554 [December 



Under these circumstances the mind naturally reverts to the idea, 

 that the diiFerence in the gall is caused by the difference in its location, 

 whether in the bud, or in the wood, or on the surface of the leaf, of 

 the same species of willow, and that the two supposed distinct species 

 of Cpci<lom,)/ia are in reality identical. But on the very same species of 

 Willow, S.humilis, there occur two galls, *S'. rhodoides n. sp. and S.gna- 

 pJmlioides n. sp., differing indeed in size, but constructed upon precisely 

 the same principle, both of them always solitary, both of them monotha- 

 lamous, and both of them formed by a similar deformation of the ter- 

 minal bud of a twig. Although each of these two galls may be recog- 

 nized at the first glance, and no two galls can be more clearly distin- 

 guished by several sharply-defined characters without any intermediate 

 u-rades connecting them, and I have examined hundreds of each to 

 satisfy myself of their perfect distinctness, yet the $ imagos proceeding 

 from these galls, and which are undoubtedly the authors of the galls, 

 because the larva and pupa live in the central cell, and 1 have actually 

 bred them from pupae extracted from the central cell, are undistinguish- 

 able when placed side by side, except by a slight difference in size, though 

 the average number and structure of the joints of the % antenna may 

 possibly be different. The larvae, too, are alike even when placed side 

 by side ; the pupje are precisely alike, even when placed side by side, and 

 the only characters, that 1 can discover, to distinguish the two species are 

 their size, their widely distinct galls, and the fact that the pupal cocoon 

 of the first is about 2 h — 3 times as long as the mature larva, and the pupal 

 cocoon of the second is from « as long again to twice as long as the 

 mature larva. Whence we may draw the general conclusion, that in 

 order to separate satisfactorily what are undoubtedly distinct species of 

 Ct'cidom^ia, it is necessary to study them, not only in the imago state, 

 but also in all their preparatory states, and to describe the galls with 

 the greatest precision. 



On reviewing the value of the characters to be drawn from all these 

 sources, with especial reference to the Gall-gnats of the Willow, which 

 are the only ones that I have carefully studied, I have arrived at the 

 following results : — 1st. The egg in all species where I have observed 

 it is uniform in shape, being constructed precisely as Osten Sacken de- 

 scribes it, but instead of being ''orange-yellow or whitish" it is always 

 sanguineous, {Dipt. K A. p. 180 ) Hence it appears that the egg 



