14 



ART. 13. K. HOXOA, S. RHIMTZIT AXD S. IvüSAKABE 



coiituiii Lcnn.s, which can ikjI pnjpci-ly he rijusidcred as the chaugc oi" 

 elasticity, ff this be the case, tlie change of elasticity by magnetization 

 is only apparent. 



In conclusion, it inay l)e noted that the reci])rocal relations 

 between torsion and magnetism, as ionnd ])y a<,'tual ex])eTiments, Avill l)e 

 loiind t«) be nl" parainounl inipoi'tanee in llie tlieorv oi" magnetosti'ietion. 

 We may conveniently place the results of our ex])eriments in the fol- 

 lowing statements parallel Avith those of ])i-evioiis investigators : — 



MAGNETIZATION TO TWLST. 



TWIST TO MAGNETIZATION. 



((') The torsion of iron do- 

 creases in all niai;iietizin,u' 

 fields. 



(/>') The torsion of nickel in- 

 creases in weak fields. 



((; ) The torsion (jf nickel de- 

 creases in stroim fields. 



(a) The magnetization of iron 



decreases by twisting for nil 



magnetizing fields. 

 (/>) The magnetization of nickel 



increases by twisting in weak 



fields. 



((•) The magnetizatioji of niclad 

 decreases by twisting in 

 strong fields. 



A similar reciprocal relation would probably exist in the case of 

 cobalt. The actual verification of the relation will he undertaken in 

 the near future. 



We have to express our cordial thaidvs to Professors II. Nagaoka, 

 and A. Tanakadate for valuable suggestions in the carrying out of 

 the jiresent experiment. 



