240 PROCEEDINGS OF THE ACADEMY OF [1889. 



These comparisons show that in the Nimrctvidce, and more es- 

 pecially in Dinictis, there is a strange association of both feline 

 and viverrine features, together with others, which are more primi- 

 tive than those which occur in the living representatives of either 

 of these families. In order, therefore, to understand the systematic 

 relations of the extinct family, we must assume either that these 

 animals are not related to the felines at all, but resemble them 

 merely as the result of an independent and parallel development, 

 or that the Felidce and Viverridce are related. In view of the 

 numerous and extraordinary resemblances between the cats and the 

 higher Nimravidce, which no one has ever called in question, the 

 weight of probability is strongly in favor of the second view — 

 namely, that the cats are a highly specialized branch of the primi- 

 tive group which also gave rise to the Viverridce. If we admit the 

 connection between the Nimravidce, and the Felidce, that between 

 the latter and the Viverridce seems necessarily to follow. 



With regard to the White River genera of this family, Schlosser 

 has remarked as follows (No. 15, pp. 589-90) : " Unter den Fleisrli- 

 fressern treffen wir hier zum ersten mal katzenahnliehe Formen, 

 Drepanodon, Dinictix, Buncelurus und Hoplophoneus, doch soil nur 

 das Gebiss und der Schadel mit jenen der Katzen nahere Ueberein- 

 stimmung zeigen, das Skelet jedoch vielmehr an jenes von Hunden 

 erinnern. Hochst wahrscheinlich handelt es sich hier um einen 

 volli£ erloschenen Formen kreis." These statements are due to a 

 misunderstanding of my account. The skeleton of Dinictis and 

 Hoplophoneus is not like that of the dogs, but it does resemble in 

 many important respects, as does also the skull, that of the primi- 

 tive dog-viverrine group, of which Cynodictis may be taken as a 

 tvpical example. So numerous and important are the correspond- 

 ences between the dentition, skull and skeleton of Dinictis, the 

 lower viverrines and the ancestral dogs, that it seems impossible to 

 avoid the conclusion that they are all very closely related. 



Schlosser has not as yet discussed fully his reasons for separating 

 the Felidce so widely from the other Carnivora, and deriving them 

 independently from the creodonts, not having reached that portion 

 of his great work. Some hint of the grounds for his views may be 

 gained from a recent short paper (No. 15, p. 590), where criticising 

 Filhol's opinion that Stenoplesictis, Proailurus, etc., form the transi- 

 tion in mi Cynodictis to the cats, he says: "Dies ist jedoch wenig 

 wahrscheinlich, da cine so weitgehende Umgestaltung unmdglieh 



