IRON AND STEEL IN A ROTATING MAGNETIC FIELD. 
741 
and depend only on the value of the hysteresis at that moment. Hence the small 
evanescent variations may be readily detected, whereas by the cumulative methods 
they would be merged in the general average. 
There are a few points concerning the effect of time and sudden changes which are 
interesting. As has been mentioned, the curves were taken with steadily increasing 
current or speed, a short time being allowed before a reading was taken. 
When, however, a sudden increase is made in the current, the value of the 
hysteresis is higher than the normal for that current and the value slowly decreases 
until it reaches the normal steady condition. The effect is not large, about 3 or 4 
per cent-., but is by no means regular. The time of its disappearance is usually a few 
seconds and the effect is greater at medium and high induction than at low induction. 
The reverse effect is shown in suddenly decreasing the current, the hysteresis having 
too low a value, and not rising up to its normal for upwards of a minute, the arrival 
at the normal being slower than with a rising current. 
A similar effect is observed in changes of speed. A rapid rise in speed gives an 
increased value of the hysteresis, which dies away in a few seconds. But a rapid 
reduction in speed gives a diminished value of the hysteresis which lasts often more 
than a minute, during which time the deflexion is very irregular, jumping up to the 
normal and going back again. The change is also greater than auy of the preceding 
ones, and is frequently as much as 10-15 per cent. 
The effect of rapid makes and breaks in the magnetising current in quick succession 
is to increase the hysteresis. It recovers its normal value after a time but the readings 
are very irregular during the recovery. In all these effects, the machine is kept con¬ 
stantly running, and they occur more or less at all speeds. They are, however, not 
regular enough in amount to allow of any connection with the speed being established. 
Effects somewhat similar to these have been noticed with alternating currents. 
Tomlinson mentions (‘Proc. Roy. Soc.,’ Dec. 1889), that repetition of the cycle reduces 
the hysteresis, and he names it the “ accommodation ” of the molecules, supposing 
that the molecules become arranged in a manner in which it is more easy for them to 
reverse, and it is not surprising that with the more uniform motion of a rotating 
field, the same effect should be observed. The effect of suddenly diminishing the 
speed is more obscure; but in these experiments it seems to be due to a disarrange¬ 
ment and an absence of recombination for a short time, giving rise to great irregu¬ 
larity of hysteresis. The length of time over which the effect lasts is remarkable, as 
it often extended to over 1,000 cycles. It was completely stopped by a temporary 
stoppage of the current, while Tomlinson found that even slight mechanical shocks 
were sufficient to stop the effect of t£ accommodation.” 
The effect of reducing the current is probably due to the same cause. The molecules 
are arranged in regular order, and it is some time before they disarrange themselves, 
while during the process the hysteresis is lower than the normal. An increase in the 
current is not immediately followed by complete arrangement of the molecules, and 
