504 Professor H. C. H. Carpenter [March 7, 



crystallization occurs. In the case of pure iron it has been shown by 

 Chappell and Goerens that the minimum temperature for the strain 

 hardening of iron is about 500° 0. It is quite certain that the 

 minimum temperature for a steel containing • 9 per cent carbon will 

 be higher than this. How much is not known, but let the figure 500° C. 

 stand for this particular steel. In hardening the steel is quenched 

 from a temperature which should not be higher than 800' C. The 

 important interval therefore in the quenching operation is between 

 800° C. and 500° C. In one of Benedicks' experiments, to which 

 reference has been already made, the time taken to pass through this 

 interval was exactly one second ; in the case of a tool of about one 

 inch diameter it will take longer for the centre to pass through this 

 interval, but the probability is that it will not take more than two to 

 three seconds for the skin of the tool to pass through this interval. 

 In such a case, therefore, our view is that the vitreous amorphous 

 layers formed will certainly not have time to re -crystallize, but will 

 be preserved, and hardening will result. It would follow, then, that 

 the final cause of hardening by quenching is exactly the same as that 

 of hardening by cold working. In the case of tool steels and other 

 alloys of a similar nature, the actual stresses which are set up in 

 quenching, which give rise to the twinning or slipping, are brought 

 into action only in the very brief period during which the mass is 

 cooling ; therefore, while the metal is undergoing inter-crystalline 

 slip and the amorphous layers are being formed, the chances for these 

 to re-crystallize become less and less as the cooling proceeds. Indeed 

 it may be that a large proportion of the slipping which occurs takes 

 place at moderately low temperatures below that at which the amor- 

 phous layers are able to re-crystallize. Put briefly, then, our theory 

 of the hardening of steel is, that it is due to the formation of hard 

 vitreous-amorphous films of a solution of carbide in y iron formed 

 along the slipping planes of the austenite. 



So far as I am aware, there is only one property of quenched 

 steels which is not accounted for by this theory, namely, that all 

 quenched steels are magnetic. It has not been shown that amorphous 

 7 iron is not. magnetic, though it is not at all probable, since the 

 property of magnetism is almost certainly associated with a definite 

 orientation of the molecules in iron and steel. Admitting, as may be 

 done on present evidence, that ferro-magnetism is a property of 

 a iron, our theory can be brought into harmony with the fact that 

 all quenched steels are magnetic, by adopting McCance's view that 

 some 7 iron passes to a in the quenching of all steels. 



The third modern theory of hardening to which I wish to allude 

 was put forward some six months after the two theories which have 

 been briefly outhned. It is due to Mr. J. C. Humfrey.* This 

 theory rests upon the assumption that a hard amorphous phase is 



Transactions of the Faraday Society, 1914, pp. 35-43. 



