CHAP, xm.] THE CAT'S PLACE IN NATURE. 489 



(6) The bulla is, at least in existing species, divided into two 



chambers, which are not placed completely one behind the 

 other ; 



(7) The bulla is most prominent towards its inner, not towards 



its hinder border ; 



(8) The bony meatus auditorius externus is not produced an- 



teriorly into an outstanding process ; 



(9) There may be a small bone to the penis ; 



(10) There are no conspicuous scent-glands ; 



(11) There are only thirteen dorsal vertebra). 



The FELID^; differ further from the TRUE HYJEXAS (i.e., the 

 Hya3nas as distinguished from Proteks) in that : 



(1) Their teeth are less powerfully formed for crushing ; 



(2) The bulla is divided by a septum ; 



(3) The carotid foramen is generally still less conspicuous. 



They further differ from Proteks in that : 



(1) Their teeth are much stronger and larger relatively, and 



more closely approximate ; 



(2) The chambers of the bulla are not quite one behind the 



other. 



The FELID^E differ from the CRYPTOPROCTiDyE in that : 



(1) They are quite digitigrade ; 



(2) Their skull is relatively shorter and rounder ; 



(3) Their premolars are (except in Archcelurus) less numerous ; 



(4) The chambers of the bulla are not quite one behind the 



other ; 



(5) There is no ali-sphenoid canal, save in some extinct genera ; 



(6) The supra-condyloid foramen is of moderate size ; 



(7) The metacarpals and metatarsals are relatively longer ; 



(8) The pollex is very small, and the hallux a mere rudiment ; 



(9) The naked pads of the feet are much less extended ; 

 (10) The bone of the penis is small. 



25. We have now before us as complete a statement as the 

 author can give of the relations which exist between the cat's family 

 and all other living organisms whatsoever. 



As to the subordinate groups contained within the cat's own 

 family, i.e., its genera, we saw in the last chapter that all known 

 cats living and extinct can be arranged in eleven sets of kinds, to 

 which the names Felis, Cyncelurus, 2Elurodon, Archcelurus, Dinictis, 

 Nimravus, Pscudcehtrus, Hoplophoneus, Pogonodon-, Maclicerodm, and 

 Eusmilus, have been given. We now see what was meant by saying 

 that these groups have each the value of a " genus." As to the 

 relations which exist between the feline genera, we now also see that 

 the exceptional characters presented by Dinictis and Arch&lurm are 

 peculiarities which cause those genera to have a certain resemblance 

 to the Viverrine family. 



In the last chapter we recognized the fact that an extreme 



