194 



Professor Josqoli John Thomson 



[Jan. 19, 



difference between the electrodes a potential divider of 100 resistances 

 of 10 ohms each was used. The current through the tube was measured 

 by a D'Arsonval galvanometer, and the potential difference between 

 the terminals by a Weston's volt-meter. 



Some of the most interesting features of the discharge are very 

 prominent when the temperature of the platinum is high, say 1400° C, 

 and the pressure of the gas low, less than * 01 mm. of mercury. The 

 discharge is light blue, and its spectrum shows the mercury hnes and 

 the band spectrum of nitrogen. In this case the relation between 

 the current and the potential difference is represented by a curve hke 

 Fig. 2, the ordinates representing the current and the abscissae the 

 potential difference. G is the point at which the luminosity begins. 

 In the case we are considering, when the wire is very hot and the 

 pressure low, the change from the dark to the luminous discharge 

 takes place very abruptly, an increase of the potential difference by 

 y^oth of a volt being often sufficient to convert a discharge, where 

 no light could be detected, even in a darkened room, to one where the 

 light was quite bright. When luminosity appears, there is a very 

 rapid increase in the current ; in some of the experiments an increase 

 in the potential difference of jiroth of a volt increased the current 

 forty-fold. At this stage the thermo]iilc showed that there was no 

 increase in the temperature of the ])latinum when the himinosity 

 appeared. We sliall see later on that it is possible by using large poten- 

 tial differences to get such large cun-ents through the tube that the 

 platinum becomes appreciably warmer hy the passage of the current. 



One point which I think very suggestive is the abruptness with 



