METEORIC AND ARTIFICIAL NICKEL-IRON ALLOYS. 
101 
amount of internal strain may arise. It is in the already strained material that the 
transformation of the 27 per cent, alloy has to begin. 
But if the material is subsequently heated to a temperature not high enough to 
cause the 6 per cent, alloy to begin to undergo the magnetic transformation in the 
reverse direction, the internal strain in this alloy will be diminished. The effect of 
such heating will be to make more homogeneous the distribution of iron and nickel 
throughout the material (Section VI., § 12, p. 92). A sufficient number of heatings 
would convert it from a heterogeneous mixture of alloys containing different percentages 
of Ni losing and gaining their magnetic properties at different temperatures into a 
homogeneous alloy magnetic at all temperatures between 15°C. and 600° C. 
When the temperature is reached, during cooling, at which the 27 per cent, alloy 
(imbedded in the now more homogeneous alloy) tends normally to become magnetic, 
its transformation—retarded at first, perhaps in a way analogous to the depression of 
the freezing-point of water by pressure—may become possible. Indeed, when it is 
remembered (cf Hopkinson and Guillaume) that an irreversible alloy rich in nickel 
increases by about 2 per cent, in volume as it changes from the non-magnetic to the 
magnetic condition, it is clear that an appreciable part, at least, of the observed 
retardation must be due to an effect of the kind just mentioned. 
§ 12. The views expressed in the above and in the preceding paragraphs were 
confirmed by some experiments on the effect of thermal treatment upon the density 
of the meteoric iron, in which it was found that changes in density corresponding with 
the changes of permeability produced by heating to different temperatures, as above 
described, could be observed ; but these experiments are omitted for the present from 
considerations of space. 
It may be mentioned, however, as showing the advantages of the thermomagnetic 
method of analysis, that a change in the taenite which produces an alteration of over 
50 per cent, in the permeability of the meteorite is not enough to produce a change of 
more than 0T per cent, in the density. 
§ 13. If the conclusions already given are correct, it is possible to discuss with some 
confidence the conditions under which the characteristic structure of meteoric (octa¬ 
hedral) iron has arisen. 
Suppose that the alloy is comparatively poor in nickel, fie., that it contains only a 
little more than the minimum amount necessary for the existence of Widmanstatten 
figures, and suppose further that the structure is acquired during cooling of the alloy 
from a comparatively high temperature. 
The material will be a homogeneous solid solution of iron and nickel at temperatures 
above 700° C. Below this temperature the solution is metastable, but spontaneous 
crystallisation (see Section VI., § 1, p. 78) cannot occur until a temperature of about 
600° C. is reached. The solution then becomes labile, and spontaneous generation of 
nuclei can take place. If the subsequent rate of cooling is extremely slow the 
crystallisation will proceed (see p. 79) around the first nuclei. 
