D O C U M E N T 2 3 3 M A R C H 1 9 2 6 2 4 3 cause of the low intensity of the beam. In the end, I made do with obtaining as large an interference length as possible, purely empirically, by shifting the diaphragms & lenses, for which I could prove that its limit is not caused by Doppler compo- nents (pp. 18 & 19 of the analysis).[3] The uncertainty in knowing the effective α caused me not to delve deeper into these questions in the publication. It could very well be that for Hβ the spectral null positions at 12 & 15 cm coincide with the null positions determined by Doppler components. I followed with the greatest interest the suggestion you kindly conveyed to me.[4] On the precondition of your approval, I would like to conduct this investiga- tion with the setup I have used. The optical image at the grating may cause some problems. I have already performed similar experiments—albeit without thinking of the goal that you have come up with—when I was trying to produce interference of the light coming from different parts of a decaying canal ray using two slits (p. 15 bot- tom). I also used gratings as such slits. Those experiments came out negatively, I assumed, as a result of beam scattering. I have had on my list of experiments yet to be done for a longer while already an experiment that, with regard to your train of thought, would probably have to be performed at even greater precision, because then an arbitrarily strong light source can be employed & I would like to share it with you. The light quantum should be divided into short segments cτ by means of a Kerr cell, which allows the light to be steered “free of inertia.”[5] (III) From a light source Q (Hg line 546 mμ) let light fall through a Kerr cell KZ onto a lens L of the Michelson interferometer. A peri- odic voltage is applied to the Kerr cell electrodes, which during the one half-period allows the light through, but during the other half blocks the light. then becomes of the wavelength λ of the electric oscillation applied to the Kerr cell. If λ is chosen = 20 cm, then for a distance between the mirrors , the first vanishing of the interference should occur, which for Hg 546 mμ would very certainly be measurable for some multiples of this distance. I would be very pleased to receive your valued response to this experiment. Respectfully and sincerely yours, E. Rupp τ 2 -- - c τ 2 -- - λ 2 -- - a2 a1 – 10 =