lecting stations too near each other. Usually they are not less than fifty miles apart. In some instances it has been possible to revisit important regions and make further collections which would more precisely establish the limits of ranges, or give more extended data on the variation of species in transition areas. As already noted, this cross-country field work has been supplemented by year-round collections from many localities.
Since the coördination of taxonomic studies depends upon an accurate understanding of the Latin binomials that have been used for previously described species, I have made a thoro study of the holotypes—the single specimens which are absolute criteria of the correct application of each published name—for 76 out of the 78 American species of the genus. Practically all of these holotypes were studied after this study had advanced toward completion and when it was possible to make critical comparisons with my own material. Of the 15 European forms in this genus, I have seen types of four. The types of seven are not in existence, and the remaining five are established by types which I have not seen. Wherever I have not seen the types, the original descriptions are quoted in full in the systematic portion of this study.
Finally, among the material resources available for the present study mention should be made of the technical assistance and museum equipment that has been available for several years chiefly thru the interest of the Zoölogy Faculty and the Administration of Indiana University and grants elsewhere acknowledged from the Waterman Institute of Indiana University, from the Elizabeth Thompson Science Fund, and from the Bache Fund of the National Academy of Sciences.
These are the bases of the present study of the genus Cynips. This is our warrant for undertaking an interpretation of species in this group of insects. We offer no further extenuation of the fact that our conclusions are not in accord with systematic work that has been based on more meager material. We believe it no coincidence that our conclusions more nearly accord with those of Dunn who studied 12,600 specimens of the 86 forms of the salamanders of the family Plethodontidae, or of Mickel who studied approximately 10,000 specimens of the genus Dasymutilla. In handling even such large numbers of individuals of a limited group of spe-
