372 Annals of the South African Museum. 



not produced in dorsal view, bent downward in a triangular projection 

 near the end in lateral view ; dorsal carinae raised towards the end, 

 rather densely pubescent in distal half. Inferior appendage about 

 two-thirds the length of the superior (much longer than in tristis), 

 slightly narrowed distally and rather deeply excised at apex. 



Wings comparatively narrow ; two rows of cells between CX and 

 Cue, in hind wing. Very slightly tinged with greyish yellow ; in 

 hind wing a deeper yellowish, diffuse cloud between M 4 and anal 

 margin. Membranule brownish black, whitish at extreme base. Costal 

 vein lined with ochreous. Pterostigma very dark brown, almost 

 black. 



Abd. 65 + 5'5, hdw. 53, pi. 4 mm. Length of segments 3, 13 ; 4, 10 ; 

 5, 10 ; 6, 9 ; 7, 5 mm. Breadth : end of 3, 2'3 ; of 5, 2'8 ; middle of 8, 

 4 mm. 



HEMIANAX (Sclys, 1883). 



The single species of this genus is not widely different from Anax. 

 The original definition was based on the absence of supplementary 

 lateral carinae in the abdomen, and a triangular instead of truncate 

 outline of the inferior appendage in the male. Under that definition 

 a second species, the Australian A. papuensis, would belong to 

 Hernia nax. But the supplementary cariuae are rather different 

 between species, being sometimes reduced to a few terminal segments, 

 sometimes only faintly indicated (A. immaculifrons). The shape of 

 the inferior appendage is obviously not a character of generic value. 

 The genus could be retained by applying the neural detail given in 

 the table page (after Karsch !) as distinctive, although even this 

 character is not of great weight. The field between (7, and Cu.-, in 

 hind wing is variable to some extent, even individually, and specimens 

 approaching the Hemianax condition may occur in various species of 

 Anax. The, venation of the single Hemianax, as reproduced in Plate 

 V, fig. 1, shows a wonderful example of highly elaborate specialisation, 

 not by reduction of elements (as in many other Odonata), but by 

 differentiation of the single elements : strengthening or weakening of 

 veins, relative disposition of the various morphologically important 

 forks and ramifications, fusion of cross-veins to form supplementary 

 sectors in the various principal fields, etc. In all those points the 

 Hemianax wing surpasses still the already highly specialised Anax 

 neuration ; and it may well be asserted that Hemianax has the most 

 beautifully organised wing of all living Odonata. I suppose that a 

 mechanical expert (which I am not) would find great delight in a 

 study of this marvellous work of Nature. 



