62 PERFECT SOCIETIES OF INSECTS. 
convenient for their reunion. No sooner does any one 
discover a little chink in the floor, through which it can 
pass below, than it returns to its companions, and, by 
means of certain motions of its antennss, makes some of 
them comprehend what route they are to pursue to find 
it, sometimes even accompanying them to the spot; 
these, in their turn, become the guides of others, till all 
know which way to direct their steps*. 
It is well known also, that ants give each other in- 
formation when they have discovered any store of pro- 
vision. Bradley relates a striking instance of this. A 
nest of ants in a nobleman’s garden discovered a closet, 
many yards within the house, in which conserves were 
kept, which they constantly attended till the nest was 
destroyed. Some in their rambles must have first dis- 
covered this depot of sweets, and informed the rest of it. 
It is remarkable that they always went to it by the same 
track, scarcely varying an inch from it, though they had 
fo pass through two apartments; nor could the sweeping 
and cleaning of the rooms discomfit them, or cause them 
to pursue a different route>. 
Here may be related a very amusing experiment of 
Gould’s.. Having deposited several colonies of ants 
(F. fusca) in flower-pots, he placed them in some earthen 
pans full of water, which prevented them from making 
excursions from their nest. When they had been ac- 
customed some days to this imprisonment, he fastened 
small threads to the upper part of the pots, and ex- 
tending them over the water pans fixed them in the 
eTEubers sly > Bradley, 134, 
