1804.] 557 



occur where the right and left antenna of the same individual % vary by 

 one joint, as has been noticed by Loew of C. ehrynopsidls Lw. (^Dipt. 

 N. A. p. 204.) Similarly, the % antenna of C. solidaginis Lw. is de- 

 scribed by Loew, probably from only a few specimens, as 22 or 23- 

 joiuted, (2-)-20 or 2-(-21,) but in one S which I bred myself of that 

 species it is distinctly 20-joiuted, (2-f-18,) thus showing a variation of 

 2 or •> joints ; arid, according to Mr. Herrick, the number of joints in 

 the antenna of the Hessian fly (C. destructor Say) varies from 16 to 19 

 or 2^-14 to 2-(-17. (Harr. LiJ. Ins. p. 570.) To avoid ambiguity, it 

 may be stated here that in the Grall-guats the long basal joint or scapiis 

 is counted as two joints, from the homology of allied families, though 

 to the eye but one joint is discoverable. As to the joints of the 9 

 antenna, I have found it impossible to count them with any precision 

 either in the recent or the dried specimen, owing to their being so 

 short and towards the tip so nearly cylindrical. On the other hand the 

 structure of the S antenna, as regards the comparative length of the 

 pedicels and verticils, is very constant ; but unfortunately it does not 

 differ at all in the different species that form galls on our willows, 

 though in other species, e. g. (J. soliddglnis Lw., it differs considerably; 

 and the same may be said of the neuration, with the single exception 

 of the structure of the anterior branch of the 3rd longitudinal vein, 

 which differs a little in some few species, the differences being nearly 

 constant. It may be worth while here to remind the student of the very 

 necessary caution given by Osten Sacken, "not to mistake for a vein a 

 longitudinal fold which generally exists between the 2nd and 3rd longitu- 

 dinal veins." {Dipt. lY. A. p. 175, note.) This fold is exceedingly 

 puzzling at first, and seems to foreshadow the interpolated vein between 

 the 2nd and 3rd longitudinals, which occurs either simple or forked in 

 the second Section of Cecidom,yidse,.^Anaretina. Even Westwood has been 

 apparently deceived by its simulating a vein so completely, for he figures it 

 along with the true veins. (^Intr. II. p. 518, fig. 3, and compare Dipt. N. 

 A. p. 174, figs. 1 — 5.) The % genitals may, and I think do, afford some 

 good specific characters ; but these characters are almost microscopic, dif- 

 ficult to describe without good figures, and become evanescent in the 



riable; but in % Prionus imhricornis Lin., which has an anomalously large num- 

 ber of antennal joints, the number varies, even in the right and left antenna of 

 the same individual, from 18 to 19. 



