518 



THE CAT. 



[CHAP. xv. 



varying conditions of life, but which we know to be sometimes very 

 differently formed in different members of one and the same family. 

 Yet we must accept their evidence or none. It is the only evidence 

 which is largely available, nor will there be much danger of serious 

 error in making use of it, if the caution here offered as to its defec- 

 tive nature be duly borne in mind. 



Felis. 



Mustclidae. 



Irsectivora. 



Ursula;. Canidse. 



Otoc yon. 



Cryptoprocta. 



Viverridse. 



Hyscn 



ictis. 



\ 



Amph 



Procelurus? 



Galcc yuus. 



iretos. 



Cyno 



icyon. 



Cyno don. 



Gym 



lictis. 



Arctocyon. 



Pter 

 Hyaen 

 Provi 



Falseonictis. 



odon. 

 odon. 

 verra. 



8. If we can only conjecture with more or less probability what 

 were the older mammalian ancestors of the cat, we are still less able 

 to determine the line of its descent through non-mammalian ances- 

 tors. The structure of the shoulder girdle of the Monotremes, may 

 be held to point towards a reptilian origin of the Mammalia ; but 

 the position of the ankle-joint and the constant development in 



