722 MONOGRAPHS OF NORTH AMERICAN RODENTIA. 
trustworthy character of white nose and ears in var. nzger, though open to 
many exceptions, is the one mainly to be relied upon in distinguishing the 
two varieties. There is apparently a gradual and quite marked increase in 
size southward in these two forms, so that size alone fails to be distinctive, 
especially in the central portion of the Atlantic States, or wherever their 
habitats meet. 
Var. dudovicianus is generally readily separable by its coloration from 
either var. cinereus or var. niger. The rusty-bellied style of var. cinereus, 
however, is often quite inseparable from specimens of var. dudovicianus from 
northern localities. Particularly is this the case when Pennsylvania speci- 
mens of the rufous type are compared with Ohio examples of var. Zudovici- 
anus. There are, for instance, two specimens before me without labels, and 
hence from unknown localities, which I cannot positively refer to one of these 
varieties rather than to the other. There is little, if any, difference in size 
between vars. cinereus and ludovicianus, although the latter is pretty constantly 
appreciably smaller than var. niger, as shown by a comparison of the forego- 
ing diagnoses of the three varieties. The external measurements are, unfor- 
tunately, nearly all from skins, and are hence not very trustworthy. Taking 
the skulls as a standard of comparison, var. niger averages appreciably the 
larger, four skulls of niger averaging 2.75 against 2.58 in both vars. Zudovice- 
anus (eight skulls) and cinereus (three skulls). 
GEOGRAPHICAL VARIATION.~-In varieties cinereus and niger, the only 
appreciable feature of geographical variation is the before-noted gradual 
increase in size southward. A large series of specimens, from numerous 
localities, might show that other variations correlate with differences of 
locality. My material illustrative of these two forms is unsatisfactorily 
scanty. Var. dudovicianus, on the other hand (thanks to more abundant 
material), shows strongly marked geographical variations in color, but I fail 
to notice any well-marked geographical variation in size. Judging from the 
skulls, the northern specimens are rather the larger, yet the largest skull of 
the series is from Brookhaven, Miss. The measurements taken from the 
skins show also a rather greater size northward, the largest specimens 
being from Ohio and Fort Randall, Dakota, if length of body be taken 
as the standard; if, on the other hand, the length of the hind foot be 
adopted, in consequence of being less susceptible to variation resulting from 
taxidermy, the southern specimens average nearly or quite as large as the 
