OBSERVATIONS ON SOME SOUTH APEIOAN TERMITES. 349 



liours — apparently refusing to enter the dry surrounding 

 soil. 



IV. OBSERVATIONS ON THE NESTING HABITS AND 

 GENERAL ECONOMY OF CERTAIN SPECIES. 



Hodoterraes transvaalensis sjj. n. PI. XXV, figs. 2-7, 

 PI. XXVI, figs. 1-5. 



During the pi'ogress of the excavations, of which mention 

 has been made, much care was taken to acquire some know- 

 ledge of the subterranean work of this species. The oppor- 

 tunity was excellent, as all of the ground involved was 

 inhabited by it. Upon the whole, however, the results were 

 meagre. 



A great many hives were exposed, but only two of these 

 were inhabited. The majority were in various stages of 

 dilapidation and others appeared to have had their general 

 contours destroyed by flooding'. The evidence collected went 

 to show that ordinarily the species voluntai'ily deserts its 

 hives and later fills them up with excavated particles of earth ; 

 using them, in fact, as dumping grounds. It is quite 

 possible that desertion is sometimes due to the collapse of 

 the f]"agile interior structure. 



The inhabited hives were both similar, and were in reality 

 central granaries in which the hay harvested is finally 

 deposited. They are doubtless also nurseries, but neither a 

 queen, nor eggs, nor particularly small callows were found in 

 either. The surrounding earth was explored in the hope of 

 tracing some special compartment in Avhich eggs were 

 possibly incubated, but nothing of the sort was found, 

 although many cubic yards of earth were removed. 



There was ample evidence that during the excavation 

 before the hives were reached the insects left them ; one, in 

 fact, was wholly deserted. In the other, which was broken 

 into somewhat suddenly, numerous and quite white callows 

 wei-e found ; none, however, were particularly small, the 



