HYPERMETAMORPIIOSIS IN BUTTERFLIES. 807 



Iiistimec's have tliiis been given in every one of tlie larger gronps of 

 buttcrHics to show the universality of this feature in the development of 

 the caterpillar ; many of the changes are gradual in their appearance, and 

 grow more complicated with advancing life ; but the more important 

 changes between the different stages of a caterpillar's life occur at the first 

 moulting; that is, those features of the young caterpillar not possessed by 

 the adult are those which it has brought with it from the egg, and which 

 are lost when its embryonic skin is cast. We should, therefore, naturally 

 suppose these peculiarities to have some reference to its condition in the 

 egg ; but this view can hardly be maintained, for certainly the most appro- 

 priate condition for a caterpillar in the egg would be entire absence of cloth- 

 ing or a uniform covering of silky hairs, conditions which are exactly the 

 ones which do not occur. On the contrary, in every instance we can find, 

 caterpillars which at maturity are naked or nearly so, or clothed uniforndy 

 with hair, when newly hatched, bristle with tubercles or are supplied with 

 cumbrous serrated or spiculiferous hairs. Some other explanation nuist 

 therefore be sought. 



From the great diversity in the character of the metamorphoses of in- 

 sects and the nature of the differences between them, it has been argued 

 with great force and reasonableness, and the argument is now generally 

 accepted by naturalists, that all of these metamorphoses and especially 

 such complete changes as are undergone by a butterfly during its varied 

 life from egg onward, are acquii'ed characteristics, gradually gained in the 

 struggle for existence by adaptive devices ; that is to say, there were 

 perfect insects before there were caterpillars and chrysalids, or indeed 

 larvae and pupae prof)erly so called of any kind. This is not the place to 

 discuss or even to state more fully the grounds for this conclusion. But 

 granted its correctness, we might reasonably look among insects in which 

 metamorphosis had been carried furthest for some indications of its still 

 greater development. Now inasmuch as the chrysalis is as complete an 

 example of adaptation to its purpose as could well be conceived, where be- 

 tween the two ends of the scale could be interpolated any further meta- 

 morphosis than in the growing and feeding caterpillar with its frequent 

 changes of skin ? And if metamorphoses were originally acquired to fit 

 the creature for the contrasted circumstances of its earlier and later life, can 

 we be far wrong in assuming that the conditions of existence are very 

 different to a young creature just from the egg, so small as hardly to be 

 seen, and its full grown self one hundred fold as large, able to pass from 

 an exposed to a concealed position rapidly? If not, why are webs and 

 other manufactured concealments so much more common in early than in 

 later life? And why, with the Pamphilidi, which live in concealment all 

 their lives, are only the younger stages provided with long, recurved 

 bristles on the terminal segments, that they may rapidly withdraw them- 



