Hlectriaty from a Nernst Filament. 249 
It will be seen from the curves that the negative leak ina 
vacuum increases very rapidly with the temperature. Pre- 
cisely similar curves have been obtained by O. W. Richardson * 
representing the variation with the temperature of the satu- 
ration negative leak at low pressures from platinum and carbon 
filaments. 
(5) (a) Variation of the positive leak with the temperature in 
a vacuum. 
(6) Relative magnitudes of the positive and negative leaks 
at different temperatures in a vacuum. 
(a) In these experiments on the positive leak the currents 
were measured by means of an electrometer, the currents 
being usually too small to be appreciable on a galvanometer. 
Table VIII. shows the variation of the positive leak with 
the temperature in a vacuum. 
Txapne V 11T 
March 22. Filament Il. Potential on filament 
= + 200 volts. 
Pressure, Temperature, Saturation Current. | 
millimetres. ne: 1=10—10 amperes. | 
(6 fo 1250 4:8 
1285 iH 
(U0 Uy ree Bee 1305 15 
1357 21 
@0OG6 Hicks 3: 1400 33 
1445 55 
1475 7A 
1520 123 
OOOO eee. | 1550 158 
1615 267 
The results given in Table VIII. are shown diagrammatically 
in curve A, diagram 8. 
Itis seen from this curve that the positive leak in a vacuum 
increases rapidly with the temperature. Comparing this 
curve, however, with those in diagram 7, we notice at once 
that the rate of increase of the positive leak with the tempe- 
rature is much smaller than that of the negative leak. 
(b) The absolute values of the positive and negative leaks 
at the various temperatures given in Tables VII. and VIII. 
may not be compared, for the filament and the thermocouple 
were not the same in the two cases. In order to see how the 
ratio of the negative to the positive leak varied with the 
* Phil. Trans. A. vol. cci. pp. 497-549 (1903). 
