Theory of Magnetism. 241 
to the current-strength 7 multiplied by the number of windings, 
for which we may put the length of the ring 2p7 nearly. 
The resistance which opposes itself to this magnetizing 
_ force is directly proportional to the length of the bent iron 
rod (that is, again, 27rp), and inversely proportional to the 
cross-section and to the magnetic conductivity of the iron, 
which may be denoted by Wy. Consequently the magnetic 
moment of the iron ring in each of its cross-sections is 
12pm 
(ay 
which expression denotes the same as the previous one, 
=igvr const., 
M=ils const. 
To investigate the crucial question whether the magnetism 
which is produced by a magnetizing force in an iron bar or in 
an open horseshoe is also inversely proportional to the total 
resistance of the magnetic circuit, I had a horseshoe made out 
of iron bar 20 millim. thick, and bent at right angles. The 
legs of the horseshoe were 70 millim. long, and each was sur- 
rounded by a spiral 35 millim. long, consisting of from 126 to 
130 windings of insulated wire 1 millim. thick. The straight 
back of the horseshoe was provided with an induction-coil of 
1160 turns of wire, 0:2 millim. thick. A prismatic piece of 
iron, of the same cross-section as the horseshoe, could be used 
to make it into a closed metallic circuit. The legs of the 
magnet projected 20 millim. beyond the coils. 
The experiments were made by momentarily reversing the 
direction of the current through the magnetizing coils by 
means of a suitable commutator. The strength of the 
current before each reversal was measured by determining 
the difference in potential between the terminal screws of 
the magnetizing spiral by means of a torsion-galvanometer 
with many coils of thin wire. By interposition of resist- 
ances or by using shunts the desired current-strength was 
easily obtained. The current which the reversal produced 
in the induction-coil was led through the coils of a nearly 
aperiodic reflecting-galvanometer. ‘The deflection then mea- 
sured the double magnetic moment produced in the magnet 
by the acting current. By taking the precaution to reverse 
the current several times after each change of current-strength 
before making the measurement, concordant results were ob- 
tained even with the greatest differences of current-strength. 
