﻿404 Sir W. Ramsay and Dr. Spencer on Chemical and 



the other, between which a powerful magnet was inserted, as 

 indicated in fig. 3. The values obtained are given in Table II. ; 

 similar values were obtained by using a Soddy's gra ting- 

 instead of the slits. 



To EARTH 



Table II. 

 Zinc amalgam plate. Electroscope positively charged, 





Time of discharge. 



Without magnet 



With magnet 



3 clivs. in 37 mins. 

 3 divs. in 96 mins. 





By the method illustrated in fig. 2, the electroscope was 

 discharged whether excited positively or negatively ; the 

 negative charge, however, was much more quickly dissipated 

 than the positive, and both discharges were more rapid than 

 in the former case. Some experiments were made by this 

 method, using plates coated with sulphides of the metals, and 

 here it was generally found that the discharge due to the 

 sulphide was more rapid than that due to the metal itself. 



3. At this stage in the research, owing to difficulty in 

 maintaining constant the source of illumination, the arc was 

 abandoned in favour of a Cooper-Hewitt mercury lamp, made 

 by Schott and Genossen of Jena, using their Uviol glass, 

 which is transparent to the ultraviolet spectrum of mercury 

 vapour. The light intensity of this lamp is constant for a 

 steady current ; this in all subsequent experiments was 

 maintained at three amperes. 



A series of measurements was made both with the plates 



