2 ART. 13.-K. HONDA, S. SUIMIZU AND S. KUSAKABE: 



in opposite directions, the position of the mirror remained unchanged. 

 A magnetizing coil was placed co-axially witli tlie upper wire. If 

 the rigidity of the wire were changed by magnetization, the mirror 

 would rotate in either directi(jn. Barus found that for soft iron and 

 steel, the change of rigidity was 0.2-1% and 0.08% respectively. In 

 his later exixiriment, he observed a change of rigidity amounting to 

 \% for soft iron. With a similar arrangement, H. D. Day'^ in- 

 vestigated the same subject in iron. He found that the change of 

 rigidity increased with the field, and that it became generally less as 

 tlie initial twist was increased. The maximum value obtained was 

 only 0.8%. 



In the experiments of liarus and Day, the tensile stress, which 

 was found to produce a n(3table effect, would complicate the change to 

 be sought for. Moreover, the lower wire was not perfectly free from 

 magnetization, and the mirror would not give the perfect differential 

 effect. 



The experiment of J. S. Stevens-^ for iron and steel rods gave 

 an increase of rigidity by magnetization. The change amounted to 

 'l.''\% for soft imn and 0.4o% for steel in a field of lo8 C.G.S. 

 units. It also increased with the magnetizing force. In his ex- 

 |)erimenl, the length of the magnetizing coil was much less than that 

 of the rod. so that tlie magnetization was far from being uniform. 



'1. Our meth(jd of twisting the ferromagnetic rod was the same 

 as that of Professor Nngaoka"'* used for studying the elastic constants 

 of rocks ; but the sensibility of the apparatus was lOG.O times greater 

 for the same scale-distance. 



The front and side views of the apparatus are given in the an- 

 nexed figures. 



1) Day, Elcctrioiau, 39. 180, 18'J7. 



2) J. S. bicvoas, Pliy. Kfv. (3) IQ, lui, lÜÜO. 



3) riiil. Ma-r. 50,53, 1000. 



