METEORIC AND ARTIFICIAL NICKEL-IRON ALLOYS. 
89 
material will differ from that of complete crystallisation represented by AB (fig. 27) 
only in the respect that very thin layers of solid solution will be present. All of the 
alloy will be magnetic, with the exception of thin intervening gaps of non-magnetic 
material. The permeability variation for a considerable range (below a temperature 
at which the rate of crystallisation has become inappreciable) will be precisely the 
same as when the material is in the state corresponding to the curve A a. At each 
temperature the permeability will he less in a constant proportion than when the 
material is wholly magnetic. 
The extent to which this conclusion is verified is shown by the following data :— 
Nickel Steel (see Curves, fig. 21) Cooled from 689° C. 
Number of experiment. 
(75). 
(76). 
(77). 
(78). 
(79). 
(82). 
F 
4-53 
2-90 
1-80 
1-39 
0-85 
1-39 
Fo 
— 
3-58 
2-42 
1-74 
1-11 
1-8 
f/Fo 
— 
0-81 
0-75 
0-80 
0-77 
0-77 
The values of /x 0 were obtained from the curve joining points 63 to 69 (Nickel Steel, 
fig. 18), by interpolation, for temperatures corresponding to (76), (77), &c. 
In the second case of fig. 27, V., in which the heating is interrupted at a comparatively 
high temperature, the change in the degree of crystallisation will proceed in a different 
way. There is at first a considerable amount of solid solution comparatively weak in 
nickel, and the temperature is relatively high. During the early stages of the cooling 
there will be a considerable amount of crystallisation round the existing nuclei, and 
the solution round these nuclei will become comparatively rich in nickel. The main 
portion of the solution will, however, remain weak in nickel, and will become labile 
when that temperature is reached which is represented by the point of intersection 
of the horizontal line through r and the curve C'B'A {cf. fig. 26, II.). Subsequently 
crystallisation will proceed from this solution exactly as when uninterrupted cooling 
took place from a temperature higher than C. 
But the portions of the solution from which crystallisation took place during the 
stage represented by rs will have become of similar content in nickel to the solution 
existing at q in the first case, and will not become labile until a low temperature is 
attained. The permeability variation, for a considerable range below a temperature 
such as f ( fig. 27, V.) will be the same on cooling from r as on cooling from C, but the 
permeability itself will he less in a constant proportion at each temperature in the 
former case than in the latter. This will happen, in spite of the greater amount 
of total crystallisation represented by the curve rst' in comparison with the curve 
YOL. CCVIII.—A. N 
