Peabody Museum

of Natural History Yale University

New Haven, CT 06511

Me 29G4K

—_—

JAN 19 anton

LIBRARIES |

(Received 27 April 1983) Abstract

Two species of the middle Eocene palae- anodont (Mammalia) genus Metacheiromys, M. marshi Wortman, 1903 (=M. tatusia Osborn, 1904) and M. dasypus Osborn, 1904 (=M. osborni Simpson, 1931), are recog- nized as valid. The previously described Palaeanodonta (Edentata) are reviewed and a new subfamily, Propalaeanodontinae, is erected for Propalaeanodon. Propalaeanodon is the plesiomorphous sister-group of the remaining metacheiromyids. The Epocoicotherlidae is a paraphyletic family in- cluding the monophyletic Epoicotherlinae and a number of poorly known genera that can not presently be assigned to distinct subfamilies.

Key Words

Metacheiromys, Palaeanodonta, Edentata, Propalaeanodontinae, indeterminate Pri- mates, fossil mammals.

Introduction

The Palaeanodonta are an extinct, archaic group of Paleocene to Oligocene, eden- tatelike mammals of uncertain affinities

© Copyright 1984 by the Peabody Museum of Natural History, Yale University. All rights reserved. No part of this publication, except brief quotations for scholarly pur- poses, may be reproduced without the written permis- sion of the Director, Peabody Museum of Natural History.

Postilla. Number 192

27 March 1984

ew of the Palaeanodonta

Robert Milton Schoch

which, with one exception, are known solely from the early Tertiary of the Rocky Moun- tain region of the western United States. The one possible exception is a palaeanodont specimen reported from the Oligocene of West Germany (Heissig, 1982). Metacheiromys Wortman, 1903, was the first genus of palaeanodonts to be de- scribed; however, this genus has never been photographically illustrated and has not been critically reviewed since Simpson’s (1931) classic study. In the past decade there has been a resurgence of interest in the Palae- anodonta and a number of new species have been described (West, 1973; Rose, Bown and Simons, 1977; Rose, 1978, 1979; Heissig, 1982; Rose and Emry, 1983). However,

no systematic review of the known genera and species of palaeanodonts has been at- tempted. Here | revise the species-level taxonomy of Metacheiromys and illustrate the genoholotype and the more important re- ferred specimens housed in the Peabody Museum of Natural History (including speci- mens which have been newly discovered in the Peabody Museum Collections). Finally, | briefly review the known Palaeanodonta.

Abbreviations

Institutions

AMNH American Museum of Natural History, New York City

YPM Vertebrate Paleontology Collec-

tions, Peabody Museum of Natural History, Yale University, New Haven

2 Metacheiromys and the Palaeanodonta

Postilla 192

YPM O Osteology Collections, Peabody Museum of Natural History, Yale University, New Haven

Statistics

CV Coefficient of Variation

M Mean

OR Observed range

SD Standard deviation

All measurements were taken with a Helios dial caliper and are expressed in millimeters.

Systematic Paleontology

CLASS Mammalia Linnaeus, 1758, p. 14 SUBCLASS Theria Parker & Haswell, 1897,

p. 448

INFRACLASS Tribosphenida McKenna, 1975,

Os 25)

SUPERCOHORT Eutheria Gill, 1872, p. 1

COHORT Edentata Cuvier, 1798, p. 142

ORDER Palaeanodonta Matthew, 1918, p. 620

Discussion

The two families Metacheiromyidae Wortman (1903, p. 347) and Epoicotheriidae Simpson (1927, p. 285) have generally been regarded as closely related and have been placed together in Matthew's (1918) subor- der Palaeanodonta (Colbert, 1942; Simpson, 1945, 1959; Rose, 1978, 1979; but fora contrary view, see Emry, 1970). As Rose (1978, p. 665) has recently noted, the Palaeanodonta are recognizable by the fol- lowing suite of characters:

1) Distinctive, large canines which show normal occlusion; |.e., the lower canines occlude in front of the upper canines (Colbert, 1942).

2) The progressive reduction in size and number of cheek teeth.

3) The progressive reduction and loss of enamel on the cheek teeth.

4) Development of a posteriorly-placed medial buttress on the mandible with an internal mandibular groove beneath it.

5) A fossorially adapted skeleton.

Matthew (1918) originally proposed the

-alaeanodonta as a suborder of the order

Edentata, but also suggested that it might show affinities with both the Pholidota (pan- golins) and Xenarthra (Edentata). Subse- quently, most authorities (e.g., Simpson, 1931, 1945; Colbert, 1942; Gazin, 1952) placed the Palaeanodonta in the Edentata; ; Simpson (1931) argued strongly against pholidotean ties for the Palaeanodonta. In 1970, Emry described a new manid, Patriomanis, and discussed the then known palaeanodonts and pangolins. He concluded that the manids were derived from the metacheiromyids and thus placed the palaeanodonts in the order Pholidota. How- ever, Emry dropped the use of the term Palaeanodonta ‘‘since the mutual affinities of the three families are not clearly known” (Emry, 1970, p. 507). Rose (1978, 1979) has recently suggested, on the basis of new ma- terial, that the Xenarthra, Pholidota and Palaeanodonta together form a monophyletic clade relative to other eutherians, but the re- lationships within this trichotomy (i.e., which two groups are more closely related to each other relative to the third) is as yet unre- solved. Szalay (1977) came to essentially the Same conclusions on the basis of a recon- sideration of the astragalocalcaneal complex of these three groups. Here | adopt Szalay’s (1977) higher level classification of the Palaeanodonta and tentatively consider this group to be a distinct order which is possibly referable, along with the orders Xenarthra Cope (1889, p. 657) and Pholidota Weber (1904, p. 412), to the cohort Edentata Cuvier (1798, p.' 142):

FAMILY Metacheiromyidae Wortman, 1903, p. 347

SUBFAMILY Metacheiromyinae Wortman, 1903; Os 84:7

Metacheiromys Wortman, 1903, p. 347

Type Species

Metacheiromys marshi Wortman, 1903 (=M. tatusia Osborn, 1904).

3 Metacheiromys and the Palaeanodonta

Included Species

The type species and M. dasypus Osborn, 1904 (=M. osborni Simpson, 1931).

Distribution

Middle Eocene (Bridgerian) of the Bridger Formation, Bridger Basin, Wyoming.

Revised Diagnosis

Metacheiromyines distinguished by the fol- OnP lee

Ve he canines large, compressed (with triangular bases) and enamel-bearing; postcanines small (vestigial), single-rooted; most of length of jaw edentulous; shelf for lower horny plate not grooved, narrower and less conspicuous than in Palaeanodon; and os- sified bullae complete.

lowing features: dental formula

Metacheiromys marshi Wortman, 1903, p. 347

(Figs. 1—5; 7A, C; 8; 9) Synonymy

Metacheiromys marshi Wortman, 1903,

p. 347, figs. 105—108 (non fig. 109) Metacheiromys tatusia Osborn, 1904, p. 165 Metacheiromys marshi: Simpson, 1931, psc0s, tigs: SA; 113A; 19

Metacheiromys tatusia: Simpson, 1931,

bE SOS: TIGS 1,386; |S3B—F, 15A, 18, 20, 21 (non fig. 22)

Metacheiromys tatusia: Emry, 1970, fig. 30C Metacheiromys sp.: Szalay, 1977, fig. 12B

Lectotype

YPM 12903, left dentary with root of canine and alveoli for first two postcanines (Fig. 1A—C). Lectotype selected from the original type material by Simpson (1931, p. 304).

Postilla 192

Paralectotypes

YPM 12903 (the following specimens may pertain to the same individual as that rep- resented by the lectotype), right maxilla fragment bearing a broken canine (Fig. 1E, F): left upper canine (Fig. 1G, H); glenoid cavity of right scapula (Fig. 1K—M); proximal and distal ends of right humerus (Fig. 2); proximal two-thirds of right ulna (Fig. 3A—C); axis of second cervical vertebra which is missing the odontoid process (Fig. 11, J); skull fragments; vertebral fragments (mostly cervical, thoracic and caudal centra); rib fragments; and distal end of radius.

Original syntypes subsequently excluded from the genus and referred to the order Primates, genus indeterminate (Simpson, 1931, p. 304): YPM 12908, right acetabular part of pelvis (Fig. 7B); proximal and distal ends of right tibia (Fig. 6).

Horizon and Locality of YPM 12903

Middle Eocene, probably Bridger B at Grizzly Buttes, Bridger Formation, Bridger Basin, Wyoming.

Discussion of YPM 12903

As Osborn (1904) and Simpson (1931) noted, all of the specimens catalogued under YPM 12903 were presumably collected together (although the collector and date are un- known, presumably collected by a Yale Sci- entific Expedition in the early 1870s), but represent parts of at least two individuals without duplication of any parts. Thus, YPM 12903 includes parts of the skull, dentary, forelimb and vertebral column of a metacheiromyid and parts of the pelvis and hindlimb of a primate.

Holotype of Metacheiromys tatusia AMNH 11549, skull, right dentary and the

majority of the skeleton (described and illus- trated by Simpson, 1931).

4 Metacheiromys and the Palaeanodonta

Horizon and Locality of AMNH 11549

Collected in 1903 by Albert Thomson from middle Eocene-aged strata, Bridger Bo at Grizzly Buttes, Bridger Formation, Bridger Basin, Wyoming.

Specimens Referred to M. marshi in the Peabody Museum

YPM 13500, proximal end of right humerus; fragments of distal ends of right and left humeri; proximal and distal ends of both femora (Fig. 4B, C, G); proximal ends of both tibiae; and miscellaneous vertebrae: col- lected by J. W. Chew in 1873 from middle Eocene-aged strata, probably Bridger C, at Dry Creek, Bridger Formation, Bridger Basin, Wyoming.

YPM 13501, fragments of pelvis, includ- ing parts of right and left ischia (Fig. 9); fragments of left femur; proximal and distal ends of right and left tibiae (Fig. 5); proximal half of right fibula (still in matrix); partial ar- ticulated tarsus and metatarsus (Fig. 8A, B) of right pes; miscellaneous phalanges and sesamoids, including an ungual phalanx of the ?manus (Fig. 8D—F); poorly-preserved vertebral centra of the thoracic and lumbar series; and first seven caudal vertebrae pre- served articulated and partly embedded in matrix (Fig. 9). According to Simpson (1931, p. 305), YPM 13501 originally included the last two sacrals. Traces of glue and a fresh break are evident on the anterior surface of the first preserved caudal of YPM 13501, but the two sacrals mentioned by Simpson (1931) have presumably been lost since he described this specimen. YPM 13501 was collected by the Yale Scientific Expedition of 1873 from middle Eocene-aged strata, prob- ably Bridger B, but possibly Bridger C, at Grizzly Buttes, Bridger Formation, Bridger Basin, Wyoming.

YPM 13502, posterior left dentary frag- ment with base of ascending ramus (Fig. 1D); miscellaneous vertebral centra including an axis missing the odontoid process and a number of caudals; glenoid cavity of right scapula; proximal ends of both humeri; distal

Postilla 192

end of right humerus; a nearly complete sac- rum with partial left ilium and acetabular part of pelvis (Fig. 8C); proximal end of right femur; distal ends of right and left femora (Fig. 4A, D, H); proximal ends of right and left tibiae; distal end of right tibia; and mis- cellaneous isolated foot bones, including an ungual phalanx: collected by the Yale Scien- tific Expedition of 1873 from middle Eocene-aged strata, Bridger B at Grizzly Buttes, Bridger Formation, Bridger Basin, Wyoming.

YPM 13503, broken vertebral centra, in- cluding axis of second cervical vertebra miss- ing the odontoid process; proximal end of right humerus; distal fragments of right and left humeri; proximal end of right femur; dis- tal end of left femur; proximal end of left tibia; right ilium and acetabular part of pelvis (Fig. 7A); proximal end of left ilium and acetabular part of pelvis; and incomplete sacrum (Fig. 7C): locality data the same as for YPM 13502.

YPM 40066, proximal end of right femur; distal end of right femur; proximal end of right humerus: no locality data; this speci- men bears YPM Accession number 1072 which indicates that it was in O. C. Marsh's possession in 1877; presumably from the Bridger Basin, Wyoming.

YPM 40067, miscellaneous vertebral centra and rib fragments; glenoid cavities of both scapulae; partial left humerus; part of shaft of right humerus; left ulna (Fig. 3D, E); proximal half of left radius (Fig. 4E, F); mis- cellaneous foot bones and bone fragments: collected by J. F. Page from middle Eocene-aged (Bridgerian) strata of the Bridger Formation in the area of Grizzly Buttes, Bridger Basin, Wyoming, 6 Sep- tember 1871.

Revised Diagnosis Smallest known species of Metacheiromys;

two lower postcanines approximately Sub- equal in size.

5 Metacheiromys and the Palaeanodonta

Metacheiromys dasypus Osborn, 1904, p. 164 (Fig. 10)

Synonymy

Metacheiromys dasypus Osborn, 1904,

p. 164

Metacheiromys dasypus: Simpson, 1931, p. 306, figs. 3C, 4—8, 10-12, 14, 15B, 15C, iG ralh7,, 22

Metacheiromys osborni Simpson, 1931,

pm: S06; 11g. 3D

Metacheiromys dasypus: Emry, 1970, fig. 28C

Holotype

AMNH 11718, skull, jaws and partial skele- ton.

Horizon and Locality of the Type

Collected by Walter Granger in 1903 from middle Eocene-aged strata, Bridger Bp at Grizzly Buttes, Bridger Formation, Bridger Basin, Wyoming.

Holotype of Metacheiromys osborni

AMNH 12119, right dentary and anterior part of left dentary.

Horizon and Locality of AMNH 12119

Collected by P. Miller in 1904 from middle Eocene-aged strata, Bridger Dz at Lone Tree, Bridger Formation, Bridger Basin, Wyoming.

Referred Specimens

YPM 40068, proximal end of left humerus; proximal end of left femur; proximal end of right tibia; partial right astragalus (Fig. 10A, C); right cuboid (Fig. 10B, D); miscellaneous phalanges, other foot bones and bone frag- ments: collected by G. G. Lobdell from mid- dle Eocene-aged (Bridgerian) strata, Bridger Formation, in the Grizzly Buttes area, Bridger Basin, Wyoming, 4 September 1871.

Postilla 192

YPM 40069 (YPM Accession number 511), head of right astragalus; two incom- plete metatarsals; first phalanx of second digit of left pes (Fig. 10E, F); first and second phalanges of second digit of right pes: col- lected by S. Smith and J. W. Chew on 17 September 1873 from middle Eocene-aged Strata, probably Bridger B, Bridger Forma- tion, Bridger Basin, Wyoming.

Revised Diagnosis

Largest known species of Metacheiromys, approximately 75% larger than M. marshi; second lower postcanine much smaller than first.

Description of MVetacheiromys

The known morphology of Metacheiromys |s adequately described in Wortman (1903) and Simpson (1931). However, these works are illustrated only by simple line drawings. Therefore, | supplement Wortman’s and Simpson's descriptions by photographically illustrating the genoholotype and selected re- ferred specimens of Metacheiromys (Figs. 1—10). Except for differences in absolute size and relative size of the second lower postcanines, the two species are identical in known morphology.

Discussion of Metacheiromys

In 1903, as part of a study of the Eocene mammals in the Marsh Collection of the Yale Peabody Museum, Jacob L. Wortman de- scribed a new genus and species of ‘‘pri- mate,’’ Metacheiromys marshi, based upon YPM 12903. Primate affinities for Metacheiromys were based primarily on the tibia (Fig. 6) which is distinctly primate in morphology (see discussion by Wortman, 1903, p. 351). He allied his new taxon with the extant aye-aye of Madagascar, Daubentonia (=Cheiromys). Wortman was under the impression that the materials catalogued under YPM 12903 pertained to a

6 Metacheiromys and the Palaeanodonta

Postilla 192

ee eee eee eS SEE

single individual. However, as noted above, YPM 12903 is a composite of at least two individuals of approximately the same size and without duplication of any elements: parts of the head, forequarters and vertebral column of a palaeanodont and parts of the hindquarters of a primate. Thus, he was mis- led as to the affinities of Wetacheiromys. Also in 1903, an American Museum of Natural History field party led by Walter Granger discovered two partial ‘‘edentate™ skeletons in the Eocene Bridger Formation of Wyoming, AMNH 11549 and AMNH 11718 (Simpson, 1931). Upon study, it was quickly recognized that these skeletons per- tained to Wortman’s genus, that Metacheiromys was an ‘‘edentate”’ and that YPM 12903 was a composite specimen. Osborn (1904) published a brief (three un- illustrated pages) announcement clarifying the status of Metacheiromys marshi and coined the names Metacheliromys tatusia and M. dasypus for AMNH 11549 and AMNH 11718 respectively. In 1918 Matthew described in detail the closely related genus Palaeaonodon and discussed the relation- ships of the Metacheiromyidae. However, It was left to Simpson (1931) to describe, illus- trate and discuss the genus Metacheiromys. Osborn (1904) distinguished M. dasypus as being much larger than M. marshi, and M. tatusia as being smaller than M. marshi. As Simpson (1931) noted, M. dasypus Is con- siderably larger (by about 75%) than M. marshi/tatusia and is also distinguished by a relative reduction in size of the second lower cheek tooth. Simpson (1931) also noted that the lectotype of M. marshi is only slightly larger and heavier (more rugose) than the holotype of M. tatusia. Indeed, Simpson (1931, pp. 305-6) stated that “it is quite possible that a large series of specimens would show M. tatusia and M. marshi to in- tergrade completely.’’ However, Simpson (1931) retained these species as distinct. Furthermore, Simpson described a fourth species of Metacheiromys, M. osborni, on the basis of a single dentary, AMNH 12119, that is ‘’slightly longer than in M. dasypus (about 10%), stouter, canine larger, bone

more swollen around alveolus” (Simpson, 1931, p. 306).

| believe that the specific distinctions be- tween M. marshi and M. tatusia, and be- tween M. dasypus and M. osborni are ex- tremely dubious. Rose (1978) has suggested that palaeanodonts may have been sexually dimorphic, as is the extant edentate anteater Myrmecophaga (Grasse, 1955). However, even without the presence of pronounced sexual dimorphism, the differences seen be- tween the presumed species within each pair (M. marshi/tatusia and M. dasypus/ osborni) are extremely slight and suggest that they may be due to individual variation (confounded in part by differing ontogenetic ages of the individuals concerned). In anal- ogy, within the single species of the extant cingulate (Edentata), Dasypus novemcinctus (the nine-banded armadillo), as represented by osteological specimens in the Peabody Museum of Natural History, | found as much variation in size and robustness among indi- viduals as is seen between M. marshi-M. tatusia and M. dasypus-M. osborni (Tables 1 and 2). Furthermore, | have found Simpson's (1931) allocations of small specimens of Metacheiromys to either M. tatusia or M. marshi to be unreplicable. YPM 13501, a specimen which Simpson referred to the smaller M. tatusia includes some limb bones which, if anything, are slightly larger than some which he referred to M. marshi. Simpson referred only three specimens to M. dasypus /osborni, two of which are the holotypes. The third specimen, a few frag- ments of a hind foot of Metacheiromys (YPM 40069: Fig. 10E, F), shares no ele- ments in common with the two type speci- mens.

Taking all of the above considerations into account, | recognize only two species of Metacheiromys, M. marshi Wortman, 1903 (= the junior subjective synonym M. tatusia Osborn, 1904) and M. dasypus Osborn, 1904 (= the junior subjective synonym M. osborni Simpson, 1931).

With this synonymy in mind, the bio- stratigraphic distribution of Metacheiromys can be reviewed (cf. Simpson, 1931, p. 307).

eee eS EE ES EE EEE ESSE EE EEE Eee

7 Metacheiromys and the Palaeanodonta

Table 1

Metric data on Dasypus novemcinctus.

Postilla 192

YPM O Length upper cheek Length lower cheek Number tooth row tooth row

Right Side Left Side Right Side Left Side 2888 23a) 2S 25.5 24.5 2334 25.1 25.0 26.2 26.0 2739 2397) D5) DOW. D3) 2740 26.5 26.4 28.1 Dil) 4474 26.1 25.8 26.4 Dies 9505 22.5 2228 22.6 22.9 10065 24.7 24.3 26.2 25:85 OR 22.5—26.5 22.8—26.4 22.6—28.1 22.9—27.3 M 24.61 24.50 25.46 2 WO SD 1.42 1.30 1.93 1275 CV IY 5S VAS 6.32 Lowest observed value as a % of highest observed value 84.9 86.4 80.4 83.9

Depth of mandible between 4th & 5th cheek teeth

Right Left 5.4 5.4 5.6 5.5 5.1 5.2 6.8 6.8 6.2 72 5.3 5.0 6.3 6.2 5.1-6.8 5.0-7.2 5.81 5.90 0.63 0.85 10.84 14.41 75.0 69.4

Table 2

Metric data on Metacheiromys.

Depth of dentary at

Length of lower cheek

mental foramen tooth row

M. marshi, type 6.1 3.8) M. tatusia, type 5.0 3.8 Lower value as a %

of higher value 82.0 Sy! M. dasypus, type 8.1 6.4 M. osborni, type Q.7/ 5.9 Lower value as a %

of higher value 83.5 O22

8 Metacheiromys and the Palaeanodonta

In the Bridger Basin, Wyoming, both M. marshi and M. dasypus are known from as low as the Bridger Bz (Matthew, 1909). The highest known stratigraphic occurrence of M. marshi is probably in the Bridger C, whereas M. dasypus is known as high as the Bridger Dz level.

A Review of the Known Palaeanodonta

In this section | review the published record of palaeanodonts and briefly discuss the sa- llent characters of each genus.

ORDER Palaeanodonta Matthew, 1918, p. 620

FAMILY Metacheiromyidae Wortman, 1903, p. 347

Included Genera

Propalaeanodon, Palaeanodon and Metacheiromys.

Discussion

The Metacheiromyidae appear to be a dis- tinct clade within the Palaeanodonta. {Character-state polarities can be assessed for palaeanodonts by making the reasonable assumption that Amelotabes (see below) is generally the most primitive known palaeanodont (Rose, 1978).] Early in the course of their evolution, the metacheiromyids reduced the postcanine teeth to small, cylindrical pegs, with blunt, rounded crowns that lacked enamel, and were separated by short diastemata. These character-states were converged upon by derived epoicotherlids. The metacheiromyids retained the symplesiomorphy of a relatively long dentary [which may have been covered by a horny plate (derived)] and they are united by the synapomorphous feature of lacking occlusion between the postcanines and thus they do not show occlusal wear on these teeth as do the epoicotheriids (Rose, 1979).

Postilla 192

SUBFAMILY Propalaeanodontinae, new Sole Included Genus

Propalaeanodon.

Diagnosis

Metacheiromyids with seven lower post- canines.

Discussion

The Metacheiromyinae, Palaeanodon and Metacheiromys, are united (relative to Propalaeanodon) by the further reduction in size and number of the postcanines and form a tight clade. Propalaeanodon is the plesiomorphous sister taxon of Palaeanodon and Metacheiromys and thus is here placed in the subfamily Propalaeanodontinae, of equal rank with the Metacheiromyinae.

Propalaeanodon Rose, 1979, p. 2 Type Species

Propalaeanodon schaffi Rose, 1979, p. 3. Included Species

Only the type species.

Distribution

Late Paleocene (late Tiffanian) of the Bighorn Basin, Wyoming.

Discussion

Propalaeanodon, the oldest and most primi- tive metacheiromyid, is distinguished by the possession of seven lower postcanines which are all single-rooted except for the fourth which is double-rooted. The crowns of the postcanines are higher and more pointed than in Palaeanodon and the posterior medial buttress of the mandible is prominent, but less so than in Palaeanodon.

9 Metacheiromys and the Palaeanodonta Postilla 192

Propalaeanodon is known only from the holotype dentary and possibly two left hu- meri which Rose (1979) tentatively referred to this genus. Except for differences in size, these humeri are similar morphologically to the humeri of Palaeanodon and Pentapassalus (Rose, 1979).

SUBFAMILY Metacheiromyinae Wortman, 1903, p. 347, new rank

Included Genera Palaeanodon and Metacheiromys. Discussion

The metacheiromyines are united by the fur- ther reduction in size and number of the postcanines (five in the dentary of Palaeanodon and two in the dentary of Metacheiromys) relative to Propalaeanodon.

Note on Authorship of the Subfamily

Under the present rules of the International Code of Zoological Nomenclature (Stoll et al., 1964), the first author of any family-group taxon (e.g., tribe, subfamily, family, super- family) is credited with authorship of all coordinate family-group taxa. Previously it was customary (and this is still often done) to credit authorship of a name to ‘‘he who first published it with exactly the spelling adopted in a given case” (Simpson, 1945,

p. 31; cf. Van Valen, 1966, p. 109). | find the present rules ‘frequently unjust and confus- ing’’ (Simpson, 1945, p. 31) and would prefer the use of Simpson's (1945) criterion of au- thorship cited above.

Palaeanodon Matthew, 1918, p. 621 Type Species

Palaeanodon ignavus Matthew, 1918, p. 622.

Included Species

The type species, P. parvulus Matthew, 1918, p. 640 and Palaeanodon sp. (Rose, TS7SMh9Z9y 181):

Distribution

Late Paleocene and early Eocene (Clarkforkian-Wasatchian) of the Bighorn Basin, Wyoming and possibly the Wasatchian of New Mexico (Simpson, 1931).

Discussion

Palaeanodon, well known from cranial and postcranial material (see especially Matthew, 1918; Rose, 1978) is distinguished by the fol- CaAkOA oleae canines large, oval in cross-section at the base and enamel-covered; postcanines small, peglike with rounded crowns and separated by short diastemata; posterior portion of jaw edentulous; large, posteriorly-placed medial buttress and dis- tinct internal mandibular groove present on the dentary; and bullae incomplete.

P. parvulus is distinguished from P. ignavus by its slightly smaller size (20—30% smaller).

lowing features: dental formula

Metacheiromys Wortman, 1903, p. 347 Discussion

This genus Is discussed in detail above

(pp. 2—8).

FAMILY Epoicotheriidae Simpson, 1927,

p. 285

Included Genera

Epoicotherium, Xenocranium, Tetrapassalus,

Pentapassalus, Amelotabes, Alocodontulum and Tubulodon.

10 Metacheiromys and the Palaeanodonta Postilla 192

Discussion

The genera which have been included in the Epoicotheriidae (listed above) appear to form a paraphyletic group which includes both primitive forms such as Amelotabes, that lack any shared and derived character-states for the family, and a few advanced forms which are united relative to the metacheiromyids and plesiomorphous epoicotheriids on the basis of synapomor- phies of the skull as the Epoicotheriinae (see below). Thus, as used by previous authors (e.g., Simpson, 1927; Colbert, 1942; Rose, 1978) the Epoicotheriidae has been diag- nosed on the basis of shared primitive character-states, especially of the lower jaw (Rose, 1978, p. 669). Here | provisionally re- tain the Epoicotheriidae as a family for all of these forms; but | separate out Xenocranium, Epoicotherium, Pentapassalus and possibly Tetrapassalus as the Epoicotheriinae. The remaining genera, Amelotabes, Alocodontulum, and Tubulodon, are presently too poorly known to adequately assess their phylogenetic relationships and are here regarded as Epoicotheriidae incer- tae sedis.

SUBFAMILY Epoicotheriinae Simpson, 1927, p. 285, new rank

Included Genera

Pentapassalus, Epoicotherium, Xenocranium and possibly Tetrapassalus.

Discussion

The epoicotheriines are united as a clade by the following synapomorphies: extremely short and posteriorly broad skulls that are triangular in dorsal and ventral view with high and transversely wide occiputs, inflated braincases, and medium to large, ossified tympanic bullae. Epoicotherium, Xenotherium, and possibly Tetrapassalus,

are further united by the derived character-

state of fusion of cervical vertebrae two through five.

Pentapassalus Gazin, 1952, p. 32 Type Species

Pentapassalus pearci Gazin, 1952, p. 32. Included Species

The type species and Pentapassulus woodi (Guthrie, 1967, p. 23).

Distribution

Early Eocene (Wasatchian), La Barge fauna, ‘Upper Knight beds,’ Sublette County, Wyoming (Gazin, 1952) and early Eocene, Lysite Member of the Wind River Formation, Fremont County, Wyoming (Guthrie, 1967).

Discussion

Gazin (1952) distinguished Pentapassalus by the following suite of features: dental for-

? mula aad teeth similar to Tubulodon (al-

though note that in the two holotypes and sole known specimens of Pentapassalus the crown morphology is worn off) except that M, alone of the lower teeth is two-rooted or bilobed; P, and Mz possess a single, conical root; skull short and broad (similar to Epoicotherium); bullae ossified, well inflated and posterior in position; and pterygoids not continuous with bullae. Pentapassalus wood! (originally described as a species of Palaeanodon, but referred to Pentapassalus by Rose, 1978) is distinguished from P. pearci by its slightly larger size (10-25% larger in most linear dimensions of the post- crania). The cheek tooth rows of P. pearci and P. woodi are the same length, but the canines in P. woodi are twice the size of those in P. pearci. As Rose (1978) noted, this is suggestive that P. pearci and P. woodi are the female and male respectively of a single species.

11 Metacheiromys and the Palaeanodonta Postilla 192

The holotypes of the two species of Pentapassalus include skulls and partial

skeletons (Gazin, 1952; Guthrie, 1967; Emry,

1970; Rose, 1978, 1979). The postcranial skeleton of Pentapassulus |s extremely simi- lar in morphology to that of the metacheiromyids.

Epoicotherium Simpson, 1927, p. 285 Synonym

Xenotherium Douglass, 1905, p. 204 (pre- occupied by Xenotherium Ameghino, 1904, p. 114)

Type Species

Epoicotherium unicum (Douglass, 1905, p. 204).

Included Species

The type species and possibly Epoicotherium sp. of Heissig (1982).

Distribution

Early Oligocene (medial Chadronian), ‘“‘McCarty’s Mountain,’’ Madison County, western Montana and the White River Group in the Flagstaff Rim area of central Wyoming (Emry, 1970, 1973; Rose and Emry, 1983); possibly the early Oligocene near Pap- penheim, West Germany (Heissig, 1982).

Discussion

For over 75 years only a single, nearly eden- tulous skull, the holotype of Epoicotherium unicum, had been described (Douglass, 1905; Simpson, 1927). Heissig (1982) has recently described an edentulous left mandi- ble with five alveoli from the early Oligocene of West Germany which he assigned to ‘Epoicotherium sp.'’ Rose and Emry (1983) subsequently described new material of Epoicotherium from North America including a complete skull, partial left dentary, miscel-

laneous vertebrae and the nearly complete forelimb. Comparing Heissig’s (1982) de- scription and illustration of the German specimen to Rose and Emry’s (1983) de- scription of the dentary of Epoicotherium confirms that Heissig may be correct in his identification, and thus this may represent the first known non-North American palaeanodont.

Epoicotherium bears an upper canine, five upper postcanines, one lower incisor, a lower canine, and five lower postcanines (Rose and Emry, 1983). Simpson (1927,

p. 285) distinguished Epoicotherium by the following features: ‘‘depressed snout, domed occiput, slender but complete zygomata without sub- and post-orbital pro- cesses, large completely ossified tympanic bullae, with which are ankylosed the pterygoid plates, there being no hamular processes, and cylindrical, one-rooted cheek-teeth without enamel.’’ Rose and Emry (1983) have since demonstrated that Epoicotherium actually bore a very thin enamel layer on at least the canines and postcanines. The axis and following three cervical vertebrae are fused in Epoicotherium and throughout the skeleton bears numerous subterranean, fossorial adaptations (Rose and Emry, 1983).

Xenocranium Colbert, 1942, p. 3

Type Species

Xenocranium pileorivale Colbert, 1942, p. 3. Included Species

Only the type species.

Distribution

Early Oligocene (Chadronian-?Orellan), Brule and White River Formations, White River

Group, Wyoming and Nebraska (Rose, 1978; Rose and Emry, 1983).

12 Metacheiromys and the Palaeanodonta

Discussion

Xenocranium, originally described from a single skull and mandible, was distinguished by Colbert (1942, p. 3) by the following fea- tures: ‘‘much larger than Epoicotherium, with a depressed snout and an elevated, greatly broadened occiput; orbit without postorbital processes; bullae transversely elongated and prominent below level of the basicranium; very large, bulbous expansions of the squamosal external to the bullae and at the posterior termination of the zygomatic arch; mandible shallow, with a broad, inter- nal shelf; teeth simple and peg-like, without enamel; five cheek teeth in the maxilla and six in the mandible, of which the anterior (canine) In each jaw is much larger than the rest of the teeth.”

Rose and Emry (1983) have since rede- scribed the skull and mandible, and have de- scribed the anterior half of the axial skeleton and most of the forelimb of Xenocranium. Rose and Emry (1983) demonstrated that the

dental formula of Xenocranium is oe (i.e, there |s actually a small incisor anterior to the canine in the lower jaw). The snout of Xenocranium (missing in the holotype) |s flat- tened and expanded anteriorly. As in Epoicotherium, and possibly Tetrapassalus (see below and Fig. 11), cervical vertebrae two through five are fused in Xenocranium and the skeleton is more highly specialized for subterranean, fossorial habits than in any other known palaeanodont (Rose, 1978; Rose and Emry, 1983).

Tetrapassalus Simpson, 1959, p. 2 Type Species

Tetrapassalus mckennai Simpson, 1959, jon 2.

Included Species

The type species and TJetrapassalus proius West; 1973, p. 929.

Postilla 192

Distribution

Middle Eocene (Bridgerian) of the Bridger Formation, Bridger Basin, Wyoming.

Discussion

Tetrapassalus, described only from the two holotype partial lower dentitions and dentary fragments, is distinguished by the following features: lower canines large; cheek teeth consisting of four subequal, cylindrical, peg- like postcanines without closed roots or root division, but showing distinct occlusal wear; and posterior medial buttress slight. T. proius iS distinguished from T. mckennai by its smaller size (approximately two-thirds the size of T. mckennai).

A maxilla and two fused cervical verte- brae series are tentatively referable to Tetrapassalus and are discussed below.

cf. Tetrapassalus sp. A Synonymy

cf. Tetrapassalus sp.: Rose, 1978, p. 672, text-fig. 5

Distribution

Middle Eocene (early Uintan) of the Tepee Trail Formation, northwestern Wyoming.

Discussion

Rose (1978) described and illustrated a right maxilla (AMNH 10215) of a small epoicotheriid which may represent Tetrapassalus, or perhaps more likely, a new genus and species. AMNH 10215 bears a large, pointed canine and four small, single- rooted postcanines (although the second Is missing from the specimen) with enamel- covered crowns, except for the occlusal sur- faces (Rose, 1978).

13 Metacheiromys and the Palaeanodonta

Postilla 192

cf. Tetrapassalus sp. B (Fig. 11)

Synonymy

?Order Edentata: Robinson, 1963, p. 6, un- numbered fig.

Epoicotheriid: West, 1973, p. 931

“fused cervical vertebrae:'’ McKenna, 1975, p. 29

“Tetrapassalus or a. closely related form:" Rose and Emry, 1983, p. 42

Referred Specimens

YPM 14919, the fused centra of vertebrae two through five (Fig. 11): collected by L. LaMothe from middle Eocene-aged strata, probably upper Bridger beds, of the Bridger Formation at Lone Tree, Wyoming in 1873.

YPM 16872, a specimen identical to YPM 14919 but preserving only the centrum of the axis and part of the centrum of vertebra three (this specimen is listed by Robinson, 1963, but | have not been able to locate it in the YPM collections): collected by J. Heisey from middle Eocene-aged strata, upper Bridger beds of the Bridger Formation at Lone Tree, Henry's Fork, Bridger Basin, Wyoming, on 20 June 1874.

Discussion

The suggestions by Robinson (1963) and West (1973) that YPM 14919 and YPM 16872 represent the fused cervical vertebrae of an epoicotheriid are corroborated by the recent description of fused cervical vertebrae in Epoicotherium and Xenocranium by Rose and Emry (1983, p. 42) who refer YPM 14919 and YPM 16872 to ‘'Tetrapassalus or a closely related form’ on the basis of size. The cervicals are unfused in the metacheiromyids (e.g., see Fig. 11, J; the axis of Metacheiromys marsh).

Epoicotheriidae Incertae Sedis Amelotabes Rose, 1978, p. 659

Type Species

Amelotabes simpsoni Rose, 1978, p. 660. Included Species

Only the type species.

Distribution

Late Paleocene (Tiffanian) of the Bighorn Basin, Wyoming.

Discussion

Amelotabes simpsoni, known from a single dentary and partial lower dentition, |s distin- guished by the following features: mandible shallow, robust with medial buttress and internal mandibular groove; dental formula Ue 2 ae P5-M, double-rooted; cheek teeth with thin enamel on occlusal surfaces; Pz, with large protoconids, small, low paraconids and slight protocristids; P3 with small talonid heel; mo- lars with primitive eutherian cusp pattern; molars low crowned with bulbous and poorly defined cusps; protoconids largest and high- est cusps on the molars; and hypoconulids, cristids obliquae and cingulids absent.

; canines large; P, single-rooted;

Alocodontulum Rose, Bown and Simons, 1978, p. 1162

Synonym

Alocodon Rose, Bown and Simons, 1977, p. 1 (preoccupied by Alocodon Thulborn, 1973, p. 93)

Type Species

Alocodontulum atopum (Rose, Bown and Simons, 1977, p. 2).

Included Species

Only the type species.

14 Metacheiromys and the Palaeanodonta

Distribution

Early Eocene, upper ‘’Gray Bull’ beds, lower Willwood Formation, Bighorn Basin, Wyoming.

Discussion

Alocodontulum is known only from the genoholotype, an incomplete premaxilla- maxilla and partial upper dentition. Rose, Bown and Simons (1977, p. 2) diagnosed this taxon as follows: ‘‘Upper molars longer than wide, and uniquely specialized; M' with median longitudinal furrow, several cusps arranged In line on lingual and buccal borders of crown, and no enamel on top of crown. M' slightly larger than M?, both much larger than M8; M8 greatly reduced, single- rooted, with very low and bulbous crown. P? premolariform, with large paracone, rudimentary metacone, small low protocone. P' tiny, Single-rooted. Canine of moderate size, triangular in section.”

Rose, Bown and Simons (1977) assigned Alocodontulum to the Epoicotheriidae primar- ily on the basis of the following similarities that it shows to either Pentapassalus or Tubulodon or to both: enamel reduction on the teeth (occlusal surfaces of the cheek teeth lack enamel); similar scalloped margins on the molars of Alocodontulum and Tubulo- don; upper molars longer than wide; M$ peg- like; and upper canine triangular in cross sec- tion with the anteromedial face honed and lacking enamel.

Tubulodon Jepsen, 1932, p. 264 Type Species

Tubulodon taylori Jepsen, 1932, p. 264. Included Species

Only the type species.

Postilla 192

Distribution

Late early Eocene (Lostcabinian) of the Lost Cabin Member, Wind River Formation, Wind River Basin, Wyoming.

Discussion

Originally thought by Jepsen (1932) to be re- lated to the extant aardvark (Orycteropus, Tubulidentata), Tubulodon, known only from partial lower dentitions (Jepsen, 1932; Guthrie, 1971) is distinguished by the follow- ing features: cheek teeth similar to those of Amelotabes with low and marginally situated cusps; three lower molars present (P, only known premolar); known teeth double- rooted with hypsodont, columnar crowns; cheek teeth with thin enamel on the sides and lacking enamel on the occlusal surfaces. The genoholotype of Tubulodon has micro- scopic tubules present in the teeth and Jepsen (1932) considered these tubules to indicate a relationship to the Tubulidentata. Colbert (1941) argued that the tubules seen in Tubulodon are unlike those seen in Orycteropus and allies, whereas both Gazin (1952) and Rose, Bown and Simons (1977) have observed similar tubules in the teeth of other Eocene mammals. Rose, Bown and Simons (1977; cf. Peyer, 1968) note that dentine tubules are present in the teeth of most mammals and these authors suggest that tubules may merely be more readily seen in some fossil mammals in which the enamel is thin (e.g., Tubulodon, Alocodontulum and Pentapassalus).

15 Metacheiromys and the Palaeanodonta_ Postilla 192

Conclusions

Fourteen species, ten genera, three subfami- lies and two families are formally recognized within the order Palaeanodonta (Table 3). The earliest known palaeanodonts are Amelotabes and Propalaeanodon, both from the Tiffanian of the Bighorn Basin, Wyoming. Whereas Amelotabes appears to be the most primitive known palaeanodont, and may thus approximate the ancestral palaeanodont morphotype, Propalaeanodon bears derived metacheiromyid features. The latest and most derived palaeanodonts are Epoicotherium and Xenotherium from the early Oligocene. Traditionally the Palaeanodonta have been regarded as a wholly North American group; however, the recent referral of a specimen from the early Oligocene of West Germany to Epoicotherium sp. (Heissig, 1982) suggests that further collecting may extend the geo- graphic and stratigraphic range of this poorly known mammalian order.

Acknowledgments

| thank John H. Ostrom (YPM) for allowing me to study specimens in his care and Mary Ann Turner (YPM) for help in locating speci- mens and locality information. | thank John H. Ostrom and Bruce H. Tiffney for critically reviewing the final manuscript. | wish to ex- press my appreciation to Cynthia B. Pettit, John W. Pendleton and Deborah A. Shea for their encouragement while | was working on this paper during the spring and summer of 1982.

Table 3

A classification of the Palaeanodonta.

Order Palaeanodonta Family Metacheiromyidae Subfamily Propalaeanodontinae Propalaeanodon (1 species) Subfamily Metacheiromyinae Palaeanodon (2 species) Metacheiromys (2 species) Family Epoicotherlidae Subfamily Epoicotheriinae Pentapassalus (2 species) Epoicotherium (1 species) Xenocranium (1 species) Tetrapassalus (2 species) Epoicotheriidae incertae sedis Amelotabes (1 species) Alocodontulum (1 species) Tubulodon (1 species)

16 Metacheiromys and the Palaeanodonta Postilla 192

Literature Cited

Ameghino, F. 1904. Nuevas especies de mamiferos, cretaceos y terciarios de la Republica Argentina. An. Soc. Cien. Argentina (Buenos Aires) 56—58:1—142.

Colbert, E. H. 1941. A study of Orycteropus gaudryi from the