OBSERVATIONS ON SOJ[K SOUTH AFRICAN TERMITES. 369 



In this connection it is not urged that the mound-shafts are, 

 in any sense, flues through which currents of air are always 

 in circulation. 



Strongly as the earth particles are cemented together, 

 mounds are subject to weathering, and their surfaces have 

 to be repaired. As colonies ai*e subject to decimation, more 

 particulai'ly by aardvarks, it follows that the mounds of 

 weakened communities, and those uninhabited, weather 

 rapidly. In the weathered stage grass creeps over them, or 

 seeds of large plants, which have no great opportunity to 

 become established in the grass, And a suitable nidus and 

 flourish. Then the mound becomes the home of a fresh 

 colony which reconstructs to its own accommodation. ' It is a 

 succession of such phases which brings about in the course 

 of time the huge nest-sites and oases, or park formation. 



The exploration of a large nest-site involves a considerable 

 amount of labour. With a view to ascertaining whether such 

 a nest-site bore evidence of successive habitation, one below 

 the medium size was selected and thoroughly examined at 

 Mt. Edgecombe, Natal (PI. XXYI, fig. 17). This stood in a 

 wattle plantation, the bush having been removed some seven 

 years previously when the land Avas put under wattles. 



The features of this nest-site were as follows. It was a 

 large accumulation of soil, rising in the centre to a rough 

 cone, and it presented, at two points on the circumference, 

 aardvark burrows which penetrated inwards for 4 and 6 ft., 

 going down to a depth of 2 and 3 ft. below the natural level 

 of the ground. The whole mound, including the central cone, 

 was perforated by galleries containing living termites, those 

 of the cone resembling the typical ramifying galleries found 

 in a normal mound. The whole of this mound was removed, 

 and the nest of its occupants was found on the western slope 

 well above ground level. 



No other colonies or traces of cavities were found, but 

 there can be no doubt that such previously existed and were 

 destroyed by the aardvarks. The site of the nest of the 

 existing colony indicated that it originated fi'om a pair which 



