History Home | Book Catalog | International Catalog of Sources | Visual Archives | Contact Us

Oral History Transcript — Dr. Norman Kroll

This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.

This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.

Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.

Access form   |   Project support   |   How to cite   |   Print this page


See the catalog record for this interview and search for other interviews in our collection


Interview with Dr. Norman Kroll
By Joan Bromberg
At LaJolla, California
March 29, 1987

Transcript

Bromberg:

Let's start talking about magnetrons, then.

Kroll:

All right.

Bromberg:

When you worked on the Rising Sun Magnetron, and you worked on asymmetric rather than symmetric magnetrons, I'm interested in the way in which many of the papers that I asked about have to do with classical electromagnetic theory, and the fact that in some sense you started almost as an electrical engineers, except that electrical engineers were not really doing microwaves at this point.

Kroll:

Well, I would put it the other way around, that electrical engineering has turned out to be applied physics, and that all that kind of work in those days was being done by physicists.

Bromberg:

So is it the case that you wouldn't be reading engineering journals at this time?

Kroll:

That's right.

Bromberg:

And the people that you were working with were all physics department people.

Kroll:

Yes. The Columbia Radiation Lab was, I would say, the nucleus of Columbia Radiation Lab was — well, it was organized by Rabi, who I think you know, was also part of the obvious leadership group at the MIT Radiation Lab, and I think he just wanted to have some representation of that kind of activity at Columbia. That's my impression about it. You can find out more by talking to Dr. Rabi. But the nucleus of the people there, I believe, were people who had worked with him on the electron beam, Jerry Kellogg and Sid Millman. And there were some young Columbia faculty members like Willis Lamb and Arnold Nordsieck.

Bromberg:

What were you reading as you did this work, do you have any recollections? For example, I think of van der Pohl as somebody that people were reading, and yet I think of him as an engineer.

Kroll:

Well, let's say I was an innocent, and not very wide ranging graduate student at that time. I think that should be clear from my account of myself. I tended to have very broad interests, and I would say I still have, but back then, I really wasn't dedicated to physics in particular. It was in fact only at the last half of my senior year at Columbia that I really decided to concentrate on physics. And so I had one year of graduate study in physics and that's all, and I had taken one course in electromagnetic theory. I hadn't taken any upper division electricity and magnetism in college. So — but I had taken an ESMWT course in microwave.

Bromberg:

What's ESMWT?

Kroll:

Engineering, Science and Management War Training. I think it was called the ESMDT (?) at the time I took it, but that D changed to W after the war started. Well, the war had started already, but still it was originally D. that was right after I graduated from Columbia. I should have mentioned that. I didn't. But I guess it was given out in Yorktown, and it was the summer of '42, and that's the first, my first real introduction to — now I'm confused, now I think about it. There were really two episodes in the summer in which I got some extra training in what I would call microwave physics. One was the summer that I spent at Brown, in that Applied Mathematics Program which you may or may not know about.

Bromberg:

That still exists ?

Kroll:

No, but it was — and then I think organized it at Brown, and there were some very distinguished people, like the day I heard lectures from Leon Brillouin and Tamarkin.

Bromberg:

And what summer was that?

Kroll:

I'm not sure. That may be the summer of '42. It was the summer of '42 and the summer of '43, and one of those summers I went to Brown, the other I spent at NYU.

Bromberg:

And the NYU was the ESMWT course.

Kroll:

That's right. And that was certainly a course on microwaves. But there was some material in electricity also and microwaves, I mean the Brillouin lectures were also on microwaves at Brown, and there were lectures by Tamarkin on partial differential equations. ...

Bromberg:

So I'm not to picture you as running to the library and reading avidly on electromagnetic theory?

Kroll:

Absolutely not. You can't picture me running to the library and reading avidly ever. It's not my style. I'm primarily a listener and a doer. Not much of a reader.

Bromberg:

So what you're learning is as you work and as you talk to all these people.

Kroll:

Right, as I go to lectures, that's an important course of learning for me. Anyway, I probably took a course, as I said, I probably took the graduate course in electromagnetic theory at Columbia during that first year of graduate study, but I hadn't had the usual kind of undergraduate things. At Rice I had had a two year course in introductory physics, and the second year was electricity and magnetism, but it was at a lower level than the typical upper division course. It was a lower level than what was given at Columbia. But I guess what I'm trying to say is that I wasn't an experienced worker in electromagnetic theory at all when I went to the Radiation Laboratory, but I did have this expert kind of training in microwaves that I got from these courses that I mentioned.

Bromberg:

I asked, in that list of questions, kind of a reverse question, as to whether you have a feeling that that early microwave training had a permanent effect on the way you did any of your subsequent physics. You were really immersed in electromagnetic theory for a while. You were just telling me that electromagnetic theory in the classical quantum sense forms kind of a big component.

Kroll:

Well, it would be reasonable to assume it has.

Bromberg:

Did people who worked in the 1 centimeter radar, you said — now, that had of course its problems with absorption in the atmosphere.

Kroll:

Yes.

Bromberg:

Did that enter into what was going on at the Radiation Lab as you were working on these things?

Kroll:

Well, the exact peak of that, I guess it was water vapor absorption, was not really know at the time the work started. I, of course, as I said, I was a young and rather innocent graduate student, and I'd had no exposure to research whatever, probably less than one would have had if it had been a different time, because all the active research people were out of the classroom, and doing other things. But looking back, I would say I was less savvy than many of my fellow graduate students, even in terms of what was going on in physics, as distinct from what was going on in the classroom, because I'd really taken a lot less physics than other people had taken. So I went to the Radiation Lab. I was, I think, the first thing that happened when I got there, when I took that job, was, I was shown a microwave magnetron, the MO, and somebody explained to me how it worked, and it all sounded perfectly reasonable to me, and he showed me the big magnets and modulators, and how you tested magnetrons, so I started testing magnetrons, like all the other people in my situation. I didn't think much about it at the time. Well, I began thinking about, you know, in the environment I began thinking about it very quickly, but it wasn't as if I were making any conscious choices, and I certainly wasn't choosing the wavelength. The wavelength had been chosen. At that stage of my life, I certainly wasn't wondering why they didn't worry about atmospheric absorption. The reason for going to shorter wavelengths was obvious. I certainly understood that, the interest in getting higher resolution. You had to have reasonable sources of microwave radiation before you could investigate the absorption situation, because while the absorption was quite significant, if you wanted to use radar on a large scale, it was in fact a very small absorption and so it was not easy to measure. The sources and techniques that were available at that time was a very broad line and quite a weak one, so one of the things that happened at the Radiation Lab during that period was that an absorption experiment was done. That was I guess, Stanley Autler was the person, a student, who was involved in that, and Lamb did the theory, and I think in fact (Willis) Lamb not only did the theory, but he really invented the method for doing the absorption experiment. There was a large copper echo box, they called it, a big thing, you know, like 12 by 12 by 12, about that big, 10 x 10 x 10, and copper, and one — let's see. You sent microwaves into it, and you compared the level of radiation in the box — see, first of all you had to measure the level of radiation in the box, which was not all that easy. I think they had many detectors. And you'd measure the level of radiation in the box sealed, and then with holes of various sizes opened in the sides, and the idea was to compare the absorption, that is , the loss rate from the box through the holes in the side, which you could calculate very easily, by a rather simple photon argument, and with the absorption caused by the gas. In fact, I had some ideas about how one might — somewhat different from the ones they had, about how one might, another way one might do it, and there's a slight reference to that in Lamb's paper on the subject, and I remember, it involved, I don't know the details, but it did involve the use of silicon. I remember getting this sample of silicon from somebody, and making some measurements on it. So I recall the method I proposed had a kind of a rather more elegant theoretical appeal to me, but I think in actual fact the method used was much better. But once it became apparent, which was really after the shooting was over, so to speak, that the wavelength had been chosen at the worst possible spot, the old K band, one and a quarter centimeters, was redefined as the KU and the KL bands, and one well above and well one below that resonance. So the 8.6 mm band I guess is the KL band, or it could be KU, I don't remember, and the other was 1.65. Those became the official bands of interest, since the 1 1/4 centimeter which was very close to the Beetle? line, which I guess was 1.32, like that. I don't remember. 1.28.

Bromberg:

When you were working with Lamb on the magnetron, was that a close collaboration, or was that more like you were working out Lamb's ideas, or how did that work ?

Kroll:

Well, the first — I mean, as I said, I started out by testing tubes. I'll repeat some of the things I said before, but in somewhat greater detail. I started out by testing tubes, but rather soon after I had been there, they — well, rather soon after I had been there, the idea of the Rising Sun Magnetron appeared, and I don't know whose idea that was. My guess is that it was perhaps Millman or Millman and Nordsieck, I'm not really sure whose basic idea of making the resonators unequal in length it was. But I think that they might even have tried a — in any event, Lamb had done an equivalent circuit theory, very simple equivalent circuit theory of the symmetric magnetron, which I had seen. He had given — you know, there were little talks they would give to educate the people in the laboratory. And Lamb asked me, and I guess as I indicated, I guess they knew about me as a graduate student so they must have thought that I was the right guy to ask, they asked me to work out a similar equivalent circuit theory for the Rising Sun Magnetron, and I looked at that for a day, and I decided it wouldn't work, the way he asked me to do it, and he must have had a library of papers on various things — there were these papers by Goldstein and Clogston, which I looked at, and it seemed to me that the method that they used, which was not a simple equivalent circuit method, but a kind of value matching field theory, I looked at that and I decided that would work, and so I in fact worked it out. I had trouble getting Lamb to listen to me at first, because he said — he wanted me to do what he told me to do. I said, "It won't work," and I explained why it wouldn't work, and he listened to me. So then I showed him what I had done. I'd actually gotten the basic equations. So that was fine. The notion that once you had a formula, that that wasn't the end of the problem, that you actually wanted to get numbers was something I hadn't yet learned. So I said, "Well, here it is. That's the formula." So then a little later, he got some of the calculated results of the formula. He did it himself, I remember. It was a great big plot of frequency versus the ratio of the large to small resonators, and so that was an important educational experience, because I learned that one was actually interested in the answers. And there were some things about it which, you know, I looked at it, and then it occurred to me that it would be nice to plot it in a way that might be more understandable, that one might see what was going on better than it seemed to me that plot did, and I made the simple change of plotting the wavelength instead of the frequency, and then it all became clear, as to exactly what was going on.

Bromberg:

Was that the first research of that sort that you did, where you were really working at a level I would guess a PhD candidate was working at that point, getting your own results, seeing ways to do it better than the more senior person.

Kroll:

Yes. Absolutely the first. That's why it was very important in my — it's the first time I had the slightest glimmer of what physicists did, actually. Well, that turned out to be very useful, because it was immediately apparently what was happening, and it really hadn't been before that. The whole physics of the Rising Sun Magnetron. And then another feature which emerged which made it clear that Lamb's original approach would not have worked was something called the — well, it was related to the fact that voltages in the cavities are not equal, and that leads to what was referred to as the Zero Mode, and there's something called the Valley of the Shadows in the Rising Sun Magnetron, where the efficiency goes down, and all of that got clarified as a result of that work. So it made a major contribution to understanding how Rising Sun Magnetrons worked. So — well, all of that, of course, improved my status in the laboratory quite a lot.

Bromberg:

When was that, if you remember, in terms of when the war ended?

Kroll:

Well, I think all of that was pretty early. I think that was probably within the first year of my working at the Radiation Lab. I went to work in the fall of '43, and I think it probably happened in that academic year.

Bromberg:

Of course, Lamb started out pretty late in the Radiation Lab. He probably started right about when you did.

Kroll:

It's probably the case, yes, and I'm glad you made that remark, because what I think happened, now that you tell me, is that Lamb was not in the Radiation Lab when I went into it. He came in somewhat later. And it was as a result of his coming in later that I got asked to do this.

Bromberg:

As a result of his just becoming acquainted with the problems?

Kroll:

Yes. He joined the Radiation Lab, began working on these problems, and then he approached me. That's actually the way that I happened to get picked. He knew me because he'd had me in courses. I think I'd taken both mechanics and electrodynamic theory with him. So that's — now that I think of it, that's how I got taken out of the testing brigade. Well, I still did some testing, but I think it was recognized that I could do theoretical work on the subject, so that's mostly what I did after that. There were — well, there were a lot of things that happened. The paper that actually turned out to be my dissertation was a generalization of that work, because subsequently, other people in the laboratory began, you know, if it's nice to make two resonators unequal, maybe there are other configurations which would be better, and so I was shown a couple of other configurations that people were interested in, so I said to myself, well, as long as we're doing this, why don't I look at the situation in general? That again was very quick. I think it took me on the order of two weeks to do that and write up the main outlines of, so to speak, the whole problem. And that's in one of the Quarterly Progress Reports. I remember having that — I wasn't part of the senior staff, so I didn't get the Quarterly Progress Reports. They weren't issued to me. And I thought by that time I deserved to get it, and I —

Bromberg:

They were just available in whatever reading room you had?

Kroll:

Yes. I said, "In view of the fact that I wrote three-fourths of this, shouldn't I be on the list?"

Bromberg:

And so they put you on the list?

Kroll:

I don't think so. There was one of the graduate students who was on the senior staff, that was Arthur Ashkin, and it was I think because he was perhaps the first graduate student that was brought into the project, and he was very good, in fact. He had also done some quite important things with regard to the development of the previous magnetron they had been working on, but nevertheless, I saw no reason why he should be on the senior staff and I shouldn't be. But I guess by now they had on the order of a dozen people in my position, and they seemed to think they would have to put them all on that status, or none, so I don't think I ever succeeded in getting the Quarterly Progress Reports issued to me regularly.

Bromberg:

Now, I'm trying to think how to make the most profitable use of our time. I want to talk about field theory. Unfortunately, I don't know how much Sam already feels is well known and how much he wants that's new. You said in this interview that you started working on field theory with Halpern, but you really started to learn field theory when you worked with Lamb.

Kroll:

That's not what I said exactly. (or, is that what I said exactly?)

Bromberg:

You said a little bit about your work in field theory, during this period, when you didn't yet have your PhD, so let's talk a little bit about that now. You worked with Halpern on the double Compton Effect. What is a double Compton Effect? Is it two photons? (or: Double Quantum Effect.)

Kroll:

No, one photon in, two out.

Bromberg:

I think — I'd like to know what kind of work you did, but also a little bit what the atmosphere was like at the time. That must have been '46, '47, just drawing up to the Shelter Island Conference, where Lamb and Retherford (?) are working together, and you must have been in pretty close contact with them.

Kroll:

Yes.

Bromberg:

OK.

Kroll:

Well, there are a number of questions. Do you want me to —

Bromberg:

— I'd like to know, as you started to learn field theory, what kinds of points of view you were working with, whose work was the most important that you were following at that point, and what you were discussing with Lamb and Retherford, what kinds of reactions people had to the work you were doing, that kind of thing. At least from your point of view.

Kroll:

All right, I'll tell it entirely from my point of view.

Bromberg:

Yes, in fact, it really needs — I mean, if it's from somebody else's point of view, we'll ask somebody else.

Kroll:

Right. Right. Well, so, I'll repeat some of the things which are already down there. As I had said, I had done all this magnetron work, and it clearly was work that could be used for a PhD dissertation, but the truth is, I knew absolutely no current physics, since, as you commented, it was all classical electromagnetic theory. I had had none of the kind of quantum background that people normally get as an undergraduate today, nor had I taken a course in atomic physics, nor had I taken a course, ever taken a course in nuclear physics. While I was working at the Radiation Lab, I did take the standard graduate course in quantum mechanics, and one of the things we had to do was write a term paper, and we really — it was not that good a course in quantum mechanics, and we really didn't do the quantum theory of radiation either, so I did my term paper on the quantum theory of radiation.

Bromberg:

Do you happen to remember what text you were using at that point?

Kroll:

I think we used elementary things like Rojansky.

Bromberg:

It wouldn't have been as elementary as something like Richtmyer and Kennard, though?

Kroll:

I think Rojansky is more elementary than that. But the [???] shift didn't exist, for example. There may have been — there were more books than Rojansky. There was Ruark and Urey. But I'm not sure that was quantum mechanics. The book I really remember is Rojansky, but I remember there was also a McGraw Hill book and I just don't remember which one it was. I just remember the amber color of the cover.

Bromberg:

Dull green khaki?

Kroll:

Yes. And the truth is, since I was very distracted by my work in the Radiation Lab, though I was only supposed to be working half-time, and other things that distract a young man at that time in his life, I didn't work very hard in the course, and so I didn't learn that much quantum mechanics even then.

Bromberg:

Rabi is the person who seems to have intervened at this point in the effort to get you to do more modern physics.

Kroll:

Well, he wanted me to do a dissertation on something besides this magnetron work, and suggested I work with Halpern. So I had, in addition to the quantum mechanics course, I had read Fermi's articles, and in my paper — I mean the famous Fermi article on quantum field theory that I read in preparation for writing this term paper — and Fermi did everything in a box, quantization in a box. It was obvious to me the box had nothing to do with it, so I tried to do it without the box, and with continuous distributions rather than discrete distributions, and — well, I thought I made some good progress on the problem, although I obviously didn't solve it completely. I didn't really have the mathematical tools at the time.

Bromberg:

You know, I think I had a lapse — that's your term paper we're talking about ?

Kroll:

Yes, that's right. It's only relevant because it is a connection to the quantum theory of radiation.

Bromberg:

Who was teaching that?

Kroll:

Joe Keller. I think he subsequently went to Iowa State. I have a feeling he's not living any more. There's another Joe Keller in mathematics whom I also know, but it's not the same one. So I went to — I was still very, let's say, green in terms of knowledge about what had been done in the field, what had been done outside — the Fermi articles, after all, were very old, and then there was a book by Heitler called THE QUANTUM THEORY OF RADIATION which I also had, or at least acquired for the purposes of working with Halpern, and so, Halpern said that he didn't believe a statement in Heitler's book, to the effect that higher order processes in quantum electrodynamics were really negligible, and that's why he asked me to do the double quantum effect. He thought that the argument that Heitler gave, which said that the double quantum effect would be negligible compared to the single one was not a correct argument. So I just sat down and began calculating it. I didn't know anything. I got long complicated formulas, and didn't know quite what to do with them, so I looked at a special case. And some special symmetry which made the calculation much simpler, and it had a big effect. I remember telling Halpern about it. He said, "I knew it, I knew it!" He was gleeful.

Bromberg:

Was he right?

Kroll:

Well, to continue the story. So I shared an office with Lamb in the Radiation Lab. He had of course another office too, but he was in the Radiation Lab most of the time. My only office was in the Radiation Lab. And while I was doing this, he was in fact conceiving of and developing the apparatus for the Lamb Shift measurement.

Bromberg:

Is it fair, since you were both in the same room, to think of you as talking together and exchanging information on what you were currently doing?

Kroll:

Some, yes. I would say neither of us were very garrulous types, so — but we did talk.

Bromberg:

So you were really in on what was going on.

Kroll:

Yes. I was very much aware of the planning of the Lamb-Retherford experiment. There were these Quarterly Progress Reports, and maybe by this time I was getting them. Whether I was or wasn't, I remember reading his proposal for this experiment, and being very impressed by it. I was very impressed by the ingenuity. Again, it took me a long time to learn what physics was about, and as I mentioned earlier, I think it was well after I got my degree that I really began to understand what it was about, and it hadn't occurred to me that one ought to doubt the predictions of the Dirac Theory. Of course, as soon as I read that, it seemed very clear to me, even very important, and I thought the method that he proposed to do it was very ingenious, and of course it involved a lot of physics I didn't know about, like his use of the electron junction off tungsten as a detector, something totally new to me. I didn't know he knew so much. If I'm painting a picture of a naive kid, that's what I was. So yes, I was very aware of it. I watched the apparatus getting built, and the initial troubles in making it work, also the initial measurements with which he got this big shift. Since the effect was so large, and much larger than anybody expected, it was very obvious that something correct and very important had been done. In fact, I remember Norman Ramsey coming into our office, and looking at Lamb's first curves, and Norman looking at Willis and saying, "Well, let me congratulate you on your Nobel Prize." So the fact that he would eventually get the Nobel Prize for it was anticipated very, very early. At the time of the very first measurements.

Bromberg:

Now, you were going to tell me about this calculation.

Kroll:

With Halpern, yes.

Bromberg:

And whether or not —

Kroll:

You distracted me.

Bromberg:

I'm sorry. I think I'm allowed to distract you a little bit.

Kroll:

I think you're supposed to. I think you're supposed to. But do you want to return to that?

Bromberg:

I want to return to that, because for one thing, I'm just curious as to whether this big high order effect turned out to be correct.

Kroll:

I was going to tell you what happened. So then — so I told Lamb what I got. And he was surprised, and he then asked me a couple of questions about it, and it just — I think that was always characteristic of me, it took very — a very tiny push to make me look at things in a different way. Here what I had calculated was perfectly correct, but the special case that I had chosen was not representative, but very special, and in fact, it had a very negligible effect upon the actual cross-section, and in fact, it's possible that as a result of that, I — that helped me look at the situation in a much more general way, and — that is, I was no longer so defeated by this complicated expression. I learned how to look at it and see what it was really telling me, and it was then that I suddenly recognized that it was a case of the infrared catastrophe and I told that to Lamb, and that he believed, and he thought that was an important result, that indeed you did get an infra-red catastrophe from the two-fold Compton (quantum? ) effect, which clearly was not generally known at the time. Well, Nordsieck was also at the laboratory, and he was one of the authorities on the infra-red catastrophe. There's the famous Bloch-Nordsieck theory of the infra-red catastrophe. So I told Nordsieck about it also, and he said, "Of course you get an infra-red catastrophe. "That is, he thought it was clear that you would, and so that was fine, and then there was the question of what to do about it. That is, since the essence of the infra-red catastrophe is that although you appear to be getting a very large effect, you really in practice don't get a large effect, because it's always cancelled out by the virtual effects which reduce the single quantum effect. What happens is the following. I'll explain it to you.

Bromberg:

The infra-red catastrophe just means that your low frequency things add up to being infinite?

Kroll:

That's correct, but what is the physical significance of that? This is really —

Bromberg:

What I'd like to do is to get some sense of how your results actually were. I mean, you got the result that there was an infra-red catastrophe, but Nordsieck said, "Sure. " What really eventuated from all this? Did it turn out that these results, that you didn't come up with anything new in this calculation?

Kroll:

There was another paper at the same time by a man named Eli [???] who — which I didn't know about at the time, it may even have come out after I started to work on it — which I felt had gotten the results that I had gotten, namely, that there was an infra-red catastrophe in the double Compton effect, but to that extent it was a new one. It was something that other people were doing at exactly the same time.

Bromberg:

So Nordsieck was just being sort of smart aleck at that point.

Kroll:

No, I think it's often true that things which are new and which have not been said before, when somebody says them to you, you say, "Yes, that's very plausible." You say, "Oh yes." If anybody had asked me, that's what I would have said, but nobody had asked me, so to speak. Physics is full of things like that. It's just that he didn't have to be convinced. I think it's true, if anyone had asked him, would there be an infra-red catastrophe in the double Compton effect, he would have thought about it and rather quickly would have said yes, because the infra-red catastrophe was something he knew about. But in my case, it was something I learned about as a result of doing this work. I didn't know about it at all before I did it. I had heard of it, but it wasn't something I knew about. I mean, the words infra-red catastrophe were words that I had heard, and it became clear to me that I had one, and then I did read about it. I read the Bloch -Ramsay paper, and in fact I did a Bloch-Nordsieck light calculation for the double Compton effect. And — well, if you read the joint letter which Halpern and I wrote, Halpern is still talking in terms of cutoff, and that's a way of dealing with the intra-red catastrophe. Anyway, I didn't feel that the cutoff was the right way to do it. If you meant an experimental cutoff, that would have made sense, but I didn't understand that that's what he had in mind, if it was. I didn't feel that he understood the situation very well, actually. That could be wrong. I've learned to be a little bit more mild than I was then. But —

Bromberg:

Now, this is something I don't know much about, but this was really the very heart of the kinds of problems people were considering at that time, wasn't it? How to handle infinities?

Kroll:

Well, yes. In fact, the difficulty with the cure for the infra-red catastrophe was that the virtual effects which you subtracted, which you combined with the — let's say, the virtual effects on the ordinary Compton effect which you would combine with the double Compton effect to get a finite result, would only work in the infra-red by bringing in the virtual effects you —always more than the ultraviolet catastrophe with them also. And yes, that's why, that's the origin of the remark that that was in fact a very good problem to be working on at that time. And if I had been smarter and better, I would have recognized that this was a great opportunity to try to understand the ultraviolet catastrophe. But I knew that people had been working on the ultraviolet catastrophe for 20 years by that time, maybe 10 years but in any event, I didn't think I should expect to solve that problem as part of my PhD dissertation. I thought that was more than — I wasn't ambitious enough, to be perfectly frank. So I decided I did not want to try to write up a paper. I didn't see how I could, in a reasonable length of time, write a paper which I would like on the subject, which I felt would have something new on it. The fact that I had done this work independently of Eliezer was not a sufficient reason for me to try to use it as a dissertation. Things are very different nowadays, I should say. There are so many people working now that many things are done semi-independently in different places, and everybody gets credit for them. People use them for dissertations. But I was very priority -sensitive at that time, and I was very unwilling to steal ideas from people. So anyhow, I decided not to use it.

Bromberg:

Then what happened? I mean, is this already about the time of the Shelter Island Conference? Is this already about the time that the Lamb Shift is coming out?

Kroll:

It's all going on at the same time. I mean, I was working on this dissertation with Halpern, and —

Bromberg:

— in the summer of '47?

Kroll:

No, we're in '46. And Lamb is working on the Lamb Shift, and Nathan Nelson is measuring hyperfine structure, and measures the Lamb Shift, and then there was a paper by (Gregory) Breit. Or not a paper, I don't know whether he wrote a paper or not, but Breit thought he had an explanation of both the Nathan Nelson effect, as I recall, and the Lamb Shift, and he suggested that the magnetic moment of the electron might be — that the nominal magnetic moment of the electron might be playing a central role there. I think, in fact, he made a mistake. I think he — my recollection is that he estimated, he sort of took the Nathan Nelson result and tried to relate it to the Lamb Shift, and he got a much bigger effect than he should have. However, the question of, does the electron have half over two pis its magnetic moment was a very good question to ask at that time, and I think as a result of Breit's bringing it up, the Kusch-Foley experiment got started.

Bromberg:

And was this something that you were personally —

Kroll:

No, I knew all this was happening, is all. I was just in the room.

Bromberg:

OK, good. Breit wasn't at Columbia, was he?

Kroll:

No, but physicists talk. I don't know when Lamb's measurements first became known.

Bromberg:

I was thinking that was the summer of '47. I don't know, I was thinking about from reading papers, I had the general impression —

Kroll:

Let me see, the double Compton effect — in the fall of '46, and the Eliezer paper was published in '46, I don't know whether it was published before or after I started working on it. As I told you, I'm not a good reader. I in fact think it was Lamb who called the paper to my attention, after I had told him about the infra-red catastrophe results, and so that happened some time, I would say, in the late fall or early winter of '46. Then — well, all during that year, these various other experiments were being done. I noticed that the Letter with Halpern was submitted in May of'47. And Lamb's first paper was published in the fall of ‘47. He wrote in the summer, it's probably the summer — his paper was probably published in the summer. Do you have the date?

Bromberg:

No, I did not make a note of the date.

Kroll:

Then it was probably summer.

Bromberg:

I think he already in that paper mentions the Shelter Island Conference, so it must be after, and that was about some time during the summer, maybe August, '47.

Kroll:

When was the Shelter Island Conference?

Bromberg:

Well, it was in the summer, but I believe maybe it was earlier than that. Because the place that they used for the conference was a resort, and it might have been that they used it in the early spring before people took it over. Maybe they were using it before the vacation crowd came. So in other words, I don't really know. Some time in the summer of '47.

Kroll:

Let me check on the date of Bethe's paper. I think what we're probably heading towards is, how I got involved in calculating the Lamb Shift.

Bromberg:

That's certainly something we want to get into.

Kroll:

Anyway, it is my recollection that the Lamb Shift experiment was done in the '46-'47 period, and that I was, so to speak, one of the first to know about it. Then, I remember, Lamb went to the Shelter Island Conference, and on the way back from the conference, there's this famous story that (Hans) Bethe worked it out, and I remember seeing the Bethe paper, again.

Bromberg:

Probably even a Letter, he probably wrote before —

Kroll:

Oh, I think Bethe wrote what was basically his little PHYSICAL REVIEW paper. That was the first thing anybody saw. And I was very impressed by it. It didn't occur to me that I could have done it. Then, let's see, there's a reference here to a talk that Lamb gave at Brookhaven.

Bromberg:

Yes, a question as to whether you had a copy of that lecture.

Kroll:

I don't have a copy of it, but I do think I heard it. And I believe I got involved in the problem shortly thereafter. That was when I actually got involved.

Bromberg:

By the way, you mention as people you were talking with Halpern and Lamb and Nordsieck. Is there anyone else you had close conversation with on problems of this sort, or other kinds of close communication with ? This is still pre-Institute.

Kroll:

Yes. Well, '46 and '47 was when the Lamb Shift was — the experiment was initiated and it was measured. I don't at this moment remember exactly when the magnetic moment was measured. Do you know that? When that came out, or when the Nathan Nelson experiments came out? I have that Schwinger compendium of reprints, but I don't think it's in that. I don't think I have it here. I'm just trying to get some of the timing straight. But I'll tell you as I remember it right now, which may be wrong. That is that all those experiments, the Nathan Nelson experiment and the Lamb Shift experiment, were going on at the same time, and they both got strange results at the same time, except it wasn't clear what the Nathan Nelson experiment meant because the effect was very small. On the other hand, it was clear that the Lamb Shift result was very fundamental, because it was a big effect. And it was as a result of those, of the preliminary experimental results from those two experiments which suggested the anomalous magnetic moment measurements. I don't remember exactly when they came out, and I don't remember in particular whether they had a number before the alpha over two pi was predicted. They may very well not have. So that would make that quite a bit later. But all those questions could be answered, and probably are, in Schwinger.

Bromberg:

There's a problem asking questions for other people, you don't really know what —

Kroll:

— you can't fill me in, when I get stuck in this way.

Bromberg:

Well, what about the point of view you decided to take, and I was going to ask in particular how you related your work to Oppenheimer's calculations on infinite electron self-energy.

Kroll:

Well, as I think I have told you, it has always been my tendency to read as little as possible, to get into a problem and start working it out myself, and Lamb had already done some very important things in the methodological development that we used. I mean, he had the —

Bromberg:

— he studied with Oppenheimer, of course, so he would probably know Oppenheimer's work pretty well.

Kroll:

Yes. Well, just to set the scene, Bethe's calculation was non-relativistic. He conjectured that if one did it relativistically, that the result would turn out to be finite instead of divergent, and so, basically Lamb set out to confirm that, which I imagine Bethe may have done too, I don't know. I don't know what his role was in doing the relativistic theory, since I don't think his name is attached to any relativistic calculations of the Lamb Shift. But, so, I mean, Lamb was one of the people trying to do it. There were others like Schwinger and (Richard) Feynman. But in any event, Lamb had a method which he though was viable, which is what he talked about at this Brookhaven lecture. Now, what I think I had been doing that summer was continuing to work on the double Compton effect, because the paper, the Letter that I submitted with Halpern had been submitted in late May, and it seems to be quite reasonable that I would have continued to work on it over the summer, and when Lamb suggested that I join him on this Lamb Shift calculation, that seemed to me to be an incomparably more interesting thing to do, and I don't know quite when I had decided to use they money (?) for work on my dissertation, but at some time probably during that summer. And one can ask why it took me a year to get my degree after I made that decision, in view of the fact that the paper was already published. The answer is that at Columbia you had to take final examinations. I had to study for those. Then you have to arrange for a defense of your dissertation, and all of that occupied that last academic year at Columbia, which was '47-'48, but what I did physics-wise during that period was, apart from the details of getting my PhD, was work on the Lamb Shift calculations.

Bromberg:

How did you and Lamb divide up that work?

Kroll:

Well, the key problem — there were sort of two key problems. One is that there were a lot of integrals to do, and we decided we would do them all individually, to decide what the integrals are.

Bromberg:

That is, you'd both do them and then compare.

Kroll:

That's right. So there was that division of labor. There were some methodological issues. One had to do with how to do the vacuum polarization calculation, and that was problem, because we were using a very different approach from the old one, and I did make a methodological contribution to that.

Bromberg:

Why did you decide to use a different approach?

Kroll:

Well, because the whole calculation was being done by a different approach, and we wanted to use a consistent treatment. I mean, it wasn't obvious, as it would be using Feynman diagrams, that the vacuum polarization and the fluctuation were really completely separable. The way we were doing it, they didn't look that separable. They looked like they were all part of the same thing. And although the Grueling and I guess Serber calculations had been done many years before, on the vacuum polarization, and in fact Lamb had believed when he did the shift measurement that that's what he was going to find, just the Ewing effect, just 27 megacycles instead of a thousand. So the experiment was sort of designed to be able to detect 27 megacycles.

Bromberg:

I see. That's interesting.

Kroll:

Well, that's, in a way, what he looked for originally was what I would call the high frequency.... (off tape)

Kroll:

Yes, I think I was going to say, he was measuring, his first measurement was of the high frequency transition, which was around 13,000 megacycles. That was from that state to the P3 half state, and since we're looking for a split between the P half and the S half, since they were supposed to be degenerate, one wouldn't have had any idea what frequency to look at. But then, when the measurement was made, I mean, the discrepancy between the measured interval and the predicted interval was quite large, although one didn't have that direct a measurement of the separation between the two P levels. One just had the old spectroscopic ones. The S to P separation was a thousand megacycles less than the theory. But that told you what frequency to look at, for the direct P to S, P to half S transition, so then the experiment was redesigned to measure that also, but it may well be that the first measurements reported were just of the transitions through the 3 level (three halves level); that could well be the case.

Bromberg:

Now, we were talking about the novelties in your calculations with Lamb, and also about the way the work was divided, and we were talking about the methodological problems. Did we say what is relevant to say about that whole calculation? What was new?

Kroll:

Well, the method was quite unique, actually. We were the only ones who used that method. It was a method, used the old fashioned field theory method.? There was the problem of what to do about the infinities, and we looked, we tried that in two different ways. One was simply by direct subtraction, since the S half of the three half levels was exactly degenerate, we simply calculated the difference in their electromagnetic energy levels, in the hopes that the infinity would be the same for both of them, and so that was, I would say, the basic method that we used. The difference was finite. Maybe in principle the difference could have turned out to be infinite. And one was certainly not guaranteed that that kind of subtraction would work, and in fact it would have turned out to be infinite if it were not the case that the expectation value of kinetic energy is the same for both states. They could have had the same energy, and the kinetic and potential energies could have been different for the two states, in which case that method would not have worked, but we knew that the kinetic energies were the same. We also knew that the infinity was supposed to be a mass renormalization effect, so we expected the difference to cancel, and it did. And that gave us an unambiguous result which proved to be the right one. We tried to look at it another way, as well, which was to calculate various operators, and analyze the coefficients of those operators. Basically we were trying to see what ambiguities there could be in the calculation. And that part was pretty much my contribution. That was one of the things I added to it. I won't say it had any long term impact, because ultimately the covariance methods turned out to be a better way of doing these things. I also thought of the Feynmann regulator method during that time, and even had tried it. It wasn't quite the regulator method. I mean, that's part of the large regulator method but — that's the Pauli-Villars regulator method but this — they did the regulator method.

Bromberg:

That's the Pauli?

Kroll:

Pauli-Villars regulator method. But there was a Feynmann precursor of that. That was also something that I had thought of and tried. I never mentioned it to anybody until after somebody else did it, because it seemed to me so illogical and so arbitrary, so — anyway, that's an aside.

Bromberg:

And that was part of the research, but not part of the publication.

Kroll:

That's right, part of my doodling.

Bromberg:

Now, at the end of that paper, you said in here I think that you heard the summer series of Schwinger lectures. Do I remember that correctly?

Kroll:

Yes, that's right.

Bromberg:

What impact did that have on the way you were thinking at that point, or how did that interact with your work ?

Kroll:

Well, let me say a little bit more about that last year at Columbia. Another thing that I mentioned was the fact that during that year, Oppenheimer had come to the Institute for Advanced Study to be its director, and that brought some recent young PhDs — and it was decided to have a series of joint Columbia - Institute for Advanced Study seminars, and I believe we may have had one at Columbia. After that — I'm not sure we had even one at Columbia, but the original idea, I believe, was alternate weeks, but the facts were, it was every week at Princeton. So I got in the habit of going down, I think weekly, always Lamb I think and sometimes other people as well, sometimes Rabi, sometimes Foley would go. I remember some of them, but not very many, I'm sorry to say, but I used to go down quite regularly, I think, and I got to know the people there. I got to know Oppenheimer. I got to know Sol Epstein and Hal Lewis who were there at that time. I think Leslie Foley was also there at that time. And there was an important piece of work done during that period, that was related to this overall problem, by Lewis and Epstein. I don't think they published jointly on it, but I think it was on the same subject. And it had to do with the resolution of a mystery associated with an old paper by Sid Dancoff. Is that something you remember from Schraeber's(?) work?

Bromberg:

No. In fact, I would like to know, before you get onto the actual subject, more of who was there. Who came to those seminars? Everybody who was at the Institute, more or less? I mean, if we read the Institute rolls, we would know who was there ?

Kroll:

Well, I mentioned a few people. It was, I think, the first year Oppenheimer was there, so it wasn't a very big establishment.

Bromberg:

Was Karplos(?) already there?

Kroll:

No. This was before either of us had gotten our degrees. I'm talking about the year before, when I was just commuting to the Institute. And no, I think the three — Oppenheimer's students, so to speak, who were there, were those. I don't think there were any others. I imagine that graduate students from Princeton and faculty from Princeton also went. I didn't really know those people. I don't remember them anymore. And (Abraham) Pais was probably there also, but again, I am not sure he was there then. I do not quite know. I believe Pais was there then, because Pais had also been doing work on the self-energy problem. I remember hearing him talk about it.

Bromberg:

I guess the thing I'd like to know most here is really a very personal point of view. What interactions or what ideas had most effect upon you, because let's presume that Sam knows or has talked to these other guys, and I think we can presume that, he's a terrific scholar, and what he really wants to know is the effect on your work and your thinking, to the extent you remember them.

Kroll:

Well, I thought the Lewis-Epstein result was important, because it cleared up an old mystery, which was why — I had talked about infra-red catastrophe and Brumstalling and trying to save that by calculating the radiative corrections (or rate of corrections ?) through elastic scattering. That had actually been tried and it hadn't worked, because the infinity remained and I even think the idea of interpreting it as a mass had occurred to Dancoff and it didn't come out that way. And the truth of the matter is, he just made a mistake. I think basically Lewis found a mistake, and I think he left out one set of transitions in the perturbation theory. In those days you did perturbation theory by doing what were called transition schemes instead of Feynmann diagrams, and it wasn't that easy to — you could easily leave one out, and he left something out, and when one put that in, it indeed turned out that that one could interpret the infinity as a mass renormalization. That was important because, you know, one wanted the theory to be consistent, and it was fine that you could fix up the Lamb Shift by introducing the idea of mass renormalization. You could also fix it up as we did. One really didn't have to think of mass renormalization to just subtract the self-energies of two degenerate states, finding that the difference was finite. Also, there was the Luttinger calculation of the anomalous moment which was in a way based on the same idea, in which —

Bromberg:

That was happening during that year?

Kroll:

It happened slightly later, but the ground state of an electron in a uniform magnetic field is independent of the magnetic field strength, and that's due to a cancellation between the orbital angular momentum and the spin angular momentum. But that's true only if the electron has no anomalous moment. But because that's true, it turned out that the radiative correction to the energy of that state was also finite, because the energy being zero was independent of the mass of the particle. The orbital and spin moment don't exactly cancel, when the radiative corrections are taken into account, so that automatically exhibited the finite anomalous moment effect. So mass renormalization didn't have to be the correct explanation of the infinities, even though that's what — the words one liked to use. And the fact that the Dancoff calculation had been inconsistent with the idea of mass renormalization therefore was a real blemish which was removed by this work of Epstein and Lewis.

Bromberg:

What made you decide to go to the Institute? Was this coming out of this year of interaction down there?

Kroll:

Oppenheimer asked me.

Bromberg:

And what made him ask you?

Kroll:

I guess he saw me coming down, and he must have said, "Who's that guy?" And somebody must have said, "It's somebody you should ask to come down." I mean, it was known that I was calculating the Lamb Shift, and so he had invited me down and was prepared to support me, although I didn't realize that, and in the meantime I had applied for a National Research Council Fellowship, which I got, so that's how I supported myself at the Institute. The whole business of postdoctoral appointments was something I didn't know about, and I think probably had not existed before the war. I mean, there had been only things like these National Research Council Fellowships. So I didn't imagine it was going to be any different after the war, and so I applied for one of those and got it. It was very natural to go to the Institute. It was a famous place where Einstein was, and supposed to be where the best theoretical physicists were, as far as I knew.

Bromberg:

Now, the Schwinger lectures, is that something that we should return to ?

Kroll:

It is something that we should return to. It was of course known, all the time we were doing this work, that we had serious competition, and in fact, Schwinger was certainly the first person to get the α over 2π, and some-time during that year, when we were calculating the Lamb Shift, we learned that Schwinger had calculated the anomalous moment of the α over 2π, and as I said, I don't know whether the measurement preceded the — came before Schwinger had that result, or after. I know the measurements were initiated before he had the result. I don't know the rest. We could easily find that out, but I don't know at the moment. It was known that he was — that he believed that the right way to do mass renormalization was, to take care of these infinities, was by developing quantum electrodynamics in a covariant way, and so I mean, we were aware of the fact that he was doing that, and at a certain point, — there was another conference. Wasn't there something called the — do you remember? There was a Shelter Island Conference, then there was a Pocono Conference, and there was another one whose name I don't remember.

Bromberg:

I don't remember either. There were three.

Kroll:

And it was at the second conference, which I believe occurred during this '47-'48 academic year, — I think it was the Pocono Conference, I think it occurred in the spring, there may have been two Pocono Conferences — anyway, I remember Lamb coming back from that conference and saying —

Bromberg:

You weren't at that one either?

Kroll:

No, I was never invited to any of them, as a matter of fact. I think of the younger people, the only one who got invited was Dyson, who got invited to the third one. I don't think any of us lowly characters got invited to the others. Anyhow, when Lamb came back from the Pocono Conference, he had presented the results that we had by that time, and we had more or less finished the confirmation. We hadn't ironed out all the theoretical points. But I think he presented what we had done, and Schwinger presented what he had done, and Feynmann presented what he had done, and that conference Feynmann remembers with some bitterness, as you've probably heard, since the person who got all of the — first of all, the person who got all the time, I think Schwinger got six hours to present what he did, and it was sort of Oppenheimer's reaction that Schwinger had done it, and everybody else should sort of give up and learn what Schwinger had done. They listened to Feynmann but there was no real reaction to what he presented.

Bromberg:

So Lamb is telling you all this, I assume, when he comes back.

Kroll:

What Lamb told me was that Schwinger had a completely Covariant theory, and that he was evidently very impressed by it. He discussed with me some of the points that had come up, with reference to our work. Schwinger had a very awesome reputation at that time. I'd been hearing about Schwinger ever since I'd become a graduate student, because he was one of Rabi's glories, having discovered Schwinger, and then of course Schwinger was also doing electromagnetic theory calculations at MIT, and I knew about that, and had even seen some of the work. So it was kind of recognized that he was one of the leading theorists of the day, and I wasn't sufficiently vain or bold to think I should be competing with him. At least not then! But anyway, the connection with the Ann Arbor study was that that was the place to go to learn what Schwinger had done, so I went with that in mind.

Bromberg:

And how did you find it? What was your reaction?

Kroll:

Well, you know, Schwinger is an extremely polished lecturer, and the theory had the great formal elements. I was very impressed with it. I hadn't heard anything about Feynmann's work, although I'd heard stories about it. I worked with his diagrams. I had heard Feynmann talk about action at a distance theory when I was at the Institute, and he was a graduate student or an assistant professor at Princeton at the time, and — or maybe neither of those things, but he did give a talk at the Princeton Colloquium when I was at the Institute and attended Princeton Colloquia.

Bromberg:

So you heard Feynmann really after the Schwinger lectures? Coming later? Well, if the lectures were that summer, maybe I've got my chronology wrong.

Kroll:

Well, I may have mine wrong. Let me see. I believe that the Schwinger lectures were the summer of '48.

Bromberg:

Which is the summer before you go to the Institute.

Kroll:

To the Institute, and the reason I think that, it has to be true, because that famous bus ride in which Dyson worked out the relationship between Schwinger and Feynmann took place — Dyson was also at Ann Arbor — took place after the Ann Arbor lectures. In fact, what happened after the Ann Arbor lectures was that I went to visit my family in Texas, with my — I guess my wife, I was going to say newly married, but we were married three years by then, it wasn't so new. We still hadn't finished the Lamb Shift paper, and while Dyson was using his time usefully to work out the connection between the Schwinger and Feynmann theories — I really couldn't have done that, because I didn't know the Feynmann theory, though in principle one could have adopted the approach which Dyson used to do it quite differently, but anyway, what I was occupied doing was finishing that paper. I went home to Texas. My father had given me a car as a PhD present, and Sally and I drove through the West to visit some areas of the West I had never visited before, and she hadn't done any traveling at all, and I turned the car over in Wyoming, and — my brand new car! And I was stuck in Rollins, Wyoming, for several days while the car was being fixed so it could be drive, and that is where I finished the paper.

Bromberg:

Were there people you met at that Institute, besides listening to Schwinger and liking his lectures, was there anybody there who had an important interaction with your own work? Was Dyson somebody you were listening to a lot at this point?

Kroll:

When I went to Ann Arbor, I don't think I knew anybody. I don't even known if I knew Dyson.

Bromberg:

Didn't you meet anyone ?

Kroll:

I met Sy Leventhal(?). Used to play tennis with him. I think it was Sy Leventhal. I met Ed Lennox, who also went into biology. (crosstalk)

Bromberg:

— your tennis game — I'm primarily interested in people who went into physics or —

Kroll:

Then I heard a lecture by Frank Yang, and by a man named Falkov, both talking about angular correlations. They were also sort of in the graduate student stage.

Bromberg:

If I'm bringing up things that are unimportant, tell me and we'll go on to something else. You know, I don't have to tell you this actually, but what we're looking for is anything that we can't find in the published material, anything that isn't easy to find but that is in a person's memory and stays there and is not going to be found by any other kind of documentation, so I think that these questions all have to do with the way you approach field theory, and anything that influenced your approach is really what I mean.

Kroll:

Yes. Well, the answer is that I do not remember interacting very strongly with people vis-a-vis Schwinger's lectures. That doesn't mean I didn't. I really don't remember. And I may very well have met and talked to Dyson while he was at Ann Arbor, but I don't have any very clear recollection of it.

Bromberg:

Apparently he made a big stir the next year, when you were at the Institute.

Kroll:

Oh, that is the famous bus ride.

Bromberg:

I see.

Kroll:

When he went West on a bus, and began thinking about what he had heard, but Dyson had gotten his degree at Cornell, and had certainly been exposed to all of Feynmann's ideas and Feynmann's methods, and having heard Schwinger, he began thinking about the relationship between them. Feynmann's original approach — Feynmann has always said that at the time he did that work, he didn't know quantum electrodynamics, didn't understand it, and he had obtained (retained?) his methods from his space-time action at a distance approach to field theory, and so the two looked very different, but the thing which Dyson knew, which none of the others knew, was what Feynmann's formulas looked like, and Schwinger's formulas didn't look that different, is the point, and I think a person seeing both those formulas would naturally think that there was a way of relating them. So as I say, on this bus ride, he figured out just what the connection was, and I think as soon as I arrived at the Institute, I heard, or as soon as I saw Lamb again, I saw Lamb shortly after I arrived at the Institute, because we still hadn't submitted our paper and I think I brought him the results of my final work on the subject, which took place in Rollins, Wyoming —

Bromberg:

— Wyoming on the way to Texas?

Kroll:

No, it was when I was going — you see, I went to Texas, and then from Texas I was driving back to the Institute, and I was visiting some national parks, and I wanted to visit the Grand Tetons, and I was on my way to the Grand Tetons when I turned the car over. Again, after my finishing the paper, we proceeded to the Grand Tetons and enjoyed the rest of the trip. When I arrived at Princeton, everybody asked me if a bear had attacked the car and made it look like it looked. Since all I did was make it drivable. So anyway — I heard from Lamb about Dyson's bus ride, and how he related the two theories, and so that was almost the first thing I did when I got to the Institute. But I wrote this little paper on the Covariant Hamiltonian formulation of field theory, which is in my publication list, a little abstract. I did that. That was a direct outgrowth of the — what I had heard at Ann Arbor. And then, as I said, I heard about Dyson's work. He had written up the first paper, which I read. He may even have given, I think he even — he certainly gave a lecture on it. There is a question, in those questions, having to do with Oppenheimer. Oppenheimer, Bethe and Dyson. I was in Oppenheimer's office when all that happened.

Bromberg:

Oppenheimer, Dyson and Bethe.

Kroll:

Yes. It's question number 13, here.

Bromberg:

Right.

Kroll:

I don't remember Dyson lectures, although there may have been Dyson lectures. What I do remember is a meeting in Oppenheimer's office, in which Dyson explained what he had done in the presence of Bethe and Oppenheimer. I had I guess read, maybe from Dyson, accounts of the very hostile reception which Dyson had gotten from Oppenheimer.

Bromberg:

Did you observe that yourself?

Kroll:

I did not observe it. I would say, what I observed was great open-mindedness. I was one of the very few people who was asked to be in that room, so I guess I must have — anyway, I was there. Dyson presented what he had done, which I found very clear, but I didn't quite know that, I thought what Dyson had done was clear and that it led to a computational scheme which was much simpler than Schwinger's, since it combined many different kinds of terms into a single expression. But as he was going through it, Bethe said, "Oh, and that's where the Feynmann diagrams come from.” So Bethe saw as he derived his formulas involving the Feynmann-Green function that indeed, what Dyson was presenting was in fact going to lead directly to the Feynmann diagrams, and then I think Bethe may have gotten up and made a few remarks about that, showing, amplifying that a little bit, and Oppenheimer was very receptive. I didn't see any hostility at all, and it seemed to me that at the end of that meeting, that Oppenheimer was fully convinced.

Bromberg:

And you didn't see any great role of Bethe as the catalyst? Because that's the other thing that is —

Kroll:

Well, all I saw was that one meeting. In a sense Bethe was a catalyst, because Bethe approved and Bethe knew — again, I don't know how well Bethe knew what Feynmann had done. Feynmann had — there's a natural tendency to be skeptical of sort of very unconventional sort of ad hoc theories, in which you don't maybe quite understand what the basic premise was, and the derivation of a complicated formula is a matter of what at the outset appeared to be a non-quantum approach to the problem. So I could easily imagine that there was a lot of skepticism when Feynmann presented his work at the Poconos, and there are all complicated formulas. It's very easy to not understand them, to not remember what you heard, and Feynmann didn't write any of them, so I have no reason to think that Oppenheimer had a very clear picture of what Feynmann was really doing. He may have. I don't know. But Bethe of course knew what Feynmann was doing very well, again because they were at Cornell together, and so when Dyson presented this work, Bethe recognized the connection, and I remember Bethe saying, "I think this is just wonderful." That was Bethe's comment. Certainly Oppenheimer was sold after that. That he had been unsold before that is something I was not aware of. I do think very early on that Oppenheimer handed me Dyson's paper and told me I should read it. It was clear that he certainly thought it was worth reading very early on. Now, there was a second Dyson paper, in which he showed how to do higher order renormalizations. Oppenheimer was extremely positive about that paper, and he also handed me that paper to read, and I may be mixing up those two things a little bit, in my memory, because the second paper went beyond what anybody else had done.

Bromberg:

Well, shall we stop here for the moment?

Kroll:

Sure...well, like I was around when the Lamb Shift experiment was being thought of and put together and all of that, I was also around when Townes thought of the maser, and all that sort of thing, so I was very aware of all that that was going on. Everybody was rather closely coupled and all these ideas got discussed quite extensively.

Bromberg:

Townes said that there was a lot of skepticism about whether he could get the narrow line he thought he could. Do you recall anything like that? That people thought that the uncertainty principle wouldn't allow —

Kroll:

I don't remember being skeptical.

Bromberg:

And in fact, it seems to me that Lamb was one of the people who was surprised that the maser worked. So that he might have been a skeptic of some sort.

Kroll:

I don't remember any very critical view of it. I thought that the separation scheme that he used, which I didn't understand at the time, of mono molecules, was very ingenious, and I thought the story was quite plausible. I would say the connection of masers with classical oscillators was something I was certainly not aware of, and I recognized that connection. I was sure it would work. And that's of course, ultimately that was one of the things Lamb did, was work on that connection, the classical theory of the laser.

Bromberg:

Of course you were also very much involved, or at least somewhat involved with trying to get short wavelengths, so that —

Kroll:

— yes. Yes, but there are things, I can think of something else that I didn't understand then, that I should have understood better. Among the — yes, that's right, we were all trying to get shorter wavelengths, and Townes had this idea of using molecular resonators, and that sounded hard to me, but I certainly thought it was a good idea, and you can say, why wasn't I doing that? It's a good question.

Bromberg:

No one was doing that, or practically no one.

Kroll:

But there's another thing Townes was doing, which was the Cerenkov oscillator. Now, the Cerenkov oscillator is the same as the free electron laser. And in fact, there's a whole program at Dartmouth now on Cerenkov oscillators. They work very well. But Townes never got it to work very well. However, had I understood free electron laser theory then as I came to understand it later, I would have understood that Cerenkov oscillator idea much better, and I might have even been able to have helped.

Bromberg:

Was Motz's — now, Motz was doing a little bit on that at this point. Was that something that people were looking at? He was talking about his undulatory (crosstalk) Cerenkov radiation.

Kroll:

Well, certainly the Motz-Nakamura paper is the free electron laser. There is no real difference. But I didn't know the Motz work. There was a man named Coleman at Illinois who was working on undulatory radiation, and I knew about him, but it's my recollection that there was nothing, that was not an oscillator, that that was just a Doppler shift radiation. But that could be wrong. I'm just telling you what I remember. Have you come across Coleman?

Bromberg:

It seems to me that I have come across review articles of his on the whole subject of producing radiation in the millimeter or sub-millimeter range.

Kroll:

Because I have not been able to find the references to him myself, because it's come up in other contexts, since I think he may have even visited the Radiation Lab once. That was the one name that I actually knew in this business, of what I would call undulator radiation.

Bromberg:

I think that the articles I have seen have been in things like TRANSACTIONS of the I Tripe E microwave theory technique.

Kroll:

Yes. Somehow his name never got to be very prominent in that business, but he is the first one, in fact the only one that I remember from that period. It was only when I really began looking back into it, after I had known independently — well, it wasn't just that it was independent. I did add something to what they had done. But it was only when I began looking into the antecedents of Madey's work that I became aware of Phillips' work. The name Motz was familiar. I don't remember quite, but maybe the name Motz was familiar because it was in the original Madey-Schepman and Fairbanks paper. There was a Madey, Fairbanks, Schwepman paper on the free electron laser, in which they referred to the Motz-Nakamura work, and in which they said that what they were doing was quite different and basically quantum, not classical.

Bromberg:

I actually haven't read much past the middle sixties. I have read, looked at some of Motz's papers without really looking at the Madey papers. At any rate, you were all in this business of going down to millimeter and sub-millimeter waves.

Kroll:

That's correct. That's right, and I was doing it conventionally, making shorter and shorter wavelength magnetrons, and we got down to 3 millimeters and that was as far as we got. But there were in fact other microwave approaches I did not pursue, which other people did develop subsequently. I had really left the field, you know, the —

Bromberg:

— yes, that's what I was going to ask you. There were a few papers coming out, but I didn't know exactly what your emotional investment was.

Kroll:

Well, what happened was, after I got into quantum electrodynamics, I decided that the only thing worth doing was very fundamental physics, and that I shouldn't spend any time on problems other than the most basic kinds of problems.

Bromberg:

Now, you told Finn that you were doing some Radiation Laboratory problems for summer salary.

Kroll:

That's right.

Bromberg:

But sometimes people found that an alternation between fundamental and applied science is useful to them in some way, because you'd done so much applied science, I just wondered whether —

Kroll:

No, I agree with that completely, but you're asking me what my attitude was about what I was doing. My attitude was that I should be spending all of my time on fundamental physics, and I didn't even consider — I didn't really consider that the applied work that I was doing was worthwhile, in a physics sense. It was just sort of applied utilitarian work.

Bromberg:

Bread and butter work.

Kroll:

Bread and butter work, and yes, I worked at the Radiation Lab for summer salary. I seem to remember having done that for a long time, but I carefully checked my records, and I decided that I didn't do it for a long time at all. I did it once. That's the summer of '51. After that, I began to go to Brookhaven, and the first Brookhaven paper, and then afterwards Columbia got a theoretical complex and I was able to go on that. The summer salary mode of operation had come into being, and I guess it had come into being just a little earlier than I realized, so I didn't take advantage of it right away, but I really just lost a year, I would say. However, it is the case that while I was at Columbia, half of my academic salary was paid by the Radiation Lab. I certainly still found that work interesting, so I did keep my hand in and I more or less was responsible for the magnetron work there that continued during the fifties.

Bromberg:

Now, you said that you were teaching in this field, and I should think that the magnetron work — I mean, you were teaching microwave electronics.

Kroll:

I think I only taught it twice, as I said. I may have taught it one other time.

Bromberg:

Was there an interplay between teaching and the magnetron work? Did you have graduate students?

Kroll:

I had a graduate student. I had a graduate student, two graduate students in fact. One was Walter Strauss, who — this, sort of late magnetron paper that you mentioned was with Walter Strauss. And another one was a man named Zed Frankel. Frankel was actually an electrical engineering student, who fits into your belief that it's really engineering. He did a dissertation in the Radiation Lab with me. But that it's really physics is shown by the fact Frankel eventually became a nuclear physicist, and is now dean of sciences at the Weizmann Institute.

Bromberg:

Well, it is a kind of hobby horse I ride, to try to understand how engineering and physics interpenetrated or failed to.

Kroll:

But to some extent, I didn't consider the development of short wavelength sources of radiation as normal an enterprise as solving the problems of elementary particle physics, and so my emotional orientation was entirely towards that, and I really wasn't looking for clever ideas for getting down to shorter wavelengths. Something that I could think of in an afternoon was fine, but it would have to be something that was — it had to sort of just happen. On the other hand, if I'd been thinking very hard about it, I still might not have invented the maser. There were plenty of people who were thinking very hard about it who didn't. I might not have understood how a Cerenkov oscillator really worked. It's a very simple classical way of seeing how they work which I now see very clearly, but I certainly didn't understand it then.

Bromberg:

Now, how did you get to this work with Lamb that we were talking about?

Kroll:

The story behind that is that Townes and Cederholm had all done this experiment in which they set a new limit on the possible size of either drift and that was published and that was that. I read the paper. I thought it was very nice. And of course I knew there was no ether drift, so I didn't consider it all that fundamental a question, but still, I understood that it's always nice to confirm things to another factor of 10 accuracy, and I thought that was very nice. At a certain point, Townes walks in and hands me a letter which Lamb has written to him, in which Lamb has claimed that the experiment does not have any bearing on the ether drift, it doesn't make any difference, if there was one you wouldn't find it by this experiment. He asked me what I thought of it. I think it was a matter of, you know, a day or two. I looked at it and I said, "He's wrong." And I worked out what is basically the theory in that letter which we wrote, and I wrote to Lamb about it, and he agreed and withdrew it, and that was that. That's all there was to it, but it was not something, you know, why publish it? Nobody had raised the question in the first place. And there was this character whose name I forget, who —

Bromberg:

He's mentioned in that article.

Kroll:

Yes, and in fact there's a long colloquy between us after the article, who wrote a letter to the IEEE, years later, in which he made more or less the same argument as Lamb had made, and so again when that letter appeared, Townes showed it to me and said, "Maybe we should publish that paper after all.” So I would say, really out of politeness and a kind of affection, since I would never have done it if Lamb hadn't done it wrong in the first place, I asked him if he wanted to join me in the publication. So we agreed to do it, so we published that Letter, after which Lamb was very sorry he published it, because he found the conflict and arguments with the electrical engineer very irritating and demeaning, and didn't like engaging in that kind of polemic or colloquy, whatever it was, in print, so he did write to me that he was sorry he ever listened to me or allowed himself to be associated with it.

Bromberg:

That's interesting.

Kroll:

That's the story.

Bromberg:

I should think you did him a favor by —

Kroll:

— he wasn't mad at me. I guess he was just irritated at the situation, and I think the guy never did surrender, but I don't think his views on the matter gained any acceptance. I don't know who the guy is. His address was Menlo Park. But the world is full of people who think they understand relativity.

Bromberg:

Yes. I think that my memory has finally gone, the last month or two, I just can't remember his name. Oh here, I made a note of it — K. Linehan.

Kroll:

K. Linehan, that's right. It was K. Linehan's letter which led us to publish.

Bromberg:

Well, then Lamb was already at Stanford by that time?

Kroll:

No, when he wrote that letter to Townes, he was already at Oxford. He was Professor of Physics at Oxford. He went from Columbia to Stanford, from Stanford to Oxford.

Bromberg:

So when he was doing his laser theory, he was already removed from you, I guess.

Kroll:

Yes. Lamb and I, our paths have crossed from time to time over the years. Whenever we meet, we always have a lot to talk about.

Bromberg:

Let's talk about the IBM consultancy.

Kroll:

All right.

Bromberg:

That's interesting. Now, many of the Columbia people of course got involved with IBM because it was right there. Is that pretty much the path you followed?

Kroll:

Well, before I consulted for IBM, I had consulted for an outfit called Anterex, and that was directly on magnetrons, and then that ended, because Anterex had financial problems and decided to reduce costs, and I was one of the ways that they reduced costs. I was approached by IBM, I suppose on somebody's recommendation, probably Townes but I'm not completely sure that that's how it happened. It could also have been Kusch. But somebody recommended that they approach me, so why not was my attitude. Certainly professors were not that well paid. I was quite willing to give it a whirl. And I had never heard of a parametric oscillator or frequency divider or any of those things before I went there. This von Neumann scheme was described to me, and I was told one needed a frequency divider, and I said, "What's that? I thought you needed harmonics." (?) So I was told what a frequency divider was and I said, "Aha," and then I was told that they could have two phases, and they wanted to switch between them. So that's again all it took. When I get pointed in the right direction, I sometimes move very quickly.

Bromberg:

Who were you involved with there, do you remember?

Kroll:

By Havens was the guy who gave me the job. Yes, Byfield Havens, his name was.

Bromberg:

Because the person first mentioned to me —

Kroll:

Or maybe it was Byron Havens. You said it was Rolf Landauer. Yes. I don't remember what the connection between Landauer and Havens was at the time. Havens was a fairly senior person, and Landauer was, I believe, not at Watson Labs. I think Landauer was perhaps at some other IBM place, Poughkeepsie, yes, eventually it was Yorktown Heights, but I think when they started, it wasn't Yorktown Heights.

Bromberg:

My memory is that Yorktown Heights came into being around '59.

Kroll:

Yes, that's my memory also, that it came into being while I was still consulting for IBM, but not when I started.

Bromberg:

OK, what did you work on with them?

Kroll:

Well, I think that paper that I wrote with Palocz on the fast switching frequency divider was my response to that question. The question was, how do you make a fast switching frequency divider? I naturally began applying what I knew. So I invented one, which was really based on the idea of a reflex klystron, and worked out the theory of its switching currents. I don't think that was ever done, because solid state devices were invented about the same time. I was a little too late.

Bromberg:

And this "Properties of Propagating Structures, with Variable Parameter Elements,” which by the way didn't really — is that ?

Kroll:

Which one is that?

Bromberg:

No. 26. Is that also an IBM?

Kroll:

Yes, that's an abstract, I think, actually I'm not sure what that is, yes, I guess it is, because this is a technical report. It's probably a summary of that technical report. Here it is.

Bromberg:

And you also did other papers for IBM.

Kroll:

One more.

Bromberg:

And that was the one that I, from Gelb(?) and McNichol —

Kroll:

Yes, that's right. They were trying to develop a device based upon that technical report that I just pointed you to. That technical report was never published. It should have been. It got a nasty review from a referee, and I didn't know in those days that if you got a nasty review from a referee, you fought it, and so I just didn't bother to try to get it published. If it happened to me now, it would have been published, I assure you!

Bromberg:

I see. It's part of the IBM —

Kroll:

— had the slightest problem in getting published, and so I got these nasty remarks, I — again, it was part of this whole attitude I had, that I didn't think that this kind of work was of any professional value to me anyhow. My true occupation in life was to solve the problems of elementary particle physics, and who cared whether I got a publication on the parametric amplification of —

Bromberg:

This is already quite late. These are coming in the late fifties, early sixties. And you're still set in this sense, at that late date, you're already starting to work for IDA or with IDA —

Kroll:

Well, again, I thought that was something I was doing for, like I do music, you know, something outside my main trade.

Bromberg:

By the way, didn't you find at all that there was a — well, I use the words cross-fertilization, but there was an interaction between your work on these kinds of more device-oriented problems, and your work on quantum electrodynamics? Maybe it gave you a deeper insight into classical electrodynamics which carried over, or I don't know what ? dealing with equations and — ?

Kroll:

No, there is one other aspect of all of this. I always had a kind of a Renaissance Man approach to all intellectual activity, not just physics, and I wanted to be able to understand everything that was going on in physics, so certainly all of this applied work was very consistent with that outlook and contributed to it, and it was, you know, one of my sources of pride that I could write papers in very many different fields, and that I could jump into a field I'd never been into and do something useful quite quickly, and — but I didn't have a tremendous drive to see everything that I did in print, is what it was. I didn't. It was really only, what I regarded as really important is what I think of as fundamental physics, and what I think of as fundamental physics is particle physics, and it was there that I would have wanted to publish everything that I thought was significant, and I think would have been more inclined to fight with the referee then.

Bromberg:

Why is it that you were able to jump into many fields and quickly do something useful? Is there an approach that you like to take to things that makes it possible quickly to get to where you can do something useful?

Kroll:

I can't answer the question. It's one of my talents. It has the defect that I'm not very energetic about, once I start something like that, I'm not at all energetic about exploiting it. There are a lot of things where I've been really one of the first people to do anything, in which I have written my initial paper, and decided that I've done what's interesting on the subject, and just have not pursued it after that.

Bromberg:

Still, people approach things in various different ways, and it would just seem to me a priori, although it might not be true, that some people would come into a new field and just try to read everything that's been done, and that you are really taking a different tack.

Kroll:

That's right. I read nothing, basically nothing.

Bromberg:

We've got some kind of creative principle here that —

Kroll:

I really look at it as a new problem, which I somehow see presented to me, in one context or another, and bring to it what I already know. That's not terribly efficient sometimes. Very often it's all been done. Besides which, the work that other people have done can save you a lot of…

Kroll:

…just something would happen that would get me interested in some item, that I would then see that I had something to say about…

Bromberg:

I guess, I just wanted to find out, you say you tried to be sensitive to what's going on. Is this sensitivity to — you're not reading, you're not going to the library —

Kroll:

— right. I do read PHYSICAL REVIEW. That's one thing I read. At least I turn the pages.

Bromberg:

So part of that sensitivity is just seeing what's going on in the pages of —

Kroll:

— and I go to meetings sometimes, talk to people, go to colloquia and seminars and often have to arrange them myself. I certainly know enough about what's going on now to know who to invite. On the other hand, you know, there are many, many thousands of pages published every month in physics journals. Nobody can read all that. Part of the difficulty of keeping up in the field has just been this enormous proliferation of information which is available. You could spend all of your time reading, yet you couldn't read it all.

Bromberg:

Is there any difference between being in New York and being out here, in that respect? Were you able to follow better there, or because there was more — ?

Kroll:

First of all, physicists travel a great deal, and I travel less than they do but I still travel quite a bit, and you know, there are six seminars a week plus a colloquium in our department. I don't go to all of them.

Bromberg:

So the attitude is, really, no.

Kroll:

No. What is desirable is to have a substantial group of active and fairly successful research people working around you, and in particle physics, particle theory, there hasn't been that much here, and I would say, there are undoubtedly places where there's more of that than there is here. But as far as physics in general is concerned, it's a very active department, and the particle theory isn't bad, it's just, if you go to SLAC there are 25 people there working on particle physics, and that's a more stimulating atmosphere than it has been here. But I used to go to SLAC quite regularly. I've spent several sabbaticals there.

Bromberg:

That should shed some light on how you work, how you think, a little bit. The paper on optical parametric amplifiers comes out very shortly after you get here.

Kroll:

Yes.

Bromberg:

And I also noticed that there wasn't any obvious, or there didn't seem to me to be any obvious origins that I could see, for that paper. How did it originate?

Kroll:

I don't know.

Bromberg:

Now, of course, that's the one paper I happen to have with me, and you begin by talking about the new work that's just coming out in nonlinear optics, people like —

Kroll:

I think it was a straightforward matter of being stimulated by, let me just say, it was certainly stimulated by Frankel's paper. And I don't remember what the Jordlane(?) paper was about, whether that was his parametrics —

Bromberg:

Index matching, I think.

Kroll:

Well, index matching was certainly the key idea, and the fact that you could get the same effect by putting the beams at an angle was very relevant to what I did here. Yes, I see, the second harmonic work of Franken? and the role played there by index matching, and then there was this, I guess, mixing, by Jordlane. So I really combined, first of all, the mixing idea with what I had learned about parametric amplification, to realize one could make an oscillator that way. That was one element in it. And then the question was, why? And then I recognized that you could make it tunable by using the angles. I think it was, that's what I would call a classical example of research that was stimulated by things I read in the literature. In fact, without talking to anybody about it. And my past experience. And why that happened then, I don't know.

Bromberg:

Now, at this point, just to go back a little bit, you had started working with IDA, and one of the things you had participated in shortly before were these meetings on applications of lasers at the Department of Defense.

Kroll:

That's correct. There could well have been a connection, because it meant I was thinking about lasers.

Bromberg:

Do you have any memory of those meetings?

Kroll:

I remember the first meeting. I'm talking about the Culver Committee now. I think Bloembergen was at that meeting, and there was something called the Bloembergen Cube that was the subject of many discussions. I think Bloembergen had pointed out that if you had a cubic meter of glass, and you excited all the energy levels in it that would ultimately lase, I suppose it was glass, with neodinium, it was probably neodinium glass, if you excited all the neodinium in the glass, the amount of energy stored in that cubic meter was very enormous and one ought to be able to do something with it. That was sort of the starting point. In fact, the straw man, with pre-electron lasers, when you talk about how much energy you have stored in the storage ring, if you could only scrape off a small percentage of that, you'd have a very high-powered device — a similar kind of argument. Just basically, you showed that the energy scales were such in laser devices that you could imagine producing enormous output powers.

Bromberg:

So that Bloembergen calculation was a big —

Kroll:

It was a back of the envelope estimate. It was a way of pointing out the potential potency of lasers as sources of very high-powered radiation.

Bromberg:

And I was told — you say you remember the first one, you don't remember the second — I was told that the outcome was an emphatic recommendation to Sebini that he pursue laser — looking into laser weapons.

Kroll:

Yes, I think that's correct.

Bromberg:

And then, aside from those meetings, was any of your other work aimed at lasers that early on?

Kroll:

No. I do remember telling Charlie about this, angle tuned parametric oscillator, in a restaurant in Washington, so —

Bromberg:

So it was published, but you had already been —

Kroll:

Well, I'm always a little show to publish things.

Bromberg:

I see, so months could elapse.

Kroll:

Years. Maybe it could be years. I think I was relatively fast then, because I had just moved to La Jolla and I was not teaching, and I needed to be doing something, and I had gotten that idea, so that was — well, I think it was probably '62 when I talked to Charlie about it, so it was not coincident with the Culver Committee. That was '61, wasn't it?

Bromberg:

It was the end of '61. Smith just got his glass laser to lase in the fall of '61. The committee met the end, around Christmas.

Kroll:

OK, so it's all very close together. I moved to La Jolla February 1, 1962, and I started working on this as soon as I got here. I had done it by the middle of the spring, and I told Charlie about it.

Bromberg:

Also with respect to this Culver Committee, you had been working or did some work with IDA on microwave beams. Did you do any work on microwave beam weapons? I don't know what you're supposed to tell me and what you're not.

Kroll:

No, I'll answer the question. I had written this memorandum on microwave breakdown. In my previous interview, I amplified a little bit more about that work. The origin of that paper was the claim by proponents of microwave beam weapons that the breakdown strength of air was much larger than the figures given in the literature, and that the reason was that in all the measurements that were done in laboratories, there were always walls, and the walls had certain effects which made the breakdown occur much earlier, and that if you had a beam out in space, the breakdown strength would be much greater, and there had been a previous calculation by —

Bromberg:

It sounds like —

Kroll:

Well, any proponent of a scheme will always — you don't get off the ground if you aren't optimistic. And the problem had been approached before. There was — someone, I think Brueckner, maybe colleagues of his, had also done some work on the subject, independently of and unknown to me. But I was asked to look at it, as part of my Jason work. So I wrote this paper on microwave breakdown, in which I analyzed the effects of the walls. I used a lot of scaling arguments to make it unnecessary to do a calculation from the basic physics, scaling from various experiments that had been done, and concluded that one could predict quite accurately what the breakdown strength was, and that it supported the view of the pessimists.

Bromberg:

Did you already bring that kind of consideration to the Culver Committee meetings? That would be a factor of some sort.

Kroll:

Well, the breakdown strength increases with frequency. It goes up as the frequency squared, so you would expect a much higher threshold. That's countered, to some extent, by the fact that you can focus a laser beam to a much smaller spot than you can focus a microwave beam, so that in the end, you might run into the same problem.

Bromberg:

I'm just wondering, that kind of study became very important shortly. I'm wondering if you have any memory of whether people were already talking about laser breakdown and stuff like that.

Kroll:

By the time of the 1960 summer study, that was a major topic.

Bromberg:

I see.

Kroll:

Nonlinear effects. Breakdown is not the principal nonlinear effect one has to worry about. A comment that I've made times, about the irony of governmental technical work, is that the things don't get finished and wrapped up, and that's in part related to the fact that funding disappears and in part related to the fact that things don't get published, so that a lot of money and a lot of effort is spent on a particular issue. The issue goes away because the program dies, or for one reason or another one doesn't pursue it, to the point of definitiveness. Then it's totally forgotten, and it never happened. And then when the issue comes up again, these ideas keep resurfacing and resurfacing, people start all over again. They have no memory that this problem was ever looked at before. This whole issue of atmospheric propagation and nonlinear effects in the atmosphere, which is currently a very hot issue in the free electron laser weapons business, is — was an issue that has been thoroughly reviewed and discussed, every time laser weapons have been seriously considered, and the first time was 1966. It came up in every laser summer study I've been associated with, and in many committees that I have had to do with in between. So the answer is, it goes all the way back to the beginning.

Bromberg:

By the way, Bob Sidell, who has been working on classified sources on this question about something called the Committee and Project Seasaw — is that?

Kroll:

Yes, Seasaw I know about, and I certainly have been to Seasaw meetings, although I'm not one of the people who's been strongly involved in Seasaw, but I know about Seasaw. The committee I never had anything to do with. That was a microwave weapon committee, wasn't it ?

Bromberg:

Here I'm just, somebody else is speaking through my mouth. He just asked me to ask this.

Kroll:

I do think he says microwave. That's probably right.

Bromberg:

I have a version of the question right here, so —

Kroll:

Yes, it says "Committee studied the microwave beam weapon."

Bromberg:

Yes, I think it's transcribed.

Kroll:

I believed the microwave beam weapons were nonsense at that time and I probably still do. So I had really no interest in participating in it. I don't know if I was asked, but —

Bromberg:

This last paragraph, and the footnote.

Kroll:

I don't see any reference here to, oh —

Bromberg:

— that was a Seasaw reference.

Kroll:

Well, there is a Seasaw in the Committee reference, and I — the answer is, I never had anything to do with that committee, nor have I had anything to do with microwave beam weapons, except to offer evidence that they weren't a good idea.

Bromberg:

But at the time, at any rate, in '61, the laser beam weapons seemed to be, because of its breakdown problems, a less bad idea. What else can you tell us about the '63 laser studies? First of all, you said that this is where the question of breakdown and other nonlinear effects in propagation —

Kroll:

Well, it was where thermal blooming was invented, I believe. There's a reference here which, in that thing you gave me — yes, I forget these things, but when I see them here — yes, this instability of an intense optical beam, December, '63, that was the thermal blooming paper, and that has been a persistent and important problem in beam weapons ever since it was first observed. That was basically Brueckner's idea, and he got us all involved in it. Now, it's — as it has turned out, it's a difficulty that you can ameliorate. That is, you can have a beam weapon which avoids the problem, but it's a constraint. It's a pain in the neck. And it's important to choose a frequency in which the absorption is very low. See, it could be that 10 percent absorption would be fine, because you get 90 percent of the power on your target, you don't mind losing the 10 percent, and if you could make something that was twice as efficient, it would be worth losing the 10 percent, if you were working in a region where the absorption was 10 percent. However, if you absorb 10 percent of the beam, the thermal blooming will destroy the beam totally, and so you actually have to absorb a very tiny fraction in order to be able to propagate a well-focussed beam through the atmosphere at all.

Bromberg:

And Brueckner was important at this summer study in raising this issue, is that what ?

Kroll:

He was the director of that particular summer study.

Bromberg:

And he was the one who raised that particular physics issue?

Kroll:

Yes, he raised that physics issue, yes. Yes.

Bromberg:

And were there other issues that other people were prominent in bringing forth? Or other interactions?

Kroll:

Oh, there were lots of them. The trouble is, I don't remember them. But there were many nonlinear effects known by that time. I believe the Raman effect was known by then, the Raman laser was known by that time, so that you would have to worry about the stimulated Raman effect was already known, and that happens to be the main nonlinear effect you have to worry about.

Bromberg:

I see.

Kroll:

It's more important than thermal blooming. And so I would say, it still appears to be the main effect, and was recognized in that first summer study.

Bromberg:

At that particular point, one didn't know whether thermal blooming or stimulated Raman effect would be the determinant one? Or one already had a sense that SRS was going to be the really big ?

Kroll:

Well, the difference between the two is that you can reduce thermal blooming by slewing, and by using short pulses, and that's not nearly as effective for the stimulated Raman effect, because the stimulated Raman effect is a much more gradually developing process.

Bromberg:

What does slewing mean?

Kroll:

You're tracking a missile, you slew, you move the source of the radiation following the missile. That means that the time that you spend heating the intervening atmosphere is, can be a small fraction of the time you spend illuminating the thing, since you're always heating different parts of the atmosphere, so — see, the first thing you would do, suppose you were aiming at a fixed target, if you get no effect then, then you relax. You don't have to worry about it. And you find out the effects. If the effect then is so bad that you couldn't imagine doing it, then you begin to worry. Then you begin to think about things that could make it better. And slewing makes it better. And pulsing it makes it better also, although that means you put more energy in in small batches, but it means that the air doesn't have a chance to expand, and the fact that you use more energy makes it worse, the fact of the short time scale so the air doesn't have to move — the more energy makes it worse, the fact that it hasn't time to expand makes it better, you have to calculate which wins, whether you gain anything or not.

Bromberg:

Were these calculations already being done at this summer study in '63?

Kroll:

I don't remember. Some of it was certainly done then, and it was certainly done — once these problems were recognized, there were people working on them all the time. Contracts would be let to various technical supporting groups to study these problems, and they would report on their results at the meetings, and I think the situation was quite well understood, let's say, by the late sixties. And it's just that all of that gets lost. It means that it's being done again now.

Bromberg:

Now, something else is coming in here, and that is, plasma physics. When did that begin to enter into — did that come about because you were at San Diego?

Kroll:

Yes, that came about explicitly because I was at San Diego, because mostly there were plasma physicists at San Diego, which they were not at Columbia, and I got involved in that paper with Ronald Rofstok(?) because he came into my office and began talking to me about the problem. There is, in my parametric amplification paper, that tunable laser paper, there is a footnote on doing light by light scattering by optical mixing. And since the experimental observation of light by light scattering was one of those great challenges to experimental physicists, because light by light scattering in a vacuum is a very — do you have that paper ? That's the one. I'll show you the footnote, I think it's footnote 9.

Bromberg:

Is this it, footnote no. 9?

Kroll:

Yes.

Bromberg:

How does this fit in with what you were telling me about Rofstok?

Kroll:

Well, there was a paper by (Phil) Platzman and (Sol) Buchsbaum and Tzoar(?) on light by light scattering in a vacuum, I mean light by light scattering in a plasma, and you will notice that, if you read that, that I replaced the problem of observing light by light scattering by observing three way mixing, and so Rofstok, whose name is also Norman so we say "Hello, Norman" to each other, —

Bromberg:

He read that and came in?

Kroll:

I don't remember why he discussed it with me. Maybe because light by light scattering is something people in quantum electrodynamics are supposed to know about, so he came in and began discussing this paper with me, and I believe that I pointed out that you could make such measurements much more easily by doing three way mixing, and that was the origin of that paper.

Bromberg:

And the AEC contract that it sites, was that his contract? I always think of people when they come to a new place, like from Columbia to San Diego, going through a little period of renegotiation to get their contracts together.

Kroll:

I had a very singular experience, along those lines, which is that I have never written an initiating contract proposal in my life.

Bromberg:

But your work is supported on contracts.

Kroll:

All the time, that's right. Some people are just lucky.

Bromberg:

Does that mean you just got added onto his contract, or…?

Kroll:

No. No. I don't remember which contract it was done under, but when I was at Columbia, first, I was born in physics working for the Radiation Lab, and since the Radiation Lab supported work on quantum Electrodynamics — after all, the Radiation Lab paid for the Lamb Shift Experiment — there was no reason why I couldn't work on quantum electrodynamics for the Radiation Lab. Then there was a blanket theoretical contract, which I think Bob Serber negotiated at Columbia, for all the people who were in theoretical physics, so I could go to Brookhaven and work on meson physics and nucleon physics and nuclear physics under that contract. Before I came to UCSD, Brueckner, as part of his great move to make this a great place, had negotiated a number of broad contracts, including one on nuclear and particle physics, which was split into nuclear physics and particle physics, and he was doing nuclear physics really himself exclusively at that time, and when I arrived, he just handed me the particle physics half, so I've had a particle physics contract ever since. It would have been more natural for this particular work to be supported by the plasma physics contracts which then existed, so it was probably supported by that contract, but the subsequent paper that I wrote on the Puerto Rican Conference paper was I believe almost certainly supported on my own contract. I'd have to check the actual contract numbers to know.

Bromberg:

That was Paper 38.

Kroll:

And that paper has an interesting history also.

Bromberg:

It was supported by an ARPA project Defender.

Kroll:

OK, I see. Then the —

Bromberg:

— it struck me, I guess, as peculiar. That's why I noted it because it seemed to be plasmas, nonlinear optics and plasmas.

Kroll:

Yes. There was an applied physics project which Keith had negotiated, associated with something called IPAP, Institute of Pure and Applied Physics, which was an organized research unit at UCSD, and I believe that that work must have been done under IPAP's sponsorship. So that was part of the additional contract which Keith had initiated.

Bromberg:

No, the optical parametric amplifier, you said that no one that you were in contact with took that work up in an applied way. With the plasma, for example, this business of shining in two light beams, was that taken up in a more systematic way?

Kroll:

It was actually three light beams, two whose frequency differs by the plasma frequency, and that excites a plasma oscillation, and there's a third one which we then used to scatter off the plasma oscillation, to observe that scattering.

Bromberg:

Was that then tried out? What was your relationship there to the experimentalists?

Kroll:

Well, that work was supposed to be an experimental proposal, as a way of measuring plasma densities. The title of the paper was "Three Way Mixing as a Plasma Density Probe.” That was and is an interesting problem in plasma physics, in fact in general diagnostics is a key problem in plasmas, and so that was certainly its orientation. Was it ever done? Yes, it was eventually done, and in fact, I had a devil of a time getting the person who does it to recognize the fact that he's just doing an experiment that we proposed there. I make complaints.

Kroll:

It is in fact the same experiment, but he had a different reason for doing it, and therefore doesn't like to acknowledge the fact that it is precisely what was suggested in this paper, and it worked fine.

Bromberg:

See, the point of that question is to see, in this period, which is the mix-sixties, what kind of relationship you were having with plasma experimentalists.

Kroll:

Well, the answer is that the plasma community certainly did not jump to pick this up, and the measurements that I'm referring to are very recent. They have to do with the beat wave accelerator. The idea of, this business of actually exciting a plasma oscillation by the beat between two optical beams whose frequency differs by the plasma frequency is something that, as far as I know, has only been done experimentally in the context of the beat wave accelerator, and so, it was Chan Cho Shih who did this work, and so he did it, he succeeded in exciting a large plasma oscillation in this way, and demonstrating that it was there by scattering off of it. An issue of key importance in Cho Shih's work was what limits the size of the plasma oscillation, and that was work with Rosenbluth and Lieu which came after our paper, and that was the focus of his interest, whether he could excite a really high intensity wave, and whether he could confirm the limitations which the Rosenbluth and Lieu paper had, and since that was his motivation, that's why he somehow doesn't see the connection with our original work. He would have done the experiment even if we hadn't done that original work. Still, I think he ought to acknowledge it.

Bromberg:

But then with all the plasma physics research for example going on in the fusion group at Livermore, was that a group that you were particularly in touch with? If you were sponsored by the AEC, did they drag you into conferences?

Kroll:

No. And I don't think this method was ever actually used, for the measurement of plasma densities, although it works for that. It was essential in Cho Shih's work that he had a plasma density which matched the frequency, and he had another way of determining what the plasma density ought to be, but again that was a theoretical inference, and the fact that it worked was a demonstration that the plasma density was what he thought it was. So one could argue that he really did use this method to confirm that the plasma density was what he deduced it should be from other considerations.

Bromberg:

What about the paper you gave at Puerto Rico? Did that bring you into the AEC plasma fusion community ?

Kroll:

Not at all, because, again that paper has a curious origin. When we tried to publish that paper, the —

Bromberg:

We're talking about the one that was published from the Puerto Rico seminar?

Kroll:

We're talking about the one that was published before, the Rostoker paper. When we tried to publish that paper, we got terrible static from the PHYSICAL REVIEW LETTERS and had a terrible time getting it published, because the referee kept saying, we should be getting an answer which was closely related to the answer in the Platzman, Buchsbaum and Tzoar paper. The referee thought you could deduce what we should get from that paper, and that paper got a temperature dependent. If you did that, if you took what they said we should get, you would have gotten a temperature dependent result and a much larger result than we got, and we were convinced that they were wrong, and so we had a long argument, and we claimed that we had a perfectly good theory. They didn't point out anything that was wrong with our theory. All they were doing was using another theory to deduce what they thought the result ought to be, and if they wanted to publish a paper saying this other theory gave this different result, that was their privilege, but they had no business making an unpublished deduction and claiming ours was wrong without directly attacking what we had done. Well, we eventually won. But the issue that they had raised was an interesting one to me, that is, exactly why was their argument wrong? And that's what this Puerto Rico paper explains. It explains the relationship of that calculation to this calculation, and why the kind of inference you might have made from that work is in fact incorrect. So that was the primary — in other words, it provides a great deal more physical insight into the nature of those related nonlinear processes.

Bromberg:

Now, Puerto Rico was the site of a great deal of conversation, I understand, and partly, I assume, on nonlinear processes.

Kroll:

Yes. There were a number of papers on it. Mine was one of them. There was another one, a very closely related one, by Bayman, I think it was, at Illinois. I may have the name wrong but I think that's who it was. I occasionally run into aphasia problems.

Bromberg:

I know a Gordon Bayman.

Kroll:

That's who it is.

Bromberg:

But I didn't think he was at the Puerto Rico Conference.

Kroll:

I think he was. I think he wrote a paper very closely related to the one that I presented. It immediately precedes it. If you want to shut that off, I'll get the book…

Bromberg:

Was this an important meeting for you in any way? I know there was a classified meeting, and the unclassified —

Kroll:

The answer is, not at all. The truth of the matter is, I was invited to give a talk on nonlinear processes in plasmas to this meeting, which I accepted, and I didn't know what I was going to talk about. I began sweating and sweating, and finally I sat down and I did that paper in one week, under the gun of having to talk about something.

Bromberg:

And under the provocation of the referees, as you described it.

Kroll:

Well, that was sort of over by that time, but I did want to understand the issue. So that's how that came about. I did have a student working on the problem actually at the time, but he was accomplishing absolutely nothing.

Bromberg:

What was the relation of your paper then to what he was doing?

Kroll:

Well, I think I probably thank him in the paper for discussions but I think he was a student who never survived.

Bromberg:

So the discussions in Puerto Rico were no big deal from your point of view.

Kroll:

I'd never been to Puerto Rico before. It was interesting to visit it, and I liked hearing the other papers, but again, this was not something that was central to my main interest in physics.

Bromberg:

Even at this point, we have to say that the work you were doing on lasers and plasmas and nonlinear optics and all this propagation was still not central.

Kroll:

From an emotional point of view, that's correct, yes.

Bromberg:

What about intellectually, as you viewed what other people were doing? I have a feeling that might just come from my ignorance of the physics that what Townes was doing was very different from what Bloembergen was doing, and so I wondered whether people at that point were recognizing a Townes school, a Bloembergen school, and were thinking that this way or some third way was the right way to do this kind of thing? Is that just too articulated a view to take of what was going on then?

Kroll:

Well, Townes and Bloembergen are very different kinds of physicists. I think that's true. I think Townes was another person who was strongly motivated to do a large variety of things. After he left the laser field he went into astrophysics, and that's a measure of his greatness, that he could do that at the level he does it. Of course he's done very important work in astrophysics.

Bromberg:

Let me go back to another meeting, and that's the one where you were giving lectures and other people — was that an interaction of interest?

Kroll:

Well, again, I went to because I was asked. It seemed to me like a few weeks in the French Alps would be nice. And I was supposed to have my lecture notes completely written out when I arrived. I arrived with absolutely nothing, and I had to prepare my lectures day by day as I was giving them, so I was in a semi-panic the entire time, getting them out, so I didn't have time to enjoy it like I was supposed to.

Bromberg:

For example, Lamb was just finishing or just publishing his semi-classical theory of the laser, and then I think, I believe this is true, was claiming that you can't do it with semi-classical theory of the laser, you have to do quantum theory. I'm not sure about that, but certainly the question of whether you should do semi-classical or quantum theory was part of what was going on, and I'm interested in whether you were involved in that because somehow you were involved in both halves — you were involved in quantum electrodynamics, you were involved in doing very classical theories.

Kroll:

I wasn't involved in those controversies, I would say, although I knew what they were doing. I'm trying to remember whether I overlapped Lamb. Certainly Lamb heard my lectures. I don't know that I heard all of his. I remember Lamb commenting, after I finished my series of lectures, that I had removed a lot of mysteries from the subject. I gave a discussion of line width problems which I don't think I had seen before anywhere.

Bromberg:

I'm sorry?

Kroll:

Line width problems, how they — from a nonrelativistic quantum electrodynamic point of view. I think those are the mysteries he was talking about.

Bromberg:

And with Glauber, did you have this coherence —?

Kroll:

Well, of course, I thought his coherent states were interesting, though I don't know how appropriate it is to associate them with him, but he certainly publicized them. It's hard for me to believe that those were not well known before, I mean, as states of the harmonic oscillator, although… and optics was certainly original with him. Maybe not entirely original. I really don't know. I use those, discuss those in my lectures, and some of the relationships of them to other ways of doing things. But my remark about being unprepared is related to the fact that I didn't have too much time to talk to people while I was there, and had a very rigid — you lectured every day, so I had to give a lecture every day and then start preparing the next day's lecture.

Bromberg:

Is there anything that went on that was important for you in that?

Kroll:

It was an emotional strain. That's what I remember about it. I was alone. I was lonely. I was cold.

Bromberg:

Quite a contrast with your expectations. All this theorizing about the laser, the nature of laser light, the nature of the fluctuations that you were getting out — you were working on this directly, about the way to theorize about lasers, whether you could used a classical field or had to use a quantized field, this was all going on. This wasn't stuff you were working on.

Kroll:

I never thought it was terribly interesting. It's probably more interesting than I think it is, but I think it's probably not as interesting as some people think it is either. It's probably somewhere in between. This is talk about squeezed states now. To me it's a fancy name for something that's been very well known for a long time. There are people who say it enables you to overcome certain kinds of quantum limitations in measurements. Well, maybe it does. My approach to that sort of thing would be not to talk about squeezed states as some important idea which might be used to overcome quantum limitations in certain kinds of measurements. My approach would be, there's a certain kind of measurement I would like to do, but the conventional way of looking at it keeps me from measuring this important quantity to the accuracy I'd like to measure it to, and I need to know it to that accuracy. How can I do it? Then I might, addressing that problem, I would regard squeezed states as a tool if they applied to that problem. And if somebody succeeds in making a measurement which one needs to make, which one needs the squeezed state technique actually to do, I'll think that's a worthwhile application. But as a kind of in principle thing, I think it's a big overblown.

Bromberg:

Now, if I wanted to apply this kind of thinking backwards, should I be saying, if there's something you want to know for which you need quantum electrodynamics and laser theory then you would think, all right, let's use quantum electrodynamics, but just to take up the problem in its general sense is nonsense? Is that a correct way to —?

Kroll:

Well, that's a way of making it clear that the statement I'm making is too extreme. The fact of a harmonic oscillator, if you measure the position very accurately, the momentum is known poorly, but that periodically it always returns to the now, and even though the spread of the momentum then makes the spread of the position grow rapidly, it periodically comes back to the now state that you started with. That's something one learns in the first course one takes in quantum mechanics. It's something that I teach every time I teach quantum mechanics. So that's an old idea. It seems to me the point that being made is that one can realize that situation experimentally, which doesn't surprise me at all, and the interesting issue is whether one can use it for anything, to me. The difference between your example and that is that something which is sort of elementary quantum mechanics, and a new general theory like quantum electrodynamics, seem to me to be different things. All that one has done is really to re-name something that is an elementary part of quantum mechanics by calling it squeezed states. That somehow makes it better. It's a little like calling the Hubitron a free electron laser. Hubitron — that's what Phillips called it. But renaming it the free electron laser obviously has enormous positive effects upon its development. So has renaming these states squeezed states.

Bromberg:

It's interesting that packaging can make such a big difference.

Kroll:

Yes, that's a good — besides history of physics, which should be sociology of physics.

Bromberg:

I don't mean to leave it out. Well, if you thought this not terribly interesting at the time, you would not have followed it all that much.

Kroll:

That's right. I'm sure it's more interesting than I thought then and than I think now, but it's not something I've gotten involved in.

Bromberg:

Well, I just want to get those things that you were watching and having opinions on at the time.

Kroll:

I think if one succeeds in focusing on a really interesting issue, then an elegant formalism for getting at that issue becomes something that would appeal to me, interest me — even squeezed states could. But in the absence of focusing any of those things down on an interesting issue, I don't find them very interesting.

Bromberg:

Let me change the subject and go to this question — the transition that gets you to the gas dynamic laser. You had another laser summer study, summer '67, and at that point you got the gas dynamic laser, and I'm wondering how that struck people.

Kroll:

Well, as you noted, the Culver Committee was in response to the neodynium glass laser, very quick response, and that was followed by the 1963 summer study. The gas dynamic laser was recognized as a comparable breakthrough, and it was perfectly natural that there would have been a summer study to respond to that possibility.

Bromberg:

What were the principal things that were going on in that?

Kroll:

It's really to the extent that I can remember them, and the answer is, very little, even though I ran that study.

Bromberg:

Did you call it, or they asked you?

Kroll:

I was a member of the Jason steering committee then. I don't like to run studies, but I considered that one of the duties of members of the steering committee was to run studies. I was obviously the person to run that study. I didn't see how I could not do it.

Bromberg:

How come you were the person to do it?

Kroll:

Because I was the person on the steering committee who knew the most about that sort of thing. About lasers in general. I had participated in the previous laser study. I had written papers about lasers in between, and nonlinear processes, which were obviously involved, and I was the natural person to run it. It was a perfectly reasonable thing for Jason to be doing that summer. Part of what the steering committee does is choose the projects for summer studies. It seemed a very reasonable, natural thing for us to do. How could I not do it?

Bromberg:

Do you have any recollections of the principal problems people defined there, or new points of view some people brought?

Kroll:

Well, in all laser studies one is interested in high power and high efficiency. I'm certain we worked on issues of power and efficiency. Instabilities, all these devices are subject to serious instabilities, once you get these very high powers in them, they don't want to oscillate the way you want them to oscillate. There are many other ways of doing it. Surely we studied things of that sort. And then again, those old problems of propagation in the atmosphere. It's a different frequency, instead of the neodinium glass frequency of 1 micron, it's 10 microns, so we reviewed all those things. Practical applications were discussed. I know Townes did a paper on fog dispersal with gas dynamic lasers. Townes — on the use of gas dynamic lasers to disperse fog at airports. Those were the kinds of things. I'm sure I could remember much more if I would look at the old report. I would also believe that most of its unclassified, but I don't have it and I don't remember more than I've told you. Also transient — my subsequent papers on transient effects are to some extent related to that meeting, although I had done work on transient effects before.

Bromberg:

How were they related?

Kroll:

Well, there is the question of whether you can ameliorate these nonlinear effects by not giving them time to build up, and so you don't look at them in a steady state situation, you look at them in a situation in which you have very short pulses.

Bromberg:

You had done that before for IDA?

Kroll:

I had done some work on that before, and there was more work, and my work with Kelly was related to that. Kelly was a member of that particular summer study. Those papers with Kelly in part stem from that meeting, that's right. Some of the methods that were involved in that work, I had already, in that paper on hypersonic, basically stimulated Rewana (?) effect paper, that had also been done in the transient regime as well as the steady state regime, and so — because there too we were really concerned with fracturing materials with short pulses, so the transient regime was of interest, and so I've had some background in how to attack problems of that kind, and that background was very much involved in Paul and my subsequent work in that area.

Bromberg:

Did he come out to Berkeley just to work with you?

Kroll:

Well, my recollection is, I believe I first met Paul at [???] and I think he's one of the people who took notes on my lectures, actually, one of the people who, since I had brought any lectures, there were people who wrote them up.

Bromberg:

I wonder if he was one of the people who invited you to Puerto Rico, since he is one of the editors of that volume, one of the organizers of that conference?

Kroll:

Possible. But no, my recollection is that I invited Paul to that summer study. That's '67. But then he came to various summer studies after that. And I think that PHYSICAL REVIEW paper is probably something we did by correspondence afterwards. We'd really run much of the initial theory.

Bromberg:

I see, so it was done right then.

Kroll:

Some of it. There are some major contributions to that work that I really would attribute to him. The theory is applied to a wide variety of processes, and some of those processes were certainly his contribution.

Bromberg:

So you said the relationship between quantum and classical theory is how you got into the free electron problem. You also said in another interview that the IDA free electron laser studies also had something to do with the way you got into the free electron problem.

Kroll:

I don't think we discussed that really.

Bromberg:

There's a brief mention, IDA was doing these free electron laser studies very early on, long before you published your free electron laser papers.

Kroll:

But the free electron lasers are first mentioned in Jason work in the 1974 Jason laser summer study, and I wrote the section on free electron lasers. The reason I wrote the section was because I had a graduate student working on free electron lasers already, and this was a case of bringing academic work to Jason, rather than the other way around, and the reason I had the graduate student working on the subject was that I had read — I read very little but sometimes I read something which has an effect on me, and I have people send me preprints all the time, and I look at the titles. Occasionally I read one, and I read that one, and it was on the free electron laser.

Bromberg:

This is Wayne McMillan?

Kroll:

Wayne McMillan was the student. The paper that I'm talking about was by John Madey, Schwettman and Fairbanks, and it was published in the IEEE TRANSACTIONS on nuclear science. They probably had written other things about it, but that was the particular one that I saw. Well, I looked at the paper. I think the title "Free Electron Laser" got to me.