446 University of California Publications in Zoology | Vou. 17 
Rainier, Washington, averages larger than rufa; and californica, 
inhabiting the Canadian zone of the Sierra Nevada-Caseade moun- 
tain system in California, is considerably larger than phaea, found 
in the coastal Marin County near San Francisco Bay. Although 
this tendeney toward variation in size is slight, its apparently uniform 
association with difference in altitude or life-zone indicates its possible 
significance. 
It has been considered (C. Hart Merriam, 1899a, p. 21; Taylor, 
1916c, p. 501) that the ‘‘mountain top’’ subspecies of Aplodontia 
(columbiana, rainiert, californica) are more closely related to each 
other than to any other forms of the genus. This is possibly true of 
columbiana and rainieri (see pl. 28), but the balance of the evidence, 
as derived from the present study of rainiert and californica, favors 
rather the view that these mountain subspecies are more closely related 
to lowland or neighboring forms than to each other. Their general 
similarity in size and certain other characters would seem to be due 
to parallelism. 
Geographic isolation appears to be intimately associated with 
speciation in the group. Although material at hand is not sufficient 
to demonstrate each step in the process, it is enough to suggest that 
the ranges of all the described subspecies of rufa, except nigra and 
phaea, mosculate at one point or another. Geographic variation is 
continuous, though shght and very gradual, in the intergrading forms. 
The most strikingly colored form is the apparently completely isolated 
Aplodontia rufa mgra of Point Arena, California. 
It is to be noted that whereas on the northern coast of California 
three well-marked forms of Aplodontia are found, on a longer coast- 
line in the state of Oregon there occurs but one. The three, communi- 
ties of Aplodontia on the coast of northern California are separated 
from each other by considerable gaps, while along the coast of Oregon 
the animals appear to be continuously distributed. 
It seems to the writer that the geographical distribution of the 
different subspecies of Aplodontia, as well as the degree of develop- 
ment and nature of the characters separating them, indicate that 
among the factors possibly concerned in their differentiation, geo- 
graphic isolation is, at least, one of the most important. 
