234 Mr. J. M. Baldwin on the Behaviour of 



From the total iron loss I, the sum of the static hysteresis 

 and the calculated eddy-current loss was subtracted, and thus 

 the kinetic hysteresis was obtained. This kinetic hysteresis 

 increased with the induction and with the frequency. 



3. In the present series of experiments, in which the mag- 

 netic intensity ranged from H=l*5 to H = 001 and B from 

 600 to 2*5, the hysteresis areas could have been determined 

 only for the higher inductions by the statical methods, but 

 the wave-tracer of Professor Lyle * affords a convenient 

 method of obtaining the total iron loss, when the magnetic 

 field is produced by an alternating current of any desired 

 frequency. The portion of this loss of energy due to eddy- 

 currents set up in the iron can be approximately calculated, 

 and subtracting this from the total iron loss, the loss due to 

 hysteresis is obtained. 



Two specimens were experimented upon, in one of which 

 (a bundle of iron wires) the eddy-current loss was compara- 

 tively small, while in the other (a rod of Lowmoor iron 

 0'3 sq. cm. in cross section) the eddy-current loss was con- 

 siderable. For the former specimen it was found that for 

 very low inductions the lag of phase of the induction behind 

 the magnetizing force was very small, and also that the 

 permeability was practically uninfluenced by the frequency 

 of the alternations. In the latter, where the eddy-currents 

 are considerable, the resultant intensity inside the rod will 

 no longer be in phase with the current in the solenoid, and 

 so the resultant induction will always lag behind the mag- 

 netizing current. 



Assuming that the induction at any point is in phase with 

 the magnetic intensity at that point, as the experiments with 

 the wire bundle show, and also that the permeability has a 

 constant value for low inductions, equations can be obtained 

 connecting the apparent permeability and the lag in phase 

 of the induction behind the magnetizing current with the 

 frequency. 



Let OF (see figure) represent the resultant flux across the 

 central section of the rod. The variation of the flux will 

 produce an e.m.f. in a circuit round the rod in quadrature 

 with OF and behind it in phase. This e.m.f.. produces a 

 current, which in turn produces a flux FM or IF (where 

 IF = FM) along the rod. This flux will also, on the assump- 

 tions which have been made, be at right angles to OF. Then 

 01 is the flux which would have passed along the rod if there 

 had been no eddy-currents. 



* T. K. Lyle, Phi]. Mag. [G] vol. vi. p. 549 (1903). 



