Magnetic Induction in Iron and other Metals. 461 



By substitution as before in (15) we get as the harmonic 

 constituents : — 



D 3 e- 3 V! 8sin ^ H -35°-l); 



D 5e - 3 V| sin (10. H -54°-5). 



Using (18) we get 



D 1= ex 1-132; 

 D 3 =cx 0-204; 



T> 5 = cx 0-144. 

 Or, by identification of amplitudes : — 



ex ri32xe~ 3 ' 4 v 4=1235, 



Again, 



.-. c=173t>. 

 ex 0-204 Xe~ 3 'V| =203, 



Lastly, 



.-. c = 2220. 



-3-4. /6» 



cx 0-144x6 v r«s = 78, 



.-. c=1390. 



As in the case of the previous curve, the two values of a*, 

 viz. 4*0 and 3'4, agree fairly well ; of the values of c only 

 the first is practically identical with that given above, viz. 

 1736 with 1700. 



The results for these two curves, approximate though they are, 

 would appear to suggest the possibility of obtaining the value 

 of x and of the saturation intensity of magnetization for a 

 material by help of the results of harmonic analysis of the 

 curve (I, H) when H is varied cyclically. 



That all the values of c are of the same dimensions I con- 

 sider to be sufficiently satisfactory, in view of the approximate 

 nature of the harmonic analysis applied by me to Dr. E wing's 

 curves, and in consideration also of the severity of the test 

 which is applied when we extract from any curve by har- 

 monic analysis such high period harmonics as the second and 

 fourth. 



