it was easily fired, either by a burning
lens, or the approach of red-hot iron on the outside of the phial in
which it was contained, and that any part of it being once fired, the
whole was presently reduced to ashes; provided it was previously made
thoroughly dry, which, however, it is not very easy to do.

With this preparation, I found that this paper burned freely in all
kinds of air, but not in _vacuo_, which is also the case with gunpowder;
and, as I have in effect observed before, all the kinds of air in which
this paper was burned received an addition to their bulk, which
consisted partly of nitrous air, from the nitrous precipitate, and
partly of inflammable air, from the paper. As some of the circumstances
attending the ignition of this paper in some of the kinds of air were a
little remarkable, I shall just recite them.

Firing this paper in _inflammable_ air, which it did without any
ignition of the inflammable air itself, the quantity increased
regularly, till the phial in which the process was made was nearly full;
but then it began to decrease, till one third of the whole quantity
disappeared.

A piece of this paper being put to three ounce measures of _acid_ air, a
great part of it presently turned yellow, and the air was reduced to one
third of the original quantity, at the same time becoming reddish,
exactly like common air in a phial containing smoking spirit of nitre.
After this, by the approach of hot iron, I set fire to the paper;
immediately upon which there was a production of air which more than
filled the phial. This air appeared, upon examination, to be very little
different from pure nitrous air. I repeated this experiment with the
same event.

Paper dipped in a solution of mercury, zinc, or iron, in nitrous acid,
has, in a small degree, the same property with paper dipped in a
solution of copper in the same acid.

4. Gunpowder is also fired in all kinds of air, and, in the quantity in
which I tried it, did not make any sensible change in them, except that
the common air in which it was fired would not afterwards admit a candle
to burn in it. In order to try this experiment I half exhausted a
receiver, and then with a burning-glass fired the gunpowder which had
been previously put into it. By this means I could fire a greater
quantity of gunpowder in a small quantity of air, and avoid the hazard
of blowing up, and breaking my receiver.

I own that I was rather afraid of firing gunpowd