330 AMERICAN HYMENOPTEKA. 



Netiroterus umbilical un n. sp. 



The galls of this species are found in great numbers on the under 

 surface of the leaves of Q. bicolor. Four or five hundred have been 

 counted on a single leaf. They are small, circular, flattened and 

 concave, with a minute conical elevation in the centre of the con- 

 cavity. They are about .07 of an inch across and .05 in depth, and 

 the pit is only .04 across. Beneath the conical elevation lies the 

 minute larval cell. Except the concavity described, which is 

 smooth, the surface is covered with short, stiff hairs, much like the 

 under surface of the leaf. The galls are easily detached, but leave 

 an indentation which is seen as a flattened elevation on the upper 

 surface of the leaf. The peculiar form of these galls suggests the 

 trivial name. 



The gall-flies are all females, and the entire body is shining black 

 and highly polished. 



Antennae short and very slender, except the first and second joints which are 

 very thick, the second quite as thick as the first, but a little shorter and less 

 tapering towards the base, the remaining are very slender, the third as long as 

 one and two together, the others (4 to 13) short, subequal, the last three forming 

 a thickened club with obscure articulations. The polished thorax without hair 

 or grooves. The scutellum small. The fovse wanting. Abdomen as dee]) as 

 long, and in shape subtriangular, like most of the species of this genus. Legs 

 clear, dark, shining brown, and in some individuals nearly black, and in all cases 

 lighter at the joints. Wings hairy; veins pale. Areolet small but distinct. 

 Radial area open, long and narrow and forming at its base with the second trans- 

 verse a perfect parallelogram. Cubitus slender, equal throughout, and reaching 

 quite to the first transverse. The first quite dark brown, the other veins pale or 

 colorless. Body .06, antennas .04, wings .07. 



Described from twenty- five specimens in my collection. 



I collected these galls in countless numbers for many years, and 

 made numerous unsuccessful attempts to rear gall flies from them. 

 Flies in vast numbers were bred from them, but all bore the unmis 

 takable marks of cynipideo-as parasites. These repeated failures 

 led me to wonder if there might not be, after all, species bearing 

 all the characters of the parasites that were true gall makers. It is 

 true that I knew of no such species, and I did not forget that proof 

 of this negative sort and hardly less in amount led Dr. Fitch to 

 publish a winged parasite as the originator of his Cynips ficus galls, 

 which, later, Dr. Walsh discovered were produced by the subapter- 

 ous Biorhiza forticornis. I, too, had reared millions of parasites 

 from the galls Neuroterus floccosus B. before a single gall fly was 

 di>covered. Visiting on a warm morning in Spring an oak that 



