smoke_ as being of the same nature, and capable of ignition. But the
smoke of common fuel consists of two very different things. That which
rises first is mere _water_, loaded with some of the grosser parts of
the fuel, and is hardly more capable of becoming red hot than water
itself; but the other kind of smoke, which alone is capable of ignition,
is properly _inflammable air_, which is also loaded with other
heterogeneous matter, so as to appear like a very dense smoke. A lighted
candle soon shews them to be essentially different from each other. For
one of them instantly takes fire, whereas the other extinguishes a
candle.

It is remarkable that gunpowder will take fire, and explode in all kinds
of air, without distinction, and that other substances which contain
_nitre_ will burn freely in those circumstances. Now since nothing can
burn, unless there be something at hand to receive the phlogiston, which
is set loose in the act of ignition, I do not see how this fact can be
accounted for, but by supposing that the acid of nitre, being peculiarly
formed to unite with phlogiston, immediately receives it. And if the
sulphur, which is thereby formed, be instantly decomposed again, as the
chemists in general say, thence comes the explosion of gunpowder, which,
however, requires the reaction of some incumbent atmosphere, and without
which the materials will only _melt_, and be _dispersed_ without
explosion.

Nitrous air seems to consist of the nitrous acid vapour united to
phlogiston, together, perhaps, with some small portion of the metallic
calx; just as inflammable air consists of the vitriolic or marine acid,
and the same phlogistic principle. It should seem, however, that
phlogiston has a stronger affinity with the marine acid, if that be the
basis of common air; for nitrous air being admitted to common air, it is
immediately decomposed; probably by the phlogiston joining with the acid
principle of the common air, while the fixed air which it contained is
precipitated, and the acid of the nitrous air is absorbed by the water
in which the mixture is made, or unites with any volatile alkali that
happens to be at hand.

This, indeed, is hardly agreeable to the hypothesis of most chemists,
who suppose that the nitrous acid is stronger than the marine, so as to
be capable of dislodging it from any base with which it may be combined;
but it agrees with my own experiments on marine acid air, which shew
that, in many cases