22 



Indiana University Studies 



bers. From 44 galls of variety consohrinns I obtained 70 per 

 cent males; and Patterson obtained 90 per cent males for 

 austriorl This bisexual generation lays fertilized eggs in 

 year-old acorns, the next generation produces a naked, seed- 

 like, acorn gall, is agamic, and takes two or more years to 

 develop. Superficially the insects and galls of the two gen- 

 erations are so different that this has appeared as extreme a 

 case of heterogeny as we knew. The differences, I now real- 

 ize, are largely superficial. Upon making the comparative 

 descriptions of the insects I find that they differ in only a 

 relatively few respects ; it has been simple to make a descrip- 

 tion which would cover either generation of the species. Upon 

 further examination, especially of younger stages of the galls, 

 I find that they are much the same except for one being naked 

 and on the acorn, while the other has a woolly covering and 

 occurs on young stems, leaves, or flowers, but both have at 

 center similar seed-like capsules enclosing nearly as large 

 larval cells. I have suggested (1920, Bull. Amer. Mus. Nat. 

 Hist., XLII, p. 373) that heterogeny in Cynipidse is an ex- 

 treme development of seasonal dimorphism. At that time I 

 questioned whether one could ascribe the differences in the 

 galls of the two generations of operator primarily to the dif- 

 ferences in the plant tissues attacked ; but now I am readj^ to 

 believe such may be the case. If, as Cosens has shown (1912, 

 Trans. Can. Inst., IX, p. 374), no new plant structures are 

 developed in the formation of a gall, one should not expect 

 the normally hairless acorn to produce as woolly a gall as 

 the normally pubescent or hairy young stem, leaf, or flower 

 tissues. 



The agamic generation has been bred for only one variety, 

 operator, but I have acorn galls which probably represent the 

 agamic generations of every one of the varieties here de- 

 scribed. The following references are to the bisexual forms 

 of possibly undescribed varieties : 



Cynips q. operator Walsh, 1864, Proc. Ent. Soc. Phila., II, 

 p. 494. Thompson, 1915 (Walsh record), Amer. Ins. 

 Galls, pp. 11, 30. 



This species has further been recorded (Beutenmuller, 1913, 

 Bull. Brook. Ent. Soc, VIII, p. 103) from Ottawa, and Penn- 

 sylvania. I do not know what varieties occur in these locali- 



