On the Seasonal Dimorphism of Butterflies. 5 3 



not perhaps be produced by the accelerated de- 

 velopment caused by a high temperature, and the 

 retarded development caused by a low temperature. 

 Other factors belonging to the category of external 

 conditions of life which are included in the term 

 " climate" may be disregarded, as they are of no 

 importance in these cases. The question under 

 consideration is difficult to decide, since, on the one 

 hand, warmth and a short pupal period, and, on 

 the other hand, cold and a long pupal period, are 

 generally inseparably connected with each other ; 

 and without great caution one may easily be led 

 into fallacies, by attributing to the influence of 

 causes now acting that which is but the conse- 

 quence of long inheritance. 



When, in the case of Araschnia Levana, even 

 in very cold summers, Prorsa, but never the Le- 

 vana form, emerges, it would still be erroneous to 

 conclude that it is only the shorter period of de- 

 velopment of the winter generation, and not the 

 summer warmth, which occasioned the formation 

 of the Prorsa type. This new form of the species 

 did not come suddenly into existence, but (as ap- 

 pears sufficiently from the foregoing experiments) 

 originated in the course of many generations, during 

 which summer warmth and a short development 

 period were generally associated together. From 

 the fact that the winter generation always produces 

 Levana, even when the pupae have not been ex- 

 posed to cold but kept in a room, it would be 



