Voles from Yunnan. 147 



have eitlier been entirely eliminated from the teeth of Evo~ 

 iomf/s, or which are represented in the latter only by more 

 or less obscure vestiges visible only in the earlier stages of 

 wear. 



In some other respects Eothenomys and its allies have 

 proceeded much further than Kvotonujs. The cheek-teetli 

 have become completely hy})sodont and rootless, while in 

 Evotomys they develop roots in adult stages of wear. The 

 increased hypsodontyis especially well marked in Caryomys, 

 in which the maxillary capsule of m^ rises up in the floor of 

 the orbit as a great mound, blocking the mouth of the splien- 

 orbital fissure ; and in all three genera it results in a marked 

 lingual displacement of m-i by the shaft of the incisor, in its 

 passage from the lingual to the labial side of the molar 

 bases between ?«2 ^^^^ ^^s- The beginning of such a dis- 

 placement of m^ is seen in Evotomys ; its culmination in 

 true Microtus. Much of that confiuency of the dentinal 

 spaces which is so characteristic of the cheek-teeth of Evo- 

 tomys is retained in Eothenomys and Antetiomys ; but in 

 Caryomys it is replaced by a more complete alternation of 

 the inner and outer prisms, the triangles being rather tightly 

 closed_, as in adult teeth of the higher voles. 



Externally, the only noteworthy advance upon the con- 

 dition found in Evotomys seems to be shown by the reduction 

 of the mammary formula from 2 — 2 = 8 to — 2 = 4^ the 

 pectoral mammae being completely obliterated, as in 

 Fitymys. 



Two other genera related to Evotomys may be briefly 

 mentioned, viz., Craseomys and Aschizomys. Craseomys 

 should be regarded, perhaps, as a mere subgenus of Evo- 

 tomys ; its species show a great increase in size, with large 

 massively built skulls and powerful teeth. The cheek-teeth 

 have become very hypsodont, although they still develop 

 roots in old age. At that moment when the molar crowns 

 have attained their greatest vertical development (i.e., just 

 before the roots begin to be formed) the capsule of m^ pro- 

 trudes from the floor of the orbit, exactly as in Caryomys ; 

 but later^ as the crown wears away, the bony mound is re- 

 absorbed and sinks, the maxilla then resuming the appearance 

 it has in ordinary Evotomys. No character has been found 

 to indicate that Craseomys is other than a relatively highly 

 specialized offshoot from normal Evotomys. 



Aschizomys, Miller, represented by a single species from 

 the Siberian shore of Bering Strait, is apparently a most 

 remarkable descendant fj'om some old species of Evotomys. 

 In skull and in the pattern and lightness of the cheek-teeth 



10* 



