the Photoelectric Effect. 579 



An examination of these curves leads to a number of 

 interesting conclusions. 



(1) The maximum energy, expressed in volts, is a linear 

 function of the frequency of the exciting light, within the 

 limits within which the results of the experiments are 

 consistent. 



(2) The curves appear to be almost symmetrical with 

 respect to their maximum ordinates. This shows that if the 

 maximum energy is a linear function of the frequency, the 

 average energy, which is equal within the limits of experi- 

 mental error to the most probable energy, also bears a linear 

 relation to the frequency. 



(3) The curves intersect the voltage axis at finite angles, 

 both at the end which corresponds to zero energy and at the 

 end which corresponds to the maximum energy. 



(4) If the wave-length is increased to a certain value, 

 which is in the neighbourhood of X = 27 in the case of 

 platinum, the distribution of velocity curve will degenerate 

 into a straight line coincident with the current axis. The 

 photoelectric currents in this region were too small to permit 

 of such a curve being obtained, but the experiments show 

 that this limiting wave-length, which we may call Ao? has a 

 particular meaning. It is the longest wave-length that will 

 produce any photoelectric effect from the metal under inves- 

 tigation, and the electrons emitted by this light are emitted 

 with zero velocity. This wave-length A , or the equivalent 

 frequency v , is that constant which has already been alluded 

 to which determines the photoelectric properties of the 

 metal. 



Analysis of the Distribution of Velocity Curves. 



Since the curves in fig. 3 show that the mean velocity of 

 the emitted electrons is very close to the most probable 

 velocity, it might be thought that the mean velocity should 

 be the characteristic velocity received by the electrons under 

 the primary influence of the light. The deviations from the 

 mean would then be due to collisions with other electrons 

 in the matter traversed, some of the emitted electrons losing 

 and others gaining energy in this way. The results can, 

 however, be equally well interpreted on the view that all the 

 collisions in the interior of the matter result in a diminution 

 of the energy of the emitted electrons. We shall limit the 

 discussion to the case of a beam of light incident normally 

 to the emitting surface. 



Let each electron which is set free from its parent atom 

 or equivalent system in the interior under the influence of 



