The evaporation of brine is slightly less rapid than that of ordinary pure water, and the boiling point of brine varies with the amount of NaCl present in solution, from 100·21 deg. when only 1 per cent. NaCl is present, to 108·99 deg. when the solution contains 29·4 per cent. of NaCl. A saturated solution of refined table-salt (_i.e._, a solution containing 26·4 per cent. NaCl) has, at normal temperatures, specific gravity 1·2. Salt crystals have specific gravity 2·167 at a temperature of 17°. The salt which separates at high temperature contains no water of crystallization. But when the thermometer falls much below -15° C. the crystals have the composition NaCl.2H₂O and are prismatic in shape. When heated, these give up their water of crystallization and take the simple composition NaCl. Pure sodium chloride is not deliquescent (_i.e._, it does not dissolve and become liquid by absorbing moisture from the air), but, owing to the presence of minute quantities of magnesium chloride (one of the most deliquescent substances known), all except the most refined table-salt appears to be so to a slight extent. Even the finest table-salt is slightly hygroscopic, its crystals absorbing as much as ·6 per cent. moisture from a damp atmosphere. In some of the mines of Cheshire and Austria the very fine saline dust that is diffused through the atmosphere is found by the miners to be extremely irritating to the eyes and lungs, but all the more usual kinds of salt are sufficiently hygroscopic to indicate plainly the condition of the atmosphere. Sodic chloride melts at a very high temperature, and at a still higher temperature it evaporates, while at white heat it forms thick clouds. It would be supposed that in the same ocean areas, the proportion of the salt contents, except where marked differences in temperature occur, would be fairly constant, but it has been demonstrated that, even where masses of water of varying densities are superimposed upon each other, no very complete process of diffusion takes place between them, and practical salt-makers are familiar with differences in density which occur in different parts of the same salt pan. The hardness of a mineral depends upon the degree of cohesion of its particles; but although no unit of hardness has been determined upon, and therefore no accurate method of measuring hardness has been arrived at, minerals have been approximately classed in a comparative table of ten substances, of which talc is placed at one end and diamond at the other. In this table, rock salt appears in the second place, and its hardness is estimated at 2·5. Its cohesion or power of supporting pressure is, therefore, about twice as great as that of bricks, and the practical advantage of this property is fully employed in rock-salt mines, where galleries and roofs are supported upon pillars of salt.