Page:Encyclopædia Britannica, Ninth Edition, v. 13.djvu/356

 340 IRON 1561, p. 341). Early in the 18th century Reaumur inves tigated the character of the process, and found that under similar conditions a bar of iron of O2 inch in thickness was carbonized in six hours to the same extent as another bar of the same metal of CHS inch in thickness in about thirty-six hours. The crude &quot; blister steel &quot; produced by the cementation process (so termed from its blistered surface) is often simply cut into pieces, piled, heated to a welding heat, and forged, when it is converted into &quot; shear steel &quot;; or this process is repeated, when it becomes &quot; double shear steel &quot;; but when a perfectly homogeneous product is required it is melted in crucibles, when it becomes &quot; cast steel &quot;: this process was introduced by Huntsmann about 1740. The nature of the chemical changes taking place during cementation have been often regarded as somewhat uncertain ; but there seems to be little room for doubt that the action in the ordinary cementation process is mainly due to the occlusion of carbon oxide (formed by the action of the air in the pores of the charcoal) in the iron, and its decomposition by the metal into carbon and an iron oxide, which is subsequently again reduced by a second portion of carbon oxide, thus Fc x O y + yCO = Fe* + 2/C0 2, the two changes going on simultaneously. The escaping carbon dioxide, which penetrates through the metal less readily than does carbon oxide, and hence is apt to accumu late in certain parts, is probably the cause of the blistering of the surface of the steel often observed, especially with pud dled bars containing small quantities of ferrous silicate dis seminated through them ; Percy has shown that fused homogeneous metal free from interspersed slag does not give rise to blisters on cementation. Certain hydrocarbons, e.g., paraffin vapour and coal gas, will carbonize iron heated therein, and the manufacture of steel by cementation in the latter has been patented by Macintosh (vide infra). Probably in these cases the carbon comes from the direct splitting up of the hydrocarbon, with elimination of hydrogen ; but pos sibly the acieration is due to carbon oxide present in the coal gas or formed from the paraffin vapour, &c., by the action of iron oxide disseminated through the bars or adherent to their surface. Many cyanogen compounds, especially ferro- cyanide of potassium, when applied to iron in a heated state convert it exteriorly into steel (case hardening), and it has in consequence been supposed that nitrogenous substances are essential to the carbonization of iron by cementation, and that nitrogen is an essential constituent of steel. The evidence in behalf of this is, however, at present unsatisfactory*; on the other hand, charcoal rich in alkalies, or a mixture of char coal powder with a little lime and soda, will carbonize iron submitted to cementation therein more rapidly than charcoal more free from alkalies ; and, as these conditions are those favourable to the formation of alkaline cyanide from the nitrogen of the air, there is some reason for supposing that the carbon in the steel formed under such circumstances (like that produced in case hardening by means of ferro- cyanide) is more or less derived either from cyanogen separated from the cyanide and occluded by the iron and gradually decomposed with formation of carbon, or from some other reaction of iron upon the cyanide. Accordingly nitrogenous organic matter, such as animal charcoal, leather, horn, &c., is often mixed with the charcoal used for cementation with a view to facilitating the conversion into steel by the formation of gaseous carbon compounds with the simultaneous presence of nitrogenous vapours. The theory that carbon oxide is the source of the carbon com municated to wrought iron during cementation, appears to have been first propounded by Leplay in 1846 (Ann. deChim. etPhys, [3] xvii. 221), at a time when the properties of metals and other bodies in absorbing gases ( i. e., the phenomena of occlusion] had not been so well studied as they have been subsequently. Leplay appears to have considered that the carbon oxide splits up directly into carbon and carbon dioxide, the latter becoming again transformed into carbon oxide by the surrounding charcoal, and to have left out of consider ation the intervention of the iron in becoming alternately oxidized and reduced. Other chemists have considered that by direct con tact with carbon combination of the iron therewith takes place, the carbon thus taken up by the outer layer quitting that and combin ing with the next layer, and so gradually travelling inwards, the outer layer recombining with more carbon as fast as it parts with carbon to the under layer, and so on throughout ; the carbon thus traversing the iron by a process somewhat akin to that by which a drop of mercury in contact with a piece of gold (or certain other metals) gradually passes into and permeates the mass, this being in short a kind of capillary action exerted upon a solid substance. Percy s observation (Metallurgy, &quot;Iron and Steel,&quot; p. 109) that char coal after being intensely ignited will not carbonize iron when air is excluded by means of hydrogen (although it will do so to some extent if still containing matters capable of being driven off by heat) negatives the possibility of the carbon being taken up by direct con tact by this hypothetical kind of chemical union between solids, or solvent action of one solid on another ; it may be that carbon deposited on the outer layer by the chemical action of the iron on carbon oxide, cyanogen compounds, carburetted hydrogen, &c. , permeates inwards by this supposed diffusive process; but the known phenomena of the absorption of gases by colloid bodies, diffusion, dialysis, occlusion, &c. , as elucidated by Graham and his followers, render it wholly unnecessary to suppose that any such action takes place, and do away with all experimental grounds for supposing that it can take place. In order to cany out the process of cementation, the bars of iron are placed in a firebrick box or chest several feet long, layers of charcoal and iron being alter nately piled in until the box is filled, when a luting of fireclay or of the sandy ferruginous mud produced in grinding and polishing steel articles after manufacture, termed &quot;wheel swarf,&quot; is applied so as to close up the upper part of the box and prevent access of air ; two or more such chests are then arranged under the arched roof of a chamber erected over a fireplace in such a way that the flames from the fire pass under and lap round the sides of the chests, and impinge upon the roof, the gases escaping through orifices in the roof into a conical chimney built over the whole, the chamber constituting in fact a kind of furnace somewhat like a glass house or pottery kiln, the flame passing upwards from the bed instead of laterally from a fireplace at the side as in the ordinary reverberatory furnaces. Trial bars are arranged in the mass of charcoal in such positions that they can be withdrawn from time to time, and the progress of the operation examined by fracturing the bars after cooling, and seeing when the core of malleable iron disappears; from seven to ten days heating according to the amount of carbonization required (averaging about 1 per cent. ) is generally allowed, with a total charge of some 10 to 20 tons of iron in the furnace. When the requisite carbonization is attained the fire is raked out and the chests allowed to cool ; the blister steel is then either melted down into cast steel, or converted into shear steel by piling and forging, &c. According to Boussingault a material diminution in the amount of sulphur present takes place during cementation ; thus he found malle able iron specimens containing 6 - 012 to 015 per cent, of sulphur yielded steels containing only 005 to O OOG per cent, of sulphur. Indications in the same direction but not to so great an extent have also been observed by others ; no noticeable effect, however, is pro duced on the silicon, phosphorus, or manganese originally present, as far as the irregular way in which traces of cinder are always inter spersed throughoiit bars of wrought iron will permit conclusions to be drawn. The following analyses indicate the effect of cementation on Swedish bar irons : Analyst Pattinson and Stead. 11. S. Bell. Hoop G L liar. Steel. Hoop L Bar. Steel. Danne- mora Bar. Steel. 99-298 0-470 0-120 0-037 0-035 0-032 0-008 98-571 1-200 o-ioo 0-OCG 0-027 0-030 0006 99-000 0-220 0-044 0-052 0-Olfl 0-008 trace 98-699 1-210 0-044 0-028 0-013 0-006 trace 99-471 0-352 0-075 0-050 0-027 0-025 98-603 1-250 0-072 0-035 0-022 0-018 Copper 100-000 100-000 100-000 100-000 100-000 100-000 In consequence of the phosphorus originally present remaining unchanged, only the purest brands of iron as free as possible from these ingredients are converted into cementation steel, often known as &quot;tool steel,&quot; commanding a high price in consequence of its physical properties, the most valuable of which are enormously deteriorated by minute quantities of sulphur and phosphorus. The process of cementation in an atmosphere of coal gas as patented by Macintosh of Glasgow consists of exposure of the bars of iron hang ing vertically in a cylindrical chamber, the walls of which are kept