THE EUTECTOID TRANSITION POINT OF CARBON STEELS. 179 



pushed into an outer copper tube of similar length, about 2 cm. in diameter. This 

 latter tube was supported within the magnetising solenoid by means of further layers 

 of asbestos paper wrapped round it loosely. 



Experience showed that this arrangement gave a nearly uniform temperature 

 throughout the specimen for different steady currents in the heating coil. To obtain 

 the maximum degree of uniformity the rods might have been made shorter, but they 

 were required for other experiments in which their length was of consequence. 

 Moreover, our object was rather to compare the temperatures at which certain changes 

 took place in the different steels than to find with extreme accuracy the absolute 

 values of these temperatures. 



The rods occupied successively almost exactly the same position in the furnace and 

 the thermocouple was so mounted, in a small porcelain tube, that its junction when in 

 use was at a fixed distance from the end of the specimen. 



The magnetic measurements were made by means of a sensitive and suitably damped 

 quartz-fibre magnetometer provided with the usual compensating arrangements. 



The data necessary for the deduction of the intensity of magnetisation from the 

 scale-readings were recorded ; but the magnetisation is expressed arbitrarily in terms 

 of these scale-readings only, in order to avoid laborious reductions which would have 

 added little to the value of the conclusions drawn. 



The coefficient of self-demagnetisation of the "rods" (roughly 0'09) was determined 

 by comparing the magnetising solenoid fields required to produce given intensities of 

 magnetisation in the 0'85 per cent, rod with those required to produce equal intensities 

 in a ring of the same steel. For the present purpose, however, it is sufficient to give 

 only the fields due to the solenoid and not the effective fields within the rods. 



The procedure in the first series of experiments was as follows. Each of the alloys 

 in turn was placed in the magnetising solenoid. It was demagnetised and a current 

 of 1 ampere, producing a field of approximately 50 C.G.S. units, was then passed 

 through the solenoid. This current was kept constant while the temperature of the 

 specimen was varied. The latter was raised slowly from air temperature to about 

 850 C. in each case and then slowly lowered. Corresponding readings of deflection 

 and temperature were taken over the whole range as the temperature rose and fell. 

 The results obtained at temperatures above 660 C., which are all that are required for 

 our present purpose, are collected in figs. 1 and 4. The former contains the observa- 

 tions made during heating : the latter those made during cooling. The observations 

 for the different alloys are denoted by different signs according to the scheme shown 

 on the figures. 



The scale of the ordinates differs from one alloy to another. The distances of the 

 different specimens from the magnetometer were unequal, being such that all of them 

 gave approximately the same deflection at the air temperature before heating began. 



The figures include, for comparison, results obtained with a rod of nearly pure iron 

 of the same size as the others and examined in the same way. For this specimen the 



2 A 2 



