1919] on The Hardening of Steel 495 



empirically to the formula Fe24C. According to the views of later 

 carbonists this hardness is the property of this solid solution, which is 

 stated to be amorphous, and the whole of the carbon is regarded as 

 having been trapped by quenching and retained in solid solution. 



Modern research has shown that this theory is right in its 

 assumption that the whole of the carbon is dissolved in a properly 

 quenched steel, but in their desire to disprove the allotropic theory 

 of hardening the exponents of this theory have gone to the unwise 

 extreme of denying that the allotropic forms of iron have any 

 influence on the hardness of tool steel at all. When it is remembered 

 that iron carbide is practically insoluble in a or /? iron, but is only 

 soluble in y iron, it is certainly unscientific to deny that y iron 

 has any share in the hardening process. Moreover, the theory as 

 stated in the above form is both incomplete and contradicted by 

 certainly one fact. It is certainly not correct to say that the con- 

 stituent formed on quenching a tool steel above A^ is amorphous, if 

 by that is meant that it is wholly amorphous ; the structure of a 

 quenched tool steel is in the main crystalline, although, as will be 

 subsequently suggested, some amorphous material may be present. 

 The carbonist theory does not explain the martensitic structure of a 

 quenched tool. If there was nothing more in the hardening process 

 than is contained in the carbonist theory, the structure of a quenched 

 steel should consist of polyhedral crystals, because this is the typical 

 structure of a solid solution. Further, the hardness of quenched 

 steel is much greater than that which can be explained merely by 

 the retention of iron carbide in iron. Some other factor must enter 

 in. Finally, the fact established by McCance, that on adding carbon 

 to iron the hardness of the quenched steel increases up to • 7 per cent 

 carbon, and remains constant between this point and 1*18 per cent, 

 receives no explanation from the carbonist theory, and appears to 

 point to the conclusion which McCance has already drawn, that at 

 any rate some of the hardening must be due to the iron itself, and 

 that there is a definite limit to the amount of hardening that iron 

 can stand. 



These two theories mainly held the field for the best part of 

 twenty years, and anyone who reads the history of this stimulating 

 period in the development of metallographical thought cannot. I 

 think, avoid the conclusion that the allotropists made a serious 

 mistake in laying such exclusive emphasis on the function of iron, 

 while the carbonists made an equally serious mistake by denying, that 

 iron had anything to do with hardening at all. 



I pass now to consider some important research work, which, 

 although it was not carried out on iron or steel or indeed any other 

 alloy of iron, has nevertheless profoundly influenced the views of 

 metallographers with regard to the causes of the hardening of steel 

 by quenching. This is but one instance of many which the history 

 of scientific discovery provides, that the great advances in any par- 



