6 ART. 12.-K. HONDA, S. SHIMIZU, AND S. KUS\KABE : 



Here H denotes the external field and T the tension per square 

 millimeters. The calculation of —rr from Steven's results for a piano- 

 forte wire gives 1.9x10"- in a field of 40 C.G.S. units, which 

 approximately agrees with the results for soft iron. 



From these tables, we see that in iron and nickel steel, the 

 magnetization considerably increases the modulus of elasticity, the 

 amount of the chang^e for a jriven load increasino; with the ma^rnetizinfr 

 field. The increase also varies witli tension ; it decreases as the tension 

 is increased. Wolfram steel shows a small increase of elasticity ; 

 under a constant field, the increase reaches a maximum as the tension 

 is increased. In nickel, the elasticity decreases in the weak fields and 

 increases in the strong. The change is also a function of the 

 tension ; in weak fields, the diminution reaches a maximum and 

 then frraduaJlv decreases as the tension is increased. In stronsf fields, 

 the increase becomes less and less and at last changes its sign with 

 the increase of tension. The field in which the change of elasticity 

 vanishes becomes greater as the tension is increased. 



The third is the method of flexure. The advantao;e of this 

 method lies in the fact that the dift'erential effect can be measured, by 

 suspending a weight at the middle of a ferromagnetic bar and then 

 measuring the change of depression caused by magnetizing it. Such 

 a bar elongates or contracts by magnetizition, while its thickness 

 diminishes or increases ; but as we shall soon indicate, the lateral 

 elongation or contraction will be very small compared with the 

 change of depression due to that of elasticity. 



Hence, uf these three methods, that of flexure is the most suitable 

 for studying the change of elasticity. We therefore used this method 

 to investisrate the said effect and also t(3 test the results of the 

 elongation method just referred to. 



