4.6 



THE AMERICAN NATURALIST 



[Vol. LII 



lated the rule, that when two forms coexist in the same 

 locality and are able to interbreed, they do not produce 

 intermediates; but when the forms are geographically 

 separated as local races, crosses between them result in 

 a series of intermediates. Bateson says: "In this apho- 

 rism there is a good deal of truth." "We think that the 

 rule expresses the difference between a non-adaptational 

 chance mutation and the adaptational change in the fac- 

 torial quantities which may lead to a similar-looking, but 

 physiologically different character. This character, al- 

 though, like the non-adaptational one, is itself of no 

 selective value, is the result of a general physiological 

 change which does have, a selective value. 



This will become still more evident if we return once 

 more to the study of the gypsy-moth. In studying the 

 relations of the different geographic races as character- 

 ized by the multiple-allelomorphic characters in question, 

 we find that these characters are paralleled closely by dif- 

 ferences in the life-cycles. Without going into details, 

 we may state as a fairly general rule that the races with 

 high degrees of pigmentation in the later stages are the 

 ones which show a fast development, comparatively short 

 larval life and a long period of hibernation. The light 

 races have a comparatively long larval period and a cor- 

 respondingly short period of hibernation. The former 

 races, furthermore, inhabit the areas where a long and 

 cold winter occurs, while the latter are endemic in places 

 which have a hot summer, early spring and mild winter. 

 One might think that these different characteristics were 

 simply the direct effect of temperature conditions. But 

 that this is not the case is shown by the constancy of the 

 differences when the races are bred in a different climate 

 and also by experiments on the physiology of hibernation, 

 which have convinced us that the time relations of the 

 life-cycle are— of course, within the limits of fluctuation— 

 a heritable trait of rhythmic character. These facts show 

 where the adaptational character of the differences of the 

 geographic races lies : the adaptation which fits the differ- 



