808 THE BUTTERFLIKS OF NEW ENGLAND. 



selves within their hiirrow-likc nests? May we not look iijton tlie distinc- 

 tive clothini( wliioh juvenile and mature caterpillars of buttci-flics possess, 

 features common to all tiie group, as the first step toward a further stage of 

 metaniori)liosis — in short a iiypcrnietauiorpiiosis, akin in nature, but not 

 yet in degree, to that so well known to entomologists in the early stages 

 of many grubs of the Coleopterous family Meloidae ? 



Tiiat this peculiar investiture and form of caterpillars in their final em- 

 bryonic stage, as they leave the egg, is adaptive, a form of protection 

 against the dangers that will beset them when they leave the egg is I think 

 tolerably clear. In the first place caterpillars having in their later life 

 special provision against enemies are practically unprovided with that class of 

 safeguards in their earliest stage ; thus I have never been able to induce any 

 caterpillars of the Papilioninae while in their first stage to protrude the 

 osraateria, with which nevertheless they are provided from birth and which 

 they use freely in after stages. Again, the single feature common to the 

 appendages of the vast number of embryonic larvae is that these are appar- 

 ently hollow tubes, having presumed connection with basal glands, and 

 thus serving as channels for the conveyance of fluids, probably protective, 

 from such glands ; for tiiese appendages nearly always flare at the extrem- 

 ity, being more or less trumpet shaped, and often can be seen to hold a 

 droplet of fluid within the embrace of the lips of the trumpet. 



It is immaterial to our present purpose whether the embryonic or the 

 mature stage of investiture and form in caterpillar life be looked upon as 

 the elder ; we only wish here to urge that a hypermetamorphosis is now 

 in the process of origination through tlie differentiation of the larval char- 

 acteristics, of the same kind and following the same direction as that 

 which has resulted in metamorjihosis from the earlier uniform conditions 

 of structure and life to the later. Considering, however, the comparative 

 uniformity of the embryonic and the diversity of the adidt characters, with 

 the naturally more widely varied conditions of life to wliich the adult must 

 be subjected, it may be regarded as probable that the hypermetamorphosis 

 is rather in the direction of the evolution of the later ontologic stages. 

 In further support of this view is the fact that the peculiar protective 

 structures of the first stage — the trumpet-tipped dermal appendages — 

 endure throughout life in the lowest family, the Hesperidae, possibly also 

 in some Picrinae, are lost at various stages in other Pierinae and in the 

 Papilioninae, and never surpass the first stage in the highest family, the 

 Nymphalidae, and probably do not in the Lycaenidae. 



That in other insects also the first stage of larval life is thus diflPerentiated 

 from the remaining stages I have elsewhere pointed out. Presumably this 

 form of hyijermetamorphosls will be found to exist in nearly if not all larvae 

 belonging to groups, tlie most of whose members live in exposed situations, 

 and wliich have anything more than the most "incomplete" metamorphosis. 



