THE PBYL.0GENY OP THE LEPIDOPTEROUS PUPA, 89 



cavity in the earth, a silken cocoon, a space in the timber in which the 

 larva has fed, the centre of a gall, or what not. In all cases it is neces- 

 sary for the imago to make its escape from this cavity, and it always 

 effects this by biting or gnawing an opening by which to do so. 

 Having thrown off the pupal integument, it remains within the pupal 

 nidus until all its parts have become hard and mature, a process 

 often occupying many days. It then bites its way out, being provided 

 with suitable jaws, in some cases, jaws that are of no use to it what- 

 ever but for this purpose (Cynipidae). 



If we now pass to the Neuroptera with complete metamorphosis 

 (Sialidae, Hemerobiidae, Panorpidae), we find again, as always, that 

 the helpless pupal state has to be passed in some protected nidus, but 

 the escape therefrom of the imago is accomplished in quite a different 

 manner. It is, indeed, the imago that escapes from the cocoon, but it 

 does so clad in the pupal skin, and, to break a way out it, uses its own 

 jaws, but armed with the hard chitin of the pupal jaws. It throws off 

 the pupal skin after emergence from the cocoon, and has to undergo 

 expansion of wings and hardening of surface after that, a process, 

 however, that takes place rapidly. No association of this method of 

 escape from the cocoon with that taking place in the bees and beetles 

 has been suggested, nor any probable derivation of one from the other. 

 These Neuroptera all have jaws as imagines, and it is by aid of these 

 that they escape from their cocoons, but the necessary hardness is 

 given by the pupal jaws that enclose them. 



These families lead directly to two others — the Trichoptera and 

 Eriocraniids, which are really intermediate between the Neuroptera 

 and Lepidoptera, the former being often treated as a family of 

 Neuroptera, or as a separate order. They differ in pupal characters 

 from the Neuroptera in this important respect that, escaping from the 

 cocoon in apparently the same manner, i.e., as an imago, encased in 

 the pupal skin, they do so without aid of any jaws of the imago (the 

 imago being without mandibles), but entirely by aid of the pupal jaws, 

 energised in a manner different from that in which insect jaws are 

 usually worked. 



We begin here, then, with the lepidopterous pupa in the Erio- 

 craniids. Unfortunately, that of the Micropterygids, a still lower 

 family, is unknown ; but even so, though the Eriocraniids are acknow- 

 ledged Lepidoptera, it must be admitted that, on pupal structure, they 

 have almost more claim to be Neuroptera, and, in any case, are closer 

 to the Trichoptera than to any typical forms of either Neuroptera or 

 Lepidoptera. 



The problem of the bees and beetles as to how to escape from the 

 cocoon, readily solved by them by the use of the imaginal jaws, is here 

 complicated by the added condition that imaginal jaws are not to exist, 

 and the solution is found in using the pupal jaws. It is convenient to 

 speak as though the problem was set first and the solution found after- 

 wards, in reality, of course, the solution {i.e., the new habit) was hit upon 

 first, and found to be useful in opening up other methods of existence, 

 and so persisted and became further altered and developed. The 

 problem then of how to escape from the cocoon is solved among 

 Hymenoptera and Coleoptera, by the mature imago biting its way out ; 

 among the Neuroptera by the immature imago breaking its way out, 

 still encased in the pupal skin, but still using its own jaws for the 



