three types of insects at Meadville is, however, the picture of segregation from a hybrid population which continues to proclaim its parentage. Similar series of erinacei are in our collections from many other localities.
There is, of course, a ready explanation of a past contact of wheeleri and derivatus. If wheeleri was in existence during the Pleistocene glaciation, its range must have been pushed southward at least as far as southern Indiana and the Ohio River, and still further south in the eastern mountains. If the range of derivatus at that time was comparable to the range of the present-day species, wheeleri first hybridized with derivatus in the Ohio Valley and in the valleys adjacent to the southern Appalachians. But as the glaciers retreated to the north, wheeleri retreated with them, leaving derivatus far to the south, and a tremendous area between where the hybrid wheeleri x derivatus found its opportunity to breed and interbreed until it had acquired the uniformity which warrants its present recognition as a species.
The Pleistocene origin of erinacei finds confirmation in three other hybrid species of Cynips in the same Northeastern area of the United States. These species are Cynips fulvicollis, C. gemmula, and C. macrescens. The detailed data are presented in the systematic portion of this study. They parallel the case of erinacei. These four cases account for all of the stocks of Cynips which are known to have penetrated far enough into the Northeast to have developed sub-Canadian varieties which would have been affected by the Pleistocene glaciation. These four cases are the only ones among the 93 species of Cynips which we now have reason for believing of hybrid origin, except for C. advena, of the Cumberland Highlands, which we shall show in a moment to date also from the Pleistocene. Of all the areas occupied by Cynipidae in the United States, this Northeastern area, the Eastern mountain country, and areas immediately adjacent to these are the only ones which had a Pleistocene history that would have provided the opportunity for the multiplication of hybrid individuals and which would have offered a subsequent isolation sufficient for the origin of new species. The apparent restriction of the hybrid Cynips to those areas seems confirmation of our explanation.
The fifth case C. advena, which we have mentioned, in-
