220 
MR. A. Lr. HUGHES ON THE 
the velocity was again I'Ol volt. The selenium was left for about fifty hours in 
contact with the residual gases which came out from the charcoal (:^ '01 mm.). 
There was iio change in tlie velocity. Then air at 200 mm. was admitted for 15 
minutes. There was no change in the velocity even after this. It is clear that the 
photo-electric properties of selenium are not nearly so much affected by contact with 
air as those of zinc or cadmium. 
yVs selenium would oidy emit electrons with X 1849, it was not possible to find Ic and 
V., for selenium. If we assume selenium to have approximately the same h as Bi 
and Sh (Avhich are in the next group), we find that V,, for selenium is about 4"8 volts. 
Arsenic .—Four distillations of arsenic were made. The effects of the first two were 
so much smaller tlran the effects with other metals that velocity measurements were 
not undertaken. Some pure arsenic from another source was obtained, but still the 
unusually small leak was obtained and so the effect appeared to be real. There was 
no photo-electric effect with X 2537. The effect set in between X 2386 and X 2330. 
With X 2257 the effect was far smaller than with other metals. The effect with 
X 1849 was considerably smaller than with other metals, but the difference was nothing 
like so great as in the case of X 2257. It was impossible to determine the emission 
velocities with any accuracy. One distillation gave ‘44 volts for X 2257 and 1’50 volts 
for X 1849. The other distillation gave '33 volts for X 2257 and 1’40 volts for X 1849. 
Even when the greatest precautions were taken, these values could not be determined 
to within 1 volt. Hence we can only say that k lies between 3'6 x 10“^’’’ and 3'8 x 10“^^, 
and that is of the order 4‘4 —4‘6 volts. 
Zinc Chloride .—Zinc chloride was heated in the quartz furnace in the vacuum for 
some time to drive off moisture. Then the disc was lowered in order to receive a layer 
of the distilled salt. There was not the slightest trace of any photo-electric leak even 
wiien the unresolved light from the mercury arc fell upon the disc. An unexpected 
effect was then observed. Ordinary undried air was admitted to the apparatus at a 
pressure of 760 mm. for 15 minutes. On evacuating again, a slight photo-electric 
leak was obtained. Under similai’ conditions a zinc plate would emit a current of the 
order 10^ times as great. This leak from the zinc chloride after contact with air 
increased by about 80 per cent, in the course of an hour. The plate charged up to 
about ’3 volt when the unresolved light from the mercury arc fell upon it. There was 
not enough effect to use the monochromator. After the exposure to air, the surface 
of the zinc chloride, previously quite dry, appeared slighly moist. 
Quantitative experiments on the total photo-electric effect were not made, but, 
as far as one could judge from the rate at which the plate charged up, the effect was 
much the same for all metals except arsenic. The reduction in the total effect for 
cadmium after the admission of oxygen was certainly far greater than the differences 
between the effects for the metals distilled in vacuo. 
8. Discu.ssion of the Results. —(i.) It is clear from these experiments that the energy 
with which the fastest electrons emerge from a plate illuminated by ultra-voilet light 
