GEOGRAPHICAL DISTRIBUTION AND DESCENT. 



37 



ungulates (Artipdactyla), but these relations 

 do not suffice to represent any special stock. 

 All the affinities that have hitherto been 

 suggested break down in face of one slight 

 objection, namely this, that the supposed 

 ancestors belong to more recent strata than 

 their assumed descendants. There is only 

 one exception. The members of the genus 

 Dindceras, gigantic animals from the Middle 

 Eocene of Wyoming and Colorado at the foot 

 of the Rocky Mountains, are older than the 

 mastodons. But does that suffice to entitle 

 us to regard these forms as constituting the 

 primitive stock of our proboscideans? I do 

 not believe it. It appears to me difficult 



to bring these animals, furnished with horn- 

 like bony excrescences, numerous very 

 small molars, and enormous canines and 

 no incisors, into connection with the pro- 

 boscideans, in which the incisors play so 

 important a r6le and the canines are always 

 absent. 



To sum up, the Proboscidea form a separate 

 order, which has some affinities to the Un- 

 gulata, which was formerly spread over the 

 whole breadth of the mainlands of both hemi- 

 spheres, but which is now in process of rapid 

 decay, since its members are all extinct with 

 the exception of two species living in the 

 tropics of the Old World. 



