Robert Marshak - Session II

Weiner:

Was this work ever published?

Marshak:

Well, I think that after the war years a series of books was published on the MIT Radiation Lab work. But I’ve never tracked them down, I’ve been so busy since then. I believe Nathan Marcuvitz of Brooklyn Poly Tech edited the volume on our calculations...

Weiner:

There was a series of OSRD reports. It was probably published in that series.

Marshak:

Yes.

Weiner:

But you don’t have this as one of your publications.

Marshak:

No, I guess I ought to add it to the list.

Weiner:

Same way as the neutron diffusion papers. But those were ultimately published in science journals whereas the radar work was not.

Marshak:

The neutron diffusion papers were handled differently. After the war, I brought together about ten papers on neutron diffusion into a review article for the Reviews of Modern Physics in 1947 and did not try to publish them separately. But I never got around to publishing the radar papers.

Weiner:

Where you worked—was it in the Radiation Laboratory itself or did you work with your group in a separate place?

Marshak:

Well, during the summer months of 1942 we had offices in the Radiation Laboratory, and we worked right in the Radiation Lab. But during the school year, I stayed in Rochester except for trips during college vacations. Every vacation I went to the Radiation Lab: I remember the Christmas of ‘42 because I had met my future wife in the Spring of ‘42, and we had many telephone calls, but I spent all the time during that Christmas vacation working away on Radiation Lab problems. But when it was not vacation I was in Rochester teaching and trying to help the students complete their degrees (e.g. Herbert York completed his M.S. degree under my supervision) in preparation for my departure for Montreal in April 1943.

Weiner:

It was planned that far ahead.

Marshak:

It was planned, I would say roughly, on the order of six months before departure.

Weiner:

Was Slater involved? Did you have any contact with John Slater? Was he in that particular...?

Marshak:

He was not in the so-called Bethe group, but presumably he was connected with the Radiation Lab in some other way.

Weiner:

I can check it out.

Marshak:

Goudsmit was very active. There were many people at the Radiation Lab. I did not get to know too many people because of my brief visits to the Radiation Lab, and because the rest of the time was spent at my university. I would send my reports to Bethe who would then transmit them, for whatever value they might have, to the main lab.

Weiner:

Since your main work was in the Bethe group, it wasn’t a question of coming into contact with large groups of people you hadn’t had much to do with before.

Marshak:

That is basically true. During the Summer of 1942, of course, I listened to seminars and got around a bit. But essentially Bethe accepted responsibility for telling us what some of the key problems were, and we picked what we thought we would like to work on and kept the research activity moving.

Weiner:

It was really one summer.

Marshak:

One summer, and long distance at Rochester.

Weiner:

Then I guess the next thing is 1943. A couple of things happened: you got married, and I know you went to Montreal. But wasn’t it your plan to go to Cambridge in 1943?

Marshak:

No. What happened was that George Placzek, I guess some time in 1942, probably in the Fall of 1942, said that he needed some theoretical people for an atomic project in Canada, under the control of the British government, but housed in the new unoccupied medical wing of the University of Montreal. My understanding actually is that Weisskopf originally was supposed to go to assist Placzek, but he was not an American citizen and felt this might create some problems, leaving the U.S. during the war and so on. I guess he must have told Placzek to contact me, which he proceeded to do. I finally agreed that perhaps this was something that would be useful for the war effort. And so I joined the theoretical group of this Canadian project but as an employee of the British government. There were many Free French there and indeed the Director of the Montreal Laboratory was initially Hans Von Halban. It was here that I met Pierre Auger, Jules Guéron (who later became head of Euratom), and so on. That’s where Bruno Pontecorvo was and the British scientist, Alan Nunn May. My recollection is that the only other American was Ernest Courant, who was persuaded to augment the theoretical staff, as was I. I stayed in Montreal just about a year. Now, are you interested in knowing about the scientific work at the time?

Weiner:

Yes, in Montreal, because I have an incomplete view of it. First of all, although you were an employee of the British government, was your responsibility to Groves essentially?

Marshak:

No, I was considered an employee of the British government and on loan. I vaguely remember that permission had to be obtained from the director of the U.S. Scientific Personnel Office under Vannevar Bush. These formalities were handled at the highest level through Von Halban and our Washington officials, and it was arranged.

Weiner:

So the radar was forgotten about.

Marshak:

Yes, and so I moved on from the radar project to the Montreal project. In many ways, the war projects in which I was involved were diverse and very interesting from an intellectual point of view. I had quite a variety of exposures, and I think it was very helpful in terms of later scientific work to have these contacts and to work on these different types of problems. Now, when I came to Montreal....

Weiner:

First of all, you got married.

Marshak:

Oh, yes, I must not forget that. We were married April 18, 1943, and our honeymoon consisted of driving to Montreal to find an apartment so I could get started on my new job. My wife then returned to finish up her school work. She was a teacher of primary grades. She had committed herself to stay until the end of that school year, until June 1943. We found an apartment in a new housing project in Montreal which would be ready by July and so I moved into a sort of bachelor’s house for several months. One of the inhabitants of that house was Alan Nunn May, who later turned out to be one of the well-known spies. But in that lies another story. As far as the work was concerned, Placzek had become quite interested in neutron diffusion problems. That’s why he was a natural leader for this group.

I think he had some interest in this subject before the war because the slowing down of neutrons was of interest for neutron measurements. That is, you produced neutrons at a certain energy and you inserted some hydrogenous material; this material then acted as a moderator because the neutrons would undergo elastic collisions with the hydrogen nuclei and thereby lose energy by means of elastic collisions. You would work out the diffusion equation and find out what the distribution of slow neutrons was as a function of initial energy and so on. Hence it was natural to ask Placzek to head up this group. When I came there, he said, “Well, we ought to improve some of the theoretical methods in neutron diffusion,” and I started looking at those problems. I decided that I would start first with the rigorous Boltzmann equation and see what happens after you make various approximations that yield the simple diffusion equation in the first instance. That’s how I discovered the “spherical harmonic” method. The spherical harmonic method did have the virtue that it led naturally to the so-called diffusion approximation.

The diffusion approximation was not good enough for the bomb because its dimensions were of the order of the mean-free path. To put it another way, if the dimensions are large compared to the scattering mean free path, then the diffusion approximation is a good one; otherwise, one must do better. I found that if I just kept on expanding in spherical harmonics, I could develop additional equations which were the next approximations and the convergence was surprisingly rapid. For example, I found that under realistic conditions, I could get a very big improvement in the predictions of the relevant parameters by expanding to the third spherical harmonic. Now, actually, the question was: How did I know at Montreal that I could make such good predictions because there were no direct experiments? Well, in order to check this point, I developed the variational method for neutron diffusion problems. In this connection, I had learned something from Schwinger in the Radiation Laboratory. He had developed a variational method for electromagnetic problems which was very elegant: I had remembered that variational methods are very powerful and converge extremely rapidly. And so I extended the variational method to neutron diffusion problems. I then could compare the results of the spherical harmonic and variational methods for the same problem. These two methods were further refined by the Montreal group, Carson Mark, B. Davison and so on.

Carson Mark wrote many papers on the spherical harmonic method and Davison later published a book which gave a great deal of attention to the variational method. The interesting thing was that when I came to Los Alamos a year later (and I’ll explain why I left Montreal to go to Los Alamos in a little while), I immediately told the theoretical division about the spherical harmonic method. For some reason, the reports from Montreal had not reached Los Alamos, or if they had, they hadn’t registered. But I think they hadn’t reached it, because there were great complications in transmitting reports. General Groves did not like the Montreal project. He was very suspicious of the director, and he did not make available, for example, Canadian uranium for this Canadian project for the experimental group. Hence there must have been also some confusion in the way the reports were being transmitted. I’m not sure. It may be that our reports reached Los Alamos and they didn’t read them. But in any case, when I arrived at Los Alamos, I immediately told Bethe about this spherical harmonic method, and it turned out that this was a more powerful method than those developed there, by smart fellows like Feynman, Serber and so on. I remember Bethe saying, “You’ve already paid your way by bringing this method here.” So the spherical harmonic method gave pretty darned good results even though I never claimed that it was a terribly elegant method. It was a natural extension of the diffusion approximation, and I guess one had to have the courage to continue calculating. It was more rapidly convergent than one had any right to expect, and it was certainly extremely useful in making predictions for small nuclear systems. And so you can understand why I was rather pleased with the invention of the spherical harmonic method. Moreover, being aware before the war that there were stellar atmosphere problems which up to that point had only been treated in the diffusion approximation, I said to myself: “When I get back after the war, I’ll apply this method to stellar atmospheres.” However, here I was scooped pretty badly by Chandrasekhar, because he was not involved in a secret atomic project, and he had read some papers of Giancarlo Wick, who had developed another version of the spherical harmonic method after I did.

Weiner:

Where was this?

Marshak:

In Italy. Chandrasekhar had read Wick’s paper and was taken with it. I think the Wick method was originally developed for a plane situation but by the time he wrote paper number 5, Chandrasekhar had become interested in the spherical case, which is actually the one I started with, because the A bomb is spherical! Chandrasekhar proceeded to quickly apply Wick’s method to a large number of problems in stellar atmospheres and probably wrote 50 papers in this general area during the next five years. I think they’ve been collected in a book. So that after the war ended, there was no point in my trying to apply my spherical harmonic method to stellar atmospheres. I did publish one paper in the Physical Review in 1947 on the spherical harmonic method, because, despite Chandrasekhar’s many papers, there were some differences worth pointing out between the two methods. As I said, because I developed my version originally for the sphere, the boundary conditions that I used were somewhat different in my method than in the Wick-Chandrasekhar method; for certain types of calculations, my approach was somewhat superior; so there seemed to be merit in publishing at least one paper on the spherical harmonic method. But with that I stopped.

Thus I published three papers in the Phys. Rev, on my Montreal work. I should emphasize that these three papers on “The Spherical Harmonic Method Applied to the Sphere,” the “Variational Method for the Asymptotic Neutron Density”, and the “Milne Problem for a Large Plane Slab with Anisotropic Scattering”—-all refer to monochromatic neutrons. What you are doing is calculating the distribution as a function of the direction and coordinates in a substance of scattering material, whether the geometry is plane spherical, whether the basic scattering process is isotropic in the center of mass system or anisotropic, etc. Well, each of these papers was a distinct contribution. The variational method I have already referred to. This was the first time the variational method had been applied to neutron diffusion problems. And again the Montreal group, after I left, carried out this program much more extensively. I used the simplest trial functions—I think constants for the functions—and got answers correct to three significant figures. The Montreal group went on later with one extra function, an exponential, and obtained answers correct to seven or eight significant figures. It’s a very powerful method.

Weiner:

In that paper, “Variational Method for the Asymptotic Neutron Density,” when finally published you mentioned Schwinger and Davison. You thank Schwinger for conversations and Davison for cooperation. What was the...

Marshak:

Well, Schwinger I mentioned. I don’t think I actually discussed the details with him because I was in Montreal, but I thanked him because I had taken a lead from his work in the application of variational methods to electromagnetic theory, so I felt...

Weiner:

But you mentioned conversations, so I wondered. Maybe it was while you were writing it up. Were you in Los Alamos?

Marshak:

I was in Los Alamos but Schwinger was not and the work was classified. I just felt generally grateful because he had been so successful in applying the variational method to electromagnetic problems and I had had conversations with him at the Rad Lab on radar problems. I do remember specific discussions with Davison, because I just used physical intuition to argue that my approach would give an extremism, and Davison later picked that up and gave a mathematically rigorous proof.

Weiner:

Which Davison was that?

Marshak:

This is B. Davison who later published a book. He was a British mathematical physicist who had come to Montreal towards the end of my stay, and then he stayed on until the war ended and perhaps beyond. He later published a book on neutron diffusion problems, and in that book gave detailed discussions of fine points in these variational methods, including his own work where he demonstrated rigorously the mathematical correctness of my hypothesis, which I didn’t bother doing. I thought that physical intuition was sufficient to argue for the existence of this extremism. With regard to the third paper, the Milne problem, there’s a famous little book by Wiener-Hopf applying elaborate methods of complex variable to the calculation of the asymptotic behavior of a semi-infinite plane slab with isotropic scattering. I was able to generalize the Wiener-Hopf method to anisotropic scattering, and I sometimes give this paper as an illustration of the fact that I could be pretty sophisticated in mathematics if compelled to! That paper represented the peak of my mathematical sophistication.

Weiner:

It involved complex variables?

Marshak:

Yes. These are the types of methods that later became important to dispersion theory in particle physics. However, my interests in particle physics have always been in somewhat different directions, and I’ve never gotten around to serious use of my old knowledge of complex variables to deal with particle theory problems. Now, the next paper was a review article-I developed quite a few things there in that Montreal year. I would say that was a very fruitful year. A year after my arrival in Montreal, in April 1944, I left for Los Alamos. And the last Montreal problem was “Slowing Down of Neutrons Through Elastic Collisions with Nuclei”. Well, you have, say, a monoenergetic source of neutrons and you put it in a medium, a physical medium, whatever it is—carbon, or hydrogen or a mixture of elements—and you want to do fairly rigorous calculations as to what the distribution in energy and as a function of the coordinates is. Fermi had been very interested in that problem and developed again a diffusion type of approximation which was called the “aging” theory. Again, he derived the key term in the expression by his own physical type of argument. There also existed a rigorous formula for the slowing down of neutrons in hydrogen.

I think Uhlenbeck had derived it by means of an elaborate mean-free path method—I would say a “kinetic theory” method. Well, I think it was Placzek who introduced me to the usefulness of the Laplace transform method. And I was an eager beaver working on these problems, and used the Laplace transform method to obtain a rigorous solution. I could recapture the Uhlenbeck formula in a few lines with the Laplace transform method. This formula, by the way, using the same method, was independently discovered by Schwinger. I don’t know whether he ever published it, but he told me he had also derived it. I worked on other types of problems with the Laplace transform method in energy loss and Fourier transform method in space. The point is: the Laplace transform essentially converts a slowing down problem, where the energy is changing, to a monoenergetic problem with absorption in Laplace transform space.

So if you can solve the latter problem, then you can invert the Laplace transform solution and get your answer. You can use Fourier transform methods to convert from ordinary physical space so that you get Fourier-Laplace transforms of your density functions, and then you can try to invert and so on. Well, this led to a variety of results, most of which I think were fairly useful in terms of giving better answers for slowing down lengths and for things of that sort that could be compared directly with experiment. After the war I decided that there was no point in trying to publish these papers separately, and I wrote a Review of Modern Physics article that incorporated all of these papers as well as others that were relevant. It was a sort of complete summary of the situation in slowing down theory at that time. I was told later, in 1956 by Landau the first time we visited the Soviet Union- - their work was declassified—that my review paper was extremely valuable to the Russians in their work.

Weiner:

You had done the review paper as a consultant at Los Alamos in the summer of ‘46. Had you returned to Los Alamos specifically for that purpose?

Marshak:

Yes, I had forgotten that. Los Alamos had asked me to be a consultant for a while, and there were quite a few papers that still had to be finished for Los Alamos. Let me say this: not all the results, my own results published in my review paper, were based on my Montreal work. In particular, the latter part of that review article, having to do with the asymptotic neutron density, which was a very difficult problem, were things that I worried about at Los Alamos. So when Los Alamos asked me to write up all the papers in connection with my work there—and I left Los Alamos in February 1946 to come to Rochester—I took along a lot of classified documents and worked in my spare time (I did not pick up on the unclassified research, but tried to fulfill my obligation to Los Alamos while I was teaching at Rochester). I wrote up quite a few papers in Rochester but did not finish everything. Hence, in the summer of ‘46 I went back to Los Alamos: I made a deal with DuBridge during that Spring semester of ‘46 so that I could spend part-time writing up these papers. In the summer of ‘46 Los Alamos gave me a summer salary so that I could spend full-time writing up Los Alamos-connected work and this is one of the things I finished there. For the sake of completeness, I included quite a bit of the Montreal material.

Weiner:

In this paper you mention specifically that slowing down of neutrons plays a role in the moderation of cosmic ray neutrons in the atmosphere; the shielding of high-voltage accelerators against neutron background; and then pile design. So it wasn’t merely a report on wartime research, but it was written with some nuclear physics applications in mind.

Marshak:

I did see the general picture. And I think this was the justification used as an argument for declassifying the RMP paper. Certainly S. A. Korff, for example, who had worked on neutrons in the cosmic radiation—certainly he would find my article relevant to his work. Indeed, it was relevant wherever there were neutrons being slowed down—in the atmosphere, shielding problems, reactor design. So it was certainly a correct statement, but I personally did not then try to work out these additional applications.

Weiner:

Now I’d like to talk about Montreal in general, about the atmosphere there, the relationships, the special role that you had as an American within a British group and what you perceived as going on. There were a number of interesting personalities in their own right. There were also a number of historic things that happened. And this is all by way of background to your leaving too. I don’t know whether one thing had anything to do with the other. It’s a year’s period, and I’m curious about it.

Marshak:

Well, maybe I should speak to the last part first. This was an international group. As I said earlier, there were British citizens there. Alan Nunn May, Bill Arrow, who was in chemistry, with whom my wife and I became very friendly; Frank Jackson, who was an administrative officer, and his wife—we became very friendly with them; the Free French group I mentioned earlier; the Canadians: Phil Wallace, who is now at McGill; and George Volkoff (of Volkoff and Oppenheimer fame on the subject of neutron stars) was on the project—he was until recently chairman of the physics department at the University of British Columbia. So it was a very interesting group of scientists and we would see many of them socially. With regard to the work itself, our theoretical group, which had Placzek as its leader, consisted of Volkoff, Wallace, Courant, myself and a few more—it was not a large group; we saw a lot of each other.

But as far as I can tell from the list of publications, most of the papers were essentially done by myself without too much collaboration. I think in many cases I thanked Placzek, and he certainly was a stimulating colleague and in the initial stages very important in terms of defining what the problems were. He was always a good consultant. But, for example, in the spherical harmonic method I went beyond what other persons had done before; ditto for the variational method. Placzek was always very interested in the Milne problem with the usual Wiener-Hopf method. He liked that method very much. He knew neutron diffusion theory well and he certainly introduced me to the general problems. Now in terms of interaction with the experimentalists at Montreal, there was not too much going on because the experimentalists were stymied by the fact that they could not do experiments without a uranium pile. They complained that they could not get the uranium because General Groves mistrusted the director, von Halban, and would not release any Canadian uranium for experimental purposes. The U.S. had signed a contract with the Canadian government to obtain all of its uranium, and the U.S. would not release any uranium to the Montreal project. And there were all kinds of complications with reports. We did not get too many American reports.

Weiner:

You mean reports from American scientists?

Marshak:

Right. For example, I don’t recall ever seeing any reports from Oak Ridge. I’m not talking about Los Alamos, which was understandably off bounds-no Los Alamos reports went out to the other projects because they involved atomic bomb work. But, for example, I think we developed much better methods than Oak Ridge in neutron diffusion, and they didn’t know about our stuff, as far as I could tell, and we didn’t know what they were doing. So Groves put a stop to the exchange with the other projects because of his suspicion of the Montreal Lab.

Weiner:

Why do you think he had this feeling about Halban? Because of his German background?

Marshak:

Perhaps.

Weiner:

Or because of his style?

Marshak:

Perhaps both. I was a little too young, you know, to be taken into those inner confidences. I was only 26 or 27. I heard some stories, but I did not get to know many senior scientists until the end of the year, but then I was off to Los Alamos. But the explicit reasons I wouldn’t be sure of. Groves just was suspicious of Halban who had fled, for example, from the Joliot-Curie Lab with his bottle of heavy water. I guess Joliot-Curie’s political views were known publicly by that time—that he was fairly leftist, not yet a Communist, as far as I know.

Weiner:

Not yet. He declared during the war that...

Marshak:

But Joliot-Curie was pretty leftish oriented. I think the fact that von Halban came from the Joliot-Curie lab perhaps was relevant. And then partly it may have been von Halban’s personality, which didn’t please Groves - they didn’t click.

Weiner:

How would you characterize him?

Marshak:

Well, von Halban was a fairly self-confident and rather arrogant fellow, with a tremendous ego and a sort of aristocratic attitude. I guess Groves perhaps was not accustomed to that sort of style. I might say that underneath Halban’s exterior was an extremely warm and generous person. My wife and I had a direct experience of his kindness a few months after my arrival in Montreal. When my wife rejoined me in June 1943, she suddenly discovered that she had serious eye trouble - that she had an unusual case of cataracts. It was a question of really pinning down the cause and trying to arrest the disease, and it then became a challenge to find real specialists to examine my wife. When Halban heard about the problem, he immediately offered to set up appointments with New York specialists—German doctor friends of his; he thought that they were the best in the world. Halban also arranged to pay for all the medical bills, because we were not covered for that sort of thing. So this was extremely unusual solicitude for another human being. I had no special relationship to Halban except, of course that Placzek was a good friend of his, and I was in Placzek’s group.

Weiner:

How did this come out?

Marshak:

Well, we went and saw some of well-known German doctors of all kinds. It turned out their view of things was not the view of American specialists whom we came upon somewhat later. Their treatment was much too drastic and could have been dangerous. But that’s not the point of the story, of course. The point is that von Halban pitched in in a way that was beyond the call of duty and exhibited that part of his character that perhaps was hidden by his rather arrogant approach to everyone around him. Now, whether it’s a combination of personality, a combination of the Joliot-Curie lab, I’m not really sure. I never discussed this problem with Groves, and I think you would get that story somewhat clearer from someone like Lew Kowarski.

Weiner:

How about the relationship of von Halban with the other people in the group? Was it a smoothly functioning thing or was it in a state of shambles after a while? I know there was considerable difficulty, which may have been due to the attitude that Groves had toward the group and its lack of real definition in terms of mission. Or could it be attributed to personality questions within the group? I don’t know if any of that occurred when you were there. It may have occurred later.

Marshak:

Well, I decided to leave because I felt that as an American it was rather foolish for me to stay in a laboratory where the experimental division was not able to perform crucial experiments and I could not have much interaction with the experimentalists. I felt, after a year, that I had pretty much exhausted the interesting problems in neutron diffusion, and I had been in touch with Bethe who, by that time, was indicating that he really could use more people at Los Alamos. I mean I could have gone to Los Alamos right at the start, but there wasn’t that sense of urgency in early 1943 of having so many scientists at Los Alamos, and since Placzek, I presumed, had checked his invitation to me with Bethe I decided to go to Montreal initially. By the beginning of 1944, there were some crash programs at Los Alamos, and Bethe had indicated that the theory division really could use more people and he would like me to come. And I figured that I would be more useful at Los Alamos than continuing at Montreal. So I made the decision to leave in January or February of ‘44 and I left in April, something like that. But just a few weeks before I left - and I guess just around February or March after I had made my own small decision - a major decision was made to replace Halban by John Cockcroft as director of the Montreal lab. And so Cockcroft actually appeared on the scene a few weeks before my departure, which had already been scheduled, and I had no idea what effect his presence would have. But I came to understand in later years that once Cockcroft came on the scene, General Groves regained confidence in the director, released the uranium and that laboratory became very busy.

Weiner:

How about relationships with other people there? You mentioned people who later became notorious as a result of their involvement - Alan Nunn May, Pontecorvo. Was there any particular thing? You just mentioned this group...

Marshak:

The only thing, again from hindsight, you might connect with Alan Nunn May’s later behavior was his great interest at Montreal in preparing a bibliography on Russian work in uranium fission. I recall that May called my attention to some papers that the Russians had written on spontaneous fission and that sort of thing. But at that time this interest seemed very natural.

Weiner:

Turner I guess, at Princeton had written a review article in 1940 on fission.

Marshak:

But it was clear that Pontecorvo and Alan Nunn May were particularly frustrated because they were co-leaders of the experimental physics group and could not do much without a uranium pile, so they kept trying to make calculations about just what they would do and getting some instrumentation assembled in the hope that a positive decision could be made.

Weiner:

You wife was with you for not quite a year up there, then. After you left, Kowarski came. Do you recall anything about Kowarski?

Marshak:

No, I don’t really. I do not remember Chadwick particularly. I remember Chadwick visiting Los Alamos but not Montreal. Appleton visited, as I indicated earlier. I do not think people kept coming to Montreal. There was not much going on.

Weiner:

This is a good time to talk about Placzek. You mentioned an unrelated part about his account of his experience in the Soviet Union. You know, he went there originally as a refugee?

Marshak:

Yes.

Weiner:

But when he talked with you about this, this was at Cornell?

Marshak:

This was at Cornell. I first met Placzek in 1938 when he arrived at Cornell, and he came there after being expelled from the Soviet Union. Apparently, he and Weisskopf were giving serious consideration to settling down in the Soviet Union, because they did not have positions in the U.S. say in 1935 or ‘36-‘35, say—and they thought, I guess, that they might have to find berths in some other country than the United States. They knew that Great Britain was not a final stopping point. They had tremendous respect for Landau, and Landau was director of the theoretical institute in Kharkov. He had, I think been sent there around 1931. So Placzek and Weisskopf came to Kharkov about 1935, and then the purges started a year or so later. And when Placzek arrived in Cornell in 1938, he was full of stories about how the entire physics institute at Kharkov had been jailed, including Landau, and they had been sent off to Siberia, and so on and so forth. And I must say that those of us at Cornell who heard this story were incredulous. I mean we found it very difficult to believe mass jailing and so on. And from Placzek’s general anti-Soviet attitude we thought that he was not exactly fabricating the stories but just exaggerating. We would ask for the evidence, and he would sort of repeat a smallish number of facts, because he really did not know what happened in Siberia, so he could not build up a strong case - just a few simple damaging statements.

And somehow this did not strike us as being convincing; remember that Placzek’s story unfolded at the time of the Spanish Civil War when most of us were on the side of the Spanish Loyalists, and again we did not have available to us the information about the role of the Communists in Loyalist Spain - some of the damage they actually did at that time. And since the Soviet Union was supposedly supporting the Loyalist side, we could not reconcile this behavior with those statements of Placzek. But I must say in later years I learned that Placzek was absolutely right when I came back to the Soviet Union in 1960 as a member of a team under the McCone-Emelyanov Agreement in Atomic Energy. Five of us visited a half dozen high-energy laboratories in 1960 - Bob Bacher, Bob Wilson, Ed Lofgren, George Kolstad from the AEC and myself - we visited Kharkov, and we were the first Americans there since the purges in 1936. This was 1960, 24 years later, and my third visit to the Soviet Union since 1956. But the Russians had studiously avoided showing us Kharkov. Even when Pief Panofsky had asked at the 1959 Rochester conference in Kiev whether he could visit Kharkov - they refused to allow him to go. It became clear in 1960 why they refused. But the reason we got Kharkov on our 1960 itinerary was that we insisted on Kharkov. Because the Russians wanted very badly to see Stanford on their trip to the U.S., so they finally agreed. And when we got to Kharkov I was able to establish by talking to some of the people there that essentially everything Placzek said in 1938 was correct.

Weiner:

Did you talk freely about it?

Marshak:

Not openly - only privately - freely but privately.

Weiner:

And these were people who had witnessed it or survived it?

Marshak:

Who were in Kharkov at that time and survived it. There was someone in low temperature physics, for example, who died in Siberia - in a labor camp - who had done a very nice low temperature susceptibility measurement of the magnetic moment of the proton whose name was Shubnikoff. And of course we learned that Landau had been jailed. Landau was released after one year because he refused to do physics in jail - thanks to Peter Kapitza who interceded on his behalf.

Weiner:

There was a Martin Ruberman who was in low temperature who was in the Soviet Union as a refugee and who left in the purges, but had a different view from his colleagues who left. He supported the Soviet Union.

Marshak:

From hindsight, Placzek was basically telling the truth about what had happened in the Soviet Union. It was not an embellishment. Stalin’s purges were very very serious, as we have learned since, of course, in later years.

Weiner:

There were still conferences throughout the ‘30s, I guess. I know Bohr was invited.

Marshak:

Well, one reason Landau and his group were so suspect was that Landau was well known and had been in Germany for a brief period. He knew Dirac and Bohr and so on and these people visited Kharkov. And so the authorities could say, “You’re dangerous. You have these international contacts.” So they just jailed the whole laboratory.

Weiner:

Well, to return, now you’re talking about leaving Montreal, and in April of ‘44 you go to Los Alamos. You’ve already mentioned one of the things you did when you got there. You talked with Bethe about the work you’d done which they didn’t know about, which made a difference to their work. Let me ask you: how did you find coming at that time? Los Alamos was reasonably well established and relationships were established and functioning. The housing was set out. What was your impression, coming on the scene anew? You didn’t have too much knowledge of what was going on. I’m curious about the intellectual atmosphere and the social atmosphere involved.

Marshak:

Well, what happened was that of course the housing was pretty much first come, first served. In recent years, Don Hornig and his wife have recalled that they got a good apartment through an error. Brief story - we had arrived in the morning, but somehow our name was not recorded. The Hornigs arrived in the afternoon; they got the next apartment, which was a good one, and the one after that was terrible which we got. We had to move to a place called Morganville, where there were adequate but quite primitive conditions compared to some other apartments. Actually, we didn’t stay in that housing during our entire stay in Los Alamos. I don’t know whether it was six or nine months, but when a much better apartment became available on so-called Bathtub Row, I guess because at that time I was a deputy group leader, we were assigned a much better apartment. I guess newcomers had to make do for a while, and everyone tried to help out, As far as the work was concerned, it wasn’t too bad because I was placed in Weisskopf’s group as deputy head. There was a sort of accelerating pace of new arrivals.

It sounds rather late in the sense that we arrived in the spring of 1944, but the whole pace was very fast. I would say that more than half of the scientists arrived after me. Placzek arrived at Los Alamos perhaps a month before Alamogordo, in June 1945, and Fermi and many of the Chicago people arrived rather late. So it started at a very low base and the numbers increased over a rather short time period because one was moving towards the deadline. For example, when I arrived, not only in neutron diffusion had all the problems not been solved, but there were completely new areas of research that had not even started. Another example, one of the chief areas in which I got involved after becoming familiar with the situation was the effects of radiation on the shock hydrodynamics of the gadget. This means that if you have the implosion due to the TNT, which makes the bomb super-critical, then you have a tremendous chain reaction resulting in a very high temperature at the center and the material moving out. Now, depending on how the shock wave travels out, the material can move very fast and the super criticality can last a short time - and not all of it will be used up in the explosion, so you won’t have as big an explosion.

Or it will not move out so fast, and you can develop a much greater efficiency. So the question was what happens. We had at Los Alamos G.I. Taylor, of Taylor stability fame, who was a consultant there, as was John von Neumann, Niels Bohr, and well, you name it. Everyone knew about the Taylor instability and the Hugoniot-Rankine relations for shock waves. But no one had considered what would be the effect if you had a temperature of 50 million degrees, how that would affect the shock hydrodynamics. And so that seemed like an important problem to look into, and I developed some approximate methods - using similarity transformations - and found that there was a very big influence of the radiation wave on the shock wave. It was during that time that I developed what was called then and for some time afterward the “Marshak wave”, which was this radiation wave. About 10% of that work was published in an article in 1958 (Physics of Fluids, Volume I). Bethe got very interested in that problem. There again, his reservoir of critical talent was very helpful in sharpening up things. An amusing sidelight: there was a pool in Oppenheimer’s office where everybody put in a dollar and put down his predictions for the TNT equivalent he expected in the Alamogordo test. And most of the members of the theoretical division put down higher figures, including myself, because we expected that the radiation would increase the efficiency. I mean we had some rough estimates.

The reasons we weren’t absolutely sure was that in the mad rush to meet the Alamogordo deadline, we tried to get IMBs to do the detailed computations on the radiation effect. I went off - took a trip to New York. But the IBM machines were not quite adequate for the complexity of the problem, there were delays, and one had to depend on some approximate calculations done at Los Alamos more or less by hand. So one didn’t want to stick one’s neck out. Actually, that Alamogordo test gave about 20,000 tons of TNT equivalent. I put down a figure of 14,000 tons because I thought maybe the radiation would give about that amount. The irony is that Rabi won that bet. He didn’t know about the radiation. He was just lucky. He came in and as a general optimist, guessed the 20,000 tons and he won. To finish my answer - this work on radiation and shock waves was a new area for me. I’d never been involved before in shock wave research and I guess I really haven’t been involved in it since.

Weiner:

How did it feel to be applying your basic skills to a question that had never occurred to you before and wasn’t allied to your natural inclinations in terms of taste for physical problems? When you did it, did it give you any feeling of having arrived in a sense, to be able to switch, to state to yourself that I really know something and I can apply it in different fields?

Marshak:

Yes, well, I mean first of all, in terms of the moral dimension, of course, we all felt morally correct at that time because we were afraid the Germans would get the bomb and we simply had to beat them out. We didn’t think in terms of dropping the bomb on the Germans, I don’t think. We thought more in terms that if we had it, we would make a threat, an ultimatum - and then the war would be over.

Weiner:

Did you have any doubts about that?

Marshak:

A little, a little. We certainly gave thought to the extent that we had absolutely no moral compunction. We talked a little about what would happen after the war. But I think hardly anyone questioned the need to go full steam ahead in order to beat out the Germans on the basis that if Hitler won the war, it would be a horrible world. Now, in terms of the actual gratification, of problem-solving, I must say yes, it was sort of interesting to flex your intellectual muscle and try new types of problems. I got involved in quite a few different problems and moved around during the war, partly I guess because of my background in astrophysics.

I had excellent colleagues at Los Alamos: I had Ed Lennox, who is now one of the big shots in molecular biology, at the Salk Institute, as my direct assistant. I was deputy group leader of Weisskopf’s group and consulted on a regular basis with Weisskopf, the leader. Another member of our group was Paul Olum who is now chairman of the department of mathematics at Cornell. One of the people whom Bethe assigned to me was a fellow in Army uniform, whom Bethe felt was rather difficult and asked me to tame, and that was Richard Bellman, who became a well-known mathematician. Another member of my group was another mathematician called Milton Wing, who is in mathematics at the University of New Mexico. The point is: we tried some difficult problems. For example, when Bethe said Bellman was a pretty conceited young man and “you’d better take care of him,” I assigned him some very tough problems which required very sophisticated math. I tried to get Bellman to solve them, which sort of did tame him because the problems were pretty difficult. I mean we looked into quite a few things.

In connection with the test, one was interested in the neutron effects when the bomb exploded: what the neutron distribution would be at varying distances from the explosion? Where you had an earth and an atmosphere, the problem became rather complex. And before the Alamogordo test we worked the solution out using Laplace transform methods; the paper was declassified and published in the Philosophical Magazine right after the war [paper No. 21 - “Laplace Transform Solution of Two-Medium Neutron Aging Problem”, with R. Bellman and M. Wing.] There was a great variety of problems. From a selfish viewpoint, I considered myself fortunate in that I had been on three projects. I got a chance to dig into interesting problems on each project.

Weiner:

Let me ask about the general atmosphere of the Los Alamos laboratory within your group and various sub-groups. You indicated that communication with Bethe was direct. What about large scale communication of physics problems that cut across the various special assignments? Was there much of this kind of a seminar atmosphere?

Marshak:

Yes, very much of it. Oppenheimer believed in it, and it was done on a rather large scale. In particular, there was something called the Coordinating Council, which met every week and consisted of all the leaders and deputy group leaders in the entire laboratory. So you really heard what was going on over the entire place. There was excellent communication within the laboratory. There was no question about that.

Weiner:

How would you characterize Oppenheimer’s role? Was this the first time you’d seen him?

Marshak:

No, I had met Oppenheimer once before the war. Weisskopf had arranged the meeting, I think, somewhere in Greenwich Village, because Oppie was considering the hiring of a post-doc. I guess it was after I had arrived in Rochester as an instructor; it was not clear whether I would continue after the one year. Weisskopf thought Schwinger might come the next year and perhaps Oppenheimer would want to take me for that year. So I met him. Schwinger was uninterested and so I stayed in Rochester as instructor. I just figured out why I never became a post-doc? Bob Sachs got the position with Oppie.

Weiner:

What was your impression of Oppenheimer? Had you had any prior impressions of him? That is, how did you regard him in terms of personality and of power?

Marshak:

Oh, my own impression was of a very brilliant man and exciting personality. He had a fondness for Weisskopf, and so there was frequent communication with Weisskopf. Weisskopf would tell me what Oppenheimer was doing and what was happening. So I sort of had knowledge of his existence on a rather continuous basis. At Los Alamos there was no question that he was master of the whole situation and was highly respected. Certainly, in all of his performances, in the Coordinating Council and on other occasions, it was clear that he knew what was going on across the board. As a deputy group leader, I think a couple of times I was over in a larger party at his home. But he had more frequent meetings with the division heads and socialized with them: the Bethes, Bob Wilsons, Bob Bachers and others. Bacher and Bethe were two members of my Ph.D. committee at Cornell. And he was fond of old friends from Berkeley - like the Serbers. Also, it depended a little on to the extent to which Kitty Oppenheimer befriended various wives. My wife never became very friendly with Kitty Oppenheimer, and so socially we didn’t see very much of them, just a little. In the lab it was through the Coordinating Council and so on that I got a good glimpse of Oppie. But my wife and I became very friendly with the Bob Wilsons and the Bob Bachers. We became very friendly with the Cyril Smiths and the Peierls. So I would say we got to know many of the senior people in the laboratory during that period.

Weiner:

And your social life would involve that circle?

Marshak:

Yes. I don’t know - some people might have considered us social climbers in the sense that the average age of the persons whom we would see more would be older than us by a few years. But I had at least position as a deputy group leader, so to that extent I got to know some people through the Coordinating Council. It was also members of the Coordinating Council who were invited to go down to see the Alamogordo test. I was one of a group of about a hundred!

Weiner:

You were there?

Marshak:

Oh, yes, I was there. So in a sense I was a member of the inner group of about a hundred. I was probably pretty far down on the totem pole in terms of seniority. Many of the friendships that we developed at Los Alamos became rather permanent friendships: for example, the Wilsons and the Marshaks have continued to be close friends. We continued our friendship with the Serbers for many years. But, of course, people whom you don’t see very frequently in later years, like the Smiths, who went to Chicago and then Boston - you lose contact with.

Weiner:

You were there a total of less than two years.

Marshak:

Less than two years. We left in January of 1946. But it was a very rapid-moving two years - I mean a very full two years, extremely so. Because we worked probably on the order of six days a week. We’d just take Sundays off.

Weiner:

At that time what about children?

Marshak:

We had no children. My wife was teaching.

Weiner:

Where?

Marshak:

In Los Alamos.

Weiner:

Now, there are two other things that I want to touch on. There may be a lot more. One is your account of the Alamogordo test in July 1945. And then the other is: when it came time to make the decision as far as the actual use and deployment of the bomb, various kinds of patterns of thinking were set up and what your perception of that was and what your recollections are? So I don’t know if these are tied together or not. One is the Alamogordo test, which I think is a story in itself, although it’s been described before, and the other is your point of view and what you did and what you saw.

Marshak:

Well, in many ways, since the test was not supposed to hurt anyone at that point, we approached it as a very exciting experiment - to see whether the several years of intense activity on the part of so many people was going to pan out. In other words, we were not somber. I think we were thoughtful. We wondered what the future had in store, particularly on the way back after we realized that it was a pretty big shot. I mean we didn’t know, of course, that it was 20,000 tons, but everybody sort of agreed, just from the external appearance, that the test was a huge success. So on the way back we talked more about what the future would be. Going down there, I guess we talked less about it. Well, as you say, you have had many descriptions.

I don’t know whether it’s worth repeating...I do remember Bill Laurence of the N.Y. Times was nearby. I don’t know whether you’ve heard the story. We had the flash and then there was a noisy reverberation after the sound had hit the neighboring mountains and came back — and Laurence wondered what caused the noise. We all joked about that. We were all told before the test to watch out for rattlesnakes and were given special instructions about our eyes and all that. But I think those of us who were observers at a certain number of miles probably didn’t go through the excitement of the small group that was involved in the assembly and in the tower. For them it must have been an even more momentous affair, although we realized that this was going to be an epoch-making event. It was on the way back from the test that we worried about what was going to happen and whether there was any way of preventing the use of the bomb. Of course, the Alamogordo test was after V-E Day, and we realized that the only victim could be Japan. So immediately after the test took place, everybody became concerned...Up to the time of the test, everyone, I think, continued to work very hard to try to settle whether implosion bombs were serious threats. But VJ day was the turning point. Everyone became very political and started worrying about the consequences, and that’s when the Los Alamos Association of Scientists was organized, and some of us really tried to be helpful. For example, Leonard Schiff, Eldred Nelson and I wrote a little book - about a hundred pages long - called Our Atomic World to help educate the public...It was published by the University of New Mexico Press, which was nearby and which was a bit of a mistake in terms of getting national coverage for the book. Actually I think it wasn’t a bad little book. It essentially put together many of the arguments which the scientists generated at that time about the nature of the atomic bomb - Is there a defense, can there be a monopoly - leading up to the overall conclusion that we must have international control of atomic energy. You might say that the day after VJ day the vast majority of Los Alamos scientists joined the scientists’ movement.

Weiner:

You wrote the book at Los Alamos?

Marshak:

Oh, yes.

Weiner:

I thought this was written when you came back. I was thinking of the petitions - at Chicago and the Franck petition to the President about not dropping the bomb.

Marshak:

No, it was written at Los Alamos within weeks after VJ day. I did not see those petitions. I don’t think those petitions reached Los Alamos. I think perhaps General Groves, with Oppenheimer’s concurrence, did not allow these petitions to be circulated at Los Alamos.^[1] So we did not know about them. But we started the Los Alamos Association independently right after VJ day. Because, you see, until the war ended, we were not allowed to leave the Santa Fe area except for very specific reasons like visiting family; you had to stay within the Santa Fe area. I think General Groves’ action was part of that overall freeze on communication with the outside world.

Weiner:

So the political activity that did jell was not regarding the dropping of the bomb. It was in regards to post-war controls.

Marshak:

I think in all honesty that is true...I do not recall a major public activity at Los Alamos taking place to stop the dropping of the bomb. That is, somehow the control of General Groves, and the military, was so complete and time was so short that we hardly began our discussions before the war was over. I guess people outside Los Alamos thought they could exert political pressure on the military, but we were essentially the instruments of the military; and I guess pretty much immobilized in terms of moving out from that framework of thinking. I think we all expressed the hope and the fervent desire that the Japanese war would end without the use of the bomb. But, after all, people like Morrison went and helped to assemble the bomb on Tinian - under orders from the military. I am not being critical of Phil Morrison or anyone else.

The fact is we did not have enough information - we heard arguments that there would be a million casualties, and so on...And this was the way to end the war. And, partly to some extent, I guess most of us thought that Oppenheimer was the leader, and he would do right. But it is true that I don’t recall a vigorous effort - say a house-to-house canvassing - at Los Alamos to try to prevent the dropping of the bomb, even though I’m sure, if a poll were taken, the overwhelming majority would have been opposed to the dropping. I wouldn’t be surprised, as I think more about it, that this business of complete exhaustion after the test played a role, that we sort of went off into the mountains for the crucial couple of weeks - July 16th till the beginning of August - while the decisions were being made by others.

Weiner:

We can check with Alice Smith’s book of this period, A Peril and A Hope.

Marshak:

What does she say?

Weiner:

Well, that’s what I don’t recall. (See Smith, Peril and a Hope, pp 55-78).

Marshak:

Incidentally, right after V-J Day, I received a letter from Little, Brown and Co. inquiring if I wanted to write a general book about Los Alamos. I said I didn’t care to do so, but I turned the letter over to my wife, who assembled a team of women - Alice Smith, Jean Bacher, Jane Wilson - and others who wrote chapters for a book. Their book is in existence - never published, because Little, Brown decided that there were not enough “climaxes” in the book.

Weiner:

Is it a manuscript?

Marshak:

I think at one time we had a copy of it. Charlotte Serber had a copy of it. I’m not sure where it is now. Maybe we have it.

Weiner:

We should make a survey, and see if Serber has a copy at home, or who does.

Marshak:

This might have some interest for the archives.

Weiner:

I would think so.

Marshak:

Parts of it were used for later articles by other people who heard about it. But the women wrote this - their side of the story.

Weiner:

Now let me get on the story of your book, with Schiff and who else?

Marshak:

Eldred Nelson.

Weiner:

Eldred Nelson. How did it come about that the three of you decided to come together and write a book on that particular subject?

Marshak:

Well, certainly after V-J Day - I’m trying to be as accurate as I can - , we were rested up and the war was over. We got very concerned about what had happened, terribly concerned, and had nightmares, I guess, and realized that we were all implicated in an awful moral dilemma. And spontaneously the whole laboratory started having large meetings and trying to generate ideas for action that would prevent a recurrence of such a catastrophe. You know, we were in a sense shell-shocked and realized what had happened and how much responsibility we had for it. So the association was formed, and all kinds of committees were set up to perform various chores. I think I was asked to chair a small fact-finding committee or something of that sort; and I think Leonard Schiff and Eldred Nelson were interested and we decided to produce this brochure. [Our Atomic World, U. of New Mexico Press, 1946. Publication was delayed until June 1946, according to Smith, Peril and a Hope, p. 302.]

Weiner:

Our library should have a copy of it.

Marshak:

I may have it somewhere, I think - the one remaining copy. And we can at least have it Xeroxed.

Weiner:

It’s small enough.

Marshak:

I lost the cover to it, which had a Kodachrome shot of the Alamogordo explosion with the mushroom cloud. It’s in libraries.

Weiner:

We may have it for all I know. (Available in AIP Niels Bohr Library).

Marshak:

I know it’s in the library here. [pause in recording]

Weiner:

The book was a product of this group. Now, I’ll have to read this book, and I might then have some questions on it. But the purpose of it was to reach a popular audience…

Marshak:

Right.

Weiner:

...to tell what you were able to about the nature of this thing and to inform public opinion on it.

Marshak:

Right.

Weiner:

Did you have in mind any specific political action in terms of legislation or anything?

Marshak:

Oh, yes. The Los Alamos Association became very active lobbying in ‘46 for the May-Johnson Act.

Weiner:

But when it was first formed...did you go into it with that in mind?

Marshak:

Well, it was not in mind that specifically, but I think it was in mind in terms of selling the world on the calamities connected with nuclear warfare and to take steps to avoid it...for example, we argued for international control of atomic energy. Of course, there wasn’t a specific Baruch Plan - it was too early - but the idea of the book was to make the main points that the bomb is a weapon of mass destruction, there is no defense, there is no monopoly here - ergo, you must have international control. In that book we actually stated that any large country that wanted to make a major effort and had the scientific and industrial power could construct a bomb in about five years. And that was in the Fall of 1945 that we wrote it. Shortly thereafter, General Groves said in Congressional testimony that he didn’t think the Russians could have the bomb in less than 25 years. As you know, the Russians had the bomb in four years from that time, in 1949; and I always half-jokingly say Klaus Fuchs helped them to reduce the period by one year. This is perhaps not incorrect in terms of what Fuchs knew, because he was certainly fairly close to some of the key ideas as a member of Peierls’ group at Los Alamos and knew all about the implosion method. So I mean the writing of the little book was essentially our committee contribution. We were willing to take on this extra job of writing a hundred pages in reasonably popular language for political reasons - namely, to further the cause of international control of atomic energy.

Weiner:

Did you split it up into sections?

Marshak:

Yes, in a way...There were nine short chapters, and we each wrote about three chapters. The book went into not only the negative side of the dangers, but also the positive side of the opportunities of atomic energy. We talked about atomic piles and radiations for medical purposes, - industrial uses, things of that sort - as well as the damage at Hiroshima and Nagasaki.

Weiner:

Now, was the Association of Los Alamos Scientists a pretty popular, sizeable group?

Marshak:

Yes.

Weiner:

Was it quite open? Was there neutrality or hostility on the part of the leadership?

Marshak:

Well, I think, after V-J Day, Oppenheimer felt that he was in a very strong position in Washington, that perhaps he could accomplish as much as the entire association. But he did not oppose the formation of the Los Alamos association, as far as I recall. Edward Teller was not keen on the association. I remember some votes - 499 to 1, with Teller casting the one negative vote (Oppie did not attend - he was probably in Washington).

Weiner:

What would the votes have been on - within the association or among...?

Marshak:

Well, for example, that vote may have been on whether to support the May-Johnson Act. .I think towards the end of the war there were discussions of the mechanism for post-war control of atomic energy. I think the May-Johnson bill must have been introduced in the fall of ‘45, wasn’t it?

Weiner:

I think so. [Bill was introduced 4 Oct. 1945.]

Marshak:

Right, because it took some time for the scientists to go into action, but many of them went into action, certainly by early ‘46. So there were some discussions as to stands to take, and maybe Edward wasn’t quite sure about his position. You know, Teller was always a rather individualistic fellow. I remember some opposing arguments that he offered. But no one read much into them at that point in terms of his later role. I mean we all knew Edward to be a very bright and imaginative scientist, and if he wanted to press his different views - that was fine. This was before he got involved with the Oppenheimer hearings. He was always very independent.

Weiner:

By the way, the records of the Los Alamos Ass. are preserved in the University of Chicago, along with the other records of the groups of the period. Let me ask another question about the discussions that I’m sure took place in Los Alamos, about getting back to peace time physics, other than political discussions about the scientists’ responsibility. From what I’ve heard, there was a good deal of speculation and job-swapping and promotions without people having left Los Alamos, and this kind of thing. I don’t know if we’ll have time, but I want to talk about that general transition period when it was quite clear that you would be going back and that you would be picking up physics again. I’d like to know what the expectations were, what the problems were, and what did you have in mind to prepare for it.

Marshak:

Well, at that point I received a letter from DuBridge urging me to come back to Rochester. I’m not sure now of the sequence - but I believe...I came back as an assistant professor, certainly in early ‘46, - I mean February of ‘46. And I guess it must have been an offer I had from Harnwell at the University of Pennsylvania in the spring of ‘46 that led to my promotion to associate professor. I do not think that it was because Weisskopf was leaving that DuBridge said, “Now we better make him associate professor in order to keep him.” I did not do much job hunting in the sense that DuBridge wrote to me at a very early stage and said that he was coming back - as he had to Weisskopf - that he was coming back as chairman of the department, that he was not going to pay any attention to any administrative offers even though he had been the director of a laboratory that had grown to more than 2000 people and he personally had become a national figure. He wrote to us in the Fall of 1945 - it must have been - and said that he was planning to build up a glorious department, had many ideas and strong support from the Rochester administration, and would we come back? Well, my wife was from Rochester. She had a sick mother with whom she wanted to keep in close touch. Her mother had had a stroke. So I did no shopping around at all. That sounded fine. With DuBridge as a nationally known chairman, it looked like an exciting post-war period ahead.

On the other hand, Weisskopf… I think that basically the reason that Weisskopf began shopping around was he had come from Vienna; he had a more cosmopolitan attitude toward the nonscientific part of one’s life. I mean he liked music and high culture. And I think he did regard Rochester as a somewhat provincial town, which it is. And he had not been treated very well by the U of R. As I’ve indicated earlier, he still was an assistant professor when the war ended. Hence, so he was rather receptive to offers, and how much shopping around he did, I don’t know. Even though we were in adjacent offices, he didn’t share every confidence with me. So one fine day he informed me he had an offer from MIT as an associate professor. And, of course, I immediately wrote to DuBridge and possibly even called him and urged him to keep Weisskopf, that Weisskopf had grown in stature (of course, he was well known before) and that he should try to keep him. Of course, it didn’t require my urging, but in any case he offered Weisskopf an associate professorship. But then Weisskopf presented that to MIT, as I said earlier, and they quickly came back with a full professorship; so then DuBridge matched that. But, by that time, Weisskopf decided to accept the MIT job. Now, whether, if DuBridge had immediately come back with a full professorship offer after the first MIT offer Weisskopf would have stayed at Rochester — I don’t know. It’s quite possible he would not have. But he certainly became impatient with protracted negotiations and I’m sure he felt he had a clear conscience in leaving Rochester.

So he left for MIT. And, so you see, Weisskopf’s departure in a sense thrust me into a leadership role in the department. Because I then became the only theorist, and the other people who were coming back to Rochester - still a small group of people, Barnes and Van Voorhis were experimentalists. And so DuBridge, I guess, was forced to fall back on me as chief advisor. And I remember how we started working to try to get a replacement for Weisskopf, and some awfully good people were offered positions by DuBridge at that time. For example, he tried to get Casimir; he tried to get Uhlenbeck; he tried to get any number of people - and they all turned him down. Again, I would guess that, apart from the special attraction of the Caltech offer, he would have accepted any decent university presidency at that point. I always had a sense that DuBridge was terribly discouraged; that, despite his national position, he could not, after three or four attempts, get a replacement for Weisskopf. And so when the Cal Tech offer came along, he decided to leave. He left by the summer of 1946. In other words, we overlapped just one semester after the war. By the Fall, he was gone. And he brought in as chairman someone from the outside, George Collins, but that starts a new era. Collins promoted me to associate professor on the basis, I guess, of the University of Penn offer, which was an associate professorship at a decent salary. So I now was the only theorist at the associate professor level, and we’ll go on from there.

Weiner:

When we pick up tomorrow there are two things I want to talk about at Los Alamos, about the refresher course in physics that some people gave...

Marshak:

Yes, I gave some of those lectures.

Weiner:

And then the other thing was: what you felt that physics would be like in addition to the personal job-hunting and the confirmation of existing jobs, what this new era really meant in terms of expectations in physics, how it would be supported, and also what problems...

Weiner:

This is Side 2 of Tape 2, and today is June 16th 1970. We’re starting our recording session in the morning at 9:30. The thing that we left off on yesterday was my question about the discussions at Los Alamos when it was apparent that people would be returning to their own universities or to other universities - to talk somewhat about the expectations that people had in terms of what physics would be like, what their lives would be like, what kinds of intellectual problems seemed to interest them and what was done in an organized or an informal way at Los Alamos to help people prepare for this transition.

Marshak:

Well, those are many questions. Certainly, right after V-J Day, when the war was really over, one started thinking about returning to one’s former place of employment or a new one. Since most of the people at Los Alamos were from universities and quite good universities, in general, most people, I think - I didn’t take a count, but my impression is so - a substantial number returned to their own universities. There were some changes. We mentioned the one of Weisskopf, changing over from Rochester to MIT. But, for example, Bethe returned to Cornell; Bacher returned to Cornell. I’m trying to think of the division leaders. Oppenheimer returned, I believe, to Cal Tech, for a while, and it was only the following year that he was offered the directorship of the Institute for Advanced Studies. Bob Wilson went to Harvard, and he was not there before, so he changed; but I think in one sense he was maturing very rapidly and became one of the very successful young physicists, having become a division head in the interim, so that he was entitled to a much higher position than the one he held when he came to Los Alamos a couple of years earlier. Actually he went to Harvard just for a while before he was persuaded to join Cornell as a replacement for Bacher, who by that time had gone to Caltech as provost under DuBridge, the new president of Caltech.

I think it was clear that the return to the same institution of some of these now fairly distinguished people or one could say notorious people - because everyone associated with Los Alamos carried some of the aura of having participated in the dramatic ending of the war - I think the consequence was that pretty fancy prices were extracted from the university administrations for the support of their physics activities. I believe that Bethe and Bacher did extract a very fancy commitment from the then President Day of Cornell, which, in later years, caused a considerable amount of trouble apparently in terms of the university deficit and so on. And certainly DuBridge came back (of course he was not at Los Alamos) from the wars to Rochester with an extremely high salary for those days and again a commitment for support of a major increase in the size of the physics department.

Weiner:

These commitments for increases were from university funds?

Marshak:

In the case of Cornell, they were university funds, because at that time it certainly was not clear yet that there were going to be major sources of funds from the federal government, although I guess people were beginning to talk about it in order to maintain the scientific momentum of the war years. And Vannevar Bush, I guess, was beginning to write his Endless Frontier. It came a little later, but I think there were certainly many conversations...

Weiner:

There was the Steelman Report, I think.

Marshak:

There was the Steelman Report - that’s right. There were certainly many discussions. For example, the Office of Naval Research, I think, very quickly made a decision, perhaps by the end of ‘45, and certainly by the beginning of ‘46, to try to put a considerable amount of money into the support of basic research, because there was no other agency. The AEC was not in existence, the NSF was certainly not - that was years later. I think there was a Captain Conrad - he later died of leukemia - who was a very far-sighted individual in the Navy, who, I think, sparked the decision by the ONR to start supporting basic research. The ONR, for example, was the agency that gave money for the Rochester synchrocyclotron, which was the first post-war large cyclotron started and completed.

Weiner:

Really? I didn’t know that. When the time comes I’d like to get into that.

Marshak:

So I think many people returned to their own institutions with more or less rising expectations and commitments of substantial support from their universities - if not in terms of specific funding, then in terms of hoping somehow to get funding through a variety of sources, once the agencies were set up. But I think probably some of these institutions, through their senior scientists who were prime movers in the Washington planning to create the agencies, perhaps hoped to get some outside funds as part of the total contribution to their research. Now, at Los Alamos itself, one began to worry about the younger people and how much they had lost in terms of normal training, and a temporary university was set up at Los Alamos, a very informal one, no credits, But many people volunteered to give lectures. I know that I and Hans Bethe (I say “I” first, because I think I gave most of the lectures, but he gave some) gave a course in classical electrodynamics. Those notes were written up, and at one point we thought of publishing them but never got around to it.

Weiner:

Have you kept them?

Marshak:

Yes, I have a copy somewhere. There were many other courses, and people like von Neumann, who was thinking ahead to the role of computers in science and society, were giving a series of lectures on his thinking for the future. I remember hearing some very interesting and novel thinking from von Neumann in some of those lectures, and many people attended the lectures. In other words, there was very great informality. I mean both senior and junior people would attend lectures when they were interested. Or there was a very informal student-teacher relationship, where senior people would give lectures for more junior people on fairly standard subjects. This activity went on pretty intensely for several months.

Weiner:

As a kind of open activity which was advertised throughout the laboratory?

Marshak:

I believe so, yes.

Weiner:

And people had the leisure and the freedom to stop what they were doing at three in the afternoon to attend the lectures?

Marshak:

Well, I think it was all after V-J Day, so that people didn’t feel the compulsion that they simply had to spend all their working hours on the war work. The understanding was that everyone would try to finish up their reports. But the outward movement was accelerated by some water shortages at Los Alamos in the Fall of 1945 because the laboratory had reached a very large size - 5000 inhabitants - and the water supply had never been planned for that number. One soon realized that one was being helpful, being a Boy Scout, to leave the laboratory as soon as possible to reduce the demands on the water supply. As I recall, this was an element in our getting out at the beginning of January. And since I didn’t have to report in Rochester until February 1, my wife and I detoured through Mexico a couple of weeks with Martin Deutsch and his wife. In any case, that’s an interesting story in itself - the excitement in Mexico -, to hear some lectures about the A Bomb, which Deutsch and I gave. But in any case, many of us accepted the commitment, that even if we left the laboratory, we would try to finish up our outstanding reports - for the people who remained and the new people who would come. I took that commitment very seriously and, as I indicated earlier, did spend pretty much the entire Spring semester of ‘46 writing up those reports. I remember well because they were classified reports and I had to work in the attic of our house, working away there, locking the door when I wasn’t working on the reports and I finished in June of ‘46.

Weiner:

Before we get away from Los Alamos, what kinds of physics problems were on people’s minds?

Marshak:

Well, now, that’s a very interesting question. You see, we were so busy...Those were rather brief years - for me almost two, for others three. The maximum time I guess, was about three and a quarter years. And these were very intense years. One really concentrated completely on the war work and trying to get that bomb built before the Germans. So one was not keeping in very close touch with other scientific work, and I guess there wasn’t too much being published, although there were papers in some of the European journals. But I don’t remember myself trying to keep up with the no classified scientific work. I was just very preoccupied trying to do the job for the lab. And I think most other people were in the same position. But, as we started thinking about going back, then of course many of us were sorting out in our own minds what type of science we would be doing. There was one important development for the future of high energy physics. One person who was giving a lot of thought to the post-war period was Ed McMillan, who was at Los Alamos and who invented the principle of phase stability, independent of V. Veksler. Apparently, Veksler had done it somewhat earlier. In later years, I was a trustee of the Atoms for Peace Awards, and we gave one of the awards jointly to McMillan and Veksler for this discovery. And, incidentally, I might say that Arthur Compton before his death - he was a trustee for Atoms for Peace - in a letter was the first to suggest that this be done. And we considered Compton’s nomination after his death, and I personally pushed it very hard, but it originated with Arthur Compton - the idea of honoring McMillan and Veksler. The reason that this was so important and in a sense opened up the whole era of laboratory high-energy physics was that there were limitations on the normal approach to the construction of accelerators. Some of these limits had been spelled out by Bethe and M.E. Rose in papers just before the war, and EM. Lawrence had started a large 184-inch cyclotron before the war to try to beat their limit.

He was trying, by brute force so to speak, to push up to as high an energy as possible. I forgot now the exact energy. It might have been 100 MeV or 150 MeV or something like that - nothing like the 340 MeV, that was finally achieved with that machine. And what happened was that Ed McMillan - I think he left very quickly after the war ended to start working, feeding his ideas to the Radiation Laboratory, to the Berkeley group; and modifications were made. As you know, the cyclotron magnet was used for the Manhattan Project during the war. After the war, Ed McMillan got together with E.0. Lawrence and they made changes and it became a synchrocyclotron. The word “synchro” reflects the new idea of phase stability. But, of course, this was just a gleam in Ed McMillan’s eye. In one sense I don’t think there were tremendous expectations for a very exciting life in laboratory high-energy physics, because one wasn’t sure whether Ed McMillan’s idea would work. It looked good.

Weiner:

It was discussed.

Marshak:

He discussed it with an inner group of people. He might have given a lecture on it. He certainly discussed it with some of his immediate colleagues, certainly the experimental people. And it looked good. But I don’t think it was so certain that one immediately started planning for huge machines. But I think it went rather fast. That is, as soon as he got back there, in a matter of months maybe, the rumor spread that it seemed as if it was going to work. It looked very good and so on. Because I know that in the spring of ‘46, with DuBridge here as chairman at Rochester, we started talking about a 150 MeV synchrocyclotron, as did Bob Wilson, who by that time had accepted a position at Harvard. The decision was quickly taken that Rochester and Harvard would build an identical synchrocyclotron for 150 MeV. And Wilson and Barnes both went to Berkeley to learn the tricks. We later changed the Rochester machine to 240 MeV for reasons that are connected to the two-meson theory, but I’ll come to that in a little while. So the answer was that some of us who were oriented towards theoretical physics were open-minded about the future. Nuclear physics was, of course, the general framework of our thinking. I guess I planned to go back to nuclear physics. Bethe planned to go back to nuclear physics. I had been involved in meson theory and beta decay theory - so I was interested in that sort of thing. But I would say that the efflorescence of high-energy physics - the golden age that followed - was something that was not very clear to us when we left Los Alamos.

Weiner:

What did you see as the biggest problem in physics in terms of a current problem, whether or not it was a new problem? You communicated that it couldn’t very well have been a new one because you certainly weren’t keeping up with the literature. Or whether it was an unfinished problem that you had to put aside. I know that when one picks up a publication from that period...

Marshak:

I’m trying to recall. Of course, it is now 25 years ago. I know that one of the things that interested me at that time was the question of what the nucleon-nucleon scattering cross-sections should be at higher energies, and high energies in this case was about 100 MeV. And I started worrying about things like: would there be a velocity dependence of nuclear forces, and that’s the title of a little note that I wrote with Hartland Snyder. I came on it first myself and then got together with Hartland. The reason for that was that I started spending summers at Brookhaven, beginning with the summer of ‘47. The summer of ‘46 I spent at the General Electric Research Lab, and that’s another story. But I started worrying about the problems of nuclear forces at high energies. That would be the two-nucleon interaction at high energies. I remember suggesting to Julius Ashkin, whom I had persuaded to come to Rochester - he was one of the bright young theorists in the Feynman group at Los Alamos - that he might look into the polarization effects of the tensor force at high energies. He has some papers under his own name in that area and I was looking at the velocity dependence of the two-nucleon interaction. So that was one area of interest to me. The other things that came along very shortly, you see, was the Shelter Island Conference of June 1947.

Weiner:

I’m not going to let you get into that.

Marshak:

We won’t get into that right away, except there was a precursor of that and a very exciting and, I would say, epoch-making paper that came out, at the beginning of ‘47. It was on work done during the war but published at that time. And that was the paper by Conversi, Pancini and Piccioni, wherein they reported that the negative cosmic ray mesons were not all absorbed in carbon, which they should have been, but there was substantial decay. A large percentage - like a half - decayed. The people who first realized how startling this development was, were Fermi, Teller and Weisskopf. They wrote a paper just before the Shelter Island Conference that pointed out that the Italian experiment implied a factor of something like 10 discrepancy between the strength of the meson-nucleon interaction that one expected and the strength that was implied by the observations of the Italian group.

Weiner:

Which one expected for the Yukawa meson?

Marshak:

That’s right. That is, if the Yukawa meson was supposed to have strong coupling with the nucleon in order to be responsible for the nuclear force, then once the negative Yukawa mesons are in K orbits of carbon, they should very quickly be captured and not decay in carbon. So you should not see any decays in carbon, contrary to experiment. Even though the Italian experiment was rough, they had magnets and so were able to discriminate between positive and negative mesons. That was a very crucial experiment. I note that this important work was done with cosmic rays since no high energy machines were available as yet. That is, the Berkeley machine didn’t start up until ‘48, and the Rochester synchrocyclotron didn’t start up until ’49. And the electron machines not until ‘49 or ‘50.

Weiner:

But meanwhile the smaller cyclotrons were in operation. They had been put back in operation in some cases and the ones that were almost completed before the war were completed. But all of these were at low energy ranges...

Marshak:

Oh, yes. All these were at energies too low to produce even mu mesons, which had a mass of about 100 MeV. So that as long as you were under 100 MeV, you just could not produce them. If you have proton-proton collisions, roughly speaking, you need twice the energy. But if you have a proton-nucleus collision and take account of the Fermi momentum inside the nucleus - Fermi energy is 25 MeV - the threshold energy for meson production is lowered to a little over the meson mass energy itself rather than twice the value. So, as long as you, say, were below 125 MeV, you had no chance of producing the cosmic ray mesons in any way. And so cosmic ray physicists had a temporary monopoly.

Weiner:

Before we get too far ahead with this, when did you first become aware of the Italian cosmic ray work and the Fermi-Teller-Weisskopf calculations? [pause in recording] We decided that we’re going to stop at this point for a minute and then go right back to the early period just after leaving Los Alamos. You mentioned casually that you went with the Deutsches to Mexico - there were some speeches you were asked to give on atomic energy?

Marshak:

Yes. It seemed as if the American Embassy in Mexico thought there would be some good will engendered if a couple of scientists from Los Alamos, with all the public notice and the big headlines everywhere, if they gave some talks to Mexican audiences. And so I gave several, and Martin Deutsch gave some and in one case I think we joined together to talk to the student body of the University of Mexico - several thousand. There was, of course, tremendous interest and headlines in the Mexican newspapers that Los Alamos scientists had come to Mexico to reveal the secrets of the A Bomb! One amusing story - at the talk given at the University of Mexico, one of the students queried how heavy the bomb was, and of course we had to stick very closely to the Smyth Report; so I answered between 2 and 200 pounds. And that evening at a big dinner party in our honor, the waiters passed out so-called atomic cocktails, which had several colors. I asked how many liqueurs were in the cocktail and the waiter said, “Between three and ten”! So there was great camaradie in those days, and everybody was very optimistic about the future.

Weiner:

Had you gone to Mexico on vacation, and because you were there, the Embassy involved you in this?

Marshak:

Yes. My wife persuaded me that we would never be so close to Mexico again. We had no anticipation in those days that physicists would travel quite so much. So, both from the point of view of economics and proximity, we drove down to Mexico City; and the Embassy learned about our coming and thought while we were there, we could cooperate and give some lectures. I gave lectures at the University Club in Mexico City, to the American staff at the Embassy, the University of Mexico, the students and so on - about four or five talks. So that’s how it happened. Let’s see: so we came back to Rochester from Mexico to meet the February 1, 1946 deadline.

Weiner:

You mentioned that when you returned to your position, times had changed. You mentioned that the spring semester of ‘46 was the time that DuBridge was trying to keep Weisskopf but was unsuccessful. You also mentioned that this was the time when the plans for the synchrocyclotrons came about. How did DuBridge or you know about this? By that time, had McMillan published?

Marshak:

No, but I think Sid Barnes played an important role there, because he had come to Los Alamos, perhaps in the spring of ‘45 from Oak Ridge. As I said, many people were coming in from all the projects. Near the end of the war, they were finishing their work on the projects and many joined in to give a final push to the Los Alamos project. And Barnes, I think, was one of those who listened very carefully - since he was one of those who had built a cyclotron, liked to build machines - had listened very carefully to Ed McMillan, because he had once spent some time at Berkeley - he was quite keenly aware of Ed McMillan’s phase stability considerations. So when he came back to Rochester, he went after DuBridge to see if Rochester could not obtain a synchrocyclotron - a modest one, not as large as the Berkeley one, but a serious one. And that’s how 150 MeV came to be chosen. Barnes also knew Bob Wilson quite well, so the two of them must have been talking together. I was not involved in those early discussions, because I was busy with other things - finishing reports for Los Alamos, for example. I was involved later in changing the energy. I think Barnes persuaded DuBridge that the wave of the future for physics was to get into these bigger machines, that Rochester had a tradition of building machines at an earlier stage - since the small cyclotron at Rochester was the third in the country in the 1930s - Berkeley was number one.

Cornell built the second machine in the 30s, a very small one. It could almost not be counted. So Rochester was really number two in terms of the 7 MeV proton cyclotron. Barnes argued that we should again be number two after Berkeley with a synchrocyclotron. And DuBridge listened and had contacts with the ONR and got them interested in supporting such a machine for Rochester. At the same time, Wilson presumably got the ONR interested to support a similar machine for Harvard. And then Barnes and Wilson proposed to achieve economics by working together and essentially designing a single machine in two copies. So that’s how it all started. But it was later that year, after the Conversi-Pancini-Piccioni paper, which destroyed the idea that the cosmic ray mu meson (the one that was known; of course it wasn’t called a muon in those days) of 100 MeV mass could be produced as a single particle, that it became clear that 150 MeV would be inadequate to do meson physics.

Weiner:

When the synchrocyclotron was planned, I can understand Barnes’ approach - that is, he had built it and at least had that in his background and would be interested in the physics of the machine and the new principle itself. That doesn’t say he had in mind what the machine would be used for. Where did that come in? I mean those who were working in meson physics in the early stages, I can understand them thinking this way. But how was this...?

Marshak:

How was the number chosen?

Weiner:

The energy?

Marshak:

That’s a very good question. Well, that choice was made just on the basis indicated earlier, that if the cosmic ray meson of 100 MeV was the Yukawa meson and was strongly coupled to nucleons, then you could produce it singly; and you needed a threshold energy, say, of 125 MeV with nuclei. And if you had 150 MeV, you had a little to spare. At the same time you had to think about the cost. That is, a 150 MeV synchrocyclotron at that time cost about $1.5 million and that was quite a bit of money. So one had to have the lowest energy possible in terms of the cost but a high enough energy in order to have some guarantee of being able to do meson physics.

Weiner:

But the idea of meson physics was coupled in from the very beginning?

Marshak:

Oh, yes.

Weiner:

The idea in building the machine?

Marshak:

Oh, yes, the idea was to produce the Yukawa meson, yes. Of course, you would do high-energy nucleon-nucleon scattering, but you also hoped to produce the meson responsible for nuclear forces which was thought to be the cosmic ray meson.

Weiner:

That’s interesting.

Marshak:

That’s where we provided some of the theoretical input.

Weiner:

That’s what I’m getting at, because I didn’t think ordinarily the people who were doing cyclotron work were thinking about meson physics. It doesn’t show up in anything that they had done prior to that time. They were conscious then of these kinds of questions?

Marshak:

Well, I talked with Barnes. We talked, and I indicated that on the basis of the usual assumption, that 150 MeV might do it but that it was doubtful after the Italian experiment. And he indicated from his side that he would try to push it above 150 MeV, but he might not get the machine because it would be too costly. So a compromise had to be reached. Now, while we are at it, I will just say that the reason we went to 240 MeV was in a sense based on an incorrect argument. Namely, by the beginning of ‘47, when the Italian paper appeared, which indicated that something was rotten in Denmark about that cosmic ray particle - that it was not strongly coupled, I could see immediately that this, in terms of the normal arguments meant that you could not produce them singly -. In my own mind, I then said: “Well, maybe they’re produced in pairs.” To think of another reason at that time just didn’t occur to anyone, or didn’t occur to anyone in the United States. [The two-meson idea had occurred to the Japanese physicists - for other reasons - unbeknown to the rest of the world until Nov. 1947 (see below)]. And therefore...I got very excited...You see, Gregor Wentzel during the war had continued the work on the meson pair theory, the spin 1/2 meson coupled in pairs to the nucleon, and had shown some pretty features in the strong coupling version that theory (I had worked on the weak coupling version) and how you could explain the anomalously small cosmic ray cross-section for scattering.

And so I thought: “Well, if you had meson pairs, then this would explain the Italian experiment.” That is to say, if the coupling of the cosmic ray meson was through pairs, then one could avoid the discrepancy. A pair of mesons could be strongly coupled, and a single one could not be absorbed. In other words, two mesons had to be produced together. That would explain the low scattering cross-section, and everything was just fine. On that basis, I suggested that the energy had to be at least, say, 225 MeV and Barnes himself was a restless soul and always interested in moving to the frontier of a field. So whether he came to me one day and asked, “Suppose we had a bigger machine?” or whether I went to him and said, “After the Italian paper, of course you can’t do meson physics,” I certainly told him: “You won’t do meson physics” with a 150 MeV synchrocyclotron. And after he mused about it, Barnes said, “Well, perhaps we could build a bigger machine” and he and George Collins persuaded the ONR to go to 240 MeV. The fact is that the Rochester machine energy was increased chiefly on the basis of the argument that at least you’ll stand a chance of producing cosmic ray mesons in pairs... Now, the amusing thing is that these arguments were transmitted by Sid Barnes to Bob Wilson. But Wilson decided that they were not weighty enough as far as he was concerned, and he stuck to the 150 MeV energy for the Harvard machine, and they got a 150 MeV machine. They never had a chance to do meson physics. For a slightly increased cost, we did go for the 240 MeV. The ONR did agree to go higher in their financial support. And we did some crucial meson physics at Rochester, like measuring the spin of the π meson.

Weiner:

The ONR accepted that argument.

Marshak:

Essentially, Urner Liddell of the ONR was sympathetic with the Rochester strategy and he provided the additional funds. As you know, Rochester did produce mesons, because the π meson, the heavier meson, existed, and was the Yukawa meson. And the 240 MeV energy was enough to produce it singly and make possible some interesting meson physics.

Weiner:

The interesting thing about this is the planning. I notice that the Italian paper was published in Physical Review in ‘45.

Marshak:

‘45?

Weiner:

Yes. It’s dated October ‘45 from Rome.

Marshak:

But when was it actually published?

Weiner:

My notes say Physical Review, Nov. 68, 1945. I’m pretty sure that was it.

Marshak:

I personally hardly read an unclassified publication until the Fall of ‘46. That is, I was still bound to Los Alamos through June of ‘46, finishing up those reports, apart having those occasional chats with my Rochester colleagues. And then during the summer of ‘46 I was at the General Electric Research laboratory, where Hans Bethe, Richard Feynman and I were consultants, giving lectures on neutron diffusion and other types of atomic energy problems. The General Electric Company had made a decision to go into atomic power in a big way, with the help of some key young people like Harvey Brooks and Henry Hurwitz. Actually, Harvey Brooks and Henry Hurwitz wrote up my lectures on neutron diffusion (later published in Nucleonics) which I gave at GE that summer. Hans Bethe gave other types of lectures, and in general, we talked with many people about the future of atomic power. As a matter of fact, I think it was during that summer that GE became involved in building a high-energy accelerator. Jim Lawson was interested and thought that the General Electric Company might actually make money if they got into the accelerator business.

Weiner:

Is that the one that had the false anti-proton that came out of there?

Marshak:

Well, I think it was a machine with very little iron and so on. Lawson had his own technique. But they gave it up after another ten years or something of that order. They built it and ran it for a while. But in any case, that summer I was sort of getting out of the neutron game - first, by finishing up reports and secondly, by passing along the message to the heathen. And I guess it was really only in the fall of ‘46 that we started buckling down and getting back into the physics that was completely unrelated to neutron diffusion or atomic power or similar applications.

Weiner:

Or civilian control of atomic energy.

Marshak:

Not quite. On the political side, we continued to maintain different levels of interest. I think the first year after the war Bob Wilson was elected chairman of the Federation of Atomic Scientists. So that first year he was pretty busy with scientific politics. The following year I was chairman of the F.A.S. and I had to spend a lot of time outside physics research. I would say the first year I had a chance to get back into civilian physics was from the Fall of ‘46 through the Spring of ‘47, when the Shelter Island Conference took place. And it was during that year, I guess, that we began to fully appreciate the discovery of the Italian group and get thinking about it, and when the theorists in general - Weisskopf and the others -, whenever we saw each other, would wonder about the important implications of it. Now, who first recognized this paper as being so important, that’s hard for me to say. It would be interesting to know just when the paper by…

Weiner:

Fermi-Teller-Weisskopf?

Marshak:

Yes.

Weiner:

That was in ‘47 — Physical Review, Volume 71. It’s on page 314 and 315. By the way, pull out the volume, please, because there’s something else that relates later to Conversi-Pancini and Piccioni...Because they’re in the same volume on page 209. They have another paper, which says that for graphite “delayed disintegration electrons are observed to be about equally abundant for positive and negative mesotrons.”

Marshak:

Well, that’s crucial.

Weiner:

But it was the other one, the earlier one, which was the basic paper, because in that paper they said specifically — and you cited it later in your two-meson hypothesis - they pointed out “the greatly different behavior of negative and positive mesons so that the prediction of Tomonaga and Araki seems to be confirmed experimentally.” And their prediction was “on account of the electrostatic interaction with nuclei, the capture probability should be for negative mesons in dense material much greater than the decay probability.”

Marshak:

Sure, that’s the point. Yes. The reason is, you see, that the first paper...This is very interesting. The first paper is not the critical one. The first paper just points out that there’s a difference between positive and negative mesons, which is what you expect. This is an agreement with the Yukawa theory. And they’ve measured the effect. It’s the second paper. That’s the whole point. The first paper did not register. Even granting that it took a month for it to come to Los Alamos - December ‘45; well, many people would have left but one would have talked about the big discovery pretty early. So it was not until the second paper, which came out in February of ‘47 that the excitement started. That’s why I was saying early ‘47. So my recollection was correct there.

Weiner:

I see. Well, let’s pick up that thread then. In that case, the ‘47 paper came out - what month was it?

Marshak:

They say, “In a previous letter we first gave an account of an investigation of the difference in behavior between positive and negative mesons.” You see, that is what’s expected if the meson is strongly coupled. Then it is expected that the positive ones will be repelled by the positively charged nucleus and decay while the negative ones will be captured into orbits. Once they are in the orbits, they are quickly absorbed and they should not decay. If the mes1s are strongly coupled, then they should be absorbed in 10 seconds. So the strange thing was that you actually had decays in carbon. In other words, that the difference in behavior was not great enough. The difference should have been such that positive ones decay and negative ones disappear. But actually negative ones also decayed. So the difference in behavior that was found in iron was not found in carbon as well.

Weiner:

Which was something that they didn’t recognize.

Marshak:

Yes. I think the full implications perhaps were not fully appreciated by the experimental people. It was the paper of Fermi, Teller and Weisskopf, three powerful men that really underlined the very serious discrepancy between the Yukawa idea and this experiment.

Weiner:

The interesting thing about this is that on February 7th 1947, their paper was received. The timing is important here. That’s the reason I was going into it, and you’ve clarified it. This means that in early ‘47 these things were coming to the fore - their paper and the Fermi-Teller-Weisskopf comment on it. This then should be the time that you had another discussion with Barnes about the machine, and so you would expect that the change in the energy goal for the machine came about some time in the spring of ‘47.

Marshak:

That’s sound right.

Weiner:

Which was the same time that you were beginning on the two-meson theory.

Marshak:

Well, the two-meson theory was invented right at the Shelter Island Conference, which was in June, early June of ‘47.

Weiner:

That was because the other experimental results were coming in.

Marshak:

No.

Weiner:

No, no, you were unaware of it. All right, we’ll talk about that. Things were moving very rapidly.

Marshak:

Things were moving very rapidly. What happened — and these are interesting points to mention perhaps — so far as the Rochester synchrocyclotron was concerned, I think I fed in the level of advice that I indicated earlier; and that was enough, together with Barnes’s persuasive powers on Urner Liddell, who was then connected with the ONR, to get them to agree to let Rochester build a bigger, a 240 MeV machine. Apart from that, what happened was this: Oppenheimer received a grant, I guess, from the National Academy of Sciences. And in particular, a non-physicist called Duncan MacInnes from Rockefeller University (he’s in that picture that I have) was trying to help rebuild American civilian science. There was a decision made to try to activate various areas, and I guess Oppenheimer persuaded him that theoretical physics should be part of the action, and he ran this conference. He was the chairman of the Shelter Island Conference, the first conference, which actually was held at Shelter Island. The next two were called Shelter Island Conferences, but they were actually held in different places near New York City — one in the Poconos and one at Oldstone-on-the-Hudson. Now, in sending out the invitations, it was limited to 25 persons, and I was fortunate to receive such a invitation.

I believe that Weisskopf was instrumental in getting me on the list. I think Weisskopf told me that he suggested to Oppenheimer that I be invited, because I did not know Oppenheimer too well, as I indicated earlier, during the Los Alamos days. He knew who I was, and we had talked some but it was not a close relationship. But Weisskopf and Oppenheimer did have a very close relationship. So Weisskopf, even though he was no longer at Rochester, suggested my name as a person to invite. At any rate, Oppenheimer sent out a memorandum indicating the types of questions we might discuss at this first meeting. The two major ones were the Lamb Shift and its implications, and the second one was this meson dilemma. And he also, in order to I think stimulate prior thinking and discussion at the meeting, indicated (I think it was a paragraph) that the meson discrepancy was a very serious matter and perhaps one had to give up detailed balancing. In other words, the problem was that here you were producing these mesons with large cross-sections in the upper atmosphere — and this was known from the Marcel Schein experiments - and somehow when you observed these same mesons at sea level — which was the Italian experiment — they did not interact strongly with matter. And that meant that detailed balancing, which relates inverse processes, might have to be given up, which was a very revolutionary idea. So that was Oppenheimer’s spur to the discussion before we came to the meeting, as far as the second problem was concerned. Well, then the meeting took place.

Weiner:

This would account for the title he gave to the meeting: “The Conference on the Foundation of Quantum Mechanics.” In the sense that you’re talking about something that’s fundamental and giving up detailed balancing and so forth. I don’t know whether or not he had the title as a general kind of title or whether he had something specific in mind.

Marshak:

I think the reason that he gave it that title was that he really thought of it as a sort of comprehensive conference which would discuss in a broad way the foundations of quantum mechanics, because there was a talk by von Neumann, who presented a completely new approach to quantum mechanics. I guess I’ve mentioned the two major topics, perhaps because they were the ones that led to the most progress by the end of the conference. But Oppenheimer, I think, originally had a broad conception of the conference and simply made some suggestions as to interesting topics we might discuss. I’m quite sure that he did not think of these two topics as limiting the agenda. Other subjects were discussed. But I believe he mentioned both of these in the original call or summons to the meeting.

Weiner:

You have that somewhere?

Marshak:

I think I have it somewhere.

Weiner:

I notice also that Kramer’s, Oppenheimer and Weisskopf were supposed to play the role of discussion leaders.

Marshak:

Yes. The picture does not have Kramers in it, and I don’t remember Kramers as being there. Do you have a definite listing?

Weiner:

I have a definite list for the Poconos but not for that.

Marshak:

I must establish that. I was a little vague in my article, because I could not get hold of it. Certainly Kramers was there in spirit if not in body, because he had made the suggestion to someone in the group that perhaps the Lamb effect had something to do with self-energy. It was a qualitative idea that was certainly in the back of certain people’s minds who were interested in that particular subject, Weisskopf must have known about it, and certainly Oppenheimer knew about it. But there was a chance he was there even though I don’t quite remember him as being there. I guess my mind was focusing more on the meson problem rather than on the electromagnetic infinities.

Weiner:

With Oppenheimer’s provocative statement about what we might have to give up, in the introductory memorandum for the meeting, did that stimulate your thinking?

Marshak:

I would not say that I came to the meeting with any idea that the way to resolve it was two mesons. I mean I came to the meeting ready to listen to all ideas, although I personally had been keenly aware of the Fermi-Teller-Weisskopf paper. I thought it was extremely interesting and realized that this was a major dilemma. I certainly was much less aware of the subtleties of the argument that went into the other topic, on the Lamb Shift. I had not been following that quite so closely. I guess when I made the decision in ‘39 not to work with Weisskopf in quantum electrodynamics, I didn’t follow that area quite so closely. And I guess it’s fair to say that I never did write a paper that had any value on the renormalizable aspects of quantum electrodynamics. Somehow I didn’t take to that particular subject. I certainly was more interested in the meson area when I came to the conference, but I didn’t have the two-meson theory in mind.

Now, what happened was that this was not the first topic discussed. The Lamb shift was discussed more fully, and I think several other topics - as I said, Von Neumann’s new approach to quantum mechanics and some other subjects. But the meson problem came on at a certain point, in the latter half of the conference. And Oppenheimer presented again his point that perhaps one had to make a major change in our approach to quantum mechanics. And I think Weisskopf and Teller summarized their quantitative arguments. Rossi was there as a sort of experimental advisor for cosmic ray physics just as Rabi was there for atomic physics. Rabi knew all about the Lamb-Rutherford work and the work on the anomalous magnetic moment of the electron. And Rossi knew about Schein’s work on the upper atmosphere. So there was a discussion going on, and everyone was very stumped as to what the resolution of the Italian experiment was. I think it was at that point that Weisskopf came up with the suggestion that maybe the way to resolve the dilemma was somehow to have the nucleon produced in an excited state, so that when a meson is produced, it is not the same as when it’s absorbed. It was a rather obscure suggestion.

Weiner:

A meson-pregnant nucleon.

Marshak:

Yes, something like that. Well, I guess it was shortly after that that I suggested; it seems to me that the simplest way to resolve the whole difficulty is to have two mesons, to have the heavier meson produced in the upper atmosphere and have it decay going through the atmosphere into the lighter one, and then the lighter one has the weak interaction. There was a little discussion of that suggestion and considerable interest. I remember Feynman getting very excited about it and saying, “We’ll call it the marshon.” And it had a sort of immediate, fairly good reception. But, of course, one had to work it out a little more, and one had to see two mesons. So essentially at this conference the idea was born of having two mesons. The basic idea was that a heavier one would be produced and would be the Yukawa meson, which would decay to a lighter one and which would then be the cosmic ray meson that is observed at sea level.

Weiner:

You were thinking of a positive meson at the time?

Marshak:

No, both charges. I don’t understand why you ask this question.

Weiner:

There was the idea of positive decay in the paper later. You talked about the positives decaying and the negatives being captured.

Marshak:

Oh, the reason for that - perhaps I should explain it - is that the nucleus is positively charged. So a positive meson won’t get close enough, you see, to be absorbed, whether it’s weakly or strongly coupled. It’s only the negative ones that get captured in the orbits and you get the mesic atoms, that you then get decays or absorption depending on the strength of the interaction of that meson with the nucleons in the nucleus. And the competition between the transition probability for nuclear absorption and the natural decay probability, decides whether it’s going to decay or be absorbed. In other words, if the transition probability that you calculate, which goes as the square of the coupling constant, is large compared to the transition probability for the decay, then it will be absorbed. That is, when a meson is in a certain orbit, it has a certain probability of being right at the nucleus, the square of the wave function at the origin gives you the probability density at the nucleus. That’s why it has to be essentially in an S state, which gives non-zero probability of being at the nucleus. And if that probability is larger than the probability for its decay, then it will be absorbed.

That was the problem: when you calculate the probability density of a negative meson in the K shell of carbon and put in the coupling constant that you derive from nuclear forces, the meson should be absorbed in 10-20 seconds. However its lifetime is 10-8 seconds. So the ratio is going to be 1012 seconds. So you should not see any decays in the Italian experiment if the meson is a Yukawa meson. But if the sea level meson is weakly coupled - that is, the coupling comparable to the strength of the coupling given by the theory of weak interactions, then it makes the probability for absorption of the mu meson, which is weakly coupled, comparable to decay. So you see an equal number, about half of them roughly, decaying in carbon in the Italian experiment. But otherwise, if it were the other way, you shouldn’t see any negative mesons decay in carbon. You shouldn’t see any decay in hydrogen.

Weiner:

This kind of explanation that you’re giving still follows the argument that you were making in the actual paper.

Marshak:

Yes, well, nothing has changed.

Weiner:

But what I’m asking now is: how much of this did you discuss at the meeting itself, at the Shelter Island meeting? And how much of the argument that you’re developing now was worked out in more detail after the meeting for the paper?

Marshak:

I would say that the basic ideas were presented there. In a word, you have to break the link. That is, the way to overcome the detailed balancing argument was not to have the same particle participate in both processes. That is, if a particle produced in the upper atmosphere is the same particle as the one at sea level, since you know the nuclear cross-section is very large, the meson has strong coupling in the upper atmosphere. And if it’s the same particle at sea level, it should have a nuclear cross-section or a nuclear transition probability when it’s absorbed. And then Oppenheimer’s point that you have to give up detailed balancing is correct. But if you have a particle produced in the upper atmosphere with a nuclear cross-section and it’s the Yukawa particle, and it has time enough in the atmosphere to decay; then most of the particles at sea level are the decay products of Yukawa mesons. If the decay products do not have any strong interaction, then you can understand the Italian experiment. So Oppenheimer’s point was well taken. Something had to break the detailed balancing argument. The answer is no, you don’t give up detailed balancing. That would change everything in physics. That would be too revolutionary and you don’t give it up so easily. But the resolution of the dilemma was another example when it was more economical to invent a new particle and retain a law of physics. So in a sense the two-meson theory was intended to save a well-known law of physics - namely, detailed balancing, which was well established.

Weiner:

That had happened with the neutrino earlier.

Marshak:

Sure. The neutrino was an example of a particle being introduced to save the conservation of energy, momentum and angular momentum. But this was another very important law of physics, detailed balancing - that’s microscopic reversibility and so on - and it was sort of basic to quantum mechanics that the square of matrix element, whether you take it between state A and B, or between B and A, should be the same. And by introducing another particle one could eliminate the need for giving up microscopic reversibility. But, of course, the question was whether this was the way nature solved the paradox.

Weiner:

Were you thinking on your feet as you were presenting it?

Marshak:

Yes.

Weiner:

It wasn’t a question of going to your room one night and...

Marshak:

No, it all happened right at the meeting.

Weiner:

So as the discussion developed, the idea occurred to you...

Marshak:

I find many good ideas for me get generated in interaction with other people. That’s why I always like to talk physics with other people. I find the friction of ideas is very helpful to me, and I try as much as possible to express my thanks. Most of my papers, except for the early formal papers on neutron diffusion, that were primarily mathematical, are co-authored for that reason. For example, it was in talking to a student that I picked up the simple way to measure the spin of the pi meson, which was crucial a little later. This is why I always like to talk to students or colleagues, to exchange physical ideas.

Weiner:

Were you called on to make a presentation or did it come out in discussions?

Marshak:

It was just general discussion. Remember, it was a small group. That was Oppenheimer’s idea to limit the conference to 25 people. He was basically the leader of the discussion. He would throw something on the table that he thought was interesting, and everybody could bring up other points, of course, within the limitations of time. This was an item that was on the agenda and after he had made some remarks there was a general discussion.

Weiner:

Weisskopf suggested at Shelter Island the meson-pregnant nucleon hypothesis to get around this difficulty?

Marshak:

It was he who in a sense brought the discussion a step further. It was after his remark that it occurred to me that his was a rather clumsy hypothesis and that the two-meson hypothesis was much cleaner.

Weiner:

Had there been any discussion of his hypothesis prior to your getting up, or were you next?

Marshak:

Very little. I would say...there was a little discussion, not much. And then when I presented my remarks, there was a better reception. As I indicated earlier, I do recall Feynman saying, “That’s very interesting, and we ought to call it the marshon.” Now, I am not sure whether he said it to the whole group or after the discussion ended and he came up to me and made that observation. There was a good deal of discussion, I think, but not discussion which said, “This is it,” because there was no evidence at that point for the second meson; so it had to be considered a hypothesis.

Weiner:

I hope you don’t mind my trying to recreate the atmosphere here just to get a picture of what these kinds of moments are like in terms of how you come to an idea and how you develop it. Now, when you did get up and talk, was it a continuous kind of thing, where you just didn’t offer a sentence or two but where you developed the entire argument? Were you able to do that without interruption?

Marshak:

Well, I think I gave the basic points.

Weiner:

And then there was a response to it and then you replied again?

Marshak:

Something like that. It didn’t come full blown, say, in the way the paper finally emerged, because actually within several weeks the darn thing was discovered.

Weiner:

There are some questions on the dating again. Their article (the article we’re talking about was Lattes, Muirhead, Occhialini and Powell) was in Nature 1947, I think a few weeks before the Shelter Island Conference.

Marshak:

Oh, yes, that’s quite possible, but...

Weiner:

May 24th, '47. Had you seen this?

Robert Marshak discusses his work on meson theory in the context of his work with the Federation of American Scientists

Marshak:

No, no, that’s the point. This is part of the story. It’s a sort of amusing story. You see, by that time, by the end of April 1947, before the Shelter Island Conference, I had been elected chairman of the Federation of American Scientists (FAS) (in the early days FAS was Federation of Atomic Scientists, but the change took place at about that time) So I was now being brought into political discussions on atomic energy and so on. And towards the end of June 1947, President Hutchins of the University of Chicago, who was very interested in these matters, had arranged a little conference at Lake Geneva in Wisconsin — for scientists and social scientists — to discuss international control of atomic energy. The Emergency Committee of Atomic Scientists, Einstein and Szilard, had been very much involved in these matters. Einstein was not there. Szilard was there and some other members of the Emergency Committee — Harrison Brown.

They were sort of a small elite committee working more or less in parallel with the big mass movement of the Federation — the younger scientists, say — of the Federation of American Scientists which had about 3000 members. Anyway, as the newly-elected chairman, I was invited to attend that little conference and Phil Morrison was on the same train with me from Chicago to Lake Geneva. He said to me...I guess he had heard about the Shelter Island Conference, probably Bethe had reported about it at Cornell, because the Lake Geneva meeting was only two or three weeks later. But Morrison said, "You know, there’s a very interesting article in Nature by Lattes, Muirhead, Occhialini and Powell. There seems to be another meson: there are a couple of photographs showing there’s another meson." And when I heard this, I jumped. I said, "My God, this must be the second meson I was talking about at Shelter Island." And perhaps I explained to him what I had said. I’m not sure that he really had heard about the two-meson theory. I don’t think there was a library at Lake Geneva, but a few days later, as soon as I got back to Rochester, I got hold of the latest issue of Nature, which had just arrived. The issue was dated earlier — May 24 — than the Shelter Island conference but in those days they were not using air mail to send journals. So that issue certainly reached the United States later...Well, I believe it reached the United States after the Shelter Island Conference. [arrived at U. of Rochester library, June 13, 1947).

Weiner:

Shelter Island was June 2nd through the 4th, and the publication date of Nature is May 24th, so even with air mail, it was close.

Marshak:

Well, certainly nobody knew about it at the Shelter Island Conference, absolutely no one. Otherwise, it would have been brought up by Rossi or Oppenheimer, and I never would have had to suggest it. Everyone will fully confirm this. I don’t think there’s any question about that. My impression is that it came by sea mail, and it took two or three weeks, arriving — in libraries in the middle of June; and Morrison saw it just before he left Cornell. It sort of jibes. We could check the Lake Geneva dates, which were towards the end of June. [June 18–21, l987] So, as soon as I got back, I found the two photographs, and I was immediately convinced that I had hit the jackpot! I had decided when I left the Shelter Island Conference that I ought to write something up on the two-meson hypothesis — a little note — but I had been sidetracked by atomic politics and the Lake Geneva meeting. But, as soon as I got back from that meeting, I decided that the FAS be damned, I was going to get that paper written up. So I worked on it and visited the General Electric Research Lab where Hans Bethe was a consultant for the second summer. I guess that summer I was invited to Brookhaven.

Weiner:

In '47 you were a visiting lecturer at Brookhaven.

Marshak:

For several of those early summers I went to Brookhaven. But I went to talk with Bethe in the Summer of 1947, because I still regarded him as my wise professor, and also, even more importantly, I wanted to make use of his knowledge of cosmic ray data. The part of that paper to which he contributed in an important way...I don’t have that...

Weiner:

I have it right here. I had it Xeroxed.

Marshak:

...was this part about the underground muon data setting an upper limit on the heavy meson lifetime. Actually, we missed the boat a little, because it could have been used to deduce the lifetime. It turned out that instead of the limit, later on, Ken Greisen showed that it really gave the value. But in any case, that’s why Bethe’s name is associated with the paper, because he did help me with part of it and, I thought, sufficiently to justify putting his name on it. But the theoretical calculation of the lifetime was mine, and it has this amusing sidelight: when I did the calculation...I considered what spin to assign to the heavy meson, but there was no clear-cut decision that one could make at that time whether it was spin zero or spin 1/2. Also, there was no clear-cut decision that one could make about the cosmic ray meson - the muon. Let’s now call the two mesons by their present names, muon and pion. One knew from cosmic ray data that the muon spin could not be greater than 1/2 - because otherwise one would have obtained too many electromagnetic bursts. This was shown in a paper by R. Christy and S. Kusaka many years earlier.

So the choice was between spin 0 and 1/2 for the muon. And then, of course, one could play games with the pion. It might be spin zero or 1/2. Of course, one could take something bigger. But it and could not both be 0 or both be 1/2. So one had to make a choice there. That’s where I let myself become a little lazy in the sense that I had earlier worked out the formalism for a spin 1/2 meson pair theory and just took spin 1/2 for the heavier meson and made an estimate, from the competition between the absorption in the K shell of carbon and the decay, as to what the lifetime should be for the decay of the pi-meson to the mu-meson - I arrived at a number which is given here as 10^-8 seconds. In the paper, we said “This value can easily be in error by a factor 10 or more in view of the crudeness of the calculation and the choice of a special model.” Now, what I meant there was: to take spin 1/2 for the heavier one and spin zero for the lighter one. The interchange of spins would not yield a very different answer as was shown soon thereafter by R. Lattes and Christy. The lifetime, of course, had not been measured at that time. There were only those two pictures that the Powell group had found. These are, of course, acknowledged in our paper and actually used to get an estimate on the lifetime, but it’s not a very useful estimate.

The best estimate was from the underground muon penetration measurement, which came very close to the experimental value 2 x 10^-8 seconds, measured later. Well, I wrote up this paper and for, some reason, I ran into Weisskopf and George Uhlenbeck at Ann Arbor. I showed them the paper and they thought it was premature to publish it. After all, there were only two pictures and so on. But I felt good about it. It just added up to me as being right, and the two pictures, of course, made me extremely confident that there were two mesons with the postulated properties. In any case, I wanted fast publication on our two-meson paper because I thought it was somewhat important. In those days, the way you could get fast publication in Physical Review was to send something you called a Note, which appeared after all the other articles in the regular part of Phys. Rev.; it was not a Letter to the Editor because it was longer - but there was no page proof that you looked at. And so I asked for that treatment, and our paper was published that way, I think, in the Sept. 15, 1947 issue of Phys. Rev.

Weiner:

It still took from July 29th when it was received until September 15th. That’s not exactly rapid publication. The page number is 506. Here we are.

Marshak:

It’s the last article before the letters. The shortcut procedure was not widely known. It was known by some old-timers, for example, Steinberger used this device to beat out Art Roberts on the spin of the π meson; but that’s another story. Steinberger got his paper into the same issue as Roberts, but he got it in as a Note. I think the procedure has been abolished in recent years.

Weiner:

You did get a Russian result along the way before it was published. You cite the Russian paper which came even more recently than the other summarizing some cosmic-ray evidence.

Marshak:

That was just a courtesy.

Weiner:

It didn’t support your hypothesis?

Marshak:

No, no. The Russians were constantly finding lots of intermediate mass particles...Alichanyan and his group - but they had small numbers of particles with all kinds of masses. Maybe some of the heavier mesons are pions but they had many more than are known even now. It was pretty poor stuff. But in the case of the Bristol work, these were nuclear emulsion events, and they looked very good, although they made a very serious mistake in the estimate of the mass of the heavier meson. They thought, at first, it was 1.7 times as much as the muon mass, and in that lies another story, but I’ll get to that maybe later. Let me say what happened after that. It’s sort of jumping a little ahead just in connection with the two-meson theory and shows how scientists are as sensitive as humans everywhere, and explains the sort of controversy that was engendered by my two-meson theory as late as 1965 at a conference in Kyoto. One part you might like to read. What happened was this: My paper was published September 15, and in January of ‘48, at the New York meeting of the American Physical Society, Oppenheimer came up to me, and said he had just received a copy of a paper by Sakata and Inoue in which they had also proposed a two-meson theory, and that paper had been published in 1946 but had just reached the United States because of the war.

After the war, the Japanese journals only started reaching the U.S. at the beginning of ‘48. Of course, I was interested to read the paper to see how much they had done even though my paper was already published. Okay, that’s just natural. Well, now here was a very strange situation. They had gotten worried about the discrepancy of a factor of 100 in the scattering cross-section of the cosmic ray mesons, which seemed to be smaller than predicted, and had said the way to explain the discrepancy between the large production cross-section and the reduced scattering cross-section by a factor of 100 - was to have two mesons. But they only had a factor of 100 to play with, and so essentially both mesons were strongly coupled, one just slightly less coupled than the other, so that the lifetime they were predicting for the heavier one to decay into the lighter one was 10^-21 seconds. In other words, the π-μ decay was not brought within the framework of weak interactions, which is what the 10^-8 lifetime does; and we know now π-μ decay is part of the universal weak interaction. But essentially the Japanese had two mesons which were both strongly coupled, one a little less than the other. However, in calculating this lifetime, they said, “Well, let the spin of the heavy meson be 0 or 1 and the light spin 1/2.” So after that, they argued that they had been the first to develop the two-meson hypothesis which explained the Bristol experiment.

Weiner:

But it was published so far in advance.

Marshak:

Yes. And furthermore, when it was established that the spin of π was zero and the spin of μ was a half, they said, “Of course, we’re the ones who did it.” Well, they felt so strongly about the matter that they refused even to acknowledge my paper in their publications. And I tried to explain to Yukawa, who was in the United States and with whom I became quite friendly, just what the differences were and what had happened and he was supposed to have explained it to the Japanese. But their grudge kept on for 20 years. And they thought it was an example of American scientists’ chauvinism and our overbearing attitude to them. And Sakata made a big issue of it at the 1965 conference in Kyoto, a small international theoretical conference to honor Yukawa, which I helped them get because I was a member of the IUPAP High Energy Commission; I encouraged them to have this conference, which was supposed to be organized as a larger conference with at least a hundred participants, but which Yukawa, under the pressure of Sakata and Taketani, reduced to 25. This small meeting was then used as a platform to make a strong denunciation of those who had tried to take credit for the two-meson theory. I answered Sakata in what, I thought, was a rather fair way. I said, “Okay, as far as priorities are concerned, your paper was published first, but your publication did not reach the United States until after mine was published.

That can be documented. So there was no attempt to steal your idea. As far as the intrinsic merits of the theories themselves are concerned, you were on the right track in terms of thinking about two mesons, but they were two strongly coupled particles, one slightly less strongly coupled; and your lifetime is off by a factor of 1013. It’s true that in your calculation you guess that the spin is integral (0 or 1) for the heavy meson and spin 1/2 was for the light one. On the other hand, I had the basic ideas correct - that the π was strongly coupled and the muon was weakly coupled to the π, weak in the same sense in which all the weak interactions are weak. I had the right lifetime as a result. But I inverted their spins, although it was certainly clear and could be shown trivially that reversing the spins did not change the prediction of the lifetime by more than a factor of two. So in one sense we were both right and both wrong. We each had part of the story. And let’s forget about our disagreements.”

Weiner:

You were there. This was a confrontation there.

Marshak:

Yes, and I answered in that vein. It’s in the record. Then the Japanese became friendly. Two years later, Sakata came all the way to Tokyo to pick me up to take me down to Nagoya to give a lecture and so on. So I think we made peace on the two-meson theory.

Weiner:

It had been smoldering. Had he ever written to you personally?

Marshak:

No, he never wrote to me personally, but Yukawa let me know how upset they were, and in publications, you could see how they were treating it - essentially trying to neglect our paper and always saying that the Japanese had discovered the two-meson theory. So if one looks at it from the point of view of scientific credit, the π and μ particles were discovered so soon after the Shelter Island Conference that I never had a chance to publish the paper as a prediction, and it looks as if it came after. Only the 25 people at the Shelter Island Conference heard me in early June, and my footnote in the paper makes it clear that the theory was proposed before our knowledge of the experimental discovery. As you indicated earlier, the Bristol paper was published in Nature at the end of May. Perhaps, in these days such situations could not develop because probably somebody would cable the results to you. We have such good communication now that the Bristol discovery would undoubtedly have been cabled, and we probably would have known about it by the time of the Shelter Island Conference and we would have been spared all that excitement.

Weiner:

By the way, after the Shelter Island Conference and after being alerted to this by Morrison and having seen the article, were you in any communication with Powell regarding further results?

Marshak:

Oh yes. Actually, what happened was that in November of 1947 I was in Europe for my first visit.

Weiner:

It was already printed.

Marshak:

That’s right. But I had sent Powell a copy of our paper, and I visited his laboratory two months later. Actually, I came to Paris for a meeting that was really connected with my position as chairman of the Federation of American Scientists but then stopped by Bristol to discuss their work and my paper. When I visited Bristol, we had some very interesting discussions. And I really do believe that they did not fully appreciate what they had discovered. I was prepared because of the theoretical approach, having arrived at two very different kinds of mesons on theoretical grounds, and could point out to them some of the nuances. In other words, they just had the photographs with one meson decaying into another. The implication that the heavier one should be the Yukawa meson and the lighter one the muon was perhaps not fully appreciated.

I think I helped to clarify the situation because they did not have too much theoretical backup. I mean the Bristol group was a pretty empirical group. They did not have a theorist associated with their lab. This is my recollection and is supported by Powell’s review article in 1950. I could also say very quickly that by the Fall of ‘47 or the Spring of ‘48, Lattes had come to Berkeley and helped them find the π meson in Ilford emulsions at Berkeley, because apparently they were not developing the plates correctly or something like that. It was after Lattes’ arrival that Berkeley saw the copious production of pions in their 340 MeV machine. And I remember that in May or June of 1948 there was a little conference at MIT on meson physics and other new developments; Alvarez was claiming that they were producing mu-mesons directly in the target - in other words, as if the muons were strongly coupled. I argued vigorously against this claim because the whole picture seemed clear to me, and I had written a little note while at the Institute for Advanced Study, entitled “On Mesons π and μ”, in ‘48…

Weiner:

It’s dated the 23rd of December ‘48.

Marshak:

I might even refer to Berkeley in that note. But in any case it was clear to me that if mu-mesons were produced directly in the target, the two-meson theory could not be correct. Conversely, if that was the explanation of the Bristol observations, then the experiments of the Berkeley machine were incorrectly interpreted. And Alvarez felt very strongly about how carefully those experiments had been done. But later he had to withdraw the claim because those muons were not produced directly in the target. Indeed, due to the magnetic field, the π’s were produced and the muons were the decay products and, due to the magnetic field, the muons came back to the target. In other words, my convictions about the two-meson theory were so strong that I was willing to tackle experimentalists who thought that the muons were directly produced. I might say in the same connection that during the visit to England in November 1947, 1 spent several days at Blackett’s laboratory in Manchester.

And in the discussions with Rochester and Butler, they told me about some results they had obtained with cloud chambers exposed to cosmic rays. They told me two major things. One was that they had seen muons produced directly in the plates, and the argument was that they could measure the ionization accurately enough - they had magnetic fields - so that they could distinguish between a pion and a muon. And I said, “I don’t believe it. You must be making a mistake. Because you cannot produce muons directly.” This was even before the Berkeley misunderstanding. The other finding that Rochester and Butler told me was that they had two events with strange particles having masses of about a thousand MeV. I had become skeptical of their method because I couldn’t believe that they were seeing mu-mesons produced directly and hence I didn’t take seriously enough those first two photographs of the V particles, which were a major discovery.

That is, they had seen the first two examples of the strange particles, which were only established later. Rochester and Butler really had the first two pictures, and I saw them with my own eyes in November of ‘47, but didn’t give very much thought to them, because I had become a skeptic because of their claims about the mu-mesons, which I just didn’t believe could be produced directly. I reasoned that if they were making mistakes on muon identification and thought they had sufficient accuracy to distinguish π and μ mesons, then maybe they were also making mistakes in the other two photographs and therefore one shouldn’t be spending a lot of time thinking about the V particles. But if they had only made the one claim about the V particles, perhaps some of us would have started thinking about the strange particles as early as November 1947 and Rochester and Butler would have received a Nobel Prize! Another sidelight that you might be interested in, is that after discussing the pion and muon situation with Powell and his co-workers (Mrs. Powell was an active member of his scanning group), I had one pair of nylon stockings left - I had brought along many pairs of nylon stockings and other goodies for friends in England and France, because it was only two years after the war, and I asked Mrs. Powell - whether she would like to have the remaining pair of nylon stockings. She said, “Sure.” And, quite independently, at the end of my visit, I said, “You know, this is so fascinating, these π-μ decays, could you send me some plates with these events on them?” By November 1947 they were getting much more data; Bristol only had two in the original publication, but this was six months later, and they were moving along. So, sure enough, a few weeks later some plates arrived, and on a sheet Mrs. Powell indicated where the decays were - there were about 10 cases.

And I turned the plates over to Bernard Peters and Helmut Bradt, who were preparing experiments for the new synchrocyclotron. At first, I said to Peters, “Would you like to look for these π-μ decays? After you find them, I’ll tell you where they are.” He was not quite sure about it, but I asked Bradt and he immediately agreed. (Bradt died a couple of years later, unfortunately, as a result of an operation.) Bradt started looking, and he was soon joined by Peters, and the two of them identified all the π-μ decays and they learned the technique. At that point, they said, “Well, this is really quite interesting. What do we do now?” This was before the Berkeley people went on the air in terms of discovering π mesons. So the only thing you could do was cosmic rays if you wanted to study more massive particles. Well, I had heard that Frank Oppenheimer (then at Minnesota) had a small cloud chamber going up in balloons and asked him if we could piggy-back some nuclear emulsions in those balloons, and he said, “Sure”. So I suggested to Bradt and Peters that they contact him. And, lo and behold, the first plates that were brought back had the heavy primaries, and that’s how Bradt and Peters discovered the heavy primaries in the cosmic radiation. Frank Oppenheimer had some signs of them in his cloud chamber, but it was not as clean as the nuclear emulsions. Then, of course, once electron-sensitive emulsions were made - at about the same time - nuclear emulsion studies of the primary cosmic radiation in the upper atmosphere became a marvelous game. And that’s how Bradt and Peters got started in cosmic rays through a pair of nylon stockings given to Mrs. Powell! And they, of course, got so fascinated that they both dropped the large cyclotron - and we had to replace them - and they started doing some very nice work. That’s how our cosmic ray group was built up. It was a couple of years later that Mort Kaplon, who was doing a theoretical thesis with me, was assigned as research assistant to Bradt and Peters, because I was so fascinated by their work; I thought I’d rather have Kaplon work with them than do more calculations for me. I had enough students at the time. So, anyway, the November 1947 trip to England was a very fruitful one.

Weiner:

There are a number of things about that. It’s interesting that when you wrote the two-meson paper or when you introduced the idea, you didn’t have experimental evidence to back it up. And yet in 1940 you had criticized Bethe’s paper on the neutral meson by pointing out that there was no particular experimental evidence for his approach, and you could demonstrate in fact that you could have a theory which didn’t depend on unobservable. I’m not saying that you’re going back on yourself, but it’s an interesting kind of approach that you took with your own theory that you were sure of.

Marshak:

Well, let’s put it this way: the two-meson hypothesis was produced under duress, so to speak. I mean people were giving up detailed balancing and that was strong medicine. So one had to invent something new. In the case of the Bethe paper earlier, it was just that I thought I could give an alternative theory using only the observed particles that had equally good features. But Bethe’s neutral meson theory and my meson pair theory were crude perturbation calculations; in th2two-meson theory I was trying to explain a factor of 1012. I don’t think it’s terribly contradictory.

Weiner:

Do you think that without the Powell results you would have published your two-meson paper.

Marshak:

I think I would have published a short note. I would say that is not absolutely certain. My inclination at the time, as I recall it, was to write it up during the course of the summer; but I had already taken on the position as chairman of the F.A.S. and realized that I had to spend time on that. So I was not giving the preparation of the paper a high priority. But when Morrison told me about the Nature article, I sure went to work fast because it was clear that I was onto something very big. You see, the Powell discovery came as a bolt out of the blue, as far as I recall. No one had mentioned, as far as I recall, at the Shelter Island Conference that such experiments were even underway - in other words, that there was a Powell group in England doing these experiments.

Weiner:

He had apparently been working during the war. He was one of the few people not involved in the British war work.

Marshak:

Right. And he seemed to be refining the nuclear emulsion technique, and he sort of pulled a fast one in that he very shrewdly exposed the plates on a mountain top and nature cooperated and they got those two pictures.

Weiner:

Do you recall the impact of the Powell paper itself? In other words, without the theoretical explanation, just the reaction to the paper that Morrison mentioned. When Morrison told you about it, was he very excited about it? Or did he tell you about it only because he had heard that you had done something?

Marshak:

I’m not absolutely sure. It may be that on the train I started telling him about the two-meson idea. The other alternative is also quite conceivable to me...Phil Morrison is a perceptive person in picking up new ideas and getting excited about new things that have happened. It’s quite possible that it went like this: “By the way, have you heard or seen that paper with the two-meson photographs?” and then he told me a little and I said, “No, I haven’t. Tell me about it.”

Weiner:

Did you get the feeling, say, during the summer while you were thinking about writing it up that other people were talking about the Powell experiments? In other words, if Oppenheimer sensed a crisis in theory, I would expect a group of theorists would certainly be very anxious to get the jump on anything, experimental or otherwise, that would relate to this question. I was just wondering about this.

Marshak:

You might wonder in a sense why weren’t there some other papers of people who didn’t know about the two-meson theory, who were not present at the Shelter Island Conference, say, and got to know about the Powell paper - why didn’t they write papers which appeared at about the same time? Because in a sense there was a delay...our paper was not published until September.

Weiner:

I don’t know. I haven’t made the search to see. [Physics Abstracts (1947) doesn’t show any discussions of the subject between Powell’s paper May 24, 1947 and yours Sept. 15, 1947.]

Marshak:

That’s an interesting question.

Weiner:

That would be a symptom of excitement and of interest, and I just don’t know whether...

Marshak:

Partly that may be due to the fact that it was the beginning of summer and people went off on holiday, and partly, I guess, it was that during the summer I discussed the subject at Brookhaven, so it became clear that I was involved in the early stages. It certainly had an impact pretty quickly in terms of the planning for machines. Let’s put it this way; I think the Powell paper by itself maybe didn’t register quite so much in terms of what it meant directly for the change in strategy for the machines, what could be produced or not. I think perhaps - I may be immodest here - it may be not until my paper was published that the Bristol work was put into a broader framework, that there were Yukawa mesons after all that could be produced. I mean, as soon as soon as you read our paper, you realize that there are particles that you would expect to produce in accelerators. And that’s what the Bristol work was telling us. I have the impression quite distinctly, as I mentioned earlier, that when I spoke to the Bristol group in 1947, they were not very clear in their own minds as to just what they had discovered, and I was able to help sort of it out. And you will find in the review articles on mesons that Powell wrote in 1950 that he refers to agreement with our theory. But this connection got lost in the shuffle; after Powell’s experimental discovery was better known - he got the Nobel Prize for the discovery and then Yukawa got the Nobel Prize for the pion and that sort of took care of everything. People like Lee Haworth in later years gave speeches in which they tried to place the two-meson paper in a proper framework but I think the Japanese claim to the two-meson theory confused the situation and my work got buried because the experimental discovery happened so fast. Certainly, by the Fall of ‘47, everyone was talking about the π-meson as the one that would be produced in machines, and that’s when it became clear that the 240 MeV Rochester machine would produce π-mesons. Indeed, by early 1949, I drew up a complete experimental program for our people at Rochester - that is, I drew up a program with several pages of interesting experiments to be done, in my opinion; this became, in a sense, a blueprint for the experimental program at Rochester.

Weiner:

At the Pocono Conference, the 30th of March to the 1st of April, 1948, the Proceedings show that some people raised questions and arguments against some of your ideas on the origin of mus via decay of pions.

Marshak:

I forget that. What were those?

Weiner:

You might take a look at this. This is from Serber’s talk on Berkeley meson experiments. (going through papers)

Marshak:

Serber came from Berkeley, and he may be reflecting at that point this Alvarez business that I was talking about.

Weiner:

It’s a little cryptic. The reason I raise this is to find out how the idea fared. Always in retrospect.

Marshak:

It seems to have done pretty well!

Weiner:

...it worked out to be true. But you wonder how long it took for complete acceptance. This is an indication that...

Marshak:

Yes, well, this was the flak that I think was connected with the Berkeley result that I was referring to earlier.

Weiner:

Your discussion with Alvarez?

Marshak:

Right. Because, you see, this meeting was in March of ‘48, and I think that the MIT meeting was maybe a month later. And Serber was at Berkeley at that time and was very close to the experimentalists. You notice here he’s trying to save muon pair production and argues that one can explain the Italian experiment through this muon pair business, which was in a sense what I was suggesting right at the start as a possible way. That’s the basis on which I was arguing for the increased energy. So partly he was coming back to that point of view and arguing that somehow the Powell discovery had nothing to do with this situation. He’s clearly wrong here. It wasn’t obvious he was wrong then because the accelerator data were just coming in from Berkeley. But he clearly wasn’t connecting the Powell discovery with the Berkeley result. I guess the way I can summarize the situation then is as follows: If you believed the Berkeley experiments - that they had seen muons produced directly -, if that was correct, then my theory was wrong, because in my theory muons should not be produced directly. They were only decay products.

Now, the only way you could save the Italian experiment is to have muon pairs. So a muon pair production process - that is, production of muon pairs with a large cross-section - is consistent with the Italian experiment and is consistent with the Berkeley experiment but has nothing to do with Powell...So at this point Serber was reflecting the fact, I think, that Berkeley thought they were seeing muons directly produced, and he was willing to give preference to muon pair production - possible with the Berkeley machine - over the other interpretation and just disregard Bristol. For me to guess Serber’s thoughts on the basis of these few sentences is difficult, but it’s clear to me that he must have been aware of the Alvarez problem, because he’s trying to have muon pairs produced directly.

Weiner:

This is ‘48. When was it that Oppenheimer’s talk with you about the Sakata thing took place? This is an interesting thing. Maybe he had a vague understanding of the whole thing.

Marshak:

I don’t think so. In many ways, one can say that the Sakata-Inoue approach is close to our hypothesis because it’s two mesons related by a decay...

Weiner:

I wasn’t suggesting that he was thinking of that as an alternative explanation, but I was just curious...

Marshak:

Well, the Sakata-Inoue paper was brought to my attention in January of ‘48 at that American Physical Society meeting. So it was several months before this.

Weiner:

And before the Berkeley results?

Marshak:

Yes. The Berkeley results, I think, came in the spring just before this Pocono meeting, in the early spring of ‘48, because Lattes was still in Bristol when I was there in November of ‘47, and the machine went on the air probably toward the end of ‘47.

Weiner:

In the paper, “On Mesons π and μ,” you say... That was the 23rd of December of ‘48 - that’s already late - where you talk about the Berkeley group confirming your own prediction.

Marshak:

Yes. You see, then they had found the error, and that, of course, was a real triumph because when you make the experimentalists withdraw data, then of course it makes your theory look better.

Weiner:

Now, how about the other two-meson ideas? That’s the wrong way to put it. Schwinger had another meson in mind at one time. How did this relate and how were you aware of it and did people raise that as an alternative?

Marshak:

That’s a good question, and I think enables me to try to clarify the early history. When the meson theory of nuclear forces was first proposed by Yukawa, he introduced charged and neutral mesons and I guess even introduced spin zero. So his first version was pretty close later to the truth (except that the pion was pseudo-scalar not scalar). But when one started calculating, on the basis of the methods of the mid ‘30s, one would calculate, say, to second order perturbation theory - where one nucleon is assumed to emit a meson and the other one reabsorbs it - one found a Yukawa potential. Well, these Yukawa potentials would have, say, a l/r³ dependence (r is the distance between two nucleons) and when you try to calculate the properties of the deuteron with such a potential, it blows up. You don’t get finite energy values. So it means that you have to cut off the potential at very small distances. Well, we know now it’s cut off by a repulsive core.

But in those days people started speculating that perhaps one could cut it off by introducing another meson with a different coupling and possibly a somewhat different mass, so that at very small distances the 1/r³ terms cancelled; and you ended up with a 1/r term, which you can use. When you put that into the Schrödinger equation, you get finite answers. So to summarize it, I would say a whole class of models was proposed to salvage the Yukawa theory from the singularity at the origin. There was a Moller-Rosenfeld mixture, two kinds of mesons with the same mass. There was the Schwinger theory with two mesons of different mass. But all those mesons, no matter how many you introduced, were all strongly coupled. There might be a slight variation in the coupling constant but the mesons were really all strongly coupled. This is why I object somewhat to Sakata and Inoue claiming that they had this two-meson theory, because they pretty much were using two strongly-coupled mesons. There was a factor of reduction, but it was quite small. That is reflected in fact that the decay time for it π-μ they predicted was 10^-21 seconds, which is of the order of the nuclear time scale - a little longer but just longer by a factor of 100, which naturally is related to the factor of 100 reduction in the scattering cross section that was being explained. You don’t get into the weak interaction domain until you get lifetimes of at least 10^-10 seconds or longer. That is, weak coupling reflects itself in the fact that the lifetimes are longer for decay with the same amount of energy. And the π-μ decay has a 10^-8 lifetime corresponding to a decay energy of 140 MeV. You must really talk about one meson which is a hadron and another meson which is a weakly interacting particle (lepton). Perhaps another way to put it is: all those two-meson theories were essentially talking about hadrons. What I was talking about was one hadron and one lepton. The mu meson is a lepton and not a hadron. It’s not strongly coupled. The strong interaction is very different. Sakata and Inoue were on the right track but they were still talking about both mesons being hadrons. So when they say they had the right theory because they picked integral spin for the heavy meson and spin 1/2 for the light one (as I said, I reversed the spins for the purposes of the calculation), they forget that they never mentioned in all their claims that they were getting 10^-21 seconds for π-μ decay. I had to point this out to them in our discussions. They wanted to sweep under the rug the fact that they were basically talking about two hadrons and not about one hadron and one lepton. So the question is: who was more correct? This is probably an uninteresting question. But the basic point was that in our two-meson theory, one was a hadron and one was a lepton. And that’s what distinguishes it from all these other two-meson theories.

Weiner:

It’s interesting - the whole background and ramifications and preliminaries of that. Let’s talk about the timing. There are a couple of things I want to cover - if not today, then another time. One of the things I wanted to talk about that fits into this whole period is your total reaction to the Shelter Island Conference. It occurs to me that here you are...You were 30 at this time?

Marshak:

Yes, 30, pretty old actually.

Weiner:

You’re at a pretty distinguished conference, and this is a pretty small group called by Oppenheimer and some other distinguished people. You’re included because your abilities and work are thought appropriate, as you demonstrated at the conference. I’m curious about your reaction to being included and your reaction to the whole participation, because for many of the people at the conference they claimed that this was one of the most important and most successful conferences they’d ever been to in their lives. And this may have been to get a grant for the next conference, but in the letter where Oppenheimer describes his reactions and in the letter with Dirac sent to Oppenheimer, certainly the point was made that there was something very special about this meeting. I’m curious about your reaction in terms of the meeting itself, but also at that stage of your life. This was the group that attended. Darrow was there. [looking at photo]

Marshak:

Yes, he got in as secretary of the American Physical Society and Duncan MacInnes was there.

Weiner:

And Van Vleck and Breit...

Marshak:

It isn’t clear why Van Vleck was invited. Well, the foundations of quantum mechanics - the point was to broaden the participation.

Weiner:

Schwinger, Weisskopf

Marshak:

Feshbach was there.

Weiner:

And Feynman.

Marshak:

Pais was there.

Weiner:

And Wheeler looking much younger.

Marshak:

Oppenheimer was a very young man. Von Neumann was there, I think. You must make prints of this or they’ll be lost.

Weiner:

We’ll arrange for that.

Marshak:

Pauling was there. Foundations of quantum mechanics. You see, that’s how the broadening took place. That’s why Von Neumann was there and Bethe. That’s Dave Bohm.

Weiner:

He looks something like Dyson in a way.

Marshak:

Uhlenbeck, Rossi, and this is Duncan MacInnes and Willis Lamb.

Weiner:

Oh, yes, I see Teller is back here.

Marshak:

Altogether 23. See, there are a couple of participants missing in the photograph. Kramers perhaps and maybe Fermi was one of them. Perhaps Kramers and Fermi came for a day...

Weiner:

I think Fermi was invited and didn’t come. I’ve seen Oppenheimer’s papers in the Library of Congress - all the materials, but I couldn’t photocopy them without permission and I was in a hurry, and so I took only very quick notes, which are more extensive for some reason on the Pocono Conference than they are on this.

Marshak:

You know, you’ve just given me a sense in the last few minutes that maybe the papers connected with the Rochester Conference are about as important as these papers, and we should really save them.

Weiner:

Well, I’m assuming they have not been destroyed.

Marshak:

I haven’t so far. I thought I might throw out some of them.

Weiner:

I’m glad you brought it up. It hadn’t occurred to me that they would be destroyed. They should be saved, and not only for nostalgic values but for real documentation. Anyway, my question about the Shelter Island Conference - and that picture helps bring back the atmosphere - as concerns your reaction to being included, and then about the style of the conference itself.

Marshak:

Well, the style of the conference itself was delightful. It was very informal, and Oppenheimer was very stimulating and did a marvelous job running it. It was undoubtedly an extremely successful conference. It was after this conference that Bethe went home and made the first calculation of the Lamb Shift, a non-relativistic calculation, using an approximate cutoff and got very close to the correct answer, a thousand megacycles. This then led to the Schwinger calculations and then the Feynman calculations which were done relativistically and that were the highlights of the next two conferences: I think Schwinger in the ‘48 and Feynman in the ‘49 conference. So that this certainly was a major triumph. In one sense, from the short-term theoretical point of view, the two-meson theory, I guess, did not have equal status, because once you had the main ideas, one didn’t know how to calculate with strong interactions. That problem is still with us. I mean, in our two-meson paper, we could make qualitative predictions or do semi-quantitative calculations, but it didn’t lead to that tremendous calculation program which characterized quantum electrodynamics in the ensuing few years and which cleaned up the divergences within the framework of the renormalization program, and led to predictions that could be tested to many significant figures. So the first consequence of this conference was to give tremendous impetus to the creation of QED.

I would say the second consequence, the two-meson theory, in my humble opinion, was qualitatively just as important and certainly sparked the replanning of the experimental high-energy accelerator program which opened up the new era of meson physics. But in terms of elegant theoretical developments, it was not the immediate starting point of major successes because the strong interactions were too involved. So if one wanted to, you could compare, say, my two-meson theory to Bethe’s non-relativistic calculation, which might have been comparable at that point; his work led to the Feynman-Schwinger-Tomonaga-Dyson contributions and, indeed, the whole renormalization program. But there was not that comparable quantitative development after the two-meson theory. The fact is that one has been struggling ever since ‘47, for 23 years, trying to get a handle on the strong interactions - current algebra, dispersion theory, Regge poles and we still do not have a good dynamical theory. It was clear right at the outside that one should not take seriously strong interaction calculations done with perturbation theory - the approach that worked so well in quantum electrodynamics. Pauli and Wentzel tried strong coupling models, and in my meson physics book, published in 1952, I gave results from strong coupling theory as well as perturbation theory with many qualifications. I tried to emphasize the phenomenological results in meson physics - like the work on the spin and parity of the neutral and charged pions - which only depended on general principles like Lorentz invariance and conservation laws. But the fact is: a meson physics book written in 1952 could not compare in mathematical elegance with, say, a book written at that time summarizing the developments in quantum electrodynamics.

Weiner:

One was a book polishing off a field and the other was a book opening one up.

Marshak:

Yes, the basic point was that we were dealing with hadrons. Once you get into pion physics, then you’re dealing with strong interactions. And we knew that when you had a coupling constant of 15, that to expand in powers of 15 is not a very good thing. And then what do you do? So with strong coupling theory, one tried to expand in inverse powers of 15, but that’s not so simple. It sounds good, but it’s really very difficult to invert the expansion parameter, and then one tried all kinds of other tricks. So that while I was aware of the mathematical differences in elegance, somehow I myself enjoyed moving in this area which was mathematically less satisfactory but was, shall we say, more pregnant with interesting experimental possibilities and theoretical ideas.

Weiner:

It seems that you were at a perfect place for it, because here was a machine being built simultaneously. A machine which you could influence by your senior position as a theorist in the department and by the fact that the particular area of physics in which you were interested required the machine, and, in fact, required your advice if it was going to make an impact. It was fortuitous.

Marshak:

Well, of course, those were gratifications. I mean Serber was the theoretical man at Berkeley, and he was giving lectures on “Serber Says”. The Rochester synchrocyclotron was the second in the country, the first post-war machine, coming on the air. Incidentally, Cornell was building a 300 MeV electron synchrotron, and Bethe and I had talked about these things, and we each had reasons for pushing our respective machines. This worked out fine because one was an electron machine and one was a proton machine. He had arguments as to why he would prefer an electron machine, and I had arguments as to why I would prefer a proton machine. I don’t want to claim victory over my old professor - but it was essentially a technical difference that made it possible for the Rochester machine to turn out more novel results than the Cornell machine in those early years: the proton intensities were much higher than the electron intensities and the larger production cross-sections gave higher intensity secondary pion beams. Also the very interesting large nucleon polarization at higher energies was discovered at Rochester. Hence, the Rochester accomplishments with the 240 MeV synchrocyclotron turned out to be greater than with the old Cornell 300 MeV machine, which of course was later replaced.

Weiner:

Did they start operation about the same time?

Marshak:

Roughly about the same time. I think ours was ahead by a year. ‘49 was when Rochester started. A while ago, you asked me how I felt personally about being at the first Shelter Island conference. Well, I would say that this conference was quite crucial in my career. In a sense, the “spontaneous generation” at the conference of the two-meson theory probably helped a great deal, because soon, by ‘49, I was a full professor. The bigger quirk came a year later when I was asked to serve as chairman, but that’s another story. So I felt in the late 40s that perhaps I was making progress in my scientific work. Of course, I was past 30 by this time, but all of us had lost years because of the war. I felt that in many ways I was fortunate during the war years in that I had received exposure to many different types of problems and actually had made some reasonably original contributions - not in any deep physics because after all the war problems usually involved classical physics. It was more a question of developing mathematical techniques to get answers quickly. I was reasonably pleased with how far I had succeeded there, but the two-meson theory was probably the first serious physics contribution. I would say this was a turning point in many ways, because I became very interested in this area and started to help with the total planning, as I said earlier, of the experimental program for the Rochester department.

Weiner:

Were you at the 1946 Princeton conference, the 200th anniversary of Princeton?

Marshak:

No. I only heard of it.

Weiner:

There were other things too, but it was basically physics. Bohr was there.

Marshak:

I guess the invitation list was based on the older generation of well-known physicists. Perhaps Schwinger was there, because he was certainly well known.

Weiner:

Feynman was there, but that probably was because of his connection with Princeton.

Marshak:

Well, Feynman was clearly a rising star. He was a fascinating fellow and showed obvious brilliance. In terms of actual scientific contributions at Los Alamos, it is difficult for me to say; but it was clear that he was a very deeply creative thinker although I don’t think in terms of civilian contributions that he had any major accomplishment to his credit by ‘46, but they soon came.

Weiner:

I don’t know how much time we have, but let me tell you the kinds of things I have in mind. The Federation, the FAS, is a whole story in itself. I’d like to explore that. I don’t know if we can, in the time available, find out how and why you got involved, what the major issues were in that period. You know, there’s never been an adequate history written of the Federation. The papers are preserved in Chicago along with other papers of the Los Alamos group and the Cambridge Association of Scientists and so forth. Alice Smith only covered the immediate post-war year. But maybe that would be a good thing to end on, because it sort of fills out that period of 1940 up to about ‘48.

Marshak:

If you want to do that, okay. Perhaps, it’s better than just continuing through a few more years of physics. Okay, fine. Before doing so, I’d like to say a word about the remaining Shelter Island Conferences if you want. It might be good to do it, because between now and the next time I may come across the papers on the Federation and have some concrete material to give you.

Weiner:

In the remaining little while, you are going to comment on the other two Shelter Island conferences. The first was held in Pocono Manor, Pennsylvania, and the dates were the 30th of March to the 2nd of April, 1948. And the other was in Oldstone-on-the-Hudson in 1949. The title of the Pocono Conference was “Conference on Theoretical Physics,” and we do have the proceedings, such as they are.

Marshak:

Yes. Well, I think you can see what was covered from the table of contents. There were several topics at this conference, but the chief one had to do with quantum electrodynamics, which was developing at a rapid pace. If I had not glanced at the table of contents, I would have recalled as the chief event the unfolding of Schwinger’s theory of quantum electrodynamics - his approach to explaining the Lamb Shift and the anomalous magnetic moment of the electron and so on. I think the proceedings must have a very thick part of it devoted to Schwinger’s work and I see now that this is indeed the case. It looks as if Schwinger has taken pages 9 to 45, which is 36 pages out of a total of 80 pages. And Feynman was also beginning to report on his work and made his first presentation: it’s much shorter - only about 10 pages. And then there’s a discussion led by Niels Bohr at this conference, and my recollection is - and it probably will be confirmed by reading the proceedings - that he was very skeptical of Feynman’s approach. He objected to Feynman’s view of the electron going backwards in time but thought that the Schwinger approach was just fine. So that at this conference, Schwinger passed with flying colors but Feynman’s very unusual way of looking at quantum electrodynamics was running into lack of acceptance...

Weiner:

That’s from Bohr. What about some other people?

Marshak:

Well, Feynman’s ideas, presented relatively briefly, were so novel that, I think, to some extent Bohr’s reaction, which was not based on a deep study, was shared to some extent by others. It seemed a very strange way to approach quantum electrodynamics. Also, as I recall, Feynman had not completed his work, and there was a problem as to whether he could do something about vacuum polarization. There was even a chance that Schwinger and Feynman would differ in their predictions of the Lamb Shift by something like 30 megacycles. Vacuum polarization gave that sort of contribution. So that the two approaches were not considered as necessarily equivalent at the time. And, to me personally, certainly Schwinger’s approach did seem to be more rigorous, and Feynman’s approach seemed to require a willingness to use language in unusual ways. I guess many of us realized that Feynman was nobody’s fool, must have been getting the right answers, and one was keeping an open mind. I don’t think Feynman had published much before this meeting.

Weiner:

I’m not quite sure, but I thought it was later than ‘48.

Marshak:

Schwinger’s approach was so much more systematic, and sort of led from one premise to the next conclusion and so on, that I think one was more impressed by his work.

Weiner:

Was Tomonaga at the meeting?

Marshak:

No, Tomonaga was not at the meeting. I believe no foreign scientists were at this meeting who were not already visiting the country.

Weiner:

I have a list of invitations that were sent - to Bohr, Dirac, Heitler. Where was he at the time? Germany?

Marshak:

No, he was in Switzerland.

Weiner:

The rest of the people who appeared were Van Vleck, von Neumann, Weisskopf, Wentzel, Wheeler...They were all here.

Marshak:

Well, you see, both Bohr and Dirac were visiting the Institute for Advanced Study. They did not come from out of the country for this particular meeting.

Weiner:

The conference was a bargain, because the first conference cost less than a thousand dollars. The Academy made $3000 available for the Pocono Conference, but Oppenheimer said they probably wouldn’t need it. He ended up using about $1500. It was really a bargain. You couldn’t have done it with those kinds of funds and still bring in...

Marshak:

Somehow one didn’t think it through at that point...I don’t know why Oppenheimer didn’t become more venturesome and offer trans-Atlantic fare. One thing was that plane travel was not that good, although I did fly across in November of ‘47. It took about 14 hours to get to Paris, stopping in Gander and so on. So I guess it certainly was being used. But it cost money, and I guess no one wanted to invest that sort of money. So, the people who came, like Bohr and Dirac, were persons who were already visiting the U.S. Yukawa was not at this meeting. He might have been at the ‘49 meeting, because he was at Columbia, and I think he came to one conference. He was a visiting professor at Columbia...

Weiner:

The only tie with the Japanese was that when Oppenheimer returned from the meeting he received a letter...

Marshak:

From Tomonaga, yes.

Weiner:

...from Tomonaga giving a report of his own work.

Marshak:

Yes, I don’t think Oppenheimer would have had any reason to invite Tomonaga because he didn’t know about his achievements at the time the invitations went out. In any case, that would have been a very costly trip, would have used up half of his budget. You see, in the early years of the Rochester conferences, the way I brought people was again not to use direct money but to start providing MATS (Military Air Transport) transportation, which we could get free, if the air force or the ONR cooperated with us — said they would allocate a certain number of seats for our foreign guests at a nominal cost of $50 per person.

Weiner:

What, in general, did you want to say about the Pocono and the Oldstone conferences?

Marshak:

Do you have the comparable proceedings of the Oldstone Conference?

Weiner:

No, I was going to ask you. I’ve been asking around to see if we could locate it.

Marshak:

Well, maybe I’ll find them in the next few weeks during my move to New York City. On the Pocono, let me just say here that it’s clear that Serber was beginning to feed in the results of the Berkeley experiments and I guess a big emphasis was being made on the pions being observed in the Berkeley machine. Rossi was improving his analysis of meson decay using cosmic rays. You’ll notice that the first two papers were to bring us up-to-date on the experimental results, This was a procedure which Oppenheimer introduced, which was of course very valuable and which I followed in a more extended form in my Rochester conferences except that I had experimentalists attend on an equal basis with theorists. Here Rossi was serving as the experimental rapporteur. I see that Dirac was telling the group about magnetic monopoles which he was interested in for a long time. But the rest of the conference was on quantum electrodynamics, mostly Schwinger and to a lesser extent Feynman.

Now, at the Oldstone Conference, Feynman was the hero. That is, he had completed his calculations and shown that he was getting completely equivalent results to those of Schwinger. He understood the vacuum polarization problem and so he basically confirmed Schwinger’s work in quantitative terms. For that reason, the Oldstone Conference was not that exciting. Feynman presented a new method and, of course, it led later on to other important developments. But I think the general feeling at Oldstone was one shared by Oppenheimer that the three years of “Shelter Island” conferences had served their purpose. The original intent of these conferences was to take stock of what were some of the outstanding problems in theoretical physics - foundations of quantum mechanics and what have you - and that purpose had been achieved. In any case, Oppenheimer decided to terminate these conferences with the Oldstone Conference.

Weiner:

Let’s just list the kinds of things we want to get into next time. It seems to me we ought to get the full story of the Rochester Conferences. But before that, could we pick up the thread of some of your own papers, some of your own work in beta decay that we haven’t talked about.

Marshak:

I was going to suggest perhaps that at some point you might send me a very rough memo before your next visit and we can try to tie together some of the topics. I’ll at least try to get some of the papers together.

Weiner:

Well, beta decay — I’d like to pick up on that because it ties in with later work.

Marshak:

Yes, it led to the universal theory.

Weiner:

And then you refer to the Rochester Conferences. But then you go ahead and talk about your students here and the theoretical school you developed and the continuing links with the synchrocyclotron. And then the FAS story, the international scene — all of those in full dimension. We have at least a need for a similar stretch.