HEREDITY. 17 



lately given so much attention tb those cases in which it is 

 claimed, that selection could modify the quality of genes. 



Castle especialy professed a belief in the power of selection 

 to change the nature of the genes. The controversy between 

 Castle on one side, and Johannsen, Eas^ and ourselves on the 

 other, does not concern a minor point, but it touches the very 

 nature of the genes themselves. If it were true what Castle 

 claimed, that selection can shift the influence one identical gene 

 exerts over one identical developmentcil process, we would be 

 no further in respect to an insight into the nature of the inher- 

 ited than the authors of the vitalistic theories. It will be 

 remembered, that one phase of the controversy concerned the 

 effect of selection on the extent of pigmentation on Hooded 

 rats. From the fact that all Hooded rats have one gene less than 

 solid-coloured animals. Castle drew the conclusion, that Hood- 

 edness was a imit-character, but he further concluded from the 

 same fact, that all Hooded rats were geno-tjrpically identical, 

 and that therefore the proof that selection modifies the hood- 

 edness, the extent of pigmentation, also proved that a gene 

 had been modified in its quality. 



To fully understand the case, it must be i;emembered, that 

 hoodedness is recessive to solid colour, in other words that the 

 Hooded rats are identical in one respect only, namely in the 

 non-possession of one definite gene. It should be easy to under- 

 stand, that the absence of such a gene does not imply purity 

 for possible other genes which may influence the extent of 

 pigmentation, but that the difference between animals with 

 and without each of these genes need not be as marked as that 

 between hooded and solid-coloured ones. It could very well 

 have been impossible to demonstrate any of the possible genes, 

 which make the difference between dark-hooded and light-hood- 

 ed rats. As we did succed in demonstrating one of such genes, 

 it may seem strange to allude to the possibility which confront- 

 ed us at the start, of not being able to demonstrate any of the 

 genes overlooked by Castle. But, it is obvious, that it is a thank- 

 less task, to repeat another experimentator's work over and 



2 



