102 



THE AMERICAN NATURALIST [Vol. LI1 



result from the relative length of the terminal pigment- 

 less zone which is present in these hairs. The ventral 

 hair of the desert race also appears to be somewhat longer, 

 or at least of a softer texture, than that of the others. 



In the case of the ventral surface, like that of the dorsal, 

 these differences relate to averages rather than to indi- 

 viduals. Likewise, it is of interest to note that within 

 each race there is little or no correlation between the 

 dorsal and the ventral shade. I have frequently graded 

 a considerable row of mice of a single race in respect to 

 the shade of the dorsal pelage, and found, on turning the 

 animals over upon their backs, 6 that the order of arrange- 

 ment did not correspond with the ventral gradations of 

 shade. 



Another differential character of these races is the de- 

 gree of lateral extension of the ventral white area of the 

 body, or, conversely stated, the ventral extension of the 

 dorsolateral pigmented area. The colored and uncol- 

 ored regions of the body come together abruptly along an 

 irregular lateral line extending from the snout to the tip 

 of the tail. In the desert race, more of the white ventral 

 region is usually to be seen in side view than is seen in 

 the darker races. The gradation of the other three races 

 among themselves is less obvious. 



This degree of extension of the colored area relates not 

 merely to the body but to the appendages. In the darker 

 races an elongated tongue commonly extends down upon 

 the fore-limb, in some cases even to the hand, while in 

 sonoriensis such a ventral projection is usually little de- 

 veloped. The graduation of our four races in regard to 

 this character corresponds to that noted in respect to 

 shade. Similar conditions are observable on the hind 

 limbs, particularly upon the ankle, where the pigmented 

 hair may extend as far as the heel, or may fall short of 

 this in varying degrees. The ease of the tail will be dis- 

 cussed separately. 



A hair character which seems to be peculiar to sonori- 

 ensis, among the races here considered, is the presence of 



e Fresh specimens, not skins, are used for most of these comparisons. 



