126 Mr. E. A. Owen on Change of Resistance of Nickel 



comparison wire of thin platinum was mounted in a similar 

 way on another ebonite plate. 



The whole was fixed in a trough of mica just wide enough 

 to contain the specimens. The trough was held horizontally 

 in a wooden stand resting on a slab of plate glass which was 

 separated from the fixed table on which it stood, by a thick 

 layer of paraffin wax. The trough, and with it the thin wires, 

 could be withdrawn from between the poles of the magnet to 

 a distance of about 80 cm. by lowering the top part of the 

 stand and turning it in its socket through 180°. 



The whole of the apparatus including the cable, bridge- 

 table, galvanometer, specimen, and all connecting-wires were 

 thoroughly insulated from the earth. 



(2) Method of taking Readings. 



The specimen was demagnetized by reversals, and then 

 removed to a distance of about 80 cm. from the pole-pieces. 

 When in this position, that is, when the specimen was in zero 

 magnetic field, the balancing point was obtained on the bridge 

 wire with the rider Rj (see fig. 2). The specimen was then 

 turned into the field. Preliminary experiments had shown 

 that the magnetic change of resistance in both nickel and iron 

 reached a maximum value in a certain field, and the current 

 required to produce this field had been carefully determined. 

 This current was now passed through the magnet coils and 

 the balancing point obtained on the bridge with the rider R 2 . 

 The specimen was aaain demagnetized and the same opera- 

 tion repeated until about a dozen readings were taken. The 

 mean of these readings gave the maximum step on the bridge- 

 wire. If £c is the balancing point in zero field, and x that in 

 the field, then (x— #o) gives the maximum step. 



The specimen was then permanently placed between the 

 poles. Different currents were sent through the magnet 

 alternately with the current which gave the maximum step, 

 and the balancing points on the bridge wire found in each 

 case, — R 2 for the fixed current, and R x for the other currents. 

 The varying currents were not turned on regularly in 

 ascending or descending order of magnitude, but quite 

 irregularly. The mean of two consecutive readings with the 

 fixed current was taken as the balancing point for this 

 current. Let it be x. If a? ' is the balancing point for any 

 other current, then (#—#</) i s the amount to be subtracted 

 from the maximum step to give the step for that current. 



The riders were moved about by means of a long ebonite 

 rod, so as to avoid bringing the hands near the bridge-wire. 



Changes of resistance were examined for fields ranging 

 from 500 c.G.s. units to 20.000 c.g.s. units with the pole- 



