1864.] 
496. 
The % has an abdomen as long as the $ and but for the absence of the 
“ ventral valve’’ and of the sheaths of the ovipositor might be easily 
mistaken for 9 . In all the above variations intermediate grades occur. 
On comparing 33 S S bred in May from the gall q. ficus with the above 
23 S S , their size varies within exactly the same limits, but the range of 
the coloration of the former is not so extreme. None have the femora, 
so far as can be seen, vittate above with brown-black, but several have 
the hind tibiae thus vittate, which does not occur in those bred from 
spongifica. It is observable that Dr. Fitch describes 0. q. ficus as 
having its “ hind shanks dusky.” The wing-veins are almost always 
brownish. Evidently the two broods of insects are identical, as the 
markings of the legs are in Cynipidae quite an inconstant character. 
On comparing 699, bred in May from the gall q. ficus^ with the de¬ 
scription of the 9 (?) by Osten Sacken, they differ as follows :—The an¬ 
tennae are 13-jointed, the last joint i as long again as the preceding, (not 
“ 14-jointed.”) The scutel is not even “ slightly reddish” at tip, and 
it was only occasionally so in the % % bred from spongifica galls. The 
abdomen is black, immaculate, and joint 2 never occupies more than 4 
of its lateral length exclusive of the peduncle. The hind femora are 
scarcely “ infuscated,” but the hind tibiae are vittate above with brown- 
black. The wing-veins are rather fine, but almost always brownish. 
Length 9 -69 —.10 inch.—It may be added that the 2nd abdominal 
joint in all the specimens % 9 describes dorsally a circular arc of 45°, 
that the tip of the “ventral valve” is unarmed and in an angle of 30°, 
and that the sheaths project at least .02 inch above the line of the 
back. As these very conspicuous sheaths are not mentioned in the 
description, and as it is expressly stated that the 2 nd abdominal joint 
conceals all the following ones, which is a peculiarity of the % sex, I 
infer that Baron Osten Sacken, having only a single specimen to guide 
him, mistook % for 9 . In any case he has, as it seems, unless I have 
incorrectly identified my species with his, miscounted the antennal 
joints, which are % 15, 9 ^'"^5 whereas he gives them as 9 14. They are 
peculiarly difficult to count in this species. 
Judging from the brief description of C. q. pisum Fitch, that fiy 
also, as well as G. q. ficus Fitch, appears to be identical with the Gruest 
gall-fiy, Iscviventris 0. S. It has already been stated that Osten Sacken 
obtained what is-apparently a true Grall-fly from the gall q. pisum Fitch, 
