INTROD UCTION. 17 



flies of Xortli America, such as many species of Polygonia 

 (in F. iiiterrogationis tliey are largely seasonal, the latest 

 brood being all of one type), Phyciodes tharos^the species of 

 Pieris, and especially Jplildides ajax. The latter instance 

 is the more remarkable, because the three forms (marcel- 

 lus, telamonides, and ajax), though sequent in the order 

 named, do not strictly represent distinct broods, since the 

 earlier emerging individuals of the first brood are marcel- 

 lus, the later-appearing individuals of the scDue brood are 

 telamonides, while the subsequent broods, of which there 

 arc several, are ajax. 



Distinct climatal differences, whether temperature or 

 moisture (or both), are unquestionably the prime cause of 

 seasonal dimorphism, the former in temperate, the latter in 

 tropical, regions. The first has been practically proved by 

 experiment, the latter by the correspondence of the ^^he- 

 nomena to that of temperate climates and their synchro- 

 nism with the dry and w^et seasons. 



Many cases of dimorphism are compound. Instances of 

 this have already been given; indeed, most cases of dimor- 

 phism involve some distinct element, such as season or lati- 

 tude, or temperature in some form. Thus, Jasoniades 

 glaucuSy which exhibits dimorphism in the female, does so 

 only in the south, for the dark form of the female (in 

 which the conspicuous normal stripes of the male are ob- 

 scured) occurs but rarely north of Pennsylvania, although 

 there is a distinct tendency in botli sexes to a broadening 

 of the darker markings and the partial suppression of the 

 yellow in high northern latitudes or their equivalent, as 

 among the White Mountains of New Hampshire. A sim- 

 ilar instance occurs in Everes comyntas with the boundary 

 limits of the dark female at about the same place. 



Nearly all the above instances of dimorphism where it is 

 not of the simplest kind (whether seasonal or not) may be 

 termed polymorphic, since more than two types of individ- 



