EVOTOMYS 259 



24-8 mm. ; on the other hand, specimens already 27-1 mm. long 

 may still have iti^ in vigorous growth ; but all the larger speci- 

 mens examined have the cement spaces of m^ closed and the 

 roots more or less well developed. These facts probably indicate 

 that in this as in other sub-species individuals may vary consider- 

 ably in size, the range of variation in skull length probably 

 amounting to about 3 mm. in any given subadult or adult stage 

 of growth. 



Cheek-teeth in adult stages of wear essentially as in the typical 

 subspecies. In young individuals m^ is much more complex than 

 it is as a rule in equally young E. r. rufocmius, having four salient 

 angles on each side with an occa-sional trace of a fifth outer 

 salient angle. Very often in such young specimens the first 

 inner infold is relatively shallow, so as to leave the first pair 

 of triangles, following the anterior loop, more or less broadly 

 confluent with each other. As age advances and wear exposes 

 deeper portions of the crown upon the grinding surface, all the 

 triangles become substantially closed, the posterior or fourth 

 salient angles on each side die out, and the posterior loop is 

 gradually shortened until the tooth acquires exactly the form and 

 appearance it has in adult specimens oi E. r. rufocanus (Fig. 82). 

 In lower molars also the dentinal spaces of //ij and mj are more 

 confluent in young stages of wear than they are in middle age, 

 when the triangles become substantially closed as in the adults 

 of other forms of the species. Ephemeral complications of the 

 fore-part of Wj are sometimes to be observed in the young (Fig. 

 78, 1, 2). In extreme old age certain of the dentinal spaces open 

 out again and the salient angles acquire that peculiar rounding so 

 characteristic of the adult teeth of normal species of Evotomys. 



For external and cranial measurements, see tables at end of 

 volume. 



Remarks. — A careful study of the magnificent material which 

 the British Museum owes to the generosity of the Duke of Bed- 

 ford leaves no room for doubting that E. smithii, E. " hedfordicB ,"' 

 E. '■ andersoni" and E. '' niigatw " belong to one form and if the 

 material representing each of these nominal species had arrived on 

 one and the same day all would, in my opinion, have been inevit- 

 ably treated as one form. Voles are always difficult animals to 

 deal with, and E. rufocanus is an especially difficult vole. A 

 species in which the skull looks quite adult when the condylo- 

 basal length measures only 23-7 mm. and then proceeds to grow 

 to about 29 mm., its progress being marked by great changes in 

 form as well as in size, offers a series of fine pitfalls for the energetic 

 systematist. To make the matter more difficult, an adult pelage 

 is acquired and sexual maturity is attained at a very early stage 

 of growth. In these circumstances long series of specimens from 

 single localities afford no absolute safeguard against error ; indeed, 

 as in the present case, they may help at first to mislead. In one 

 locality where voles are breeding fast, and where there is perhaps 



