When a steel spring is forcibly bent till it break, it requires less power to bend it through the first inch than the second, and less through the second than the third; the same I suppose to happen if a wire be distended till it break by hanging weights to it; this shews that the particles may be forced from each other to a small distance by less power, than is necessary to make them recede to a greater distance; in this circumstance perhaps the attraction of cohesion differs from that of gravitation, which exerts its power inversely as the squares of the distance. Hence it appears that if the innermost particles of a steel bar, by cooling the external surface first, are kept from approaching each other so nearly as they otherwise would do, that they become in the situation of the particles on the convex side of a bent spring, and can not be forced further from each other except by a greater power than would have been necessary to have made them recede thus far. And secondly, that if they be forced a little further from each other they separate; this may be exemplified by laying two magnetic needles parallel to each other, the contrary poles together, then drawing them longitudinally from each other, they will slide with small force till they begin to separate, and will then require a stronger force to really separate them. Hence it appears, that hardness and brittleness depend on the same circumstance, that the particles are removed to a greater distance from each other and thus resist any power more forcibly which is applied to displace them further, this constitutes hardness. And secondly, if they are displaced by such applied force they immediately separate, and this constitutes brittleness. Steel may be thus rendered too brittle for many purposes, on which account artists have means of softening it again, by exposing it to certain degrees of heat, for the construction of different kinds of tools, which is called tempering it. Some artists plunge large tools in very cold water as soon as they are compleatly ignited, and moving it about, take it out as soon as it ceases to be luminous beneath the water; it is then rubbed quickly with a file or on sand to clean the surface, the heat which the metal still retains soon begins to produce a succession of colours; if a hard temper be required, the piece is dipped again and stirred about in cold water as soon as the yellow tinge appears, if it be cooled when the purple tinge appears it becomes fit for gravers' tools used in working upon metals; if cooled while blue it is proper for springs. Nicholson's Chemistry, p. 313. Keir's Chemical Dictionary. MODERN PRODUCTION OF IRON.