Triacanthagyna and Gynacantha ^ 



genera may be recognized by the following characters: Upper piece of the 

 arculus equal to or larger than the lower piece, Rs forked, separated from 

 Rspl by at least three cells ; Sc ending at the nodus ; median space free ; mem- 

 branule of hind wings reduced, with only a slight extension on the anal 

 margin ; eyes in contact for a greater distance than the antero-posterior 

 dimension of frons and vertex combined; males with the anal triangle and 

 auricles well developed, and no dorsal teeth or well developed carinae on 

 segment lo; females with two or three long inferior spines on lo. 



Constricted and iinconstricted abdomens. — These terms have been used 

 rather loosely and indefinitely or relatively by authors. The third segment 

 of the abdomen is the region involved. On the second and succeeding seg- 

 ments, except the last two or three, on each side of each tergite two distinct 

 longitudinal carinae can be detected. One of these is the lateral carina and 

 below this, on the margin, is the ventral carina. Both may be armed with 

 small denticles. On the second segment of the male the ventral carina 

 parallels the extreme margin of the genital fossa and often bears denticles 

 of various forms. The lateral carina on the same segment in its anterior 

 part is modified into the auricle. On the third segment the relative position 

 of the lateral and ventral carinae can be accurately described and figured, 

 and serves as a measure for the degree of constriction of the segment. If 

 the carinae are widely separated and approximately parallel the segment is 

 unconstricted. If the carinae fuse at about the level of the transverse carina 

 and. anterior to that point, are joined or are parallel and in close juxtaposi- 

 tion the segment is constricted. If the carinae approach at the level of the 

 transverse carina so they are separated by a distance possibly half the dis- 

 tance of separation near the apex of the segment, the segment may be defined 

 as slightly or somewhat constricted. By defining this character in terms of 

 the position of the carinae it is believed ambiguity or uncertainty in the future 

 can be avoided, and especially in cases where the abdomen is not clearly 

 constricted or not constricted. 



The trifida group {Triacanthagyna) : This group of four species has 

 given me the most difficulty in this study. All of them had been confused 

 under one name trifida, and to add to the confusion TriacantJiagyna needhami 

 is a synonym of the true trifida. The females so far knov/n, strange to say, 

 are more easily recognized than the males. In the latter the appendages are 

 so similar that in three of the sj^ecies I detect no difl^erence and in the fourth. 

 caribhea, dififerences can be detected, probably, only by direct comparison 

 with other species. Collections south of the equator will probably reveal 

 additional species as practically all of the material seen by me is from more 

 northern localities. Although the species are so similar I am nevertheless 

 convinced, after many hours spent studying them, that the four are distinct. 



The nervosa group {Gynacantha) : In the same way in Gynacantha, the 

 nervosa group offers some serious difficulties. In the trifida group no varia- 

 tion within a species in the form of the male abdominal appendages was 

 detected. But among certain species of the nervosa gron^, especially nervosa, 

 bifida, croceipennis, litoralis and interioris, the superior appendage seems 

 somewhat flexible. Flattened out, there is a distinct angle on the inner 



