On the Kathode Fall of Potential in Gases. 359 



p. Grlow. E. 



6 J 298 



Apparatus exhausted and refilled. 

 5 f 304 



Apparatus exhausted and refilled. 

 2 298 



2 i 296 



2 298 



More gas admitted without exhausting. 

 4 297 



4 296 



The mean result is 298 volts, which agrees so closely with the 

 value 300. found by Warburg, that no further experiments were 

 made. 



Oxygen. The gas was made by heating permanganate of potash, 

 and passed through a set of Geissler bulbs of sulphuric acid and over 

 lumps of potash and phosphoric anhydride before reaching the dis- 

 charge tube. The permanganate was twice recrystallised, dried, and 

 heated till it fell to a fine powder, before being introduced into the 

 apparatus, in order to ensure the absence of moisture, and so diminish 

 the risk of formation of volatile manganese compounds. 



The degree of purity of the gas was tested by the spectrum of the 

 discharge, and it proved a difficult matter to get rid of the nitrogen. 

 By repeatedly exhausting and filling the apparatus with oxygen 

 while the glass was kept hot, the nitrogen lines were rendered very 

 faint, but they were not entirely removed even after continuing the 

 operations for a fortnight. The hydrogen lines soon disappeared 

 entirely, but the nitrogen lines were always faintly visible when the 

 current was strong enough to make the kathode red hot. As, how- 

 ever, a very small amount of nitrogen is sufficient to make the nitro- 

 gen lines far brighter than the oxygen lines, the quantity of nitrogen 

 present must have been extremely small. 



In the observations recorded below, from two to six readings with 

 different current strengths were taken at each pressure. The mean 

 result is given for each set. 



p. E 



11 370 



7 371 



4 363 



2J 370 



H 373 Mean, 369 



The tube was then heated, exhausted, arid refilled several times, 

 and a similar set of readings taken. 



VOL. LXIII. 2 D 



