38 Messrs. I\. Honda and T. Terada on the Change of 



rigidity, from the change of twist caused by magnetization, 

 provided relation (1) hoids. 



The validity of (3) depends upon relation (1), which is 

 usually employed in the theory of magnetostriction. It was 

 thought, therefore, that it would be interesting to test ex- 

 perimentally how far the relation actually holds. Accord- 

 ingly the present investigation was carried out in the following- 

 order : — 



I. Experiments on the change of elasticity : 



(a) Measurement of the change of elasticity by the 



elongation method. 

 (/>) Measurement of the modulus of elasticity, 

 (c) Comparison of the results with those of the flexure 



method. 



II. Experiments on the change of rigidity: 



(a) Measurements of the change of rigidity by the 



oscillation method. 



(b) Ditto by Barus's method. 



(c) Ditto by the ordinary method. 



The specimens tested in the present experiments were 

 Swedish iron, tungsten -steel, nickel, and nickel-steels of 

 different percentages. They were generally tested in the 

 form of wires. For the change of rigidity, the rod was also 

 used in the case of nickel and iron. 



§ 2. Experiments on the Change of Elasticity by 

 Magnetization. 



(a) Measurement of the Change of Elasticity by the 

 Elongation Method. 



Since the elongation due to the change of elasticity is a 

 small fraction of the total elongation, it was desirable to 

 devise some differential method. For this purpose, the elastic 

 elongation at no field was compensated by that of a non- 

 magnetic wire, and the differential elongation due to magnetic 

 field was observed. As shown in the annexed figure 1, the 

 magnetic wire to be tested and the compensating copper 

 wire were hung side by side in the vertical magnetizing 

 coil cc. To both ends of the wires were brazed brass rods bb, 

 whose elongation could be neglected. The upper rods were 

 hung separately on two horizontal brass beams a a' resting on 

 pairs of knife-edges, the distance between which could be 

 adjusted at will. To the lower rod attached to the- copper 

 wire, a light carriage m for a mirror system was clamped. 

 Two fine spiral springs s s of german-silver attached to the 

 sides of the carriage supported horizontally the axis of the 



