88 THE CANADIAN ENTOMOLOGIST, 



17. Joint 5 longer than 6. ... cynosbati, rosce, auraiitiacum^ qitercifex^ 



S^fraxini. 



5 and 6 equal, or 6 longer armeniacum (Calif, and Kans.), coryli 



[of King and Reh, Lymani^ persicce^ (Canada, under 

 [glass, and from Dr. Reh, cf King, pruhwsiim^ 

 \iliiercitronis (Ariz, and Calif.), viiii of King and 

 [Reh, robi?iice} (Tempe, Ariz., on osage-orange), 

 ]^que?'ci/ex (of Signoret), takachihoi (Japan), Cana- 

 \iiense (Maine, on elm), sp. on tulip tree, R. I. 



18. joint 3 conspicuously longest, 5 conspicuously shorter than 4 or 6, 



7 shortest of all ; 8, 5 and 2 about the same length ,. wc/o;/^//^!;-/^^. 



3 longest, but not long, 5 and 8 about equal, and not much shorter 



than 3, 4 conspicuously shorter than 3 or 5 . . . .prunastri {^ya^nct). 



3 and 4 subequal, or sometimes 3, sometimes 4, a little the longer. . 19. 



4 longest, much longer than 3, 3 and 5 about equal .... hortensict. 

 3 longest, and rather long, 4 conspicuously longer than 5 ; 5 being 



conspicuously shorter than 4, but a little longer than 6 20. 



3 longest, 4 and 5 equal or subequal 21. 



19. 5, 6 and 7 shortest, and equal or 



almost stibsimile (Chihuahua) and berberidis. 



6 and 7 shortest and equal, but 5 conspicuously longer. . . .hortensicE. 



20. 8 short, shorter than 4, 7 a little shorter than 6 quercifex. 



8 longer, longer than 4, 7 a little longer 



than 6 persiae ? (Canada, on peach under glass, cf. King). 



21. 3 very long, much longer than any other joint 22. 



3 not very long, usually quite short ; 4 shorter 



than 5 Cocke7'eUi\ rosarum of King and Reh, Marchali. 



22. Joints 4, 5, 6, about equal riifuluui. 



5 longer than 4 or 6 '. '. Cockerelli. 



The occurrence of the same ^-pecies in several different places in the 



above table shows the great variability of the antennae of these insects ; 

 yet I do not believe for a moment that this variability is indiscriminate, or 

 that the antennie are useless for purposes of identification. They must, 

 however, be used cautiously in this genus, and in conjunction with other 

 characters. 



I do not expect to pay much more attention to this genus myself; for 

 Mr. J. A. Sanders, with much better opportunities than I possess, is about 

 to beg'n an investigation of it; and he will undoubtedly make many things 

 clear which have been obscure. 



