attempting to weld too great a faggot of square bars at once, as in that case it frequently happens that at the centre the bars are not welded at all. A few bars should therefore first be welded together, and when soundly joined, other layers of bars may be packed around this central core: the process being continued until the required size is attained, which can thus be effected with perfect success. Another method consists in first making a round core or heart, and packing around this bars of a V form; this method is adopted frequently for forging railway axles, and it was also employed in the manufacture of the monster gun at the Mersey Ironworks. From a paper by Mr. Clay, we find that the method employed in forging this gun was as follows. The gun was built up in seven distinct layers or slabs, the forging occupying seven weeks, and it was found that the metal, after being worked, was improved in strength rather than deteriorated, by the long exposure to great heat. The chief points to be considered by the designer of the gun were, to obtain sound weldings, to place the iron with its fibres in the proper direction for resisting the most severe strain to which it could be exposed, and to take care that while one part of the forging was being worked, other portions were not wasted under the action of the furnace by burning or crystallization.

The first operation was to prepare a core of suitable dimensions and nearly the whole length of the gun. This was done by taking a number of rolled bars about six feet in length, and welding them together, and then drawing them out until the proper length was obtained: a series of V-shaped bars were now packed round the core, heated in a reverberatory furnace, and forged under a large hammer, Another series of bars was next packed on, the mass again heated, and worked perfectly sound. Another longitudinal series of bars was yet required over the whole length of the forging, after which the work was about fifteen feet in length and thirty-two inches in diameter, but requiring to be augmented to forty-four inches at the breech, tapering down to twenty-seven at the muzzle. This was accomplished by two layers of iron, placed in such a manner as to resemble hoops laid at right angles to the axis of the mass, and after two more heatings and careful weldings, the forging of the work was complete. After each important addition, a securing heat was given to prevent flaws.

A great deal has been written at various times on the crystal

lization of wrought-iron under the action of heat long continued, more especially when the metal is allowed to cool slowly, and also when the metal is subjected to the action of blows frequently repeated; some experiments have, however, tended to disprove the theory of crystallization by heat; but we have seen bars of iron, originally of a tough or fibrous character, snap with a force far below the calculated resistance of the material, the fracture exhibiting a beautiful crystalline texture.

The forms of iron ordinarily obtainable in commerce are as follows: square, round, elliptical, rectangular, semicircular, segmental, channel, T, H, and L iron bars, also plate and sheet iron; the thinnest being that employed for the manufacture of tin plates.

It is unnecessary here to dilate upon the forging of copper, it being only necessary to observe that it is worked at a low temperature, and that when it is wrought cold it is necessary occasionally to anneal it by heating.

Besides the tools already mentioned, various tongs, and also various special forms, are frequently required to execute hollow or other work, which cannot conveniently be wrought upon the anvil; these are called stakes. Among the swages occur some having at their extremities a conical or cup-shaped recess, intended to complete the heads of rivets; these are termed snaps, and are sometimes worked by machines termed rivetting, machines, being attached to piston rods, acted on by pistons working in cylinders of large diameter but with a very short stroke.

CHAPTER III.

ON MOULDING AND CASTING.

FOR many purposes it is found convenient to produce articles from molten metal, by a process termed casting, which consists in pouring the metal in a fluid state into a cavity which corresponds to the form of the article to be produced. Several methods of producing these cavities are in use, but we shall here confine our attention to the manipulations included under the head of green or baked sand-mouldings, loam-moulding, and moulding for chilled castings. We will first speak of green sand-mouldings.

The first operation to be performed, when it is proposed to make a green sand-casting, consists in making a model or pattern of the article to be produced. This may be made of wood; it must be in form similar to the required object, but tapered so that it may admit of being readily removed from the sand in which the casting is to be made; and it must also be larger than the finished article, in order to allow for contraction in cooling, and also for the removal of material in producing finished surfaces. The contraction is, for iron, about a tenth of an inch to the foot, and for brass one-eighth of an inch may be allowed. All apertures in the intended casting are produced by pieces called cores, fixed in the mould; these are retained in position by being made longer than the apertures to be produced, the excess of length being inserted into recesses formed in the sides of the mould. These recesses or hollows are produced by protrusions upon the pattern or model, such protrusions being called core prints.

Cores are also used, under some circumstances, for the produc tion of undercut recesses.

We may perhaps best illustrate the manner in which the casting is produced, by taking an example, and describing the process required for the completion of such example. Let us suppose that the poppet-head of a lathe, which is of the form shown, Fig. 9,

having an aperture running through the whole length of the upper cylindrical part, is required. The pattern will be of the form shown at Fig. 10, being furnished with core prints, as shown at Fig. 10.

Fig. 9.

each end of the cylindrical part. The process of moulding is conducted in the following manner. Two boxes, having neither top nor bottom, but capable of being fitted together by means of pegs fixed in lugs on one frame, which fit into apertures in lugs on the other frame, together called a flask, are used to contain the sand of which the mould is to be made. One flask is taken and placed with the lugs downwards upon a smooth slab, and filled with moulding sand, which is firmly rammed down. The flask may then be inverted, the sand being retained in the frame by its cohesion and adhesion to the sides of the flask; but when the latter is large, it is, for greater security, furnished with transverse bars. After the frame has been inverted, the upper surface presents a smooth and level appearance, and in the centre of this a hollow is scooped resembling the form of the article to be cast. In this the pattern is bedded in a horizontal position, being sunk in the sand to as nearly as possible half its thickness; powdered charcoal or coal-dust is now sprinkled over the whole surface, and the upper part of the flask adjusted in position; it is then filled up with sand, which is firmly rammed down around the pattern. The two parts of the flask can now be separated, the adhesion of the sand being procured by the layer of charcoal dust mentioned above. The impression formed in the sand of the upper flask is smoothed and repaired, where necessary, by trowels of a suitable form. The sand placed in the first frame is now broken up, the frame which served as the top of the flask placed with the cavity uppermost, the pattern placed in the cavity, the empty frame fitted on, and the whole filled up as before. The flask is then again taken in pieces both cavities repaired where necessary, and openings made from

the cavity in the upper frame, through to the surface of the sand. These are called gates, or gits, and serve, some of them, for the admission of the molten metal, and others afford egress to the air in the cavity and the gases generated by contact of the hot metal with the sand. The core, previously made, of tough loam with chopped straw or other filamentary material, and dried, is now inserted in the core prints; the flask is then put together, the two parts being secured by pins or wedges passed through apertures in the extremities of the pegs fixed to the lugs of the lower frame. The metal may then be cast. When the casting is sufficiently cool the mould is broken up, the superfluous metal knocked of; and when the casting is quite cool, the false seams are cut off, the core cleaned out, and the hard sandy coating rubbed smooth with a piece of oven-coke.

We may now mention a few particulars to be observed in the general preparation of moulds. Ample space for egress of gases must be allowed, wherefore it is desirable to pierce the sand to within a small distance of the cavity, by means of a stiff wire; also to form a sufficient number of gits. The sand should be of open texture, but of a binding character, otherwise the casting will be apt to scab,-that is to say, there will be a liability in the sand to scale off the surface of the mould, and rest on the surface of the casting. If sufficient egress be not allowed for the air, blow-holes will occur within a short distance of the surface of the casting, thereby materially reducing its strength. It is usual to tap the pattern with a hammer in order to loosen it, previous to withdrawing it from the mould, thus preventing the risk of damage to the mould, and for this purpose wires are sometimes screwed into the pattern, which protrude through the surface of the sand and the upper part of the flask. Very heavy patterns may be removed from the sand by the united efforts of several men, each lifting the pattern with one hand while with the other he taps it with a light hand-hammer.

If the sand be used too damp, hard places will be formed in the casting, thereby adding materially to the difficulty of subsequently working the metal.

Moulding in baked sand is conducted in a manner similar to the above; but the sand is used in a more moist condition, the mould' being subsequently dried in a suitable furnace. Moulding in loam