110 



species the spring form, or something resembling it ; and spec- 

 imens of this form were not rarely accompanied by specimens 

 such as are called suffused when they occur naturally. Mr. Ed- 

 wards suggests that the chrysalids producing such specimens in 

 nature may have been subjected to such extreme cold, over a 

 long period, as would be produced by a snow or ice envelope, — 

 a suggestion which is rendered more probable from the fact 

 that these examples of suffusion are most common in cold or 

 mountainous countries, and, so far as I am aware, are unknown 

 in the tropics. 



In this connection we may again refer to Mr. Edwards' ex- 

 quisite illustrations, where Jie discusses the melanism of the 

 female of Papilio turnus. He notices that the dark form of 

 the female is restricted to districts where a summer generation 

 is possible, and is thus inclined to think the cause of its original 

 appearance " in some way climatal," and brings forward some 

 evidence against Dr. A. Weismann's theory that the dark 

 female has superior attractions for the male. Mr. Edwards 

 does not find the dark form more prevalent in one brood than 

 in another, and has obtained dark females from a yellow 

 mother and vice versd ; not however with the same frequency, 

 having found only one instance of the former, which, as the 

 male is always yellow, he does not think surprising ; but that 

 more yellow females are not produced Irom dark mothers — he 

 has reared two out of twenty-three — indicates, he believes, an 

 amazing energy in the dark form, and implies a time when the 

 yellow female will wholly succumb to the other, throughout 

 the region now inhabited by both. He would further explain 

 the probable greater abundance of dark forms in the west than 

 in the south, by the larger proportion of insectivorous birds in 

 the prairie region, and the association of this butterfly with a 

 larger number of yellow-colored swallow-tails in the south. 



This insect, possessing two forms of female, one differing 

 from the male and the other resembling it, is only one example 

 of a large class, which I have recently discussed, attempting to 

 show, in the first instance, that sexual dimorphism, or antigeny, 

 is of two kinds, colorational and structural : the first prevailing 

 in females, the latter in males. Colorational antigeny is divisi- 



