440 Experiments with Electron Currents in Different Gases. 



On testing with the electrode E (fig. 1) a vigorous emission 

 of ultra-violet light was found to set in simultaneously with 

 the appearance of the glow. With pressures sufficiently high 

 to give a visible discharge under the comparatively low 

 potential differences, we have not been able to detect any 

 ultra-violet emission at potentials under that at which the 

 glow sets in. In fact, we have never been able to discover 

 any effect on the nickel electrode E with potentials across 

 AB — >-C of less than 11*5 volts. Presumably this is due to 

 the absorption of the single line 2536 of the low-voltage 

 spectrum by the mercury vapour in the connecting tube ; 

 but it may be that the frequency of this line is not high 

 enough to excite the photoelectric emission from the nickel 

 electrode used. We have not tested this point. 



With high pressures there is no discontinuity in the 

 current when the luminosity sets in, such as is observed at 

 lower pressures. This is indicated by the numbers in the last 

 row of the table on p. 430. 



A number of observations giving results of interest were 

 taken in the early stages of the experiment before the 

 tungsten filament had been thoroughly glowed out. Under 

 these conditions the discharge in the Greissler tube showed 

 the spectrum of hydrogen quite strong, but no other lines 

 except those due to mercury. It was then found that with 

 a low pressure of mercury vapour a considerable emission of 

 ultra-violet light set in at a potential difference considerably 

 under that necessary later, when the hydrogen had been got 

 rid of. This effect set in quite consistently between 16 and 

 17 volts, and at first increased rapidly with rising potential 

 difference. The rate of increase then fell off until, finally, the 

 effect varied little, if at all, with the potential difference. This 

 is illustrated by the following numbers taken with a mercury 

 pressure of about O'OOl mm. and an uncertain amount of 

 hydrogen : — 



Volts across A B->C 5 10 15 16 17 20 25 30 35 40 55 65 



Microamps across AB->C. 13 31 45 52 60 68 70 70 71 ... 



Photoelectric current ] 

 from E (scale divs. in I 5 11 25 33 50 58 75 76 

 30 sec). I 



It will be observed that as the potential rises the discharge 

 current becomes constant sooner than the photoelectric 

 emission. The thermionic currents were too small for the 

 kick to develop under the conditions of these observations. 



Wheatstone Laboratory, 

 King's College. 



