1865.] 339 



Gynipidse according to Ilartig's system, is apparently due, besides the 

 imperfection of the system itself, to the peculiarity of the American 

 fauna, which may require the establishment of several new genera. 



The great prevalence of oaks in this country and the great number 

 of their species, would justify o priori the inference that Gynipidse are 

 very abundant. That they are not only abundant, but also very different 

 from the European representatives of this family, may be inferred from 

 the following facts: — 1st. The group consisting of the species G. spon- 

 gifica, ilicifolise, coelebs, singularis, etc.. quite numerously represented 

 in North America, does not exist in Europe. 2d. The agamous genus 

 Cynips, in the restricted sense of Hartig, is much more numerous in 

 Europe than in America, as among 08 species of true gall-flies ( Psen ides). 

 described by Hartig, and in the last publication of Giraud, 39 belong- 

 to that genus; whereas among nearly 50 North American Psemdes at 

 present known, only /our, perhaps Jive, as we shall see hereafter, belong- 

 to Cynips Hartig. 



In the introductory remarks to this paper, I have already expressed 

 the conviction that before the North American fauna of Gynipidse is 

 better known, but especially before tin mystery of tin reproduction of 

 Cynipidx is solved, it would be premature to attempt a final distribution 

 of the family in genera. Indeed, if we were not acquainted with the 

 fact that C. aciculata is nothing but a dimorphous <j> of C. spongifica, 

 we would probably have located it in a different genus, on account of 

 the peculiar structure of its abdomen, the number of joints of its an- 

 tennae, etc. 



Admitting, therefore, as announced in the introduction, only three 

 genera of psenidous Gynipidse, Cynips, Rhodites and Diastrophus, all 

 that I can offer at present, as to the subdivision of Cynips, is confined 

 to a few hints, as follows : — 



1. G. strobilana 0. S., C. tubicola 0. S., C. globulus Fitch, and C. 

 centricola 0. S., are the only American species at present known which 

 Hartig would have located in his agamous genus Cynips. (The two 

 former species I have communicated to Dr. Reinhard; of the two latter. 

 I judge by analogy.) Although C. clavula Bassett has a pubescent 

 thorax, a character peculiar to Cynips Hartig, I am not sure whether 

 it can be considered as belonging to this genus. All these species 

 occur within the group of the white and chestnut oaks (Q. alba, prinus, 

 obtusiloba, etc.) 



2. G. seminator Harris, which I have communicated to Dr. Reinhard, 

 is an Andricus ; judging by some structural analogies, C. petiolicola 

 Bassett and C. fusiformis 0. S., perhaps also Cfutiilis (C. papillata), 



