Classification of existing Felidse. 347 



H, yaguarondi and H. geoffroyi are widely divergent 

 species in cranial characters, the former departing the 

 most, the latter the least, from the typical feline type. In 

 H. yaguarondi the anterior nares are subvertical ; the post- 

 orbital processes of the frontal are short and depressed, but 

 the malar process is longish ; the zygomata are weakly 

 arched, with the suborbital portion less salient above the 

 carnassial and the squamosal root very long from before 

 backwards, with its posterior border strongly inclined 

 forwards and outwards ; the valley between the bulla and 

 the glenoid ridge is wider; the occipital area is narrow 

 across the mastoids, and the crest is less strongly inclined, 

 backwards and inwards ; the temporal crests always form a 

 wide lyriform area. 



On account of the difiFerences in the skulls the two species 

 might be regarded as generically or subgenerically distinct, 

 but the difficulties of definition would in that case be great 

 on account of H. 2Jurdinoides and its allies occupying in 

 many cranial respects an intermediate position between the 

 two, as Thomas pointed out in 1903. 



The names yaguarondi and eyra were given, I believe, to 

 colour-phases of one and the same species, the former being 

 blackish with speckled hair, the latter red. Neither form 

 shows pattern in the adult. If newly born kittens show 

 pattern at all, which by the analogy of Lynx caracal is 

 uncertain, it may be predicted that it will be like that of 

 H. purdinoides, guigna, or geoffroyi. It may be added that the 

 skull of H. yaguarondi differs widely from that of Puyna 

 concolor and completely negatives the claim of relation- 

 ship that has been made between the two species on account 

 of the absence of pattern in the adults. 



Some of the smaller spotted species of the pardinoides and 

 tigrina types of this genus, with the skull smooth and 

 rounded, recall on superficial inspection the smaller species 

 of Prionailurus like bengalensis, more particularly in the 

 large size of the orbit, the shortness of the muzzle, expanded 

 maxilla, and the vertical truncature of the anterior nares. 

 The two, however, differ as follows : — 



The P. benffcilensis-ty^e. 



Postorbital processes long, often 

 confluent. 



Preorbital constriction narrowest 

 midway along orbit, usually well 

 behind maxilla and nasals. 



Masseteric ridge advancing to 

 middle of malar, i of malar. 



23* 



The H. pardinoides-ty^e, 



Postorbital processes short and 

 subspiniform. 



Pi-eorbital constriction narrowest 

 in front of centre of orbit and close 

 behind maxilla and nasals. 



Masseteric ridge near lower edge 



