64 
MR. S. W. J. SMITH ON THE THERM0MAGNETIC ANALYSIS OF 
Fig. 21. Nickel steel. 
Reheating to 810 C. interrupted at 690° C. 
Variation of // during subsequent cooling to 15° C. and reheating to 660° C. 
and then allowed to cool, with stoppages for the measurement of permeability. The 
results are shown in the points 76 to 79 (fig. 21). The air temperature permeability 
was now even lower than after previous coolings from 705° C. and 750° C. respec¬ 
tively (cf. 79, 42, and 57). The ring was then heated beyond the critical point and 
then allowed to cool to the temperature of the air. The permeability was raised 
about 10 per cent, by this process, but was still comparatively low. Reheating to 
660° C. did not seem to raise the permeability (cf. 85 and 88), but the effect of 
heating to a lower temperature, e.</., 600° C., was not observed. The latter treatment, 
as can be inferred with certainty from the earlier and later experiments, would have 
raised the permeability at the air temperature. 
The sets of observations 75 to 79, 36 to 42, 50 to 57, 9 to 13, and 20 to 26 show 
that the air temperature permeability after cooling from beyond the temperature of 
maximum permeability is greater in proportion as the temperature from which the 
cooling begins approaches that at which ferromagnetism has disappeared. 
From the above results it can be inferred that the permeability does not vary 
