1864.] 567 



Be this as it may, one thing is quite clear. It is impossible that, 

 in one and the same genus of insects, some species, as Harris be- 

 lieved, should spin a silken cocoon and transform into the pupa state 

 inside that cocoon, without moulting any larval integument, by a cer- 

 tain anomalous budding process, and that other species should spin no 

 cocoon, become detached from the larval integument without ceasing 

 to be still larvae, and then transform inside that detached larval integ- 

 ument by the same budding process as the others. It is undoubtedly 

 true, for I have verified the fact myself, that some Coccinellidse trans- 

 form to pupa inside the larval integument, and some moult it in the 

 normal manner ; this is anomalous enough, but it is not so utterly ano- 

 malous as the Harrisian theory.* But the climax is reached, when it 

 is proved by the observations of Winnertz and Osten Sacken, that 

 several other species of the same genus exude their cocoons from the 

 general surface of their bodies, thus giving three totally different me- 

 thods of forming the pupal envelop in the same genus — spinning, 

 moulting and exuding!!! It is very true that the pupal envelop, 

 in the Hessian Fly and in the Gall-gnats that exude their cocoons, 

 is much more dense and leathery than in the Grall-gnats of the Willow 

 and in the Wheat-midge ; but that is merely a question of mode and 

 degree, not of principle, and is probably due to the fact, that in the 

 Gall-gnats of the Willow the pupa is completely protected by a dense 

 mass either of wood or leaves, and does not therefore require a robust 

 cocoon, while the Wheat-midge ordinarily goes under ground to assume 

 the pupa state, though a few transform in the ear of the wheat. That 



"••■In Chilocorus, as stated by Westwood and as I have myself observed In C. 

 bivulnerus Muls., the larval integument is retained whole by the pupa; in the 

 European Coccine/la Argus it is retained, but widely split open along the back, 

 thus showing an intermediate grade between the anomalous transformation of 

 Chilocorus and the normal transformation of most other Coccine/lidce. (Westw. 

 Intr. pp. 397 — 8.) But there can be no possible intermediate grade between a 

 cocoon spun by the mouth of a larva, and the puparium of a true coarctate 

 pupa, which is formed ovit of the indurated integument of the larva, the two 

 things being radically and fundamentally distinct. In Anthrenus {Dermestidce), 

 which also retains the larval integument when it transforms to pupa, there is a 

 similar slit made along the back of it; but whether this is also the case in 

 other Dermestide genera which retain the larval integument when they trans- 

 form to pupa, {Megatoma and Tiresias,) is not stated. (See Westw. Intr. pp. 159, 

 161.) 



