304 



K. HONDA : Oi^ THE MUTUAL INFLUENCE 



large increase of magnetization must have taken place in the above 

 experiment, because the demagnetizing force in the direction of 

 circularly magnetizing field does not exist. That this was not 

 the case is to be ascribed to the magnetic œolotropy of the nickel 

 tube. The tube was made of a plate 6. '5 mm. thick. It was well 

 annealed both before and after the hole was bored ; but the 

 seolotropy was, notwithstanding, still to be traced. This and the 

 geometrical shape of the tube are the causes of the difference 

 in longitudinal and circular magnetizations as shown in the 

 following table : 



TABLE XII. 



These numbers plotted against H are given in Fig. 13. 

 The longitudinal magnetization in the above table was measured 

 by the magnetometric method, that determined by the ballistic 

 method being almost coincident. It does not seem probable that 

 the aeolotropy can be entirely effaced simply by the process of 

 annealing. The result here obtained is confirmed by the ex- 



