Oskar Klein – Session V

Kuhn:

I think the thing I’d most like to start out with is really questions about Copenhagen in the period before the coming of the new quantum mechanics.

Klein:

Yes.

Kuhn:

Now, one thing I’m particularly interested in is the thing you mentioned in your first talk with us and then again in your remarks last Thursday afternoon. That is the dispersion problem — that Bohr had said something to you about this in your first talk with him in 1918. My impression is that what he then talked to you about was the inadequacy of the Sommerfeld-Debye approach, that is, the difficulty that arose when you got the orbital frequencies instead of the radiative frequencies. Now, what I really then wanted to ask you is what then happened to that problem in Copenhagen. Though there are lots of places to look we haven’t had time to look, there’s no evidence that I know of for a real continuing attempt to do anything on that problem in Copenhagen until close to 1924 and the Bohr-Kramers-Slater paper.

Klein:

I think you are quite right. I think this was always in Bohr’s thoughts, but he had so very many things, you know, his head was full of ideas; and the one came up and the other came up and there came new experiments so that I think this was postponed, partly because there were so many other problems to solve and partly because he didn’t think them right. But after talking about this in a general way, should I say a few words more about that? I think one important point in the background was the paper by Debye, who had tried to give a dispersion formula for the hydrogen molecules by making such mechanical calculation from a ring model. And Bohr stressed then very strongly that that could not be possible on account of the relation between dispersion and the absorption and emission lines, which he thought quite fundamental. And then in that connection, I believe that he already mentioned the possibility that energy should only be conserved statistically. I think that he had that already at that date.

I cannot be actually too certain, but I see a little bit of the road in connection with these ideas. But it is hard to be certain of such things. You may know from other publications, or there may be letters. He may have written letters about it. The next thing in this line which I remember had to do I think with a visit of Ehrenfest to Copenhagen. I imagine it was sometime in the fall of 1920, because I think Rosseland was there also. And where be mentioned some discussions with Einstein about the quantum paradoxes and where they were using the term “Gespensterstrahlung”, had you seen that? And I think that was again in this line. But, you know, Bohr didn’t like light quanta and photons and he tried to avoid them, so that meant that he was a bit in favor of regarding the radiation field as something still more imaginary than you would do it now; it was on a different basis than the atoms. Then I don’t remember any discussions about this theme. He mentioned of course quite often the quantum paradox and the difficulty of light quanta. I remember that at that time he stressed very much some remarks by Lorentz, who had pointed out that there was no way, which was consistent with what we know about superposition of light, to concentrate light energy into quanta.

To concentrate them would be to concentrate it in other ways as you do when you make sharp rays. So that would be against the superposition. Now Einstein hoped for a time, you know, to have something intermediate between the superposition principle and particles. But that never led to anything. He thought that there might be deviations from the superposition. But then I think that the real start came during the time that I was absent in the United States and Slater came to Copenhagen. Slater had some of his own ideas about radiation processes, and Bohr was somewhat critical of it. But I think that that inspired him [Bohr] to take up his own thoughts. I think that he often talked to Kramers about them. I don’t know quite if Kramers started it then or if he thought of it later; but, you know, an important point in Kramers’ work was that he really found a way to connect these ideas with a very beautiful treatment of classical mechanics, which he then translated by means of the correspondence principle into quantum relations. That really was a great progress in spite of the fact that the general ideas were Bohr’s. But Kramers made a great progress there. And that, I think, was what then remained after it was proved that the general philosophy of that Bohr-Kramers-Slater paper was not correct.

Should I say a little more about my next memory on this question? That was when I returned from the United States in the summer of ‘25. Then a little time after, Bohr invited me to come over to him at Tisvilde. I spent some days there in August, and then he said that already before the Geiger-Bothe experiments came out he had got very great doubts. I’m not certain, but I think the Geiger-Bothe experiments were known to him then. He had doubts about this theory and he meant that that was too cheap a way to connect things in quantum theory. He had some special arguments which I cannot remember, but they were connected with a paper he wrote then on (alpha particles). I think I looked up that paper without finding any explicit mention of this, but it must have been in his head at that time. Kramers was a bit depressed because he was really strongly engaged in that paper. I met him also at that time.

Kuhn:

In the Ehrenfest-Bohr conversations around 1920, was the Gespensterfeld brought out at all explicitly by the problem of dispersion; or was this more generally the photon problem?

Klein:

I think mere generally the photon problem. I don’t remember that the dispersion was talked of on that occasion.

Heilbron:

Do you remember any details of that? Was it just that the Gespensterfeld guided the photon?

Klein:

Yes, I think there was no detail given, but Einstein tried to defend the photons as something real. At the same time one had to have superposition and that was this paradox. That came out very strongly in this paper of Bohr-Kramers-Slater.

Kuhn:

Bohr himself told us in the few discussions we had with him of the conversation with Einstein in Berlin in 1919. I think he went to Berlin that winter.

Klein:

No! Did he go to Berlin in 1919? Because the other day I had some discussion with Lise Meitner and she thought that he came there in ‘21 first, but I believed it was in ‘19 or ‘20.

Kuhn:

It could be ‘20 rather than ‘19. A number of people have mentioned it, though I’m not sure that I’ve ever checked it against the letters here. Either ‘19 or ‘20.

Klein:

I believe ‘20, but she was very strong for ‘21. Could it be that he went twice? No, I wasn’t with him. I was with him later in Gottingen in ‘22, but I heard about it. At the time his complementarity philosophy was not formed yet. I think that at that time there was largely agreement between him and Einstein, although Einstein stressed the quantum more. What did you get?

Kuhn:

Well, you’ve hit now on exactly what was bothering me. He told us a story that sounds as though his complementarity philosophy was more developed at that tine than I have any real reason to suppose it was. He said that Einstein had said to him, “We must find out whether light is waves or whether it is particles—settle this issue once and for all.” And then he had said, “But that’s not something that can be done; how would you legislate the wave effects out of existence if it were a particle, or the particle effects out of existence if it were a wave?” This sounds somewhat as if he may have been reading back in his memory views that came later. And it, in any case, had a very different sound from what you have said about his relatively firm insistence that light must be waves.

Klein:

Yes. Now I remember that he stressed very strongly that the definition of light quanta had to do with the frequency of the light, and the frequency could not be defined except by experiments which proved the wave theory or superposition. I believe he told me a bit about this. And, I remember that in that connection he mentioned Lorentz; but I don’t remember if he at this or any time had met Lorentz. Do you know when he was at the Solvay Conference for the first time? I remember that he went there in ‘27. That was in the time of complementarity, but was there any earlier Solvay Conference to which he went? You remember, he wrote himself the history of the Solvay Conferences.

Kuhn:

I don’t off-hand remember. It’s easily checked. I would be surprised if ‘27 had been his first one. There was, I know, one earlier one he was to have gone to at which Ehrenfest then read his contribution because at the last minute he could not go.

Klein:

Yes, because I remember that quite often he mentioned that Lorentz had pointed out these difficulties of concentrating light; and that he said, in his own way of putting it, that the frequency was only defined through such experiments that one couldn’t say that light quanta existed independently of this. One could not say that he then had a really complementary view, although he had, so to say, the seeds of it. Because I think that he exaggerated, so to say, the nonreality of the photon. I think we would say now and also after the discovery of the Compton effect or after the Geiger-Bothe experiments that the photon is about as real as other things. [laughter.] So I think Einstein and Bohr together represented complementarity, and not Bohr quite alone. Although he had more, of course, of the philosophy; I mean, he always stressed the necessity of looking into the deep abyss between the classical way of looking at things and so on, while Einstein tried in some way to have a continuous development of the classical ideas. That was quite in his head already at that time and earlier. Do you think that he really thought that these discussions about complementarity—he didn’t use the name complementarity; I don’t remember if he did even in 1927—started so early? I think they were prepared in some way so early because there were several problems and he already had similar viewpoints.

Kuhn:

My guess would really be that, particularly because of the role that Einstein himself had in Bohr’s thinking later, that he was probably remembering that conversation as though he had expressed a view that really only developed in that form later; although it was certainly true that, as you say, there were all sorts of discussions including the fact that somehow or other physics had to learn how to support these incompatibles, or find a synthesis which would embrace them.

Klein:

Yes, I believe that you are right in that, and I think an argument in favor of it is that I remember very clearly from ‘27 when he came back from the Solvay Conference that he was really then very deeply disappointed that Einstein wouldn’t accept his philosophy. He believed that he would accept it because he at that time thought that Einstein had almost given it by means of his earlier work in quantum theory. So I think that he knew much earlier that [he and Einstein) did not agree on details, but I think he believed that their general view of the theory was always similar. I don’t think that he realized that Einstein really was opposed to the fundamental idea of quantum. I remember he was very shocked and rather disappointed at that time. He always spoke about Einstein with the greatest admiration as if Einstein understood everything. Then he saw that there was a limit.

Kuhn:

Your trip to America was very fruitful for science and America, but it makes problems for me. We keep coming up to the Compton effect, which really I think reached Copenhagen mostly after you had left for America. Now, my real question is: To what extent do you think that effect may really have been decisive in this change or interchange of attitude toward the photon? … This problem was no longer being brushed aside as much as it had been before say, ’22. I’m terribly interested in locating the source of that transition.

Klein:

Now, how does Heisenberg then explain the paper of ‘24 with Bohr, Kramers and Slater?

Kuhn:

He would say that that was an index of what I here have in mind. I don’t mean to say that by 1924 it wasn’t perfectly clear that everyone in Copenhagen believed in the quantum. But I still mean that sometime around this time, I think I mean in the sense you yourself have said, one recognized that there were problems in the area of needle radiation, and so forth; but somehow or other those would wait ‘til later. They were not yet at the forefront of attention in Copenhagen; people were doing other things.

Klein:

I think they were coming up more and more strongly in these years. Partly because there were several failures of this multiple periodic theory. And then Compton’s discovery and the interpretation of the Compton effect. And then there was a sudden attempt to reconcile the different facts by means of that Bohr-Kramers-Slater theory which was then to say that you have strict superposition for this field which does not represent the energy and only induces transitions in the atom, so that would mean that there would be no such real energy states in the electromagnetic field. And that led to the giving up of the strict energy principle and of the momentum principle. I think that these things were very much in Bohr’s head there. And I think also a further reason behind it comes later, in the year ‘25 - ‘26. I mean the beginning of ‘26. Then Bohr used always to point very strongly to the very paradoxical way of quantum theory when looked on from a classical physical point of view, so that he used such different symbolic expressions as that the quantum means a knot on [existence]. The knot may be moved, but it couldn’t be circumvented. And that was before Heisenberg’s paper on the indeterminacy principle came, so that he was always trying to get those things done and he was talking of them in very general terms. But I think that the concrete way of expressing them was missing. But I believe that they were very strong in his head just during these years, ‘24 – ‘25.

Heilbron:

Did he doubt at all the experimental findings of Compton?

Klein:

No, I don’t think he did that, and I think he was very impressed by the interpretation also. But you see then in this Bohr-Kramers-Slater paper the interpretation was not so different from that one given later on the point of view of wave mechanics; it was only the way of introducing probability that was different. The idea was then that the electromagnetic waves induced transitions of the electron and at the same time new waves were emitted. That is not so different from the way in the Bohr-Kramers-Slater paper only there it is taken in a way too literally that the waves didn’t represent any energy because they meant this probability. And we now know that we can start from the one point of view or the other, they are just different ways to express it. That became clear after Dirac’s first paper on the radiation theory. And, Slater also very soon wrote a little paper himself. Now I am going a little too fast in time, I think, because now I am at about the time when Bohr really began complementarity. In ‘27 Slater wrote the little paper where he independently had come to about that view—that was just regarding the problem about electromagnetic fields and photons. It would perhaps be good to find, that; I think it must be among Bohr’s reprints.

Kuhn:

We shall certainly look that up. That’s a paper I don’t know, though it has been mentioned on one or two occasions.

Klein:

I thought it was a rather impressive one, because Bohr had those ideas clear at the time, but he had no connection with Slater; Slater was in the United States then; but then there came this paper which was very much shorter and didn’t go into everything, but the main trend of it was very similar to complementarity.

Kuhn:

In talking about the problems that build up to make this problem of reconciling the two views more important by ‘24 than they had been before, you mentioned also the increasingly acute realization that you couldn’t handle, by the multiple periodic techniques, the atomic problems that had been worked on. I think in this connection particularly of the conviction of—I don’t know which paper it first comes out in so I’m not sure when to date it from—Bohr that the problems of the interaction of light and matter as in the Compton effect are the same problems and go just as deep as the problems of interaction of electron with electron within atom. Born points to this conviction very significantly and attributes it to Bohr. Now tell us about that. Where did that idea come from? What discussions do you remember of it?

Klein:

There I think the trouble is just what you mentioned. I was not in Copenhagen at that time when it was most mentioned.

Kuhn:

Do you remember any conversations toward that idea before you left Copenhagen?

Klein:

They came now and then, for instance this very early work by Rayleigh when he first gave the Rayleigh formula by using the electromagnetic field as a system of oscillators. You knew, of course, of that? Debye, I think in 1911 or something like that, made the derivation of Planck’s formula on that basis. If I don’t remember Ehrenfest wrongly, he mentioned in a conversation when he came to Copenhagen that he had already done that earlier, which was not known to Debye. Is there an earlier paper of Ehrenfest where he mentioned this? You know, I mention this also because conversations of Ehrenfest and Bohr became more and more intense after 1919 when Bohr went to Kramers’ doctors disputation in Leiden. That, I believe, was the first time he met Ehrenfest — at least they got to be very close friends then and had very many discussions. Ehrenfest always tried to go down on those questions which were unsolved while Bohr tried often to use some general words to just postpone them and so he avoided them. But then Ehrenfest wanted to go right to them always. This similarity between light and what one called matter was not very clear during those years, and I think that the main ideas came up from other places, for instance de Broglie’s idea about the waves.

Kuhn:

Well, I think it’s pretty clear that before de Broglie this notion that whether there were similarities between light and matter was in a broader sense —. But the question I’m looking for is that there is the similarity of principle between the problem of handling the interaction of an electron with the electromagnetic field or with light and the problem of handling the interaction of an electron with a second electron within the atom.

Klein:

But that means, if one speaks technically, that one should quantize the electromagnetic field in a similar way as one quantized the mechanical motion. Already at that time the point of view was known and appeared, if not through Ehrenfest, in a paper by Debye, which was also I think 1911. It was known the whole time. I remember myself thinking of it when I saw some paper about the Hamilton-Jacobi equation in a continuous medium; I think with respect to electromagnetic waves; I believe it was Hasenohrl. I never did anything with it; but I remember a plan to do it, which would have been then a quantization of the electromagnetic field and using this for calculating the interaction also, which then could not be carried out strictly at that time but approximately and which was, of course, carried out by Dirac much later.

Heilbron:

Was it connected perhaps also with the argument Bohr liked about the time and the definition of the stationary states? When one had the interaction of two electrons in an atom, the electric field would vary something like with the same frequency as an incoming light wave so that there would be some difficulty there. Was that at all part of the argument?

Klein:

I think there are some connections, but don’t know if one could say so. Bohr then regarded the atom, (the completed atom), as something in the stationary state when there was sufficient time and then electrons had to fit in that. I remember when he began those studies of higher atoms and the question of the Ramsauer effect and such things came in, then he thought that for instance if you have a lithium atom then you have the two electrons inside. Then he was already thinking of exchange possibilities. An electron goes in and collides as in the Franck-Hertz experiment and that was bound and the other electron was thrown out. But I think he regarded all this as more or less one process, but that analogous phenomena came in there as those which came with free electrons. But in this case, when it was a stationary state, then indeterminacy would only come in by means of interactions from outside which had not to do with the atom itself. I don’t know if this had to do with your question or not.

Kuhn:

Yes. Was the Ramsauer effect really a bothersome problem in Copenhagen?

Klein:

Oh, yes. The solution came after wave mechanics when Faxen and Holtsmark were in Copenhagen. Holtsmark had been studying it experimentally and thought a bit about it. And then Bohr had the idea that it could perhaps be explained in a similar way as interference phenomena which you have in a glass ball; then there are phase relations. Faxen, who knew the mathematical facts very well, and Holtsmark, who knew the phenomena, got together and wrote it up. I think they were asking Bohr about some (theme) and he proposed this collaboration. That gave a very important contribution to it.

Kuhn:

But in this earlier period, when there was concern about the Ramsauer effect, was it felt again to be one of those fundamental breakdowns, like dispersion, say, like the helium atom; or was it just an anomaly?

Klein:

I’m not quite certain. It was felt as something very difficult, and there were many attempts to see if one could understand it. I believe that Bohr would have classified it as a quantum phenomenon. I believe that he would have said that the whole Franck-Hertz phenomena were already a typical quantum phenomenon which one could really not describe quantitatively although one could account for the energy relation there. This (Ramsauer effect) contained a further feature which one could not account for in the usual way. But I think that he compared it a little bit, when he then came to the higher atoms, with what the electrons are doing on the inside of the atoms. I believe that he thought also of such possibilities as phase changing, but that was all with mechanical pictures and not with optical pictures at that time.

Heilbron:

When did the Ramsauer effect become a subject of interest to Bohr? It was two or three years after the effect was discovered, I think, when they first began to pay much attention to it.

Klein:

Do you remember when it was discovered?

Heilbron:

1920.

Klein:

Oh, I think he got interested in it immediately because in those years there was much talk about it. I met Hund last week; and I am reminded that in Gottingen when he was a very young man, he had tried to work out some model for the Ramsauer effect. He told us of it—I think not in a lecture but in a little talk he gave us. I have forgotten if Bohr was there. I was asked to be there with Franck, but there were only a few people present, and he thought himself that that was not correct. I don’t exactly know what it was, but it was an attempt.

Heilbron:

I wanted also to go back for just a minute to this very interesting thing you said about the lithium atom—that Bohr was thinking of the possibility that electrons should exchange there. Do you know if this was connected at all with these earlier papers of Lande and Born; were the electrons confined to octants on the surface of a sphere and then they collided at the boundaries of the octants with one another in order to maintain their complicated mechanisms?

Klein:

I don’t think I ever read those papers, but there were several people who tried models; Langmuir at the time had something of that kind and Lewis had some such ideas. I think Bohr read these papers and tried to understand them. I think that was about the period when he developed the structure which led to the explanation of the periodic table, but I think he was then rather opposed to such static models because he wanted always to have it connected by correspondence to classical mechanics. It was very difficult to have anything like these octets and things of that kind. In some ways he was impressed with Lewis especially, who could connect very much of the chemical evidence by this. I think this impressed him although he didn’t believe that one should express it in that way.

Heilbron:

You mentioned last time that Bohr, in those years about 1920, would talk even in terms of the regular geometrical solids.

Klein:

Yes, he did for a very short time. I remember one summer. That might have been inspired by some of these works. Had they appeared then? The summer of ‘20 I was not there; the summer of 1921 I was for a time up in Tisvilde at the end of the summer. It might have been then.

Heilbron:

What did he mean by these?

Klein:

Oh, that was speculation which he made outside of his real attempt. He knew of course from these chemical regularities of the importance of the number “8” and that in some way there were closed groups in the atom. That began to come quite empirically. But the real basis then was the behavior of the spectral terms, you know, which made him change the quantum numbers first for the sodium atom and lithium atom, and which led him gradually to the Pauli principle. That was the real thing behind it. But then he began to speculate if there could be symmetries among the orbits. In earlier years, of course, he had these rings where he wanted to place the electrons in the regular way and then they circulated round. You remember the later pictures of the orbits, after the things began to come out, were also nicely arranged but in the three-dimensional way. I think these were just little speculative things which were to be taken more symbolically. Also when they then came out in the other [three-dimensional] way, it was of course already also rather symbolic. The main thing then in the later work about ‘21 was to connect the chemical properties of the atoms with the spectral properties in the line of the correspondence principle. Of course he had very great success there. There was such a mixture of theory and the empirical which was very typical for Bohr’s beginnings always.

Kuhn:

This question that keeps coming up again and again in our conversations with you and with other people is that of the repeated entry—just how early it starts I don’t know—of the idea of energy non-conservation. One sees it in the Bohr-Kramers-Slater paper. You were quite certain that it was discussed in some work that was done with you. It comes up again with Dirac over problems of the electromagnetic field, and it comes up with the continuous beta spectrum and I think other places besides. Looking at some of the correspondence, I gradually developed the feeling that Bohr really wanted to get rid of energy conservation. It does seem almost a guiding principle-the first place he will go.

Klein:

It’s very curious; I cannot answer it. What I say may give some hints at it. One thing that we were speaking of before is that he didn’t want to think that light quanta were as real as atoms, and the backing for that was of course that they couldn’t be defined without the waves. Therefore he wanted to keep the wave theory as something rigorous for light. That I think was probably the beginning of it. Another thing must of course enter in—that energy conservation didn’t belong to those physical laws of which he had the very definite opinion that they were quite fundamental. That may, of course, be a little bit a matter of chance. There were certain things that he thought must be quite fundamental. I remember when he was talking, about Ehrenhaft’s attempts to divide the electron, he said that of course we had to doubt many things but we must believe something; and he believed that the electron was —. But apparently the energy principle for him didn’t belong to those necessary foundations.

I wonder if perhaps we would come nearer if we would talk about Pauli’s attitude toward this because you know that both times when Bohr came to doubt the energy principle, Pauli was very stoutly opposing him… Pauli was very strongly against it. I think he felt rightly very proud because the rest of us I think were rather doubtful; we had no very clear idea. Of course we knew that the energy principle was something practically very important; but if it should therefore belong to those things which were rigorously taken over into quantum theory, we were doubtful. I remember I took that paper by Kramers, Slater and Bohr—I got it while I was at Ann Arbor—very seriously. I remember I was invited to give a lecture about it by Langmuir at Schenectady and talked about it there and tried to give the arguments as they were given in the paper; but Pauli was very stoutly opposing it.

Of course I didn’t see Pauli then but heard it later both from him and from Bohr. Now I think that Pauli, because of his work on relativity theory earlier, was very much aware of the fact that the energy principle and change of the zero point of time and the symmetry aspect of that were quite closely connected; so that if you believe that you have something invariant, like the Lagrangian or Hamiltonian or something, then the energy principle means that the absolute time does not come in. Then, if one has that very clear and knows that in thermodynamics there is no special thing about time, although it might seem so, and all these things, then I think one would, as I did later when I really realized these things, think that energy and momentum principles would be the last things to give up. But I don’t think that these connections were uppermost in Bohr’s mind. If one had asked him, he would perhaps have known it; but for him, I think the energy and momentum principles were a very important empirical law though many important empirical laws, like most of classical mechanical theories, were limited. Therefore one would think that that was also limited—although the limitation must not be so that it came in contradiction, with the ordinary experiments. That would be the same problem as lay behind the correspondence principle, that of agreement in the limit. Now I think there came up other doubts also.

When you took this statistically, I think you might have had fluctuations which really would be apart from the fact that you could observe single processes like Geiger-Bothe did. I think it has been pointed out sometime and I believe there would have been difficulties with fluctuations also. But that was not clear. I think that it was not that he wished to give up the energy principle, but the quantum paradox pressed on all of us to give up something. It was not quite clear what one had to give up, and he would not give up superposition with regard to light. Then later on with the beta rays…he tried to see that one could not go so far in respect to classical ideas. I think that at that time he did not realize that quantum electrodynamics would be so strictly valid as it is. He thought of a much earlier limit; I remember many discussions about that. The limit was of this description where you took the relativistic problem seriously and especially—and that was the main thing—you took the electric charge seriously, and thus the test bodies would always contain an integral number of elementary charges. Then he thought that that would lead to another kind of indeterminacy; and I think that was the background for his again coming to doubt the energy principle, apart from the empirical difficulty with the beta spectrum.

Kuhn:

How early did he discuss this question of the discreteness of the charge in connection with quantum electrodynamics?

Klein:

I think that was in the late twenties. About ‘29, ‘30 or so that began to come up in connection with Dirac’s theory and, of course, this hole theory idea, which at first seemed rather strange. I don’t think he believed in that during these years.

Kuhn:

You know, on the energy non-conservation argument, there is one very likely step to it or connection with something else that Bohr believed in. I wonder whether he ever brought this out explicitly. This is just the pre-complementarity argument about the complementarity of time and energy. If the interval of time available is too snail, then the energy itself is not well defined. Now, this would certainly suggest that for processes which occupy only a very short time, that in some sense one cannot define energy within this time interval; so that is only then in some sense over longer intervals of time, which would mean statistically for the processes occurring in the interval, that one comes to a point where the energy itself is sufficiently defined to give the conservation a sharp meaning. This at least would tie parts of the argument together. I wondered whether Bohr had in fact ever so tied them together.

Klein:

He was very clear about this limit in the definition of energy, but I don’t think that came very explicitly into this question because the beta decay period is very long; so that everyone should think that if there were no other reason, the energy should be defined. I think the point in question was rather that bodies are constituted by charges. There is, though my memories are not certain, a lecture which he gave at Como; and I think there he tried to put in what he at that time could. By the way, do you remember he wrote a letter to Nature, which he drew back again, about this; and that must exist, of course. Have you seen it?

Kuhn:

I haven’t, but I think it does exist.

Klein:

I think there he puts it as clearly as he could at that time. Then of course Pauli wrote to him about the neutrino. You heard that the other day about how Bohr did not answer—Weisskopf told that I remember also that Bohr either showed me or told me about that letter, and I remember that I teased Pauli a little bit because at that time Pauli used to say that experimental physicists only try what theoreticians send them on, and I said it would be irresponsible of him to bring out such an idea with that background. [Laughter]. I don’t think we took that very seriously at that time when it first was brought up. Then of course afterward when it entered that Pauli was very clever, and that he always kept the energy principle was right. But I believe that with Bohr it was more that he had no such strong idea about the energy principle. He had a stronger idea about superposition.

Kuhn:

When you say superposition here was that the phrase that Bohr himself used from an early time; or was it simply interference that they spoke of?

Klein:

Oh, I think he used both, but it was in optics. I don’t think he had thought of it in connection with mechanics.

Kuhn:

There is a sense in which the whole phraseology of superposition tells more about the structure of the theory than simply saying we must have interference. This was one of the reasons why I asked whether they used the word superposition.

Klein:

One used of course in those times the superposition principle in optics, and I think he sometimes used it and sometimes pointed directly to it.

Heilbron:

You mentioned experimental physics just a moment ago, and there is another line of questioning. Bohr seemed to feel a little suspicious always of experimentalists and experimental results. Is that a true statement?

Klein:

Of course he was in special cases, but he had a rather close connection with experimental physics from the beginning so that he thought of himself often as an experimental physicist so that when mathematicians began to talk with him about exotic mathematical formulas, then he liked to point out that he was only an (experimental) physicist. So I think that his doubts were more of the special kind than of the experimental physicists themselves. Of course with his theoretical background, he might have stronger arguments for doubts; you know, special examples, as hafnium for instance.

Heilbron:

The reason I asked was that in one of the letters I think he wrote to you urging you to go to Ann Arbor, or in support of your application to go to Ann Arbor, he points out that it is fine for you to go because American experimental physics is so good. I was wondering whether this was a generally held point of view.

Klein:

Oh, I think he meant that theoretical physics should keep as close a contact as possible with experimental physics and he’d try to do it.

Heilbron:

He particularly recommends—

Klein:

Oh, you see the background was that I had no position at home, and I asked him if he knew something perhaps about England where I could go. I wanted to marry at that time and had no position. He said that Colby had just been to Copenhagen and asked him if he could have someone; but then he thought that I would and could stay home because I had gotten the docentship at Lund; he didn’t know that there was hardly any possibility of getting anything but very little pay for that. He didn’t know that; he thought I would stay there; and so he hadn’t said anything. Then when I mentioned this, he said that England would probably be difficult, but he had just got this questioning about America and he would gladly recommend me for that. Then I think he wanted to say the nice things, which weren’t bad, of course, rather (???) All that was very kind; they made very good things there. You know in Lund I had had a little to do with Erik Hulthen, who also did band spectra work; and I had been a little bit interested in that. In Ann Arbor such things were carried out on a greater scale than there.

Kuhn:

Who did you find in America to talk with about the sorts of problems you were interested in? Was it a time of isolation for you?

Klein:

Oh, I tried to talk a little but there were not many. You see, Dennison was there—quite a young man. He took his doctor’s degree, but I think I may have talked a little with him. I remember once Ehrenfest came on a visit; and we had very interesting talks about things; but on the whole this came a little at the side of my regular work, which was then to teach. They were very kind to me; because I had to teach first a general course on the quantum theory, where only the people of the institution came. Then the next year I got some elementary teaching also, but at the same time I gave a course in electromagnetic theory which helped me very much because I had to learn things. So these more speculative things I had rather at the side. Then Bohr came there, and I had just done a work on crossed fields. I think I told you that the idea behind it was that I wanted to get behind the molecular things. I wanted to be of some help with them, and I knew very little of molecules.

Kuhn:

The very fact that you say, “Well there was Dennison, but he was very young and just getting into it”; and then that the other two people you mentioned that you could discuss problems with were both visitors from Europe indicates that there really was not much contact to be made for you.

Klein:

I tried more to interest myself in their problems; but since I had a very slight background there, it didn’t work. I don’t think I was much help. I was a little help to Dennison. I got into this crossed field work to see if I could understand something more behind these questions of molecular spectra, which was their main subject. Now the time was rather short, and I got into very many specialties of my own.

Heilbron:

In view of the fact that Bohr was, when he came, so different from the surroundings in Michigan; I wonder if you might have preserved any clear memory of the kind of things he was saying about quantum physics then—where it was going, the troubles it was in?

Klein:

When he came in ‘23, there was not very great change. It had only been a short time since I had seen him. But then he had been most of the time at Amherst College, and Frank Hoyt was with him and assisting him. He went to Chicago; and so I met them both in Chicago; then he came to us in Ann Arbor. But I think it was just the spectral problems. You remember, he gave in England the Guthrie lecture, and I think that was shortly before. Do you remember the date?

Kuhn:

It may well have been ‘23.

Klein:

I think there were just these somewhat advanced spectral problems; and there were many puzzles also about both the spectra of higher atoms; and he had got a very good way to find them, to give a theoretical background for this more general Rydberg formula of the terms in spectra. And then, there were all these questions about the magnetic fields and the anomalous Zeeman effect which were very puzzling at that time. And then he was very interested in them also. But they were far from clear at the end of ’23. But then Lande had of course contributed very importantly. Then Pauli also, I think it must have been about that time or even the year before when Pauli told the story which I think I also heard from Harald Bohr. You may have seen that Harald Bohr met Pauli in the street—[General assent.]. So these problems were going on very strongly in ‘22, ‘23, ‘24, and led gradually to the Pauli principle in ‘24.

Kuhn:

Was Bohr already, when he lectured in America, talking much about Unmechanishe Zwang?

Klein:

With the core? I think he was talking about that also. I’ve forgotten if it was not even earlier. I rather believe it was already a little earlier that that started—that the inner core should in some way behave in a way which didn’t fit the Larmor theorem, which was very strange.

Kuhn:

I wonder about the fact that you were then only in America for two years. Had you intended to come back when you could have a job from the beginning?

Klein:

That was for more practical reasons. I had hoped always to get a job at home sometime; and then when Bohr was over in Ann Arbor at the end of ‘23, then I talked with him about it and very, very kindly he got into connection with the professor of mechanical and mathematical physics in Lund, (Ekman), with whom I had been docent for a term, and asked him about possibilities there. And then, he said if I wanted to stay near Copenhagen; I could have a fellowship there; so that I think the plan was that I should come back. At that time I was only appointed for one year at a time, but I think the idea was that I would probably come back [to Michigan] but have a stay in Copenhagen either for a half-year or for a year, I’ve forgotten that. But then, after I came back—I made a little visit to Copenhagen as mentioned—my wife and I went up to Stockholm where my mother lived just before going to Copenhagen. While we were there—very soon after—I got some kind of stomach influenza; and that developed to a very long jaundice, which took several months so that the whole was postponed; and I got the fellowship then for a later time; so that we didn’t come back to Copenhagen until March, 1926.

Then I still didn’t feel very well. Then all those things came rushing in—Schrodinger and everything. I had had ideas myself in that line, so that everything became very difficult. A little before that time Randall had very kindly asked me to come back and take a new position for three years, but I didn’t dare to accept it for three years. I said that I would come back next year. But then I felt it rather difficult there, so that after a time Kramers understood that, and he must have mentioned it to Bohr because Bohr came and said that if I wanted to stay next year I could also do it. I felt badly because I thought that I ought to go back to the United States but then it was settled that I would not go back at that time. And then it went one year after another; then they got other people to Ann Arbor, partly through me. Colby came over in the summer of ‘26 and visited us in (???) and then I just had been in Leiden, where I was invited for a month. I met Goudsmit and Uhlenbeck there and thought that they were very good and very nice. When he was asking about people, I recommended any one of them, and then he engaged both of them and also Laporte, so that they got a very good group. Then I felt better because first I felt badly that I had not followed what I had said about coming back. It was really very difficult.

Heilbron:

I wonder if you might revert for just a moment to some ancient history. In our first discussion we had, I think, you talked about a visit that Sommerfeld paid to Copenhagen and some lectures he had given at Lund in 1919.

Klein:

It was in Lund. I don’t think that be came to Copenhagen; but it was in Lund; and the Lund people invited Bohr and then also the rest of us who were working there; so that I went with Bohr. They were very interesting works. Then already the problems of the anomalous Zeeman effect were very acute.

Heilbron:

That must have been quite a new thing though.

Klein:

Yes, but it was discussed. I remember that Sommerfeld gave a lecture about it, that Sommerfeld made some extrapolation of the terms so that the first was so, the second so, and he thought the rest would be so. I remember I was sitting beside Bohr, and Bohr smiled a little and said to me he didn’t believe that.

Heilbron:

Were there any other, or any direct exchanges between Bohr and Sommerfeld that you recall about the quanta.

Klein:

Yes, I don’t know if it was mainly then or mainly perhaps later—because he came to Copenhagen also—but I know that very friendly relations came about between them. You know, Sommerfeld had made some very important progress with respect to quantization and especially application to x-rays. Part of this was already done but not published by Bohr—about (series) spectra and so—and Bohr showed Sommerfeld a paper which he had never published; you know about that. Then Sommerfeld very strongly said to Bohr that that should be published together with his collected papers, which I think were at about that time. There was a man with the name Stintzing who asked him if he could translate his papers into German, and they came out. Do you remember the year? It must have been that year.

Kuhn:

Twenty or ’21.

Klein:

Oh, it was as late as that. So it might have been then at a later time when he met Sommerfeld and this came up. But I know that they got very soon into very friendly relations although Bohr was rather critical against certain of Sommerfeld’s ideas, and it took some time before Sommerfeld came to realize that the correspondence principle was important.

Heilbron:

Yes, we wanted to discuss that. But, before we do, I think you mentioned in that talk that Bohr had met Sommerfeld in Munich in 1914, or—?

Klein:

I don’t think so. It may be so. But I don’t think I knew it. Maybe he mentioned this himself in his historical paper. I believe so because it sounds familiar to me.

Kuhn:

I don’t want to cut off that line. I wanted to ask one thing about the Bohr-Sommerfeld relation which has been very puzzling to me but seems to be extremely well documented. I talked with Gerlach, and he told me the following story. We’re aware somewhat of the differences in approach between Bohr and Sommerfeld. This is an example of it, except that it seems to me to be terribly hard to understand. Gerlach showed me a letter that had been written to Edgar Meyer; the first part of it was written the day the apparatus was all set up; and the last part of it was written the next day, when the apparatus had run for the first time.

Klein:

You mean the Stern-Gerlach experiment. I had forgotten the date; it was made, I think, when Bohr came to Gottingen in the summer of ‘22.

Kuhn:

Yes, in fact the paper is published relatively early in ‘22, and I think this letter is actually still in ‘21, I think it’s before Christmas. Now, the first time they ran the apparatus, they didn’t get any separation; and this led Gerlach to put in the letter, “therefore Sommerfeld is right.” Then they refined the apparatus. By this time Gerlach had spared himself, had gone to Rostock and then they did get separation. And Gerlach said, “I remember this very clearly; as I ran this refined apparatus for the first time and did get separation, I immediately said to the graduate student who was working with me, ‘You must go immediately to the telegraph office and send a telegram to Stern, saying, ‘We’ve run it again; Bohr is right.” The strange thing about this is not that there had been a difference between Bohr and Sommerfeld about the Stern-Gerlach experiment, but that it should have gone that way. It seems just the reverse of what one would expect in terms of what one knows about their attitudes toward space quantization, which was much more real I think for Sommerfeld than it was for Bohr.

Klein:

I am very astonished also.

Kuhn:

There seems to be no question that this is right. There is contemporary written evidence for this. Yet nobody I have talked to has been able to say what would be the difference that would make “no separation” the Sommerfeld and “separation” the Bohr interpretation.

Klein:

That’s very curious.

Heilbron:

This letter, the documentary evidence, could not have interchanged the names?

Kuhn:

If there’s no other explanation, one may finally be reduced to that, but I think it seems very unlikely.

Klein:

Have you asked Heisenberg?

Kuhn:

Yes, and he is just as surprised as I am. He says he’s got no idea, but you in a sense were with Bohr. You were with Bohr during the years before Heisenberg got here and therefore might have direct evidence on this.

Klein:

I cannot see—Franck might have known. Have you asked Franck about it?

Kuhn:

I didn’t ask him about this. No, in general, he remembers very little about the experiment and its reception.

Klein:

He might know about this. And have you asked Stern?

Kuhn:

I’ve asked Stern about the experiment, and his recollection of what was expected and so on and what came are very much less detailed than Gerlach’s. I have not asked him this question in particular because I didn’t have this question to ask when I talked with him.

Klein:

You see, Stern had very deep theoretical interests already at that time; you know, he studied with Einstein earlier. So I would rather believe that—and Bohr was a really very close friend of Stern. I remember when he came to Gottingen, Bohr was so happy because he liked him so very much. He had no close personal relation to Gerlach, so that I think that Stern would probably be the best one to ask. I heard from Mrs. Pauli that she met him recently; I don’t think he’s at Zurich now.

Kuhn:

He may still be in Zurich at the moment; he lives in Berkeley, so we will have the opportunity.

Klein:

Oh, that will be easy for you. Now I must say (???) in those times, but I am particularly at a loss to see why Bohr and Sommerfeld should have different views on that point.

Kuhn:

Well, I could understand their having different views if those views were the other way ‘round. I would expect them actually to have the same view. But if Bohr had thought that this would not work and Sommerfeld had thought that it would work, that I could also understand. But to have them crossed in that way I’m finding extremely puzzling.

Klein:

Yes. You mean because Sommerfeld, already with the hydrogen atom, thought that there would always be quantization of that kind while Bohr thought that the quantization was only there if you had a field. (I would rather believe that both, and that if there would be any difference, it would be on the other side.)

Heilbron:

It might be something about the experimental possibilities of the situation. If one, say, expected three instead of two lines, then perhaps then one might have thought that they would be just too smeared to distinguish if there were really three. Maybe Bohr’s greater feeling for experimental possibilities might have made him distrust the possibility.

Klein:

Yes, there were very soon discussions about the time (of change). I’d think Stern particularly might know. Bohr discussed that very much with Stern. There were experiments made then. One person you might ask also, although he came later into the method, is Rabi. Rabi was also a theoretician really when he started the things. He came into the experimental things, so far as I know at least, because he came to Pauli; and then he met Stern. But that is very strange. Bohr then used this very much, you know, in later times also, to define the stationary state. So he was very, very interested in it.

Kuhn:

It may just be that the names are switched and that this was Gerlach’s misunderstanding of the opposition. There are ways out, but it’s still striking enough that I’m still looking for another way to make sense of it.

Klein:

But still I would believe that Bohr also would believe in the positive results at that time. I think he regarded this as an important proof of the general ideas of quantum theory, as far as I remember. I don’t think he was startled by the result in any way; I think he was happy with the result.

Heilbron:

Perhaps this would be a convenient place to ask a question which occurred to me in reading over the transcripts. You say there that you and Bohr met Pauli and Heisenberg for the first time in Gottingen—that you thought more of Pauli’s and Bohr thought more highly of Heisenberg’s abilities. Then, when Pauli afterwards came to Copenhagen, he was quite surprised to find that physics was much more difficult; the quantative method was much more difficult in Copenhagen than it had been in Gottingen.

Klein:

You see, he has told that story himself. You remember where he wrote and told that Bohr and I had laughed at him when he said that, and he didn’t understand why, but he understood it later.

Heilbron:

I wondered if you could supply a little more detail in both aspects of those remarks. Why, for instance, was it Pauli that you felt—

Klein:

Why, I saw more of Pauli so that may be the main reason; and I saw much more of Pauli than I saw of Heisenberg at that time; and Bohr had a very long talk with Heisenberg. I think I saw somewhere, I think it was the newspaper yesterday; and it corresponds quite to my memory also, that Heisenberg said that he had uttered something in the discussions where he doubted something that Bohr had said. I’d forgotten that part, but he said that himself. But then afterwards they had a very long walk and talk, and I remember when Bohr came back from that he said to me that Heisenberg understands everything. He was really impressed. And I had seen little of Heisenberg then so that I had no negative view at all, no especial view there; but I was very impressed by Pauli.

Heilbron:

What particularly bothered him about the Copenhagen difficulties?

Klein:

Bohr suggested that he come to Copenhagen. It may be that I had asked Bohr about it first; I don’t remember whether it was that way or if he thought of that himself. At least I thought it would be a very nice thing. Then we were both there, and Bohr said that to Pauli and asked him if he could, and Pauli said that he would be very happy to come, but he thought it would be difficult because of the Danish language—physics wouldn’t be difficult, he said. Then he said himself that Bohr and I smiled a little, and he didn’t understand why. But then he came into the anomalous Zeeman effect and all these things where he found very much that he didn’t understand.

Heilbron:

He had not begun to become involved in anomalous Zeeman effect before he went to Copenhagen?

Klein:

He might have thought a little about it; but that became, so to say, his main problem; and he took that so seriously that he was quite impressed that he couldn’t find the solution to it at those times.

Kuhn:

When he said, “language may be difficult,” do you think he meant simply the Danish language? He did not mean the Copenhagen way of discussing problems?

Klein:

No. It was the language. He was very proud then that he learned it. I remember that he came there sometime in the autumn of ‘22; and then I spent that time in Stockholm. From February I came to Lund, and then I came over to Copenhagen sometimes and saw him, and he came over to Lund once. But then, the summer I spent in Denmark; and he had been at home on summer vacation; and I met him—I cannot tell the exact time—at the station when he came. Then suddenly, just at the first he said, “Ich glaube dass ich jetzt besser Danisch spreche denn ich habe inzwischen daruber nachgedacht.”

Kuhn:

I have one other question that grows out also along these last lines from the thing you said about the Gottingen ‘22 meeting. You reminded us quite rightly that this was just after the Ladenburg work, I think before the Ladenburg-Reiche formula.

Klein:

I think he already had something. He was there and told something about it. I’ve forgotten if he had published it.

Kuhn:

Yes, I think the first of the Ladenburg papers on this subject was out. Now, the question I want to ask you is: You said also, and again I think quite rightly, that the Bohr-Kramers-Slater paper is rapidly discarded insofar as it’s about non-conservation of energy. The idea of virtual oscillators and of extending the correspondence principle by use of virtual oscillators is something that comes out of that paper that is terribly, terribly important. The idea of virtual oscillators people could have taken from the Ladenburg work also. My impression is that they didn’t, that it was only after the Bohr-Kramers-Slater paper with its virtual oscillators. I wonder a little why they didn’t take them a little more seriously when Ladenburg used them.

Klein:

That’s an interesting question. Now, one difference, as far as I remember, between Ladenburg’s and Kramers’ was that Kramers had this mechanical correspondence treatment while Ladenburg just took it empirically in the way which Bohr had done in his own; so that what Ladenburg did was experimentally very interesting to Bohr; but I don’t think there was any idea which was foreign to Bohr. That’s a little aside of the question, but I wanted to mention it because Bohr had that in the old times already. But then there were these very nice experiments and the way to connect them with a theory which Ladenburg made, and I think that Bohr was very happy for that. But Kramers went, of course, further when he developed this correspondence treatment where, in the mechanical action-angle variables, he showed how these amplitudes came; this was then so important for Heisenberg.

Kuhn:

That is not quite what I mean, because although certainly the Kramers dispersion formula is a correspondence principle, he derived the formula in a way that Ladenburg would not at all.

Klein:

I realize that I didn’t answer your question; because I only wanted to say what I thought was the main difference between the two. But the question—I really cannot answer. The theoreticians would have been of course more impressed by Kramers’ because that gave a theoretical connection which very soon gave very important results. You remember that Pauli drew the conclusion from this that one could treat the second order perturbation by means of an electric field of an atom and get a rigorous formula—it was one of the first rigorous formulae in quantum theory. So I think it was natural that theoreticians thought more about Kramers’ paper than Ladenburg’s. But I think that Ladenburg was always quoted in those early papers. I think that Bohr and Kramers, too, quoted Ladenburg, and it was clear that be played a role in this.

Kuhn:

Let me direct your attention for a moment away from the Kramers dispersion formula and back to the Bohr-Kramers-Slater paper in which the idea of the Virtual oscillator and of a correspondence principle approach involving the virtual oscillator appear. Now, my feeling is that after that paper, in spite of the Geiger photon work, people do try very hard to use virtual oscillators in formulating the correspondence principle. The Kramers dispersion formula is only one example of this. Now, somehow or other they might have been started a little earlier trying to formulate this with virtual oscillators if they had taken the Ladenburg work more seriously. So it’s about reactions to Ladenburg’s use of a formal virtual oscillator and why it was not pushed further than it was—

Klein:

Don’t you think it had more historical than logical reasons? When Heisenberg came to Copenhagen, the Kramers work was very new; and he got into close connection with Kramers and they even published the second part of the paper together. Then the next really important step was Heisenberg’s generalization which was quite based on the Kramers work. I think that Heisenberg must have known Ladenburg’s work because he was also in Gottingen, but I suppose that he really came into the things from Kramers. It may just be historical chance that none of them took it up independently of Bohr and Kramers, just from Ladenburg as they might have done. Is there any connection between Ladenburg and earlier things by Bohr? I don’t believe that there is much because I don’t think that Bohr ever published anything on these earlier ideas. It would perhaps be nice to look up Ladenburg’s paper to see if he mentioned Bohr in it in some way. He doesn’t? No. Then he certainly came to it quite independently. So I think you are quite right that if Kramers had not been there, people might have started from Ladenburg. Maybe Heisenberg might have made the correspondence treatment which Kramers did.

Kuhn:

Well, I asked Heisenberg a question of this sort. Heisenberg said, “Well, you know, the Ladenburg paper was not a correspondence principle paper at all, and it was much more like the Einstein approach, which was not being taken very seriously.” He thought that for that reason nobody had paid very much attention. Really it was lumped somewhat with an Einstein-like approach, was off the main route and resisted by some people. And then, although they were interested in experimental results, it did not seem to be theoretically useful until one saw how to do it as a correspondence principle treatment, which people did not try to do initially because they just didn’t see that as a way to handle it.

Klein:

That’s rather near to what I believe, except I hadn’t thought of the idea that Ladenburg s could have more to do with the Einstein view. How would that connection be? Ladenburg’s looks very much like wave theory?

Kuhn:

I think, though, it was simply that it was meant that this was like the Einstein view. I’m not entirely satisfied with this for somewhat the same reason, which was one reason I hoped to get you to speak to it. I think what Heisenberg had in mind is that it’s an un-analyzed Einstein coefficient.

Klein:

Yes, so that it was not connected to any mechanics. Yes, I quite agree. I thought of Einstein’s light quanta but you meant un-analyzed coefficients. Still it was a very important step. Now the development went so fast, but if there had been no correspondence treatment,… that step might then gradually have developed into this which Kramers made. They now all came so swiftly one after the other, and there was the other source that Kramers knew these things from Bohr, of course…But it I think for Bohr it would have been an alternative. I think that one reason was that things went so swiftly at that time. But in some ways it should have been almost as impressive as the Einstein coefficients themselves, and they were of course very impressive. They meant very much to Bohr.

Kuhn:

My next question is a very general one and of a very different nature. We have, of course, talked to a number of people about this, but none who worked with Bohr as early as you did. I wish you would sort of sit back and relax and tell us just what it was like working with Bohr. Talk to us about Bohr, about working with Bohr, about the techniques of writing things out for Bohr, about the nature of the sort of collaborative efforts that that involved. I know of no other working relations quite like the relations of Bohr and his temporary assistants, and I’d love to hear you talk about what that was like.

Klein:

Yes. Of course I witnessed very much of that and Kramers told me much, because you know that Kramers was an extremely good assistant to Bohr, like Rosenfeld in later years; and he had the gift to enter into Bohr’s ideas swiftly and with enthusiasm and love, you can say; and then at the same time he had a very natural way to react a little critically in a kind way immediately; and I think Rosenfeld has a similar way. So they could further Bohr’s work in a very important way. Then when I came—Bohr meant very much to me, and I think he thought at some times that I could help him a bit, but I could do neither as Kramers nor as Rosenfeld so that there was much less of direct results when I was his assistant than when Kramers and Rosenfeld were. Now, it had a little to do with the problems and how things were. You remember that Kramers came first. Then Bohr was in the situation which he stated in the introduction to his collected papers when he writes about that unpublished work—where he came into troubles because there began to come so much from Sommerfeld and other people, and he tried to catch up on that, and it took him time. But that was just the state in which Kramers came to him. You know, Kramers was there—he came—right after he had become professor in Copenhagen; and Kramers was then, his first assistant. They were working very hard and with very great concentration on this, and he had just gotten out the first part of these big academy papers when I came and was working on the second. The third and the fourth parts, though, didn’t come out in quite that order. Part 2 was almost ready at that time and came out soon. Are you thinking now of the more human side of this?

Kuhn:

Well, really of both. You’ve said already a great deal, and I’d be very glad to know anything you can tell me about the way the two of them worked together or about how you worked with him. I’d also be glad if you’d say more about the respect in which you think Kramers was able to make a sort of contribution that you feel you were not able to make. Elaborate on that.

Klein:

Yes, I’ll try to gradually perhaps. The main way he did his work, as you know from what Courant told us the other day also, was trying to express the ideas. For him writing was a very difficult thing; and he always used to say that for him form and content were one, and he could never say that he understood a thing before he had expressed it. But that meant again that he had a very great difficulty in expressing some things. I don’t think all of these difficulties were deep; some had to do with just more superficial things; but some were very deeply connected with the fact that his way of working out the ideas was trying to express them. And then he thought; perhaps he was a little (???) with the idea than he really was; and then he tried again. So very much of the work I witnessed, where I tried to help him, began with his starting to talk a little bit about the things—he used to go round and round—and then he began to dictate so that one sat with a pencil and a block and he dictated and one wrote a few words. And then, after he had come a little way, he began to look at what was written, began to say, “no, no, that doesn’t do,” and then started again.

The great problem was when should you make suggestions and when should you not make them. I think that was a “phase” problem because sometimes (you did harm) when you started to make a suggestion; sometimes you could help. But I think that Kramers and Rosenfeld were very helpful in finding the right suggestion at the right moment, and that was very difficult because sometimes that would lead to the other thing. Then also when he got into difficulties, suddenly he stopped and began to talk about the things. Then of course the assistant had to try to think about the problems also and began also to talk, and then the question was if the assistant had really understood him so well that what he said would be helpful or not. Sometimes it might; sometimes it might not at all. Then when he got tired he said, “We’ll go for a little walk.” And he’d begin to talk about the problems during the walk. If it was in town, he went and bought some chocolates or bought some small cigars or something like that; and there were always little interruptions of that kind. Then he started again. When the thing came nearer to publication, it was of course very, very difficult.

In the beginning—the first year I was there—he had no secretary, so that sometimes he asked me to copy out the writings so that they could be read, because the first sketches were very difficultly written. Sometimes, and especially in earlier years, Mrs. Bohr had copied very much for him; and sometimes his brother Harald would copy; so that he had lots of manuscripts which I think must be copied with, different hands. Such a copy was usually only a sketch of a work, and usually that was (???) all over and started anew again. I hardly know any work of his which wasn’t written very many times in that way and often with a great development of the ideas in the meantime. It was Franck who told that nice story about H. M. Hansen telling him there was a young man who didn’t publish his papers until he had been writing at them for several years. That was earlier than I knew him. That must have been much earlier than Franck knew him because Franck knew him after me. I think it must have been around 1912 or something. The external side of it was often very funny; I remember once when I had my family out to the country and he accompanied me to the train at Osterport. He had something he began to think about. We had some paper, and I was standing there. There was a system then where the conductor would collect tickets at the head of the stairs to the platform. We stood up there, he dictated, and I had to look with one eye down the stairs to see if the train had come, and as soon as the train came I had to run down. I remember once when we were working on the Planck celebration paper. That was ready then also very shortly before the last moment. We went together to Hovedbanegaard because there the post office was open until 12 o’clock in the night. It must have been a little before 12 o’clock when we got there. I think it was really very little before so that we ran there. A student from the institute came up to us, now the mathematician (Lauritsen), who must have known this, but he said he saw two people running and he got curious to see who they were.

Kuhn:

When one’s job was officially that of taking dictation and there was talk, would Bohr suddenly stop and say, “Now write this down,” or would you try to write down everything that was said? Did he select things that he wanted—

Klein:

No, he rather made a difference, [made it apparent], when he only talked. Sometimes one would try to write, and he would say, “No, no, not now.” He used to make a difference between dictating and talking.

Kuhn:

You would sometimes start to write something down and he would say, “No.”

Klein:

Yes, so that the dictating was always meant to become a manuscript, but it was always very long before it came. The worst I had was in the summer of ‘27 working on complementarity. We were out in the country then, and I came every day to him, and he began to dictate. The next day that which he had dictated was not used, and we began over. But the background there was peculiar. I think he was really very overworked, and of course that whole year before had been very strenuous. And then, in the spring, after Heisenberg’s paper, he had got his main ideas about complementarity; and then he tried to write them perhaps a little too early. Partly he was very tired and partly the ideas were still not at the stage to be written, so I think there were several reasons why he didn’t succeed there. Did I tell you about that this past winter, how that happened then and how this paper was finally then published in Naturwissenschaften? There is a story about that.

Kuhn:

You told something about it, but perhaps not the whole story.

Klein:

I mean the outward story because when this happened it was, I suppose, one of the optimal occasions of his writing many manuscripts; and we were starting manuscripts after proofs and such. And finally Harald Bohr got him to write out a little short manuscript after that summer before he went to Como, but then he draw that back again. Then he came together with Pauli later at Como, after the meeting I think. He got then a better manuscript, and that was the basis for the paper. The first was just a short summary because he couldn’t—

Heilbron:

Was Pauli good at this, too? Was he as good at this amanuensis business as—

Klein:

He was very good. Perhaps he was not so good, or wouldn’t have been so good perhaps in the long run because I don’t think he would have been patient enough to do this as for instance Rosenfeld now does it. But he was very good; and on a special occasion like this when they were together, he grasped the main ideas though they were then rather new to him. He had talked very much to me from the beginning, but now he had a new reaction on them which I think was very inspiring for him. Pauli would also be good as an opponent as soon as things were not clear. I think that I myself was mere easily persuaded to think that perhaps things were as good as they could be at present and would see if one doesn’t come deeply into the things with time, finally. Pauli had the right impatience, I think, on certain occasions.

Kuhn:

What room was there for creative contribution in this role as assistant?

Klein:

Oh, Bohr got quiet. Bohr was very kind when one came up with an idea. Sometimes one happened to come upon an idea which was wrong and which he believed first; and that was the worst situation, of course; and then he discovered, or somebody else discovered, that it was not right afterward; and then one felt rather badly. Sometimes one had an idea which was not so wrong, but he didn’t accept some of them. But sometimes there was a correlation, and then he was always very kind and very helpful. There was perhaps a little too little time to work things out because one was kept busy there. So I made most of my papers during my later time in Copenhagen when I was there in those five years beginning in ‘26. Then most of the papers I wrote myself were when Bohr was away; but of course much of it was thought out and discussed when he was there; and he took very kindly interest in it. As a very strong personality, of course, he could not leave off his own; or if he did for a tine, it was not so good. I sometimes felt badly that I took his time in that way when he took too much interest in my things, especially when they did not lead to anything. But in principle, and I think often in fact, he had very much of that spirit which is told about Moses when somebody came and said that there were some people who were prophesying. They thought that he would be very angry, but Moses said that he hoped that the whole people would be prophets. I think Bohr had much of that. But all these things were a little bit complementary, of course.

Kuhn:

Did Kramers talk to you about it? The fact that one is missing this early Bohr-Kramers work leaves information about this relationship out except for what may be in the correspondence. So there is just no way to re-capture that relationship. It’s a great pity. So I wonder what Kramers may have said to you about his own early work with Bohr.

Klein:

I remember that he once said that when Bohr first had got this correspondence point of view quite clearly, so that he began to know how one could estimate transition probabilities (to normal) states, was when he came back from a walking trek in Jutland. Then he told Kramers about it. Now, Kramers had a very good mathematical and mechanical background, so he began to calculate such Fourier coefficients (they had in that). Then he said that Bohr was not quite happy first because I think he had planned to do it himself. Afterwards he was, and he supported Kramers very much, but I think his first reaction—which was very human, of course, and he was a very young man then—was that when one has an idea one wants to work it out. But Kramers, who had sort of got in half-way as does a dog, I think felt very badly about it. When he told me about it, I thought it was very natural.

Bohr was a very great man; why should he be superhuman? But then he supported Kramers very much. But there was one thing which I think, Kramers sometimes felt more than I, much more than I. When Bohr got an idea in his head which was in a way polemic, then he could talk very strongly about it as if the one that he talked to was the opponent. When Kramers came back from Holland after he had offered his disputation, Bohr went with him; and Bohr was very glad for his work once he had done it. He had done very fine work, you know. Bohr went home; and then I was there in the summer; and Bohr told me very nicely about the things— Then Kramers got typhus and was very ill for many months and didn’t come back until the beginning of December or something like that. I remember the first talk we had. Then Bohr was talking very strongly about some point which really wasn’t connected with (???), and that was not so easy for Kramers, because he came right in the middle of it, and Bohr accused him in a way. Bohr didn’t mean it at all, but you know he could be extremely enthusiastic in that way. I don’t remember what it was; but you know he had always, even in late years, this way of trying to convince people. He wasn’t satisfied if he couldn’t.

There was a story from his childhood which you may have heard when his mother heard some very violent sounds from the boys’ room and saw that Niels was on top of Harold saying, “You won’t believe; you won’t believe!” And they were discussing some theoretical question. And he had a little of that way all through life. And he had it a little bit with me because I couldn’t quite understand, or I couldn’t share, his political ideas in later years. He took that always very, very strongly and seriously. Then afterwards he was extremely complimentary that way—afterwards he was the most tolerant person in the world. But just during the moment when he wanted that, then he could say very critical things and strong things about other people also whom in reality he admired, and he wanted to be as kind as possible with them. In that way Pauli I think played on the whole a good role because Pauli was, of course, in some things extreme. He was a very nice man, but he had a very extreme way of expressing himself; and Bohr usually wanted to express himself very cautiously and very politely.

While Pauli made a point of not expressing himself very politely, which was hard for people who didn’t know him; but his friends usually took it very well. We tried to react in a similar way to him, which he didn’t always like. That’s a little about the externals of Bohr’s way. Kramers told me also a little bit of how they had written. There they were out in the country, I think; and then they had some kind of cupboard with some drawers in it. There was a comedy in Denmark about a man who wrote poems for various occasions, and he has these arranged in a cupboard so that if there is some special celebration like a wedding or a funeral, then he says to his assistant, “Take one out of “skuffe tre”, [drawer three], or something like that. But “skuffe tre” was, in this comedy, always for the worst poems; so that then Bohr and Kramers called that drawer where they put the papers on which they were working “skuffe tre.” He used to walk very much with whatever young people came then in those times, like Kramers and me. I told the other day of a long walk we had; there he went many times like that. Then sometimes he walked and he wouldn’t say a word; and when I tried to talk a little, he just continued walking. And then, suddenly, he’d begin to talk about the subject and talk in a very, very lively way and very much, and he made very many observations. One thought that he wouldn’t though. I remember one of the early times when we went out to a pension in North Sealand to work. We were sitting at a table with some people there, and he sat as if he were very deep in thought and didn’t say a word during the whole meal. And then afterwards he began to tell me about what that person on the left of him had said. He had listened to everything.

That is a little bit about his way. He was so many, many-sided, you know. His family was very eager that he shouldn’t overstrain himself. He had done that, he told me, very strongly at an early age. He said that before that time he hadn’t known what it was to be tired, I think it was when he was working with his doctor’s dissertation. Then I think he went to Norway for a long vacation, and after, that he knew what it was to be tired, but he was extremely strong. Often he would be quite tired, would go out to the country for two or three days and then come back quite fresh again. But his family was very eager that he should not be overstrained so therefore we all felt very badly when we induced him into some very intense conversation. I remember when we were out in the country at a pension, he would meet somebody there in the drawing room who began to take up some literary or philosophic subject; and he got eager and talked for a very long time. And then he was tired for the day after. He said many interesting things in this connection.

Heilbron:

How many hours a day would he work?

Klein:

I think usually I would come in the morning sometime between nine and ten. And often…one sat and waited for him because he got caught by somebody else. Then one would work until lunch time. It would start in again sometime after lunch. Often he was interrupted, but I think it was usually not so very long so that one got home to dinner. Sometimes, of course, and that happened more with the unmarried people than with the married, he could work into any time of the night also. That didn’t so much happen to me because when I was alone with him, it was mostly either here he had his family very near or in Copenhagen after I was myself married. So there were not many nights when I was there. That happened to some people. Some have such recollections of his going on until very late in the night.