236 Lord Rayleigh on the Behaviour of Iron and Steel 



carried out, the operation was probably advantageous as obli- 

 terating the influence of the previous history of the iron 

 core. 



The compensation was in the first place adjusted so that no 

 displacement could be detected, whether the resistance was 

 infinity or 2007 ohms*. This, of course, was in the position 

 of the reversing-key denoted (— • ). When the iron and the 

 compensating-coil acted in tSe same direction ( + ), the dis- 

 placement was 8 divisions. 



In Table I. the first column gives the total resistance of 

 the circuit in ohms, and the second gives the reciprocals of 

 the first, numbers proportional to the current or magnetizing 

 force. Repetitions of a cycle are shown on the same hori- 

 zontal line, for greater convenience of comparison. Thus the 

 first application of current +197 gave the reading 242 ; a 

 second application, after the cycle +197,0, —197,0, gave 

 241J. After two of these cycles had been completed, the 

 current +326 gave the reading 245. To the readings as 

 entered a small correction to infinitely small arcs has been 

 applied. The letters R, L in the first column indicate the 

 alternative positions of the battery reversing-key. It will 

 be seen that very nearly the same numbers are obtained 

 on repetition of a cycle, and that even the first application 

 of an increased force gives a normal result. 



The first question which suggests itself is the law connecting 

 the magnitude of a current with the alteration of magneti- 

 zation caused by its reversal. The quantities under conside- 

 ration are exhibited in Table II., where the first column gives 

 the current {%) and the second column the displacement (y) 

 due to reversal. The relation between % and y is well ex- 

 pressed by the formula 



^=-•0053^ + 1-072^, (1) 



of which the whole of the second member is shown in column 5, 

 and the two parts separately in columns 3 and 4. Column 6 

 gives the differences between the observed displacements and 

 those calculated from the formula ; they do not much exceed 

 the errors of observation. 



It will of course be borne in mind that the magnetization 

 exhibited here is additional to the part rendered latent by the 

 compensating-coil, and that the existence of the small linear 

 term may be attributed to a defective adjustment of that coil. 

 The calculated value of y for the step from infinite resistance 



* For greater delicacy, recourse was had to the " method of multipli- 

 cation/' assisted by a pendulum, as already described. 



