258 
PHYSICS: R. C. WILLIAMSON 
Proc. N. a. S. 
photoelectric leak from the gauze and walls, which was unavoidable, and 
also at the lower accelerating voltages to thermelectrons from the nozzle 
of the jet. The breaks toward positive values resulted when the positive 
ions, which were produced in the jet by the radiation, acquired sufficient 
energy in the accelerating field to enable them to reach the electrometer 
electrode against the retarding potential. When a screen was inserted 
to cut out wave-lengths less than 3100 A, the negative leak was reduced by 
a small amount, while the positive breaks disappeared, showing that the 
positive currents depend upon the presence of wave-lengths which are 
less than 3100 A. It may be suggested that the positive current is due to 
flOO 
I 
I 
zs 
-50 
A/o 
No 
5cr ze/7 
scrun 
"^000 , 5000 ^ 4000 A. 
tlinmnim/meng/'/is TrmmMijy^cnm 
FIG. 2 
ionization in the jet by collision of photoelectrons from the gauze, and that 
the disappearance of that current upon screening out wave-lengths below 
3100 A results from the correspondingly lower initial energy of emission 
of these photoelectrons. The electrons reach the jet of vapor with their 
initial emission energy plus at least Y a- Therefore, even though the 
initial energy were zero in the screened curves, we should have positive 
ionization by collision when Y a is greater than about 8.2 volts or twice 
the ionizing potential. And the pairs of curves for V,. greater than 4.1 
volts, would show no marked decrease of ionization with screening. On 
the contrary, the pairs of curves are of the same type for retarding poten- 
tials of from 1 to 10 volts, and there is no indication in the screened curves 
