THE EXTERNAL MORPHOLOGY OP THE LEPIDOPTEROUS PUPA. 63 



lateral movement being possible. In Aporiinae, the 5th and 6th ab- 

 dominal segments are movable, in the Pierinae and the Bhbdocerinae 

 only the 5th abdominal segment is movable, whilst in the Antho- 

 cliarinae the pupa is solid and immovable. With regard to this, 

 the movable incision of Anthocharis belia shows a remarkable transition. 

 This species is double-brooded, one section, known as var. aiisonia, 

 feeding up in the spring and appearing in the imaginal state in June, 

 the second overwintering as a pupa. In the pupae producing the 

 summer imagines the incision between the 4th and 5th abdominal 

 segments is movable, in the winter pupae this is fixed and the pupa 

 solid (Ent. Record, vi., p. 288). The Nymphalid pupa possesses only 

 lateral movement like the Pierids, and, like them, varies in the number 

 of movable segments, being restricted to two incisions in the Argyn- 

 nids, Vanessids, &c, to one in Enodia hyperanthus, the pupa being 

 solid in Hipparchia semele and Melanaryia yalathea. The Euplceine and 

 Brassolid pupae, too, have also lost all power of movement whilst the 

 Lycaenid pupae are solid and possess hairs and bristles. 



The pupa of Troides amphrysus, if disturbed or blown upon, makes 

 quite a loud noise by the movement of the abdominal segments one 

 over the other, which noise is so loud that it is probably sufficient to 

 scare away some of its enemies (Niceville). 



The pupae of Lepidoptera vary much in adaptation to their sur- 

 roundings, and Chapman has pointed out that they afford important 

 taxonomic and phylogenetic characters. With regard to this we have 

 already stated that some pupae are much more generalised than others, 

 the pupa-libera and the pupa-incompleta being more ancestral than the 

 pupa-obtecta. The connection between the most generalised lepidop- 

 terous pupa-libera, as exhibited in Erioerania, in which the large pupal 

 jaws enable it to free itself from its cocoon, and those of other orders 

 of insects with active mandibles, has led Chapman to consider that 

 their evolution has largely been dominated as to how the pupa could 

 escape from the cocoon without the aid of active pupal jaws. He points 

 out that in the Coleoptera and Hymenoptera the imago is perfected 

 within the cocoon, not only throwing off the pupal skin there, but 

 remaining till its appendages have become fully expanded and com- 

 pletely hardened when the imaginal mandibles are used to force an 

 outlet of escape. He further points out that, in many cases, the 

 mandibles are of no use whatever to the imago, except in this one 

 particular, and he cites the Cynipidae as the most striking instance 

 known to him. In the Hemerobiidae, Trichoptera, &c, that spin silken 

 cocoons the pupal mandibles are used to cut a way out of the cocoon 

 by the pupa. In the Myrmeleonidae the pupa partly emerges from its 

 cocoon as does the lepidopterous pupa-incompleta, so that the funda- 

 mental methods of obtaining freedom from the cocoon, viz., by pupal 

 mandibles and partial emergence, as exhibited in the more generalised 

 Lepidoptera, are also in use among other orders, presumably more 

 generalised than any Lepidoptera. 



Although the Eriocraniids are the only Lepidoptera known to have 

 mandibles with which the cocoon can be ruptured and which (assisted 

 by the vermicular action of the abdominal segments) can be used also 

 to drag the pupa through any superincumbent earth, yet most of the 

 pupae-incompletae possess a sort of beak or hard process adapted for 

 breaking open the cocoon. In all these cases, the pupa emerges from 



