SOJIK NOTES ON THE PUl'AL MOULT OF LEl'IDOl'TERA. 117 



Of this, however, there can be no doubt, that the pupa is black, that it 

 is not bhick by any intensity of the oi-dinai-y brown colour of chitin, 

 but by the presence of a special pigment, and that it must be so with 

 some special useful object in view. There can, further, be little doubt 

 that the larval pigment we arc considering is similar to that of the 

 pupa, and is connected with it in some way. We n)ay be tolerably 

 certain of one other fact, r/c, that the material for this pigment is present 

 in the superficial layer of the chitinous covering of the pupa at 

 moulting, whilst it is still green and soft, and that the pigment is 

 formed from it by some chemical change on exposure to air and light, 

 probably by oxidation. This material for the formation of pigment is 

 probably formed from some constituents of the epidermic cells that 

 break down to liberate the effete superficial layer from the dermis 

 beneath, and is probably at once absorbed by the new layer of epidermal 

 cells forming on the surface of the dermis. I have just used the word 

 " probably " a good many times, but the chemistry and physiology of 

 the matter is so likely to be something similar to that I have sketched, 

 that one "probably," if I could have properly spread it, would have, I 

 think, met the case. Assuming this to be so, I think we may be able to 

 explain the occurrence of these black patches. 



During the larval life, the wing lies in an invaginated pocket of the 

 dermis, but at the date of the change to pupa it does not do so, but 

 lies immediately beneath the effete skin that is about to be thrown off", 

 and, therefore, one of the changes that occur at the end of larval life is 

 the disappearance of the double fold of dermis, between the epidermis 

 and the wing. Now, when the outer layer of this frees itself from the 

 epidermis, it no doubt does so by the same process as that which develops 

 pigmentary material elsewhere, but here the dermis below has itself also 

 to disappear, so that this material is not at once appropriated by the 

 proper new cells beneath, and there is a certain excess of it, so to speak, 

 free. This then permeates the effete cuticle, reaches the surface, and under- 

 goes the oxidising, or other process, that converts it into pigment. There 

 is no other portion of the larva where pigmentary material might thus 

 be set free. But in this situation it is difiicult to see how it could be 

 avoided, without a different physiological process having been evolved 

 for this little area. If we knew the precise date at which the wing 

 became disencapsuled, we could better judge as to the probability of 

 this being the correct hypothesis. It would also throw light on the 

 subject to know if other instances of the same coloration of the larva 

 were confined to cases in which the pupa has pigmental colouring, and 

 does not occur where the ordinary chitinous brown only is the pupal 

 colouring. 



In S. libatrir, and other transparent larva), certain portions of the 

 pupa are seen, before moulting takes place, to have already assumed a 

 brownish chitinous colour, and, in many pupje, innuediately after the 

 moult, there are portions already brownish, whilst the greater part of 

 the pupa is green and soft. In S. lUnttiir, before the moult, a slight 

 ruddy tint is observable over the prothorax, and a pink spot 

 appears on the second thoracic segment, just behind the first spiracle. 

 During, and just after, the moult, the colouring of the prothorax is 

 seen to affect precisely that part of it that is exposed in the pupa, 

 whilst a slight tinting exists over both the meso- and nietalhonix. 

 The pink spot is seen to be the posterior lip of the thoracic spiracle, as 



