THEORIES OF EVOLUTION 119 



The results of anything but differential death 

 among these alternatives savor much more strongly 

 of isolation than of selection. Wagner,^ in his 

 elaborate presentation of the isolation theory, 

 emphasized geographical isolation as a factor in 

 the evolution of species, but it is evident that 

 other things than space may isolate organisms. I 

 have cited elsewhere the case of Plebeius melissa, 

 one of the small Lycaenid butterflies. The males 

 appeared in Dickinson County, Iowa, during the 

 period of my observations, six weeks earlier than 

 the females. Males were abundant in July and 

 were present during the last two weeks of June. 

 Females appeared in small numbers toward the 

 end of July and were abundant in August. This 

 ease involves a seasonal isolation which would 

 tend to eliminate any characters possessed by the 

 earlier males exclusively. The insects are so deli- 

 cate that they must be destroyed accidentally in 

 large numbers, and it is doubtful that even a mod- 

 erate percentage of the very early males have an 

 opportunity to mate. 



Great variation in size also has an isolating effect, 

 since extremely large and extremely small indi- 

 viduals of opposite sex may be unable to mate. 

 This is obviously true in the case of some insects 

 in which the abundance of food is reflected in the 

 size of the adults, and is conspicuously so in such 



^ Entstehung der Arten durch rdumliche Sonderung, 1889. 



