period of 0,08 second in the uncompensated field we were able to 
reach over the whole length of the eylinder and the leads periods 
ranging from 10 to 20 seconds. From this it follows that a com- 
pensation from 1 to 7/,°/, has been reached. 
As to the arrangement used for the compensation a brief description 
will suffice. In fig. 1 R and / are resp. a frame and a boop wound 
with copper wire through which a current from a battery of accu- 
mulators was made to flow. The regulation of the intensity of the 
current was effected by means of resistance boxes with shunts for 
the finer regulation. 
When in this way the greater part of the earth-field had been 
got rid of, the remaining part was compensated in the following way: 
At a distance of about 3 m. a rather large permanent magnet 
was placed at the height of the middle of the cylinder S. This 
magnet was turned in the horizontal plane in which it lay till the 
above described apparatus had reached a period of 20 seconds, so 
that the compensation near the middle of the cylinder could be 
considered as sufficient. The compensation at the extremities of the 
cylinder, however, was not reached at the same time and I had 
therefore further to effect it without disturbing that in the middle. 
For this purpose small permanent magnets of equal moment were 
placed in horizontal direction vertically above and beneath the cylinder 
(not shown in the figure). These magnets could be turned about a 
vertical axis and shifted in vertical direction. To ensure the equality 
of their magnetic moments, they were made of pieces of a 
hardened and magnetized knitting-needle, from which a bundle was 
formed. By taking equal numbers of pieces of equal length equi- 
valent magnets were obtained. 
§ +. Compensating alternating field. 
The compensation which was reached in this way and which we 
sball denote by I was not good enough however for sufficiently 
eliminating the source of error to which attention has been drawn 
and for making the weak effect appear that was sought for. 
In order to neutralize the “residual field” as it will be called, the 
following artifice was used (ID. To the iron cylinder S a small 
magnet J/ (magnetized piece of a watch-spring) was fixed in hori- 
zontal direction and a horizontal coil (the ‘compensating coil’) was 
placed with its axis along a line perpendicular to the magnet and 
passing through it. This coil could be shifted in the direction of 
its length. The same current that passed through the windings was 
also made to flow through this coil. It is evident that now, with 
83* 
