202 BRITISH BUTTERFLIES. 



as the cocoon obscures matters, and to break the cocoon disturbs the 

 whole arrangement. In one instance, nothing to represent a girth 

 can be found ; in the others, quite apart from the cables forming a 

 cocoon, are certain threads crossing over at the posterior border of the 

 mesothorax, irregular in quantity and varying in their fastenings, but 

 unquestionably representing a girth. One specimen "spun up" on 

 the lid of the box, and had on pupating nothing to support it but the 

 " cocoon," which nevertheless' holds it, somewhat tremulously, in 

 place. This cocoon is, to all intents and purposes, a girth. It has 

 rather irregular and diffuse fastenings at each end, but across the larval 

 dorsum there pass, from a place about opposite the wing spines, a couple 

 of threads across the front of the mesothorax, and another across it 

 further back, or really for the most part on the metathorax, and two 

 more over the 1st abdominal ; these are all, however, connected directly 

 with the same origin beside the middle of the mesothorax. There is 

 another thread not belonging to this set, and therefore clearly merely 

 " cocoon " across the 3rd abdominal segment. Since there are no 

 cremastral hooks, that is, since the hooks have became obsolete too, 

 we must assume that the girth is not in process of development, but of 

 degeneration, or rather, perhaps, has developed into a cocoon, without 

 having quite forgotten its original form and purpose. After pupation, 

 a removal of the leaves above the pupae, left them quite free, except 

 that one pupa has a thread of silk, by means of which it can be lifted, 

 attached to the meso-metathoracic incision (Chapman). Hellins notes 

 that his fullfed larvae did not spin any girdle for pupation, although 

 they went to the cover of their cage for that change, and he adds that, 

 when it took place, the larval skin remained fixed, whilst the pupa 

 had fallen down. Bayward notes (in litt.) that, when full-grown, the 

 larva generally draws together three or four leaves of the foodplant 

 by spinning round them a few threads of silk, and in this exceedingly 

 slight cocoon undergoes its transformation to the pupa. Of about a 

 dozen larvae reared in confinement in the spring of 1907, all but one 

 chose this position for pupation, and the single exception spun a 

 slight pad of silk on the side of the bottle containing the foodplant, 

 and supported its thoracic segments by a girth composed of some 

 three or four silken strands attached at some distance from each other 

 to the glass side of the bottle, but coming close together over the 

 thorax ; in this horizontal position, the larva pupated safely, and the 

 imago emerged in due course, leaving the empty pupa-case still 

 attached to the side of the vessel. 



Pupa. — The newly-formed pupa has the wings very transparent, 

 and all the tracheae of the nervures of both wings very distinct ; there 

 are fine hairs all over the dorsum ; it has no sign of cremastral hooks, 

 the end of the pupa being smooth and rounded. The pupa is pale 

 green, lighter and transparent over the thorax and wings, rather 

 denser and slightly olive over the abdomen ; the head and terminal 

 segments slightly flesh-tinted ; no marks, except a darker dorsal line 

 (really the dorsal vessel) down the abdomen ; the skin is apparently 

 glabrous, but under a microscope shows a good many very fine hairs 

 (dark) 0-06mm. long and less, and a few finely stellate hairs, which 

 are little more than half the length of the ordinary hairs. The general 

 surface has a reticulation of fine raised lines, wider apart than in the 

 Theclids, running more frequently longitudinally and without any 



