THE MAGNETIC PROPERTIES OP IRON. 
537 
E.M.F. have also been omitted from the diagrams as unnecessary. In the results, as 
shown by the derived curves, it should also be stated that no correction has been 
made for the induction through the coils which did not pass through the iron. 
The curves in hgs. 1-6 show the rise of current through the circuit of the electro¬ 
magnet with different impressed E.M.F.’s. The full line curves were obtained after 
the magnet had been several times magnetized in the same direction as that produced 
by the current, while the dotted line curves show the rise of current when the magnet 
had been previously magnetized in the opposite direction by a current of the same 
maximum strength. The iron circuit was closed during these experiments, and had 
a mean length of about 265 centims. It will bo observed from these figures that the 
time required to complete the magnetization is nearly inversely proportional to the 
E.M.F. applied to the poles of the magnet, and, also, that the shape of the curve 
changes considerably with change of impressed E.M.F. The time required for the 
current to reach different percentages of its maximum value is shown in figs. 7 and 8. 
The ordinates of these curves are proportional to the time required, and the abscisspe 
to the impressed E.M.F. It appears from these curves that for any particular 
percentage of the maximum current there is always a particular voltage which 
requires greatest time. The curves also indicate that above a certain voltage, higher 
than the percentage considei’ed, the time required to complete the magnetization will 
become nearly constant. Figs. 9 and 10 show the values of the coefficient of induc¬ 
tion for different values of the current flowino’ throuo’lj the maD-netizino' coil, and the 
eftect on this coeflficient of the degree of magnetization ultimately reached. The most 
noticeable feature of these curves is the gradual change in the value and position of 
the maximum. This depends largely on the amount of residual magnetism previously 
existing in the magnet, and consequently, on the maximum value of the current 
which had previously flowed through the circuit. The exact value of the maximum 
coefficient is somewhat difficult to obtain by this method when the current produces a 
reversal of magnetization, because of the extreme smallness of dCldt at the point of 
the curve where this maximum occurs. No particular importance is attached, there¬ 
fore, to the apparent maximum of maxima shown in fig. 10. That there is a 
maximum of maxima, as shown in fig. 10, is, however, possible when we consider the 
cause of the change in position and magnitude of the maximum value of L from the 
point of view of the magnetization curve. It is well-known that this curve begins to 
rise somewhat slowly, and has a point of zero curvature corresponding to a com¬ 
paratively small value of the magnetizing force. When the residual magnetism is 
great, and in the same direction as that which the current produces, the results show 
that the curve never attains to so great a steepness, and attains the greatest steep¬ 
ness later than if no residual magnetism existed. When the residual magnetism is of 
opposite kind from that produced by the current, the maximum steepness seems at 
first to increase with increase of residual magnetism, and afterwards to diminish; the 
position of the point of greatest steepness corresponding to greater and greater 
MDCCCXCIII.—A. 3 Z 
