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



is precipitated in the form of ultra-microscopic particles, which are 

 known as troostite. On raising the temperature the troostite passes 

 into another variety known as sorbite, and at still higher tempera- 

 tures well-segregated pearlite results. The gradual tempering and 

 softening of hardened tool steel is caused by the precipitation of iron 

 carbide in these forms together with soft a iron, and as to this no 

 difference of opinion exists. 



Secondly, all tool steels when properly quenched, and thus 

 hardened, are magnetic. As austenite is non-magnetic, and as, so 

 far as is known at present, the only form of iron which is magnetic 

 is a, it follows logically that some a iron is present in hardened tools, 

 and therefore that martensite is austenite which has undergone some 

 change. 



Up till recent times there were two chief theories of hardening, 

 which were known as the allotropic and the carbonist. The former 

 was mainly a French, the latter chiefly an English theory. What- 

 ever be the verdict of history on these two theories, there can be no 

 question that each has been justified by the volume and character of 

 the research work to which it has led. The two schools of thought 

 were in many respects sharply opposed, but in the hght of the fuller 

 knowledge which exists to-day, it must be admitted that each has 

 made a substantial contribution to scientific knowledge. The 

 allotropic theory is in the main due to the late M. Osmond, who was 

 the first to investigate the critical points of iron and steel by a 

 delicate pyrometer. The history of the development of this theory 

 is particularly interesting, and must be briefly reviewed if a correct 

 perspective of its merits is to be obtained. When M. Osmond first 

 investigated the coohng of pure iron he only detected the point Arg ; 

 as has been already mentioned, this is accompanied by a large 

 evolution of heat. He called the iron above Av.^ /?, that below it a, 

 and upon this he developed the following argument :^There are two 

 kinds of iron, the atoms of which are respectively arranged in the 

 molecules so as to constitute hard and soft iron, quite apart from 

 the presence or absence of carbon. In red-hot iron the mass may be 

 soft, but the molecules are hard ; let this be called /3 iron. Cool 

 such red-hot pure iron, whether quickly or slowly, and it becomes 

 soft ; it passes to the a modification because there is nothing to 

 prevent its doing so. If, however, carbon is present and the steel be 

 rapidly chilled, the following result is obtained : a certain proportion 

 of the molecules are retained in the form in which they existed at a 

 high temperature — namely, the jS modification, which is hard — and 

 hard steel is the result. According to this view, then, the hardness 

 of quenched steel is due directly to iron and only indirectly to 

 carbon. 



Subsequently, when M. Osmond had perfected a more refined 

 method of taking cooling curves he discovered the critical point at 



