

THE EUTECTOID TRANSITION POINT OF CARBON STEELS. 203 



coefficient of the magnetisation now. remains positive until the structural change 

 point is reached. 



It would seem, therefore, that by suitable adjustment of the field in any particular 

 case, the thermomagnetic method can be used to indicate the change point, known 

 to metallurgists as Ac,, with an accuracy that can scarcely be exceeded in measure- 

 ments by any other method. 



The experiments of fig. 13 followed immediately after those of fig. 12. The 

 material was cooled from F to G, then heated again to H, cooled to I (about 640 0.), 

 then heated to J, cooled to K, and finally heated to L. 



The observations from I to J show that the lag during heating (the existence of 

 which is again clearly shown) returns to its full value when the material is allowed 

 to cool to about 640 C. before it is re-heated. The series JKL goes a stage further 

 and shows that the same is also true when the cooling is stopped at about 705 C. 



Such observations indicate another thermomagnetic method of determining the 

 temperature at which, during cooling under given conditions, the eutectoid change is 

 complete. They also corroborate the hypothesis that the lag during heating is due 

 to surface effects. 



20. The Equilibrium Temperature. 



If the interpretation of our results which has been given is correct, they show that 

 (subject to limitations due to variations of pressure) iron crystals, carbide crystals, 

 and a solid solution of the two of uniform (eutectoid) composition cannot coexist 

 without change except at a definite temperature. 



According to the results obtained with the 0'85 per cent, steel, shown in figs. 3, 

 12, and 13, this temperature lies between 725 and 730 O. 



Fig. 9A* contains the results of an attempt to find the equilibrium temperature by 

 means of the steel containing 1'2 per cent, of carbon. The method was the same as 

 before, except that we used a field of 200 C.G.S. instead of the previous fields of 25 

 and 50 C.G.S. respectively. The thermocouple had been calibrated more recently and 

 was probably rather more trustworthy than that used in the earlier experiments. 



The point Q on the " interrupted heating " curve is that at which the temperature 

 coefficient of ss with respect to is the same as at the corresponding temperature 11 

 on the (uppermost) curve of continuous heating. It is presumably, therefore, to a 

 first approximation, the temperature at which the change from solution to re-crystal- 

 lisation began during cooling. 



The point P on the " interrupted cooling " curve is similarly that at which the 

 temperature coefficient of ^ with respect to 6 became the same as at the corre- 

 sponding temperature on the uppermost curve. It is, therefore, to be regarded as 

 the point at which change from re-crystallisation to solution began during heating. 



It happens that the temperatures corresponding with the points P and Q are 



* See p. 195. 



