666 SCIENCE PROGRESS 



family {Apternodontince) of the Centetidce — a reference which, if 

 trustworthy, is of great interest and importance from the point 

 of view of geographical distribution. 



The phylogeny of the Felidce forms the subject of an important 

 article by Dr. W. D. Matthew, in vol. xxviii. pp. 289-316 of the 

 Bulletin of the American Museum of Natural History. It is stated 

 that the majority of the extinct members of the family, including 

 the oldest species, are characterised by a more or less pronounced 

 development of the upper canines into long, flat-sided tusks. 

 These sabre-tooths, or machaerodonts, date from the Lower 

 Oligocene, but typical cats with relatively short upper canines 

 are not known till the Pliocene. The early sabre-tooths are 

 divisible into two series, one characterised by the length and 

 slenderness of the tusks and the large size of the protecting 

 flange on the lower jaw, and the other by the shorter tusks and 

 smaller flange. Hoplophoneus and Dinictis respectively represent 

 the two series in North America. The derivation of the large 

 Pliocene and Pleistocene sabre-tooths from Hoplophoneus has 

 been accepted, but the relations of modern cats to Dinictis have 

 been overlooked. The evidence seems to indicate that the 

 Dinictis phylum led directly into modern Felidce^ the canines 

 having reverted from the machaerodont specialisation to the 

 normal type of carnivorous mammals. The series Dinictis^ 

 Niniravus, Pseiidcelurus, and Felis forms a direct succession, 

 structurally and geologically. 



Dr. Matthew considers that the origin of the family cannot 

 be carried back further than the Oligocene sabre-tooths ; deri- 

 vation through ^lurotherium, a genus based on the milk- 

 dentition of a species of the same group, from the Eocene 

 creodont Palceonictis being inadmissible. 



On the other hand, it may be noted that Mr. R. O. Peterson 

 has described, under the name of Daphcenodon {Mem. Carnegie 

 Mus., Pittsburg, vol. iv. No. 5), the skeleton of a dog-like 

 carnivore of the size of a leopard from the Miocene of Nebraska, 

 which, together with the older Daphcenus, he regards as in some 

 degree intermediate between dogs and cats, although the skull 

 and teeth are essentially dog-like. In many respects Daphcenus^ 

 of which the skeleton is known, is very cat-like, especially in the 

 long, leopard-like tail. A cat-like feature is the partially 

 retractile structure of the claws. In concluding his description, 

 Mr. Peterson observes that the model "is instructive, as it 



