combination of the metals with other matters, and in that way shall arrive at the safest means to convert such combinations into oxides.

f. Metals and Sulphur.-Metals combine very readily with sulphur, and such combinations are called sulphurets. Iron especially has great affinity for sulphur, and does not part with it even in the highest heat. The process by which sulphur combines with metals, is analogous to the process by which oxygen combines with them. Sulphurets burn; they emit light and heat; and, in all their chemical properties, are almost identical with the oxides. They are distinguished from the oxides by their metallic lustre. They are sometimes translucent, as, for instance, sulphurets of mercury and zinc. Very few sulphurets can be reduced by carbon; but almost all of them by adding alkalies, or a metal which has a stronger affinity for sulphur. This is the case with the sulphurets of copper and lead. If these are melted, and metallic iron added, the iron will combine with the sulphur and revive the metals. Metallic oxides are reduced to sulphurets by adding sulphuretted hydrogen, or sulphuretted carbon; and perhaps this is the manipulation by which, in the laboratory of nature, where sulphuretted hydrogen abounds, metallic oxides are daily reduced. Sulphurets can be reduced by heating them in an atmosphere of hydrogen; by which means sulphuretted hydrogen is formed; this application, however, is very limited, and does not apply to iron or copper. The most common way to reduce the sulphurets, is to transform them into oxides, and then to reduce the oxides. This is most safely done in the chemical laboratory.

Sulphurets are transformed into oxides by roasting and calcining. The material should be pounded to powder, and then heated with access of the atmosphere. Great care should be taken that the mass does not melt; for if this happens, the operation is a failure, and must be repeated. The largest quantity of sulphur escapes as sulphurous gas; and the metal remains in the highest state of oxidation. One part of the sulphur is generally converted into sulphuric acid, and remains with the oxide; another part remains with the metal, and is detected by adding an acid. This especially happens with the sulphuret of iron. Such remains of sulphur can be removed by adding alkalies, or washing with water, the latter of which extracts the sulphates and carries them off; but in case a part of the sulphur is left in the form of sulphuret, the whole mass should be roasted until it is properly oxidized. One way of con

verting the sulphurets into oxides, is very important, and deserves attention. As above mentioned, some metals are not very easily converted into oxides; to effect this conversion, we should melt them together with alkalies, or with oxidized bodies which have a great affinity for water. This law applies equally well to the sulphurets. Sulphurets are, like the metals, very compact, and their atoms are not exposed to the influence of oxygen or any other matter unless when dissolved. If we melt the sulphurets together with alkalies, which, besides dissolving most sulphurets, have a great affinity for water, the aggregate form of the sulphurets is destroyed, and the atoms offer their poles to the poles of other matter; and if heated in the mean time, most of the sulphur is expelled either as sulphurous acid or sulphuretted hydrogen. The rest of the sulphur is generally converted into sulphuric acid, and remains with the alkali. Barytes and lime are in this case powerful agencies; more powerful than even the alkalies. The more permanent sulphurets, melted together with chloride of sodium, are very quickly transformed into oxides. All other salts will act in the same way; and nitrates even better than chlorides; but nitric acid is not so permanent as chlorine, and would be more expensive. This behavior of the sulphurets with alkalies and the salts, is particularly applicable to iron, and may be productive of benefit to the careful manipulator.

If we consider the great affinity of the metals for sulphur, particularly iron, whose affinity is very strong, and consider further the injurious effects of sulphur upon iron, we shall be very cautious in preparing and selecting our ores, for it frequently happens that sulphur exists in ore where we least suspect it; it is not only injurious to the metal, but to the manipulation in the blast furnace. We should, therefore, pay attention to the perfect oxidation of the ores, before we make any use of them. When describing the manipulation in roasting ores, we shall allude to this subject again.

g. Metals and Phosphorus.-Most metals combine readily with phosphorus, especially iron, though not so readily as with sulphur. Carbon is necessary, in almost every case, to produce a combination of phosphorus with metal. Phosphorus is easily expelled by roasting a phosphuret without carbon; but if carbon is present, the phosphorus adheres very strongly to the metal, and its evaporation is difficult. Just so it is with sulphur, for the same manipulation which removes sulphur will remove phosphorus. Phosphorus presents to us no greater difficulties than sulphur. The main difference

between them is the different effect they have upon iron; sulphur makes iron hot-short, and phosphorus cold-short: but phosphorus is advantageous in the blast furnace; the reverse is the case with sulphur.

h. Metals and Carbon; Carburets.-Iron, lead, and potassium have great affinities for carbon; but if simply carburets of iron are to be smelted, the expulsion of carbon may be easily effected by roasting the ore.

i. Metals and Acids frequently exist in native ores. The Halogen combinations of chlorine, bromine, iodine, fluor, are either evaporated, or combine mostly with the oxides, that is, the alkalies, whose metals are to be smelted; they are never injurious. Sulphates are more dangerous, for the sulphuric acid is, in the presence of carbon, decomposed, and leaves the sulphur in connection with the metal. This remark applies equally to Phosphates. Nitrates are not in the least dangerous; for a small heat, with the presence of carbon, decomposes them. Carbonates sometimes require a strong heat, as well as a long time, to be decomposed, which is particularly the case with iron; and as carbonates are most generally protosalts, we never get the higher oxides directly. This makes the roasting of carbonates of iron very difficult; for, if the heat is strong enough to expel the carbonic acid, it is generally strong enough to melt the magnetic oxide, or the protoxide, together with foreign matter. Borates are injurious to the metal, but very advantageous in the furnace. Silicates are directly of no use, particularly those of iron, but may be converted into oxides by being melted with alkalies, and then oxidized. Tellurates, Arseniates, Antimoniates, Wolframiates, Titanates, and Manganates, are very easily converted into oxides, and but slightly injurious in the manufacture of iron. We meet with the whole of these compounds of metals and acids in the native hydrates of the oxides, for in case the ore or hydrate is a decomposition of a salt, the acid is never entirely removed; and should either of the above acids have access in any way to an oxide of iron, we shall surely detect it in the hydrate. Of all the hydrates, that of iron is the most apt to retain acids, partly on account of its electro-positive character, but mainly on account of its forming a great variety of basic compounds, which are more or less difficult of solution. Such basic salts are then mechanically mixed with the hydrates, and are the cause of forming hydrates from the oxides of iron. By all means, therefore, hydrates should be roasted.

XV. Roasting of Iron Ore.

Whether an iron ore should be roasted, is a question which very seldom arises; at least this question seldom ought to arise. With the exception of the red impalpable oxide, the whole body of iron ores require roasting; even the specular iron ore, if it is very compact; but the best oxide, if too compact, works badly in the furnace. All other ores should be subjected to calcination. Some iron masters are in the habit of using the hydrates raw, but this should not be done where clay ores are smelted, for these tend to blacken the tuyere; or where the hydrates contain either chlorides or phosphates. In the latter case, the pig metal will be cold-short, if there is too much phosphorus. Under all circumstances, however, it is best to roast the ores if we expect good metal and well-regulated furnace operations.

The object of roasting ores is either to produce higher oxidation, or to expel injurious admixtures. In both cases, liberal access of atmospheric air is required; we should, therefore, so arrange our roasting operations, as to fulfil these conditions, from which it will appear that different ores require different treatment. To explain this more fully, we shall take a review of the various ores.

a. Magnetic Oxide of Iron.-This ore is very compact, heavy, and of an almost metallic appearance; to open the textures of the ore, to make it more porous, lighter, and to oxidize it more highly, it should be roasted; sulphur is frequently combined with it. This ore melts into a slag by a cherry-red heat; we should, therefore, avoid a high heat, for a melted clinker is useless and injurious in the blast furnace, and a melted mass cannot be oxidized by common

means.

b. Hydrated Oxide of Iron, Brown Oxide, Hematite, Bog Ore.— This whole class ought to be roasted, not for the purpose of oxidation, but in order to drive off the acids, and destroy sulphurets and phosphurets, for all the ores of this class contain more or less injurious matter. This ore will bear a high temperature in roasting, if there is no foreign matter mixed with it; but of this it is very seldom free.

c. Carburets of Iron are to be roasted, partly on account of the sulphur which they frequently contain, and partly for the expulsion of the hydrogen which is generally combined with the carbon. The roasting of this ore is easily effected.

d. Sulphurets of Iron.-These, of course, require roasting, if designed for the manufacture of iron; the manipulation is difficult, and requires more than usual attention and time.

e. Phosphurets of Iron, where they happen to be mixed with the oxides, should be roasted, if we expect medium qualities of iron; but if the quality is no object, and cheapness the aim, then phosphurets, in their raw condition, will answer.

f. Arseniurets of Iron.-If iron ores contain arsenic, it is best to roast them; arsenic does not injure the metal; but if the top or shaft of the blast furnace works cool, there is sometimes danger of choking at the top, or of scaffolding at the lining above the boshes.

g. Chlorine contained in iron ore does no harm whatever, and may be considered beneficial in roasting.

h. Sulphates of Iron should be carefully roasted with liberal access of air. This will apply also to

i. Phosphates.

k. Carbonates require careful treatment. In the furnace they melt before carbon has any influence upon them; and if there is any admixture of foreign matter, the carbonates are very apt to produce but a small quantity of white iron, with black cinder. The roasting of carbonates is difficult; the best means of roasting them are, low heat, and, if possible, access of watery vapors, partly to carry off the heavy carbonic acid gas, and partly to prevent a too high temperature; for, if the heat is too strong, the carbonate melts together with the oxide, and forms a black cinder.

All other ores are easily calcined; they require no particular

attention.

It is evident that, as the qualities of these ores are different, they should receive different treatment; and the question which meets us is, what arrangement, in each particular case, will best enable us to arrive at the highest perfection. For roasting ores, there are three distinct modes of manipulation-ovens, piles, and rows. Each arrangement may be considered perfect for a particular kind of ore; but each is not equally applicable to all varieties of ore. We must modify our manipulations according to circumstances, in order to produce appropriate results.

Under all circumstances the ore to be roasted should be broken into pieces as small as those usually put into the blast furnace, say two or three inches; if we neglect this, of course we cannot expect a good result, for it is obvious that large pieces will not receive heat and oxygen through their whole body so soon as smaller pieces;