OBSERVATIONS ON SOME SOUTH AFRICAN TERMITES. 371 



the brood matures, and as it increases, the nest is enlarged 

 by the making of cavities around the original cell ; and these 

 increase in number and size with the growth of the colony, 

 whilst the original brood-chamber becomes the queen-cell 

 and is enlarged from time to time to accommodate her 

 growth. Evidence of this latter contention is supplied in the 

 case of a nest examined for me at Barberton. This was 

 indicated by a low and comparatively small mound, and the 

 queen-cell (PI. XXVIII, ligs. 8 and 4) taken from it is a 

 quite small, thin-walled structure, with a long diameter of 

 60 mm. and a height of 10 mm. ; it is exceptionally entire, 

 having no shelves bracketed to it. The soldiers and workers 

 of this colony were normal in size and plentiful, but the 

 queen was only 40 mm. long. 



The soil removed by a young colony in ni;iking the hive- 

 chamber and driving tunnels is no doubt scattered loosely 

 over the surface — this is usually what happens where this 

 species is found feeding over the open veld on dry grass or 

 the droppings of animals. The mound, it is thought, is only 

 built when the hive-cavity has been enlarged upwards so 

 nmch as to be threatened by subsidence ; in short, when it 

 becomes necessary ; and it is because of this that really 

 small and new mounds are never found. It is difficult to 

 explain why in certain parts mounds are not built; but, as 

 the parts of the country (the Witwatersrand) where there is 

 some evidence of this are high in elevation (5 to 6000 feet) 

 and bleak in winter, it may yet be shown that the species 

 takes advantage of the naturally stony nature of the countiy 

 to construct its hives beneath or amidst sheltering rocks, when 

 the presence of a mound would become non-essential. In this 

 connection it is to be noted that "svlien the species nests under 

 tiled, stone, or brick-flagged floors it extends its nest upwards 

 to the floors. Where these are not laid upon a strong cement 

 bed, subsidence frequently occurs ; indeed it is no exaggera- 

 tion to say that tbe characteristic unevenness of old brick 

 floors to rooms and stoeps in Pietermaritzburg, w'here 

 natal en sis abounds, is entirely due to this termite. 



