March, 191;.] FORBES : HydRIOMENIN^ OF THE UNITED StATES. 45 



As the primary purpose is if possible to furnish an aid to the identi- 

 fication" of North American material, I have based my primary tabu- 

 lation on the male sex; in Hulst both sexes are necessary for identi- 

 fication. For the same reason I have avoided as far as possible names 

 whose application is, or has lately been in dispute, believing that even 

 a fair proportion of names that would prove synonyms would cause 

 less confusion than a very few, even, which have now or lately been 

 applied to more than one genus. Thus I have dropped a couple 

 of Cramer's and Hiibner's names, misidentified by Hulst, and have 

 referred to the species under dispute by names which may prove 

 synonyms on further study, but which I believe at least unambiguous. 



The arrangement of the Hydriomeninse has been made fairly com- 

 plete for both sexes, as a large proportion of the genera can be de- 

 fined on other than secondary sexual characters. I have followed 

 the general European usage and treated as a single genus the central 

 mass of species, which show no clean-cut structural characters. An 

 analysis of such characters as I have been able to appreciate, has been 

 added in a note. Some time this genus (Hydriomcna) will be divided, 

 but not along any lines yet proposed. Hampson's character of the 

 course of the middle discocellular of the hind wing is the most nearly 

 natural of any yet proposed and, in our fauna at least, is correlated 

 with the presence or absence of coremata in the male, but it fails in 

 the caesiata group. Silaceata also belongs by habitus to the group 

 which it least resembles structurally; as the coremata are weak in it 

 it may be a true intermediate. It belongs to a considerable Oriental 

 group, whose genitalia have not been examined. 



Hulst's discussion of the characters used in classification leaves 

 comparatively little to be said. I have found the hind tibial pencil 

 less useful than he did, largely because it is often evanescent (as 

 noted by Pearsall in the genus Epimccis). Like most secondary sexual 

 characters, if not used cautiously it will separate closely related 

 species. The tuft of the thorax in the Hydriomcna group is probably 

 a natural character of a certain importance. Practically it is useless, 

 as a very large proportion of available material is rubbed; in fact a 

 large proportion at least of Hulst's Cocnocalpes possess the tuft. At 

 this point I have fallen back on a very unsatisfactory character, the 

 wing-pattern, to separate Stamnodes, and have let the more typical 

 Ca-iioca!pes (which in fact have the thoracic tuft) fall back into 

 Hydriomcna. 



