22 Mr Zeleny, On Convection Currents and on the [Oct. 31, 



force from these parts. Ions reach these places, however, either 

 by the convection currents described in § I. or by diffusion or by 

 secondary radiation. 



In the results obtained with this apparatus, the fall of potential 

 at or near the plates was found to be less than that hitherto 

 observed, and especially no approach could be made to such high 

 values as those described by Child (loc. cit). 



There is good evidence of the lack of accidental leakage, for 

 with the same potential difference between the plates the nature 

 of the potential curve was found to be independent of the absolute 

 potential of the plates. 



The following table gives the results obtained for the potentials 

 of some points near to the zero plate, when the plates were 1*215 

 centimetres apart and were kept at a difference of potential of 

 320 volts. 



Table I. 



These results are shown graphically in fig. 4 where the line AB 

 represents the uniform gradient existing when the rays are not 

 acting. 



For the same distance between the plates and smaller voltages 

 which were tried down to 40 volts, the fall in the first -^ mm. was 

 found to be correspondingly smaller. 



It is seen that in the first -^ mm. from the plate the change of 

 potential is about as large as in the following -^ mm., and it thus 

 appears that the fall of potential near the electrode is abnormally 

 large. 



The results do not permit us to say however whether this is 

 due to a charge adhering as it were to the metal surface or whether 

 it is due to the existence of a charge in the gas within the first 

 ■£q mm. from the electrode. 



If it were possible to measure the electric force acting upon 

 the plates during conduction such a distinction could be made by 

 comparing the potential gradient thus found at the plate with 

 that obtained in the above experiments for points near to it. 



