102 Mr. E. D. Cope on the 



question may not be so early caducous, and rely on the re- 

 stricted diagnosis. It is thus apparent that the genus Dyso- 

 dus is distinguished from Synagodus by the absence of two 

 premolars from each jaw. While the genera agree in other 

 respects, their typical species are very different. 



This genus probably diverged from that now represented 

 by Synagodus at a comparatively late period. Although it 

 exhibits a greater degree of dental reduction than that form, I 

 admit that the possibility of its having come off from Canis 

 rather than from Synagodus is worthy of consideration. This 

 is suggested by the fact that the remaining (first) tubercular 

 molar of the inferior series is, in D. pravus, more like that of 

 the species of Cants in all respects — among others, in having 

 two roots. 



In D.2)ravus the superior third premolar is sometimes shed, 

 like the incisors, leaving the formula, I. %' C. -f; Pm. \] 

 M. i|r. I have excluded this character from the generic 

 diagnosis, as in the case of the incisor and superior tubercular 

 teeth, because they are at the present time unstable ; that is, 

 the parts in question are in process of metamorphosis. When 

 characters are thus variable, they cannot be used as the bases 

 of natural divisions ; but when they are stable we are com- 

 pelled to recognize them. The characters which I have 

 included in the diagnoses of Synagodus and Dysodus I have 

 thought to be of this character ; and I am by no means sure that 

 the absence of the superior incisor teeth should not be placed 

 in the same category. But none of these characters, whether 

 stable or unstable, can be regarded as monstrosities, such as 

 multiplied digits, fissured palate, &c. They are, on the con- 

 trary, in the direct line of numerical succession of parts 

 already represented by the genera of Canidse and of all digiti- 

 grade Carnivora. This, as already stated, consists in the 

 reduction in the number of the teeth and their tubercles, 

 forming a series which, commencing with the generalized 

 extinct type Amphicyon , approaches more and more nearly to 

 the Felidae. In the inferior sectorial, the genus Dysodus 

 approaches nearest of all Canidse to some of the earliest genera 

 of cats, as Hoploplioneus (although easily distinguishable) ; 

 while in the reduction of its premolars it approaches the 

 modern forms of that family. In the early shedding of the 

 incisors it reaches a condition not found m any Carnivora, but 

 one which marks the extreme of development of the ungulate 

 mammals in various lines, e. g. Euminantia, Omnivora, and 

 Amblypoda. 



Dysodus pravus, sp. nov. 

 This species, which is known as the Japanese sleeve-dog, 



