42 Messrs. K. Honda and T. Terada on the Change of 



was then completely demagnetized, and the zero position of 

 the scale tested. Another weight of 1 kilogram was added 

 and demagnetized, and the same processes were repeated. 

 In this way a set of observations corresponding to the different 

 weights was taken. 



The change o£ elasticity was next measured. We applied 

 an initial load to the pan, demagnetized the wire by reversals, 

 added and removed a weight of 1 kilogram, and adjusted the 

 compensation until the deflexion due to the additional weight 

 was as small as possible. The demagnetization was then 

 carefully effected. The small deflexion due to the additional 

 weight for no field was recorded. Then we applied a mag- 

 netizing field and observed the deflexion due to the addi- 

 tional weight. After the demagnetization the deflexion due 

 to the same load for no field was again tested. The difference 

 between the two deflexions for no field and for the magnetic 

 field gives the change of elasticity due to magnetization, 

 provided we know the total elongation caused by the addi- 

 tional weight of 1 kilogram. To avoid any accidental error 

 in observing these deflexions, the deflexions for several cycles 

 of loading and unloading were, in each case, recorded, and 

 their mean was taken. The initial weight was next increased, 

 and the same procedure repeated. The compensation was 

 always readjusted for each different initial weight* 



Our object was partly to test the validity of equation (3), 

 and since the change of elasticity was considerably affected 

 by tension, the initial and the additional tensions were so 

 chosen as to exactly correspond to those in the experiments 

 of the magnetic change of length under constant tension. 



In the calculation of the percentage change of elasticity 

 due to magnetization, the values of elasticity were taken 

 from the results of our , experiments specially undertaken for 

 this purpose. 



The specimens tested and their dimensions were as 

 follows : — 



Substances. Length. Diameter. 



Pure nickel 73*25 cm. 0*928 mm. 



Commercial nickel 71*90 0*930 



Swedish iron 71*87 0*904 



Tungsten steel 73*76 0*886 



28*74 per cent, nickel-steel... 73*80 0*964 



50*72 „ „ „ ... 73*60 0*880 



70*32 „ „ „ ... 73*25 0*892 



The experimental results are given in the following 

 pages. 



