swarm leaves.
_Fig. 4_ when a queen has been perfected and left. _Fig. 5_ is a cell
where its occupant has been destroyed by a rival, and removed by the
workers. It will be perceived that each finished queen's cell contains
as much wax as fifty made for the workers.

LIABILITY OF BEING DESTROYED.

In any stage from the egg to maturity these royal insects are liable to
be destroyed;--if honey fails from any cause sufficient to make the
existence of a swarm any way hazardous, the preparations are abandoned,
and these young queens destroyed; (I would here request the reader not
to condemn me for telling more than I can prove, until he has had the
whole story; in the swarming season, I will give further particulars.)

DRONES DESTROYED WHEN HONEY IS SCARCE.

When an occurrence like the above happens, the drones next fall victims
to the failure of honey. A brief existence only is theirs; such as are
perfect, are destroyed without mercy; those in the chrysalis state are
often dragged out, and sacrificed to the necessities of the family.
Such as are allowed to hatch, instead of being fed and protected as
they would be if honey was abundant, are allowed, while yet weak from
the effects of hunger, to wander from the hive, and fall to the earth
by hundreds. These effects attend only a scarcity in the early part of
the season. The massacre of July and September is quite different. The
drones then have age and strength--an effort is apparently first made
by the workers to drive them out without proceeding to extremes; they
are harassed sometimes for several days; the workers feigning only to
sting, or else they cannot, as I never succeeded in seeing but very few
dispatched in that way; yet there is evidence proving beyond doubt that
the sting is used. Hundreds will often be collected together in a
compact body at the bottom of the hive; this mutual protection
affording a few hours' respite from their tormentors, who do not cease
to worry them. In a few days they are gone, and it is a hard matter to
tell what has become of them, at least the majority. If the hive in
September is well supplied with honey, a portion of the drones have a
longer lease of life given them; I have seen them as late as December.
In some seasons, when the best hives are poorly supplied with stores,
the ensuing spring the bees will rear no drones, until the flowers
yield a good supply. I have known one or two years in which no drones
appeared before th