ious enough to retain the
impression until the next day and later; it is scrupulously faithful,
for it guides the column by the same path as on the day before, across
the thousand irregularities of the ground.

How will the Amazon behave when the locality is unknown to her? Apart
from topographical memory, which cannot serve her here, the region in
which I imagine her being still unexplored, does the Ant possess the
Mason-bee's sense of direction, at least within modest limits, and is
she able thus to regain her Ant-hill or her marching column?

The different parts of the garden are not all visited by the marauding
legions to the same extent: the north side is exploited by preference,
doubtless because the forays in that direction are more productive.
The Amazons, therefore, generally direct their troops north of their
barracks; I seldom see them in the south. This part of the garden is, if
not wholly unknown, at least much less familiar to them than the other.
Having said that, let us observe the conduct of the strayed Ant.

I take up my position near the Ant-hill; and, when the column returns
from the slave-raid, I force an Ant to step on a leaf which I hold out
to her. Without touching her, I carry her two or three paces away from
her regiment: no more than that, but in a southerly direction. It is
enough to put her astray, to make her lose her bearings entirely. I see
the Amazon, now replaced on the ground, wander about at random, still,
I need hardly say, with her booty in her mandibles; I see her hurry
away from her comrades, thinking that she is rejoining them; I see her
retrace her steps, turn aside again, try to the right, try to the left
and grope in a host of directions, without succeeding in finding her
whereabouts. The pugnacious, strong-jawed slave-hunter is utterly lost
two steps away from her party. I have in mind certain strays who, after
half an hour's searching, had not succeeded in recovering the route
and were going farther and farther from it, still carrying the nymph in
their teeth. What became of them? What did they do with their spoil? I
had not the patience to follow those dull-witted marauders to the end.

Let us repeat the experiment, but place the Amazon to the north.
After more or less prolonged hesitations, after a search now in this
direction, now in that, the Ant succeeds in finding her column. She
knows the locality.

Here, of a surety, is a Hymenopteron deprived of that sense of di