1902.] 31 



where the median sector arises at one-half that distance: As regards the number 

 of antenodals, Heliocharis and Palceophlebia are typical Calopterygince. 



(4). The position of the nodus varies in both the Calopterygince and the 

 Agrionina, but in general it is farther from the base of the wing in the former than 

 in the latter. Among the first-named it is probably nearest the base (one-third the 

 wing-length) in Amphipteryx and in Chalcupteryx, among the second group it is 

 farthest from the base (more than one-third the wing-length) in Archilestes. No 

 Calopterygince have the nodus as near to the base as one-fifth the wing-length, but 

 many Agrionince have it in that position. 



On the whole, therefore, I believe that Thaumatoneura is rather 

 Agrioninc than Galopterygine in its affinities. 



The question next arises, to what Agrionince is it most closely 

 related ? The existence of a " true " pterostigma, of supplementary- 

 sectors, of forcipated superior appendages in the male, as well as the 

 shape of the quadrilateral, place Thaumatoneura in the legion 

 Podagrion, Selys. The nearest ally therein is Paraphlebia, Selys, of 

 Mexico and Central America, which of all others of this legion has 

 the most complicated reticulation, and which, like Thaumatoneura , 

 has two submedian cross-veins (two nervules in the lower basal cell), 

 the nodus at one-fifth the length of the wing, the post-stigmatical 

 area densely reticulated, the arculus at the second antenodal, the 

 nodal sector arising at ahout the middle of the length of the wing, 

 the subnodal at about the nodal point, the median before the nodus 

 [in P. hyalina, teste Brauer], and very short inferior appendages in 

 the male. Differences between Thaumatoneura and Paraphlebia exist 

 in the number of antenodals, the length of the quadrilateral, the 

 shape of the pterostigma, the exact relative positions of the sub- 

 median cross-veins, and the point where the wings cease to be 

 petioled. None of these differences, however — after what has been 

 said on the number of antenodals— seem to me to be of greater than 

 generic value. 



But Thaumatoneura raises other questions more difficult of 

 solution. Any one who accepts my conclusion that this genus is more 

 Agrionine than Calopteryrjine must also abandon the old distinction 

 between these two subfamilies based on the number of antenodals. 

 Failing this, the next available character is that offered by the point of 

 origin of the median sector, and this gives us the following result: — 



Median sector separating from the principal nearer to the arculus than to the nodus 

 (half-way in Palceophlebia). 

 Antenodals five or more, cross-veins in the postcostal area beginning before the 

 level of the apex of the quadrilateral (except in some Diphlebia). 



CALOPTEBYGIN-E. 



