METEORIC AND ARTIFICIAL NICKEL-IRON ALLOYS. 
55 
for the measurement of permeability) to the temperature of the air. The resulting 
permeability was again very low (see 72) and practically equal to that obtained after 
cooling from 770° C. and 720° C. respectively (see fig. 12). 
Reheating to 850° C. interrupted at temperatures between 700° C. and 800° C. 
Variation of // during subsequent continuous cooling to air temperature. 
On fig. 12 there are plotted also the results of a much later set of observations 
(168 to 171) which lead to the same result as the above sets. They serve further to 
show the constancy in the behaviour of the material. 
Apart from showing that very low and nearly equal values result from uninterrupted 
cooling from a temperature between 700° C. and 800° C., the numbers just given show 
that Guillaume’s qualitative diagram does not represent exactly what happens in the 
present case ; since, if it did, the permeability at ordinary temperature in each of the 
above cases should be the same as that obtained after uninterrupted cooling from 
above 850° C. 
The experiments 74 to 80 (shown in fig. 11) prove that, when cooling from 850° C. 
is interrupted near the temperature at which the permeability ceases to rise, 
subsequent reheating is accompanied by irreversible change of permeability. The 
nature and extent of this irreversibility is shown further in fig. 13. From the 
experiments 82 to 86 it is seen that, after a second reheating, the state of the 
material is practically reversible and the hysteresis with respect to temperature has 
almost disappeared, being ten times smaller after the second reheating from 490° C. 
than after the first (cf. 31 to 34 and also experiments described later). 
Fig. 13 shows also the result of the first attempt to trace exactly how heating 
interrupted in the neighbourhood of 600° C. results in an exceptionally high 
