252 Mr. G. Owen on the Discharge of 
TABLE X. 
Pressure=764. mms. 100 ohms in series with the filament. 
Time in minutes 
(approx.). Leak 
0 +55 
15 +13 
30 —30 
The initial positive leak with no potential on the filament | 
does not come back cn allowing the filament to remain in the 
tube for some days. Positive leaks may, however, always 
be obtained by charging up the wire to potentials greater than 
about +70 volts. The “steady ” positive leak thus obtained 
also decreases slowly with long-continued heating of the 
filament. The negative leak, on the contrary, diminishes 
but very little with the time. From the following Table (XI.) 
the rates of the decrease of the positive and negative leaks 
may be compared. The leaks were measured at the com- 
mencement of sets of observations taken on each of the dates 
given. 
ee a ee Se, a 
TABLE XI. 
Filament I. fully glowing, at atmospheric pressure. 
Date, Positive Leak under |Negative Leak under 
1904. +600 volts. —600 volts. | 
February 27 ......... 136 x 1078 95x107° | 
February 29 .. .... 75x 10-8 92x107° 
Mardi” }..0) 0 45x 10—§ 90x 10-° 
Between Feb. 27th and March 1st the positive leak has 
thus diminished 67 per cent. In the same interval the 
negative leak has only diminished about 5 per cent. It is 
of course to be understood that the initial positive leak with no 
potential on the filament had been once and for all destroyed 
by continuous heating on dates previous to Feb. 27th. 
The positive leak decreases also with the time in a vacuum, | 
as may be seen from Table VI., where the variation of the 
