rface of the liquid.

Mr Cavendish and Dr Priestley were the first inventors of a proper
apparatus for collecting these elastic fluids. That of Dr Priestley is
extremely simple, and consists of a bottle A, Pl. VII. Fig. 2. with its
cork B, through which passes the bent glass tube BC, which is engaged
under a jar filled with water in the pneumato-chemical apparatus, or
simply in a bason full of water. The metal is first introduced into the
bottle, the acid is then poured over it, and the bottle is instantly
closed with its cork and tube, as represented in the plate. But this
apparatus has its inconveniencies. When the acid is much concentrated,
or the metal much divided, the effervescence begins before we have time
to cork the bottle properly, and some gas escapes, by which we are
prevented from ascertaining the quantity disengaged with rigorous
exactness. In the next place, when we are obliged to employ heat, or
when heat is produced by the process, a part of the acid distills, and
mixes with the water of the pneumato-chemical apparatus, by which means
we are deceived in our calculation of the quantity of acid decomposed.
Besides these, the water in the cistern of the apparatus absorbs all the
gas produced which is susceptible of absorption, and renders it
impossible to collect these without loss.

To remedy these inconveniencies, I at first used a bottle with two
necks, Pl. VII. Fig. 3. into one of which the glass funnel BC is luted
so as to prevent any air escaping; a glass rod DE is fitted with emery
to the funnel, so as to serve the purpose of a stopper. When it is used,
the matter to be dissolved is first introduced into the bottle, and the
acid is then permitted to pass in as slowly as we please, by raising the
glass rod gently as often as is necessary until saturation is produced.

Another method has been since employed, which serves the same purpose,
and is preferable to the last described in some instances. This consists
in adapting to one of the mouths of the bottle A, Pl. VII. Fig. 4. a
bent tube DEFG, having a capillary opening at D, and ending in a funnel
at G. This tube is securely luted to the mouth C of the bottle. When any
liquid is poured into the funnel, it falls down to F; and, if a
sufficient quantity be added, it passes by the curvature E, and falls
slowly into the bottle, so long as fresh liquor is supplied at the
funnel. The liquor can never be forced out of the tube, and no gas can
escape thro