606 [December 



I have observed in two successive autumns, that many of these galls, 

 especially the large, potato-like ones, are already bored by holes of the 

 same size as those made by C s. hatataa ; and on placing a large num- 

 ber of such bored galls next spring in a separate breeding-jar. I obtained 

 from them in considerable numbers the same 3 parasites which I bred 

 at the same time in very large numbers from the unbored galls, but no 

 Cwidomyia. Hence I infer either, \»t. that a few C. s. batatas come 

 out in the autumn and the rest not till the following spring, which actu- 

 ally occurs with the Wheat-midge, (see above p. 568), and is a com- 

 mon thing with many other insects, or 2nd. that the species is double- 

 brooded like the in([uilinous C. albovittata n. sp., which is contrary to 

 the analogy of the other (lall-gnats of the Willow and does not harmo- 

 nize with the fact of the bored and unbored galls producing the same 

 identical 3 parasites, or 3r(Z. that there is some parasite or inquiline. 

 hitherto undiscovered by me, which infests these galls and whose natu- 

 ral time for assuming the imago state is in autumn. Of these three 

 hypotheses, which are all possible. I decidedly incline to the first. 



No. 13. Gall S. verruca, n. sp. — On S. humilis. A small, monothalamous, ir- 

 regularly spherical, greenish yellow gall, .07 — .13 inch in diameter, growing 

 the latter end of August from the midrib or some of the principal veins of the 

 leaf, half of it projecting from each side of the leaf. The upper side is flattish 

 or with a minute point or nipple, the lower side branches out into a ragged, 

 wart-like excrescence, whence the specific name, which later in the season 

 bursts open so as to afford an exit to the insect. When cut into in August, the 

 external wall of a few galls is found to be rather woody, enclosing a central 

 cell, in which lies the larva; but the majority of them-are still solid. From 1 

 to 12 are found on a single leaf, several of them being often confluent, but with 

 their internal cells, when they have any, separated by a thin partition, and 

 with separate warts to each on the under side of the leaf, which afterwards open 

 separately. Rather abundant, but local near Rock Island. Described from 38 

 affected leaves. 



Larva. — By August 26th, in a few of the galls, the larva is .07 inch long, of the 

 usual oval shape, orange-colored, and with the breast-bone suborbicular, small 

 and indistinct. In the others the larva is not developed, nearly the whole in- 

 terior of the gall being solid. Those that I attempted to breed all dried up in- 

 side the gall before November, the leaves having been kept too dry ; but from 

 the structure of the gall itself and the analogy of similar Cecidomyidous galls 

 on the oak, {Symvietrica 0. S. and Quercus pi/ulce Walsh,) I infer that they go 

 under ground to transform into the pupa state. 



Pupa and imago unknown. 



No. 14. Gall S. semen, n. sp. — On S. nigra. A minute, monothalamous, hoi- 



