CH. VIII.] 



OF AN ELECTRON 



89 



the vessel airtight so that it might be exhausted and 

 yet allow the ultra-violet light to pass, was employed, 

 and connected to the expansion apparatus, fig. 12, 

 instead of the vessel A. The current ~Neu was 

 measured just as in the original experiment. 



FIG. 13. J. J. Thomson's auxiliary apparatus for the counting experi- 

 ment. Ultra-violet light passes through the quartz plate CD and 

 through a layer of liquid, which keeps the air saturated, and which 

 constitutes one electrode, to a clean zinc plate K, which constitutes the 

 other, and which is kept negatively electrified. Connexion through 

 the tube L is made with the expansion apparatus shown in fig. 12, this 

 being employed instead of the vessel A in that figure. Then when 

 the sudden measured expansion is caused, a fog is condensed round the 

 negative ions which have sprung into being in consequence of the 

 electrons thrown off by the ultra-violet light; and the rate of settling 

 down of this cloud is then measured. 



A great many precautions must be taken, because 

 there will be some residual cloud found even when 

 electrons or intended nuclei are not present. Positive 

 ions and other stray or undesired nuclei if present 

 can be eliminated by aid of their different behaviour. 

 A differential observation is generally necessary ; 



