1384 THE BUTTERFLIES OF NEW ENGLAND. 



only two generations, are plural brooded in the southern states ; and so far 

 as our still very insutticient information goes, it tends to support the state- 

 ment that in general a brood is added in each great faunal belt, the change 

 occurring not far from the delimiting lines between the faunas. This is 

 perhaps most marked in the Paniphilidi, which in the Canadian fauna are 

 almost invariably niouogoneutic, and seem in the Carolinian to be as 

 commonly trigoneutic. 



Now the effect of temperature is seen not only in the histories of our but- 

 terflies, but in many cases in their very structure ; the first brood of a 

 given buttei-fly for any year differing from the subsequent brood or broods, 

 sometimes to a marked degree: so much so, that in not a few cases they 

 have been first described as distinct species. These differences are in the 

 main confined to the color and patterns of the wings, but are sometimes 

 seen in the size and form of the wings themselves, in the abdominal ap- 

 pendages of the males and in one instance at least in the clothing of the 

 head. Where there are several broods in the year, no differences what- 

 ever have been detected between the members of the broods subsequent to 

 the first ; it is only the first which differs from all the others alike, showing 

 pretty clearly that it is cold acting upon the creature in the preparatory 

 stages of its life, and not heat, which is the prime agent in this class of 

 distinctions ; in support of this we also find in a few cases differences be- 

 tween the earlier and later appearing members of the first brood, the later 

 members showing an approach toward the summer forms. 



This would seem to be a strong objection to the theory advanced by 

 Weismann that the winter-form so called, that is, the first brood of 

 seasonally dimor|)hic butterflies, is the primeval form of the species. 

 Considering the \'a8t preponderance of butterfly varietv in the south and 

 the almost universally intimate alliance of the northern butterfly fauna to 

 the southern, there can be no doubt that the butterfly is par excellence a. 

 creature of the tropics and the northern fauna in no proper sense endemic, 

 but a colonial offshoot from the southern. It would seem to follow that 

 the seasonal dimorphism of the north (imitated, so to speak, in the south 

 in the dry and wet season butterflies, as de Niceville has shown) is a 

 phenomenon superinduced in the colonist by the new conditions in which 

 it finds itself; and the summer type should therefore be looked on as the 

 normal and primeval. 



It is a significant fact that in most, perhaps all, butterflies which show 

 marked seasonal dimoi-phism, the hiemal condition is that of the chrysalis ; 

 for it points directly to the conclusion, borne out by experiments in the 

 laboratory', that the divergence of the spring from the summer type is the 

 effect of cold upon the insect in the chr3'salis state. These experiments 

 have not been so extended nor so complete as not to leave the matter still 

 open to investigation before a final conclusion may be warranted : but 



