Amphidon, AND

ROUGIER, GUILLERMO W., SPURLIN, BARTON K. & KIK, PETER K., 2003, A New Specimen of Eurylambda aequicrurius and Considerations on ‘‘ Symmetrodont’ ’ Dentition and Relationships, American Museum Novitates 3398, pp. 1-15 : 11

AMPHIDON AND ‘‘SYMMETRODONTS’’

Simpson (1929) considered the possibility that the type of Eurylambda (YPM 13639) was, in fact, an upper molar of Amphidon , instead of Tinodon , and was his original attribution. These caveats were his main reasons for erecting a new genus for YPM 13639. Crompton and Jenkins (1967), however, determined that the wear facets of Eurylambda and Amphidon did not match, rendering Simpson’s 1929 attribution unlikely.

The type and only known specimen of Amphidon superstes seems to be better interpreted as an amphilestid with worn­down molars rather than a typical ‘‘symmetrodont’’ ( Rougier et al., 2001). There are five molariforms preserved in the type and only specimen (YPM 13638) traditionally interpreted as p–last m1–4. The first preserved tooth shows little wear in contrast with the heavily worn posterior molariforms, indicating that this element had a deciduous predecessor and that it erupted later than the more distal teeth. A deciduous predecessor would make this tooth, by definition, a premolar (Clemens and Lillegraven, 1986; Luckett, 1993). The morphology of this tooth, however, agrees closely with that of the first molariform of Amphilestes and Phascolotherium in showing five cusps that are fairly symmetrically arranged, as well as a faint basal cingulum (the smaller cusp b is damaged). Replacement of molariforms is known in the putative amphilestid gobiconodonts Gobiconodon and Hangjininia (Jenkins and Schaff, 1988; Godefroit and Guo, 1999); additionally, the tooth identified as m 1 in zhangheotheriids has a deciduous predecessor (personal obs). Therefore, the dentition preserved in Amphidon can be interpreted as m1–m5, which agrees well with the almost universal presence of five molariforms among amphilestids. The somewhat triangular aspect of the crown of Amphidon is, in our opinion, caused by the extreme wear of cusps b and c, and by the labial bulging of the base of cusp a. A few other mammals have been included in Amphidontidae ( Yabe and Shikama, 1938; Trofimov, 1980; Yadagiri, 1985; Krusat, 1989), with Manchurodon ( Yabe and Shikama, 1938) and Gobiotheriodon ( Trofimov, 1980, 1997) based on the most complete specimens. These materials, however, do not shed additional light on the morphology of the Amphidontidae , if there is such a group at all. Manchurodon has been lost since the brief original description. It is unclear if the dental morphology described really corresponds to a full buccal view or simply to the exposed labial surface of the molariforms. In the published source, there is very little to link Manchurodon with Amphilestes . The affinities of Manchurodon are at present better left unresolved.

Gobiotheriodon ( Trofimov, 1980, 1997) is known by a fairly complete lower jaw with three molariforms and an attributed upper tooth in a fragment of maxilla. We had the opportunity to study this specimen at the Paleontological Institute, Moscow. Gobiotheriodon is indeed a symmetrodont, but it resembles more closely the recently discovered Zhangheotherium from Liaoning (Hu et al., 1997, 1998) than Amphidon . As in Zhangheotherium , the lower jaw is very slender with multiple mental foramina (four or more), and the molar cusps have their bases separated, forming individual cones. The presence of individualized cusps contrasts with the condition in spalacotheriids where the main molariform cusps are united by crests at the time of eruption. In zhangheotheriids, and presumably also in Gobiotheriodon , wear removes substantial parts of the crown to obtain matching surfaces between upper and lower molariforms. We think, therefore, that Amphidontidae as a distinct family of Mesozoic mammals is probably unwarranted, and that it is based on an artificial grouping of badly preserved and poorly known fossils.

In addition to Amphidon , there are two other amphilestids described from the Morrison Formation, Phascolodon ( Simpson, 1925a) and Aploconodon ( Simpson, 1925a) . Both of these are known by the type specimens only, in both cases partial lower dentitions. We cannot rule out that Eurylambda represents the uppers of an amphilestid, al­ though the poor matching of wear facets between Amphidon and Eurylambda makes it unlikely (Crompton and Jenkins, 1967). Based on the orientation of the wear facets on these amphilestids and their crown outline we would expect the upper molariforms to be similar to those of Gobiconodon , with the main cusps forming a broad triangle, the teeth somewhat broad buccolingually, and the parastylar hook absent. Thus, we support the traditional view that Eurylambda is likely the upper dentition of Tinodon .