408 Dr. Kichardson on the discharge of Electricity 



only be due to changes in the surrounding gas produced by- 

 heating the pentoxide-bulb. 



Throughout this paper, in describing the leaks the unit of 

 current is taken as 3*3 X 10 -n ampere. In measuring the 

 leak the platinum wire was charged to a potential of + 80 

 volts, and the neighbouring electrode connected to one 

 quadrant of an electrometer, the other being earthed. The 

 effect of heating the pentoxide-bulb is indicated by the 

 following table : — 



Table I. 



Time. 



Temperature of 

 P 2 5 bulb. 



Current. 



Wire at 



+ 80 volts. 



Wire at 

 -80 volts. 



3.38 



18° C. 



•35 



•036 





3.50 



94° 0. 

 94° C. 

 94° C. 



20-4 

 55-2 



025 



3.58 



358 





4.06 



124° C. 

 125° 0. 



268 



•025 



4.10 





4.16 



4.20 



145° C. 

 144° C. 



87 : 5 



•030 



4.26 



164° C. 



133-0 







The table shows that the positive leak from the hot 

 platinum is greatly increased by something given off when 

 the P 2 5 is heated, whereas the negative leak is unaffected. 

 As a matter o£ fact, the values o£ the negative leak given 

 above are just about equal to the insulation leak obtained 

 with the wire cold, so that there is no evidence of the 

 existence of any negative leak from a platinum wire in 

 phosphorus vapour at this temperature. Throughout the 

 experiments described in this paper, the temperature of the 

 hot wire was always about 600°-700° C. The table also 

 shows that the value of the positive leak increased with the 

 temperature of the P 2 ^5 DU ^ D : an y irregularities in this 

 respect are probably to be attributed to the fact that the 

 temperature of the bulb was not absolutely steady, but was 

 sometimes going up and sometimes going down as the 

 observations were taken, so that the effect in the testing 

 arrangement might lag behind a good deal. 



