The plan we adopt is to excite the electro magnet by the whole or a part
of the current which is to be measured. Since this current varies, the
power exciting the core of the electro magnet must also vary; and since
we require the core to have as nearly as possible a permanent magnetic
force, we are brought face to face with the question, whether an electro
magnet can be constructed that has a constant moment under varying
exciting currents. This question has been answered by the well known
experiments of Jacobi, Dub, Mueller, Weber, and others. To get an
absolutely constant magnetic moment, is not possible, but between
certain limits we can get a very near approximation to constancy.

[Illustration]

The relation between exciting power and magnetic moment is very
complicated, depending not only on the dimensions and shape of the core
and the manner of winding, but also on the chemical constitution of the
iron of the core. It is not possible, or at least it has hitherto not
been found possible, to embody all these various elements into an exact
mathematical formula, which would give the magnetic moment as a function
of the exciting current; but the above mentioned experiments have shown
that within certain limits, and in the neighborhood of the point of
saturation, the relation between the two is that of an arc to its
geometrical tangent. It will be seen that for large angles the arc
increases much slower than the tangent; that is, for strongly excited
cores, a very large increase of the exciting current will produce only
a slight increase of magnetic moment. If Mueller's formula were correct
for all currents, absolute saturation could only be reached with an
infinite current. Whether this be the case or not, it is certain that
the greater the exciting current the less will a variation in it affect