522 Mr Campbell, Discontinuities in Light Emission. 



period of the electrometer varied in the ratio of 25 to 1, it is clear 

 that the prediction of theory that the fluctuations are independent 

 of the period of the measuring instrument, so long as it is suf- 

 ficiently small, is justified. 



Table II. 



F is potential on electrometer needle, s sensitiveness, C capacity ; 

 6y ^ is in scale diAdsions, deduced from not less than 300 observations 

 in each case. 



80 0-24 0-64 5-4 7-0 1-9 0-37 0-39 



(2) Table III shows the variation of Qt^ with 'p, when p is 

 changed by alterations in the resistance R. Only one set of 

 measurements is given, but other sets with other values of the 

 remaining constants show exactly similar features. The last 

 column gives the value of the quantity which should be constant, 

 if the theory which has been given is correct : it will be observed 

 that it is approximately constant for the higher values of K, but 

 shows a marked increase for the lower values. 



H 

 0-336 

 0-304 

 0-205 

 0-155 

 0-110 

 0096 

 0-032 

 0-027 



s = ^x 10^ i=84. 



I 9. The explanation of this deviation from theory has not been 

 discovered, but it appears to be connected with an increase in the 

 measured capacity of the system with a decrease in resistance, a 

 change to which reference was made in the previous paper. The 

 resistance R was measured by passing a known current through 



