i)(ji) [December 



quiescent specimen, it projects a little from the anterior extremity of 

 the body, I infer that its use is to abrade the interior of the gall, and. 

 by the irritation thereby produced, promote the growth of the gall and 

 cause a flow of sap which is to form the food of the larva. As no solid 

 faeces are found in the cells of Cecidomyidous larvse, it is evident that 

 those larvEB cannot devour the solid substance of the gall, and their 

 mouths seem entirely too soft and membranous to produce any material 

 abrasion in the interior of some of the more woody galls e. g. »S'. siliqtca. 

 In confirmation of the above idea, it may be stated that I found in No- 

 vember a single larva of C. s.strohiloides n. sp.,with one of the thorns 

 of its Y-shaped breast-bone absent, and apparently broken oif short at 

 the bifurcation. The breast-bone can scarcely be used for locomotive 

 purposes, as Osten Sacken doubtingly suggests; for if it were, we should 

 surely find it in other Dipterous larvas besides those of the Grall-gnats. 

 Whatever be its use, it must be something specially connected with the 

 habits of the Gall-gnats, otherwise we should find it elsewhere. In the 

 larva of another widely distinct Dipterous gall-maker, Irijpeta mlida- 

 yinis Fitch, there exists no such organ, but the mouth terminates in a 

 robust, horny, black, emarginate piece, which probably subserves the 

 same purpose that I believe to be subserved by the breast-bone of the 

 larva of the Gall-gnats. 



As to the pupal cocoon of Cecklomyia^ Winnertz, as quoted by Osten 

 Sacken, " positively denies that the larvae spin this cocoon ; according 

 to his observation, the latter is, so to say, exudtd by the larva. He 

 found that larvae, which had fastened themselves to a leaf, were encir- 

 cled within twenty-four hours by a white halo, consisting of tiny, thread- 

 like particles, which seemed to grow somewhat like crystal-needles ; the 

 larva during this time remained perfectly motionless. The cocoon is 

 perfected within a few days, and even then, under a strong magnifying 

 power, no genuine thread is perceptible." {Dtpt. JV. A. p 184.) I 

 believe that it is in this manner that the pupal cocoon of all Cerido- 

 myia is formed, i. e. that it is not spun by the larva, but secreted in a 

 glutinous form from the general surface of its body. I have observed 

 that the thin, filmy cocoon of such species of Willow Gall-gnats, as reside 

 in a gall composed internally of the closely appressed and over lapping- 

 leaves of the deformed bud, (C. s. brassicoidcs n. sp., C. s. strohiloides 

 u. sp., C. s. rhodoides n. sp. and C. s. ynaj^hafioides n. sp.) is almost 



