On the Seasonal Dimorphism of Butterflies. 85 



It is not here my intention to enter into the 

 ultimate causes of metagenesis ; in this subject we 

 should only be able to advance by making vague 

 hypotheses. The phenomenon of seasonal dimor- 

 phism, with which this work has mainly to deal, 

 is evidently far removed from metagenesis, and it 

 was to make this clear that the foregoing observa- 

 tions were brought forward. The characters com- 

 mon in the origin of metagenesis are to be found, 

 according to the views previously set forth, in the 

 facts that here the faculty of asexual and of sexual 

 reproduction is always distributed among several 

 phyletic stages of development which succeed each 

 other in an ascending series (progressive meta- 

 genesis of Haeckel), whereas I find differences only 

 in the fact that the power of asexual propagation 

 may (in metagenesis) be either newly acquired 

 (larva of Cecidomyid] or preserved from previous 

 ages (Hydroida). It seems that in this process 

 sexual reproduction is without exception lost by 

 the earlier, and remains confined solely to the 

 most recent stages. 



From the investigations on seasonal dimorphism 

 it appears that a cycle of generations can arise in 

 an entirely different way. In this case a series of 

 generations originally alike are made dissimilar by 

 external influences. This appears to me of the 

 greatest importance, since seasonal dimorphism is 

 without, doubt closely related to that mode of re- 

 production which has hitherto been exclusively 



