Instead there is an emission of a comparatively small number of dull particles, and there is clinging to the wheel closely a heavy band of a deep, rich red color. This red streak is distinctive of the presence of tungsten.

By testing various pieces it was soon observed that different quantities of tungsten gave different sizes of red streaks; as tungsten decreased the width of the band diminished and the number and brilliancy of carbon sparks increased. As little as .10 tungsten will show a fine red line amidst a brilliant display of sparks, and it soon became possible to determine so closely by the streak the quantity of tungsten present that the ordinary analyses for tungsten became unnecessary, except in occasional important cases where analysis was used merely to confirm the testimony of the wheel.

Self-hardening steel, then, is a steel which, owing to the presence of manganese and tungsten, hardens when quenched in quiet air, and which retains its hardness almost up to a red heat.

It may be forged between the temperatures from orange to bright orange; it cannot be worked safely outside of this range. The more quickly it is quenched the harder it will be; and it may be annealed so that it can be machined readily. Therefore it is not self-hardening; it simply has all of the properties of carbon steel modified profoundly by tungsten and manganese. If a piece of this steel will not harden sufficiently by cooling in the air quietly, that difficulty may be remedied by cooling it in an air-blast; if quenching in an air-blast will not give sufficient hardness, the steel had better be rejected, for quenching in oil or water means almost certain destruction.

As stated before, the range of temperature in which self-hardened steel can be forged safely is much smaller than for a high-carbon steel; it is harder at this heat than carbon steel and not so plastic, so that it requires more care and more heats in working it to tool-shapes.

This steel is so sensitive that it often occurs in redressing it that it will crumble at a heat that was all right in the first working. This difficulty may be remedied by first cutting off the shattered part with a sharp tool,—it must be cut hot,—then heating the piece up to nearly a lemon color, heating it through without soaking it in the fire, and then allowing it to cool slowly in a warm, dry place. After this treatment the steel may be heated and worked as at first. This treatment does not anneal the steel soft, because the heat is not continued long enough, and the cooling is not sufficiently slow; it does relieve the strains in the steel, so that it is plastic and malleable.

This treatment is good in any high steel which has become refractory from previous working.

Self-hardened steel is not as strong in the hardened condition as good high-carbon steel; it has not been used successfully for cutting chilled cast iron, for instance. If made hard enough to cut a chill, it is so brittle that the cutting-edge will crumble instead of cutting; if the temper be let down enough to stop the crumbling, the steel will be softer than the chill, and the edge will curl up instead of cutting.

Owing to the retention of hardness at a higher temperature than carbon steel will bear this steel is capable of doing a great amount of work at high speed, so that for much lathe-work it is cheap at almost any price.

Owing to its brittle, friable nature its use is limited to the simpler forms of tools, and to a narrower range of work than is possible with carbon steel.