METEORIC AND ARTIFICIAL NICKEL-IRON ALLOYS. 
105 
II the above conclusions are correct it would seem that the eutectic cannot be 
definitely detected microscopically in artificial alloys only because of the absence of 
sufficient segregation, but it is still possible that the thermomagnetic method (of 
which the advantages have been pointed out) may reveal its presence. 
§ 3. The thermomagnetic data, for various artificial alloys, have already been 
discussed, but there is a further series of experiments to which reference should be 
made (of Honda and Shimizu, ‘Phil. Mag.,’ 1905, pp. 548 and 642). Messrs. Honda 
and Shimizu have not, to my knowledge, given any interpretation of their results, 
but they have determined the thermomagnetic properties for temperatures between 
about —186° C. and 800° C., and fields up to about 400 C.G.S. units, in the case of 
eleven nickel-iron alloys. 
For the present purpose the most important data are those for a 36 per cent, alloy, 
and for a set of six alloys containing percentages of nickel ranging from 29'24 
to 24'04. Unfortunately there are no similar data for alloys containing lower 
percentages of nickel. 
According to the views stated above, the 36 per cent, alloy will give rise on cooling 
to mixed crystals almost wholly of the nickel-rich type. 
Of the others, the 29^ per cent, alloy, containing a little more than the eutectic 
percentage of nickel, will give first mixed crystals of the nickel-rich type, but, 
subsequently, when the eutectic temperature is reached, a eutectic mixture of nickel- 
rich and nickel-poor crystals. The 24 per cent, alloy, on the other hand, containing a 
little less than the eutectic percentage of nickel, will give first mixed crystals of the 
nickel-poor type, and, subsequently, at a lower temperature, the eutectic mixture. 
§ 4. In seeking for the interpretation of the results obtained, it is necessary to 
consider the extent to which nickel-rich and nickel-poor crystals are likely to be 
thermomagnetically distinguishable. 
A study of the permeability curves for nickel-iron alloys, at the air temperature 
after cooling from a high temperature, shows that these alloys can be divided 
magnetically into two classes. 
When the percentage of nickel is 35 or more the alloy is magnetically soft. The 
“ second stage ” in the magnetisation curve is soon attained and the maximum 
permeability occurs when the field intensity reaches 2 or 3 C.G.S. units. 
When the percentage of nickel is 25, or less, the alloy is magnetically hard. The 
hardness decreases as the nickel content is reduced. When the percentage of Ni 
is 25 the maximum permeability is not attained until H = about 50 C.G.S units, but 
even when the percentage is as low as 5, a field intensity of, roughly, 10 C.G.S. is 
required. 
It is clear, therefore, that mixed crystals containing about 7 per cent. Ni and forming 
at or near the temperature of the air must, although their exact properties are not 
deducible a priori , be distinguishable magnetically from mixed crystals containing 
' not less than 37 per cent, of nickel. 
VOL. COVIII.—A. 
P 
