98 
MR. S. W. J. SMITH ON THE THERMOMAC4NETIC ANALYSIS OF 
The early evidence of the presence of this alloy (shown in fig. 9) proves that much 
of the mixture is intimate. 
When all the taenite had been converted into the artificial 27 per cent, alloy, the 
behaviour of the material as a whole would be as follows :—The permeability would 
rise continuously during heating at first, would reach a maximum in the neighbourhood 
of 400° C., would then fall comparatively rapidly until the rate of increase of 
permeability due to increase in temperature of the 6 per cent, alloy (of which the 
meteorite is mainly composed) over-balanced the rate of loss due to increase in 
temperature of the 27 per cent, alloy. The magnetism of the latter would finally 
disappear at about 490° C., and, if the heating were then discontinued, the 
permeability would remain lower than at the corresponding temperature during 
heating until, towards the temperature of the air, the permeability would become 
subject to a comparatively rapid increase. 
§ 8. Inspection of fig. 14 (2nd winding) shows that this is almost exactly what is 
observed after the material has been heated a large number of times. The only 
essential difference is that there is still at first a small decrease in permeability, 
which ceases below 100° C., and which would be accounted for if there were still in 
the taenite small quantities of alloys containing slightly more than 27 per cent, of 
nickel. 
On the other hand, when the material has been heated only a few times, the 
variation is different, as shown in the curves for the 1st winding. Here there is a 
pronounced decrease of permeability between 400° C. and 450° C. as before—pointing 
to the presence of a considerable amount of 27 per cent, alloy; but there is also 
another pronounced decrease during the first stages of the heating from 15° C. This 
is obviously what would occur if, besides the artificial 27 per cent, alloy, there were 
still appreciable quantities of alloys containing amounts of nickel intermediate 
between 27 and 40 per cent. 
In the lower curve (cf. Experiments, a, b, c, Section IV., § 3, p. 42) the decrease 
continues approximately to 300° C., and hence in these early observations there is 
detectable the alloy containing at least 37 per cent, of nickel (see fig. 23). 
j 9. Of special significance is the peculiarity, above referred to, of these earlier and 
of the later curves, namely, that the percentage decrease of permeability at lower 
temperatures during heating is much greater when the permeability follows the upper 
curve than when it follows the lower. 
This peculiarity will be seen by comparison of the experiments a and b, 33 and 34, 
1st winding, with the experiments 1 and 2, 39 and 40, and also of the experiments 
111 and 112, 2nd winding, with 119 and 122. 
These results are in complete accord with the interpretation of the magnetic 
phenomena which is advocated. In the lower curves the permeability is relatively 
small, because (see § 11, below, p. 100) most of the irreversible 27 per cent, alloy has 
not undergone the transformation into the magnetic state. There are, in consequence, 
