334 MONOGRAPH OF THE GENUS SYNTHEMIS, 



enlarged at 6 and 7, and slightly pinched again at 8-9 (except in 

 S. cyanitincta, where 7-9 are uniformly broad). In the second 

 group(figs.9-12) the superior appendages are very short, and 

 slightly forcipate; the corresponding abdomen is shorter, not very 

 pinched at 3-4 (except S. nigra). In the third gronp(tigs. 13-14), 

 the superior appendages are of medium length, slender and straight; 

 tlie corresponding abdomen is exceedingly slender and cylindrical. 

 (In the coloured Plates, figs. 1-7 belong to group 1, 8-11 to group 

 2, 12-13 to group 3; S. nigra, fig.9, has an abdomen varying in 

 shape, but the most slender specimen has been figured in order 

 to contrast more strongly with >S'. guttata, fig. 8). 



As already shewn, there is very little iu the wing-venation 

 which is really constant enough to be of value in subdividing the 

 genus. There is, however, one character, not yet mentioned, 

 which seems to me to be of considerable importance, and that is 

 the form of the membranule in the hind wing. Nearly all Cordu- 

 liincH possess this membranule as a rather narrow, elongated, and 

 often darkened, membi'ane at the base of the anal vein; in the male 

 reaching some distance along the border of the anal triangle; in the 

 female attached to the corresponding curved anal border. Now 

 in the two %^%ciQ% S . Jlavoterminata &.\\di. S. olivet, the membranule 

 is so reduced as not to be visible to the eye, and the anal triangle 

 of the male has developed a slight convexity of its outer border, 

 instead of being slightly concave, as it is when the membranule 

 is present. These two species form group 3 above-mentioned. 

 So that, with their .slender bodies, thin straight appendages of 

 medium length, and this peculiai-ity in wing-venation, they are 

 sufficiently distinct from the main body of species to warrant a 

 new genei'ic name. There is also another important difTerence 

 between them and the rest, and that is the size of the front of 

 the head, which is only about half the width of that of any of 

 the remaining species. I propose, therefore, on the strength of 

 these characters, to separate out S. Jlavoterminata and S. olivei, 

 and to place them in a new genus, Chorlsthemis.* The tj'pe of 

 this genus will be S. Jiavoterminata Martin. 



* Greek X'^P'-^ without, in allusion to the absence of membranule. 



