650 TRANSACTIONS OP SECTION B. 



not given off as H 2 S, and that a part either passes off as S(CH 3 )., or remains 

 behind with the solution as some organic product — was tentatively believed as 

 indicative that the sulphur is chemically associated with the carbon and the iron. 

 Levy, 1 who has done much good work in the endeavour to determine the rela- 

 tions which exist between iron, carbon, and sulphur, in the alloys of these 

 elements, states, as the result of his research, that there is no conclusive evidence 

 of any chemical union. 



In his tabulated results showing the amount of sulphur evolved presumably 

 as S(CH„) 2 on dissolving iron, carbon and sulphur alloys, the maximum is 

 O06 per cent., but the average is very much less. 



Schulte, on the other hand, had found that from 1 per cent, to 12 per cent, of 

 the total sulphur is evolved as an organic sulphur compound ; and Bischoff found 

 an even greater quantity. 



The results are apparently conflicting, and it is evidently obvious that more 

 research is required in this direction. 



It has been shown by Arnold and McWilliam, and confirmed by others, that 

 carbide of iron does not decompose into graphite and iron during the annealing 

 of steel until it segregates into relatively large masses. Taking this as a basis 

 Mr. Levy has advanced an explanatory hypothesis as to how it is that sulphide of 

 iron prevents the decomposition of carbides in white irons. He had found that 

 during the solidification of irons free from silicon and manganese, but rich in 

 sulphur, ' the sulphide separates at a temperature in the neighbourhood of 

 1130° C, together with, and as a component of, the austenite-cementite eutectic, 

 forming a triple austenite-cementite-sulphide eutectic, the cementite component 

 of which is interstratified with a jointed pearlite (by decomposition of austenite) 

 sulphide one.' He stated that ' The presence of iron sulphide in the eutectic intro- 

 duces intervening layers, which may partly ball up on annealing, but even then 

 leave sulphide films between the cementite crystals ; these act almost as emulsifiers, 

 preventing the coalescence of the cementite portion, which is apparently a neces- 

 sary preliminary to its decomposition into free carbon and iron. These layers and 

 films are so persistent, even on slow cooling, as to retain their position between 

 the cementite crystals, until the metal has cooled well below the temperature of 

 decomposition, so that an iron which might otherwise become grey is retained, 

 even on very protracted cooling, in the white form, by sulphur as sulphide; 

 0-25 per cent, sulphur being sufficient for this purpose under the moderately 

 protracted cooling conditions of the research. It is not improbable that the 

 mechanical force exerted by sulphide, on separation and cooling, may also 

 prevent the physical conditions necessary for carbide decomposition, which, as 

 is well known, is accompanied by considerable expansion.' 



It is to be noted that Mr. Levy's argument is based on the effect of the 

 sulphide films in the eutectic, preventing the segregation of the cementite into 

 relatively large masses, which, as he expresses it, 'is apparently a necessary 

 preliminary to its decomposition.' 



His conclusions were based on the examination of hypo-eutectic alloys contain- 

 ing not more than 275 per cent, carbon and free from massive plates of cementite. 

 Whilst admitting that his conclusions may be correct, as applied to the 

 eutectic, some other explanation would be necessary if decomposition did not 

 occur when a considerable quantity of massive cementite initially were to form 

 in the alloy. 



That stable massive cementite can be so obtained in iron sulphide alloys I 

 shall presently show. 



If it could be shown that sulphur in some form of combination with the iron 

 and carbon does crystallise with the carbides, and that such mixture or solid 

 solution is stable and not readily decomposed, it would be reasonable to conclude 

 that the sulphur is responsible for the stability. 



It has been suggested that silicon in iron decomposes the carbides according 

 to the following chemical reaction: 3Si- 2 Fe 3 C = 3 Fe,Si-2C. The only objec- 

 tion to this explanation is that the silicon is not free in cast iron, as was proved 

 by Turner, and, moreover, as will be shown presently, it is combined with iron 

 in solid solution before the carbide is decomposed. 



1 Journal of the Iron and Steel Institute, No. 2, 1908. 



