PAIRING OF ANTS. 251 



large female ants near the surface, but in no re2:ular 

 apartment. He deposited one of these clusters in a 

 box, with some earth, under which they concealed 

 themselves, still keeping together, but did not exca- 

 vate any chamber. "Some time after," he adds, 

 " three or four of these females laid a few eggs, but 

 did not seem to take any great notice of them. For 

 curiosity, I placed in the box a cell of workers, of the 

 same species ; and it was surprising to observe what 

 fondness was expressed. The common ants imme- 

 diately surrounded the females, took care of the eggs, 

 and in a short period made an apartment in the earth 

 fit to receive them. It may also be observed that 

 there were no common ants (workers) in the hills 

 where I found the above clusters. In all probability 

 they were originally large ant-flies, which, having 

 been expelled their colonies, and not faUing victims 

 to their adversaries, associated together in this man- 

 ner, and survived the winter.*" The concluding 

 conjecture shows that Gould mistook the efforts of 

 the sentinel-ants to detain the females for forcible 

 means of expulsion ; so different does the same cir- 

 cumstances often appear when observed through the 

 medium of preconceived notions. 



Huber, in the same way as we have repeatedly 

 done, enclosed several impregnated females in ves- 

 sels filled with moist earth, in which they constructed 

 apartments; laid eggs, of which they took great care; 

 and, though they could not vary the temperature of 

 their habitation, reared some of their larvae till they 

 were of tolerable size, but which perished from Ru- 

 ber's own neglect. " I afterwards," he continues, 

 " placed some other females in a similar apparatus, 

 and delivered to them some pupse of workers, to as- 

 certain if their instinct would teach them to open the 

 covering in which they were enclosed. Although 

 * Gould's Account of English Auts. 



