29(1 II. XAfiAOKA 



cliango sign by twisting, ft first oociirn^d tn mo tluii iliis \crv 

 oxtriiovdinarv result must 1x' dun id some defect in tlie aiT;ingeiueiit. 

 but (;aiT'lul exnminatiou disc<)\'erod uo fl;i\v. The <|iiesii(in n:irm-;ill\- 

 siijro-esteil itself, was this pheuDmennu a fnucriou «if llie Icmd as well 

 as of the Iwisr? lleiiee as a iH'Xt step, the weiü'lit was increased li\- 



•5 ' 



IlKi 



the e\']ieriuictU Was ]iei-fnrnied in the usual inaiiuer 



'Idle result is shewn in Fig. '2. lierez the ett'ect is <|uite siuiilar 

 to the former, the differences being i>nl\- dilferenees of detail. The 

 mtu'cli of magnetization witli positive twisting is sensilily the same 

 as in the farmer case ; bui (bu'ing negative twisting, the o[)|»osiie 

 maü'iieti/.ation has increased to more than 10 times its amount in tlie 

 iirst ex|>erimeiit. Howexer. the rate of recovery has very much 

 diminished, tuid even after a twist of I8tl°. the magnetization is far 

 from reaching the former value. 



The same experiment was then performe(l with I he load incre;ised 

 to .">.14 kgs. The residt is shown in Fig. .'i. There again we tind 

 the opposite mati'uetization still more increased. Indeetl. the two 

 kinds of inaoTietization do not diifer nuich in intensity at the two 

 exti'pme twists, although the initial or ])ositive magnetization is some- 

 what preilominant. In the two former experiments, there was a 

 (hstinct ten<lencv towards rcco\ ery of positive magnetization as the 

 \\ii-e was twiste<l mon' and more in the negati\c din^ciion. Here, 

 lenvever. no such tcndcncv shows itself except in the diminished rale 

 of growth of negatixc magnetization. 



Still increasing the load, we see that the curve (Fig. -i) becomes 

 nearlv sNinuictrical with respect both to i he line cif/cro magnetization, 

 and the line of zero twisting. In this case we find further (hat tlie 

 rani;'<' of flu' change of magnetization has considerably diminished. 

 Tdie slight excess of the initial magnetizatioji over the other still shows 

 itself, a fact which is probably to be referred to the direi'tion of the 

 matiMK'tizing foi-ce. 



