342 K. Honda and S. Shimizu on Change of Length of 



By making use of fig. 4 the curves showing the relation 

 between the change of length and the tension under constant 



Fiar. 4. 



— 































2 





































-4 



£ 



^ 



^\ 



s 2( 



\Q 



3( 



)0 



4( 



)0 



5(10 



6(10 



7(10 



^ 



^ 



























H 





-^ 







<** 



?^ 



s?. 



^ 

 3^- 

















-e, 













^"^ 





















-8 































































field are obtained ; some of them are given in fig. 5. We 

 learn from these curves that the effect of loading on the 

 magnetic change of length is not proportional to the load. 



Fiar. 5. 



f- 



2 



xio" 6 















 2 



\, 



2( 



100 



M 



100 



T 







** 









A- 









/ Oo 













^? 



£go 







Generally speaking these results coincide with those of 

 Bidwell. In our case the reduction of the initial elongation 

 by tension is far greater than that with BidwelFs wire. The 

 smallest tension at which the elongation vanishes is about 

 four times greater in the latter case than in the former. 

 The discrepancy perhaps arises from the fact that our specimen 

 is comparatively soft as regards magnetization. 



5. Wolfram Steel. — Fig. 6 represents the result for Wolf- 

 ram steel hardened by stretching ; the anomaly in the 

 change of length for the steel was already pointed out by 

 Prof. Nagaoka and one of us. This anomaly gradually dis- 

 appears when the tension is increased, and at a tension of 



