INTRODUCTION. 11 



palpi from four to six jointed. The maxillary lobes are broad and membranous, 

 while the ligula is fleshy, and either rounded or square at the end. There is a com- 

 plete costal cell, while the subcostal cells are incomplete. The egg is of large size, 

 and increases in size as the embryo becomes more developed. The larva is a short, 

 thick, fleshj% footless grub, with the segments of the body rather convex. When 

 hatched they immediately attack the interior of the gall, which has already formed 

 around them. Many species transform within the gall, while others enter the earth 

 and there become pupae. 



Like the Aphides and certain other insects, the females often repro- 

 duce parthenogeuetically, viz, they lay eggs without having paired with 

 males, the latter not being at the time in existence. Thus the late B. 

 D. Walsh * discovered that the autumn brood of a gall-fly {Ci/nips quer- 

 cus-aciculata) consisted entirely of females which laid eggs, producing 

 the following spring both males and females which were originally re- 

 ferred to a supposed distinct species (Gynips quercus-spongifica). Hence, 

 after several experiments Mr. Walsh declared that ''the agamous 

 autumnal female form of this Cynips (C q.-aciculata) sooner or later 

 reproduces the bisexual vernal form," and is thus "a mere dimorphous 

 female form" of C. q.-spongifica. It was reserved for two other Ameri- 

 can students of the gall-flies to establish the fact that an alternation 

 of generations takes place in these insects. The case is thus stated by 

 Mr. L. O. Howard, in Psyche (iii, 329, June 24, 1882). 



America may justly claim the credit for the discovery of this most Interesting fact 

 of alternation of generations among Cynipids. Kiley, in the interjected remarks in 

 his article on "Controlling Sex in Butterflies" (American Nat., Sept., ly73, v. 7, p. 

 519). was the first actually to establish the fact beyond all perad venture, asM. Lichten- 

 stein points out; yet Bassett, four months previously (Can. Entomologist, May, 1873, 

 vol. 5, p. 93) had stated, in the following words, the theory which Adler has so fully 

 verified : "From all the above facts I infer that all our species that are found only in 

 the female sex are represented in another generation by both sexes, and that the two 

 broods are, owing to seasonable differences, produced from galls that are entirely 

 distinct from each other." In this article Bassett has just missed the actual proof in 

 two instances. With Vynips q. -operator he had observed the females of the vernal brood 

 ovipositing in acorn cups and producing the gall q.-operatola of Riley's MS.; but 

 he failed to rear the flies from these galls and so missed the complete proof. In the 

 case of C. q.-batatus Bass., lie had bred the sexual forms from leaf galls, and the agamic 

 females from twig galls, but had not actually observed the females of the former in the 

 act of ovipositing in the twigs ; thus again missing the proof. Riley, however, as he 

 tells us in his published note, succeeded in breeding the agamic females of q. -operator 

 from the acorn galls; thus, in connection with Bassett's observation of the oviposi- 

 tion, completely establishing the fact of alternation. So the credit should be joint. It 

 is, in fact, much like the well known case of Siredon and Amblystoma, in which the 

 credit should be divided between Baird and Dumeril. Dr. Adler very excusably 

 overlooked this note of Riley's. Walsh, in his earlier articles, came no nearer the 

 actual state of the case than to prove that two females, formerly described as dis- 

 tinct species, may belong to the same male. 



Independently of and subsequently to the work done in the CTnited 

 States, Dr. Adler, of Germany, also discovered and satisfactorily 



* American Entomologist, ii, 3.30, October, 1870. 



