the branches of a WurarsroNe bridge. If now the resistances in the 
latter had been regulated and the pendulum passed from one extreme 
position to the other, the breaking of the contact caused a current 
to flow through the galvanometer during the time rt. From the 
sudden deflexion the length of the time t may be deduced. The 
result was about zt sec. 
§ 6. The phase of the effects. 
Now the phase of the different effects will be considered in details. 
Fig. 3 may be of use for this purpose. It represents as seen from 
im above the iron cylinder C, the 
small magnet AB fixed to it, 
\ the mirror Z, the compensating 
E coil A, the telescope £ and 
the scale S. We shall call the 
direction of rotation shown by 
the arrow the positive one. To 
this rotation corresponds a shift 
of the cross-wire over the scale 
c to the side marked with +. 
We shall speak of a positive 
se. 2S reversal of the current, when 
it gives rise to a north pole at the 
Fig. 3. upper end of the cylinder. When 
all has been arranged it is easy to determine the direction in which 
the pendulum must move if such a positive reversal shall take place. 
The molecular currents existing after a positive reversal have 
a direction opposite to that of the arrow. If these currents consist 
in circulating negative electrons, the moment’ of momentum’ to 
which the reversal gives rise has the direction of the arrow. The 
couple acting on the cylinder must be opposite to this moment and 
we may conclude: At a positive reversal the first effect consists in 
an impulse in the negative direction. Of course it would be in the 
positive direction if we had to do with positive electrons. 
We may remark here, that strictly speaking the effect we are 
seeking for consists in two equal impulses, the first of which 
(breaking of the current) takes place at the beginning of the time 
rt while the second (closing of the current) begins at the end of 
this interval and lasts during the period r'. On account of the short- 
ness of rt and t/ we may say however, that the two impulses coincide. 
It ought to be observed that, with a view to the appreciable 
