CYDNIDAE OF THE WESTERN HEMISPHERE — FROESCHNER 353 



the European genera but did not establish any named categories 

 upon the results. Unfortunately, his choice involved the close ap- 

 proximation of the basal halves of the Sc+R and M, a feature 

 which apparently occurs with some irregularity in the Cydnidae so 

 that even otherwise closely allied genera may not agree in this char- 

 acter, though they may agree in it with more distantly related forms. 

 The results of the present study of venational features (figs. 165-169) 

 permit forming the following tabulation, which agrees with the results 

 obtained above from the trichobothria. 



la. Sc + R recurved at apex to meet M (fig. 1G8) Amnestinae 



16. Sc + R straight, connected to M by a more or less strongly oblique cross vein. 

 2a. Cross vein r-m very strongly oblique so that ]\Il + 2 leaves radial cell 



basad of fracture in Sc + R (fig. 165) Scaptocorinae 



26. Cross vein r-m not so strongly oblique, Ml + 2 leaving radial cell beyond 

 fracture in Sc + R. 

 3a. Vein M with a spur or lobe projecting into radial cell at its midlength. 

 4a. Three veins arising independently from apex of radial cell (fig. 169). 



Garsauriinae 



46. Two of the three veins arising at antero-apical angle of radial cell, 



third one from posteroapical angle (fig. 166) . . . Sehirinae 



36. Vein M without trace of spur or lobe projecting into radial cell (fig. 



167) Cydninae 



Thus, with evidence drawn from two nonadaptive characters of the 

 Cydnidae it is possible to establish five subfamilies. Unfortunately, 

 both of these features are somewhat difficult to use, either because 

 of their small size or the fact that the}" are normally hidden from 

 view. But a survey of the other characters of these insects shows that 

 a much more usable key to the subfamilies can be based on certain 

 more conspicuous characters. Such a key to the subfamilies follows. 



General key to the subfamilies of Cydnidae 



1. Clavi meeting beyond short scutellum and forming a commissure almost as 



long as scutellum (fig. 2 ) Amnestinae (p. 628) 



Clavi not meeting beyond scutellum, not forming a claval commissure . . 2 



2. Anterior tibia strongly cultrate, much produced beyond tarsal insertion so 



that tarsus appears to arise at middle of tibial length (fig. 3), 



Scaptocorinae (p. 365) 

 Anterior tibia not cultrate, tarsus arising at or very near apex of tibia . . 3 



3. Pronotum with a lateral, submarginal row of setigerous punctures; tarsal II 



subequal in diameter to I and III Cydninae (p. 377) 



Pronotum without a lateral, submarginal row of setigerous punctures; tarsal 

 II distinctly narrower than I and III 4 



4. Antennal II as long as or longer than I; pronotum without fine, distinctly 



impressed subapical groove Sehirinae (p. 354) 



Antennal II less than half as long as I; pronotum with fine, distinctly im- 

 pressed subapical groove paralleling anterior margin (fig. 65). 



Garsauriinae (p. 364) 



