David Dennison - Session II

Kuhn:

…I would like to get from you more systematically some notion of your scientific curriculum at Swarthmore, that is what subjects you knew. Also I would like to know more about G.E.; you told me about the work with Langmuir and that you had been passed on to [A. W.] Hull and I would like to go on with that. It might be as well to take the Swarthmore curriculum first.

Dennison:

It is certainly true that at that time the things that were taught would be thought of as being quite elementary now. In mathematics there were the subjects, starting with the freshman year, of solid geometry and trigonometry and a course called ‘College Algebra’ which was very well taught by a Doctor [R. W.] Marriott and he had a great deal of influence on me. Then in the Sophomore year I began with calculus which I found tremendously exciting, and shortly thereafter a course in what we would call intermediate mechanics and this was very enlightening to suddenly see how the differential equations would give various forms of motion. Dr. Marriott was very anxious to give me as much as possible and he did give me what would have been one or two courses in which I was the only member, the only student, essentially in the beginning transformations of theoretical physics or mathematical physics, Gauss’ Theorem and Stokes’ Theorem and Green’s Theorem and a few things of this sort. I am not sure that I really comprehended them to any great extent but at least I thought that I was somewhat familiar with them at this point. In physics the beginning course was not particularly helpful to me; I survived it, I was always interested really in physics, but it just didn’t mean very much to me I’m afraid. It was only later when I saw it a little bit more through mathematics that it became meaningful. Then during my Junior year, which would then have been in 1920, I had the course in the Bohr theory and I think it did mean quite a lot to me although surely would not have been the revelation that it was to people who had grown up firmly rooted in classical theory. I was prepared to accept new things at this stage, and I was of course very much fascinated in what it did and in its giving of the Balmer spectrum, and the spectrum of hydrogen. I don’t remember that I had many other courses in physics.

Kuhn:

Do you remember how far you carried the Bohr atom?

Dennison:

Probably just through the spectrum of hydrogen. This was not as I recall it written at all, it was just the man, I think Dr. [H. C.] Hayes was the man, and he just lectured on it. I am probably wrong, though, that it was Hayes. About this time there was a graduate student from the University of Michigan who came to be on the staff at Swarthmore; his name was Dr. Winthrop Wright, and he was very helpful to me in these extra activities, loaning me a microscope and various things that I wanted to have. I think he was a fair inspiration, although by this time I was already set in my interests. But he was later then to be very helpful in my coming to the University of Michigan.

Kuhn:

How much chemistry, astronomy, other sciences, had you had?

Dennison:

I thought, from my home kitchen experiments in chemistry and from the books that I had read, that I knew general chemistry. I am sure I didn’t but I thought that I knew it and I would not take any further courses in that. The man who was the head of the department was somewhat annoyed at my attitude. I wanted to take a fairly advanced course in physical chemistry — this would have been about my Junior year — and I was sure that I could do it all right, but be forbade me to take it. The man who was teaching it, Dr. Creighton, was however much more enthusiastic and Creighton told me to come and audit the lectures and also that he would allow me to take the examinations and would grade them for me although I couldn’t get it for credit, which I did, probably just out of spite to show that I could have, and did. This was all the chemistry I had. I don’t think I really had any courses in astronomy. I would probably have been much too snobbish or something.

I thought I knew astronomy, having done professional observing and reduction of these plates and having read a great deal in the library, so I would not have taken it anyhow. I guess that about finishes Swarthmore really. As I told you I got the knack of being able to do any course and get a perfect grade on it. There was also one course I took in electrical engineering which was very interesting to me too, I think really interesting because I have always had a certain interest in applied work too, in spite of being a theoretical physicist. Back to Schenectady [i.e., General Electric]. The summer was not quite half over when I had the feeling that I was just getting too much of the calculation end of it so I objected and Langmuir agreed that this was true. He was very interested in getting results done but he agreed that I was quite right and so he handed me on to [A. W.] Hull. Hull at this time was in the middle of his powder X-ray structure, crystal structure, and he was very kind to me. He started me out on a number of experiments doing new things. One was a metal, probably cesium or something like that; it had to be done in vacuum, and I never really managed to do it.

The other one though was the ice one, [paper #1], and I did this in several different ways. Hull had charts made up at that time so that once you got the hang of how it went you could take this crystal powder photograph and could compare it with the various charts and could then determine both the form of the structure and the parameters and all the rest of it which I did. Now comes one of the nicest things. The truth was that this was Hull’s work; I know now that I didn’t really understand it at all, not really deeply. I could not have reconstructed any of these charts, so I was only a technician. But Hull was very nice and said that I should publish this, publish it under my own name and so he helped me to write it up and checked everything on it. I remember I wanted to make some speculations, for example, that the structure showed something about why water expanded in the last few degrees before it solidified and he said that was a little bit too problematical and I had better leave that one out, which I did. It was a great thrill, and I have often thought this is one of the nicest things that an older man could do to a young man and I am not sure I have ever been able to repay anyone just this way but I have tried at times to do things of this sort because it really started me professionally. That was essentially the end of that summer.

Kuhn:

To what extent was Hull’s own approach to structure in this period entirely a classical approach? To what extent was he concerned at all with bringing in quantum mechanical properties?

Dennison:

I would say not; his was a purely classical approach. The only one that I knew there who was very much concerned with the quantum aspects was Langmuir. He was trying to calculate essentially ionization potentials, the work necessary to put things together and so on. Also, my opinion now at least, and shortly after I had been there, was that Langmuir did not deeply understand what he was doing either, that he was taking a formula and applying it but didn’t have the kind of deep knowledge that for example anyone like Bohr would show to such problems as these.

Kuhn:

Did Langmuir take spectroscopic evidence at all into account or did he continue to be particularly interested in, or pointing toward, the chemical factors?

Dennison:

He was pointing towards the chemical factors. The sorts of things he was calculating I think were, as I say, ionization potentials, the work necessary to take an atom apart or a molecule apart. Nor was this particular phase of it that I saw in his calculations very close to let us say valence theory; it was more the detail of what the energy relations were. That was what he was looking for at that state. He was very widely interested in lots of things though. He showed me an article at that time that he had written on contact potentials and it opened my eyes a good deal to what these things were about and I was very interested in it.

Kuhn:

Would you say that he felt that the phase integral technique in his hands was going pretty definitely to provide the answer, that it was only a matter of finding the particular model which would make these things work out?

Dennison:

I think so. I think it was to some extent a kind of black magic business with him. Here was a formula, which had been developed by Sommerfeld and [W.] Wilson. This looked as though this was the way to do it; he wasn’t really questioning the formula or questioning anything about that, he was just trying to apply it and see whether this would then work out. But you see also his difficulty then was that he didn’t have any firm feeling about the mechanics of it, so it was picking out an orbit and then making this calculation, the integral of pdq. I was doing numerical integrations around a particular path, but there was this trouble that he didn’t have any deep ideas as to what or why this was going on. So it didn’t come to anything I don’t believe. The next summer, this would be 1921, right after my graduation from Swarthmore, 1 did not do anything, I loafed. The summers following that, namely 1922 and 1923, I went back to Schenectady. During most of this time I worked for Dushman and Dushman was interested in thermionic emission and we were doing things at that time that had to do with work functions and to do with Langmuir’s theory. When they ‘thoriated’ a filament, i.e., put in thorium oxide, and then flashed the filament, the thorium would be reduced and come to the surface; there were all the work functions of the thorium on the surface, mono-molecular layers and this kind of business. And I did a lot of that with various substances, in fact I think it is published in a short note of which I have no reprints or anything, but I am sure that this was done.

Kuhn:

Was your contribution here largely experimental or theoretical?

Dennison:

It was experimental, really. I am sure there must be notebooks there, they did exist, that have these data in them, and the interpretation of it. I did talk with Langmuir quite a bit during this time and I know there were various problems about fields, electric fields, and he would quote from Maxwell and in fact I think he had Maxwell’s book there, Electricity and Magnetism. Later one time in talking with Hull I expressed my admiration for Langmuir and what a very fundamental guy he was, and he knew all these things, he knew all about Maxwell theory; and Hull, a very patient and sweet tempered person, nevertheless kind of snorted, and he said “You know where he got that book?” I said “No”. He said, “That’s my book, he borrowed that, he had never known anything about it until he read it out of my book.” [Laughter] And Hull I think was probably the more profound scholar of the two, but I was well impressed by both of them. I’ll make the personal remark here that Dushman eventually I think became a little bit apprehensive of me because I could do things, in theory at least, much better than Dushman could. I would make predictions and he was always very astonished that they mostly worked. I think he was a little bit apprehensive, as I say, of just how this came about. When I was just about to receive my degree in l924 I was offered a position at General Electric Company in their research laboratory. I really didn’t think about it very strongly because I was just sure this was not what I wanted to do. I felt I wanted to do much more fundamental physics and so I turned it down and I know Dushman was quite annoyed at the time and wrote me a letter about it which fortunately doesn’t exist anymore. But I was so sure that this was not what I wanted to do. And it was quite right, it wasn’t the thing for me to do.

Kuhn:

What is it that annoyed him? The implication that they were not doing fundamental work?

Dennison:

Yes. The notion that I was going, I suppose, a little bit swell-headed, I had set my sights too high, something of this sort. The reason that I went here to Michigan was simply that Dr. Wright, who had been a graduate student there, wrote to Prof. Randall and suggested my name for a teaching fellowship. In those days they were very careful about all appointments. I went to the Physical Society Meeting in New York and met Randall there; he tried to tell me a little bit about what they were doing — just beginning to get the interpretation of the HCl infrared spectrum. I am sure I didn’t understand too much of it, but at any rate he was well enough impressed to offer me a teaching fellowship, some $800 for an academic year which seemed to me like a tremendous amount of money, and I came here in the fall.

Kuhn:

… Were there other places you also considered going, or did that whole thing work out so quickly and well for you that you didn’t even really think of alternates?

Dennison:

I think this is correct — that I didn’t really think about alternates. Perhaps the fact that I had lived here and that when I went to Swarthmore I had been very lonesome for this place and for my old friends here, who had of course all disappeared by this time, was probably an underlying thing. I probably am someone who has fairly strong roots and a certain amount of homesickness. This probably accounts for the fact that I have been away a good deal. Often times people that do have very strong roots will nevertheless spend time away but always are terribly glad to get back.

Kuhn:

Where else would you say one could have gone to get a good physics education in that period in this country?

Dennison:

Well, of course knowing now, the places would be Cal Tech — was Cal Tech in existence in those days?

Kuhn:

I believe it was in existence, it was certainly in existence.

Dennison:

It was well recognized at least by l924, that I know. Now whether it was in 1921, I think so, but I’m not entirely sure. When did H. A. Lorentz go there, do you know?

Kuhn:

I’m not sure; I think it is a little later.

Dennison:

I doubt it. I think it was the very early twenties. I think that’s right. I don’t believe I really thought very deeply about this question. In a sense other people were looking after me; Wright and Randall were doing this, and as you’ll see later on people still did it for me. So I really didn’t have to worry too much about it. When I came here I was very much attracted by Prof. Colby, who had a good deal of influence on me. Colby was the only one that gave any good courses here; he was giving thermodynamics that fall and I wanted electricity and magnetism which I had not had. Now saying this brings to mind another thing. Wright at Swarthmore during my last year had taught me thermodynamics from Planck’s book Thermodynamics, a rather classical book, and I thought that I knew thermodynamics. Again, I don’t think I really did, that is, I don’t think I was sufficiently mature; I just knew the outsides of it, but I didn’t know it deeply. I had the feeling that I didn’t want to take any more classes than I needed, so I didn’t go to Colby’s thermodynamics because I said I had had it. The courses which I had, in fact throughout the whole time, would be just completely ridiculous and insufficient now. There was one course in mechanics which was not a very good course, but it did have the advantage that at times it referred to Whittaker, and I then seized on Whittaker and this notion of the Hamiltonian, this was just absolutely wonderful. I was thrilled to pieces all over it, and in fact I think I’ve always liked mechanics better than any other subject.

Kuhn:

Did the course itself follow through, on Hamiltonian techniques?

Dennison:

No, no.

Kuhn:

Did you follow it far enough to be doing Hamilton-Jacobi theory?

Dennison:

Hamilton-Jacobi I’m not so sure. It may have been just Hamilton equations. But I would try out every problem with every new method that would come up, so while the class, which was undoubtedly a very small class, might be doing some vibration problems or something, I would pick it up. It was quite an elementary class really; in vibration it would get as far as what is called (Blackburn’s) pendulum. This is just a two-degrees of freedom problem and a very special one. It did not get into small oscillations as such, I didn’t have that. I took a very inadequate course in kinetic theory which actually never went much beyond just the Maxwell-Boltzmann distribution and the distribution with respect to components of velocity, total velocity and things of that sort.

Kuhn:

Was there any book used in that?

Dennison:

There certainly must have been, and it was probably Jeans’ book. But we didn’t go any distance in it. I took an intermediate course in heat which again had some laboratory, but not much; a few mathematics courses, one in potential theory — at least my recollection of it was that it was pretty slim going. Professor Zweit, Z-w-e-i-t [sic. should be Ziwet) who was then well over seventy, a mathematician, gave a course in hydrodynamics. This was quite good; this I really got something out of. For example he told us about vortex rings and had us read some of the original literature on it; it was quite good, and really opened my eyes. The textbook was the standard one, Lamb’s Hydrodynamics. So there were a few courses that were good but not too many this first year I was a teaching fellow. I was very suspicious of taking courses, wanting to do things on my own.

Also there was a course in complex theory that was unbelievably bad. Incidentally this has been my characteristic, with its good and bad points, that I have always had pleasure in doing things by myself — with my own fingers so to say. And I am not terribly excited about what anyone else does, which means that while I do read the literature this is not the way I have ever gotten results. It has many good points, and it gives me a great deal of pleasure; it has many bad points because there will be things that I ought to know that I don’t know. It was quite a contrast later when Uhlenbeck and I joined the staff here in 1927; his power always comes from his knowledge of physics and everything that has been done, and I had no such knowledge at all. Nevertheless we both would achieve results, but it is a difference of method. This explains why I didn’t want to take classes. The second year one of the principal things that I did was to work with Prof. [E. F.] Barker and I was reading and translating Sommerfeld’s book, probably about the third edition, maybe the second…

Kuhn:

It was 1922 so it was the third edition.

Dennison:

Every two days I would come in and talk with Prof. Barker for an hour or more and give my version of a chapter or portion of a chapter.

Kuhn:

Having written it out in English in the interim?

Dennison:

No, just writing notes on it essentially, but it did mean I had to translate it of course.

Kuhn:

You don’t mean you were preparing a translation.

Dennison:

No, no. I had to read it and it was the physics that I was discussing with him. That’s really what it amounted to.

Kuhn:

Was he reading it simultaneously, and for the first time?

Dennison:

Yes. Probably not for the first time. I would guess that this is a case where neither of us was leading the other. He could often supply some things on physics that I didn’t know and I perhaps could supply at times some of the points of analysis that he didn’t know. This was a very fruitful thing because I spent a good share of the year on this and I really did know it at the end of that time.

Kuhn:

Was it known to other people hereabouts, had it been a model already in the earlier editions, or was it just beginning to take hold?

Dennison:

It was certainly just beginning to take hold. I would think that probably Barker was one of the few people who would have really known anything much about it.

Kuhn:

What about Colby?

Dennison:

Colby might have all right. Colby undoubtedly did. I forgot to mention that the second year Colby was on sabbatical, and this is why I never got electricity and magnetism at all. I’ve forgotten where he worked. He sent me at times problems from the lectures that he was listening to there and I worked them all out and sent them back; I admired him very much. He was also on a recruiting mission and he brought back Oscar Klein with him. And this was then the third year and a very fruitful one for me. I think I picked up the problem of methane because I was interested in molecular problems. They lay close enough to both quantum theory and mechanics so as to just be the sort of thing I would like to do.

Kuhn:

Was there a carry-over there from the G.E. work? You had been so much concerned with molecules all along, and I just wondered how this had developed.

Dennison:

I don’t really think it was a carry-over directly; no, I don’t think so. I was very much fascinated by the charts, the spectra shown in the book by [W. N.] Coblentz. You know, that’s a huge, fat volume; and he has perhaps a hundred different things in there. Somehow or other I had the feeling that these must be telling something. But what they were saying was completely unknown at that time. Then [J. P.] Cooley’s work was on methane; this seemed a very good place to start.

Kuhn:

Tell me about the background of that work. Was he a student here?

Dennison:

Yes, he was a graduate student at that time, my generation. You see, the work here had been the work of Randall and (Imes) on HCl. And as you have undoubtedly picked up, this was entirely Randall’s work. (Imes) played the same role in this that I played in what I did with Hull. (Imes) took the measurements and worked the shutter on the galvanometer and things of this sort, but it was Randall’s work, his design of the spectrometer, his knowledge that it was important and everything else. Then [W. W.] Sleator had been doing some work on water vapor, so completely confused that no one had the slightest notion what it might be about. And they were then beginning, however, to pick up some of the more complicated things. The first one of these to be taken, I think, was methane. Why? Probably because in Coblentz’ book it is one of the very simplest of the spectra, showing essentially just two absorption bands. I worked with Klein and Colby, and I would come in, I suppose every other day, and show them all the things I had calculated up to that point; we would talk about it, and then I would go away and come back. At the same time I was also reading Gibbs’ book on statistical mechanics, statistical theory, with Klein. We were going over that together. But the main thing was the work on methane. I must have had fairly good luck with that because a great deal of it was sketched out, and indeed calculated, and here my memory says probably sometime in November, late November when Bohr visited here.

He came no doubt because Klein had been his student, he came here and gave a lecture. And Bohr was very much interested in what I was doing and talked with me a good deal about it, and in fact set me straight on a few things. We had forgotten some of the frequency factors in doing intensity, and he told me what to do about this, which I did. At that time it was decided that I would go to Copenhagen. Just how I was to go, this was another question. But I was absolutely positive I was going; there was just no question. I didn’t know where I was going to get the money, but I was going, that was it. I continued to work during the rest of the year, and I had occasion, as a matter of fact, to look over my thesis, a while back. It has two parts to it; the first is roughly what was published. Even that is really not right in many respects. All the analysis is correct, and the way of doing it is correct, but I did not have enough chemistry to realize that in a molecule of this sort the valence forces were the important ones. A physicist, in general, without chemical training will always think about central forces; that’s why I did. Klein and Colby made no objection to it, and it would have been unnatural for them to, I’m sure.

Kuhn:

You’d have had a bad time at that particular time knowing what to put in for valence forces if you’d taken them more seriously, wouldn’t you?

Dennison:

Probably. Actually, what I did though, eventually, was to put in the most general type of force. There are five constants for the general potential, and I had them. But I was always thinking that, they were really essentially central forces with perturbations on them. I think, guided largely by this, I then made a wrong identification of some of the frequencies. That is not a bit surprising now because only two are active. The other two are inactive. One of them is made by a very high-order perturbation slightly active, which I didn’t know, but there is a little bit of a wiggle that appears in the Coblentz curve and I correctly identified that one. The other is totally inactive, but I identified it, however, with a wiggle in Coblentz’ curve. This was silly because I should have just stuck by my guns that it was completely inactive and wouldn’t be there. But I thought here was a wiggle, and it might be, and so I did. That was wrong. If I had been more of a chemist, I probably wouldn’t have done it. At any rate, there it was. So the identification was only partly right.

Kuhn:

How long was it before that was straightened out?

Dennison:

It was a little while. It probably was about 1930, and it may have been people like Urey who did it…The second part of my thesis is concerned with the electronic structure of the thing. How could one obtain the force constants out of the electronics?

Kuhn:

These are the force constants for the molecular problem?

Dennison:

Yes, that’s right, for the molecular problem. I didn’t try to get all of them, but I did try to get the symmetrical one, which would be the easy one. And actually it was not too horribly stupid. What it was, was to take circular Bohr orbits, of which some fell inside the hydrogens, and consequently were not affected by the charge, which I thought was smeared into a sphere; and some were Bohr orbits which went outside the hydrogen. So this was really essentially an inert-gas model, in which some orbits were enclosed within this thing and were therefore around, so to say, an ionized nucleus, and the others were on the outside and therefore a normal —. Now this meant patching up orbits, you see, in a war that is frightful, but —

Kuhn:

What did you use for the orbital plane?

Dennison:

No, this was just like for an inert gas, for neon for example. These were not in a plane. These were independent orbits. It was really following Bohr’s atomic structure, which was just beginning to develop at that time. Enough had filtered back especially through his student Klein, so I knew about this.

Kuhn:

Are these really then molecular orbitals rather than orbitals for —

Dennison:

Yes, they are molecular orbitals. My model was essentially a carbon nucleus with its charge, then a sphere with charge 4 at the distance of the hydrogens. So I was putting together pieces of orbit on the outside of an unionized atom with pieces on the inside of an ionized one.

Kuhn:

I was trying to take these to be perturbed atomic orbitals, which is why my question about the plane came in.

Dennison:

This was all of course absolutely wrong and not wave mechanics or anything of the sort. The answers that came up for the force constants were not too far off; I’ve forgotten now, but they’re off by about the square root of two or something like that. And this has been somewhat of a lesson to me that I’ve thought of many times since. Namely, when anyone makes a calculation now of electron orbits, of electron energies in a molecule, it comes out with the right order of magnitude for these things. This, in my opinion, is not at all a test of the correctness of the model which he has used, because it just lies in the dimensionality. It’s the dimensionality, the way in which forces must fall off with distance because they are essentially Coulomb-like, and the dimensions of a molecule which will inevitably give you the right order of magnitude of constants. So I think people have often been fooled to think that they have the right method because it coughs up about the right answer, and it may not be the right method at all.

Kuhn:

Why is it that that work wasn’t published?

Dennison:

I regarded it as less certain. I published the first one. I suppose when I went to Copenhagen and began learning more I regarded it as an interesting attempt, but really quite amateurish; so I never did think about publishing it, actually. Also, there was in the original version of my thesis an error of logic which I didn’t pick out, nor did Klein or Colby. I originally had only four constants, and not five. The fifth one comes in because in a structure like this you can have tensions, that is to say all hydrogens may repel each other but all be attracted towards the carbon. This gives you a tension, which is then a first-order effect. When you make a displacement from equilibrium, this tension, because of the geometry, enters again in first-order, giving a second-order effect, which is what you want, that is something depending on the square of the distance from equilibrium. This I only realized when I later worked on carbon dioxide in Copenhagen, towards the end of my first year there. I then wrote back here and said that there was this error; they wrote to the Astrophysical Journal and got the thing stopped in time, and I got the correct version to them. The published version is right in this respect, although the identifications are wrong. Still it is entirely correct this way. I was also always interested in this problem of intensity. Why not? I was trying to extract all the information I could from the molecule. The fact that experimentally it was hard to obtain intensities didn’t particularly concern me. For some time in later publications I always brought out this problem of intensity. Of course, I was doing it very badly because there was a question as to what charge one should use on the nuclei, and there is the question as to whether the electrons are following the nuclei or not. I took what was the simplest beginning assumption, which in the case of methane you might think would be roughly right, that the electrons made a spherically symmetric distribution; the hydrogens and carbon moved in this distribution. Therefore I at least was thinking that it was mainly the hydrogens which would carry the electric moment. Wrong, but still not —

Kuhn:

Yes, but at least it’s helpful. It’s exactly what led me to ask you the question about the plane, thinking of these electrons as being, in the first approximation, around the individual points of the nuclei in the molecule. In the same way, it led me to ask you, in the script, [outline, p. 5] why you hadn’t expected the electrons to move along and had therefore been surprised to find that the effective charge was quite so small as it turned out to be.

Dennison:

This problem of intensity was a very long struggle. It took me some time to realize why it was that the experimental things were so far off. In fact, this was not until 1928, which I’ll come back to. At this time Klein suggested that I write a preliminary version which I would send on to Copenhagen and which is the one which is mentioned in this letter. It was very interesting. I was sure that I was pretty good about writing. I was positive that I knew how to write. So I wrote a version of the thing up, and Klein sat down with me, and every single sentence was wrong, had to be revised and changed. ‘Exactly what do you mean by this, and what’s the shade of meaning here?’ and so on. Everything, piece by piece. I remember at one session I was beginning to be a little annoyed. I said finally, “Professor Klein, you would write it one way; I will write it another way. This is my way of writing it.” Klein said, “This is nonsense. There’s only one way to write a paper. If you have something to say, there is only one way; there are no two ways of writing it.” This actually, to a certain extent, was so banged into my head that I tend to believe it. I tend to believe it now that a good paper is one in which — in a sense it’s a little bit like an analytic function — each sentence has a little bit of what was past and a little bit of what is coming, so that from a portion of it you almost can reconstruct the rest. This is an exaggeration, of course. But, nevertheless, there is really only one way of writing well.

Kuhn:

Do you think that in retrospect, or did you think at the time that Klein was usually right?

Dennison:

He finally dinned it into my head that this was the way to do it. When I later came to write the article, I was sure that Klein would make lots of objections, because the article was an amplification of the thing I had sent to Bohr. Actually I had apparently learned my lesson, because there were very few changes that Klein suggested. He was very insistent on how it was to be done, and I think I really learned a great deal of that. I hope I’ve never lost it. Later, in Copenhagen, Kramers was very kind to me, and he also went over papers with me with considerable care, and he helped finish up what Klein had really essentially worked out with me. During the spring I became engaged, and I wanted to be married in the summer, and to bring my wife to Copenhagen.

Kuhn:

Had you known your wife for a long time?

Dennison:

Only for a year. She was a student at Michigan State. I learned to know her by accident, through a cousin of mine. I proposed that she should come with me, and the fact that she had only had two years there and would have to give up the rest of her education seemed to me completely inconsequential. I hadn’t the slightest belief in the efficacy or validity of a college degree, or of any degree, and I think I never have had, really. I have complete skepticism about what any of these things mean. Her father was kind of shocked about this and not very happy. Helen perhaps at times has been a little unhappy not to have finished, but I don’t think really. I also think my own opinion is just the same actually; one learns things by oneself, really. Whether someone says you have passed a certain course is not of any great importance. Through Colby’s suggestion I applied to the National Research Council for a fellowship to go abroad. Up to that time they had been sending people abroad on these National Research Council fellowships. I didn’t go to the Washington meeting of the Physical Society when this thing was decided, and I waited here for word from Colby, who had gone. I was quite sure that I would get it. Very sure of myself, you see. Word didn’t come, and then Colby came back, and he said that I had been allotted this fellowship all right, but their policy had changed, and I would have to have it in this country. The suggestion was that I would go out to Cal Tech. I was absolutely indignant. Of course not! I was going to Copenhagen; that’s where I needed to go for my best work, and I was certainly never going to accept a substitute for it.

Kuhn:

Did Colby say anything about what had led to this decision? In the first place, was this an across-the-board decision?

Dennison:

This was across the board. It was just a complete change of policy.

Kuhn:

Why do you suppose that was done? Was there anything said that would illuminate it?

Dennison:

Not that I know of. Whether it was a wave of nationalism that hit the committee at that time, whether they had had any bad experiences, I just don’t know. But at any rate this was the decision.

Kuhn:

This was spring of ‘24?

Dennison:

Spring of ‘24, that’s right. So I wrote them and told them to go roll a hoop. [Laughter] I don’t remember exactly how I said it; I’d be very much interested to see the letter now. I suspect that I was probably very polite and thanked them for having given it to me, but that it was my plan that I needed to go to Copenhagen and I would go. This was certainly pretty high-banded. I had about $1700 that I had somehow or other managed to scrape up over the years — gifts and various things — that was in the bank. So I would use this to go with. So I was going. I’ve always had people who have been guardian angels, I guess. This somehow or other became known to Dr. Miller in Swarthmore, and he probably represented to Aydelotte, who was by this time president of Swarthmore, that here was a young man who should be supported. Aydelotte stirred himself with the new General [International] Education Board which was the Rockefeller Foundation, but it was called a General Education Board Fellowship. They were just beginning to award them, and they did award me one. My recollection is that it was approximately $1800. It was less than $2000, at any rate. I was delighted, although I suspect not astonished. [Laughter] I was so positive this was the thing to be done, there was just never any question in my mind. I [knew] somehow; I’d always done it. In this letter to Helen’s father in which I explained things a little bit, it was on the basis of the money that I had that I was sure that we could do it for one year, but I wanted to stay at least two. Consequently, I would have to borrow money for the second year. This seemed a perfectly good investment to me and that it was the proper thing to do, which indeed it was. So we went to Copenhagen. Just to finish up the fellowship things once and for all, at the end of one year Bohr wrote a letter recommending my re-application for a second year, which was then given to me. At the end of the second year I was not quite sure whether I wanted a third year or not. We came back here during the summer of ‘26. Professor Randall felt that he was not at that time in just the right position to start theoretical physics here. He suggested that I go back for a third year, and he got some money from the then newly created Engineering Research — I don’t think it was called an institute — but it was the first applied physics end engineering work that was done here, sponsored mainly I guess by industry.

Kuhn:

That was at Michigan?

Dennison:

That was at Michigan. He got a small amount, I think of the order of a thousand or maybe twelve hundred dollars. But the other one had run a little bit over, what with vacations and so on, so that it was enough to carry us through the third year all right.

Kuhn:

You were already determined in your own mind to come back to Michigan? It wasn’t a matter of working for other teaching positions at home?

Dennison:

It seemed again like a more or less natural thing, and Randall said that he wanted me to come back. Again, I don’t think I really thought too much about it; it just was a natural thing to do. I’ll say again what I said earlier: This work here was the first theoretical dissertation that had come out, and I think they were all a little astonished about it.

Kuhn:

Did you have any trouble getting it accepted?

Dennison:

No, I don’t think so. There was no trouble. For the three-hour oral examination I had had to prepare myself in various things, including electricity and magnetism. I don’t really think I understood it very well, but at least I knew it well enough so that Colby thought. I understood it. It all went very well.

Kuhn:

Did you have to prepare yourself in experimental work also?

Dennison:

No, this was not necessary. No, it was understood that I wouldn’t be an experimentalist.

Kuhn:

… You’ve said on several occasions how close you were to Colby and how influential he was. On the other hand, I get the impression that until the third year you never had any course work with him.

Dennison:

That’s correct.

Kuhn:

But he clearly influenced you before then because you were corresponding with him, sending problems back and forth in the second year. What was the nature of that connection?

Dennison:

A little hard to tell. I suppose it was just the very great admiration of a young man for a very accomplished and a very knowledgeable sort of person. Colby was very good in theoretical physics, although not intrinsically an innovator. Again, he was one of these people who at that time, at least, knew a very great deal of physics, although as I say, was not a creator himself, not really. Of course, it was very thrilling to have someone who had been a student of Boltzmann, who really had connections with the great people in the past. Also his interests as a person, interests in music and literature — we had lots of very nice conversations. I think this was essentially it.

Kuhn:

Did you see him at his home, or at his office?

Dennison:

Sometimes at his home, but mostly at his office, I think, and I would discuss some things with him. It’s hard for me to be so positive about this; I can remember being up in his office and talking with him, but I can’t really remember exactly what we talked about.

Kuhn:

…You spoke of his trip in your second year as being in some part a recruiting trip.

Dennison:

Yes.

Kuhn:

Michigan made a policy of this, at least from that point on, and was more oriented to Europe sooner than a number of other places. I wonder whether you were conscious of that movement, and to what extent it may in fact have had something to do with your own fiercely set determination to get to Europe.

Dennison:

Probably very little. It might have, but I had such a firm notion that this was what I was going to do that —

Kuhn:

Was some of that acquired from Colby do you suppose?

Dennison:

It could be; it certainly was strengthened by that. But, you see, as I told you, the fact that my father and mother had done this, it just seemed to me that this was the thing to do. I was not going to stop my training with getting a doctor’s degree. I really wanted to go on to the limit of my capabilities. This was just a natural thing to do. Randall had a tremendously forward-looking plan with respect to this department. He’s really a wonderful guy. He was the one who fastened on Colby, and with Colby’s connection with Europeans then began recruiting people from there. You see, he brought Klein here. Klein only stayed two years. Then he brought Laporte. This was somewhat of an accident, really, because Laporte had been in this country, I think at the Bureau of Standards. He came here, and I personally think that Laporte’s coming had not been part of Randall’s plan. But Randall did want to have several people at once; he realized this was the proper way to do it. He sent Colby over at about the time of this last year that I was in Europe, and it was to find one or two young people. Actually he fastened onto two, Goudsmit and Uhlenbeck, and myself. Laporte, as I say, I think was not part of his original plan but he was able to get him, and this was just fine. So we had the four of us, and we really made, of course, quite a considerable nucleus here. He supported us tremendously; I know it must have cost him quite a lot with respect to the opinion of many of the more subtle people in the department. We got quite rapid promotions; and then this notion of the summer symposia, this was Randall’s notion. This was all Randall. He first had thought that the thing to do was to bring good experimentalists here. He tried this during the years when I was in Europe, and it didn’t go particularly well. Then he fastened that it was theorists that would really do it.

Kuhn:

[He had attempted to] bring good experimentalists from Europe?

Dennison:

His first idea, I think, was just anyone in this country. I think that the distinction was not between this country and Europe, but it was first a good experimentalist. He had someone here from Harvard, I believe. He also had Kemble here; this was in the years when I was away. Kemble was a theorist. He saw that he got so much more out of Kemble than he had out of the experimentalist who had been here that he went after theorists. In those days there were so few theorists in this country that it was then inevitable that he would go abroad to get the best ones, and that’s the way it worked out. He was very successful.

Kuhn:

…You finished your thesis in ‘24, presumably in the spring of ‘24. It didn’t appear until September ‘25; there’s no receipt date in the Astrophysical Journal.

Dennison:

I guess that’s right… I think that there was actually of the order of a six months’ delay in this, and that because of the fact that I had found the error in Copenhagen, it probably was not received by them until the spring of ‘25. That’s probably the way it worked out. I think that was undoubtedly it; there was at least a six months’ delay between when you put it in and when it came out.

Kuhn:

Apparently Cooley’s article must have been held up too, waiting for you —.

Dennison:

It must have been held up too. He was probably annoyed, but I don’t know about it at any rate.

Kuhn:

Now, do tell me again what you said at lunch on your choice of the Astrophysical Journal.

Dennison:

What my recollection is, as told me by Colby and Randall, was that just at that period the Physical Review was fed up on spectroscopic papers, and that they were not receiving them. Consequently one had to look elsewhere. I think that the Physical Journal — well, I thought maybe it was a rather new journal then, but I guess that’s wrong. At any rate, they were receiving them, and the papers were being sent; at that time these molecular spectroscopic papers were being sent there… It was very curious to have it come out in the Physical Journal, certainly, at that time.

Kuhn:

Yes, I’m certainly surprised to see it there.

Dennison:

And then we continued for a number of years; there are a number of articles that appear from this. And then apparently molecular spectra became respectable again, and the Physical Review accepted them, and so then I sent them there. Incidentally, this will come later but the one on the rotation of molecules, which was sent to the Physical Review, as far as I know was the first article on matrix mechanics to appear in the Physical Review.

Kuhn:

Yes, I think that’s very likely true.

Dennison:

There were no others.

Kuhn:

Since you’ve brought this up, I was curious to see an article from Copenhagen in the Physical Review, because I think that may very well be the first article from Copenhagen that appeared in the Physical Review.

Dennison:

That might be, indeed.

Kuhn:

Bohr characteristically sent material to the Phil. Mag. or to Nature, sometimes to the Zeitschrift. Do you, remember what determined that choice?

Dennison:

My sending it to the Physical Review? No, not really, except one of the articles that I wrote in Copenhagen went to the Phil. Mag. because this was a journal that Bohr was accustomed to publish in, and this seemed quite reasonable to me to do it. I don’t think I had any real feeling of wanting one journal more than another, just whichever one would publish them rapidly and the ones that physicists were reading. Also, in those days reprints were very important. I don’t think they’re very important any more, but they were at that time. People sent them around a great deal, and this was very important to get them.

Kuhn:

You went straight to Copenhagen?

Dennison:

That’s right.

Kuhn:

Do you remember when you got there?

Dennison:

Yes, pretty much. Let’s see, we were married on the 24th of August — Oh, Heavens, the 28th of August — which meant that we left about two weeks later by the Danish-American steamship, the (Heliola), and arrived in Copenhagen pretty much towards the end of September. I was terribly naive; Bohr had been so very cordial to me, I rather expected somebody might come down to the boat and meet us; of course, they didn’t. They don’t do that sort of thing very often. I hadn’t really made too much in the way of plans, finally ended up in a hotel recommended by someone on shipboard at the last minute, and went to the Institute the next day. They were very nice to me. This was a period when it was awfully difficult to find places to live. We started out in a Pension, for pretty nearly two months, I think, before we found any place at all. Copenhagen was wonderful. But it also was a competitive place. Heisenberg was there that fall —

Kuhn:

Was he there for the whole fall?

Dennison:

This I don’t remember, but he was there for some little while. Now I also think he was probably not there in the first days when I came, but within a month I think he had then come to Copenhagen again. I was by no means sure what sorts of things I wanted to do… I did not go there with any feeling about wanting to continue with this molecular work. I had read all of the papers of Bohr, and I think had understood them quite well, and had read as well most of the other important people; Pauli, for example, some of his had come out. I was very much interested in physics [and] the development of physics at this time. I did not have, I think, the fundamental background training which I should have had, so I’m not sure that I was as useful to Bohr and the people at that time as I wished I could be. It was highly competitive. Bohr was working and he was talking with people like Heisenberg, and it was a question not only of the physical ideas but of very great technical excellence, of past knowledge. I knew the things that were in the papers of Bohr and others, but I didn’t have, I think, as profound a background knowledge as I should have had for this. It had its moments of being discouraging.

Kuhn:

When you say that it was competitive, what were people competing for? The chance to talk with Bohr?

Dennison:

Partly, yes, a chance to talk with Bohr or work with Bohr. Bohr was helpful, but I’m not sure that I was in a position to be as helpful to him at that time as I would have liked to be. For example, I can remember that I proposed working on the anomalous Zeeman effect at that time, and Bohr was a little discouraging. He said that Pauli had just spent a complete year on this with him, and they had tried every avenue and had simply failed everywhere. And so he was a little discouraging. This is the kind of competition that I mean. The problems which were staring people in the face at that time were difficult problems and they had very fine people working on them.

Kuhn:

When you speak of the problems you thought you’d like to work on, though for one reason or another you didn’t, what else would you pick out as the problems of this period?

Dennison:

Slater had just been there. Bohr and Slater had written an article, of which the principal thing that I remember at this moment was that Bohr was seeking whether a way out of the difficulties might be, for example, in denying conservation of energy. This was something that I think Bohr had returned to a good many times with a feeling that this was the key to it, that perhaps some of these great conservation laws were not absolute. It never came to anything, but I think he was always prepared for the notion that they might break down in some fashion or other. This was one of the problems, to continue with this, you see. But again this had been studied pretty carefully by Bohr and Slater, and they had reached as far as they felt able to go.

Kuhn:

I’m not quite certain how the relevant experiments would date relative to your arrival, but I would guess that they were a little after your arrival. The paper may on the one hand have gone as far as it could go, but there was also then fairly rapid evidence from Bothe and Geiger and then Compton and Simon, that it wouldn’t do. I’m terribly curious to know anything you can say about the response to those experiments and the feeling about that paper, which indeed was a terribly important paper although it, in any literal sense, was almost immediately shown to be wrong.

Dennison:

I’m afraid I don’t remember very much about it actually. I started to work with Bohr on some of these things, but at that moment he was perhaps just casting around for any new approaches to things. I didn’t have the luck to hit on anything at this point that would fasten his attention. I know we talked of various things, but none of them seemed good enough to come on. After a month or two, as I say, it was kind of discouraging, and then it occurred to me to try some of the methods that I had used on methane for some other molecules which were somewhat simpler: carbon dioxide and ammonia and things of this type. I proposed this to Bohr, and I think Heisenberg was there at the time when I did it, and they felt that this would be a very good thing for me to work on. So I did. An historical note: eight or ten years earlier Bjerrum, a Danish chemist, had done a very fine paper on carbon dioxide. I went to see him; he was a relatively old man at that time, and he couldn’t really tell me anything, but it was very interesting to talk with him in his laboratory and see what he had done. He felt that he had gone as far as he could go, and he stopped at this point. It is a very good paper though, incidentally. So I continued working. I was somewhat apprehensive in this period as to whether any creative ability that I had might dry up. I was then, and for a good many years afterwards, very apprehensive of falling into a rut of some sort, very anxious to keep out of this, and to keep fresh approaches to things. I also felt that if I wasn’t able to do at least one piece of work a year, I was pretty badly failing in my awn estimation. The article on these molecules again was wrong in its interpretation. It was right in method. And I suppose I shouldn’t really say this, but I do think that the work on methane and on these other molecules which did for the first time give the correct methods of doing them, was the real start of all of the tremendous development that has taken place in molecular structure since that time. If you think of the thousands of papers that now exist, they really essentially all use this starting notion that came out of it.

Kuhn:

This is not in the least to suggest that that isn’t true, but another terribly important paper that is obviously a bit later than this, but whose historical role in view of its sort of mathematical purity I’ve been very uncertain about, is the Born-Oppenheimer paper.

Dennison:

This was indeed a very fine paper. It validated the whole method, of course, of treating molecules. I’m afraid that in a way I took this for granted. I was very glad that this paper existed, but it didn’t really bother me particularly. It seemed to me that it was evident that the way to approach it was to put in these empirical force functions and to find out, to make the interpretation of what was actually found experimentally. I’m not sure that it really had a profound influence on me because I just took it as self-evident that this must be so.

Kuhn:

That makes terribly good sense to me. I have some feeling that it may very well, on the whole, have, just about that role. It validated things that people either were doing anyway or would have had to do anyway when they came to those problems; it made it all legitimate. But if it had been impossible to make it legitimate, you would still have had to work that way.

Dennison:

Yes. Later I’ve gone back over this and thought of it a bit, and in the talk that I gave in Copenhagen last summer the first part was essentially on what are the fundamental things behind molecular structure. This paper, I think, is not too bad. It will probably be published, but only as a kind of internal paper because it was a closed conference and it was done that way. But sometime you might be interested to read it when it does finally appear. I continued to be very much interested, of course, in all the developments, read every paper that came out in the fairly new journal Zeitschrift fur Physik — this was, of course, one of the things everyone read.

Kuhn:

Had you been reading that regularly before you went to Europe?

Dennison:

I think so, yes. It seemed to me to be the livest one, that’s where all the new things came out. Yes, certainly I did. There were lots of interesting people at the laboratory. Nishina was there; he was at that time an experimentalist. He was using X-ray spectra to detect the small changes in the curvature of the potential functions, the energy functions, of the Bohr theory. It was the last bit of cleaning up on the old Bohr models, which were about to be swept aside. I think it was at that time that Bohr had his fortieth birthday, and there was a celebration. Pauli was there and had a cake made which had the orbits on the frosting and had a quotation from Goethe which I don’t remember very well, but something to the effect that “All important things are wrong.” [Laughter] It was already saying that this was all wrong but nevertheless extremely important, which of course one was beginning to know was so. Matrix mechanics was beginning to come out; I was certainly very much impressed with Heisenberg’s approach.

Incidentally, this is, I think, very interesting too, to notice with Heisenberg how the theme of observability has run through so much of his work. Just as Bohr had always this feeling of being willing to examine all kinds of catastrophic things, so I think this idea that it had to be observable has been an important thing for Heisenberg. Kramers was the one who was, I think, initially interested in this notion of systems which had constraints in them. He talked to me about it and I think it was through him that I got interested then in the article on the rotation of molecules, which used this method of constraints, which has, I guess, completely disappeared. There’s no trace of this at all anymore. But it was a question, you see. You have free particles: what are the rules on the matrices? Okay, but now suppose there are constraints: what are the rules there? Then there was a question of showing that here are certain sums of these that are still correct, and then I could make use of those in the rotation of molecules.

Kuhn:

You think that Kramers had already been interested in the problem of reformulation of the commutation conditions in the case of —

Dennison:

This particular modification with constraints, that’s right.

Kuhn:

That didn’t grow out of an attempt to find a way to handle your problem?

Dennison:

I don’t think so. I think he had this notion and then talked to me about it, and I think I must have seen that this was just what I wanted to use for molecular problems. But Kramers was very much interested in this, always was, and was very helpful to me in all kinds of things.

Kuhn:

By the time you arrived in Copenhagen [at the beginning of Dennison’s second year in Copenhagen], the group must already have known of Heisenberg’s very first paper, though I don’t think it would have been out yet. But I think he’d already talked to people there about it.

Dennison:

Yes he had.

Kuhn:

I’d be terribly curious about any early discussions in Copenhagen or any discussions in this whole year about matrix mechanics, whether this seemed pretty clearly the answer at a fairly early stage.

Dennison:

I’ll give my version which, of course, may or may not be correct; this is just a recollection, namely, that during the first part of the time when I was there, Heisenberg was more concerned with other, and in a way, less important problems. Now I don’t even remember exactly what they were, but they were spectroscopic problems of some sort or other, and he was groping for ways of dealing with them. My impression is that he was not in Copenhagen at the time when I first saw, either through a manuscript or something or other, his article on the matrices, and that I studied it pretty much by myself, actually, not so much through discussion.

Kuhn:

Now, of course, before the end of the year — I mean now the calendar year, not the academic year — one also had the Born-Jordan and the Born-Heisenberg-Jordan papers. Those you obviously also saw.

Dennison:

I certainly must have studied them all too, and I don’t remember that they made as much impression on me. The matrix mechanics paper made a tremendous impression.

Kuhn:

When you say the matrix mechanics paper, you mean this very first paper of Heisenberg’s?

Dennison:

Yes, the one of Heisenberg in which he was able, for example, to obtain the energy of the oscillator, and in which he really introduced the matrix multiplication. This was a very important —

Kuhn:

Although he didn’t know that’s what it was.

Dennison:

Probably not, probably not.

Kuhn:

That comes really with the whole —

Dennison:

With Dirac.

Kuhn:

Well, the commutation laws come with Dirac; the notion of the explicit tie to its being an older thing known as matrix multiplication really comes first with the Born-Jordan papers, which go over the same ground now in explicitly matrix form, introduce the commutation relations; then it’s carried still further in the third of those three papers, in which Heisenberg is back in the group again.

Dennison:

You see, this is one of the things which was my weakness. I did not have the mathematical background to have ever heard of matrices. This was the first time I’d ever heard of them or of the notion of matrix multiplication. In fact, to show how naive I was, the notion that multiplication might depend on the order was one which I really just didn’t have at all.

Kuhn:

There weren’t very many physicists who did, including some with a great deal more mathematical training. Of course, with Born-Jordan it comes immediately to the center of things.

Dennison:

Sure. Then they know, they knew. Well, probably the Born-Jordan made less impression on me than this initial one of Heisenberg’s, because this opened it up; here was something that I’d never have thought of and had never seen before, and suddenly it appeared.

Kuhn:

Now what about other people at Copenhagen?

Dennison:

Well, Hund was at Copenhagen, and Hund, of course, during this last year was doing a good many interesting things in molecular structure. He did really some very fine things, and I built a great deal on his work. As a person, he was not an awfully agreeable guy to work with. I don’t just know why this was, perhaps some psychological trick or something. He had a tendency to be a little scornful of other people’s work, probably coming from an inner weakness of some kind. I don’t know. He was also very scornful of Ehrenfest’s work, and I know he and Sam Goudsmit had some very bitter words at one time because of this. Of course, Ehrenfest was Goudsmit’s teacher. Well, just to enumerate some of the people who were there: L. H. Thomas was there; he’d gotten a fellowship. I talked with him a good deal. A Swiss, Werner Kuhn, whom I met in Copenhagen last year; someone told me he had died this fall.

Kuhn:

I think that’s right. Yes.

Dennison:

It must have been very suddenly. Then there was an Indian, [B. B.] Ray, who I think has also since died. There were a number of Japanese there: Nishina; Sugiura, a theorist; and Kimura, an experimentalist who got tuberculosis while he was there. He was very well looked after by the laboratory, and he recovered. I met him in Japan a year and a half ago; he’s now a big shot in atomic energy work there. And a Japanese, Hori; just what period he was there I don’t know, but he certainly must have been there during the spring of ‘27. That’s when I principally remember him.

Kuhn:

Was it at Copenhagen that he did the work you referred to?

Dennison:

Yes, at Copenhagen. I think that’s right. At any rate, he was there and it was from him that I learned of his work on the moments of inertia of the molecule.

Kuhn:

…I have some feeling, though it’s on very inadequate evidence, that in practice, except perhaps for Kramers, there wasn’t an awful lot of interest in matrices at Copenhagen. Is there anything in this?

Dennison:

I think you’re probably right if you mean by this the application of matrices to problems. The importance, however, of Heisenberg’s work and of the concepts that were coming up, I think were thoroughly appreciated and were very strong. This much I think is correct.

Kuhn:

Do you remember any discussions of them, colloquia or private discussions with people?

Dennison:

No, I don’t really remember them. There were not so very many colloquia when I was there, and I had discussions with various people, but again I suppose because of my trick of working mainly by myself, I have never gotten as much from discussions as I would have gotten otherwise. It’s this business of enjoying doing things by myself always.

Kuhn:

Were there lectures there that you went to?

Dennison:

No, no, there were no lectures. And the colloquia were really quite infrequent, and I gather much more infrequent than they were at a later time, after I had left, during the next, so to say, great rebirth during the nuclear phase of the work. I think you mentioned a paper that was on carbon dioxide, written in German [paper No. 5]. This pretty much arose from a misunderstanding. I had seen, I think it was Schaefer, and saw what Schaefer was doing on carbon dioxide, and I had understood that we were going to more or less publish something jointly on it. And so I did write this article which I was never very sure about. I didn’t think there really was much content to it, and it was done, in a way, because I felt that Schaefer was waiting for an article from me in order to publish his stuff. Now actually he wasn’t; he had gone ahead about publishing it anyway, so this one was somewhat of a mistake. I don’t think it had really enough essential content; I wasn’t ready to write anything at this time, but I felt it was something I had to do and so I did it. So this one we just write off actually.

Kuhn:

About the time you got to Copenhagen, and even before the Heisenberg work, there is the other great big break, which is spin.

Dennison:

Yes.

Kuhn:

And that goes on while you’re there. You also pointed out that Thomas was there. That is the year of the Thomas factor of 2.

Dennison:

Yes, that’s right.

Kuhn:

I wonder if you have recollections relating to that.

Dennison:

I’m afraid I don’t have very much. I remember Goudsmit’s being there all right and talking with him a bit, and the fact of the importance of this was quite clear. I think that perhaps by this time I had become so interested in the molecular problems that I didn’t have quite the complete openness that I had when I first came to Copenhagen. When I first came it really wouldn’t have mattered which problems I had taken up; any of them were perfectly all right. But having gotten into the molecular problems, I think I was just taken up with this work, you see, and in a sense did not pursue these others so much. So I don’t have much to offer, really, about it. I was very much influenced by Heisenberg’s matrices and was very anxious to go on with them and to understand molecular problems. We came back in the summer of ‘26 and then returned, and we went next to Zurich.

Kuhn:

…You were in Europe for two years, and were in Copenhagen steadily for those two years, were you?

Dennison:

Yes, essentially. Vacation during the summertime in Switzerland, but essentially in Copenhagen.

Kuhn:

In the spring of your [second] year the Schrodinger papers begin to come out… Did those make initially a deep impression?

Dennison:

Yes, yes. Certainly. Perhaps not as much as the Heisenberg one had, but still a very deep impression, that’s right. They certainly did. Why did the Heisenberg make — as I look at it now, why was that so important? I really don’t know; it just struck me as so important to suddenly look through things. I really don’t know the answer to it, although I certainly studied the Schrodinger ones very carefully and knew about them, was prepared to work with them, to go ahead.

Kuhn:

Would you say that in general in Copenhagen they created more interest, or simply more follow-up, than the matrix approach had?

Dennison:

I’m afraid I just don’t have too much to offer here actually. In a way I’m afraid I probably am inclined to minimize what they were, because, as I say, I remember so well studying the Heisenberg one and what an impact this business of non-commutation meant. Yet I certainly must have thoroughly worked with the Schrodinger ones because I began using these methods right away. But it just is a little blank as to just what it did at that time.

Kuhn:

…You speak of the fortieth-birthday cake and that everything really important is wrong. In some places, and I think perhaps more clearly than anywhere else this would have been true in Gottingen in this period, it had become totally clear to people that there had to be a really, really fundamental break. It wasn’t any longer just that there were conflicts between quantum and classical, or that certain problems just weren’t coming out. It was to the extent that, as you know, Born called his Atommechanik ‘Volume I’ and says, “We’ve got to change all of this drastically, and that’ll be for Volume II.” I wonder to what extent this whole idea, in the first place that this wasn’t right, was new to you when you came to Copenhagen, and in the second place, how deeply not right; what that seemed to mean to you and other people there, to what extent the notion of anything so drastic as the Heisenberg approach really came as a shock even though you’d known things weren’t right.

Dennison:

You’ve partly put your finger on it, all right. From what I had studied here, I had the feeling that a lot of Sommerfeld’s work was right and was fundamental. It seemed to work in so many places; it seemed so very powerful. The general attitude in Copenhagen, I think, was more, “Ja, it’s remarkable that it works, all right, and interesting, but it isn’t right, and this is not the way to proceed further, to try to make little refinements on the Sommerfeld approach.” Perhaps one was even a little scornful about it; it’s too childish; it’s essentially missing the real point of what’s happening here. This seemed a bit shocking to me. However, I think I was not shocked in the sense that I wanted to hold on to it at all. I felt undoubtedly it [the criticism] was correct. Perhaps I was just a bit shocked at the fact that Pauli so clearly was saying that all these wonderful things that Bohr had been doing with atoms and which agreed so well in so many details were all essentially wrong and were about to be swept aside.

Kuhn:

How clear was it to you that this also meant “and was about to be swept aside”? Put it this way: it’s not clear to me that Pauli would have meant that… To the extent that the correspondence principle was really something philosophically fundamental, one could say that these things are wrong without saying that they’re about to be swept aside, because one could hope to simply keep on narrow —. The whole idea of the extended correspondence principle was one that provided, if you like, even a philosophical justification for working with the wrong — with no notion that anything so drastic as the Heisenberg or the Schrodinger approach would be the next step, but rather that the continual refinement and revision that had done so much over the three or four preceding years would continue.

Dennison:

When I say “wrong” and imply that they felt it was about to be swept aside, I think I am really thinking of it as a literal, detailed picture, which after all many physicists were believing at this time. It hadn’t been many years since Bohr himself in a sense was believing it; at least he was drawing all of these pictures and was inferring what was happening with the energy levels as you went from atom to atom. It was this kind of detailed thing which I think people thought was wrong. The importance of the correspondence principle at this moment was, in my recollection, as a guide to the way in which it was going to be swept away and to any new theories that would come. It may well be that at this moment they had no very good idea of which way it was going to come. This was, of course, one of the things Bohr often said, that if you only knew even the direction of the next important discovery, then the next important discovery was already made. I’m not sure whether it was really known at this time, but there seemed to be just a conviction that this particular approach was finished. There was no more to be done here and there was something else to come. The correspondence principle, I think, was probably held as a guide for the new things which would come, which I think is correct too. There aren’t so very many people who believe in the correspondence principle nowadays. I do find myself — this is just an aside — that many times questions are asked which I can find the answer to very rapidly through correspondence principle considerations, whereas the next generation, which has never been through this, will not find the answer easily at all. Between several choices as to what happens in a given situation, they will often be quite at sea and perhaps pick the wrong one, when the correspondence principle is often very helpful and tells you that it must be this way and not any other. This gives perhaps a little of the feeling of the way things were. I still don’t quite hit what you’d like in detail.

Kuhn:

This is all right. I don’t in any case want to say it myself because I’m trying to fish in areas and simply see what kind of reaction — this is certainly responsive to my question. Do you remember any discussions in Copenhagen or, for that matter really, back at Michigan, that revolved in the pre-Schrodinger or immediately post-Schrodinger period, about the problem of light as a wave and as a particle? This is, of course, exactly the, area of the Bohr-Kramers—Slater paper.

Dennison:

During the first year it was certainly a great puzzle, the first year that I was in Copenhagen. Let me say that the general feeling here before then, when I was a graduate student, would have been, more along the lines that Maxwell theory was successful, but was pretty old-fashioned. Photons and photoelectric effect and things of this sort were certainly the proper way of describing it. It just about stopped there. In other words, omit the fundamental feeling of the real difficulty with this situation. When I was here as a graduate student, the profound antagonism between particles and waves was not appreciated. I certainly didn’t know, I’m sure, at that period, just the fundamental thing that to define a wave you have to have at least a number of wavelengths. I’m certain this had never occurred to me at all.

Kuhn:

So the whole problem of interference of photons, which you do a little bit later for matter waves, really was not an acute problem and indeed you were quite aware of it as a graduate student.

Dennison:

No. The two things, as I say, were separate. Here was Maxwell theory — ‘yes, this was right; it would give you all of these phenomena.’ It was, however, a pretty old-fashioned sort of thing, kind of dusty. At the same time, however, here were the new things. Light really wasn’t a la Maxwell; it really was something else. Completely lacking in my mind at that time was the appreciation, so to say, of the profound inconsistencies in just talking about photons and still giving them, say, an energy hυ. What under the sun does υ mean under these conditions? This I don’t think was borne in on me. This, however, must have been much more closely borne in during the first year in Copenhagen. And I remember that we went down to Switzerland during the summer of ‘25, and Pauli was at that time in one of the German universities.

Kuhn:

He was at Hamburg, I think.

Dennison:

Hamburg. That’s right. And we stopped there, and I went around and looked him up at the Institute. My German was pretty bad, and I think we tried it in Danish, which was not very good either. I asked him this question: ‘What did he now think? Was light waves or particles?’ And I can remember this kind of cynical laugh of his, you see. What was it? I don’t think he made any answer at this point at all. That it was a profound question, that’s right; but that there was as yet any answer, no. At least he didn’t talk about it. But it was certainly one of the things which was very much concerning the people at Copenhagen in that period… I was probably a bit chagrined at this rather cynical laugh of his. Of course, Pauli could be a little bit sharp at times. Actually, to me he was very nice always. This has nothing to do with anything except it just shows that Pauli was sometimes very nice to people too. When we finally left Copenhagen, I think to go to England in the last year — Pauli was in Copenhagen — our train left fairly early in the morning. I can still see, just before the train left, Pauli came walking down the platform, eyes just practically slits, you know, because he never got up early in the morning, to see us off. The most wonderful thing. We were terribly impressed because he never got up ordinarily, you know, and he was just barely able to see his way down the platform to say goodbye to us. [Laughter] He was very nice.