50 THE UNIVERSITY SCIENCE BULLETIN. 



two, denoting the closer relationship of uhleri to sanguinolenta 

 than to cinerea. These three forms agree also in having a 

 small cedagus built on the same pattern but differing in minor 

 details. 



Thus it was found that representatives of these three sub- 

 genera, though each with its own characteristic genitalia, yet 

 by these organs alone could readily be divided into three dis- 

 tinct groups. In view therefore of this, combined with the 

 differences in the nymphs and the adults, it has been thought 

 best to raise Kirkaldy's subgenera to generic rank. And this 

 leads us to believe that with similar studies in other groups, 

 similar changes, one way or the other, will be forthcoming. 



The above shows the value of such studies in determining 

 generic differences. It has been found throughout the work 

 that they are also of great value in specific determinations 

 within the genus. So far we have not run across a single case 

 where we could not find specific differences in the genitalia of 

 the species of any genus. It is true, however, that in some 

 genera, such as Idiocerus, these differences may be very slight, 

 and further and careful study must be given to them before 

 they could be used very much in separating the species. Even 

 here, however, it has been found that they have some value, for 

 such species as Idiocerus verticis and nervatus can readily be 

 distinguished by the structure of the cedagus. 



Furthermore, we believe these genitalia will help to settle 

 questions as to the specific or varietal rank of certain forms. 

 Illustrations of this were readily found among the Typhlocyhini 

 as well as among other groups. Thus it was found that Ery- 

 throneura obliqua had a constant form of internal genitalia. 

 When its variety fumida, however, was examined, it was found 

 that in no way could it possibly be considered as belonging to 

 the same species, for the differences both of styles and cedagus 

 could not possibly be as great in mere varieties of the same 

 species. In the styles it was found that the terminal tooth of 

 the latter was invariably much longer and more slender, while 

 the cedagus of the latter was distinctly bifid apically as com- 

 pared with the bluntly apexed cedagus of the former. Then 

 when the variety dorsalis was examined, the style was found 

 to be very different apically from that of the preceding two 

 forms, while the cedagus had a pair of very large and con- 

 spicuous lateral processes of which in the two preceding forms 



