Kifisey: Studies of Cynipidse 



5 



veloped distinct varieties on each host. In this case the 

 amount of host variation is usually less than that of geographic 

 varieties, but sometimes it is more. In every case the varie- 

 ties of a single species are confined to either white or to black 

 oaks, altho closely related species may occur on oaks of both 

 groups. 



Many undescribed varieties have been previously over- 

 looked because of a charming lack of knowledge of faunal 

 areas and possible factors of distribution. It would be the 

 rankest sort of farce to delimit faunal areas on the basis of 

 the published records of distribution of Cynipidse, based on 

 determinations which neglect varietal differences! We have 

 yet to learn that some of the distribution work has no better 

 foundation than such careless taxonomy. It now appears that 

 each variety is restricted to a limited, more or less continuous 

 geographic area ; in no case does a single variety extend over 

 more than one faunal zone, and in only two or three of the 

 Cynipidse I have studied does more than a single variety of one 

 species occur in a single faunal area. These areas I have of 

 necessity in part defined by studying Cynipidag themselves, 

 when the several species of a region prove to agree most re- 

 markably in the extent of their range. A consideration of 

 the ranges of other organisms has been of some help in these 

 interpretations, but there is not much trustworthy literature 

 available. For the Pacific Coast good work on oaks has been 

 done by Jepson (1910, Mem. Univ. Calif., II, pp. 202-241), and 

 there is a remarkable study of the distribution of beetles by 

 Van Dyke (1919, Ann. Ent. Soc. Amer., XII, pp. 1-12). My 

 studies of Cynipidse very largely bear out Van Dyke's conclu- 

 sions as far as we have covered the same fields. Undoubtedly 

 many factors have contributed in limiting the distribution of 

 cynipid species. Host distributions do not appear to have had 

 much if any positive influence, altho acting negatively to some 

 extent, as when oaks are entirely absent from a region. I 

 must admit some prejudice for believing that the past geologic 

 history of an area is the most potent factor in present distri- 

 butions. The new^er country, geologically, appears to possess 

 the newer varieties. All of these questions I shall consider in 

 detail after a further study of Cynipidse. 



The genera of Cynipidse have always needed revision, and 

 the introduction of two terms new to American literature is 



