E. L. Troxell — American Bothriodonts. 337 



absent, a long diastema separates C^ from P^, the pos- 

 terior nares are far forward, and the skull is narrow. 



^pinacodon deflectus (Marsh). 



(Figs. 6, 7.) 



Holotype, Cat. No. 11802, Y. P. M. Oligocene ( Titanotherium beds), 

 near Deadwood, South Dakota. 



This species,^ ^ made by Marsh in 1890,^^ can now be 



Fig. 7. — ^pinacodon deflectus (Marsh). Holotype. Cat. No. 11802, 

 Y. P. M. A, erown view of lower molars and premolars. B, top and side 

 view of lower canine. X l/^- 



more clearly defined. As contrasted with A. americanus 

 (Leidy), one sees on M^ that the transverse groove ends 

 in a square, is not pointed between the para- and meta- 

 cones, and the edge is crenulated or cleft deeply; the 

 cingulum on the posterior side of the hypocone is faint 

 or absent, while the main posterior cingulum is continuous 

 with the ridge from its apex; an internal basal cusp or 

 cingulum appears on the hypocone and strong ridges also 

 from the protocone descend into the valley. The proto- 

 cone is not conical but is selenodont due to its heavy 

 ridges ; the protoconule is near the protocone. 



Carrying this contrast with A. americanus to the second 

 molar, we find in A. deflectus the very strong internal 

 basal cingulum extending quite around the hypocone ; on 

 the right M^ the protocone and hypocone are joined by a 

 strong ridge and on this tooth there is a wide groove 

 between the meso- and metastyles. P"^ is a heavy tooth 

 with strong cingula and styles; the cingulum, which is 

 posterior, leads around the deuterocone and not onto it, 

 leaving the latter truly a cone, round in cross-section. 



" From the meagre description of A. rostratus (Scott) ( Journ. Acad. Nat. 

 Sci., Phila., 9, appendix, p. 536, 1894) , these distinguishing features may be 

 noted: P* of this species has a cingulum upon the innej* side of the crown 

 and the buttresses or styles are less conspicuous. The molar length is equal 

 to that of A. deflectus, 71 or 72 mm. A close comparison of the types of 

 these two species is not practicable at this time, but additional points should 

 be had of A. rostratus in order to separate it clearly from A. delectus. 



^O. C. Marsh, This Journal (3), 39, 525, 1890. 



