1296 
$ 9. Conclusions and remarks. 
1. The way, in which, both in the figures I—V and VI—XIII, 
the phases succeed each other agrees with the foregoing considerations. 
Simultaneously with the minimum of the amplitude I observed the 
phase in which, at the moment considered the wire is in its extreme 
position. The further we are away from the minimum, the more the 
phase approaches that of effect 2. 
2. The change in phase that is observed at the passage through 
the minimum (If and IV, XI and XIII) is caused by our having, gone 
too far with the compensation II. After having been neutralized, 
effect 2 reappears with opposite phase. It consisted initially in the 
action of the residual field on the horizontal component of the 
magnetization, and finally in the oppositely directed action of the 
compensating coil on the small magnet fixed to the eylinder. 
3. In the experiments HI and XII a minimum is reached while 
at the same time the phase is that of effect 1. Now in one of these 
cases effect 8 must have had the opposite phase, so that we may 
conclude to the existence of the effect we were seeking for. As in 
XU the amplitude (0,7 em) was smaller than in III, this latter 
experiment must have been the one in which 3a, and 1 had the 
same phase. It is seen by this that the total effect 38a, } has had 
the same phase as 3a alone. 
4. For the minimum amplitude of the oscillation of the wire we 
may take 2 em before the change in azimuth (see experiments IT] 
and IV) and afterwards 0,7 cm. The first value corresponds to the 
sum of effects 1 and 8, the last to the difference of these two effects. 
Therefore we may estimate the first effect at half the sum of the 
measured distances viz. 1,4 em and the third effect at 0,6 or 0,7 
em. Of course the result 1,4 does not give much more than the 
order of magnitude. 
5. Taking this value for the deflexion on the scale caused by 
effect 1 we may make an estimate of the magnetic moment of the 
cylinder. 
According to (J) half the distance over which the wire is seen 
to oscillate, is given by 
3.195, 
if S is the velocity produced by one sudden impulse. Hence 
0.7 : 
= 99 em/ 
-= 0.22 eu fone 
TSI 
Le» 
which must be corrected to 
SIOE ee 
