38 F. B. SUMNER AND R. R. HUESTIS. 



difference is about as great, for example, as that between P. inani- 

 culatus sonoricnsis and P. jnaniculatus gainbeli. 



The material comprised in these studies included 208 specimens 

 of the parent races, eremicus and fratercnlus, 199 F : and 147 F 2 

 hybrids. No bodily measurements were made of these mice, nor 

 were skeletons saved. Skins were, however, prepared according 

 to the method long practiced by the senior author. 1 Later, the feet 

 were removed from the dried skins for a study of the pigmenta- 

 tion of the soles. After removal, they were first soaked for a day 

 in distilled water, next transferred to 70 per cent, alcohol, then to 

 equal parts of alcohol and glycerine, and finally to pure glycerine. 

 After this treatment, the feet were graded according to a scale 

 previously adopted by the senior author. 



The series of skins upon which these studies were made were 

 superior in two respects to those earlier employed by us in the 

 study of Peromyscus. In the first place, the numbers were larger, 

 both for the pure races and the hybrids. Secondly, the animals 

 were skinned at a somewhat greater age, eight months being 

 adopted as a minimum instead of 6 l / 2 months, as had been the 

 previous practice. 2 Thus the hair of these mice seldom showed 

 any traces of the second molt, having passed into the third (first 

 "adult") pelage. 3 



From these prepared pelts small samples of skin, with the ac- 

 companying hair, were removed with a sharpened cannula, having 

 a diameter of about 1.5 millimeters. The hairs were subjected by 

 the junior author to quantitative determinations of two sorts, 

 (i) They were grouped, on the basis of their size and pigmenta- 

 tion, into four somewhat arbitrarily distinguished classes, and the 

 number of each class comprised in the sample was counted. (2) 

 The length of the subterminal (" agouti ' 5N > band and of the ter- 

 minal region, distal to this, were measured for ten of the " class 

 A " hairs in each sample. Likewise, the total length of the long- 



1 Sumncr, Jonrn. E.vp. Zool, Aug., 1909 (p. 101) ; described more fully by 

 Collins, Jonrn. Expcr. Zool., Aug., 1923 (p. 48). 



2 A few which died at an age less than 8 months were included, provided 

 that the hair of these showed no trace of molting. On the other hand, some 

 others were several months older than this minimum age, owing to their hav- 

 ing been kept for breeding purposes. 



3 At the age adopted for the earlier series (6^ to ; months) a certain pro- 

 portion of animals had not fully completed the second molt. 



