193 



XI. — On Superposed Magnetisms in Iron and Nickel. By Professor C. G. 



Knott, D.Sc. (Plate XXIX.) 



(Bead 2nd July 1883.) 



The experiments which form the subject of this paper are designed, in the 

 first place, to test the relation pointed out by Maxwell* between Joule's 

 discovery of the lengthening of iron in the direction of magnetisation, t and 

 Wiedemann's later researches into the twisting of iron under the influence of 

 longitudinal and circular magnetisations,! and, in the second place, to investi- 

 gate the corresponding properties of nickel. 



According to Joule's discovery, an iron bar or wire lengthens in the direc- 

 tion of magnetisation, and contracts in directions at right angles thereto. 

 The extension is greater for a stronger magnetising force, and, if the metal is 

 subjected to traction in the direction of lengthening, is smaller for a greater 

 traction. In the experiments to be described a wire was fixed at its upper 

 end, and stretched vertically by means of an appended mass. It passed cen- 

 trally through a glass tube of nearly the same length, round which a helix of 

 wire was wound. The length of the helix was 343 centimetres, and the total 

 number of coils 196. A current passed through the helix magnetised the wire 

 longitudinally. At the lower end of the wire was fixed a short copper wire, 

 which dipped into a pool of mercury. By this means a current could be passed 

 along the wire so as to magnetise it circularly. The twist produced under the 

 joint influence of the longitudinal and circular magnetisations was measured by 

 the deflection of a spot of light focussed upon a millimetre scale after reflection 

 from a mirror attached to the lower end of the wire. Both the magnetising 

 currents were measured on a Helmholtz tangent galvanometer. 



The method of experimenting was as follows : — One of the currents was 

 kept steady, while the other was varied through a considerable range. When 

 both currents were flowing the free end of the wire came to rest in a definite 

 position, which was registered by the reading en the scale. One of the currents 

 was then reversed, and a second reading obtained. The difference between 

 these readings was approximately four times the angle of twist. By successive 



* See Maxwell's Electricity and Magnetism (2nd edition, vol. ii. § 448). The first edition comesto 

 a wrong conclusion, in consequence of a misprint in Wiedemann's Galoanismus (1st edition, Bd. ii. § 491). 

 See also Cheystal's article on "Magnetism" in the Encyclopcedia Britannica (vol. xv. pp. 269, 271). 



+ Sturgeon's Annals of Electricity, vol. viii. p. 219 ; and Phil. Mag., 1847. 



\ Wiedemann's Galoanismus, 1st edition, Bd. ii. § 491. 



VOL. XXXII. PART I. 2 I 



