224 
MR. A. Ll. hughes ON THE 
producing a photo-electric effect. It was shown in a previous research* that the 
vapours zinc ethyl, tin tetrachloride, and carbon disulphide, could not be ionised by 
ultra-violet light,! although all these coinpoinuls contain elements which are photo¬ 
electric when uncoml)ined. Tliese results are easily interpreted on the view that the 
valency electron is the electron concerned in the photo-electric effect. When combina¬ 
tion takes place the valency electron is attracted more strongly to the molecule of the 
compound than it was to the electropositive member when uncombined. One could 
get a measure of the affinities of different elements for zinc, by observing the shortest 
wave-length necessaiy to cause ionisation in various zinc compounds. The experimental 
difficulty is that the critical wave-lengths will usually be beyond the region that can 
be conveniently used in experiments. 
(ix.) The inside of the apparatus surrounding the illuminated plate was covered 
witli the same layer of soot throughout the experiments. As carbon is of marked 
electronegative properties it is probalde that the fall of potential across the surface 
layer, which may be present, is not more than a fraction of a volt. To correct for 
this, the values of Vy should all be altered by this small amount. This correction 
should be given by an experiment in which the inside of the case as well as the 
illuminated plate is covered with the same distilled metal. 
9. Variations in k and V^^witli Atomic Volume .— Pohl and Pkingsheim found from 
their investigations on the alkali metals tliat tlie photo-electric effect may be divided 
into two parts—the “ normal” effect and the “selective” effect. In the normal effect 
the emission of plioto-electrons is determined solely by the amount of light absorbed, 
and the ratio of the photo-electric current to the energy in the light increases 
continuously with decreasing wave-length. The emission of photo-electrons in the 
selective effect only occurs when a component of the electric force in the light beam 
is in the plane of incidence. The ratio of the photo-electric current to the energy in 
the incident light rises to a very well-marked maximum for a certain wave-length 
which varies from metal to metal. Tins maximum becomes less defined and moves 
towards the ultra-violet as the element becomes more electronegative. A selective 
effect has been observed for Rl), Li, Na, K and Ba, and the maxima correspond to 
wave-lengths within the region X 5000 to X 2000. The question arises as to whether 
the velocities of the electrons produced in the two effects are identical or not. If the 
two effects originated in different systems we sliould not expect a coincidence. Even 
if they are fundamentally the same, it is not self-evident that the velocity of the 
electrons produced near the maximum should obey the same law as the velocities at 
two points well away from tlie maximum. The selective effect is generally regarded 
as a resonance phenomenon, and therefore the emission velocities near the maximum 
of the effect might possibly possess unusually high values. No selective effect has been 
* Hughes, ‘Proc. Camb. Phil. Soc.,’ XVI., p. 378, 1911. 
t The shortest wave-length is given in the paper as X2300. It should be corrected to X 1849, which 
was discovered later. 
