332 



Profs. J. J. Thomson and R. Threlfail. [May 6, 



Fig. 2. 



of the mercnry different readings could be obtained by moving it 

 about, the surface of the mercury was therefore illuminated from 

 behind by a parallel and horizontal beam of light which passed 

 through an alum cell to avoid any heating effect. 



The electrodes were either platinum or aluminium, generally pla- 

 tinum ; in some of the tubes these were fused into small pieces of 

 glass tubing, so that only the tips of the electrodes were exposed to 

 the nitrogen. Before being sealed the tube was connected with the 

 pump and the gas supply in the way shown in fig. 3. After being 



Fig. 3. 



cleaned and dried, and the gauge filled either with sulphuric acid or 

 mercury, the tube was pumped out and filled with nitrogen, and this 

 process was repeated several times ; when the pressure was very low 

 the tube was heated to as high a temperature as it would stand with- 

 out softening, in order to drive off any gas that might be on the 



