OBSERVATIONS ON SOME SOUTH AFRICAN TERMITES. 409^ 



seldom there is not quite a large percentage of deserted 

 mounds. Why so many mounds are deserted is very baffling. 

 Dr. Warren tells me that, in the course of his study of the 

 biology of this species, he endeavoured to sustain his colonies 

 by watering the nests through the dry winter; but, not- 

 withstanding this attention, many became deserted. Dr. 

 Warren was fortunate enough to see the action of desertion, 

 and by his express wish his account is incorpoi^ated here. It 

 should be stated, however, that the observation was made 

 prior to our full acquaintance with the radiating ga.Ueries. 



One day during the winter of 1913, whilst visiting the 

 ground where he had a series of nests under frequent 

 observation, he noticed a series of small holes 4 to 6 iu. apart 

 and a quarter of an inch in diameter arranged in straight 

 lines all radiating from the mound (PI. XXIX, figs. 16, 

 a, h). Upon looking into these holes it was seen that the 

 termites were hurrying backwards and forwards in an agitated 

 stream. Quite unexpectedly, the insects — nymphs, workers, 

 and soldiers — poured out of the radiating channels, scattered 

 over the surface, and before long died. Some that were 

 taken and placed in test-tubes with moist earth lived a fort- 

 night longer. In addition to this abnormal behaviour, small 

 holes about a quarter of an inch in diameter vvere formed 

 irregularly over the mound and more particularly on the 

 sides. These holes might be plastered up again with moist 

 pellets ; when open they were always fringed by a border of 

 soldiers within, whose closely placed heads, projecting beyond 

 the edge of the hole, formed a couspicuous reddish border to 

 the aperture, as viewed from the outside. 



Subsequently at the sources of the Crocodile River, and 

 again at Platrand and at Pretoria, the writer found, during 

 the winter of 1914, many nests showing these radiating liues 

 of holes. The natural inference was that they were simple 

 perforations in the roof of the galleries ; but upon investiga- 

 tion it was found that the galleries had been deliberately 

 altered and the floor there raised towards the soil-level, the 

 apertures being immediately over the raised part of the runway. 



