328 E. L. Troxell — American Bothriodonts. 



little resemblance to Anthracotherium magnum Cuvier, 

 in form or size. Briefly its distinctive features are as 

 follows: absence of cingulum except on the anterior 

 side ; the bases of the proto- and hypocone equal ; these 

 two cones joined by two ridges inclosing a deep short 

 segment of the transverse groove; the proto-conule 

 weak; the para- and metacones most prominent of all 

 the moderate cones ; ridges from them joining one from 

 the mesostyle to form a Y as in Octacodon; the para- 

 style very prominent, with a sharp cusp, and widely 

 separated from the mesostyle; the metastyle almost 

 obsolete. The dimensions are: antero-posterior, 19.5 

 mm. on outer border, and transverse, 21.5 mm. on anterior 

 side. 



Octacodon Marsh. 



This genus, having for its genoholotype 0. valens 

 Marsh, should also include, it seems, Heptacodon gibhi- 

 ceps Marsh and Anthracotherium harense Osborn and 

 Wortman,^ for reasons set forth on the following pages. 

 All are of the Protoceras beds and all conform in impor- 

 tant details to the genoholotype. 



Octacodon valens Marsh. 



(Fig. 2.) 



Holotype, Cat. No. 11860, Y. P. M. Upper Oligocene (Protoceras beds), 

 Hermosa, South Dakota. 



The type specimen consists of the third upper molar 

 (see ^g. 2) and an incisor, and was described by Marsh as 

 follows :^ 



' ' The tooth represented . . . may be regarded as the type of the 

 present genus and species. It is the last upper molar of the right 

 side, and is in fine preservation. The slight wear of the tooth 

 shows that the animal was adult. There are five main cusps in 

 the crown, two on the posterior half, and three on the anterior, the 

 antero-median cusp being the smallest. On the outer margin of 

 the tooth are three prominent buttresses with conical summits, 

 making in all eight prominences on the crown, which feature has 

 suggested the generic name. 



^ H. F. Osborn and J. L. Wortman, Bull. Am. Mus. Nat. Hist., vol. 6, 221- 

 223, 1894. The referred specimen of these authors, A. cwr^um (Marsh), No. 

 1039 American Museum Collection, can not be considered the same as Sep- 

 tacodon curtis Marsh (see Heptacodon above) , but should in all probability 

 be placed with Octacodon, cf. 0. gihMceps, in spite of its earlier age. 



' O. C. Marsh, This Journal (3), 48, 92, 1894. 



