450 
DR. W. MAXSERGH VARLEY ON THE PHOTO-ELECTRIC 
previously made as to the connection between the potential at which ionisation by 
collision commences and that necessary for a discharge to pass, from a consideration 
of the sparking potentials in these gases at atmospheric pressures, the potential 
gradieiit necessary for a discharge to pass in hydrogen being only 0'55 times that 
recpiired in air. 
A com])ai'ison of the curves for liydrogen in figs. 4 and 9 will show the consistent 
character of the results obtained. 
For pressures down to a tentli of an atmosphere the currents in carbon dioxide are 
approximately 175 times as great as those in hydrogen for all potential gradients for 
wliich the current lies on the more horizontal part of the current-E.M.F. curve, 
that is for those currents nearest approaching to saturation. The currents in air are 
greater tlian tliose in liydrogen, though not nearly so great as those in carbon 
dioxide. . At lower pressures (see figs. 7 and 8) the ratio becomes much less, at 
18 millims. pressures the current in carbon dioxide being less than 1'5 times that in 
hydrogen, as before, for |)oints lying on the less steep portion of the curve. 
At still lower ])ressures the ionisation by collision introduces complications even 
with small potential difierences, and soon becomes the dominant factor in determining 
the currents until the very lowest pressures are reached. At pressures below about 
0’005 millim. the method employed ceases to l)e satisfactory, as the corpuscles given 
off ly the cathode are shot riglit through tlie meshes of the gauze. Some observa¬ 
tions wliicb were made, using a gauze of rather wide mesh, are given in the following 
talde as showing some apparently remarkable results whicb were traced to this 
cause:— 
Table XL 
Potential clifferonee between 
the electrodes in volts. 
Current in arbitrary units. 
2 
52 
8 
109 
40 
124 
120 
148 
200 
138 
280 
126 
360 
112 1 
The experiments were made in air at a pressure of 0‘024 millim., the distance 
separating the electrodes being 2 millims. They show that beyond a certain 
potential an increase in the potential difference between the electrodes produced a 
decrease in tlie current. That this was due to the corpuscles being shot through 
tlie meshes of the gauze as soon as the potential gradient was big enough to give 
them a sufficiently great velocity was shown by substituting a much finer gauze, 
