332 [May 



number of European insects of this family. In view of the compara- 

 tively small number of known N. A. species and of the large additions 

 which we can reasonably expect very soon, it would be premature to 

 attempt now the reform above alluded to. This conclusion is still more 

 justified by the consideration that we may likewise hope within a short 

 time, to see new light thrown upon the still mysterious sexual ques- 

 tion of this family, a discovery wbich will naturally affect its systema- 

 tic distribution. 



In agreement with these motives, I have adopted in this paper only 

 three pxenidous genera, easily recognizable by their general appearance, 

 even in the absence of a definition, the more so, as each of them is con- 

 fined to a different genus of plants. These genera are : Gynipa (in the 

 wider sense, in contradistinction of the genus Cynips Hartig, in the 

 narrower sense, reputed agamous), which is confined to the oak ; Rho- 

 dites, all the known species of which live on the rose, and Diastrophus, 

 hitherto obtained from galls on Rubus and the allied Potentilla. As 

 to the subdivision of the first and largest of these genera, I have con- 

 fined myself to a few hints, leaving it to the future student either to 

 follow them out or to controvert them. 



The same views about the unseasonableness of a reform of the sys- 

 tem of Oynipidx have induced me to retain the primary subdivision of 

 the family, adopted by Hartig and based upon the shape of the radial 

 cell and the position of the areolet This subdivision, insufficient as it 

 is. affords the advantage of being almost coincident with the subdivi- 

 sion into Psenidve and Inquilinse, based upon the habits of the insects. 



Hartig divides the gall-inhabiting Cynipidx into two sections: 



I. With a narrow radial area, the areolet being opposite its basis. 



II. With a broad and stout radial area, the areolet being beyond its 

 basis. 



The former are generally Psenidse ; the latter Inquilinse; the excep- 

 tions, mentioned by Hartig or observed by others, will be noticed at 

 the proper places in the sequel. Besides, most of the former (with the 

 exception of Rhoditcs), have an open radial area ; most of the latter 

 (with the exception of several Aulax), have it closed. All these dis- 

 tinctive characters may be useful, as long as taken in connection ; but 

 how vague Hartig's definition of the two sections is, becomes very 

 striking, if the radial area of Rhodites is compared with that of any spe- 

 cies of the other section ; of the two, it is certainly the former which 

 deserves to be called broad and short; moreover, it is closed, whereas 

 the area of Aulax sylvcxtris is open. 



