128 Mr. H. Nagaoka on the Effects of Torsion and 



It is very difficult to determine the exact loading at which 

 the reversed phenomenon of reverse polarity makes its 

 appearance in various fields. Assuming, however, that with- 

 in small ranges of load, the decrease in the intensities of 

 magnetization is proportional to the amount of loading, I 

 obtained the following values for the stress which must be 

 applied so as to effect the reversal of polarity of the particular 

 nickel wire placed in various fields : — 

 For ^^ -34, W = 

 $= 2-47, W = 





4-94, 



6-71, 



8-06, 



11-92. 



W = 

 W = 



w= 



77 kilogs. 

 217 „ 

 421 „ 

 654 „ 

 783 „ 

 1120 „ 



centim.^ 



^=13-85, W=1220 „ „ 



^ = 15-78, W = 1530 „ „ 



.0 = 23-50, W = 2110 „ „ 



.i^ = 33-54, W = 2830 „ „ 

 Plotting the curve, with -5 for abscissa and W for ordinate, 

 we get the annexed figure : — 



W 



1500 















y 











X 















y 











/ 













/ 













/ 















S5 ^ 



This seems to show that, for moderate strengths of field, the 

 load at which the wire begins to show reversed polarity is 

 nearly directly proportional to the strength of the field. For 

 very weak and strong fields this rule does not seem to hold. 

 It must be remembered, of course, that all these peculiarities 

 are for one particular twist only, and that it is possible that 

 quite a difi"erent series of effects might exist for other twists. 



