Mekriam.] 



Pliocene and Quaternary Canidae. 



281 



restoration is not justified, as the outer portion of the heel was 

 unbroken. On fig. 4 the writer has indicated in an unbroken 

 line a third suggestion as to probable form, beginning the rest- 

 oration at the point a where the break in the heel occurs. This 

 shows the tooth to have a form somewhat similar to that of Pt 

 in Hyaenognathus , although not so broad. Cope classified 

 Borophagus on the assumption that it had four inferior premolars, 

 though he suggests doubt of this at one point in his description. 

 This is, however, improbable, as the presence of a Pt 

 corresponding in size to what he considered as the large P3 

 and accompanied by a heavy sectorial would mean the elitnina 

 tion of Pt, just as has occurred in Hyaenognathus and in the 

 hyaenas. 



3 4 



Figs, land 3. Hyaenognathus pachyodon, ■ f. c, alveolus of canine. 

 Figs. 2, 4 and 5. Borophagus diversidens Cope, X f. c, alveolus of 

 canine, a, point on P T where fracture begins. (After Cope.) 



Hyaenognathus is evidently allied to Borophagus and it is not 

 impossible that future investigation may show generic identity. 

 The differences in the minor details of form in the premolars 

 indicate, however, that the types are specifically quite widely 

 separated, Hyaenognathus being the more specialized. Several 



