300 BRITISH BUTTERFLIES. 



but projects from it at an angle of, perhaps, 45°. A gentle push 

 will remove the pupa entirely from its larval skin, the latter adher- 

 ing firmly to its moorings, in most instances. Newman states that larvae 

 in his possession changed to pupae of a pale, semitransparent, brown 

 colour, unattached either by a belt or by anal hooks, at the bottom of 

 the receptacle in which the larvae had been fed. Bowles observes that 

 the fullfed larvae usually spins a loose web, and thus encloses itself, 

 changing to a pupa, which is often unfixed, and at other times making 

 a slight anal connection. Pabst says that the fullgrown larva 

 leaves the bush or tree on which it has been feeding and pupates 

 upon the ground among dry leaves and grass, the pupal period lasting 

 three weeks. Rossler says that pupation takes place on the ground 

 under leaves, etc. Gillmer states (in litt.) that Heissler informed 

 him that " the larva fastens itself with a slight thread to a leaf 

 or twig, but so weak is the thread that it breaks when the pupa is 

 touched." He adds : " This statement, made by Heissler, disagreeing 

 with that of Bossier, led me to obtain a supplementary account, in 

 which Heissler states that ' the larva spins only a very weak girth, the 

 larval skin in very many cases remaining clinging to the anal end of the 

 pupa.' " Steinert states that " the pupa has the cast skin covering its anal 

 end, but there is a terminal bunch of rust-coloured hairs ; the skin, how- 

 ever, may be separated from the pupal end without harm resulting, 

 and the loose pupa will produce a perfectly-formed imago. The pupa 

 is perfectly incapable of movement, and one can only tell that it is 

 alive by its change of colour before emergence." L. Newman states 

 (in litt.) that the larvae go to ground to pupate, and do so under 

 moss or dead leaves, spinning the faintest possible semblance of a 

 silk pad, with a few frail silken threads that are hardly a belt ; the 

 pupa, however, obtains little support from these, and falls away from 

 them at the slightest disturbance. Harwood says (in litt.) that the 

 larvae do very little spinning, but usually manage to attach themselves 

 to the underside of a leaf to undergo transformation. Having carefully 

 examined the larvae attached to the underside of leaves ready for 

 changing, and also some freshly-changed pupae, Harwood adds that 

 he could find no trace of a silken girdle, nor is it easy to 

 make out how they attach themselves, for the amount of silk they 

 spin appears to be insufficient, but one observes that, in the case 

 of the pupa, the cast larval skins remain attached to the anal ex- 

 tremity. Head adds (in litt.) that the larvae spin a slight web on a 

 leaf, and pass a fine thread or two over their back ; they prefer to get 

 into a curled- up leaf, and, in fact, appear sometimes to draw the leaf 

 slightly round them, but they spin very little silk, and the pupal con- 

 nection with the spinning is of the very slightest. Kaynor writes (in 

 litt.) : " Twenty-two larvae of H. betulae received from Monk's Wood, 

 in June, 1907, I placed in two cardboard boxes, each one foot high. 

 Nineteen of the larvae formed loose surface cocoons, fastened with 

 slight silken strands, and partly open at the top, placed on the 

 surface of some oak leaf-mould. The other three, however, 

 attached themselves to the surface of the cardboard boxes towards 

 the top — one being actually in the northeast corner, 12 ins. from 

 the ground, the other two about 10 ins. up. I examined these 

 three pupae carefully with a strong lens, and quite failed to detect any 

 trace of a silken girth or fastening of any kind. The pupae seemed to 



