358 Dr. J. W. Capstick. 



The parts of the apparatus were sealed together without glass taps 

 or india-rubber connections. 



The difference of potential to be observed is that between H 

 and K when the current is passing between G and K. There is no 

 need for accurate adjustment of the distance between H and K, for 

 in the dark space the potential gradient is very slight for two 

 or three centimetres, so that all that is required is that the end of 

 H should be outside the negative glow. In the present experiments 

 it was about a centimetre from the end of the kathode. 



In the earlier experiments the difference of potential was measured 

 by a bifilar quadrant electrometer, whose constant was determined by 

 means of a battery of Clark cells. In the later experiments a Kelvin 

 multicellular voltmeter was used. 



The current through the discharge tube was supplied from 600 

 storage cells, a lead pencil line of variable length, drawn on a slate, 

 being included in the circuit, to vary the current strength. 



The current is not always continuous, but sometimes consists of a 

 rapid succession of discharges, and when this is the case the observed 

 kathode fall is not generally constant. Hence a telephone was 

 inserted in the circuit to show when the current was continuous. 

 With the elementary gases the telephone was generally silent, but 

 with the compound gases the humming was so persistent as to come 

 near wrecking the work. 



Hydrogen. Warburg found the kathode fall in hydrogen to be 

 800 volts, but it seemed desirable to repeat the experiments, in order 

 to find the degree of concordance that can be obtained by different- 

 observers. 



The gas was obtained from palladium that had been saturated 

 with hydrogen prepared by the electrolysis of dilute sulphuric acid. 

 It was purified by solid potash and phosphoric anhydride, and the 

 discharge tube contained a piece of sodium, to destroy the last traces 

 of moisture. The apparatus was repeatedly exhausted, heated, and 

 refilled with hydrogen, whilst the electrodes were kept red hot by a 

 strong current, to expel occluded gases. 



The strength of the current was not measured during the experi- 

 ments, but it was varied by altering the resistance in the circuit, so 

 as to cover a varying length of the kathode with the negative glow, 

 and so to show whether the tube was large enough to allow the dis- 

 charge to pass without hindrance. The table below gives details of 

 the measurements. 



The first column gives the pressure of the gas in millimetres of mer- 

 cury, the second gives the fraction of the kathode that was covered 

 with the glow, and the third gives the kathode fall in volts. 



