Kinsey: Studies of Cynipidx 



83 



other experience than four months in a course in taxonomy, 

 fully distinguished between undetermined material of dasy- 

 dactyli and pacificus, two of the most closely related species. 



The galls of each species, however, are very distinct, 

 Heteroecus, ''many homes", referring to this characteristic. 

 All of the galls conform to a single type. The gall is the best 

 data we have as to the physiology of a gall wasp, as I have 

 shown before (Bull. Amer. Mus. Nat. Hist., XLII, p. 365) ; 

 and that this measure is sometimes extremely fine is abun- 

 dantly illustrated with many of the varieties I describe in 

 this paper. Many species also in this paper show that either 

 the morphology or the physiology may exhibit the greater de- 

 gree of variation, or that in other instances the amount of 

 variation may be equal for the two sorts of characters. There 

 can be no objection, then, to recognizing as a distinct unit a 

 group defined by abundant morphologic and physiologic char- 

 acters, where the species are best distinguished by physiologic 

 data, and the varieties of a species by morphologic data. It 

 may be objected that the insects should define species and the 

 galls the varieties. If one should attempt such an arrange- 

 ment he would be completely confounded, I think, by the diffi- 

 culties in choosing a particular structural character to unite 

 forms; the utilization of several such characters in combina- 

 tion would be impossible, for a large number of combinations 

 exist, with few coincidences. Such a method would combine 

 forms with distinctly different galls, and have to ignore facts 

 of distribution. By utilizing the galls in deciding species 

 lines, the conclusions reached are confirmed by some morpho- 

 logic characters of the insects, and especially by distribution 

 data. 



This genus presents an instructive instance of the need 

 for considering characters of several sorts in taxonomy, in 

 contrast to the practice among too many taxonomists! With- 

 out the galls as guides, it is probable that a solution of the 

 situation in this group would not yet have been reached. 



Whether to interpret this group as a species or a genus is 

 largely a matter of individual opinion and convenience. Cer- 

 tain it is that we must recognize three sorts of relationships : 

 the unity of the whole group, the divisions marked by the 

 galls, and within each division another division indicated 

 largely by insect characters and confirmed by a reasonable 



