alts, occur in a contrary sense during their
cristallization. Caloric is disengaged at the instant of their assuming
the solid state, which furnishes an additional proof of salt being held
in solution by the compound action of water and caloric. Hence, to cause
salts to cristallize which readily liquify by means of caloric, it is
not sufficient to carry off the water which held them in solution, but
the caloric united to them must likewise be removed. Nitrat of potash,
oxygenated muriat of potash, alum, sulphat of soda, &c. are examples of
this circumstance, as, to make these salts cristallize, refrigeration
must be added to evaporation. Such salts, on the contrary, as require
little caloric for being kept in solution, and which, from that
circumstance, are nearly equally soluble in cold and warm water, are
cristallizable by simply carrying off the water which holds them in
solution, and even recover their solid state in boiling water; such are
sulphat of lime, muriat of potash and of soda, and several others.

The art of refining saltpetre depends upon these properties of salts,
and upon their different degrees of solubility in hot and cold water.
This salt, as produced in the manufactories by the first operation, is
composed of many different salts; some are deliquescent, and not
susceptible of being cristallized, such as the nitrat and muriat of
lime; others are almost equally soluble in hot and cold water, as the
muriats of potash and of soda; and, lastly, the saltpetre, or nitrat of
potash, is greatly more soluble in hot than it is in cold water. The
operation is begun, by pouring upon this mixture of salts as much water
as will hold even the least soluble, the muriats of soda and of potash,
in solution; so long as it is hot, this quantity readily dissolves all
the saltpetre, but, upon cooling, the greater part of this salt
cristallizes, leaving about a sixth part remaining dissolved, and mixed
with the nitrat of lime and the two muriats. The nitre obtained by this
process is still somewhat impregnated with other salts, because it has
been cristallized from water in which these abound: It is completely
purified from these by a second solution in a small quantity of boiling
water, and second cristallization. The water remaining after these
cristallizations of nitre is still loaded with a mixture of saltpetre,
and other salts; by farther evaporation, crude saltpetre, or
rough-petre, as the workmen call it, is procured