718 
PROFESSOR F. G, BAILY OR THE HYSTERESIS OF 
by the spring becomes equal to the torque due to the hysteresis and eddy currents 
in the armature. The rate at which energy is dissipated in the armature is then 
easily measured, for 
Power absorbed = torque X 2 77 - speed of rotation, or 
Energy absorbed per revolution = 27- X torque. 
Since the restoring force of the spring was found to be proportional to the angle 
of rotation, the hysteresis and eddy current loss per revolution is proportional to the 
deflexions of the armature as shown by the reflected beam of light. The reflected 
beam of light was focussed on to a semicircular translucent scale concentric with, and 
perpendicular to, the axis of revolution of the electro-magnet. The readings of the 
movement of the spot of light are, therefore, proportional to the energy absorbed per 
revolution in the armature. 
The springs were calibrated by measuring the deflexion produced by a small 
weight at the end of a light steel arm attached to the armature. They were 
calibrated frequently and were found to change only very slightly, even after 
considerable damage by being wound up, as once or twice happened. The hysteresis 
per revolution is thus measured directly in terms of the force of gravity acting 
through a known space. 
If Z is the length of the arm, 
m is the mass of the weight and equivalent weight of the arm, 
v is the volume of the iron, 
d is the deflexion caused by the weight. 
Then a deflexion of 1 division against the spring = -7i "’ j ~ -ergs per cub. centim. 
per revolution of energy loss in the armature. 
Determination of Eddy Currents. 
The only correction of any importance that is required is the loss due to eddy 
currents. In most tests on hysteresis at a high speed of alternation the eddy 
current loss has introduced an element of some uncertainty, and therefore great 
attention was paid to this point. There are two ways in which errors may be 
caused. Eddy currents will be set up in the armature plates by the movement 
of the lines of force through them, which will produce a torque on the armature in 
the same direction as that due to hysteresis. There may also be eddy currents set 
up in the framework of the apparatus which will modify the magnetic field through 
the armature, so that it will have a value slightly differing from that produced by 
the magnetising current alone when the magnet is standing still. This, however, 
may be shown to be very small. In the rotating parts there will be no eddy 
currents, since the direction of the lines of force will travel round with the electro¬ 
magnet. In the fixed spindles there will be currents set up due to the leakage 
