90 
ME. S. W. J. SMITH ON THE THERMOMAGNETIC ANALYSIS OF 
C'B'A, because of the well-known effect of gaps in a magnetic circuit. The amount 
of crystallisation at f is greater than at f, for example, but the excess results from 
crystallisation round nuclei, which is accompanied by the production of layers of non¬ 
magnetic material which surround these nuclei. 
O 
These conclusions can be tested by examination of the relation between the 
permeability data obtained during the cooling of the nickel steel from 810° C. and 
from 750° C. respectively. 
The data are given in the tables (p. 46) in the experiments from 20 b to 28 and 
from 50 to 58. The relation between the permeability variations in these two series 
of observations is shown graphically in fig. 20 (nickel steel), in which the unnumbered 
dotted curve corresponds with that drawn through the points 21 to 29 in fig. 18. 
The quantitative relation between the data is shown in the table below :— 
Nickel Steel (see Curves, figs. 18 and 20). 
Number of experiment. 
(50). 
(51). 
(52). 
(53). 
(54). 
(55). 
(57). 
(58). 
M 
0-395 
0-76 
1-705 
1-49 
1-38 
1-245 
0-97 
1-37 
M 0 
— 
0-1 
2-45 
1-96 
1-80 
1-63 
1-255 
1-80 
m/Fo 
— 
7-6 
0-70 
0-76 
0-77 
0-76 
0-77 
0-76 
The values of /x 0 were obtained, by interpolation from Experiments 21 to 29, for the 
temperatures corresponding to 51, 52, &c. 
From these data it is seen that below 460° C. the permeability variations in the 
two cases are practically identical, and that the value of the permeability in the one 
case is about three-fourths of the corresponding value in the other (cf. Section IVI, 
§ 8, p. 62). 
It is also seen that the observations 50, 51 (fig. 20) agree with the view that 
crystallisation round crystals already present takes place above 650° C. The rapid 
rise of permeability between 51 and 52 is consistent with the hypothesis that the 
bulk of the solution present at 50 began to crystallise from the labile state at about 
650° C. This temperature is approximately that at which a permeability curve, 
corresponding to a constant amount of crystallisation and passing through 50, would 
cut the curve through 21 and 22 (see fig. 18). 
§ 11. The explanatory value of the theory is still further confirmed by the 
measurements, after interrupted cooling, of the permeability at the temperature of 
the air. 
It was noticed frequently (see Section IV., §§5,7 and 8) that the permeability at the 
air temperature was less after interrupted cooling than when the cooling from beyond 
