288 Annals of the South African Museum. 



feiTuginous colour passes gradually to Mack, the final result being a 

 uuicolorous blackish pterostigma. 



9 . Very similar to male in colour and pattern. Abdomen com- 

 paratively shorter and much stouter ; lateral yellowish l)au<l broad on 

 segments 3-7, although interrupted by the terminal blackish-brown 

 rings ; 8-10 dull reddish brown with thin greyish pruiuosity. The 

 two colours of pterostigma more sharply defined than in male. 



S , Abd. 44 + 1-5, hdw. 29, pt. 2 mm. ? , 45, 33, 2-5. 



This and the preceding species are so exactly alike in proportions, 

 structui-al characters, venation and colour pattern that the possibility 

 remains open of their being forms of but one species. The present 

 material excludes the eventuality of longlcauchi being the immature, 

 tessehita the mature colour of male (as shown for fasciata, ante), but 

 the two might well be geographical races or subspecies. 



I Br. Subfamily AGRIONINAE. 



Of the Sc'lysian " Legions," three are represented in the present 

 faiina (after removal of Chlorolestes from the " Legion Podagrion " to 

 the Lest inae, as discussed on p. 268). The "Legion Platycnemis " 

 includes, in de Scly's own papers, the genera Allocnemis, Chlorocnemis 

 and Metactiemis. Disimroneura is representative of the " Legion Proto- 

 neura." The remaining genera Pseudagrion, Oeriagrion, Enallagma, 

 Ischnura and Agriocneniis are the South African representatives of 

 the " Legion Agrion,"' one of the largest and most homogeneous units 

 of the entire Order Odonata. We must admit that at various points 

 the limits of the " Legions " as given by de Seh's are doubtful. The 

 present paper is not, one where the questions here arising might he 

 profitably discussed ; nevertheless, small as the number of genera 

 represented in our fauual limits may be, there are two points where 

 accepted views appear in an unfavourable light : When we confront 

 the venation of Allocnemis, Chlorocnemis and Disparoneura (Plate 

 VII, figs. 7, 8, 9) on one side, Metacnemis (Plate VII, fig. 10) on the 

 other side, there can be no doubt about the great analogy between the 

 three of the former group in the region of tlie quadrilateral and 

 the conditions of Cu 2 show a natural line of reduction from 

 Allocnemis through Chlorocnemis to Disparoneura. Further, the three 

 have in common the peculiar shape of head, much restricted in the 

 frontal region with the eyes very strongly prominent ; the terminal 

 appendages of males are built on the same plan ; finally, the colour 

 scheme and its ontogenetic development in Disparoneura is about the 



