SOTTII AFRICAN TERMITES. 27 



dorsal or ventral longitudinals ; and almost invariably issue 

 from these trunks between the arms of the Y-shaped trachea) 

 of the second and third legs, fig. 48 (PI. V), there would 

 appear to be little relation between the termite system and 

 that of the hypothetical type. ]\lore particularly is this so 

 when it is realised that the ever-present spiracular trunks 

 lie between the dorsal and ventral longitudinal trachete, when 

 the two latter are both present. Further, Avhilst there is 

 some evidence to show that the basal connections of the 

 principal tracheae tend to migrate in a posterior direction, the 

 trachea which is attached to the base of the cubital in many 

 termite species and which precedes the median rib of the 

 wing is not necessarily a migrant medial trachea. Where 

 this condition occurs, the evidence all goes to show (a) that 

 the medial trachea has disappeared and (b) that there are two 

 principaKcubital stems (see, in particular, Micro termes 

 in cert us, Section IX y). 



The theory of specialisation proposed by the authors (Corn- 

 stock and Needham) is rather arbitrary. Specialisation by 

 addition takes place only by "a nuiltiplication of the number 

 of the branches of the principal veins," whilst specialisation 

 by reduction follows the atrophy and coalescence of veins. 

 " In nearly every case," they state, " we found the reduction 

 of the pre-anal area accompanied by a similar tendency in the 

 anal area, or, if a reduction has not taken place, there is no 

 increase in the number of veins in this area, the tendency 

 being towards the production of a fewer-veined wing." It is 

 enough for me to add that there is nothing throughout the 

 work under reference bearing upon specialisation by reduc- 

 tion which the ontogeny and final condition of the termite 

 Aving does not seem to confirm. But the venation of the 

 termite wing is such that it cannot be derived from the 

 hypothetical type except by methods that are difficult and 

 unacceptable. If anything, the study adds to the palajonto- 

 logical evidence and supports " the (juite generally accepted 

 view that the primitive insect wing had many wing veins " 

 (C. & N.) 



