Oral History Transcript — Niels Bohr
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
See the catalog record for this interview and search for other interviews in our collection
Interview with Niels Bohr
by Thomas S. Kuhn, Leon Rosenfeld, Erik Rudinger, and Aage Petersen at Prof. Bohr's Office, Carlsberg, Copenhagen, Denmark
Bohr:Now first of all you would like to know something about my preparation. I would like to say that in doing this we are going to have some notes, we can't do that today so directly. But I could just say that I had, before I went to England, some preparation through writing this book about the electron theory of metals. And I'm also not prepared to go through that; you see, that would be an independent thing. I will only say that I made some efforts on the basis of classical theory to give a logical treatment, and as full a one as possible. And I found that there were a lot of errors in the earlier things. They're not so important, but, to give an example, some tried to calculate the number of electrons from the diamagnetism of metals. However, it came out of my work that, according to the classical theory, the electrons will not give rise to any diamagnetism at all. There were a lot of things of that kind. Now I think that to make this accurate is a question by itself.
Rosenfeld:What, was the incentive to take up that subject?
Bohr:That is also very interesting. That was actually a thesis for an examination paper. It was not the intention that one should do so much with it. The suggestion came from Christiansen, and then I went into it very carefully. I had most of it in the examination paper, but then it was, a few months later, made into a dissertation. It is difficult to go into any of that now. May I only say that I started on the Lorentz theory, but tried to make assumptions more general. I took another law for the collisions between the electrons and the ions. There I found out a lot of very curious things. But I'm not going to tell that about the flow of heat. -- Do you know about it? It is a very odd thing, but actually the flow of heat is larger than one would expect. One would expect the energy transferred to be 3/2 kT. ...Actually the flow is 5/2 kT per molecule. Do you know that? It is of course not the intention, but I could just tell you --. [Bohr goes to the blackboard to illustrate this point which is touched upon briefly in a footnote on page 70 of the Danish edition of his thesis. What follows is a reconstruction of the explanation Bohr gave at the blackboard.] If we have a box filled with a gas at temperature T, and if we move it from one side to the other like this: [figure can be found in transcript] then one will say that we have transferred an energy of 3/2 kT per molecule from one side to the other. But now when considering a flow of gas, one actually has a thing like this: [figure can be found in transcript] where the gas is contained in the tube between the moving pistons A and B, and the pistons are themselves connected by the rigid rod C. Due to the pressure, energy is delivered to the gas at the left side and taken away from it at the right side. A simple calculation now shows that the net result is a transfer of 5/2 kT per molecule across the boundary due to the flow of the gas. However, kT per molecule returns from the right side to the left due to the tension in the rigid connecting rod C. Thus, the net result is indeed 3/2 kT. If we now return to the first figure, by a similar argument one can show that when we move the box there will be a tension in the sides, s. The result is that due to the gas itself an energy of 5/2 kT crosses the boundary, but kT goes back in the other direction due to the tension in the sides of the box. I'm sorry to express myself so badly, but the point is of great importance. In the Thomson [Lord Kelvin] effect it is important to know the flow energy. And J.J. Thomson had calculated it by saying that it is 3/2 kT times the amount of (???). But this amount is dependent on the forces between the electrons and the molecules, and therefore, you can get anything you like for it. That was just something to show; there are lots of such things on every page of the thesis. But as regards these thermoelectric effects, one doesn't have a thermal-equilibrium. Therefore, it is mixed up with a loss of entropy due to heat conduction and the (???) effect. But then one gets that these thing can be separated, and quite generally. And that is actually really what later on Onsager has done. I don't know whether Onsager has more, but I don't think he has more. It is exactly this. And it turns out that one can reduce, under very general conditions, the flow of things into an integral equation. That integral equation has a certain symmetry in the core; therefore, it can be solved. And (the other) has also, and the solution has the same symmetry, and from that you get these relations. Now all these things are very stupid, and. I hope another time to tell you about it.
Kuhn:When was this work actually finished?
Bohr:That was actually finished just before I went. I started on it in ‘10, and I finished it in the first month of '11. And then I went over to Cambridge In autumn of '11. And, so it has been a very short time. But now the reason. The main preparation was that Jeans and Rayleigh had shown that if you take classical theory, then you have a Distribution of energy -- which is really no distribution of energy; it just doesn't converge. Jeans thought that one actually in the experiments had no equilibrium. And that was wrong you see. I went into it very thoroughly as regards the absorption and emission. Then it comes out that there is a paradox -- that the relation between absorption and emission is something very regular on the classical theory. But they didn't calculate it correctly for the absorption, but only for the emission. Then they find that the emission drops down. But the point is that that is not an actual equilibrium. Because if yea are thinking of an equilibrium, then, of course, you would say that there is infinite energy in it. And now how is one to avoid it? Then if one says (what is the theory of energy), therefore the electrons will move independent of that, and that before that you don't have a large amount of energy. And then just by calculating wrong -- it was nothing else -- Jeans and Thomson saw that they could get Planck's formula, or something similar to Planck's formula. They thought Planck's formula was then not the real thing, the real equilibrium. But it is wrong. The only point I will try to make accurate is this. If you treat an equilibrium problem, or at any rate consider an equilibrium problem, and if you then assume something and then find that you get some reasonable results under these assumptions, then it is a real equilibrium. And it is absolutely real. And that is the only way one does in physics. And therefore, the whole thing was calculated wrong -- there was nothing else about it. And, therefore, I knew very well that the quantum theory was (very valuable). Now other people perhaps also thought so, but there was, at any rate, this discussion. And so when I came to England, I knew that the quantum theory had to be taken seriously. Now I'm very sorry because this is just --. But we can always go into this paper sometime.
Kuhn:When you say that you knew that the quantum theory had to be taken seriously ... Can you say more about what seemed to be the heart of what had to be taken seriously?
Bohr:To get Planck's formula one had to make a definite change in the classical theory; that is the point. And, therefore, this thing here, where Jeans played such a part, was from the point of view of the electron theory. And I only felt and showed that this was no way of getting out of the difficulties. I hope we may so through all this some day, but it is a bit away from the main line, and it is not to be taken too seriously. But I came over to England and then I learned about the Rutherford discovery. That gave the reason to really think of a definite model of the atom. There we came -- Hevesy was there, you know. You will also ask him someday about things. It was so that I became clear about the atomic number and about the radioactive displacement laws. That was very interesting. First of all it was mixed you see. Namely this: Hevesy was there, and Hevesy told Fajans and Soddy about it, so that (they practically) knew that there was something. But their own attitude to it was entirely different. Fajans considered that this was against the Rutherford discovery. "Clearly because," he said, "If an alpha particle goes out then there is a change in the valence. But the valence is associated with the outer part of the atom. Therefore, the alpha particle must also come from the outer part of the atom." It was terrible, you see, but it was so. Ana he wrote a whole article -- do you know it?
Kuhn:No, but I've seen references to it. I knew it only from what you had said earlier.
Bohr:Yes. Oh, but you must read it because it is very definite and very clear that he thinks that it is all against Rutherford. And Soddy was the same. This is a little bit difficult, and therefore, we must also again ask Hevesy. But actually it (all came out at a meeting in London in '13, in the summer of '13.) Soddy had announced his paper. And he explained to Hevesy that it was the same as Fajans. But Hevesy said that I had written about it, and Soddy gave it up, and didn't say a word about it. But they both thought that it was completely against Rutherford. And later on (they said that it is directly) the same as Rutherford's, but that's not true. Now in all these things, you see, we must be very careful -- but that is fairly factual. Similarly in the case of the atomic number. Actually van Brock made a real mess of it. He had many elements because he put all the elements in a series and didn't understand that there were isotopes. Therefore, he got up to very large numbers of elements, and so on. But that is more or less immaterial.
Rosenfeld:The discussion about the number of electrons in an atom was also taken up at, I think, about the same time by Thomson in his experiments on the scattering of X-rays. Did you use that in your arguments. Or did it come later?
Bohr:No. First of all I'm not sure if some of it came later -- some of it was earlier also. But there was the question of the accuracy.
Rosenfeld:No. I meant only that the number of electrons was of the order of the atomic weight.
Bohr:Oh, that, of course, everyone knew, and I knew all of Thomson's work. No, the point was that now we had it accurately, now it was clear that the electron number was the number in Mendelejeff's table. Yes, that was the point. So then there was no doubt about anything.
Kuhn:On this same point, I wanted to ask you about the alpha scattering paper. When you treat oxygen in the last part of the paper, you compute 18 from the scattering measurements and say, "Now on Rutherford's theory it ought to be 16." But would most people have said already that on Rutherford's theory it ought to be 16? I mean, this grows out of your own analysis?
Bohr:Yes. Because you see actually the Rutherford work was not taken seriously. We cannot understand today, but it was not taken seriously at all. There was no mention of it in any place. The great change came from Moseley. But before Moseley's work there was absolutely nothing about the Rutherford thing. I could have put it all in this paper on alpha scattering. But thought that I would come out with a paper soon. And so when I said 18 or 16, it means that I knew it was 16, but I was satisfied if one would get 18.
Kuhn:I put my question badly, because it's clear that you do know it's 16. What I meant to ask you ö, would Rutherford, for example, have know in this period, or just before, that it was 16?
Bohr:You see, that is a very curious thing because -- you see Rutherford (already had very much to do). He was writing a big book. As I wrote in the article, shortly after I came I went to him and just explained that there were these immense possibilities of simplicity. And he thought that it was doubtful, that it was extrapolations. So it wasn't taken seriously at all. So I thought I'd better do it myself. The essential point is that this thing was independent of the whole business of the quantum theory. That was just an accident that it was me who saw at once how it was with the numbers and with the displacement law and so on. But then I actually, after I had written about the alpha particles, went back to Copenhagen. I only stayed in Manchester for three months. And then I worked on it in Copenhagen. And then I --
Bohr:Yes, but you have jumped now. When you say you worked on it -- there are so many problems here tied together.
Bohr:Yes, but you see I worked on enormous problems. I thought of whole thing. I wrote a very brief paper really. I even worked on magnetism, because it was a great problem that the hydrogen molecules was (diamagnetic). And I thought perhaps things are very much more difficult, and so on. Then, just half a year later, I just saw how it had to be arranged. Out of the series spectra -- as they stand -- one could just see how the radiation came from an atom when the atom is formed by the electrons being bound more and more firmly. I say it very badly -- we'll do it all over, you see. But this was the point. And then I came to know this very dramatic question of helium. That was a very interesting thing. First of all Rydberg had taken the view that it was hydrogen. He felt that there would have been a very great difficulty if there were these various series in the other elements and in hydrogen only one series. Rydberg was not thinking that hydrogen could have one electron and had a very special position in the table. We must do this over, but I can just tell you of the reaction when I first wrote about it. Then there was a meeting in Gottingen. And in Gottingen they were very dissatisfied with the whole thing. First of all they concluded that it was a mistake, that the whole thing was wrong, that there was a factor of two missing. But next Runge said that he wanted to say that he had long been interested in the spectra and thought that one should find a way to interpret them. And, therefore, he very deeply deplored that the literature about spectra should be contaminated with a paper like that. It vas very interesting. We were such close friends later on. But so it was; they absolutely were against it. But the point was that then I saw that one had this connection with the classical theory through the correspondence. ... Then one gets the (classical spectrum out of it. Then one has the whole spectrum, and that you cannot do in hydrogen with this new spectrum. So it had to be something.) That was really the main reason.
Rosenfeld:When you speak of a factor of two then, there is indeed, in your first paper, a place in which there is a factor of two, which is introduced in a way, which must have been very surprising for readers at that time.
Bohr:Yes. Absolutely, absolutely, you see. That was first of all written to resemble the other things. But then I just say, "Now we'll make this assumption," and then I make it completely correct and no doubt about the factor of two. ...
Kuhn:But I was also much struck by this.
Bohr:But that is entirely because you have not read it.
Kuhn:Oh, this paper I've read very carefully.
Bohr:Yes, but you see I say this, and then I go through the correspondence, and then it's necessary to have that out.
Kuhn:Well, you give a hint of a correspondence argument when you first introduce it. You say that it's the average between a state of no motion and the rate of rotation in the --
Bohr:Yes, but that is not the correspondence, you see. That is only to get into it.
Bohr:It's really done very, very accurately. .That was what Göttingen misunderstood; therefore, they didn't believe it. Would you just say what it is?
Rosenfeld:Oh, yes, here it is, and he [reads] "Let us now assume that during the binding of the electron, a homogeneous radiation is emitted of a frequency equal to half the frequency of revolution of the electron in its final orbit."
Bohr:Yes. Then one goes back and one makes it into a proper correspondence. That was just the stupidity of the way of writing it.
Rosenfeld:Yes. Then when you start from the postulate you show that in the limit you get equivalence with the classical motion. So it is quite true that when you read the whole paper very carefully and think about it, then you understand it. But the Göttingen people probably stopped, you see, when they saw this assumption and read no further. I suppose that must be psychologically -- ...
Kuhn:We must work back, of course again and again to this paper. I am impressed with how very rapidly we have gone from the thesis to this paper. And I would like it if, now or later, we could go back to Manchester, and talk more about where the individual Problems here come up. I am curious to know, for example, whether you had even heard of Rutherford' s atom before you got to Manchester; whether there was any talk of it in Cambridge?
Bohr:Oh, yes. And I have also written about this. First of all Rutherford came down to talk at a Cavendish dinner. That was a very interesting moment because that was the time when one first had the Wilson photographs. And Wilson had taken some photographs where one does see these bendings of alpha rays. And Rutherford was so enthusiastic about it. But, of course, he said that he knew it was so. He had based the whole theory of the atom on it. But that one ever should directly see it, that was something for which he was completely unprepared. I have written that [in the Rutherford Lecture] fairly clearly, haven't I?
Kuhn:You speak of his address and of his enthusiasm about the Cloud chamber; now it wasn't clear from this whether he had talked about the model at that time or not.
Bohr:No, you see, of course I may also have done it very badly. But you see I went up to Manchester a few weeks after. ... I must say one thing about this lecture. We had started to write accurately about the literature, putting in the reference to his paper about the atoms. And we put all of our references at the end of the paper. And then we felt, at any rate I felt, that it would not be right if one wrote about Rutherford, and the references were all one's own papers. Because there were so many points that referred to our papers. So I left it all out. We can give you the references. It really disturbed the thing, but there was nothing else to do, I thought. You know you were in it?
Bohr:No, but this point you see we must make quite clear because that is really so. ...
Kuhn:Were you, Sir, convinced of the likely validity of Rutherford's atom from first hearing?
Bohr:Yes. I was absolutely convinced of it. Because it was clear, you see, when one looked upon it from that way --. (1 don't know) what (they) mean to be absolutely convinced of it, but I was, you see. I felt that everything fell in line. Rutherford was not sure himself; but I felt that everything now was (aligned) so that one knew the number of electrons in the atom, one knew the isotopes. And, of course, our point was -- as I've written down there -- that the weight of the nucleus meant that they couldn't effect the spectral lines or anything. Of course we must go and do it accurately, so I think really perhaps these things one should not even read. One shall perhaps not type this up. You see that is the next thing. But that is only a question of time. No, but that's very interesting. Then I could also say one thing more, and that is about Moseley. Have you seen the letter from Moseley to me?
Kuhn:No, I've not read the letter.
Bohr:Yes, because he just asked what I thought about it from, the point of view of the spectra. I mean the whole thing was clear after the helium spectrum. But then things came very interestingly. And there was the very interesting work of Kossel's very soon after, where he actually found also the relations between absorption and emission and the further development of the shell structure of the atom.
Now I've made a complete mess of this. But the point was, above all, what the correspondence meant. My view was that one had to consider the use of the quantum theory as a generalization of the classical theory. And that was, of course, a view which was taken generally later on, but not exactly that way. And there was a very interesting discussion with Sommerfeld about things. And Sommerfeld also, when the paper came out, calculated the dispersion of hydrogen. But I objected to that, and my first letters to Sommerfeld dealt with that. ... The rate of progress was various. First of all it went tremendously when Sommerfeld's own things used the further quantization. (But the rate of progress was over dispersion) because that was the only possibility of making use of the classical theory. And, therefore, that was a very great point. It was very interesting that Lorentz had, always said that we have no dispersion theory, that there is no connection with dispersion. But that I have written in the thing on the Solvay meetings and we hope when you come back that we shall have that.
Kuhn:Well, you have already let us look at a typescript of it.
Rosenfeld:How did you come to examine the spectra?
Bohr:The spectra was a very difficult problem. There were two different schools -- those in England, and then there was the school of the spectroscopists. And one thought that this is marvelous, but it is not possible to make progress there. Just as if you have the wing of a butterfly then certainly it is very regular with the colors and so on, but nobody thought that one could get the basis of biology from the coloring of the wins of a butterfly. So that was a way to look at it; I'm sure that that was not the way that Rydberg looked at it. At any rate it was known that there was something of the sort. And I discovered it, you see. Other people knew about it, but I discovered it for myself. And I found then that there was this very simple thing about the hydrogen spectrum. I was just reading the book of Stark, and at that moment I felt now we'll just see how the spectrum comes.
Kuhn:Was this at Manchester that you were reading Stark?
Bohr:No, no that was later in Copenhagen -- that was half a year after. It was in January, I think, of '13. I had always a lot to do because I was really an assistant of Martin Knudsen when I came back. But then I wrote this very big paper about everything, you see, just to see how it went. But only until I got this clue. And then I started writing about that (and left it all).
Rosenfeld:Do you still have a manuscript of that paper.
Bohr:I'm not sure I have; I will look for it. I'm really getting more and more disturbed about it, [the interview], and I don't know what we should say now.
Kuhn:Already a great many things that surprise me and help me a great deal in thinking about these developments have come out, and I think that's the way it should be. I have a question in this same general area. How did the paper on the alpha collisions start? Had you actually started that in Cambridge?
Bohr:No. That was in Manchester. Darwin had a paper on the absorption of (these things). But he calculated it out, saying that if you have an atom, then, if you go outside the atom, there will be no absorption. Or, similarly, in the atom, if you co outside the core of an atom. That's not true. That's a very interesting point that if you have an atom -- [Bohr goes to the backboard to illustrate.] So that was really a technical point which brought me into it, but the treatment was not proper you see.
Kuhn:Did anyone suggest that it would be good to follow this take this further, or did you take this on yourself when you read Darwin' s paper?
Bohr:Yes. I just did it. And it must have been a very busy time, you see, because actually I was working with all hinds of things in Manchester. I was there only three months. This was a large amount of work, but I got it out when I was there.
Kuhn:Do you know how the wonderful idea of utilizing the frequency of the electrons as a measure of the atomic force comes in? In that paper you introduce a collision-time parameter and then compare the collision time with the natural frequencies of the electrons in the atom in order to decide which ones you can ignore. There has been no precedent for that; and I wondered how that whole approach to the problem of the atom had developed.
Bohr:First of all, you see, we cannot expect anything; the problem was new. That was that. Thomson had used the same calculation as Darwin's, but Darwin had used the Rutherford atom. And I just felt it was different. That was clear, and Darwin was apparently in agreement, and so on. Now that, of course, has been changed due to quantum mechanics. But that's something else, and that's also a much smaller change. Do you know I have written about these things in the later years?
Kuhn:I have not read that later paper, but I know of its existence. I know that this was not your own last concern with the problem. But there are, as you yourself said, there are a number of hints in the paper of the development of the early stages of the ideas that are going to come out again in the papers in which you do the constitution of atoms and molecules. So the development of the elements in that paper are of themselves of great concern.
Bohr:Yes. You see that paper really gives an idea -- if one reads it with your interest -- of the way I looked upon the Rutherford atom when I was in Manchester. There is absolutely no hint about the origin of the spectra.
Rosenfeld:You had not yet thought of introducing the quantum?
Bohr:No, but you see the spectra is this: everybody (thought of a number of states in matter.) But to say that if you have such two states, then the radiation is the difference, that was not in my mind; as far as I know, it was not in anyone else's mind. It was so that actually there is a little humorous discussion with Einstein in the paper I wrote about Rydberg. You know Rydberg's work, perhaps.
Kuhn:I don't know it in detail. I have read almost none of his papers, but I know of him.
Bohr:I have written here a paper called, "Rydberg's discovery of the spectral Laws." (I made it short). But I went into it, and its most beautiful.
Kuhn:There is, to me, another particularly interesting thing in the scattering paper, and it again occurs in connection with the discussion of oxygen. You remember, Sir, that you had been using measurements of dispersion in order to identify the frequencies of the electrons in the atom. When you get to oxygen, you can do four of the sixteen electrons this way, but you need estimate of the frequency for the other twelve electrons. This you do in a way that, I think, is particularly striking and is quite now in your own work and, I should think, quite new and different from anything that had happened previously at Manchester. You use Whiddington's results to find the velocity of the electron that will just excite characteristic radiation -- which has not, in fact, been measured for oxygen. Then you compute the energy, of that electron and then you equate it to Planck's constant times the frequency of the corresponding electron. Do you remember where you had gotten that technique?
I can show you that. It's the paragraph that begins here: "We know very little about the higher frequencies --" and goes through to there. You see there is a use of the quantum to get an electron frequency, and it is certainly not standard use.