334 PROFESSOR J. A. EWING ON THE MAGNETIC QUALITIES OF NICKEL. 



influence of compressive stress, for it is easy to fix the lower end of the bar in the 

 yoke and apply weights directly, or by a lever, to the upper end. The arrangement 

 adopted in the present experiments is shown in fig. 11.* The sample under test was a 



Tig. 11. 



Arrangement for testing magnetisation of nickel under compression. 



bar of nickel supplied by Messrs. JOHNSON and MATTHEY (which was found on analysis 

 to contain 075 per cent, of iron). It was 10 cms. long, and was turned to a diameter 

 of 0'656 cms. The yoke was of soft wrought iron, with a cross-section on either side 

 of 67 square cms. The lower end of the bar was supported in the yoke by resting on 

 the end of a screw-bolt ; on the upper end a short plunger of wrought iron pressed, 

 and through this the desired stress of compression was applied by means of a lever 

 (fig. 11). The clear length of the sample, within the yoke, was 5 cms. Over this there 

 was wound a magnetising solenoid of 250 turns, inside of which there was a small 

 induction coil wound close to the metal. The magnetisation was determined by 

 reversing the magnetising current while the induction coil was connected to a ballistic 

 galvanometer. To find the residual magnetism the magnetising current was broken 

 after reversal, and the effect of this break was subtracted from half the effect of the 

 reversal. In every case several reversals were made before a measurement was taken ; 

 and the process of demagnetising by reversals t was resorted to whenever it was 

 necessary to get rid of residual effects of previous magnetisation. 



In the first place, the nickel bar was examined in the rather hard rolled or drawn 

 state in which it was supplied, by applying magnetising forces which were raised, step 



the bar was cut at the middle of its length, to allow an induction coil to be slipped out. This must have 

 had the effect of making his measurements of susceptibility and retentiveness somewhat lower than they 

 would have been had the bar been continuous. In the present experiments there was no joint in the bar 

 itself, but there were, of course, joints between the ends of the bar and the yoke. These must have had 

 some influence of the same kind, though less in amount, from the fact that the bar's ends were sunk a 

 good way into the yoke, to give a large surface of contact. 



* The figures are numbered consecutively with those in the former paper (' Phil. Trans.,' 1888, A., 

 p. 325). 



t ' Phil. Trans.,' 1885, p. 539. 



