CONTEMPORARY ADVANCES IN PHYSICS 325 



short after the first reduction in magnetizing field following the 

 attainment of the maximum applied field (the one which I above 

 called Hq), returns to Hq, and alternates the field several times between 

 Hq and the inferior value Ho — AH, he finds that the magnetization 

 settles down to a routine ot alternating between definite values /o and 

 /o — A/. The limiting value of the quotient Al/AH, for small values 

 of AH, is known as reversible susceptibility. It is a function of /; 

 that is to say, if we select any particular value of I we always get one 

 and the same value of Al/AH when we impart that value of I to the 

 metal, whether by mounting to it along the initial curve or the "ideal" 

 curve or coming to it along any hysteresis-loop. 



If the hysteresis-loop is described very rapidly and continuously, it 

 retains its shape surprisingly closely until the frequency is raised into 

 the hundreds of thousands. The initial permeability is still more 

 nearly unafi^ected by rapidity of variation of field, remaining sensibly 

 unchanged until the range of radiofrequencies is reached and passed. 

 In the range of light-frequencies, however, it is reduced to unity.* 



Magnetostriction 



"Magnetostriction" is the clumsy name given to the divers very 

 inconspicuous strains in a magnetizable body, brought about by the 

 process of magnetizing it. As they are exceedingly small — a variation 

 of any linear dimension amounting to four parts in one hundred 

 thousand would be ranked as a remarkably big one — and as magnetiz- 

 able materials are usually investigated in the form of long thin rods, 

 the change in the length of such a rod resulting from a magnetic field 

 applied parallel to its axis ("Joule effect") is the only magnetostrictive 

 change which is often mentioned. Changes in the dimension normal 

 to the field do, however, take place; a rod which expands lengthwise 

 in a longitudinal field will contract sidewise, and vice versa. It used 

 to be thought that the change in length just compensates the change 

 in thickness, so that the net change in volume would turn out to be 

 nil; but this turned out too simple to be true. A wire exposed to a 

 longitudinal field and traversed by an electric current will twist itself 

 (the "Wiedemann effect"). This occurs because the impressed field 

 and the circular field due to the current itself are compounded with 

 one another into a resultant pointing slantwise to the axis, so that 

 any particular "line of force" can be visualized as winding in a helix 



* In certain materials there is said to be a "magnetic viscosity," because of which 

 the magnetization continues to vary for an appreciable time after an alteration in 

 field is made and ended. The observations upon this are much confused by eddy- 

 currents, and the question is still under debate. 



