216 On Magnetomotive Force. 



If we assume a = *6 from analogy to sound and electricity, 

 ratio = '020. 

 « is not likely to be greater than this ; it may be less, as the 

 case is like that of a tube with side resistances removed to a 

 certain extent. 



The conclusion is : The magnetic induction of a permanent 

 magnet may be supposed to be produced by a magnetomotive 

 force derived from permanent Amperian currents, acting 

 through the resistance of the steel. 



In the case of the steel examined, this resistance was 5^ of 

 that of space, if ot='6. If a. be less the resistance will be 

 less in proportion. 



Meyer (Wiedemann, Ann. xviii. p. 233) has determined 

 the magnetization-function k of hard steel (p. 245). He finds 

 generally values varying from 2 to 3 for small magnetizing 

 forces, but in some cases as much as 9 or even 12. These 

 correspond, according to our formula /j, = 4:7rk, to the follow- 

 ing values of jjl : — 



h . . 2 3 8 12 



At..' 25-1 37-7 100-6 150*8. 



According to our result fju would be 50 for the above steel. 

 The number is quite uncertain, as the application of the co- 

 efficient a is as yet hypothetical. But it shows that the 

 hypothesis is not in contradiction with known values. De- 

 terminations of all the quantities involved require instruments 

 of a more accurate kind than those I have hitherto employed. 

 These are being constructed for other purposes ; and I hope to 

 examine the matter further. 



The assumption of the existence of magnetomotive force 

 and resistance in permanent magnets appears to be the neces- 

 sary consequence of Faraday's comparison of the permanent 

 magnet with the voltaic battery immersed in water. It is 

 the simplest assumption by means of which the facts can be 

 represented. 



On the old theory the assumption of rigid magnetism might 

 be modified to suit the facts by assuming the magnetism to 

 be elastic instead of rigid. Suppose, then, that when the 

 magnet is cut up into spheres, or disks say, the demagnetizing 

 forces of Maxwell (ii. 57) act on the elastic magnetization. 

 There will be a temporary diminution. 



It will be understood that I regard these demagnetizing 

 forces as arising out of the fictitious quantities created by 

 analysis for the purpose of compensating the errors of the 

 original hypotheses of Poisson's theory. I therefore prefer 

 the method here indicated. This has the advantage of re- 

 taining permanent elements in the permanent magnet ; and it 



