Oomphine Dragonflies from South America, 135 



light no differentiating cliaracters of greater value than the 

 £ollo^yinfr : — 



G. calveriu 



(1) Meso-metathorax with a 

 second antehumeral green stripe, 

 about half as long as the first, very 

 narrow, linear in front, almost 

 touching posteriorly the ante-alar 

 ridge, lying a little above the 

 humeral suture. 



(2) The pale basal colour on 

 abdominal segments 4-7 inter- 

 rupted mid-dorsally, and therefore 

 divided into spots. 



(3) Expansion of inferior lateral 

 margin of segment 8 smaller, not 

 rouuded ; that of 9 not very pro- 

 nounced, parallel and co-extensive 

 with the long axis of the segment. 



(4) Upper anal appendages with 

 the two inferior processes more 

 obtuse ; the superior subapical 

 sin nation longer. 



(o) Fore wings with 18-19 ante- 

 nodals, 10-13 postnodals. Hind 

 wings with 13-14 antenodals. 



(6) Abdomen 37'o mm. ; hind 

 ^ing 28 mm. 



G. camposi. 



(1) Meso-metathorax without a 

 second antehumeral green stripe. 



(2) The pale basal colour on 

 abdominal segments 4-7 not inter- 

 rupted mid-dorsally, and therefore 

 not divided into spots. 



(3) Expansion of inferior lateral 

 margin of segment 8 larger, strongly 

 convex ; that of 9 more pronounced 

 than in calverti, regularly convex. 



(4) Upper anal appendages with 

 the two inferior processes more 

 acute ; the superior subapical 

 sinuation shorter. 



(5) Fore wings with 20-22 ante- 

 nodals, 13-14 postnodals. Hind 

 wings with 16 antenodals. 



(6) Abdomen 43 mm. ; hind 

 wins: 32 mm. 



It was realized that differences of this description might 

 I'ise their significance, if sufficient material of both s]:)ecies 

 became available for study, and, with a view to throwing 

 further light upon the question at issue, Ur. Calvert prepared 

 and sent me camera lucida drawings of the penis and accessory 

 genitalia of G. camposi (Pi. VII. figs. 8 & 9), for comparison 

 with the corresponding structures in Kiiby^s type. Tiie 

 liamules did not seem to be conspicuously different, but, when 

 the penis of G. calverti came to be dissected out and com- 

 pared with the figure of G. camposi, a state of things was 

 disclosed which removed all reasonable doubts on the score 

 of specific distinctness. The camera lucida drawings repro- 

 duced as figs. 7 and 8 show the remarkable difference in form 

 and })roportion of each of the three joints of which the penis is 

 composed, and especially the enormous disparity in the length 

 of the lateral loi)es of the third joint. Tlie vesicle of the 

 penis, too, is widely different in the two species. It may be 

 pointed out that the drawing of G. camposi appears to have 

 been made from the penis extended, but still in situ, while 

 that of G. calverti was made from the organ afier ii had been 

 dissected out. 



