[M‘LENNAN | ELECTRIC SCREENING IN VACUUM TUBES 87 
1. PRELIMINARY EXPERIMENTS. 
In these experiments an effort was made to work under precisely 
the same conditions as those selected by Professor Thomson. 
The form of tube used, as already mentioned, was that shown in 
Fig. I. The Faraday cylinder D which served as a screen for the elec- 
trode C, was kept connected to earth by means of a fine platinum wire 
passing through the tube at H. The plate which closed the upper end 
of this cylinder was made of aluminium about ‘04 mms. in thickness. 
As glass, even for low voltages, is not a good insulator, care was 
taken to prevent any leak from the electrode over its surface by melting 
wax on the tube at a and a’. Tests made from time to time through- 
out the experiments showed that this insulation sufficed to maintain any 
charge given to the electrode when the tube was not excited. 
The tube was kept connected to a mercury pump throughout the 
investigation and was excited by an eight inch spark length induction 
coil running under a tension of eight volts. The exploring electrode 
was joined to a quadrant electrometer and this instrument together 
with the connecting wire was surrounded by an earth connected con- 
ductor in order to screen off electrostatic action. 
Under these conditions, it was found that on passing a discharge 
through the tube the electrometer indicated no action until a pressure 
of about one millimetre of mercury was reached. At this pressure, 
with the coil joined to any two of the three terminals A, B and D, posi- 
tive or negative charges given to the protected electrode gradually leaked 
away. At lower pressures the electrode C slowly acquired a negative 
charge. This charge, however, did not go on increasing when the 
discharge was passing, but after a time reached a limiting value and 
then remained stationary. 
With still lower pressures the same effect was observed but the 
value of the limiting charge increased and was more quickly reached. 
With pressures so low that cathode rays could traverse the bulb of 
the tube, a momentary discharge sufficed, when A acted as a cathode, 
to give a deflection beyond the range of the electrometer, a value which 
indicated a charge of at least five or six volts. 
Although this statement represents in a general way the results ob- 
tained, it was exceedingly difficult to trace uniformity in the effects. 
It frequently happened that the electrode C, instead of receiving a nega- 
tive charge, received a positive one, this being especially the case when 
A and B were the terminals and neither connected to earth. 
Sec. III., 1900. 6. - 
