ded stone-ware mixed together, and used in
the same manner. This dries and hardens by the fire, so as to form a
true supplementary retort capable of retaining the materials, if the
glass retort below should crack or soften. But, in experiments which are
intended for collecting gasses, this lute, being porous, is of no manner
of use.

In a great many experiments wherein very violent fire is not required,
the reverberatory furnace may be used as a melting one, by leaving out
the piece called the laboratory, and placing the dome immediately upon
the fire-place, as represented Pl. XIII. Fig. 3. The furnace represented
in Fig. 4. is very convenient for fusions; it is composed of the
fire-place and ash-hole ABD, without a door, and having a hole E, which
receives the muzzle of a pair of bellows strongly luted on, and the dome
ABGH, which ought to be rather lower than is represented in the figure.
This furnace is not capable of producing a very strong heat, but is
sufficient for ordinary operations, and may be readily moved to any part
of the laboratory where it is wanted. Though these particular furnaces
are very convenient, every laboratory must be provided with a forge
furnace, having a good pair of bellows, or, what is more necessary, a
powerful melting furnace. I shall describe the one I use, with the
principles upon which it is constructed.

The air circulates in a furnace in consequence of being heated in its
passage through the burning coals; it dilates, and, becoming lighter
than the surrounding air, is forced to rise upwards by the pressure of
the lateral columns of air, and is replaced by fresh air from all sides,
especially from below. This circulation of air even takes place when
coals are burnt in a common chaffing dish; but we can readily conceive,
that, in a furnace open on all sides, the mass of air which passes, all
other circumstances being equal, cannot be so great as when it is
obliged to pass through a furnace in the shape of a hollow tower, like
most of the chemical furnaces, and consequently, that the combustion
must be more rapid in a furnace of this latter construction. Suppose,
for instance, the furnace ABCDEF open above, and filled with burning
coals, the force with which the air passes through the coals will be in
proportion to the difference between the specific gravity of two columns
equal to AC, the one of cold air without, and the other of heated air
within the furnace. There must be some heat