SOUTH AFRICAN TERMIThJS. 79 



enough, the reproduction of a pliotograph of the wings of 

 this species (fig. 6, PI. Y\, Holmgren's ' Terniitenstndien/ II) 

 shows such an extra very clearly. 



One feature of the Rhinotermes wing which the trachea*, 

 as studied, do not throw any light upon, is the marked extent 

 to which it is ridged. This is so much the case that some riljs 

 are not readily differentiated from the ridges. However, 

 those between the offsets of the cubitus probably arise as they 

 do in Calotermes, and the numerous oblique ridges between 

 pseudo-media and radius may arise from tracheal branchlets 

 of the former that had not developed in the wing-sacs 

 studied. 



IXe. PSAMMOTERMES ALLOCERDS 8ilv. 



Plate X, figs. 167-173. 



In Rhinotermes it has been shown that the medial trachea 

 and therefore the media is normally wanting ; there is some 

 evidence to show that it may be occasionally present. This 

 is also the case in Psammotermes, and later it will be 

 shown that in Termes natal en sis an extended examination 

 of wing-sacs and Avings has failed to show any trace of a 

 medial. Upon the other hand, the medial (and media) is 

 regularly present with Calotermes and Cryptotermes, 

 and although it tends towards decadency in the latter there is 

 ample evidence of what may be described as an effort to re- 

 establish itself as one of the principals. The conclusion one 

 comes to is that in certain generalised termites the medial has 

 become suppressed by the regular and strong development of 

 a duplicated cubital. In Calotermes and Cryptotermes 

 it has been seen that any deviation of the medial from its 

 field or any atrophy that occurs is made good by the cubital. 

 In short, the cubital seems ever ready to assume the chief 

 aerating route of the wing-sac. 



The following remarks upon the wings of Psammotermes 

 are introduced only to illustrate the aggressiveness of the 

 cubital. They are based upon the examination of twenty 



