Two New Oligocene Desmostylians anci a Discussion _ *^ of Tethytherian Systematics

DARYL P. DOMNING, CLAYTON E. RAY, and MALCOLM C. McKENNA

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY .

•>... ^ ^Jik. SERIES PUBLICATIONS OF THE SMITHSONIAN INSTITUTION

Emphasis upon publication as a means of "diffusing knowledge" was expressed by the first Secretary of the Smithsonian. In his formal plan for the Institution, Joseph Henry outlined a program that included the following statement: "It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge.'" This theme of basic research has been adhered to through the years by thousands of titles issued in sehes publications under the Smithsonian imprint, commencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series:

Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to the (Marine Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Folklife Studies Smithsonian Studies in Air and Space Smithsonian Studies in History and Technology In these series, the Institution publishes small papers and full-scale monographs that report the research and collections of its various museums and bureaux or of professional colleagues in the world of science and scholarship. The publications are disthbuted by mailing lists to libraries, universities, and similar institutions throughout the world. Papers or monographs submitted for series publication are received by the Smithsonian Institution Press, subject to its own review for format and style, only through departments of the various Smithsonian museums or bureaux, where the manuscnpts are given substantive review. Press requirements for manuscript and art preparation are outlined on the inside back cover.

Rotjert McC Adams Secretary Smithsonian Institution SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY • NUMBER 59

Two New Oligocene Desmostylians and a Discussion of Tethytherian Systematics

Daryl P. Domning, Clayton E. Ray, and Malcolm C. McKenna

SMITHSONIAN INSTITUTION PRESS City of Washington 1986 ABSTRACT

Domning, Daryl P., Clayton E. Ray, and Malcolm C. McKenna. Two New Oligocene Desmostylians and a Discussion of Tethytherian Systematics. Smith­ sonian Contributions to Paleobiology, number 59, 56 pages, 23 figures, 1986.— A new genus, comprising two new species of desmostylians, is de­ scribed from marine Oligocene deposits of the Pacific Northwest. Behemotops proteus, new genus, new species, is based on an immature mandibular ramus and apparently associated skeletal fragments from the middle or (more likely) upper Oligocene lower part of the Pysht Formation of Clallam County, Washington. A related new species, Behemotops emlongi, is founded on a mandibular ramus of an old individual and a mandibular fragment with canine tusk from the uppermost Oligocene (early Arikareean equivalent) Yaquina Formation of Lincoln County, Oregon. The two new species are the most primitive known desmostylians and compare favorably with the primitive Eocene proboscideans Anthracobune and Moeritherium, and to the still more primitive tethythere Minchenella from the Paleocene of China. For many years the Desmostylia were widely regarded as members of the mammalian order Sirenia before being accepted as a taxon coordinate with the Sirenia and Proboscidea (Reinhart, 1953). On the basis of cladistic analysis we go a step further and regard the Desmostylia as more closely related to Proboscidea than to Sirenia because the Desmostylia and Proboscidea are interpreted herein to share a more recent common ancestor than either order does with the Sirenia. This analysis also suggests that the common ancestor of the Proboscidea and Desmostylia (but not the Sirenia) had suppressed P5 and the original last molar. These characters may be conver­ gent with some other mammals. The Superorder Tokotheria McKenna, 1975, was originally thought to be characterized by loss of both P5 and M3. However, because early sirenians do not show these losses, they may have occurred independently in the common ancestor of proboscideans and des­ mostylians and in various other tokotheres. The late Paleocene genus Minchenella Zhang, 1980, from China, is a suitable candidate to be the common ancestor of both the Desmostylia and the Proboscidea. It possesses a small entoconid II on M:<. The Eocene genus Lammidhania Gingerich, 1977, from Pakistan, and the late Paleocene and/ or early Eocene Chinese and Mongolian phenacolophids had not acquired an entoconid II on M^ but are otherwise similar to Minchenella and the anthra- cobunids. The Asiatic occurrence of phenacolophids, Lammidhania, Min­ chenella, and anthracobunids suggests an Asian origin for the Proboscidea and is in accord with the exclusively Pacific distribution of the Desmostylia. We believe that desmostylians were amphibious herbivores that fed on marine algae and angiosperms, and that at least the earlier taxa depended to a large extent on plants exposed in the intertidal zone.

OFFICIAL PUBLICATION DATE i.s handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. SERIES COVER DESIGN: The trilobite Phacops rana Green.

Library of Congress Cataloging in Publication Data Domning, Daryl P. Two new Oligocene desmostylians and a discussion of tethytherian systematics. (Smithsonian contributions to paleobiology ; no. 59) Bibliography: p. Supt. of Docs, no.: SI 1.30:59 1. Behemotops proteus. 2. Beheniotopsemlongi. 3. Desmostylia. 4. Proboscidea, Fossil. 5. Paleontology—Oligocene. 6. Paleontology—Washington(State)—Clallam County. 7. Pa­ leontology—Oregon—Lincoln County. 1. Ray, Clayton Edward. II. McKenna Malcolm C. 111. Title. IV. Series. QE701.S56no. 59 [QE882.D45J 560 s [569'.6] 85-600322 Contents

Page Introduction 1 Acknowledgments 4 Abbreviations 5 Class MAMMALIA Linnaeus, 1758 5 Order DESMOSTYLIA Reinhart, 1953 5 Family Uncertain 5 Behemotops, new genus 6 Behemotops proteus, new species 6 Behemotops emlongi, new species 23 Relationship between Behemotops proteus and B. emlongi 31 History of Desmostylian Systematics 33 Characters Used in Phylogenetic Analysis 37 Comparisons with Early Proboscidea and Minchenella 38 Moeritherium 38 Anthracobune 43 Minchenella 44 Implications for Eutherian Dental Homologies 45 Status of the Tethytheria 46 Desmostylian Lifestyle 47 Conclusions 48 Literature Cited 49

ill

Two New Oligocene Desmostylians and a Discussion of Tethytherian Systematics

Daryl P. Domning, Clayton E. Ray, and Malcolm C. McKenna

Introduction tion of these fossils was made by Barnes et al. (1985). Douglas Emlong's Promethean prowess in dis­ The three specimens to be described herein covery of unprecedented vertebrate fossils, alike are from marine Oligocene deposits ofthe Pacific in beds where many, few, or no collectors pre­ Northwest (Figures 1-3). The first to be found, ceded him, is well known to specialists having from the Yaquina Formation of coastal Oregon, personal knowledge of his activities (Ray, 1977). consists of a massive tusk with a bit of poorly Only a handful of his specimens have thus far preserved bone at the anterior end of the man­ been described (Coombs, 1979; Emlong, 1966; dible (USNM 186889; Figures 16 and 18). At Munthe and Coombs, 1979; S.L. Olson, 1980, the time of its discovery in 1969 it was regarded 1981; Olson and Hasegawa, 1979), but many are by Emlong (field list and pers. comm. to Ray) as under study. Tragically, the flow from the possibly representing a land mammal but more wellspring of these riches ended abruptly on 8 likely a "new and very aberrant desmostylian." June 1980 with Emlong's death (Ray, 1980), but The second specimen to be found, discovered in his already towering reputation as a fossil finder 1976 in the Pysht Formation on the Olympic will be progressively and justifiably widened with Peninsula in Washington, is an immature half every added publication of the results of studies mandible with apparently associated postcranial in progress on the "Emlong Collection." The fragments (USNM 244035; Figures 4-11, 12A- purpose ofthe present communication is to make D, 14A,c, 15A,B,E,F). It was thought by Emlong known several of his more remarkable and pro­ to be a desmostylian or possibly a land mammal, vocative discoveries: putative desmostylians, although he also believed that the one molar much more primitive than any previously known exposed in the field resembled those of sirenians and forging hitherto "missing links" (E.C. Olson, (field list; pers. comm. to Ray, 1976). The third 1981) with primitive proboscideans. Brief men- specimen found by Emlong, a half mandible of an old animal with only Ms preserved (USNM Daryl P. Domning, Department of Anatomy, College of Medicine, 244033; Figures 12F, 16, and 17), collected in Howard University, Washington, D.C. 20059. Clayton E. Ray, Department of Paleobiology, National Museum of Natural His­ 1977 from the Yaquina Formation of Oregon, tory, Smithsonian Institution, Washington, D.C. 20560. Malcolm was described in Emlong's field list as a desmos­ C. McKenna, Department of Vertebrate Paleontology, The Amer­ ican Museum of Natural History, New York, N.Y. 10024. tylian or a land mammal and as elephant-like. In

1 SMITHSONIAN CONTRIBUTIONS TO PALEOBIO OGY

-PYSHT FM.,TYPE SECTION-

-REFERENCE SECTION,UPPER MEM6ER,TWIN RIVER FM. STRAIT OF JUAN DE FUCA —JUANIAN STAGE. TYPE SECTION LIRACASSIS ><^7VI ZONE.TYPE SECTION LIRACASSIS /feXZONE, UPPER PART OF | TWIN RIVERS FM. REFERRED,LOWER PARTOF TWIN RIVERS FM

GETTYSBURG

FIGURE 1.—Index maps of Pacific Northwest: A, sketch map

MILES "^CANADA of part of western Canada, Washington, and Oregon, show­ 0 10 20 30 40 50 ing some major place names, localities mentioned in text, UNITED STATES' and location of enlarged area shown in B; B, detail of part of Twin Rivers and Disque 7.5-minute quadrangles, USGS, along north shore of Olympic Peninsula, showing type- locality of Behemotops proteus, and other relevant localities and boundaries discussed in text; data primarily from Dur­ ham (1944:113, fig. 6), Brown and Cower (1958:2502, fig. 5), Rau (1964:027, pi. 1), AddicoU (1976a:98, fig. 3), Snavely et al. (1978:A118, fig. 8), and Moore (1984a:719). Spelling "Twin Rivers Formation" is that of Durham only.

to resemble those of the specimen from the Twin River [USNM 244035]. This Oligocene specimen is far larger and heavier and I am sure it is a great find, whether desmostylian or land mammal. It came from the Corn- wallius horizon, but is not Cornwallius. It may be related to that giant tusk [USNM 186889] from the Yaquina Formation, and is not far from that area. I am afraid to expose much of the specimen, so I am largely guessing. Emlong's instant intuitions of affinities, al­ though based on unprepared specimens, virtually no literature or comparative material, and almost no formal training, proved in this case as in many others to be uncannily perceptive and to fore­ a letter to Ray of 27 March 1977, two days after shadow our own more belabored conclusions. the discovery, Emlong commented as follows: However, it should be mentioned that his and I stopped at Seal Rock . . . and found the most interesting our views have not been universally accepted by thing of all—a giant desmostylian-like mandible, nearly colleagues who have examined these specimens complete, with teeth [only one as it turned out] that seem between 1969 and now. NUMBER 59

FIGURE 2.—Index maps of coastal Oregon: A, sketch map of part of northwestern Oregon, showing some major place names and location of enlarged area shown in B; B, detail of part of Yaquina and Waldport 15-minute quadrangles, USGS, along central west coast of Oregon, showing location of Seal Rock State Wayside and other localities discussed in text, including location of enlarged area shown in c; C, detail of area including Seal Rock State Wayside, showing localities for Behemotops emlongi and Arretotherium. SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

FIGURE 3.—Correlation of beds containing Behemotops proteus and Behemotops emlongi with some relevant systems of chronology; after Armentrout et al. (1983, chart). Wavy lines at top and bottom indicate continuation of unit beyond limits of chart.

ACKNOWLEDGMENTS.—Of course we are in­ vision of casts from their respective institutions, debted above all to the late Douglas Emlong, we thank Lawrence G. Barnes, the Natural His­ whose perseverance in the field gave us the fas­ tory Museum of Los Angeles County; Philip D. cinating fossils considered herein. Gingerich, the Museum of Paleontology, Univer­ The specimens were skillfully prepared by Ar­ sity of Michigan; J. Howard Hutchison, the Mu­ nold D. Lewis, and the photographs made by seum of Paleontology, University of California; Victor E. Krantz. Figures 1, 2, 5-11, and 14-19 Charles R. Schaff, the Museum of Comparative were made by Lawrence B. Isham, including the Zoology, Harvard University; Richard H. Ted- exquisite drawings for Figures 5-7. Figures 22 ford, the American Museum of Natural History; and 23 were made by Lisa Lomauro. The re­ Mary Ann Turner, Peabody Museum of Natural maining figures were made by Mary Parrish. History, Yale University; and Robert M. West, For the loan of comparative specimens or pro­ then of the Milwaukee Public Museum. NUMBER 59

Warren O. Addicott and Parke D. Snavely, Jr., ABBREVIATIONS.—The following abbrevia­ both of the United States Geological Survey, tions are used to identify the institutions listed:

Menlo Park, California, have shared generously AMNH Department of Vertebrate Paleontology, their special knowledge of the source beds of the American Museum of Natural History, fossils and of biostratigraphy and correlation in New York, N.Y. the Pacific Northwest in general. Kristin Mc- AMNH CA Department of Mammalogy, Comparative Dougall, also of the USGS at Menlo Park, pro­ Anatomy Collection, American Museum of vided foraminiferal analysis of rock samples, as Natural History, New York, N.Y. LACM Natural History Museum of Los Angeles did William A. Berggren of the Woods Hole County, Los Angeles, Cal. Oceanographic Institution. For the privilege of MCZ Museum of Comparative Zoology, Harvard examining Japanese desmostylian material, in­ University, Cambridge, Mass. cluding many unpublished specimens, as well as NMNH National Museum of Natural History, Smith­ for valuable discussions and superb hospitality, sonian Institution, Washington, D.C. NSM National Science Museum, Tokyo, Japan Domning is particularly grateful to Yoshikazu UCMP Museum of Paleontology, University of Cali­ Hasegawa, Yokohama National University; No- fornia, Berkeley, Cal. rihisa Inuzuka, University of Tokyo; Masaichi USGS United States Geological Survey Kimura, Hokkaido University of Education; Os- USNM former United States National Museum, col­ amu Sakamoto, Saitama Prefectural Museum of lections in the National Museum of Natural Natural History; and Yukimitsu Tomida, Na­ History, Smithsonian Institution, Washing­ ton, D. C. tional Science Museum, Tokyo. James M. Clark YPM Peabody Museum of Natural History, Yale generously gave access to his manuscript on a University, New Haven, Conn. new species of Paleoparadoxia from Point Arena, California. Jeheskel Shoshani generously shared with us his data and ideas on paenungulate char­ Class MAMMALIA Linnaeus, 1758 acters and relationships. We have also profited greatly through discussion of sirenians, probos­ Order DESMOSTYLIA Reinhart, 1953 cideans, and desmostylians in general and of our Family Uncertain particular specimens with Lawrence G. Barnes, Kishor Kumar, Earl Manning, Adele Panofsky, DISCUSSION.—The family-level taxonomy of Roy H. Reinhart, Charles A. Repenning, R.J.G. desmostylians is at present unsatisfactory. The Savage, and Andrew Wyss. Further, Barnes, Re- original Family Desmostylidae Osborn, 1905, penning, and Savage have read the manuscript was supplemented by the Cornwalliidae Shikama, critically and have supplied us with important 1957 (emended from the original spelling "Corn- unpublished information. Although the end walliusidae" by Shikama, 1966:153), which was product has benefitted greatly from the assist­ created to accommodate Cornwallius Hay, 1923. ance of all ofthe above-mentioned colleagues, its When Reinhart (1959:94) generically separated remaining deficiencies are of course our respon­ "Cornwallius" tabatai from the type-species C. sibility. sookensis and assigned the former to a new genus, Financial support, especially for field work, Paleoparadoxia, he also erected a third family, was provided in part by the Smithsonian Institu­ Paleoparadoxidae (emended from "Family Pa­ tion through the Smithsonian Research Foun­ leoparadoxia" Reinhart, 1953); however, he re­ dation and the Walcott and Kellogg funds. tained Cornwallius, sensu stricto, in the Desmos­ Domning's visit to Japan was generously financed tylidae. Shikama (1966) instead placed both by the Yamagata Prefectural Museum. Cornwallius and Paleoparadoxia in his Cornwalli­ The sequence of authorship was determined idae. Apart from these inconsistencies already in by lot. the literature, the few and dissimilar genera mak- SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY ing up the order could arguably be arranged in Behemotops proteus, new species almost any arbitrary number of groups, from a single, all-embracing Desmostylidae to mono­ FIGURES 4-11, 12A-D, 14A,C, 15A,B,E,F typic families for each genus. However, until desmostylian diversity and phylogeny are better HOLOTYPE.—USNM 244035, right mandibu­ understood, we prefer to reserve judgment on lar ramus of immature individual with DP4, Pi the familial assignment of the taxa described (or DPi), and P-^-Ms; and, probably from the herein, and we suggest a temporary moratorium same individual, the distal half of the right femur, on new family-level arrangements within the or­ a proximal fragment of the right tibia, and two der. phalanges. Field number E76-14, collected II March 1976 by Douglas Emlong. DIAGNOSIS.—Lower canine probably smaller Behemotops, new genus than in B. emlongi (described below); premolars TYPE-SPECIES.—Behemotops proteus, new spe­ large, high, not molariform; P3 with one domi­ cies. nant cusp (protoconid), P4 with two (metaconid INCLUDED SPECIES.—Behemotops proteus and B. and protoconid); metaconid of P4 not twinned; emlongi, new species. P:, and P4 double-rooted, unlike B. emlongi; DP4 DIAGNOSIS.—Desmostylian differing from trilobate. other members of the order in having seven ETYMOLOGY.—From the Greek sea-god Pro­ lower postcanine cheek-teeth, without marked teus, son of the sea-goddess Tethys and sea-god diastemata; Pi (or DP]) large, caniniform, pro­ Oceanus; old man of the sea and herdsman of cumbent, single-rooted; P9 large, procumbent, the sea-calves (seals) of Poseidon; also able to with root partially or completely divided; molars assume different forms; Latin, masculine, noun brachydont, bunodont, with four principal cusps in apposition. In allusion to (among other things) neither cylindrical nor appressed, and forming a its tethytherian affinities and marine habitat, and square; metaconids of lower molars not twinned; the pronounced ontogenetic changes in its dental all permanent teeth in use together at maturity; morphology. and lingual surface of mandible lacking swelling TYPE-LOCALITY.—South side of Strait of Juan at rear of tooth row. de Fuca, on north shore of Olympic Peninsula, ETYMOLOGY.—From the biblical (Job 40:15- Clallam County, Washington; some 34 km (21 24) Hebraic (and Greek, Latin, and English by miles) west of Port Angeles and approximately adoption) b'hemoth, plural, "great beast," thought 3.6 km (2.2 miles) east of mouth of East Twin by many etymologists and zoologists (including River; 1.6 km (1.0 miles) east and 290 meters Linnaeus, 1758:74, whose latinized Behemot is (950 feet) north of SW corner. Sec. 19, T. 31 N, used herein) to refer to the Nilotic hippopotamus R. 9 W, Twin Rivers Quadrangle, 7.5-minute (and, by others, including Maglio, 1973:2, to the series, USGS; 48°09'38"N, 123°53'55"W; elephant, mammoth, etc.); or b'hemah, singular, wave-cut bench 15 meters (50 feet) north of cliff "beast," conjectured by some to be derived from face (Figure 1). a (pos.sibly artificial) Coptic term p-ehe-mau The postcranial elements came from the same "water-ox"; plus -ops, Greek, suffix, masculine, bedding plane as the mandible but were sepa­ like or similar aspect; in allusion to the Egyptian rated from it horizontally by approximately 6 source of many specimens of tethytheres and to meters (20 feet). the hippopotamian habitus and proboscidean af­ HORIZON.—In place in northwesterly dipping, finities ofthe fossil animal. In any case, we agree concretionary, silty, gray mudstone, within the with Lydgate (1412-1420, volume 2: page xvii) lower part of the type section of the Pysht For­ that the animal's name "doth in latin playne mation (Snavely et al., 1978:A118, Al 19). expresse A beast rude full of cursednesse." AGE AND CORRELATION (Figures IB, 3).—The NUMBER 59

5CM J I

FIGURE 4.—Behemotops proteus, holotype, USNM 244035, from lower part of Pysht Formation of Washington, diagrammatic representation of right mandibular ramus in labial aspect, showing interpretation of dental loci.

rocks from which the holotype of Behemotops vant biostratigraphy and correlation all recom­ proteus was collected are by definition within the mend a more extended discussion of the subject type section of the Pysht Formation (lower part), than would otherwise be warranted. An intro­ explicitly assigned to the Echinophoria rex Zone duction to the broad regional biostratigraphic (now Liracassis rex Zone). The latter is coeval framework can be obtained from the recent pub­ with the Matlockian and "lower" Zemorrian lications by Addicott (1981), Armentrout et al. stages. We regard the lower part of the Pysht (1983), Marincovich (1984), and Moore (1984a, Formation as middle or (more likely) late (but 1984b). For a discussion of the position of the not latest) Oligocene in age. In terms of North boundary between the Oligocene and the Mio­ American Land Mammal Ages, this would imply cene in both marine and continental deposits, see that B. proteus is Orellan or Whitneyan in age. Berggren, Kent, and Flynn (in press). Although the formational assignment and gen­ The Twin River Formation (now Twin River eral age of the holotype are certain, the nature Group) was named by Arnold and Hannibal of deposition, internal conflicts in the biostrati­ (1913:584, 585) for rocks exposed on the coast graphic literature, limited exposures, complexity from "about three miles east of Twin River west of the geology (including folding and faulting), nearly to Pysht Bay," thus including the type- and continuously evolving concepts of the rele­ locality of Behemotops proteus. The locality lies SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

FIGURE 5.—Behemotops proteus, holotype, USNM 244035, from lower part of Pysht Formation of Washington, right mandibular ramus, in lingual aspect. DP4 in original position as found, prior to preparation of specimen; P^ and P4 shown in present state of specimen with bone removed dorsal and lateral to teeth; compare Figure 8. Note opening of coronoid canal above crypt of M3. Scale approximately 2 cm. between Durham's (1944:113, figs. 5 and 6) lo­ Durham (1944) both to his Echinophoria apta and calities A 3683 and A 3684, in a section that he to his E. rex zones. Our locality lies approximately explicitly assigned to the lower part of the Twin 366 meters (1200 feet) east of USGS foramini­ River Formation (Twin Rivers formation of Dur­ feral collecting locality f 11802 and 91 meters ham) and to his Echinophoria rex Zone. (The (300 feet) west of f 11803 (= Durham's 1944, Echinophoria rex Zone should now be called the locality A 3684), both of which localities were Liracassis rex Zone, and the Echinophoria apta assigned to the upper member ofthe Twin River Zone, the Liracassis apta Zone, according to Formation and their faunas to the upper part of Moore, 1984a:719. In the following discussion the Zemorrian Foraminiferal Stage by Rau we retain the older notation of the references (1964:G9, G27, table 6, pis. 1 and 4). cited.) However, Brown and Gower (1958:2502, On the basis of molluscan biostratigraphy of fig. 5) redefined the Twin River Formation so another rock unit, the Lincoln Creek Formation that the reference section for their upper mem­ of southwestern Washington, Armentrout ber, between East Twin River and Murdock (1975:25-29) established the Matlockian mollus­ Creek, includes our locality. Nevertheless, they can stage, which he divided into a lower and an noted (1958:2510) without disagreement that upper zone, equivalent, respectively, to the their upper member included strata assigned by Echinophoria rex and overlying E. apta zones of NUMBER 59

FIGURE 6.—Behemotops proteus, holotype, USNM 244035, from lower part of Pysht Formation of Washington, right mandibular ramus, in labial aspect. DP4 in original position as found, prior to preparation of specimen; bony ramus shown at stage of preparation of Figure 9; P3, P4, and M3 shown as visible in present state of specimen with bone removed to show teeth in labial aspect. Scale approximately 2 cm.

FIGURE 7.—Behemotops proteus, holotype, USNM 244035, from lower part of Pysht Formation of Washington, right mandibular ramus, in occlusal aspect. DP4 shown adjacent; Pi (or DPi), P;(, P4, and M3 shown as if fully erupted and in occlusal position; compare Figures 4, 5, and 8- 10, for true position of these teeth. Opening of coronoid canal, adjacent to posterolabial corner of M3, shown only to indicate its presence, as it would have shifted in position through remodeling of bone as M3 erupted. See Figure 5 for true location as preserved. Scale approximately 2 cm. 10 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

FIGURE 8.—Behemotops proteus, holotype, USNM 244035, from lower part of Pysht Formation of Washington, right mandibular ramus, in lingual aspect. Specimen photographed prior to removal of any detectable bone in preparation, and prior to exposure of mandibular foramen and additional preparation on opening of coronoid canal. the Olympic Peninsula. Addicott (1976a) estab­ 3691, inclusive, and geographically between A lished molluscan stages for the Neogene of Ore­ 3691 on the west and A 3690 on the east (Dur­ gon and Washington. His earliest stage, the Ju- ham, 1944, fig. 6). Our locality lies some 1.6 km anian, shares its type section with the E. apta east of locality A 3691 in westerly dipping strata, Zone, and has its base in the sea cliff approxi­ thus well below the base of the Juanian Stage mately 2 km east of the mouth of East Twin and within the reference section of the E. rex River (not 3 km as stated erroneously in Addi­ Zone. cott, 1976a:97, 98; Addicott, pers. comm. to A sample of the enclosing matrix from the Ray, 21 Sep 1978), between UCMP localities A holotype of 5. proteus, USNM 244035, was pro­ 3680 and A 3691 (see Durham, 1944, fig. 6). cessed for Foraminifera by Kristin McDougall of Addicott's (1976b, fig. 2) generalized section of the USGS, yielding one planktonic and 27 the upper member (= Pysht Formation) of the benthic taxa, indicative of an Oligocene, late Twin River Formation erroneously includes the Zemorrian age. She stated, "This fauna is quite lower half of the upper member of the Twin similar to that found by Rau (1964) in samples f River Formation in the Echinophoria apta Zone 11801 and f 11802. The faunas suggest cool (Addicott, pers. comm. to Ray, 8 Mar 1982). temperatures in a protected upper bathyal to The section from locality A 3691 downward outer neritic environment" (USGS Report on should be referred to the E. rex Zone, including Referred Fossils, shipment number 0-76-8M, all localities numerically between A 3681 and A sample Mf 3256, 24 Aug 1976). Snavely et al. NUMBER 59 11

FIGURE 9.—Behemotops proteus, holotype, USNM 244035, from lower part of Pysht Formation of Washington, right mandibular ramus in labial aspect. Specimen photographed prior to removal of any detectable bone in preparation.

(1980, fig. 15) have presented a paleogeographic shore from Pillar Point State Park eastward to map showing the distribution of late Eocene and 3.5 km west of Low Point. This eastern limit is Oligocene shelf and deep-water marginal seas in in the SE corner (not SW as stated erroneously the Pacific Northwest during the time of depo­ in Snavely et al., 1978:A118), Sec. 19, T. 31 N, sition of the Makah, Pysht, Alsea, and other R. 9 W, whereas our fossil locality is in the SW broadly correlative formations of the region. corner, in the lower part ofthe Pysht Formation. The regional geology of the Olympic Penin­ Armentrout (1977; 1978; 1981:140) re­ sula, including major, previously unpublished re­ stricted his Matlockian Stage to the E. rex Zone, sults of mapping by Snavely and associates in the leaving the succeeding E. apta Zone for the Ju­ area of concern herein, has been portrayed in a anian Stage, thus eliminating the overlap be­ map at the scale of 1:125,000, with a synopsis tween the two stages. As now restricted, the and bibliography (Tabor and Cady, 1978). The Matlockian Stage is wholly beneath the Juanian relevant strata are therein mapped as the upper Stage. member of the Twin River Formation. Snavely As originally proposed, the Juanian repre­ et al. (1978) raised the Twin River to group rank sented the basal Neogene, lower Miocene mol­ and named formations for its upper members, of luscan stage in the Pacific Northwest, with its which the Pysht Formation is uppermost. The base at the Oligocene-Miocene boundary, at 23- type-locality of the Pysht Formation is the section 24 million years ago, within the upper part of exposed for some 18 km in the cliffs and on the the Zemorrian Benthic Foraminiferal Stage (Ad- 12 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

FIGURE 10.—Behemotops proteus, holotype, USNM 244035, from lower part of Pysht Formation of Washington, right mandibular ramus in occlusal aspect. Specimen photo­ graphed prior to removal of any detectable bone, and prior to additional preparation on opening of coronoid canal.

dicott, 1976a:98). Subsequently, however, the upper boundary of the Juanian has been re­ garded as almost or exactly coeval with that of the Zemorrian Stage, with the bases of both at 29 or more million years ago. Thus, the Juanian Stage is now usually regarded as late Oligocene (pre-Aquitanian) in age (Addicott, 1977:163, fig. 3; Allison, 1977:876; Armentrout, 1981:140, 142, 145; Armentrout et al., 1983), although Allison (1976; 1978), Allison and Marincovich (1981:4), and Moore (1984b:4) regard the Ju­ anian as mostly late Oligocene and partly early Miocene in age. Armentrout's restricted Mat­ lockian molluscan stage (= E. rex Zone, = below the uppermost part of the Zemorrian Foramini­ feral Stage), representing time from some 29 to 32 million years ago, is thus earlier in age than latest Oligocene (Armentrout, 1981:145), or is in fact early Oligocene, 33-38 million years old (Armentrout et al., 1983, chart; our Figure 3). ASSOCIATED FAUNA.—Douglas Emlong discov­ ered numerous specimens of vertebrate fossils in the Pysht Formation. The majority of these are skulls of archaic cetaceans under study by R. Ewan Fordyce and others and as yet alluded to only briefly in the literature (Whitmore and Sanders, 1977:310, 311, fig. 2; Fordyce, 1981:1028, 1033). A single specimen from the same area has provided the basis for a new genus and species of penguin-like pelecaniform bird, Tonsala hildegardae Olson, 1980. This material is thought to be similar in age to B. proteus, but

FIGURE 11.—Stereophotographs of casts (whitened) of some postcanine teeth of Behemotops proteus, holotype, USNM 244035, from lower part of Pysht Formation of Washington: A-D, RP:, in anterior (A), medial (B), po.sterior (c), and lateral (D) aspects. E-H, RP„ in anterior (E), medial (F), posterior (G), and lateral (H) aspects. i-L, RM? in anterior (i), medial 0), posterior (K), and lateral (L) aspects. Scale 1 cm. NUMBER 59 13 14 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY faults introduce an element of uncertainty, as the incompletely formed), as preserved (me­ direction of throw of the faults and the relation­ taconid) Length of crown (31-7) ship of the residual accumulation of fossils to the Anterior width of crown (19.0) bedrock and the faults are unresolved. There is Posterior width of crown (21.0) in any case no basis to suppose that any of the material postdates the Pysht Formation. DESCRIPTION.—The holotype, USNM MEASUREMENTS.—The following measure­ 244035, is generally well preserved, with sharply ments of the holotype right mandibular ramus defined surfaces on the bones and all teeth. For of Behemotops proteus (USNM 244035), are in the teeth the only exception is some fracturing millimeters. Those in parentheses are approxi­ in the posterior lobe of DP4 (Figure 10) and a mate, based on incomplete, damaged, or incom­ single, major, anteroposterior fracture with some pletely accessible parts. Those for all alveoli or offset and compression in the crown of M] (Fig­ dental crypts are as preserved, in all cases at least ures 10, 12c). For the bones, the lingual surface somewhat below the real alveolar border. See of the horizontal ramus of the mandible (Figure "Description, Dental Formula" for identification 8) and the caudal surface of the femur (Figure of dental loci. I 4A) and much ofthe tibial surface are curiously eroded (chemically?) and pitted. This destruction of bone apparently removed the thin lingual walls Maximum length of specimen as preserved 235 of the crypts of the unerupted P3, P4, and M3, Depth of ramus below M) 56.5 creating windows through which the initial prep­ DP2 Maximum diameter of crypt (alveolus?) (8.4) aration was done to expose these teeth. The Minimum diameter of crypt (alveolus?) (5.5) phenomenon seems to have affected exactly the DP3 Maximum anteroposterior dimension of (16.3) combined alveoli parts of the specimen that remained covered by DP4 Anteroposterior length of crown (26.8) the remnant of a primary concretion of slightly Width of anterior lobe of crown 10.2 different color and texture than the enclosing Width of medial lobe of crown 11.8 secondary concretion. (15.1) Width of posterior lobe of crown Further preparation was done to expose more PI (or DP,) Maximum height of crown, measured on 39.9 of the crowns of the unerupted teeth, by remov­ medial surface ing much of the labial and dorsal walls of their Maximum diameter of crown (19.7) crypts. The unerupted teeth have been main­ Minimum diameter of crown at same level (12.2) tained in their original positions as found, and P2 Maximum diameter of crypt (23.3) casts have been prepared of their crowns to en­ Minimum diameter of crypt (alveolus?) at (14.2) able viewing in direct occlusal aspect. edge Maximum height of crown 26.0 Dental Formula (Figure 4): Considerable phy­ Anteroposterior diameter of crown 20.8 logenetic significance attaches to the identifica­ Transverse diameter of crown 14.4 tion of dental loci in Behemotops proteus, in part P4 Maximum height of crown 21.5 because of divergent specializations and emphasis Anteroposterior diameter of crown 23.1 or de-emphasis of given loci in related taxa. For Transverse diameter of crown 15.8 Ml Maximum height of crown (metaconid) 14.5 example, inferior tusks are developed from sec­ Length of crown 24.1 ond incisors in moeritheres and other probosci­ Anterior width of crown (18.0) deans, but from canines in desmostylians. Siren­ Posterior width of crown (17.5) ians are less critical for such comparisons because M, Maximum height of crown (metaconid) 21.2 their retention of five premolars in the known Length of crown 31.7 Anterior width of crown 23.7 Eocene taxa, unique among Tertiary placental Posterior width of crown 23.6 mammals (Domning, Morgan, and Ray, 1982), M, Maximum height of crown (base of crown (15.5) sets them phylogenetically apart from desmosty- NUMBER 59 15

Hans and proboscideans, which primitively re­ (AMNH CA 2423) of the living pygmy hippo­ tained only four premolars. The interpretations potamus, Hexaprotodon liberiensis, that died at a of B. proteus presented herein rest in part on comparable stage in its ontogeny. Thus, the DP3 evidence from B. emlongi from the Yaquina For­ of Behemotops proteus was probably double- mation of Oregon, described later in this paper. rooted. Only the deepest apical parts ofthe DP3 Some aspects ofthe dental formula and succes­ alveoli are preserved, because a considerable, sion in the holotype of Behemotops proteus seem although unknown, amount of bone is missing certain. There is no reason to question the iden­ anterior to DP4 along the dorsal, alveolar-mar­ tification of Ml-Ms as conventionally understood ginal edge of the horizontal ramus. (but see below, "Implications for Eutherian Den­ Anterior to the P3 locus the situation becomes tal Homologies"). Mi is considerably smaller than less clear because of the loss of bone and the M2 and M3, and is already substantially worn, absence of all teeth except one (interpreted be­ whereas M^, although fully in occlusal position, low as P] or DPi). We regard the large, simple is virtually unworn, and the crown of M3 is in­ alveolus anterior to the P3 locus as in fact the completely formed and remains in its crypt (a ventral remnant of the crypt for a large, fully small window to which had opened in the bone formed, but probably unerupted P2. There is at above it). The early emplacement and attrition depth a slight vertical crest in the bone on each of a relatively small M] is a common feature in side of the alveolus, suggestive of incomplete bunodont herbivores, and is strongly pro­ subdivision into anterior and posterior moieties nounced in the desmostylian genus Paleopara­ of the root that occupied it. On the anterolabial doxia (Figures 12E, 19). The identification of DP4 border of this large crypt or alveolus is an apical is also definite, because it was found, deeply remnant of a much smaller alveolus, probably worn, in place immediately anterior to and in for the anterior root of DP2. The anteriormost contact with Mj. Its posterior root, although tooth present (forming the anteriormost pre­ broken, is still in place anterior to Mi. Moreover, served part of the specimen) is a large, procum­ DP4 is trilobate, again as in many herbivorous bent, essentially caniniform tooth, with its crown mammals, for example artiodactyls (see Figure and at least much of its root fully formed but 20), phenacodonts (West, 1971, fig. 1), some lying in its crypt deep within the mandible at the specimens of Moeritherium (YPM 34764) but not time the animal died. This tooth, which we re­ especially so in the specimen figured by Andrews gard as Pi or DP], looks as if it would also fit the (1906:110, fig. 43), Phiomia serridens (Andrews, crypt of Pj reasonably well, suggesting that the 1906, pi. 18), and Deinotherium and Gomphother- P9 also was a relatively simple tooth. Unfortu­ ium (Lartet, 1859, pi. 13: fig. 4c; pi. 14: fig. 4c. nately, the holotype of B. proteus retains no ves­ See also Frick, 1926; 1933). If DP4 and M1-M3 tige of the canine or incisive loci. Enough bone have been identified correctly, then the tooth of the horizontal ramus remains beneath Pi (or occupying the crypt directly under DP4 (and the DPi) to demonstrate that the canine or any other anterior part of M]) must be P4 and the tooth in enlarged anterior tooth could not have extended the crypt immediately anterior to it must be P3. to the rear past Pi (or DPi) at the stage of The two alveoli, subequal in size and circular in ontogeny represented by the holotype. There is cross section, lying one each anterior and poste­ apparently no room for such a tooth, and, fur­ rior to the principal cusp of P3, are for the two ther, there is no hint of growth ofthe mandibular roots of DP3. Between and slightly lingual to ramus in anticipation of the eventual emplace­ these alveoli is what appears at first sight to be a ment of such a tusk. Rather, the symphyseal third alveolus but which may be in reality the region may have resembled that of Hexaprotodon, apex ofthe P3 crypt below (Figure 10). A similar with the Pi (or DPi) of B. proteus occupying the resorption window can be seen in a specimen position ofthe canine in Hexaprotodon. 16 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY NUMBER 59 17

Dentition: The terms developed by Tobien the crown into two unequal sections, a narrower, (1978 and elsewhere) for the description of mas- flattened, posteromedial one and a broader, todont molars, including Moeritherium, can be rounded, anterolateral one. Later in ontogeny applied readily to the molars of Behemotops, and this tooth almost certainly would have rotated are noted herein throughout the description. clockwise (as viewed dorsally) more than 45° All teeth except M3 show finely to coarsely about its longitudinal axis, so that the crests crenulated enamel, wrinkled or pustulose, except would have become anterior and posterior, the where smoothed by wear. smaller, flatter surface lingual, the larger, more The only deciduous tooth preserved is the convex surface labial, and the greatest diameter deeply worn, trilobate DP4 (Figures 4-10). Wear anteroposterior. It would have stood much has reduced the occlusal plan to a series of three higher than P3 (and perhaps somewhat higher interconnected, subcircular rings of enamel, in­ than P2) when fully erupted. creasing in size posteriorly. The cusps of the P2 (Figures 4, 7, 10) is represented only by its upper deciduous premolars of Moeritherium crypt (alveolus). The crypt is similar enough to (Schlosser, 1911, pi. 13: fig. 9) have a similar the shape of Pi (or DPi) to suggest a very similar round outline at the stage of wear demonstrated; tooth. However, its orientation, if that of Pj (or an only minimally trilobate lower deciduous pre­ DPi) resembled it, would imply rotation of Pi (or molar has been reported by Andrews (1906:110, DP|) as suggested above. A single root was pres­ fig. 43). However, DP4 of YPM 34764 is trilo­ ent, but traces of fusion of two roots, perhaps bate. The anterior lobe of DP4 in B. proteus is separated at some earlier phylogenetic stage, are bordered anteriorly by a distinct precingulid. indicated on the alveolar walls. Pi (or DPi) (Figures 4-10) is represented by a P3 (Figures 4-8, 1 1A-D, 12A) is dominated by well-formed but unerupted tooth. The tooth is a single, high, bluntly conical cusp (protoconid), simple, unicusped, essentially caniniform, single- with a weak posterior crest. Low on its anterolin­ rooted, robust, and has a bluntly rounded tip gual slope is a well delimited, subcylindrical, reminiscent of the principal cusp of P3 in the somewhat recurved cusp (paraconid); on its pos­ same specimen. There are faint longitudinal terolingual slope is a weaker, lower, less inde­ crests on opposite sides of the crown from ap­ pendent cusp (metaconid); at its anterolabial proximately 7 to 13.7 mm down from the tip of base, labial to the paraconid, is a small cingular the tooth, from which point the crests each give cusp, which, with a meager shelf at the anterior way to a single row of small cuspules or coarse base of the paraconid, represents a precingulid. crenulations of subequal size, best exposed on Small cusps on the posterior slopes of the proto­ the medial surface of the crown where the cus­ conid and metaconid near their bases may rep­ pules are some six in number. These crests divide resent the hypoconid and entoconid, respec­ tively. Posterior to that and spanning almost the breadth of the tooth is a strong postcingulid FIGURE 12.—Stereophotographs in occlusal aspect of some consisting of some five small cusps, the labialmost inferior postcanine teeth of Behemotops proteus from lower three of which are largest and subequal to one part of Pysht Formation of Washington, Paleoparadoxia ta­ another in size. A tiny, marginal, basal cusp oc­ batai from Izumi locality, Japan (Shikama, 1966:12), and Behemotops emlongi from lower part of Yaquina Formation cupies the base of the crease delimiting the pro­ of Oregon: A-D, Behemotops proteus, holotype, USNM toconid and possible hypoconid. The anterior 244035: A, RP,; B, RP4; c, RMi-RM^; D, RM3. A, B, and D and posterior roots are separate, in contrast to are photographs of casts (whitened), because original teeth those of B. emlongi. cannot be viewed in exact occlusal aspect; compare Figures P4 (Figures 4-8, 1 1E-H, 12B) is slightly larger 4-10. E, Paleoparadoxia tabatai, cast of neotype, USNM 26375, LP3-M3. F, Behemotops emlongi, holotype, USNM than P3, but not so tall. It is dominated by two 24033, LM3. Scale 1 cm. bluntly conical cusps of almost exactly equal SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY height. The lingual of these (metaconid) lies conid. A low, worn, independent cusp occupies slightly posterior to a point directly medial to the the center of the occlusal surface of the crown. labial cusp (protoconid), and is simple, with only This cusp lies adjacent to the anterolingual ter­ two minor subsidiary cusps, one each at the base mination of a. cr. 2 and on the anterior slope of of its anterolingual and posterolingual slopes; the the transverse valley. It may represent a conelet metaconid is not twinned. The metaconid is em­ subsidiary to the metaconid but, if so, it is pos­ braced labially by the more complex protoconid, teriorly displaced. A low, broad hypolophid ex­ with its strong, cuspidate paracristid and proto- tends down the labial slope of the entoconid. cristid curving about the anterolabial and poster­ The presence of the hypolophid, together with olabial parts of the base of the metaconid. The the nature of the wear on the crown, excessive paracristid terminates in a poorly differentiated on the anterior walls of the hypoconid and en­ and somewhat bifid paraconid near the middle toconid and on the posterior walls of the proto­ ofthe anterior wall ofthe crown. A well-marked conid and metaconid, impart a mildly lophodont and cuspidate precingulid lies below the para­ character to this tooth, not clearly evident in the cristid along the anterolabial base of the crown. little-worn Mj and even less so in the unerupted Posterior to the metaconid and protoconid are M;,. The crenulated postcingulid rises from its two low cusps (entoconid and hypoconid), the lingual and labial extremities to a central hypo­ latter the larger, lying posterior to the termina­ conulid. The anterior and posterior roots are tion of the protocristid. The entoconid and hy­ well separated labially, but are closer together poconid together are bordered posterolingually, lingually. posteriorly, and posterolabially by a strong M^ (Figures 4-10, 12c) is almost an exact but postcingulid consisting of five to six low cusps of enlarged replica of M]. It is almost unworn and varied size, shape, and height. The anterior and therefore reveals the features of the crown posterior roots are apparently still separate, in clearly. The base of the crown forms a bulbous contrast to those of B. emlongi. collar, especially distinct on the anterior half of Ml (Figures 4-10, 12c) is by far the smallest the tooth. The cingulid, although variably devel­ ofthe molars, being only slightly larger in crown oped, is interrupted only lingual to the entoconid area than P., but more nearly rectangular. The and labial to the hypoconid. The metaconid is basic plan of the bunodont, brachydont tooth is not twinned. Anterior to the postcingulid and simple; it consists of four independent, subequal, side by side posterolabial to the entoconid and major low conical cusps, the protoconid, meta­ posterolingual to the hypoconid are two small conid, hypoconid, and entoconid, each occupy­ cusps on either side of the midline of the crown ing a quadrant of a square, to which attach strong (at least the worn labial member of the pair is pre- and postcingulids. The cingulids are not identifiable in Mi, and the lingual member could continued on the lingual and labial walls of the be present but obscured by wear and faulting of tooth, although suggested by a bulbous expan­ the tooth). These cusps are similar to the conelets sion of the anterolabial part of the base of the of Gomphotherium (Tobien, 1978, fig. 1) but are protoconid and by a pustulose shelf at the labial not placed quite so far forward. The cristid ob­ outlet of the valley between the protoconid and liqua (a. cr. 2) of M- supports two small cuspules hypoconid. The conical shape ofthe protoconid separated by a notch representing the transverse is modified slightly by a very weak paracristid valley. These cuspules are similar to those of Mi, (anterior crescentoid of first pretrite, a. cr. 1) but unworn. The anterior of the two lies at the that extends anterolingually but lacks an identi­ base of the posterolabial slope of the metaconid, fiable paraconid. The metaconid is not twinned. and might be regarded as its conelet, analogous A cristid obliqua (a. cr. 2) occupies an analogous in position to that ofthe entoconid. The postcin­ position on the anterolingual slope of the hypo­ gulid rises to a median apex as in M|. Insofar as NUMBER 59 19

can be seen, the roots are as in Mi. each column to fill the spaces between the col­ M3 (Figures 4-8, 1 li-L, I 2D) is basically simi­ umns. The clearest example of this that we have lar in construction to Mi and M2 insofar as its seen is in NSM 5600, the skull and mandible of incompletely formed enamel cap reveals. Its Desmostylus described by Yoshiwara and Iwasaki roots had not formed at the time ofthe animal's (1902). On plate 2 of their paper an unerupted death. However, the crown does show some right upper molar ("M2") is shown. Examination uniquely interesting features. For example, all of the actual specimen, now further prepared, four principal cusps are simpler, higher, more discloses the condition described above, the tooth nearly cylindrical than bluntly conical, and thus consisting of separate, very slender columns (Fig­ more widely separated from one another. The ure 13). Thus, there is no reason to doubt that a metaconid is not twinned. There is no evidence similar although less extreme process of cusp of a paraconid or paracristid (a. cr. 1). The cristid thickening took place during the dental ontogeny obliqua (a. cr. 2) is sharply defined, as is a com­ of B. proteus and no reason to suppose that the plementary crest (hypolophid?) extending anter­ difference in M3 cusp thickness between the lat­ olabially from the entoconid. The anteriormost ter and B. emlongi is of any major taxonomic ofthe two low cuspules seen on the cristid obliqua importance. Due allowance for additional enamel is a low, trihedral cusp, similar in position to a deposited on the sides (and to a lesser extent, on more rounded cuspule in the same position on the tips) of the cusps of B. proteus would bring M2, and to the worn cuspule on Mi, but less the dimensions of its M3 into tolerable agreement clearly tied to the metaconid. The postcristid is with those of B. emlongi. sharply defined, and posterolabial to the ento­ Osteology: The mandible of the holotype, conid it possesses a low, sharply pointed conelet. USNM 244035, is robustly proportioned; its hor­ A small crest descends along the posterolingual izontal ramus was certainly thicker than imme­ base of the hypoconid. The crenulated postcin­ diately suggested by the specimen as preserved, gulid shelf is broader anteroposteriorly than in in view ofthe postmortem loss of bone over much Ml or M2, and lends an angular, V-shaped outline of its lingual surface. A remnant of bone below to the posterior end of the tooth. The crown is the posterolingual corner of M2 and the thickness not sufficiently formed to show the condition of ofthe ramus at Pi (or DPi) and P2 indicates that the cingulid elsewhere. the horizontal ramus was originally several mil­ Apparently, the odd "bunostylodont" nature limeters thicker over much of its expanse (Fig­ of the M3 of Behemotops proteus must be attrib­ ures 8 and 10). The ventral margin ofthe man­ uted to its incomplete ontogenetic development; dible is essentially straight as far as it is preserved. at maturity it would presumably have had more Two small mental foramina open adjacent to the swollen cusps like those ofthe other molars. This crypt of P3 (Figure 9); undoubtedly others were is indicated by two pieces of evidence. First, M3 present more anteriorly. The ostensible foramen contrasts with all other teeth of the specimen in midway below DP4 (Figure 9) is in fact a window not having wrinkled or crenulated enamel; this adjacent to the tip of the crown of P4. The suggests that the outer layers of enamel had not ascending ramus of the mandible has its anterior yet been deposited. Second, at least in Desmostylus and (as far as preserved) posterior margins nearly it is certain that the columns of the molars at­ vertical. The rounded articular condyle is ele­ tained nearly their full height long before reach­ vated well above the plane of occlusion; the an­ ing their final diameter. The developing molar, terior margin of the ascending ramus is nearly therefore, consisted of a group of high, slender straight and approximately perpendicular to the columns, initially not connected at their bases plane of occlusion. The coronoid process is and often found isolated. Only in later stages of broad, smoothly curved, and has a posterior development was sufficient enamel deposited on hook. The mandibular foramen lies midway be- 20 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

FIGURE 13.—Desmostylus hesperus, NSM 5600, from Togari, Gihu Prefecture, Honshu, Japan, originally described by Yoshiwara and Iwasaki (1902); right upper molars in labial aspect. Note slender and distinctly separated enamel columns of unerupted and incompletely formed posterior molar, in contrast to thick, appressed columns of fully formed and erupted anterior molar. tween the anterior and posterior margins of the thought to be part of the holotype. The distal ascending ramus and directly posterior to the half of the femur (Figure 14A,C) displays an an­ crypt for Mi (Figure 5). The mandibular canal teroposteriorly narrow shaft expanding distally passes forward from the foramen, under the into a broadly flattened extremity with a broad ventrolateral edge ofthe developing M3, and less patellar facet; the narrowness of the shaft is than 4 mm from it. A coronoid foramen (coro­ strongly reminiscent of the condition in Paleo­ noid canal) pas.ses through the base of the coro­ paradoxia. Its incompleteness and the erosion of noid process at the rear of the developing alveo­ the bone on its caudal surface make precise as­ lus of Ml (Figures 5 and 8). sessment of its proportions impossible. The distal Very little can be said about the postcranial epiphysis was not coossified with the diaphysis, skeleton, represented only by a few pieces as it is slightly displaced. The tibia is represented NUMBER 59 21

FIGURE 14.—Behemotops proteus, holotype, USNM 244035, from lower part of Pysht Formation of Washington, distal part of right femur, in lateral (A) and cranial (c) aspects; Paleoparadoxia tabatai from Izumi locality, Japan (Shikama, 1966:12), cast of neotype, USNM 26375, distal part of right femur, in lateral (B) and cranial (D) aspects. 22 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

-f^s:

V*%

5 CM

FIGURE 15.—Behemotops proteus, holotype, USNM 244035, from lower part of Pysht Formation of Washington, phalanges, lacking proximal epiphyses, in dorsal (A, B) and palmar (E, F) aspects; Paleoparadoxia tabatai from I/umi locality (Shikama, 1966:12), cast of neotype, USNM 26375, phalanges, in dorsal (c, D) and palmar (G, H) aspects. NUMBER 59 23

by a proximal fragment with separate proximal Quadrangle, 15-minute series, USGS (Figure 2). epiphysis. The broken cross section of the dia­ Same horizon as USNM 244033, Yaquina For­ physis, some 60 mm from its proximal end (some mation, from the foot of the sea cliff, approxi­ 80 mm with the epiphysis in place), indicates a mately 43 meters (140 feet) north ofthe isthmus broad, anteroposteriorly flattened tibia with a joining Tourist Rock to the mainland. The isth­ low tibial crest and very shallow fossae. The two mus to Tourist Rock lies at latitude 44°29'50"N, phalanges lack proximal epiphyses. They are longitude 124°05'00"W. strikingly broad, flat, and splayed distally (Figure HORIZON.—In coarse grit layer of lower part 15A,B,E,F), also reminiscent of Paleoparadoxia. ofthe Yaquina Formation (Snavely et al., 1976). AGE AND CORRELATION (FIGURE 3).—The hol­ Behemotops emlongi, new species otype and referred specimens of Behemotops em­ longi, USNM 244033 and 186889, are from the FIGURES 12F, 16-18 Yaquina Formation of western Oregon, assigned HOLOTYPE.—USNM 244033 (Emlong field to the Juanian Stage (Addicott, 1976a:99; Ar­ no. E77-21), nearly complete left mandibular mentrout, 1981:141) and are somewhat younger ramus, considerably damaged in region of sym­ than the holotype of B. proteus from the lower physis and posterior border of ascending ramus part of the Pysht Formation of northwestern and condyle; only M3 present, but with partial or Washington. The area has been mapped by complete alveoli of all adult teeth; collected by Snavely et al. (1976), who noted that the only Douglas Emlong, 25 March 1977. formation exposed in the area is the lowermost DIAGNOSIS.—Lower incisive alveoli three in part ofthe Yaquina Formation. number, subequal in size, with round cross sec­ Most of the Yaquina Formation is usually as­ tions; lower canine (probably unlike that of B. signed to the Zemorrian Stage, with only the proteus) greatly enlarged, tusk-like, procumbent, uppermost part possibly referable to the Sauce- laterally compressed; longest diastema between sian Stage (Snavely et al., 1969:38; Rau, P2 and P3, shorter than anteroposterior diameter 1981:81; Armentrout et al., 1983, chart). The of either tooth; P3 and P4 alveoli simple, housing exposures in the vicinity of Seal Rock have been single-rooted teeth, unlike B. proteus; anterior identified explicitly as the lower part of the Ya­ half of mandible massive; symphyseal region quina Formation (Emlong, 1966:2; based on broad, shovel-like. pers. comm. from Snavely) and thus pertain to ETYMOLOGY.—For the late Douglas R. Em­ the Zemorrian part of the formation. long, collector extraordinary. ASSOCIATED FAUNA.—The Yaquina Forma­ TYPE-LOCALITY.—183 meters (200 yards) tion in Lincoln County, Oregon, has produced, south-southeast of Elephant Rock (Figure 2c), in addition to the two specimens of Behemotops Seal Rock State Wayside, Lincoln County, Ore­ emlongi, a fairly rich fauna of pinnipeds, desmos­ gon; Sec. 25, T. 12 S, R. 12 W, Waldport Quad­ tylians, and cetaceans (Ray, 1977:428, 429). One rangle, 15-minute series, USGS (Figure 2); inter­ of the cetaceans, Aetiocetus cotylalveus Emlong, tidal bench 76 meters (250 feet) west of the cliff 1966, was collected from the upper part of the face. Yaquina Formation, approximately 0.8 km (0.5 REFERRED SPECIMEN.—USNM 186889 (Em­ mile) north of Seal Rock State Wayside, and long field no. 555), much fragmented, poorly additional skulls are now known from the Ya­ preserved anterior part of right mandibular ra­ quina Formation but are not as yet described mus with root of P3 and essentially complete (Whitmore and Sanders, 1977:317; L.G. Barnes, canine tusk; collected by Douglas Emlong, April pers. comm.). Other Desmostylia from the Ya­ 1969. Seal Rock State Wayside, Lincoln County, quina Formation include specimens of Cornwal­ Oregon; Sec. 25, T. 12 S, R. 12 W. Waldport lius, under study by Reinhart (1975; 1982:550, 24 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

FIGURE 16.—Diagrammatic representation of mandible of Behemotops emlongi from lower part of Yaquina Formation of Oregon, mature individual, in occlusal (A) and left lateral (B) aspects; based on both the holotype, USNM 244033, and referred specimen, USNM 186889. Degree of posterior divergence of mandibular rami and length and shape of incisors (but not the fact of their presence as indicated by alveoli) are based primarily on analogy with Paleoparadoxia. NUMBER 59 25

FIGURE 17.—Behemotops emlongi, holotype, USNM 244033, left mandibular ramus from lower part of Yaquina Formation of Oregon, in lingual (A), labial (B), and occlusal (c) aspects. 26 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY fig. 1), collected from the same area and horizon somewhat damaged. See "Description, Dental (lower part of the Yaquina Formation) near Seal Formula" for identification of dental loci. Rock as the specimens of Behemotops emlongi. The USNM 244033, holotype, left mandibular ra­ original material of Cornwallius sookensis is from mus: the Sooke Formation on Vancouver Island, in large part at least coeval with the Yaquina For­ Maximum length of specimen as preserved 396 Maximum height of mandible at coronoid process 222 mation (Durham, 1944:113; Addicott, 1976a: Depth of horizontal ramus below M) 114.1 99). Depth of horizontal ramus behind M.s 128.3 Also from the same horizon in the lower part Maximum width of jaw at symphysis as preserved (119.3) of the Yaquina Formation is the only land mam­ Breadth of jaw behind M3 (35) Canine tusk alveolus, maximum dorsoventral (75.9) mal known thus far from the formation, a frag­ diameter ment of a maxilla with deeply worn M'""' and M^ Canine maximum transverse diameter (29.4) (USNM 187125; Emlong field no. 291) of an Alveolar length anthracothere (Ray, 1977:431). Richard H. Ted- P, (or DP,)-M:, (208?) ford has examined this specimen and has pro­ P.-M:, (176) vided us with the following statement (7 Apr P:,-M, (135) 87.8 1983). M,-M, P] (or DPi) (alveolus strongly inclined forward) (36.7) The teeth, although worn and broken, retain the diag­ Pj 24.9 nostic divided mesostyle and loss of paraconule found in P:i 21.5 only the genus Arretotherium among known North Amer­ P4 16.1 ican anthracotheres. The dimensions of the M^ of USNM M, 20.3 187125 are: length, 24.6 mm; width across the mesostyle, M, 23.0 25.8 mm. All three of the described species of Arreto- Ms 40.6 therium[A. acrirf^ni Douglass, 1902; A. leptodus{Maixhevi, Alveolar width 1909); and A. fricki Macdonald and Schultz, 1956] have P, (or DP,) (18.6) upper molars of similar size and proportions, so it is not PL' 15.6 possible to determine the precise affinities ofthe Yaquina P. (12.3) anthracothere without further evidence. It is closest in P4 11.4 size to the holotype of A. leptodus from the late Arika­ M;, (anterior) 6.9 reean. As now recognized (Macdonald, 1956, 1963), the M:i (posterior) 18.0 genus is confined to the early Miocene (late Arikareean Diastema between P-^ and P^ 17.0 through early Hemingfordian) and seems to succeed the M:, closely related late Oligocene (Whitneyan through early Maximum height of crown (metaconid) 18.0 Arikareean) genus Elomeryx. However, an extension of Length of crown 37.6 the geological range ofthe genus, in the form of a species Anterior width of crown 24.2 like A. leptodus, into the early Arikareean seems indicated Posterior width of crown 28.6 by unpublished material from Nebraska, South Dakota, and Wyoming contained in the Frick Collection at the USNM 186889, referred anterior fragment of American Museum of Natural History. These records right mandibular ramus: would push the range zone of Arretotherium into the late Oligocene, in agreement with the assignment of a late Combined breadth of alveoli ofthe three incisors (63) Oligocene age for most of the Zemorrian and approxi­ Maximum depth (as preserved) of I^ alveolus (55) mately equivalent Juanian stages. Maxinmm mediolaleral diameter of f, alveolus (17.9) Maximum anteroposterior diameter of I3 alveolus (H.8) MEASUREMENTS.—The following measure­ Maximum depth (as preserved) of F, alveolus (33.5) ments of the holotype and referred specimen of C>anine tusk Behemotops emlongi are in millimeters. Those in maximum length (as preserved) 267 maximum dorsoventral diameter (63.5) parentheses are approximate, based on incom­ (near alveolar margin) plete or damaged parts of specimens. Those for maximum transverse diameter (43.1) all alveoli are as preserved, in all cases at least (near alveolar margin) NUMBER 59 27

Ps root tion of crests on each side. There is a short anteroposterior diameter (23.8) diastema anterior to P3, perhaps resulting from transverse diameter (15.7) maximum length (as preserved) 43.1 the progressive forward tilting of P2 and Pi (or DP]) in accommodation to their position dorso- DESCRIPTION.—The holotype and referred medial to the alveolus ofthe massive canine tusk. specimen of Behemotops emlongi, USNM 244033 The alveolus of P2 is larger than those of P3 and and 186889, as well as specimens of other taxa P4, anteroposteriorly elongate, simple, and has from the same horizon in the Yaquina Forma­ crests on its lingual and labial walls, reflecting tion, exhibit a peculiar preservation. The bone indentations in the root of P2. The alveolus of Pi is weak and, especially in USNM 186889, has (or DPi) is similar in size to that of P2, strongly grit from the enclosing matrix pressed into its inclined forward, and simple, with no indication surface as if to become almost an integral part of of subdivided roots. The alveolus of the canine the bone (Figure 18). The alveolus ofthe canine tusk is very large, laterally compressed (possibly tusk of USNM 244033 may well have been com­ in part postmortem), and extends posteriorly to pressed laterally after burial. A part ofthe lingual a point at least below Mi. There are vestiges of wall of the alveolus had disintegrated; for this the deepest parts of the simple alveoli of Ii and reason during preparation the mandible was I2, and possibly of I3, but the amount and quan­ strengthened in that area with fiberglass (Figure tity of the bone preserved in this region would 17A,C). The poor preservation of surface detail be inadequate for secure interpretation were it in both specimens leaves some doubt as to the not for the existence of USNM 186889. If cor­ detailed character of the anterior premolar al­ rectly interpreted, the apices of the alveoli of veoli (Figures 17c, 18c), and makes illustration these three teeth converge in a triangular ar­ of the incisive alveoli impractical. rangement. Dental Formula (Figures 4 and 16): The hol­ The bone in USNM 186889 is poorly pre­ otype and referred specimen of Behemotops em­ served and meager, but is just sufficient to pro­ longi are compatible with the interpretation of vide a reliable basis for establishment of the the dental formula of B. proteus and add infor­ anterior mandibular dental formula. P3 is repre­ mation on the permanent incisors and canine. sented by a remnant of a robust, forwardly tilted, USNM 244033 has its well-worn Ms in place. Its simple root, 43.1 mm long as preserved, and broadly exposed roots consist of a transversely transversely subdivided into subequal anterior widened anterior root and a larger posterior and posterior moieties by slight lateral indenta­ root, triangular in cross section with the apex of tions. There is no indication of a diastema ante­ the cross section posterior, under the talonid. rior to P3, but the adjacent shattered, poorly Some of the weak, thin-walled bone of the alveo­ preserved bone is very likely displaced lingually. lar margins of all teeth has been lost in preser­ If this bone and the tooth were swung labially vation or preparation, making their size and into line with the Pi (or DP]) and P2 alveoli, a character somewhat conjectural. The alveoli of diastema would open anterior to P3. The alveoli My and Mi indicate transversely widened ante­ of P2 and P] (or DPi) are essentially similar to rior and posterior roots subequal in size but with their counterparts in USNM 244033, as far as the anterior root longer in M2 and the posterior preserved. Each lies anteriorly inclined along the longer in Mi. The alveoli of the roots of Mi are dorsolingual surface of the canine and separated comparatively small, shallow, and convergent ap­ from it only by a thin alveolar wall. The alveolus ically, indicating a small, possibly senescent Mi. of P2 is very incomplete, and inadequate to re­ The P4 alveolus is shallow, simple, and ovoid in flect subdivision of the root if such is the case. cross section; that of P3 somewhat deeper, more The alveolus of Pi (or DPi) indicates a much elongate anteroposteriorly, and with the sugges­ smaller, shorter root than in Pi (or DPi) of 28 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

lOCM 1 I

FIGURE 18.—Behemotops emlongi, referred anterior fragment of right mandibular ramus from lower part of Yaquina Formation of Oregon, USNM 186889, in lingual (A), labial (B), and dorsal (c) aspects. NUMBER 59 29

USNM 244033. Perhaps the most important fea­ persistent growth, maintained at least into old ture of USNM 186889 is the presence of alveoli adulthood. The tusk has a thin (perhaps 0.5 mm) of three incisors, strongly indicating that the tusk sheath of enamel extending entirely around its is a canine as in Paleoparadoxia (Figure 19). The circumference, from the worn tip proximally for alveolus of the first incisor is much fractured and some 95 mm. The exact proximal limit of the obscured by matrix; that of I2, although highly enamel crown is difficult to define because the incomplete, is 55 mm deep as preserved and deposition of enamel apparently terminated ir­ suggests a straight, robust tooth of rounded cross regularly in streaky continuations of longitudinal section. The alveolus of I3 is similar, but its ribs and wrinkles, which are apparent on the preserved part is only some 33.5 mm deep, in­ crown wherever wear facets or polishing have dicating a shorter tooth. not removed them. There is a large, subplanar Thus, Behemotops emlongi from the lower part wear facet truncating the crown obliquely. This of the Yaquina Formation of Oregon retained a facet would have been produced, not by occlusal complete, inferior, adult dentition of three inci­ wear, but by wear against a substrate, presumably sors, one canine, four premolars, and three mo­ in feeding. There is on the dorsal part of the lars. We infer that B. proteus from the lower part medial surface of the crown at its widest part of the Pysht Formation of Washington did so as what appears to be a facet of occlusal wear (pre­ well. Paleoparadoxia, a more derived desmosty­ sumably produced by shearing action with a su­ lian, had one less premolar. perior incisor), recognizable over a length of Dentition: As in Behemotops proteus, all pre­ some 47.5 mm and a maximum width of 7.5 mm. served teeth of both specimens of Behemotops There is no evidence of cementum on the tusk. emlongi show the enamel finely to coarsely cren­ Except for the root of P3 in USNM 186889, ulated, wrinkled, or pustulose, except where P1-M2 are represented only by alveoli in Behem­ smoothed by wear. otops emlongi. For information on the size of these The inferior incisors are known only from teeth, see the discussion of the dental formula their incomplete alveoli in USNM 244033 and and the measurements (pages 26, 27). 186889. These indicate three incisors, approxi­ A rather heavily worn M3 is the only tooth mately similar in size, straight and subcylindrical preserved in USNM 244033 (Figure 12F). It is in shape, perhaps 15-20 mm in their maximum essentially similar to the molars of the type spec­ diameters. imen of Behemotops proteus, especially Mi and M2, The inferior canine, actually preserved in but differs from the incompletely formed M3 of USNM 186889 and represented by its incom­ that specimen in having less stylodont principal plete alveolus in USNM 244033, is a greatly cusps. As is demonstrated by Mi ofthe holotype enlarged, strongly procumbent, laterally com­ of B. proteus, USNM 244035, thickened enamel pressed tusk. It is widest dorsally (anatomically, can be seen in M3 of B. emlongi, USNM 244033. posteriorly) and narrowest ventrally (anatomi­ Its cingulid is continuous labially, being especially cally, anteriorly). The ventral narrowing is ef­ strong adjacent to the hypoconid. Two small fected largely by development of single, comple­ conelets are present near the posterolingual base mentary, broad longitudinal channels on either of the hypoconid and the posterolabial base of flattened surface, deeper on the medial side. The the entoconid, respectively, but are heavily worn. tusk is essentially straight in dorsal or ventral Before wear, the heel was evidently a transverse aspect, but gently curved in an open S-shape in crest composed of several small crenulations. lateral or medial aspect, with the alveolar end The metaconid apparently is not twinned. downturned and the extruded end upturned Osteology: The remnant of the bony ramus of (Figures 16 and 18). The base ofthe tusk is open, USNM 186889 is osteologically useful primarily with a deep, conical pulp cavity, indicative of in revealing the procumbency of the canine tusk 30 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

FIGURE 19.—Paleoparadoxia tabatai, cast of neotype, USNM 26375, from Izumi locality (Shikama, 1966:12), middle Miocene of Japan, left mandibular ramus, in lingual (A), labial (B), and occlusal (c) aspects. NUMBER 59 31 and the incisors, and the broadly scoop-like sym­ tops proteus and B. emlongi can be directly com­ physeal region. Both of these characteristics are pared. However, these include the size and mor­ confirmed in USNM 244033, which preserves phology of M3, the postcanine dental formula, most of the bony ramus. The most conspicuous the form ofthe ascending ramus, and to a certain character ofthe ramus is the massively expanded extent the form of the symphyseal region of the anterior end (Figure 16), reflecting the relatively mandible. As discussed at length above, the den­ huge canines and battery of six incisors. The tal formulae ofthe two species are not demonstr­ under-surface of the symphyseal region is ably different except for the fusion of the roots broadly flattened, almost planar. If this surface of P3 and P4 exhibited by B. emlongi (and by more is oriented perpendicular to the sagittal plane, as advanced desmostylians). However, the peculiar it almost inevitably was in life, then the canine M3 morphology of B. proteus is attributable at was compressed almost exactly in the vertical least in large part to incomplete development; plane and the postsymphyseal body of the jaw the ascending rami are not significantly different; was canted strongly inward dorsally, as in Hex­ and the symphyseal region of B. proteus could aprotodon. There is a single, large, mental fora­ well have been broad and scoop-like as in B. men adjacent to the canine alveolus and lying emlongi. The length of the incomplete M3 crown ventral to the P2 alveolus (Figures I 6B, 17B). The in B. proteus is 31.7 mm, compared with 37.6 postsymphyseal ventral margin of the horizontal mm for the complete M3 crown in B. emlongi; the ramus is nearly straight, but the angular margin difference is even within reasonable limits of is missing, as is the thin posterior margin of the intraspecific variation. ascending ramus. If one assumes that little of this Only four differences, two of which are in­ latter margin is lost, the anterior and posterior ferred rather than clearly demonstrated by the margins are essentially parallel and inclined specimens at hand, suggest to us that specific slightly forward. The condyle is incomplete but distinction is warranted: greater adult mandible its position well above the plane of postcanine size, size and position of the canine tusk, and occlusion is clear. The mandibular foramen fusion of the roots of P3 and P4 in B. emlongi. In opens just above the level of the dorsal border each of these characters B. emlongi is more de­ of the bony horizontal ramus and approximately rived than B. proteus, but the matter is clouded midway between the anterior and posterior mar­ by the immature condition of the only known gins of the ascending ramus. The coronoid canal individual of B. proteus. (which descends from the rear of the tooth row We base our conclusions about projected man­ to open above the mandibular foramen), if pres­ dible size in part on growth of this bone in Recent ent, is tiny and obscured by poor preservation. Hippopotamidae, the closest living morphologi­ The coronoid process is broad, flattened, cal analogs of desmostylians. An immature Hex­ smoothly rounded in profile, and inclined some­ aprotodon (USNM 271019, Figure 20) with M2 what anteriorly. The bony ramus posterior to the partly erupted and DP4 heavily worn (therefore base ofthe canine tusk is relatively thin, and the dentally slightly younger than the type of Behem­ postcanine dentition of modest size, in contrast otops proteus) has a mandible 20 cm long, com­ to the massive, broad, scoop-like muzzle with pared to 27.5 cm in an adult (USNM 302054) large teeth indicated anterior to that point. with worn M3. An immature Hippopotamus (USNM 162976) with M. completely erupted Relationship between Behemotops proteus and DP4 heavily worn (comparable to the type of and B. emlongi B. proteus) has a mandible 46 cm long; the largest adult mandible we measured (USNM 123387) Unfortunately, there are few points of anat­ was 59 cm long. If a similar relationship between omy on which the present specimens of Behemo­ growth and tooth eruption existed in B. emlongi 32 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

0 5CM J I I L

B

FIGURE 20.—Hexaprotodon liberiensis, modern pygmy hippopotamus, USNM (Division of Mam­ mals) 271019 from the National Zoological Park, mandible of immature individual, in occlusal (A) and right lateral (B) aspects. NUMBER 59 33

(an animal intermediate in size between Hexa­ should be consulted for details to that date. We protodon and Hippopotamus), its juveniles should present only a synoptic coverage here. have reached 75%-80% ofthe adult mandibular The desmostylians were first made known by length by the stage of tooth eruption seen in the O.C. Marsh (1888) on the basis of some material type of B. proteus. However, the mandible of the of Desmostylus itself from marine Neogene de­ latter is only about 55% as long as that of the posits of Alameda County, California. Marsh re­ adult B. emlongi. ferred Desmostylus to the Sirenia. Flower and As noted in our description of Behemotops pro­ Lydekker (1891), on the basis of Marsh's work, teus, there does not appear to have been space in next placed Desmostylus in the Halicoridae (= the mandible below Pi (or DPi) for an enlarged Dugongidae) and from then until 1953 Desmos­ canine tusk like that of B. emlongi. Rather, any tylus was generally regarded, sometimes with a canine or incisor tusks that were present must query, as a sirenian. have been medial to Pi (or DPj). It seems highly The second major find of a desmostylian fossil unlikely that enough tusk growth or mandibular was reported from Japan by Yoshiwara and Iwa­ remodelling could have taken place in the time saki (1902), who described and figured the an­ remaining until eruption of M3 for B. proteus to terior part of a skull and both lower jaws of a take on the form of B. emlongi. specimen of Desmostylus. They believed their find Principally for the latter reason, we prefer to to be some sort of proboscidean, based in part regard the Washington and Oregon animals as on a letter from H.F. Osborn. Osborn had ex­ representing separate species of a single genus, amined photographs of the specimen and had pending additional knowledge of their anatomy read a brief description of the skull, sent to him and ontogeny. by Yoshiwara and Iwasaki. In their paper Yoshi­ As Hirota (1981), Reinhart (1982:554), and wara and Iwasaki made no mention of Marsh's Shikama (1966:131) suggested, and as Reinhart description of Desmostylus, so presumably they (1959:92) hinted but then denied, some or all were unaware of it. They believed their find to desmostylians probably were sexually dimorphic. represent a new genus, but did not supply a new Writing about the proboscidean allies of the des­ name. Although Osborn had "informed" them mostylians, Osborn (1936:183) claimed the same. that the skull belonged to a proboscidean, Yosh­ iwara and Iwasaki demonstrated that it was not In all the known Proboscidea there is a marked disparity between the male and female incisive tusks both in length like deinotheres or elephantids and therefore and in diameter. The adult female tusks never fully attain would have to represent a branch from the prim­ the length ofthe adult male tusks, but a still more striking itive proboscideans, near the origin of that order difference is their slenderness of proportion and diame­ from among the other ungulates. They also men­ ter. tioned some similarities to Sirenia. In the same Frick (1933:507, 574, 581, 632, 650) also year, however, both Osborn and J.C. Merriam recognized that the Japanese specimen was ref­ mentioned sexual dimorphism in gomphotheres erable to Marsh's Desmostylus (Osborn, 1902). and mammoths. Nevertheless, in spite of these Schlosser (1904) regarded the Japanese specimen examples in proboscideans, we believe that the as definitely sirenian. observed differences in morphology between Be­ hemotops proteus and B. emlongi are too great to Osborn (1905:109) placed Desmostylus in a be of sexual origin alone. monotypic family Desmostylidae and stated that it belonged in either the Sirenia or the Probos­ cidea. Merriam (1906; 1911:412) regarded Des­ History of Desmostylian Systematics mostylus as a sirenian, possibly requiring its own VanderHoof (1937:170-177) thoroughly re­ family, and reinforced the suggestion of relation­ viewed the desmostylian literature, and his work ship between sirenians and proboscideans. 34 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Abel (1914:213; 1919:830; 1920:445), in con­ VanderHoof (1937) countered Abel's interpre­ trast to his later work, concluded that Desmostylus tations in detail and supported inclusion of des­ was more closely related to the Proboscidea than mostylians in the Sirenia as the suborder Des­ to the Sirenia, although it is not clear why he did mostyliformes. Sickenberg (1938), however, ar­ so. Later Abel (1922:381; 1923) abandoned his gued strongly against a desmostylian-sirenian re­ view of proboscidean affinities of Desmostylus in lationship. Gregory (1951:428, 801-803) re­ favor of a bizarre notion that it belonged to the tained Desmostylus in the Sirenia but gingerly mammalian subclass Allotheria (= Multituber- suggested "remote derivation from such a prim­ culata). He persisted in this belief even after itive proboscidean as Moeritherium." examining a skull of Desmostylus at the NMNH According to Shikama (1966:151), an earlier (Abel, 1926); then, and later (in Weber, publication by H. Kishida (1924) assigned Des­ 1928:xiii, 44, 85), he placed the family Desmo­ mostylus to the Marsupialia. We have not seen stylidae in the Monotremata, suggesting a possi­ Kishida's work. ble relationship between them and multituber- Ijiri (1939) considered Desmostylus to be an culates. Still later Abel (1933:875) elevated them ungulate "in the broadest sense" but not a mon- in rank, naming an order Desmostyloidea within otreme, multituberculate, marsupial, or sirenian. the subclass Multituberculata. He evidently re­ Reinhart (1953) proposed the order Desmo­ garded them as multituberculates to the end of stylia, essentially an elevation of Osborn's Des­ his career (Abel, 1944). Although it has priority mostylidae and Hay's Desmostyliformes to still over Desmostylia Reinhart, 1953, Abel's ordinal higher taxonomic rank. Reinhart's (1959) revi­ name apparently has been overlooked com­ sion of the Desmostylia led him to believe that pletely during the more than 50 years since its the desmostylians, sirenians, and proboscideans creation. No useful purpose would be served by are closely related paenungulates, but that the resurrecting it. Adherence to priority in names desmostylian stem separated from the other two of suprafamilial taxa is not required, and in this orders in the Paleocene. On the basis of postcran­ case it would not be in the interest of stability. ial evidence, Reinhart noted that the desmosty­ Therefore, we strongly recommend retention of lians could not be descended from known siren­ the shorter, more euphonious, and entrenched ians because desmostylians were still capable of name Desmostylia. locomotion on land. Similarly, dental evidence Hay (1915), followed by Matsumoto (1918), led him to conclude that Moeritherium (Figure placed the Desmostylidae in the Sirenia, although 21) was already too advanced along the probos­ he emphasized that the Desmostylidae were very cidean path to have been a desmostylian ances­ different from other (true) sirenians. Later Hay tor. Thus, the Desmostylia were shifted from (1923:109) created a suborder Desmostyliformes their former status as a sirenian subdivision and to contain the Desmostylidae alone among sir­ were given taxonomic equality with both the enians, placing all other sirenians in a suborder Proboscidea and the Sirenia. McKenna (1975:42) Trichechiformes. In the same paper and in a later dubbed this unresolved trichotomous group succeeding one (Hay, 1924:7) he corrected in of paenungulate orders the mirorder Tethyth­ detail the misinterpretations of cranial sutures eria. The cladistic analysis given below (Figure on which Abel's assertion of multituberculate 22) resolves the trichotomy and indicates that affinities were largely based. Winge (1924:187, 188; 1942:214) regarded Desmostylus as "undoubtedly a lateral offshoot of FIGURE 21.—Stereophotographs in occlusal aspect of infe­ rior postcanine teeth of Moeritherium trigodon from Djebel- the oldest manatids." Winge also provided some el-Qatrani Formation of Egypt: A, YPM 18181, RP^-M^; B, pithy comments about Abel's theory of multitu­ YPM 18098, RM.-M^; c, AMNH 13437, RP., P3, Mj-Mj. berculate affinities of the desmostylians. Scale 1 cm. NUMBER 59 35 36 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY within Tethytheria the Desmostylia are more superorder Protungulata in the sense of Simpson closely related to the Proboscidea. (1945). Later (Thenius, 1969:584-589, 631) he Sera (1954) believed that Desmostylus, sireni­ did not employ the group Protungulata, but con­ ans, and proboscideans were derived from an- tinued to regard the Desmostylia as "Huftiere" omodontine therapsids. excluded from the superorder "Subungulata" Thenius (in Thenius and Hofer, 1960:189, (Sirenia, Proboscidea, Hyracoidea, and Embrith- 190, 196, 197), on the basis of meager similarities opoda). and geographic separation between probosci­ Ijiri and Kamei (1961:27) thought that des­ deans and sirenians on the one hand and des­ mostylians were especially close to perissodactyls mostylians on the other, placed the latter in the and artiodactyls, as did Shikama (1966:153).

TETHYTHERIA

SIRENIA PROBOSCIDEA DESMOSTYLIA

/ S"

./ / / / x

73-77

69-71

67,? 68

<1 = parallelism

100,13-15

FIGURE 22.—Cladogram of Tethytheria (Sirenia, Proboscidea and Desmostylia). An unnamed taxon within the Tethytheria comprising Proboscidea and Desmostylia is holophyletic on the basis of characters 13 and 15, plus characters 10 and 14, which are subject to convergence in Sirenia and, as interpreted herein, in other tokotheres. Character 35 is also evolved conver­ gently in Sirenia. See p. 37, 38 for explanation of characters 1-92. NUMBER 59 37

Minkoff (1976) suggested that desmostylians ward to arise labial to M2 (coronoid process not pre­ served in Dor el Talha barythere). should be placed with the Amblypoda rather 27. Ascending process of palatine bone disappears from than the Paenungulata. orbitotemporal fossa (bones fused and poorly pre­ served in Dor el Talha barythere). Characters Used in Phylogenetic Analysis 28. Condyloid foramen lost (not preserved in Dor el Talha barythere skull). The position of this character on the The following list of characters was used in the cladogram is uncertain, but close to the point indicated. construction of the cladogram depicted in Figure 29. Front end of skull much shortened, with orbits for­ 22. The numbers correspond to those character­ ward, anterior to P^. Infraorbital foramen under orbit. izing the various clades of the cladogram. 30. Lacrimal bone lost. 31. Postorbital process of frontal reduced. 1. Anterior border of orbit lies forward of M' 32. Cranium tubular. 2. Zygomatic process of squamosal bone expanded far 33. I3 lost. laterally. 34. Nasal opening retracted and proboscis developed 3. Pachyostosis and osteosclerosis. (interpretation). 4. Bilophodonty incipient. 35. Frontal/premaxillary contact occurs (nearly does so in 5. I' enlarged to become tusk. (See Savage, 1977, for Moeritherium; probably does so in Dor el Talha bary­ postulated parallelism in this character in Sirenia.) In there skull). trichechids and some dugongids the tusk is secondarily 36. I^ lost (not determinable in Fayum type oi Barytherium lost. grave, but so interpreted in Dor el Talha barythere 6. Rostrum deflected. skull; Savage, 1969:170). 7. Petrosal separate from skull anteriorly and posteriorly. 37. Size larger (and still larger). 8. Atlas vertebra modified (Savage, 1977:347). 38. C lost (not determinable in Fayum type of Barytherium 9. Mandibular dental capsule exposed posteroventrally. grave, but shown in Dor el Talha barythere skull; 10. DP5 (= Ml new) retained without replacement in Savage, 1969:170). adults. 39. Ii lost. 11. Hind limbs reduced. (Condition unknown in Prorasto­ 40. I2 more enlarged. mus, Sirenavus, and some other fossil taxa. Protosiren 41. Scapula with enlarged coracoid process. still apparently retained a large functional femur. How­ 42. Stance graviportal. ever, Eotheroides libycum had a reduced pelvis (An­ 43. Mandibular ramus with ventral protuberance anterior drews, 1906:119; Sickenberg, 1934:94).) to the level of P2 (protuberance faint in Dor el Talha 12. External auditory meatus wide. barythere). 13. M3 with hypoconulid shelf transversely broad, but the 44. P^ and P"* crowded. P'' transverse, lacking metacone hypoconulid still central. A small entoconid II (a new (not determined in Dor el Talha barythere). entoconid situated behind the true entoconid) can be 45. Scapular spine either loses acromion and metacromion present adjacent to it lingually (Minchenella) and tiny or they are short (Dor el Talha barythere). Supraspi­ cuspules can be present labially. nous fossa greatly reduced. 14. Former last molars lost. 46. Distal end of humerus expanded (not determinable in 15. External auditory meatus high, nearly enclosed ven­ Dor el Talha barythere). trally by mutual contact of squamosal post-tympanic 47. 1' lost. and postglenoid processes. 48. Ethmoid foramen shifted rearward beneath crista or- 16. Ms with two definite cuspids at rear: a labially displaced bitotemporalis (ethmoid not preserved in Dor el Talha hypoconulid and a large entoconid II. barythere skull). 17. Ms entoconid II somewhat posterolabially shifted. 49. Cranium pneumatized (possibly this character is mis­ 18. P4 with enlarged, high hypoconid and entoconid. placed; it may also characterize barythere-like probos­ 19. P, (or DPI) double-rooted. cideans). 20. I2 enlarged. 50. M' and Mi trilophodont. 21. P' (or DP') and P, (or DP,) lost. 51. P2 (but not DP2) lost. 22. C reduced. 52. p2 (but not DP^) lost. 23. C, lost. 53. Lower tusks and mandible sharply downturned. 24. I^ enlarged. 54. Skull with deep rostral trough. 25. Anterior end of jugal bone reduced (condition un­ 55. Scapular acromion and metacromion reduced. (See known in Dor el Talha barythere skull). character 45.) 26. Base of coronoid process of mandible shifted far for- 56. Paroccipital process enlarged. 38 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

57. Paroccipital process further enlarged. 86. Cusp height increases. 58. Occipital condyles highly placed. 87. Molar cingula lost. 59. Postmetaloph ornamentation of M" and M' reduced. 88. Supra.symphysial depression reduced. 60. Metapodials long. Femur shortened. Foot functionally 89. Rear molars very high-crowned, with enamel extend­ tetradactyl. ing below gum line and into alveolus. 61. Skull roof shorter and narrower. 90. Zygomatic process broadened. 62. Subsidiary median styles occur on P^ and P"" 91. Sagittal crest lost. 63. I^ more enlarged. 92. Extra cusps occur on molars. 64. External auditory meatus ventrally closed (see charac­ ter 15). 65. M", M2, and M3 trilophodont. 66. Horizontal component occurs in tooth replacement. Comparisons with Early Proboscidea 67. P2 roots fuse. and Minchenella 68. Passage (postzygomatic foramen of VanderHoof, 1937:178, figs. 9, 11) present through squamosal from From early in the history of their study, des­ external auditory meatus to roof of skull. mostylians were generally considered to be odd 69. P3 roots fuse. sirenians. Sirenians themselves were consid­ 70. P4 roots fuse. Ijiri and Kamei (1961) state that X-ray photographs of the Izumi specimen of Paleoparadoxia ered to be related to Proboscidea, a concept that tabatai show P, and P4 to be double-rooted. Although dates back at least to de Blainville (1816, 1836 that might be the case at depth in the mandible, the [1834]) and perhaps to Linnaeus (1758). Lin­ roots of each tooth are fused at the level of their naeus (1758:33, 34) had placed Elephas and Tri­ emergence from the alveolus at the base of each pre­ chechus next to one another within his order molar crown. X-ray photographs ofthe roots of P3 and P4 of a specimen of Paleoparadoxia from the lower Bruta, but of course he had also added sloths, Miocene of Point Arena, California (Phillips et al., anteaters, and pangolins to the mix. Unlike Cu­ 1976:152; Clark, in prep.) show a longitudinal differ­ vier and most authors ofthe succeeding century, entiation interpretable as a relic of fusion (A. Panofsky, Linnaeus did not ally the Sirenia with the Ceta­ pers. comm. to the authors). cea. Gervais (1855) pointedly abandoned de 71. Canines enlarged further to form procumbent tusks. Blainville's scheme of a sirenian-proboscidean C| enormous. special relationship, emphasizing aquatic habitus 72. CJ somewhat angular in cross section rather than suboval. and neglecting the shared features claimed by de 73. Pi lost. (This and the following two characters apply Blainville. Although some workers in the late to all previously recognized desmostylians.) nineteenth century championed sirenian-probos­ 74. Pj reduced. cidean affinity (e.g., Kneeland, 1850; Kaup, 75. All cusps on posterior cheek-teeth become desmosty- 1855), most taxonomic treatises made no special lodont. attempt to relate the two or in fact argued against 76. P;i paraconid lost. 77. P^-Mn hypoconulid and entoconid II enlarged, espe­ the relationship and maintained the cetacean cially entoconid 11. connection (e.g., Owen, 1859; Gray, 1866; Gill, 78. Lower incisors become rectangular and flat. 1871, 1872, 1873; Trouessart, 1879, 1898; 79. M-.. with extra cuspid between and labial to protoconid Flower and Lydekker, 1891; von Zittel, 1893). and hypoconid. Moeritherium.—A new claim of sirenian rela­ 80. Molar cingula reduced. 81. Mandibular symphysis becomes elongate. tionship to proboscideans was provided when the 82. Sagittal crest reduced. fossil genus Moeritherium was made known from 83. At most, only canine tusks and one pair of lower the Fayum deposits of Egypt by Andrews (1901b, incisors remain. 1901c, 1902, 1904a, 1906). Andrews correctly Remaining premolars lost in adults (but in young in­ placed Moeritherium in the Proboscidea but sug­ 84. dividuals 3 upper loci and at least 1 lower locus are occupied by deciduous or permanent premolars). gested in his discussion ofthe pelvis that Moerith­ Medially positioned bony swelling occurs al the rear of erium and a primitive sirenian (his ''Eotherium'") 85. the dentition. with unreduced pelvis and femur might have had NUMBER 59 39 a common ancestor at some earlier time in the tusks, was conclusively in favor of genealogical Tertiary (Andrews, 1906:119). Sickenberg relationship of Moeritherium to Palaeomastodon (1934:94) later referred the sirenian pelvis in and more advanced elephant-like genera, not to question to a different sirenian genus, Protosiren. Sirenia. This is perfectly true, but Osborn Winge (1906, 1924, 1942:26) arranged the un­ (1909:139) was aware of it and had not claimed gulates in a genealogical scheme in which arsi- otherwise. Possibly influenced by Winge (1906), noitheres first differentiated from a menisco- Osborn (1910:200, 203, 558) retained the Moer- theriid source. The arsinoitheres were then sup­ itheriidae in the Proboscidea, but (Osborn, posed to have divided into hyracids and elephan­ 1910:204) took a new tack by claiming that "we tids, the latter giving rise to the Sirenia. Accord­ may look for other radiations of the probosci­ ing to Winge (1942:148, 149, 211, 212), the dean stock in Africa; possibly the river-living Sirenia are merely aquatic proboscideans. Winge sirenians may prove to be one of these radia­ considered Moeritherium to be the most primitive tions." then-known proboscidean, with Barytherium in­ Gregory (1910:368) stated: terposed between it and the deinotheres. Osborn (1907:15) included Moeritherium in the Probos­ The genus [Moeritherium] represents a very primitive offshoot from the Proboscideo-Sirenian stock. Its denti­ cidea and placed the orders Proboscidea and tion and certain other characters indicate a nearer alli­ Sirenia next to each other among the nine un­ ance with the Proboscidea than with the Sirenia, but it is gulate orders that he recognized in his classifi­ far more primitive than any other known representative cation. In a paper on the feeding habits of Moer­ of either order. itherium and Palaeomastodon, Osborn (1909:140) Ten years later Gregory (1920:180, and erra­ suggested that "Moeritherium [sic] is an offshoot tum and addendum) continued to regard Moer­ of the Proboscideo-Sirenian stock, with slightly itherium as a proboscidean, but pointed out that nearer kinship to the elephants than to the Sir­ its lacrimal bone, if present, was not like that of enians [sic]." Osborn took both real and supposed later proboscideans and that in this respect Moer­ aquatic modifications of Moeritherium to imply itherium resembled sirenians. With regard to the that it was somewhat more sirenian-Iike than origin of the latter, he stated that "the Sirenia, other Proboscidea, but he did not conclude that although highly specialized for aquatic life, show Moeritherium was a sirenian, although he did state special resemblances with Moeritherium in the that "Moeritherium [sic] is closer to the Sirenians skull (including the orbital region) and dentition, [sic] and less close to the Proboscidea than has and are generally regarded as a derivative of the hitherto been supposed." Osborn's paper was proboscidean stem" (Gregory, 1920:245). objected to by Andrews (1909), who was, of Abel (1914:191-213; 1919:826; 1920; 1933: course, thoroughly familiar with both Moerither­ 899) not only placed Moeritherium in the Probos­ ium and the Egyptian Tertiary sirenians. How­ cidea, but also placed Barytherium there as well, ever, based upon what Osborn actually wrote, stating that Barytherium might be some sort of Andrews (and also, apparently, Gregory, side branch from the early deinotheres. Abel 1920:245) misinterpreted Osborn's point. An­ thought that the Proboscidea and Sirenia are drews noted that sirenians with tusks (i.e., many closely related, at times (in Weber, 1928:425; dugongids) have evidently obtained them by en­ 1933:899, 904) following von Zittel (1925:246) larging V, and he also pointed to derived features in arranging the two groups as suborders, along of the ear region that Moeritherium shared with with hyracoids and embrithopods, within an or­ other proboscideans. He argued that the aquatic der Subungulata. habitat of Moeritherium had produced some sim­ Osborn (192la:2) strongly supported probos­ ilar adaptations, but that the dentition, particu­ cidean affinities for Moeritherium and divided the larly the enlargement of I^ and I2 to become Proboscidea into Moeritherioidea, Dinotherioi- 40 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY dea (sic), Mastodontoidea, and Elephantoidea. less close to the Proboscidea than has hitherto These coordinate categories were believed to be been supposed." of either subordinal or superfamilial rank. Os­ Osborn (1923:1) once more included the born no longer seems to have regarded the sir­ moeritheres in the Proboscidea, referring to enians as derivatives of proboscidean stock, al­ them (1925:20, 21) as "small, amphibious pro- though he stated (Osborn, 1921a) that the facial mastodonts." and cranial proportions of Moeritherium are anal­ Simpson (1931:264) included the moeritheres ogous to those of the Sirenia. Osborn omitted as one of four superfamilies of the order Probos­ consideration of the barytheres in his two 1921 cidea: Moeritherioidea, Dinotherioidea (sic), papers and two years later specifically excluded Mastodontoidea, and Elephantoidea. The bary­ them from the Moeritheriidae (as Moeritheri- theres were maintained at ordinal rank and the inae; see Osborn, 1923:1). order Sirenia was classified in accordance with That Moeritherium might not have been so Hay's (1923) division of the sirenians into two aquatic an animal as had been supposed by var­ suborders: Trichechiformes and Desmostyli­ ious previous authors was suggested by Matsu­ formes. moto (1923:105). Moreover, an extensive analy­ Although he did not believe Moeritherium to sis of the similarities between Moeritherium and be ancestral to other proboscideans, Osborn hyracoids, sirenians, and proboscideans brought (1936:22) maintained the genus in a monotypic Matsumoto down heavily in favor of probosci­ suborder of the Proboscidea, Moeritherioidea, dean relationships for Moeritherium. coordinate with Deinotherioidea, Mastodon­ Petronievics (1923) not only thought Moeri­ toidea, and Elephantoidea. As he had held fifteen therium was a member of the Proboscidea, but years earlier, Osborn (1936:24, 48) believed the also believed it to be ancestral to Palaeomastodon. facial and cranial proportions of Moeritherium to He misinterpreted Osborn (1909) by claiming be analogous, not homologous, to those of the that Osborn had held that Moeritherium was "an Sirenia. At one point in his monograph Osborn offshoot of sirenian stock allied to Proboscidea" (1936:39) did indeed list the sirenians, moerith­ (Petronievics, 1923:58). Petronievics (ibid.) also eres, and proboscideans as coordinate groups, accused Osborn of holding the same views in but elsewhere on the same page he placed Moer­ Osborn's (1919) paper on Palaeomastodon. Curi­ itherium within the Proboscidea. ously, Osborn's (1919:266) paper did indeed In his famous two-volume compendium on the contain the following statement: "In 1909^ Os­ Proboscidea it is noteworthy that Osborn (1936, born pointed out that Moeritherium is to be re­ 1942) omitted the barytheres from considera­ garded as a terrestrial form of the Sirenians [sic] tion. Although mentioned several times (volume (manatees and dugongs) in no way directly re­ 1:51, 53; volume 2:1424) in passing in this enor­ lated to the Proboscideans." (The footnote to mous and otherwise comprehensive monograph, this passage reads "2. Osborn, 1909, 332." Here the barytheres, then known from a single Eocene "332" refers to paper 332 in Osborn's life-long Fayum species, were evidently considered to be­ sequence of publications, i.e., Osborn (1909) of long to a monotypic order, Barytheria Andrews, the present paper's bibliography, not to some 1904b, and were therefore excluded from Os­ unknown paper of 1909 containing a page 332.) born's work. Osborn had evidently reached this Here, however, Osborn (1919) clearly misquoted conclusion as early as 1905 (Osborn, 1905:112) himself We have not been able to find such a and apparently he continued to hold the same statement in Osborn's (1909) publication. The view in 1921 (Osborn, 1921a, 1921b, 1921c). nearest that Osborn had come to such a stance In his earliest discussion of Barytherium (as in 1909 was his statement, already quoted, that "Bradytherium" grave; see Andrews, 190Id, er­ Moeritherium was "closer to the Sirenians [sic] and ratum) Andrews (1901a) had regarded the an- NUMBER 59 41 imal as a Deinotherium-Vike proboscidean, as did tions of more advanced Proboscidea. Abel (1914). But in 1904 Andrews viewed the A nearly complete skeleton of Moeritherium Barytheria as an amblypod subgroup coordinate from the Egyptian Fayum was placed on display with the Dinocerata. By 1906, however, Andrews at the Yale Peabody Museum in December, once more considered the barytheres to belong 1963, and was figured and discussed briefly by to the Proboscidea as family Barytheriidae, in­ E.L. Simons (1964:14). No detailed description certae sedis (Andrews, 1906:172). Four years of the specimen has yet appeared, but the skele­ later, Osborn (1910:200, 559; and also 1921a, ton was refigured by Tobien (1976, fig. 8). Si­ 1921b, 1921c) continued to regard the bary­ mons noted that the elongate skeleton and var­ theres as deserving ordinal rank. It is curious ious features of the skull were suited to aquatic that Osborn, even if in agreement with Andrews life and he stated that "the creature can hardly (1904b) in that he regarded the barytheres as have been close to the line of elephant ancestry." non-proboscidean, made no attempt to compare However, no attempt was made by Simons to ally them with moeritheres or other proboscideans Moeritherium with any other mammalian group. in his 1936-1942 monograph. Thus, Simons' comments about the distinctive­ In his masterly classification of mammals, ness of Moeritherium, like Deraniyagala's re­ Simpson (1945:132-134) maintained the Moer­ marks, are based upon symplesiomorphies and itherioidea as a monotypic suborder within the possible autapomorphy rather than on synapo- Proboscidea, coordinate with Deinotherioidea, morphous features. Later, however, Simons Elephantoidea (with which Simpson combined (1968:3) alluded to "certain postcranial resem­ Mastodontoidea), and Barytherioidea. The name blances" between Moeritherium and the Desmos­ Barytherioidea was coined by Simpson partly tylia. because of a change in rank of Andrews' Bary­ Similarly, Tobien (1971, 1976) kept Moerith­ theria to that of a suborder, partly in order to erium well away from other proboscidean ances­ agree with the suffixes of other proboscidean try, but noted only autapomorphous and symple- suborders, and partly because Andrews' name siomorphous characters. He implied, but did not "Barytheria" was preoccupied by Barytheria document, sirenian relationships for Moerither­ Cope, 1898 (p. 123), a term that Cope had used ium. By 1978, however, Tobien no longer sug­ for toxodont notoungulates. gested sirenian affinities, simply referring to Until 1955, Moeritherium continued to be re­ Moeritherium as a paenungulate remotely related garded as an unquestioned primitive probosci­ to Palaeomastodon and Phiomia (Tobien, dean, but in that year Deraniyagala (1955:15, 1978:199,200). 16) separated Moeritherium from the Proboscidea Savage (1971:220) believed that "Moeritherium and placed the various species of moeritheres in and Barytherium are, if not true proboscideans, a new order, Moeritheria. This was done because basically close to the Proboscidea; nothing allied of the presumed lack, in moeritheres, of the to this order occurs outside of Africa until the trunk, believed by Deraniyagala to characterize Miocene." Savage believed the ancestry of the Proboscidea alone. Thus, Moeritherium was ostra­ Proboscidea to have been probably from un­ cized on the basis of retention of a primitive known early Paleocene African condylarths. character, not on the basis of derived features Maglio (1973, 1978), Coppens et al. (1978), possessed uniquely or shared with any other and (for all intents and purposes) Harris (1978) group of mammals, e.g., Sirenia. Deraniyagala restricted the Proboscidea to Elephantoidea by failed to show that Moeritherium is anything other excluding not only the moeritheres, but also the than a primitive offshoot from proboscidean barytheres and deinotheres. These semantic joint ancestors that had not yet become equipped with maneuvers served only to obfuscate matters by a trunk and the associated anatomical specializa­ sweeping the problem of interrelationships of 42 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY these animals under the rug, as Deraniyagala had from Dor el Talha, Libya, and again attempted done with Moeritherium in 1955. a cladistic analysis of the systematic position of Coppens and Beden (1978:333) stated that Moeritherium. Once more he concluded that the "other authors, such as Tobien (1971), also ex­ genus is more closely related genealogically to clude [moeritheres] from the Proboscidea and the Proboscidea than to Sirenia or Desmostylia. classify them among the sirenians." Indeed, To­ However, he linked Sirenia with Proboscidea on bien (1976:157) mentioned "many authors" the basis of reduction of the mastoid apophysis holding such a view. But only Tobien seems to and the presence of lophodonty, both weak char­ have written anything at all about Moeritherium acters. actually being a sirenian; those authors who In our own analysis, Moeritherium is regarded thought Moeritherium to be related to the Sirenia as a primitive proboscidean because it shares ten did so because of their belief that the Sirenia derived features with barytheres, deinotheres, originated from early proboscidean stock of and elephantoids (characters 4, 20-28 of clado­ some sort. Other authors have remarked only gram: Figure 22). about the difficulty of retaining such an aquati- That Moeritherium is not directly ancestral to cally adapted animal within the Proboscidea. other proboscideans is suggested by five unique Stating that their comments were based on a characters that would have to have been reversed personal communication from one of us (Dom­ if Moeritherium were to have given rise directly ning), Coppens and Beden (1978:333) noted that to proboscideans such as Palaeomastodon and Moeritherium does bear some striking similarities Phiomia (Characters 29-33 of cladogram: Figure to the order Desmostylia. However, these simi­ 22). After the departure ofthe phylogenetic line larities were not discussed by them. Domning leading to Moeritherium, Barytherium, deinoth­ himself (1978a:573, 574), in an article published eres, and elephantoids developed nine shared- in the same book as that of Coppens and Beden, derived features (characters 34-42 of cladogram: stated simply that "the desmostylians were for­ Figure 22). However, the loss of I', I^, C', and merly included within the Sirenia but are now Ii had not yet occurred in the earliest known accorded ordinal status in close alliance with proboscidean, an unnamed taxon (the Brezina moeritheres, proboscideans, and sirenians." Cop­ animal) from the early Eocene of Algeria (Mah- pens and Beden briefly considered the question boubi et al., 1984). The Brezina animal is there­ of whether Moeritherium should be regarded as a fore not a barythere; rather, it is the plesiomor- branch from early Sirenia, but they ultimately phous sister-group of all "higher" proboscideans concluded merely that moeritheres were de­ and could be ancestral to them, in keeping with scended from basic "subungulate" stock. They its age and geographic position. left the ordinal affinities of moeritheres as an Moeritherium has been reported from five (pos­ open question. sibly six) sites in the northern half of Africa and Van Valen (1978, fig. 3) depicted both Des­ one in southern Asia (Tobien, 1971; Coppens mostylia and Sirenia as descendants of Probosci­ and Beden, 1978; Tassy, 1981; Coiffait et al., dea, but he provided neither evidence nor dis­ 1984): cussion of his conclusions. Tassy (1979) reviewed the relationships of 1. Fayum Basin, Egypt. Eocene and Oligocene. See An­ Moeritherium on the basis of both new and old drews (1906). material. On the basis of cladistic analysis he 2. Dor el Talha, Libya. Eocene. See Arambourg and Mag- concluded that Moeritherium is more closely re­ nier (1961); Savage (1969, 1971). 3. 60 km NE of Gao, Mali. Eocene. See Arambourg, Ki- lated to the Proboscidea than to the Sirenia. koine, and Lavocat (1951); Tobien (1978:194, 195). Two years later Tassy (1981) published a de­ 4. M'Bodione Dadere, Senegal. Lutetian. See Gorodiski tailed description of an Eocene Moeritherium skull and Lavocat (1953). NUMBER 59 43

5. Khenchela, northeastern Algeria. PEocene. See Gaudry the holotype of Behemotops proteus in 1976 one (1891); Schlesinger(1912). of us (Ray), convinced that it was desmostylian 6. Nementcha Mountains, eastern Algeria. Late Eocene. but equally convinced that comparison with An­ See Coiffait et al. (1984). 7. Harudi, Kutch, India. Lutetian. See Sahni and Mishra thracobune (and "Pilgrimella") was warranted, (1975). spent much fruitless and frustrating time wan­ dering among the artiodactyls and perissodactyls At several of the African sites as well as in Asia until a chance conversation with Earl Manning the identification of specimens as Moeritherium dehorned the dilemma. Manning's identification may be misleading and might better be stated as of Anthracobune as a Moeritherium-Vike animal was "moeritheriid" or "primitive proboscidean-like also generously made known to R.M. West, who animal" rather than given as an identification to was the first to publish on the matter (West, the generic level. For instance, according to Go­ 1980:518; 1983). West placed Anthracobune in rodiski and Lavocat (1953:316), the Eocene ani­ the Moeritheriidae. During the long interval mal from Senegal is only about half the size of from 1940 to 1980, Anthracobune (with "Pilgri­ Moeritherium gracile. mella") had masqueraded as an artiodactyl (Pil­ Inasmuch as the Harudi specimen on which grim, 1940; Gingerich, 1977; Coombs and the Indian identification is based is a small moer- Coombs, 1977; and most other authors), a per- ithere-like sacrum, it is likely, as implied by West issodactyl (Coombs and Coombs, 1977:303; (1980:520), that it represents Anthracobune or a 1979), and a phenacodontid condylarth (Van very closely related genus (see below) rather than Valen, 1978, fig. 3). Wells and Gingerich (1983) Moeritherium. assigned it to a new family Anthracobunidae A sometimes quoted early reference to an­ within the Proboscidea, and (based on an exam­ other Asian occurrence of Moeritherium is in er­ ination of the specimens of Behemotops reported ror. Pilgrim (1912:15) referred what he thought herein) suggested that the Desmostylia, as well as was a fragmentary upper molar to MoeritheriumQ) the Moeritheriidae and the Sirenia, may be de­ sp. The specimen was found near Khajuri, Bugti rived from anthracobunids. The fact that Anthra­ Hills, Baluchistan, Pakistan, in what is presently cobune occurs in southern Asia, rather than in thought by R.L. Bernor (pers. comm.) to be the Africa, was doubtless a major cause of its long continental equivalent of the lower part of the neglect in discussions of the phylogenetic origin Gaj Formation (early Miocene). Later, Osborn of proboscideans and their possible affinities to (1936:79) identified the specimen as a P^ prob­ the desmostylians. ably referable to Trilophodon pandionis (?= Gom­ Anthracobune (sensu lato) shares a derived fea­ photherium angustidens of present terminology), ture (character 16 ofthe cladogram: Figure 22) a common Siwalik species. Possibly the animal with Moeritherioidea, Barytherioidea, Deino­ might represent the poorly known genus Hemi- therioidea, and Elephantoidea and is therefore mastodon Pilgrim, 1912, recently discussed by the sister-group of Proboscidea as classified by Tassy (1982:239, 240). Simpson (1945). M3 has two definite cusps at Anthracobune (including Pilgrimella and Joza- rear: a labially displaced hypoconulid and a large ria).—The closest relative of Moeritherium and entoconid II. other proboscideans, Anthracobune pinfoldi Pil­ That Anthracobune itself was not the direct grim, 1940 (p. 129) (congeneric with Pilgrimella ancestor of the African moeritheres and other pilgrimi Dehm and zu Oettingen-Spielberg, 1958 proboscideans is attested by three autapomor­ (p. 33), ^nd Jozaria palustris Wells and Gingerich, phous features (characters 17-19 of the clado­ 1983 (p. 125)), was known for nearly forty years gram: Figure 22). before Earl Manning recognized its true affinities Neither could the known anthracobunids have in the late 1970s. In the first stages of study of been ancestral to the Sirenia {pace Wells and 44 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Gingerich, 1983), whose earliest members retain other, similar mammals. However, phenacolo­ five premolars as a primitive condition. This phids are essentially bilophodont mammals in condition has nothing to do with the fact that which the M3 hypoconulid is small and merged one lineage of sirenians (manatees) much later with a lingually steeply ascending, wide, posterior evolved supernumerary molars (Domning, cingulum (McKenna and Manning, 1977, fig. 1, 1982). legend, character 5). Minchenella (as a supposed Ishatherium subathuensis Sahni and Kumar, phenacolophid) was recognized by Gingerich and 1980, described as a sirenian, may also be a Russell (1981:237) to be closely related to Moer­ specimen of Anthracobune. itherium and Anthracobune, and was regarded as Gingerich and Russell (1981) maintained both "the most plausible ancestor of Anthracobuni­ Pilgrimella and Lammidhania as separable from dae" by Wells and Gingerich (1983). Anthracobune and stated that the type specimen Comparison of Minchenella grandis with An­ of Ishatherium subathuensis is similar to an upper thracobune pilgrimi (identified by West, 1980, as molar of Pilgrimella. Gingerich and Russell A. pinfoldi) has been made possible by means of (1981) included Anthracobune, Pilgrimella, and casts prepared by P.D. Gingerich and R.M. West, Lammidhania in the Moeritheriidae as probosci­ respectively. The two species prove to be re­ deans. West (1983) likewise assigned Anthraco­ markably close in morphology, the former almost bune (= Pilgrimella) and Lammidhania to the certainly lying close to if not actually within the Moeritheriidae. Wells and Gingerich (1983), ancestry of the latter. Minchenella grandis lacks however, placed these three plus Ishatherium and the very autapomorphies of Anthracobune (sensu their new genus, Jozaria, in the Anthracobuni­ lato) that set the latter apart from Moeritherium dae. West (1980) had previously regarded Lam­ and other, more elephant-like, proboscideans midhania as an artiodactyl. (characters 17-19 ofthe cladogram: Figure 22; Sahni and Mishra's (1975) reference of a fossil see above). In fact, Minchenella grandis also lacks proboscidean-like sacrum from the middle character 16 ofthe cladogram (Figure 22), which Eocene of Kutch, India, to the Moeritheriidae umiQS Anthracobune pinfoldi with its moeritherian may thus be basically close to the mark, but the and more advanced proboscidean allies: M^ with genus involved may well be Anthracobune or a both a labially displaced hypoconulid and a large closely related genus rather than Moeritherium entoconid II. itself (West, 1980:521). However, Minchenella grandis shares the fol­ Minchenella.—Described as a member of the lowing character not only with Anthracobune and family Phenacolophidae, Minchenella Zhang, the Proboscidea, but also with Behemotops and 1980 (p. 257) was originally given the name more advanced Desmostylia (character 13 of the Conolophus Zhang, 1978 (p. 268), a junior hom­ cladogram: Figure 22): M3 with hypoconulid onym of a living iguanid lizard of the Galapagos shelf transversely broadened to form a cuspidate Islands. The type-species, Minchenella grandis crest in which the hypoconulid is still central. A (Zhang, 1978:268) occurs in the Upper Paleo­ small entoconid II (a new entoconid situated cene Datang Member, Nonshan Formation, Lo- behind the true entoconid) lies lingually adjacent fochai Group, Datang Commune, Nanxiong to the hypoconulid. County, Guangdong, People's Republic of For these reasons Minchenella appears to us to China. It is known from lower jaws only, so that be a late Paleocene Asiatic possible ancestor of skull characters and postcranial features do not both the Proboscidea (including Anthracobune) yet enter into deliberations about its affinities. and the Desmostylia (including Behemotops). We Zhang (1978) assumed the genus to be a member see no reason why Minchenella should not be ofthe family Phenacolophidae (considered to be regarded as ancestral to both of these mamma­ condylarths); she made no comparisons with lian orders (Figure 23). At present, therefore, we NUMBER 59 45

PLEISTOCENE-RECENT -2 PLIOCENE -5

other Proboscidea | Paleoparadoxia Deinot he res Desmostylus

MIOCENE —25 B. emlongi | Cornwallius Behemotops proteus

OLIGOCENE

Minchenella PALEOCENE

FIGURE 23.—Simplified phylogram of Proboscidea and Desmostylia. believe the Desmostylia to be more closely re­ edentates, macroscelideans, lagomorphs, pholi- lated to the Proboscidea than to the Sirenia. The dotans, possibly rodents, and a few extinct taxa Sirenia appear to us to have branched away from from the Cretaceous and early Cenozoic. He ancestors more primitive than Minchenella. The hypothesized that "all tokotheres share or further Sirenia, like many other groups of mammals as modify a postcanine dental formula consisting of divergent as marsupial diprotodonts and placen­ dP} P| Pi Pl dPt Ml Mi. DPi is the tooth usu­ tal pyrotheres, also became bilophodont early in ally called Ml." Domning, Morgan, and Ray their history, but early sirenian genera retained (1982) pointed out that primitive sirenians (be­ five premolars and lacked the peculiar elevated lieved to be tokotheres in McKenna's classifica­ external auditory meatus that characterizes des­ tion) retained five premolars (contrary to Fox, mostylians and proboscideans in which the skull 1983:21, and others) in addition to three molars. is known. For a contrary opinion, see Wells and This condition was shown especially clearly in a Gingerich (1983). mandible of Protosiren sp. from the middle Eocene of North Carolina (USNM 214596). However, comparison of this specimen and other Implications for Eutherian Dental Eocene sirenians with the holotype of Behemotops Homologies proteus (USNM 244035) suggests that adult re­ McKenna (1975:37) proposed a new superor­ tention of DP"" and DP5 together with loss of the der Tokotheria to include all eutherians except last molar, while not seen in the Sirenia, may yet 46 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY be characteristic of the Desmostylia and possibly the then unnamed Xenungulata), Proboscidea, other groups of tokotheres. Domning, Morgan, Embrithopoda, Hyracoidea, and Sirenia (includ­ and Ray (1982) concluded that McKenna's ing the desmostylians). Separate superorders co­ (1975) hypothesis was false if the Tokotheria ordinate with the Paenungulata were maintained were taken to include the Sirenia, but, alterna­ for perissodactyls, artiodactyls, carnivores, and tively, this could also be interpreted to mean that for a scrap-basket group combining condylarths, sirenians are not tokotheres. notoungulates, litopterns, astrapotheres, and In USNM 214596, a molariform DPr, is pre­ aardvarks. That Simpson's (1945) arrangement ceded by a F^ and P4 that closely resemble, re­ is unnatural has long been recognized, but pro­ spectively, the P3 and P4 described above for gress in understanding the detailed interrelation­ Behemotops proteus (USNM 244035). Each P3 has ships of all these mammals has been very slow. a single, high, conical cusp flanked by much The taxon Tethytheria was created by Mc­ smaller cusps and pre- and postcingulids. Each Kenna (1975:42) for a restricted group drawn P4, in contrast, has a transversely oriented ante­ from the ranks of Simpson's paenungulates. rior pair of large cusps and a much smaller pos­ Tethytheres were defined as comprising the co­ terior pair, together with variously arranged ordinate orders Proboscidea, Sirenia, and Des­ small cuspules and crests. mostylia. It will be remembered that the desmos­ The DP4 of Protosiren is unknown. However, tylians had been removed from the Sirenia and in Prototherium veronense de Zigno, 1875, a sir­ given ordinal rank by both Abel (1933:875) and enian from the late Eocene of Italy, DP4 is trilo­ Reinhart (1953:187). No attempt was made by bate in outline (Sickenberg, 1934, fig. 28b) and McKenna to resolve the trichotomous genealogy is accompanied by a molariform DP5 and a dou­ implied by the use of three coordinate ordinal ble-rooted DP3 similar to the P4 described above. taxa, nor was such an attempt made by McKenna The trilobate DP4 closely resembles that of Be­ and Manning (1977, fig. 1). However, in the hemotops proteus (USNM 244035). Domning present paper (Figure 22) we support the view (1982) has pointed out that the DP3_5 of Prototh­ that Proboscidea and Desmostylia share a more erium also resemble the three anteriormost teeth recent mutual ancestor than either of them does of living Trichechus, which he considers to be with the Sirenia. If our current view is correct, their homologues. Trilobate DP4S are also found Desmostylia have therefore been shifted from a in primitive proboscideans, phenacodonts, and former special sirenian alliance (e.g., Simpson, artiodactyls as noted in the description of B. 1945) through a neutral position (trichotomy) to proteus on page 000. We suggest that these tri­ a synapomorphous liaison with the Proboscidea lobate deciduous teeth in the fourth postcanine and Minchenella based upon characters 10, 13, position may be homologous in all these taxa. 14, and 15 of the cladogram depicted in Figure This would imply that the molariform teeth in 22. Early Sirenia did not share these features and the fifth postcanine position in most of these had not yet remodelled their dental formula groups, traditionally termed Ml, are actually (characters 10 and 14 ofthe cladogram). P5 and homologues of the sirenian DP5, and that the the former last molar may have been lost inde­ sirenian Ms has no homologue in the other pendently in several groups of mammals. groups. The polarity and distribution of these Novacek (1982) united Proboscidea and Sir­ character states require further study. enia as a monophyletic group whose sister group is the Hyracoidea. However, the characters used are not convincing. For instance, the phenaco­ Status of the Tethytheria lophids are not excluded by his character 71 and Simpson (1945) coined the term Paenungulata a squamosal contribution to the glenoid region at superordinal rank for the following orders: of the skull (part of composite character 73) is Pantodonta, Dinocerata, Pyrotheria (including surely plesiomorphous. His incomplete dado- NUMBER 59 47 gram did not include the Desmostylia and did land and venture into the cold water (and possi­ not deal with the presence of five premolars in bly the surf) only for some energetic payoff. early sirenians. Further, it seems likely that they would have been tied strongly to the shore for reproduction, Desmostylian Lifestyle again as with pinnipeds, even if they were capable of aquatic copulation as are pinnipeds and hip­ Desmostylians were difficult to visualize as liv­ pos. ing animals as long as they were regarded as The desmostylian diet is still controversial. sirenians and until sufficient skeletal material VanderHoof (1937:194) and, more recently, became known. It was long supposed that they McLeod and Barnes (1984) have supposed that lacked functional hind limbs. Even now that ex­ desmostylians fed on mollusks or other benthic tensive skeletal material is available, some ques­ invertebrates. This supposition may in part be tion remains regarding the manner in which the the basis for the otherwise inexplicable compar­ limbs could have supported the body (see Inu­ ison of desmostylians to walruses (Thenius, zuka, 1984, for a critical review and a radical 1969:586; 1980:49). Walruses do not crush or new interpretation). In our opinion, desmostyli­ masticate shells, but feed by suction (Fay, ans were undoubtedly amphibious, but more 1982:167-172). Further, there seems no basis in suited to terrestrial locomotion than pinnipeds. osteology, dentition, or locomotion (cf. Gordon, Analogies have often been drawn between des­ 1981) of walruses to warrant special comparison mostylians and hippopotami, especially after to desmostylians. complete skeletons of the former were discov­ Most writers, including Reinhart (1959:103) ered (e.g., Abel, 1914:212; Matsumoto, 1918; and Domning (1978b:l 13), have considered des­ Reinhart, 1953, in diagnosing the order; Then­ mostylians to be herbivores. Certainly, we see ius, 1960:196; Romer, 1966:254, 1968:201). nothing in the dentitions of at least the more The remarkable resemblances in size, build, and primitive, brachydont genera that would suggest particularly jaw and dental structure (cf. Figure a diet radically different from that of probosci­ 20) between desmostylians and hippopotami sug­ deans, hippos, pigs, or other plant-eaters. Shi­ gest analogous lifestyles as well. kama (1966:188), while not excluding inverte­ We believe, however, that to suppose that des­ brates from the diet, suggested that Paleopara­ mostylians (like hippos) fed mainly on land and doxia and Desmostylus were adapted for "brows­ resorted to the water chiefly for rest or other ing" and "grazing," respectively. Domning activities, would carry the analogy too far. Des­ (1978b: 114) extended this idea by speculating mostylians have been found only in marine and that Desmostylus specialized on seagrasses such as never in freshwater or terrestrial deposits; hence Zostera and especially their rhizome systems, they probably never strayed far from salt water. while brachydont taxa primarily ate benthic al­ If they fed on the seashore and sought shelter in gae. These food sources should have been avail­ the water, they would face the problem that the able throughout the northern parts of the des­ available "shelter" was a much higher-energy en­ mostylians' range (Japan, Sakhalin, Kamchatka, vironment than the lakes, rivers, and estuaries Alaska, British Columbia, Washington, Oregon, frequented by hippos. Except for the most pro­ and California), even during the relatively cool tected bays and inlets, coastal waters ofthe North Oligocene, and thus could have provided a high­ Pacific would not seem to provide a safe place road for desmostylian dispersal around the mar­ for such large animals to rest. It is difficult to gin of the North Pacific. conceive of a selective regime that would have If we visualize primitive desmostylians as feed­ resulted in terrestrial feeders occupying such a ing on benthic algae and other marine plants rigorous environment. Rather, it seems more along North Pacific shores, a major potential likely that, like pinnipeds, they would rest on adaptation that suggests itself is intertidal feed- 48 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY ing. The substantial Pacific tides regularly expose dence to settle the controversy over diet. Mean­ vast amounts of fresh, attached plant material in while, we remain convinced that desmostylians addition to fragments washed ashore. It would were littoral marine herbivores. be natural for a moerithere-like herbivore to exploit these resources before swimming out to Conclusions feed on submerged subtidal plants, with the at­ tendant risks of surf, currents, and hypothermia. The tethytherian order Desmostylia, although Even after desmostylians had evolved their lim­ all its known representatives are marine mam­ ited adaptations to aquatic locomotion, the inter­ mals, is the sister-group ofthe order Proboscidea tidal zone would still offer a safer, more accessi­ rather than of the order Sirenia. A new genus ble, and less energetically costly source of food. and two new species of primitive desmostylians The peculiar limb structure of desmostylians from marine Oligocene rocks of the Pacific (Shikama, 1966, 1968; Inuzuka, 1984) surely Northwest are described in this paper. They help indicates something other than purely terrestrial to span the morphological gap between the Pro­ locomotion, but it seems unnecessary to assume boscidea and Desmostylia, as do the primitive, that they needed to forage exclusively underwa­ late Paleocene tethytherian genus Minchenella ter. We suspect that desmostylians, at least in and various primitive Eocene proboscideans de­ their earlier evolutionary stages, supplemented scribed recently from both Africa and Asia. To­ their subaqueous diets with intertidal plants. gether with Minchenella, the proboscideans and Sheep in certain of the Orkney Islands, which desmostylians thus form an unnamed monophy­ feed principally on intertidal algae (Hall, 1975), letic group whose earliest presently known rep­ may provide an unexpected modern analog for resentatives occur in terrestrial late Paleocene the earliest desmostylians. rocks of southeast Asia. The relationships of this The procumbent incisors and canines of des­ clade to the Sirenia and to other paenungulates mostylians seem well suited to forking up masses are still not known in detail from paleontological of vegetation, detaching plants from rocks or evidence, nor is it known when the desmostylians sand, or uprooting mats of rhizomes as suggested first entered the sea and spread along the shores long ago by Matsumoto (1918:66, 70). In partic­ of the North Pacific Ocean as far as Mexico. ular, the wear on the ventral side of the tusk in Probably the transition was in the Eocene. In the the referred specimen of Behemotops emlongi sug­ first stages of assuming their marine habitat these gests frequent contact with abrasive substrates. interesting and even bizarre amphibious herbi­ Microwear on the cheek-teeth, examined by vores probably fed to a large extent on intertidal scanning electron microscope, may provide evi­ plants. Literature Cited

Abel, O. America Special Paper, 184:17-37, 6 figures, 5 1914. Die vorzeitlichen Sdugetiere. viii -I- 309 pages, 250 plates. figures. Jena: Gustav Fischer. Allison, R.C. 1919. Die Stdmme der Wirbeltiere. xviii -f 914 pages, 669 1976. Late Oligocene through Pleistocene Molluscan figures. Berlin and Leipzig: Vereinigung Wissen- Faunas in the Gulf of Alaska Region. In Interna­ schaftlicher Verleger Walter de Gruyter. tional Congress on Pacific Neogene Stratigraphy, To­ 1920. Lehrbuch der Paldozoologie. xvi -I- 500 pages, 700 kyo, fapan, May 16-21, 1976 [abstracts of paper], figures. Jena: Gustav Fischer. pages 10-13. [Reprinted 1977 in Proceedings of 1922. Desmostylus: ein mariner Multituberculate aus the First International Congress on Pacific Neogene dem Miozan der Nordpazifischen Kiistenregion. Stratigraphy, Tokyo, 1976:313-316, 1 figure, 1 Acta Zoologica, 3:361-394, 5 figures, 3 plates. table.] 1923. Uber einen Multituberkulaten aus dem Miozan 1977. Late Paleogene and Neogene Molluscan Faunas der Nordpazifischen Kiistenregion. Palaeontolo- ofthe Gulf of Alaska Region: Correlation, Paleo­ gische Zeitschrift, 5(3):213-225, 5 figures. climatology, and Paleozoogeography. Geological 1926. Neue Untersuchungen iiber Desmostylus, einen Society of America Abstracts with Programs, 9(7):876. Monotremen aus dem Tertiar der pazifischen 1978. Late Oligocene through Pleistocene Molluscan Kiistenregion. Verhandlungen der Zoologisch-Bo- Faunas in the Gulf of Alaska Region. Veliger, tanischen Gesellschaft, Wien, 74/75:134-138. 21(2): 171-188, 2 figures, 2 tables. [Meeting of 18 November 1925; sometimes cited Allison, R.C, and L. Marincovich, Jr. as year of publication.] 1981. A Late Oligocene or Earliest Miocene Molluscan 1933. Saugetiere (Palaontologie). In R. Dittler, G. Joos, Fauna from Sitkinak Island, Alaska. United States E. Korschelt, G. Linck, F. Oltmanns, and K. Geological Survey Professional Paper, 1233: 11 Schaum, editors, Handworterbuch der Naturwissen- pages, 3 figures, 3 plates, 6 tables. schaften (second edition), 8:859-944, 132 figures. Andrews, C.W. Jena: Gustav Fischer. 1901a. Fossil Vertebrates from Egypt. The Zoologist, series 1944. Studien iiber vergrosserte Einzelzahne des Vor- 4, 5:318-319. dergebisses der Wirbeltiere und deren Funktion. 1901b. ijber das Vorkommen von Proboscidiern in un- Palaeobiologica, 8(1/2): 1-112, 40 figures. tertertiaren Ablagerungen Aegyptens. In Tage- Addicott. W.O. blatt des V. Internationalen Zoologischen-Congresses, 1976a. Neogene Molluscan Stages of Oregon and Wash­ Berlin, 6:4, 5. [16 August 1901. Not seen. See ington. In A.E. Fritsche, H. Ter Best, Jr., and footnote 1 of Andrews (1901 c:401) and also Har­ W.W. Wornardt, editors. The Neogene Symposium, ris (1978:330, bibliography).] pages 95-115, 6 figures, 5 plates, 1 table. San 1901c. Preliminary Note on Some Recently Discovered Francisco: Society of Economic Paleontologists Extinct Vertebrates from Egypt, I. Geological Mag­ and Mineralogists (Annual Meeting, Pacific Sec­ azine, new series (Decade IV), 8(9):400-409, 4 tion). figures. 1976b. New Molluscan Assemblages from the Upper 190Id. Preliminary Note on Some Recently Discovered Member of the Tvs'in River Formation, Western Extinct Vertebrates from Egypt, II. Geological Washington: Significance in Neogene Chronostra­ Magazine, new series (Decade IV), 8(10):436-444, tigraphy. United States Geological Survey fournal of 4 figures, plus erratum for Andrews, 1901c. Research, 4(4):437-447, 7 figures, 2 tables. 1902. Preliminary Note on Some Recently Discovered 1977. Neogene Chronostratigraphy of Nearshore Ma­ Extinct Vertebrates from Egypt, III. Geological rine Basins of the Eastern North Pacific. In Pro­ Magazine, new series (Decade IV), 9(7):291-295, ceedings ofthe First International Congress on Pacific 3 figures. Neogene Stratigraphy, Tokyo, 1976, pages 151-175, 1904a. Further Notes on the Mammals of the Eocene of 4 figures. Egypt, I. Geological Magazine, new series (Decade 1981. Significance of Pectinids in Tertiary Biochronol­ V), 1(3): 109-115, 2 figures. ogy of the Pacific Northwest. Geological Society of 1904b. Note on the Barypoda, a New Order of Ungulate

49 50 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Mammals. Geological Magazine, new series (Decade Berggren, W.A., D.V. Kent, and J.J. Flynn V), 1(10):481,482. In press. Paleogene Geochronology and Chronostratigra­ 1906. A Descriptive Catalogue of the Tertiary Vertebrata of phy. In N.J. Snelling, editor, Geochronology and the Fayum, Egypt; Based on the Collection of the the Geological Record. Geological Society of Lon­ Egyptian Government in the Geological Museum, don, Special Paper. Cairo, and on the Collection in the British Museum Blainville, H.M.D. de (Natural History), London, xxxvii -I- 324 pages, 26 1816. Prodrome d'une nouvelle distribution systema­ plates, 101 figures, frontispiece. London: British tique du regne animal. Bulletin des Sciences, par la Museum (Natural History). Societe Philomatique de Paris, series 3, 3:105-124. 1909. The Systematic Position of Moeritherium. Nature, 1836. [Classification presented in 1834]. In F.-E. Guerin, 81:305. Dictionnaire pittoresque d'histoire naturelle et des Arambourg, C, J. Kikoine, and R. Lavocat phenomenes de la nature, 4:619. Paris: Au Bureau 1951. Decouverte du genre Moeritherium Andrews dans de Souscription. le Tertiaire continental du Soudan. Comptes Ren­ Brown, R.D., Jr., and H.D. Gower dus Hebdomadaires des Seances de I'Academie des 1958. Twin River Formation (Redefinition), Northern Sciences, Paris, 233(l):68-70, 1 figure. Olympic Peninsula, Washington. Bulletin of the Arambourg, C, and P. Magnier American Association of Petroleum Geologists, 1961. Gisements de Vertebres dans le bassin tertiaire de 42(10):2492-2512, 9 figures. Syrte (Libye). Comptes Rendus Hebdomadaires des Coiffait, P.-E., B. Coiffait, J.-J. Jaeger, and M. Mahboubi Seances de I'Academie des Sciences, Paris, 1984. Un nouveau gisement a Mammiferes fossiles d'age 252(8):1181-1183. Eocene superieure sur le versant sud des Ne­ Armentrout, J.M. mentcha (Algerie orientale): decouverte des plus 1975. Molluscan Biostratigraphy of the Lincoln Creek anciens Rongeurs d'Afrique. Comptes Rendus de Formation, Southwest Washington. In D.W. I'Academie des Sciences, Paris, series 2, Weaver, G.R. Hornaday, and A. Tipton, editors, 299(13):893-898, 7 figures. Paleogene Symposium and Selected Technical Papers, Coombs, M.C. pages 14-48, 7 figures. Long Beach, California: 1979. Tylocephalonyx, A New Genus of North American American Association of Petroleum Geologists, Dome-Skulled Chalicotheres (Mammalia, Perisso- Society of Economic Paleontologists and Miner­ dactyla). Bulletin ofthe American Museum of Natural alogists, and Society of Economic Geophysicists History, 164(1): 64 pages, 30 figures, 12 tables. (Annual Meeting, Pacific Sections). Coombs, W.P., Jr., and M.C. Coombs 1977. Cenozoic Molluscan Stages of Oregon and Wash­ 1977. The Origin of Anthracotheres. Neues fahrbuch ington. Geological Society of America Abstracts with fiir Geologic und Palaontologie, Monatshefte, Programs, 9(7):882, 883. 1977(10):584-599, 5 figures. 1978. Progress Report on Cenozoic Correlations, West 1979. Pilgrimella, a Primitive Asiatic Perissodactyl. Zoo­ Coast North America. Geological Society of America logical fournal of the Linnean Society, 65(3): 185- Abstracts with Programs, 10(7):360. 192, 2 figures. 1981. Correlation and Ages of Cenozoic Chronostrati­ Cope, E.D. graphic Units in Oregon and Washington. Geolog­ 1898. Syllabus of Lectures on the Vertebrata. xxxv -(- 135 ical Society of America Special Paper, 184:137-148, pages, 66 figures, errata sheet. Philadelphia: Uni­ 3 figures. versity of Pennsylvania. Armentrout,J.M., D.A. Hull,J.D. Beaulieu, and W.W. Rau, Coppens, Y., and M. Beden editors 1978. Moeritherioidea. In V.J. Maglio and H.B.S. 1983. Correlation of Cenozoic Stratigraphic Units of Cooke, editors. Evolution of African Mammals, Western Oregon and Washington. Oregon Depart­ pages 333-335. Cambridge: Harvard University ment of Geology and Mineral Industries Oil and Gas Press. Investigatio7is, 7: iii -t- 90 pages, 1 chart. (-oppens, Y., V.J. Maglio, C.T. Madden, and M. Beden Arnold, R., and H. Hannibal 1978. Proboscidea. In VJ. Maglio and H.B.S. Cooke, 1913. The Marine Tertiary Stratigraphy of the North editors. Evolution of African Mammals, pages 336- Pacific Coast of America. Proceedings ofthe Amer­ 367, 4 figures, 1 table. Cambridge: Harvard Uni­ ican Philosophical Society, 52(21 2):559-605, plates versity Press, 37-48, 12 tables. de Blainville. See Blainville, H.M.D. de Barnes, L.G., D.P. Domning, and C.E. Ray Dehm, R., and T. zu Oettingen-Spielberg 1985. Status of Studies on Fo.ssil Marine Mammals. Ma­ 1958. Palaontologische und geologische Untersuchung­ rine Mammal Science, 1 (I): 1 5-53, 9 figures. en im Tertiar von Pakistan, 2: Die mitteleocanen NUMBER 59 51

Saugetiere von Ganda Kas bei Basal in Nordwest- in the Paleocene Mammal Pararyctes. Canadian Pakistan. Bayerische Akademie der Wissenschaften, Journal of Earth Sciences, 20:19-22, 1 figure, 1 Mathematisch-Naturwissenschaftliche Klasse, Ab­ table. handlungen, new series, 91:1-54, 9 figures, 3 Frick, C. plates. 1926. Tooth Sequence in Certain Trilophodont Tetra- Deraniyagala, P.E.P. belodont Mastodons. Bulletin of the American Mu­ 1955. Some Extinct Elephants, Their Relatives and the Two seum of Natural History, 56(2):122-176, 27 fig­ Living Species. 161 pages, 13 figures, 48 plates, 15 ures, 3 tables. tables. Colombo: Ceylon National Museums Pub­ 1933. The Remains of Trilophodont-Tetrabelodont lication. Mastodons. Bulletin of the American Museum of Domning, D.P. Natural History, 59(9):505-652, 38 figures, 12 1978a. Sirenia. In V.J. Maglio and H.B.S. Cooke, editors. tables. Evolution of African Mammals, pages 573-581, 1 Gaudry, A. figure. Cambridge: Harvard University Press. 1891. Quelques remarques sur les Mastodontes a propos 1978b. Sirenian Evolution in the North Pacific Ocean. de I'animal du Cherichira. Memoires de la Societe University of California Publications in Geological Geologique de France, Paleontologie, Memoire No. Sciences, 118: xi -I- 176 pages, 37 figures, 18 plates, 8, 2(1): 1-6, 2 plates. 27 tables. Gervais, P. 1982. Evolution of Manatees, a Speculative History. 1855. Histoire naturelle des Mammiferes, avec Vindication fournal of Paleontology, 56(3):599-619, 9 figures. de leurs moeurs, et de leurs rapports avec les arts, le Domning, D.P., G.S. Morgan, and C.E. Ray commerce et I'agriculture, II: Carnivores, Probosci- 1982. North American Eocene Sea Cows (Mammalia: diens, Jumentes, Bisulques, Edentes, Marsupiaux, Sirenia). Smithsonian Contributions to Paleobiology, Monotr'emes, Phoques, Sirenides et Cetaces. 344 52: 69 pages, 34 figures, 4 tables. pages, 286 figures, 40 plates. Paris: L. Curmer. Douglass, E. [Plates not in sequence and not continuously num­ 1902. Fossil Mammalia ofthe White River Beds of Mon­ bered; see pages 343, 344.] tana. Transactions of the American Philosophical So­ Gill, T. ciety, new series, 20:237-279, plate 9, 1 map. 1871. On the Characteristics of the Primary Groups of Durham, J.W. the Class of Mammals. American Naturalist, 1944. Megafaunal Zones of the Oligocene of North­ 5(association number, September, 1871): 1-8. western Washington. University of California Pub­ 1872. Arrangement of the Families of Mammals, With lications, Bulletin of the Department of Geological Analytical Tables. Smithsonian Miscellaneous Col­ Sciences, 27(5):101-212, 7 figures, plates 13-18, 1 lections, 11(1): vi -I- 98 pages. [Publication no. 230 map, 10 tables. ofthe Smithsonian Institution.] Emlong, D.R. 1873. On the Affinities ofthe Sirenians. Proceedings of 1966. A New Archaic Cetacean from the Oligocene of the Academy of Natural Sciences of Philadelphia, Northwest Oregon. Bulletin ofthe Museum of Nat­ 25(2):262-273. ural History, University of Oregon, Eugene, Oregon, Gingerich, P.D. 3: 51 pages, 15 figures, 7 tables. 1977. A Small Collection of Fossil Vertebrates from the Fay, F.H. Middle Eocene Kuldana and Kohat Formations of 1982. Ecology and Biology of the Pacific Walrus, Odob- Punjab (Pakistan). Contributions from the Museum enus rosmarus divergens I Niger. North American of Paleontology, University of Michigan, 24(18):190- Fauna, 74: vi -I- 279 pages, frontispiece, 138 fig­ 203, 5 figures, 2 plates. ures, 50 tables. Gingerich, P.D., and D.E. Russell Flower, W.H., and R. Lydekker 1981. Pakicetus inachus, a New Archaeocete (Mammalia, 1891. An Introduction to the Study of Mammals, Living and Cetacea) from the Early-Middle Eocene Kuldana Extinct, xvi -\- 763 pages, 357 figures. London: Formation of Kohat (Pakistan). Contributions from Adam and Charles Black. the Museum of Paleontology, University of Michigan, Fordyce, R.E. 25(ll):235-246, 5 figures. 1981. Systematics of the Odontocete Whale Agorophius Gordon, K.R. pygmaeus and the Family Agorophiidae (Mam­ 1981. Locomotor Behavior of the Walrus [Odobenus). malia: Cetacea). fournal of Paleontology, journal of Zoology, London, 195:349-367, 9 fig­ 55(5): 1028-1045, 4 figures. ures, 1 table. Fox, R.C. Gorodiski, A., and R. Lavocat 1983. Evolutionary Implications of Tooth Replacement 1953. Premiere decouverte de Mammiferes dans le Ter- 52 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

tiaire (Lutetien) du Senegal. Compte Rendu Som­ Japanese; English abstract, pages 25-27; unpub­ maire des Seances, Societe Geologique de France, 15 lished, unedited English translation available to (Seance du 7 D6cembre 1953):315-317. us.] Gray, J.E. Inuzuka, N. 1866. Catalogue of Seals and Whales in the British Museum. 1984. Skeletal Restoration of the Desmostylians: Her- Second edition, viii + 402 pages, 101 figures. petiform Mammals. Memoirs of the Faculty of Sci­ London: British Museum. ence, Kyoto University, biology series, 9(2): 157-253, Gregory, W.K. 27 figures, 11 plates, 11 tables. 1910. The Orders of Mammals. Bulletin of the American Kaup, J J. Museum of Natural History, 27:1-524, 32 figures. 1855. Beitraege zur naeheren Kenntniss der urweltlichen 1920. Studies in Comparative Myology and Osteology, Saeugethiere; Ueber die Stellung, die das urweltliche IV: A Review of the Evolution of the Lacrymal Genus Halitherium und seine Verwandte im System Bone of Vertebrates with Special Reference to einnehmen muss. Volume 2, 23 pages, 7 plates. that of Mammals. Bulletin of the American Museum Darmstadt: C.W. Leske. of Natural History, 42(2):95-263, 196 figures, Kneeland, S. plate 17, 3 tables, erratum and addendum. 1850. The Manatus Not a Cetacean, but a Pachyderm. 1951. Evolution Emerging; A Survey of Changing Patterns Proceedings of the American Association for the Ad­ from Primeval Life to Man. Volume I, xxvi -I- 736 vancement of Science, 3:42-47. pages (text); volume II, viii -1-1013 pages (illustra­ Lartet, E. tions). New York: Macmillan. 1859. Sur la dentition des proboscidiens fossiles [Dino- Hall, SJ.G. therium, Mastodontes et Elephantes) et sur la distri­ 1975. Some Recent Observations on Orkney Sheep. bution geographique et stratigraphique de leurs Mammal Review, 5(2):59-64, 1 figure, 3 plates, 1 debris en Europe. Bulletin de la Societe Geologique table. de France, series 2, 16:469-515, plates 13-15. Harris, J.M. Linnaeus, C. 1978. Deinotherioidea and Barytherioidea. In V.J. Mag­ 1758. Regnum Animale. In Systema Naturae per Regna lio and H.B.S. Cooke, editors. Evolution of African Tria Naturae; Classes, Or dines. Genera, Species, cum Mammals, pages 315-332, 3 figures, 6 tables. Characteribus, Differentiis, Synonymis, Locis (10th Cambridge: Harvard University Press. edition), 1: 823 pages, emendanda and addenda. Hay, O.P. Stockholm: Laurentii Salvii. Lydgate, J. 1915. A Contribution to the Knowledge ofthe Extinct Sirenian Desmostylus hesperus Marsh. Proceedings 1412-1420. Chronicle of Troy. See Oxford English Dic­ tionary, Bibliography, page 49, Oxford: Clar­ ofthe United States National Museum, 49:381-397, endon Press, 1933. plates 56-58. Macdonald, J.R. 1923. Characteristics of Sundry Fossil Vertebrates. Pan- 1956. The North American Anthracotheres. yowrna/ of American Geologist, 39(March): 101-120, figures 4, Paleontology, 30(3):615-645, 16 figures. 5, plates 7-9. 1963. The Miocene Faunas from the Wounded Knee 1924. Notes on the Osteology and Dentition of the Gen­ Area of Western South Dakota. Bulletin of the era Desmostylus and Cornwallius. Proceedings of American Museum of Natural History, 125(3): 139- the United States National Museum, 65(8): 1-8, 2 238, 30 figures, 31 tables, 2 maps. figures, 2 plates. Macdonald, J.R., and C.B. Schultz Hirota, K. 1956. Arretotherium fricki, a New Miocene Anthraco­ 19H1. [Problems on Paleoparadoxian Mandible {Paleo­ there from Nebraska. Bulletin of the University of paradoxia tabatai).] Fossil Club Bulletin, 14:9-15, Nebraska State Museum, 4(3):53-58, 3 figures. 7 figures, 2 tables. [In Japanese; English abstract, Maglio, V.J. page 15.] 1973. Origin and Evolution ofthe Elephantidae. Trans­ Ijiri [Iziri], S. actions of the American Philosophical Society, new 1939. Microscopic Structure of a Tooth of Desmostylus. series, 63(3): 149 pages, 50 figures, frontispiece, Proceedings of the Imperial Academy, Tokyo, 18 plates. 15(5): 135-138, 6 figures. 1978. Patterns of Faunal Evolution. In V.J. Maglio and Ijiri, S., and V. Kamei H.B.S. Cooke, editors. Evolution of African Mam­ 1961. [On the Skulls of Desmostylus mirabilis Nagao from mals, pages 603-619, 6 figures, 4 tables. Cam­ South Sakhalin and of Paleoparadoxia tabatai (To- bridge: Harvard University Press. kunaga) from Gihu Prefecture, Japan.] Earth Sci­ Mahboubi, M., R. Ameur, J.Y. Crochet, and J.J. Jaeger ence, 53:1-27, 22 figures, 6 plates, 1 table. [In 1984. Earliest Known Proboscidean from Early Eocene NUMBER 59 53

of North-west Africa. Nature, 308:543-544, 3 tributions in Science, 351: 42 pages, 180 figures. figures. Munthe, J., and M.C. Coombs Marincovich, L., Jr. 1979. Miocene Dome-Skulled Chalicotheres (Mammalia, 1984. Neogene Molluscan Stages of the West Coast of Perissodactyla) from the Western United States: North America. Palaeogeography, Palaeoclimatol­ A Preliminary Discussion of a Bizarre Structure. ogy, Palaeoecology, 46(1-3): 11-24, 4 figures, 1 ta­ Journal of Paleontology, 53(1):77-91, 7 figures, 1 ble. table. Marsh, O.C. Novacek, M.J. 1888. Notice of a New Fossil Sirenian from California. 1982. Information for Molecular Studies from Anatom­ American Journal of Science, 35:94-96, 3 figures. ical and Fossil Evidence on Higher Eutherian Phy­ Matsumoto, H. logeny. In M. Goodman, editor, Macromolecular 1918. A Contribution to the Morphology, Paleobiology Sequences in Systematic and Evolutionary Biology, and Systematic of Desmostylus. Science Reports of pages 3-41, 2 figures, 2 tables. New York and the Tohoku Imperial University, series 2 (Geology), London, Plenum Press. 3(2):61-74, plate 22. Olson, E.C. 1923. A Contribution to the Knowledge of Moeritherium. 1981. The Problem of Missing Links: Today and Yes­ Bulletin ofthe American Museum of Natural History, terday. Quarterly Review of Biology, 56:405-442, 2 48(4):97-139, 11 figures, 5 tables. figures, 1 table. Matthew, W.D. Olson, S.L. 1909. Observations upon the Genus Ancodon. Bulletin of 1980. A New Genus of Penguin-Like Pelecaniform Bird the American Museum of Natural History, 26(1): 1- from the Oligocene of Washington (Pelecani- 7. formes: Plotopteridae). Natural History Museum of McKenna, M.C. Los Angeles County Contributions in Science, 1975. Toward a Phylogenetic Classification ofthe Mam­ 330:51-57, 5 figures. malia. In W.P. Luckett and F.S. Szalay, editors, 1981. A Third Species of Mancalla from the Late Pli­ Phylogeny ofthe Primates, pages 21-46, 3 figures. ocene San Diego Formation of California (Aves: New York and London: Plenum Publishing Cor­ Alcidae). Journal of Vertebrate Paleontology, poration. l(l):97-99, 3 figures. McKenna, M.C, and E. Manning Olson, S.L., and Y. Hasegawa 1977. Affinities and Palaeobiogeographic Significance 1979. Fossil Counterparts of Giant Penguins from the ofthe Mongolian Pa\eogeT\e Genus Phenacolophus. North Pacific. Science, 206(4419):688, 689, 2 fig­ Geobios (Lyon), Memoire Special, 1:61-85, 1 figure, ures. 5 plates. Osborn, H.F. McLeod, S.A., and L.G. Barnes 1902. A Remarkable New Mammal from Japan; Its Re­ 1984. Fossil Desmostylians. In B. Butler, J. Gant, and lationship to the Californian Genus Desmostylus, CJ. Stadum, editors. The Natural Sciences of Marsh. Science, new series, 16(409):7l3, 714. Orange County. Memoirs of the Natural History 1905. Ten Years Progress in the Mammalian Palaeon­ Foundation of Orange County (Newport Beach, Cal­ tology of North America. In Comptes Rendus du ifornia), 1:39-44, 6 figures. 6""" Congres International de Zoologie, Session de Merriam, J.C. Berne, 1904:86-113, 7 figures, 15 plates. 1906. On the Occurrence of Desmostylus, Marsh. Sci­ ence, new series, 24(605): 151, 152. 1907. Evolution of Mammalian Molar Teeth to and from 1911. Notes on the Genus Desmostylus of Marsh. Uni­ the Triangular Type, including Collected and Revised versity of California Publications, Bulletin ofthe De­ Researches on Trituberculy and New Sections on the partment of Geology, 6(18):403-412, 11 figures. Forms and Homologies of the Molar Teeth in the Minkoff, E.C. Different Orders of Mammals, ix + 250 pages, 215 1976. Mammalian Superorders. Zoological Journal ofthe figures. New York: Macmillan. Linnean Society, 58(2):147-158. 1909. The Feeding Habits of Moeritherium and Palaeo­ Moore, E.J. mastodon. Nature, 81(2074): 139, 140, 2 figures. 1984a. Middle Tertiary Molluscan Zones of the Pacific 1910. The Age of Mammals in Europe, Asia, and North Northwest. Journal of Paleontology, 58(3):718- America, xx -I- 635 pages. New York: Macmillan 737, 10 figures, 1 table. Company. 1984b. Molluscan Paleontology and Biostratigraphy of 1919. Palaeomastodon, the Ancestor of the Long-jawed the Lower Miocene Upper Part of the Lincoln Mastodons Only. Proceedings of the National Acad­ Creek Formation in Southwestern Washington. emy of Sciences, 5:265, 266, 1 figure. Natural History Museum of Los Angeles County Con­ 1921a. The Evolution, Phylogeny, and Classification of 54 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

the Proboscidea. American Museum Novitates, 1: 15 1940. Middle Eocene Mammals from North-west India. pages, 4 figures. Proceedings of the Zoological Society of London, 1921b. Adaptive Radiation and Classification ofthe Pro­ 110B(l/2): 127-152, 3 figures, 1 plate. boscidea. Proceedings of the National Academy of Rau, W.W. Sciences, 7(8):231-234, 1 figure. 1964. Foraminifera from the Northern Olympic Penin­ 1921c. Evolution, Phylogeny, and Classification of the sula, Washington. United States Geological Survey Mastodontoidea. Bulletin ofthe Geological Society of Professional Paper, 374-G: 33 pages, 2 figures, 7 America, 32:327-332. plates, 6 tables. 1923. New Subfamily, Generic, and Specific Stages in 1981. Pacific Northwest Tertiary Benthic Foraminiferal the Evolution of the Proboscidea. American Mu­ Biostratigraphic Framework—An Overview. Geo­ seum Novitates, 99:4 pages. logical Society of America Special Paper, 184:67-84, 1925. Final Conclusions on the Evolution, Phylogeny, 1 1 figures. and Classification of the Proboscidea. Proceedings Ray, C.E. of the American Philosophical Society, 64(1): 17-35, 1977. Fossil Marine Mammals of Oregon. Systematic Zo­ 3 figures. ology, 25(4):420-436, 2 figures. [December 1976 1936. Moeritherioidea, Deinotherioidea, Mastodonto­ issue, mailed 8 February 1977.] idea. In Proboscidea, a Monograph ofthe Discovery, 1980. Douglas Ralph Emlong, 1942-1980. [Obituary.] Evolution, Migration and Extinction of the Masto- Society of Vertebrate Paleontology News Bulletin, donts and Elephants of the World, 1: xl -I- 802 120:45,46, 1 figure. pages, frontispiece, 680 figures, 12 plates. New Reinhart, R.H. York: The American Museum Press. 1953. Diagnosis of the New Mammalian Order, Des­ 1942. Stegodontoidea, Elephantoidea. In Proboscidea, a mostylia.yowrna/ of Geology, 61(2):187. Monograph of the Discovery, Evolution, Migration 1959. A Review ofthe Sirenia and Desmostylia. Univer­ and Extinction of the Mastodonts and Elephants of sity of California Publications in Geological Sciences, the World, 2: xxvii -I- pages 805-1675, frontis­ 36(1): 146 pages, 19 figures, 14 plates (paged). piece, figures 683-1244, plates 13-30. New 1975. New Discoveries in the Order Desmostylia. [Ab­ York: The American Museum Press. stract.] i4/n^ncan Zoologist, 15(3):808. Owen, R. 1982. The Extinct Mammalian Order Desmostylia. Na­ 1859. On the Classification and Geographical Distribution tional Geographic Society Research Reports, 14:549- of the Mammalia, Being the Lecture on Sir Robert 555, 1 figure. Reade's Foundation, Delivered before the Univerity of Romer, A. S. Cambridge, in the Senate-House, May 10, 1839; To 1966. Vertebrate Paleontology, viii -I- 468 pages, 443 fig­ Which is Added an Appendix "On the Gorilla," and ures, 4 tables. Chicago and London: University "On the Extinction and Transmutation of Species." of Chicago Press. 103 pages, 9 figures. London: John W. Parker 1968. Notes and Comments on Vertebrate Paleontology, viii and Son. -I- 304 pages. Chicago and London: University of Petronievics, B. Chicago Press. 1923. Remarks upon the Skulls of Moeritherium and Pa­ Sahni, A., and K. Kumar laeomastodon. The Annals and Magazine of Natural 1980. Lower Eocene Sirenia, Ishatherium subathuensis, History, series 9, 12(67):55-61, 2 figures, plate 2. gen. et sp. nov. from the Type Area, Subathu Phillips, FJ., B. Welton, and J. Welton Formation, Subathu, Simla Himalayas, H. P. four­ 1976. Paleontologie Studies of the Middle Tertiary nal ofthe Palaeontological Society of India, 23/24(for Skooner Ciulch and Callaway Formations at Point 1978-1979): 132-135, 3 figures. Arena, California. In A.E. Fritsche, H. Ter Best, Sahni, A., and V^P. Mishra Jr., and W.W. Wornardt, editors. The Neogene 1975. Lower Tertiary Vertebrates from Western India. Symposium, pages 1 37-1 54, 5 figures, 5 plates. San Monographs of the Palaeontological Society of India, Francisco: Society of Economic Paleontologists 3: 48 pages, 6 figures, 6 plates. and Mineralogists (Annual Meeting, Pacific Sec­ Savage, RJ.C;. tion). 1969. Early Tertiary Mammal Locality in Southern Li­ Pilgrim, Ci.P,. bya. Proceedings of the Geological Society of London, 1912. The Vertebrate Fauna of the (iaj Series in the 1968(1657): 167-171,2 tables. Bugti Hills and the Punjab. Memoirs of the Geolog­ 1971. Review of the Fossil Mammals of Libya. In C ical Survey of India, Palaeontologia Indica, new Gray, editor. Symposium on the Geology of Libya, series, 4(2): 1-83, 31 plates. pages 215-225, 1 figure, 1 table. Tripoli: Fac- NUMBER 59 55

ulty of Science, University of Libya. American Museum of Natural History, 59(5):259- 1977. Review of Early Sirenia. Systematic Zoology, 293. 25(4):344-351, 2 figures, 1 table. [December 1945. The Principles of Classification and a Classifica­ 1976 issue, mailed 8 February 1977.] tion of Mammals. Bulletin of the American Museum Schlesinger, G. of Natural History, 85: xvi -\- 350 pages. 1912. Studien uber die Stammesgeschichte der Probos- Snavely, P.D., Jr., N.S. MacLeod, and W.W. Rau cidier. Jahrbuch der Kaiserlich-Koniglichen Geolo­ 1969. Geology of the Newport Area, Oregon. The Ore gischen Reichsanstalt, Wien, 62(1):87-182, 10 fig­ Bin, 31(2):25-48, 6 figures, 2 tables. Portland: ures, plates 6, 7. Oregon Department of Geology and Mineral In­ Schlosser, M. dustries. Snavely, P.D., Jr., N.S. MacLeod, H.C. Wagner, and W.W. 1904. [Review of] S. Yoshiwara and J. Iwasaki: Notes on Rau a New Fossil Mammal. Neues Jahrbuch fur Miner­ 1976. Geologic Map of the Waldport and Tidewater alogie, Geologic und Palaeontologie, Referate, Quadrangles, Lincoln, Lane, and Benton Coun­ 1904(2):464, 465. ties, Oregon. United States Geological Survey Mis­ 1911. Beitrage zur Kenntnis der Oligozanen Landsiiu- cellaneous Investigations Series Map, 1-866: 1 sheet. getiere aus dem Fayum: (Agypten). Beitrdge zur Snavely, P.D., Jr., A.R. Niem, N.S. MacLeod, J.E. Pearl, Palaontologie und Geologic Osterreich-Ungarns und and W.W. Rau des Orients, 24(2):51-167, 1 figure, plates 9-16, 1 1980. Makah Formation—A Deep-Marginal-Basin Se­ table. quence of Late Eocene and Oligocene Age in the Sera, G.L. Northwestern Olympic Peninsula, Washington. 1954. La posizione zoologica del genere mammale mio- United States Geological Survey Professional Paper, cenico Desmostylus. Archivio Zoologica Italiano [for 1162-B: 28 pages, 16 figures, 2 tables. 1953], 38(2):485-506, 4 figures. Snavely, P.D.,Jr., A.R. Niem, andJ.E. Pearl Shikama, T. 1978. Twin River Group (Upper Eocene to Lower Mio­ 1957. On the Desmostylid Skeltons [sic]. Natural Science cene)—Defined to Include the Hoko River, and Museum (National Science Museum, Tokyo), Makah, and Pysht Formations, Clallam County, 24(1/2):16-21, 2 plates. Washington. United States Geological Survey Bulle­ 1966. Postcranial Skeletons of Japanese Desmostylia. Pa­ tin, 1457-A:A111-A120, A124-A132 (in part), laeontological Society of Japan Special Papers, 12:1- figures 8-10. 202, 116 figures, 12 plates, 31 tables. [Errata Tabor, R.W., and W.M. Cady appeared in Shikama, 1968.] 1978. Geologic Map ofthe Olympic Peninsula, Washing­ 1968. Additional Notes on the Postcranial Skeletons of ton. United States Geological Survey Miscellaneous Japanese Desmostylia. Science Reports of the Yoko­ Investigations Series Map, 1-994: 2 sheets. hama National University, Section II, Biological and Tassy, P. Geological Sciences, 14:21-26, 5 figures, plates 3- 1979. Relations phylogenetiques du genre Moeritherium 6. Andrews, 1901 (Mammalia). Comptes Rendus Heb­ Sickenberg, O. domadaires des Seances de I'Academie des Sciences, 1934. Beitrage zur Kenntnis Tertiarer Sirenen. Memoires Paris, series D (Sciences naturelles), 289(2):85- du Musee Royal d'Histoire Naturelle de Belgique, 63: 88, 1 figure. 352 pages, 52 figures, 11 plates. 1981. Le crane de Moeritherium (Proboscidea, Mam­ 1938. 1st Desmostylus eine Sirene? Palaeobiologica, malia) de PEocene de Dor el Talha (Libye) et le 6(2):340-357. probleme de la classification phylogenetique du Simons, E.L. genre dans les Tethytheria McKenna, 1975. Bul­ letin du Museum National d'Histoire Naturelle, se­ 1964. Yale Peabody Museum. Society of Vertebrate Paleon­ ries 4, section C, 3(1):87-147, 15 figures, 6 plates. tology News Bulletin, 70:14, 15, 1 figure. 1968. Early Cenozoic Mammalian Faunas, Fayum Prov­ 1982. Les principales dichotomies dans I'histoire des Proboscidea (Mammalia): une approche phylo­ ince, Egypt, Part I: African Oligocene Mammals; genetique. Geobios (Lyon), Memoire Special, 6:225- Introduction, History of Study, and Faunal 245, 8 figures, 2 tables. Succession. Bulletin of the Peabody Museum of Nat­ ural History, Yale University, 28:1-21, 103-105, 1 Thenius,E. 1960. Huftiere (Ungulata). In E. Thenius and H. Hofer, figure. Stammesgeschichte der Sdugetiere, pages 188-267. Simpson, G.G. 1931. A New Classification of Mammals. Bulletin of the Berlin: Springer Verlag. 56 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

1969. Stammesgeschichte der Saugetiere (einschliesslich Kohat (Pakistan). Contributions from the Museum of der Hominiden). Handbuch der Zoologie, 8(2):369- Paleontology, University of Michigan, 26(7): 117- 722, figures 394-715. 139, 6 figures, 3 tables. 1980. Grundziige der Faunen- und Verbreitungsgeschichte West, R.M. der Sdugetiere. 375 pages, 112 figures, 5 tables. 1971. Deciduous Dentition ofthe Early Tertiary Phen- Stuttgart: Gustav Fischer. acodontidae (Condylarthra, Mammalia). American Tobien, H. Museum Novitates, 2461: 37 pages, 23 figures. 1971. Moeritherium, Palaeomastodon, Phiomia aus dem 1980. Middle Eocene Large Mammal Assemblage with Palaogen Nordafrikas und die Abstammung der Tethyan Affinities, Ganda Kas Region, Pakistan. Mastodonten (Proboscidea, Mammalia). Mitteilun­ Journal of Paleontology, 54(3):508-533, 1 figure, gen aus dem Geologischen Institut der Technischen 5 plates, 5 tables. Universitdt Hannover, 10:141-163, 10 figures, 1 1983. South Asian Middle Eocene Moeritheres (Mam­ table. malia: Tethytheria). Annals of Carnegie Museum, 1976. Zur palaontologischen Geschichte der Mastodon­ 52(16):359-373, 6 figures, 1 table. ten (Proboscidea, Mammalia). Mainzer Geowissen- Whitmore, F.C, Jr., and A.E. Sanders schaftliche Mitteilungen, 5:143-225, 52 figures, 1 1977. Review of the Oligocene Cetacea. Systematic Zool­ table. ogy, 25(4):304-320, 10 figures, 1 table. [Decem­ 1978. The Structure ofthe Mastodont Molar (Probos­ ber 1976 issue, mailed 8 February 1977.] cidea, Mammalia), Part 3: The Oligocene Masto­ Winge, H. dont Genera Palaeomastodon, Phiomia and the Eo/ 1906. Jordfundne og nulevende Hovdyr (Ungulata) fra La- Oligocene Paenungulate Moeritherium. Mainzer goa Santa, Minas Geraes, Brasilien, Med Udsigt over Geowissenschaftliche Mitteilungen, 6:177-208, 26 Hovdyrenes indbyrdes Slaegtskab. E Museo Lundii figures. (Copenhagen), 3(1): 239 pages, 9 plates. Trouessart, E.-L. 1924. Pattedyr-Slaegter, III: Ungulata, Cetacea. 270 pages, 1879. Primates (Simiae, Prosimiae, Chiroptera), 1. In frontispiece. Copenhagen: H. Hagerups Forlag. Conspectus systematicus et geographicus Mammalium 1942. The Interrelationships of the Mammalian Genera. Volume 111. Ungulata, Cetacea. [Translated from tam viventium quam fossilium; Catalogue systema­ Danish by E. Deichmann, Ph.D., and G.M. Allen, tique, synonymique et geographique des Mammiferes Ph.D]. 308 pages, frontispiece. Copenhagen: CA. vivants et fossiles. 93 pages. [Privately published; Reitzels Forlag. extracted from Revue et Magasin de Zoologie, Yoshiwara, S., and J. Iwasaki 1878.] 1902. Notes on a New Fossil Mammal, yowrna/ of the 1898. Tillodontia, Ungulata, Sirenia, Cetacea, Edentata, College of Science, Imperial University, Tokyo, Japan, Marsupialia, Allotheria, Monotremata. In Catalo­ 16(6): 1-1 3, 4 figures, 3 plates. gus Mammalium tam viventium quam fossilium (new Zhang, Y.-p. edition), 2(4/5): pages 665-1264. Berlin: R. 1978. [Two New Genera of Condylarthran Phenacolo­ Friedlander & Sohn. phids from the Paleocene of Nanxiong Basin, VanderHoof V.L. Guangdong.] Vertebrata PalAsiatica, 16(4):267- 1937. A Study of the Miocene Sirenian Desmostylus. 274, 1 plate, 1 table. [In Chinese, English sum­ University of California Publications, Bulletin of the mary, page 274.] Department of Geological Sciences, 24(8): 169-262, 1980. Minchenella, New Name for Conolophus Zhang, 65 figures, 2 maps. 1978. Vertebrata PalAsiatica, 18(3):257. Van Valen, L. Zigno, A. de 1978. The Beginning of the Age of Mammals. Evolution­ 1 875. Annotazioni Paleontologiche; Sirenii fossili trovati ary Theory, 4(2):45-80, 7 figures, 2 tables, 8 plates. nei Veneto. Memorie del Reale Istituto Veneto di Weber, M. Scienze, Lettere ed Arti, 18:427-456, plates 14-18. 1928. Systematischer Teil. In Die Sdugetiere, Einfiihrung Zittel, K.A. von in die Anatomic und Systematik der recenten und 1 893. Vertebrata (Mammalia). In Handbuch der Palaeon­ fossilen Mammalia (second edition), 2: xxiv -I- 898 tologie, 1: Paldozoologie, xii -(- 799 pages, 590 fig­ pages, 573 figures. Jena: Gustav Fischer. ures. Miinchen and Leipzig: R. Oldenbourg. Wells, N.A., and P.D. Gingerich 1925. Text-book of Palaeontology, III: Mammalia. [Revised 1983. Review of Eocene Anthracobunidae (Mammalia, by Max Schlosser; translation revised, with addi­ Proboscidea) with a New Genus and Species,yoz- tions, by Sir Arthur Smith Woodward.] viii -1-316 aria palustris, from the Kuldana Formation of pages, 374 figures. London: Macmillan and Co. REQUIREMENTS FOR SMITHSONIAN SERIES PUBLICATION

Manuscripts intended for series publication receive substantive footnotes are required, use the short form (author, brief title, page) review (conducted by their originating Smithsonian museums or with the full citation in the bibliography. offices) and are submitted to the Smithsonian Institution Press Footnotes, when few in number, whether annotative or biblio­ with Form SI-36, which must show the approval of the appropriate graphic, should be typed on separate sheets and inserted imme­ authority designated by the sponsoring organizational unit. Re­ diately after the text pages on which the references occur. Exten­ quests for special treatment—use of color, foldouts, case-bound sive notes must be gathered together and placed at the end of covers, etc.—require, on the same form, the added approval of the text in a notes section. the sponsoring authority. Bibliography, depending upon use, is termed "Literature Cited," Review of manuscripts and art by the Press for requirements "References," or "Bibliography." Spell out titles of books, articles, of series format and style, completeness and clarity of copy, and journals, and monographic series. For book and article titles use arrangement of all material, as outlined below, will govern, within sentence-style capitalization according to the .rules of the lan­ the judgment of the Press, acceptance or rejection of manuscripts guage employed (exception: capitalize all major words in English). and art. For journal and series titles, capitalize the initial word and all Copy must be prepared on typewriter or word processor, subsequent words except articles, conjunctions, and prepositions. double-spaced, on one side of standard white bond paper (not Transliterate languages that use a non-Roman alphabet according erasable), with 1 VA" margins, submitted as ribbon copy (not to the Library of Congress system. Undertine (for italics) titles of carbon or xerox), in loose sheets (not stapled or bound), and journals and series and titles of books that are not part of a series. accompanied by original art. Minimum acceptable length is 30 Use the parentheses/colon system for volume (number): pages. pagination: "10(2):5-9." For alignment and arrangement of ele­ Front matter (preceding the text) should include: title page ments, follow the format of recent publications in the series for with only title and author and no other information, abstract page which the manuscript is intended. Guidelines for preparing bibli­ with author, title, series, etc., following the established format; ography may be secured from Series Section, SI Press. table of contents with indents reflecting the hierarchy of heads in Legends for illustrations must be submitted at the end of the the paper; also, foreword and/or preface, if appropriate. manuscript, with as many legends typed, double-spaced, to a First page of text should carry the title and author at the top page as convenient. of the page; second page should have only the author's name Illustrations must be submitted as original art (not copies) and professional mailing address, to k)e used as an unnumbered accompanying, but separate from, the manuscript. Guidelines for footnote on the first page of printed text. preparing art may be secured from Series Section, SI Press. All Center heads of whatever level should be typed with initial types of illustrations (photographs, line drawings, maps, etc.) may caps of major words, with extra space above and below the head, be intermixed throughout the printed text. They should be termed but no other preparation (such as all caps or undertine, except for Figures and should be numbered consecutively as they will the undertine necessary for generic and sfiecific epithets). Run-in appear in the monograph. If several illustrations are treated as paragraph heads should use period/dashes or colons as neces­ components of a single composite figure, they should be desig­ sary. nated by lowercase italic letters on the illustration; also, in the Tabulations within text (lists of data, often in parallel columns) legend and in text references the italic letters (underiined in copy) can be typed on the text page where they occur, but they should should be used: "Figure 9b." Illustrations that are intended to not contain rules or numbered table captions. follow the printed text may be termed Plates, and any components Formal tables (numbered, with captions, boxheads, stubs, should t)e similarly lettered and referenced: "Plate 9b." Keys to rules) should be submitted as carefully typed, double-spaced copy any symbols within an illustration should appear on the art rather separate from the text; they will be typeset unless otherwise than in the legend. requested. If camera-copy use is anticipated, do not draw rules Some points of style: Do not use periods after such abbrevi­ on manuscript copy. ations as "mm, ft, USNM, NNE." Spell out numbers "one" through Taxonomic keys in natural history papers should use the "nine" in expository text, but use digits in all other cases if possible. aligned-couplet form for zoology and may use the multi-level Use of the metric system of measurement is preferable; where indent form for botany. If cross referencing is required between use of the English system is unavoidable, supply metric equiva­ key and text, do not include page references within the key, but lents in parentheses. Use the decimal system for precise meas­ number the keyed-out taxa, using the same numbers with their urements and relationships, common fractions for approximations. corresponding heads in the text. Use day/month/year sequence for dates: "9 April 1976." For Synonymy in zoology must use the short form (taxon, author, months in tabular listings or data sections, use three-letter abbre­ year:page), with full reference at the end of the paper under viations with no periods: "Jan, Mar, Jun," etc. Omit space between "Literature Cited." For botany, the long form (taxon, author, initials of a personal name: "J.B. Jones." abbreviated journal or book title, volume, page, year, with no Arrange and paginate sequentially every sheet of manu­ reference in "Literature Cited") is optional. script in the following order: (1) title page, (2) abstract, (3) con­ Text-reference system (author, year: page used within the text, tents, (4) foreword and/or preface, (5) text, (6) appendixes, (7) with full citation in "Literature Cited" at the end of the text) must notes section, (8) glossary, (9) bibliography, (10) legends, (11) be used in place of bibliographic footnotes in all Contributions tables. Index copy may be submitted at page proof stage, but Series and is strongly recommended in the Studies Series: plans for an index should be indicated when manuscript is sub­ "(Jones, 1910:122)" or "...Jones (1910:122)." If bibliographic mitted. ^ %

^ X

il^'S^^^'

^.y^"-. fm