FROM THE CURRENT LITERATURE

POLARIZATION OF AN ELECTRON BEAM DURING SCATTERING

One of the essential consequences of Dirac’s theory—the polarization of a beam of fast electrons as a result of scattering—had until recently remained essentially unconfirmed experimentally. Quite numerous specially designed experiments on double scattering of an electron beam led to contradictory and not very reliable results. Therefore the preliminary communication by Shinohara and Ryu*), who obtained better agreement with the predictions of the theory, is of considerable interest.

Fig. 1. Diagram of the experimental apparatus.

The scheme of the experiment is clear from Fig. 1. A beam of electrons from the accelerator was directed onto a gold foil set at an angle of 45° to the beam and serving as the polarizer. The electrons scattered in this polarizer foil at an angle of 90° to the direction of the initial beam (the “transmitted” beam) were selected by a diaphragm system and fell on the analyzer target, made, like the polarizer, of hammered gold foil of thickness \(5 \cdot 10^{-6}\) cm and set perpendicular to the beam of scattered electrons. Two Geiger counters measured the number of electrons scattered by the second foil in directions corresponding to a scattering angle \(\theta_2\), approximately equal to 78°, and lying in the plane of scattering of the primary beam \(BPA\) (\(\varphi_2 = 0^\circ\) and 180°; Fig. 2). As the authors indicate, such an arrangement of the analyzer ensured the automatic exclusion of a number of errors, primarily differences in the origin of the secondary scattered beams (both beams are “transmitted”), as well as errors connected with possible changes of voltage in the accelerator. The latter is especially

*) K. Shinohara and N. Ryu, Phys. Rev. 75, 1762 (1949).

important, since the intensity of the electron beam scattered by atomic nuclei is very sensitive to changes in the energy of the scattered electrons. The slits that selected the electron beam, as well as the scattering chamber,

Fig. 2.

were made of aluminum in order to reduce the number of parasitic electrons, as well as the emission of secondary X-rays. Lead blocks, not shown in Fig. 1, shielded the counters from X-radiation. Control experiments showed that these precautions were sufficient. The branch pipe to which the counters and the analyzer foil were attached could rotate as a whole about the axis \(a\) (Fig. 1), and thus the two counters could change places. This made it possible to eliminate the influence both of differences in the sensitivity of the counters and of unavoidable differences in their geometrical arrangement.

According to the theory, the intensity of the beam of electrons that has undergone scattering in the polarizer \(P\) through an angle \(\theta_1\), and then has been scattered by the analyzer \(A\) through an angle \(\theta_2\) at an azimuth \(\varphi_2\) relative to the plane \(BPA\) of the primary scattering (Fig. 2), is proportional to

\[ 1+\delta \cos \varphi_2, \]

where \(\delta\) is a certain function of the angles \(\theta_1\), \(\theta_2\) and of the electron energy. Thus, placing counters I and II at azimuths \(\varphi_2\) equal to \(0\) and \(180^\circ\) relative to the plane \(BPA\), and measuring the intensities \(I_1\) and \(I_2\) of the electron streams entering these counters, it is easy to find, for the given values of \(\theta_1\), \(\theta_2\), and the electron energy, the value of \(\delta\):

\[ \delta = \frac{I_1-I_2}{I_1+I_2}. \]

The table gives the values of \(200\delta\) measured by Shinohara and Ryo, as well as those calculated theoretically by various authors. It is evident from the table that the experiments of Shinohara and Ryo convincingly testify to the presence of polarization of the scattered electron beam. The authors believe that the degree of polarization and its growth with increasing electron energy are in good agreement with the predictions of the theory. They explain the reduced measured values of \(\delta\), as compared with the calculated ones, by the difference of \(\theta_2\) from

Table

Measured and calculated values of \(200\delta\) for electrons of different energies

) N. F. Mott, Proc. Roy. Soc. A135, 429 (1932).
) J. H. Bartlett a. R. E. Watson, Phys. Rev. 56, 612 (1939).
) H. S. W. Massey a. C. B. O. Mohr, Proc. Roy. Soc. A177**, 341 (1941).

\(90^\circ\), as was assumed in all theoretical calculations. The error limits indicated in the table correspond to the probable statistical error.

R. G.