SOUTH AFRICAN TERMITKS. 81 



radius. They also agree in that the membrane derives its 

 main support fi-om a pseudo-media and cubitus. Tlie principal 

 point of difference is that, inset between the bases of the two 

 outer ribs, there is a short auxiliary, the rudiment of a sub- 

 costa. Other differences that may be cited are: (a) the absence 

 of an anal from the hind wing; (b) of struts between radius 

 and costa (occasionally one such strut is to be found in the 

 wings of T. swaziae); (c) the oblitetated condition of the 

 "reticulation"; (d) the frequent presence of inner branches 

 (sectors) to the radius. The general features of the Termes 

 wing are illustrated in figs. 174, 175 (PI. XI) ; but in these 

 no attempt has been made to indicate the obliterated reticu- 

 lation to which reference has been made, and both wings 

 exhibit sectors to the radius. 



Considerably over 100 wings of one nest series, together 

 with a number of minor lots, were searched without avail for 

 abnormalities of venation, and this is remarkable insomuch 

 that the tracheation of the wing-sac is so extensive and in some 

 particular features curiously variable. The full development 

 of the trachea; is traced in fig. 176 (PI. XI) and parts of the 

 scheme are drawn upon a larger scale in figs. 177 and 178. 

 As in Hodotermes and Rhinotermes this tracheal display 

 originates in the main from two roots or basal counections, 

 the radial and cubital. Figs. 179 and 180 are sketches 

 showing the genesis of wing tracheae, 179 of the meso- and 

 180 of the metathorax. Here the normal condition appears 

 reversed, a greater development having taken place on the 

 cardinal of the metathorax. That this is unusual may be seen 

 from fig. 181, wherein not only is the mesothoracic wing-sac 

 the larger but the post-costal is present, whereas it has not 

 yet put in an appearance in the metathoracic sac. In this 

 particular instance the post-costal is in the position that is 

 normal for Hodotermes, Calotermes, and Crypto- 

 termes. The second development is that of a sub-costal, 

 and later the anal may appear. The only principal trachea 

 that is ever absent is tlie last mentioned, as in figs. 183 and 

 185. It may be present in an atrophied condition, fig. 184, 

 VOL. 4, PART 1. 6 



