COLOR MUTATIONS IN MICE OF PEROMYSCUS 309 



be tested. Of the remaining six, two were entered as 'probably 

 yellows/ though 'atypical' (p. 297), while four were entered as 

 'probably not yellows.' They were, however, recorded as being 

 much paler and more buff than the average gambeli. 



These six specimens were mated in various combinations, and 

 some twenty-five descendants were born.^^ The latter were a 

 rather nondescript lot. Five of them were pallids — a result 

 which was not unexpected, since even some of the 'normals' 

 were doubtless heterozygous for the pallid factor. It is of 

 interest that these F3 'pallids' included the dark-eyed specimen, 

 with darker pelage, referred to above (p. 304). The others, for 

 the most part, differed rather widely from the average normal 

 gambeli. A few might have passed for paler (more buff) repre- 

 sentatives of the latter subspecies, and had nothing in their ap- 

 pearance to suggest the 'yellow' variety. One or two looked as 

 if they might be 'atypical yellows.' The majority, while not at 

 all uniform, were of a curious, somewhat ruddy, appearance, 

 having little resemblance to the typical yellows of either of the 

 chief strains discussed above, but also probably unlike any 

 wild mice which we have trapped. Even some of the descendants 

 of the F2 pair which had been listed as 'probably yellows' showed 

 no nearer approach to the yellow type than the others. One 

 of these, indeed, is entered as 'medium gray.' 



This case is, of course, complicated by the fact that we are 

 dealing with a mongrel combination of three subspecies, as well 

 as with two different color mutations. The 'pallids,' as already 

 stated, sprang from a sonoriensis-rubidus cross, while the 'yellows' 

 were of pure gambeli stock. This intermixture of subspecies 

 does not, however, affect the clear segregation of the pallid and 

 the albino factors. It is only the 'yellow' factor (or factors) 

 concerning which there is any question. 



These difficulties cannot be cleared up until, 1) the genetic 

 behavior of the 'yellow' color variety has been far more thoroughly 



^* These births occurred during the protracted absence of the senior author. 

 Upon his return, it was not in many cases possible to distinguish the children of 

 a given pair from their grandchildren, or from broods resulting from the mating 

 of parents and offspring. If the parent mice had been pure recessives, however, 

 this fact would, of course, have made no difference in the result. 



