MA'JXETTZATTOX OF XTCKEL. 



289 



'Pho nhovo ivadiiiî^s rodiK^od to ahsi')lntn units arc sliowii plotted 

 in Fi.ü'. 1- I Ik' aniomit of load per s<|. (mi. was S2 k.u's.. and the 

 twist of ]R()° corresjionded t<i that of .DZ.Sö radian per cjii. 



Kxaniiniiii;- the figure we see that the first effect of twisting is to 

 increase the niagnetizati<Mi. Tiie rate of increase is rapid at first, but 

 graduallv falls off as the magnetization attains its greatest value at 

 the niaximuüi twist. During the process of untwisting, the magne- 

 tization diminishes more rapidly than it increased during twisting, so 

 that for every position of twist, the magnetization during twisting is 

 greater than the magnetization during untwisting. The diminution 

 of magnetization goes on even after the wire has passed the original 

 position from which the twisting was begun, until the apparent mag- 

 netization is at length reduced to zero. This happens very soon after 

 the original position of the untwisted wire has been reached, as the 

 process of untwisting is continued as twisting in the negative direction. 

 But now as this negative twisting is continued, the polarity of the wire 

 chaufn's sifin. a verv striking fact indeed. As the twisting is continued 

 on towards -180°, this negative magnetization passes through an 

 arithmetical maximum, becoming finally almost zero. As the wire 

 is lieing brought back to the position from whidt it started, the mag- 

 netization liraduallv recovers nearlv its oriü'inal value as sliown in the 

 figui-e. 



Xow i-easoning from analogy, we should ex])ect to obtain Ity such 

 twisting and untwisting a curve of the form given in Fig. ."), since 

 the behaviour of nickel with rcfi-ard to the effect of stress in maij-ne- 

 tization seems to be just opposite to that of iron (see Sir. AV. Thomson's 

 figures for iron, I'liilosophicaJ Transaction 1S79). lîut in this experi- 

 ment, the cia-\(' of magnetization seems to be no resemblance at all to 

 any figured liy Sir. AV. Thomson ; and then there is the very curious 

 fact that tlic magnetization of nickel in a steady field can be made to 



