placed in the fire, or both are heated separately. The steel must be brought up to the proper temperature as rapidly as possible and excluded from the air; it is best done with charcoal and good coke, since coals, on account of the fact that they contain sulphur, produce a thin layer of sulphide or sulphate of iron, which prevents proper welding.

As a welding mixture, Th. Rust recommends. 41.5 parts of boracic acid, 3.5 common salt, 15.5 prussiate of potassa, and 8 calcined soda ash.

Habich prescribes 7 parts of anhydrous prussiate of potassa, 2 calcined soda ash, and more or less burned borax, according to the nature of the steel. Ermer recommends to dissolve in water 8 parts of borax, 1 salammoniac, 1 yellow prussiate of potassa, and to evaporate the solution at a low heat to dryness. When strongly heated, violent explosions may occur by the formation of chloride of nitrogen. Another method is as follows: Borax is fused with 1-10th of its weight of sal-ammoniac, and to the vitreous mass the same quantity of burned lime is added. Still another employs 8 parts of heavy spar, 1 part of gall of glass, and 1 of black oxide of manganese.

In welding, at first, light, then heavy blows are given, so that the slag may escape from the joints, whereupon the outer surfaces are united.

Since hard steel is tempered (after hardening) sooner than soft, and the latter sooner than iron, the various kinds of steel do not always exhibit the same degree of hardness, although they may show the same tempering colors. There appear small differences, inasmuch as a brand cooled at a bright yellow heat may become as hard as one cooled when of a straw yellow color; or another one may get as hard when violet as one that has been dark blue. In some cases, especially when a particular hardness is required, as is desirable for the edges of astronomical and philosophical instruments, and when the steel is rich in carbon, it may be proper to conduct the tempering at such a low temperature that no colors appear at all. And in order that the operator should not be subject to delusion in observing the change referred to, the steel should have a shining, and sometimes polished, surface, and be uniformly heated.

Since the colors owe their appearance to the formation of an exceedingly thin superficial skin of oxide, it is evi dent that the steel, when withdrawn from the fire, does not retain its first color, but there appear other colors in consequence of a subsequent oxidation by the air, until the steel is sufficiently cool. Of a certain color, one can only judge with certainty by examining the conditions under which it occurs. If two pieces of the same steel are heated until the yellow color appears, and if one is

withdrawn, it may become in the air purple, violet, and finally blue, while the other piece assumes the same colors in the fire. However, if both pieces when blue are dipped into water, they acquire different degrees of hardness, that is, the one which turned blue in the air will be harder than the one left in the fire. Hence it follows that proper caution must be observed in this respect, and steel must either be cooled rapidly, when the right color appears in the fire, or it must be withdrawn at a preceding color, if the desired shade is to appear after tempering.

In tempering scythes and similar tools, they are stuck in a layer of hot sand or hammer slag, spread on a heated plate, and sometimes only the hot sand is spread over them. For sword blades, for instance, the thicker parts are heated by a red-hot piece of cast-iron having the proper shape; and if the edges are to be harder than the other parts, they may be rubbed with a potato or beet.

Parkes has proposed the following alloys for tempering baths. They are suitable in some particular cases, and their temperature should be maintained near the melting point, without over heating:

The fracture of hammered or tilted steel is often oblique, angular and rugged, and the broken surface presents a quantity of small and sharp points. On the other hand, the fracture of rolled steel is more even and the grains are rather rounded in shape.

Of two kinds of cast steel possessing the same hardness and the same fineness of grain, the purer is the more malleable, and the difference is the more appreciable as the percentage of carbon is greater.

Steel articles which have warped during annealing, had better be slightly heated for the straightening process which precedes hardening. The proper temperature is that which allows of handling the articles with a thick leather glove.

Steel should be brought up rapidly to the desired temperature, because a slow and protracted heat changes its molecular structure, and diminishes its tenacity and malleability.

Screw and key files, cut at the edges only, and other thin and flat articles, should be filed or ground lengthwise before hardening, in order to diminish breakage. The furrows produced by cross filing or grinding cause many breaks during the hardening process.

When, for nearly finished articles, somewhat out of shape, the iron hammer cannot be employed, they are straightened upon a wooden block with wooden mallets. In this case, the steel must be heated until it acquires a blue, violet, or pigeon-throat color, otherwise, by the hardening process, it will resume its previous deformed shape.

A thin paste with water, of 75 parts of fine wood ashes and 25 of fat clay without sand, and applied not too thickly with a brush upon steel, is a good protection against the action of the fire, and does not change the nature of the metal.

Scythes are hardened in hot, and sometimes boiling, baths of tallow mixed with a small proportion of rosin. When steel is hardened by dipping it into mercury, its grain becomes finer than when any other cooling compound is employed.

Steel does not require tempering when, as by watchmakers, it is hardened by pressing it into a block of cold lead.

Steel blades which become curved by hardening, are straightened cold with hammers, the striking surface of