The diminution of the tensile strength by a too strong hardening of a highly carbonaceous steel is, however,' much more rapid than the corresponding decrease in consequence of the content of carbon being too large in the unhardened steel, but this difference is easily explicable by the great tension which a strong hardening must produce in the highly carbonaceous steel. When a hard steel is too strongly hardened, the larger pieces in particular readily break in pieces of themselves, which again is a natural consequence of the fact already mentioned, that the limit of elasticity in this case nearly coincides with the ultimate tensile strength, and therefore when the resistance of the inner layers against the contraction of the outer becomes so strong that the limit of elasticity is exceeded, fracture of the hard steel readily takes place, instead of the extension which would have occurred in the outer layers of a less hard steel. The same correspondence between the influence of hardening and an increased content of carbon also prevails in respect of ductility, hardness, and fineness of grain.

The first of these properties diminishes and the two others increase with the content of carbon, and corresponding changes are produced by hardening.

Another proof that the effects of hardening depend on the oft - mentioned compression is afforded by the behaviour of burnt iron in hardening. Burnt iron, as is well known, is the name given to an iron which, through too long - continued or strong heating, has had the opportunity of assuming a crystalline texture, with the brittleness which accompanies it on account of diminished cohesion of the crystals. The disposition to such a crystalline segregation is less in proportion as the iron is both more mixed with cinder and freer from certain substances; but the more carbon, and in particular the more phosphorus, it contains, the greater is the liability of the iron to be burned, and the more care ought therefore to be taken with the heating, if it is not in consequence of this distribution into crystals to fall in pieces utterly, or at least crack, as soon as the drawing begins. An iron as good as free from these substances can, without danger of burning, be heated to the strongest welding heat, but with an increase of carbon in the iron all the more care must be observed in the heating, and this is rendered necessary in a much higher degree by an increased content of phosphorus in the iron.

So long as the content of carbon in the iron is quite small, however, this detrimental influence of phosphorus is still rather limited, more particularly if the iron contains at the same time a good deal of manganese; but the greater the content of carbon in the iron, the more is the detrimental influence of phosphorus increased, and the first requisite of a really good steel is, therefore, that it contains as good as no phosphorus. An iron whose disposition to burn is so great that this detrimental change cannot in general be avoided, has from old times been called "cold - short," from the brittleness caused by its crystalline texture. But in this connection it must be kept in mind that the disposition to burn, increasing with the content of phosphorus, is not only counteracted by the presence of manganese and the absence of carbon, but also that the iron molecules in the puddled iron are intercalated with layers of a fine interspersed cinder, which is unfavourable to the formation of the coarse crystals which cause the brittleness.

A certain content of phosphorus is therefore not so detrimental to the puddled iron, loose in its texture and mixed with cinder, as to the cinder - free ingot - iron, and if by the adoption of a suitable treatment the production of even the least sign of crystalline texture be prevented, a content of phosphorus rising to 1/3 Per cent. does no great harm to the iron, so long as it is in the non - crystalline condition just mentioned; but the difficulty is to avoid the crystalline texture in the phosphoriferous iron, and, if success is attained in this, to prevent the formation of crystals in the case of a possible future re - working of the iron in a warm state. Now, whether an iron which, from one cause or another, has a tendency to burn, becomes, after a certain heating, burned or not, depends mainly on the degree to which it is afterwards drawn out; fur the more an iron, which, when heated, has begun to be crystalline, is afterwards drawn out in a warm state, the less is the danger that the crystalline texture will remain in the fully - drawn iron. In this way it is explained why a greater degree of drawing out, and thus also larger ingots, are requisite for a more than for a less phosphoriferous iron.

If, however, an iron after a certain heating, followed by drawing, still appears to be coarsely crystalline or burned, this burning can frequently be removed by heating the iron anew to a certain welding heat, properly adjusted to its content of carbon and phosphorus, succeeded by a new drawing out; but we must not make too sure that we can in this way always remove the burning or cold - shortness. In complete correspondence with this, experience has shown that burning can be removed by a corresponding heating, followed by hardening instead of by drawing; and this circumstance affords a new proof of the correctness of the view, that the effects of hardening must depend on the compression caused by the. contraction, as it, like the drawing out, can remove the crystalline texture. In close connection with this, doubtless, also stands the circumstance that the hardening of an iron which contains much phosphorus but little carbon can even increase its ductility, for the somewhat crystalline texture of a phosphoriferous iron may be destroyed by hardening, whereby again its ductility is greatly increased.

"To the older observations noticed above has lately been added a new experience, which further confirms the correctness of the view that the effects of hardening depend on the compression caused by the contraction. The experience now referred to is that ingot - metal free from blow - holes can, without drawing, and merely by hardening followed by a new heating to redness, become quite equal to ingot - metal that has been drawn out. It is, at least if the content of manganese be not all the greater, only when the percentage of carbon exceeds 0.3 that the ductility suffers through hardening any loss endangering the strength of the material, and it therefore appears probable that in all the cases where special ductility is not demanded, it may not be necessary to reheat to redness castings poor in carbon after they have been hardened. The higher the temperature at which the reheating takes place, provided, however, it does not exceed a full red heat, the more is the ductility increased; while, on the other hand, the limit of elasticity and the ultimate tensile strength are thereby diminished until they descend to nearly the same minima which characterise ingot - iron or ingot - steel of the same composition which has been heated to redness after hardening.

By modifying this reheating, on the other hand, it is possible to bring the hardened castings to intermediate stages of these qualities, quite as is the case with the hardened, drawn iron or steel. The consequence of this must be that a drawn iron or steel - above all, when the drawing, as in the case of hammering, has been continued to a low temperature - loses in limit of elasticity and ultimate tensile strength by heating to redness followed by slow cooling.; while an undrawn ingot may increase to some extent in these properties, although in a smaller proportion, according as the degree of heating was less and the cooling slower. On the other hand, by sufficiently rapid cooling or actual hardening, the oft - mentioned properties are increased in the material to a higher degree than can be attained merely by drawing; but by renewed beating this excess can be again removed."