46 THE GROWTH OF GROUPS 



second more attentive observer may be indisputable. 

 Let us suppose that he is right. The red tint must be 

 due to some material influence which is present in every 

 hair of the animal. One cannot say more or less than 

 this in regard to the blackness, although the latter con- 

 stitutes a difference which is obvious at a glance. One 

 cannot say that the redness is trivial and of no importance, 

 but it is permissible to choose deliberately the black 

 element for observation and to neglect others that are 

 less visible. This is the method adopted here. The 

 characters selected because of their visibility are as 

 follows : 



(1) Melanism. Some rats are black, like the one 

 shown in the frontispiece. Among these, some have no 

 trace of brown in their fur ; black, iridescent green, and 

 slaty blue being the only appreciable tints. Others, 

 though appearing to be black, yet have a few brownish 

 hairs in their fur. Two litters raised in captivity from 

 a cross between a thorough black and a brown were of 

 this mixed kind. Both the pure and mixed black are 

 found wild. 



(2) Albiventralism. Some rats are pure white on the 

 ventral surface. In the great majority the fur of the 

 under side is coloured, each hair being grey with a light 

 brown tip. Rats were received however, from many 

 parts of India, in which the fur covering the belly, breast, 

 throat and lower jaw, as well as the inner side of the limbs, 

 was pure white. This whiteness was sharply marked off 

 from the coloured sides, as in many other vertebrate 

 animals. 



(3) Caudal albinism. As a rule the tail of a rat is 

 deeply pigmented from base to tip. The pigment, which 



