184 F. Leininger: Relation of Electric Charges transported 
When coupled differentially, the two coils of the Wiede- 
mann galvanometer were adjusted to equality of resistance, 
and insulated from one another by two thick plates of glass. 
If only the front coil was in use, the sensitiveness of the 
instrument with a scale distance of 2 metres was 3°2 x 1077 
ampere per scale-division. 
In order to screen the galvanometers as much as possible 
from the magnetic effect of the motor, they were set up as 
far as possible from it. 
In the later experiments, moving-coil galvanometers were 
used, by Siemens & Halske and Hartmann & Braun, of 
Frankfort-on-the-Main. The last-named instrument had a 
resistance of 660 ohms, and to this was added a further resist- 
ance of 5000 ohms. In addition to this, a shunt of 50 ohms 
was used across the galvanometer. This instrument served 
for the current measurement, and had a sensitiveness of 
9°3x 10-1, which was fairly constant, with a scale distance 
of 1°65 m. 
For the measurement of the radiation, a Siemens & Halske 
moving-coil galvanometer was employed, having a resistance 
of 10,000 ohms; a shunt of at first 200, and later 100 ohms 
was used. The sensitiveness of this instrument was very 
constant, and amounted to 5:1 x 107!° ampere with a scale 
distance of 1‘8 m. The source of electrical energy was a 
self-exciting Wimshurst influence-machine having two movable 
ebonite plates 40 cms. in diameter, and driven by a small 
electro-motor supplied with current at 12 volts.. It gavea 
spark-leneth of 14 ems., and when directly short-circuited 
through the galvanometer, a current whose average value 
was about 2°14 x 10—° ampere. 
The tube and the influence-machine were set up in an 
earth-connected zinc box. All the leads to the galvanometers 
and tube were of bare copper wire stretched on sealing-wax 
supports. Only the voltmeter wire was, on account of its 
high potential, embedded in a thick glass tube filled with 
paraffin wax. . 
Mertsop I. 
Account of Experiments. 
I first determined the ratio of the convection-currents due 
to cathode or canal rays to the main current, by measuring 
‘in succession the current received by the collecting electrode 
from the charged particles of the rays, and then, by applying 
a shunt of 50 ohms, a fraction of the main current. In 
measuring this ratio, the electrode next to the collecting 
electrode was made an anode in the case of cathede rays, and 

