10 



THE AMERICAN NATURALIST [Vol. XLVI1 



than the part taken as the index is affected. Weismann's 

 theory, on the other hand, seems as a rule to identify each 

 character with a special determinant for that character, 

 and his meaning is clear when it is remembered that the 

 process of development on Weismann's view is a process „ 

 of sorting out of the determiners of the germ-plasm into 

 different regions of the body. The factorial hypothesis 

 makes no such assumption, but refers differentiation to 

 the interaction of the parts on each other — every cell 

 retaining the full complex of the original germ-plasm. 

 Hence the possibility of the far-reaching effects of any 

 change in the germ-plasm ! 



II 



The presence and absence system of nomenclature 

 (aside from its implications as to what is meant by pres- 

 ence and absence) has till the present time justified itself, 

 when properly interpreted, by its usefulness. It seems 

 to me that as a system of nomenclature it may be used, if 

 one so desires, quite apart from the idea, that a loss in 

 a character involves necessarily a loss in the germ-plasm. 

 I can bring forward one clear case at least that seems to 

 me difficult to explain if absence is taken literally to mean 

 the loss of a factor from the germ complex. I refer to a 

 mutation "backwards," which in the older terminology 

 meant reversion, or atavism. 2 In my pure cultures (at 

 rare intervals) individuals have appeared like the orig- 

 inal progenitors of the stock. I have not scrupled to put 

 aside this evidence, because contamination, even with 

 extreme care, will occasionally occur ; and even if a rever- 

 sion had occurred there would be no way of proving that 

 it was such and not contamination. In fact, eosin first 

 appeared in white-eyed stock and seemed to arise through 

 reversion, but at the same time it seemed so improbable 

 that this could happen that I tried to account for its ap- 

 pearance in a roundabout way. Now I should say that 

 the factor w reverted to W. 



