brick, and, besides weakening, exerts an injurious influence in the puddling furnace. Plumbago is frequently mixed with fire-clay in making crucibles. It is added, not for the purpose of enabling the crucible to resist a higher heat, but for the purpose of preventing it from breaking, when suddenly cooled and heated. Such a crucible does not resist as much heat as one made of silicious matter; but the latter is very liable to break during or after the first heat. The plumbago crucible, if well made, will endure ten or twelve heats in melting iron.

b. Artificial sandstone is an article very little used in this country; but as cases may occur in which it can be of service, we shall devote a short space to its consideration. At many establishments on the Continent of Europe, artificial sandstones are used instead of fire brick, and strange as it may appear, they are used for hearth and boshes in blast furnaces. They can be made from any good, coarse, silicious sand, which does not melt in a high heat. Sand generally contains some foreign matter, particularly lime, of which even river sand is not free. Such matter must be removed. The surest method of procedure is, in all cases, to burn the sand, pebbles, or native sandstone, in a cherry-red heat; the lime, which it may contain as a carbonate, may thus be burnt, when it can be removed by pounding and washing. In large iron manufactories, this branch of the business is quite an extensive one, and the mills for pounding and grinding receive considerable attention. Where coarse pure sand cannot be obtained, which is frequently the case, particularly in the coal regions, silicious or white river pebbles may be employed; or, if these can not be had, white coarse sandstone, or millstone grit. This is burnt, pounded, and washed. The artificial sand, thus derived from pebbles or stone, is mixed with about one-fourth or one-sixth of its amount of fire clay, or with just a sufficient quantity to keep the mass together after being dried. The finer the sand has been pounded, and the more tenacious the clay, the less of the latter which will be required. Before the clay is mixed with the sand, it is burnt and pounded. Clay should be burnt under all conditions, for raw clay contains a large amount of water. If once burnt, it will not absorb so much moisture as it previously contained. If fire brick is made from clay containing a large amount of water, or from green clay, it will be porous; for the water which is evaporated from the interior of the brick of course occupied a certain space. If the pounded sand is too coarse, or if the grains are round, spaces will be left between them, to fill

which a large amount of clay will be required; in such a case, the stone will not glaze well, when exposed to heat. Therefore, the artificial sand and burnt clay are moistened with as little water as possible, and mixed together thoroughly; the latter object may be best effected by means of edge-wheels. The damp mass is formed into bricks in the common way. These bricks may have any form most conveniently adapted for our purposes. In rolling mills, specific forms are desirable for the various corners, angles, and arches of the puddling and reheating furnaces. After being stored under an open shed, and air-dried, they are ready for use. It would be a vain attempt to burn such an artificial sandstone, for the highest heat of the reheating furnace would scarcely glaze it, if the iron did not form an alkali; by this means, a glazing at the surface of the interior is effected. Such stones, if not cheap at first cost, are easily cut, easily laid, easily removed, and no loss arises from spales or bats, for any unglazed remains of the stone are easily transformed into new brick again. However useful a material in the rolling mill, they are less adapted for the fireplace, and the lining of the hearth of a puddling furnace. The least touch with an iron tool will destroy them; but where no mechanical or chemical action is exercised, they are equal, if not preferable to the best fire brick.

Of such artificial sandstones, the hearth and boshes of blast furnaces are built, and are said to answer well. If the raw material has been carefully managed, the statement may be true. We know that, in the coke blast furnace of Silesia, a similar mass is used, which is pounded in moist; the hearth forms a single sandstone. This kind of hearth is very durable, and commonly endures a blast of twenty-four months, and occasionally a blast even of four or five years. In the preparation of the mass, less reliance can be placed on its composition than upon the careful mixing of the compound. Air-dried stones may be made of pure fire-clay; but these are not so durable as the sandstones of which we speak. These, as we have stated, ought never to be made of green clay, even if the clay is of the best quality. Clay and silex, be it observed once for all, are the only serviceable materials for fire-proof stones.

c. The joints of fire brick, sandstones, and artificial stones are to be thoroughly filled by mortar. The mortar ought to form a kind of solder between one stone and another, and may, for this reason, be more fusible than the bricks or stones themselves. Pure fire-clay is not a good mortar, for it cracks in drying, and leaves spaces,

which occasion the destruction of the stone. A mixture of fireclay and fine sand is preferable, not because it does not melt, but because it shrinks less in drying. The very best mortar for hearthstones and fire-brick is made of a mixture of fire-clay and finely pounded blast furnace cinder; this mortar will cement stones and bricks firmly.

III. Conductors of Heat.

In concluding this chapter, we shall make a few remarks on the capacity of matter to conduct and reflect heat, so far as this subject has any relation to the object of our work. Metals are the best conductors and reflectors of heat; therefore, for its preservation iron or copper pipes are evidently unserviceable. To conduct hot air in iron pipes is a violation of established principles, but, unfortunately, we cannot substitute for such metal any material of less conducting power; the same remark is applicable to steam-conducting pipes. A boiler of fire-clay would be useless; for, in addition to its weakness, a great deal of fuel would be required to raise steam in it. This applies equally well to air-heating pipes. An iron roof on a puddling furnace would not answer, even though the iron should not melt; for the furnace would not retain sufficient heat to work the iron it contains. It is a good arrangement to place pipes vertically, if we desire to retain the heat; but if we wish to conduct it, by the medium of the pipe, from the fire to the interior of the pipe -an object we seek to secure by employing heating apparatus and steam boilers—it is entirely wrong thus to place them. In these cases, heat is conducted by contact, and by the motion of gases; for this reason, we should employ the best conductors of heat, and put these in such a position as to offer the largest surface to the moving gases. We may consider this surface extended, if the heated particles of air can change position among themselves. This object may be effected by exposing a convex surface to the current of the hot gases.

The reflective capacity of matter depends, in some measure, on color and polish. A bright surface will reverberate more heat than a dull surface. This fact we may observe in a new puddling or reheating furnace, for a furnace with an unglazed roof can not be well heated. No furnace works well until its whole interior surface is glazed. If it were not possible to glaze fire brick, the strongest heat would not make a furnace sufficiently warm for puddling or heating. A roof of carbon, black and velvety, would not heat a puddling furnace red hot; from which we may infer that white fire brick or

stones produce a higher temperature at those points accessible to their reflected heat than brick or stones, which are of a dark color. From this, we may easily understand why a furnace hearth gets cold-unaccountably to him who does not bear in mind the factwhen we smelt black cinder in the blast furnace, and why an excess of limestones or gray cinders draws the heat into the hearth. Other circumstances being equal, the heat will be greatest at those points which exhibit the brightest color, or polish, and at a concave surface whence the reflected rays of light are thrown into a focus. With the latter part of this theory all are practically acquainted who understand the nature of the lens.

From these important considerations, we conclude that dark fireproof stones are an unserviceable material; that a glazing of the stones and fire brick is essentially necessary, particularly in the fire-chambers and hearths of the puddling, reheating, and blast furnaces; that the roof and bottom of a puddling furnace ought to form the two surfaces of a lens, so as to throw the highest heat above the bottom, and between the bottom and roof; and that the roof of a reheating furnace should be as straight as possible, so as to produce a uniform heat over the entire bottom. By reference to these laws of physics, we are able to understand the influence exerted by steep boshes in a blast furnace, and to know where the heat is thrown, when the boshes are drawn from the widest part of the furnace down to the tuyere; this is frequently done, and answers an excellent purpose. A greatly tapered hearth will throw the highest heat above the tuyere, while a cylindrical hearth will retain it just at the tuyere. From this, we may understand that a furnace whose hearth is injured does not work well; for, in that case, particularly where there are more tuyeres than one, a lens is formed by the concavities of the tuyeres, and the highest heat of the furnace is thrown below them, at which point it has the effect of reducing cinder and impurities along with the iron. These observations enable us to understand why ores of a refractory nature do not work well in a narrow hearth, and why fusible ores work most profitably in a high cylindrical hearth and flat boshes. In the former case, we require a high preparatory heat in the stack; in the latter, a very low heat.

CHAPTER X.

MOTIVE POWER.

The consideration of this subject belongs to those works which treat specifically of mechanical science; still, as motive power is extensively applied in iron manufactories, a few remarks in relation to it may not be without interest and profit.

Lack of power is one of the worst evils to which an iron factory can be exposed. If power is deficient, nothing works rightly; everything is in disorder. Sometimes we are compelled to modify operations, and reconstruct apparatus, where it does not bear a certain relation to the engine or the waterwheel. The amount of power necessary for the different branches of the business varies according to circumstances. Where the blast machines are well constructed, a power of sixteen horses is required at a charcoal furnace, a power of forty at a coke furnace, and a power of sixty at an anthracite furnace. By means of the latter power, applied in a rolling mill, with squeezers, we ought to produce 100 tons of small and coarse rod, and hoop iron, or 200 tons of rails and heavy bar, per week. A sheet iron factory is never profitable if it lacks motive power; in fact, the profits of such an establishment chiefly depend upon its successful application. The power exerted by a waterwheel cannot be calculated except by means of a dynamometer; we refer to those who are interested in this subject to works on hydraulics, whence they can obtain all necessary information. Waterwheels are seldom used in iron manufactories; the application of waste heat to the generation of steam renders steam-engines, in the opinion of all experienced manufacturers, a much cheaper element of power. But the purchase of steam-engines requires great caution; inattention to this matter has resulted in great loss to establishments otherwise well arranged. One horse power is considered equal to 33,000 pounds, lifted one foot high in one minute. Many other equivalents have been proposed by various engineers; but this, established by James Watt, is generally adopted. Where