194 BRITISH BUTTERFLIES. 



by eating a large, round hole in the centre of its upper surface, leaving 

 the rest of the shell untouched, and, for some days, in feeding, eats 

 only into the substance of a leaf of Lotus, Ornithopus, etc., either 

 from the upper- or the underside, leaving the opposite skin as a small, 

 pale, transparent, whitish spot, or by devouring the flowers which it 

 can eat entirely ; its colour matches well with the dull tint of the 

 bluish-green Lotus leaf ; later, the transparent blotches become much 

 more conspicuous (Hellins and Buckler). Chapman observes (in litt.) 

 that, when newly-hatched, the larva eats a small hole in the leaf of 

 the foodplant, thrusting in its head and clearing out the parenchyma, 

 like a Coleophorid larva, just in the same way as do the young larvae 

 of Agriades coridon and other Lycaenids. He further adds that they 

 differ, however, from the Coleophorid larva, in that, at this stage, 

 they make a very small hole and reach the extremities of the cavity 

 they dig out by means of the extensile neck, and not as in Coleophora 

 by taking in the forward segments of the body. This seems to be true 

 of those species that mine leaves at this stage, and that do not appear 

 to like the cuticle as food, it is not so, however, with the larvae of 

 such species as Callophrys rubi that burrow into flowers and make a 

 hole big enough to admit the thorax. A first stage larva of Agriades 

 coridon, Polyommatus icarus, or Strymon pruni on a slide is apt to 

 display this extensile neck, and, when it does so, has a very distorted 

 appearance. When well-fed in this instar, the larva of P. icarus assumes 

 a certain amount of colour and marking, although its tint is still pale 

 ochreous, and the markings are in the form, of dark brown mottling, 

 making a sort of band down the middle of the dorsum, two between 

 dorsal ridge and spiracles, and one subspiracular. The larva does not 

 appear to eat its cast skin at the end of the first moult as appears to be 

 the custom with that of A. coridon. Some newly-hatched larvae placed 

 on flowers of Lotus corniculatus, presented two methods of feeding 

 according to the sizes of the flowers attacked ; on those that are 

 rather buds, they merely bore a hole and introduce the head and 

 part of the thorax, and feed on what can be so reached ; on what are 

 almost flowers, they enter by boring a hole and getting quite inside 

 and living on the stamens and other interior parts ; the small larvae 

 do not seem to like fully-opened flowers. The difference of procedure 

 is, no doubt, due to two equally cogent facts ; in the bud, there is no 

 room for the larva inside without eating all the material en route, and 

 this material is also young and nutritive, in the larger bud, there is 

 room inside, and the outer envelopes, being more mature, is not so 

 succulent and nutritive. Chapman further notes that, after the 

 moult, at the end of the 3rd instar, the larvae look so small that 

 one suspects another moult must take place, but they feed up rapidly 

 and grow very fast in this stage, and soon take on a more mature 

 appearance. Some spring (hybernating) larvae fed on Lotus cornicu- 

 latus up to the final instar, were then given flowers of I le.v europaeus, 

 a diet they seem to prefer greatly to the leaves of L. corniculatus. 

 There is, of course, a very great deal of difference in the habits of the 

 summer and over-wintering larvae of this species, the former some- 

 times feeding up in about six or eight weeks, the latter feeding slowly 

 until in the 3rd instar, when hybernation takes place, and continues 

 till about the end of February or early March. Chapman observed 

 some kept out of doors all the winter (under glass, but without heat), 



