ounce of purified nitre into a glass
retort for distillation and made use of a bladder, moistened and emptied
of air, in place of a receiver (Fig. 3). As soon as the nitre began to
glow it also began to boil, and at the same time the bladder was
expanded by the air that passed over. I proceeded with the distillation
until the boiling in the retort ceased, and the nitre was about to force
its way through the softened retort. I obtained in the bladder the pure
fire-air which occupied the space of 50 ounces of water. This is the
cheapest and best method of obtaining fire-air.

* * * * *

+38. Fifth Experiment.+--I took a silver solution prepared with acid of
nitre, and precipitated it with alkali of tartar; I washed the
precipitate thus obtained and dried it. I then placed this calx of
silver in a small glass retort on the open fire for reduction, and
fastened an empty bladder to the neck. The bladder was immediately
expanded by the air which passed over. After the end of the distillation
I found the calx of silver half melted together in the retort, with its
metallic lustre; however, as I had effected the precipitation with
alkali of tartar, and this is always united with a quantity of aerial
acid which attaches itself to the calx of silver in the precipitation,
so this acid was necessarily present also in the bladder. This acid was
removed from it by milk of lime (Sec. 30, letter _i._), and there remained
behind one-half of pure fire-air.

+39. Sixth Experiment.+--I precipitated with alkali of tartar a solution
of gold which was made with _aqua regia_; I reduced in the foregoing
manner the washed and dried calx of gold. I obtained in this case the
same fire-air, except that no aerial acid accompanied it. This is not to
be wondered at, because the saturated solution of gold effervesces with
the alkali, which does not take place with the solution of silver.

+40. Seventh Experiment.+--It is likewise known that the red precipitate
of mercury regains its flowing condition without the addition of an
inflammable substance. Since mercury, however, really loses its
phlogiston as well by means of vitriolic acid as of the acid of nitre,
it must necessarily assume this again as soon as it recovers its
metallic property.

(_a._) I added a solution of alkali of tartar, drop by drop, to a
solution of corrosive sublimate. I washed the brown-red precipitate
obtained, and dried it; then I placed it