92 HYMENOPTERA. 



is impossible. This is especially true of the group of species 

 most nearly related to P. a'esso7ii and viodestus. 



Fortunately there is a group of characters in the males 

 which usually makes separation in that sex a simple matter. 

 These are the characters of the genitalia ; the genital arma- 

 ture and the retracted seventh and eighth ventral plates. 

 These organs are very charateristic and easily recognized, 

 and as a rule are extremely constant. In the present work 

 over two hundred dissections of these have been made, and 

 in nearly all cases it was evident at a glance to which species 

 they belonged. In only one species are they really variable, 

 and then but slightly. This is the species modestus which is 

 extremely inconstant in all characters. 



By means of the genitalia the genus is naturally divided 

 into three very distinct groups or divisions, which include 

 all the species except the peculiar basalts. These I have 

 called the "Modestus Division ", the " Cressoni Division" 

 and the " Asininus Division ", after three characteristic spe- 

 cies in them. The males can also be separated into these 

 three groups by the character of their face marks, as will be 

 seen later. The more similar the species are in external 

 characters the more similar are their genitalia, or vice versa, 

 and the relationship seems to be especially close between 

 the face markings and the genitalia, without much regard to 

 sculpture or size. There are, however, some exceptions like 

 P. coloradensis, in which the face marks are very different 

 from, and the genitalia almost identical with some other 

 species. 



The genus Prosopis is apparently rapidly evolving and the 

 path of the developments can often be quite plainly followed. 

 The species modestzis, again, is peculiar in this respect. It 

 seems to be in the process of breaking up into distinct groups, 

 and this apparently affects all its characters. The series of 

 figures of the seventh ventral plate (69-72), and the face 

 marks (93-100) of the species will illustrate this point. In 

 a few instances, on the other hand, the evolution has not 

 affected the genitalia, but has exhibited itself in other char- 

 acters. P. episcopalis, for example, is practically identical 



