INTEODTJCTION. xxxv 



general acceptance. Schlosser objects to it as receiving very little 

 support from Palaeontology, and regards the Dogs as very closely related 

 to the Bears, a view which receives support from both Gaudry * and 

 Lydekker f. Scott also deems the Bears and Dogs to be nearly allied, 

 while he regards the Civets and Hysenas as being more allied to the 

 Weasel-group than to the Felidce ; while Schlosser considers the Cats to 

 be most widely separated from all the other groups of existing Car- 

 nivora, and to have had an origin independent of them. Garrod, on 

 account of the form of the brain, represented the Canidcs as an off- 

 shoot from the Felida. 



Such conflicting opinions suffice to make plain, to everyone who 

 reflects on them, how speculative and uncertain such phylogenetic 

 statements are. 



Maintaining, then, still that system of classification for the Carni- 

 vora which we before made use of, it but remains for us to note here 

 the characters by which the Canida, or Cynoidea, differ from the 

 Arctoid Mammals on the one hand, and from the ^Eluroids on the 

 other. 



That the Dog-group (excepting domestic forms) is singularly uniform 

 in structure compared with the others, will be evident if we compare 

 the amount of divergence between C. lupus and Otocyon megalotis, 

 with the great contrast which exists between such species as a Lion 

 and a Mongoose amongst the ^Eluroid forms, and between a Raccoon 

 and an Otter amongst the Arctoids. 



The characters by which the Canidte differ from the Arctoidea are 

 the following : 



They are always digitigrade. 



They possess a smooth auditory bulla which tends to be divided 

 internally by a bony septum, which nevertheless remains very in- 

 complete. 



* See his ' Les Enchain ements,' chap. ix. 



f Paljeontologia Indica, ser. 10, p. 202 ; and Cat. of Fossil Mammalia, part i. p. 106. 



i Notes on the Osteology and Systematic Position of Dinictis felina, p. 242. 



See Proc. Zool. Soc. 1878, p. 377. 



