no. 2089. NOTES ON NEOTROPICAL DRAGON FLIES— WILLIAMSON. 617 



genus.. In 1886, in his Kevision du Synopsis des Agrionines, the 

 American agrionines lacking Cu 2 were considered a genre Protoneura, 



2 subgenera, Microneura (with long legs with long and numerous 

 bristles) and Protoneura (with short legs and moderate, less numerous 

 bristles) being recognized. Under this last, 10 species are described, 

 and a careful grouping of the species, based on venational characters, 

 is worked out. These groupings serve admirably as an aid in the 

 identification of species, but, as might be expected, larger material 

 shows that in a very few details they are artificial. For example, 

 the very different wings, figures 4 and 7 of plates 41 and 42, would 

 fall together; and figure 10, plate 42, combines characters in such a 

 way as to exclude it from both of De Selys's groups. 



Doctor Calvert 1 describes (or redescribes) 6 new species, bringing 

 the Central American species up to 7, for which 7 species groupings 

 similar to De Selys's, but amplified by other venational and other 

 than venational characters, are given. This summary covers briefly 

 about all that has been written on the relationships within the 

 genus of the 16 known species. 



The study of British Guiana material collected by B. J. Kainey 

 and L. A. and E. B. Williamson led to a study of the groupings 

 within the genus in relation both to the known species and to new 

 species in this material. The conclusion has been that groups of 

 generic rank are recognizable, and that it is desirable at this time 

 to separate Protoneura into 4 genera. Venational differences 

 between closely related agrionine genera are much more marked in 

 the legion Protoneura than in the legion Agrion, for example. In 

 the present paper calverti, corculum, and amatoria (see figs. 1, 2, 

 and 3, pi. 41), for example, are considered congeneric, though they 

 differ in two striking venational characters — the proportions of the 



3 antenodal costal spaces, 2 and the ending of Cu^ In the characters 

 of Cu t corculum is like calverti; in its antenodal costal spaces it is 

 like amatoria. If calverti and amatoria were separated generically 

 the question of the generic position of corculum would be in doubt, 

 and two decisions, in the absence of other determining factors, would 

 be possible: One of the two characters could be regarded as the 

 more important and corculum could be associated generically with 

 one or the other species; or a new genus could be erected for corculum. 

 That is to say, with our present knowledge, the existence of the 

 species corculum, having one venational character of calverti and 

 another venational character of amatoria, makes the generic separa- 

 tion of calverti and amatoria impractical. Were the number of 



i Biologia Centrali-Americana, Neuroptera, pp. 140-145, April, 1903, and 394-397, Nov., 1907. 



2 The first antenodal costal space is the space from the wing haso to the first antenodal; the second ante- 

 nodal costal space is the space between the antenodals; and the third antenodal costal space is the space 

 from the second antenodal to the nodns. 



