Silica is the earth to which we have referred, as being susceptible of fusion when mixed with the oxide of iron. Silica, also, when mixed with the other earths, renders them more fusible than is its own mixture with oxide of iron. Hence it may be stated as a general rule, that ores which do not contain silica, cannot be decomposed without the addition of that earth. The most of our American ores contain silex in sufficient abundance; hence it is usual to add to them, in the process of reduction, carbonate of lime, which is called _flux_. Did not the ore contain silica, this would not produce its effect, and a due admixture of the three earths, silica, alumina, and lime, appears to be necessary to cause the most advantageous results. The remarks of Karsten on this head are new and worthy of attention. "It is upon the choice and the just proportion of the flux, that the profit of the manufacturer in a great degree depends. Employed in too great quantities they fail in the important purpose of giving to the scoriæ a proper consistence. It is very difficult to fix their proportions exactly, and, in truth, these ought to vary with the manner in which the furnace works; but a proportion determined for a state of the furnace when the temperature is neither too high nor too low, is usually adopted. "Chemists and metallurgists, have endeavoured to determine the degree of fusibility of the earths when mixed with each other; but their researches have shed but little light upon the management of blast furnaces. We are, in spite of them, still compelled to have recourse to experience. Far, however, be it from me to depreciate the attempts of Achurd, Bergman, Chaptal, Cramer, &c.; they are valuable at least, in pointing out the road that is to be pursued in the experiments. "It follows, in general terms, from these experiments, that lime, silica, alumina, and magnesia, are infusible when not mixed with each other; that no mixture of earths is fusible without the presence of silica; that the fusion of the oxides of iron cannot take place by the addition of any simple earth other than silica; that ternary mixtures are more fusible than binary; that quaternary mixtures vitrify even more readily, and that the oxide of manganese promptly determines the liquefaction of all the earths.