7. The next important question is, what influence will any mixture of foreign matter have upon the iron ore, so far as the absorption of carbon is concerned? To answer this we would simply say, that, according to science, and experiments in the laboratory, clay possesses the greatest affinity for carbon; next in order is silex; and then lime. This classification, however perfectly true in theory, is not confirmed by practical results. In this case, theory and practice appear to be at variance with each other; but when we take into consideration that the mechanical form of the matter is the cause of this difference in the results arrived at by theory and practice, this apparent exception to a general rule of chemistry is explained. Iron manufacturers generally consider calcareous ore the most favorable of all ores for the manufacture of gray or foundry metal. Clay ore, and then silicious ore, come next in importance. But we should be cautious how far we base practical results upon this experience, for it frequently happens that the theory which we have deduced from practice fails; and from this failure great losses ensue. The above practical rule is applicable only where the lime or calcareous ores are, as is generally the case, already mixed with foreign matter, and where silicious and argillaceous ores are in their purity. Experiments, practically confirmed, made by Mushet, and related in his papers on iron and steel, clearly prove that clay has the greatest affinity for carbon; next to clay comes silex, and then lime. A low temperature and very little fuel will revive iron from a mixture of clay and oxide of iron; but all the iron the mixture contains will not be revived, because clay is infusible by itself, and retains some particles of iron, and of course carbon. The iron is retained as an element, or radical, of an alkali. A stronger alkali is necessary, by combining with the clay, to thrust the iron from its hiding-place. This affinity of the carbon and iron for the clay might be dissolved, if the aggregate form of the clay would permit the formation of larger globules of iron; for these, following the law of gravitation, would separate in spite of affinity. Nearly the same thing happens with a mixture of silex and oxide of iron, with this difference that silex does not absorb carbon so readily as clay, and does not revive iron by so low a temperature, and with so little fuel. But if no alkali combines, at the proper time, with the clay or silex, neither would yield all its iron, even though revived and carbonized. Of these earths, lime is the very last which absorbs carbon and revives iron; but then it precipitates all its iron at once, because carbonate of lime is fusible

by itself, and will, when concentrated into a melted slag, squeeze

the iron out.

m. Agreeably to these principles, clay ores will require a low temperature in the upper part of the stack. We should endeavor to extend this temperature to as low a depth as possible. This will prevent the precipitation of iron before any lime is sufficiently heated to receive the clay, and will consequently prevent the combination of iron and clay into an aluminate, from which it is difficult to separate the iron. Silex or silicious ore is very nearly of the same character, but will permit of a higher degree of heat, without much danger. With calcareous ore we may raise the heat as high in the stack as we please, without endangering the result. These principles, deduced from theory, coincide exactly with experience at the furnace. If we smelt pure calcareous ore-not what is commonly called the limestone ore, for this is generally a precipitate of iron upon a limestone bed, and contains very little lime-we need a strong heat in the furnace, and an abundance of fuel. The reason of this is that the upper heat of the stack and the action of the reviving gases are entirely lost, for lime and limestone ore condense little or no carbon. We thus find that pure calcareous ores are not the most profitable; and we shall make a better use of the fuel, if along with it we mix silicious and clay ore. In this way, we shall not only derive greater profit from the gases, but a lower temperature of the stack will enable us to secure many advantages. Foreign admixtures are thus shown to be unaccompanied with injurious results; but this principle cannot be extended to a chemical admixture or combination. Chemical compositions of silex, clay, and lime are of very difficult decomposition; the very fact that their texture is so close, is the reason why no carbon can penetrate and combine with the oxygen of the iron. This is applicable also to forge and puddling cinder, and to clinkers, and to ore very hardly burnt.

n. From these statements, it is evident that a proper mixture of different ores will be beneficial, so far as the use of fuel is concerned; and that the more closely and intimately the ores are mixed, the better will be the result. A medium temperature is a security that the furnace will work well, and guarantees economy of fuel and a favorable product. Where proper mixtures of foreign matter are already contained in the ore, the most profitable work may of course be expected. Ores of this kind are frequently met with in the coal formations, as precipitates upon limestone or clay. This is the case

at Huntingdon county, Pa., and at other places. The out-crop ores of the anthracite coal series, as well as the Western coal fields, exhibit generally this composition. A great majority of the Western furnaces, such as those at Hanging Rock, and at many places along the Alleghany River, work these ores.

We have, we think, sufficiently proved that the aggregate formthe mechanical composition-of an ore has an important bearing upon the operations of a furnace; but it is obvious that the chemical relations must be still more important. To arrive, by the surest and shortest method, at a clear and comprehensive conclusion, we shall describe the particular behavior of each kind of ore.

o. If we charge a furnace with unroasted magnetic ore, the ore will sink with the coal charges unaltered until it arrives at a certain point, when it will melt into a more or less liquid slag. This slag will pass through a column of hot coal, when a portion of the iron will be revived; another portion will combine with silicious and aluminous matter, and form cinder, which is lost. The iron which results is not gray. The carbonates, and other compact and heavy ores, exhibit the same peculiarities. If limestone is charged along with the ore, a large quantity of iron will be revived; still, a great deal of iron is lost. In no case should we expect gray iron; for, though it should happen that some carburetted iron has been formed in the furnace about the hearth, yet so long as the cinder contains protoxide of iron, the carbon from the gray iron in the hearth will be absorbed, and iron from the cinder revived. The latter is the case when the ores contain foreign matter; but, if the ores contain little or no foreign matter, there will not be sufficient cinder even to protect the iron from the influence of the oxygen of the blast. In this case, the iron must of course be white. The ores may be compact or porous. The result is, in both cases, the same; for, if carburetted iron is formed in the upper parts of the furnace, without a protecting cinder, it will be white before it arrives in the crucible. Satisfactory results cannot be obtained from these ores, unless we have a warm furnace, and unless the heat is raised to a considerable height in the stack.

p. If an iron ore contains foreign matter, and if this matter is a single earth, in itself refractory, the mechanical form of the ore may be the most advantageous; but the metal which results will always be white. When a furnace is charged with clay ore, the ore will, in its descent, absorb and condense carbon. When the carburetted metal arrives within reach of the blast, the carbon will be absorbed

by the carbonic acid, and the iron will arrive whitened in the crucible; the remaining iron yet in the clay will be highly carburetted; but the clay cannot melt and protect the iron. The result is white iron; and, if no limestone is present, an aluminate of iron, as cinder. What we have stated is applicable, in most respects, to silicious ore; also to calcareous ore, with this difference, that, in the latter case, no protoxide of iron is needed to flux the cinder. If, by applying an excess of fuel, we try to revive all the iron from the ore, or, at least, to revive it in greater quantity, then, with clay as well as silicious ore, we receive a tenacious cinder in the hearth below the tuyere, which retains the globules of iron on its surface. If a dark gray carburet comes down, it will soon become white iron within the influence of the blast. Should the cinder not be sufficiently liquid to permit the iron to pass through it, the iron will oxidize, and form protoxide to the cinder, until it effects a passage. The necessity of fluxes is thus clearly seen.

q. If we charge a furnace with poor ore, with an admixture of a refractory character, in a state of fine, impalpable aggregation, as is generally the case with clay ores, and particularly the case with some silicious ores, the iron will be revived by a comparatively low temperature, and for this reason will combine with a large amount of carbon. But this carbon cannot be retained, if the original globules of iron are exposed to the direct influence of the blast; for these grains of melted metal are so small that they can pass through only a very liquid cinder. Should the cinder not be sufficiently liquid, the resulting metal will be white. This is another reason why the smelting of gray iron from clay and some silicious ores is so difficult. To arrive at desirable results, it is advisable to have fine clay ore along with silicious ore. This ore revives a portion of its iron by a low heat, and is, of course, highly carburetted. If the iron produced descends, and finds, on its way, silicious ore ready to deliver iron, it will combine with it, and form a larger mass. If this combination, in its further descent, comes in contact with a calcareous ore, which, under ordinary circumstances, would not liberate iron, the carburetted iron of the clay and silicious ore will draw with it a portion of iron from the calcareous ore; this augmented combination will resist the influence of the blast, and by its ponderability will work, with greater readiness, a passage through the melted cinder below the tuyere. The remaining iron in the clay ore, which, in most cases, amounts to half the original quantity, will be inclosed

in the unaltered piece of ore until it arrives in the hearth below the tuyere. If, at that point, it meets with silicious or calcareous ore, both of which are in the same condition, the different earths, being in a high temperature, will combine, form a liquid cinder, and squeeze the iron out. The iron, having been protected from the blast by the refractory cinder which surrounds it, is now perfectly protected against the blast by the melting cinder, composed of the foreign matter of the different ores.

The case which we have described seldom happens, for there are few clay ores which do not contain a portion of silex; few silicious ores which contain no lime, or magnesia, or clay; and scarcely any calcareous ore which does not contain a portion of clay or silex. The above is a theoretical case, brought forward merely to illustrate a principle. There is a possibility that similar coincidences may exist in practice; but they can happen only very seldom.

r. Experience has clearly proved that, of all ores, those which flux themselves are the most profitable. That is to say, any mixture of ore, or any individual ore, which produces good metal, and a liquid cinder free of iron, is more profitable than those ores which require the interference of art. What constitutes a good cinder, we shall investigate hereafter. We shall confine our attention at present simply to the iron, and to the operations which take place in the furnace. If clay ore, as already explained, yields a portion of its iron very readily, we may infer that this is grayer than any other portion, because carbon combines more easily with iron at a low than at a high temperature: but this carburetted iron is destroyed on account of the refractory quality of the clay. If the clay, mixed with the ore, should contain a portion of lime and silex, its refractory character would be diminished, and the carburetted iron in the inside of the fragment of ore would be more perfectly protected against the influence of the blast. If the carburetted iron thus protected, should find an alkali in the cinder, below the tuyere, waiting to receive the foreign matter, it will descend with scarcely any loss of carbon. From this it is evident that we may expect gray metal from mixed clay ores, if lime or any alkali is present in the hearth; but not otherwise. If the foreign admixtures of the ore are not of such a nature as to form a liquid cinder, the cinder must be made sufficiently liquid by the addition of flux, or by the loss of a portion of iron.

In reality, there are few purely clay, silicious, or calcareous ores.