On the emission of light hydrogen atoms

When a beam of Canal-Rays passes into a high vacuum, it is well known that the light-intensity falls exponentially to a constant value which, if the pressure is sufficiently low, is very small indeed. The experiments of Wien, and the theory developed by him, showed that the observations could be equally well accounted for on either of two hypotheses, namely that the light from each atom falls off exponentially at the observed rate, or alternatively, that all the energy from any one atom is emitted in a very short time indeed, and that what is measured is essentially the decay-curve of those atoms that are in the initial state for the radiation in question. As to whether the maximum emission (on either theory) occurred exactly at the point of excitation, however, these experiments provided very little evidence; the excitation was in fact, so far as can be determined, distributed continuously over a region about which all that can be said is that it was bounded more or less definitely on one side by the back of the cathode, and it was possible that the curves for a point-excitation would really resemble, for instance, the activity-curve of an intermediate radioactive substance initially absent from a preparation. In fact plausible grounds for believing that this should be the true form were ready to hand, since the atom might very well be expected to spend some time in radiating infra-red lines before it reached the initial state for the radiation observed; on this point all that could be inferred from Wien’s curves was that the interval must at most be appreciably smaller than the half-period of the exponential fall. This interval is not the same as the “Yerweilzeit” considered by him; the Verweilzeit is the basis of a demonstration that the curves due to point-excitation should be exponential, while the present interval would have the effect that they should not. In order to decide whether there was in fact any delay of this nature at all, it seemed desirable to try whether non-luminous canal-rays, some distance after their entry into the vacuum, could be excited anew to luminescence at approximately one known point only, so that any interval between excitation and emission could be directly measured. The experiments devoted to this object have been somewhat interrupted, and it is further necessary to discontinue them, at least for a time, once more. The present paper is thus an interim account, containing a description of the method that has been worked out and the result that has been obtained; there are, however, a number of further questions which it was hoped to decide, and which presumably can be decided by the same method, that are still unsettled.