412 CLAUDE FULLER. 



fall from the nest during November. This is also the case for 

 the Transvaal. It has also been noticed that such swarming- 

 takes place after rain has fallen, and that the full complement 

 does not leave the nest at one and the same time. The 

 crescent-shaped openings in the crusts of the mound through 

 which the winged insects emerge, and the facts that these 

 are closed up aiid subsequently persist for quite a long time 

 afterwards, have also been repeatedly observed. 



When a nest is broken open suddenly numerous represen- 

 tatives of all four castes are exposed ; but shortly such 

 workers as can get out of sight do so. The soldiers, however, 

 swarm over the broken surface, the smaller predominating, all 

 lifting their heads aggressively and discharging viscid fluid. 

 So eager are these little creatures that, where a perpendicular 

 break is made, hundreds of them fall and form a mass at the 

 base. After nightfall the breach is repaired by the workers, 

 and in a little time the mound may be built up to its full 

 propoi'tion again. When a calamity like this befalls the 

 nest, the bulk of the workers attended by some soldiers, and, 

 if present, all the nymphs, retreat into the radiating galleries ; 

 hastening away from the zone of danger. 



Upon the whole this species is very immune from attack, 

 and it is extremely rare to see nests much despoiled. In 

 some, small holes are burrowed, 3 or 4 in. in diameter, most 

 probably by meerkats ; but whether as a foe of the termite, 

 or simply for the sake of making a lair, has not been deter- 

 mined. Haviland^s notes state that jackals appear to feed 

 largely on the species. This seems merely a conjecture. It 

 has also been said that the aardvark destroys the nest, but if 

 this is really so it is extraordinary that I have seen absolutely 

 no evidence of it. 



The species to some extent escapes the attack of true ants. 

 In certain parts of the Pretoria district where many nests of 

 E. bilobatus are intermingled with those of trinervius, 

 whilst every nest of the former species has one ants' nest 

 associated with it, and presents ample evidence of the 

 constant siege going on, nothing of the sort obtains with the- 



