CnAP. I. ANT COMMl^NITIES. 31 



iiiteiidence over the others ; but tliis fimctioii is 

 entirely unnecessary in a community where all work 

 with a precision and regularity resembling the 

 subordinate parts of a piece of machinery. I came 

 to the conclusion, at last, that they have no very 

 precisely defined function. They cannot, however, be 

 entirely useless to the community, for the sustenance of 

 an idle class of such bulky individuals would be too 

 heavy a charge for the species to sustain. I think they 

 serve, in some sort, as passive instruments of protection 

 to the real workers. Their enormously large, hard, 

 and indestructible heads may be of use in protecting them 

 against the attacks of insectivorous animals. They 

 would be, on this view, a kind of ''pieces de resist- 

 ance," serving as a foil against onslaughts made on the 

 main body of workers. 



The thu'd order of workers is the most curious of all. 

 If the toj) of a small, fresh hillock, one in which the 

 thatching process is going on, be taken off, a broad 

 cylindrical shaft is disclosed, at a depth of about two 

 feet from the surface. If this be probed with a stick, 

 which may be done to the extent of three or four feet 

 without touching bottom, a small number of colossal 

 fellows (Fig. 3) will slowly begin to make their way up 

 the smooth sides of the mine. Their heads are of the 

 same size as those of the class fig. 2 ; but the front is 

 clothed with hairs, instead of being polished, and they 

 have in the middle of the forehead a twin ocellus, or 

 simple eye, of quite different structure from the ordinary 

 compound eyes, on the sides of the head. This frontal 

 eye is totally wanting in the other workers, and is not 



