1664 
tube B, so that the plumb-line was seen exactly half way between 
them; these wires were to serve afterwards for bringing the pin S¢ 
accurately into the axis of the apparatus and thus ensuring, that 
the centre of the sphere B lies on the same vertical as the centre 
of the spherical shell. The two centres could further be brought 
to the same level by making use of a mark which was made on 
the small pin in the form of a notch, whereas the tube B was 
provided with a divided seale; the exact position of the mark with 
respect to the scale in order that the spheres might be centred was 
ascertained by a few vertical measurements *). 
The oscillating system is set in motion by turning the knob 
K through a small angle. In order to keep this angle within suitable 
limits a lever is attached to the knob carrying two screws at ils 
extremities; these screws come up against the plate which carries 
the knob, and are so adjusted that in the one extreme position of 
the lever the oscillating system assumes a desired position of equili- 
brium, whereas in the other it is turned through a suitable angle. 
By turning the lever from the position of equilibrium into the other 
and back again, the swings about the equilibrium are of the desired 
amplitude. 
3. The determination of the logarithmic decrement of the swings. 
The oscillations were observed by mirror-reading. For this purpose 
the oscillating system at the level of the glass windows carried a 
small ring with frame into which was slid a flat mirror Sp. The 
reading-arrangement was of the special form used in the physical 
laboratory at Leiden?) consisting of telescope, glass scale divided into 
mms., illumination-lamp and hollow cylindrical mirror, and was 
mounted at a distance of about 1.5 metre from the apparatus on a 
separate stone column. 
The decrement of the oscillations was deduced from the observation 
of the maximum-deflections on both sides of the position of equili- 
brium, which was in the middle of the scale. The graph of the loga- 
rithms of these deflections in the successful observations gave an 
1) It was unnecessary to give very minute care to the centring of the sphere 
(similar so that required in ZeMPLÉN’s experiments: Ann. d. Phys. (4), 19, p. 783—1966), 
as the presence of the outer sphere at the most gives rise to a small correction 
(comp. previous papers) and the latter being obviously a minimum for centred 
spheres a small excentricity could not have an appreciable influence. The inner 
sphere could be centred by eye to a few tenths of a mm. and this accuracy was 
more than sufficient. 
2) Comp. Communications n°. 25. These Proceedings April 1896. 
