1902.] on the Electric Conductivity of Iron and Nickel. 



The changes of resistance are determined as follows : — The nickel is 

 first thoroughly demagnetised by the process of reversals ; the reading 

 cc of the bridge is then taken as soon as the equilibrium becomes 

 steady. *The magnetising circuit is then closed, and after reversing 

 the current a number of times to ensure reaching a point on the curve 

 of ascending reversals, the new reading x is taken. Finally, the 

 magnetising current is broken, and the residual reading x determined. 

 Then Ax = x - x is the " step " on the bridge wire due to the in- 

 duced magnetisation, and Ax r = x' - % is the corresponding residual 

 step. In taking the readings, the sliding contact piece is moved along 

 the bridge until the point is reached at which no permanent deflection 

 of the galvanometer needle takes place when the bridge current is 

 made or broken. This was found to be the most satisfactory way of 

 eliminating the thermo-electric effects in the bridge wire. Before each 

 set of readings the specimen was reduced to a neutral state by the 

 process of demagnetisation, thus avoiding the errors due to slow 

 changes of temperature. 



The fractional change of resistance A^> in the nickel wire P corre- 

 sponding to the step Ax is calculated from the formula 



AP A + B 



where o- is the resistance per unit length of the bridge wire. A slight 

 correction is required for the resistance of the copper leads connecting 

 P to the bridge. 



A preliminary experiment was made to test the effect of the magnetic 

 field on the resistance of the copper comparison coil Q. The nickel 

 coil was disconnected from the bridge, an equal resistance of German- 

 silver being substituted. Only the copper coil Q remained in the 

 solenoid, and as no change of resistance was observed, it was concluded 

 that in copper the effect is negligible. 



The magnetisation curves were obtained by the ballistic method. 

 Seven pieces of the nickel wire, each 65 cm. long, were enclosed in a 

 thin glass tube, round the middle of which was wound a ballistic coil 

 of 200 turns of fine copper wire. The glass tube was placed in the 

 axis of the magnetising solenoid, and the coil connected to a ballistic 

 galvanometer provided with telescope and scale. The galvanometer 

 was standardised before and after each set of readings by means of a 

 standard solenoid and secondary coil. 



The true magnetic force H in the nickel is given exactly by H = 

 H' - NI where H' is the magnetic field, calculated from the known con- 

 stant of the solenoid and the strength of the magnetising current, and 

 N is the demagnetising factor for the bundle of wire. In both the 

 resistance and magnetic apparatus the value of N was less than 0*0005, 

 so that this term could be neglected. 



VOL. lxxi. d 



