On the polarisation of electrons by scattering

Abstract

The possibility of demonstrating the spin of the electron directly has been discussed since the first growth of wave mechanical conceptions, and the ascription of a magnetic moment to the electron to explain doubling of spectral lines. Bohr has shown, however, that any experiment designed to do this by deflection in a magnetic field, in any manner similar to that used by Gerlach and Stem to measure the magnetic moment of atoms, must fail owing to the operation of the Uncertainty Principle. It does not seem, however, that this will hinder us from demonstrating an asymmetry in the electron by means of scattering experiments; the method is closely analogous to the asymmetric scattering of polarised light, and was in fact used by Barkla to show the polarisation of X-rays. The earlier experiments which were carried out on the polarisation of electrons by Wolf, by Joffé and Arsenieva and by Davisson and Germer all led to negative results. The very careful work of Davisson and Germer should be specially mentioned, in that they reflected electrons from the faces of single crystals at angles corresponding to coherent scattering of the waves. Working up to a velocity of 120 volts they found no asymmetry in the scattering more than ½ per cent., which was within the limit of error of the apparatus. Darwin has shown, however, that no polarisation is to be expected in this type of experiment.