ART. 5 GENERIC REVISION OF THE FGSSORIAL WASPS — PARKER 3 



assigned to this feature. My examination of these structures in the 

 several genera was not extensive enough to warrant the drawing of 

 broad conclusions, but it was carried far enough to show that in tlie 

 genus Bembh', in which the typical numbers are four segments for 

 (he maxillary and two for the labial palpus, the variation in the 

 number of segments in the palpi of some species is such as to lessen 

 the value of these structures for generic purposes. Variation in 

 vepect to the number of segments in the palpi was also found in sc»me 

 species of Steniolla and Microhemhex, but the number of individuals 

 examined in each genus was too small to warrant any other state- 

 ment than that variation does occur. In those genera in which the 

 typical number of segments for the maxillar^^ palpus is six and for 

 the labial four, I did not find variation in the limited number of 

 individuals examined. Since these numbers, however, are regarded 

 as indicating primitive conditions, this lack of variation is readily 

 explained. It was this search for variation in the mouth parts that 

 led to the discovery that the number of segments in the palpi of 

 Monedula chUensis Eschscholz is five for the maxillary and three 

 for the labial instead of six and four as had heretofore been supposed. 



Since in this tribe (Bembieini) the ocelli are much distorted or 

 are reduced to cicatrices. I have found the anterior ocellus providing 

 one of the most reliable characters on which to separate tlie tribe into 

 genera. Although I have found within a genus some A-ariation with 

 regard to the extent to which the reduction or distortion of the 

 anterior ocellus has been carried, I have found no variation in the 

 form which that reduction shows or to w'hich the reduction is tending. 

 Another character in which I have found little or no variation within 

 a genus is the pattern of the spatha of the male genitalia. With the 

 exception of Thernpon and Tiichostictia, in which genera the spatha 

 It almost identical in form, each of the other genera has its own 

 peculiar form of the spatha. I have refrained from using this 

 character in my keys for distinguishing the genera since there is no 

 corresponding character in the female, but for all that, it is a char- 

 acter that must be taken into consideration. In addition to the char- 

 acters cited above, I have made use of the pubescence of the eyes, the 

 form of the dorsal border of the clypeus, and of certain differences 

 m the venation of the wings. 



In the accompanying table I have attempted to show my concep- 

 tion of the relationship existing among the several genera included 

 in this revision. I have not, however, included in the table the 

 genus He7nidida, for the simple reason that I have had no opportu- 

 nity to study a single specimen of Monedula singularis Taschenberg. 



'J'he difficulty met in an attempt to show exact relationships among 

 genera by a linear arrangement becomes apparent at once. If. as 

 is generally conceded, the presence of unimpaired ocelli labial palpi 



