332 PROFESSOR J. A. EWING AND MR. G. C. COWAN. 



susceptibility was 1'7. With each addition of load this was reduced, until it fell 

 with 18 kilos, to 07. These numbers should be compared with those given for iron 

 by Lord RAYLEIGH,* who has shown that under very feeble magnetising forces the 

 susceptibility of that metal has a finite and sensibly constant value. For one speci- 

 men of unannealed Swedish iron he gives 6 '4, and for another specimen 6 '8 as the 

 initial value of K, which is constant only up to a magnetising force of about 0'04 c.g.s. 

 units. Thus in nickel the initial susceptibility is much less than in iron (just as the 

 maximum susceptibility, at a later stage in the magnetising process, is much less), but 

 the range of magnetic force within which a sensibly constant value applies is 

 immensely greater. 



Magnetisation of Impure Nickel. 



A few supplementary experiments were made with a specimen of cheap commercial 

 nickel wire, 0'154 cm. in diameter, which was found to contain about 4 per cent, of 

 iron. A piece 41 cms. long was annealed and was subjected to cyclical magnetisation, 

 with results which are shown in fig. 10. The chief difference between this curve and 

 that of fig. 2 is the higher limit to which 3 tends in the present case, which is no 

 doubt to be ascribed to the presence of iron in this impure specimen. A set of 

 readings of residual magnetism were afterwards taken, and these, along with the ratio 

 of residual to induced magnetism, are also shown by curves in fig. 10. The maximum 

 ratio of residual to induced was 0'74. The dissipation of energy by hysteresis in the 

 cycle of fig. 10 was 12,600 ergs. The effects of stress on the magnetic qualities of 

 this wire were also examined, and were found to agree in all general features with 

 the effects observed in purer samples of nickel, which have been described above. 



* ' Phil. Mag.,' March, 1887. 



