312 Chemical Effects of the Electric Discharge. 



current never exceeded '0052 ampere. The mean value o£ 

 the pressure may be taken to be 3 or 4 nuns. 



The phosphorus pentoxide in the bulbs was carefully 

 watched for the first symptom of absorbing water, and it was 

 observed that the phosphorus pentoxide under the cathode 

 was distinctly soiled by water-vapour before that which was 

 under the anode appeared to be soiled at all. And later, 

 when the latter showed also the yellow effect of absorbing 

 water-vapour, the former continued to maintain a deeper 

 colour. 



The experiment proA^ed that the chemical activity was 

 greater in the neighbourhood of the cathode than elsewhere 

 in the discharge. 



[After obtaining this result some quantitative experiments 

 were made. The values of Ap/AQ for several pressures were 

 determined and compared with values obtained with the 

 previous apparatus when the distance between electrodes did 

 not exceed 1*5 cm., due allowance being made for change of 

 capacity. 



This comparison showed a large increase in the chemical 

 combination per coulomb, presumably due to dissociation 

 within the positive column, and led to the present 

 investigation.] 



We may apply the curves of fig. 2 to obtain a rough 

 estimate of the relative amounts of water-vapour contributed 

 by the positive column and by the region of the cathode fall 

 of potential during the experiment just described. 



Taking the mean pressure to be 3*3 and the mean current 

 •00275, our curves give, for D = ll-6. Ap/AQ = 3"5. Hence 

 W = 10*5. Again, taking D about 1 cm., so as to eliminate 

 the effect of the positive column, we obtain Ap/AQ = 2 about, 

 and W = 6, as the contribution of the cathode fall. The 

 quantities of water- vapour formed in the positive column 

 and in the cathode fall were therefore in the ratio 4*5 : 6. 

 The former would diffuse about equally into both bulbs ; of 

 the latter by far the greater part would go to the bulb 

 beneath the cathode which thus would receive about four 

 times as much as its companion. 



I must express my thanks to Professor J. S. Townsend, 

 to whom I am greatly indebted for some most valuable 

 suggestions. 



