212 
MR. A. Ll. hughes ON THE 
(b) Reflection of Light .—Let ABC, fig. 4, represent the velocity distribution curve 
from the illuminated plate. If a little light gets to the inside of the case, by reflection 
or otherwise, then the curve ahc will be the velocity distribution curve for the current 
from the case to the plate. Experimentally the sum of these two curves is observed. 
A' is the potential to which the system will charge up, and this is less than the real 
value A of the maximum velocity. It is clear that the difference AA' increases as the 
ratio of the reflected effect y to the direct effect x increases. 
(c) Efleet of an Electron approaching the Boundary OhUqueJy .—Consider photo¬ 
electrons leaving the point P (fig. 5) with equal velocities and in all directions. We 
shall see that the velocity distribution curve is not ABC (fig. 6) hut AhC, which 
w JV" 
implies, on the usual interpretation, that all velocities from A down to zero are 
present. Let Vq be the potential required to stojD the electrons travelling normally. 
An electron starting in the direction PN' will describe a parabola. The potential 
difference, which will allow this parabola to graze the plane NN' is Yq cos^ 0. If we 
assume that the electrons are emitted equally in all directions, then the number of 
electrons emitted between the cones 6 and 6 + d6 is proportional to sin 6. Hence a 
number n sin 6 dO of electrons apparently have a velocity equal to and greater than 
Jo 
Yq cos^ 6, and therefore the relation between the numher possessing velocities above a 
certain apparent velocity and that velocity is parabolic. The experimental velocity 
distribution curve would be A6DC and not ABDC. 
The usual experimental arrangements are not so simple as in this case, but it easily 
follows that whenever an electron approaches a boundary obliquely then the potential 
difference just necessary to stop it is less than that potential which corresponds to its 
actual velocity. 
(d) The Effect of Magnetic Fields .-—-A weak magnetic field (such as the earth’s 
field) is often suflicient to impose a considerable curvature on the path of a photo¬ 
electron. This would cause many electrons to approach the boundary with an 
increased tangential component of velocity and so give rise to the effect discussed in 
the last paragraph. Some of the slower electrons would never get away from the 
Illuminated plate, for in the absence of an electric field, they would describe complete 
