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

Oral History Transcript — Dr. Martin Harwit

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. Martin Harwit
By David DeVorkin
At National Air & Space Museum
June 20, 1983

open tab View abstract

Martin Otto Harwit; June 20, 1983

ABSTRACT: After surveying Harwit's (b. March 9, 1931) family background and early education, the interview concentrates on his graduate education at MIT (PhD, 1960, physics) and his career in physics at Cambridge University as a NATO Fellow (1960-1), at NRL as a NSF Fellow (1961-2), and, principally, at Cornell as assistant and associate professor of astronomy (1962-8), professor (1968- ), and chairman of the department (1971-6). The interview covers a broad range of his scientific interests: galaxy and star formation; comets; infrared optics, especially relating to detector technology; infrared astronomy; rocket astronomy; history and
philosophy of science; and educational astronomy.

Transcript

Session I | Session II | Session III

DeVorkin:

Dr. Harwit, the last time we ended with your discharge from the service in January, 1957, and your move to MIT. Your were about to recall your experiences in an astronomy course at Harvard under Cecilia Payne Gaposchkin.

Harwit:

The course actually was a seminar that was to be given jointly by Mrs. Gaposchkin, Dr. Thomas Gold, who has just come to this country from England, and Fred Whipple at Harvard. The first lecture I attended was given by Mrs. Gaposchkin, who was phenomenal in her recollections of star names and details of the light curves of different variable stars. She spent the whole hour and a half or so that particular session talking about these different stars — if I remember correctly, without any notes at all. This was very discouraging for me, because I had never had any elementary course at all in astronomy, and it made me feel that all of the other graduate students who were sitting in that audience must of course know what all this was about.

Since my interest in astronomy had been brought about by reading Fred Hoyle's popular book, FRONTIERS IN ASTRONOMY, which was really an astrophysicist's book rather than an astronomer's book — that is, a book that concentrating on the physical processes, and understanding of what was going on, rather than worrying about peculiarities of the individual sources — I thought that perhaps I had been mistaken in deciding I ought to be able to do astrophysics on the basis of that book, and that there was much more to the subject, that Fred Hoyle had perhaps diluted it down to such an elementary level that I had completely lost perspective of what the field was all about.

With that sort of depressing thought in mind, I started walking down the hall, and came to the office of two students whom I had chatted with briefly before. One of them was Frank Drake, and the other one was Nannilou Dieter, who were sharing an office at the time, not far from the lecture room. I mentioned to them that this probably wasn't for me after all. They said, no, no, I should really give it another try, that the following week Tommy Gold was going to be arriving. He had spent, I think, a semester at Cornell University before, and he was an astrophysicist, and had been a colleague of Hoyle's at Cambridge. He most likely would conduct his seminars, at least, in a way that would be much more understandable to me.

So I did go back to all the subsequent seminars that semester, and I found very soon that I could understand Gold and Whipple well, and even Mrs. Gaposchkin's work seemed less formidable after I had found out where I could read up about these subjects. I decided to stay on, and later took further courses. I had to take three courses for credit at Harvard in order to get a minor degree satisfied at MIT. Then later after I was done with courses, I also started working on a small problem with Tommy Gold, who had become a professor there, replacing Bart Bok, who had gone to Australia at the time.

He was interested in the surface of the moon at the time, and transportation of dust on the surface of the moon, to smooth out the surfaces and fill in craters. Since there is no atmosphere on the moon, there was some question about how dust would be transported. There is no wind; the erosion processes that act on earth wouldn't be active there, no water and no wind.

DeVorkin:

Was this part of your course work?

Harwit:

No.

DeVorkin:

When was this actually?

Harwit:

This was the following year, I think sometime in 1958. Gold, I think, only stayed at Harvard through the spring semester of 1957-58, or maybe 1958-59. I think he was there only a year and a half, or two and a half, and then was called back to Cornell to head up the Center for Radio Physics and Space Research there.

But in the meantime I had another course with him on magnetohydrodynamic problems, which were interesting him at the time. He was very much interested in the earth's magnetosphere at the time. I started begging and borrowing apparatus at MIT that would allow me to test an electrostatic transport mechanism of dust on the lunar surface. There was some question about whether the solar wind pounding the surface of the moon could charge up grains occasionally to the point where they would repel each other, and would become airborne, and then would float down slopes into the craters to fill them up, or into the rills and fill them up.

This was an experimental effort where I built a little vacuum tube on my own. I was very proud of it, actually, at the time; I blew the glass and bedded the electrodes in it and everything. Did it from scratch. It was really quite nice, and at MIT they had all these facilities where you could do that. Then I asked for permission to use their vacuum setups that they had there, pumped out this tube, and placed dust in it, then applied the voltage and the whole thing went, puff! You could really see these particles — I had put some graphite particles in there, and various things like that.

DeVorkin:

I'm interested in what you based your model of the lunar dust on? Hanait: Well, I tried various different things. I think, if I remember correctly — none of that was ever published — but I think I took some olivine and made olivine dust out of that just abrasively. I think I also took some soot that I deposited on a cold glass surface from a candle or something, and scraped that off.

At least some of the materials seemed to move apart when you irradiated them. Often, though, you could only get the effect to work once. You would charge it up and the whole thing would go, poof. But if you did it again, it wouldn't work, because I didn't have any way of discharging these things quickly, because I only had electrons to do it with. It was an electron gun really that I had built. But I recall Tommy Gold giving me a call one time. He had never seen this. I had to tell him over the phone what was happening, and I had gone up to tell him about it at Harvard.

I guess I was working down at the lab one evening. I don't know whether I was working on my thesis topic or what, but Marianne, my wife, called me, and said Tommy Gold had just called from a party where he was, and he was leaving town the next day, and was going to be away for the summer and then would be up at Cornell, rather than back at Harvard. He really wanted to see this, so I said I'd go and set up the system. I think he appeared around midnight and I let him in the door, down at MIT. He came in and I demonstrated this effect, and he was quite impressed by it.

DeVorkin:

He wasn't your advisor then? Who was your advisor?

Harwit:

No, I'll come to that in a minute.

DeVorkin:

This was not part of your thesis?

Harwit:

This was not a thesis. No, this was just a sideline. I had scrounged around and gotten all the equipment that was needed. The high voltage supply and stuff like that. At MIT it was really very nice, because you could borrow all kinds of things. I also was interested in discoloration effects, and so I was able to get people to put samples of olivine into the Van de Graff, and into the cyclotron beam there to see if they got darkened or not. Some of them did, and then there was the question of whether it was the heat that did it or the particle bombardment, because the time scales are different for the solar wind and for these particles. The energies were the same, but in order to get a reasonable total darkening, you had to of course contract millions of years' worth of time into just a few seconds worth of beam, and so I wasn't sure whether the beam just heated up things too much.

DeVorkin:

So all of this dealt with the nature of the lunar surface?

Harwit:

That's right, yes.

DeVorkin:

I know Tommy Gold had his theory of the prehistoric glazing, wondering whether the surface of the moon reflected the past history of the sun.

Harwit:

Possibly, yes. I don't remember the glazing, but certainly the darkening of the lunar surface and the smoothness of the dust distribution, the smoothness of the craters, suggested erosion processes, and those were the ones we were looking at.

DeVorkin:

Had you been reading his papers by this time?

Harwit:

Oh, sure. Yes.

DeVorkin:

So what was his reaction to your apparatus?

Harwit:

Well, he was very pleased about it, and in fact, said he would be going to Cornell, and he would like me to consider Cornell, if I needed a job after getting out, which was quite pleasing actually, because I still had quite a long time to go on my thesis work.

DeVorkin:

Did you have any other contacts with astronomers or astronomy at Harvard?

Harwit:

I used to see people up there. I met Whipple through this course, and a lot of the students, for example, Richard Teske was there at the time. He later went to Michigan. I lost track of him. One person who was very influential was Larry Mertz, who was extremely interested in optics, and one of the first people in this country to do Fourrier transform spectroscopy and to take it seriously.

Although a student at Harvard, he was also vice president of Block Associates, a small company that Myron Block had built up in Cambridge. Myron Block was a consultant for MIT at the Naval Supersonic Laboratory there. Since I needed a job for the summer, somehow I was put in contact with Myron Block, I think through Larry Mertz. I don't remember exactly how that went, but at any rate, I started working for Myron Block that first sunnier I was at MIT. That was the summer of 1957-58, and there were three people in that company at the time. Larry Mertz was not working actively.

There was a young guy about 25, 26, about my age, Fritz Zernicki, who was Dutch, and the son of the physicist, Fritz Zernicki, who had won a Nobel Prize for optics, and who was dead by that time. So Block, Zernicki and I were l all working in a former grocery store not far from the Harvard Observatory on Sherman Avenue, I think it was, not far from the firehouse there. The Blocks lived upstairs in this ramshackle house. But that was the beginning of what eventually became a rather prosperous optics firm, Block Associates, which then was bought out and had offshoots, called Block Engineering and Digilab. These were small sub-companies.

So, at any rate, I started working that summer on an interferometer that IBM had commissioned, a Michelson interferometer with retro-reflecting, corner-cube mirrors that were thought to perhaps lend the instrument greater stability under vibrations, something that the IBM people were interested in, because they were thinking of possibly using an instrument like that for running milling machines very accurately by counting fringes. I put that instrument together.

DeVorkin:

This wasn't explicitly a space interest.

Harwit:

It wasn't a space interest, not at all. I built that interferometer and the main problem was, every time a truck passed by outside, the floor, which was a feeble wooden floor, shook so much that the fringes would vibrate all over the place and get lost. And we would have to wait until things calmed down again. There was a work bench which I was using which was an old milk cooler, one of these old ice chests with a sliding top. The whole thing was fairly primitive. There just wasn't enough money around for the company to do anything better than that.

DeVorkin:

This was all done on contract, and IBM was not supplying that much money?

Harwit:

Yes, that's right. But at least it paid me for the summer and that was nice. I got the system to work, which meant that the first phase of the contract had been satisfied and we could get money for the second phase. That was sort of amusing, because the IBM people came in, together with the intermediary who had arranged for the contract, and who had to be there, because I guess he was being paid a percentage of the contract. He was a fairly heavy set guy who perhaps must have weighed 200-250 pounds. And every time he approached the instrument the floor would bend so much that the fringes would disappear. Myron Block, who was a rather outspoken guy, started yelling at him to stay in the corner of the room (laughs) so he wouldn't bend the floor, and eventually the IBM people did see the fringes and they were satisfied; and so I guess it worked out.

Then I stayed with Block Associates for summer work and various part-time work until I was finished with my studies at MIT. That really was one of the first optics jobs that I had. Most of the work that I have done since then has involved optics in one form or another. Although Larry Mertz was a contemporary of mine in age, I always felt that I was influenced by him, perhaps more than any other single person in the overall outlook on optics that I have had. For example, he first explained to me this whole advantage that you can get through multiplexing. That is, in a Michelson interferometer, which can be used as a spectrometer, the fringe spacing is different for different wavelengths. If you Fourrier analyze the light and dark patterns of the fringes, you can recover the spectrum of radiation passing through the instrument. In contrast to a slit instrument, for example a grating spectrometer where you are looking at one wavelength at a time, when using individual element detectors in the infrared.

DeVorkin:

Were you working in the infrared?

Harwit:

Well, I wasn't working in the infrared yet at the time. I was going to be after that, but Larry was thinking of building an interferometer for the infrared for astronomical purposes. In the infrared one didn't have photographic plates to work with, so one couldn't look at all the colors all of the time the way that one can by photographing an entire spectrum. So in the infrared, what you would normally have had to do would be to look at one color at a time, one wavelength at a time by turning a grating or turning a prism in a fixed-slit instrument. This is a laborious and slow process.

If you have a Michelson interferometer, you can look at essentially all of the radiation half of the time, or half of the radiation all of the time, because all of the radiation is passing through. Half of the time on the average there is negative interference, and you get zero light coming through, and half of the time on the average there is positive interference, and all of the light comes through at any wavelength you want. If you only have a single detector available, you can gain a great deal of sensitivity by looking at all of the radiation all of the time, essentially, or all of the radiation half of the time by means of this interferometer. This is the idea of multiplexing, as this is called.

DeVorkin:

That's how it's described optically?

Harwit:

Yes, optically you talk about mutiplexing, or instead of with an interferometer, you can do it by other means also. Larry Mertz was looking at different mechanical ways of getting similar schemes; for example, by building multi-slit instruments where the radiation would pass through different combinations of open and closed slits before being spectrally recombined again to fall on a single detector.

DeVorkin:

Didn't Redman and Griffin have some sort of a spectrometer that worked that way?

Harwit:

I don't recall it. It's quite possible, quite possible.

DeVorkin:

They created grids that let particular lines through, which were then imaged onto one photoelectric cell. You could move the grid around to achieve minimum intensity, maximum intensity; and the position of the grid told you the radial velocity of the object.

Harwit:

I see. Yes, there have been for quite awhile, special masks that one would make up, for example, to look at carbon monoxide lines, or carbon dioxide lines. If you knew what the spectrum was going to look like, then you could look at all of these lines simultaneously, and gain in sensitivity by going through the kind of procedure that you just mentioned. But the techniques that Larry Mertz was interested in were more general; that is, they could be applied to any kind of a substance, an arbitrary spectrum. He was making up grids that had been suggested by Marcel Golay, a tremendously ingenious person who had around 1948-49 suggested ways of multiplexing with grids; and then also worked out mathematical schemes that would optimize the decoding of these patterns.

I mention all this because later on I spent quite a number of years working in that area, and I'll talk about that later. Myron Block actually wasn't working for the company full time. He was working full time down at MIT as a consultant, as I mentioned before, at the Naval Supersonic Labs. He put me in touch with the people down there at the Naval Supersonic Lab. Later on when I needed to have a sponsor for equipment that I needed at MIT for my thesis work, the people at the Naval Supersonic Lab were very helpful and actually paid for all of the equipment that I needed.

DeVorkin:

So what of your thesis? Who was your advisor?

Harwit:

All right. My thesis advisor was William P. Allis, a plasma theorist. He had done plasma research, discharges, ionized gases, all kinds of things, long before plasma research became fashionable. He essentially was the spiritual father of the plasma group at MIT. Because of the respect everybody awarded him, he really was almost the head of the group. Titularly, I think Sanford Brown was the guy who supervised the experimental work, and ran the day-to-day affairs of that group. But between the two of them, they supervised that group of post-docs and students.

DeVorkin:

How did you become his student?

Harwit:

What happened was that about the time that I was looking around for a thesis topic, I became aware of the work of Hanbury-Brown and Twiss on a new type of stellar interferometer that they had constructed in which they used two telescopes to look at a star, and then correlated the signals that came from photomultipliers that were illuminated by these two telescopes. As they increased the distance between the two telescopes — one of the telescopes was movable — they found that the correlation in the signals obtained with these two photomultipliers dropped to zero.

DeVorkin:

It is time correlation?

Harwit:

This is a cross correlation taken at one and the same time. Suppose you get a current going through these two photomultipliers tubes, then the amplitude of these two currents is multiplied. When the two telescopes are very close together then you will find that this product is high as you separate the telescopes the product diminishes. The reason for that is that when they are close together, they don't resolve the star, and you therefore get the same flux falling on both telescopes. When they are far apart, they see the stars from somewhat different angles, and for sufficiently separated angles the fluxes no longer are correlated.

DeVorkin:

That angle is a very small fraction of a second of arc?

Harwit:

That's right, yes. In fact, the whole point was to try to measure angular diameters of stars that way.

DeVorkin:

I never have understood how they could perceive such small angles with such crude equipment.

Harwit:

Well, the separation between the telescopes that's required for most of the bright stars that they look at is of the order of meters. So if you had two telescopes, separated by tens of meters perhaps, you only had to measure the separation between the telescopes accurately. As long as you had the light squarely focused in each of the telescopes on the photomultiplier tube, it didn't matter. Nothing else mattered very much, so actually the set up could be quite crude. I think initially they may just have used searchlight mirrors in fact, as light collectors.

DeVorkin:

Yes, then they used segmented mirrors, I know, for the big interferometer.

Harwit:

That's right, I think, which they later put together in Australia.

DeVorkin:

Yes. Well, where was the first one?

Harwit:

The first ones were constructed in England. Harbury Brown and Twiss both were at Manchester. They had first used the technique at radio wavelenghts, and there had been no controversy about that, but when they started using the technique in the optical region, then there was a furor, and a somewhat acrimonious debate in Nature where people claimed that Hanbury Brown and Twiss had made an error in photon statistics. That controversy was then resolved in favor of Hanbury Brown and Twiss by Edwin Purcell at Harvard, who pointed out that this extra correlation that one would get when the telescopes were close together was a result of the Bose-Einstein fluctuations, that a photon gas would undergo, and had to do with stimulated emission.

DeVorkin:

Now, this is something I'm still not clear about. It's fascinating.

Harwit:

Well, what happens really is that at high temperatures, particularly at longer wavelengths, atoms in the atmosphere of a star can emit radiaton stimulated by radiation that's incident on the atoms. What happens is the stimulated photon, the photon that is given off as a result of this process, quantum mechanically has to have absolutely identical, or indistinguishable properties from the photon which stimulated the emission. And so, what happens now is that you get a pair of photons with exactly identical properties — within the Heisenberg Uncertainty Principle, which in this case also turns out to be the diffraction limit — traveling through space for thousands of years, perhaps.

Now when they reach the two telescopes, then as long as the telescopes are close together, within the diffraction limit, then the telescopes are indistinguishable to these two photons, and then one photon can go into one telescope, and one in the other, and they will simultaneously produce a current pulse. When you take these telescopes apart to the point where they can resolve the star, at just that point also the two photons can distinguish the two telescopes. The telescopes no longer are within the diffraction limit of the star.

Now if one of the photons enters either of the mirrors, the other one can't enter the other telescope any more; because the two photons travel together within a certain angle, which is prescribed by the diffraction limit — the Uncertainty Principle. They can't come any further apart by that. If you separate the telescopes, as seen from the star, further than that small angle, you can no longer have one photon entering one mirror and the other simultaneously entering the other one. It will miss it. The photons will be too close together.

DeVorkin:

Is this the aspect that people were objecting to?

Harwit:

That's right, yes. The mistake was that people who were objecting to it, had forgotten about the stimulated emission, and they felt that the photons in a beam ought to somehow be randomly distributed in this beam, and then there was no reason why there ought to be an additional correlation when the telescopes were close together, as contrasted to what they would have when they were far apart.

DeVorkin:

Who were the major antagonists?

Harwit:

There was a two-man team of E. Brannen and H. Ferguson, Peter Fellgett and RC. Jones were against that. I should look that up. (checks its.) I guess Fellgett had been against it, and somebody by the name of R. Clark Jones, who worked for Polaroid Corporation.

DeVorkin:

This is on page 10 of your thesis.

Harwit:

Yes, and on Page 16.

DeVorkin:

What was your thesis then?

Harwit:

Okay, so for my thesis then, what I wanted to do was to show that this excess fluctuation which Hanbury Brown and Twiss postulated, or had noticed, really was the same thing as the Bose-Einstein fluctuation. This excess fluctuation of a Bose gas had never been measured in the laboratory for a source in thermal equilibrium; and this was about 50 years after Einstein had suggested this whole business.

So it seemed to me that if one could do it with stars, one might be able to do it in the laboratory also. Take a source where you knew what the temperature was, and then verify this — I guess it was called the Einstein-Fowler fluctuation by measuring this cross-correlation that Hanbury Brown and Twiss had obtained, and which was a measure of the total fluctuation in the photon gas. There is a total fluctuation, because the photons are not coming randomly. If you had just one detector, for example, then the arrival times of the photons from a star would not be random. They wouldn't be random, because occasionally instead of single photons coming, you get pairs or triplets.

DeVorkin:

As long as it is below the diffraction limit.

Harwit:

If you have one telescope, yes, you are always within this limit. Then there isn't just a completely random arrival time of the photons, because occasionally you get two or three coming together in clusters through the stimulated emission. Then you will get an extra big pulse, which gives you this excess fluctuation. Rather than coming randomly, they are more strongly correlated in time. So I thought I would like to measure that. I went up and talked to Purcell at Harvard, and at that time I thought that this was something different, or additional to the Hanbury Brown and Twiss effect, or maybe an adjunct. My mind wasn't quite clear about this. We had a long discussion in which I disagreed with him. It wasn't until I was really finished with my thesis about two years later that I understood what he had meant, and of course he was right. He is just a very smart guy.

DeVorkin:

Now, had you talked about this already with Allis?

Harwit:

No, what happened then was that I started looking around for someone at MIT who would let me do this as a thesis. All of the experimentalists I talked with, said, oh, well, that's interesting, but how would you like to work on this or that problem which I could support. Then I'd go on to the next one. At Michigan I had somewhat unpleasant relations with the professors I had worked with, especially the second one, Cyrus Leventhal, whom I mentioned in our previous interview.

I wanted to be sure that this time I would not have any runins with whatever professor I worked with, and so I took opinion polls of fellow students on which professors they thought were reasonable to to work with, and which ones one should try to avoid. Eventually I came to William Allis, who, although he was a theorist, said he was willing to be thesis supervisor if I could find somebody who would support me. He thought it was kind of a nice thing to try to do. I could find somebody who would be able to provide the equipment for it.

DeVorkin:

By the way, do you recall who won and who lost that popularity contest for thesis advisors?

Harwit:

I don't remember really. I think Sandy Brown was not recommended very highly, although I'm not sure. There is a lot of graduate student gossip which isn't always right. Sandy Brown certainly produced some awfully good students, and so I'm not even sure that I'm right in saying that. But I did go into see Brown, and he was one of the people who said, how would you like to work on my project instead, and I didn't want to do that.

DeVorkin:

What was his project?

Harwit:

I don't know, something in plasmas.

DeVorkin:

Your thesis then was a project in experimental physics that would support an important astronomical technique that was still controversial, or was it to study that technique, to learn more about it?

Harwit:

Yes, it was partly that, but I can't really say that I was that interested in the astronomy. I saw it as a possibility perhaps for doing astronomy later with the technique. But I was in the physics department and I really wanted to do something that the physicists would consider a thesis, and which also interested me, because it was a basic physical process that ought to be checked out and verified in a laboratory.

DeVorkin:

So Sputnik didn't have much effect on you?

Harwit:

Well, it did. It certainly was interesting and exciting; but not a direct effect, no, I don't think so.

DeVorkin:

At that point you weren't thinking of altering your career to take advantage of this?

Harwit:

I did think that I would ultimately like to go into astrophysics, once I got done with my thesis; but the immediate problem was to get through thesis work.

DeVorkin:

But it wasn't necessarily space physics or space astrophysics that you were pointed to as a result of Sputnik?

Harwit:

Well, perhaps indirectly, yes, because what happened with the thesis project, was that I found that it would be easiest to carry it out in the infrared, and that immediately led to the idea that perhaps I could do space things in the infrared later on.

DeVorkin:

So that is the origin of your infrared interests.

Harwit:

Yes.

DeVorkin:

Well, how did it develop then?

Harwit:

What happened was that I of course had to find somebody then who would supply the instrumentation. One of the things about these correlation techniques was that they also were very sensitive infrared detection schemes. If, say, you wanted to look at an airplane plume in the infrared, and there was a lot of detector noise at the same time, then if you looked with two different detectors and compared their results, correlated them in some way, you would be able to reduce the effective noise and be left with a positive detection than if you had only had a single detector.

At the time the people at the Naval Supersonic Lab were interested in airplane detections; in fact, that's what Myron Block was working on for them. He had done a lot of infrared work before. He was supplying them with all kinds of infrared advice, and also hoping, I think, to at the same time build sensitive infrared instrumentation for the military. So I went to them and suggested that they might be interested in this technique (the Hanbury Brown Twiss effect in any case for a source at a given temperature ought to be stronger at longer wavelengths, in the infrared, in this case), and so I said that I would build the instrumentation if they paid for it.

They could use it for airplane detections, and I could use it for measuring the photon cross-correlation and we would share the correlation apparatus. They would have their own infrared detectors and I would have my own infrared detectors. That's in fact what happened then.

DeVorkin:

What range of infrared were you talking about?

Harwit:

I was working with lead sulfide detectors at the time. Those were the most sensitive things available, and that allowed you to go out to two or three microns. There were some detectors that were being started at that time for longer wavelengths. This was around 1957, but they weren't all that sensitive yet. I didn't know much about them, and the lead sulfide detectors looked as though they would be able to do what I wanted to do.

DeVorkin:

Okay. Let me turn the tape over.

DeVorkin:

This is Tape No. One, Side No. Two. What were these other detectors?

Harwit:

I think they were mercury-doped germanium detectors that the military had started using. Then there was a new type of detector that came out around that time also, zinc-doped germanium detectors which the military with their characteristic flair called zip-detectors, i.e., zinc-doped.

DeVorkin:

Were these considered to be classified at that time?

Harwit:

I think they might have been, I'm not sure. There were also copper-doped germanium detectors; all these various dopings allowed you to go to different limiting wavelengths. The mercury-doped, I think allowed you to go out to perhaps 12 microns; the copper-doped to maybe 28, 29, if I remember correctly; and the zinc-doped to maybe 40. That's why the zinc-doped detectors were interesting, because they went out to a longer wavelength.

DeVorkin:

Were these things around MIT? Or were they things you just heard about?

Harwit:

No, I heard about them. I think MIT may have had some of them. Many of them required liquid helium cooling, which was at the time still a difficulty; and the lead sulfide detectors I was able to cool just to dry ice temperatures.

DeVorkin:

Not liquid nitrogen?

Harwit:

No, I used dry ice-alcohol, or dry ice-acetone mixtures, to work with Actually, one should not use acetone because it is flammable. The liquid was just there to conduct the heat otherwise, the dry ice alone wouldn't have conducted the heat.

DeVorkin:

Where would I go, to your knowledge, to find out about the origins of the mercury-doped germanium detector or any kind of germanium detectors?

Harwit:

I don't know. There is a book called HANDBOOK OF MILITARY INFRARED TECHNOLOGY.

DeVorkin:

I've got that. Okay, I'll look at it.

Harwit:

I think you might find something in an older edition of it.

DeVorkin:

Let's move on. What interesting experiences did you have in working up your thesis? Did you have any glitches?

Harwit:

Well, it was difficult because I was working by myself. on the other hand, they had a good electronics capability at the Naval Supersonic Lab. And so they were able to help me out that way. One of the things that was difficult was that I really was only able to work very late at night sometimes, and on long week ends, because the wind tunnel had to operate in the evenings, because of power consumption. I think it used something like one-third of the capacity of the Cambridge, Massachusetts power stations. It really requires a lot of power, so that they would run fairly late in the evenings.

During the daytime there were too many people walking around, so there would be a lot of noise that would interfere with the correlation techniques. The trouble with correlation techniques is that they are undiscriminating. Anything that will affect both detectors of the system will give you a correlated pulse. So you have to try to work out schemes that will discriminate against these unwanted sources of correlated signals. The very first step is to try to work in thebuilding when nobody else was there, Christmas time, New Year's, late at night, and so forth. Eventually I managed to get the thesis through, but it took quite a while. In the meantime I had decided I definitely wanted to go into astrophysics.

DeVorkin:

Can you put your finger on anything that influenced you then?

Harwit:

I think the person who undoubtedly influenced me most after the initial reading of Fred Hoyle's book was Tommy Gold, because he was at the time a fairly brash young man. I think he was about 38 when he came to Harvard. And he took Cambridge by storm. He'd be invited down to MIT to give colloquia and he was in Washington a lot. He was written up in the newspapers. Walter Sullivan would call up and get his opinions for the New York Times, and so forth. He was very outspoken and jovial always. So people liked that, and he didn't mind what he said particularly.

DeVorkin:

I heard he had come to Harvard without knowing that he was going to be made director of the radio facility there at Harvard, because Bok was leaving. Did he ever comment?

Harwit:

I'm surprised to hear that, because one of the first things he did was to persuade the British physicist, Jelley, who at the time was working on masers, to come to Harvard and build a hydrogen maser together with the Harvard people who were interested in maser work, I think, people like Bloembergen, and maybe Norman Ramsay, so I don't know. Ed Purcell was also interested. What they wanted to do was to have a maser for 21 cm. hydrogen measurements. That was carried out very soon after Tommy came there, so I wouldn't be surprised if he hadn't known about his appointment ahead of time. One would have to ask him directly.

DeVorkin:

I think he's already been asked. That's just something that I had in the back of my mind. Funny things were happening at Harvard. This was toward the end of Menzel's tour there. Did you have any contact with Menzel?

Harwit:

I used to see him from time to time, but I dcn't know if he knew me. I never did get close with him, or anything like that. I've talked with him, I think, but not in any official way.

DeVorkin:

So you weren't aware of Harvard politics or anything like that?

Harwit:

No, not too much. The only thing that did strike me about Harvard is that I felt the people there were not very good. I didn't think they knew enough physics, except for Tommy Gold, who had a fantastic intuitive knack for how things might be, and whom I really found very appealing to work with. Later on the same outspokenness that made him appealing also, I think, won him a lot of enemies.

DeVorkin:

In the Cambridge community?

Harwit:

No, I mean, at NASA Headquarters, for example, because he didn't mind criticizing something off the top of his head, if he felt it was wrong. NASA is very much an organization which counts on team playing and team players. If you're not a team player, they kind of feel, to hell with you. Later on he had a fairly substantial decline, because I guess he had said so many things about so many people that got back to them, or that was in the press, and that was taken in a way that he perhaps didn't always intend. He was quick to say, rubbish, and say that somebody was wrong, and foolish about it, and so forth. So as time went on, people I think just turned against him.

DeVorkin:

How did you feel about that kind of a person?

Harwit:

Well, at the time, as a student, he certainly was a glittering personality. As I say, he took Cambridge by storm. Everyone was inviting him, and people wanted his opinion. If I needed to borrow equipment for this moon dust thing, I just had to say that I was working with Prof. Gold at Harvard on this, and would it be possible perhaps to borrow something like that, and people would be very helpful. So one really only had to mention his name and people would be pleased to help him out. If I was a student who was working with him and was polite enough, they didn't mind letting me do those things that I was trying to do with him. So that was fine. I didn't know, at the time, that the things he said in quips would be taken very seriously by people and hurt their feelings, and turn them against him.

DeVorkin:

Who else controlled it? Was it Frank Drake?

Harwit:

No, Bill Gordon in electrical engineering had built it, had already been at it for several years when Tommy Gold came, and resented Gold's interference as he saw it. But Gold, too, was a very smart engineer. His degree was in electrical engineering at Cambridge. He did have good suggestions from time to time. Of course there can't be anything more annoying than good suggestions coming from somebody whom you don't want to listen to. A lot of times I suspect Tommy was right and Gordon might have been wrong; and maybe the other way around also a lot of times. But the ones that Tommy would have talked about of course were the times that he was right and his advice was not taken.

Eventually he forced Bill Gordon out, and that of course upset the other people in electrical engineering. Cornell lost, as a result of that error, a tremendously powerful ionospheric group that included Henry Booker and William Gordon went to Rice; and a number of other people went with Booker — to San Diego where the University of California was just building up a new graduate school at La Jolla.

DeVorkin:

He never did that to you?

Harwit:

No. He's always been very generous, in fact, I think, to a fault to people who work with him; so that even if he had people who were absolutely horrible, he would defend them. So his barbs usually went to people who were his superiors. He would say all kinds of unkind things about the way that things had been run stodgily in Britain, much to the amusement of everybody over here. Of course in England he became persona-non-grata, but he already was, I suppose; and eventually that happened also in a number of other areas. In addition to that, unfortunately, once he came to Cornell, he also really wanted to be a director there in the sense of British directors. He had been made director of the Center for Radio Physics and Space Research.

When I eventually went to Cornell, I found that he really had started having a lot of fights with people at Cornell, in order to gain control of the Arecibo Dish. It was essentially 99% of the Center. If you didn't control that, being Center director really wasn't anything very much, because it was the man who was director of Arecibo who controlled essentially all the toys.

DeVorkin:

Well, let's not get to Cornell just yet.

Harwit:

Yes, but anyway, the point is that Gold was very influential on me at the time.

DeVorkin:

Through personal contact?

Harwit:

Through personal contact. I had worked with him. I had seen him in classes. We seemed to talk the same language.

DeVorkin:

No specific problem areas, other than the lunar dust by this moment.

Harwit:

Yes, and then I had done a term paper on something in magnetohydrodynamics, which never really amounted to anything much, but he thought it was kind of interesting.

DeVorkin:

So he obviously had taken a liking to you?

Harwit:

Yes. I think he considered me a highly promising student at that time. Later on — it may even have been that evening when he came down to MIT to look at things, and asked me if I would like to come to Cornell sometime — I think I mentioned to him I would like to do a post-doc. When he was there I had found a mistake in a paper that Dennis Sciama did on galaxy formation in the steady state universe. I mentioned that to Gold who had had Sciama as a student He thought a lot of Sciama also, who is a good man of course. So he suggested that I recalculate that, which I did, and I came out with a different result. I think he wrote Sciama about it. Sciama denied that I had done it right.

But eventually I was going to be writing that up as a paper. I mentioned to him — again I think it may have been that last evening — that I might like to go to work with Sciama at Cambridge. So Tommy said, well, why go work with Sciama — Sciama was a fairly young guy at the time still — why not go work with Fred Hoyle. I never even dared to think of that. Here was the great Fred Hoyle whose book had influenced me into taking an interest in astrophysics, and the thought of reading the book one year and going to work with him three years later was just staggering. At any rate, I said that would be nice, and I'd let Tommy know when I was reaching the end of my thesis. At that point I heard from somebody that there were these new NATO postdoctoral Fellowships that were available.

DeVorkin:

You don't know who you heard it from?

Harwit:

No, I don't remember, from some fellow student, I think. I knew a number of students that I had been in the Army with at Eniwetok. One of them was an anthropologist at Harvard. One of them was in architecture school at Harvard. One of them was in the business school at Harvard and two of them were in the business school at MIT. We all had been in the same tent at Eniwetok, so we used to get together periodically.

DeVorkin:

Did you hear about it through them?

Harwit:

I think I may have heard about it through Loring Brace, the anthropologist.

DeVorkin:

The postdoctoral fellowships were certainly well known.

Harwit:

No, that was the first year they had been offered, I think, by NATO.

DeVorkin:

By the time I got around they were pretty well known, in the mid-1960's. Had you considered any other alternatives? was there anything in your thesis that might have kept you at MIT, any problems?

Harwit:

Not really. It was one of those things where I think, when Marianne and I were first talking about getting married, I mentioned to her that physicists don't earn very much, but that one did get to travel, and that sometime we might want to go to Europe. So we had always thought it might be nice to get a postdoctoral position in Europe. Then when I heard about this, I wrote in immediately, and applied for one. I also wrote Fred Hoyle. Tommy Gold gave me a recommendation, and I imagine, William Allis would have also. I don't remember whom all I asked. But I got the Fellowship, and then the problem was that I had to begin within 12 months, and I finished my thesis finally in the 11th month. It was a little bit harrowing.

DeVorkin:

Harrowing from what standpoint?

Harwit:

I couldn't get the thesis to work out properly. It was just a very fine measurement to make at the time. Nowadays it would be fairly straightforward. The detectors just weren't that sensitive at the time, and the electronic techniques weren't that well worked out.

DeVorkin:

We have already identified this as your first major activity in infrared work. Can you say that this was an influence in your later research? Or is there something broader that you can identify?

Harwit:

Yes. I remember that I was doing infrared work, and I was starting to do infrared work, even while Tommy Gold was still at Harvard. There was one afternoon seminar, I remember, where we were talking about various things. One of the things which was interesting at the time was that people were getting interested in the possibility that molecular hydrogen might exist in interstellar space. The trouble with molecular hydrogen is that it's very difficult to detect, because for ultraviolet measurements you would have had to go above the atmosphere.

Even then, the same ultraviolet radiation that the molecules absorb, which would lead to the spectral identification, also dissociates the molecules. It's difficult to see the ultraviolet absorption. Certainly you can't see it through thick interstellar clouds, because the dust there will absorb the ultraviolet. Even nowadays you only see the molecular hydrogen in very thin wisps, and you know that while you are looking at it, it is also being destroyed.

DeVorkin:

Spitzer was making these predictions at that time.

Harwit:

Yes, Spitzer was. I don't know who did what when. I know that Tommy was talking about molecular hydrogen; and Max Krook at Harvard also had something on molecular hydrogen. Since I had done infrared spectroscopy when I was at Michigan. I knew that you could also see the vibration spectrum of molecular hydrogen, although it was a quadripole transition, and therefore, very weak. But in the seminar at Harvard I mentioned that one ought to be able to do the infrared vibration and rotational spectrum.

I think Fritz Ziwicky in fact had also suggested looking at around 85 microns for a transition, the ortho-para transition which has to do with the spins. Molecular hydrogen can exist in two different states. One of them is called the ortho; one of them is called the para, and it has to do with the way that the nuclear spins are oriented. Now actually that transition is so strongly forbidden that you would never see that, I think; or you would certainly see others a lot sooner.

DeVorkin:

Was the radio spectrum understood at that time?

Harwit:

There is no radio spectrum for molecular hydrogen at all. There just isn't any.

DeVorkin:

Someone told me that a lot of Spitzer's predictions were verified by radio interferometry before Copernicus. So that's incorrect.

Harwit:

That, I think is maybe incorrect. What had been done before was that Karl Heiles at Berkeley had looked at the correlations of 21 cm. atomic hydrogen absorption or emission, and dust concentration in the galaxy. Where the dust was strongest, he always found that there was a lack of atomic hydrogen. And so by inference, one decided it ought to be molecular hydrogen.

DeVorkin:

There would be no other reason for that deficit?

Harwit:

No. But any rate, later on when I came to Cornell, just to sort of follow up that particular story, one of the first things that I decided I wanted to do was to look for the infrared spectrum of molecular hydrogen, because up to then there had been no molecular hydrogen detected anywhere at all.

DeVorkin:

So that's already 1962, 1963.

Harwit:

That's right, yes. But that interest dated back to a 1958-59 seminar up at Harvard. There were a lot of people talking about molecular hydrogen, and that there might be some around in interstellar space. In fact, there might be quite a lot around. And so what would be the best way to look for it. Most people were talking about the ultraviolet spectrum, and some people about the infrared, as I mentioned, but nobody was talking about actually doing the infrared stuff, as far as I know. I thought, well, maybe this could be something I could try to do.

DeVorkin:

Was that with the understanding that you would have to put it on a satellite or a rocket sonde, or a balloon or something?

Harwit:

Actually that could be done from the ground because the vibrational transitions are in the near-infrared, and it turns out that there are some transitions, for example, one at 2 microns, which falls right into an atmospheric window. I think other people were suggesting similar things; but I don't know of anybody else who was actually going to make the infrared measurements.

But to go to England, I thought I didn't want to do experimental work there. In the first place, I thought I could do the sample calculations that people were doing at the time, not the highly mathematical things of the type that Chandrasekhar was doing; but there were a lot of theoretical papers in the style of Gold and in the style of Hoyle, where one was trying to estimate what physical processes could play a significant role. That required nothing much more than order of magnitude estimates very often.

When there were five or 10 or 15 different processes competing, the whole point was to try to see which two or three were the most important ones to influence things. Many of the astronomers were not sufficiently familiar with modern physics at the time to do that. So this was a natural place to go in and maybe do some reasonably interesting stuff which didn't require a very high proficiency of mathematical techniques.

DeVorkin:

For your NATO application, did you write basically a research prospectus?

Harwit:

I don't know what I had to write there, but I think I may have said that I wanted to work with Hoyle, because I had done this work on galaxy formation and the steady state universe, and I wanted to follow that up. I don't recall.

DeVorkin:

Well, you got the Fellowship.

Harwit:

I got the Fellowship. By the time I got to England I only had about a month's worth of work to do on the first of these papers, and Fred Hoyle at the time was in California, I think. He didn't get back until I had been there about a month. So I finished up that paper by myself and eventually sent it in to the Monthly Notices of the Royal Astronomic Society.

I think Fred may have sent it in for me, and it was eventually published there. When Fred came back, he was getting interested in comets. He suggested that maybe we could work on a comet problem together. So I went off and started reading about comets. I knew nothing at all. I decided that what I ought to do was to read the first book that came along, and that was one that had been written by Ray Lyttleton, who also was at Cambridge.

So next time I came back I talked with Fred, and he said, that's the wrong book to read. Ray Lyttleton's theory is believed only by himself and nobody else. So I had to go back, and then I started reading stuff that Fred Whipple had done. Eventually Fred Hoyle and I wrote two papers on plasmas in comet tails, to try to explain how you got these several 100-kilometer/sec velocities in the tails of comets. We wanted to know how these high velocities were reached by ionized gases in the tails of comets, and Ludwig Biermann had at that time suggested an interaction of the solar wind with comet plasmas. We worked out in more detail some of the mechanisms. Also, Hannes Alfven had done some work on comet magnetohydrodynamics which I think we didn't agree with. In any case, we wrote up two papers that dealt with comet tail accelerations.

DeVorkin:

So you never did anything on galaxies?

Harwit:

Well, we did one other thing. We did a paper on the possible origin of intergalactic bridges. These were threads that connected galaxies that Fritz Zwicky, and then also a Russian astronomer by the name of Vorontsov Vel'yaminov had compiled catalogs on. In fact, Vorontsov Vel'yaminov's catalog at that time was completely new. I don't know how we got onto that; but I think I may have seen the catalog in one of the Cambridge libraries and started talking about it with Fred.

DeVorkin:

What was the mechanism that you suggested?

Harwit:

I don't even remember any more, but there were questions of whether these could be stars, or whether this would be synchrotron emission.

DeVorkin:

But this wasn't considered to be effective collisions at that time?

Harwit:

Oh, the formation of the bridges?

DeVorkin:

That's right, Yes.

Harwit:

Probably, no, in fact, it could be some sort of collisional process that pulled the material out in threads. I forget now what we did exactly, but I think it was a gaseous process that we looked at. I would have to look it up.

DeVorkin:

That's fine, but the important thing about your year at Cambridge was that you worked with Hoyle, and you were doing interesting problems. But were you settling into some sort of a career style at that point? Were you looking for one?

Harwit:

Well, sure I was looking for one. In answering that, let me talk about another man whom I met, and that was Hermann Bondi. I got a note one day from the people down at the Monthly Notices in London that they would like me to come and give a talk on this paper at one of their regular monthly Friday evening sessions, the RAS meetings. And so I prepared that. I never had met Bondi, but everybody told me that he was a very sharp critic.

He at the time was one of the two secretaries of the Society who always sat up on the podium, and evidently, I was told, mercilessly criticized all the speakers. He was just a tremendously sharp fast thinker, just like Tommy Gold, and could come up with criticisms on the spur of the moment. About a week before I was supposed to give this talk I saw a little announcement on a bulletin board in Cambridge, saying that Bondi was going to be giving a talk about a week after my talk was due at the RAS at King's College London — where he was — on exactly the same topic; namely, galaxy formation and the steady state universe.

So here I had the double threat of somebody who was working on the same problem and at the same time also had this reputation for being very critical. So I read everything conceivable that Bondi had written on the topic of the steady state universe, and prepared myself for all the possible counter-arguments that he might come up with, and had counter-arguments to his counter-arguments all prepared. I went down there and gave a talk which I think was pretty good. Of course, he immediately lit into me afterwards, and I lit back into him. And that went on for maybe 10 minutes or so.

DeVorkin:

This was during the public discussion?

Harwit:

During the public discussions.

DeVorkin:

Some of this must be recorded in the Observatory.

Harwit:

I think so, yes, although they do it very briefly.

DeVorkin:

That's right.

Harwit:

All of this was of course to the enjoyment of the fellows who were sitting there, because people like nothing better than sort of a parrying back and forth. I went back and sat down afterwards and then there were a couple of other speakers. Then when the session ended I went up and introduced myself. Bondi said, well, you know, I really enjoyed this. That's why I invited you to come down here.

It turned out that he had actually seen my paper when it was submitted and found it interesting, asked to have me invited down, and then got to thinking about it. He asked would I come down next week to London, because he was going to give a talk. The paper I had submitted had made him think about galaxy formation; and he thought he had some ideas on it, and he would like me to come and listen in. We talked a while and went down to one of these Wimpeys — a hamburger stand — and had a bite to eat. Then he went off to Surrey and I tock the train back up to Cambridge.

But when we were talking over supper, we talked about the relative merits of theoretical and experimental work. I said I really thought that I wanted to go back to do observational things, because there were a lot of things one could start doing now, with space work and instrumentation, and I was feeling a little bit ill at ease with theoretical work, because you could always argue everything on both sides of the fence, and you really weren't getting any real answers. I was kind of discouraged by that. His reply was that the experimentalist always came up with results that they swore by and then it turned out two or three years later that they were all wrong. Experimental answers were not the last word either, and that they vacillated back and forth just as well.

DeVorkin:

Bondi has written position papers like that in Vistas and other places like that.

Harwit:

It wouldn't surprise me, yes.

DeVorkin:

Was he quite adamant about it? Was he trying to encourage you to go in theory?

Harwit:

I think he wanted me to stay in theory, perhaps, yes. I said, well, you know, I enjoyed doing theory, but it did worry me that it seemed to me you could go on for years and years without ever knowing for sure whether something was right or wrong.

DeVorkin:

Did Bondi talk about his opinion of doing astronomy from space at all?

Harwit:

I just don't remember that. But I think what he was referring to perhaps was these number counts by (Martin) Ryle which were changing. At any rate, the mention of Ryle brings to mind a meeting the following Monday after this Friday session. At that time Fred Hoyle was associated with the Department of Applied Mathematics and Theoretical Physics. The center he was to found hadn't been built yet, and wasn't going to be built till many years later. In fact, it wasn't even being talked about, I think. I didn't know about it at any rate.

The Department of Applied Mathematics and Theoretical Physics was also housed in the Cavendish Laboratories; so I would go there for tea every afternoon with everybody else, and also in the mornings, for coffee. It was a great place where you could talk with people like Pippard, who did solid state physics, thermodynamics, that kind of thing, a very smart guy; and Ziman, who actually now is a sociologist of science. But at the time he was a solid state theoretical physicist, who then went on and took a professorship at one of the other universities. I forget which one. He is actually well known.

DeVorkin:

There were teas twice a day?

Harwit:

The morning one, I think, was called coffee, and wasn't quite as well attended. The teas, almost everybody went to, including quite often Sir Neville Mott, who was the Cavendish professor at that time. Some of the people from Ryle's group who attended that meeting on Friday at the RAS came up at the tea the following Monday and said that Ryle would very much like to meet me.

So I was ushered into his presence, and he warmly shook my hand. Nobody had ever introduced me to Ryle. But Ryle appeared clearly happy that here was another person who was attacking the steady state theory, because my paper was on the fact that you would have to have a pretty damned dense steady state universe to get galaxies to form at the rate that was required to reproduce every feature of the universe within the time allotted by the expansion. So he considered this an attack on the steady state theory, which he also was attacking himself. I was actually quite disgusted by that whole thing. That was the first and last time, it turns out, that I ever talked with Ryle.

DeVorkin:

Really!

Harwit:

Well, I never had an occasion to again. But in contrast to Bondi — whose view I had attacked, and who immediately became open and friendly, and regarded our disagreement as a purely scientific dispute — Ryle took it as a personal dispute immediately, and saw me as a member of his group then. This sort of welcoming into his fold was just more than I could take, and at the time I still was fairly idealistic and thought that this was just completely uncalled for. I didn't want to have anything to do with that sort of thing at all. I much preferred the technical antagonism of Bondi, to the personal friendship of Ryle.

DeVorkin:

This of course comes out in other places. It seems to be just his style.

Harwit:

That's right, yes. Ryle was strongly outspoken, bitter against Americans, bitter against the theorists at Cambridge, bitter I think, against Australians. His whole attitude seemed to be terribly bitter about so many different things. And none of his people, except for Shakeshaft, seemed to ever come to any of the other astronomical colloquia at Cambridge. There were three groups at the time, the Department of Applied Mathematics and Theoretical Physics; the Observatory group on Madingley Road where they did optical work; and then the radio group of Ryle. Ryle's people worked pretty much as a self-isolated group, which was too bad. He wanted to have his own theorists who would work with him, and maybe come up with the results he wanted. It was just his style. I would call it paranoid, probably, as much as any scientist that I have every had a chance to observe — at some distance, admittedly.

DeVorkin:

So he in part created the atmosphere that you are identifying at Cambridge?

Harwit:

Yes. Well, he took everything terribly seriously, and acted paranoid. Tommy Gold with his outspoken, off-hand jovial criticism of course goaded that kind of personality intolerably, I imagine. Other people could let it run off their backs, but Ryle couldn't. The same is true, I think, of Bondi's approach. It was similar to Tommy Gold's, but not quite as disrespectful, I suspect.

DeVorkin:

What of Hoyle and Ryle during the time you were there?

Harwit:

Well, they also didn't get along. Now Hoyle is the type of person who doesn't give a hoot for other people's opinions. He doesn't make jokes about them the way that Tommy does. But if he disagrees he'll say so quite openly in public, make no bones about it whatsoever. I never had the feeling that he enjoyed that as much as Tommy enjoyed it; but he would also call people fools, if he thought they were fools.

DeVorkin:

Publicly?

Harwit:

Yes, maybe not quite as openly. He wouldn't say they were fools; he'd just say, I think they are dead wrong.

DeVorkin:

Yes. Okay. Let me turn the tape.

DeVorkin:

This is Tape No. Two, Side No. One on Martin Harwit, June 20, 1983. So, it sounded like the atmosphere at Cambridge then, as in many other years, was quite complex.

Harwit:

Yes.

DeVorkin:

How did you like it? How did you feel as a part of it?

Harwit:

I had a great time, for personal reasons more than anything else. The people left me alone. We had rented a tiny cottage in a little village nine miles south of Cambridge, a charming little town called Duxfor. There were nice neighbors there. There was a pub. There was no place in the Department of Applied Mathematics and Theoretical Physics where they had room, so I worked at home. I'd go in every once in a while to see Fred Hoyle maybe at two-three week intervals.

DeVorkin:

So you did not spend that much time with him?

Harwit:

No, he was fairly busy, but what would happen is that I'd call up and we'd make an appointment. I'd go to his home; he always worked at home. His wife would prepare a cup of tea, and bring out a couple of cookies, and we would sit there and talk for an hour or hour and a half. I'd tell him what I had done, and then he would suggest other things that we might do next, and maybe sketch some of them out. I'd go off and start working on those. Sometimes they'd work out; and sometimes I'd come back with something else, instead, having found that that didn't work.

hen we'd discuss it again. What I needed from him mainly was the critical acumen of somebody who had been in the field for a while, to make sure that I wasn't overlooking a lot of things. Don't forget, my formal training in astrophysics was limited to those three courses I had taken. So there were huge areas that I didn't know anything much about at all. I wanted to make sure that I didn't go off the deep end on something. So that was, as far as I was concerned, his main function.

He also was just a very good sounding board, and for me that was enough, really. I didn't expect him to do any more work than that. On the things that we published together, I wanted to be able to show to myself that I could do this kind or work; so I really wanted to have a collaborator who would do the minimum in a way, but keep me from making the worst possible mistakes. We had a very nice working arrangement. Without discussing it, we apparently agreed on what our roles ought to be.

DeVorkin:

This was at a time, 1960-61, when quasistellar radio sources were just being measured, and still not very well understood. Is this before the three degree measurements?

Harwit:

Yes. That wasn't going to come out for another five years or so.

DeVorkin:

Did you talk at all with Fred Hoyle about quasars?

Harwit:

I don't recall it, and in fact, I'm not even sure whether I was that aware of them. The radio work had been done at Manchester, so the chances of it being mentioned all that much in Cambridge were fairly small perhaps among the radioists. I knew Anthony Hewish, and Baldwin; and I talked with them from time to time; Shakeshaft, especially. I knew those people and I enjoyed talking with them. They were really quite sensible when you talked to them individually.

DeVorkin:

But the optical work had not really hit yet. It was just being done.

Harwit:

Yes, I think some of the optical identifications were being worked on by David Dewhirst, whom I also met at the time and talked with. He was very nice. He showed me the Madingley Road Library of which he was in charge at the time, and allowed me to work there. So we got to be on very cordial terms.

DeVorkin:

Yes. But that research didn't loom large in your own work?

Harwit:

No. I used the library system a lot there. I used the Madingley Road Library, the Philosophical Library, and sometimes the Cambridge University Library, which was an awful pain to work in, because they kept all their books on the shelves in the order in which they had arrived, and the catalog system was in huge leatherbound volumes, instead of a card file. It wasn't impossible to work there, but it was tedious to find things. Let me still mention Bondi one more time, because I did go down to this talk of his then, which really was just a sketch of what he thought might be done.

DeVorkin:

What is the date on that approximately?

Harwit:

It must have been in the late fall of 1960, that I had gone down there. His talk was, as I said, the following week after mine. After his talk he asked me whether I would like to work our the details of the set of non-linear equations that needed to be solved. Non-linear differential equations were not my metier but I thought maybe I could try it. Eventually I did solve the problem. It turned out I could do it in a closed form by doing something which didn't require me to solve the equations specifically. It just turned out that I could do the thing, and I sent that to Bondi. He was rather pleased with that.

DeVorkin:

Even though there wasn't any unique solution to the problem?

Harwit:

No, it did come out that there was a unique solution to this. I think it turned out that the density you would need in order to duplicate galaxies within the time allotted was something like five times the amount that was likely to be there in one of the relativistic models that Hoyle had worked out for the steady state theory. So Bondi liked that and he said I should publish it. He felt he hadn't done enough work on it, and so then I went ahead and wrote it up and published that also, as a follow-on paper.

Hoyle wasn't involved in that one at all. Since then I've always been on good terms with Bondi. For me it was just a very productive interaction. One of the things that was was interesting about that time was that Ryle came out with a second log N/log S curve, number of sources as a function of brightness. Initially there had been a slope of that plot which had a value of minus three and a half (-3 1/2), and at this point he had reworked it with his group and they had come up with a slope of minus two-point-two (-2.2), and the steady state was predicting a slope of minus one-point-five (-1.5). On the way to the RAS meeting where this was to be announced I saw some hawkers who were peddling evening newspapers out on the street right in front of Burlington House in London, and the headline was: "The Bible was right." This was because Ryle had already given the newspapers his results before announcing them to the Society, and saying that the steady state theory was out. The world had a definite beginning; therefore, the Bible was right.

DeVorkin:

Did Ryle say the Bible was right?

Harwit:

No, that was the headline.

DeVorkin:

That was the inference? I

Harwit:

Yes.

DeVorkin:

Did you buy a copy?

Harwit:

No. I don't recall buying one. I don't even know if the headlines actually said that. I think the headlines said it, but I think certainly the hawker was shouting it.

DeVorkin:

Was it the London Times?

Harwit:

No, it was an evening paper. I don't know whether it was the Express or what. I just don't remember. One would have to go back to that date. But then there was the most acrimonious debate I think I have ever witnessed. Ryle gave his results and said that the steady state theory was completely out by these figures. Then in the debate Bondi immediately took the word up and said he congratulated Ryle on this very nice piece of work. He was glad to see that the slope had come down from -3.5 to -2.2, and he had no doubt at all that with a little extra work that they would come down to the value of -1.5, which the steady state was predicting. He felt that this was really a great advance, that Ryle had made. Ryle was absolutely furious, he was red in the face. It was a real cat fight.

DeVorkin:

(laughs) Oh no! But this must have made people laugh openly.

Harwit:

It was frightening, I thought. It was really just so unseemly; here were these great scientists going at each other in the most disgraceful way. At any rate, that impression was quite vivid. I certainly never wanted to get involved in that kind of a fight with people. It was, I think, somewhat British. I don't think in America I've ever seen any fights quite as bad; although in Germany I've seen people going at each other in essentially similar ways, ad hominem, much more than we do in the United States.

DeVorkin:

You certainly weren't aware of it at the time; but were you concerned that this was going to be an inevitable part of your professional life? And did it turn you off?

Harwit:

No, I don't think so. I had seen enough of how things were done in the United States that I didn't feel that this was going to be any kind of problem; but it just struck me that this wasn't the way one ought to be doing science.

DeVorkin:

Yes. Sometime during your NATO year you had to think about what you were doing after that time?

Harwit:

Yes.

DeVorkin:

How did those plans develop?

Harwit:

Let me talk about one other person still before I do that. There were a number of other people who were at Cambridge at the time who were interesting. One of them was Schluter, a German, who had done some work on plasmas, and who then went back to the Max-Planck Institute for Extraterrestrial Physics in Munich and became the director of the Plasma Lab there. He then also worked on thermonuclear processes that they were trying to develop in Germany. But at the time he was a well-known person in astrophysical plasmas. Among the students the person I talked with for hours on end was Linden-Bell. He was very smart, obviously head and shoulders above everybody else at that time, I think. Among other students there at the time were N.C. Wickramasinghe, J.V. Narlikar, Joan Crampton, S.M. Chitre and Faulkner. They were students of Hoyle's.

DeVorkin:

Was David Edge there at all?

Harwit:

I've talked with him, and I think he may have been there a little bit earlier than I. I certainly don't remember him from that time.

DeVorkin:

I'd like to develop some understanding of how you eventually came to write COSMIC DISCOVERY; how you became interested in history, interested in the discipline, and developed a greater perspective than that of a pure practitioner? I thought maybe early contact with David Edge might be of interest.

Harwit:

Not at all. In fact, at the time I had no interest whatsoever in history. There were two things that I was thinking of doing if I went back to the United States. Before I left MIT, the head of the department there, Nathaniel Frank — he had written some texts with John Slater — had said that if I wanted to come back to do astrophysics at MIT, they might be interested. I think they were just going to get started there. So I wrote to him, I think. In the meantime he had been replaced as chairman of the department, and his successor wrote back and said they really weren't that interested any more, but maybe the people in what by then was the Department of Aeronautics and Astronautics might be interested. Those were the people whom I had worked with at the supersonic wind tunnel.

So I think I wrote them also. But Tommy Gold came through England in August of 1960, called up in Cambridge, and asked me if I could meet him in London. I said, well, I couldn't do that because Marianne was just in labor at the time, in the room next door (laugh) in our cottage. They didn't have enough room in the Cambridge hospitals. They would have taken her because she was an American lady, they said. But she wasn't going to have any of that. They just didn't have enough room in the hospitals at the time, so people were having babies at home quite often.

DeVorkin:

With a doctor?

Harwit:

With a midwife, and eventually the midwife came about 10 minutes before Alex, our first son, was born. I was trying to brush up on what I ought to do in case the baby was born before that, but it didn't come to that. Anyway, I told Gold I couldn't go down. But he said, if I wanted to come to Cornell, that he would be glad to have me there. I applied for, and got, an NSF fellowship then to go to Cornell for a year. After I had accepted that, I mentioned it to Hoyle, and he said, oh, that was too bad because he had thought of asking me to stay on another year. Of course, I would have loved to have stayed on another year, but I didn't see any way of getting support in England. I think, in the meantime he had found a way to support me. But I had already committed myself. So we went off to Cornell at that point.

DeVorkin:

So that was it then, a contact through Gold, and the offer of a one-year NSF postdoctoral Fellowship.

Harwit:

Yes, but it was a nine-month Fellowship.

DeVorkin:

But was it a research Fellowship?

Harwit:

Yes.

DeVorkin:

Still no teaching duties or anything of that sort?

Harwit:

No.

DeVorkin:

Let me start by asking, was this considered to be only a one-year appointment, or was there a chance that you would have been able to stay longer? What was in your mind as you moved to Cornell?

Harwit:

I really didn't have any plans. I had thought of possibly going back to MIT, because there had been talk of an assistant professorship there. I wasn't too terribly keen about going there as an assistant professor in aeronautic and astronautic engineering.

DeVorkin:

Yes. So your weren't ready or interested in jumping into the space program or whatever at that point?

Harwit:

Yes, I was, but I felt that the astronautics, was a gimmick. They really were not doing astrophysics or any of that kind of thing. They were still in the rocket and aircraft business, rather than in the scientific application of these vehicles. So I felt that I'd probably be misplaced there. I started working on a problem of interstellar matter while I was at Cornell. I worked pretty much by myself on that, and eventually published it by myself also. About the middle of the winter, or maybe earlier than that even, Tommy Gold asked me, whether I might be interested in an assistant professorship there. There also was an assistant professorship here. There also was an assistant professorship open at Johns Hopkins in the Physics Department. I did pursue the MIT position enough so that they made me an offer, as I remember. But I then decided I'd stay at Cornell.

DeVorkin:

And why so?

Harwit:

I didn't like Johns Hopkins. The person who invited me to come down there was sort of a stick-in-the-mud. I forget his name. I really wanted to go speak with John Strong, whose work I had always admired. He wasn't in town, but the people over in physics didn't seem to be too keen on him. He had a laboratory of his own there, and there seemed to be friction.

DeVorkin:

Is he still alive?

Harwit:

Yes, I think so. He went to the University of Massachusetts at Amherst later on. I have never met him yet.

DeVorkin:

He must be quite old by now?

Harwit:

Yes, he's been retired maybe 10 years, I think, so he may be in his late 70s or so. But he had done interesting things in optics in a style that I was interested in. He always had done very nice things. At MIT I was worried about what I have just mentioned; Cornell, on the other hand, had Ed Salpeter, and Phil Morrison was there at the time. Hans Bethe was there. Tommy Gold was there. It seemed to be a place that had promise, and they wanted me. Besides, I liked the layout. It's a nice place to live and bring up children; so all of those were factors. So what happened then was, I got an intermediate six-months appointment as a research associate from March 1962, when my Fellowship ran out, until fall 1962 when I would start being a professor.

DeVorkin:

What degree of contact did you have with Salpeter, Morrison, Bethe?

Harwit:

Oh, I'd see them from time to time. I saw Tommy Gold more than the others perhaps, because he was my titular boss. The others all were in the physics department; whereas, my appointment came to be in the Astronomy Department.

DeVorkin:

Right. By now, with the early Mariner flights and other infrared studies developing, 1962, although this might still be early, I know that there were some ground-based infrared studies of the lunar surface. Was Tommy Gold still very actively involved in that? And were you interested in it at all?

Harwit:

Tommy had some interest. His whole argument on the dusty composition of the lunar surface on the heat capacity and heat conductivity of the surface material, which he thought was only compatible with dust. He turned out to have been right. I didn't have that much interest in that, actually. Tommy and I have only done one paper together ever, with Mike Werner, a student of mine at the time. And it was something that had to do with the possibility of water vapor on the moon. It was a small sort of thing that we just happened to do together.

My main concern at that point was to work on things that interested me. Tommy had a lot of interests that he wanted people to work on. Again, as I mentioned, his viewpoint of what a director ought to be was somebody who brought the ideas in and then had people working on those. If that was going to be the way things were run, I wasn't going to stay at Cornell very long, I figured.

DeVorkin:

Why is that?

Harwit:

Because I had no interest in working on somebody's else's problems. It just wasn't the sort of thing that would be fun to do, for me at least. I don't know of anybody else who ever got a one-year appointment as an assistant professor at Cornell, but I certainly did. And I'm not quite sure why that was, but at any rate, that was the appointment I had. As soon as I had the appointment I also started looking around for other things that I might be doing a year later.

DeVorkin:

You must have had a teaching assignment the first year?

Harwit:

Yes.

DeVorkin:

What did you teach?

Harwit:

I taught cosmology with a very bright graduate student at the time, Peter Goldreich, who was really a phenomenon, and it was a tremendous amount of fun to do that together.

DeVorkin:

Was he your graduate student?

Harwit:

No, he was Tommy Gold's student, and I think Salpeter and Phil Morrison played a big role in his education, also. But he and his wife actually lived in the house of the Golds in a little apartment there. So his contact with Tommy was quite strong. He would give Tommy a hard time all the time, because he was just tremendously smart and mathematically he could walk rings around Tommy, whose strength is mainly good intuition. So we did that together. I think it was the same year that I started, or maybe the next year, teaching Beginning Astrophysics.

DeVorkin:

What textbooks did you use; do you recall?

Harwit:

There wasn't one at the time that I thought was useful; so I started making up notes and that went on for many, many years. I kept refining and honing the notes, giving them to the students, asking them to tell me what they found difficult or not understandable, making up problems for them, having them work them out. Eventually that gave rise to ASTROPHYSICAL CATS, the book that finally came out in 1973. But by that time I had been working on it eight years or so.

I liked teaching that course, because since I hadn't had the formal education on astronomy, I had to teach myself the material somehow. Each year I would try to introduce one new section, one new topic that I didn't know anything about, and that I then had to learn. The first time around, of course, it was always a disaster for the students. Eventually I got to the point where I could do all of those things, and it was really very good training for me, and I think it also helped the students some. This eventually led to ASTROPHYSICAL CONCEPTS which is used now, I think, more than any other astrophysics text.

DeVorkin:

It also serves as an introduction to professional astrophysics, rather than providing a purely descriptive account of the subject.

Harwit:

It treats astrophysics the way that the physicists teach solid state physics, or atomic physics at a junior-senior or first year graduate student level. Quantitatively, but in general, the aim was to leave out complicated calculations, and expose the student to the material in a form where he would notice the physics of it and be able to calculate order-of-magnitude effects; or calculate something out correct within a factor of two, rather than within two or three percent.

The point was that in many astrophysical processes, you know the basic conditions so poorly that it doesn't really help you to calculate something with that accuracy. What you want to know is whether one effect or another effect is the dominating one. If you can show that effect A is 10 times larger than effect B under even the worst conditions, you just forget about effect B altogether, except as perhaps a minor disturbance. I wanted the students to get that kind of a feeling, to home in on the important physics in a particular situation, and not get confused too much by calculating five or six different processes all to an accuracy of one percent, only to notice that the last three are really a factor of a million off, and won't play a role at all.

DeVorkin:

Yes. Did you find that that turned you off, when you found things like that?

Harwit:

Yes. It was kind of interesting. One of my colleagues, Ken Greisen, whose son was taking my course and telling him about it, came by one time grinning, and said, I hear the students are having a hard time understanding why you are having them work out the answers numerically, instead of just deriving a formula at the end; why you actually have them substitute numbers in, because in the physics courses they tended to just come up with a final equation. They figured, well, at that point anybody can plug in the numbers. But I actually wanted them to take that last step of plugging in the numbers to see whether something was plausible, whether what you calculated by this particular process was anything like one observed.

DeVorkin:

And more often than not, they found it wasn't.

Harwit:

Well, it depended. They might find that one process would do it and the other one didn't; but if you just had the formulae for the two processes, there really was no way of looking at these equations and knowing which was bigger. You really had to put the numbers in, which was pretty easy to do. But you had to take that last step.

DeVorkin:

To what degree would you say your book, ASTROPHYSICAL CONCEPTS, constituted your own formal education in astrophysics.

Harwit:

Very much, yes. A lot of the stuff I had never seen before, so it was really very good training, and also a lot of hard work. I spent thousands of hours on that book, literally.

DeVorkin:

Not every physicist who enters into astronomy builds a textbook, but they do teach courses in astrophysics. To what degree are you aware of your experience as being typical of how physicists learn about astronomy in a professional mode, such as teaching courses. Is is atypical, do you feel? Or is it common?

Harwit:

Well, I don't know. It's kind of hard to say. I was pleased a year ago or so, when Roger Hildebrand, whom I visited at Chicago, to give a talk there, mentioned that he was teaching something in astrophysics, and he had taught himself out of my book. It was designed for just such people, more or less. I had figured that there were a lot of people within NASA, and there were a lot of students who had a good physics background, but had never really been exposed to astronomy, and if I could explain the astrophysical problems to them in terms of basic physics, they would be able to jump right into that, and find it relatively easy to make the transition.

DeVorkin:

So you feel that is the source of the utility and popularity of your text?

Harwit:

I think so, yes. I think so.

DeVorkin:

Do you have any idea how many copies have been sold?

Harwit:

Over the years, I don't know, maybe 10,000, something like that. It was sold by the Library of Science Book Club at one point. So there were maybe four or five thousand sold right there. Right now it is selling at the rate of another three-hundred a year or so, world-wide. I think that represents pretty much the number of people who are taking astrophysics courses, because they might decide to go into the field, or at least would like to know about it peripherally.

DeVorkin:

Yes. How would you compare it to the book by Smith and Jacobs that came out a little after yours?

Harwit:

Yes. Smith and Jacobs is at a much simpler level. I think this book is at about the highest level before you start specializing.

DeVorkin:

Yes. That was my feeling. (Off the tape: In other words, your book would be the book one would read before reading the ApJ, so to speak?)

Harwit:

Yes. That was one of the things, I always said that I'd like to get students up to the point where they can read an abstract in the Astrophysical Journal and know whether it's important or unimportant, not understand the details of the article itself, but at least, have an idea of why that article was written and what sort of problem it tries to solve, how important that problem is, that kind of thing.

DeVorkin:

Did you start teaching in 1962-63 with the ultimate goal of producing a textbook; or did that just evolve?

Harwit:

No, that came much later. I don't know at what point I decided on writing the book. I really didn't know until the last two or three years whether I would go through with it. One of my colleagues, Kurt Gotfried, quoted someone as saying that no matter where you quit in writing a textbook, you always are ahead, because there is always so much more left than you think, that you still haven't lost as much perhaps as you ultimately are going to lose (laugh). But that first year then, there were two things I wanted do in research. One of them was to apply to the National Science Foundation for funds so that I could start building instrumentation to look for molecular hydrogen in the infrared.

DeVorkin:

And that was ground-based?

Harwit:

That would have been ground-based, yes. But at the same time also, I really wanted to do something from space, and I mentioned that to Tommy Gold at one point, maybe even before I started teaching. He said I really should talk to Herb Friedman, if I was going to do space work; he was the tops, he felt.

DeVorkin:

What did Tom Gold try to say to you about space? Did he encourage you? Did he try to discourage you?

Harwit:

He actually was somewhat discouraging. He really wanted to have somebody who would work on problems he wanted to have worked on. He brought in Mukul Kundu for a while, because he had interests in solar flares. That didn't work out too well. I don't know exactly what happened there, but Kundu left after a couple of years. He didn't like Ithaca too much, I think. It was cold.

DeVorkin:

He's here in Maryland now?

Harwit:

Yes, we've been in contact over the years. I just saw him again at NRAO a few weeks ago. Then Gold brought in Ian Axford who worked on magnetospheric problems, and who really did an awful lot with Tommy. And Tommy appreciated that and then promoted him very quickly.

DeVorkin:

That was what Gold was looking for. He was looking for a number of younger theorists who were willing to work on his problems.

Harwit:

I think so, yes. Gold didn't have the manipulative abilities. His strengths are that he can hear about a pulsar and in his mind imagine what must be going on. He works with conservation principles, with symmetries, with fundamentals, but if you have ever read any of his papers, there are no equations in them. They are theoretical, but they are not mathematical. He is just extremely good at that. I don't know anybody better. That was really the strength of the Bondi-Gold combination when they were collaborating. Bondi was this fantastic applied mathematician, and Tommy Gold had this good insight. Bondi also had good insight, but I guess that team was just very strong.

DeVorkin:

He allowed you to do this sort of thing, if you were determined to go through with it? He wouldn't try to cut you off?

Harwit:

Well, my feeling was that I only had a year's appointment, so I could damn well do what I pleased the following year.

DeVorkin:

So this was during your NSF year?

Harwit:

I'm not quite sure. I was looking through the correspondence a few months ago, and I think what must have happened is that I was in Tommy's office one time talking about wanting to do space work at Cornell. And he said, well, If I wanted to do that, I should work with Herbert Friedman, or talk to Herb.

DeVorkin:

Even though there was the Center for Radio Physics and Space at Cornell.

Harwit:

We didn't have anybody. It was in its infancy. There wasn't anybody there doing space work.

DeVorkin:

Even though the title was there?

Harwit:

Yes. That was just a come-on to get NASA money, I think.

DeVorkin:

Sagan wasn't there yet?

Harwit:

No, he didn't come until much later. Frank Drake wasn't there either. In our department there was a young man named John Cox, who now is at Colorado. I haven't met him since he left. He was an assistant professor. I think Tommy misjudged him and didn't renew his assistant professorship. He did a very nice two-volume set of books on pulsating variables.

DeVorkin:

With Giuli and Cox.

Harwit:

It could be. Cox was quiet, and Tommy wanted someone flamboyant. I don't know what. I actually got Cox's job when I was hired. felt badly about that, because I thought he was pretty good. Then there was an older man by the name of William Shaw, who couldn't get along with Tommy at all. It had been a two man department. Cox has been hired by Shaw. Tommy and Shaw never talked. I had heard from somebody once that Tommy never gave him a raise in the whole time until he retired, because Shaw had told him he wasn't going to cooperate at all. Shaw taught virtually all of the undergraduate courses, except for the one that I taught.

DeVorkin:

I wonder if this is the same Shaw of Shaw and Boothroyd?

Harwit:

Yes. Boothroyd had been at Cornell. Shaw had done some lunar work years and years and years before. He resented the university bringing in somebody from the outside to run the department, but it is clear that he would never have done anything much himself.

DeVorkin:

Yes, but in the great upsurge of activity in the late 1950s or early 1960s, wouldn't you say, a lot of departments were experiencing this kind of an overturn?

Harwit:

I don't know. It's possible. I really haven't followed that at all.

DeVorkin:

It is similar to stories I have heard elsewhere. There was an older teaching department, then in comes the astrophysics and the physics.

Harwit:

Yes. I think it must have been that the physicists at Cornell — Bethe, Salpeter and Morrison all had interests in astrophysics, as did Greisen — felt that around Shaw one couldn't build up an astronomy department.

DeVorkin:

So, Salpeter was in physics all the while.

Harwit:

Yes, he has always been paid out of physics, although he now is half time listed as astronomy and half time physics, but his salary still comes out of physics.

DeVorkin:

Well, let me identify, if this is appropriate to do so, two things. First, Tommy Gold suggested you go talk with Herbert Friedman, that's clear enough. But can you identify what eventually decided you to devote yourself to space work, and in addition to Tommy Gold's reactions, what were the reactions of other colleagues? How did they see your career going, if you were to start working in space? Did people, again, try to encourage you or discourage you? Harwit.

I wasn't too terribly worried about any of that really. I knew I wanted to do space work, because I knew that the atmosphere was just a terribly messy thing in the infrared. I thought if one was going to go into the infrared — and there was a lot of money in space work — then one ought to try to go directly into space, rather than fitting the atmosphere. When I mentioned that to Tommy, he actually picked up the telephone and called Herb Friedman in Washington and arranged for me to go down and see Herb. When I got there I talked to some of the people. Herb had wanted to get into the infrared; and I of course, wanted to get into rocketry. He said there was a new program that the NSF was running, what were called E.O. Hulburt Fellowships in honor of E.O. Hulburt who had been the superintendent of that division for a long time, and who had started space work.

DeVorkin:

These were funded by NSF?

Harwit:

These were funded by NSF, yes. I i

DeVorkin:

And not by NRL itself?

Harwit:

No.

DeVorkin:

And not by NASA?

Harwit:

Not by NASA, either.

DeVorkin:

Isn't that peculiar?

Harwit:

Herbert Friedman has good contacts all through Washington, and he had persuaded the NSF that they ought to put some money into that. That was before NSF totally withdrew from space work. That only happened about a year or two later. I went down there as the first Hulburt Fellow, and enjoyed that very much. It was terribly hard work. I never worked as hard. We tried building a first payload, which was going to go piggyback on an Atlas. I described all of that in the interview with Kondracki. Since I described it there, maybe I shouldn't talk about that any more.

DeVorkin:

Right. This will be dovetailed with that one. But I do have a few questions about the NRL year that we did not cover, that I would like to. Let me turn this over.

DeVorkin:

This is Tape No. Two, Side No. Two. First of all, you said you were the first Hulburt Fellow. There have since been other Hulburt Fellows, and I am wondering if you are aware of who they were?

Harwit:

Paul Feldman came as a Hulburt fellow. And I think Richard Henry came as one also. They came very shortly after I did. There were a lot of others, but I don't know.

DeVorkin:

Were people there earlier like Kupperian and Boggess there under other auspices?

Harwit:

Kupperian and Boggess had left by the time I got there, and had joined NASA. But they were staff members.

DeVorkin:

So they weren't these fellows?

Harwit:

No.

DeVorkin:

Okay, fine.

Harwit:

That was started around 1963 when I went there.

DeVorkin:

Had you been aware at all of Friedman's ultraviolet work where he had detected, or thought he had detected these large ultraviolet halos around stars?

Harwit:

I don't know how much I had been aware of his work before. I know I started reading up on what he had done once I became interested in going down there. I don't know whether I had heard his name before Tommy Gold had mentioned him, maybe once or twice. But I hadn't taken that much of an interest in him.

DeVorkin:

Had you had any impressions of how science was done at NRL from the knowledge of this type of detection? Were you aware of the fierce competitions in x-ray observations?

Harwit:

Well, those were just barely coming on. Don't forget that the AS&E group only discovered the x-ray stars in 1962. I went down to NRL in the fall of 1962, and the AS&E effort had just made a big splash at that point. But it wasn't clear what they had found, if anything, because they had such poor pointing information that one really didn't know very much of what was going on.

DeVorkin:

Yes. Did you experience any of the contacts between the AS&E group and the NRL group while you were there?

Harwit:

Not that much really. While I was there they dial prepare for the rocket shot which looked at the Crab Nebula and the lunar occultation. I thought that was really a very neat thing, and a very impressive effort to prepare for on seven weeks' notice, I think. They found there was going to be an occultation and they jumped and took advantage of it. That was really the first location of an x-ray source. It is true that what the AS&E people discovered, later on turned out to be an x-ray star, but at the time they didn't know what they were seeing yet. They just knew there was something in that part of the sky, and it was many tens of degrees across, if I remember correctly.

DeVorkin:

Yes, Scorpio X-1?

Harwit:

Yes, in fact, they may have seen a combination of Scorpio X-1 and the Crab Nebula, again if I remember correctly.

DeVorkin:

That's a rather large chunk of the sky.

Harwit:

Well, Yes. I don't know exactly how it went. There were a lot of speculations about what it was, and successively, the NM and the ASO group narrowed down the field in which the Sco X-1 source was. Friedman was telling me one time that, I think, they had narrowed it down to within a degree or so from the initial positioning that the ASO Group had done. I think the ASO group with Oda from Japan managed to narrow it down even more. It certainly was a collaborative effort. It's clear in retrospect that the AS&E group has to be credited with the first extra-solar system verifiable detection. The NRL people might have had an earlier one, which Friedman has told me about, that he and Kupperian mentioned some place; but they weren't able to verify it, and it may have been that it was a transient, or the second time they looked the source was below the horizon. But they had had one flight in which they felt that they saw an additional source. Kupperian at that time went to NASA, and they just never followed it up.

DeVorkin:

Did some of this excitement stimulate you to continue in the field, and possibly do the same thing in infrared?

Harwit:

Sure. That was the whole point, yes. I was interested in the infrared. When I went down and talked with Friedman we settled what could be done: a small telescope that would be cryogenically cooled. That suggested background observations. Hoyle had once mentioned that there was quite a lot of missing radiation that ought to be around some place, if you just looked at the total amount of hydrogen that had been converted into helium in stars.

DeVorkin:

You mean from the big bang?

Harwit:

No, he was a steady state theorist. If you look at the amount of radiation that gets converted from hydrogen into helium, and you have a steady state universe, even a universe which is roughly flat and expanding at a rate that we see, then you can calculate that there ought to be approximately the same amount of background radiation somewhere else. That is still true, if the bulk of the helium we see nowadays, or that is existing in the universe now, is due not to the early big bang, but to the helium formation in stars. One doesn't really know just what the relative fractions are, because some of the helium of course still will be stored inside these stars at the moment. So from that point of view, I was interested in looking for background radiation, because it might perhaps clear this up. You would see this then as starlight which was red-shifted into the near-infrared.

DeVorkin:

Yes.

Harwit:

It is kind of interesting; last week I got a paper from a Japanese group which has done a survey of the near-infrared at a sensitivity level which is much higher than we were able to do in the 1960s. In between there hasn't been any effort of this kind at all, with the exception of a still classified Air Force project which Steve Price told me about two or three days ago. But at any rate, they have come up with a signal which is rather high, and would suggest an active phase of galaxy formation, with very bright galaxies at that time, probably too bright to reproduce the hydrogen helium conversion. They might have had to have formed black holes there, if they are correct. But that was the kind of thing we wanted to do with background studies.

DeVorkin:

Let me ask you about more mundane things, not so much the science, but the NRL atmosphere for doing science. You now had been at a good number of places. Now you were again at a military center of research. How would you contrast it, first with the academic atmosphere that you were used to, and then with the Chemical Corps atmosphere that you experienced when you were in the service?

Harwit:

The Chemical Corps was Mickey Mouse compared to NRL. It was bush league stuff. They didn't have people who were world class. NRL and, I think, maybe National Bureau of Standards were the only government laboratories I knew of at the time that were world class. Since then the Government establishments have become quite a lot stronger it seems to me. But during my military career I had gone into a number of different labs just to visit, and they really all were very low quality, just disheartening low. I thought NRL was on a par with university research.

The one thing that they lacked, perhaps, was that they didn't have a strong theoretical background. Friedman would come back from talks that he gave at various universities, and then tell people what the theorists there had told him was interesting to look for, or ways to interpret the data. That's why I think he and Tommy Gold hit it off, because he would listen to the theorists. Tommy was always the kind of person who was glad to get in on things like this, and make suggestions.

DeVorkin:

That is quite interesting. You have answered pretty much both sides of the question. That brings up an interesting one about Tommy Gold, considering what you had said before about him. At this time, did he have any interests for taking over a bit of the space program for his own interests, and did he see you as an entering wedge?

Harwit:

I think that's quite possible. At any rate, while I was at NRL I was given tenure at Cornell.

DeVorkin:

That's interesting, because in our talk with Kondracki you had not had any flights while you were at NRL. You were really just learning the ropes and doing a lot of hard work. What was the criterion for getting tenure at Cornell, or was it easy (laughs)?

Harwit:

Well, in the first place, it was easier at the time. In the second place, I had done reasonably good theoretical work, both by myself and with one of the leaders in the field, Fred Hoyle. Thirdly, it was pretty much a sellers' market. Astrophysicists were in demand. I mentioned to Peter Goldreich one time, in passing, that I wasn't sure I'd go back to Cornell if I didn't get tenure.

I was a bit incensed at getting a one-year professor appointment, but I figured that that worked both ways. If they gave me only a one-year appointment then I had the freedom to leave. I never brought up the subject with Tommy, I don't believe. But of course Peter lived in their house, and he always was goading Tommy, so it may well have been that he mentioned it. I don't know. At any rate, I got tenure.

DeVorkin:

Yes, that's interesting. Had you considered ever staying at NRL. Was there a possibility there?

Harwit:

I don't really know. I think Friedman really did want to have people coming in and going. He was a person who liked quite a lot of control also, and I think he wanted to have more control over the infrared group, and probably eventually to have one of his own people in there.

DeVorkin:

Yes, as you said last time.

Harwit:

Yes, and I really wanted an academic environment. I enjoyed teaching. Probably because my father was a professor, I thought I might also like to be one. Also I used to sometimes get annoyed at the way the military ran things. They could suddenly cut out all overtime, and that happened a number of times while we were down there, and you couldn't keep on working at the rate you had been.

DeVorkin:

Can you explain that? Do you mean they would come in with some special requirement, some crash program that you would have to do?

Harwit:

We were working night and day to try to meet deadlines. The commander at the laboratory would suddenly decide that there was not going to be any more overtime paid. This was at the time when really there was quite a lot of money around, so I don't know what the reason was. I think he probably thought people were abusing it. Work would cease for a few days until the dust settled. We would work a regular 8 hour day, or the technicians would. The scientists would continue working and then we would get a hearing and explain what the problems were. And then, yes, okay, we would be allowed to go ahead and work overtime again. That sort of thing happened a couple of times.

DeVorkin:

Trying to meet the launch date of the Atlas, for example.

Harwit:

Yes, that's right.

DeVorkin:

Those were pretty hard and fast deadlines. Let's clarify what kind of detectors you first started using. And then if you could, could you give me a brief review of the history of infrared detectors that you are familiar with?

Harwit:

Well, it has been quite a while. We were looking at a number of different detectors for the near infrared with the liquid nitrogen payloads we were going to fly down there. Indium atiminide was one that figured strongly; Indium arsenide was just coming in at that time as a possibly faster detector than lead sulfide. We were playing around with both types. Eventually we went with Indium arsenide.

DeVorkin:

How do you contrast these to the bolometric germanium-doped detectors?

Harwit:

Okay, there are two types of germanium-doped detectors. There are the bolometers and also the photoconductors. The bolometers really measure heat. The photoconductors are sensitive to a particularly wave-length range of photons, and will have a spectral response where some wavelengths are responded to very strongly and others not at all. The bolometers have a flat response across the spectrum. At that time bolometers were only just barely coming in, and I think Frank Low was the only person who was using them. I think he was at NRAO at that time, if I'm not mistaken.

DeVorkin:

Before TI?

Harwit:

No, he had been at TI, and at some point he transferred to NRAO at Frank Drake's suggestion, to concentrate on building bolometers for astronomy. Soon thereafter he went to Rice. I am not quite sure when. I think he must have gone to Rice around 1963 or 1965. I didn't know him at the time.

DeVorkin:

We're talking about the fact that balometers were for the thermal infrared. Could mercury-doped germainium crystals be used both for photoelectric and bolometer detectors?

Harwit:

You use a different doping for the bolometers than you do for the photoconductors. The photoconductors were what we were using. The photoconductors were very fast detectors. Their response time is something like a 10-millionth of a second. At that time it was immeasurably fast. The bolometers typically had response times of a fraction of a second. With the rockets we chose to go with the fast response detectors, because one was going to be sweeping out relatively large portions of the sky. The first rockets we were going to use didn't have any pointing control. They were just going to cone around.

DeVorkin:

This is what we have here.

Harwit:

No. This one has a pointing control. But the first ones that we flew at NRL didn't have any pointing controls. They were just despun, and then were going to yaw around. So different parts of the sky were going to come in and out rather quickly, and we wanted to make sure that we could identify where we were looking and what we were seeing at a given time, in case we were seeing sources rather than just background radiation. Even then I had started doing some theoretical work also to try to determine what we would be seeing.

One of the things that I calculated at the time was what the contribution of the zodiacal light was going to be in the infrared. I was quite surprised to see that in the mid-infrared it was going to be the dominant limiting source for infrared observations, if you didn't go out of the solar system altogether. That was quite surprising at the time. But while we were building up these things, I also did some calculations on potential sources that we would see, because none of that had been done before.

DeVorkin:

Yes. This was around 1963-64, still before the 2.2 micron survey, before any comprehensive surveys. Had you had any contact with other early infrared people, such as Frank Low, Neugebauer, Leighton, Marts.

Harwit:

No, not really. I guess there was a symposium on infrared work at Liege in the fall just before I went to NRL. I think I met Low and Neugebauer there, but I had never heard of Neugebauer before, and he hadn't done anything yet either. Murray and Wildey at Caltech were the people I had heard about. They were using mercury-doped germanium photoconductors for some planetary work they reported on, and that was interesting at the time.

DeVorkin:

Westphal was working with them at the time?

Harwit:

And Westphal also, yes.

DeVorkin:

Yes, they had been working on the moon with detectors that were considered to be secret. They were just provided to them. They had contact with China Lake, the Naval Ordinance test station, which made the detectors available to them. What kind of contact did you have at NRL to make detectors available to you, and did you find that you ran up against any kind of classified technology screen?

Harwit:

If I remember correctly, Friedman never wanted to get anything that was classified. He wanted to be able to publish, and so he never got into that at all. It didn't seem necessary anyway. Their detectors needed lower temperature cooling, so if we were going to stay with liquid nitrogen it didn't make much difference. We also were interested in and flew gold-doped germanium which is an infrared detector out to 7.5 M.

DeVorkin:

Did you obtain your first infrared photoconductors – you always worked with photoconductors — through NRL?

Harwit:

We just bought them.

DeVorkin:

These were off the shelf?

Harwit:

They were made up especially for us.

DeVorkin:

Who made them?

Harwit:

Different detectors were made by different places, and I don't even remember that.

DeVorkin:

Is there anything else, about your NRL Year that you feel is important in understanding your future in infrared work?

Harwit:

I think we covered most of that in the session with Hank Kondracki, if I remember correctly.

DeVorkin:

Concerning this particular payload right here, the one that is in your office and that we are collecting, the one, for the record, that was flown in 1967, is it fair to say that it is a direct outgrowth of your NRL work?

Harwit:

I, think so, although by that time we were trying to do liquid helium cooling. As I said, we tried to cool this to liquid helium temperature, tut then failed, and flew it with liquid nitrogen.. But yes, it definitely is a direct outgrowth.

DeVorkin:

What did you learn from this infrared instrurrment that allowed you to make improvements later?

Harwit:

I guess, to keep the geometry of the telescope very simple, no intersecting cylindrical surfaces because you had too many strains. We also learned to try to keep heat paths as long as possible. By that I mean, that you have liquid helium separated from room temperature by only a few centimeters, and what you want to do is to somehow artifically increase the resistance to heat traveling in from the out-side to the liquid helium. For me it was a good engineering education where one learned about properties of materials that were strong, had either high or extremely low heat conductivities, and certain electrical properties and so forth. I enjoy building things.

There is this abominable phrase that one always hears Robert Oppenheimer being quoted as having said: "technologically sweet". But you know, there is such a thing as saying all of a sudden, gee, this is clever, this particular thing is economical. It's clever and you make the best use of available materials without going out and doing something terribly expensive. I felt that a lot of the engineering that we did was of that nature. So I enjoyed that.

DeVorkin:

Had you had the intention, as you were building these sounding rockets, eventually to apply the same designs to satellites?

Harwit:

Yes, absolutely. That was the whole point. I knew that there was no sense in fooling around with 5-minute observing times. That was always the goal, and I put in a number of unsolicited proposals, each of which got refined. Other people did, too, and I used to read proposals that industry made. At that time people like TRW would come around and show you proposals that they would like to make, and try to solicit your support for them.

DeVorkin:

This was in the mid to late 1960s.

Harwit:

That's right.

DeVorkin:

In doing so, did you ever get close to actually getting on a flight, or which flights did you actually go for, which kind of satellites?

Harwit:

They were IRAS satellites.

DeVorkin:

So you were proposing satellites rather than experiments on pre-existing programs?

Harwit:

On satellites, yes. None of them came through until IRAS came through.

DeVorkin:

What were the rationales for not being supportive?

Harwit:

I think NASA felt that they were not ready yet for them, and I didn't have the best relations with the other infrared astronomers.

DeVorkin:

Why was that?

Harwit:

I'm not sure, but NASA would have advisory panels throughout the 1960's, that usually had somebody from Caltech. It started out with Guido Munch, who was replaced by Gerry Neugebauer, when Gerry became known better. Frank Low was on the panels initially, and Nick Woolf was on there later. He was replaced by Ed Ney.

DeVorkin:

Wasn't he Caltech?

Harwit:

No, he was Minnesota. He got into the infrared from cosmic ray work.

DeVorkin:

What's the link?

Harwit:

He just wanted to switch.

DeVorkin:

It wasn't a programmatic link?

Harwit:

No, he told me that he thought there were too many people coming in or something. He wanted to do something else, whatever it was, he switched.

DeVorkin:

It's the same Ney that was doing cosmic ray work in the Naugle group that was active there since the 1940s and 1950s. Is Ed Ney still active?

Harwit:

Yes, as far as I know he is. He is still in Minnesota. He was sick a year ago or so, had a heart attack or something.

DeVorkin:

You are actually identifying the other infrared groups. 1,

Harwit:

Right, and they would meet first with Nancy Roman, and then I guess there was somebody called Ernie Ott. And then Nancy Boggess too over infrared efforts. Bill (William F.) Hoffmann came on later also. He had worked with Nick (Neville J.) Woolf on the earliest balloon work. Nick then dropped out of that very quickly.

DeVorkin:

Nick Woolf was in the infrared?

Harwit:

Yes, Nick Woolf was in the infrared for a while. He was at Princeton and he started building a small balloon for infrared telescope, which eventually Bill Hoffmann took over and stayed with.

DeVorkin:

At Princeton, or some place else?

Harwit:

No, at NASA Goddard in New York.

DeVorkin:

Didn't John Strong have any kind of infrared project?

Harwit:

Yes, he did also. For some reason or other, he never seemed to come to these meetings. Maybe he wasn't supported by NASA. Being in balloons, he might have been supported perhaps by National Science Foundation. I am not really sure.

DeVorkin:

The Air Force or Navy, also.

Harwit:

Yes, he had a little Air Force support actually. Time after time there would be a request to prioritize the different types of efforts. The prioritization always ended up with rockets at the bottom of the list. People felt it was a waste of money. They were quite expensive. I was getting $100,000 a year while other people were getting maybe 10 or 15 to do ground based work or even airplane or balloon work. The grants were small. It was just expensive to do rocket work, more so than ground-based work. Everything has to be put together much more expertly because it's got to work.

DeVorkin:

You recalled when we had the interview with Kondracki that Nancy Roman had encouraged you after your return to Cornell to continue with the infrared rocket work. Did she continue to encourage this?

Harwit:

Yes, she did, although I think after a while she may have had doubts. These panels always put rocketry down at the bottom of the list, after ground-based work and after ballooning. There usually were two or three people in each of the other fields, and it seemed to me that Gerry and Frank both agreed that ground based was very important. Frank had an interest in ballooning, and so did Nick Woolf or Bill Hoffmann. So those fields would get good grades. I was the only one in rockets, and the results we were getting initially were not all that spectacular. We had a lot of failures. We had a lot of problems, and people would say, oh well, these rockets are really not necessary. What we need is a satellite, and so an infrared satellite would always get high grades. These people all felt somehow the way to a satellite was through ground-based work. That was the natural way to go.

I felt that technologically you had to go through rockets. I still feel that the cost overruns would have been much smaller, perhaps unnecessary. IRAS had pretty big cost overruns. If there had been more experience with rockets, if they hadn't been cut out by NISA, I think this could have been avoided. There was much less rocket experience in the infrared than there had been in the ultraviolet or in x-ray astronomy before those satellites were built, and technically the liquid helium conservation in the telescope was a much more difficult problem than any of the others. It is just difficult to do. We were finding out how to do the thing by doing it. We actually got some Air Force support later on. The moment we got the Air Force support the NASA support was cut out. I think maybe they had been thinking of cutting it out, but didn't want to leave us stranded. When they found that they were off the hook by having the Air Force support us, they cut us out.

DeVorkin:

When was that?

Harwit:

This was around 1969, I think.

DeVorkin:

Was that after the Mansfield statement?

Harwit:

Well, the Air Force could support us, because they had an ARPA-funded survey they were supposed to conduct and they hadn't been able to get their own payloads to work. For a very small amount of money they were able to get us to rejuvenate our own payloads and fly them.

DeVorkin:

A small amount to them, I take it.

Harwit:

Yes, they were giving us $60 or $70-thousand for a flight, and I don't know what they got, but they must have been spending tens of millions, or at least several million dollars a year.

DeVorkin:

What were they flying; were they also flying Aerobees?

Harwit:

I think they had Aerobees also. They may have had Black Brandts, or Nikes. I'm not sure.

DeVorkin:

What was the Air Force's interest in the infrared?

Harwit:

I think they wanted to make a map of the sky so they would know what the natural sources looked like in order to then be able to track militarily interesting targets, or to decide what spectral range perhaps was optimal to track militarily interesting things, rather than stellar, I never found out.

DeVorkin:

I see. Okay, let me ask also about the difficulties of rocketry research. You noted in your interview with Kondracki that you needed a successful rocket shot for promotion at Cornell. You said you were serious about that. How serious?

Harwit:

It was just that actually, yes.

DeVorkin:

Well, did he say, if you can't get a rocket shot, get out and do something you can do?

Harwit:

No, it wasn't that.

DeVorkin:

This was about the time you did get promoted to a professorship in 1968?

Harwit:

Yes. In the first place, I had only been an associate professor for four years when I got promoted to full professor, and so to that extent there was nothing that prevented them from waiting the full length of time before making the decision. But I mentioned to you that Ian Axford was there. He was about the same age as I was and was working on something with Tommy, and he got promoted much faster.

I felt that I was running a program that really was much more demanding, was bringing in money, and was supporting students; whereas, he was really just working with Tommy and having students out of Tommy's grants, and so forth. From the point of view of enriching the department, I felt that I was really carrying my share of the burden and that that ought to be appreciated. There was also the fact that this was a difficult problem, and it was an important problem. There wasn't anybody else who was doing it or able to do it. Also I was doing my share of undergraduate teaching, in fact, of the people who were active in research, I was the only one who was doing any undergraduate teaching.

I felt I really was pulling my weight there, and that ought to be rewarded, that they ought to have enough confidence in me to promote me. Tommy Gold at that point said he felt that I really ought to have a successful shot in order to make it easy to get me promoted. Then just about then we did have one that was reasonably successful.

DeVorkin:

The successor to this one? The first one that was actually helium cooled?

Harwit:

Yes.

DeVorkin:

Let me just ask you some quick questions to finish this session. First, I would like to know how and when you became involved in the NASA airborne program, and was this a migration from rocket sondes to airborne, or did you do both at once for a while?

Harwit:

No, that was pretty easy, actually. The NASA people almost in the same breath said, we won't support rockets any more, but if you would like to do airplane work we'll support it.

DeVorkin:

When was that?

Harwit:

It must have been around 1972, 1973, I think.

DeVorkin:

So Frank how's prototype had already flown for a few years?

Harwit:

Oh yes, that had flown for a while.

DeVorkin:

The 36-inch hadn't been built yet?

Harwit:

No, there was discussion of that coming up, and in fact, I sat in on a panel on that fairly early on, but then wasn't actually involved in any of the construction considerations.

DeVorkin:

So it was a forced migration from one technique to another.

Harwit:

Yes, absolutely. Well, there were two things at the time that were of some interest. We could continue doing the rocket work a while with Air Force support, and then also NASA was interested through Bill Hoffmann, who was working as a NASA employee, in building a helium-cooled balloon payload. Jim Houck and I talked about how we should divide things up, and I told him he had the choice. I would do whatever he wasn't going to do and he decided he wanted to continue with the rockets, so he did that.

DeVorkin:

Yes. That's at Cornell.

Harwit:

At Cornell, so he flew a couple of more rocket flights with his students, which I had nothing to do with, and I started working with Bill Hoffmann on a collaborative program to build a liquid helium cooled balloon telescope. He is still flying that, actually.

DeVorkin:

Were you actually in the project when he started flying it?

Harwit:

We reported together on a first flight, and then I withdrew. It just became impossible to work together, because he had moved to Tucson from New York City, and I felt we weren't really helping out enough at Cornell. In fact, I felt quite embarrassed about how bad a showing we had made. I had inherited a student of his as a post-doc and it turned out that the person was untrustworthy. It was an embarrassing situation for both Bill and me, and we decided somehow we'd pull it out as much as possible. One of my students went and worked with him for two years in Tucson, and finished his Cornell degree while working out there. As far as I was concerned, doing a preliminary paper that we got our names on, was sufficient reward for several years of agony. Just technically we hadn't done all that much.

DeVorkin:

I see. Okay, I think it is a good time to stop. Just for the tape, the last session will include your airborne work; and then a review of space science at Cornell: its origins to the best of your knowledge of major groups, major successes and failures,competition for funds with Aericebo, and also, the role of Carl Sagan when he came. I'd like to know what he did at Cornell. And a few questions on Space Telescope, how closely you followed the developments, and your advisory roles in NASA. Here is a question I can ask right now. Have you ever heard of the Wildey effect? After W ildey at Mt. Wilson, Caltech.

Harwit:

I don't remember. It sounds vaguely familiar. But can you remind me?

DeVorkin:

Well, in the early 1960s there was some criticism of the infrared data that was coming out of Mt. Wilson, and it was attributed to the Wildey effect.

Harwit:

I don't remember that.

DeVorkin:

Okay, were you aware that there were problems with the infrared data?

Harwit:

The background studies we were doing had so little to do with the background studies of ground-based people that when they got together and started comparing notes, I usually didn't have any contribution to make.

DeVorkin:

Okay, fine. That is all I need to know.

Okay, this finishes this session. Thanks a lot.

Session I | Session II | Session III