484 Changes in the Molecular Condition of Iron. 



Of the three consequences of a purely mechanical expla- 

 nation of the processes involved in the hardening of steel 

 by sudden cooling, that which relates to the distribution of 

 density has been shown to be correct ; the second, concerning 

 the distribution of hardness, could not be demonstrated abso- 

 lutely, although it is not to be concluded that it may not be 

 correct for thicker bars ; and the third, which requires an 

 increase of volume produced by sudden cooling, corresponds, it 

 is true, to the behaviour of most kinds of steel, but is strikingly 

 contradicted by the behaviour of other kinds. The mechanical 

 explanation in question assumes, lastly, a parallelism be- 

 tween density and hardness, since it explains an increase in 

 hardness by an increase in density. We have seen, however, 

 that density and hardness by no means increase or decrease 

 together. 



Hence we arrive at the conclusion that in the tempering of 

 a steel bar, besides the mechanical, purely physical, process 

 of sudden contraction, another change also, of chemical nature, 

 takes place. This consists in a combination between the free 

 carbon and the iron. 



Barus* arrived at the same result some years ago, by 

 means of other experiments and considerations. He considers 

 the combination of carbon as the chief reason of hardness, 

 which may, however, be somewhat modified by the mechanical 

 process. 



If, however, we inquire further into the reason of the 

 combination of the carbon, we easily see that the reason can 

 only be sought in the mechanical process, which appears then 

 as the primary cause of hardening. It appears to me that 

 it is essentially great pressures, produced by the sudden con- 

 traction of the red-hot steel plunged into cold water, which 

 bring about the combination of the carbon and the iron. 



This same idea has been already expressed in a paper pub- 

 lished twenty years ago by Caronf. Experiments are there 

 described which are concerned with the behaviour of suddenly 

 cooled steel, of hammered steel, and of slowly cooled steel 

 when treated with acids. The result obtained was that the 

 residue of carbon was inappreciable in the first case, more 

 considerable in the second case, and largest in the third case. 

 Caron concludes from this that the steel hardened by sudden 

 cooling contained its carbon almost altogether chemically, and 

 that the change produced in the most complete manner by 

 sudden cooling may be, partially at least, brought about by 



* Wied. Ann. vii. p. 383 (1879). 



t Comptes Rendus, lvi. pp. 45, 211 (1863). 



