1910-11.] Influence of Temperature on Carbon Steels. 507 
final value zero, the specimen was left devoid of residual magnetism and of 
magnetic history. The specimen was now allowed to cool by a suitable 
amount, when the cooling was again arrested and a further test carried out ; 
and so on. In this way a series of magnetisation curves was obtained 
from which permeability-temperature curves corresponding to various field 
strengths were deduced. 
It is to be carefully noticed in connection with w.hat follows that the speci- 
men was submitted to the action of the alternating magnetic field (in what 
follows I shall refer to this process, by means of which the specimen is rendered 
neutral, as a “ process of reversals ”) after and not before each test, with 
its preliminary thermal change. Thus a change in temperature intervened 
between each test and the previous application of the process of reversals. 
Now, it is not generally known that purely thermal treatment, no matter 
what temperature is reached in the process, develops in the specimen a 
peculiar state which renders additional precautions necessary. The bearing 
of this fact upon magnetic testing has been investigated by Dr J. G. Gray * 
and Mr A. D. Ross, who have shown that in order that a magnetisation 
curve yielded by a specimen at a particular temperature (following upon a 
temperature change) should be characteristic of the material at the new 
temperature, it is necessary that the specimen should be submitted to a 
process of reversals at the new temperature. 
The importance of attending to this point will be evident from the 
following results, which were obtained on testing a specimen of hard steel 
by the magnetometer method. The specimen was first tested at room 
temperature after having been rendered neutral, when the following read- 
ings were obtained : — 
© 
Magnetising current in amperes . 
0-2 
0-4 
0-6 
0-8 
1-0 
1-5 
Magnetometer deflection 
32 
73 
117 
167 
221 
365 
The specimen was now submitted to a process of reversals at 15° C. ; 
the temperature was then raised to 105° C., and a further test carried out 
with the following results : — 
Magnetising current in amperes . 
0-2 
0-4 
0-6 
0-8 
1-0 
T5 
Magnetometer deflection 
40 
88 
139 
195 
253 
410 
* “On Magnetic Testing,” Phil. Mag., Jan. 1911. 
