ORDER CARNIVORA. 



H37 



of the Upper Eocene and Lower Miocene of the Continent, 

 and also of the White-river Miocene of North America, where 

 it has been described as Galecynus. In this genus the dental for- 

 mula is usually the same as in Cam's, but in some cases (as in that 

 genus) there are only two lower true molars. The teeth (figs. 13 19, 

 1320) resemble also those of the Viverridce, the carnassials being 

 never very long, and the cusp-line of the lower one transverse ; and 

 the humerus has a foramen. The auditory bulla is inflated, but 

 according to Dr Filhol, has no septum. This genus is, indeed, 



Fig. 1320. 



-Left half of the palate of Cynodictis longirostris ; from the 

 Upper Eocene of France. 



one closely connecting the Canoids and Viverroids ; and since it is 

 pretty evident that the Ifycenidce and Felidce are divergent branches 

 from an early Viverroid stock it is probable that Cynodictis repre- 

 sents a type not far removed from the one which has given rise 

 to several of the more specialised Carnivores, and is itself derived 

 from Amphicyoti, or an allied type. 



Family Miacid^e. — This family is provisionally adopted to in- 

 clude the Eocene genera Miacis and Didymictis, which appear to 

 be primitive forms allied to both Canoids and Viverroids, but which 

 Professor Cope placed among the Creodonta. Miacis ( = Uintacyon 

 and Viverravus) occurs in the Upper, or Bridger, Eocene of North 

 America, and according to Dr Schlosser, who places it with the 

 Canoids, in both the Lower and Upper Eocene of France. Didy- 

 mictis (Zimnocyon), which occurs in the Puerco, Wasatch, and 

 Bridger Eocenes of North America, is included by Dr Schlosser 

 in the Viverridce. Dromocyon is noticed on page 1453. 



Family Viverridce. — The Viverroids are comparatively small 

 Carnivores, showing such close affinities in one direction with Cyno- 

 dictis among the Ursidce, in another with the Mustelidce, and in a 

 third with the Hycenida and the Felidce, that their accurate defini- 

 tion is quite impossible. The skull in existing forms generally 



