immersed
in water, and again weighed. The piece of gold in Mr Brisson's
experiment lost by this means 3 gros 37 grs.; and, as it is evident
that the weight lost by a body weighed in water is precisely equal to
the weight of the water displaced, or to that of an equal volume of
water, we may conclude, that, in equal magnitudes, gold weighs 4893-1/2
grs. and water 253 grs. which, reduced to unity, gives 1.0000 as the
specific gravity of water, and 19.3617 for that of gold. We may operate
in the same manner with all solid substances. We have rarely any
occasion, in chemistry, to determine the specific gravity of solid
bodies, unless when operating upon alloys or metallic glasses; but we
have very frequent necessity to ascertain that of fluids, as it is often
the only means of judging of their purity or degree of concentration.

This object may be very fully accomplished with the hydrostatic balance,
by weighing a solid body; such, for example, as a little ball of rock
cristal suspended by a very fine gold wire, first in the air, and
afterwards in the fluid whose specific gravity we wish to discover. The
weight lost by the cristal, when weighed in the liquor, is equal to that
of an equal bulk of the liquid. By repeating this operation successively
in water and different fluids, we can very readily ascertain, by a
simple and easy calculation, the relative specific gravities of these
fluids, either with respect to each other or to water. This method is
not, however, sufficiently exact, or, at least, is rather troublesome,
from its extreme delicacy, when used for liquids differing but little in
specific gravity from water; such, for instance, as mineral waters, or
any other water containing very small portions of salt in solution.

In some operations of this nature, which have not hitherto been made
public, I employed an instrument of great sensibility for this purpose
with great advantage. It consists of a hollow cylinder, A b c f, Pl.
vii. fig. 6. of brass, or rather of silver, loaded at its bottom, b c f,
with tin, as represented swimming in a jug of water, l m n o. To the
upper part of the cylinder is attached a stalk of silver wire, not more
than three fourths of a line diameter, surmounted by a little cup d,
intended for containing weights; upon the stalk a mark is made at g,
the use of which we shall presently explain. This cylinder may be made
of any size; but, to be accurate, ought at least to displace four pounds
of