200 DR. S. W. J. SMITH AND MR. J. GUILD : A THERMOMAGNETIC STUDY OF 



temperature of the furnace is lowered after solution has begun. The immediate effect 

 is to lower the rate of supply of heat to the material, and therefore to lower the rates 

 of solution of the carbide and the iron at the respective surface films. 



Before the lowering, the rates of diffusion of these were sufficiently rapid to keep 

 their concentrations from rising in the regions where they were dissolving. After it, 

 they will be more than sufficient, and these concentrations will fall, with the result 

 that, even when the temperature of the material is lowered, solution may still 

 go on. 



In practice, as will be seen from the figure, the rate of solution soon becomes very 

 slow, and the temperature has not fallen very far before the magnetisation begins 

 to rise. 



The temperature at which solution ceases will be above the true eutectoid point, 

 unless the diffusion has sufficed to make differences of concentration negligible within 

 the solution which remains. It will thus tend to be nearest the true value (and, 

 incidentally, lowest) when the layers of solid solution are never very broad, i.e., when 

 the heating is interrupted very soon after solution has begun. 



It will be seen that this inference is corroborated by the experimental results. 



It remains to explain why, as in the figure, the temperature at which the lost 

 magnetism begins to be recovered rapidly is lower the higher the temperature at 

 which the heating is interrupted. 



Suppose that re-crystallisation has begun. At one stage, marking the point 

 where concentration differences produced during solution have been approximately 

 obliterated by the reverse changes during re-crystallisation, the layers of solution 

 will be nearly of eutectoid composition. Their thickness will depend upon the amount 

 of solution that took place before the furnace temperature was reduced. 



At this stage the specimen will be at the eutectoid temperature. When it is 

 passed the difference in composition between one surface film and the other will 

 change in sign and will become greater as the temperature falls. At the same 

 time the thickness of the whole layer will decrease. Ultimately, a temperature 

 will be reached at which the ratio of concentration difference to layer thickness 

 is sufficient to permit the rapid rate of diffusion which rapid re-crystallisation 

 requires. 



If the layers of solid solution are very narrow when, during cooling, the eutectoid 

 temperature is passed, the concentration difference required for rapid diffusion will be 

 relatively small, -i.e., the temperature of rapid re-crystallisation will be relatively high. 

 If they are very wide, as happens when the solution is nearly complete before heating 

 is interrupted, that temperature will be relatively low. 



The temperature of rapid increase of magnetisation should thus decrease pro- 

 gressively from its highest value when solution is arrested almost as soon as it has 

 begun to its lowest value when solution is not arrested at all. 



This is exactly what happens experimentally as the curves of fig. 2 show. 



