Oral-History:Norman Ramsey (1995)
About Norman Ramsey
Norman Ramsey is a physicist best known for his development of the separated oscillatory field method of measurement and for his work in magnetic resonance. He received his bachelor's degree in pre engineering at Columbia University. Ramsey attended Cambridge University on fellowship, ultimately receiving his Ph.D. in physics at Columbia with I.I. Robby, and wrote his dissertation on magnetic resonance and measurement methods. After a post-doctoral appointment at the Carnegie Institution in Washington, Ramsey worked briefly at the University of Illinois in 1940. During World War II he worked on radar and magnetron development at MIT's Radiation Lab; later, he became an expert advisor to the Secretary of War, working first at the Pentagon and then at Los Alamos, where he worked on bomb shape testing at Dahlgren proving ground. After the war, Ramsey returned to Columbia, where he and Rabi established the Brookhaven National Laboratory with Robby. Ramsey was head of the physics department at Brookhaven and professor at Columbia until he moved to Harvard in 1947, where he continues to work as a professor.
The interview begins with the circumstances of Ramsey's education at Columbia and Cambridge and then provides some detail about Ramsey's doctoral work with Robby on magnetic resonance and measurement. Ramsey discusses his work at the Rad Lab, the Pentagon, and Los Alamos during World War II, and then describes the establishment of Brookhaven National Laboratory and his work at Harvard, including his participation in the development of the separated oscillatory field method and its various applications, discussing the relationship between technology and its applications and the roles physicists and engineers play in the development of applications. Throughout the interview, Ramsey compares and contrasts physicists' and engineers' work habits, motivations, and general philosophies, noting that at many times in his career he as a physicist functioned as an engineer, usually when he had to make his own apparatus for a particular experiment. He suggests that there are, overall, too many similarities between physics and engineering for a real acid test to determine who fell into which category. After treating several of his research projects from the 1960s and 1970s, the interview concludes with a discussion of Ramsey's use of IEEE journals for background information related to engineering.
Also see Norman Ramsey's 1991 interview.
About the Interview
NORMAN RAMSEY: An Interview Conducted by Andrew Goldstein, Center for the History of Electrical Engineering, May 9, 1995
Interview #253 for the Center for the History of Electrical Engineering, The Institute of Electrical and Electronics Engineers, Inc.
This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.
Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, Rutgers - the State University, 39 Union Street, New Brunswick, NJ 08901-8538 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.
It is recommended that this oral history be cited as follows:
Norman Ramsey, an oral history conducted in 1995 by Andrew Goldstein, IEEE History Center, Rutgers University, New Brunswick, NJ, USA.
Interview: Norman Ramsey
Interviewer: Andrew Goldstein
Place: Cambridge, Massachusetts
Date: May 9, 1995
Childhood and Undergraduate Education
From some of your earlier interviews, I feel like we have a good picture of your early background. So I would like to just start with your professional career.
Well, you presumably have my background, and my earliest education, so let me start with very briefly on my undergraduate work. My first interest in physics, as far as I can determine, started somewhere before I was ten, reading a chapter of a book of my father's on chemical engineering. He was an Army ordinance officer. He had been a commanding officer of Picatinny Arsenal in New Jersey. He wasn't a general then, he was a major then, and later he was commanding general at Rock Island arsenal. In any case, as a small boy I liked to read whatever books I got a hold of, and he had a book on chemical engineering. One chapter was on atomic theory, a pretty feeble version of atomic theory at that time, essentially on the simple Rutherford model. But that fascinated me. Reading about physics in that engineering book was probably the first thing that really interested me in that subject, though I didn't realize that at the time. Then I skipped two grades in school. Usually when we moved we moved from better schools to worse schools — so they thought — and I skipped at least two years, which was fortunate for me because that meant I got one of the last Ph.D.s before World War II instead one of the first ones after. There was about an eight-year interval of time difference between those two.
In any case when I graduated from Leavenworth High School in Kansas, I was too young to go to West Point and too young to go to MIT according to the rules. My MIT friends later said, "Oh, there would have been exceptions." But my father said, "No, you don't ask for exceptions." So they were going to send me to a place called Trinity Academy in New York City, to more or less fill in a year and make up for the two years I had skipped. The headmaster there, to my great appreciation, decided after talking to me awhile that I would probably be a nuisance in his school — be bored with what they were doing. He recommended that I go to Columbia, which I did that same afternoon. The college term was already something like two weeks underway, but that was in 1931, in the middle of the Depression, and like any fee-paying student, I was welcome, so I was immediately admitted.
I went around that same afternoon and bought my books and showed up the next day for class. I signed up initially for pre-engineering. But the way pre-engineering was taught at Columbia at that time (1931) pretty much emphasized the use of tables and formulae and deemphasized trying to understand what the tables and formulae were all about. That wasn't really what I wanted to do. Then I made an excessive flip in the opposite direction. I was a mathematics concentrator most of the time I was in college. I didn't even know physics was a subject that one could work in. It was maybe a course one took, but I didn't even know it was a profession. That possibility didn't even surface till about my senior year. Then Columbia gave me a traveling scholarship, a fellowship to Cambridge University for two years. And as a disappointment to the math department, I said I would be happy to accept, provided that they would let me use the scholarship to convert from mathematics to physics, which I did.
I was in Cambridge at a very good time. The world's centers for almost all physics and certainly for nuclear physics was Cambridge with Lord Rutherford, J.J. Thompson, Dirac, Cockcroft, etc., twenty or so very outstanding physicists. Because of the difference in the level of undergraduate training at that time, I really received a second bachelor's degree from Cambridge because that is where you do the course work. Then I came back to Columbia and started immediately doing research. It only took me two years to get my degree from then on. So it only really took four years in all to get my degree. I don't think I lost any time, and I benefited by having a good foundation in theoretical physics as well.
Rabi and the Magnetic Resonance Method
There I had a piece of very good luck. In writing one of my essays for my tutor at Cambridge, I had written about nuclear magnetic moments. I was looking into the work that was being done by I.I. Rabi measuring the magnetic properties of nuclei by deflecting atoms in an inhomogeneous magnetic field. I was fascinated by this work and also very much attracted by the fact that Rabi, though he was doing experiments, was also doing theory, and very much understood both. I decided to go back to Columbia and work with him. When I first told him that I wanted to work on molecular beams, he was somewhat discouraged. He said he would be happy to have me work with him, but he felt that maybe the field of molecular beams was a bit finished. One could measure magnetic moments to no better than one percent accuracy since one had to measure an inhomogenous magnetic field, which could be determined much more accurately than that. So, maybe there wasn't so much to do other than work hard for a little bit of improvement.
Nevertheless, I said that's what I wanted to do. About three months later, just as I was finishing my apprenticeship work and learning my way around the lab, Rabi invented the molecular beam magnetic resonance method, which was the very first magnetic resonance of any kind: the predecessor of NMR and the predecessor of MRI and all of these. The fundamental work was really done on that method as far as the magnetic resonance portion. There were great advances from each of those steps. When we were doing that work together, I had the lucky break of doing the first Ph.D. thesis on magnetic resonance of any kind. But at the time that we did that research, we certainly did not realize we were going to anticipate NMR chemical analysis, and we did not anticipate magnetic resonance imaging. We wouldn't have dreamed of that. There weren't CAT scans or anything like that at the time. Nevertheless, our work was a major start, and these were the fundamental originating papers on that.
We also had the good luck to open a new field and a total new method of measurements. Instead of to one percent accuracy, we could make measurements to a part in ten or hundred thousand easily by essentially using an oscillating magnetic field and looking at a resonance. When the frequency of the oscillator was equal to the intrinsic frequency of the atom in the magnetic field, then we could detect that by the change in the beam intensity that came through the apparatus. This worked very well. We had a whole field to open up. Sort of intriguingly, as far as my Ph.D. thesis went, we originally worked with a total of four people: I.I. Rabi, J.M.B. Kellogg, Jerold Zacharias, and myself.
Ph.D. Work on Protons and Deuterons
We were trying to measure the magnetic moment of the proton and the deuteron, and we were expecting to get a rather clearly defined resonance pattern. A peak that goes up and comes down, which our friends working with Rabi at about that same time, or just a little bit before, had found in the case of something like lithium chloride resonance; they got what they expected. To our horror we got a mess, just things all over. We thought first that something was wrong with the apparatus, and we couldn't even see a good way of measuring really accurately the magnetic moment of the proton and the deuteron. Then we discovered instead of doing experiments with H2 and D2, if we did it with HD we got some rather nice, sharp resonance and from them we could get the magnetic moments.
At Columbia, at that time, there was a rule that a Ph.D. thesis had to have a single author; it couldn't be shared authorship. That always created the necessity of inventing an uninteresting problem for the thesis student to work on. It was decided investigating this mess would be a good thesis project for me. It seemed fine to me. The following summer I began this study. I fairly quickly found that we were using much too much radio frequency power. Every time I reduced the power by a factor of 2, I should have done it by a factor of 10 and it would have gone quicker; the resonance pattern got better. In fact it became clear instead of it being just a mess, it was six rather sharp peaks in the case of H2, and for the first time that showed I was getting a spectrum. It was really spectroscopy we were doing at radio frequencies rather than just measuring magnetic moments. They are related, but nevertheless we thought of it initially in the first form and were not expecting any structure. This could be interpreted, we all discussed that together, in terms of the magnetic interaction between the two protons and the interaction of the protons magnetically with the rotating magnetic field of the molecule, as well as the interaction with the external field we'd put on. I then did the same thing with D2, where we obtained a moderate central peak but the resonance continued off to the side. It didn't quite die away, and it went out much further even when it went in H2, which seemed to be unlikely for a magnetic interaction. Since the magnetic moment of the proton is bigger than the deuteron, it should have been proton much further out.
At that time it was clear we were also discovering something new in D2. It was also becoming clear that was the important part of the experiment. We talked a little about that before, that it could be a quadrupole moment with the deuteron. The deuteron, which is a heavy hydrogen nucleus, instead of being spherical-shaped, as everyone had presumed, was more football-shaped. This, in turn, meant there had to be a different kind of nuclear force, a so-called "tenser force." So then we decided these discoveries were too much for a single-authored Ph.D. thesis, since we all had built the apparatus together. So with full agreement on all sides, we decided to share the results of that work.
Another thing I had discovered was that I could measure magnetic moment due to the rotation of the molecule. So that became my formal thesis project. We did the quadrupole work together, and it worked out extremely well. It was a major discovery. There is always a great place in research for luck, and this was clearly one of my first big lucky breaks. I came into that lab at just the right time; it was just the right lab; and we had a powerful new technique, and just happened to make a great discovery, or two great discoveries, in the process of it.
We were doing fundamental physics research, but there was engineering involved when we had to make apparatus; it was a crude apparatus, not very well engineered. Physics and engineering have very many things in common; many physicists become engineers and many engineers become physicists. Both of them do many similar things. This is the case for an experimental physicist. He would say a physics thing is finding the problem, but how to do it is a very large degree engineering, at least engineering on a crude scale.
WWII and Initial Work on Radar
After that work, I was a post-doc at the Carnegie Institution of Washington for one year. There weren't many post-docs at that time, but I got one of them along with Jim Van Allen, later famous for discovering the Van Allen Belt in space, who was the other post-doc that year. Then following that year I was married and got a job at the University of Illinois. I had probably the shortest career at the University of Illinois of anyone who didn't rob a bank. I lasted about five or six weeks.
This was in 1940. The big Nazi invasion of France was that summer, and things were looking very dim internationally. The U.S. was not yet formally in the war, but the Nazis seemed unbeatable. France had just fallen completely, and it looked as if Britain would soon follow suit. Consequently there was much worry. The British had, at the end of that summer, sent over a delegation to try to see if they couldn't persuade some of the American scientists to do research on radar, which they then called radio-location.
Actually, radar was independently invented by about five different places. I don't know which was one first. I think that the Office of Naval Research invented radar in a very early form, but the U.S. wasn't at war, so they didn't push it so hard. The French ship The Normandy actually had a radar iceberg detector before the war. All of those were long at wavelengths, at least several meters long.
But the British had a real threat of war, so they went all out in capitalizing on radio and particularly developed this very effective "chain" system. Namely, they had big tower radars all along the coast that would detect German bombers as they came in. They left their Spitfires on the ground until the attack was about to begin, and then they could put them in the right place to attack the Nazis, so it meant each Spitfire was worth many of the others. It's generally agreed that the winning of the Battle of Britain was a joint accomplishment between the RAF and the radar that was used. Incidentally, at that time, they at that time called it "radio location," as did we when we first got started at the M.I.T. Radiation Laboratory.
The British also knew that the Nazis would probably move to night fighters because the radars were proving very effective. They were shooting down a possibly large number of the attacking planes in the daytime, and in preparing for that change, the British were trying to develop night fighters. They had a crude form of them, fairly long wavelength and two matching beams to tell where the airplanes were. The pilot and the gunner worked together on this, but it was pretty crude. They realized this was indeed the case, that they needed better resolution. You couldn't have put a bigger antenna on the airplane, but you could have a shorter wavelength. So there was a big desire to get shorter wavelengths. They had started to develop ten-centimeter radar. One key thing which they had developed was a high-power magnetron which was developed by Randall, Marcus Oliphant and Boot. They developed this magnetron, and so they brought a couple of magnetrons with them to the U.S. I wasn't at the first meeting, but they came and met with some of the members of the already-then-being-formed National Defense Research Council that Vannevar Bush helped form. The head of the microwave division was a man named by Alfred Loomis, who had a private lab in Tuxedo, New York. He was a wealthy man and had a private lab developing radar. He was head of the microwave committee.
The microwave committee met with the British and were very impressed by what the British had done, and how much it meant for the war. It was quickly decided by the fairly senior scientists who were attending, I.I. Rabi, E.O. Lawrence, and others, that this was something that the U.S. should push even though we were not yet in the war. It was decided that a group of very good scientists should meet together initially for about three months at MIT and learn the techniques of radar and then go back to their own universities to develop the technique further.
One of the people who attended that meeting was Wheeler Loomis. (This is a different Loomis, not Alfred.) Wheeler Loomis was the chairman of the physics department at the University of Illinois, and therefore my boss. For a combination of reasons he urged me to go right away to the Radiation Lab, even though I had only been at the University of Illinois about five weeks. I had expected that I would probably get drafted anyway since I was the right age. During World War II they eventually did a pretty good job of making good use of scientists who were drafted, but that was not so true of World War I. I agreed to go to the new laboratory, even though we'd bought all our new furniture and expected to spend the rest of our lives there in Illinois.
Beginnings of the Radiation Lab at MIT
I left after about five weeks or so and went to the MIT Radiation Lab. I arrived there very early; I had badge number 14. I think the first eight or ten badge numbers all went to the microwave committee. I know my wife was the first wife from out of town to come.
There weren't more than about half of a dozen of us at the first meeting. Then we got together, we were told a little of what radar was about. There was a very good man from the British mission, E.G. Bowen, "Taffy Bowen" they called him. He kept providing us with information, and, of course, getting information back. I remember the first meeting; there were not more than six or eight of us. We had to decide on what we should call the laboratory. We thought we might call it Radio Location Laboratory. We never even heard of the word radar up to that time. We actually had originated from the British, and only later, after the U.S. was in the war, did we realize that the Navy had been also doing such research and calling it "radar." We then shifted, and everybody ended up calling it radar. In any case, we decided not to call it the Radio Location Laboratory; that would be a dead giveaway of something very important. We had never heard the term "radar," so we decided we would be deliberately misleading to the Nazis and also take the opportunity to honor Ernst Lawrence, who headed the Radiation Lab at Berkeley and had been one of the principle recruiters for the new lab. We decided we could for several reasons call the new lab the Radiation Laboratory at MIT. One reason being that it was radiation that we were studying, not in the nuclear sense, but electromagnetic radiation. The second reason was that it would be a little bit of a tribute to Lawrence, who had been instrumental in getting us recruited. Thirdly it would be wonderfully misleading to the Germans who might think from the name of the laboratory that we were working on something foolish like a nuclear bomb. For these reasons we chose Radiation Laboratory as the name, which stuck.
The laboratory proved to be very effective. Initially I worked with Rabi in the magnetron group developing an apparatus to measure the magnetron power. We were terribly ignorant at the time; we didn't even know much about microwave or impedance diagrams or things like that. We made some stupid mistakes in the beginning as we were developing, but we learned fast. Then I was in charge for a while of the magnetron group and later for most of my time at the Radiation Lab, I was in charge of what was called the advanced development group. The advanced development group used to jokingly say, "We are fighting World War III." We did not insist that our projects had to be ready in a short time, although obviously there was an incentive to be useful in World War II. We started working on the three-centimeter magnetrons as contrasted to ten-centimeter, and this work went along pretty well so we began developing three-centimeter hardware.
There we had good fortune from the point of view of development that new techniques were required. Although rectangular waveguides had been used before in microwaves, they were always fairly big ones. They were not much better than transmission lines, because one could make good transmission lines without too much power loss. Whereas when we got down to three centimeter wavelength, transmission lines at high power were no very good so we decided to go all out for microwave wave guides. This meant that, although there had been some developments in microwave hardware, there was not very much before our time, so we had to develop new components.
There is one that I subsequently wished that I had patented, but we didn't do patenting then. This invention was made done jointly with Shep Roberts. Previously when people wanted to connect two waveguides together, they'd just squeeze them together, put a clamp around the outside. But such joints tended to spark in between and it was sort of a nuisance and a lot of loss occurred. We decided we could make a little flange and a choke in the flange so that it would be the right impedance for the two connected together. This worked like a charm, and is still extensively used. If I only had a penny for each one I would be a wealthy man, but I don't. Anyway, I don't think in any case I would have had the penny. It would have gone to MIT probably.
Now all this radar work was really engineering. At Columbia, doing work on nuclear magnetic resonance, it was fundamental research. But in this case at MIT we were serving as engineers. At that time (not now), there was a tendency in engineering education to overly emphasize tables and rules and not so much what the fundamental principle was. This education made some of the engineers have greater difficulties adapting to totally new things whereas the physicists were a little more used to doing unfamiliar tasks. Although there were some excellent engineers at the lab, I'd say that the dominant leadership came from physicists. But we all got along very well. The engineers who liked to work in that kind of environment and physicists got along extremely well together. I would define my work during the war as engineering work, including the development of radar. We developed systems, and out of the systems we started to make our first applications. My group took on initially two applications. One was for night fighter aircraft interceptor at three centimeters, which eventually became APQ 13 something or other. I could never remember the numbers; it's probably given inside the radar book. It was actually manufactured by Bell Labs and Western Electric.
Then we also worked on an air-to-surface radar, again at three centimeters, which had a big advantage over the ten centimeters. The key interest there was detecting submarines which were proving to be a menace on the east coast. One of the main ways of detecting them was this: they could travel underwater with batteries, but they had to charge those batteries sometime. In order to do so, they would put their periscope up, preferably at nighttime, to get the air needed for the battery charging. The periscope is a very small target, but it could be detected. However, the shorter the wavelength the better defined is the pattern and hence the greater the signal to noise. So 3 cm was very much better for submarine detection. Then we also did some experiments navigating around the harbor and taking radar pictures of Nantucket, which you've probably seen. It may seem at ten-centimeters, but the picture was much, much better at three centimeters. The pictures are in the book of the Radiation Lab, named something like Five Years of Radiation Laboratory. We have pictures on that and the same thing on Cape Cod. It was quite clear that for conspicuous landmarks, we could use it for navigation, and presumably even for automatic bombing. But we ran a few experiments where it wasn't so conspicuous, and we decided if we had a suitable good map and were trained (which we weren't), we might even be able to navigate with 3 cm radar over land. We had a miserable time trying it; we kept missing. But we thought it was worth trying and it did work out eventually. It became what was initially the H2X system, and then again some official name (maybe APQ 13) when it went into production at that time; it was all engineering work.
Work at Pentagon
The laboratory management thought that the Navy was doing an excellent job of getting the radar into use. The Navy needed it badly and it had easy ways of using it. The Air Force was reluctant to use new radar and they weren't getting much ordered, and weren't using it much. So I was asked to go down to the Pentagon building with the official title of Expert Consultant to the Secretary of War. Actually my immediate supervisor, Edward L. Bowles, an engineer from MIT, was the Expert Consultant to the Secretary of War and headed the offices; but all of us had that title. I was assigned to the Requirements Branch, which had to decide what sort of equipment was needed. This proved to be quite effective; we got a lot of things done. Things weren't happening before. People were always discouraged when new radar came up for consideration. They noted that about fifty-three signatures to get it approved, and it required setting up whole new training programs, etc. So I agreed to go and said, "Well I can carry the papers around and get the approvals myself," and I did so.
Probably the most effective opportunity occurred over one weekend. A young lieutenant, I don't remember his name, came in panic into my office one Friday afternoon. He just had been asked by his superior officer, probably a captain, to prepare a first draft of the five-year procurement plan for radar for the Army Air Force; he had to have it Monday morning. He didn't know beans about radar, and so the two of us got together all weekend. I had a list of what were the possible radars; he had a list of what the airplane procurement program was going to be. We sat with our feet on the desk all weekend, and he would say, "We are going to get so and so many B-29s" and I would say, "Every one ought to have a tail warning radar and then probably one 3 cm imprecise bombing radar. But one out of ten ought to have an eagle radar which is a very high-resolution bombing device invented by Alvarey, which had a much bigger antenna." And we went through the weekend making informed guesses. I think we ended up having decided about a two-billion-dollar program, that we decided on that weekend. I was told later that basically everybody up the line said, "Oh, that's ridiculous. Some items are 10% too high and others 10% too low." But basically this was the procurement program from which almost all of the Army Air Force radar was bought. I am sure my name doesn't appear anywhere on the document, and I doubt that a young lieutenant's name was on it either because that is not the way the military operates. It was probably officially prepared by some general at the top.
Along about that time, I had been there about a year or so, some staffers from Radiation Lab came down and to propose some new radar. I found myself at a meeting sort of saying to them, "Gee, that's wonderful. But you know it takes fifty-three signatures to get them approved; you have to set up a training program." About that time I said, "Hmm, that sounds like the statement that people who are blocking things were saying when I first came down here." I therefore decided my time for effective work in the Pentagon building was about a year. I felt I really should stop fairly soon after that.
Los Alamos and the Bomb
At about that same time, Robert Oppenheimer met with Ken Bainbridge and myself in Washington to persuade us to go to Los Alamos. It was a very exciting project, so I agreed to do so. Then I was told, "Don't say a word. General Groves has unlimited power, and you will be immediately be detached from your present office." But what he didn't realize was that Edward Bowles had similar power. Each of them reported only directly to the Secretary of War. I had learned out of experience that when one was in powerful position, in which the person can get his way always, you must never lose a battle. If you lose one battle, you probably lost future battles. Each of them I think felt that way, so although I was asked to join the lab somewhere around March 1943, before the lab at Los Alamos was even started, I didn't hear a word officially that I was to go. I also had agreed with Oppenheimer that I would start some tests at the Dahlgren Proving Ground about shapes of possible bombs. I had Dahlgren make so-called sewer type bombs. It looked like a sewer pipe welded together to each end, a small-scale model of what might have been dropped to see if they came down head over heels, or what happened. The answer was some of them did come down head over heels. I supervised these tests, part time maybe half a day a week.
Well, July came along and I still had no word. Just as my wife and I were leaving for a three-day vacation in the Shenandoah Valley, I got a phone call at home from Bowles, who finally had to say something. I think he expected me to say that I was willing to stay, and so he was very disappointed when I agreed that I probably should go. Then came the problem of a "face-saving" device, which was invented: namely that I should go to Los Alamos and be a regular staff member out of there, but that I should be paid out of Bowles' office and be officially an expert consultant at Los Alamos, in principle reporting back to the Secretary of War if things were badly run. If I had done that it would have undercut the lab, and I didn't. I didn't expect to either. So all during the war I was officially not an employee of the University of California, as all the rest of the people were, but was with the Secretary of War.
They wanted me to take advantage of my experience with the Air Force in both airplanes and the operations of the organization. So I was put in charge of so called "Delivery Group," which had the responsibility of converting a device that could just have a nuclear explosion to one that could fit in an airplane and be dropped without breaking apart. It was primarily an engineering program that I directed. My group managed about fifty percent of what later became the Sandea Corporation. We had quite a small group to do all the things that they were doing. That was my principle activity. I was also the liaison with the Air Force. We set up program for all the necessary testing at Los Alamos. We had three test B-29s assigned to us — one test B-29 assigned to us initially — and then we supervised a bunch of tests that we had at Muroc Dry Lake. In those tests we also made a discovery important for the war in Europe. We had high altitude airplanes. We wanted to test for bomb shapes to see if they would fall head over heels. As a control for our tests we dropped a bunch of standard five-hundred-pound bombs, the bomb most used in Europe, and to our surprise they missed the target terribly. But we had high-speed cameras photographing them, and we could see that the tail fins bent on the way down. When they were dropped from fifteen thousand feet, no problem, but above something like the fifteen thousand, the tail fins distorted and the bombs tumbled. These were the standard bombs that were being used. Parsons, who was the head of the Los Alamos ordinance division, felt at last we could contribute something to the war on-going by this discovery. So this was reported through channels to Washington, and only a year or so later did we discover that for atomic bomb security reasons this message had not been passed through to the right people. It was eventually discovered independently about a year later by the 8th Air Force, in observing how badly the bombs were falling. So excess security can do harm.
We continued test programs at Muroc but then when the 509th bombardment group was organized, the airborne test program was attached to it. I managed the test program and other activities there. Initially we did work at Wendover, Utah, that was the 509th base. As the project advanced, my group had to make arrangements to get the right equipment overseas. Initially we did not know where we might go so we prepared three different so-called kits. The kits included eight or ten buildings and many things under what we called "Kit bomb assembly." (We had one set up and then it became apparent that the first one, and the only one, got set up was at Tinian, in the Pacific.) Dick Ashworth was my deputy and a Navy commander, and he then went out early to establish a base in Tinian. He got us the Seabees former camp which was very nice for us. We had to build special buildings and special air-conditioned assembly buildings out there. I stayed at Los Alamos up through the Trinity test at Alamogordo but then left the next day for the Pacific. I supervised the technical part of the bomb assembly and had to report on the reliability of the assembly operator. Both bombs brought an end to the War and an end of my atomic bomb engineering activities.
Brookhaven and Harvard
I then returned to Columbia with Rabi to get our physics experiments started there. We also helped establish Brookhaven National Laboratory, which we wanted. Rabi and I were essentially the inventors of that laboratory. After a year or so at Columbia I became head of the physics department at Brookhaven for around six months: half-time. I was also a professor at Columbia, then I was lured to Harvard, and I have been here since 1947.
When I got to Harvard I thought I knew how to make a molecular beam resonance apparatus (every student thinks he knows better than his teacher) which we were just using for about ten times better than the best previous by making the apparatus ten times longer. But I knew that this was not easy to do because we had made our apparatus a little longer at Columbia, and it got worse rather than better. That's because the magnetic fields were not uniform so the resonances would be one frequency here and another frequency there, and the whole thing would be messed up together. But I thought I knew how to make the magnetic fields better, so I set out to do so with my first three graduate students at Harvard. After Los Alamos, Robert Wilson came to Harvard and then he was lured away to Cornell and I was probably appointed in his position and perhaps as a result of that, I was asked by the department to be also Director of the Harvard cyclotron, which he had designed during its construction and early operation. Bob Wilson was a very good designer, and so I took over that. Lee Davenport was a very excellent deputy; we got that running. In fact, I think this is now probably the oldest and largest cyclotron that is still operating. It is now used for proton therapy, particularly for brain tumors, and it is still probably the most effective treatment for that. Hospitals are now building cyclotrons just for medical purposes. In any case, I was involved with building the cyclotron along with developing the molecular beam apparatus.
Separated Oscillatory Field Method
Then I began having a problem on the molecular beam apparatus. My method of making the uniform field didn't seem to be working very well, and this had me worried and wondering what to do. It suddenly occurred to me that maybe instead of having the oscillatory magnetic field extend throughout the magnetic field region, if I put it in two lumps at each end, then we would only see the average magnetic field, which would be fine; if it was just the average that determined the frequency rather than having the resonance at one frequency in one place and then a different frequency at another. I tried the method and it worked. This is the basis of the separated oscillatory field method which indeed proved to be much better. I guess like most inventions when they are really successful, they always turn out to be better than they were expected to be. Usually when you have a new development, sometimes it doesn't work at all, or if it does work, it usually works worse than you think. Once in a while there is one that works better than it was supposed to. This was the case here, because I realized, after inventing it only for the purpose of averaging the magnetic field, that in addition it also made the resonance lines intrinsically a factor of two better in addition.
It also had the benefit of getting rid of second-order Doppler shift when you made two resonance fields at the same frequency and with no phase shift between them. There was no second order Doppler effect. It also enabled us to work at very much higher frequencies for the same accuracy. If you used a Rabi method single-oscillatory field, when the wavelength of the radiation was shorter than the length of the total magnet, then the oscillations in different regions were out of phases. On the other hand with the separated oscillatory field, all one had to do was to have coherency in each of the separated regions, which could be very short. Those had to be less than a wavelength and this was easy. There were several other advantages. Everything seemed to work in our favor on that one, so this method proved to be very effective. My initial interest in that method was not for engineering but for fundamental measurements of atomic and nuclear properties, but it produced the best atomic clocks and frequency standards.
But when you get involved with a project like that, do you find yourself needing to address engineering concerns?
All the time addressing engineering problems. In a certain sense the separated oscillatory field was an engineering problem, namely it was a problem of trying to make the field uniform and not succeeding. It was a different way of accomplishing the same goal; I would say, "Yes, it is very heavily engineering." Though my motivation was at that stage was pure physics.
When you say that sometimes results are better than you planned, it sounds like what you do is try to solve a specific technical problem. You are not always thinking in terms of deep principles and then you get some fortunate fallout.
That's right. This one was better than I expected in two ways, well, many more than even two ways. One of the ways was, it just worked better than I thought. It had these advantages: you could use it shorter wavelengths. In addition to the purely scientific reasons for which I invented the method, it also had many engineering applications. It had applications to accurate frequency standards and atomic clocks. I didn't even know there was a problem about clocks initially. My wristwatch was pretty good, even though it was not a quartz watch at that time.
Experience with Electrical Engineering
I am not sure that I caught where you became acquainted with some of the specialized knowledge of electrical engineering?
I would say probably that depends on the specialized knowledge of electrical engineering; I probably learned this at the Radiation Lab. I first learned about impedance diagrams and standing waves there, and learned about it the hard way. I don't know if the first Radiation Lab device is still around, but it should be a trophy to stupidity. The first Radiation Lab device actually designed at the lab was actually manufactured at Harvard because there weren't any machinists available to us at MIT. It was designed jointly by I.I. Rabi, Ed Purcell and myself. It was to measure the power on the first British ten-centimeter magnetrons. We were going to do this by having a resistor in it, which was water-cooled, a little hollow tube with conducting materials on the outside. We did know enough about transmission lines to know that there were both a reactance and a resistance — two components that you had to worry about.
So we felt that there ought to be two adjustments of some kind in order to get increase maximum power into the device. We made it and did indeed measure the power, and we got pretty good measurements out of it. We thought for awhile we had made another interesting discovery, which was that you only needed one adjustment. Anywhere we set one stub tuner, we could do equally as well by balancing it with the other. It was another two or three weeks before we realized that by total accident we had put these two stub tuners exactly half a wavelength apart, so whatever one did, the other could indeed do the same thing. In fact, to do the tuning with a tuner at a fixed location it takes three adjustments to tune, or two and moving it along in position. It showed we started toward total ignorance, but we then soon found —
Actually, the thing that was very helpful to us was E. U. Condon, the physicist who wrote Condon and Shortley's Theory of the Atomic Spectra. He came to the lab and wrote a guide to microwaves for us on ditto paper. We read through that, and we saw what mistakes we had been making, and began to learn about microwaves, and then learned more about other electronics. Then we learned how to use impedance diagrams, and we then learned about the circular diagram and how to use it and used it very greatly. I think within six months probably, there was more knowledge about waveguides at MIT Radiation Lab than any other place, but there was surely ignorance when we started.
Were all the physicists at Rad Lab able to adapt so easily?
I suppose some of them were pretty good about adapting. We had a regular colloquium and when one of us learned about impedance diagrams he told the other. I am not the one who discovered the impedance diagram; somebody else did.
Relation of Physics and Engineering
Was it an effort to suspend your interests in investigating what the fundamental phenomenon involved?
No, I don't think there was any effort to do that. Basically, the war was on and going badly during most of that time, and so there was plenty of incentive to try to get better weapons. I think everyone was really interested in something that would be applicable soon. In fact, our group (Advanced Development) got a certain amount of kidding for the fact that we were probably working on World War III rather than World War II. And World War II was more than we could handle. No efforts were made to discourage fundamental research, but nobody was doing it.
So it sounds like you're claiming there is not much difference between, or you see many similarities, I should say.
Well, many similarities, but there is a difference in motivation. I would say that is a key difference; the difference in motivation can change with the given person. The basic motivation, probably, of a physicist is to understand something or discover a new phenomenon, but basically one that adds to understanding. And also, yes, if it's fundamental, it usually has applications. A good engineer has his eye more clearly on the goal. By and large, we physicists don't usually have our eye so much on the goal. Now when you once have an idea, let's say developing a separate oscillatory field method, you obviously put your eyes on the goal of that. But in having the original idea, it isn't so at all. I was on the goal of trying to make better measurements. So there is that difference. The measure of knowledge that you have to carry is really rather similar in the two cases. In the case of the physicist, he needs more fundamental atomic physics, theory of quantum mechanics and whatnot, which the engineer does not need so much. He needs some quantum mechanics, a lot of things are quantum mechanical devices nowadays. The key difference is almost one of motivation and of course, what your background training has been. But I think the good researchers in both fields, when they have reason to want the other background and if they have changed their motivation, I think they do very similar things.
Creativity in Physics and Engineering
So you don't think that there is a different aspect to the creativity of the individuals?
I think there is a lot of similarity. Creativity is a very strange thing. That's what we were talking about in my last interview. In the first place, different people create ideas in very different ways; it is just a matter of individual tastes. It is not so much whether it's an engineer or not. The fundamental creativity process is not always rational. Knowing whether to throw an idea out or keep it is a very much a rational process. But to create a new idea you need something to stimulate you to get you thinking in a new direction. In the case of my separated oscillatory field method, I was stimulated by this worry that I was probably not going to succeed in making my field uniform enough, so our apparatus was not going to be as much better as I thought it should be. I didn't promise people I would do that, not to worry. In addition I needed a stimulus to make me think in a different direction. That is one of the key problems for creativity in any field. There is a tendency to always think in the same direction. If the idea is going to be creative and new, it requires thinking in a new direction. It is not easy; you need to stimulate yourself somehow to do it.
In the case of the separated oscillatory, it actually got stimulated by my giving a course in optics where there is a device known as the Michelson Stellar interferometer, which one of my professors at Cambridge University described in a rather dramatic fashion by saying, "If you had a telescope and were looking at a star that is very bright, so you don't have to worry about light gathering — and if you didn't have quite enough resolution to tell whether it is a singular or double star — if you take a can of black paint and paint over the middle of the telescope, get twice the resolution. You could then tell whether it was single or double." I was actually giving this somewhat rather dramatic discussion of it to my class, and it suddenly occurred to me that although I was happy to get twice the resolution, I wasn't very much concerned about that. But it occurred to me that if you paint over the middle of the telescope with a can of black paint, it must not depend very much on the quality of the glass underneath the paint. Maybe I could do something analogous that wouldn't depend upon the quality of the magnetic field, if I put radio frequency fields only at the ends. I wasn't clear what the application would be, but it started me thinking in that direction. It's miscellaneous things that make you think in a new direction. Sometimes it can be jokes you think of, something that just seem funny, and then maybe it's not just funny. Also I think it requires effort.
One of the most inventive physicists I know for finding new ideas — many of them — is Luis Alvarez. He wrote an autobiography which I wish I had read thirty or forty years earlier, in which he said that his father, who was a famous doctor at the Mayo Clinic, gave him some good advice. (His father had a column in the newspaper on health before such things were so popular.) His father told him when he was going into physics that he should spend at least one evening a month just sitting in his most comfortable reading chair and just thinking — not reading, just thinking. Thinking about what he is doing, what the meaning of it is, and what he ought to be doing, what is the interpretation? Thinking. I believe it is a very good advice, because thinking is hard and you have to do something to put yourself into doing it. Sometimes it can occur by accident, sometimes it takes hard thought. I think this is true. You have to get a little untrapped from too much prior knowledge.
Physics and Practical Applications
A fascinating question is how science and technology are related; can you comment on that based on your experience as a producer of both?
They are very closely related. I think most of the technological things in the fundamental sense have come from science. That has usually been a totally new direction in which you look, because that is not what an engineer ordinarily should be putting much effort on. I think the next steps of the beginnings of applications can come from both. I think most good scientists are quite happy to see applications of their work. Though I am primarily a physicist, I enjoy the fact that my methods are used in atomic clocks. I have even published papers with the atomic clock application. I write about it and I don't avoid it. I am willing to take a detour on practical application.
That's something I'm curious about: do you have an instinct about how far you are willing to go to in reducing something to practice?
I guess it is a matter of time. I would be happy to do it all the way, if it is not going to take too long. But I have got a lot of other things I want to do. I don't want to do applications so much that it starts interfering with my fundamental work, so that I am very delighted to have someone take over. On the other hand, when I had an idea I wanted to push along, at various intervals I have gone fairly far in that direction. It depends also on the willingness of somebody else to pick it up. Sometimes no one else is so inclined so I had to pursue it a little further or it would not have been picked up. Even though the main reason for doing it is more engineering in direction. Ordinarily, I think the difference would come that if someone else is going to pick it up and would probably do it pretty well, I am happy to withdraw. That's been true of all my projects.
I have been involved in several other time-keeping projects like the atoms hydrogen maser. Now that project was interesting also. When I invented the separated oscillatory field method, I think atomic clocks were furthest from my mind. However, by the time I invented the hydrogen maser, I did know about atomic clocks. I realized that there was a need for greater stability, and I was probably partly stimulated by that need to invent it. Well, again it was fundamentally to study hydrogen, which is a very interesting atom to study. Also there was the fact that it might even be a more stable clock than cesium. I am interested in both, but I would say it has been years since I have contributed in a major way to the application. I have written about them and encouraged some of the later developments, particularly the more fundamental ones like laser cooling and things of that kind. If I think of something useful, I tell my good friend of Bob Vessot. If I think of a better wall coating than Teflon I will write him a note, saying that he ought to look into it!
When you look at a case like the laser or the transistor, do you think these are clear-cut scientific innovations or do you think that they belong more in the engineering tradition?
They are both. At one time I think I had personally in my possession all of the really successful, solid-state electronics devices in the United States. This was when I came back from England, and we had sent Radiation Lab radar to England. I went there primarily to tell about our work in three centimeters. Dale Corson and a very good friend of mine had taken a U.S. ten-centimeter radar to show how much better it was than the British radar because it had a ground-to-grid triode which would show them that radiation lab system would receive signals further away than the other. They set up in Harwell, next to one of the British radar, and they didn't see receive signals three times further away than the British. Instead, the British received signals three times further away. I worked with them a little bit while they were there and it soon became apparent what was happening. There had been a U.S. test showing the ground-to-grid triode better, apparently, than a crystal for detection; this was because that test was done with crystals that had been burned out by having a poor TR box. I was the first one to go back to the U.S.; I was given by the British three or four good quality crystals and a couple of TR boxes which I carried in my pouch. We had to go via Lisbon at that time, a little bit disconcerting with it in my pouch, changing planes with the British overseas airways being the next gate to Lufthansa. Nevertheless, everything came out fine and I came back with those crystals.
At this point crystals began to be taken seriously in the United States, but they weren't understood at all. The British were ahead of us on that one. I think the next step arose during the war. There was a lot of effort to try to understand crystals, but it was not very successful initially; it was still pretty much witchery. The reason was that it took extremely high purity to make good crystals. It is very minute amounts of impurities but not more than a certain amount of impurities that was the difference between N and P type. All of these things weren't known then. With that understanding, then it became possible for the next step, which was a transistor. I think that was done at Bell Labs, probably with more engineering motivation. At least one of the key people in that, Bardeen was primarily a theoretical physicist, and Shockley was a physicist. But certainly they had engineering interests. I think by the time crystals started being understood, it was a fairly obvious thing that maybe somehow one ought to be able to make a transistor.
They were exploiting crystal properties that weren't understood in the interest of developing their device. The fact that you see people doing that — physicists like Bardeen were — make it difficult for me to really see what the acid test is to know if someone is an engineer.
Well, I don't think that there is an acid test. I think both have too many similarities for there to be an acid test. It was a very interesting problem of fundamental physics as to why you get these differences of crystals? Some would work well and some wouldn't work well, it seemed to be magic. I think that anytime there is magic, there is an interesting fundamental physics problem as to what it is and it is an interesting applied physics problem if there is going to be some use for it. Both of these were together on this one.
In your own career can you think of examples where you have come on situations like this where you have to weigh the needs of getting your apparatus to work versus understanding more deeply what's happening?
Yes, all the time. And in fact one of the arts is picking out what you want to when a new device like the hydrogen maser first gets it going. There are very strange peculiarities in the process of having a device like that newly turned on, it probably relates to chaos phenomenon and whatnot. The peculiarity may be an interesting fundamental thing to look at, on the other hand it may not. There are many things for which one just has to guess. Many people don't realize the extent that is true. They think of science and engineering both as very exact subjects: you just straightforwardly plod ahead and do it. Many people don't understand why different scientists argue with great vigor as to whether something is a good device or a bad device or whether one should build an SSC or shouldn't. The real problem is when it is something new, nobody really knows. And you just have to guess. Now some people guess better than others.
Speaking of the SSC, you know one common argument for a project like that is the spin-off technology. It rose up from fundamental research. The atomic clocks are a wonderful example from your own career. Are there other examples from your research?
Well, yes, many of them. For example, much of the computer devices got started from particle physics. The first flip-flops were actually used for nuclear physics, counting radioactivity, Williams and others; these are now incorporated in computers. A big push for pattern recognition and high power computing has come from particle physics. For example, data transmission has been pushed by particle physics. Everybody is talking these days about the Internet and electronics communication; it is not at all surprising that the WWW, the World Wide Web, is centered at CERN which is a nuclear physics establishment. They had to get a lot of data out to people, and they were data experts. Now inversely, the particle physics community has benefited tremendously from the engineering developments. In computers, one of the biggest users and biggest stimulators in particle physics.
Let us take the Harvard cyclotron as another example. This was developed for nuclear physics and it is no longer interesting for the nuclear physics. The physics department still runs it as a favor to the medical school and it is used very extensively for radiation therapy. Another example is magnetic resonance. I was involved with magnetic resonance in its earliest phases. I also worked papers on its applications. I wrote the first theoretical paper on chemical shift, it is the chemical shift that provides chemical analysis; you can identify different chemicals with it. I wrote half of a dozen papers on that subject. The first one was called nuclear magnetic shielding in molecules. If you have a molecule and you put on a magnetic field, the magnetic field at the nucleus is not the same as the field you apply externally. Because when you put on a magnetic field it will induce a current. Now if the molecule is a pure atom, it is very simple. The current is just a circulating current, and Willis Lamb, before World War II, had developed the theory of the magnetic shielding in an atom by calculating the magnetic fold from the electron circulating around when you put on the magnetic field, which provide an opposite magnetic field.
I realized that on-going molecules that the shielding would be different for the same atom in different molecules. When I wrote the first paper magnetic shielding. I had a nice general theory, but it was a little hard to calculate numerically, and in any case we were stuck with a problem, that we could make measurements of the resonance frequency better than a part in ten million, but we had this correction which you couldn't calculate with better than a part into a hundred thousand. So you lost accuracy. And worse still it was even different in different molecules. In turning that around in the other direction, that means that the shielding enables you to tell what molecule you are looking at, and identify it. That has been really the big power of the NMR, this nuisance thing from my original point of view. The change of viewpoint shows, in the naming of the two papers I published a month or so apart, the first one I called, "Magnetic Shielding" and the second one I called, "Chemical Shifts."
My last big engineering activity was more on the management end. I was for sixteen years president of the University's Research Association that constructed and operated Fermi Lab, which is a big facility. It was the biggest accelerator at that time and a very successful one. Incidentally, it had an interesting combination. I mentioned that Harvard cyclotron was designed by Bob Wilson, and I was in charge of the building of it as director at the lab after he went to Cornell. That was the first instance that we cooperated. We almost forgot because we never talked to each other about it since he went to Cornell before I came to Harvard. He had done his bit first; we did our bit next. But the two of us did cooperate in a totally different fashion many years later when I was the president of the University Research Association that operates Fermi Lab, and we hired Bob Wilson, an excellent man, to be the Director of the lab.
Then we collaborated very closely there for the construction of a very big project. That was dominantly an engineering problem for fundamental scientific purposes. There also had to have fundamental scientists in the project because you want to build the right kind of machine. The real engineering was in a certain sense done by Bob and his staff; I was more of a manager which was a good arrangement. Bob is a very clever designer. I think I am reasonably good at ideas, and reasonably good at appreciating good design, but I am not very good at drawing pictures or mechanical designs. Bob is a professional sculptor as well as an excellent physicist, and he just has a talent for not only making machines work well but also making them look good. But I can appreciate that, and so the two of us worked very well together. In fact that was the phone call that interrupted us just now, it was from Fermi Lab.
Developing Experimental Equipment
When you were talking before about the influence of equipment for physicists, I was wondering how you have handled this problem in your career and maybe with people you have worked with. If you have an idea for an experiment that you would like to try but you don't necessarily have the equipment that you need at hand — certain detectors aren't accurate enough — what are the range of possibilities? Will the physicist often launch out and try to develop the new and improved?
They will if necessary. I think the physicists are very willing to do whatever is necessary. They are prepared, in a certain sense, even to do many things badly. They can do many things, but they don't always do them that well. I try to do them reasonably well, but mainly so they work. A large fraction of my equipment I have built myself or my students have built it. But also for many things, we buy them. For example, things that are purchased are computers and computer-controlled experiments. Nobody would dream of trying to make his own computers for that purpose. If you can possibly afford it you buy them.
You spoke before about missing out on the patent for the joints and the waveguides. If you develop some equipment in order to use it in an experiment, do you make an effort to pass it on?
Not very much. I make an effort to pass it on, publish it and whatnot, but not much effort to patent it. In some respects, patents are not all that good for really fundamental developments because they are good for only something like fourteen years. Ordinarily it takes about fourteen years before the invention really takes hold. I didn't patent anything on the separated oscillatory field; I didn't patent anything on that. I did have some patents during the war because the government insisted on it. I can't imagine anything that could earn me less money than my patent, which has now expired, on a tail fin for an atomic bomb. Quantity production of that is low.
But I did actually patent the hydrogen maser along with Dan Kleppner. We got a little bit of money out of it, but it was about fourteen years before it was widely used. We said right away, "It would be great for radio astronomers, they should use it." Well, they said they had never paid more than twenty thousand dollars for a clock; they couldn't see their way to doing more, but now they all have hydrogen masers. However, the extensive use began only after our patent expired. I am not sure on this, but I have a feeling maybe in software there is an advantage, because I think you can get a copyright there that is good for a long time, maybe a hundred years.
The rules keep changing, but I think that could be over sixty-seventh the longest time, maybe a hundred now.
Well, no, I think they have been probably shortening it if it is changing. Anyway, it is a long time compared to fourteen years. So for fourteen years it really has to be something that has very immediate application. As far as I can tell from the history of the laser, it helps to have invention under dispute for most of the time the product is being developed and finally be settled in your favor after twenty years of debate. Then it has another fourteen years to go after the patent is finally granted.
How does the approach differ when one is working out a physics problem as compared to an engineering problem? One hears a lot about the importance of physical intuition when you are thinking about physical systems. If the same sort of intuition harnessed when you are working on an engineering problem?
I think so.
Do you maybe visualize an analogous system?
Yes. And the main things in both of them are probably just ideas, something new and different to think of. It is physical intuition that helps you, but just physical intuition alone isn't enough, especially if you don't think of anything new.
Research on Molecular Beam Devices
Can we cover some of the research that you have done in the 1960s and 1970s?
Okay. That was a very active field period. Basically, my primary basic research was being done with these precision methods I had developed. I was really using the devices, using these molecular beams devices we measured interactions in molecules. I discovered a number of new interactions that hadn't been previously been observed and how these interactions changed with temperature, how they changed with the vibrational, rotational states of the molecules. We did very fundamental tests looking for a pair of symmetry properties of particles, of so-called parity and time reversing symmetry. Everybody assumed there was parity symmetry, that is left-right symmetry. In our paper, Purcell and I jointly pointed out that in fact parity symmetry had to be tested. It wasn't obvious that it was true for nuclear force; we did an experiment, but it didn't show up. We looked for a neutron electric dipole moment, but didn't find it. It turned out that there are two kinds of nuclear forces: we were looking for a minute effect on the strong forces, but we should have been looking for a huge effect on weak forces. We had bad luck, but it was a good fundamental search and started a new direction.
And the same thing occurred in my test of time reversal symmetry. I directed the Ph.D. thesis research of about eighty-four graduate students. Some of them did routine measurements, very valuable but slightly more routine of molecular properties from which we could then work out molecular and nuclear structure. Then on occasion we looked for some fundamental new forces. For example, we were able to look for a long-range tension force between two particles.
Are there performance improvements in your equipment in apparatus that made certain measurements or experiments possible?
In the first place the separated oscillatory field made a huge difference, and then there were other improvements after that.
Once you developed the separated oscillatory field, were important improvements in performance after that?
Well, yes. There were important improvements and better ways to interpret the results. We could use computers and also do analysis with computers.
The data reduction tools?
Not only the data reduction, but theoretical analysis of what you have, theoretical interpretation. We are now still carrying on with a neutron electric diode pole moment experiment in Grenoble, France using very slow neutrons as a test of time-reversal symmetry and that's all automated to run by computers. There are many changes in techniques.
Engineering Literature and IEEE
Thought of another way of getting to this question of how engineering relates to your work as a physicist. That is to ask you what sort of papers you find useful to you published in engineering literature, in IEEE journals. What do you get from that literature?
That is a good question, and I guess there are two answers. Basically the main journals I read are probably not engineering ones, they are mostly Physical Society journals, science magazines, biology. I am interested in all of science, and of course that includes some engineering, but not much. And I must admit that I find just for general interest, and also needing to keep up, I do read the IEEE Spectrum. I read that on the subway coming in. They had a quite interesting article on a subject I was not up on at all; I read the whole story on the high-definition television and the new grand alliance system being worked on here, which I found very interesting. I need to know that kind of background. Sometimes out of that I get a new idea. I never can tell what will give me an idea. Other than that, I am usually looking up explicit articles. I frequently look up things in one of the IEEE journals, but usually it is for something specific. I don't read it with regularity. I often read transactions pertaining to time and frequency.
Are there any questions describing analytical tools or perhaps new devices?
In the case of IEEE, it is most frequently new devices.