4 BULLETIN 123, UNITED STATES NATIONAL MUSEUM. 



made after consideration of both primary and secondary characters 

 in the light of the added information derived from a study of the 

 male genitalia and to a limited extent of the larvae, only split within, 

 and not across, the limits of their genera. The only weakness of 

 their system lay in the equal value they attached to differences in 

 fore and hind wing variation. The erection of the family Spargano- 

 thidae is untenable on the character given (7 and 8 of fore wing 

 stalked) for the stalking of veins 7 and 8 of the fore wing does occur 

 in several genera in the Olethreutidae, and these can in no way be 

 united with Sparganothis^ in spite of the pectinate hind wing. 



PHYLOGENY AND CLASSIFICATION. 



It is the author's conviction that in the Olethreutidae venational 

 changes of the hind wings are the characters of most fundamental 

 import. On these characters — with the exceptions of a stalked 6-7, 

 easily and at different places derived from the normal approximate 

 condition of these veins, and the united 3-4 derived equally readily 

 from a stalking of 3-4 — the genera fall into larger natural groups 

 to which, for convenience of handling, subfamily names are given. 



In separating the genera of the Eucosminae I have considered as 

 nearly as possible all the external structural characters of the moth, 

 including secondary sexual modifications. My purpose has been to 

 arrange the species in their natural order, putting together those 

 most alike in genitalic structure and general habitus and separating 

 them into groups according to their development from the general- 

 ized type. These groups I have designated as genera, defining them 

 on any characters that would serve to identify them. In this family, 

 strange as it may seem, it is necessary to know what species consti- 

 tute a group before the taxonomic value of any single character can 

 be established. Once we know what species constitute a group, any 

 character or combination of characters will serve to identify it, even 

 though such a character unsupported may not of itself justify ge- 

 neric separation. This applies to venational almost as strongly as it 

 does to secondary sexual characters. Of the latter I have been able 

 to use the male antennal notch (figs. 3a, 4a) and certain male sex 

 scalings on the hind wings {Proteoteras^ Grocidosema, Rhopohota^ 

 figs. 6, 7). These characters, though independently acquired in dif- 

 ferent places, indicate an advance from the primitive type and corre- 

 spond with other progressive characters in the genitalia. This, of 

 course, does not mean that all the species having an antennal notch, 

 for example, belong together, any more than that all the species 

 having veins 7 and 8 of fore wing stalked belong together, but I 

 think it does mean that they have developed further than those 

 without the notch and should be separated from them, particularly 

 as such a character when once acquired would not be easily lost. 



