238 Lord Rayleigh on the Behaviour of Iron and Steel 



is perhaps as small as could have been expected. It is fair to 

 conclude that, if the compensating-coil could have been per- 

 fectly adjusted for a Yery small step (the actual step was 

 scarcely small enough), the uncompensated effects visible with 

 larger currents would have been expressible by a quadratic 

 term simply. 



The currents (x) given in the tables are reduced to C.G.S. 

 measure when divided by 1 6 . On the same system the mag- 

 netizing force is 



8-2xl(T 4 x#; 



so that the force due to the strongest current referred to in 

 the table is 1*2 C.G.S. , or about 7H. When the current is 

 reversed, the change of magnetic force is of course the double 

 of this quantity. 



In extending the definition of susceptibility to cases in 

 which the force is not very small, we might proceed in more 

 than one way. If we take the ratio of the change of mag- 

 netization to change of force when the force is reversed, we 

 are following good authorities; and we get a definition which 

 is at any rate consistent with the definition necessary when 

 small forces are concerned. The values of k for different 

 forces are not given by a direct comparison of the numbers in 

 Table II., since the magnetometer-scale is arbitrary ; but we 

 may find for what force the susceptibility is (for example) the 

 double of that applicable to infinitely small forces. 



Table II. 



Current, 



X. 



Displacement, 



y- 



■0053a:. 



l-072.r 2 . 



-0005307. 

 +1-072* 2 . 



Diff. 



99 



1 



052 



1-05 



0-5 



+0-5 



197 



Si 



10 



42 



3-2 







326 



10 



1-7 



11-4 



9-7 



-f-0-3 



483 



22£ 



2-6 



25 



22-4 



-0-2 



934 



88£ 



4-9 



93-7 



88-8 



-0-3 



1149 



136 



6-1 



141-5 



135-4 



+0-6 



1298 



174 



69 



180-6 



173-7 



+0-3 



1493 



231 



79 



238-9 



231-0 







For this purpose we must note that the conjoint effect of 

 the magnetization due to current 50, simply applied or re- 

 moved, and of the compeusating-coil, was 8 divisions, of 

 which half is due to each cause. The effect of the coil for a 



