops and Condylura; a bony tube is absent in the Urot- richini, Scapanulus, and Scaptonyx. In none of the species does the stapedial ar- tery or bony tube completely fill the stapedial fenestra and apparently the bony tube does not contact the stapes. Concomitant with the de- velopment of an ossified stapedial tube is the closure of the pyriform fenestra. It is open in the Urotrichini, Scaptonyx, and Scapanulus and closed in the Talpini, Condylura, Parasca- lops, and the Scalopina. Stroganov ( 1945), in a comparative study of the auditory ossicles of the living Talpidae, grouped the moles into six taxa; the Desmani- dae, Talpinae, Scalopinae, Condylurinae, U rotrichinae and U ropsilinae. It is not clear in his discussion whether he used the terms Urotrichinae and Uropsilinae interchangeably. Parascalops, Scapanulus and Neurotrichus were not available to him. I have made no dis- sections of the auditory region of these moles but the form of the malleus is visible in the skulls available. The malleus of Neurotrichus resembles that of Urotrichus figured by Stroga- nov. Parascalops and Scapanulus resemble each other in the form of the malleus being in- termediate in morphology between the Sca- lopina and Condylura. Proportionately the manubrium and processus gracilis bear the same angular relationships to the articular sur- face as in the Scalopina, but the posterior mar- gin of the lamina is sharply angulate as in Con- dylura. The apophysis orbicularis is intermedi- ate in size between Condylura and the Scalo· pma. DENTITION: There has been much attention given to the dentition, particularly the number of teeth, as a means of classification. This has led to a proliferation of genera which, in its extreme form, resulted in a large number of monotypic genera. Other workers ( Osgood, 1937; Jackson, 1915; Palmer, 1937; and Schwartz, 1948) have noted the variability of the number of teeth and have synonymized 18 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 mts prs l pac pal pre A pae B Figure 5. Nomenclature of upper teeth, A; Scapanulus oweni, left M2, B; Uropsilus soricipes, left M1, C; U. soricipes, right P 2• l-length; e-ectoloph; le-labial cingulum; pac-paracrista; pae-paracone; pal-paraconule; pc-paracingulum; pe-post ectoflexus; pf-protofossa; pre-protocone; prs- parastyle; mc-metacingulum; mss-mesostyle; mt-metacone; mtl-metaconule; mtlf-metacon- ule flange; mts-metastyle. many of the named genera. The pendulum swung completely the other way in some groups so that the Talpini is represented by only one living genus, Talpa (Schwartz 1948). Eller- man and Morrison Scott ( 1951) exceeded even Schwartz in that they recognized only three living species of Talpa as compared with the six of Schwartz. Valid as this limited approach may be, there are other less variable character- istics of the postcranial skeleton which can be used to evaluate some of the other generic units. These postcranial characters will be touched on under the appropriate category. I shall not attempt to summarize the dental formulae of the living and known fossil moles except where need be in the systematic discus- sion; these may be found in the above named sources and elsewhere. There are few gross dentitional modifications exhibited by the tal- pines as a group. The hypertrophy of the 12 occurs independently in at least two lineages. The hypertrophy of P1 seems to be naturally related to the Talpa complex as a whole (Scap- tonyx here included). Extreme molar hypso- donty only develops in two lineages ( one ex- tinct) of the Scalopina ( those with the hyper- trophied 12 ) but is incipient in a third (Scapa- nus latimanus). The reduction in the number of antemolar teeth is invariably associated with a general shortening of the rostrum. This short- ening of the rostrum has apparently occurred again and again in divergent lines of moles and is of little taxonomic significance itself but may be very useful in evolutionary progressions when the systematic position of the form can be delineated on other criteria. Condylura, Para- scalops, Scapanus, Talpa, and Scaptonyx all contain species with dentitions of the complete eutherian dental formula. Upper teeth ter- minology is illustrated in Figure 5. There are two major molar patterns in the lower dentition which may, in general, be char- acteristic of a subtribe, tribe, or subfamily. The first type consists of a pattern in which the crista obliqua of the M2-M:~ joins the trigonid directly or just labial to a well developed me- tastylid. The postfossid is usually open in this case ( see Fig. 6A). In the second case, the crista obliqua joins the protolophid and the metasty- lid and is usually weak or absent. In this pattern 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 19 I ent pae pac aac hye hyf A pa] ac I co ets B Figure 6. Nomenclature of lower teeth, A; Scapanu- lus oweni, left M2, B; Neurotrichus gibbsii, left M2 • I-length; aac-anterior accessory cuspid; ac-anterior cingulum; co-crista obliqua; ent-- entoconid; entc-entocristid; ets-ectostylid; hye-hypoconid; hyf-hypoflexid; mt--meta- conid; mts-metastylid; pac-posterior acces- sory cuspid; pae-paraconid; pal-paralophid; pc-postcristid; pef--prefossid; pre-protoco- nid; prl-protolophid; ptf-postfossid. an entocristid is frequently developed which tends to isolate the postfossid ( see Fig. 6B). lntergradation between these types is not un- common, although a tendency toward one or the other is usually discernible in moles with relatively higher crowned ( mesodont) teeth. A comparison of selected dental features is presented in Table 2. It might be inferred from the table that the upper canine is double-rooted and somewhat enlarged. Only Scaptonyx and the Talpini have reduced incisors and hyper- trophied upper canines. VERTEBRAE: Vertebrae other than the sacral vertebrae are seldom identified in fossil re- mains except when associated with an articulat- ed skeleton. The number of vertebrae in the var- ious body regions has been noted to vary. The cervicals invariably number seven. Thirteen thoracic vertebrae bearing functional ribs and six lumbar vertebrae were noted in Urotrichus, Neurotrichus, Parascalops, and Talpa. Of the two specimens of Condylura examined, one had the above condition and the other bore a ves- tigial and nonfunctional rib on the fourteenth thoracic vertebra, thus leaving only five lum- bars. Scapanus, Scalopus, and Talpa (Ma- gera) also had five lumbar vertebrae and four- teen thoracic vertebrae with functional ribs. Slonaker ( 1920) reports six sacral vertebrae for Scapanus and Scalopus and lists a num- ber of records by others which tend to sup- port his view. C. A. Reed (1951) and others (Leche 1883) observed only five sacral verte- brae and my observations of available talpi- dines support this view. Gupta ( 1966) also con- sidered five as the typical talpidian number but noted the variable fusion of the first caudal vertebra to the sacrum in Scalopus. The aqua- tic talpine Condylura has seventeen to twenty- one caudal vertebrae, sixteen were counted in Urotrichus, and the fossorial moles have the lowest counts: Parascalops, thirteen; Scapa- nus, thirteen to fourteen; Scalopus, ten to eleven; Talpa (Talpa), ten. MANUBRIUM: The manubria of Scaptonyx, Con- dylura, and Urotrichini, have large alae orig- inating medially and the ventral ridge forms a deep keel ( Campbell, 1939, Figs. 3-5). Scap- tonyx and the Urotrichini exhibit anterior ex- pansions on the alae which serve as the origin for part of the M. costoscapularis ( subclavius of Campbell). In the Scalopini, "the heightening of the keel and the decrease in width (both real and pro- portional) is carried still further." ( Campbell, 1939, Figs. 6-8). The Parascalopina have a dorsal ridge between the two alae which is lack- ing in the Scalopina. In the Scalopina and Tal- TABLE 2 Comparison of Selected Dental Features Neurotrichus Urotrichus Condylura Scapanulus Parascalops Scapanus Scalopus Talpa Scaptochirus Scaptonyx Relative size of upr C1&I1 I = C I>C Size of upper II enlarged enlarged 12 > II > C enlarged I>>C I>>C I>>C I>>C KC KC KC hypertrophied hypertrophied hypertrophied hypertrophied small small enlarged No. of roots on upr canine 2 1 1 2 1 1 1 2 2 K.—less than >—greater than > >—much greater than A—Antemolars where homologies are questionable—anterior tooth=Al :,:—homologies questionable or teeth of variable occurrence withinor among species No. of roots on upr P4 2 3 3 3(?) 3(?) 1 1 3 3 of Recent Talpid Genera No. of Relative size of Lower PI roots on lwr PI anterior lower antemolars Metastylid of lower M2 small or absent* 1 Al > A2 absent small or absent* 1 Al > A2 aosent normal 2 C>I1^I2 present normal or absent* 1 12 > 11 present normal 1 12 > 11 present normal or absent 1 12^11 absent small or absent 1 12 >> 11 absent hyper trophied 2 PI >> 11-3 absent hyper trophied 2 PI >> 11-3 absent enlarged 2 PI >> 11-3 present IS3 O to h I Co o I 8 o © o 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 21 pini5 this ridge is replaced by a groove for the anterior vena cava which divides posteri- orly. This vein is double in Parascalops and Scapanulus although the dorsal ridge may be perforated by a foramen. In the Talpini the groove for the vena cava leads off posteriorly to the right instead of dividing and the clavi- cular facets are inclined at nearly a 45° angle to the shaft, with the ventral margins anterior- most. The reclined articular facets are also typical of Condylura, Scaptonyx, and the Uro- trichini. In the Scalopini the clavicular facets are nearly vertical. The manubrium of Scapanulus has not been previously described. It most closely resembles that of Parascalops. As in Parascalops there is a dorsal ridge between the two alae. The anter- ior vena cava is divided by the dorsal ridge, and a foramen perforates the ridge near the alae. The length of the manubrium is relatively shorter than that of Parascalops. It seems appropriate at this point to men- tion the unique presence in Scapanulus and Parascalops of a tetrahedral heterotopic hone ( Fig. 7) wedged in between the ventromedial spine of the clavicle and anterior basilateral portion of the manubrium. These heterotopic bones are relatively larger in Scapanulus and the manubrium is correspondingly narrower anteroventrally than in Parascalops. These bones were not observed in any of the other moles examined nor was there any particular space available for them in the articulated skel- etons. In Parascalops the anteroventral extrem- ity of the manubrium is proportionately wider in response to the relatively smaller heterotop- ic bones. In Scapanus and Scalopus, which lack these bones, this portion of the manubrium is still broader. The development of such hetero- topic bones may be related to the shift of the clavicle articulation from an oblique to a ver- tical position with respect to the manubrium. Apparently, there is some ventrally directed force on the proximal end of the clavicle which 5 The sternum bone of Scaptochirus primitivus Zdansky figured by C. C. Young 1934 (Fig. 2A) appears to be the first phalanx of the second digit of a bird. m A th m B Figure 7. Anterior view of the right clavicle in ar- ticulation with the manubrium, A; Scapanulus oweni, B; Parascalops breweri. h-heterotopic bone; m-manubrium (right half); th-tetrahed- ral heterotopic bone; vp-ventral process. would, unchecked, tend to disjoint the clavicle. In the non-scalopines this force is nulified by the dorsolateral position of the manubrium articular facet. MESOSTERNUM: The mesosternum of moles is said by Slonaker (1920) to consist of a vari- able number of parts. Flower ( 1870) lists the number of sternebrae as high as five although his figure clearly shows only three. All my ob- servations bear out the fact that there are only three sternebrae in all the moles examined. Three to five sets of ribs may tenuously contact the last sternebra. CLAVICLE: The clavicle of the U rotrichini and Scaptonyx is roughly at least twice as long as deep (Campbell, 1939, Figs. 13-14). There is a well developed ventral process and no reflec- tion of the ventromedial spine. The dorsal prominence is very pronounced owing to the I 22 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 oblique angle of the sternal articular facet in relation to the long diameter of the clavicle. The clavicle of Uratrichus pilirastris (Fig. 8B), although resembling that of N euratrichus (Fig. 8A) in many respects, differs from the latter in its narrower shaft, lower and less com- plicated spine, and more obtuse angle between the planes of articulation. Campbell's illustra- tion of the clavicle of Neurotrichus is too gener- alized to show the nature of the spine. It is ex- panded into a plate ventrally and proximally so that it nearly converts the vascular notch into a tunnel. The clavicle of Candylura (Campbell, 1939, Fig. 15) is similar to that in the Urotrichini but relatively shorter. There is an additional process on the anteromedial side for the origin of the distal division of the M. acromiodeltoi- deus ( cleido-deltoid of Campbell). The short- ness of the clavicle in Candylura compared to the Urotrichini and Scaptanyx is the result of the shortening of the distal moiety. The planes of the articular facets are decidedly not par- allel. The clavicle of the Talpini is as variable in morphology as it is in the Scalopini. The artic- ular facets are roughly parallel and the clav- icle is quite short, being as long as deep, or less so. There is a prominent hook-like ventral pro- cess in Talpa (Talpa) (Campbell, 1939, Fig. 19) .The process is simple in Talpa (Magera) ( Campbell, Fig. 20). The vascular notch which delineates the ventral process from the vent- romedial spine is not continued as a gutter pos- teriorly as in the Scalopini. In Talpa (Magera) the vena cava pierces the clavicle and the vent- romedial spine is thus undefined and the vas- cular notch is absent. The clavicle of the Parascalopina is longer than in the Scalopina but shorter than the Con- dy lurini. The gutter on the posterior side lead- ing from the vasular notch toward the dorsal prominence is usually well defined. A vascular foramen usually pierces the clavicle within the vascular notch just above the ventral process. The ventromedial spine is deflected laterally in both the Parascalopina and Scalopina by the tetrahedral heterotopic bone in the former and the basilateral buttress of the manubrium in the latter. The clavicle of Scapanulus (Fig. 7 A) is shorter and broader than that of N euratrichus and Candylura but longer and narrower than in Parascalops. As might be expected, Scapan- ulus most closely resembles Parascalaps in the general aspect of the clavicle. The notch is sharply defined by the ventral process which nearly encloses it ventrally. The ventromedial spine is broadened and reflected laterally so that the articulation of the heterotopic bone is horizontal. In Parascalops (Fig. 7B) the ven- tromedial spine is reflected to about a 45 ° angle with the horizontal as in Scalapus and Scap- anus. In the latter two genera, however, the ventromedial spine articulates directly with the manubrium rather than with a heterotopic bone. The clavicle of the Scalopina ( Campbell, 1939, Figs. ] 7-18) is even shorter than that of the Parascalopina. The ventral process is reduced to a ridge. In Scalopzts as in T alpa (Magera) the vena cava pierces the clavicle. If the living talpines are arranged in a struc- tural progession from the least fossorial to the most, several trends become apparent. As the manubrium increases in length the clavicle shortens and is rotated posterolaterally in re- lation to the manubrium.6 However, the hum- eral articular facet in all the living moles, ex- cept Urapsilzts, continues to face essentially laterally. Therefore, as the clavicle shortens, the sternal and humeral articular facets tend to become parallel. These functional changes are complicated by changes in the clavicular-stern- al articulation, but the end result in the ad- vanced burrowing moles is to create a nearly straight bony bridge from humerus to humerus. SCAPULA: The scapula in the Talpine is very narrow with the glenoid fossa nearly perpen- 6 Kistin (1929) figures Scalopus and Galemys pectoral girdles in ventral view which dramatically illustrates the effect of a lengthening manubrium and shortening of the clavicles. He states " . .. in the figure shown here, which was drawn from a dried specimen, the clavicle points more anteriorly than it should" (p. 313). In my opinion the angle of the clavicle as he illustrated it is normal. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 23 A dp hf vp vn B Figure 8. Right clavicle, A; Neurotrichus gibbsii, posterior and ventral views, B; Urotrichus (Dy· mecodon) pilirostris, posterior view. I-length; 2-length of ventral process ; dp-dorsal promi· nence; hf-humeral articular facet; mf-manu· brial articular facet; VD-vascular notch; vp- ventral process; vs-ventromedial process or spine. dicular to its long axis. In the Urotrichini the origins of the M. supraspinatus, M. infraspina- tus, and M. teres major are separated on the dorsal side by ridges of hone. The infraspina- tus fossa thus appears as a deep groove or chan- nel extending nearly the length of the scapula. The acromion process has a moderately well developed metacromion region ( see C. A. Reed, p. 537, for a correction of Campbell's, 1939, Fig. 25). The supra-scapular nerve passes under the acromion process in the usual manner and no foramen is present. Scaptonyx closely resembles the Urotrichini in the scapula but according to Campbell's Figure 24 lacks the metacromion process and has a more restricted supraspinatus fossa. The scapula of Condylura also resembles that in the Urotrichini in major features but is more angulate. In the Scalopini, "A great development of the teres major fossa and the joining of the up- raised dorsal lip of the posterior margin with the vertebral end of the spine has partly cov- ered over the end of the infra!lpinatus fossa and deepened it in the vertebral half of the bone. The supraspinatus fossa is narrow and defined only on the proximal half of the scapula. The acromion is degenerate as in the Talpinae [ my Talpini]. The tuber though sma11 is better rep- Figure 9.-Dorsal view of the left scapula of Sea· panulus oweni. ab-axillary border; ap-acro- mion process; gf-glenoid fossa ;itp-path of in- fraspinatus tendon; rf-rhomboid fossa; s- "'spine" of scapula; sh-superior border; sf- supraspinatus fossa ; snf-foramen for suprascap- ular nerve; t--tuber of the spine (broken); tf- teres fossa; vp-vertebral border. 24 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 resented than in the ... " (Campbell, 1939) Talpini. The base of the acromion process is pierced by the supra-scapular nerve. There is a rhomboid fossa between the supraspinatus fossa and the vertebral end. The scapula of Scapanulus (Fig. 9) is a blend of features seen in Parascalops and Sca- panus. The vertebral border is angulate as in Parascalops. The rhomboid fossa is separated from the supraspinatus fossa by a low ridge. The walls of the infraspinatus fossa converge anteriorly and thus eliminate tbe fossa as in Scapanus and Scalopzts before it reaches the acromion process. In other respects the scapula of Scapanulus compares favorably with that of Parascalops. In the Talpini the infraspinatus fossa is en- tirely absent. The supraspinatus fossa is corres- pondingly large. The equally large teres fossa is partially divided into two small fossae near the verteb'ral end. The acromion process is even smaller than in the Scalopini and forms a low triangular ridge which has neither a suprascap- ular notch nor foramen. There is no rhomboid fossa in the Talpini. HUMERUS: The humeri of the U rotrichini and Scaptonyx (not seen-Campbell, 1939, Figs. 35, 41) are said by Campbell ( 1939) to form a closely knit group which differs from the Des- maninae in a number of details. He lists: ( 1) clavicular facet on the greater tuberosity is en- larged to a round convex area; (2) the head has become even more elliptical; (3) the pec- toral crest has moved over to the medial side of the bone nearly in line with the teres tubercle; ( 4) the passage of the biceps tendon between the pectoral crest and the teres tubercle has formed a bicipital notch; ( 5) the teres tubercle forms a long high crest rather than a pointed prominence; ( 6) a small brachialis fossa is apparent beneath the greater tuberosity; (7) the greater development of the lateral epicon- dyle; (8) and the humerus is relatively wider. Scaptonyx differs according to Campbell from the Urotrichini in the development of a spike- like deltoid process and larger olecranon fossa. The humerus of U rotrichus ( Fig. 1 OA) close- ly resembles that of Neztrotrichus but is rela- tively narrower. The teres tubercle is shorter than in Neurotrichus. Posteriorly the ridge ex- tending from the lesser tuberosity to beneath the head is sharply angled. Anteriorly the ole- cranon fossa is more extensive than in Neuro- trichus. The long axis of the head with respect to the shaft points laterodistally as pointed out by Reed (1951, p. 154) for Neztrotrichus. With reference to Condylura, Campbell (1939, Figs. 37, 43) expressed his own and the observations of others by stating that "it is probably safe to say that the humerus of Condylura (Fig. lOB) shows greater fossorial specialization ... "than the preceding groups. Condylura resembles the preceding tribes in (1) small brachialis fossa, (2) small clavic- ular articular surface, ( 3) lateral inclination of the head, ( 4) and in comparison with the Urotrichini a distinct notch between the head and the clavicular facet and relatively well developed deltoid and epicondylar processes. However, Condylura also resembles the Tal- pini and Scalopini in ( 1) the division of the pectoral muscle scar into two distinct regions, the pectoral crest and pectoral ridge, by a right angle bend, ( 2) large olecranon fossa, and ( 3) relative width of proximal end. The humeri of the Talpini (Campbell, 1939, Figs. 50-55) and Scalopini (Fig. 10 D-F) are structurally and adaptively very similar and may be contrasted with the preceding by, ( 1) the relatively broader proximal ends, ( 2) heads parallel or medially directed with re- spect to the long axis of the humeri, ( 3) large, hemicylindrical clavicular facets not sharply separated from the bases of the heads, ( 4) large brachialis fossae, and ( 5) very well de- veloped epicondylar and deltoid processes. Campbell ( 1939) distinguished the otherwise similar Scalopini from the Talpini by ( 1) the presence of "a sharp ridge running from the medial side of the scapular facet to the distal end of the lesser tuberosity" accompanied by an abrupt change of plane between the two areas divided by the ridge, and (2) the more -------~- . - - ~- 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 25 A B c D E F Figure 10. Posterior views of right humeri. Comparison by deformed coordinate method with Scalopoides isodens as the standard, A; Urotrichus pilirostris, B; Condylura cristata, C; Scalopoides isodens, D; Sca- panulus oweni, E; Parascalops breweri, F; Scapanus sp., UO 4585 (reversed). pronounced notch between the head and the proximal edge of the lesser tuberosity. These differences Campbell correlates with the great- er development of the brachialis fossa in the Talpini. Scapanulus in part fills the gap in a struc- tural series between Condylnra and Parasca- lops. Although the humerus of Scapanulus (Fig. 11) resembles that of Condylura in gross proportions, the details of the humerus are more like those of Parascalops. The long axis of the head with respect to the shaft points slightly mediodistally. The deep notch between the trochlea and the fossa for the M. flexor digi- torum profundus ligament in Condylura is eliminated by the broadening of the trochlea as in Parascalops, the Scalopina, and the Tal- pini. The teres tubercle is relatively longer than 26 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON· No.11 I hg ··-...· .... I s I - 1-zi-- ,. \. Jf I . ~ef enp ~ -'> .. . I Figure 11. Left humerus of Scapanulus oweni, anterior and posterior views. 1-length; 2-proximal width; 3-distal width; 4-shaft width (least) ; bf-brachialis fossa; bg-bicipital groove and tunnel ; C-capitulum; cf-clavicular articular facet; dp-deltoid process; ecp-ectepicondylar process; ef- entepicondylar for amen; enp-entepicondylar process; ffd-fossa of M. flex or digitorum ligament ; gt-greater tuberosity; h-head; It-lesser tuberosity ; lf-olecranon fossa; pc-pectoral crest; pr- pectoral ridge; pt-pectoral tubercle; s-shaft; sr-"scalopine" ridge; t-trochlea; tt-teres tubercle. that of Parascalops. The angle formed by the plane of the clavicular facet of Scapanulus and the axis of the shaft is as in Parascalops and the other highly fossorial moles. The plane of the clavicular facet is not parallel to the shaft as in Condylura. The shape of the clavicular facet is also of the Parascalops type (Fig. 42B). Posteriorly there is a distinct ridge extending from the lesser tuberosity to the head. The notch between the head and the lesser tuberos- i ty is well developed. ULNA: Reed (1951) discussed the major differ- ences between the foreleg of soricids and tal- pids as based upon their structure in Sorex, Neurotrichus, and Scapanus. I shall attempt to fill in a description of some of the more strik- ing osteological features of additional genera and will review some of the features mentioned by Reed. The distal elements of the foreleg seem to remain more plastic to change than does the humerus. Three major and progressive structural dif- ferences may be observed within the Talpinae if the ulnae of moles are arranged in order of fossorial ability ( Fig. 13). The manus rotates nearly 90° with respect to the proximal end of the ulna in such a progression. Secondly the ulna progresses from a sinusoidal but rela- tively delicatelv built bone in the Urotrichini J • and Condylura to straight but very short-shafted and robust bone in the Scalopini and Talpini. It is interesting to note that the length of the proximal crest of the ulna is relatively as great in Blarina as in Scalopus. the transversely broadened olecranon process may be consid- ered as a general feature of living soricoids with a few notable exceptions. In soricids (Fig. 13A) the ulna is relatively long and straight. The olecranon process occu- 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 27 tm I ·- -S Figure 12. Left ulna of T alpa europea, lateral and anterior views. af-abductor fossa (M. abductor p. 1.) ; as-abductor scar or tubercle; rbs-bra- chialis scar; cf-cuneiform articular facet; cp- coronoid process; hf-humeral articular facet; le-lateral olecranon crest; If-lunar articular facet; me-medial olecranon crest; pa-process us anconaeus; pc-posterior crest; pxc-proximal crest; rf-radial articular facet; s-shaft; sn- semilunar notch; sp-styloid process; tm-tri- ceps area of insertion; tp-terminal process. pies a proportionately small amount of the length of the ulna. The triceps scar is restricted to the medial side of the olecranon, and is ellip- tical and perpendicular to the shaft. The abduc- tor fossa is very narrow and the posterior crest is correspondingly thick and close to the semi- lunar notch. The laterial olecranon crest is short but well defined. The medial olecranon crest is very strong and level with the poorly defined processus anconeus, and the similunar notch is broad. The distal articular surface forms a slight oblique angle with the hand. The ulna of the Urotrichini (Fig. 13B) is relatively long and sinusoidal in lateral aspect. The olecranon process is proportionately as short or shorter than in the shrews. The triceps scar is transversely elliptical and occupies only the lateral third of the proximal crest. The proximal crest forms a less obtuse angle with the shaft than in the shrews. The apparent ex- pansion of the proximal crest over that in the shrews is partly illusionary owing to the marked narrowing of the anterior platform of the olecranon process. The abductor fossa is greatly expanded over that of the shrews in re- sponse to the more posterior position of the posterior crest and a distinct abductor tubercle is present just posterior to the processus ancon- eus on the lateral side. The lateral and medial olecranon crests are well defined and close to- gether. A distinct processus anconeus is pre- sent. The semilunar notch is more restricted than in the shrews. The distal articular surface forms a silghtly more obtuse angle with the hand, but the area of articulation with the hand is almost as small as in the shrews. In Condylura (Fig. 13C) the ulna is still relatively long and sinusoidal, and the olecra- non process is still relatively short. The tri- ceps scar is poorly defined and appears to be limited to the internal half of the proximal crest; the scar is very narrow. The proximal crest more nearly aproaches a right angle to the shaft than in the Urotrichini. The relative expansion of the proximal crest is greater than in any of the other talpines. The abductor fossa and posterior crest are about as in Neu- rotrichus. A poorly defined abductor tubercle or scar is present on the base of the lateral ole- cranon crest. The lateral and medial olecranon crests are well defined. The processus anconeus is distinct and elevated; but the coronoid pro- cess is exceeding! y low. The semi lunar notch is correspondingly very broad and ill defined. The distal articular surface has rotated to nearly a right angle with the hand; and the ar- ticular area with the hand is noticeably ex- panded over that of the Urotrichini. The ulna of the Scalopina and Parascalops (Fig. 13E) bear many features in common and in part may be discussed together. The ulna is relatively short, straight, and quite robust. The proximal crest forms a sharp angle with the 28 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 .(] le c;?1 &l m((i pc I A B c E Figure 13. Left ulnae, cross section of olecranon process (top) , lateral (center) , and anterior (bottom) views. A; Blarina brevicauda, B; Neurotrichus gibbsii, C; Condylura cristata, D; Parascalops breweri, E; Scalopus aquaticus, F; Scapanus townsendii. The wedges on the bottom row indicate the area of the cross section. Abbreviations-see Fig. 12. shaft. The abductor fossa is greatly enlarged and the proximal crest forms a large blade widely separated from the semilunar notch. The semilunar notch is well defined and may form an arch slightly greater than semicircle. Largely responsible for this is the greatly en- larged and elevated processus anconeus and a smaller but distinct coronoid process. The dis- 1968 HUTCHISON: FOSSIL TALPJDAE FROM OREGON 29 tal articular surface forms nearly a right angle with the hand and is greatly expanded-pri- marily posteriorly. Most of the noticeable differences between the ulnae of these genera are centered around the olecranon process. Scalopus (Fig. 13F) may be immediately distinguished from the other two genera and indeed all other talpines by the lack of or suppression of the medial ole- cranon crest. As in N eurotrichus the triceps scar occupies most or all of the central and medial positions of the proximal crest. In Para- scalops and Scapanus the scar is large but it is relatively narrow in Scalopus. The abductor scar is elongate and forms part of the base of the lateral crest in Parascalops but it is ovoid and lies posterior and internal to the lateral lip of the humeral articular facet in the Scalo- pma. The ulnae of Talpa (Fig. 12) and Scapto- chirus7 are like those of Parascalops and Scap- anus in general features but differ in detail. The triceps scar and medial and lateral ridges of the olecranon are similar to those of Scap- anus. However, the articular surface for the humerus and radius of the Talpini is dis- placed laterally so that the radial articular process overhangs the abductor fossa to a greater extent than in the highly fossorial North American moles. The shape of the sty- loid process in the Talpini differs from that of the Scalopini in shape and relative shortness. The cuneiform articulation is not a long cylin- drical articular surface with a small terminal process as in the Scalopini but is formed by a rather short conical to cylindrical internal seg- ment and an equally strong terminal process. The ridge proximal to the cuneiform articular surface is simple in Scalopus and Scapanus but somewhat swollen terminally in Parascalops. In the Talpini this region is developed into a distinct knob. RADIUS: The radius of talpids is not as complex as the ulna and there are correspondingly ; Comments on the skeleton of Seaptochirus other than those on the skull and humerus are based on illustrations hy Young (1934). fewer outstanding differences between the vari- ous genera. The most striking differences, re- lating to the overall length and robustness · of the radius, follow the same pattern as the shaft of the ulna. The radius in all the moles is rela- tively straight, although there is a gentle broad arching of the radius in the shrews examined. The broadening and torsion of the distal end also follows the pattern seen in the ulna. Reed ( 1951) noted the significance of the unique development of the capitular process of the radius in talpids. In the Urotrichini a distinct crest extends distally from the posterodistal margin of the capitular process of the ulna (Fig. ISA). This crest is lacking in all the other Recent moles examined. The radial head of the M. abductor pollicis longis originates on the lateral side of this crest. In Condylura ( Fig. lSB) the area of origin of this muscle still faces laterally but is on the shaft itself. In all the other more ad- vanced moles, the area of origin forms a shal- low posterior facing fossa or flattened area be- tween the ulna facet and the midline of the shaft. In Scapanus (Fig. lSC), Talpa, Scalopus, and Parascalops (Fig. 14) the lateral surface of the radius is grooved where the tendon of the M. abductor pollicis crosses it. The scars of the retaining ligament for the M. extensor carpi radialis are also well developed in these genera. Neither the grooves nor the scars are developed in Urotrichus, Condylura, Neu- rotrichus, or in the shrews. The angle of the groove for the tendon of the M. · abductor pol- licis is sharper in relation to the axis of the shaft in Scalopus and Scapanus than in Para- scalops and T alpa. The proximal extremity of the capitular process culminates in a rounded or arched tuberosity proximal to the level of the ulnar articulation in all the moles exam- ined except Parascalops and Talpa. In Para- scalops the process is abruptly terminated at the level of the proximal border of the ulnar articulation. Aside from the distal expansion of the radius as in the ulna, the articular facets for the lunar and scaphoid are better delineated in 30 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 s I 2 srl Figure 14. Left radius of Parascalops breweri, medial, lateral, and posterior views. I-length ; 2-shaft length; cp-capitular process; ga-groove for tendon of M. abductor pollicis longus; gc-glenoid cavity; If-lunar articular facet; raf-fossa for radial head of M. abductor p. 1 ; s-shaft; sf-scaphoid articular facet; srl-scars of retaining ligament for tendon of M. extensor carpi radialis; uf-ulnar articular facet. A B c Figure 15. Left radii, posterior view of proximal end (top) and medial view (bottom). A; Neurotrichus gibbsii, B; Condylura cristata, C; Scapanus townsendii. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 31 the more fossorial forms. This is easily seen in the outline of the distal end in medial view. In the Scalopina, Parascalops and Talpa this out- line is distinctly scalloped whereas in Condy- lura, the Urotrichini and the shrews this outline is only a broad convex arch. The increased defi- nition of the distal articular facets in the more fossorial forms is seemingly correlated with the increasing restriction of the hand to one ( fore and aft) motion for digging. () f c fort l SCAPHOID: I believe that virtually every bone in the carpus of the various talpines is distinc- tive at the generic level and probably at the species level as well in many cases. However, of all the carpals, only the scaphoid was recov- ered frequently in the fossil samples, so that I shall restrict broad comparisons to it. Other carpals will be discussed in the systematic section. I i .·.· ~/_;__) B Figure 16. Left scaphoid of Scapanus townsendii, A; dorsal, B; ventral, C; proximal, D; distal, E; lateral, F; medial. C---central articular facet; fcrt-path of M. flexor carpi radialis tendon; fdpt-path of M. flexor digitorum profundus ligament; I-lunar articular facet; of-os falciformis articular facet ; r- radius articular facet; t-trapezium articular facet. 32 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 A 1 l Figure 17. Left scaphoid, lateral (left) , proximal (center) , and distal (right) views. A; N eurotrichus gib- bsii, B; Condylura cristata, C; Talpa europea. Abbreviations-see Fig. 16. The scaphoid in talpines provides many sig- nificant clues as to the functional adaptions of the hand. A large portion of the great profund- us ligament passes over the scaphoid. This lig- ament is larger in the more fossorial moles. The os falciformis also articulates with the sca- phoid. The divergence and number of distal articular facets is a clue to the increasing breadth of the hand related to digging. The scaphoid of talpines (Fig. 16) is shaped like a crude three dimensional "L" with the base being the segment between the more dis- tal carpals and the radius. The vertical arm of the "L" forms the portion ohhe scaphoid over which the ligament of the M. flexor digitorum profundus rides and to which the os falciformis attaches. The scaphoid of the Urotrichini as exemplified by N eurotrichus ( Fig. 17 A) is rel- atively elongate dorsoventrally. The centrale and trapezium articular facets are simple and broadly joined. The central articular facet is longer ( dorsoventrally) than broad. The sca- phoid of Condylura (Fig. 17B) is more com- pact but the distal articular facets are simple and broadly contiguous. The central facet, however, is as relatively broad as it is in all the Talpini and Scalopini. In the latter two groups (Figs. 16, l 7C) the trapezium facet increases in complexity. Most of the trapezium facet becomes concentrated on a well defined medial process. The trapezium facet in the Talpini is not distinctly notched as in the Scal- opini. The changes in the shape of the centrale and trapezium articular facets reflect both the increase in the breadth of the hand and size of the more distal carpals. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 33 t td u c I A II III IV v B 1 c Figure 18. Right metacarpals, ventral view. A; Parascalops breweri, B; Neurotrichus gibbsii, C; Scapanus townsendii. c--cuneiform articular facet; ID-magnum articular facet; mv-metacarpal V articular facet; p-phalanx articular facet; pp-proximal prominces; t-trapezoid articular facet; td-trapezoid articular facet; u-unciform articular facet. METACARPALS: The metacarpals, like the scaph· oid, offer many clues as to the breadth of the hand and in addition off er direct evidence of the shortening of the hand in response to in- creased fossorial ability. The metacarpals of N eurotrichus ( Fig. 18B) are more robust and notably shortend as compared to those of Urotrichus. The length of the metacarpals is still more or less symmet- rically arranged in decreasing length about the long, third metacarpal. Proximal prominences on the ventral side for interosseous ligaments are poorly differentiated. The lateral moiety of the distal articulation of the first metacarpal is distinctly more distally extended than the medial portion. The metacarpals of Condylura are, on the average, no shorter than in Neurotrichus. They are nearly equal in length but are slightly graded in size from the first to the fifth which is the longest metacarpal. The relatively great length of the fifth metacarpal may again be cor- related with the semiaquatic habits of Condy- lura. The proximal prominences are more strikingly developed than in other talpines. The medial moiety of the distal articulation of the first digit is more distally extended than the lateral moiety. 34 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 The metacarpals of the Talpini, and Scal- opini are extremely shortened, with the breadth exceeding the length in some cases. The longest bone is the first metacarpal with the other meta- carpals being more or less graded in size from the first to the fifth. Prominent proximal pro- cesses are developed for the interosseous liga- ments but they are not the great flaring struc- tures seen in Condylura. Within the Scalopini the medial portion of the distal articulation of the first metacarpal of Parascalops (Fig. 18A) is about equally extended distally as the lat- eral portion. In both Scapanus (Fig. 18C) and Scalopus the medial portion is distinctly more extended distally than is the lateral por- tion thus completely reversing the condition seen in Neurotrichus. This is correlated with the relatively greater development of the med- ial ( and ventral in life) portion of the hand. The phalanges show similar reductions in length and the development of proximal pro- cess as do the metacarpals but I shall not elab- orate on them here. PELVIS: In Urotrichini the first three vertebrae are fused with the ilium. The neural spines of the sacrum are separated by large fenestrae, although the anterior fenestrae may tend to close. The pubi do not approach each other closely anteriorly. The fifth vertebra has well developed transverse processes. In Condylura the first three and perhaps the anteriormost part of the fourth vertebrae fuse with the ilium. The neural spines are separated by moderately large fenestrae. Anteriorly the pubi come to within one millimeter of each other. The posteroventral extremity of the inn- ominate is strongly hooked ventrally as in the Desmaninae, providing increased insertion and leverage of the M. adductor brevis et magnus. The transverse processes of the fifth vertebra are ligamentously joined to the ischium. In Scapanulus the available innominate was severely damaged so the degree of pseudosym- physis was not determinable. The first three vertebrae are fused with the ilium. Slight bridges of bone apparently extended from the fourth and fifth vertebrae to the ischium. The interneural fenestrae are still present but are encroached upon on all margins by thin lam- ellae of bone. In Parascalops and the Scalopina the first three vertebrae fuse with the ilium; and pubi approach within one millimeter of each other anteriorly. In Parascalops the broad transverse processes of the fifth vertebrae do not contact the ischium. Interneural fenestrae are large. The posteroventral extremities of the innomin- ates are hooked as in Condylura but are not so heavily developed. In the Scalopina the fourth and the fifth sacral vertebrae are also fused to the ischium by broad bridges of bone. The in- terneural fenestra are usually represented by only a small fenestra between the fourth and fifth vertebrae. In T alpa europea the pubi are also within a millimeter of contact and the interneural fenestrae are open. As in T. (M.) latouchi the first four vertebrae are fused to the innominate but not the fifth. In Talpa (Magera) the pubi are fused anteriorly and only the posterior in- terneural fenestra remains open. HIND LIMB: The hind limb in general shows fewer striking modifications among the various talpini genera than the fore limb. Reed ( 1951) gave detailed descriptions and figures of the hind limbs of Neurotrichus and Scapanus. For- tuitously these genera represent the two main morphological groups that I was able to dis- tinguish among the genera examined. FEMUR: In the femur of the Talpinae the proxi- mal articular surface, besides extending to- ward the greater trochanter, also fans out pos- teriorly to approach the intertrochanteric notch and terminate near it in a ridge. This ridge usu- ally causes an inflection in the intertrochanteric crest or results in the poor definition of the crest in its midregion. The femur in the Urotrichini is relatively long and delicate. The femur in the Talpini and Scalopini resembles that of the Urotrichini ex- cept that the shaft is shorter and the whole femur is stouter. With one exception, in all Talpinae the condyles of the femur are con- 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 35 vergent distally. In Condylura, however, the long axis of the lateral condyle has rotated so that it is parallel to that of the medial condyle. Elsewhere in the Talpidae, I have only seen this feature within the Desmaninae. TIBIOFIBULA: The tibiofibulae fall into two well defined types. The Urotrichini type is charac- terized by a slender shaft and falciform process expended into a wide flange. Additionally the lateral process of the fibula is slender and bar- like with its dorsal extremity expanded into a hatchet-shaped structure. Condylura con- forms to this type. The Scalopini type is char- acterized by the relatively stout shaft, narrow and hook-like falciform process, and the lateral process of the fibula expanded into a triangular plate just after leaving the body of fibula. Scal- oplls, Scapanus, Talpa, and Parascalops exhib- it this type of tibiofibula. Parascalops retains a slightly expanded dorsal process on the fib- ula which is undoubtedly the homologue of the hatchet-shaped process in Neurotrichus but in overall construction the lateral process is more like that of Scapanus. The elements of the foot were infrequently found in the fossil record of Oregon and thus the descriptive osteology of modern moles is not presented here. 36 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 SYSTEMATICS TALPIDAE Brief osteological diagnoses of the tribes of the Talpinae and subfamilies which include Oregon species are presented below. The other subfamilies are diagnosed in the osteological section. Subfamily UROPSILINAE Dobson, 1883 Zygoma arched dorsally, bulla not expand- ed; manubrium simple; clavicle long and slender with articulations for humerus and sca- pula; humerus narrow, head rounded, open bi- cipital canal, no fossa for ligament of M. flexor digitorum profundus; ulna slender and with- out transversly expanded olecranon process; radius slender, no capitular process; feet elon- gate and shrew-like, ungual phalanges not hifid; pelvic girdle and hind limb shrew-like. MYSTIPTERUS Hall, 1930 TYPE : My stipterus (My stipterus) vespertilio Hall, 1930; INCLUDED SPECIES: Mystipterus(Mydecodon) martini Wilson, 1960. Mystipterus (Mystip- terus) pacificus, n. sp. KNOWN DISTRIBUTION: Middle Miocene ( Hem- ingfordian) of Colorado, Late Miocene ( Bar- stovian) of Oregon, and Early Pliocene (Clar- endonian) of Oregon and Nevada. AMENDED DIAGNOSIS: Small shrew-moles with dental formula; ? ? ? ? and/ or ? ? ? ? . P1 2133 3133 (?) absent, P2 single or double-rooted, Pa-P 4 double-rooted, P2"P4 graded in size, largest posteriorly. Lower molars with moderately brachyodont crowns, anterior, posterior and labial cingula present, entrocristid of M2-M:1 extending to or near to bases of metaconids with concomitant basining of the postfossids. Trigonid cusps of M1 crowded together, pro- toconid very low. Two mental foramina in lower jaw, posterior foramen beneath M1. Pos- terior margins of upper molars concave, P2 double-rooted and larger than P3 • Uropsilus- like humerus. M ystipterus vespertilio Hall, 1930, was erected on the basis of a single worn and slightly damaged lower third molar and asso- ciated dentary fragment. Described originally as a vespertilionid bat by Hall, it was trans- ferred to the Soricidae by Patterson and Mc- Grew in 1939. The characters they used for placement within the Soricidae are character- istic of the Heterosoricinae only. Clark, Daw- son and Wood (1964) pointed out that the fea- tures of the dentary used by Patterson and Mc- Grew were not unique to the Soricidae, and would just as well serve to align M ystipterus with such diverse insectivores as Nyctitherium, Micropternodus or cf. Myloestes. It is indeed curious that these authors did not suggest affin- ities with the Talpidae, which in the light of the geologic age is more reasonable than some of the archaic early Tertiary insectivores. Van Valen ( 1967) independently arrived at align- ment with the Talpidae but was ( understand- ably) in doubt as to its affinities within the group. Additional material of Mystipterus vesper- tilio from Fish Lake Valley, Nevada, was re- cently collected by J. A. Suthard and described in an unpublished Masters thesis. With his kind consent, this material was borrowed and is described below. Suthard noted the similar- ity between Mystipterus vespertilio and Myde- codon martini and considered them as conge- neric although actual specimens of M. martini were not available to him at the time. Wilson ( 1960) noted the similarity of M. martini with the shrew-moles comprising the Uropsilinae and Urotrichina. Although in ref- erence to these groups, he stated that "it is hardly possible in the absence of both upper dentition and the humerus to state to which Mydecodon shows the closest affinities", he favored the Urotrichina on the basis of the construction of the talonid valleys. With both of the above restrictions now removed, at least in part, it is now possible to ascertain with greater certainty the relationships of these two moles as well as the third one described below. . 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 37 Figure 19. Mystipterus vespertilio, Fish Lake Val- ley, UCR loc V6302, UCR 10048, right maxillary fragment with P3-M1, occlusal and labial views. The three species discussed below comprise the first avowed fossil Uropsiline moles either in the Old or New World. Mystipterus vesper- tilio is not assuredly known from Oregon, but its taxonomically important position to the New World uropsilines and relationship to the Oregon uropsilines in particular necessi- tate the description of the new and possibly topotypic material from the Esmeralda For- mation of Nevada. M ystipterus ( M ystipterus) vespertilio Hall, 1930 (Figs. 19 and 20) TYPE: UCMP 29604 from UCMP loc. V-2804. REFERRED MATERIAL: UCR loc. V-6302 site 1. UCR 10048, maxilla fragment with P3-M1; UCR 10249, ectoloph of M1; UCR 10246 and UCR 10248, Mi's; UCR 10044, M2-M3; UCR 10247, M2 and trigonid of M3; UCR 10245,. dentary fragment with fragment of the M3. AMENDED DIAGNOSIS: Small shrew-mole with dental formula; ? ? ? ? . Upper teeth brachyo- ? ? ? 3 dont, P2 double-rooted and larger than P3, M1 transversely broader than Uropsilus, well de- veloped metaconule flange, anteriorly situated metacone and concave posterior margin. Lower molars like Mystipterus martini except with more cylindrical entoconids on M1-M2. DESCRIPTION: Aside from the teeth, the maxilla fragment, UCR 10048, also retains three alv- eoli anterior to the P3 (Fig. 19). The anterior- most alveolus is small and is followed by two larger alveoli apparently for a double-rooted P2. The P3 is small and ovate in occlusal out- line. It is double-rooted. The single cusp is situated just anterior to the center of the tooth and is entirely enclosed by a narrow cingulum. If alveoli are a reliable indication of size, the P2 was larger than the P3. The three-rooted P4 is subtriangular in occlusal outline with a low but prominent and elongate protocone. The principal cusp, the paracone, is centrally situ- ated on the labial moiety. It is connected with the posterior extremity of the tooth by a broad sloping ridge. The cingulum is poorly pre- served in places but appears to have enclosed the entire tooth except for the lingual margin of the protocone. Both premolars are very low crowned. The M1 is also quite brachyodont and subquadrate in occlusal outline except for the typically extended metastylar region. The metaconule is situated anterior to and near the level of the metacone but the posterolingual margin of the metaconule is greatly extended posteriorly as in the soricids. This results in a deeply concave posterior margin on the tooth. The paraconule is only vaguely defined and is little more than a slight thickening of the anterior cingulum which connects the para- style to the protocone. The metastyle shows an incipient twinning. Narrow labial cingula con- nect the mesostyle to the parastyle anteriorly and to the distal portion of the metacone-meta- stylar crest. The posterior cingulum is well defined on the labial moiety of the tooth but becomes poorly defined lingually. Only the small part of the dentary that is associated directly with the molars is pre- served, thus nothing is known of the antemolar dentition. Patterson and McGrew (1939) and Hall (1930) have already mentioned that the masseteric fossa extends downward nearly to the ventral margin of the jaw. There is a men- 38 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 tal foramen below the middle of the trigonid of the M1 • The M1 is very similar to that of M. martini except in the construction of the talonid. The entoconid of M. vespertilio is more cylindrical in cross section than in M. martini, and the postcristid instead of connecting with the tip of the entoconid joins the posterolabial margin of that cusp at the same level as the posterior accessory cusp. The entoconid and postcristid tend to form a continuous wear sur- face in most moles, but this peculiar arrange- ment in M. vespertilio, as in the majority of the soricids, delays the development of a continu- ous pattern until late wear. The specimens of the second molar differ somewhat from each other. UCR 10044 (Fig. 20) strongly resem- bles the corresponding tooth in M. martini ex- cept that the talonid differs, although less strik- ingly, as does the M1 • Additionally the talonid of the M2 is relatively longer than in M. martini. The other specimen, UCR 10247, is somewhat larger ( see Table 3) and the entocristid joins the protolophid a little more labially at the mid- point of the protolophid. The entire M3 is pre- served in the type and UCR 10044. The tip of the metaconid of the type specimen has appar- ently been broken off since the time of Hall's description ( 1930). The U.C. Riverside speci- men compares favorably with the type in both size and morphology. The posterior cingulum tends to be better defined in Mystipterus ves- pertilio than in M. martini. RELATIONSHIPS: As noted by Suthard ( unpub- lished thesis) M ystipterus vespertilio bears a close relationship to M ydecodon martini of Wilson ( 1960) in the construction of the lower molars. Suthard considered the two forms to be congeneric on this basis; this arrangement is tentatively accepted. Wilson (1960) while not- ing a strong similarity between M ydecodon martini and Uropsilus nevertheless believed that M. martini bore a closer relationship to Urotrichus (Dymecodon) pilirostris (True) and Urotrichus talpoides Temminck based on the construction of the lower molar talonids. Judging from the material from other talpids on the whole, this was a good assumption. The Figure 20. Mystipterus vespertilio, Fish Lake Val- ley, UCR loc V6302, UCR 10044, right M2-M3, occlusal and labial views. upper teeth of M ystipterus vespertilio, however, indicate a strong degree of affinity with Urop- silus. The expanded metaconule shelf is unique to Uropsilus (Fig. 2A, SC) within the living talpids. Unfortunately, the lack of preserved antemolar dentition prevents further compari- son between M. martini, M. pacificus, and Urop- silus. Uropsilus and Mystipterus resemble each other in the morphology of the P 4, and overall shape, brachydonty, expanded metaconule shelf, labial cingulum, and concave posterior margin of the M1• These characters in combi- nation are unique to Uropsilus among the liv- ing moles. M. vespertilio differs from Uropsilus in the more brachyodont upper and lower teeth, shorter P4, relatively smaller P3, double- rooted P2, larger P2 than P3, well devel- oped entocristids and more lingual extensions of the cristae obliquae on the lower molars. The latter two characters, those noted by Wil- son (1960) in M. martini, tend to make basins of the talonid crowns. It was also these two features which disuaded him from aligning Mydecodon with Uropsilus. Despite these dif- ferences from Uropsilus, I believe the charac- ter of the M1 to be the single most significant feature upon which to align Uropsilus with Mystipterus. The occurrence, described below, of Mystipterus-Mydecodon lower teeth along with a uropsiline maxillary fragment and un- doubted uropsiline humerus tend to substanti- ate this otherwise tenuous conclusion. TABLE 3 Comparative Measurements of the Lower Dentition of Mystipterus vespertilio, M. pacificus and Related Forms M3 CO P2 P4 M 1 M2 M3 M1-M3 Depth of Specimen Lgth Wdth Lgth Wdth Lgth Wdth Lgth Wdth Lgth Wdth Length Jaw Co © M. pacificus UO 22437 1.79 1.18 1.47 0.92 1.67 UO 22438 (Type) 1.04 0.66 1.64 1.18 1.71 1.11 1.35 0.82 4.55 1.72 UO 22345 1.82 1.16 1.43 1.44 0.91 0.73 1.65 < UO 22346 "*j UO 22350 0.71 0.50 1.02 0.75 O UO 22567 1.60 1.62 1.12 1.19 Co UO 22570 t-i UO 22577 1.73 1.18 *~3 M. cf. M. pacificus to UO 24336 1.50 0.99 1*1 USNM 23771 1.60 1.13 1.67 1.17 1.40 1.00 4.80 1.67 "*1 USNM 23770 1.51 1.17 1.93 So o Mystipterus sp. © UO 24786 1.70 1.20 1.82 1*3t*1 UO 24787 1.60 © © M. vespertilio UCR 10246 1.53 0.93 UCR 10248 1.33 1.00 UCR 10247 1.57 0.97 UCR 10044 1.50 0.93 1.33 0.80 UO 29604 (Type) 1.18 0.77 40 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 Mystipterus (Mystipterus) pacificus, 8 n. sp. (Figs. 21-27) TYPE: UO 22438, an incomplete left dentary with P4-Ms, roots of Ps and the posterior root of the P2. HYPODIGM: UO 22437, UO 22345, UO 22346, UO 22350, UO 22601, fragmentary jaws and teeth; UO 22567, UO 22570, UO 22577, iso- lated Mi's; UO 22510, maxillary fragment with left M2; UO 22309, right humerus lack- ing distal extremity; UO 26801, proximal ex- tremity of left humerus. All specimens from the type locality. REFERRED SPECIMENS: uo 22418, maxillary fragment with P2-P4 ; UO 22516, DP.4 Both from type locality. TYPE LOCALITY: uo loc. 2465. Quartz Basin. AGE: Barstovian. DIAGNOSIS: Differs from the other species of the genus in larger size, relatively longer antemo- lar region, additional antemolar tooth, P2 dou- ble-rooted, anterior mental foramen beneath P2, paralophid more transversely aligned. Met- aconule of M2 posteriorly situated. P4 rela- tively more massive than in M. martini and there is no pronounced expansion of the heel. Lower molars similar to those of M. (Myde- codon) martini except with more cylindrical entoconids. DESCRIPTION: A single upper second molar, UO 22510, seems to be referable to M. pacificus on the basis of size and occlusal comparability with the lower molars of this species. It is too large for association with Achlyoscapter, n. gen., and too small for Domninoides. Since upper molars ascribed to Scalopoides with a high degree of certainty are also known from the same locality, this leaves Mystipterus the probable associate. The tooth (Fig. 21) is roughly rectangular in occlusal outline being wider than long. The cusps are heavily worn and the metastyle is broken off. The ectoloph is W-shaped and the mesostyle may have been twinned but, if so, this is obscured by wear. 8 From the Latin paci/icus, alluding to its Pacific Coast occurrence. Figure 21. Mystipterus pacificus n. sp. Quartz Basin, UO loc 2465, UO 22510, left M1 and maxil- lary fragment, labial, occlusal, and dorsal views. The paraconule, protocone, and metaconule are united in wear hut all are easily distinguish- able. The protocone is the largest of the labial cusps. The paraconule is about half the size of the protocone and separated from it by a deep lingual furrow. The metaconule is situ- ated posteriorly slightly posterior to the level of the center of the metacone. The base of the metacone is expanded to such an extent that this cusp and the metastyle form the posteriormost extremities of the tooth while contributing to the concave posterior margin of the tooth. A narrow paracingulum extends from the para- conule to the parastyle and a faint metacingu- lum is present posteriorly. The labial root is very broad anteroposteriorly originating from the base of the protocone to the metaconule. The distal extremity of this root is expanded still further into a hatchet-shaped structure. The root is transversely compressed. The M2 is about 1.5 mm. long and 1.94 mm. wide. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 41 Although only a small fragment of the max- illa is adhering to the M2, several significant features are preserved. A blade-like process terminating dorsal to the metastylar region of the M2 extends laterally from the maxilla. The process apparently originated in part anterior to the M2• I believe this process to be homologous to the subzygomatic process of the maxilla in U ropsilzts ( see Fig. 2A). A low crest extends from the posterior margin of the subzygomatic process dorsally to the posterior edge of a dorsally directed process. This latter process is transversely compressed. Its longi- tudinal breadth is not surely determinable on this specimen but it does not appear to have extended anterior to the level of the M2• A sharp ridge of bone extends anterior from the base of the dorsally directed process and forms along with it part of the external wall of the infraorbital canal. Very little of this ridge is preserved and its crest appears to be broken off so that I was not able to determine whether it is simply a ridge or part of the base of the infraorbital bridge. There are two probable interpretations of the above structures. One is that the root of the dorsally directed process is that of the infraorbital bridge as in the Tal- pinae, Desmaninae, and Proscalopinae, and the other that this structure is the base of a dor- sally arched zygomatic arch as in Uropsilus. The latter interpretation seems more likely to me in the light of the following facts: ( 1) there is no other apparent process suggested or evi- dent from the specimen which would give rise to a posteriorly projecting root of the zygomatic process, ( 2) the position and fragility of the root of the alleged zygomatic process is ex- tremely similar in development and position to that of Uropsilus, ( 3) the well defined infra- orbital canal shows no indication of fanning out anteriorly as in the non-uropsiline moles, ( 4) and the short ridge preserved in this speci- men anterior to the zygomatic root suggests that another process, the infraorbital ridge, is present anteriorly. Like the other specimens referred to this species, a fragment of the snout with P2-P4, UO l Figure 22. Mystipterus pacificus n. sp. Quartz Basin, UO loc 2465, UO 22318, right P2-P4, occlu- sal and labial views. 22418, is referred to M. pacifiicus by elimina- tion and by general size agreement. The three teeth are crowded together so that about half of the P3 is concealed in lateral view by the P2 and P4 (Fig. 22). The P3 is smaller than the P2 • The P2 -P3 are simple single cusped teeth. The P2 has a relatively higher and sharper cusp than the P3• The P2 is double-rooted. The P4 is badly damaged by crushing. It has the same general form and morphology as Scalopoides ripafodiator but is smaller. Figure 23. Mystipterus pacificus n. sp. Quartz Basin, UO loc 2465, UO 22516, left dP4, occlusal and labial views. The maxilla fragment (Fig. 23) contammg an upper deciduous premolar resembles the adult P 4 of UO 22418 in general proportions and size. The principal cusp is relatively low, and there is a large well developed protocone lingually. A cingulum encompasses the tooth except for the protocone. The three roots di- verge strongly as is commonly the case in de- ciduous teeth. The apex of an incompletely formed adult P 4 is discernible in the cavity 42 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 Figure 24. Mystipterus -pacificus n. sp. Quartz Basin, UO loc 2465, Type specimen UO 22438, labial and occlusal views. between the roots of the dP4. There is usually a general similarity between the permanent tooth and its deciduous predecessor in talpids in both size and morphology and it is for this reason that UO 22516 is referred to M. pacifi- cus. Little of the dentary (Fig. 24) aside from the midsection of the horizontal ramus is pre- served in the hypodigm. The antemolar region appears to have been moderately long and rel- atively deep. There are two mental foramina, one under the trigonid of the M1 and the other beneath the P2. The dental formula is not im- mediately determinable, although at least three premolars seem to have been present and are double-rooted. In UO 22350 (Fig. 25A) there is a rather large but shallow alveolus inter- preted here as the canine immediately anterior to the P2. The specimen is broken away an- terior to this but there appears to be the rem- nants of at least three more alveoli. The most posterior of these is situated rather labially and is flanked anterolingually by the remains of a second alveolus. Ventrally on the broken anterior surface of this specimen, the posterior extremity of still another alveolus may be seen. The tooth was apparently rather procum- bant and is interpreted as the Ii. These three anterior alveoli are assumed to represent the Ii to Ia. There are several other interpretations for the four anterior alveoli. The two alveo- li in front of the P2 may have held an en- larged P1 as in the Talpini. The P1 may be single-rooted. The P1 may be single or double- rooted and there were additional alveoli not in evidence. However, for the sake of descrip- tion and for subjective reasons, I shall con- sider the dental formula to be 13, C, P3, M3. The P 2 is dominated by a high protoconid encircled by a small cingulum with a very small paraconid. The roots of the P2 are longi- tudinally aligned but compressed anteropos- teriorly as are those of the succeeding pre- molars. The P4 is a larger version of the P2. The protoconid is very sharp. There is no pro- nounced expansion of the posterior cingulum into a heel; although the most posterior part of the cingulum is somewhat cuspidate. A dis- tinct groove descends the posterior slope of the protoconid which in M. martini is either absent or effaced by wear or breakage. The M1 is rather wedge-shaped with the trigonid shorter and narrower than the talonid. There are narrow anterior, labial and poster- ior cingula. The anterior cingulum is incom- plete lingually. There is a prominent postero- 1968 HUTCHISON: FOSSIL T ALPIDAE FROM OREGON 43 j B Figure 25. Mystipterus pacificus n. sp. Quartz Basin, UO loc 2465, A; UO 22350, left dentary fragment with P 2 and P 4, labial and occlusal views, dashed outlines of roots restored from X-ray photograph, B; UO 22577, unworn right M1, anterior view. internal accessory cuspule. The protoconid is the tallest cusp followed by the equally tall hy- poconid, entoconid and metaconid. The para- conid is exceedingly low (Fig. 25B), small and more transversely aligned than in M. martini or M. vespertilio. The crista obliqua intersects the protolophid about midway along its length. There is no metastylid but a low distinct ento- cristid extends to the base of the metaconid thus enclosing the postfossid. The labial cingula are interrupted by the hypoconid in the M2-Ma. The M2 is more rectangular than the M1 and the trigonid and talonid are about equal in length and breadth. The morphology of the M~ of M. pacificu,s closely resembles that of M. martini and M. vespertilio except that the trigonid is somewhat more compressed antero- posteriorly and the anterior cingulum is not as relatively large. The entrocristid abuts against the base of the metastylid which is more or less fused with the metaconid. The MH of M. pacificus is not significantly differ- ent from that of the other species. ' The two incomplete humeri ( Fig. 26) are referable to the Uropsilinae. The Quartz Basin specimens are a little longer and somewhat more robust than the humerus of U ropsilus but otherwise correspond in morphology to nearly the finest detail. A I ~ B Figure 26. Mystipterus pacificus n. sp. Quartz Basin, UO loc 2465, A; UO 26801, left proximal portion of an uncrushed humerus, anterior and posterior views, B; UO 22309, right humerus with proximal end fractured and rotated postero-dis- tally, anterior and posterior views. 44 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 4 ' { ' \ Figure 27. Mystipterus pacificus n. sp. Quartz Basin, UO loc 2465, UO 22399, right femur, posterior view. The only well preserved femur (Fig. 27) from Quartz Basin (UO 22399) more closely agrees with that of Uropsilus than with the Talpinae. The condyles of the distal epiphysis, and the tips of the greater and lesser trochan- ter are broken away but their position is well marked. The femur is longer and relatively more slender than any of the talpids aside from Uropsilus. Like Uropsilus, ( 1) the inter- trochanteric crest is smoothly curved and well defined, ( 2) the articular surface of the head does not extend posteriorly toward the inter- trochanteric crest, and ( 3) the third trochanter is short. The third trochanter is more nearly on a level with the lesser trochanter thus mak- ing the proximal end somewhat more compact as in the Talpinae. It differs from the femur of soricids in the well defined trochanters and broader proximal end. RELATIONSHIPS: The problem of whether all the elements here described belong to one spe- cies cannot be irrevocably solved until either an articulated skeleton or a number of large samples become available. The similarity of the maxilla and humerus to Uropsilus indicate that these elements, at least, belong to the same species. The femur is likewise assigned to M. pacificus on a similar basis although on less satisfactory grounds. This leaves only the lower jaw (the type specimen) in serious doubt. Scalopoides may be easily eliminated on the basis of the sampling which indicates association with other limb bones and differ- ence in gross size. Achlyoscapter longirostris n. gen. et. sp. is thus the last possibility for association with the maxilla and limb bones. However, the P 4-M1 of Achlyoscapter is not easily reconcilable with M2 assigned to M. pacificus. The occlusal relationships of the upper teeth assigned to A. longirostris agree very well with the type jaw of Achlyoscapter but not with that of M. pacificus. The M1 of A. longirostris and the M2 of M. pacificus do not compare well serially as one would expect if they represented the same species. The M2 assigned to M. pacificus occludes well with lower teeth of this species. Lastly, the occur- rence of M ystipterus lowers with uropsiline upper teeth in the two Clarendonian localities tends to support my conclusion that Mystip- terus is a uropsiline mole. The referred humerus alone is sufficient to situate M. pacificus within the Uropsilinae. The possession of a submaxillary process, dorsally directed zygomatic process and con- cave posterior margin of the M1 which are otherwise unique to Uropsilus offer further confirmation of uropsiline relationships. The presence of three double-rooted premolars in the lower jaw as compared to only two in the Uropsilus examined is nearer the theoretic- ally more primitive condition than either M. martini or Uropsilus. The presence of rather transversely elongate upper molars may like- wise represent something nearer to the primi- tive condition. The development of the trans- versely shorter M1 of M. vespertilio represents a further step toward the condition in Uropsi- lus. M. pacificus differs in the lower jaw from M. martini and Uropsilus in its probable greater number of teeth, presence of three double- rooted premolars, and relatively longer and deeper antemolar region. It is further dis- 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 45 tinguishable from Uropsilus by the converg- ence of the entocristid and crista obliqua on the metastylid, transversely longer M2 with a more posteriorly situated metaconule, and bladelike suhzygomatic process. M. pacificus is distinguishable from both M. vespertilio and M. martini by its larger size and more con- vergent entocristid and crista obliqua on the lower molars. On the basis of serial analogy the M2 of M. pacificus would seem to differ from that of M. vespertilio ( as deduced from the M1) in its relatively greater width and more posterior metaconule and concomitantly less developed metaconule flange. If the maxil- lary fragment, UO 22418, is properly assigned to M. pacificus, additional differences between M. pacificus and M. vespertilio include higher crowned teeth and reduced cingula on the P2-P3 • M. pacificus, while not the earliest record of the uropsiline moles, appears to be the most primitive on the whole. A similar situation with regard to the Quartz Basin moles is exem- plified by Scalopoides ripafodiator n. sp. de- scribed below. Mystipterus cf. M. (Mystipterus) pacific us Specimens from four Barstovian localities are referable to M. cf. M. pacificus and are described below. A single M1, UO 24336, from Red Basin loc. 2495 is indistinguishable from that of M. martini except for its slightly larger size. A jaw fragment containing the M1-Ms, USNM 23771, from Guano Ranch USGS loc. M 1042 is the size of the Quartz Basin species but has the molar morphology and relative molar size of M. martini. Another jaw frag- ment, USNM 23770, from the same locality retains a well worn M1 and has a deeper jaw than USNM 23771, but this is probably an allometric feature. The variation noted in the molar patterns and sizes (Table 3) of the specimens of M ystipterus from Quartz Basin, Red Basin, and Guano Ranch, could well be that encompassed within a single species. The lack of sufficient material from the last three localities precludes any such definite assignment. The upper dentition, variation, and morphology of the antemolar region remains the best known criterion for distinguishing the species of Mystipterus. Size ( used broadly), geologic age, and geographic distribution argue for tentative assignment of eleven of these specimens to M. cf. pacificus until better material demonstrates otherwise. Figure 28. Mystipterus sp. Black Butte II, UO loc 2500, UO 24780, left M2, occlusal and lingual views. Mystipterus (Mystipterus) sp. (Fig. 28) An upper M2, UO 24 780, very similar to the one referred to M. pacificus was recovered from Black Butte, UO loc. 2500. The antero- labial extremity is broken away but the tooth is little worn and therefore has some features not directly observable in the M. pacificus specimen. The pronounced difference between the two specimens is that the lingual sulcus which separates the paraconule and protocone is much deeper in the Black Butte specimen. The paraconule, protocone, and metaconule are distinct. The tooth is lower crowned than Scalopoides or N eurotrichus but in agreement with Uropsilus. The base of the metaconule has the same extended base as in the Quartz Basin specimen and has a concave posterior margin. The parastyle is only minutely de- veloped hut a narrow paracingulum is present. The mesostyle shows no particular indication of division. The lingual root is similar to 46 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 the Quartz Basin specimen although somewhat twisted. A small papiliform root somewhat labial to the center of the tooth is also pre- served. The metacone crests of the ectoloph are equal in length and both shorter than the para- cone crests of the ectoloph as in Uropsilus. The morphology of the paraconule is nearly identical to that of the Uropsilus pilirostris specimen available for comparison. The M2 is 1.80 mm. long and 1. 99 mm. wide. Three specimens, an M1 (UO 24786), talo- nid of an M2 (UO 25183), and a jaw fragment (UO 24787) referable to Mystipterus were re- covered from the Black Butte UO locality 2500. The size and morphology of the teeth fall within the range of variation of those re- ferred to Mystipterus cf. M. pacificus. The jaw fragment (UO 24787) retaining all the alveoli between the anterior root of the M2 and the anterior root of the P2, closely resembles that of M. pacificus. The antemolar portion shows little taper anteriorly, the alveoli are not so crowded together anteroposteriorly, and the antemolar length may have been relatively longer than that of M. pacificus. Again the serial comparison of the M2 of Mystipterus sp. with the M2 of M. vespertilio tends to distinguish these species. These speci- mens and M. vespertilio constitute the latest known record of the genus. Summary of Mystipterus relationships Wilson ( 1960) compared M ystipterus (My- decodon) with nearly all the reasonably well known and possibly related genera from Eu- rope. I, like he, have been handicapped in com- paring the North American material with the European by the lack of sufficiently detailed descriptions and figures or the lack of suitable comparative material. Mystipterus may be dis- tinguished from all the described European small talpids where upper molars or humeri are known or are nominally associated. No assuredly uropsiline moles are yet known as fossils from Europe, but it is possible that some material has been misassigned as in the American genera. Aside from the lower molars and possibly a few upper premolars, M. (Mystipterus) ves- pertilio and M. (Mystipterus) pacificus are not comparable by direct means. M. vespertilio appears to be specializing in the direction of Uropsilus in the upper dentition and the lower dentition of M. martini has already achieved the antemolar reduction characteristic of Urop- silus. The construction of the lower molars alone serves to distinguish Mystipterus vesper- tilio from Uropsilus. The development of the more open talonid of Uropsilus is not an insur- mountable barrier to the interpretation that Uropsilus is a descendant of Mystipterus, even if only American forms are considered. All the species here assigned to the genus Mystipterus are so referred on the basis of the similar morphology of the lower molars. To assign them all to the same genus without recog- nition of subgenera would tend to obscure the marked differences between antemolar regions of M. (Mydecodon) martini and M. (Mystip- terus) pacificus. M. vespertilio is tentatively aligned with M. pacificus under the subgenus Mystipterus but the largely uncomparable material of M. ves- pertilio make this arrangement largely one of convenience. The more primitively generalized and less reduced antemolar dentition of M. pacificus would tend to remove it as a direct descendant of M. martini but the sharing of at least an early Miocene common ancestor seems prob- able. The phyletic position of M. vespertilio must remain in doubt until the acquisition of more material. Subfamily GAILLARDINAE, new INCLUDED GENERA: Gaillardia Matthew, 1932. KNOWN DISTRIBUTION: Middle Pliocene (Hem- phillian) of Oregon and Nebraska. DIAGNOSIS: ( Based on Gaillardia) : An aquatic desman-like mole; lower molars with antero- posteriorly compressed trigonids and talonids, wide separation of the metaconids and meta- stylids; dentary with small coronoid and angu- ----~~-- 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 47 lar processes, angular process buttressed by anterior ascending ridge of bone; clavicle long but moderately robust; humerus with open bicipital canal, no fossa for the ligament of M. flexor digitorurn profundus, ovate head, lesser tuberosity lower than head, deltoid proc- ess separated from greater trochanter, prom- inent teres tubercle; femur with third and greater trochanters fused together, intertro- chanteric crest extends laterally to third tro- chanter, medial epicondyle spikelike, lateral epicondylar sesamoid fused to epicondyle; distal end of tibiofibula with two anterior and three posterior tendonal grooves; calcaneum and cuboid relatively short and compact; meta- carpals long and twisted. GAILLARD/A Matthew, 1932 GENOTYPE: Gaillardia thomsoni Matthew, 1932. KNOWN DISTRIBUTION: Same as for subfamily. DIAGNOSIS: Sarne as for subfamily. Gaillardia thomsoni Matthew, 1932 (Fig. 29-38) Hydroscapheus americanus Shotwell, 1956, p. 724. TYPE: AMNH 20508, right lower jaw with p3.p4 and M3 preserved. REFERRED SPECIMENS: UQ 9415, UQ 26827, two edentulous mandible fragments; UO 24810 M1; UO 25125, M2; UO 24815; trigonid of a lower molar; UO 26824 and UO 26825, two upper premolars; UO 3206, clavicle; UO 3185, uo 3505, uo 4101, uo 24811, uo 26829, three complete and two partial humeri; UO 26830, proximal portion of an ulna lacking the end of the olecranon process; UO 26831, third metacarpal; UO 4102, damaged sacrum; uo 2412, uo 2443, uo 2888, uo 2902, uo 3682, UO 3814, caudal vertebrae; UO 2698, UO 3200, UO 10381, UO 25125, two complete and two partial femora; UO 2456, UO 3792, UO 24808, UO 24809, UO 26832, distal ends of tibiofibulae; UO 3815, midshaft of tibio- fibula; UO 3376, UO 3414, UO 24807, UO 26833, calcanea; UO 26834, cuboid; UO 25127, UO 26836, two fourth metatarsals; UO 2400, UO 2415, UO 26837, distal ends of metatarsals; all of the above from UO locality 2222. Subsequent to Shotwell's (1956) original collections and description of H ydroscapheus americanus, additional specimens were col- lected among which is an edentulous incom- plete mandible, UO 9415. The specimen (Fig. 29) lacks the portion of the jaw anterior to the p3 and the extremities of the coronoid process, condyle, and angular process. The remaining portion of the dentary is essentially identical to that of Gaillardia thomsoni Matthew from the Upper Snake Creek beds of Nebraska. The two forms here are considered as conspecific. In comparison to Desrnaninae (see Schreu- der, 1940) the lower jaw of Gaillardia thom- soni exhibits a number of differences. The whole ascending rarnus slants posterodorsally rather than directly dorsal. The coronoid pro- cess is relatively shorter and much more slender and delicate. Conversely, the condylar process is longer and relatively stronger. The condylus is small, slightly ovate, and not transversely expanded. The angular process is dorsoven- trally compressed, transversely expanded, and relatively smaller and shorter than in the Des- maninae. The posterior mental forarnen is below and a little posterior to the posterior root of the M1. The anterior mental forarnen is below the alveolus of the canine (P1 of Matthew). On the labial side a ridge of bone extends from the angular process to a region somewhat above the angle of the horizontal Figure 29. Gaillardia thomsoni, McKay Reservoir, UO loc 2222, UO 9415, left mandible, left lateral view. 48 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON TABLE 4 MEASUREMENTS OF THE DENTARY OF Gaillardia thomsoni Length P1-M3 alveoli .................... .. ..................... ........... ... . Length M1-M3 alveoli ............ ....... ........ .............. .. .... ........ .. . Height of ramus (internal) behind M1 .... ......... .. ... ... .... .... . Posterior border of jaw to anterior border of ascending ramus ........ ...... ........ ..... ............... ........... ........ . Width of condylus ......... ......... .... .. ...... ....... .. ...... ................ .. . and ascending rami; this ridge is entirely ab- sent in the Desmaninae. The mandibular fora- men is situated in about the same position as in Galemys. The lower dental formula of G. thomsoni is here considered to have been com- plete with all the teeth anterior to the P s single- rooted. The anterior incisor was probably hypertrophied. B Type 12.3 6.8 3.5 4.9 1.4 I uo 9415 7.3 3.8 5.4 No.11 c In the type of Gaillardia the P a is slightly larger than the P4 which is simpler in form than in Desmana. In the Oregon specimen of the M1 ( Fig. 30A), the crista obliqua unites with a metastylid which is not joined to the meta- conid. This feature relates to the greater an- teroposterior compression of the talonid and more spacious hypoflexid. The degree of ex- pression of the major cusps is about the same as in the Desmaninae with the paraconid the smallest and the rest about equally strong. There is a very narrow anterior cingulum and small posterolingual accessory cusp. The M2 (Fig. 30B) from the Oregon sample resembles the M1 except that ( 1) the trigonid is more compressed, ( 2) the anterior cingulum is ab- sent, ( 3) the trigonid is wider than the talonid, ( 4) there is a well developed anterolingual accessory cusp. The Ms of the type of Gaillardia is heavily worn but shows the compressed trigo- nid of the preceding teeth and large hypoflexid. All the molars lack ectostylids. Figure 30. Gaillardia thomsoni, McKay Reservoir, UO loc 2222, A; UO 24810, left M1, occlusal view, B; UO 25125, right M2, occlusal and labial views, C; UO 26825, left upper premolar, lingual view. Of the isolated antemolars referred to this form, two closely resemble the P2 of Galemys in proportions. They have a large principle cusp (Fig. 30C) with anterior and posterior accessory cuspules and narrow internal cingu- lum. Pa p4 M1 M2 Ms TABLE 5 MEASUREMENTS OF THE LOWER TEETH OF Gaillardia thomsoni length width (Type) ------·· ·· ·· · ··- -····----- 1.65 1.06 (Type) --- ----- -------- ---- ----- - 1.60 1.00 (UO 24810) -- ---- ---------· 2.45 1.90 (UO 25125) --------------- - 2.59 1.90 (Type) ---------- ---- -- ------- - 2.24 1.66 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 49 Figure 31. Gaillardia thomsoni, McKay Reservoir, UO loc 2222, UO 3206, left clavicle, posterior and distal views. Shotwell ( 1956) previous! y described most of the postcranial elements of Gaillardia; how- ever, I wish to call attention to additional fea- tures not specifically mentioned by Shotwell, and in contrast to the Desmaninae. The clavicle (Fig. 31) of Gaillardia differs from the Des- maninae in that ( 1) the distal end is as wide or wider than the proximal end, (2) the hu- meral articular surface is relatively smaller, and ( 3) there is a prominent dorsally directed process on th~ distal end. This process is termi- nated by a peculiar callosity, possibly anom- alous, but also possibly the remnant of the scapular artculation which is present in this position in Uropsilus. The humerus (Fig. 32) differs from that of the Desmaninae in a number of features already listed by Shotwell ( see diagnosis of subfamily) and additionally in ( 1) head not separated from greater tuberosity by a deep channel, (2) deltoid process separated from the greater tuberosity by a broad concavity, (3) shape of pectoral process closer to Uropsi- lus, ( 4) area of muscle attachment on teres tubercle strongly oblique to shaft, ( 5) lesser tuberosity lower than head, ( 6) broader distal end, and (7) no deep notch sharply separating trochlea from the ventral margin medially. The ulna fragment ( Fig. 33) preserved few features except those in the immediate vicinity of the semilunar notch. As preserved the ulna resembles that of the Desmaninae in its narrow- ness, undeveloped coronoid process, narrow articular surface, and proximal third bowed Figure 32. Gaillardia thomsoni, McKay Reservoir, UO loc 2222, UO 26829, left humerus, anterior view. TABLE 6 MEASUREMENTS OF THE HUMERI OF Gaillardia thomsoni U04101 Length .......................... ... ............................. 18.6 Width of proximal epiphysis .... .......... .. .. .. .. 5.0 Minimum shaft width ........... ..... .................. 2.9 Width of distal epiphysis .......... ... .. ....... ...... 10.0* e Estimated measurement * Noticeable abrasion or breakage tending lo reduce measurement 0 24811 18.0 5.7 3.0 ll.3e uo 26829 20.1 6.25 3.35 12.9 50 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 Figure 33. GaiUardia thomsoni, McKay Reservoir, UO loc 2222, UO 26830, fragment of the proximal region of the right ulna, anterior and right lateral views. Figure 34. Gaillardia thomsoni, McKay Reservoir, UO loc 2222, UO 2443, caudal vertebra, dorsal, anterior (left) and posterior (right) views. posteriorly. The ulna of Gaillardia differs from the Desmaninae in that the processus anconeus is flush with anterior margin of the olecranon process, the proximal articular surface is nar- rower and more strongly convex transversely, and the abductor fossa is open and shallow. The third metacarpal resembles that of Des- mana in general but is relatively somewhat longer and the distal articular surface is con- vex transversely with only a shallow depres- sion just proximal to it dorsally. In Desmana the distal articular surface is concave distally with a prominent pit proximal to it. The third metacarpal is about 7 .5 mm. long. The fragment of the sacrum is too damaged to do more than note the complete fusion of the three vertebrae preserved. These probably rep- resent the third to the fifth sacral vertebrae and apparently had thin transverse processes such as in Galemys. The caudal vertebrae (Fig. 34) more closely resemble those of Desmana than Galemys except that the prezygopophyses are more horizontally aligned. The femur of Gaillardia ( Fig. 35) while functionally related to those of the Desmaninae differs from them in a number of features as noted by Shotwell and in that the ( 1) liga- mentum teres scar is not surrounded laterally by the articular surface of the head, (2) lesser trochanter is more widely separated from the head, ( 3) third trochanter is fused with greater trochanter to form a large plate, ( 4) inter- trochanteric crest extends to the third rather Figure 35. Gaillardia thomsoni, McKay Reservoir, UO loc 2222, UO 2698, right femur, posterior view. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 51 TABLE 7 MEASUREMENTS OF THE FEMORA OF Gaillardia thomsoni uo 2698 Maximum length -------- ------ ------------------ ------------------------- ----- 19.0 Maximum width of proximal end across trochanters --- ----- -- 8.25 uo 10381 16.6* Minimum width of shaft ------------------ --- ---- ----- --- -- --------------- 3.1 2.6 7.3e Width of distal end ------ --- ---- --- ---- ------------- ------ ------------------- 8. 7 e Estimated measurement than greater trochanter, ( 5) pos1t10n of the M. adductor brevis et magnus is marked by a prominent scar opposite the lesser trochanter on the posterolateral margin of the femur, (6) medial epicondyle is extended into a thick, dor· sally projecting spike, (7) lateral epicondylar sesamoid is fused to epicondyle, and (8) patel- lar surface is broader. As noted by Shotwell ( 1956) the tibia and fibula are fused distally and the posterior side of the distal end of the tibiofibula (Fig. 36) has three rather than two grooves. The large an- terior groove for tendons of the M. extensor digitorum longis and tibialis anticus is divided distally by a prominent tubercle on the medial side of the fibula thus causing the M. extensor digitorum longus tendon to lie entirely on the fibula and more posterior than the groove for the M. extensor digitorum tibialis anticus ten- don. The prominent lateral crural tuberosity is located near the lateral and distal extremity of the lateral malleolus. On the posterior surface the groove for the M. flexor digitorum tibialis and tibialis posticus has retracted posteriorly and the tendons are restrained from slipping anteriorly by a well developed ridge which extends laterally from the shaft. The groove for the tendon of the M. flexor digitorum fibu- laris is separated from the above by a low nar- row ridge for about a third of the length of the tibiofibula shaft. As in the desmans the large deep groove for the peroneus muscle tendons is distinct along the entire length of the combined shafts of the tibiofibula. The peroneus groove is bifurcated distally as in Desmana by a laterally compressed tubercle. The more lateral groove is slightly larger in Figure 36. Gaillardia thomsoni, McKay Reservoir, UO loc 2222, UO 26832, distal portion of right tibiofibula, anterior and posterior views. contrast to that in Desmana. The fragment of the midportion of the tibiofibula resembles Desmana in proportions except that the falci- form plate is slightly concave rather than con- vex distally. The cuboid (Fig. 37) retains all the major features of the Desmaninae and Talpinae ex- cept that it is much more compact. The appar- 52 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 TABLE 8 MEASUREMENTS OF THE TIBIOFIBULA OF Gaillardia thomsoni Length of coossified portion to base .... ..... . Distal width .. ......... .. .... ........ ..... ....... .... ......... Minimum width of shaft ......... .................... . e Estimated measurement Figure 37. Gaillardia thomsoni, McKay Reservoir, UO loc 2222, UO 26834, right cuboid, anterior and right lateral views. ent shortening correlates with the closer ap- proximation of the calcaneal and metatarsal articular facets which tends to increase the rela- tive length of the prominent proximoventral spine. The groove for the tendon of the M. peroneus longus passes around the lateral and posterior sides of the cuboid perpendicular to the axis of the cuboid instead of oblique to the axis as in other talpids and soricids. The calcaneum (Fig. 38) has the deep pit on the end of the calcaneal tuberosity for the tendon of Achilles and the well developed trochlear process and sustentaculum astragali as in Desmana. The calcaneum of Gaillardia differs from Desmana in the longitudinally elongate posterior articular surface, laterally and posteriorly displaced cuboid facet, and lack of a continuous ridge or backbone extend- ing from the end of the calcaneal tuberosity to the cuboid facet ventrally. The fourth metatarsal (Fig. 39) is elongate with the distal end rotated medially about 45° with respect to the proximal end as in the Des- maninae. The metatarsal differs from that of Desmana in the narrower proximal articular surface, transversely concave distal articular uo uo uo uo uo 3792 24809 2456 24808 26832 18+ 18.7 6.6 5.8 6.75 5.8e 6.25 2.6 2.4* 2.4 2.1 Figure 38. Gaillardia thomsoni, McKay Reservoir, UO loc 2222, UO 26833, right calcaneum, dorsal and right lateral views. TABLE 9 MEASUREMENTS OF THE CALCANEA OF Gaillardia thomsoni Length uo 3376 .................... . . uo 24807 ...... ............. . 7.8 uo 26833 .. ... .... .... ...... . 9.1 e Estimated measurement Width across sustentaculum 3.8e 4.0 4.7 surface and relatively weak ventral keel on the distal end. The transverse compression of the metatarsal is not as great as in Desmana. The two specimens (UO 25127, UO 26836) of the fourth metatarsal are 18.0 mm. and 16.2 mm. long respectively. 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 53 L. A B '.~/J-~- c D Figure 40. Left humeri, posterior views, A; Uropsilus pilirostris, B; Gaillardia thomsoni, UO 26829, C; Desmana moschata, D; Galemys pyrenaicus. Figure 39. Gaillardia thomsoni, McKay Reservoir, UO loc 2222, UO 25127, right metatarsal, proxi -- mal and dorsal views. RELATIONSHIPS: In the original description of Gaillardia thomsoni Matthew ( 1932) did not specifically include it within the Talpidae but this is implied by the nature of his compari- sons. He assumed on the basis of the alveoli that the P2 and canine were double-rooted and thus excluded it from relationship with the desmans while noting that "Condylura perhaps comes nearest, ... " My examination of the type re- vealed no particular reasons for assuming that the canine or P2 were double-rooted and indeed the converse is suggested by the size relation- ships of the alveoli. In my opinion the lower jaw of Gaillardia retained a complete dental formula as in the desmans. The postcranial skeleton of Gaillardia be- longs unquestionably to that of an aquatic insectivore in which the hind feet and tail were the main locomotor organs in water. This feature and the humeroclavicular articulation immediately suggest relationships with the des- mans. However, the functional adaptation of Gaillardia to aquatic locomotion does not nec- essarily indicate relationships with the Des- maninae. While sharing a number of func- tional similarities with the Old World desmans, Gaillardia has a great number of dissimilari- ties in every element available for comparison, and these differences are of greater degree than the differences within the Desmaninae. 54 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 The humerus exhibits some of the more striking differences on which relationships may be weighed. On the humerus, the ovoid rather than elliptical head, open bicipital groove, low lesser tuberosity, and lack of fossa for the great ligament of the M. flexor digitorum profundus are not characteristic of Old World desmans nor most of the Talpidae with the exception of the Uropsilinae. Comparison of the humeri of Gaillardia and Uropsilus ( or Mystipterus) (Fig. 40) reveals a striking similarity both morphologically and proportionately. Considering this similarity in the humerus and possibly also the clavicle, should Gaillardia be placed within the Uropsilinae? Gaillardia may be distinguished from both unquestioned desmanines and unquestioned uropsilines at least at the tribal level. Systematic assignment hinges on the question of whether burrowing adaptions of the humerus seen in the Des- maninae were inherited from a semifossorial ancestor or developed independently after semiaquatic or aquatic adaptations had been initiated. There would be no great difficulty in deriving the Desmana type of humerus from a urotrichine type, a view favored by Gill ( 1888) and Reed (1951). Campbell (1939) postu- lated that fossorial moles must have had an aquatic ancestor to survive a nonadaptive stage in the transition between ambulatory and fos- sorial stages. I agree with Reed ( 1951) that Campbell's hypothesis is unnecessary. The de- velopment of fossorial features from a uropsi- line-like or shrew-like ancestor has happened at least twice according to Reed and Turnbull ( 1965). The Proscalopinae and Talpinae both develop an elliptical humeral head, fossa for the great ligament of the M. flexor digitorum profundus, large olecranon fossa, etc., but other functional and chronological data sug- gest that both groups were independently de- rived from a uropsiline-like stock. , The same may be true for the Desmaninae whether or not Gaillardia is a relic of an early adaptive stage in the evolution of the Desmaninae. My conclusion is that Gaillardia represents a separate lineage of talpids derived from uropsiline ancestors, and adaptively convergent on Old World desmans. The magnitude of the differences between Gaillardia and Uropsilus and M ystipterus suggests the proposal of a sub- family. Subfamily TALPINAE Fischer von Yaldheim, 1817 DIAGNOSIS: Bulla ossified ( Talpini, Scalopina) or open, infraorbital foramen much larger than lacrimal foramen; zygomatic arch only slightly arched dorsally or straight; hypocone- like metaconule present or absent, anterior in- cisors enlarged (Scalopini, Urotrichini, Con- dylurini) or upper canines and lower first premolars enlarged (Talpini, Scaptonychini), postfossid of molars open or closed lingually; manubrium with greatly developed ventral keel, dorsal keel present or absent, alae large and expanded, (Scaptonychini, Urotrichina, Condylurini) or vestigial-when expanded the alae are medial; clavicle not articulating with scapula, short and stout, ventral spine present -may be vestigial in extremely short clavicles; heterotopic bones present (Parascalopina) or absent ventrally between clavicle and manu- brium; scapula-metacromion present or absent, acromoin may be reduced, glenoid fossa per- pendicular to long dimension, subscapularis and teres major fossae present; humerus rela- tively slender to very broad, proximal end broader than distal end, bicipital groove long and partly covered by fusion of the pectoral crest with lesser tuberosity, fossa for the M. flexor digitorum profundus ligament present, head laterally compressed; ulna with coro- noid process moderately to strongly devel- oped, transverse proximal crest present; radius with capitular process weak to strong; manus relatively narrow to very broad; ungual pha- langes bifid, scaphoid and lunar separate, either M. extensor carpi ulnaris sesmoid or os falciformis present; pubic symphysis absent, pseudosymphysis present or absent; five verte- brae in sacrum, three to five vertebrae fused to innominate; femur moderately expanded prox- imally, shaft normal; tibiofibula long to short, 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 55 only moderately grooved distally, fibula com- plex proximally; pes larger to smaller than manus, metatarsals unspecialized. Tribe Scaptonyc'hini Van Valen, 1967 DIAGNOSIS: Unexpanded bulla; Ii not enlarged, upper canine and P1 enlarged, manubrium with large alae and articular facets for clavicles strongly inclined; clavicle long; scapula with infraspinatus and rhomboid fossae, no meta- cromion process or suprascapular foramen. No New World genera known. Tribe Talpini Fischer von Valdheim, 1817 DIAGNOSIS: Bulla expanded; g not enlarged, upper canine and P1 enlarged; manubrium with small alae, groove for anterior vena cava undivided, articular facets for clavicles in- clined anteriorly; clavicle short, articular facets roughly parallel; scapula without infra- spinatus fossa, rhomboid fossa, metacromion process, or suprascapular foramen, teres fossa partly divided; humerus broad with head di- rected distally or mediodistally, great excava- tion of brachialis fossa; pubi close together. No New World genera known. Tribe Urotrichini Dobson, 1883 DIAGNOSIS: Bulla not expanded, Ii enlarged, upper canine and P1 not enlarged; manubrium with large alae, articular facets for clacivles inclined; clavicle long with converging articu- lar planes; scapula with metacromion process, infraspinat:is fossa and rhomboid fossa, no suprascapular foramen; humerus narrow, brachialis fossa shallow, head directed latero- distally; ulna long, sinuous; radius long and with a capitular process crest; third metacarpal longest; pubi widely separated. ? N eurotrichus columbianusn n. sp. (Fig. 41) TYPE: UO 24816, a nearly complete right mandible lacking only the tip of the coronoid and angular processes and the teeth preceding the P4 • HYPODIGM: uo 26826, P4; uo 24814, M1; uo n After Columbia River. 24832, UO 25129, M2's. All of the above from type locality. TYPE LOCALITY: McKay Reservoir, UO loc. 2222. AGE: Hemphillian. DIAGNOSIS: Small mole with dental formula: ? ? ? ? . Mandible with noticeable taper to ante- 3l33 molar region, anterior teeth procumbent, Pa-P4 double-rooted; molars brachyodont, resemble N. gibbsii but crista obliqua of M2-Ma usually joins protolophid more lingually and labial cingula better developed, molars proportion- ately larger; anterior margin of ascending ramus straight and at right angles to the hori- zontal ramus, longer condylar process and more posteriorly situated mandibular foramen than in Neurotrichus. DESCRIPTION: The dentary ( Fig. 41) is similar in general shape to that of the shrew-moles. The ascending ramus forms a sharp right angle with the horizontal ramus. Posterior to the level of the coronoid process the mandibular foramen is situated on the ventral margin of the mylohy- oid ridge. The masseteric fossa is rather deeply excavated. Although the angular process is broken off, the broken surface suggests that it was rather platelike as in Scapanulus. The coronoid process is also broken but was defi- nitely rather short anteroposteriorly. The ante- molar region is relatively short and tapered. There is one mental foramen beneath the p3 and another below the M1 trigonid. The sym- physial scar extends to the level of the P4. The three most anterior teeth, here interpreted as incisors, are represented only by their alveoli, all of which are quite procumbent. The borders of the two anterior alveoli are damaged so that the relative diameters are difficult to interpret. The Ia seems to be the smallest and is arbitarily considered here to be the tooth lost in N. gibbsii. The I 3 is followed by two equally large alveoli for the C and P1. I consider the succeed- ing two alveoli, a small anterior one followed by a somewhat labially situated larger one, to represent a double-rooted P 3 • It might be ar- 56 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON I I A c D No.11 Figure 41. ?Neurotrichus columbianus n. sp. McKay Reservoir, UO loc 2222, Type specimen UO 24816, right mandible with P 4-Ma, A; labial view, B; lingual view of ascending ram us, C; lingual view of P 4 - Ma, D; occlusal view alveoli and P4-M3. TABLE 10 MEASUREMENTS OF THE LOWER DENTITION OF ? N eurotrichus columbianus uo 24816 (Type) P 4 : Length ... ....... .... ...... .............. 1.40 Width ---------------------------------- 0.83 M1 : Length ---------------------------------- 2.07 Width ---------------------------------- 1.32 M2 : Length .. .... ............. .. ............. 2.13 Width ---------------------------------- 1.30 M3 : Length .... --- -- --- --- ------------------- 1.80 Width ---------------------------------- 0.97 M1-Ma Length -------------------------------- 5.9 C-P4 Alveolar 3.0 P2-P4 Alveolar -------------------------------- 2.5 Length Ma to Condylus ________ --------- 6.5 uo 24814 1.90 1.33 uo 24832 2.00 1.23 uo 25129 2.00 1.17 uo 26826 1.56 0.81 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 57 gued in light of the dentition of some of the other shrew-moles ( Urotrichus) that these two alveoli represent individual teeth, but their relative size and arrangement to each other and to the anterior mental foramen is so like that found in N. gibbsii that any other interpre- tation seems less likely. The P4 is elongate with a well developed heel which is the widest part of the tooth. The protoconid is inflated and centered over the anterior root. There is no evidence of a meta- conid on the type but the isolated specimen has a small metaconid about half way up the posterolingual side of the protoconid. A nar- row anterior cingulum surrounds part of the protoconid and a somewhat heavier cingulum encloses the heel posteriorly and laterally. A deep short groove divides the heel longitu- dinally. A similar groove or arrangement of flanking enamel folds is present in N. gibbsii, Uropsilus, Mystipterus, Urotrichus, and Scalo- poides. The anterior root is smaller than the posterior one. The molars are brachyodont and propor- tionately similar to Uropsilus, and other Uro- trichini. The molars appear to be relatively larger in proportion to the jaw than the above. Labial cingula bridge the hypoflexid. The crista obliqua of the MrM3 contact the middle of the protolophid, or the base of the metaconid. The other features of the molars fall within the variation of those of N. gibbsii. RELATIONSHIPS: Size and general proportions of ? N. columbianus seem to align it with the shrew-moles. The crowded and reduced ante- molar dentition tends to remove it from Achlyo- scapter as well as a number of other differ- ences. The known uropsilines all possess pos- terior cingula on the M1-M2 and differ in mor- phology of the P4, low situation of the man- dibular foramen, and in the shrew-like angular process. ? N. columbianus lacks the well de- veloped metastyles of the M2-M3 of Scalopoides , and the arms of the M2-M3 trigonids diverge more. Among the Urotrichini ? N. columbianus differs from the species of Urotrichus in dental formula, proportionately larger P4, double- rooted P3, and more robust M1 trigonid. ? N. columbianus resembles N. gibbsii in the gen- eral morphology of the P4-M3 , double-rooted P3 , and mental foramina position. It can be dis- tinguished from the living species by the sharp angle of the horizontal and ascending rami, proportionately larger molars, additional ante- molar tooth, and more posterior mandibular foramen and condyle. The above comparisons are similar to those of Wilson (1960) in comparing Mystipterus { M ydecodon) martini to the shrew-moles. Here, as there, the dental formula and small differences in molar form may give a false im- pression of relationship. The morphology of the lower antemolar teeth in these groups is of little use without corroborative postcranial ma- terial in determining the relationships of the various small moles. On the jaw material avail- able, ?N. columbianus seems to be closely re- lated to N. gibbsii. However, convergence in molar and premolar morphology has been dem- onstrated in other moles ( e.g. Scalopus-Sca- panus, Domninoides-Proscalops) and can be documented in a number of other undescribed forms. Features of the ascending ramus and crowding of the antemolar teeth lead me to suspect that ? N. columbianus is related to Scalopoides. None of the features of the teeth are so strange that they could not be de- rived from Scalopoides. S. isodens has already achieved the same dental formula and root re- duction seen in ? N. columbianus. Also the ?Scalopoides material from McKay would tend to support such a view. On the other hand ?N. columbianus would make a good temporal and structural ancestor for N. gibbsii on the basis of assigned material. The poor record of Scalopoides-like dentitions in the Clarendonian and Hemphillian does not give any clues as to the dental evolution or trends of this complex. The general relationship of the reduction in rostrum length and increasing fossorality ex- hibited by several lineages would suggest such an end product, but this generalization cannot as yet be demonstrated within the Scalopoides complex. 58 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 If some of the Scalopoides-like postcranial material and the ? N. columbianus material rep- resent the same group of animals, then a new genus would probably be in order. Since this association cannot be adequately demonstrated and inasmuch as there is no means evident to determine if the similarities signify homology or analogy, I have sought to treat the various elements separately and let the weight of the morphological features determine taxonomic assignment for each element. Tribe Condylurini Dobson, 1883 DIAGNOSIS: Bulla not expanded, Ii enlarged; upper canine and P1 not enlarged; manubrium with large alae, articular facets for clavicles inclined anteriorly; clavicle moderately long with distinct process for origin of M. acromio- deltoideus muscle, articular planes converge; scapula with metacromion process, rhomboid fossa, and infraspinatus fossa, no suprascapu- lar foramen; humerus moderately broad, head directed laterodistally, brachialis fossa shal- low; ulna long and sinuous, very broad proxi- mal crest; metacarpals slightly graded in size from the fifth, the longest, to the first; pubi close together. No Tertiary fossils known. Tribe Scalopini Dobson, 1883 DIAGNOSIS: Ii or g enlarged, upper canine and P1 not enlarged; manubrium with small alae, articular facets for clavicle vertical, groove for anterior vena cava divided posteriorly; clavicle moderately long to short, articular facets convergent to roughly parallel; sea pula with infraspinatus fossa, rhomboid fossa, and suprascapular foramen, metacromion process absent; humerus moderately to very broad, head directed distally to mediodistally, brachi- alis fossa moderately deep; ulna and radius moderately long to short; pubi close together. Subtribe Parascalopina new DIAGNOSIS: Metastylids of molars usually pre- sent; bulla not expanded; stapedial tube in- completely or poorly ossified; manubrium with dorsal ridge and paired anterior venae cavae; heterotopic bones may occur between clavicles and manubrium; humerus generally narrower and pectoral ridge terminates more laterally than in the Scalopina, concavity surrounded by pectoral ridge, pectoral crest and greater tuber- osity of humerus deep; capitular process of ra- dius flattened proximally. SCALOPOIDES Wilson, 1960 GENOTYPE: Scalopoides isodens Wilson, 1960 KNOWN DISTRIBUTION: Middle Miocene ( Hem- ingfordian) of Colorado, Late Miocene (Bar- stovian) to Middle Pliocene ? ( Hemphillian) of Oregon. AMENDED DIAGNOSIS: Small fossorial moles of dental formula: ? ? ? ? orfewer. 12 moderately 3133 enlarged, P2 single or double-rooted, p3.p4 double-rooted. Molar crowns more or less mesodont, protoconids and hypoconids angu- late, prominent metastylids present on flanks of metaconids in MrM3 , crista obliqua extending to bases of metaconids, entocristids help en- close postfossids lingually. M1-M2 subequal. M3 smaller but robust. Dentary talpine-like in shape. Two mental foramina, posterior one be- neath or just posterior to P4. P4 subtriangular and compact. Upper molars tritubercular, pro- toconule small or absent, metaconule relatively small and close to protocone, mesostyles slight- ly or not twinned. Clavicle intermediate in length between Condylura and Scapanulus, de- tails closer to Scapanulus, faint to moderate reflection of the ventromedial spine. Humerus with gross proportions similar to Condylura but details nearer Scapanus, axis of the head parallels humeral axis, clavicular articular facet crudely ovate, prominent "scalopine ridge". Ulna Parascalops-like but more slen- der. Radius more robust than in Urotrichini or Condylura, capitular process well devel- oped, distal end scalloped. Metacarpals much shorter than in the Urotrichini or Condylura but longer than living Scalopini. l 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 59 Scalopoides isodens Wilson, 1960 ( Figs. 42-45) Although Scalopoides isodens is not known from Oregon, as the genotypic species it fig- ures prominently in any discussion of the rela- tionships of the genus and referred species. The following descriptions are based upon Wil- son's ( 1960) original material and additional material sorted from the original collection res- idue. In the description of the humerus Wilson (1960) stated that the "humeri of S. isodens, in addition to proportions, agree except for minor details with those of Condylura." In the details of the humerus a much better match is found within the Scalopini (Fig. 10). Wilson did not see the humerus of Scapanulus, so a detailed comparison is called for here as well as a re-evaluation of the similarities and dissimilarities between Condylura and Scalo- poides. Of the living Scalopini , Scapanulus has the longest humerus in relation to its width and is comparable to Condylnra in this re- spect. Scalopoides resembles Scapanulus in that ( 1) the axis of the head is nearly parallel to the axis of the humerus, ( 2) the "scalo- pine ridge" of Campbell ( 1939) is present and generally similar, ( 3) the clavicular articular facet is strongly angled with respect to the axis of the humerus ( about 30° ), ( 4) the large deep channel separating the head from the greater tuberosity is absent, ( 5) the trochlea is relatively broad, ( 6) the fossa for the liga- ment of the M. flexor digitorum profundus has only a very slight anterior component, and (7) the proximal end of the bicipital tunnel is vis- ible anteriorly. Scalopoides differs from Sca- panulus in that ( 1) the plane of the head is more nearly parallel to the humeral axis, (2) the teres tubercle is shorter, and ( 3) the pec- toral crest is shorter. Scalopoides resembles Condylura in the relatively wide gap between the medial rim of the trochlea and the fossa for the ligament of M. flexor digitorum pro- fundus and the pectoral crest is more nearly comparable in length. Condylura differs from Scalopoides in that ( 1) the head is directed strongly laterally, (2) the scalopine ridge is absent ( 3) the clavicular articular facet is nearly parallel to long axis of humerus, ( 4) a deep, wide channel separates the head from the cla vicular facet, ( 5) the trochlea is rela- tively narrower, ( 6) the teres tubercle is rel- atively shorter, (7) the fossa for the M. flexor digitorum profundus ligament has a strong anterior component and (8) the proximal end of the bicipital tunnel is not visible anteriorly. The subovate shape of the clavicular facet (Fig. 42) in Scalopoides shows a much closer rela- tionship to Scapanulus and thus to the Para- scalopina on the humerus than it does to Condy- lura or any more generalized form. However, in many respects the humerus of Scalopoides may be thought of as intermediate in form be- tween Condylura and the Recent Scalopini. The clavicle is relatively longer than in the living Parascalopina but shorter than in Condy~ B c Figure 42. Lateral views of the clavicular region of the humerus, A; Condylura cristata, B; Scapanulus oweni, C; Scalopoides isodens (UK 10082 with deltoid restored from UK 10084). 60 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 lura. The ventral process is damaged on both available specimens, UK 15570 and UK 15571. The ventral process seems to have been blade- like and unelaborated. There is a ventral artic- ular facet on the ventromedial spine for artic- ulation of a small heterotopic bone. The hu- meral facet faces strongly posteriorly and slightly ventrally rather than slightly dorsally as in the living Scalopini. The gutter leading from the vascular notch to the dorsal promi- nence is only moderately developed. The clav· icle has none of the extra processes seen on the clavicle of Condylura. Figure 43. Scalopoides isodens, Quarry A Martin Canyon, UK 15572, proximal end of right scapula, dorsal and superior views. Two specimens of fragmentary scapulae are probably referable to Scalopoides isodens. The more complete one (UK 15572, Fig. 43) is smaller than the less complete specimen (UK 15573). The disparity in size is probably the result of individual age or variation. In UK 15572 the acromion process is damaged but high. The area of attachment for the acromio- clavicular ligment faces strongly anteriorly. The groove for the infraspinatus tendon or muscle is extremely deep and well defined to the region of the suprascapular foramen. The deepness of the groove and high acromion pro- cess are suggestive of the Urotrichini except that the suprascapular nerve pierces rather than runs anteriorly around the acromion process. Three ulna fragments (UK 15576- UK 15578, Fig. 44) represent the entire form of the ulna. While there are resemblances to the ' ·, .. ' ' 1 A B Figure 44. Scalopoides isodens, Quarry A Martin Canyon, A; UK 15576, proximal end of right ulna, lateral and medial views, B; UK 15577, distal end of right ulna, lateral view. Urotrichini, the majority of features place it well within the Scalopini. The ulna of S. isodens resembles the Urotrichini and Condylura in ( 1) the relative shortness of the olecranon proc- ess, (2) the very slight sinuousness of the shaft, ( 3) and the somewhat obtuse angle between the distal end of the shaft ( in the U rotrichini) and the anterior moiety of the distal epiphysis. S. isodens resembles Parascalops and Domni- noides in the short and robust shaft and the structural details of the distal and proximal articular facets. The triceps scar merges im- perceptibly with the anterior platform of the olecranon so that I could not determine on these specimens whether the triceps insertion was as large as in the Scalopini or reduced as in Con- dylura; although the former is more probable in light of other resemblances to the Scalopini. The two complete radii (UK 15579, UK 15580) and five partial ones (UK 15581-UK 15585) show close affinities to the Scalopini. The radius (Fig. 45) is relatively shorter and more robust than in Condylura or the Uro· trichini. The distal extremity is scalloped in palmar view as in the Scalopini. The capitular process resembles Condylura and the Scalopina in having an oblique rather than flat termina· tion. There is no capitular crest but there is also no distinct fossa for the origin of the M. abduc- tor pollicis longus, although its area of origin is probably the slight depression immediately anterodistal to the ulnar facet. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 61 LJ Figure 45. Scalopoules isodens, Quarry A Martin Canyon, UK 15579, right radius, medial view and posterior view of proximal end. A single proximal phalanx II or III, UK 15586, of the manus is relatively longer than that of Parascalops or the Scalopina but much too short to be confused with Urotrichini. It is definitely that of a burrowing mole. The three ungual phalanges, UK 15587, of the manus are bifid and tranversely broadened distally as in other Talpinae. The subungual process is rather deep as compared to those of the living burrowing moles but not as deep and globular as in the Urotrichini. The lateral blades of the claw extend more posteriorly as in the burrowing moles. Only one poorly preserved specimen (UK 15588) of the distal end of the tibiofibula was recovered from Martin Canyon. It resembles the better known S. ripafodiator described be- low except that the groove for the peroneus tendons is not well developed and the crural ligament is not prominently ossified. RELATIONSHIPS: The relationships of s. iso- dens to other species and the affinities of the genus are considered in the discussion of S. ripafodiator and summary below. Scalopoides ripafodiator10 n. sp. (Figs. 46-60) TYPE: UO 22488, a right mandible with 12 and P1-M3, lacking the ascending ramus, tip of the jaw anterior to the 12, and trigonid of the M1, Molars heavily worn. HYPODIGM: uo 22504-UO 22509, uo 22511-UO 22515, UO 22517-UO 22518, UO 22520-UO 22524, UO 22527-UO 22543, UO 22578, U022580, UO 22582-UO 22589, UO 26314, UO 26322-UO 26324, UO 26339, upper teeth and maxillary fragments in- cluding 11 P4, 2 P2, 20 M1, 17 M2, 2 M3 ; UO 22342, UO 22482-UO 22485, UO 22490, UO 22492, UO 22494-UO 22496, UO 22499 --UO 22501, UO 22545-UO 22546, UO 22548-U022551, UO 22566, UO 22568- UO 22569, UO 22571-UO 22576, UO 22590 - UO 22592, UO 22595, UO 22598-UO 22600, UO 26303, lower teeth and fragmen- tary mandibles; UO 22414, anterior portion of manubrium; UO 22415-UO 22417, UO 22427, uo 22431, uo 22436, uo 24839- uo 24842, complete and incomplete clavicles; UO 22354-UO 22373, UO 22375-UO 22384, UO 22387-UO 22394, UO 24844- UO 24846, humeri and fragments thereof; UO 22339, UO 22451-UO 22469, UO 2484 7- UO 24848, fragmentary ulnae; UO 22336, UO 22420-UO 22425, UO 22428-UO 22429, UO 22433-UO 22434, UO 26800, one com- plete and ten fragmentary radii; UO 22444- UO 22450, UO 24833- UO 24834, scaphoids; UO 24239-UO 22440, UO 22443, third meta- carpals; UO 22441-UO 22442, ungual pha- langes of manus; UO 22552-UO 22557, UO 22559-UO 22563, fragmentary innominata; UO 22395-UO 22396, UO 22398-UO 22399, UO 22402-UO 22408, UO 22411, UO 24836, fragmentary femora; UO 224 70- UO 224 71, UO 224 73-UO 224 79, UO 22481, tibiofibula fragments-three proximal epiphyses and seven distal ends; UO 22401, astragalus; UO 22404, UO 22410, UO 24837, calcanea. All from type locality. JO From the Latin ripa, stream bank, and Jodiator, a digger. 62 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 TYPE LOCALTY: uo loc. 2465. Quartz Basin. AGE: Barstovian. DIAGNOSIS: Small mole about the size of S. isodens, and dental formula ? ? ? ? . Lower mo- 3l33 lars generally similar to S. isodens. la less spat- ulate, P2 double-rooted, antemolar region more robust and longer. Metaconule of upper molars present, paraconule low but evident and persis- tent with wear, M2 with slight division of the mesostyle. Clavicle, humerus, radius and ulna generally similar to S. isodens, humerus rel- atively narrower, longer pectoral ridge, radius and ulna proportionately more slender. DISCRIPTION: Only a small portion of the maxilla of Scalopoides ripafodiator is pre- served in the sample; however, UO 22515, (Fig. 46) clearly shows the nature of the base of the zygomatic process of the maxilla. The relationships of this region are most like those of Parascalops breweri among the living moles and to a lesser extent like those of Scalopus aquaticus. There are two distinct crests for the insertion of the M. masseter that extend anterior from the dorsal and ventral margins of the zygomatic arch. The dorsal crest is either ab- sent or fused with the ventral crest in other talpines examined. The P8 is a simple tooth ( Fig. 4 7). The crown is essentially a single connate cusp with a small posterior cingulum. It is double-rooted with the round anterior root inclined anteriorly. The transversely wider posterior root projects vertically into the maxilla. The P4 is a symmet- rically triangular tooth (Fig. 47) with the la- bial side the longest. The paracone is centered over the center of the tooth. The posterior moi- ety of the cusp is somewhat compressed trans- versely into a stout crest. The protocone is rel- atively small but usually a well defined cusp. It is transversely compressed and essentially an enlargement of the narrow cingulum which encloses the posterior moiety of the tooth. This cingulum may connect by way of a very narrow labial cingulum with the short anterior cingu- lum which forms the anteriormost extremity of the tooth. , ' 1-------j Figure 46. Scalopoules ripafodiator n. sp. Quartz Basin, UO loc 2465, UO 22515, right M2 and as- sociated maxillary fragment, labial view. Figure 47. Scalopoides ripafodiator n. sp. Quartz Basin, UO loc 2465, UO 22580, right P3-P4, labial, occlusal, and lingual views. The M1 of S. ripafodiator (Fig. 48A) resem- bles that in the Urotrichini in general shape and morphology. The protocone is large and flanked by a smaller but distinct metaconule. The pa- raconule is minute and just anterior to the apex of the protocone on the protocrista. All evidence of the paraconule is usually elimi- nated during the initial stages of wear. Narrow paracrista are present. The metastyle is elon- gate but only weakly divided. There is a small and well defined parastyle. The four roots are situated under the metacrista, paracone and protocone and a filiform one below and a little labial to center of the tooth. The root under the metacrista is strongly compressed anteroposte- riorly while the others are oblate in cross- section. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 63 I • ¢!;(; · B l c Figure 48. Scalopoides ripafodiator n. sp. Quartz Basin, UO loc 2465, A; UO 22537, right M1, la- bial, occlusal, and lingual views, B; UO 22506, left M2, labial and occlusal views, C; UO 22520, left M3, occlusal view. The M2 of S. ripafodiator (Feb. 48B) strong- ly resembles that of S. isodens in most respects. The metaconule is larger and more distinct than in S. isodens. The parastyle which is not present in Wilson's material (possibly abraded off) is a distinct cusp. Some vestige of a labial cingulum may be sporadically present in the reentrant folds between the protocone and meta- conule and between the protocone and the para- conule. The paraconule, although about as prominent as in S. isodens, is slightly more re- moved from the apex of the protocone and less likely to be completely obliterated by wear. The paracingulum and metacingulum ( with one exception in the case of the latter) do not extend to the stylar cusps. The M2 has four roots in analogous position as the M1• lnter- radicular crests are absent in all the molars. Only two worn third upper molars (Fig. 48C) of a small mole were recovered from Quartz Basin, both appear to be referable to S. ripafodiator. They strongly resemble those of S. isodens. The paraconule is not discernible in the worn Quartz Basin specimens and the metaconule is somewhat better developed. The Ma is three-rooted with one root under each extremity. The mandible ( Fig. 49) is more robust espe- cially anteriorly than S. isodens. There is little taper to the jaw anteriorly and it is about as deep beneath the P2-Pa as under the Ma. There are two mental foramina, one under the P 4 and a larger one under the Pa or between the roots of the P2 and Pa. The symphysis extends pos- teriorly beneath the P4. There is one antemolar tooth absent in S. ripafodiator. Wilson (1960) listed the lower dental formula for S. isodens as complete based on analogy with Prosca- panus sansaniensis Lartet. He had previously thought the formula was 3, 1, 3, 3. Comparison of S. ripafodiator and S. isodens indicates that Wilson's earlier determination was the cor- rect one and that both species have the same den- tal formula. 11 The antemolar region is rela- tively longer and the teeth are not as crowded as in S. isodens. The length of the h to P4 com- prises 82 % of the M1-M3 alveolar length in S. ripafodiator as compared with only 68 % in S. isodens. The 11-h are inclined about 45° to the axis of the jaw and the succeeding teeth progressively less so. The P2-P4 are double- rooted. An X-ray of the antemolar region (Fig. 49B) reveals a long and strong h root extend- ing posterior to the level of and beneath the P2. Consequently, the roots of the canine and la are quite short. The single-rooted antemolars (Fig. 50) are of uneven size ranked in the fol- lowing order from smallest to largest, la, C, Ii, h. The Ii is incisiform and closely appressed to the fa The h is enlarged as in Scapanulus and bears a similar relationship to the 11 • Pos- teriorly the crown is rounded but anterolingual- ly it narrows to an edge thus imparting a tear- shaped cross-section to the crown. The la and C are peg-like or slightly premolariform teeth. The roots of the P2 are somewhat oblique in the 11 The homologies of the two teeth between the 12 and P 2 are not presently demonstrable. The designation of the missing tooth as the P 1 is an arbitrary one. 64 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 A 8 Figure 49. Scalopoi,des ripafodiator n. sp. Quartz Basin, UO loc 2465, Type specimen UO 22488, right mandible, A; labial and occlusal views with 12-M3,B; sketch of X-Ray photograph of antemolar region. jaw with the posterior root lingual to the an- terior one. The large protoconid is centered over the anterior root. As in the rest of the pre- molars the protoconid slants somewhat an- teriorly. There is a very small posterior cin- gulum with an incipient ridge extending an- terodorsally toward the apex of the protoconid. The roots of the P2 of S. ripafodiator are not as closely appressed as in S. isodens. The Pa re- sembles the P2 but is larger, the roots are bet- ter aligned longitudinally and the posterior cingulum and ridge are better developed. The P4 of S. ripafodiator generally resembles that of S. isodens except in the following details: ( 1) it is broader, ( 2) the protoconid is more inflated, ( 3) the posterior cingulum extends transversely the width of the protoconid before curving anteriorly, ( 4) paraconid and anterior cingulum are absent or weakly developed, ( 5) the metaconid is not discernible or very weakly so. The molars (Fig. 51) closely resemble those of S. isodens but in general have a somewhat more inflated appearance and the labial cin- gula are less distinct. In the M1 the paralophid is not straight as in S. isodens. The lingual moiety of the paralophid is bent lingually rather than anterolingually thus imparting a more compact aspect to the tooth. The anterior cingulum is variable in occurrence and expres- sion. The M2 is similar to S. isodens except as noted above and in the construction of the tal- onid. The entoconid is usually more massive and columnar than in S. isodens. The Ma closely resembles S. isodens except as noted above. Only the anteriormost extremity of the manubrium (Fig. 52) was recovered. In the features preserved it shows closest similari- ties with the manubrium of Scapanulus and Parascalops. The bases of the alae are more dorsally situated and merge posteriorly with the broad and poorly differentiated dorsal keel. The alae were apparently rather broad al- though probably not much more so than in Scapanulus. The anterior margin and clavicu- lar articular surfaces are nearly vertical. The Figure 50. Scalopoides ripafodiator n. sp. Quartz Basin, UO loc 2465, UO 22499, anterior end of right mandible with Ii-C, labial view. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 65 A B c ~ ltfJ Figure 51. Scalopoides ripafodiator n. sp. Quartz Basin, UO loc 2465, A; UO 22484, right M1, anterior, labial, occlusal, and lingual views, B; UO 22573, right M2, labial, occlusal, and lingual views, C; UO 22490, right M3 , same views as B. Figure 52. Scalopoides ripafodiator n. sp. Quartz Basin, UO loc 2465, UO 22414, anterior part of manubrium, right lateral and anterior views. Figure 53. Scalopoides ripafodiator n. sp. Quartz Basin, UO loc 2465, UO 22432, right scapula, dor- sal and superior views. anterior ventrolateral bosses are not anterolat- eral projections, situated to contact the ventro- medial borders of the clavicles as in Scapanus, Scalopus, and Talpa, but are rather posteriorly situated, suggesting that they served for the ar- ticulation of a tetrahedral heterotopic bone or cartilage as in Scapanulus and Parascalops. This is verified by the morphology of the ven- tromedial spine of the clavicle as well as the gap between the ventral articular surfaces of the clavicle and manubrium when articulated. Only a small region of the proximal area of the scapula (Fig. 53) was preserved in the sample, and its construction strongly resembles that of Parascalops. The suprascapular nerve pierces the acromion. The path of the infra- spinatus tendon is well marked by a pro- nounced groove extending from the infraspi- natus fossa to the region of the suprascapular foramen but this groove is possibly not as deep as in S. isodens. The area for insertion of the 66 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 TABLE 11 MEASUREMENTS OF THE UPPER DENTITION OF BARSTOVIAN Scalopoides FROM OREGON Specimen Scalopoides ripaf odiator Quartz Basin uo 22580 ......... ..... ... ...... .... ..... . uo 22582 ..... ... .. .... ....... ........... . uo 22578 .. ... ....... ..... .. ..... ........ . uo 22579 ......... ....... ........ ........ . uo 22584 .... ...... ......... ......... .... . uo 22585 .. ...... ........ ... ..... ...... .. . uo 22587 ........ ............... ...... ... . uo 22588. ...... ...... ........... ....... . . uo 22589 ....... .. .......... ... ....... .. . . uo 22528 ............ ...... ......... ..... . uo 22530 ........... ...... ... ........ ... . . uo 22532 .. ..... .... ...... ......... .. .... . uo 22534 ........ ..... ...... .. ...... .. ... . uo 22535 .... .. .. .. ....... ....... .... .... . uo 22536 ....... ...... ......... .. .... .... . uo 22537 ...... .. ...... .. .. ... ... ... .. ... . uo 22538. ........... ..... .... ........ ... . uo 22540 .... .............. ... ........ ... . uo 22541.. .. ..... ... ............... ..... . uo 22543 ........ ... ................ ..... . uo 22504 .. ............ ... ... ......... ... . uo 22506 ....... .... ...... ..... ........ .. . uo 22507 ....... ....... ... ..... .......... . uo 22509 .................. ... ....... ... . . uo 22511.. ...... ... ...... .... ........ .. . . uo 22514 ....... ..................... ... . . uo 22515 ......... .. ....... ... ......... .. . uo 22518 ... .... .............. .......... . . uo 22523 ....... ..... .......... ....... .. . . uo 22524 ....... ..... .................. .. . Scalopoides cf. S. ripaf odiator Red Basin uo 24311.. .. .... ... ... ..... ...... ...... . . uo 24319 ............ ....... .. ......... .. . uo 24312 ...... ... .... ........ .... .... .. . . uo 24316 .......... ........ ........... .. . . uo 24317 .... ........... ...... .... ... ... . . uo 24318 ... ...... ... ......... ...... .. ... . uo 24320 .... .................. .. .... .... . uo 24321.. ....... .. ..... .... .... ... ... .. . uo 24322 .. ........ ...... ...... ........ .. . uo 24324 .. .... .... .... .... .. ..... ..... .. . uo 24325 .. -. .... ... ...... ....... ........ . uo 24328 ......... ... ..... .... ........ .. . . uo 24329 .. ........ ....... ... ... ... ...... . uo 24330 .... ..... .. ...... .. ..... ........ . e Estimated measurement. p a Lgth Wdth 0.87 0.65 0.91 0.61 0.77 0.54 p4 Lgth Wdth 1.61 1.31 l.6e 1.07 1.66 1.33 1.48 1.08 1.68 1.29 1.64 1.32 1.47 1.2* 1.70 1.32 1.64 1.23 1.56 1.12 1.54 1.10 1.53 1.08 * Noticeable abrasion or breakage tending to reduce measurement. M1 M2 Lgth Wdth Lgth Wdth 2.55 1.67 2.45 1.70 2.53 1.78 2.40 1.72 2.58 1.72 2.45 1.73 2.62 1.65 2.50 1.75 2.52 1.73 2.63 1.69 2.65 1.85 1.88 2.35 2.12 2.36 1.98 2.22 1.97 2.35 2.09 2.40 1.79 2.37 2.01 2.37 1.96 2.27 1.91 2.36 2.01 2.37 1.52 2.46 1.55 2.44 1.72 2.46 1.65 2.33 1.57 1.82 2.03 1.90 2.22 1.76 1.78 1.69 2.05 1.62 2.06 1.70 2.17 1.67 2.04 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 67 TABLE 12 MEASUREMENTS OF SELECTED LOWER DENTITIONS OF B..\HSTOVIAN Scalopoides S. ripafodi.ator S . cf. S. ripaf odiator Measurement uo uo uo 22488 22501 22550 (Type) Ii: Length Width h: Length 0.49 Width 0.43 13: Length Width P1: Length 0.38 Width 0.37 P2: Length 0.58 0.54 Width 0.50 0.48 Pa: Length 0.72 0.81 Width 0.60 0.60 P4: Length 1.16 1.20 1.13 Width 0.81 0.85 0.78 M1 : Length 1.92 1.85 Width 1.50 1.44 M2: Length 2.12 2.16 Width 1.45 1.46 Ma: Length 1.83 Width 1.16 acromioclavicular ligament faces dorsolater- ally as in the living Scalopini with only a slight anterior component. The clavicle of S. ripafodiator closely re- sembles that of the S. isodens. The sample from Quartz Basin shows a considerable variation in size (Table 14). The smaller specimens more closely resemble those of S. isodens in details and proportions (Fig. 54B) but the larger specimens (UO 24415, UO 22431, UO 22482) are proportionately longer than the others. The difference in length occurs entirely in the dis- tal moiety thus resulting in a long plate-like ventral process ( Fig. 52A). The gutter lead- ing from the vascular notch toward the dorsal prominence is better developed than in S. iso- dens. uo uo uo uo uo 22548 22499 24300 24301 23766 0.39 0.32 0.50 0.45 0.33 0.33 0.37 0.28" 0.37 0.34 0.54 0.38 0.74 0.53 0.71 0.59 0.42 0.50 1.06 0.86 1.15 1.30 0.76 0.65 0.67 0.85 1.47 1.82 1.33 1.45 The humerus of S. ripafodiator (Fig. 55) while strongly resembling S. isodens in most features differs in a few small details. The clav- icular articular facet is somewhat saddle- shaped compared to the simple curved articular facet in S. isodens. The trochlea is relatively wider. The pectoral ridge is straight, extends farther distally, and ends in a more vertically oriented and elongate tubercle, whereas in S. isodens the ridge scar tends to be curved or broken, does not extend as far distally and ter- minates in a stout, prominent and more trans- versely oriented tubercle. On the whole the humerus is relatively more slender than S. isodens. The radius of S. ripaf odiator ( Fig. 56), rep- resented by one complete and twelve incomplete 68 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 TABLE 13 MEASUREMENTS OF THE LOWER MOLAR TEETH OF BARSTOVIAN Scalopoides S. ripafodi.ator Quartz Basin uo 22494 ··········· ···· ··········· ······ uo 22546 ·· ·················· ······· ····· uo 22549 .. ...... .... ....... ...... ...... . uo 22551 ···· ··· ········· ··· ·· ··········· uo 22500 ... .... .. ..... ..... ...... ...... . uo 22484 ···· ············ ····· ···· ······· uo 22568 ··········· ·· ······ ······ ······· uo 22571 .... ... ... ... ...... ........ .... . uo 22572 ·········· ····· ················· uo 22576 ··· ·········· ······ ······ ······· uo 22600 .... ......... ...... .. .... ..... . . uo 22573 ···· ····· ··········· ···· ··· ····· uo 22575 ..... .. ........ ...... ..... ..... . uo 22342 ······· ··· ·· ··· ···· ········ ····· uo 22485 ... .... .............. ... .... ... . uo 22490 .. ..... ........................ . uo 22496 .... .. .. .. ... ...... ....... ..... . uo 22590 ... .... ....... ........ ........ . . uo 22591 ... .. ... ..... ..... ... ... ...... . . uo 22595 ....... .................... .... . uo 22598 ······· ······ ···· ·········· ····· S. cf S. ripafodiator Red Basin uo 24304 ............ .... .... .......... . . uo 24305 ·········· ··············· ······ · uo 24310 ······· ··· •··· ····· ············· uo 24298 ....... ..... ....... ............ . uo 24299 ·· ······ ·· ·· ·· ······· ···· ······· Guano Ranch M1 Length Width 1.80 1.90 1.94 1.88 1.88 1.87 l.92e 1.93 1.81 1.69 1.75 1.15 1.43 1.36 1.45 1.42 1.42 1.43 1.45 1.35 1.28 1.31 USNM 23767 .......... .......... ...... 1.89 1.36 Beatty Butte USNM 23761 ....... ... ... ... ..... .... . e Estimated measurement. radii, resembles that of S. isodens in most de- tails but differs in general proportions. It is relatively more slender than S. isodens (Table 16) in the shaft and terminal epi physis. While the number of specimens is greater than that from Martin Canyon, none of the M2 Length Width 2.30 2.01 2.09 2.04 2.01 1.93 1.94 2.00 2.20 2.20 1.15 l.26e 1.29 1.40 1.38 1.19 1.29 1.31 1.42 1.26 M3 Length Width 1.82 1.86 1.82 1.81 1.83 1.79 1.84 1.81 1.79 1.79 1.81 1.56 1.78 1.17 0.96 1.13 1.07 1.12 1.07 1.05 1.09 1.12 1.09 1.07 0.91 1.00 ulnae from Quartz Basin are as well preserved. In features that can be compared, S. ripafodia- tor differs from S. isodens in that the lateral olecranon crests are not as well defined but <;loser together and the medial crest is not high- er than the lateral crest. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 69 B Figure 54. Scalopoules ripafodwtor n. sp. Quartz Basin, UO loc 2465, A; VO 22415, left clavicle, anterior and posterior views, B; UO 22416, right clavicle, anteroventral view and anterior view of ventromedial spine. I I Figure 56. Scalopoules ripafodiator n. sp. Quartz Basin, UO loc 2465, UO 22428, left radius, medial, lateral and posterior views. Of the moles available for comparison, the scaphoid of S. ripafodiator (Fig. 57) most closely resembles that of Parascalops in mor- phology and general proportions. There are three notable differences: the groove for the tendon of the M. flexor carpi radialis is bridged by bone to form a short tunnel, the facet for the os falciformis is more distally situated like that in Scapanus, and the posterior crest is rel- atively broader. Figure 55. Scalopoides ripafodwtor n. sp. Quartz Basin, UO loc 2465, UO 22366, left humerus, anterior and posterior views, unshaded areas with solid lines restored from other specimens. 70 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 TABLE 14 COMPARATIVE MEASUREMENTS OF THE CLAVICLES OF Scalopoides, Neurotrichus, Condylura, AND Scapanulus Scalopoides isodens UK 15570 ........ ... ... .. .... ....... .... ...................... . UK 15571 ... ................ ........ ... ......... ....... ....... . Scalopoides ripaf odiator uo 22415 ··············· ···· ········· ····· ·····--·············· uo 22416 ....... ... ... .. ...... ...... ......................... . . uo 22417 ···· ············ ···· ······· ················ ······ ····· uo 22427 .............. .............................. .. ..... .. . uo 22431 ············· ·· ·· ········ ······ ······················· uo 22436 ... .... ..... ............... .......................... . uo 24840 .. .. ...... ..... ...... ............ ..... ........ ....... . Scalopoides sp. A uo 24 761 ··· ········· ·········· ··· ······ ············ ··· ·· ·· ···· ? Scalopoules sp. C uo 25106 ·· ········ ··· -··· ·· -·- ······ ····· ···············------ uo 25107 ·· ····· ······-·-··· ··· ····· ················ ··········· N eurotrichus gibbsii ... .... ....... .... ..... .. .... ............... . . Condylura cristata ..... ... .. ..... ...... ..... ..... .. .... ...... .... . Scapanulus oweni .......... ..................... .......... ... ..... . e Estimated measurement. Maximum length 3.60 3.98 4.50 3.61 3.17 4.21 3.69 3.24 3.58e 2.80* 2.96* 4.08 6.00 3.52 Ventral process length l.80e 2.88 2.13e 2.85 1.95 l.45e 3.17 3.12 2.27 Process lg. max. lg. % 50 64 59 54.5 78 52 65 '' Noticeable breakage or abrasion tending to reduce measurement. The third metacarpal ( Fig. 58), repres~nted by three specimens in the Quartz Basin sample, is more elongate than Parascalops and Scalo- pina but shorter than in the Urotrichina. The proximal tuberosities are more poorly defined than in Parascalops and the Scalopina but are specialized toward these types. The ungual phalanges are like those of S. isodens. In the innominate there are three vertebrae fused to the ilium as in all the other Talpinae but there is no bony connection between the acetabular region and the fourth sacral verte- bra as in the Scalopina. In general robustness and other details preserved in the fragmentary innominata in the sample, S. ripafodiator agrees well with the other Parascalopina. The remains of the femur (Fig. 59A) are very fragmentary. One specimen of the proxi- mal extremity (UO 22395) closely resembles that region in Scapanulus and agrees closely with the expected size of the femur for S. ripa- fodiator. The majority of the specimens are similarly referred to this species on the basis of the relative abundance within this locality. The tibiofibula is represented only by proxi- mal and distal end fragments. The distal end of the tibiofibula (Fig. 59B) while closely re- sembling the common scalopine type has a . . : . I 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON TABLE 15 COMPARATIVE MEASUREMENTS OF THE HUMERI OF Scalopoides AND Scalopoides-LIKE MOLES FROM OREGON AND COLORADO Specimen S. isodens Martin Canyon Quarry A UK 10081 ............. .. ....... ... ......... ...... ... ........... . . Total length UK 10084 ............... ....... ...... ......... ...... ...... ....... 9.52 UK 10085 .... ..... ...... .... ... ... ...... ..... ... ............ ..... 8.96 UK 10086 .. .... ........ ................. ... .. .... ....... ........ . UK 10087 ................................ ........ ... ............ . UK 10088 ...... .. .... .. .. ......... ...... ....... .... .. ....... .... . UK 10089 .... ....... ... ... .. ...... ................ .............. . UK 10090 .... ........... ...... ....... .... .................. ..... . S. ripafodiator Quartz Basin uo 22355 ··············· ······ ··········· ····· ········ ··········· uo 22361 ··················· ······ ·· ···· ···· ················· ···· uo 22366 ···· ·························· ·············· ············ 9. 75 uo 22370 ·· ··· ··· ···· ······· ············ ················ ··· ······ uo 22377 ........ ..... .................. ..................... ... . uo 22387 ··········· ········ ·· ·········· ········ ···· ········ ····· 9.23 S. cf. S. ripaf odwtor Red Basin uo 24343 ········ ······················· ···· ····· ············ ···· uo 24249 ··· ····· ················· ···· ··········· ············ ···· Guano Lake USNM 23765 ...... .............................. .... ....... . USGS loc. M. 1041 USNM 23752 .. ............. ............... .......... .. ..... .. . USNM 23754 ... ................ ........ .. .... ................ . USNM 23755 ............. ...... .............. ..... ... ........ . USNM 23756 ....... .. ...... .... ...... .... .. .. .. ..... ......... . USNM 23757 ... ...... ... ......................... ... ......... . Scalopoides sp. A Black Butte uo 24764 ································ ······················· · uo 24765 ··············· ····· ····················· ······· ····· ··· Scalopoi.aes sp. B Black Butte uo 24784 ······ ······ ····· ·············· ················ ········· ? Scalopoides sp. D McKay Reservoir uo 24817 ... .. ........ .. .......... .. ................. .. ... .. ... . . ? Scalopoides sp. E McKay Reservoir uo 22606 .. ........ ... ................ ... ... ......... ............ 7.81 • Noticeable breakage or abrasion tending to reduce measurement. e Estimated measurement. Prox. width 6.31 6.23 6.09 Shaft width 2.06 2.31 2.31 2.29 2.16 2.28 2.01 2.33 2.26 2.17 2.24 2.26 2.2le 2.06 2.15 1.40 1.85 1.85 1.80 2.24e 1.66 2.18 1.73 Distal width 4.91 4.95 4.95 5.18 4.74 5.30 4.52 4.54 4.62 4.33* 4.56 3.91 3.87 4.50 3.20 4.95 4.10 3.67 4.73 71 72 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 I Figure 57. Scalopoides ripafodiator n. sp. Quartz Basin, UO loc 2465, UO 22446, right scaphoid, distal, dorsal, and lateral views. Figure 58. Scalopoides ripafodiator n. sp. Quartz Basin, UO loc 2465, UO 22439, left third meta- carpal, dorsal and ventral views. B Figure 59. Scalopoides ripafodiator n. sp. Quartz Basin, UO loc 2465, A; UO 22398, distal end of right femur, posterior view, B; UO 22471, distal end of left tibiofibula, anterior and posterior views. I Figure 60. Scalopoides ripaf odiator n. sp. Quartz Basin, UO loc 2465, UO 22401, left astragalus, dorsal and ventral views. rather deep groove for the peroneous tendons as in Condylura. The fused portion of the tibio- fibula is relatively long compared to the known Scalopini. The ossified portion of the trans- verse crural ligament is usually quite promi- nent and the ligament may ossify for more than half its length in some individuals (UO 22470). Of the three proximal epiphyses (UO 22481, UO 22475, UO 22479) assigned to this species on size, none of the significant diagnostic features of the falciform process are well enough preserved to enable a meaningful comparison with the other moles. The calcaneum and astragalus are not sig- nificantly different from that in other known scalopines to warrant description. Unlike the other genera, the astragalus (Fig. 60) has a very broad neck as in Scalopus. RELATIONSHIPS: Wilson ( 1960) in his discus- sion of the relationships of Scalopoides isodens, compared Scalopoides with various European genera and concluded that although it could be separated from the better known European genera, unidentified, or incompletely identified, specimens in the Sansan and La Grive faunas may be referrable to Scalopoides. He notes however that such similarity is only observable in the molars since the antemolar dentition was not preserved in the sample he examined. Con- vergence in molar form or in many of the limb bones is not uncommon in the talpids. Although re-examination of described and undescribed European materials may yet reveal similarities between North American and European tal- pids at the generic level, the differences cited by Wilson between S. isodens and various de- scribed European genera will also serve to sep- arate S. ripafodiator from these forms. 1968 HUTCHISON : FOSSIL T ALP/DAE FROM OREGON 73 TABLE 16 COMPARATIVE MEASUREMENTS OF THE RADII OF Scalopoides, Condylura AND N eurotrichus Shaft Distal Capitular length (1) Scalopoides isodens Martin Canyon UK 15579 ..... ... ........ .............. .... 7.70 UK 15580 ... .......... ......... .......... .. 7.85 UK 15581 .................................. UK 15583 .. ........ ..... ..... .... ...... .... UK 15584 ................. ........... .. ... . UK 15585 -- -···· ···········---···· ··· ····-- Scalopoides ripaf odiator Quartz Basin uo 22428 ·· ········· ·· ··· ················· · 8.40 uo 22421 .. .............. ...... ........ ... . 7.34 uo 22433 ··· ····· ·· ·· ············ ··· ······· 7.85 uo 22424 ..... ....... .... .................. uo 22425 ··· ············ ····· ············· · U022429 ... ... ........ .. ... ... .......... .. U026800 ................ ................. . uo 22420 ..... ...... ... .............. ...... uo 22422 ···· ·· ·········· ····· ···· ······ ··· uo 22423 ·· ········ ········· ·············· · uo 22434 ···· ··· ···················· ·· ··· ·· Scalopowes cf. S. ripafodiator Red Basin uo 24337 ···· ··· ··· ······ ··· ··· ······ ······ 7.47 uo 24339 .......... .......... ........ ..... . uo 24340 ···· ······· ····· ·················- uo 24341 ........ ..................... ..... uo 24342 .... ....... ... ..... ... ...... ... ... N ezirotrichus gibbsii ·· · ·-·- ·- ------······ · ·· 7.55 N eurotrichus gibbsii ·· · · · --- ------------···· 8.40 N eurotrichus gibbsii -----------------······· 7.84 Condylura cristata ............................ 11.55 '' Noticeable abrasion tending to reduce meagurement. While the conservative nature of the denti- tion and relatively long clavicle with converg- ing planes tend to align Scalopoides with the Urotrichini, the more fossorially specialized humerus, radius, ulna, and manubrium tend to situate it adaptively nearer Condylura and width process 3/1 2/ 1 (2) (3) % % 2.55 0.85 11 33 2.40 0.55* 31 2.05 2.30 0.70* 0.85 2.23 0.76 9 27 2.14 29 2.07 26 0.75 0.77 0.72 0.74 2.21 2.04* 2.06 2.18 1.92 0.74 10 26 2.12 2.03 0.85 0.75 1.57 0.55 7 21 1.71 0.53 6 20 1.84 0.53 7 23 2.84 0.96 8 14 Scapanulus. The details of humerus ( "scalo- pine" ridge, angle of the head, angle of the clav- icular articular facet, etc.), manubrium ( ver- tical anterior margin), clavicle ( evidence of heterotopic bones, simple construction), ulna (prominent coronoid process, angle and detail 74 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 of distal articular surface), radius ( scalloped distal margin) , metacarpals ( robust and rela- tively simple construction), and generalized dentition and premolar construction serve to segregate Scalopoides from the Condylurini and align it with the Scalopini, especially the Parascalopina. Scalopoides is readily distinguishable from all the living North American Scalopini in its less specialized foreleg, double-rooted pre- molars, and a number of smaller details. In foreleg specialization, it compares better with the only Eurasian living scalopine Scapanulus than with any of the North American forms, however, this is primarily an adaptive similar- ity. Scapanulus closely approaches Scalopoides in the construction of the lower molars but lacks the entocristids. The incisors are similar in form and the fa is enlarged in Scapanulus to approximately the same degree as in Scalo- poides; however, Scapanulus retains a com- plete lower dental formula and all the pre- molars are single-rooted. The upper molars of Scapanulus (Fig. 5) also differ from Scalo- poides in the prominently twinned mesostyles and relatively smaller protocones. On the hu- merus the teres tubercle and clavicular artic- ular facet are relatively larger in Scapanulus. Scalopoides does not exhibit the twisting of the greater tuberosity and head medially as in Scapanulus and Parascalops (Figs. 10 D, E). The clavicle of Scapanulus while showing many similarities to Scalopoides in overall morphol- ogy and the presence of a facet for a hetero- topic bone is relatively shorter than that of Scalopoides and the planes of articular facets are more nearly parallel. The articular planes of the clavicle of S. isodens and S. ripafodiator converge both posteriorly and ventrally to a greater degree than in any of the living Scalo- pini. The manubria of Scalopoides and Scapan- ulus strongly resemble each other in the slight degree of lateral expansion of the anterobasal extremity. The heterotopic bones of Scalo- poides, however, appear to have been much smaller than those in Scapanulus. Parascalops is a step further removed from Scalopoides in all of the above features. More- over, in Parascalops upper molars, the meta- conule and paraconule are very prominent and form a nearly rectangular lingual shelf. In the lower molars Parascalops has distinct meso- styles separated from the metaconids. The an- temolars are all present and all the premolars are single-rooted. Also, the clavicle is much shorter and the humerus much broader than in either Scapanulus or Scalopoides. The base of the manubrium is expanded anteriorly with consequent reduction of the heterotopic bones. Scapanus, while seemingly remote from Scalopoides, differs in some ways to a lesser degree from it than do Scapanulus or Para- scalops. The mesostyles of upper molars are only slightly twinned and the paraconule and metaconule are subservient to the large proto- cone. The humerus, although broader and shorter shafted than Scalopoides, does not ex- hibit the twisting of the head and greater tuber- osity of Scapanulus and Parascalops. The low- er molars are higher crowned than in Scal- opoides and lack metastylids. The lower ante- molar dentition is usually complete, and pre- molars are single-rooted but the molars have entocristids. Scalopus is still further removed morpholo- gically from Scalopoides owing to its much broader humerus, shorter clavicle,, radius, ulna, and metacarpals, greatly reduced denti- tion, and hypsodont teeth. At the species level, S. ripafodiator closely resembles S. isodens in most details. The two moles may, however, be segregated on a num- ber of features which I interpret as more primi- tive in S. ripafodiator. The longer and more robust antemolar region, double-rooted P~, peg- like l:i, and relatively longer humerus, ulna and radius all indicate the retention of a primitive feature and a more generalized species than S. isodens. Assuming that the relative time occurrence of these two species is correct, then S. isodens does not form an ideal ancestor for S. ripafodiator. That they shared a common ancestor within the genus Scalopoides seems quite probable; however, such an ancestral form seemingly would be Arikareean in age. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 75 Scalopoides cf. S. ripafodiator Several specimens of Scalopoides were re- covered from the Red Basin and southern Ore- gon localities. They are probably best referred tentatively to S. ripafodiator on gross features but differ from the type sample in their smaller size (Red Basin) and details of some teeth. All of the Red Basin specimens except the distal end of one tibiofibula come from UO loc. 2495. Specimen UO 24311 from Red Basin pre- serves two P 2 alveoli anterior to the P3 ; this single P3 has a small third root on the labial side between the two main roots. The P3-M1 differ in no consistant respects from the Quartz Basin sample except in their somewhat smaller size. The M2 may, on the average, have a slightly more distinct paraconule, slightly de- flated protocone and relatively narrower labial moiety than in the Quartz Basin Scalopoides. Only one Red Basin specimen, UO 24300, preserves a substantial part of the lower ante- molar region. Heavily worn C-M1 are pre- served, although the bone of the dentary, minutely fractured and splintered, may be dis- tored strongly. The dental formula as preserved is like S. ripafodiator and all the premolars are likewise double rooted. There appears to be a mental foramen under the Pa and a smaller one under the P2. The jaw is very shallow in the antemolar region and is similar to S. isodens but the paralophid of the M1 is bent as in S. ripafodiator. The shallowness of the jaw may be due in part to the advanced age of the indi- vidual or post depositional distortion ( a com- mon phenomenon in this locality) or both. The relationship of the posterior cingulum of the P4 to the protoconid is more like that of S. iso- dens than S. ripafodiator. In another specimen, UO 24301, from the same locality, the Pa-P4 are well preserved. A mental foramen is pres- ent beneath the P4 in this specimen. Again the P4 is very similar to that of S. isodens, although the paraconid is not as well defined and the protoconid is not as angulate. The molars are like those of S. ripafodiator except in size. Two specimens from Guano Ranch pre- served in one the PrM1 (USNM 23766) and the M1-M2 in the other (USNM 23767). An- other specimen (USNM 23761) from USGS loc. 1041 preserved the M2-Ma. The molars are like those of S. ripafodiator but the one P4 is relatively larger and longer than those from Quartz Basin. The morphology of the tooth re- sembles that of the Red Basin sample more than the type material. No complete humeri were recovered from Red Basin. The fragmentary ones-mostly dis- tal ends-are smaller than those of S. ripa- f odiator but are not otherwise easily com- parable to any of the better known Scalopoides in detail. The distal or mid-portions of six humeri (USNM 23752-USNM 23757) from USGS loc. M 1041 and one from Guano Ranch (USNM 23765) agree closly with those of S. ripafodiator in detail. One specimen (USNM 23752) is smaller (Table 15) than all the others and is comparable in size to the small Scalopoides from Black Butte to which it may be related if there are indeed more than one species of Scalopoides in this sample. Aside from Quartz Basin other foreleg ele- ments of Barstovian Scalopoides from Oregon are known only from Red Basin. The few ulna fragments are not observably different from those of S. ripafodiator from Quartz Basin. Figure 61. Scalopoides cf. S. ripaf odiator, Red Basin, UO loc 2495, UO 24376, left second meta- carpal, posterior view. The one complete and five partial (UO 24337 -UO 24342) radii, although smaller, are similar to those of the hypodigm. The sea phoid ( U O 24364, U O 24365) , meta- carpal III (UO 24379), ungual phalanx of the manus (UO 24374), and astragali (UO 24369 -UO 24370) are also like those from the Quartz Basin sample. The metacarpal II (UO 24376, Fig. 61) which was not represented in ' . ,. . . . 76 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 the other sample is shorter than the metacarpal III from Quartz Basin and relatively narrower and longer than the metacarpal II in Para- scalops and the Scalopina but considerably shorter than that in the Urotrichini or Condy- lura. The proximal tuberosities are damaged but better developed than in the Urotrichini. The distal ends of the tibiofibula from Red Basin are essentially identical to those from Quartz Basin except for smaller size. However, the crural ligament is extensively ossified in the majority of the specimens from the Red Basin collection. The three distal ends of the tibiofibula (USNM 23758-USNM 23760) from USGS loc. 1041 fall within the range of S. ripafodiator in morphology and size. Two extreme alternatives dealing with the above specimens are that at least three species are represented in the Barstovian samples of Scalopoides from Oregon or that there is con- siderable interdeme variation ( including ge- ography and time) particularly in regard to the P4. The latter view is accepted here until such time that larger samples from these and other localities are available to adequately appreciate the population differences which may be used in differentiating grossly syn- chronous and geographically proximate popu- lations or species. All the above material is thus referred to Scalopoides cf. S. ripafodiator until further comparison is possible. Scalopoides sp. A (Figs. 62, 63) Several specimens of a Scalopoides-size mole were recovered from the Black Butte locality, UO loc. 2500. They are grouped together under a single heading largely on the basis of size and similarity to better known Scalopoides. Nearly the entire proximal half of the hu- merus (UO 24765) and fragments of the shaft (UO 24764) and condyle (UO 24766) of two others attest to the presence of a Scalopoides- like mole, about the size of S. ripafodiator and S. isodens, in the Black Butte fauna. Although similar to the Miocene Scalopoides in most comparable details, the narrow groove between B A Figure 62. Left clavicles, posterior view, A; Scalo- poides sp. A, Black Butte, UO loc 2500, UO 24761 ( reversed right clavicle), B; ? Scalopoides sp. C; Bartlett Mountain, UO loc 2517, UO 25106, dorsal prominence restored from UO 25107. the head and clavicular articular facet is oblit- erated. There is some slight difference in the shape of the lesser tuberosity anteriorly al- though this is variable in structure; and the margin of the teres tubercle appears to be better aligned with the axis of the humerus. The pectoral muscle scars are too poorly pre- served for comparison. A single clavicle (UO 24761, Fig. 62B) lacking the ventral border of the ventral process and the clavicular articulation represents a mole about the size of that expected for the humeri described below. The clavicle repre- sents a mole about the size of S. ripafodiator except that it is markedly shorter. Most of the shortening compared with that of S. ripafodi- ator occurs in the lateral moiety of the shaft, thus the ventral process forms a short plate rather than a long blade. In general appearance this clavicle resembles the clavicle of Scalopini sp. A from Red Basin discussed below. There is no prominent reflection of the ventromedial spine for a heterotopic bone. The relative short- ness of clavicle is comparable to that in Scap- anulus. 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 77 l A c Figure 63. Scalopoides sp. A, Black Butte, UO loc 2500, A; UO 24 759, proximal end of left radius, me- dial and posterior views, B; UO 24774, left first metacarpal, dorsal view, C; UO 24760, left scaphoid, lateral, dorsal and distal views. Two ulna fragments (U024767, U024768) preserve neither extremity but resemble those of S. ripafodiator in comparable features. A specimen of the proximal end of a radius (UO 24759, Fig. 63A) differs from the Miocene species of Scalopoides in that the capitular process is drawn out more and supported as in Parascalops by a strong rib posteriorly result- ing from the excavation of the shaft below the ulnar articular facet for the origin of the M. abductor pollicis longus as in Scapanus. The ulnar articular facet is elongated and well re- moved from the body of the shaft. The scaphoid (UO 24760, Fig. 63C) re- sembles that of Parascalops even more closely than that of S. ripafodiator in the more prox- imal position of the os falciformis attachment, unbridged groove for the tendon of the M. flexor carpi radialis and the narrower posterior crest. However, there are a number of small diffi- cultly described proportional and morpho- logical differences which prevent this form from being a perfect miniature of Parascalops. The two small first metacarpals (UO 24774, UO 25181) are definitely those of a burrowing mole. They are as relatively compact as in Parascalops and most closely resemble this genus except that the lateral moiety of the distal articular surface is still more distally extended (Fig. 63B) than the medial one and not equal as in Parascalops or longer than the medial one as in the Scalopina. The three prox- imal phalanges (UO 24763, UO 24777, UO 25180) representing possibly as many digits of the manus, strongly resemble the phalanx of Scalopini sp. A from Red Basin in size and pro- portions, although, the Black Butte specimens are still slightly but relatively longer (Table 17). The middle phalanx ( UO 26803) of the manus is short and of the burrowing type. The ungual phalanges of the manus (UO 24769, UO 25179) are like those of other Scalopoides. The bones of the manus indicate a mole better adapted for digging than S. ripafodiator or S . isodens which is consistent with the short- ened clavicle. The bones of the manus could belong to a small Domninoides such as that . 78 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 reported from Fish Lake Valley by Clark et al. ( 1964), but comparable elements of Domni- noides are not recorded and the genus is un- known from the Juntura Formation on un- equivocal material. The general agreement in expected size for the humerus of the normal Scalopoides-like mole seems to make this assignment the more reasonable one at present. TABLE 17 MEASUREMENTS OF THE PROXIMAL PHALANGES OF VARIOUS Scalopoides-LIKE MOLES Length Length Width Width Scalopoides isodens UK 15586 ____________ 1.37 1.30 1.05 Scalopoides sp. A uo 24763 ------------ 1.43 1.20 1.19 uo 24777 ------------ 1.57 1.30 1.21 uo 25180 ------------ 1.72 1.13 1.52 ? Scalopoides sp. C uo 26807 -----------· 1.40 1.27 1.10 ? Scalopoides sp. D uo 27144 ·---········ 1.50 1.13 1.33 Scalopoides sp. B A single humerus (UO 24784) from Black Butte loc. 2500 lacking most of the proximal end except the teres tubercle would seem to be well beyond the range of variation in size rep- resented by Scalopoides sp. A from the same locality. The pectoral scars are not sharply defined but do not differ greatly from those of S. ripafodiator. ? Scalopoides sp. C (Figs. 62, 64) A few postcranial elements of a small mole about the size of Scalopoides ripaf odiator were recovered from the Bartlett Mountain local- ity. Two scapula fragments (UO 24800, UO 26805) resemble the scapulae of Parascalops and the Miocene Scalopoides. The infraspi- natus groove while well defined is narrower than in S. ripafodiator but quite similar to that in Parascalops. Two incomplete clavicles (Fig. 62A), one (UO 25106) lacking the tip of the ventral process and all of the dorsal prominence and the other (UO 25107) all of the ventral margin, are very similar in size and morphology to the clavicle of Scalopoides sp. A except that the shortening of the lateral moiety is still greater. The fourth metacarpal (UO 25108, Fig. 64) is narrower and propor- tionately longer than that of Parascalops but generally resembles it in structural detail. The proximal phalanx (UO 26807) of a manus digit is nearly identical to UO 24 763 from Black Butte except it is very slightly short- er. Two medial phalanges (UO 26808, UO 26809) and two ungual phalanges ( U O 24 798, UO 26810) of the manus are also like those from Black Butte. The proximal end of the femur (UO 26806) is comparable to that of other Scalopoides. The astragalus (UO 26812) is like that of S. ripafodiator but the neck is narrower as in Scapanus. Figure 64. ? Scalopoides sp. C, Bartlett Mountain, UO loc 2517, UO 25108, right fourth metacarpal, posterior view. Because of the great time gap between the better known Scalopoides ( Barstovian) and the Hemphillian moles, assignment of various elements to Scalopoides becomes more tenuous. Nevertheless the similarities in morphology of the above postcranial elements and general agreement in size still point to closer rela- tionship to Scalopoides than to other com- parably represented moles. ? Scalopoides sp. C and ? Scalopoides sp. D may be synonymous but since they are each known from different elements and are nearly four hundred miles apart geographically, it seems best to treat them separately. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 79 ?Scalopoides sp. D (Fig. 65) The distal portion of a humerus (UO 24817) and a shaft fragment of another (UO 25128) indicate the presence of a Scalopoides grade of mole about the size of S. ripafodiator in the McKay Reservoir locality. The shaft is more robust and the anterior orifice of the entepi- condylar foramen is more extensively pocketed than in the other species of Scalopoides. The pectoral crest terminates more proximally than in S. ripafodiator. This humerus appears to be somewhat better fossorially specialized judging from its robust- ness and pectoral crest position. This specimen and the jaw and teeth provisionally assigned to "Neurotrichus" may represent a single species. This possibility is discussed further under ?Neurotrichus columbianus. The proximal phalanx of the manus (UO 27144) assignable to this or the following species, is intermediate in proportions between Parascalops and Scalopina sp. A. ?Scalopoides sp. E (Fig. 66) Four specimens from McKay of a Scalo- poides-like mole appear to be too small to assign to ?Scalopoides sp. D. The humerus ( UO 22606) is complete except for the radial condylus, head, deltoid process, and part of the lesser tuberosity. It resembles S. ripafodi- ator in most details except on the obliteration of the groove between the head and clavicular facet, configuration of the pectoral muscle scars, more pocketed anterior orifice of the en- tepicondylar foramen, and small size. There is a vertical raised ridge anteriorly in the middle of the shaft which is part of the pectoral muscle scar complex. This is not prominently developed or present on the other Scalopoides- like humeri seen, although its expression is probably age dependent. Part of the semilunar notch and shaft of the ulna (UO 26819, UO 27145) agree in the detail of the distal margin the semilunar notch Figure 65. ? Scalopoides sp. D, McKay Reservoir, UO loc 2222, UO 24817, shaft and distal end of left humerus, anterior view. I f I I I I I ' \_) l ,, 1. \ I ' I I I \ Figure 66. ?Scalopoides sp. E, McKay Reservoir, UO loc 2222, UO 22606, right humerus, anterior view. with Scalopoides especially S. ripafodiator and Scalopoides sp. A rather than N eurotrichus or Urotichini. These specimens could be part of Scalopoides sp. D but size seems to favor a smaller mole. This represents the second or third occur- rence of a small Scalopoides-like mole with a larger Scalopoides-like mole. Each of these 80 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 joint occurrences is a small sample, so that it is difficult to determine whether there is a great deal of size variation within a single species or two species. The range of variation of the humeri in the Quartz Basin, Red Basin, and Quarry A samples suggest that the latter situa- tion is the case. Summary of Relationships of the Scalopoides-like Moles The number of good locality samples is too small to make any concrete statements on the phylogeny within the Scalopoides complex. The increased know ledge of the postcranial skeleton does align Scalopoides with the more specialized moles, the Scalopini, rather than with the urotrichine moles as Wilson (1960) postulated. Structurally, Scalopoides, espe- cially S. ripafodiator, is very close to a hypo- thetical ancestor which could have given rise to all the living scalopine moles. The reduced dentition of all the adequately known Scalo- poides species eliminates any of these forms from the direct line of descent of any of the living scalopines. Proportions of the limb bones indicate that S. ripafodiator is either structur- ally more primitive than S. isodens and the Pli- ocene species, or slightly better adapted for fac- ultative semiaquatic behavior, perhaps both. When arranged in chronological order, S. iso- dens and the postcranial elements assigned to Scalopoides from the Pliocene, suggest that in general Scalopoides became progressively better adapted for burrowing with time. This may have occurred at various times in different regions and thus may not be used as a means of correlation outside the Great Basin until equally good or better sequences can be dem- onstrated in other areas. DOMNINOIDES Green, 1956 Domninoides ( sensu lato) is turning up with increasing frequency in middle and later Ter- tiary deposits of North America, but little is yet known about its systematic position with respect to the living genera. Only Tedford ( 1961) has attempted a careful comparison of Domninoides with the Recent American genera. He concluded, rightly, that Domninoides is more closely related to Parascalops than to any other Recent American mole. PROVISIONAL DIAGNOSIS: Burrowing moles with reduced antemolar dentition; P 4 and P 2 double-rooted; Pa single-rooted. Lower molars moderately high crowned ( mesodont), post- fossids open lingually and crista obliqua of M2-Ma joining metastylid as in Parascalops, metastylids not as posteriorly situated as in Parascalops. Humerus Parascalops-like in mor- phology and proportions, except in the pectoral muscle scar V-configuration. Domninoides sp. (Figs. 67-69) Domninoides is known from teeth in two Barstovian localities in Oregon, Guano Ranch and Quartz Basin. Two incomplete jaws each containing the M1-M2, USNM 23768 andUSNM 23769 (Fig. 67), from Guano Ranch closely approach the Scalopoides species from the same locality in size. They may be distinguished from the Scalopoides on ( 1) the absence of an entocristid, ( 2) the extension of the crista obliqua of the M2 to the metastylid and its more lingual extension on the Mi, ( 3) and relatively Figure 67. Domninoides sp. Guano Ranch, USGS loc M1042, USNM 23769, right dentary fragment with M1 -M2, occlusal and labial views. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 81 I Figure 68. Domninoides sp. Quartz Basin, UO loc 2465, UO 22333, distal end of left humerus, anterior and posterior views. wider molars. All of these features are charac- teristic of Domninoides. A single, somewhat eroded M~, UO 22498, from Quartz Basin closely resembles the Guano Ranch forms in size and morphology. Several postcranial elements, mostly incom- plete, of a burrowing mole were also recovered from Quartz Basin. They are included here largely on the grounds of expected relative size and the lack of any conclusive evidence to the contrary that there is another burrowing mole in the fauna. The distal end of the hu- merus, U O 22333 ( Fig. 68), has a very stout and inflated shaft. The pectoral tubercle is very well developed and elongate as in the Scalo- pina, but the lateral edge of the tubercle is more vertically oriented than in Scapanus or Scalopus. Unfortunately, the actual surface of the tubercle is eroded to the extent that the de- tails of the muscle scars are obscured. The teres tubercle does not approach the epicondyle as closely as in Scapanus and Scalopus and is thus more like the condition in Parascalops, Domninoides, and Scalopina sp. A. TABLE 18 MEASUREMENTS OF THE LOWER TEETH AND JAWS OF Domninoides SP. U022498 USNM23768 USNM 23769 M1: Length ..... ........ ................. . . 2.00 2.13 Width ........ ...................... .. .. 1.37 1.40 M2: Length .. .. .. .. ...... .... .............. 2.36 2.30 2.23 Width ................ .......... ...... .. 1.20* 1.47 1.50 Depth of jaw below M1 metaconid .... .... .. ........ .. .. 1.87 1.73 '-' Abraded. 82 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 The fragment of the semilunar notch re- gion of the ulna, UO 22334, most closely re- sembles that of Parascalops in the shape and position of the articular facet for the radius and the M. abductor pollicis longus scar. The dorsal portion of the olecranon process imme- diately behind the processus anconeus is rel- atively narrower than in the Parascalops. Two radius fragments, UO 22340 and UO 22335, in combination lack only the distinctive capitu- lar process. The radius is relatively shorter than in Parascalops. The length of the shaft of UO 22340 is 7.2 mm. The metacarpal I, UO 22338, and metacar- pal IV, UO 22341, are relatively shorter in re- lation to width than in Scalopoides but longer than in Parascalops. Morphologically the meta- carpal I (Fig. 69A) closely resembles Para- scalops except for the presence of a shallow groove on the dorsal side immediately distal to the proximal articular facet. The metacar- pal IV (Fig. 69B) has a smaller articular facet for metacarpal V than in Parascalops which results in a more angulate proximolateral ex- tremity. The medial phalanx, UO 22337 (Fig. 67C), of the second or third digit of the man us is relatively short and resembles Parascalops, Scapanus orarius and S. townsendii in propor- tions. An ungual phalanx, UO 24835, was also recovered. RELATIONSHIPS: The dentulous elements fall within the reasonable variation in size and mor- phology of that of D. riparensis or of the D. cf. riparensis reported from the Clarendonian Fish Lake beds of Nevada. Differences in the configuration of the antemolar teeth will prob- ably prove to be useful in delineating the spe- cies of Domninoides, but this region is not pre- served in any of the Oregon specimens. The specific relationships of D. platybrachys will be in doubt until more material is found. specific reference to these or any of the other Barstovian and Clarendonian records of Dom- ninoides seems unwarranted in the light of the meager material thus far published. The postcranial elements pose different problems. More than one burrowing mole may A B c Figure 69. Domninoides sp. Quartz Basin, UO loc 2465, A; UO 22338, metacarpal I, dorsal and ven- tral views, B; UO 22341, metacarpal IV, dorsal and ventral views, C; UO 22337, left proximal phalanx of digit II or II, dorsal and right lateral views. be represented in the Quartz Basin. Some of the features of the humerus and radius suggest possible relationships to the Scalopina while other elements more closely resemble those of the Parascalopina. If the association of these elements with each other and the Domninoides tooth is valid, then we are dealing with a small burrowing mole which is more specialized in some aspects of the humerus and shortness of the radius and phalanges than Parascalops but less specialized in the relative shortening of the metacarpals. Subtribe Scalopina (Van Valen, 1967) Winge, 1917 No metastylids on lower molars; bulla ex- panded, stapedial tube strongly ossified; manu- brium without dorsal ridge, anterior vena cava divided posteriorly; no heterotopic bones be- tween clavicles and manubrium; humerus 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 83 Figure 70. Scapanoscapter simplicidens n. sp., n. gen., Red Basin UO loc 2495, Type specimen UO 24286, left mandible with 13 , P3, talonid of M1, and M2-M3 , labial and occlusal views. broad, pectoral ridge terminates medially, area surrounded by pectoral ridge, pectoral crest, and greater tuberosity shallow; capitular pro- cess of radius rounded or pointed. SCAPANOSCAPTER12 n. gen. GENOTYPE: Scapanoscapter simplicidens, n. sp. KNOWN DISTRIBUTION: Late Miocene ( Barsto- vian) of Oregon. DIAGNOSIS: Scalopine mole about the size of Scapanus latimanus with dental formula: ? ? ? ? . 12 is not hypertrophied, P 1-P 4 double- 3l43 rooted, prominent heel on all (?) lower pre- molars. Molars brachyodont, trigonids antero- posteriorly compressed. M1 metacingulum well-developed and extends posteriorly around base of metacone. Scapanoscapter simplicidens13 n. sp. (Figs. 70-72) TYPE: UO 24286, fractured left mandible with heavily worn P1, talonid of M1, and M2-Ms; mandible broken just posterior to beginning of ascending ramus and anterior to middle of the 12 alveolus. HYPODIGM: uo 24289, uo 24290, worn and unworn Mi,s; UO 24291 lingual moiety of M2 (?) ; UO 24287, UO 24288 trigonid of M2, unworn M1; UO 24294 proximal end of scap- 12 From the Greek skapane, digging tool, and skapter, digger. 13From the Latin simplis, simple, and dens, tooth. ula; UO 24292, ulna fragment; UO 24296 scaphoid; UO 23295, third metacarpal; UO 24285, UO 24377, one proximal phalanx, two medial phalanges and an ungual phalanx of manus; UO 24386, astragalus. All from type locality. REFERRED SPECIMEN: USGS loc. M 1040; USNM 23751, humerus lacking ulnar articular reg10n. TYPE LOCALITY: Red Basin, uo loc. 2495. AGE: Barstovian. DIAGNOSIS: Same as genus. DESCRIPTION: The dentary of Scapanoscapter (Fig. 70) is shaped like that of Scapanus (Scapanus), i.e., the Recent species. There are two mental foramina, one ventral to the P 4 , and the larger anterior one ventral to the PrP3 • The anterior tip of the jaw is broken but the dental formula was apparently complete. The jaw is broken through the 12 alveolus. The remains of the I2 alveolus do not indicate a notably en- larged tooth. The 13 is a peg-like tooth and the smallest tooth in the jaw. The alveolus of a larger, single-rooted canine follows the 13 . This alveolus is obliquely ovate. Of the premolars only the P1 is preserved but the alveoli of the remaining teeth indicate that the premolars in- creased regularly in size from the P1 to P 4 • All the premolars are distinctly double-rooted. The roots of the first two premolars are obliquely aligned in the jaw with the anterior root more labial. The crown of the P1 consists of a simple 84 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 Figure 71. Scapanoscapter simplicidents n. sp., n. gen., Red Basin, UO loc 2495, UO 24287, right M1, occlusal and lingual views. connate cusp centered over the anterior root. A simple heel forms the posterior moiety of the tooth. The M1 (UO 24287, Fig. 71) resembles S. shultzi in many respects but differs as follows: ( 1) it is lower crowned, ( 2) the trigonid is slightly less compressed anteroposteriorly, ( 3) a narrow cingulum is present around the an- terior and labial base of the protoconid. The M2 also closely resembles that of S. schultzi except that ( 1) a poorly defined metastylid and long tapering anterior cingulum are present, (2) the crista obliqua and entocristid more closely ap- proach each other anteriorly, and ( 3) there is no suggestion of an ectostylid. The M3 is heav- ily worn and damaged in the type but it also ap- parently resembled S. schultzi except in the presence of a narrow anterior cingulum and lack of an ectostylid. Figure 72. Scapanoscapter simplicidens n. sp., n. gen., Red Basin, UO loc 2495, UO 24289, right M1 (posterolabial part restored from UO 24289) , oc- clusal view. TABLE 19 MEASUREMENTS OF THE LOWER DENTITION AND ]Aw OF Scapanoscapter simplicidens. uo 24286 (Type) UO 24287 P1: Length ..... ............. ... . 0.76 Width ........ .... ... ... ... . 0.48 M1 : Length ..... ..... ....... .... . 2.46 Width ......... ..... ... ...... . 1.72 M2 : Length ....... ...... ........ . 2.33* Width .. ...... ......... ...... . 1.87 M3 : Length ............ ...... ... . l.83e Width .. ....... .... ... ...... . . 12-M3 alveolar length ... ... .. 12.1 P1-P4 alveolar length ... ..... 4.0 M1-M3 alveolar length .... .. 7.0 Depth of jaw below M1 .... . . 2.55 * Noticeable abrasion tending to reduce measurement. e Estimated measurement. The M1 (Fig. 72) resembles that of Scalo- poides more than Scapanus in most features other than size. It is relatively brachyodont with a distinct metaconule. The paraconule is a well developed cusp adjacent to the apex of the protocone in the unworn condition, and there is a paracingulum. In the position of the filiform root of Scalopoides there is a raised transverse ridge. The other roots are similar to those in S. ripafodiator. The living species of Scapanus do not possess prominences be- tween the roots of the molars. The M1 is about 2.8 mm. long and 2.64 mm. wide. Only the region immediately forming the glenoid cavity of the scapula is preserved. The acromion process appears to arise directly from the glenoid rim rather than removed from it by a small gap as in Scapanus and Para- scalops. The foramen for the suprascapular nerve is correspondingly nearer the glenoid border than in the aforementioned genera. However, in these features UO 24294 closely resembles Scalopus. The biceps tubercle ap- pears to arise directly from the glenoid rim. 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON TABLE 20 COMPARATIVE MEASUREMENTS OF THE HUMERI OF THE Scapanus-LIKE MOLES FROM OREGON Scapanoscapter simplicidens USGS Loe. M1040 USNM 23751 ...... ....... ............. ... ... ..... . Scapanus (Xeroscapheus) cf. S. (X.) shultzi Black Butte I U024632 .... ...... ...... .............. ... ... ...... ... . uo 24633 ······· ······ ··· ··· ········· ··· ··· ·········· West of Riverside uo 23075 ······· ·· ····· ····· ···· ·· ······· ··· ·· ···· ··· Loe. ? J untura Fm. uo 23074 ···· ·········· ·· ···· ····· ················ ··· S. (X.) proceridens Krebs Ranch II uo 8476 ······ ·· ·· ··· ··· ······ ······ ····· ······· ····· · uo 8474 ·· ··· ·· ··· ··· ··· ···· ······ ·· ··········· ······· uo 8475 ··· ···· ··· ····· ······· ······ ·········· ······ ·· uo 22603 ······· ··· ··· ·· ·· ····· ·· ·· ······· ·· ········ · Krebs Ranch I uo 4585 ·· ·· ······ ····· ····· ············ ······· ······ · W estend Blowout uo 7960 ..... ... ..... ......... ...... ...... ..... ...... . U05959 ....... ... .. .... ... ... ...... ..... ..... ....... . CIT 375 LACM 16765 ....... .... ... ....... ................ . Little Valley I uo 10816 ··· ··········· ····· ······· ·················· S. cf. S. (X.) proceridens McKay Reservoir uo 3506 .... ........... .............................. . Scapanus sp. Rome LACM 6499 ...... .. ... .... .. ..... ...... ..... .. .... . LACM6499 ....... ............................ .... . Scapanus nr ? S. latimanus Enrico Ranch uo 4585 ........... ...... .. ...... ..... .. ... .. ..... .. . . e=Estimated measurement. + = Broken measurement as preserved. Total length 12.5e 16.1 16.2 14.5e 14.6 16.2 Proximal width 10.0 11.7 12.1 10.5+ 10.9 7.7+ 12.1 Distal width 8.2 8.1 9.0 9.7 8.5 9.0 8.8 9.8 9.0 9.4 8.6 9.7 7.8 4.4 4.1 9.8 Shaft width 4.2 4.2 4.4 4.8 5.3 4.6 4.8 4.6 4.8 4.7 4.8 5.0 4.7 3.8 2.3e 4.8 85 86 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 The humerus from USGS loc. M 1040 lack- ing the ulnar articular region is about the size of that of Scapanus orarius and is nearly iden- tical to that of the living species of Scapanus in morphology and proportions except that the fossa for the brachialis muscle is not as deeply excavated. Reference to this genus and species is provisional. A fragment of the semilunar notch region of the ulna resembles that of the modern Scapanus except that the coronoid process is rather poorly developed. The scaphoid is broken and lacks all of the posterior and medial crests. The medial process is smaller than in Scapanus (Scapanus) but relatively larger than in Parascalops. The pit for the scapholunar tendon is small and re- stricted to the lateral moiety. As in Scapanus the central facet occupies a large majority of the radial trochlea distally. The metacarpal III is essentially the same as in S. townsendii and S. orarius in morphol- ogy and proportions except that the lateral proximal process is somewhat more heavily developed and thus resembles Parascalops in this feature. The third metacarpal is propor- tionately shorter than in Parascalops. The prox- imal phalanx and two medial phalanges of the manus (UO 24285 collectively) and un- gual phalanx are proportionately and morpho- logically like those of S. orarius and S. town- sendii. The astragalus is similar to that in S. (Scapanus) except that the tibial facet is more saddle-shaped as in Parascalops. Scapanoscapter simplicidens appears to be referable to the Scalopina on the construction of the talonid and close similarity of the teeth to those of S. shultzi. The upper teeth and ante- molar teeth suggest a very primitive member of the Scapanus complex which retains m a n y Scalopoides-like features. Postcranially the skeleton seems to have achieved a structural grade near that of S. orarius and S. townsendii. The double-rooted premolars are considered as a primitive character which will serve to dis- tinguish it from all the living and known fos- sil North American highly fossorial talpids. Temporally and functionally it is in a posi- tion to be a structural ancestor to either Scapa- nus and Scalopus, although the former is more probable based upon known distributions. SCAP ANUS Pomel, 1848 DIAGNOSIS: Scalopine moles with usually com- plete dental formulae. Clavicle usually not pierced by vena cava. Humerus not as broad as in Scalopus. Subgenus SCAP ANUS Pomel, 1848 GENOTYPE: Scalops townsendii Bachman, 1839. INC LUDED SPECIES: Scapanus townsendii (Bachman) 1839. S. orarius True, 1896. S. latimanus (Bachman), 1842. KNOWN DISTRIBUTION: Middle Pliocene (Hem- phillian) to Recent of Western North America. DIAGNOSIS: Medium to large moles with rela- tively long antemolar regions, P1-P4 single- rooted, I~ only moderately enlarged. Clavicle not pierced by vena cava. Scapula with infra- spinatus fossa extinguished well before reach- . . mg acrom10n process. Scapanus cf. S. (Scapanus) sp. (Fig. 73) Three teeth and some hand bones of a mod- erately large mole from Bartlett Mountain, UO loc. 2517, are probably referable to Scapanus. The lower Ma (UO 24803) is so heavily worn that the cusp morphology is obscured. The anterolabial accessory cusp is large and rec- tangular. The anterior cingulum extends part way along the paralophid from the accessory cusp. There is a labial cingulum bridging the hypoflexid. The tooth is therefore not referable to S. proceridens. The M3 is 2.13 mm. long and 1.47 mm. wide. An upper M2 (UO 24804) and M3 ( UO 26341) fall within the variation of the teeth of S. orarius or S. townsendii in every detail except for size in which they are inter- mediate between these two species. The meta- conule is not as isolated from the lingual base of the metacone by a deep notch as in S. lati- 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 87 Figure 73. Scapanus (Scapanus) sp. Bartlett Moun· tain, UO loc 2359, UO 9664, anterior portion of the manubrium, left lateral view. manus. The M2 is 2.58. mm. long and an esti- mated 2.4 mm. wide; the M3 is about 1. 7 mm. long and 1. 7 + mm. wide ( ectoloph damaged) . The anterior third of the manubrium (UO 9664) from Bartlett Mountain loc. 2357 closely resembles that of the living species except that the depth of the area for insertion of the M. pectoralis superficialis posticus is deeper, i.e., more extensive than in the modern forms. Two fifth metacarpals (UO 25102, UO 26935) are essentially identical to that in S. orarius. The three ungual phalanges of the manus (UO 25110, UO 26936, UO 26937) are also present in the sample. The sample is small but there is not a single morphological feature which would exclude the Bartlett Mountain species from inclusion in the suhgenus Scapanus or forming a possibly ancestoral species to the living forms. The Ms, in extreme wear, is not unlike S. shultzi except for the lack of an ectostylid. The material is too sparse to eliminate the possibility that these specimens represent a primitive member of Xeroscapheus, although their assignment to the subgenus Scapanus seems more likely to me. Scapanus nr.? S. latimanus (Fig. 10 F) An isolated humerus, UO 4585, from Enrico Ranch, Lake Co., is about the same size as that of the living S. latimanus and is otherwise not distinguishable from the humeri of Recent Scapanus. The range of S. latimanus includes the site of the fossil locality. The age of the Enrico Ranch material is in doubt but is at least as nld as Blancan. XEROSCAPHEUS14 n. subgen. SUBGENOTYPE: Scapanus (Xeroscapheus) pro- ceridens, n. sp. KNOWN DISTRIBUTION: Early Pliocene (Claren- donian) of California and Oregon, Middle Pliocene ( Hemphillian) of Oregon. DIAGNOSIS: Medium to large moles with hypso- brachyodont to hypsodont lower molars, and crowded antemolar teeth. Trigonids of lower molars anteroposteriorly compressed, P1-Ps single-rooted, P4 double-rooted or suggestion of fused roots, 12 enlarged. Scapula with a nar- row infraspinatus fossa which extends as a deep groove to the acromion process. Radius, ulna, and non-terminal phalanges relatively shorter than in living S. (Scapanus). Scapanus (Xeroscapheus) cf. S. shultzi MATERIAL: Black Butte II, UO loc. 2500: UO 24662, dentary fragment; UO 24749, about two-fifths of the proximal portion of the scap- ula; UO 24751, UO 24753, two humerus frag- ments; UO 24747, metacarpal IV; UO 24758, ossified palmer ligament; UO 24743, UO 24744, UO 24756, ungual phalanges of the manus; UO 24745, UO 24754, calcanea. Black Butte I, UO loc. 2337: UO 24632, UO 24633, distal ends of two humeri. West . of Riverside, UO loc. 2489: UO 23075, humerus lacking proximal end; UO 23076, distal end of radius. Juntura Formation, exact locality unknown; UO 23074, humerus lacking most of proximal end. About two-fifths of the proximal portion of the scapula was recovered. The acromion pro- cess originates slightly more anterior than in the modern species. The infraspinatus fossa continued forward as a deep gutter until pass- ing anterior to the posterior margin of the aero- 14 From the Greek xeros, dry, and skapheus, digger. 1 88 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 mion process. The biceps tubercle is developed just posterior to the glenoid rim as in Parasca- lops and is contiguous with it. In size and the above mentioned features, UO 24749 closely resembles a similarly preserved scapula from the Ricardo Formation in the UCMP collec- tions. The humeri closely resemble the modern Scapanus in details and proportions. The notch between the teres tubercle and the epicondyle is small and not like the relatively large notch in the specimen referred to S. shultzi from the Ricardo Formation. The somewhat eroded metacarpal IV is about at the structural grade and morphology of S. orarius, S. townsendii, and Parascalops. The proximal phalanx of the second or third digit of the manus and median phalanx are proportionately slightly shorter in relation to their respective widths than S. latimanus and approach Scalopus in this respect. The ossified palmer ligament, ungual phalanges of the manus, and calcaneum show no consistent dif- ferences from those of the living Scapanus. The described material of Scapanus shultzi is largely not comparable to the material from the Juntura Formation. The humerus referred to S. shultzi by Tedford (1961) is crushed and broken so that comparisons between the incomplete but well preserved humeri from Oregon are difficult. Additional material of Scapanus from the Ricardo Formation of Cal- ifornia is coming to light through screen wash- ing efforts of Mr. David Whistler at the Uni- versity of California at Berkeley. Although this material has not been studied, cursory ex- amination suggestions that the Oregon material may be referable to S. shultzi. Scapanus (Xeroscapheus) proceridens15 sp. n. (Figs. 74, 75) TYPE: UO 22508, right mandible with Iz, P1, M1-M2, but lacking ascending ramus and tip of horizontal ramus. HYPODIGM: UO 8474, UO 8476, UO 22603, one complete and three fragmentary humeri; 15 From the Latin procems, high, and dens, tooth. UO 22602, distal portion of radius; UO 26839, complete scapula; UO 8429, pelvic girdle an- terior to acetabular region; all from type lo- cality. REFERRED SPECIMENS: uo loc. 2380: uo 26921, right dentary with M1-M2, and alveoli of lz-Ma but lacking ascending ramus; UO 10816; humerus lacking proximal portion. UO loc. 2349: UO 7956-UO 7960, two dam- aged humeri. CIT loc. 375: LACM 16764, edentuluous dentary fragment; LACM 16765, humerus lacking head. UO loc. 2322: UO 8643, dentary fragment; UO 26890, poorly preserved distal end of radius. UO loc. 2516: UO 26922, dentary fragment with M2-Ma; UO 26923, head of humerus; UO 26924, ulna frag- ment; UO 26925, scaphoid; UO 26926, lunar; UO 26927, medial phalanx II or III of manus; UO 26928-UO 26929, ungual phalanges. TYPE LOCALITY: Krebs Ranch II, UO loc. 2323. AGE: Late Hemphillian. DIAGNOSIS: Deep jawed species with dental for- mula ? ? ? ? . I~ hypertrophied , l i vestigial; P, 3143 with fused double (?) root, P1-Pa single-rooted; molars hypsodont with extensive os cementum. Antemolar length proportionately m o r e re- duced than in other species. Scapula with infra- spinatus fossa extending beneath acromion process. Humerus as in living species of Sca- panus. Radius and ulna relatively shorter than in living species (?). Phalanges of manus short- er than living species. DESCRIPTION: The mandible ( Fig. 7 4) is deep and relatively short. The mandibular symphy- sis extends posteriorly to the level of the P4. Of the two mental foramina, one is located ven- tral to the P2 and the other under the P4. The tip of the type dentary is lacking but the dental formula appears to have been complete based on the referred material from Little Valley and by analogy with other scalopine moles. The L is hypertrophied, incisiform and rather ro- dent-like. The anterior leading edge of the root is sheathed in os cementum which also extends up the posterior moiety of the crown. Crowded ! 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 89 A oc Figure 74. Scapanus (Xeroscapheus) procerwens n. sp., A; Krebs Ranch II, UO loc 2323, Type specimen UO 22508, right mandible with h, P1, and M1-M2, labial and occlusal views, B; Type specimen, lingual view of antemolar region, C; Type specimen, M2, lingual view, D; Little Valley I, UO loc 2380, UO 26921, anterolabial view of antemolar region. hn·h alveolus; mc-metaconid; oc-os cementum. between the labial margins of the b and canine is what appears to be the remnant of a tiny alveolus presumably for a vestigial 13 • The crown of the canine is broken off but the root indicates that it was a simple peg-like tooth, anteroposteriorly compressed like the P1. The P1 is a simple sharply pointed tooth. All the premolars were apparently single-rooted. The antemolar teeth were all tightly appressed. The P~ root indicates a tooth the size and shape of the preceding. The alveolus of the Pa is larger than the preceding but similarly constructed. The alveolus of the P4 is nearly square in trans- verse outline. It may have housed two closely appressed roots but more likely contained a single fused root. The molars differ from all other talpids, ex- cept Scalopus, in two salient features; they are truly hypsodont (taller than long with enamel extending below alveolus rim) and they are partly enclosed by os cementum. The occlu- sal pattern is essentially the same as that of S. (X.) shnltzi except that the connection between the trigonid and talonid is broader ( this may be owing to the more advanced stage of wear) ; the trigonids are slightly more compressed an- teroposteriorly. There is no cingulum bridging the hypoflexid, and the anterior and posterior accessory cusps are relatively smaller in size. The referred mandibles from Little Valley differ little from the type specimen. The P2 and canine alveoli are narrower and the os cemen- tum is not as extensive on the molars. There is a small remnant of the I1 alveolus also in UO 26921 (Fig. 74C). A complete scapula (Fig. 75) lacking only the tip of the scapular spine tuber and a chip out of the auxilliary border was recovered from 90 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 TABLE 21 MEASUREMENTS OF THE LOWER DENTITION AND J Aw OF Scapanus proceridens uo 22503 (Type) uo 26921 uo 26922 h: Length ----·---- ---------- -- ------------ ----------- ------ 1.13 Width-- ------------------- ---------------------- --- ---- ---- 1.03 P1 : Length ----------------- --------- ------- ------ --- ------ -· 0.58 Width ------------------- ----- --------- ------ ------- ------ 0.81 M1 : Length -- ------------ ----- ---- ---- ----- ---- ---- --------- - 2.41 2.82 Width --- ---------------- -------------------------- ----- -- 1.91 2.00 M2 : Length------------- -------- ---------- --------------------- 2.64 2.76 2.66 Width ------- -- ---------------- -------- ------ ---------- -- - 2.00 1.86 1.75 M3 : Length ------------- ---- ----------------------------------- 2.36 Width -------- -------------- ------ ---------- -------------- 1.53 h-M1 alveolar length -------- --- ---- ---- ---· ---- ---------- - 12.9 13.9"* Pi-P 4 alveolar length -- -- ------------- -- ---- ---- ------ -- --- 2.95 4.0 M1-M3 alveolar length --------- ---- -- --------------------- 7.8 7.7 Depth of jaw below M1 --- ---- ------ ----- -- --·-- --------- 3.8 3.2 '' Anterior margin of I2 broken the type locality. It closely resembles the liv- ing species in the extent and development of the acromion process and the posterior position of the biceps tuberosity. The infraspinatus fossa continues forward as in S. cf. S. shultzi. Un- like the living species the dorsal opening of the infraspinatus fossa is very narrow, tapering only slightly anteriorly, and does not flare out posteriorly. The scapular spine tuber tends to isolate the posteriormost portion of the fossa opening from the rest. Of the material exam- ined this feature is partly developed in S. lati- manus and S. townsendii but absent in S. ora- rius owing to the more posterior situation of the spine tuber. A proximal fragment of the scapula from Little Valley is smaller and differs from the Columbia River specimen. In this specimen the walls of the infraspinatus fossa converge to extinguish the fossa slightly posterior to the acromion process. In this specimen all of the superficial lamellar bone has been dissolved away which may in part account for these dis- crepancies. The humeri fall within the range of varia- tion of the recent species of Scapanus in propor- tions and morphology. The brachialis fossa is deeply excavated as in the modern species. The ulna fragment resembles that of Scapano- scapter simplicidens from Red Basin ( Barsto- vian) in the relatively weak coronoid process; although, in this instance, this may be owing to the evident erosion of the specimen. The dis- tal ends of two radii resemble the modern spe- cies of Scapanus except that the groove for the tendon of the abductor pollicis longus ap- pears to be more distally situated. The more distal position of this groove is usually an indi- cation of a relatively short radius as exempli- fied by Scalopus aquaticus. The perfectly preserved scaphoid has a short and relatively broad proximal crest as in Sca- lopus. The part of the trapezium facet on the trochlear process is very small and faces dis- tomedially as in S. latimanus and Scalopus. Laterally nearly the entire groove for the ten- don of the M. flexor carpi radialis is roofed over by bone to form a long tunnel. The lunar . 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 91 I Figure 75. Scapanus (Xeroscapheus) proceridens n. sp., Krebs Ranch II, UO loc 2323, UO 26839, left scapula, dorsal view. more closely resembles that of Scalopus in the general proportions and stockiness but in the details of the carpal articulations it also re- sembles the modern Scapanus. The medial phalanx II or III of the manus is propor- tionately very short in relation to its width and is therefore quite comparable to Sea/opus in this respect. The sacrum and innominata of the pelvic girdle as preserved are fully as fused anterior- ly as in the living Scapanus. RELATIONSHIPS: Scapanus (Xeroscapheus) proceridens exhibits a number of features char- acteristic of only the most advanced North American burrowing moles. The mandible strongly resembles Scalopus aquaticus in ( 1) the short antemolar region, (2) hypertrophied I2, ( 3) hypsodont molars, ( 4) presence of os cementum, ( 5) reduction of cingula and acces- sory cuspids, and ( 6) compressed trigonids. However, Scalopus (Hesperoscalops) rexroadi or S. (H.) sewardensis, presumably ( and ad- mittedly) ancestral to S. aqnaticus, primitively possess better developed cingula and accessory cuspids than S. proceridens. The I1 and I3 are greatly reduced in S. proceridens but it is the canine and a premolar that are reported lost in S. rexroadi ( Hibbard, 1953). The cf. Sea/opus sp. from the Hemphillian Edson Quarry of Kansas described by Hibbard ( 1939) suggest that the Scalopns lineage was already well es- tablished by Hemphillian time. Of the modern species of Scapanus, S. latim- anus most closely resembles S. proceridens in ( 1) the shortening of the antemolar region, ( 2) high crowned or hypsodont teeth, ( 3) absence of anterior cingula and reduction of the labial cingula, and ( 4) hypertrophy of the I2 • S. lat- imanus also differs from S. proceridens in the ( 1) relatively longer antemolar region and mandibular sympysis, ( 2) lack of os cement- um, ( 3) presence of large accessory cuspids, ( 4) remnants of labial cingula, and ( 5) more open trigonids. The possibly somewhat older material from Little Valley, UO loc. 2516, ap- proaches S. latimanns in the presence of larger accessory cuspids and vestiges of the labial cingula. The humerus and scapula of S. proceridens indicate close affinity with Scapanus (Scapa- nus). The humerus does not show the extensive broadening as in Scalopus. The striking resem- blances to Scalopus may be considered as an example of parallelism. The proximal elements of the foreleg do not differ greatly from those of S. (Scapanus) but the distal elements tend to more closely resemble Sea/opus in their func- tional shortness. In conclusion, S. (X.) proceridens can not be considered as ancestral to any of the living Scapanus but rather an offshoot of the primitive 92 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 Scapanus stock via S. shultzi which roughly paralleled Scalopus in several structural modi- fications. Scapanus cf. S. (Xeroscapheus) proceridens (Figs. 76, 77) MATERIAL: Otis Basin UO loc. 2347: UO 26930, edentulous dentary fragment; UO 26931, clavicle; UO 26932, ulna fragment lacking the proximal crest, posterior crest and distal end; UO 26933, radius lacking capitular process; UO 26934, ungual phalanx of manus. McKay Reservoir UO loc. 2222: UO 3508, proximal end of scapula; UO 3506, humerus lacking proximal end; UO 24818, metacarpal I; UO 24819, proximal phalanx II or III of manus; UO 24818, ungual phalanx of manus. The clavicle (Fig. 76) from Otis Basin is fully of the modern type. I could see no con- sistent differences between the various living species and the fossil with the sample avail- able except that the ventral process is relatively shorter proximodistally and the clavicular (vascular) notch correspondingly is wider than in the living species. There is also a nutrient foramen dorsal to the ventral process on the anterior face of the clavicle which was not ob- served in the modern sample. The humerus from McKay is small, about the size of that in S. orarius, but the scapular fragment is that of a mole about the size of the Krebs Ranch mole. The ulna from Otis Ba- sin is too fragmentary for much comparison but the shaft appears to have been relatively shorter than in S. orarius or S. townsendii. The much better preserved radius (Fig. 77) from this locality is definitely shorter than either of the above species and is relatively as short or shorter than that in S. latimanus or Scalopus aquaticus. In details the radius (position o_f the aroove for the tendon of M. abductor polh-o cis longus, etc.) more closely resembles Sca- panus, especially S. latimanus. The metacarpal I from McKay strongly resembles that of S. latimanus but is pro- portionately somewhat shorter, although not closely approaching Scalopus in this respect. I Figure 76. Scapanus cf. S. procerulens, Otis Basin, UO loc 2347, UO 26931, left clavicle, posterior (left) and anterior views. The proximal phalanx of the manus is propor- tionately similar to that of S. ( X.) proceridens from Little Valley. The other elements are not notably different from those of the living Scapanus. Although no teeth of a Scapanus-Iike mole were found at any of these localities, advanced features of various elements especially in the distal foreleg elements, suggest a mole beyond the structural grade of S. latimanus and ap- proaching Scalopus in several respects. The reduction in the length of the radius, meta- carpals and phalanges is like that occurring or expected in S. (X.) proceridens. l Figure 77. Scapanus cf. S. proceridens , A; Otis Basin, UO loc 2347, UO 26933, right radius, lat- eral view, B; McKay Reservoir, UO loc 2222, UO 24819, right second or third proximal phalanx of manus. 1- length; 2--width. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 93 Scapanus sp. The two partial humeri (LACM 6599) from Rome ( CIT loc. 62) are about the size and morphology of S. orarius but assignment even to subgenus is of little meaning based on the material available. Summary of Relationships of the Scapanzts-like Moles The oldest of the Scapanus-like moles, Sca- panoscapter simplicidens, is in the position of providing a good structural and temporal ( Bar- stovian) ancestor to the genus Scapanus. The low crowned molars, complete dental formula, and double-rooted premolars pose no special problems to deriving all the species of Scapanus from this genus. Likewise, Scalopus could also be derived from such an ancestral form, al- though there is no direct indication of this. The phalanges and metacarpals of the manus essen- tially achieved the same degree of shortness as in the Recent S. orarius and S. townsendii. In- deed, on the basis of the known material, S. simplicidens, provides the latest species which could serve as a common ancestor for all the species of Scapanus. Tedford (1961), on the basis of teeth, postu- lated that the Clarendonian S. shultzi consti- tutes a reasonable common ancestor for the living species; however, the referred Oregon material and as yet unstudied postcranial ma- terial from the Ricardo Formation indicate that S. shultzi is off the direct lineage of the Recent species. The shortening of the distal forearm elements in S. shultzi has proceeded beyond that of the most fossorially specialized of the Recent species, S. latimanus. S. shultzi is thus well on the way toward the development of the brevirostrine S. (Xeroscapheus) proceridens of the Hemphillian. While some hypsobrachyodont moles may still occur in the Hemphillian faunas, the hyp- sodont S. proceridens seems to be the dominant species in the Great Basin and is strikingly con- vergent with Scalopus in many features, al- though surpassing it in the devolpment of os cementum. The living species all seem rather closely related with S. latimanus being the most dis- tinctive. This species occupies the largest range and variety of habitats and in many ways seems to be heading down the same structural path- way as S. shultzi. S. latimus has the greatest crowding of the antemolars ( one lost in some southern populations~ see Palmer, 1937), most compressed trigonids, highest crowned teeth, and shortest metacarpals and phalanges of all the living species of Scapanus. The more open trigonids in S. townsendii and S. orarius may be derived from the compressed condition in Scapanoscapter. Scalopina sp. A (Figs. 78-81) A few isolated teeth and skeletal elements recovered from USGS loc. M. 1043 at Snyder Creek seem to represent one species and prob- ably a new genus of advanced burrowing moles thus far undescribed. Additional matrix in the form of anthill accumulations has been col- lected but had not been picked at the time of this writing. Since available material seems in- Figure 78. Scalopina sp. A, Snyder Creek, USGS loc M1043, USNM 23773, occlusal, lingual and labial views. 94 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 sufficient to me for useful characterization of this form, I shall restrict my discussion to the description of the material at hand and tenta- tive relationships. The M1, USNM 23775, closely resembles that of the living Scapanus townsendii or S. orarius except that the protocone is more ro- bust and inflated basally. Unlike Scapanoscap- ter, the paraconule is not distinguishable and the metaconule is small and not particularly distinct. The tooth is 3.12 mm. long and 2.33 mm. wide. The M2, USNM 23773 (Fig. 78), is brachyo- dont. The trigonid is wider but shorter than the talonid and the anterior cingulum tapers and vanishes labially. The tooth is moderately worn hut the entocristid was either very low or ab- sent so that the postfossid opens lingually. The crista obliqua intersects the middle of the pro- tolophid, and there is no metastylid. These last two mentioned features distinguish it from all the other North American highly fossorial moles. The trigonid is about as compressed as in Scapanoscapter. There is a weak cingulum bridging the hypoflexid and a strong posterior accessory cuspid. This tooth is 2.88 mm. long and 1. 93 mm. wide. A lower jaw fragment, USNM 23774 (Fig. 79), of the antemolar region retains the roots or alveoli of at least four or possibly five teeth. The anteriormost alveolus is inclined about 45° and held an apparently large and long tooth, probably the h This is followed by the roots of a small obliquely oriented premolar. Figure 79. Scalopina sp. A, Snyder Creek, USGS Joe Ml043, USNM 23774, fragment of the ante- molar region of the left mandible, occlusal and labial views. Figure 80. Scalopina sp. A, Snyder Creek, USGS loc Ml043, USNM 23776, left humerus lacking proximal extremity and epicondyles, anterior view. There are a number of possible interpretations for the succeeding four alveoli. There are two mental foramina on this fragment of jaw. Re- gardless of the dental formula, the antemolar region was at least somewhat crowded. If the first alveolus preserved is that of the 12, then at least two teeth have been lost between the 12 and the first recognizable premolar. The humerus, USNM 23776 (Fig. 80), is devolped to about the same structural grade as Parascalops and Domninoides. It also resem- bles these genera in several structural details such as the hasining of the area between the pectoral muscle scars, and relatively narrow clavicular facet. On the other hand the rela- tionship of the pectoral muscle scars is roughly intermediate between that of Scapanus and Domninoides. The second and fourth metacarpals, USNM 23777 and USNM 23778, are proportionately longer than in either Parascalops or the Recent Scalopina. The fourth metacarpal resembles the same element from Quartz Basin referred to Domninoides, except in greater size, less expanded proximal tubercle and slightly longer proportions. The medial phalanx of the second, third, or fourth digit, USNM 23779 (Fig. 81), is ex- 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 95 I Figure 81. Scalopina sp. A, Snyder Creek, USGS loc Ml043, USNM 23779, medial phalanx from man us, dorsal view. tremely short in relation to its width and is comparable to or shorter than that in Scalopus. The development and lateral situation of the pectoral tubercle of the humerus, lack of the metasty lid and labial . situation of the crista obliqua on the second lower molar, small size of the conules on the upper first molar, and the highly fossorial specializations of the forearm suggest affinities with the Scalopina. Scalopini sp. A ( Figs. 82, 83) Four postcranial elements, a clavicle, two metacarpals, and a phalanx of a small burrow- ing type mole were recovered from Red Basin, UO loc. 2495. They are either too small or too large to be adequately assigned to any of the other moles known from this site with per- haps the exception of Scalopoides. The ele- ments are only slightly larger than those of Scalopoides ripaf odiator from Quartz Basin but differ strikingly in proportions. Inasmuch as the Scalopoides from Red Basin is smaller but otherwise closely resembles the Quartz Basin Scalopoides, it seems unlikely that these elements belong to this genus. The clavicle, UO 24388 (Fig. 82), is pro- portionately shorter than any of the known Miocene Scalopoides, although it does resemble the short clavicle of ?Scalopoides sp. C. The shape of the ventral process is similar to that in Scapanulus although somewhat more broadly based. There is no obvious lateral reflection of the ventromedial spine. A span of bone bridges the vascular notch from the base of the ventro- medial spine to the dorsal margin of the ventral process thus forming a foramen. The presence or absence of this bridge may be individually variable if a larger sample were known. A nutritive foramen also pierces the clavicle within the vascular notch. Posteriorly the notch continues dorsally as a deep gutter extending to the dorsal prominence. The long axes of the articular facets are more nearly parallel than they are in the Talpini. The two metacarpals are relatively shorter in relation to their width than those of Scalo- poides although slightly longer than in Sca- panoscapter simplicidens. The third metacar- pal, UO 22439 ( Fig. 83A), is about at the same structural grade as in Parascalops. The fourth metacarpal, UO 24375, is nearly identical to the fourth metacarpal referred to Domninoides from Quartz Basin except that it may be very slightly shorter. The proximal tuberosities are well defined on both specimens. A single prox- imal phalanx of the manus, UO 24378 (Fig. 83B), is essentially identical to that of S. iso- dens except for its slightly larger size. It is 1.43 mm. long and 1.33 mm. wide. The structure and details of the clavicle in- dicate that Scalopini sp. A is referable to the Scalopini. The high degree of specialization in the clavicle and hand bones eliminates the Urotrichini as here interpreted. Among the Parascalopina, the Miocene Scalopoides are Figure 82. Scalopini sp. A, Red Basin, UO loc 2495, UO 24388, left clavicle, posterior and ventral views. 96 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 B Figure 83. Scalopini sp. A, Red Basin, UO loc 2495, A; UO 24373, third metacarpal, ventral view, B; UO 24378, proximal phalanx of manus, dorsal view. less specialized but the Pliocene forms referred to Scalopoides especially the later members approach the condition in this form. The pro- portional similarity here is probably one of convergence and there are several detailed dif- ferences between the Pliocene Scalopoides-like moles and Scalopini sp. A. The discernible lateral deflection of the ventromedial spine of the clavicle for the tetrahedral sesamoid in Scapanulus and Parascalops is absent in Scalo- pini sp. A. Domninoides is not certainly repre- sented by any of the elements known from the Red Basin. There is a similarity of the manus bones to those referred to Domninoides from Quartz Basin, but the Quartz Basin specimens are only tenuously grouped together. There is so much structural convergence in various lineages of burrowing talpids, that I believe it is more realistic to leave these elements un- assigned especially in view of the lack of any associated Domninoides teeth ( or any other available mole) from the locality. General similarity to Scalopoides, Scapanulus, and Parascalops tends to align Scalopini sp. A with the Parascalopina. TALPINAE incertae sedis ACHLYOSCAPTER 10 gen. n. GENOTYPE: Achlyoscapter longirostiris, sp. n. 16 From the Greek achlys, mist, dimness, darkness, and skapter, a digger. KNOWN DISTRIBUTION: Late Miocene (Barstovi- an) of Oregon. DIAGNOSIS: Small mole with dental formula ? ? ? ? ~ i ~ ~- Mandible long with very little taper to antemolar region; antemolars uncrowded; Is to P4 roughly graded in size, largest poste- riorly; all premolars double-rooted and later- ally compressed; P2 (?) and P3 double-rooted; P4 elongated posteriorly, prominent protocone; M1 with three discernible labial cusps and well defined paracrista and metacrista. Achlyoscapter longirostris17 n. sp. (Figs. 84-88) TYPE: UO 22412, left mandible with Is-M1 and the alveoli of li-b and M2-M3, most of the ascending ramus lacking. HYPODIGM: UO 22419, fragment of right den- tary with P 4-M1 and alveoli of P2-P3; UO 22583 and UO 22344, P4's; UO 22343, left mandible fragment with M1-M2 and alveoli of P4 and M3; right mandible fragment with alveoli of M1-M3. All specimens from the type locality. REFERRED SPECIMENS: uo loc. 2465: uo 22351, fragment of upper jaw with two ante- molars; UO 22352, UO 22353, two scaphoids. UO 24335, right mandible fragment with dam- aged M1 and alveoli of P1-P4. TYPE LOCALITY: uo loc. 2465, Quartz Basin. AGE: Barstovian. DESCRIPTION: One specimen, UO 22419, (Fig. 84A) preserves the P4-M1, alveoli of the P3 and posterior root of the P2• A downward flexure of the snout is evident in the area of the P4 • The posterior alveolus of the P2 suggests that the tooth was slightly oblique in the maxillary with the anterior moiety more medial. The P3 was apparently larger but similarly arranged. The P4 (Fig. 88B) is a narrow elongate tooth with a prominent protocone which imparts a distinct triradiate aspect to the tooth. The high sharp 17 From the Latin long, long, plus rostrum, snout. 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 97 11\- B ~ I Figure 84. Achlyoscapter longirostris n. sp. , n. gen., Quartz Basin, UO loc 2465, A; UO 22419, right maxillary fragment with half of P 4 and M1 , occlusal view, B; UO 22583, left P4, occlusal and lingual views. TABLE 22 MEASUREMENTS OF THE UPPER DENTITION OF Achlyoscapter longirostris uo 22344 ······ ··· ····· ··· · uo 22583 .......... ...... . . uo 22419 ............. .. .. . M1 : UO 22419 ......... ... .. ... . Length 1.41 1.56 1.46 Width 0.89 1.04 1.40 paracone is situated anterior to the center of the tooth. A long concave crest extends from the apex of the paracone to the posterior extremity of the tooth. The tooth is enclosed in a narrow cingulum which expands somewhat at the an- terior tip and is contiguous with the protocone. The protocone extends medially and slightly anterior to the center of the tooth. There are three roots, one beneath each extremity. The lingual roots diverge from each other. The M1 is subtriangular in occlusal outline. The protocone, paraconule, and metaconule are all distinguishable on the worn tooth. The metaconule is situated close to the base of the metacone as is the paraconule to the paracone. Well defined paracingula and metacingula are present. The parastyle is either not developed or abraded away in the specimen, probably the latter. In UO 22351 (Fig. 85) part of the anterior- most lateral wall of the narial opening and two associated teeth are preserved. Immediately below the narial flange there is a round alveolus presumably for a single tooth. This alveolus is followed by a single-rooted tooth with a high transversely compressed cusp which is slightly recurved. Immediately posterior to this tooth is a larger double-rooted tooth somewhat less compressed and with the beginnings of a pos- terior cingulum but otherwise similar in form to the preceding tooth. Part of another alveolus follows this. The tall, compressed and un- crowded nature of these teeth as well as the size of the specimen suggest that they may be P1 to canine inclusive of A. longirostris. The dentary is long and not strongly tapered anteriorly ( Fig. 86). There are two mental foramina, one beneath the P2 and the other under the P4. The symphysis extends back as far as the P2. The dental formula is complete: 13, C, P4, M3. All the teeth anterior to the P 1 are single-rooted and the P1 and all succeeding teeth are double-rooted ( Fig. 86A). The Ii is inclined about a 45 ° angle to the axis of the Figure 85. Achlyoscapter longirostris n. sp. , n. gen. , Quartz Basin, UO loc 2465, UO 22351 , rostral fragment with ?C-P1, labial and occlusal views. 98 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No. 11 A B Figure 86. Achlyoscapter longirostris n. sp. , n. gen. , Quartz Basin, Type specimen UO 22412, left mandi- ble with 13-Mi, A; sketch of X-Ray photograph, B; labial and occlusal views. TABLE 23 MEASUREMENTS OF THE LOWER DENTITION AND J Aw OF Achlyoscapter longirostris uo 22412 0.59 0.30 0.64 0.31 0.60 0.33 0.60 0.35 0.67 0.39 0.98 0.51 1.61 0.98 uo 22343 uo 24335 Ia : Length ... .... .... .. .... ...... .................. . Width ···· ······---·-· ······--·-············-·-- C1: Length ....... ..... ................ ...... ....... . Width ........... ..... ... ......... ...... ...... . . P1: Length .... ..... .... .. ...... ..... .. ...... ....... . Width ...... ... ... .. .. ...... ... ... ...... .. .... . . P 2: Length .... ... ... ..... ...... ... .... .... .. ....... . Width ...... ... ....... ...... ............ ...... . . P a : Length ... ........ ........... ............ ....... . Width ...................... .. ...... .......... . . P4 : Length ............ .... ...... .. ................ . . Width ................. .. ... ... ................ . M1: Length .... ............ ... ......... ...... ...... . . Width ............ ... ....... ........ ........... . M2: Length ... .. ..... .......................... ..... . Width .... ...... ... ...... ... ... .... .. .......... . P1-P4 : Alveolar length .. .... ... .... ... ...... . . M1-M3 : Alveolar length ....... .. .. .... ........ . Total alveolar length ..... .. .... ... ......... ... ... . Depth of mandible below M1 (internal) .... ..... ...... .... ...... ...... .......... . e Estimated length. 2.63 4.36 8.32e 1.59 1.58 1.03 1.64 0.99 4.43 1.56 2.45 1.57 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 99 jaw and the succeeding teeth become progres- sively more vertically aligned to about the P2-P3. The Ii and crown of the 12 are missing but both possessed long roots, although neither seems to have been especially enlarged. The rest of the antemolars progress in size from the 13 to the P4 • The crown of the 13 is procumbent and strongly compressed transversely. The pos- terior base of the crown suggests an incipient heel. The canine is similar to though larger than the I:1• The root of the canine extends pos- teriorly beneath both the roots of the P1. The P1 is not so compressed as in preceding teeth, and there is a small posteriobasal cuspule flanked by weak cingula which. extend ventrally and an- teriorly a short distance on either side. The ba- sal cuspule is joined to the base of the protoco- nid by a short ridge. The crown of the P 1 is still rather procumbent and the apex of the protoco- nid is situated slightly anterior to the level of the anterior root. The roots of the P1 are rela- tively and absolutely shorter than those of the other premolars. The P 2 is about the same size as the P1 but is more compact, and the heel is no more developed. The protoconid is centered squarely over the anterior root. The P~ is oval in occlusal outline and has lost the wedge- shaped outline of the preceding teeth. There is a slight basining of the heel on either side of the ridge connecting the base of the protoconid and the posterior cuspule. There is a very faint sug- gestion of the metaconid on the P 3 • The P 4 in turn resembles the Pa but is more elongate. The heel is relatively larger but its features are no better defined than on the P 3 • The molars (Fig. 87) are brachyodont. The M1 is rather wedge-shaped in occlusal view and the trigonid is slightly longer but narrower than the talonid. The anterior cingulum is poorly developed or absent. A short cingulum covers the base of the paraconid anterolingually and lingually. The labial cingulum is incom- pletely developed bridging only the hypoflexid. There is no posterior cingulum but a small pos- terointernal accessory cuspid is present. The basilabial margin of the tooth is notched. The profossid is deep and cuts anterolabially thus Figure 87. Achlyoscapter longirostris n. sp., n. gen., Quartz Basin, UO loc 2465, UO 22343, left M1-M2, labial, occlusal and lingual views. restricting the paralopid to a low sharp ridge deeply notched between the parconid and proto- conid. The paraconid is the lowest cusp but well formed. The protoconid is the tallest cusp and the remaining cusps are of about equal height. The entoconid is connate without an en- tocristid or only a very small one; however, the lingual rim is somewhat raised so that the shallow postfossid is enclosed. There is no metastylid and the crista obliqua joins the mid- dle of the protolopid. The M2 is only preserved in UO 22343. The tooth is more rectangular than the Mi. The anterior cingulum is well developed but the labial cingulum as in the M1 is only developed across the hypoflexid. Also, like the M1, there is a small anterolingual cingulum around the base of the paraconid and a posterior accessory cuspid. The trigonid is more compressed an- teroposteriorly than in the M1, and the talonid is slightly longer and wider than the trigonid. A metastylar ridge is fused to the metaconid and is connected to the entoconid by a low but distinct entocristid. The crista obliqua extends more lingually than in the M1 to terminate just before the juncture between the metastylid ridge and the metaconid. The labial edges of the protoconid and hypoconid are more angular 100 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 than those of the M1. The M3 is not preserved in any of the specimens but its alveoli do not suggest a reduced tooth. The single specimen of the dentary (UO 24335) from Red Basin preserves all the pre- molar alveoli. The M1 is so badly damaged that little comparison can be made between it and the type but the anterior and posterior cingula are absent and the basal labial crown margin is notched. Both mental foramina are situated more posteriorly than in the type under the p3 and trigonid of the M1. Although specific allo- cation cannot be demonstrated on such poor material, UO 24335 is referable to A. longi- rostris on the basis of its four double-rooted premolars and construction of the M1 • Two scaphoids (Fig. 88) from a mole about the size of Achlyoscapter were recovered from the type locality. They are about 13 per cent smaller than those of Scalopoides ripafodiator and presumably quite different from those expected for Mystipterus. In general the two specimens resemble the scaphoid of S. ripa- fodiator but are transversely narrower, dorso- ventrally longer, and possess a prominent spur which projects posteriorly from the center of I Figure 88. ? Achlyoscapter longirostris n. sp., n. gen., Quartz Basin, UO loc 2465, UO 22353, right scaphoid, dorsal, medial and distal views. the posterior crest. The general features of the scaphoid suggest a manus nearly comparable to that of Scalopoides. RELATIONSHIPS: Achlyoscapter longirostris may be distinguished from all of the North American fossil and living moles by a combi- nation of its small size, brachyodont molars, complete lower dental formula, narrow double- rooted lower premolars, and elongate upper fourth premolar. There is no evident hyper- trophy of the anterior lower incisors as in all the Recent Urotrichini, Scalopini, and Uropi- linae, nor is the P1 double-rooted or enlarged as in Recent Scaptonyx or Talpini. Among the Eurasian fossil moles in which comparable material is available, "Scaptonyx" jaegeri Seeman (1938) most closely ap· proaches Achlyoscapter in having the same number of teeth, little antemolar specializa- tion, and elongate M1 with open prefossid. Notable differences of "S." jaegeri from Achlyoscapter include ( 1) the single-rooted P1, (2) more posterior location of both mental foramina, ( 3) well developed anterior and labial cingula on the lower molars, and ( 4) crista obliqua of M1 extending much further labially ( if Seeman's 1938 figures are accu- rate) .18 Regardless of these differences A. longirostris, as known, shows its greatest gen- eral similarity to "S." jaegeri, however, the types of these two species must be more closely compared than is possible with the existing lit- erature before actual relationship or simple retention of generalized characters can be evaluated. The structure of the lower molars and the small size of this species suggest affinities with the Urotrichini. The lower dental formula is unlike any Recent urotrichine but this is to be expected in a possible primitive member. Reduction of the antemolar dentition except the anterior incisors, increase in molar crown height, and compaction of the P4 and other small dental modifications could situate Achly- 18 In Seeman's illustrations (Figs. 12-15) all the molars sho"'. the same structure. This seems highly unlikely in any detailed sense for a brachyodont mole. 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 101 loscapter on the line to any of the Recent Uro- trichini. On the other hand the scaphoid, if associated with this mole, would tend to align Achlyoscapter with the Scalopini or at least place it beyond the structural grade of the Uro- trichini as used here. The conservative nature of the dentition suggests a structural notch below that exemplified by Scalopoides. Of the North American fossil talpids, Achly- oscapter provides the best available ancestor for Condylura. Like Achlyoscapter, Condylura has a complete set of brachyodont teeth, elon- gate antemolar region, double-rooted premol- ars, and elongate P4. However, Condylura is far removed from Achlyoscapter and other moles by its molariform lower premolars separated by diastemata. The lower molars of Condylura also have prominent metastylids with the crista obliqua joining directly to the metastylid rather than the protolophid as in Achlyoscap- ter. These features cannot be considered primi- tive in Condylura and could be developed di- rectly from such a form as Achlyoscapter. Un- fortunately, no moles which might bridge the gap between these two forms are yet known. TALPIDAE AND TALPINAE incertae sedis Several localities have yielded isolated upper teeth which, owing to their rarity in the samples or to the small size of the sample, are difficult to assign to a particular genus. They are therefore discussed under general head- ings and arranged by locality. There is sufficient variability within genera and species and a general similarity between many genera with otherwise widely divergent postcranial skele- tons to make specific assignments meaningless without a more complete series of the dental and skeletal material. Talpinae incertae sedis BLACK BUTTE: uo loc. 2500 (Fig. 89). The upper P\ UO 24 779, of a small mole closely resembles that of Scalopoides ripafodiator in most details. Although much of the enamel has been stripped away, the paracone appears to have been joined to the ·anterior extremity of the tooth by a cingulum. Figure 89. Talpinae incertae sedis, Black Butte, UO loc 2500, A; UO 24781, right M1, occlusal view, B; UO 24782, right M2, occlusal view. An M1, UO 24781 (Fig. 89A), is a brachyo- dont tooth with a distinct protocone, paraco- nule, and metaconule. Well developed para- cingulum and metacingulum are present as well as a strong parastyle. The mesostyle is elongate and was probably partly divided before wear. The structure of the lingual moiety of the tooth approaches Parascalops in the reduction of the protocone and prominence of the conules. This imparts a rather rectangular outline to the lingual moiety rather than a V-shaped one as in 5. ripafodiator. The roots are like those of S. ripafodiator. The tooth is 2.39 mm. long and 1.61 mm. wide. The M2, UO 24782 (Fig. 89B), is not unlike Scalopoides except that the protocone is re- duced in volume thus imparting a flatter out- line to the lingual margin. The metaconule al- though worn was not greatly separated from the protocone as it is in the M2 specimen re- ferred to Mystipterus from this locality. The roots are broken but were apparently like those of S. ripafodiator. The tooth is 1.87 mm. long and 2.00 mm. wide. In size the M1 could belong either with UO 24 782 or the referred Mystipterus M2 • I believe the former association more probable on the basis of the analogous similarity of the lingual 102 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 cusps and lack of any particularily uropsiline features. The M2 could be construed as a varia- tion of the Mystipterus M2, but the similarity between the Quartz Basin M ystipterus M2 and the one from this locality suggests a consistency of the characters stressed. Association with one of the Scalopoides-like moles or Domninoides seems probable to me. Figure 90. Talpinae incertae sedis, Bartlett Moun- tain, UO loc 2517, UO 26804 right P4-M3, postero- lingual tip of M2 restored from opposite side, oc- clusal view. BARTLETT MOUNTAIN: UQ loc. 2517 (Fig. 90). The most complete associated upper dentition, UO 26804, of all the fossil moles studied comes from this locality. Unfortunately no lower teeth of a mole of comparable size were recovered so that its affinities remain in doubt. The speci- men consists of a right upper tooth series with the P 4-M3 and a left series with P 4-M2• All the teeth were recovered as isolated specimens in a single screen of washed matrix. Since they are all in the same state of preservation and wear and two of the teeth retain a bony contact, there is no doubt that only one individual is represented. An additional P4, UO 26841, was also recovered from the site. The molar teeth indicate an animal about the size of Achly- oscapter. The P 4 nearly forms an equilateral triangle in occlusal outline except that the labial side is somewhat longer than the others. The pa- racone is transversely compressed posteriorly as in Scalopoides but the cusp as a whole is not as relatively large. A cingulum encloses the tooth except around the lingual base of the protocone and the posterior extremity and nar- rows anterolabially as in Scalopoides. The pro- tocone is a low but well defined connate cusp. Three prominent roots are present. TABLE 24 MEASUREMENT OF THE TEETH OF TALPINAE incertae sedis FROM BARTLETT MOUNTAIN Length Width uo 26841 p4 ······· ······· ················ 1.21 0.97 uo 26804 P4: right .............. .... 1.23 1.00 left ·-- ----·-··------- -- 1.27 0.97 Ml: right ........ ..... ..... 1.88 1.17 left --- -- -- ----------- -- 1.90 1.13 M2: right ................ .. 1.63 left --- ----- ---------- -- 1.47 1.60 Ma: right ............. .. ... 1.20 1.05 The M1 resembles that of the Urotrichini and small parascalopine moles. The protocone, paraconule, and metaconule are all discernible even with wear. Paracingulum and metacingu- lum are present. The parastyle forms a promi- nent shelf-like anterolabial extension of the paracrista. The posterior ectoflexus is unusual in its shallowness, and there is a short cingulum which fills in even what little concavity there is basially. The lingual cusps of the M2 resemble those of the M1 but the lingual sulcus separating the protocone and paraconule is better developed. The paracingulum and metacingulum do not extend to the stylar cusps. The mesostyle shows no incipient twining and the parastyle and metastyle are not strongly curled as in many other moles. There are four roots as in Sca- lopoides but they are interconnected to the central root by low thread-like interradicular ridges. The lingual root is transversely com- pressed and longitudinally broadened as in Mystiptems and Neurotrichus, but it is more obliquely oriented. The M3 resembles that of S. ripafodiator in most details except that the anterior ectoflexus is longer and broader. There are three roots. General features suggest affinities of these specimens with the Scalopoides-like moles but the urotrichine moles also show many of these same features. 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 103 I Figure 91. Talpinae incertae sedis, McKay Reser- voir, UO Joe 2222, UO 24812, right M2 , occlusal view. MCKAY RESERVOIR: uo loc. 2222 (Fig. 91). Four upper teeth of small moles were recov- ered from McKay. Two of these teeth are upper second molars, both different. The heavily worn P\ UO 26817, is transversely compressed so that the posterior crest of the paracone must have been quite sharp-edged. The cingula are similar to those of Scalopoides ripafodiator except that there is no cingulum on the anterior extremity and the anterolabial side of the tooth. The protocone has been sheared away with wear. The lingual root is situated close to and almost between the labial roots. The tooth is 1.50 mm. long and 1.00 mm. wide. One of the second molars, UO 24813, strongly resembles S. ripafodiator in detail except that the protocone has a shorter lingual slope, is less angulate lingually, and is enclosed by an irregular narrow cingulum. It lacks the metacingulum of N eurotrichus gibbsii and has a small cuspidate paraconule like S. ripafodi- ator. N. gibbsii either lacks the paraconule or it may be expressed by a small noncuspidate elevation of the protocrista. The roots are like those of Scalopoides. The lingual root of N. gibbsii is similar to that of Mystipterus. In addition to the major roots there are two papil- late roots, one centrally situated and the other between the lingual and postirolabial roots. The tooth is 1.80 mm. long and 2.02 mm. wide. The other second molar, UO 24812 (Fig. 91), differs from the first in the following fea- tures: larger size, large distinct paraconule as in Mystipterus sp. from Black Butte, lack of a lingual cingulum, lack of the posterior papil- late root, and somewhat more divided meso- style. It is 2.10 mm. long and 2.26 mm. wide. The M3, UO 26818, resembles that of Neuro- trichus in almost every detail except that the paraconule is discernible and similar to that in the M2 of S. ripafodiator in position and ex- pression. No cingula are present and the meta- conule does not bulge posteriorly as promi- nently as in many specimens of N. gibbsii. It is 1.40 mm. long and 2.37 mm. wide. It seems likely that most or all of the above teeth might be assigned to one or both of the small moles known from McKay, ?Scalopoides or ?Nenrotrichus. There is a significant amount of time separating the McKay fossils from the latest well known Scalopoides (S. ripafodiator, Barstovian) and the living N eurotrichus gibb- sii. The molar teeth of talpines are basically conservative but may develop many small transient or convergent modifications to ob- fuscate assignment without the benefit of a large sample or an articulated specimen. In the case of the small McKay sample, both the moles ( if indeed there are two) are similar enough in size to cause confusion of similar teeth and I will thus refrain here from making arbitrary assignments. GUANO RANCH, USGS loc. M 1042 (Fig. 92). A single M2, USNM 23772, appears to belong to a mole smaller than Scapanus but larger than Scalopoides ripafodiator. Although gen- erally similar to Scalopoides in morphology, it is higher crowned. The labial moiety is rela- tively smaller and the ectoloph is very sharp with more deeply excavated ectoflexi and proto- fossa. The tooth is 2.06 mm. long and 2.58 mm. wide. A cuspidate mesostyle is absent. This may represent the M2 of Domninoides but is hardly demonstrable on the small sample available. Figure 92. Talpinae incertae sedis, USGS loc Ml042, USNM 23772, right M2, occlusal view. 104 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 Figure 93. Talpinae incertae sedis, Quartz Basin, UO loc 2465, UO 25289, two upper right ante- molar teeth, labial and occlusal views. Talpidae incertae sedis QUARTZ BASIN: uo loc. 2465. An upper jaw fragment, UO 2258 (Fig. 93) anterior to the P4 from Quartz Basin preserves two antemolars. The anterior tooth is simple with a moderately tall and slightly inflated cusp. This is followed by an alveolus apparent- ly for a tooth of similar nature which in turn is followed by a large double-rooted tooth four times the size of the first. The large tooth is regularly ovate in occlusal view with a large double-rooted tooth about four times the size of the first. The large tooth is regularly ovate in occlusal view with a simple large cusp and small posterior cingulum. The general ap- pearance of the teeth would suggest a small mole other than Achlyoscapter but this would still leave Scalopoides, Mystipterus, and Dom- ninoides and possibly one of the other non- soricid insectivores as possible candidates for association. The proximal epiphysis of the tibiofibula, UO 22304 ( Fig. 94), from the same locality represents a mole smaller than Scalopoides ripafodiator. The falciform process is dam- l Figure 94. Talpidae incertae sedis, Quartz Basin, UO loc 2465, UO 22304, proximal end of right tibia , proximal and anterior views. l Figure 95. Talpidae incertae sedis, Bartlett Moun- tain , UO loc 2517, UO 24799, left MR, occlusal view. aged but appears to have had a large flange. In proximal view there is a small spur off the lateral condyle that is not visible in this view in any of the moles or shrews examined. The specimen could belong to Achlyoscapter or Mystiptems but there is no definite evidence as to which it should be assigned-if either. BARTLETT MOUNTAIN: uo loc. 2517 (Fig. 95). An isolated M\ UO 24 799, about the size of Scalopoides ripaf odiator was recovered from this locality. It differs from the M3 of sev- eral other small moles studied in the pres- ence of a very strong lingual sulcus separating the large distinct protocone from the para- conule. The paracingulum is very short and does not extend labially beyond the base of the paracone. There are no other cingula. The mesostyle is elongate but not noticeably di- vided. In size and analogus morphology of the entire lingual moiety this tooth strongly re- sembles the form represented by Talpinae in- certae sedis M2, UO 24812, from McKay. It also agrees with the ?Talpidae incertae sedis from this locality in size. ?Talpidae incertae sedis (Figs. 96, 97) Three lower teeth of questionable talpid affinities from two localities of different age were recovered from the J untura Basin. An Mi, UO 25289, from Black Butte locality 2500, is strikingly different from any of the tapids, living or extinct, examined. The tooth (Fig. 96A) is wedge-shaped in occlusal outline. The trigonid and talonid are about equal in length but the talonid is wider than the trigonid. The 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 105 l A Figure 96. ?Talpidae incertae sedis, lower right first molars, occlusal, lingual and labial views, A; Black Butte, UO loc 2500, UO 25289, B; Bartlett Mountain, UO loc 2517, UO 25124. trigonid is very peculiar. The protoconid is the largest trigonid cusp. It is either fused with a labially displaced metaconid or somewhat transversely elongated. There is a low connate paraconid anteriorly and labially which is con- nected to the protoconid by a narrow para- lophid. The paralophid arches anterolabially from the anterior side of the protoconid to con- nect to the labial side of the paraconid. A ridge extending from the level of the protoconid to the level of the paraconid along the lingual side of the trigonid serves to enclose the prefossid lingually so that it opens outwardly at the an- terior extremity of the tooth. The metaconid is either absent, fused with the protoconid or dis- placed posteriorly into the normal position of the metastylid. For the sake of discussion I will consider the latter to be the case. The metaconid is a tall, sharply connate cusp displaced pos- terior to the level of and not joined to the pro- toconid. The crista obliqua and short low ento- cristid join the base of the metaconid thus en- closing the postfossid. The hypoconid is the largest cusp of the tooth and more or less nor- mal in shape for a talpid. The entoconid is like the metaconid in shape. The postcristid joins the base of the entoconid posteriorly. There is a distinct posterior accessory cuspid but no posterior cingulum. A narrow labial cingulum extends from the anterolabial face of the hypo- cond to beneath the paraconid. There is a tiny ectostylid on the cingulum at the base of the hypoflexid. The labial side is nearly mesodont but the lingual side is very shallow thus the entoconid, metaconid, protoconid and hypo- conid are all about the same occlusal level but the latter two are much deeper. There is a rather small cross-sectionally ovate anterior root and a much larger subtriangular posterior root. Of the two unassigned lower molars from Bartlett Mt. UO loc. 2517, one is an M1, UO 25124 (Fig. 96B). This specimen resembles UO 25289 from Black Butte in all essential 106 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 • ' . I Figure 97. ?Talpidae incertae sedis, Bartlett Moun- tain, UO loc 2517, UO 25105 right Ma, occlusal, lingual and labial views. features but differs in detail ( 1) the tooth is higher crowned labially, (2) the labial cingu- lum is greatly reduced and absent around the protoconid, ( 3) the roots slant more posteri- orly, and ( 4) the talonid is proportionately wider. An Ma, UO 25105 (Fig. 97), probably rep- resents part of the same species as UO 25124 on the basis of size. The trigonid and talonid are about equal in length but the talonid is narrower. There is a wide and sharply defined anterior cingulum which is expanded somewhat labially into an accessory cuspid. A labial cingulum bridges the base of the hypoflexid. The trigonid is rather compressed anteropos- teriorly and the protoconid is sharply angulate labially. The prefossid is deep and opens out between the paraconid and the metaconid. The protoconid is the tallest cusp, followed by the metaconid and then by the equally tall para- conid, entoconid, and hypoconid. The hypo- flexid and postfossid are very deep and the crista obliqua, postcristid and entocristid are consequently sharply defined. The entoconid is situated in the posterolingual extremity of the tooth. The crista obliqua and entocristid join and form a slight elevation at their juncture which may be considered the metastylid. The two lophs continue as one loph anterior to the metastylid and join the posterolingual base of the metaconid. The lingual side of the tooth is not as disproportionately short compared to the labial side as in UO 25124. A slight swelling on the posterior side of the entoco- nid may represent the remains of a postero- lingual accessory cuspid. The two roots seem to have been about equally strong. The remote juncture of the crista obliqua and enterocristid, mesodonty, and sharply angulate hypoconid tend to support association with UO 25124. The possibility exists that M3 , UO 25105, and the upper M3, UO 24799, from this locality are modified second molars in a reduced dentition. RELATIONSHIPS: Little can be said as to the re- lationships of the animals represented by these three teeth. The morphology of the talonid and posterior accessory cuspules on the first molars is suggestive of talpids. The bizarre structure of the trigonid, however, is not present in any of the known talpids. The Chiroptera also have lower molars which closely resemble taplids in general morphology and isolated molars may be relatively easily confused with those of tal- pids (e.g. Mystipterus). No attempt was made to survey all of the genera of Chiroptera but the spot checking of various bat genera did not re- veal any more structural similarity of the fos- sils with bats than with the talpids. The trend toward hypsodonty is more characteristic of the Talpidae than of the Chiroptera. Among the other insectivora there are nu- merous possibilities of relationship, however none of the North American or better known Eurasian erinaceids or soricids show any par- ticular similarity with these fossils. These teeth are thus provisionally referred to the Talpidae. Assuming that the development of hypso- donty or trends in that direction are usually not reversed, the Black Butte specimen indi- 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 107 cates an older age ( as do other elements of the fauna) than the specimens from Bartlett Moun- taiil . The rarity of these forms precludes much utility in relative dating at present. TABLE 25 MEASUREMENTS OF THE LOWER TEETH OF ?TALPIDAE incertae sedis M1 M1 U025289 uo 25124 Length ·----- ---- --- -- ---- ------ ------- --- ------- ---- 1.98 2.06"" Width trigonid ------ --- ----- -- ------- -------- --- 0.87 0.93 Width talonid ------ --- -- -------- ------------- -- 1.20 1.33 Height metaconid --- --- -- ----- ---- ----- ----- - 0.67 0.67" Height hypoconid -- --- --- --- ------- ----- ----- 1.38 1.80 ,:, Accessory cuspid broken and metaconid worn. Ma uo 25105 1.77 0.77 1.07 0.77 1.93 108 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 SUMMARY The history of the talpids is perhaps more poorly understood than is warranted by the material at hand; however, tremendous gaps in our knowledge of the group allow a wide margin of error and chance for misinterpreta- tion. No one has yet postulated a general hy- pothesis of the history of the North American moles, largely because the published record is poor and in some cases misleading. I do not pretend to present a complete or cohesive pic- ture of the paleogeographic and phyletic his- tory of the group, but the new material discus- sed herein, as well as much additional material under study, does suggest a pictui·e or rather an imperfect mosaic which allows a closer approximation of that history than was hitherto possible. The earliest talpids known from North America are the Proscalopinae. Reed and Turn- bull ( 1965) suggest that the Proscalopinae, the only assuredly endemic subfamily, were derived from a generalized basal stock inde- pendently of that which gave rise to the rest of the moles sometime in the Eocene. If such a generalized stock was Holarctic in its dis- tribution in the Eocene or earlier then none of its members have been identified as such. The earliest form known from limb material, Cryp· toryctes C. A. Reed, from the Chandronian of Colorado is already highly specialized toward a fossorial existence. The Proscalopinae re- main the only known moles in North America during the Oligocene. In Europe during this time, the "true" moles or Talpinae appear and enjoy a similar freedom from competition with members of other talpid subfamilies. The Pro- scalopinae continue on into Hemingfordian ( early Middle Miocene) time, but by the close of this land mammal age they had become extinct. Before the extinction of the Proscalopinae, two other subfamilies of moles, the Talpinae and Uropsilinae, appear in the fossil record from the early Hemingfordian of Colorado (Wilson, 1960). I postulate that representa- tives of these two groups dispersed from Asia during the Arikareean (Early Miocene). Sig- nificantly, the two new genera, Mystipterus and Scalopoides, are ambulatory and only moderately fossorially specialized and thus offer seemingly little competition to the highly fossorially specialized proscalopines which occur at the same locality. In the absence of the proscalopines, several different and highly specialized moles all within the Talpinae ap- pear during the Barstovian (late Middle Mio- cene). Scalopoides and M ystipterus continue and are joined by still another talpid (Achly- oscapter) of unknown fossorial ability but probably structurally between the afore-men- tioned genera. The highly adapted burrowing moles known from at least three genera are not uncommon in Barstovian microvertebrate as- semblages. At least one of these, Domninoides, continues on into the Clarendonian (Late Mio- cene-Early Pliocene). Scapanus and probably Scalopus arise from one or more of the pre- Clarendonian genera. Mystipterus and Scalo- poides still survive at least in the western faunas. The Hemphillian (Middle Pliocene) tal- pids from the Great Plains are not yet well known but relatives of most of the Clarendon- ian forms from the Great Basin with the pos- sible exception of Mystipterus continue on into Hemphillian time. In addition, a single genus referred to a new subfamily, Gaillardia, ap- pears in both the Great Basin and Great Plains. By the close of the Hemphillian and possibly earlier, the diversity of the talpids decreased, leaving only forms related or ancestral to Re- cent genera. The ancestors or Tertiary repre- sentatives of Neurotrichus and the unique Con- dylura are not unquestionably known. The nearest living relatives of Neurotrichus are in Japan, and may represent the latest talpid dis- persal into North America. The temporal ranges and postulated relation- ships of the better known North American tal- pids are illustrated in Figure 98. The later Tertiary records of talpids de- scribed here indicate several related trends. There is a general but uneven decrease in diver- -------~- 1 9 6 8 3 J = m « a a - o e — * H U TC H ISO N : FO SSIL T A L PID A E FR O M O R EG O N 1 0 9 11so 13ft, 13HoaUBSOO "> ^ J 3Oh 1=1V Ho o O So>s- 3W> 110 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 TABLE 26 MINIMUM AND ESTIMATED MAXIMUM NUMBER of OREGON TALPID GENERA Observed Estimated Number of Age Minimum Maximum"· Localities Recent ....... .................... ............ ...... ........... 2 2 Blanc an ...... .... .. ... .... ....... ....................... ..... 1 1 1 Hemphillian ........................ .... ..... .... ... ...... 3 5 11 Clarendonian ....... .............. .... ...... .......... .. . 3 5-6 3 Barstovian ... ........... ............ .................... .. 6 8 7 * Based on specimens which may represent new taxa when better known. sity (Table 26). Two generic groups (Scalo- poides and Scapanus) show an increase in fos- sorial specialization in the reduction in length of the clavicles, manus, and distal long bones. The reduction in antemolar length in Scapanus ( and probably Scalopoides) is correlative with increased fossorial specialization. The decrease in diversity of the Oregon tal- pids from Barstovian to Blancan eliminates the non-fossorial moles (Mystipterus and Gail- lardia) as well as the slightly or moderately specialized fossorial moles ( Achlyoscapter and Scalopoides) as shown in Table 27. These chronologic trends agree with the general hypo· thesis stated by Shotwell (1967) that "the gen- eral environment of the area of these samples may be said to progress from a more humid wooded one to a condition of semiarid scrub- land at the end of the sequence". Although the fossil record of the various lineages of the Scalopini is spotty, there is no serious evidence to the contrary that all the North American Scalopini could be derived from a Scalopoides-like ancestral stock. The reduced dentition of the known Miocene and Pliocene records of Domninoides indicates that its closest structural relative, Parascalops, does not share a common ancestor later than early Barstovian. The fossi l record of Scalopus is not well documented prior to Hemphill time, but there is no indication that Scalopus or Sca- panns shared a common ancestor later than Barstovian time. The Scapanus complex shows a basic dichotomy by late Clarendonian time, with a conservative line ancestral to the living species and a highly specialized line conver- gent on Scalopus. Thus, it seems that all the Recent genera of the Scalopini can be traced as distinct lineages as far back as the Barstovian land mammal age. The diversity of the Scalo- pini in the Barstovian of Oregon indicates an adaptive radiation of the group in the Middle Miocene. The three tribes, U ropsilinae ( no tribes de- fined), U rotrichini, and Scalopini, shared by North America and Southeast Asia, indicate close faunal ties between these regions at vari- ous times during the later Tertiary. In each case, however, the immigrant appears to have been an ambulatory or only moderately fossor- ially specialized mole. Continuing studies on the talpids are yielding further information as to the type and degree of intra- and inter-conti- nental endemism in this highly specialized group of insectivores. -~------ 1968 Recent Hemphillian HUTCHISON: FOSSIL TALPIDAE FROM OREGON TABLE 27 APPROXIMATE ADAPTIVE GRADE OF OREGON TALPIDAE Ambulatory Aquatic Gaillardi,a Semiaquatic Semifossorial F ossorial N eurotrichus 111 Scapanus Scapanus Clarendonian Mystipterus ? Scalopoides Scalopoides Scapanus Barstovian Mystipterus ,:, From Nevada. Achlyoscapter? Domninoides"" Scapanoscapter Scalopoides Scalopina sp. A Domninoides taxon uo 2465 UO 2495 Mystipterus pacificus''' 12/3 M. cf. M. pacificus M. (Mystipterus) sp. Gaillardia thomsoni ?Neurotrichus columbianus* Scalopoides ripafodiator* 227/14 S. cf. S. ripafodiator 71/7 Scalopoides sp. A Scalopoides sp. B ?Scalopoides sp. C ?Scalopoides sp. D ?Scalopoides sp. E Domninoides g/\ Scapanoscapter* simplicidens''' Scapanus cf. S (Scapanus) sp. S. cf. S. shultzi Scapanus proceridens* S. cf. S. proceridens Scapanus sp. S. nr. ? S. latimanus Scalopina sp. A Scalopini sp. A 4/j Acklyoscapter1'' longirostris* 9/2 1/1 Talpidae incert. sed. Talpidae incert. sed. 2/1 ?Talpidae incert. sed. Unassigned Talpidae 11/5 * New species or genus. 1/1 TABLE 28 Occurrence and Abundance of Oregon Late Tertiary Moles barstovian UO M M 2493 1040 1041 M 1042 2/2 1/1 10/4 2/2 2/2 M 1043 clarendonian i uo uo uo i uo uo uo I2500 2337 2489 |2222 2323 2347 4/1 17/2 1/1 42/3 4/2 2/2 4/2 hemphillian uo uo uo uo uo 2349 2380 2516 2517 2357 12/2 CIT 62 15/2 1/1 16/2 2/2 1/1 5/1 7/1 1/1 3/1 4/2 1/1 8/2 5/1 2/2 2/1 8/1 2/2 1/1 2/1 2/1 8/1 1/1 2/2 BLANCAN UO CIT I UO 2322 375 12223 1/1 1/1 1/1 to to hhi en Co 1 © fcu. C* So £. t- Co © to 3 S3 Co >—. © © ft © © 5* 3 o 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 113 LOCALITY DESCRIPTIONS The following described localities are listed by institution and then numerical order. Lo- calities are described only to the level of sec- tion. More precise information is on file in the respective institutions and the University of California Museum of Paleontology, Berkeley. Method of collection is also given for each locality as an indication of sampling relia- bility. The designation of fine screen refers to industrial woven wire cloth with 24 by 24 mesh per line inch of 0.0128 inch wire, and standard screen refers to woven wire cloth with 15 by 18 mesh per line inch of 0.0114 inch wire. (UO= University of Oregon Museum of Nat- ural History; M=United States Geological Survey locality; CIT=California Institute of Technology-these collections now the prop- erty of Los Angeles County Museum). UO 2222--McKay Reservoir, Umatilla Co., sec. 35, T.2 N., R.32 E., and to sec. 2, T.l N. , R.32 E. Sur- face collected, quarried, and approximately a ton of matrix washed through fine screen. AGE: Hemphillian REFERENCES: Shotwell, 1955, 1956, 1958 A-B, 1967; Brodkorb, 1958. UO 2223- Enrico Ranch, Klamath Co., sec. 28, T.40 S., R.9 E. Surface collected. AGE: Blan can? UO 2322-Krebs Ranch I, Gilliam Co., sec. 30, T.3 N., R.22 E. Surface collected. AGE: Late Hemphillian REFERENCES: Shotwell, 1958 A. UO 2337-Black Butte I, Malheur Co., sec. 11, T.21 S., R.37 E. Quarried and roughly one and one half tons of matrix washed in the field through standard screen. AGE: Clarendonian REFERENCES: Brodkorb, 1961 ; Shotwell and Rus- sell, 1963. UO 2347-0tis Basin, Harney Co., sec. 18, T.20 S., R.36 E. Surface collected and from ant hills. AGE: Hemphillian REFERENCES: Shotwell, 1963. UO 2349-Westend Blowout, Morrow Co., on line of sec. 7 and 8, T.3 N., R.27 E. Surface collected. AGE: Late Hemphillian REFERENCES: Shotwell, 1958 B, 1961. UO 2357-Bartlett Mountain II , Harney Co., sec. 22, T.21 S., R.35 E. Surface collected. AGE: Hemphillian REFERENCES: Shotwell, 1958 B, 1963 UO 2380-Little Valley I, Malheur Co., sec. 33, T.20 S., R.43 E. Surface collected, quarried, and washed with standard screen. AGE: Late Hemphillian UO 2465-Quartz Basin, Malheur Co., sec. 33, T.24 S., R.43 E. Quarried, alxmt three tons of matrix washed in field with standard screen, and about 400 pounds washed through fine screen. AGE: Early Barstovian REFERENCES: Shotwell, 1963, 1967; Hutchison, 1966. UO 2489-West of Riverside, Malheur Co., approxi- mately sec. 9, T.24 S., R.37 E. Surface collected. AGE: Clarendonian UO 2493-Red Basin I, Malheur Co., approximately at junction of sections 10, 11, 14, and 15, T.23 S., R.40 E. Surface collected and quarried for large vertebrates, residue from casts washed through standard screen. AGE : Barstovian REFERENCES: Gazin, 1932; Shotwell, 1963. UO 2495-Red Basin II, Malheur Co., about 30 yards east of UO 2493. Same method of collec- tion as UO 2493. AGE: Barstovian, K/ A date 15.1 X 106 years b. p. REFERENCES: Gazin, 1932; Shotwell, 1963; Evern- den et al., 1964; Hutchison, 1966. UO 2500-Black Butte II, Malheur Co., approxi- mately at junction of sec. 10, 11, 14 and 15, T.21 S., R.37 E. Surface collected and about one and three quarters tons of matrix washed through fine screen. AGE: Clarendonian REFERENCES: Shotwell, 1963, 1967. UO 2516-Little Valley II , Malheur Co., sec. 33, T.19 S., R.43 E. Surface collected, quarried, and washed through fine screen. AGE: Late Hemphillian REFERENCES: Shotwell, 1967. UO 2517-Bartlett Mountain, Malheur Co., sec. 9, T.21 S., R.35 E., part of UO 2239, and CIT 107. About one and one quarter tons of matrix washed through fine screen. AGE: Hemphillian REFERENCES : Shotwell, 1963, 1967 M 1040-Malheur Co., sec. 25, T.35 S., R.42 E. Ant- hill collection. AGE: Barstovian REFERENCES: Repenning, 1967. 114 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 M 1041-Beatty Butte, Harney Co., sec. 14, T.36 S., R.29 E. Anthill collection. AGE : Barstovian REFERENCES: Wallace, 1946; Repenning, 1967. M 1042-Guano Ranch, Lake Co., sec. 34, T.38 S., R.27 E. Anthill collection. AGE : Barstovian REFERENCES: Repenning, 1967. M 1043-Snyder Creek, Lake Co. ,sec. 32, T.35 S. , R.23 E. Anthill collection. AGE: Barstovian REFERENCES: Repenning, 1967. CIT 62-Rome, Malheur Co., sec. 7, R.32 S., R.51 E. Surface collected. AGE: Hemphillian REFERENCE'S: Wilson, 1934, 1937. CIT 375-Arlington, Gilliam Co., near junction of sections 30, 31, 32, and 29, T.3 N., R.22 E. Sur- face collected. AGE: Late Hemphillian 1968 HUTCHISON: FOSSIL T ALP/DAE FROM OREGON 115 REFERENCES CITED Brodkorb, P., 1958, Birds from the Middle Plio- cene of McKay, Oregon: Condor. v. 60, no. 4, pp. 252-255 ----, 1961, Birds from the Pliocene of J untura, Oregon: Quart. Jour. Florida Acad. Sci., v. 24, pp. 169-184 Cabrera, A., 1925, Genera Mammalium-lnsecti- vora, Galeopithicia: Madrid Museo National de Ciencias Naturales, 232p. Campbell, B., 1939, The shoulder anatomy of the moles. A study in phylogeny and adaptation: Amer. lour. Anat., v. 64, no. 1, pp. 1-39 Clark, J.B., M. R. Dawson, A. E. Wood, 1964, Fossil mammals from the Lower Pliocene of Fish Lake Valley, Nevada: Bull. Mus. Comp. Zool. v. 131, no. 2, pp. 29-63 Dobson, G. E., 1882, A monograph of the lnsecti- vora, systematic and anatomical: John Van Voorst, London, pt. 2: pp. 86a-l 72, pls 8-22 Douglass, E., 1903, New vertebrates from the Mon- tana Tertiary: Ann. Carn. Mus., v. 2, no. 5, pp. 145-200 Ellerman, J. R., and T. C. S. Morrison-Scott, 1951, Checklist of Palaearctic and Indian mamals: British Museum (Natural History) , 810p. Evernden, J. F., D. E. Savage, G. H. Curtis, and G. T. James, 1964, Potassium-Argon dates and the Cenozoic mammalian chronology of North America. Amer. Jour. Sci., v. 262, pp. 145-198 Firby, J. R., 1966, New non-marine mollusca from the Esmeralda Formation, Nevada: Proc. Calif. Acad. Sci., v. 33, no. 14, pp. 453-480 Flower, H.F., 1870, Osteology of the Mammalia: London, Macmillan and Co., 344 p. Galbreath, E. C., 1953, A contribution to the Ter- tiary Geology and Paleontology of Northeastern Colorado: Univ. Kansas Paleontol. Contrib. Ver- tebrata, art. 4, pp. 1-120 Gazin, C. L., 1932, A Miocene mammalian fauna from southeastern Oregon: Carn. Inst. Wash. Pub., no. 418, pp. 39-86 Gill, T., 1875, Synopsis of insectivorous mammals: Bull. U.S. Geol., Geog. Surv. Terr., Second ser. , v.l,no.2, pp.91-120 - - - - , 1888, lnsectivora. In the Riverside nat- ural history, 5: pp. 134-156 Green, M., 1956, The lower Pliocene Ogallala-Wolf Creek vertebrate Fauna, South Dakota: J our. Pale- ontol., v. 30, no. 1, pp. 146-159 Gregory, J. T., 1942, Pliocene vertebrates from Big Spring Canyon South Dakota: Univ. Calif. Puhl. Bull. Dept. Geol. Sci., v. 26, no. 4, pp. 307- 446 Gupta, B. B., 1966, The anatomy of the posterior axial region and the hind limb of the Eastern American Mole: Mammalia, v. 30, no. 4, pp. 667- 682 Hall, E. R., 1930, A new genus of bat from the later Tertiary of Nevada: Univ. Calif. Puhl. Bull. Dept. Geol. Sci., v. 19, no. 14, pp. 319-320 Hibbard, C. W., 1939, Notes on additional fauna of Edson Quarry of the Middle Pliocene of Kansas: Trans. Kansas Acad. Sci., v. 42, pp. 457-462 ----, 1941A, New mammals from the Rexroad Fauna, Upper Pliocene of Kansas: Amer. Midland Nat., v. 26, no. 2, pp. 337-368 ----, 1941B, Mammals of the Rexroad Fauna from the Upper Pliocene of Southwestern Kansas: Trans. Kansas Acad. Sci. , v. 44, pp. 265-313 ----, 1950, Mammals of the Rexroad Forma- tion from Fox Canyon, Kansas: Contrib. Mus. Paleontol. Univ. Kansas, v. 8, no. 6, pp. 113-192 ----, 1953, The insectivores of the Rexroad Fauna, Upper Pliocene of Kansas: Jour. Paleon- tol., v. 27, no. 1, pp. 21-32 ----, 1954, A new Pleistocene fauna from Okla- homa: Pap. Michigan Acad. Sci. , Arts, Letters, v. 34, pp. 339-359 - ---, and E. S. Riggs, 1949, Upper Pliocene Vertebrates from Keefe Canyon, Meade Co., Kan- sas: Bull. Geo!. Soc. Amer. , v. 60, pp. 829-860 Hutchison, J. H., 1966, Notes on some Upper Mio- cene Shrews from Oregon: Bull. Mus. Nat. Hist. Univ. Oregon, no. 2, pp. 1-23 Jackson, H. H. T., 1915, A review of the American moles: North Amer. Fauna No. 38, 100 p. James, G. T., 1963, Paleontology and Non-marine , stratigraphy of the Cuyama Valley Badlands, Cali- fornia: Univ. Cal. Puhl. Geo!. Sci. , v. 45, pp. 1-154 Kistin, A. D., 1929, The mole clavicle: Jour. Mamm., v. 10, no. 4, pp. 305-313 Leche, W., 1883, Zur Anatomie der Beckenregion hie lnsectivora, mit besondere Beriicksichtigung ihrer morphologischen Beziehungen zu derjenigen anderer Saugethiere: K. Svenska Vet.-Akad. Hand- lingar sev. 4, v. 20, no. 4, pp. 1-113 Lewis, G. E., 1964, Miocene vertebrates of the Bar- stow Formation in Southern California: U.S. Geol. Surv. Prof. Pap. no. 475-D, pp. 18-23 ., ' ' 116 BULLETIN, MUSEUM OF NATURAL HISTORY, UNIVERSITY OF OREGON No.11 Littleton, R. T. and E. G. Grosthwaite, 1957, Ground-Water Geology of the Bruneau-Grand View Area Owyhee County, Idaho: Geol. Serv. Water-Supply Pap. no. 1460-D, pp. 147-198 Macdonald, J. R., 1963, The Miocene faunas from the Wounded Knee area of Western South Dakota: Bull. Amer. Mus. Nat. Hist., v. 125, art. 3, pp. 139- 238 Matthew, W. D,., 1924, Third contribution to the Snake Creek fauna: Bull. Amer. Mus. Nat. Hist., v. 50, pp. 59-210 ----, 1932, New fossil mammals from the Snake Creek quarries: Amer. Mus. Novitates 540, 8 p. Merriam, J. C., 1911, Tertiary mammal beds of Virgin Valley and Thousand Creek in northwest- ern Nevada: Univ. Cal. Bull. Dept. Geol., v. 6, no. 11, pp. 199-304 Osgood, W. H., 1937, Variable dentition in Chinese insectivore: Field Mus. Nat. Hist., Zool. Ser., v. 20, pp. 365-368 Palmer, F. G., 1937, Geographic variation in the mole Scapanus latimanus: Jour. Marum. v. 18, no. 3, pp. 280-314 Patterson, B. and P. 0. McGrew, 1939, A soricid and two erinaceids from the White River Oligo- cene : Geol. Ser. Field Mus. Nat. Hist., v. 6, no. 18, pp. 245-272 Quinn, J. H., 1955, Miocene Equidae of the Texas Gulf Coastal Plain: Univ. Texas, Bureau Economic Geol., no. 5516, pp. 1-102 Reed, C. A., 1951, Locomotion and appendicular anatomy in three soricoid insectivores : Amer. Mid- land Nat., v. 45, no. 3, pp. 513-671 ----, 1954, Some fossorial mammals from the Tertiary of western North America: Jour. Paleon- tol., v. 28, no. 1, pp. 102-111 ---- and T. Downs, 1958, A fossorial mammal of · unknown affinities from the Middle Miocene fauna of Nevada: Journ. Marum., v. 39, no. 1, pp. 87-91 ---- and W. D. Turnbull, 1965, The mammal- ian genera Arctoryctes and Cryptoryctes: Fieldi- ana, Geol. , v. 15, no. 2, pp. 99-170 Reed, K. M., 1961, The Proscalopinae, a new sub- family of talpid insectivores: Bull. Mus. Comp. Zool., v. 125, no. 14, pp. 473-494 ----, 1962, Two new species of fossil talpid insectivores: Breviora, no. 168, pp. 1-7 Repenning, C. A., 1967, Subfamilies and genera of the Soricidae: U.S. Geol. Survey Prof. Pap. 565, 74p. Schreuder, A. 1940, A revision of the fossil water- moles (Desmaninae): Archives Neerlandaises de Zoologie, v. 4, pp. 201-333 Schwartz, E., 1948, Reivison of the Old-World moles of the genus Talpa Linnaeus: Proc. Zool. Soc. London, v. 118, pp. 36-48 Seeman, I., 1938, Die lnsektenfresser, Fleder- mause und Nager aus der obermiocanen Braun- kohle von Viehausen bei Regensburg: Paleonto- graphica, v. 89, abt. A, pp. 1-56 Shotwell, J. A., 1955, Review of the Pliocene beaver Dipoides: Jour. Paleontol., v. 29, no. 1, pp. 129-144, ----, 1956, Hemphillian mammalian assem- blage from northeastern Oregon: Bull. Geol. Soc. Amer., v. 67, pp. 717-738 ----, 1958A, Inter-community relationships in Hemphillian (mid-Pliocene) mammals: Ecology, v. 39, no. 2, pp. 271-282 ----, 19588, Evolution and biogeography of the aplodontid and mylogaulid rodents: Evolution, v. 12, no. 4, pp. 451-484 ----, 1961, Late Tertiary biogeography of horses in the northern Great Basin: J our. Paleon- tol. , v. 35, no. 1, pp. 203-217 ----, 1967, Peromyscus of the late Tertiary in Oregon: Bull. Mus. Nat. Hist., Univ. Ore., no. 5, pp. 1-35 ---- and D. E. Russell, 1963, Mammalian Fauna of the Upper Juntura Formation, the Black Butte local fauna: Trans. Amer. Philos. Soc., N.S. , v. 53, pt. 1, pp. 42-69 Simpson, G. G., 1945, The principles of classifica- ton and a classification of mammals: Bull. Amer. Mus. Nat. Hist., v. 85, xvi + ~50 p. Slonaker, J. R., 1920, Some morphological changes for adaptation in the mole: Jour. Morph. , v. 34, no. 2, pp. 335-373 Stroganov, S. U., 1945, Morphological characters of auditory ossicles of Recent Talpidae: J our. Mamm., v. 24, no. 4, pp. 412-420 Tedford, R. H., 1961, Clarendonian lnsectivora from the Ricardo Formation, Kern County, Cali- fornia: Bull. Southern Calif. Acad. Sci., v. 60, pt. 2, pp. 57-76 Van Valen, L., 1967, New Paleocene insectivores and insectivore classification: Bull. Amer. Mus. Nat. Hist., v. 135, art. 5, pp. 217-284 Wallace, R. E., 1946, A Miocene mammalian fauna from Beatty Buttes, Oregon: Carn. Inst. Wash. Pub!. Contri. Paleontol., no. 551, pp. 113-134 1968 HUTCHISON: FOSSIL TALPIDAE FROM OREGON 117 Weber, M., 1928, Die Saugetieve. Einfuhrung in die Anatomie und Systematik der recenten und fos- silen Mammalia: Zweite Auflage, Bd. 2, Systema- tischer Teil {Gustave Fischer, Jena) pp. i-xxiv, 1-898 Wilson, R. L., 1965, Techniques and materials used in the preparation of vertebrate fossils: Curator, v.8,no.2,pp. 135-143 Wilson, R. W., 1934, A new species of Dipoides from the Pliocene of eastern Oregon: Carn. Inst. Wash. , Puhl. 453, pp. 19-23 ----, 1937, New Middle Pliocene rodent and lagomorph faunas from Oregon and California: Carn. Inst. Wash. Puhl., no. 487, pp. 1-19 ----, 1960, Early Miocene rodents and insec- tivores from northeastern Colorado: Univ. Kansas Paleontol. Contrib. Vertebrata, art. 7, pp. 1-92 Winge, H., 1917, The interrelationships of the mam- malian genera. V. I. Monotremata, Marsupialia, lnsectivora, Chiroptera, Edentata {Translated from the Danish by E. Deichmann and G. M. Allen, 1941): K¢benhaun, C. A. Reitzels Forlag, pp. i-xii, 1-418 Young, C. C., 1934, On the lnsectivora, Chiroptera, Rodentia and Primates other than Sinanthropus from Locality 1 at Choukoutien: Palaeontologia Sinica, Ser. C, v. 8, no. 3, pp. 1-60 ADDENDUM After the completion of this manuscript, unseen talpid material from pre- viously mentioned localities was examined. Dr. C. E. Ray located and sent to me the alleged Desmana moschata listed in Littleton and Crosthwaite (1957) from Idaho. The specimens are not mammalian but the pharyngeal teeth of a cyprinid ( Cypriniformes, Osteichthyes), probably Sigmopharyngodon Uyeno 1961. The talpid reported by Hibbard ( 1939 from the Edson locality in Kansas is referable to Scalopus. ; PUBLICATIONS Museum of Natural History University of Oregon Eugene, Oregon Bulletins Titles Price No. 1 Cenozoic Stratigraphy of the Owyhee Region, Southeastern Oregon; Kittle- man, L. R. et al., 45 pp., 9 plates 11 figures, (December 1965) $1.50 No. 2 Notes on some Upper Miocene Shrews from Oregon, Hutchison, J. H., 23 pp., 17 figures, (March 1966) $1.25 No. 3 A New Archaic Cetacean from the Oligocene of Northwest Oregon, Emlong, D., 51 pp., 15 figures, (October 1966) $1.50 No. 4 The Archaeology of a Late Prehistoric Village in Northwestern California; Leonhardy, Frank C., 41 pp., 17 figures, (March 1967) $1.00 No. 5 Peromyscus of the Late Tertiary in Oregon; Shotwell, J. Arnold, 35 pp., 11 figures, (June 1967) $1.25 No. 6 Ethnomalocology and Paleoecology of the Round Butte Archaeological Sites, No. 7 No. 8 Deschutes River Basin, Oregon; Roscoe, E., 20 pp., 4 figures (July 1967) $ .75 Its Own Story, The Museum of Natural History; 16 pp. no charge Geologic Map of the Owyhee Region, Malheur County, Oregon; Kittlernan, L., and others, scale 1: 125,000 ( September 1967) $2.00 No. 9 Late Tertiary Geomyoid rodents of Oregon; Shotwell, J. Arnold, 28 figures (November 1967) $1.25 No. 10 Refinements in Computerized Item Seriation; Craytor, W.B., and LeRoy Johnson Jr., (March 1968) $ .75 No. 11 Fossil Talpidae (Insectivora, Mammalia) from the Tertiary of Oregon; Hutchison, J. $1.25 No. 12 Plants of the Three Sisters Region, Oregon Cascade Range; Ireland, 0. 130 pp. $3. 75 II I•