the changes that have taken place. Yes, it is sight, but a very dull
sight, whose horizon is altered by the shifting of a few bits of
gravel. To this short sight, a strip of paper, a bed of mint-leaves, a
layer of yellow sand, a stream of water, a furrow made by the broom,
or even lesser modifications are enough to transform the landscape;
and the regiment, eager to reach home as fast as it can with its loot,
halts uneasily on beholding this unfamiliar scenery. If the doubtful
zones are at length passed, it is due to the fact that fresh attempts
are constantly being made to cross the doctored strips and that at
last a few Ants recognize well-known spots beyond them. The others,
relying on their clearer-sighted sisters, follow.

Sight would not be enough, if the Amazon had not also at her service a
correct memory for places. The memory of an Ant! What can that be? In
what does it resemble ours? I have no answers to these questions; but
a few words will enable me to prove that the insect has a very exact
and persistent recollection of places which it has once visited. Here
is something which I have often witnessed. It sometimes happens that
the plundered Ant-hill offers the Amazons a richer spoil than the
invading column is able to carry away. Or, again, the region visited
is rich in Ant-hills. Another raid is necessary, to exploit the site
thoroughly. In such cases, a second expedition takes place, sometimes
on the next day, sometimes two or three days later. This time, the
column does no reconnoitring on the way: it goes straight to the spot
known to abound in nymphs and travels by the identical path which it
followed before. It has sometimes happened that I have marked with
small stones, for a distance of twenty yards, the road pursued a
couple of days earlier and have then found the Amazons proceeding by
the same route, stone by stone: