88 DETERMINATION OF THE MASS [CH. vm. 



an electrometer ; a sufficient difference of potential 

 being maintained between them. 



And now, metrical condensation having been pro- 

 duced by the expansion appliance, and a mist formed, 

 the rate of its fall or gradual subsidence can be 

 observed by looking through the vessel at an illumi- 

 nated surface ; whence by Stokes's theorem the size 

 of each globule is known. The quantity of water 

 which had gone to form globules is known from the 

 measured amount of expansion, by a process the 

 details of which I will not give here ; and so 

 the number of such globules, and therefore the 

 number of their condensation-centres or nuclei or 

 ions, can be determined. 



If c is the observed rate of fall in stagnant air, 

 the linear dimensions of the falling drops will be 



//9//c\ 7/4 '5 x '000 18 

 r = V WJ = Vi 981 



In a given case c was observed to be 0*14 centimetres 

 per second ; hence the volume of each drop was in 

 that case . 



-7rr 3 =l-6xlO- 10 c.c. ; 

 o 



and so, if the aggregate amount of water in all the 

 drops in a given space is reckoned from the measured 

 amount of adiabatic expansion which caused the chill 

 and the precipitation, the drops can be counted. 



Professor Thomson, a few months later, repeated 

 this experiment, the ions being now produced by 

 illuminating a negatively electrified zinc plate with 

 ultra-violet light. The apparatus used is shown in 

 fig. 13. A clean zinc surface in vacuo, faced by a 

 transparent conductor through which the light could 

 shine on it, and by a window of quartz which makes 



