1864.] 561 



always indissolubly agglutinated, especially towards its base, where 

 the external air has not so much chance to dry it. to one or more 

 of the small linear-lanceolate leaves that form the interior of the 

 gall. I have also observed that the cell in which the immature larva 

 of C. s. hdtatas n. sp. resides — the gall itself lacing composed of a ho- 

 mogeneous, rather compact, spongy substance — is (July 30) rough, 

 opaque and scaly on its internal surface, while the cell of the mature 

 larva for many months before it assumes the pupa state (November 11 

 and subsequently) is glabrous and polished, without any distinct cocoon 

 as in the other species. To what can we attribute this change, but to 

 the exudation of some glutinous substance by the larva, with which it, 

 as it were, plasters the rough walls of its house ? If the cocoon of Ceci- 

 domyia was always spun by the mouth of the larva, as most hymenop- 

 terous and lepidopterous cocoons are constructed, it would surely here 

 assume the ordinary form of such cocoons when spun inside the walls 

 of a cell, i. e. an integument distinct from the walls of the cell; whereas 

 the smooth internal surface of the cell is intimately united to the origi- 

 nal rough surface, and can no more be detached from it than the finish- 

 ing coat of plaster can be detached from the first rough coat. I have 

 observed a similar smooth lining to the cell-walls of Lauoptera solida- 

 glnis 0. S., which, like those of C. s. batatas, are surrounded by brown 

 sponge. As a proof that the smooth internal surface of the gall-cell of 0. s. 

 batatas is homologous with the filmy cocoon of C. s. brassicoides, &c., we 

 find in C. s. siliqua and C. s. cornu n. sp. an intermediate grade between 

 the two, viz : the central and generally the lower portion of the cocoon 

 almost indissolubly plastered on to the smooth walls of the cell, and the 

 upper and sometimes also the lower end forming a thin, filmy diaphragm, 

 of precisely the same texture as the entire cocoon of C. s. brassicoides, 

 &c., across the mouth and sometimes the lower end also of the cell. 



From not sufficiently attending to the peculiar nature of the above 

 process, some authors have sixpposed that the pupal cocoon or " flax- 

 seed" envelop of the Hessian ^y (Ccc. destructor Say) was nothing but 

 the indurated '' skin" of the larva, i. e. that a Nemocerous Dipteron had 

 a coarctate metamorphosis like a Notacanthous or an Athericerous Dip- 

 teron ! (See Harris InJ. Ins. pp. 575-7. and Fitch as quoted tit length 

 by Osten Sacken, Dipt. M. A. p. 204.) But both Harris, and West- 

 wood, and several other authors, expressly state that, when the " flax- 



