508 



THE AMERICAN NATURALIST [Vol. XLIU 



ing here and elsewhere in animals is due to factors of the sort 

 postulated they are controlled (?) by a periodic function of the 

 hair bud. We meet here with the same problem the embryol- 

 ogist encounters in proportionate development or interrelation 

 of the parts. Whether this is a dynamic function (epigenetic) 

 or can be referred to a system of factors in the germ is a difficult 

 problem and for the future to' decide. 



Before leaving this question of yellow mice a few well known 

 facts may be stated. Yellows exist from a deep orange to a 

 pale lemon yellow. All intermediate gradations may be found. 

 Whether these are in reality a graduated series or a series of 

 overlapping conditions we do not yet know. The presence of 

 other pigments combined with yellow is also familiar to every 

 student of these mice. If yellow is due to an inhibiting factor 

 that factor must at times very imperfectly do its work of in- 

 hibiting. I have a race of sports of the house mouse with white 

 bellies, gray backs and yellow sides. The hairs on the sides 

 may be pure yellow, which should be due to the action of the 

 inhibiting factor in this particular region only, since on the rest 

 of the upper surface the hairs are ticked. Even in gray mice 

 single hairs may be yellow. 



In some yellow mice the belly is pure white. This must be 

 due to a further factor that in this region inhibits the yellow, the 

 yellow itself being the result of another inhibiting factor. In 

 other words an inhibitor of yellow must also be postulated. It 

 might be assumed of course that the white belly is due to the 

 absence of yellow in the region, but since the mouse can produce 

 yellow its absence in the belly must also be accounted for by 

 some special assumption. Again we meet with the localization 

 factor— a problem that Mendelian studies have scarcely yet ap- 

 proached except by the purest symbolic representations. 



The fact that many rodents change the color of their hair 

 according to age indicates that the physiological condition of 

 the animal is an important factor in determining its color. If 

 the mechanism of Mendelian inheritance involves only the 

 shuffling of morphological determinants, as implied in many 

 current conceptions of the mechanism of inheritance, the changes 

 that take place in the same individual are difficult to under- 

 stand unless it be admitted that temporal and local conditions 

 affect the development of the determiners. Such an admission 

 is practically equivalent to referring the development of a color, 



