STEEL.— Cementation Process. 853

iron, and that they are due to the formation of com-paratively largo volumes of carbonic oxide gas, bythe action of the carbon (taken up by the iron) uponthe particles of slag contained in the puddled bar.The slag consists largely of oxide of iron, which be-comes reduced thus, C + FeO = CO + Fe , and thecarbonic oxide being given off more rapidly than itcan be taken up by the iron, forces the soft coveringinto a blister. The theory of the process of cement-ation is not yet perfectly understood; but it is highlyprobable that the charcoal when heated producescarbonic oxide, and it may be cyanogen gas andhydrocarbons; and the iron when heated first oc-cludes some of these gases and then decomposesthem, in consequence of the affinity of the iron forcarbon. Percy has clearly shown—1st, that car-bonic oxide alone is perfectly capable of convertingiron into steel; and 2nd, if pieces of iron be heatedin charcoal, under conditions in which it is difficultfor carbonic oxide to be generated, that the irondoes not increase in weight, and is not convertedinto steel. Hydrocarbons are capable of producingcarburization; but the many attempts that havebeen made to use them have not resulted in prac-tical success.

The charcoal which has been used for one opera-tion is found to be pulverized in a curious way, andafter the process the dust is sifted out, and about anequal portion of new charcoal is added to the oldbefore it is used again. Hard charcoal, rich inalkalies, is found to yield the best results, and thisfavours the view that cyanogen may do a good dealtowards hastening the operation. Hence Tunnerrecommended the addition of a half per cent, potashand common salt to the charcoal. It is a curious factthat Blair has noticed that, in the reduction of ironore by heating it in contact with charcoal in a re-ducing chamber, the presence of a small quantity oflime facilitates the elimination of oxygen from theore, and Dodd shortened the time of cementing byusing (i per cent, lime and 2 per cent. soda. Cokepowder does not answer in place of charcoal, pro-bably on account of the sulphur which it contains;as there is little doubt that the tendency of sulphuris to prevent iron taking up carbon, and to throwit out of combination when it has been combinedwith iron.

The following complete analyses by Messrs. Pat-tinson and Stead will show the changes whichtake place in iron which has been subjected to theprocess of cementation.

SWEDISH BAR-IRON BEFORE AND AFTER CEMENTATION.

©

©

. Before.

After,

Before,

After.

99 298•470•120•037•035•032•008

98-571

1*200

•100

•006

■027

•030

•006

99-660

•220

•044

•052

•016

•008

trace

98-699

1-210

•044

•028

•013

•006

trace

Carbon,....Manganese,.Silicon, ....Sulphur,. ...Phosphorus,Copper,.

100-000

100 000

100-000

1C 0-000

In commenting on these analyses Messrs. Pattin-SON and Stead remark, that “the small differencesshown in silicon and manganese, in the bars beforeand after cementation, are due no doubt to slightdifferences in different parts of the bars, and not toany change having taken place.” This is doubtlesstrue, for it is well known that in puddled iron theslag is interposed very irregularly throughout themass. On examining the above analyses, it will beseen that the only change has been the absorptionof '73 per cent, more carbon in the case of the sam-ple of © iron, and '99 per cent, more in that ofthe © sample. Both samples contain a rather largeper cent, of carbon before converting, but particu-larly the first, which must be regarded as puddledsteel. The © iron is remarkably free from impuri-ties. Much of this blister steel is used withoutfurther remelting, the bars being simply nicked andbroken into short pieces, piled, heated, and forgedinto shape, when it is termed “ shear steel.” Fre-quently these bars are cut up again, repiled, heated,and forged again, when the product is known as“ double shear steel.” Blister steel is more readilywelded than cast steel; but the welds are rarelyperfect, and cutlery made from such steel can gene-rally be detected by a practised eye by the marksof the welds. If a knife-blade, for instance, be ex-amined carefully there will often be seen faint darklines running parallel to the edge, owing to thewedge-shape of the blade revealing the layers ofwhich the piece is made up.

Cutlery made from cast steel does not show theseparallel striae. Double shear steal is more liable tothese marks than single shear, because in the firstprocess it is customary to cover the pile with a claywash to prevent oxidation while heating, and thisadds to the difficulty of obtaining a perfect weld atthe second operation. The heating of the piles isusually performed in hollow coke fires, and the weld-ing under small tilt hammers, driven at great speedby water power. The kind of fuel used for heatingthe converting pots is one of considerable import-ance (it must be free burning, and leave no clinker),as is also the quality of the material from whichthe pots are made; and it was doubtless owing tothe proximity of good water power, highly refrac-tory materials, and first-class and cheap fuel, thatSheffield rose to be the Queen city of the steel manu-facture. It seems curious that only Swedish ironswere held to be capable of producing good cementsteel, and that even of these one or two brands wereconsidered infinitely superior to all others. Butwhen we examine the analysis of the celebrated© brand given above, we can see that few English brands, before hematite iron became so largelyintroduced, could be expected to yield so pure aproduct. There does not appear to be any reasonnow why Swedish iron should not be quite equalledby iron made from our own hematite pig. Thefollowing analysis shows clearly, that even in thepig iron the impurities are little more than in theSwedish bar, and with good puddling there is noreason why these should not be decreased to a very