Relation between Magnetic Stress and Magnetic Deformation. 19 



"On the Relation between Magnetic Stress and Magnetic 

 Deformation in Nickel." By E. Taylor Jones, D.Sc. 

 Communicated by Professor Andrew Gray, F.R.S. Re- 

 ceived January 13, — Read February 25, 1897. 



(Abstract.) 



The experiments described in the paper constituted an attempt to 

 determine whether the change of length which a nickel wire experi- 

 ences when placed in a magnetic field can be explained by means of 

 the stresses which are known to exist in consequence of the magneti- 

 sation. 



The method of calculating the effect of magnetic stresses on the 

 dimensions of a magnetised body was given by Kirchhoff and applied 

 by Cantone to the case of an ellipsoid of revolution of soft magnetic 

 material placed in a uniform longitudinal field. 



The expression given by Cantone for the elongation of an ellipsoid 

 of revolution contains terms representing the effect of Maxwell's 

 system of stresses, and also terms representing the effect of those 

 stresses which arise in consequence of the fact that magnetisation 

 depends upon strain. 



It is shown that the latter terms reduce, when the eccentricity of 



the ellipsoid is great, to -JH where cl is the increase of mag- 

 netisation of a cylindrical portion of the material caused by an 

 increase of tension tV per unit area, and H is the magnetising field. 

 It is also shown by an independent method, due to Professor J. J. 



Si 



Thomson, that the term \ H — represents the elongation of a long 



cylinder of the material, supposed uniformly magnetised, due to the 

 stresses which exist in consequence of the dependence of magnetisa- 

 tion on strain. Hence, since this term is generally numerically 

 great in comparison with the terms representing the effect of Max- 

 well's stresses, a very long cylinder of the material could be used for 

 the measurement of both the magnetic elongation and the effect of 

 tension on magnetisation. 



The specimen used in the experiments was a long wire of annealed 

 nickel. The magnetisation and the effect of increase of tension on 

 magnetisation were first measured for a series of field-strengths and 

 for two different tensions, these measurements being made by the 

 ballistic method. 



Then the magnetic elongation was measured, in independent ex- 

 periments, for a series of field-strengths and the same two tensions, 

 it being arranged that the nickel wire was in the same magnetic 



c 2 



