H 



NATURE 



[Nov. 7, 1889 



The carbon might have been presented to the iron in the 

 form of a gas capable of yielding carbon, and this element 

 would as surely have found its way into the steel. 



Margueritte,^ for instance, in 1865, repeated Clouet's 

 experiment, and showed that, although carburization can 

 be effected by simple contact of iron and carbon, it is 

 nevertheless true that in the ordinary process of cementa- 

 tion the gas carbonic oxide plays an important part, 

 which had until then been overlooked. The discovery by 

 Graham,'* in 1866, of the occlusion of carbonic oxide by 



Fig. I.— Plan of iron rrucible and diamcnd from the drawing in Guyton 

 de Mcrveau's paper. In the original, the d.amcnd and the crucible are 

 drawn, in plan, separately. 



iron, gave additional support to this theory. I am glad 

 to remember that he intrusted the experiments to me. 



The question, however, of the direct carburization of 

 iron by the diamond has never been doubted since 1815, 

 when a working cutler, Mr. Pepys,^ heated iron wire 

 and diamond dust together and obtained steel, the heat 

 being afforded by a powerful electric battery. I am 

 anxious to make this absorption of carbon in the diamond 

 form clear by this diagram (Fig. 2). 



Fig. 2 represents a glass vessel which may either be rendered vacuous or 

 may be filled with an atmosphere of gas through the tube d. An iron wire, 

 b, placed between the terminals of a battery, c, c', is heated to redness, and 

 remains glowing until it is touched by pure diamond dust, which is 

 effected by raising the cup a. The iron combines with the diamond dust 

 and fuses. 



Do not think for a moment that the steel owes its 

 hardness to the passage of diamond into the iron, as 

 diamond. I have repeated Margueritte's form of Clouet's 

 experiment, using, however, a vacuum instead of an atmo- 



' "Sur I'acieration," Ann. Chitn. et Pkys., t. vi. [4], 1865. 



2 Phil. Tians. Roy. Soc, 1866, pp. 399-439. 



3 Ibid., 1815, p. 371. 



sphere of gas, and employing the form of apparatus showr> 

 in this diagram (Fig. 3). [The carburized iron which M'as 

 the result of the experiment was thrown upon the screen.] 

 The diamond by union with iron has passed partially at 

 least to the other form of carbon, graphite, while treatment 

 with a solvent which removes the iron shows that carbon 

 has entered into intimate association with the iron, a fact 

 which leads us to the next step in the study of the relations 

 between carbon and iron. 



Hempel ^ has shown that, in an atmosphere of nitrogen, 

 iron appears to assimilate the diamond form of carbon 

 more readily than either the graphitic or the amorphous 



Fig. 3 represents an arrangement for heating the diamond and iron /« 

 vacuo. A strip of pure iron, b, is placed between two terminals, c, c' . which 

 are connected with a dynamo The vessel (of gla^s) is rendered vacucns 

 by connecting the tube (/with a Sprengelpump. 1 he iron is then heated 

 by the dynamo, and maintained glowing until all occluded gas is e.xpelled 

 from the iron, which is then allowed to cool in vacv.o. Small pure 

 diamonds, a, a , a", are then placed on the strip of iron through the 

 orifice into which the tube d fits. The vessel is rendered vacuous, and 

 when the iron is again heated in contact with the diamonds it fuses and 

 combines with them. 



forms, but directly carbon is associated with molten iron it 

 behaves like the protean element it is, and the state which 

 this carbon assumes is influenced by the rate of cooling 

 of the molten mass, or even by the thermal treatment to 

 which the solidified mass is subjected. Let me repeat, 

 all are familiar with carbon in the distinct forms of 

 diamond, graphite, and soot : all are alike carbon. It 

 need not be considered strange, then, that carbon should 

 be capable of being present in intimate association with 

 iron, but in very varied forms. 



Now the mode of existence of carbon in soft annealed 

 steel is very different from that in which it occurs in hard 

 steel. I believe that Karsten was the first to isolate, in 1 827, 

 from soft steel a true compound of iron and carbon. 

 Berthier^ also separated from soft steel a carbide of iron, 

 to which he assigned the formula FeC ; but to attempt 

 to trace the history of the work in this direction would 

 demand an entire lecture. I will only add that within the 

 last few years Sir F. Abel has given much experimental 

 evidence in favour of the existence in soft cold rolled steel 

 of a carbide, FcgC, which he isolated by the slow solvent 

 action of a chromic acid solution. His work has been 

 generally accepted as conclusive, and has been the starting- 

 point of much that has followed. 



It will occur to you that the microscope should reveal 

 wide differences between the structure of various kinds of 

 iron and steel, and I am happy to be able to give you en- 

 larged diagrams made from the drawings of Mr. Sorby, 

 the eminent microscopist, which illustrated his very 

 delicate investigations into the structure of steel. ^ 



The point I am mainly concerned with is the existence 

 of a substance which Sorby called the " pearly constituent" 

 in soft steel. This pearly constituent is closely related to 

 the carbide of iron, P^egC of Abel,* and is probably a 

 mixture of FCaC and pure iron. I have diagrammatical!/ 

 indicated its presence in Fig. 4, which will enable me 

 to summarize the work of many experimenters. The 

 diagram (Fig. 4) will serve, for the purpose of illustration, 



' Ber. der deutsch. chem. Gesellschaft, vol. xvlii. p. 998. 

 ^ Ann. des Mines, t. iii., 1833, p. 229. 



3 The reader must refer to the Journal of the Iron and Steel Institute, 

 No. i., 18S7, 255. 

 "• Proc. Inst. Mech. Engineers, January 1883. 



