220 Journal Xew York Entomological Society. f^"o'- ^i^- 



unarmed was also a prime character with many writers. Various 

 species of Pseudagenia have distinct spines on the hind tibiae, and the 

 males of certain Priocnemis have the hind tibse almost mutic. 



Thomson was the first to discover a really valuable character in 

 the furrow on the second ventral segment. This unfortunately breaks 

 down in the males of some Pseudagenia, but by using it in connec- 

 tion with a hitherto unmentioned venational character (as in the 

 following table) one may recognize two groups. Fox, and others, 

 laid much importance on the development of the labrum. In Nofo- 

 cyphus and Ceropalcs it is entirely exposed, but in many forms it is 

 more or less exposed, and in P. maura and P. mterrnpta it is usually 

 so much exposed that Ashmead put these in the group with Ceropales. 

 Mr. Rohwer informs me that Ashmead's type of Ageiuoxcnns is 

 merely a specimen of P. interrupfa. 



Kohl made a classification of the family, dividing the Pompilns 

 group up into many sections, sometimes without names. One of his 

 main characters is the position of the origin of the cubitus in the 

 hind wing, whether before, interstitial with, or beyond the end of the 

 cell beneath. I doubt if this character is stable for the European 

 forms, but certainly when applied to our species it is full of pit-falls, 

 and should be used only in connection with other structures that may 

 check it. 



Ashmead, with catholic liberality, accepted all the preceding classi- 

 fications and merged them into one, adopting in some sense nearly 

 all of the former genera and making numerous new ones. His greatest 

 mistakes were in trying to apply the character of the exposed labrum 

 in a strict sense, and in using Kohl's character of the position of 

 cubitus in the hind-wing for our American species. Many of his 

 genera represent natural groups, but his characters for them are often 

 inconstant or inapplicable. It is impossible to make a new classifica- 

 tion in a day, and in presenting that below I have used all that I 

 could find of value in previous systems and have had the fortune to 

 discover a few additional points that may be of use in arranging our 

 species. 



Ashmead's genus PonipiUnus is a good one, when made to include 

 some other forms, but his distinguishing characters are not usable; 

 however, I find that this genus, and one or two allied ones, may be 

 separated by the absence of erect hair on the metanotum. I am not 



