626 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxvi. 



related with but not ancestral to the Diprotodont division of the 

 Marsiipialia, with which division they may be now classed as an 

 Infraorder, or Superfamily, their relationship dating back to a 

 common ancestry somewhere in the Jurassic or even to earlier Tri- 

 assic times, as was suggested by Cope. 



Regarding the probable character of the food, upon which Ptilodus 

 and its related genera subsisted, the sj^ecimen from Montana described 

 above seems to throw considerable light. An examination of the 

 type of P. gracilis (PI. 70) shows Cope's" statement, that the grind- 

 ing teeth are " weak in structure," is incorrect, and his supposition 

 that it was necessary for the animal to swallow its food without 

 mastication is not admissible. On the contrary it will be seen that 

 the grinding area is comparatively very considerable, occupying 

 nearly three-fourths of the actual contact space between the upper 

 and lower dental series, and although the tooth-crowns are low they 

 are relatively broad and massive. To add to their efficiency the 

 molars are well supplied with an array of short stout tubercles, well 

 adapted to crushing and grinding small hard substances but very 

 poorly adapted to cutting or masticating meat. In specimens of old 

 individuals the much Avorn condition of the tubercles of the molars 

 as compared with that of the lower cutting-premolar suggests that 

 the latter may have been used for the purpose of cutting only soft 

 materials, such as the skin and pulp of fruit, while the molars Avere 

 employed in grinding harder substances, such as seeds. 



The evidence that Ptilodus and Phujiaidax were not carnivorous 

 in habits seems rather conclusive, but as to whether they were in- 

 sectivorous, herbivorous, or frugivorous there may still be some 

 differences of opinion. I am inclined to consider them as frugivorous, 

 since the incisors were well fitted for picking small fruits or berries, 

 while the large cutting blades of the lower premolars were admirably 

 adapted to cutting or sliceing the rinds of tough-skinned berries, or 

 to chopping up fleshy fruits held against the blunt-pointed premolars 

 of the ui)per jaw. For masticating the seeds of such small fruits 

 and berries the multituljerculate molars were amply sufficient. 



EXI'LANATION OF PLATE 70. 

 Ptilodus (jracili-i (JuUey. 

 (Typc-f^pccimcn.— Cat. No. GOTO, TT.S.N.M. All figures twice natural size.) 

 Fig. fl. Sliull with lower jaw in position, side view. 

 1). Right ramus of lower jaw, inside view. 



c. Lower jaws in normal position, viewed from above. 



d. Skull, top view. 



e. Sliull, palate view. 



/. Distal portion of left humerus, posterior view. 



"Cope, Tertiary Vertebrata, U. «. (j!oo1. Surv. Terr., Ill, 1S84, p. 170. 



