SEASONAL CHANGES AND HISTORIES. 149 



would continually feed the aestival ; yet it would 

 suffer no greater loss than by the practical steril- 

 ity of the September butterflies. Were the sea- 

 son still longer, the vernal series would become 

 double-brooded, the caterpillars from the Sep- 

 tember butterflies having time to attain half then- 

 size before hibernation ; the aestival series, on the 

 other hand, would probably by this time have 

 assumed the position our vernal series occupied 

 at the outset, and adopted and monopolized the 

 lethargic propensities of the latter. 



The preceding hypothesis was suggested sev- 

 eral years since to explain the probable effect of 

 climate on generation in butterflies ; it was not 

 deemed capable of proof, inasmuch as the butter- 

 fly does not naturally live in a climate allowing 

 of such changes. But it has received a remark- 

 able confirmation in some forced experiments of 

 Mr. Edwards of West Virginia, who has pointed 

 out that caterpillars of this species carried from 

 the north to West Virginia and raised in the south 

 acted in precisely the way that had been sug- 

 gested ;* they all went through their transforma- 



* These observations of Mr. Edwards were published in the 

 Canadian Entomologist (volume vii. p. 189 seq. ; see also vol. 

 viii. p. 161), and were not unreasonably supposed by him and 

 others to be opposed to the account of the insect I had previously 

 given. They may be so, but the author neglected to state that 

 the caterpillars raised by him were reared in West Virginia 

 south of the natural range of the insect (as in response to my 



