526 THE BUTTERFLIES OF NEW ENGLAND. 



jjassion flower). Grote and others have found it on the same. It is also 

 recorded from other Passifiorae, P. caerulea for instance. Mr. Riley has 

 also found it injurious to one of the Violaceae, — the garden pansy, Viola 

 tricolor Linn., and jVIr. Edwards fed specimens readily with violets. In 

 the west Mr. Mead found it on Sedum, one of the Crassulaceae closely 

 allied to the Passifiorae. It is also figured by Abbot on Desmodium pani- 

 culatum, a leguminous plant, and is said to occur on purslane, one of the 

 Portulacaceae. Mr. Riley received it from Norfolk, Virginia, as taken on 

 cabbage, which, however, his informant added, it did not harm, its princi- 

 pal food being the "pop apple" (probably the May-pop, the fruit of Pas- 

 siflora incarnata). Finally Abbot says it feeds on "beggars lice" 

 (Cynoglossum?). PassiHora and Sedum are evidently its favorites. The 

 Cuban species, E. hegesia (P. columbina Fabr.) which is distinct from 

 ours, has been found by Dr. Gundlach on Turnera ulmifolia. 



It is probable, as Mr. Edwards suggests, that the larva feeds by night, 

 and by day resorts to stems of bushes and higher plants, for he has found 

 it on black alder several feet from the ground ; "it travels," Mr. Edwards 

 remarks, "with wonderful rapidity and a daily journey of ten feet would 

 be a small affair." Abbot in several places speaks of the butterfly as 

 common, but the larva as rare (probably because it hides by day). 



Life history. The species is apparently triple-brooded ; the first 

 butterflies appear very early in the spring, the middle of February in 

 Texas (Belfrage), or the last of March in northern Florida (Chapman). 

 Whether these are hibernating individuals or fresh from wintering chrysa- 

 lids is not stated, but in either case they probably belong to the same 

 brood as those which appear late in the preceding autumn. In Georgia, 

 caterpillars are full grown early in May and after about eleven days spent 

 in the chrysalis stage, emerge as butterflies (Abbot) ; apparently these 

 form, properly speaking, the first brood. A second seems to appear about 

 the middle of July in Virginia and Tennessee, when the egg state, according 

 to Edwards, lasts five days (it may be as long as twelve at other seasons) , 

 the caterpillar grows to maturity in a fortnight and the chrysalis hangs a 

 week. A third — the only numerous one — appears in the middle of 

 September, becomes abundant by the first of October, and in the extreme 

 south certainly continues, although in diminished numbers, throughout 

 most of November (Chapman). It is possible, perhaps probable, that in 

 the extreme south, other broods are interpolated between these. 



As to hibernation, it seems probable : 1, that the butterfly often hiber- 

 nates ; 2, that some of the autumn chrysalids do not disclose their inmates 

 until very early the following spring; and 3, that caterpillars hatched 

 from eggs laid by the October butterflies hibernate either as soon as born 

 or partially grown, reviving in the following spring with the earliest vege- 

 tation and developing so rapidly as to transform to the May butterflies. 



