1899-] Wortman and Matthew, Ancestry' of the Canidce, etc. I 25 



large, not extended transversely, subquadrate ; (4) lower carnas- 

 sial broad-heeled, and with an accessory cusp at the postero-ex- 

 ternal corner of the trigon ; (5) canines slender ; (6) otic bullce 

 large. 



Canis iii'ostictus Mivart, and C. parvidens Mivart, both from 

 South America, belong to this genus, to which we also refer pro- 

 visionally the three John Day species, latidens Cope, lemur Cope, 

 3.r\d geisi)iaria?ius Cope. These latter seem to be directly ances- 

 tral forms, and, like all the earlier Dogs, have shorter feet, longer 

 lumbar region, and smaller brain-case than the modern species. 



The distinctions above noted may seem hardly to be of generic 

 value ; they are, however, tolerably constant, Urocyon cinereo- 

 argenteus being the only intermediate type. And if the group 

 was first separated in the Oligocene, as seems probable, it gives 

 it an ancestry that deserves full generic recognition. The alter- 

 native to uniting the John Day and modern species is to place 

 the one as a subgenus of Cynodictis, the other of Canis j the pres- 

 ent method seems, however, to recognize more clearly the actual 

 genetic affinities ; for if the living Canis parvidens and urostictus 

 are lineally descended from the John Day Cynodictis latidens 

 letnur, and geismarian7is, the two modern species are more nearly 

 related to the three Miocene species than to any living Canidae. 

 The best way to express this fact is to remove the Miocene species 

 from Cynodictis, to the typical forms of which they are not more 

 nearly related than the half-dozen or more distinct genera of the 

 Phosphorites of France, and to unite them with two modern 

 species, disregarding the considerable modernization of the latter, 

 which retain, however, an unusual amount of the Tertiary facies. 



Nothocyon urostictus {Mivart). 



Canis urostictus MlVART, Proc. Zool. Soc. Lond. 1890, 112; Monograph 

 of the Canidse, 81. 



The type and only specimen hitherto known is in the British 

 Museum. No. 391, Dept. Osteol, Amer. Mus. Nat. Hist., we 

 refer to this species and figure here. In size it is not very dif- 

 ferent from Canis azarce, but the size and characters of the 

 molars and carnassial easily distinguish it. The lyrate area on 

 top of the skull is shared by many species of Dogs. Limbs and 

 feet about as in modern species of Canis ; the lumbars appear to 



