Vol. XXX] KNTOMOI.OMCAL NEWS 295 



not have my present ability in discriminating likenesses and 

 differences in abdominal appendages. The material, upon 

 which the notes of 1906 were made, is now widely scattered in 

 collections and a re-examination of it is impossible. But tor 

 several years past 1 have had it iu mind to repeat the observa- 

 tion again at the earliest opportunity, a> 1 felt some doubts as 

 to the accuracy of my earlier conclusions. 



A chance to make this study came this year, when on June 

 8, 1919, I found G. crassus again abundant at a ripple in the 

 \Yabash where the river flows along the city corporation line 

 of Bluffton, Indiana. Only two gomphines were on the wing 

 at the ripple and these were G. crassus and f rat emus. Of the 

 former seventy-six males and eleven females, and of the latter 

 about twenty males were taken. These specimens, it may be 

 repeated, were all taken at one short ripple on the same day. 



A constant color character not hitherto employed for separ- 

 ating the males of the two species was detected. In fratcrnus 

 the sterna of abdominal segments 8 and 9 are largely yellowish. 

 In crassus the same sterna are largely brown, patterned in 

 lighter and darker brown, with the only pale area a greenish 

 or yellowish bar, or two spots, anterior to the parameres. The 

 seventy-six males of crassus were so colored and the append- 

 ages of all were identical. 



But in the color pattern of the dorsum of the apical abdom- 

 inal segments the seventy-six males of crassus showed a wide 

 variation. The following tabulation bv groups will give a fair- 

 ly accurate statement of the facts : 



Group I. A minute basal dorsal yellow spot on 8, dorsum 



of y and 10 bljick 2 males 



Group II. Like I, but spot on 8 slightly larger 12 males 



Group III. Like II, but spot on 8 slightly larger and segment 



9 with a dorsal basal yellow point 10 males 



Group IV. Segment 8 about as in III or spot slightly larger, 



spot on y slightly larger iS males 



Group V. Spot on S larger than in IV. on segment y about 



one-fourth to one-third the length of the segment 13 males 



(iroup VI. Similar to V. but spot on -.eminent ') one-half <>r 



slightly more than half the length of the segment 10 males 



Group VII. Similar to VI. but spot on segment y wider and 



reaching or almost reaching the apex of the segment... <> males 



