514 MR JAMES RUSSELL ON THE SUPERPOSITION OF MECHANICAL VTBRATIONS 



with (Proc. R.S.F., vol. xxvi. p. 33). The experimental methods being the same in 

 both cases facilitates comparison. Under both the A and B conditions the similarity 

 extends to details. 



It also appears evident that, had the experimental methods of the earlier investiga- 

 tions dealing with the effects of purely mechanical vibrations upon magnetisation been 

 such as to elucidate the various phenomena involved under the A and B conditions, 

 the results of later investigators (Ascoli, Arno, Walter and Ewing, Garibaldi, 

 Marconi, Maurain, Piola, Wilson, and others) relative to the effects of electric 

 oscillations upon magnetisation would have fallen more readily into line with each 

 other * and with the effects of purely mechanical vibrations. 



The effects, therefore, of electric oscillations upon magnetisation are essentially 

 the same as those produced by mechanical vibrations. 



Conclusion. 



The experimental results given in this paper, summarised on pages 509 to 510, and 

 discussed in their relation to magnetic hysteresis, molecular theory, and electric 

 oscillations in the pages which follow, answer more or less completely the questions 

 propounded when the effects of mechanical vibrations upon magnetisation have not 

 reached a limiting value. 



The experiments, however, were made, as has been stated, with wires subjected to a 

 small load. A few of these have been repeated without load ; wires of annealed iron 

 and nickel were soldered to the gong, and supported in the glass tube by means of loose 

 pads of cotton wool. The general relation of the curves to each other, as shown in 

 figs. [II. and V., were not found to be materially altered, although in the case of nickel 

 the increase of the induction and residual magnetisation curves is considerable owing to 

 the absence of load (see table, p. 494). 



The results, therefore, in so far as common to the three magnetic metals examined 

 in an annealed and quenched condition, may be concisely stated, without reference 

 to load, as follows : — 



(L) In all cases vibrations increase permeability (induction), but in high fields the 

 ratio B v /B approximates to unity. 



(2) The effect of vibrations superposed at all stages of the normal loop (A conditions) 

 is, generally speaking, to lessen those differences of magnetisation to which hysteresis 

 without vibrations has already given rise. 



(3) When change of field (cyclic or increasing from zero) is superposed upon 

 permanently acting vibrations (B conditions), the differential permeability is increased 

 in low fields, diminished in high fields. In sufficiently high fields vibrations must 

 delay demagnetisation. 



(4) The effects of electrical oscillations upon magnetisation supported by an 



* " Notes on the Effect of Electric Oscillations on Magnetism,'' L. H. Waltkr, Electrician, May 5, 1905. 



