METAMORPHOSIS. 81 



gon-fly quits simultaneously with the harder one, being still 

 retained by the May-fly. Here then we have the strange 

 fact of an insect's flying before it reaches the imago ; that 

 is, flying in its penultimate state. In twenty or thirty mi- 

 nutes at farthest it settles again, casts its skin, and becomes 

 a perfect imago. 



It thus appears that although, until the final ecdysis, no 

 insect arrives at perfection, yet, before that period, even in 

 the state immediately preceding, it may feed, run, and even 

 fly ; or may swim, crawl, barely move, or be without mo- 

 tion, without apparent life, or without apparent organiza- 

 tion. It appears that the apparently lifeless or quiescent 

 state may be entered without ecdysis ; that ecdysis itself 

 may be either single or double ; that the states called pu- 

 pa, in various tribes, are neither substantially nor numeri- 

 cally the same. That comparing the few insects herein 

 noticed, the fly, the bee, the cricket, the dragon-fly and the 

 May -fly, all of which represent great orders, we shall find 

 it perfectly impossible to apply, if we aim at precision, any 

 other than a numerical denomination to their intermediate 

 states ; and finally, therefore, that insects, like higher ani- 

 mals, have but three eras of existence, the foetal, the 

 adolescent, and the adult. 



As to the number of times ecdysis takes place in the life 

 of an insect, little can be said at present owing to the care- 

 lessness and imperfection of our researches ; and on this 

 account it will be found safer to count downwards from the 

 imago, than upwards from the egg. Although the contrary 

 has been asserted, and perhaps generally believed, it yet 

 remains to be proved, that the grubs of Diptera and acule- 

 leate Hymenoptera undergo any ecdysis until full grown. 

 The order Tenthredinites, on the contrary, and the Lepid- 

 optera, change very frequently, with some exceptions ; for 



