About the Production of the Continuous X-Ray Spectrum
I propose, therefore, to generalize my previous assumption so as to admit secondary beams to emerge from the scattering atom in all directions. This amounts to an averaging of the matrix elements defined in my former note over all possible directions. According to this new assumption, only beams emerging with zero velocity should give rise to completely polarized radiation, i.e., only radiation corresponding to the short wave-length limit will be completely polarized. This is in agreement with experiments of P. A. Ross. Furthermore, only radiation near this limit would give zero intensity in the direction of the incident beam; other parts of the continuous spectrum will, of course, furnish intensity in this direction because of the emergent beams forming an angle with the incident wave. This difference in the radiation from the limit and from other parts of the spectrum was already mentioned by Kulenkampff.
Copyright © 1929 by the National Academy of Sciences Communicated March 18, 1929. Read before the Academy, April 22, 1929. I am indebted to Professor E. C. Kemble for having suggested to me this generalization, and also to Professor Kennard for having emphasized the analogy of this procedure with the usual treatment in the case of line spectra. Indeed, when calculating intensities of line spectra, we have to sum over all the different orientations of the atom characterized by the magnetic quantum number m, giving one weight to each of the magnetic levels. Corresponding to this is in our case the averaging over all possible directions of the emerging beam, assuring equal weight for all of them.