METAMORPHOSIS. 215 



largely dependent upon, external conditions of tem- 

 perature, season, and climate. 



In order to explain the origin of this complicated 

 method of development, we must look to the lower 

 and ancestral forms of insect life. In the gradual 

 evolution of the hereditary impulse, as I have de- 

 scribed it, we cannot suppose the ancestral line of 

 the butterfly to have come by a simple and direct 

 manner to its present complicated method of devel- 

 opment. The simple manner of development of the 

 hereditary impulse is by constant and similar repe- 

 tition in each cycle of life with the resulting associa- 

 tions, the adding of new slight characteristics at the 

 end of each cycle, the more rapid repetition of that 

 earlier part of the cycle which has been repeated 

 most often, and thus the backward transference of 

 the newly acquired characters. The result of such 

 a simple manner of evolution of the hereditary 

 impulse, would be that the earliest part of the devel- 

 opment of an individual would be most rapid and 

 invariable, and the later part would become slower 

 and slower as it progresses toward the end of its 

 course, and would be also less certain. There could 

 be no period of slow and variable development 

 succeeded by a period of decisive and invariable 

 development. The lower orders of insects, we find, 

 follow the simpler method of development, — the 



