284 PROF. HUXLEY ON THE CRANIAL AND [-^P'"- ^ 



cusps of the lower sectorial of the Dogs. A series of changes 

 of just the same kind is observable in Gijmmira ; but the posterior 

 cusps acquu-e a much greater size, and the molars take on a procyo- 

 uoid character. 



If the cheek-teeth o^Cenie/es, Gijmnura, Oloojon, and Nasua are 

 placed side by side, it is easy to see that the first presents the 

 least-modified condition of the pattern of the crowns of the molars 

 common to the whole series. The reduction, or the less development, 

 of the addition to the inner sides of the teeth, on the other hand, 

 would give such crowns as those o? Pterodon and Hi/cenodon. 



The Insectivorous aflanities of the dentition of the lower Carnivores 

 harmonize very well with other peculiarities of the group. 



The presence of a rudimentary clavicle and of a rudimentary hallux 

 in the existing Canidse leaves no reasonable doubt that they are 

 descended from ancestors which possessed both in a complete state. 

 And the suggestion is no less obvious that their digitigrade condition 

 has resulted from the modification of a plantigrade form. Now we 

 do not know whether any of the Eocene Canidae possessed clavicles ; 

 but there is reason to believe that Amj)hicyon was plantigrade and 

 pentadactyle. 



M. Filliol has shown that Cynhyanodon had a brain more like that 

 of an Insectivore than that of a Carnivore ; and this fact is in accord- 

 ance with a rule which now rests on a pretty broad basis of induction, 

 that, in any given series of IMammals which is represented through- 

 out the Tertiary series, the oldest forms had less highly developed 

 brains than their modern representatives. 



But, if the oldest Tertiary Carnivores were pentadactyle, plantigrade, 

 claviculate, and had brains with relatively small cerebral hemispheres 

 and large, completely exposed, cerebella, one may ask, by what cha- 

 racters were they distinguishable from the Insectivora? and why 

 may not Hycunodon and Pterodon be an extreme development of 

 that type of the Insectivora which is at present represented by 

 Centetesi 



On the other hand, if the primitive stock of the terrestrial Carni- 

 vora was represented by a plantigrade, pentadactyle, claviculate form 

 with the dentition audjaw-angle of Otocyon and provided with epi- 

 pubes, we should be furnished with tliat which is at present wanting, 

 namely a link between the monodelphous and the didelphous Mam- 

 malia. According to our present system, of classification, such a 

 mammal would be grouped among the Insectivora, or as a transitional 

 form between them and the Bidelphia ; and I have long entertained 

 the conviction that the primary stock of all the groups of the mono- 

 delphous Mammalia will be found to occupy this position. 



14. It may be desirable to state in a summary form the principal 

 conclusions to which the facts stated in the preceding pages appear 

 to me to point. 



I. The existing Canidse exhibit a gradual series of modifications, 

 in the form and size of their skulls and the number and characters of 

 their teeth, from Otocyon, as the least-differentiated member of the 



