ME. G. D. HATILAND ON TERMITES. 367 



jaws, they lift themselves up and then hammer their heads 

 against the tree, producing a rattling sound. If left alone they 

 soon retire under cover ; but if one breaks into tbeir retreats, 

 out they come again in great excitement, hammering their heads, 

 opening and shutting their jaws, and discharging their milky 

 secretion. In the section of the fungus-growers to which 

 T. hellicosus belongs the workers run away to their subterranean 

 passages when the nest is being opened, whilst the soldiers stay 

 to defend the nest ; generally the smaller soldiers are more active 

 than the larger, for they run about whilst the larger occupy the 

 crevices of the nest and the cavities of the fuugus-buds, there they 

 wait and bite at anything which comes in reach. The soldiers 

 of this group can generally produce the rattling sound. In this 

 accomplishment, T. carbonarius has reached the highest stage of 

 development, for the soldiers can hammer in rhythmic unison. 

 At first a few begin irregularly, then they get into time, and the 

 others take it up. Every soldier in the exposed portion of the 

 nest stands up and hammers with his head; the blow is given 

 thrice in very quick succession, and then there is an interval of 

 two seconds. The noise they produce reminded me of wavelets 

 lapping on a shore. This trick of hammering with the head is 

 seen in only a few species ; it is clearly a modification of the 

 shaking movements so often seen in workers. 



I have not found a species without soldiers, though Dr. Fritz 

 Miiller found some in America. I have rarely found a nest 

 without "soldiers, though in T. lohatus I have done so. G-eaerally 

 the soldiers are numerous, perhaps about a fifth part as nume- 

 rous as the workers. It is only just before the last moult that 

 they take on their peculiar form. Iq many species there are two 

 kinds of soldiers, the larger and the smaller, without intermediate 

 forms. This is especially noticeable in the group of fungus- 

 growers to which T. hellicosus belongs, and in Itliinotermes, in which 

 the difference is very marked. In two species of soldiers with 

 rostrum and rudimentary mandibles {T. trinervius and T. longipes) 

 a similar condition occurs. 



To the workers I have not paid much attention. The amount 

 of colouring and chitinization is correlated with the period during 

 which they are exposed to light. A broad head, slender legs, and 

 arched abdomen goes with activity and the habit of foraging for 

 food. A narrow head, short stout legs, and fusiform abdomen 

 go with a sluggish habit. The workers not; ouly collect the 



