294 Mr. W. B. Gill on Ionization Power of 



were obtained for very much lower pressures, and the values 



X 



of — were in general very much higher : for air 100 to 



1000, hydrogen 30 to 400, carbon dioxide 100 to 1200. 



"V" 



Kememberino- that - is a function of -, it is only possible 

 to p p J r 



to compare these results in the case of hydrogen, as is done 

 in the following table : — 



X 



v 



30 



50 



70 



f Townsend 



•08 

 •054 



•36 

 •29 



VOqp 

 •59 



■P 1 Bishop 





from which it appears that the values obtained by Bishop 



are smaller than those previously obtained. Even where no 

 y v 



values of — are common, as for air, by plotting - against — 



it is possible to see that the curves obtained by Townsend and 

 Bishop cannot possibly join up, that of Bishop being again 

 the lower. To account for this discrepancy it is necessary to 

 consider the methods of finding a in the two cases. 



Both start from the fundamental relation that if n ions start 

 from a negatively charged plate, the number reaching a 

 parallel plate at distance d is given by 



n = n e ad , 



it being assumed that only the negative ions are effective in 

 producing fresh ions by collision. In both methods the n 

 ions are produced by ultra-violet light falling on a zinc 

 plate. 



The difficulty of finding n was evaded by Townsend, who 

 worked with two different distances, d h d 2 , from which, if 

 n u n 2 are the corresponding values of n, 



!ll _ e a(d x -d 2 ) 



n 2 

 an equation giving a. 



Bishop, on the other hand, obtained experimentally a 

 complete current-E.M.F. curve, keeping d constant, and 

 attempted to estimate n from the saturation portion of this 

 curve. 



