ANTHONY, FOSSIL MAMMALS FROM PORTO RICO 303 



and the posterior portion of the ramus is expanded and deep with a con- 

 cave border. 



Dentition (Plate XI, Figs. 2-3).— The teeth of Heteropsomys are of a 

 highly specialized type. They are considerably worn and the primary 

 pattern is, on this acconnt, lost, but the cro^m surfaces are none the less 

 distinctive. The incisors are small and of the ordinary curved rodent 

 type. The molars, four in number in each jaw, are all nearly equal in 

 size and have a prominent median lateral indentation or infolding- down 

 the side of the tooth, the fold being on the inside of the tooth above, on 

 the outside of the tooth in the lower jaw. Within the cro"^vn surface of 

 each tooth and more or less isolated from one another by the dentine are 

 little hollow, flattened <tubes of enamel; three in number in the upper 

 molars (only pm* and m^ present to be examined), two in the lower 

 molars. The molars are two rooted, the roots being short truncated cones 

 in appearance. 



Measurements: Total length, 69 mm. ; zygomatic width (approxi- 

 mate), 40 mm. ; width of brain case, back of zygomatic root of squamosal, 

 27 mm.; interorbital width, anterior to postorbital processes, 18.5 mm.; 

 alveolar length of maxillary tooth row, 14.5 mm.; length of diastema, 17 

 mm. ; dimensions of m^, 3.5 X 4 mm. ; greatest length mandible, without 

 incisor, 44 mm.; alveolar length of mandibular molar series, 15.5 mm. 



EEMAEKS 



Heteropsomys apparently requires no lengthy comparison Avith any 

 knoAvn genus. Its distinctness when compared with any of the mainland 

 hystricomorphs is immediately evident, and because of this fact it is not 

 advisable in this paper to attempt to place it in any particular family. 

 A possible later discovery of more material® in Porto Eico and a better 

 understanding than that prevailing now of the relationships of the di- 

 verse families of the hystricomorphs is necessary in the case of Heterop- 

 somys as well as in that of Elasmodontom,ys. 



This rodent is doubtless of a later age than either Acratocnus or Elas- 

 m.odontom.ys, judging from its position in the cave deposits and from the 

 appearance of the bone itself. While one may not be justified in consider- 

 ing it a contemporary of Isololodon, considered by Allen" as exterminated 

 by the natives in recent times, at least it could scarcely have been earlier 

 than late post-Pleistocene. 



^ Specimens with unworn teeth may reveal points on the evolution of the molars now 

 impossible to surmise with accuracy. 

 8 AlleNj I. c, p. 22. 



