538 
MR. T. GRA.T ON THE MEASUREMENT OF 
magnetizing’ forces as the residual magnetism is increased. That there is a particular 
value of the residual magnetism, which gives greatest steepness of the I’eversal of the 
magnetization curve, is possible, but it is not here considered established. The inter¬ 
pretation of the results, with reference to this point, is somewhat complicated bj the 
fact that Foucault currents no doubt produce considerable shifting of the position of 
the point of maximum, as indicated by the apparent magnetizing current. These 
currents may also modify the value of the maximum coefficient. This point will be 
more readily understood in. connection with the effect of a secondary coil referred to 
below. 
In fig. 11, a set of curves are sliowii which illustrate the rise of current when the 
previous residual magnetism is in the same and in the opposite direction, together 
with a continuation of the curve showing the rate of diminution of current when 
the battery is suddenly cut out, the magnet circuit being left closed. The dotted- 
line curves show the rise and fall of the magnetic induction for the two cases, the 
curve of ffill of current being, of course, the same for both. It will be observed, 
from the total induction or magnetization curves here shown, that the residual 
magnetism has amounted to about 60 per cent, of the whole magnetization produced. 
These, and similar curves shown further on, also illustrate the points referred to 
in connection with figs. 9 and 10, as to the variation and great difference in the 
value of the coefficient of induction, as depending on previous magnetization. Fig. 12 
shows the current and derived curves for a complete cycle of magnetization. The 
full line shows the variation of the current after reversal of the impressed E.M.F., 
while the dotted line shows the corresponding variation of the coefficient of induction. 
The dot-dash line shows the induction cycle, and the area enclosed by this curve 
gives the dissipation of energy due to the combined effects of magnetic retentiveness 
and Foucault currents. 
The curves in figs. 13-16 are similar to those in figs. 11 and 12. The impressed 
E.M.F. was 23 volts for all three sets, but the iron circuit in the magnet contained 
half a centimetre of air-space for figs. 13 and 14, and two centimetres of air-space for 
figs. 15 and 16. Comparing the latter two sets of curves with those obtained when 
the iron circuit was closed, we find as one of the most noticeable features the evidence 
they afford of the great effect a small air-space has on the magnetic retentiveness of 
the magnet. Curves (1) and (2) of figs. 13 and 15 are now nearly the same, and the 
sharp rise in the coefficient of induction has disappeared. The maximum value of the 
coefficient of induction is now nearly the same, whether the current repeats or reverses 
previous magnetization, but, somewhat curiously, seems to be somewhat less for 
reversal than for repetition. The ftill of current curves show slower fall at first, due, 
apparently, to the less firm hold of the residual magnetism and consequent larger 
initial value of the coefficient of induction just after reversal. This effect would 
probably be still more marked, were the total induction not considerably diminished 
by the air-space. Figs. 14 and 16 show the complete cycle, and thus correspond to 
