THE MAGNETIC PROPERTIES OF IROK 
539 
fig. 12, The sharp inllection in the current curve does not appear in these curves, 
, and there is evidence that, for a given impressed E.M.F., there will be a particular 
air-space, which will give the maximum time for the current to pass through zero, and 
reverse after the reversal of the impressed E.M.F. The curves of coefficient of induc¬ 
tion are now comparatively flat, and the curve of total induction is nowhere nearly so 
steep as it is for the closed magnetic circuit. There is an important difference also in 
the amount of energy dissipated in the cycle. Of course, a direct comparison of the 
amount of energy dissipated in the cycle cannot be made from these curves alone, but 
a separate set of experiments, illustrated in figs. 22-27, bring out the result that 
considerably less energy is dissipated on the open circuit for the same amount of total 
induction than is the case with the closed circuit. From the curve (fig. 26) it appears 
that the energy dissipated per cycle on closed magnetic circuit for total density of 
induction of 13,200 C.G.S. units is 92 X 10”^ ergs, while the actual experiment on open 
circuit only gives 74 X 10^ ergs, the air-space being only half a centimetre. Again, 
from the same curve, it appears that, for closed magnetic circuit and a total density 
of induction of 8610 C.G.S. units, the energy dissipated per cycle is 4 X 10® ergs, 
while for the actual experiment, with two centimetres of air-space in the circuit, it is 
-only 3X10® ergs per cycle. This is ec[uivalent to a saving of about two and a half foot- 
‘ pounds of energy per cycle. That the dissipation of energy due to magnetic retentive¬ 
ness becomes comparatively small when the distance between the poles is as much as 
half a centimetre, was proved by static experiments made for the purpose. These 
experiments showed that, for a total induction of 14,000 C.G.S. units per square centi¬ 
metre across the air-space, the dissipation, as measured by the variation of the field in 
the air-space, was only 590 X 10® ergs. The intensity of magnetization here cpioted 
was obtained with a,bout 130 volts on the terminals of the magnet, and approaches 
near to saturation of the cores. When the circuit is open, the dissipation of energy 
is largely due to Foucault currents in the iron, and, as wil] be seen from the values 
calculated and marked on the curves, it approaches towards beiiig proportional to the 
square of the induction through the iron. This is a result which we should expect so 
long as the law of the variation of the induction is not very different in the cases 
compared. Tire practical bearing of the result here arrived at on the question of open 
I versus closed circuit transformers is evident. 
The group of figs. 17-21 illustrate results obtained with the magnet used as a trans¬ 
former with equal numbers of turns on the primary and secondary coils. The coils 
were joined alternately so as to form two sets of four coils each, each pair of j^rimaries 
containing between them one secondary and vice versa. Fig. 17 is similar to fig. 12, 
except that the current is greater and the number of turns in the magnetizing coil 
less. The different curves in this set will be at once recognized. In fig. 18, the effect 
of closing the secondary coil through a total resistance, consisting of the coil and 
incandescent lamps, is shown. The curves of primary and secondary current are 
clearly marked in the figure, and can be better studied fi-om it than from a description. 
