10 University oj Michigan 



subtriangles and supratriangles all free, absence of trigonal supple- 

 ments and sectors of Rs and M4, and by having the proximal angle 

 at the point of separation of Aj from CU2+ A,, nearly a right angle, 

 and not as obtuse as the distal angle at the same point. Its 

 affinities are with the Hagenius series, and since I regard the latter 

 as a series undergoing reduction and because of the uncrossed 

 supratriangles and subtriangles in the Zonophora series, I believe 

 the latter series has undergone reduction. The genus is repre- 

 sented by three species confined to the Neotropical Region. 



5. The Hagenius series. — This is another small compact series, 

 characterized by a distinct trigonal supplement in both wings, 

 with the distal side of the triangle posterior to the attachment of 

 the trigonal supplement distinctly concave. In the other charac- 

 ters of the triangles, supratriangles, and subtriangles, this series 

 resembles the Zonophora series. But the Hagenius series is dis- 

 tinct from all others, excepting the Gomphus series, in that a con- 

 siderable reduction of the cross-veins between M,_3 and M4 has 

 taken place. 



Hagenius with a single species (unless a second species, gigas, 

 from the Oriental Region should prove to be really congeneric) 

 occurs in the Nearctic Region; and Sieboldius, with three species, 

 occurs in the Palaearctic and Oriental Regions. 



The genus Davidius has generally been associated with Hagen- 

 ius and Sieboldius. I do not know the genus well enough to be 

 certain, but I believe it belongs to the Gomphus series, being prob- 

 ably the most primitive member of that series. If this is correct 

 free triangles is not an infallible character of the Gomphus series 

 any more than crossed triangles is an infallible character of the 

 Progomphus series. 



6. The Diastatomma series. — I know the venation of the genus 

 Diastatomma only from the figure in the Monographie des Gom- 

 phines, and the venation of Lindenia and Isomma not at all. 

 Hence, I have to assume that the Legion Lindenia of de Selys is a 

 single series and that the genera I know, Gomphidia, Ictinus, and 



