614 THE BUTTERFLIES OF XEW ENGLAND. 



Desert Island (Scudder) and a})})ears to be common at Orono (Fernald), 

 as elsewhere in Maine. I have taken it on the summit of Mt. AVashing- 

 ton. In the White Mountains and about Boston it is equally common, 

 and besides I have it reported, or have taken it myself, at Norway, Water- 

 ville, Brunswick and Hallowell, Me., Milford, Thornton, Dublin, Suncook 

 and Plvmouth, N. H., Stow, Vt., Walpole, Montague, Amherst, Spring- 

 field, Belchertown, Andover, Worcester, Princeton, and Williamstown, 

 Mass., and Farmington and New Haven, Conn. 



Haunts. This butterfly is more or less local in its haunts, being found 

 only about wet meadows and bogs, where it frequents the mint blossoms, 

 according to Mr. Lintner, flies with moderate rapidity, fluttering in a 

 waving, zigzag course about two or three feet abo^'e the ground, and is 

 easily captured. 



Food-plant. The caterpillar appears to feed indiscriminately upon 

 wild and cultivated Violaceae. Mr. Dimmock writes, ''specimens which I 

 bred fed upon either ordinary turf-grass or white closer, both of which I 

 had in the box." 



Habits of the caterpillar. At eclosion the caterpillar invariably 

 gnaws its way through the upper side of the egg, afterward de^•ouring 

 either the whole or a [)ortion of the deserted shell ; its habits closely re- 

 semble, in ftict cannot be d'itinguished from, those of its congener, 

 B. myrina, including the phenomena of lethargy. 



Life history. As mentioned under the genus Brenthis, I was formerly 

 constrained to construct an hypothesis concerning the periods of this in- 

 sect (American naturalist, vi : 513-518) which seems no longer tenable. 

 I then supposed that instead of confining itself to the uniform cycle of 

 changes common to most butterflies, this insect exhibited two sets of in- 

 dividuals, each with its own distinct alternations, so that in tracing the 

 history each series would need to be treated as if it were a distinct species. 

 Later research, however, — the observations of others as well as my own, 

 extending over several years, — shows that the behavior of this insect, 

 though striking enough, does not so essentially differ from that of ordinary 

 l)utterflies. 



It is a somewhat earlier l^utterfly than its congener B. myrina. The 

 first brood of butterflies is heralded by a few individuals between the 

 4th and 15th of May, but it rarely becomes common before the end of 

 May, although Mr. Lintner once found it not rare in Scoharie, N. Y., on 

 the 13th. Occasionally it has escaped the notice of observers until toward 

 the end of May, but this may be due to the observer's not visiting the 

 proper haunts of the species previous to the date noted ; and as in certain 

 years its ad^'cnt is likely to be somewhat retarded, he might easily su})- 

 pose that the few individuals he saw in the latter half of May were the 

 first tliat had appeared. But it appears as early as the middle of May in 



