4 University of Michigan 



a\ Quadrangle short; usually nearly two or more postquadrangular 



cells ; the cubito-anal cross-vein much nearer the second than the 



first antenodal. Hctcragrion. 



b\ Inferior appendages of male rudimentary; teeth on genital valves 



of female several-rowed. Generally the hind margin of wing 



separating from the anal vein at a greater distance distad to the 



cubito-anal cross-vein than the arculus is distal to the cubito-anal 



cross-vein. Frons obtusely angled, similar in the sexes. 



Groups I and 2. 

 b'. Inferior appendages of male small, spine-like ; teeth on genital 

 valves of female one-rowed. Generally the hind margin of wing 

 separating from the anal vein at a lesser distance distad to the 

 cubito-anal cross-vein than the arculus is distal to the cubito- 

 anal cross-vein. Frons acutely angled (males of some species) 

 to right angled (females of some species and both sexes of some). 

 Wings narrower (more or less) in the males than in the 

 females. Group 3. 



a^ Quadrangle long; usually one or about one postquadrangular cell; 

 the cubito-anal cross-vein about midway between the antenodals. 

 Inferior appendages of male small, spine-like ; teeth on genital valves 

 of female several-rowed. The hind margin of wing separating from 

 the anal vein at a greater distance distad to the cubito-anal cross- 

 vein than the arculus is distal to the cubito-anal cross-vein. Frons 

 low, rounded, similar in the sexes. Group 4. Oxystignia. 



Of de Selys' grotip i only two species have been described. 

 Only one of these, ftavoviitalmn, is well known. The gronp 

 was separated from group 2 by de Selys on the following 

 characters: Head more robust; mouth more projecting; 

 abdomen shorter; legs longer; third joint of antenna almost 

 three times as long as the second joint; wings broader; hind 

 margin of wing separating from the anal vein distad to the 

 arculus; quadrangle of medium length. I have seen only an 

 imperfect male and female of flavovittatiiiii. The species is 

 more robust, with slightly longer legs than the species of 

 group 2 known to me. But these characters are not very de- 

 finite, and the venational and antennal characters, as well as 

 the sexual characters, are essentially those of group 2. I have 

 measured the second and third antennal joints of all species 



