104 
MR. S. W. J. SMITH ON THE THERMOMAGNETIC ANALYSIS OF 
should disappear. For the material should thereby be converted into a uniform solid 
solution which, on cooling, under ordinary conditions would undergo the same changes 
as take place in the artificial alloys. The disappearance of the meteoric structure 
would take place by interdiffusion between the taenite and the kamacite by which 
the original taenite layers would become eventually much poorer in nickel and the 
kamacite layers a little richer. This process would apparently be very slow, however, 
unless the temperature were raised much above the highest used in the present 
experiments in which very little interdiffusion of these two constituents seems to have 
occurred. I have attempted to show, however, from the thermomagnetic curves, that 
the temperature conditions permitted an analogous process, on a very minute scale, to 
take place within the taenite itself in the time over which the experiments extended. 
The taenite, initially a relatively coarse mixture in which, owing to the slowness and 
uniformity of the original cooling, comparatively large crystals of the two components 
of the eutectic were present, became gradually converted into a fine-grained mixture 
of the same constituents possessing appreciably different thermomagnetic properties. 
Section VIII.— Further Data from Artificial Alloys. 
§ 1. In the preceding section an attempt lias been made to establish quantitatively, 
from thermomagnetic data, the truth of the hypothesis, due originally to Foozeboom, 
that nickel-iron alloys are examples of a system giving rise to a series of mixed 
crystals of two types and possessing a eutectic, or transition point. 
The evidence is based mainly upon a study of meteoric nickel iron and, if the 
conclusions are correctly drawn, it follows that:— 
(a) If the cooling alloy contains less than about 27 per cent, of nickel, the crystals 
which form first are of the nickel-poor type and contain, when stable equilibrium is 
reached, not more than 7 per cent, of nickel. 
( b ) If the cooling alloy contains more than 27 per cent, of nickel, the crystals which 
first form are of the nickel-rich type and contain not less than 37 per cent, of nickel. 
(c) If the cooling alloy contains (approximately) 27 per cent, of nickel, there are 
formed simultaneously mixed crystals of two types, one containing not more than 
7 per cent, and the other not less than 37 per cent. Ni. The transition begins (of. 
Section VII., § 11) in this alloy at about the ordinary temperature of the air. The 
beginning of the transition in any other alloy occurs at a higher temperature. 
§ 2. Although the existence of the eutectic is to be expected below 0° C. in any 
alloy containing more than 7 per cent, or less than 37 per cent, of nickel, it has not 
yet been identified micrograph ically in artificial alloys. 
I have attempted to show, however, that it can be identified in meteoric iron in 
which the rate of cooling has probably been slow and uniform, and which has probably 
remained for a very long time at or below the eutectic temperature. In such alloys 
the segregation of the two constituents of the eutectic has proceeded to an extent 
that is magnetically discernible. 
