STEEL.— Tempering. 877

metal so made as in the scrap process to producesteel; but so far this, also, has not yet been com-mercially successful—possibly because at Towcester,where the process was carried on, there was nomelting furnace to make the directly-reduced ironinto steel at once, and as a process of iron makingit did not pay.

Physical Properties of Combinations of Ironwith other Elements.' —There is little doubt ttiat alarge number of elements may and do unite withiron, but only the action of the following need beconsidered here: — viz., carbon, silicon, sulphur,phosphorus, manganese, oxygen, copper, tungsten,and chromium. It is rarely that the opportunity isafforded for studying the action of these substancesupon iron singly. An attempt to do so was madesome years ago by the British Association; but noresults of any importance were obtained. We have,therefore, to make deductions from a study of thequalities of various combinations as they arise.

Carbon is generally considered to be the essential■starf-forming element. If iron be nearly free fromother elements, the presence of T per cent, carbon,or even T5 per cent., produces scarcely any harden-ing effect upon the alloy, as such a compound maybe plunged while red hot into cold water and remainperfectly soft and ductile. As the proportion ofcarbon increases, the hardening effect of such treat-ment becomes more and more evident, until, withanything approaching 1 per cent, of carbon, thesteel becomes very hard and brittle by sudden cooling.Even when containing about 2 per cent, carbon itdoes not quite cease to be malleable ; but it requiresgreat care in the heating to prevent the carbonbeing burnt out, and can only be forged at amoderate red heat. If small quantities Of silicon,phosphorus, or manganese be present, it is notpossible to work steel containing even this smallquantity of carbon. Manganese appears to favourthe solution of carbon in pig iron, since an alloycontaining 8-55 manganese contained also 6‘62 ofcarbon, and the carbon appears to rise with themanganese.

The action of carbon upon iron is very remark-able ; and though the effects are tolerably known,the cause of its action is even yet shrouded inmystery. Thus it is well known that when any-thing like 1 per cent, of carbon is present, the alloysuffers remarkable changes from various modes oftreatment. If it be heated and plunged into a fluidthat abstracts heat rapidly, it becomes hardened inproportion to the rapidity of this action, and thusunder the same conditions mercury causes greaterhardness than water, and water than oil. When apiece of steel has been hardened and is again heated,the hardness disappears as the temperature is raised.Thus if a certain degree of hardness is required, itmay either be obtained by first' pushing it to theextreme point, as by raising the steel to a dull redheat and plunging in water or mercury, and after-wards gradually raising the temperature of the metalto a certain point and leaving it to cool of itself ; orelse the piece may be raised to a definite tempera-

ture, and then suddenly cooled in a vehicle that onlyabstracts the heat slowly, as oil. According to theexperiments at Terre-Noire, it appears that temper-ing steel (whether containing much or little carbonor manganese) in oil about doubles the tensilestrain, while the elongation is reduced in thesame or even greater degree. The process ofproducing a certain degree of hardness is termed“tempering,” and is quite an art. In raising thesteel to the required temperature the workman isaided by the colour produced upon the surface ofbright steel by slight oxidation, as the thin filmof oxide passes through a regular series of colours,from light straw to brown, then blue, and finallyblack, through every gradation in shade. Thus itis that the “ temper,” as the degree of hardness iscalled, is determined by this colour. The followingtable (XII.) shows the colours produced in temper-ing carbon steel for particular uses:—

TABLE XII.

Temp. F.

Colour.

Temper.

430

Very faint yellow.

Lancets.

450

Pale straw.

Razors & surgical instruments.

470

Full yellow.

Penknives.

490

Brown.

Scissors and cold chisels.

5101

Brown, with, purplespots.

Axes, plane irons, pocketknives.

530

Purple..

Table knives, large shears.

550

Bright blue.

Swords, watch springs.

560

Full blue.

Fine saws, augurs.

600

Dark blue. . .

Hand and pit saws.

The process of tempering should leave the steel ofthe hardness required, with the minimum of brittle-ness. Tempering in oil appears to produce thisresult best.

No perfectly satisfactory explanation of the theoryof hardening and tempering has yet been adduced.One set of metallurgists suppose that steel containsa certain quantity of graphite, or uncombined car-bon, and that when the steel is heated this entersinto combination with the iron, and that suddencooling keeps it combined.

Probably a more satisfactory explanation is thatwhen the steel is heated it is expanded and its par-ticles put into a state of tension, which is retainedby instantaneous cooling. Chemical analysis hasnot satisfactorily proved that tempered steel con-tains more combined carbon than untempered steel,while it is well known that tempered steel has a lessspecific gravity than untempered steel, and that theprocess of hardening increases the size of the articleslightly. Thus the specific gravity of cast steel at11° C. was—Unhardened, 7-9288; hardened, 7-6578;and a bar of steel which before hardening had avolume of 20 c.c., became 20-351 after hardening.The expansion by hardening can be easily proved byrivetting together a piece of soft iron and a piece ofsteel, carefully filing the edges true. On heating andhardening this compound plate, the edges of thesteel plate will be found to overlap those of the iron,while the rivets will probably have been sheared.Hardened steel is much less readily acted upon byacids than the same steel in a soft state, and the