ANT COMMUNITIES 



of such a mode was observed before a storm, the gate 

 being shut up by pebbles, earth-pellets, and chippage 

 precisely as at night. One agricultural ant-disk was 

 watched during a rain. The harvesting workers rushed 

 for the central gate from all points of the circle, over 

 which the water was beginning to gather, and in a 

 moment the gateway was choked up by the crowd of 

 insects massed on and around it. I did not think at the 

 time this was done with the intention of closing the gate 

 and shutting out the water accumulated in the plaza, 

 but it had that effect. Possibly it may have been in- 

 tended so. 



It is probable that in such weather conditions the rain 

 that enters the nest gradually descends through the 

 storied rooms and galleries and is partly absorbed dur- 

 ing descent, and at the bottom of the nest is gradually 

 taken up by the underlying ground. It may even be 

 that the lowest cavities, both chambers and galleries, are 

 left uninhabited to receive excess of intrant rains, or 

 are vacated during wet weather, that they may serve as 

 a sort of temporary relief reservoirs. 



Livingstone [Lv. 1, p. 353] notes that the ants of Dilolo 

 (South Africa) manage to preserve their communes upon 

 plains where water stands so long annually as to allow 

 the lotus and other aqueous plants to mature. When 

 all the ant horizon is submerged a foot deep they occupy 

 little houses built on stalks of grass and placed above 

 the line of inundation. Livingstone argues that this 

 must have been the result of experience, since, had the 

 insects waited until the inundation had invaded their 

 subterranean quarters, the required soil for fashioning 

 their elevated nests could not have been obtained. 

 Some of these raised rooms were the size of a bean, others 



286 



