Electricity from a Nernst Filament. 247 
TaBLe VI. 
D Pressure in Temperature of fila- | Poten- Positive 
1 a Filament. | mms. of | ment or Resistance | tial, | Current, 
904. mercury. | ia series with it. | volts. | 1=10—-8 amp. 
March 8. if 0:22 1430° C. + 400 15 
March 22. Lay 0-018 1540° C. + 40 0°56 
+ 80 1-41 
+160 Lay, 
+240 1-72 
+400 1-67 
March 25. ITI. 764 100 ohms O | (negative leak) 
+ 80 2°8 
+160 4°9 
+240 53 
+400 64 
67 150 ohms O | (negative leak) 
+ 80 9°5 
+160 101 
+240 9°4 
+400 10°3 
32 150 ohms 0 | (negative leak) 
+ 80 o7 
+160 73 
+320 70 
of the galvanometer was only about 2 divisions, too much 
importance must not be attached to the value 1°5x107§ 
ampere given to the current in the table. In measuring this 
leak on March 8th it was noticed that after the potential of 
+400 volts had been on for a few seconds, the discharge 
suddenly changed from a steady current of 2 divisions to a 
luminous discharge. This result, combined with the other 
observations given in Table VI., shows that the positive leak 
can be saturated at all pressures provided no luminous dis- 
charge takes place. The positive ion, therefore, does not 
acquire so readily as the negative ion the kinetic energy 
necessary to ionise a molecule. 
(4) Vareation of the Negative Leak with the Temperature 
in a Vacuum. 
Table VII. shows the way in which the saturation current 
from the filament charged to anegative potential of 160 volts 
varied with the temperature. The currents were measured 
by means of a galvanometer. 
' 
