356 THE CANADIAN ENTOMOLOGIST. 



Subfamily I. — Dolicluirina'. 

 This subfamily is easily distinguished from the AmpiiUcin(? by a 

 pointed, not truncate, marginal cell, by the non-carinate clypeus, and by 

 the rounded, or at least unarmed, raetathorax. 



The group is unknown in our fauna, and is found principally in 

 Australia, or at least in the Oriental region, only a single genus Dolichurus 

 being found in Europe. 



According to Kohl, Dolichurus, like Ampulex, preys upon Blattidte, 

 which they store up in their nests as food for their young. 



Three genera fall into this subfamily, and may be recognized by the 

 characters made use of in the following table : 



Table of Genera. 

 Stigma not small, subovate ; submedian cell shorter than the median ; 



metanotum with foveolse 2. 



Stigma small, poorly developed ; submedian cell shorter than the median ; 

 metanotum with several longitudinal carinse. 



Third cubital cell smaller than the second, narrowed towards the 

 radius ; cubitus in hind wings originating a little before the trans- 

 verse median nervure, or interstitial ; metanotum above not 



broadened out laterally Dolichurus, Latreille. 



Third cubital cell quadrangular, much longer than the second ; 

 cubitus in hind wings originating beyond the transverse median 

 nervure ; metanotum above broadened out 



laterally Trirogma, Westw. 



2. Second cubital cell subtriangular, smaller than the third ; cubitus in 

 hind wings interstitial ; metathorax rounded posteriorly; clypeus pro- 

 duced medially \ claws toothed Aphelotoma, Westw. 



Subfamily II. — Ampulicinie. 

 In this family the marginal cell is more or less rounded at apex, with 

 an appendage ; the clypeus is subrostriform and carinate ; while the 

 metathorax posteriorly is truncate, with the upper angles most frequently 

 acute, toothed or produced into spines. 



Kohl would restrict all the species to a single genus, Ampulex, Jurine, 

 treating the others as sections ; but in this I cannot follow him, since I 

 believe these sections are really genera, to which names have already been 

 applied. 



Four genera have been recognized, and are distinguishable by the aid 

 of the following table : 



