No. 438.] 



AMBL YSTOMA TIGRINUM. 



399 



tive changes take place during metamorphosis ; but the changes, 

 as a whole, are plainly destructive. The beginnings of the 

 process, the resorption of tin-told and gills, are plainly processes 

 of sell -digestion of peripheral parts. I have seen cases ot early 

 enforced metamorphosis, in which it seemed that this destructive 

 action attacked the periphery, not only in fin and gills, but in 

 the legs as well. Now, without for a moment attempting to 

 explain, fundamentally, the nature of metamorphosis, it is thought, 

 none the less, that the description just given makes obvious its 

 intimate relation with nutrition. Looking at the matter thus 

 from the standpoint of physiological facts, and forgetting for the 

 nonce our teleology both "old " and " new," it becomes probable, 

 a priori, that quick starvation will be more effective than enforced 

 air breathing in causing the flabby larva to digest its loose and 

 vascular tissues and take on the more compact form of the adult. 

 I may now summarize, briefly, sufficient evidence to support the 

 conclusion that this is the case. 



All of the numerous instances observed of extreme accelera- 

 tion of metamorphosis have been obviously the result of starva- 

 tion. The three-gram adult, spoken of at the beginning of this 

 paper, was the result of a small larva, accidentally overlooked, 

 and left for several weeks without food, in a jar of water. The 

 excessively early metamorphosis of nearly fifty specimens in the 

 cool water of the fountain is also a case in point, the three or 

 four larva? that did not metamorphose made little if any growth 



supply. Metamorphosis in the tank of cold spring water was 

 likewise attributable to this cause ; several of the specimens 

 were under the size at which metamorphosis usually takes 

 place ; but there was no visible food supply ; and the adults 

 that resulted showed very evident emaciation. Many essentially 



