jOO u. nagaoka 



\v:i.s tlieii gradually increased till il attained the value i)f 29 C.Cr. S 

 units, inagnetoinetcr readings being taken at intervals, 'i'lie residual 

 magnetism was deterniincd as the magnetizing cin-reut was gradually 

 reduced i'rom its maxiuiinn xaluc to zero. Tlic iKinnal curve of 

 magnetization tlius obtained is represented in Fig. \.(ii). After tins, 

 the wire was demagnetized by reversals, and twisti-d through an 

 angle of 40°, so as to give a twist of 1° per em. I nder these slightly 

 altered conditions, the experiment was repeated. I p to the strength 

 of field .'ô == i.'-i, things were sensibly the same as in ihe former case. 

 Jjut as M was raised to fj.^, a very sudden change occurred in the 

 mao-netization. The maii'netometer i-eading rose abruntiv from 8 to 

 97 ; and then as .'ô was still further increased to (j.l, the reading rose 

 to 144. i)Ut after a!^ attained the value of 9.1, the rate of increase of 

 induced magnetism per unit of the increase of the magnetizing field 

 suddenly diminished, almost immediately falling oft' to a value which 

 remained nearly constant for higher fields. I'luit is, the magnetiza- 

 tion ciu've at these higher fields becomes almost a straight line. Thus 

 it ap])ears that twisted nickel has an enormous dinerenlial suscepiibility 

 (d^/cls;))* within a certain quite limited range of field, and that for 

 fields outside the lower and higher limits the ditferential susceptibility 

 is comijaratively small, and at high fields ])ractically constant. Curve 

 Fig. I (h) represents the case just discussed. Xo peculiar character 

 as compared with the normal curve frtj is seen at a glance. The 

 " wendej)unkt " or point of maximum susceptibility occurs in a 

 smaller field for the twisted nickel than for the untwisted. The 

 A\alue of the maximum susceptibility is also greater for the twisted 

 nickel. lUil when once the "Wendepunkt" for the twisted nickel 

 is passed, the ditferential susceptibility rapidly diminishes in such a 



* Tliis term is suggested by Dr. 0. G. Knott. Susceptibility always moans tlic ratio of 

 the iiiagnetiz.ation to the magnetizing force that is producing it; and it is convenient, 

 especially in discussing the peculiarities of nickel, to have a simple but suggestive phrase 

 for " the rate of change of magnetizatiou. per unit increase of field." 



