74 Claude Fuller 



With a sufficiently elastic description, especially as regards 

 soldier characters and the wing measurements of imagos there is 

 no need to sub-divide naialensis into sub-species nor forms. Con- 

 sequently, the forms here suggested are submitted conditionally 

 and in full realization that, in practice, they may have to be 

 discarded. 



Considering the many soldier (major and minor) variants 

 Haviland must have noticed in the numerous nest series he col- 

 lected, it is extraordinary that his descriptions of these castes are 

 so rigid. This, however, I have sufficiently dealt wdth (1915). 



M. natahnsis s.str. is essentially a Natal-Zululand termite and, 

 in that most kaleidoscopic of terrains, the species is seen in its 

 most unstable phase as regards those structural characters of 

 soldiers usually taken as providing the required critena for 

 taxonomic purposes; there is also a considerable variation in the 

 wing-length and some of the imagos I have examined from lower 

 elevations have the head slightly but regularly shorter than the 

 typical form and in these the ocelli tend to be a little more 

 prominent. Further, among the series before me are some imagos 

 caught in Durban that are remarkably yellow. 



In Natal, in the case of each nest series examined, there 

 are always some soldiers (majors and minors) with somewhat 

 thickened heads and these soldiers, as a rule, exhibit angular 

 mesonota. This variant tends to predominate among the nest 

 series brought together by Haviland and now in the Natal 

 Museum. The proportion of such variants is quite irregular, 

 tut in nest series further afield I have not found them nearly so 

 common, indeed soldiers with angular metanota are seldom, if 

 ever, found in the Transvaal. 



Apart from their proportionately longer and straighter mandi- 

 bles, the major soldiers of natalensis can be at once distinguished 

 from those of bellicosus by the less decided sculpturing of the 

 frons. This is regularly shallower and wanting in definition 

 whilst the median ridge, when well expressed is more in the 

 nature of a flatly convex mound and does not extend from the 

 fontanelle to the front margin of the clypeus. There is much 

 variety in the degree to which the sculpturing is expressed, espec- 



