118 Mr. H. Nagaoka on the Effects of Torsion and 



tudinal stress only. On this account the combined effect of 

 pull and torsion on the magnetism of nickel was still a matter 

 to be determined. 



In the following experiments 1 have examined these points, 

 and have found that the longitudinal stress produced a singular 

 effect. For weak stresses the change of magnetism came out 

 as was to be expected, but when the load exceeds a certain 

 limit this is no longer the case. The changes of magnetism 

 become gradually altered, and beyond a critical value of the 

 longitudinal stress one end of the nickel wire acquires the 

 two opposite kinds of magnetism during the torsion and 

 detorsion, notwithstanding the absolute constancy of the mag- 

 netizinp; force both in direction and magnitude. This critical 

 value of the load seems to vary with the strength of the mag- 

 netizing field, becoming greater as the field is increased. All 

 these points will be described in the following pages. 



I must here express my thanks to Dr. G. Gr. Knott for his 

 kind suggestions during the course of experiments. 



The intensity of magnetization was measured by a direct 

 magnetometric method. The magnetometer consisted of a 

 small mirror hung by a spider-thread 11 centim. long. This 

 was geometrically fixed in position on a wooden plank ac- 

 cording to Thomson's method of the hole slot, plane. Level- 

 ling was effected by three base-screws. In front of the mag- 

 netometer a lamp was placed, and the image of the slit was 

 reflected on a circular scale. Its radius was 1 metre, and it 

 was so placed that the magnetometer was just at its centre. 

 The wire to be examined was set vertically due east of the 

 magnetometer. The upper end of the wire was level with the 

 centre of the magnetometer-mirror. To each end of the wire 

 a short stout brass wire was brazed. The lower of these was 

 bent into a hook, so that a pan holding the weight could be 

 hung from it. The upper one was riveted to a strong brass 

 rod projecting from the middle of the side of a table, which 

 rested on stone piers. The nickel wire* was surrounded by 

 a magnetizing coil 45 centim. long. The resistance of the 

 coil was 19'6 ohms, and the strength of the field for a current 

 of one ampere was 138"4 C.G.S. units. The magnetizing 

 current was sent from 12 Daniell cells, and its strength was 

 adjusted by means of a liquid slide and measured by a tangent 

 galvanometer. 



The twisting-apparatus is shown in the subjoined cut. 



* This wire contained 1'7 per cent, of iron, besides small quantities of 

 carbon as impurities. 



