LYCAENINAE: TKNISEfA TARQUINJUS. 1023 



to chrysalis ; so that probahly the supply of plant-lice may have much 

 to do with their later action. If almndant. they will burrow and weave 

 their film as before ; if not, will, as Miss Morton says, "eat roads through" 

 the mass, "but still [devour them] from the under side, their backs cov- 

 ered with wool from the unlucky aphides. 1 think the wool prevents 

 eating from above, for I noticed the larvae eat the red aphides from cherry 

 from the back, or wherever they seized them." 



When the stem supporting one of these colonics of plant-lice with the 

 caterpillars feasting in their midst (75: 4;5) is jarred, the latter, unless 

 closely protected by web, are very apt to curl and fall to the ground, appar- 

 ently by intention. Or they may drop by a thread for several inches and 

 then regain their place, even though they are tolerably large, — a very 

 unusual thing for most butterfly caterpillars, as remarked by Mr. Edwards. 

 Besides the protection this and the web afford, their hairs catch the wool 

 of their devoured victims and renders them, even when moving freely 

 abt)ut, not noticeably distinct from the aphid masses. 



Life history. The localization and consequent comparative rarity of 

 this butterfly with the recent nature of our' knowledge of its habits makes 

 its life history still a little obscure. Probably there are three broods in 

 the north, four in the middle and five in the farthest southern states. 

 The winter is probably passed in chrysalis, but it is not impossible that it 

 may also winter as a butterfly, for at least battered specimens have been 

 taken on the wing very early in the year before the complete unfolding 

 of the leaves, and butterflies have been known to emerge from the chry- 

 salis as late as the last of September. 



The first brood of butterflies of the year appears in New England about 

 the beginning of the last week in May and continues upon the wing until 

 after the middle of June. The second brood appears early in July and 

 becomes abundant by the middle of the month ; search at Granby, Mass. 

 on July 28 brought to light only caterpillars in the second stage ; a fresh 

 specimen and full grown larva were taken by me as far north as Sudbury, 

 Ontario — the northernmost point yet recorded — on July 13, showing that 

 even there the second brood appears at this early date : fresh specimens 

 may still be found during the whole of July and rubbed ones during 

 August ; Mr. Fletcher found a much worn female ovipositing at Ottawa 

 on August 2 and they continue upon the wing until the third brood ap- 

 pears, about the middle of August, from which time fresh specimens con- 

 tinue to come out until near the end of September. 



One of the remarkable things about this extraordinary butterfly is the 

 rapidity of its transformations in the preparatory stages. As Edwards 

 and Miss Morton have shown, the eggs hatch in three or four days, and 

 the caterpillar not only undergoes only three moults, but makes these with 

 extraordinary rapidity, so that in specified instances only thirteen days 



