4 E. B. Williamson and J. H. Williamson 



the first cell basal to the first cell extending from Cui to the wing margin. 

 Usually there are scattered or more or less continuous double cells between 

 Cui and the wing margin distal to the first single cell behind Cu^, so the real 

 termination of Cu^ is indefinite, but it seemed that no more precise way of 

 indicating this character could be employed. . There is a real difference in 

 the development of Cu, in different species and the method of counting cells 

 as we have done indicates this as accurately as any other method. In fact 

 the only other method is to begin at the most distal double cell between 

 Cuj and the wing margin and count the marginal cells from there to the 

 origin of Cuj. Another venational character, which occurs occasionally 

 and which gives a rather peculiar venational pattern but which offers no 

 difficulty to precise description or tabulation, is the arising of Mo and R* 

 at the same level. More rarely they are separated at their origin by two 

 postnodal spaces. 



There are two colors of stigmas, black (brown in tenerals) or distinctly 

 reddish brown. Both colors exist in each group. The posterior border of 

 the hind lobe of the prothorax and the mesostigmal lamina offer specific 

 characters. The color of the face, the leg color pattern, the abdominal 

 color pattern, especially of segments 3-5, and in the female of segment 9, 

 and the color of the metasternum differ in different species. In the male 

 the anterior lamina with its posterior process, the second hamule and 

 especially the superior abdominal appendages offer specific characters. 

 In the male of remotus the anal plate is more conspicuous in dorsal view 

 than in any other species. The longer anterior lamina of species of Group 

 I, as compared with Group II, is associated with a relatively longer second 

 segment in species of Group I. The length of the ovipositor is a specific 

 character. 



Historical 



Hagen, in 1862 (in De Selys' Syn. des Agrionines, Legion Podagrion), 

 described the genus Perilestes and the species fragilis from two female 

 specimens, one from Congonhas, Brazil, the other, a smaller female, from 

 the Essequibo River, British Guiana. The specimens were in two Euro- 

 pean museums and there is nothing to indicate that direct comparison of 

 the two was ever made. At the present time there are several reasons 

 for i;hinking the two specimens are not conspecific and no reason for 

 thinking they are. In this paper the name fragilis is retained for a 

 Brazilian species. 



In the Revision du Syn. des Agrionines, Premiere Partie, 1886, De Selys 

 adds to the original generic description and describes a male and female 

 (or females) from the Amazon, referring them doubtfully to fragilis 

 which he knew only from Hagen 's inadequate description, but in the 

 final paragraph proposing the name cornuta for the Amazonian speci- 

 mens if they should later prove different from the two females described 

 by Hagen. The name cornuta {cormitiis) is retained in this paper for 

 one of the five known Amazonian species and it is assumed that the two 



