by comparing the heat disengaged with the volume of gas emitted, and
with the rapidity of chemical action. In the case of gunpowder, the
proper manipulation and division of the grains is important, because
favoring _rapid_ deflagration; but in a purely chemical explosion, each
separate molecule is an explosive, and the reaction passes from the
interior of one to the interior of another, suddenly driving the atoms
much further apart than their naturally infinitesimal vibrations.

Purely chemical explosives like nitro-glycerine, gun-cotton, the
picrites, and the fulminates, present a terrible danger from the unknown
mode of the new union of atoms, and reaction of the particles within
themselves, in spontaneous explosions happening in irregular manner.
Some curious circumstances attend the manufacture and use of
gun-cotton,[1] nitro-glycerine, and dynamite. Baron von Link, in his
system of the artillery use of gun-cotton, diminishes the danger of
sudden explosion by twisting the prepared cotton into cords or weaving
it into cloth, thereby securing a more uniform density. Mr. Abel's mode
of making gun-cotton, which explosive is now used more than any other by
the British government, includes drying the damp prepared cotton upon
hot plates, _freely open to the air_. If ignited by a flame, however, in
an unconfined place, gun-cotton only burns with a strong blaze, but
if _confined_ where the temperature reaches 340 deg. F., it explodes with
terrific violence. Somewhat similar is the action of nitro-glycerine and
dynamite, which simply _burn_ if ignited in the open air, while the same
substance will _explode_ through a very slight concussion or by the
application of the electric spark; a red-hot iron, also, if applied,
will explode them when a flame will not. With care, nitro-glycerine can
be kept many years without deterioration; and it has been heated in a
sand-bath to 80 deg. C. for a whole day without explosion or alteration. One
curious experiment is deserving of mention: If a broad-headed nail be