no 



MR. R. I. POCOCK ON 



smaller and the crowns of the cheek-teeth are higher with relation 

 to their breadth in G. antarcticus than in C. latrans. But apart 

 from these, the principal differences mentioned above are quite 

 sufficient to disprove the claim that the two species are closely 

 related. According to modern standards of classification they are 

 subgenericall}^, if not generically, distinct. 



But the characters above described tell us more than that. 

 Taking G. latrans first, it is obvious that in the carinif orm sagittal 

 crest, the angularly produced occipital crest, the position of the 

 masseteric ridge on the malar bone, and in the points alluded to 

 in connection with the upper and lower carnassials, the species 

 falls into line with the large wolves like G. occidentalis and lujnis, 



Text-fi2'. 74. 



A. Vertical view of upper carnassial of Canis latrans. 



B. Vertical view of upper carnassial of C. antarcticus. 



C. Internal view of lower carnassial of C. latrans. 



D. Internal view of lower carnassial of C. antarcticus. 



and with G. jxdlipes and G. lupaster, which, a.ccording to fancy, 

 may be called large jackals or small wolves. These resemblances 

 explain Mivart's dismissal of the cranial and dental chai'acters of 

 G. latrans with the remark, " The skull possesses no distinctive 

 characters, nor have we been able co detect any in the shape of 

 the teeth." 



On the other hand, the skull of C. antarcticus, with its lyriform 

 sagittal area and truncated occipital crest, agrees in the main with 

 the skulls of certain species or subspecies of South- American dogs 

 in the British Museum labelled G. thous ( = cancrivorus), riidis, 

 sclaieri [=7mcrotis), parvidens, lorosticttts, gracilis, and fulvipes. 

 And in the skull of a dog, perhaps referable to G. gracilis, which 



