CHAP. XV.] THE PEDIGREE AND ORIGIN OF THE CAT. 513 



extinct forms of carnivora nearly related to the cats and civets ; and 

 the ancestors of such forms, again, must be sought amongst car- 

 nivorous mammals of more and more generalized structure till we 

 come to creatures from which all mammals may be supposed to have 

 descended. What animals were the progenitors of all mammals, is 

 as yet a matter of pure speculation, and no positive judgment can be 

 formed concerning it by any prudent naturalist. Certain proba- 

 bilities, however, are evident as to this and cognate questions, but 

 before adverting to these probabilities, it will be well to recall to 

 mind some of those existing and extinct animals to which reference 

 has already been made. 



The most aberrant and generalised of all existing cats is the 

 cheetah (Ci/ucehcrus). But though this animal approaches the other 

 carnivora in that its claws are less retractile than those of other cats, 

 yet its tooth structure — its upper sectorial * — is exceptional in a way 

 peculiarly its own. Neither Cynmlurus, therefore, nor its extinct 

 ally ^hirodon, seems to help us towards tracing the cat's pedigree. 



The flat-headed cat {F. pknuccps), as has been shown (Fig. 177), 

 approximates somewhat towards viverrine forms in the large size of 

 its two-rooted first upper molar. But the extinct genera Psendcelurus, 

 Dinictis, and above all, Arcluelurus, lead us decidedly towards more 

 generalized forms, and render the descent of both Cri/ptoprocta and 

 Felis from some common Yiverrine root a matter highly probable. 



It must be borne in mind, however, that although these miocene 

 and eocene cats were thus generalized in structure, yet a most ex- 

 tremely specialized form of cat, e.g., Eiismilus, existed at the same 

 early period. But a very generalized kind of dog, Otocyon (which has 

 four premolars and three molars on either side of each jaw), exists 

 to-day side by side with dogs in which the number of teeth is much 

 less, and which are more specialized. Yet naturalists do not on that 

 account doubt but that Otocyon is a survivor of an earlier condition 

 once common to the whole group of Cynoidea. Similarly the co- 

 existence oi EusniUus with Dinictis or Archcelunis, does not detract 

 from the probability that in the last two genera we have examples of 

 the sort of animals whence all cats come. 



Zoological and palaeontological evidence, then, points to a viverrine 

 origin of cats. They seem either to be the very specialised descend- 

 ants of ancient viverrine animals, or else both cats and ^averrines are 

 the diverging descendants of an ancient, more generalized form 

 which existed in times anterior to the eocene, of which more gene- 

 ralized form no relics have as yet been discovered. 



§ 4. But the viverrine animals themselves, whence came they ? 

 In the existing creation, they are distinguished from the hycenas by 

 having two upper tubercular molars. But Professor Gaudry has 

 discovered a form (named by him Ictithenum hipparionum, which, 

 as we have seen,t is intermediate between the civets and hyaenas, 



See ante, p. 428. f See ante, p. 507. 



L Ii 



