ing 2.5 by 4.0 mm., with a distinct wall but inseparable from the compact filling of the gall. Attached to a leaf vein, on the undersides of leaves of Quercus pedunculata and Q. sessiliflora. Doubtfully identified on Q. turneri (Rolfe acc. Kieffer 1901).
BISEXUAL GALL.—Apparently very similar to that of the bisexual forms of Cynips folii (q.v.). Known only for the variety longiventris form substituta, where it is dull green in color, with a longer, gray pubescence, a more sharply pointed tip, and a smaller size which is up to 2.0 mm. in length. On the bark at the bases of older trees, less often on the two-year old stems; on Q. pedunculata, Q. sessiliflora, and Q. pubescens.
RANGE.—Known thruout Europe except in the Mediterranean area. The two described varieties, longiventris and forsiusi, are confined to more central Europe and northern Europe respectively.The typical variety of longiventris is apparently common in Central Europe. The variety forsiusi is known only from southern Finland and Denmark. The species is strangely absent from the records for Mediterranean Europe. Our only indication of the occurrence of the insect there is Trotter and Cecconi's (1904) record for Italy without more definite locality, and a doubtful record (acc. Tavares 1928:44) for Spain. There seems no available explanation to show why longiventris should differ from so many other species that have their Central European varieties matched by Mediterranean segregates.
Cynips longiventris and Cynips folii give striking data on the existence of physiologic species. The galls of the two are never to be confused, and yet the insects, in both agamic and bisexual generations, are so nearly alike that the European authors have usually accepted them as indistinguishable. Mayr (1882) thought he could distinguish the agamic females by the proportions of the twelfth antennal segments, some minor points of color, and the body length. These are the only distinctions made in the Kieffer (1901) and Dalla Torre and Kieffer (1910) monographs. Examination of good series of both longiventris and folii convince me that in antennal and color characters the insects vary more individually than they were supposed to differ specifically. The average length of longiventris is clearly under that of folii, but series again show many specimens of folii which are smaller than the larger specimens of longiventris. On the other hand, I offer the characters of the scutellum (see the description of the
