safety those
chrysalides which were at hand and readily obtainable. The latter fact
I could not ascertain, since I feared to disturb the ants at their
interesting labors; but a simple experiment served to show the
feasibility of the idea that the chrysalides were probably within
easy reach of the ants.

Taking possession of one chrysalis which was being conveyed to the new
domicile, I buried it about half an inch deep in the sand, directly in
the track over which the ants were journeying to their new residence,
and a second chrysalis I placed at a little distance from this track,
but in a spot over which numerous ants were running apparently without
any definite aim. The second pupa ant was not buried in any sense, and
was covered merely with a sprinkling of sand. The result in both cases
was negative. No attempt was made to disinter the chrysalis from the
beaten track, although numberless ants walked directly over it; and I
extricated the chrysalis five hours after its interment, and when the
busy scene of the morning had been replaced by a dull prospect, over
which only a single ant now and then hurried in a rapid fashion. The
other chrysalis was also unnoticed, despite its proximity to the
surface of the sand. Whether or not ants want a sense of smell or
other means of guiding them to the whereabouts of their neighbors or
children, is a subject difficult of determination either toward a
positive or negative result. And I am the more inclined to wonder at
the incapacity of the insects to discover their buried companions,
since they appear to be perfectly capable of detecting them at a
considerable distance above ground. When a chrysalis was placed in a
spot remote from the nest, and an ant placed within a foot or so of
the chrysalis, the insect would occasionally seem to be attracted to
the neighborhood of the object. I frequently observed that if an ant
happened to crawl within two or three inches of the chrysalis as it
lay on the ground, it appeared to become conscious of the object,
although at the same time it seemed ignorant of its precise locality.
In such a case the insect would proceed hither and thither in an
erratic fashion, but would continue to hover or rotate around the
chrysalis until it seized the object and bore it off in triumph in its
jaws. Relatively to the size of the ant, we must consider this latter
incident by no means a slight tribute to its acuteness.

The busy scene resulting from the distu