Potts — Rowland? s New Method. Ill 



In place of measuring the current by an electro-dynamome- 

 ter, in some cases a small ammeter might be used, but as the 

 current enters as a square in the energy loss, it must be accu- 

 rately determined. 



Determination of the induction : For the determination of 

 the induction several different forms of the secondary coil were 

 tried. Coils of the same size as the internal diameter of the 

 coils A and C were first used. These were found unsuitable, 

 however, as a considerable current was induced in them, even 

 when they contained no iron, and their coefficient of self- 

 induction was not negligible. The best form was found to be 

 a coil of fifty turns, just large enough to contain the iron used. 

 No measurable current was induced in this, when it contained 

 no iron, and its self-induction was negligible. The e.m.f. 

 around the circuit when K 4 is closed will be E, where 



E, = ^X -00213 xR volts 



Now let 1ST be the number of turns in the coil I and S the 

 average cross section of the iron surrounding the coil L and we 

 have the induction per sq. cm. B 



r _ Vd"x -00213 x R „„. 



B ■«• — n — - — X 10 8 c. g. s. units 



Nx4-44Xr s 



where N is the number of complete periods of the current. 



Energy loss. — The energy loss due to hysteresis is ordinarily 

 expressed as a certain loss per c.c. of iron per cycle. Energy 

 loss=rR£. 



If C 1 is the current in arm 1, H the total added resistance 

 due to the hysteresis and Foucault currents, v the volume of 

 the iron used, we have the energy loss per cycle, 



c '. t T 



since t will equal T, the period of the current. 



Experiment. — The iron used was ordinary transformer iron. 

 The plates were L-shape, and could be fitted about the coil C 

 very nicely. In the first place the uniformity of B was tested 

 for different quantities of iron. A slot was cut in the center 

 of one side of an L plate, thus dividing the side into two parts 

 which were made as nearly equal as possible. A coil of wire 

 of one hundred turns was wound on each of these parts of the 

 iron. By noting the deflection produced by the two coils in 

 succession when the iron was placed in coil A, the relative 

 value of B close to the coil A and farther away could be tested. 

 By placing in the coil different numbers of the plates and 



