246 _ Mr. G. Owen on the Discharge of 
At this higher temperature the leak attains its steady 
value at a pressure of about 7mm. The smali positive leak 
at the three highest pressures is due to the fact that this 
filament had just been newly set up. A new filament dis- 
charges for a short time after it is set up a copious supply 
of positive electricity. This point will be discussed im 
Section 6. Dr. H. A. Wilson also found the negative 
leak from a platinum wire under a potential of 40 volts to 
fall off as the pressure increases. He suggested as a 
probable explanation that the presence of the gas hinders 
the escape of the negative corpuscles from the wire. The 
same explanation may be applied to the above in the case of 
the Nernst filament. 
There is a further point of some interest in connexion with 
the leak at low pressures that may be shown from a con- 
sideration of the curves in diagram 3. The following table 
gives the leak under a potential of 350 volts at four conse- 
cutive pressures taken from that diagram. 
TABLE V. 
Pressure. Leak. 
0°035 mm. 260 
O09: +, 360 
/ 
O16 ,, 520 | 
032, 960 | 
| 
If these values of the pressure and the corresponding leaks 
be plotted, the curve passing through the four points is very 
nearly indeed a straight line. This shows that at low pres- 
sures for a considerable range the current is proportional to 
the pressure. 
(3) Variation of the Positive Leak with the Potential at 
Different Pressures. 
Table VI. gives the results of observations made at various 
dates on the variation of the positive leak with the potential 
at different pressures. 
The currents on the 22nd and 25th of March were measured 
by means of an electrometer. That on March 8th, however, 
was measured by means of a galvanometer. As the deflexion 
