conditions of heat and moisture
offered by the interior of the ant-hill are favourable to it; it is
not less certain that this arrest is due to the ants. This is shown in
a very simple manner. It is sufficient to prevent the access of the
insects to one of these chambers to cause the grains to germinate
immediately. We can only suppose some direct action of the ants, every
other hypothesis falling before this single fact: the arrested
phenomenon is produced as soon as the _Atta barbara_ no longer acts on
it. Therefore they arrest germination without rendering it impossible,
and when the moment arrives for utilising the accumulated stores,
their first care is to allow the grains to follow the normal course of
evolution. The envelope breaks, the little plant makes its appearance;
radicle and stalk come to light. But the ants do not permit the
development to go too far. The little plant, in order to grow, digests
the starch which is associated with the albumen, for it is not yet
able to draw its nourishment direct from the soil. To be absorbed and
assimilated this starch must first be transformed into sugar. This
chemical transformation being effected, the grain is in the condition
in which the ants prefer it. Like a wine-grower who watches over the
fermentation in his vat, and stops it before the wine turns sour, they
stop the digestion of the starch at this stage. If we do not know how
they retard germination, we know at all events how they render it
impossible at this later stage. It is the young plant which absorbs
the glucose, and which must therefore be destroyed; they cut off the
radicle with their mandibles, and gnaw the stalk; the germ is thus
suppressed. They have not yet finished their manipulations, which must
enable them to preserve without further alteration the provisions
which they have already rendered palatable. They bring out all their
provisions to the sun, dry them, and take them back to the barns. As
long as winter lasts they feed on this sweet flour. An anatomical
peculiarity enables them to make the most of it; their mouth is so
arranged that they can absorb solid particles and eat the albuminous
powder. In this they differ from their northern kin, who are obliged
to feed exclusively on juices.

[56] See chapter on "The Ancient Belief in Harvesting Ants,"
in McCook's _Agricultural Ants_.

I have compared the labours of these ants to those of the wine-grower.
Both of them in fact utilis