740 
PROFESSOR F. G. BAILY ON THE HYSTERESIS OF 
The machine was run at a gradually increasing speed, and the current was kept 
perfectly constant. As has been pointed out, this ensures a practically perfect con¬ 
stancy in the value of B in the armature. 
At first with the steel armature, a distinct variation was found, the sign of the 
change depending on the induction ; but the effect was traced to errors in the truth 
of the armature. When these were eliminated the result of a large number of tests 
is to show that up to a speed of 70 revolutions per second there is no regular or 
definite change in the hysteresis when tested in the rotating magnet machine. 
All the experiments were performed by keeping the current constant and unbroken, 
and increasing the speed step by step, waiting at each reading until the reading had 
become steady. 
The tests of the steel are shown in Diagram 11. There are irregularities which 
could not be eliminated, and which frequently repeated themselves at the same 
points, but on the whole there is a clear indication of no sensible change. The 
readings are corrected for eudy currents. 
The values at the maximum, or near to it, tend to be irregular, as was seen also in 
the previous curves. This may well be expected, the metal being to some extent in 
a critical condition. 
The values at the highest speeds were rendered slightly uncertain owing to 
vibration, which shook the spot of light, but there is scarcely a definite change in one 
direction more than the other. 
The soft iron shows the same kind of curves, and the same kind of small variations, but 
it may be noted that the readings given in Tables XI. to XIV. are not so regular as 
in the case of hard steel, and repetition of the series does not reproduce the same 
irregularities. 
At the value B = 17,200, four series were taken at different times; two of them 
show slightly rising values, another falling values, while a fourth gives an approxi¬ 
mately constant value. An example of a rising curve is shown at B = 7,800, but on 
repetition this gave a practically horizontal line. 
On repeating the tests with a smaller armature, the same results appear. At the 
maximum value of the hysteresis when B = 10,500, the readings were very irregular, 
as was anticipated, since the iron was in a critical condition. The same effect is 
noticed to a lesser degree near the maximum, when B = 15,000 and 17,200. 
A large number of series have been given because the actual readings show 
irregularities which might prevent a generalization ; but after constant repetition it is 
seen that the variations obey no regular law, and are probably due partially to errors 
in the machine, though to some extent there is strong ground for concluding that 
they are due to actual variations in the value of the hysteresis. 
The method is more accurate than those which entail the accumulating of the 
waste of energy, since in them the amount measured is proportional to the speed, and 
the conditions are changed. The deflections here are cpiite unaffected by the speed, 
