LAWS AND WARREN. — HYSTERESIS. 493 



Watt-balance. It may be well to mention that the use of high resist- 

 ances constructed in the usual way by double winding may lead to 

 serious errors in alternating current measurements on account of 

 the capacity effects. 



If the plug G be inserted so that a current flows in the apparatus^ 

 and no deflection is observed, then the factors of the upper and lower 

 dynamometers are equal ; if the key K is depressed, there will in gen- 

 eral be a deflection, which may be reduced to zero by adjusting r^, as- 

 a zero reading means that the average values of the products of the 

 currents in the two coils of the upper and lower dynamometers are the 

 same. It is known that the work in S bears the same ratio to the 

 work in R that the total resistance in circuit 7\ does to the total resis- 

 tance in circuit r^. If the current is alternating, the work in S is of 

 course divided between hysteresis and heating losses. No allowance 

 has been thought necessary for the work done in the suspended coils. 

 If Wg and JFg represent the work in the specimen and in R, then 



ri + 103x _P^(ri+l03) 





^2+103 ' 



J^ is given by the electrodynamometer DYN. 



To correct for the heating loss in S a direct current of any conve- 

 nient magnitude is sent through the circuit, and a second balance, r^, 

 obtained, 



P 

 1000' 



B = 1,079 ohms, r^ = 976 ohms ; so Ws= t-/^7^(''i ~ '"i^)- 



It will be seen that the Watt-balance as used in the present case is a 

 Wattmeter with an electrodynamic control. The advantage derived 

 from obtaining this control by shunting a portion of the main current 

 is that any manipulation of the circuits does not disturb the equilibrium 

 of the instrument, and that it renders it possible to use the instrument 

 in a null instead of a compensation method. 



As the hysteresis loss is a function of the maximum magnetization, 

 denoted by i?, it was necessary to provide some means by which the 

 maximum might be kept constant, and its value determined. The de- 

 vice sketched in Figure 3 was used for this purpose. In addition to 

 the magnetizing coil, the specimen was provided with a winding of 24 

 turns, which was connected in series with an adjustable resistance of 

 874 ohms during the tests, and a galvanometer of 126 ohms. Leads 

 were carried to the dynamo room, and terminated in brushes which 

 rested on the edge of an ebonite disk 11 inches in diameter, fastened 

 rigidly to the dynamo shaft ; the brushes were carried by a radial arm, 



