252 THE STRUCTURAL EVIDENCE OF EVOLUTION. 



but for the fact of the imperfection of the shear blades in all ex- 

 cept the last. This is seen in the large and prominent anterior 

 cusp and very short posterior blade. 



In all recent and most Miocene Carnivora, the sectorials are 

 reduced to one in each jaw, the exception in the latter period 

 being the genus Hycenodon, which did not continue later. The 

 posterior carnassials of the Eocene forms disappear, and in the 

 most specialized recent families HymnidcB^ Mustelidm, and Fe- 

 lidcB have scarcely any representatives. The shortening of the 

 series appears in the premolars as well, until we have the formula 

 of the FelidcB, P. m. 2 ; M. 2. 



In the inferior dentition the same process may be observed in 

 the successional modifications of the entire series. The Eocene 

 forms of carnivora frequently display more numerous sectorial 

 teeth (such as they are) than any of the existing families. The 

 important change, which is clearly indicated, is the progressive 

 extinction of the genera with numerous sectorial teeth, accom- 

 panying the increasing specialization of the sectorial tooth in the 

 genera which remain. In other words, the numerous types of 

 digitigrade carnivora which have survived are those developing 

 but one sectorial tooth (whose earliest rei^resentative is Didymic- 

 Us). The increased perfection of the sectorial tooth has been as- 

 sociated with a reduction in the number of other molars, first, 

 posterior, then anterior to it, which reduction has been accom- 

 panied by an increased relative size of the sectorial. By this pro- 

 cess concentration of the carnassial function has been gained, and 

 increased robustness of the jaws, by progressive shortening. The 

 slender form of the rami of the Eocene genera and Hymnodon are 

 much less efficient in functional use than the stout jaws of exist- 

 ing MustelidcB, Hywnidce, and Felidce. 



2. The Lopliodonts. 



Transitions between the Bunodonts and Lopliodonts are very 

 obvious, so much so as to lead to the belief that the several sub- 

 divisions of the Lopliodonts represent modifications of correspond- 

 ing types of Bunodonts, and that the two are partially '^ homolo- 

 gous groups." Both present corresponding Amoebodont and An- 

 tiodont types ; as an example of the former kind of Bunodont, 

 the mandibular molars of the genus Hyopsodus may be cited ; of 

 the latter kind, the same of the genus AchcBnodon, both the ear- 

 liest, or Eocene genera. It remains to indicate the intermediate 



