318 MONOGRAPH OF THE GENUS SYNTHEMIS, 



turbed, even when sfcruck at repeatedly. The former came and 

 went like a flash of light, so that one might truly say it was gone 

 before one realised that it had come. During a three weeks' 

 stay, I found the males absolutely uncatchable, while of the 

 females I secured only two. 



What, then, makes all this difference between two wings 

 whose venation is on so similar a plan*? It is surely that in 

 Macroniia every superfluous vein has been eliminated, and the 

 whole strength of the wing-material thrown into its most useful 

 channels; whereas in Synthemis, reduction of cells and cross-veins 

 has proceeded only a little wa}* — not so far, in fact, as in any 

 other Corduline genus. Strong evidence of this is the persistence 

 of the cross-veins in the basilar space — an area of the wing 

 Avliich, above all others, needs to be strong and compact, with no 

 lines of weakness. Stronger evidence still is the excessive 

 variability of wing-venation, which can be seen on examining 

 any long series of Synthemis — not to mention the frequency of 

 absolute "freak-venation," of a kind that I have never seen in 

 any other group. Let us examine this venation in detail. 



Cross-veins. — The number of cross-veins in the basilar space 

 varies, in different species, from one to five; in the submedian 

 space, from three to nine. The position of these, especially in 

 the submedian space, is not constant in the same species. For 

 instance, in S. regina male, with five cross-veins, there may be 

 only one beyond the arculus, or two, or one ma}'' be directly under 

 the arculus. In the triangles of both wings, which are usually 

 free, cross-veins often occur in one or moi-e of the four wings, and 

 sometimes even there ai'e two cross-veins in a triangle. 



Recession of tlie hindwing-triangle. — This line of specialisation, 

 stated by Needham to be ttever present in the ^lacromiinm, is 

 almost accomplished (quite in one or two specimens) in Si/nthemis 

 cyanitincta(P]eite vi.). Fig.5 gives the usual position of the 

 arculus, the distance from its base to the inner angle of the 

 triangle being about one-fourth of the whole length of the latter. 

 In all other species the distance is from somewhat under to some- 

 what o\er one-half. (Compare the other figures with fig.5). 



