METEORIC AND ARTIFICIAL NICKEL-IRON ALLOYS. 
49 
the permeabilities at 315° C., 460° C. and 575° C. (approximately) were subject to little 
variation under the treatment undergone. 
The observations 10 to 16 were undertaken to find, if possible, how the permeability 
varied between 100° C. and 400° C. There seemed to be no sudden or irreversible 
changes; but the permeability was now intermediate between its earlier values and those 
observed after the lapse of 3 months. No experiments were made between those 
numbered 16 and 20, except to explore without success for a possible increase of 
permeability in the neighbourhood of 300° to 350° C. 
The experiments 20 to 37 were undertaken with the object of finding again, under 
the apparently changed condition of the ring, the variation of permeability as the 
temperature was raised above the critical point, and, also, the manner in which the 
permeability reappeared as the ring was gradually cooled. 
The observations 23, 24. and 26 to 28 seemed to show that there was an appreciable 
difference between the hysteresis in the reappearance of magnetism in the meteorite 
and that in nickel steel of similar composition examined by Hopkinson. In 
Hopkinson’s curve for a 47 per cent, nickel steel (H = 0'12 C.G.S.) the portions 
of the fxO curve below 600° C. (for rising and falling temperatures respectively) would 
apparently be nearly superposable, whereas in the present case they were not. It 
was thought that the distance between the points 26, 24, &c., and corresponding 
points on the curve of rising temperatures might possibly have been less if the ring 
had been cooled more slowly. For this reason the observations were repeated (see 31 
and 32), and the ring was kept for several hours at temperatures near 550° Cl, after 
cooling, in order to find whether a gradual rise in permeability could be detected. 
The value of /x at the end of two hours was, however, in each case not perceptibly 
different from its value after two hours more, and hence, if the change in question 
was actually occurring, it was of a much more gradual nature than those represented 
in the curves described here, and probably, also, in the curves described by 
Hopkinson. 
The observations 33 to 35, and also 29, showed that the state of the material (after 
gradual cooling from about 820° C.) was practically identical with its state at the 
beginning of the observations three months before. (A preliminary heating beyond 
the critical temperature had preceded the latter observations.) 
The observation 37 showed that the permeability at 575° C. was practically the same 
as when heating took place from the state represented by Experiment 1, at which 
the permeability was much higher than at a , 29, 33, &c. 
Thus it appeared that, although there was considerable uncertainty concerning the 
nature of the variation of /x with 6, especially at temperatures below 500° C., the material 
was not one in which the properties were continually changing in an arbitrary manner 
owing to gradual oxidation, to disintegration, or to some other effect increasing with 
time and arising possibly from an unassignable cause. 
The permeability was next measured (38 and 39) after the temperature had again 
VOL. ccviii.—A. H 
