480 BELL SYSTEM TECHNICAL JOURNAL 



reflection have been made also by Cox, Mcllwraith and Kurrelmeyer, 

 by Joffe, and by Wolf. The experiment by the first-named three ^ is 

 similar in principle and arrangement to our own; the intensity of a 

 beam of electrons which has been twice reflected through 90 degrees is 

 measured while the second reflector and collector are revolved about 

 the direction of incidence of the second reflection. But in other re- 

 spects the experiments differ. The electrons constituting the primary 

 beam are jS-rays from a sample of radium, the reflectors are plates of 

 polycrystalline gold, and the collector is a point-discharge electron 

 counter. The authors report that the shielding between the electron 

 source and the counter was inadequate to suppress entirely an effect 

 due to the gamma radiation, and further that rapid changes in the 

 characteristics of the discharge point made it difficult to obtain con- 

 sistent data. The results which they publish are ratios of the cur- 

 rent received by the collector in one of the "parallel" positions to 

 that received in one or the other of the "transverse" positions, and the 

 ratios of the currents received in the two "transverse " positions. The 

 values found for the first of these ratios depart from unity by much 

 more than the probable error, and show a bias in favor of polarization. 

 The authors do not point this out, however, but lay emphasis instead 

 upon a rather slight departure from unity of the values obtained for 

 the second ratio — that of the currents in the two transverse directions. 



The experiment by Joffe is mentioned by Darwin ^ in a short article 

 on the Sixth Congress of Russian Physicists which was held last sum- 

 mer. Darwin remarks that at one of the meetings Joffe reported that 

 he had looked for a polarization of electrons by reflection, but had 

 failed to detect such an effect. So far as we are aware no report of 

 this work has been published.^" 



In the experiment by Wolf " a beam of low speed electrons (accelerat- 

 ing potentials of about 10 volts) is deflected in a magnetic field and 

 caused, while still in the field, to impinge at 45 degrees incidence upon 

 a target which in various tests was a plate of brass, a cleft crystal of 

 galena and a crystal of copper. The currents to the target and to an 

 enclosing electrode are measured as the target is revolved about the 

 direction of incidence, and are found to be independent of azimuth. 

 This result is susceptible of two interpretations at least ; it may mean 

 that the incident beam is not polarized by the magnetic field, or it may 

 mean that none of the targets serves as an analyser. The latter inter- 

 pretation, which leaves unanswered the question of polarization in a 

 magnetic field, is consistent with the result which we have obtained. 



8 Cox, Mcllwraith & Kurrelmeyer, Proc. Nat. Acad. Sci., 14, 544 (1928). 



9 Darwin, Nature, 122, 630 (1928). 



1° A brief account of these experiments has appeared recently in the Comptes 

 Rendus; Joffe and Arsenieva, C. R. 188, 152 (1929). 

 11 Wolf, Zeit.f. Phys., 52, 314 (1928). 



