2 3 2 D O C U M E N T 3 0 0 J U L Y 1 9 2 2 300. From Peter Pringsheim Berlin W., 63 Lützow St., 19 July 1922 Dear Professor Einstein, During a visit that Franck[1] paid us a couple of days ago in Berlin we discussed “our” problem more thoroughly and I think we arrived at some results.[2] At that time we tried to reach you by telephone but as we learned that you are away[3] and I do not know when I shall have an opportunity to talk to you again, I would like to write you a little about it before I forget the matter and then first have to pick my way through again. The main point seems to me to be, if all indications are not deceiving us, that the experiment concerned again does not lead to a decision between instantaneous or longer-lasting emission, instead that not only the “posi- tion effect”[4] but also the “Lorentz effect” (broadening through [cutting off] of the wave train or through phase shift)[5] can likewise occur in instantaneous emission but then, of course, another explanation is needed. First of all, it is an experimental fact that some spectrum lines that exhibit virtually no Stark effect (and the position effect probably ought to be interpreted as such) are very strongly broadened by a rise in pressure.[6] Theoretically, however, two reasons could lead to a line broad- ening even for timeless electron transitions from an increase in collisions. The first is already treated in a publication by Franck, as yet unfamiliar to me, from the Fest- schrift issue of the Kaiser Wilhelm Society.[7] It is, of course, conceivable that a collision could make the natural value of an electron’s mean lingering period on the excited orbit not be attained, but that the electron suddenly jumps back again and in so doing emits light. These, so to speak, forced acts of emission are, of course, always greatly disturbed the more frequent they become, the more must a line broadening consequently manifest itself. A second possible consideration is the following: Collisions of excited molecules should always extinguish the luminosity in that the energy is always expended in a different way. Now, one can surely gen- erally assume that, even excluding the Doppler effect, external disturbances, etc., each line has a finite breadth it is classically calculated from the radiation damp- ing the greater the damping, therefore, the shorter the emissive process, the broader the line. Something similar may also exist for quanta and, I believe, was once even assumed by Bohr: The quantum orbits are not absolutely precise, they have a certain width that matches a corresponding spectrum-line width. The stabil- ity of the electron on the excited orbit and hence its lingering period is less, the greater the scattering is against the actual precise quantum orbit. So, again, the larg- est broadening corresponds to the fastest decay whereas the transitions corre- sponding to the true center of the line, on average, occur the slowest. Thus, if every