11)19] on The Hardening of Steel 49;3 



Ai'o. As already mentioned, the evolution of heat is very much less 

 than at Arg, and for some time he was unable to decide whether Ar^ 

 was merely the termination of Arg spread over more than 100° or 

 was an independent point. Finally, when he had shown that at Ar, 

 magnetic properties appear in the iron, he decided that this point was 

 independent of Arg. The logical consequence of this was that three 

 varieties of iron had to be considered. Now comes the important 

 point — M. Osmond changed his theory of hardening, and put forward 

 the view that it was not the iron above Arg which was hard, but the 

 variety which existed between Arg and Ar2. The iron above Arg was 

 called 7 ; that between Arg and Ar2, /5 ; and below Ar^, a. 



The difference between the two theories will now be obvious. 

 Whereas in the former the hard form of iron was that which existed 

 in what we know to be the y region, in the latter it was the variety 

 which existed in the narrow temperature range between Arg and Av^- 

 Hardening was thus due to an intermediate form of iron between the 

 varieties stable at high and low temperatures. According to the new 

 theory ^ iron was still supposed to be the cause of hardness, but the 

 new /? was quite different from the old one, and progress in the 

 development of the theories of hardening would have been more 

 rapid if it had been generally realized that the original hard /S iron 

 was, according to the more accurate knowledge revealed by later 

 work, y iron. The new theory presupposed the existence of all three 

 forms of iron in a quenched tool, and according to it the act of 

 quenching permitted the retention of only a portion of the y iron, 

 while the remainder was converted into a mixture of /3 and a, the 

 former conferring the hardness, and the latter the toughness and 

 magnetic properties. 



This theory was enthusiastically adopted by many workers, though 

 certain modifications of it were put forward from time to time. 

 AVhereas the Osmond theory predicated that the /3 iron separated in 

 a carbonless form, the whole of the carbon being retained in the 

 y iron, one of the modifications put forward was that the ^ iron 

 contained some dissolved carbon. The theories, however, all had 

 this in common, that the hardness of quenched tool steel was supposed 

 to be due to a crystalline variety of iron, namely, /?, intermediate 

 between y and a. According to these theories, then, the special 

 constituent present in quenched steel, namely, martensite, is due to 

 /3 iron. 



It is possible to test this theory to-day by fuller knowledge, and it 

 will be sufficient to advance two considerations which seriously invali- 

 date it. In the first place, if the equilibrium diagram be examined, 

 it will be seen that in the case of all tool steels — that is, those containing 

 from 0'7-l*5 percent carbon — yiron passes directly to a and not by 

 way of /3 iron at all, and it is difficult, though not impossible, to see 

 why /3 iron should be formed. The usual answer made to this objection 

 is that the diagram refers only to the stable varieties of steel, and 



