matter could have escaped through the glass, we have a right
to conclude, that the weight of the substance resulting from the
combustion in form of white flakes, must equal that of the phosphorus
and oxygen employed, which amounts to 114.375 grains. And we shall
presently find, that these flakes consisted entirely of a solid or
concrete acid. When we reduce these weights to hundredth parts, it will
be found, that 100 parts of phosphorus require 154 parts of oxygen for
saturation, and that this combination will produce 254 parts of concrete
phosphoric acid, in form of white fleecy flakes.

This experiment proves, in the most convincing manner, that, at a
certain degree of temperature, oxygen possesses a stronger elective
attraction, or affinity, for phosphorus than for caloric; that, in
consequence of this, the phosphorus attracts the base of oxygen gas from
the caloric, which, being set free, spreads itself over the surrounding
bodies. But, though this experiment be so far perfectly conclusive, it
is not sufficiently rigorous, as, in the apparatus described, it is
impossible to ascertain the weight of the flakes of concrete acid which
are formed; we can therefore only determine this by calculating the
weights of oxygen and phosphorus employed; but as, in physics, and in
chemistry, it is not allowable to suppose what is capable of being
ascertained by direct experiment, I thought it necessary to rep at this
experiment, as follows, upon a larger scale, and by means of a different
apparatus.

I took a large glass baloon (A. Pl. iv. fig. 4.) with an opening three
inches diameter, to which was fitted a crystal stopper ground with
emery, and pierced with two holes for the tubes yyy, xxx. Before
shutting the baloon with its stopper, I introduced the support BC,
surmounted by the china cup D, containing 150 grs. of phosphorus; the
stopper was then fitted to the opening of the baloon, luted with fat
lute, and covered with slips of linen spread with quick-lime and white
of eggs: When the lute was perfectly dry, the weight of the whole
apparatus was determined to within a grain, or a grain and a half. I
next exhausted the baloon, by means of an air pump applied to the tube
XXX, and then introduced oxygen gas by means of the tube yyy, having a
stop cock adapted to it. This kind of experiment is most readily and
most exactly performed by means of the hydro-pneumatic machine described
by Mr Meusnier and me in the Memoirs of the Aca