On considering all the circumstances, I am inclined to believe, that the properties which distinguish steel from iron are rather owing to a peculiar crystallization or arrangement of the ultimate particles of iron, than to their combination with carbone or any other substance. In all cases where steel is formed, the mass is brought into a liquid form, or nearly approaching to it, a circumstance which allows the particles to be subject to the law of crystallization. We see that great change is made in steel by the mere _tempering_ of it, which cannot be ascribed to the loss or gain of any substance, but to some modification of the internal arrangement of its particles. Why then may not its differences from iron be ascribed to the same cause? It is allowed that steel, by being repeatedly heated and hammered, becomes iron: that is, it should seem, the change of figure disturbs the regular arrangement of the particles. And it may be further observed, in corroboration of the opinion that cast iron is capable of being made permanently magnetic, from its having been in fusion more probably than from its near approximation to steel in its component parts. The most powerful artificial magnets, after being forged of steel, are said to undergo the operation of steelifying again, before they are hardened finally to receive the magnetic virtue. SECTION 17. METALLIC ALLOYS. When two or more metals of different specific gravities are melted together and intimately mixed, they frequently enter into chemical union and form a new compound, called an alloy of the metals. These alloys often possess important properties which their constituents singly do not, and hence become valuable acquisitions to the arts. The metals thus combined may be fused together in any proportion; but if one of them greatly exceed the other in specific gravity, their intimate union is sometimes rendered difficult and even impracticable, partly from the weak affinity and partly from the gravitating principle causing the metal of least specific gravity to arise to the surface. Notwithstanding this union of metals in seemingly indefinite proportions, there are only a few proportions in which the alloys possess peculiar excellences so as to entitle them to the attention of artists. These proportions have in many instances been discovered by experience; and it only remains for theory to point out the reason for such proportions, and to suggest other proportions which may bid fair to possess desirable qualities, and thereby diminish the unsuccessful attempts for improvement in these combinations. That the metals thus alloyed constitute true chemical compounds and not merely mechanical mixtures, may be inferred from the change made in their primary qualities; such as