1286 
proper connections and a rightly chosen distance of the compensating 
coil, the alternating couple witb which it acts on the transverse 
magnet will always neutralize the forces, with which the “residual 
field” acts on the alternating horizontal magnetization of the cylinder. 
To reach this the coil could be shifted and besides there was a 
shunt Sh, the resistance of which could be regulated, so that only a 
larger or a smaller part of the whole current passed through the 
compensating coil. 
Because of the small changes in the terrestrial field, which gave 
rise to rather large changes of the residual field, the compensation 
II had to be changed continually. It could however be kept constant 
long enough to allow the observation of the Ersrnin-effect. 
In the last experiment it was found necessary to put an iron core 
in the compensating coil, as the latter could not be brought near 
enough to the cylinder. 
§ 5. The different effects. 
We shall call the effect we are seeking for the first effect and 
the effect cansed by the action of the residual field on the horizontal 
magnetization of the cylinder, combined with that of the compensat- 
ing coil on the transverse magnet the second. 
There is however still a third effect, which like the first consists 
in alternately directed impulses, acting on the cylinder at each in- 
version of the current. Indeed, when the pendulum-wire leaves one 
of the two copper wires a and 5 there elapses a certain time before 
it makes contact with the other one. During that interval + there 
is no current in the apparatus. The suspended cylinder has its 
remanent magnetism, which like the magnetization existing before 
the breaking of the current has a horizontal component and this 
component is acted on by the residual field. This action is not neu- 
tralized in the way described in 7, as during the time r the com- 
pensating coil is without current. We shall call this the ted effect a. 
Finally there is a third effect 6, the direction of which it is 
difficult to indicate without further examination. It may as well 
have the sign of the effect 3a as the opposite one. 
In order to understand the nature of this effect, we must keep 
in mind that, at each change of the direction of the current the 
magnetization of the iron passes through a hysteresis-cycle, as is 
roughly shown in fig. 2, where the abscissae represent the intensity 
of the current ¢ and the ordinates the magnetization. OA or OD 
corresponds to the remanent magnetism, which remains every time 
after the breaking of the current. 
