338 PROCEEDINGS OF THE NATIONAL MUSEUM vol. ni 



sidered to set the cydnid and corimelaenid bugs apart from the other 

 Pentatomoidea, as follows:^ 



. . . the presence of fringes of closely set, stiff, bristles at the apices of the mid 

 and hind coxae and the spiracles of the second abdominal segment being in a 

 membranous strip of the sternite, not in the heavily sclerotized portion. Mem- 

 bers of these two subfamilies have tri-segmented tarsi, and distinct tibial bristles, 

 and, with the exception of the Sehirini, have the trichobothria longitudinally 

 arranged often nearly in hne with the spiracles. The trichobothria, or delicate, 

 pale, sensory hairs (which must not be confused with the strong, dark, lateral 

 bristles which are frequently present) are two in number on each side of all 

 sternites and in the other subfamilies of the Pentatomidae are arranged trans- 

 versely, or nearly so, behind the spiracles. 



Disregarding the corimelaenids which McAtee and Malloch (loc, 

 cit.) adequately separated from the C3^dnids on the basis of the greater 

 claval exposure and the absence of "an area of smooth chitin behind 

 the eyes on the ventral surface of the head," the results of the present 

 study agree with most of those statements. They confirm McAtee 

 and Malloch's observations in the presence of the apical fringe of 

 bristles on the middle and hind coxae, the presence of the distinct 

 tibial bristles, the 3-segmented tarsi (except in Scaj>tocoris where the 

 hind legs lack tarsi) and the location of the spiracle of the second 

 abdominal segment. In contrast, the present results show that the 

 description of the trichobothrial arrangement is not true for all 

 genera in the Cydnidae. McAtee and Malloch apparently followed 

 Tullgren (1918) concerning the location of these structures in Cydni- 

 dae. Tullgren's choice of two genera for study was unfortunate 

 because both of them (Sehirus and Gnathoconus) were members of the 

 subfamily Sehirinae, which agrees with the other pentatomoids in 

 arrangement of these structures. Had he examined genera other than 

 those in the Sehirinae he would have found that other trichobothrial 

 patterns exist in the family. I have noticed that four additional 

 arrangements occur in the family, so that it is possible to divide the 

 Cynidae into five subfamilies on the basis of the trichobothria. 

 Further discussion of these subfamilies based on the trichobothria and 

 supporting characters will be found in the discussion under the family 

 heading on a later page. 



Thus, for differentiation of the Cydnidae from all other pentato- 

 moids (except the corimelaenids which were separated above) there 

 are four distinguishing features. Of these, the 3-segmented tarsus is 

 least diagnostic because it is shared with nearly all other pentatomoids. 

 The possession of distinct tibial bristles is shared with a few true 

 Pentatomidae (i.e., Strachia in the subfamily Asopinae), but if this 

 condition is restricted to a consideration of the lateral marginal row 

 of stout spines on the more or less flattened anterior tibia it may be 



'The parenthetical references to illustrations in the original have been omitted from this quotation. 



