346 THE AMEBIC AN NATURALIST [Vol.XLVII 



tors. He also points out that in tricolor dogs a spot, if 

 on the back, is black ; if on the leg, yellow. If this view is 

 correct it would seem to follow that regional differences 

 determine the color that develops, or that somatic segre- 

 gation of color factors is definite in respect to body 

 regions. 



In the case of the Norway rat, a wild gray bred to a 

 spotted animal gives offspring that generally contain a 

 white spot on the belly. It would seem, in this case, that 

 the "factor" for spotting in this region of the body is 

 dominant over the uniform coat— the other spotting fac- 

 tors may be recessive, and for their development depend 

 on the ss-factor. 3 



So long as these questions remain on such an unsatis- 

 factory basis we can do little more than adopt provision- 

 ally some such view as Castle's, or else describe the facts 

 without regard to any special theory. In the following 

 account, therefore, we shall attempt little more than a 

 description of the results that we have obtained. Our 

 description resolves itself, therefore, into the question 

 of the heredity of black-and-white somatic areas. The 

 question of whether these are overlapping areas as 

 Castle assumes or else spot centers in which color or no 

 color may occur, or both is left undecided. It is certain 

 that a spot may be large or small, and, therefore, the 

 realized pattern is variable. Possibly we may get a 

 clearer idea of this question if we look upon the spot as a 

 center from which color, if present, is more likely to 

 spread, and, if we assume somatic segregation in an 

 early stage of the embryo the extent of the spot will be 

 a measure of the extent to which a given cell containing 

 the color factor multiplies as compared with neighboring 



