METEORIC AND ARTIFICIAL NICKEL-IRON ALLOYS. 
59 
neighbourhood of 650° C. has the effect of slightly raising the air temperature 
permeability obtained on subsequent continuous cooling. Hence, under the conditions 
described, it may be said that the change in question is not strictly reversible, 
although by analogy, from results already given, it is probable that a few reheatings 
and coolings would result in subsequent practical reversibility. 
The final experiments, 182 to 198, show the effect of alterations over various 
ranges of temperature. The points 182 to 187 (fig. 16) show the irreversibility already 
noted (cf. 78 to 80). Points 190 to 194 show, over a wider range than 85 and 86, 
that the permeability becomes reversible under repetition of heating and cooling 
between about 430° C. and 680° C. It is not until the temperature is lowered beyond 
430° C. that the permeability shows any further rise on heating. The observations 
195 to 198 show, over a wider range, the same irreversible effect as occurs between 
182 and 187. 
§ 8. Hadfelcl 5’8 per cent. Nickel Steel. —(See fig. 17, &c.) 
The object of the experiments on the nickel steel was to enable a comparison to be 
made, under conditions as far as possible identical, of the thermomagnetic properties 
of natural and artificial nickel-iron alloys of similar percentage composition. The 
rings were almost identical in size, were heated in the same way, and had primary 
and secondary coils as nearly as possible alike. 
The nickel steel contained almost the same percentage of iron as the meteorite, 
but about 1 per cent, less of nickel. It contained about 0 ’7 per cent, of manganese 
and smaller quantities of carbon and silicon. All these substances were absent in the 
meteorite, which contained, however, about 0 - 4 per cent, of cobalt. 
The experiments 1 to 8, in the table of results, were made with the ring in the 
unannealed condition in which it was after construction, and serve only to show that 
the permeability begins to rise rapidly at about 500° C., and that the maximum had 
been passed at 700° C. It seems worth while to note (although the data upon which 
the inference is based are few in number) that the positions of the points 1 to 7 (see 
fig. 17) suggest a closer resemblance, below 550° C., between the properties of the 
nickel steel in its unannealed state and those of the meteoric iron than was found 
after the annealing of the former. The possible significance of this will be seen later 
(cf. Section VI., § 12, p. 92, and Section VII., § 14, p. 102). 
The experiments 9 to 13 were made to find the effect of annealing, and those from 
14 to 16 to show the effect of reheating interrupted at a temperature in the region 
in which the permeability is increasing rapidly with rise in temperature. The results 
are shown in fig. 17, in which the outer dotted curve, running from 15° C. to 810° C., 
is drawn in order that the relation between the present data and those obtained later 
may be represented in the figure. 
The ring was next annealed at a higher temperature (see 20 b), as it was doubtful 
whether, in Experiment 9, the ring had been heated to a temperature at which 
i 2 
