234 
DR. J. HOPKINSON AND MR. E. WILSON ON 
In fig. 5 the speed of the alternator experimented upon was 923 revolutions per 
minute, or a frequency of 92‘3 complete periods per second. For the purpose of 
obtaining a marked effect from the current in the armature a large current was taken 
out of the armature, and the current in the magnet winding was only 8 amperes. A 
Kelvin multicellular voltmeter placed across the terminals of the machine read 190 
volts on open circuit, and 81 volts when loaded, and a Kelvin ampere balance in the 
external non-inductive circuit read 40 amperes. The armature bobbins were coupled 
6 in series 2 parallel between the brushes, the total resistance of the circuit (R) was 
2‘52 ohms, the armature resistance alone being '55 ohm. 
Fig. 5. 
Curve E is the electromotive force curve of the machine when there is no current 
in the armature. Bx is the curve of electromotive force deduced from the potential 
difference taken between the terminals of the alternator when supplying current 
through non-inductive resistances. The curves E and B.x have been integrated, 
the corresponding integral curves being A and B respectively. 
That the ordinary theory does not fully account for the facts is easily shown. 
We have Ike = E — (Lir)\ Integrate both sides from any fixed epoch 0 to any time 
t and we have 
f R.r clt = f E dt — L (x, — x 0 ). 
JO • Q 
Each term of this equation consists of a constant part and of a periodic part. The 
constant parts must be equal and also the periodic parts. We have to deal only with 
the periodic parts. The curves A and B, fig. 5, represent the periodic parts of each 
