I l8 Bulkiin American Musei/vi of Natural History. [Vol. XII, 



and the John Day genera Ejihydrocyon, HycBtiocyon and Oligobunis, 

 all with a more reduced dental formula. They form the most 

 carnivorous division of the family, approximating the Viverridae 

 or Felidse, while Nothocyon and Megalotis stand at the other ex- 

 treme, approximating the Procyonidge. 



3. Daphcznus and Tenmocyon. 



The relationship between these two genera has been pointed 

 out by Prof. Scott ' and Dr. Eyerman.^ All our evidence tends to 

 confirm the view that Ternnocyon is a descendant of DaphcBJiuSy 

 although the gap between them is in some respects considerable. 

 Both have the same strong, round-edged, massive cusps, but in 

 Daphcenus the molars are large and low, while in Tenuiocyon they 

 are small and high-cusped. Daphcenus is just beginning to de- 

 velop a trenchant heel on m yrg- ; it has the same deep jaw, and 

 does not show any intermediates on the upper molars. 



Hypotemnodon has a somewhat problematic position intermediate 

 between Cynodictis and Daphcenus in most of the skull characters 

 but with trenchant heel on nij only. Scott derives it from 

 DaphcEtius J there is, however, a species of Cyjwdictis in the upper 

 White River, which presents a closer approximation in character 

 of teeth, and if our association of specimens is correct, has de- 

 veloped a low trenchant heel on my only. For this reason it is 

 placed provisionally in the Cynodictis line. 



Prof. Scott derives Cynodesmus also from DaphcBnus, basing the 

 derivation apparently on the presence of frontal sinuses in both. 

 Whether the presence of these in the larger species of Canidae is 

 of much phylogenetic value seems uncertain ; on other grounds 

 Cynodesmus might well be connected with Cynodictis. 



4. Prodaphcenus and Daphcenus. 



5. Uintacyon and Prodaphcenus. 

 These have already been considered. 



III. — The Eocene Ancestors of Cynodictis. 



Vulpavus palustris ' Marsh. 

 With this species we come to consider the phylum which in all 

 probability terminated in the modern genus Canis. Its oldest 



' Notes on the Canidae of the White River Oligocene. Trans. Am. Phil. Soc. 189S. 



« Amer. Geol. 1896. 



^ Amer. Jour. Sci. Aug. 1871, (p. 16 of separate). 



I 



