72 [OCTOBKR 



hard grains, resembling seeds, each having a cavity in its centre," etc. 



The flies which I obtained from these galls agree with Dr. Fitch's des- 

 cription, except that the neuration of their wings shows that they do not 

 belong to the true gall-flies (Psenides) of Hartig. The second segment of 

 the abdomen is shorter than the third, the radial area is cksed by the 

 prolongation of the subcostal vein along the anterior margin and the areo- 

 let corresponds more to the middle than to the base of the radial area, all 

 characters distinguishing them from the true gall-flies. Still, I am at a 

 loss to say to what genus they belong. I thought for some time that they 

 agreed with Amhh/notus Hartig, as defined in Reinhardt's recent Mono- 

 graph of the Figitidse (Berl. Entom. Zeitschr. 1860), but the antennae of 

 the male, which I examined repeatedly on the living and the dead insect, 

 appeared to me 15- and not 14-jointed. The % of Dr. Fitch's C. qiiercus 

 tuber has 14-jointed antennae. My 9 , like Dr. Fitch's, have 12-jointed 

 antennae, if the last, very elongated joint, is counted for one. The 9 of 

 Amhlynotus, according to Reinhardt, has 13 joints. 



My specimens (several S and one 9 ) are black, with yellowish mandi- 

 bles and a spot of the some color between them and the antennae; the lat- 

 ter brownish-yellow, more dusky towards their tip in the % ; feet yellow- 

 ish, last pair, except at the knees and tarsi, more brownish; intermediate 

 pair !also slightly infuscated on the femora; wings hyaline, thick veins 

 pale yellowish-brown; % 0.05, 9 0.09 long. 



