Oral History Transcript — Dr. William Gordon
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.
See the catalog record for this interview and search for other interviews in our collection
Interview with Dr. William Gordon
William Gordon; November 28, 1994
ABSTRACT: This interview with William E. Gordon was conducted as part of a history of planetary radar astronomy underwritten by NASA. The topics covered reflect the narrow requirements of that history. Gordon begins with a brief mention of the Waynick Memorial Lecture, before discussing the work on radio propagation and scattering that led to his dissertation work at Cornell University under Henry Booker in the department of electrical engineering. A further consideration of the dissertation work on scattering provided Gordon with the theoretical foundation for the conception of the Arecibo Ionospheric Observatory. Gordon relates how he went about designing and siting of the Arecibo facility with the help of Cornell University colleagues, then he recounts the search for project funding, which led to the adoption of a spherical antenna design devised by researchers at the Air Force Cambridge Research Laboratories. A key moment in the creation of the Arecibo instrument was the experiment, conducted by Bowles at the National Bureau of Standards, which demonstrated the validity of Gordonís scattering theory. After briefly addressing the addition of radar and radio astronomy to the Arecibo instrumentís research agenda, Gordon discusses a number of the problems he encountered after completion of the facility: the conflict with Gold, difficulties with the feed and the need for Gregorian optics, and friction with the NSF and ARPA. He also mentions his reasons for taking a position at Rice University. Gordon concludes the interview with a review of the current Arecibo upgrade and sketches of his former Cornell colleagues.
Butrica:I understand that you gave a talk on the history of the Arecibo Observatory.
Gordon:Not too long ago. It must have been in the spring  at Penn State. It was a Waynick Memorial Lecture. Arthur Waynick was in URSI [Union Radioscientifique Internationale] forever. He died ten or more years ago. These are a series of memorial lectures. My lecture at Penn State was titled something like ďArecibo From Start to Finish.Ē I donít really have a manuscript. I talked from notes. I find thatís much better in a talk. But they did do a video, and I have a videotape of the lecture. If youíd like to borrow it...
Butrica:Yes, if I could.
Gordon:The sound is somewhat spotty, and the introduction is awful from a sound point of view. But that doesnít matter. It was with John Matthews, who didnít have a mike on his lapel. I had a mike on my lapel, and I think you can understand virtually all of it. Iíd be glad to loan you a copy of the video if you think that would be helpful.
Butrica:Sure. Absolutely. Letís begin with your research on the ionosphere and how the concept for Arecibo began.
Gordon:Itís a long story, but Iíll try to keep it short. Most of my career, beginning with my Ph.D., had to do with radio scattering from something or other, including some part of the atmosphere. My Ph.D. thesis, which I did under Henry Bookerís direction at Cornell, was on what was called tropospheric radio scattering; it was sometimes called tropospheric forward scatter. It dealt with what you could do, if you set a transmitter on the ground and sent a fairly strong signal into the troposphere. Somewhere, around the curve of the Earth, out of sight of the transmitter, you could receive a signal that was scattered from the troposphere. My thesis essentially described that process. Itís a turbulent process in the atmosphere. A small amount of the radio wave energy is scattered, and itís scattered in particular directions, depending on the shape of the scatterer. What I did was calculate what the scattering would be at various angles, including the more or less forward direction, which is slightly off forward, if youíre going around the horizon. But you can calculate what the scattering would be. Then if you go out and measure it, youíve got something which agrees pretty well. Booker and I got into this because the Navy at the end of World War II did some radio propagation experiments on a ship, and they were receiving the signals well beyond the horizon when they shouldnít have. These were microwave signals. On the sea, itís difficult to sort out this because ducting occurs often, and ducting is very effective in making radio signals follow around the curve of the Earth. But in cases where ducting was not present from weather measurements made in the atmosphere, they still received signals, and they were quite different from the ducting signals. The ducting signals are essentially at free-space level. The scattered signals are tens of decibels below the free-space level. So, you can tell from the signalís strength and appearance what it is youíre looking at. At any rate, these went unexplained for a while. In a sense, in the thesis there is an explanation provided for that data and for many other propagation paths. Tropospheric forward scattering was quite a successful communications link over distances that were non-optical, that is, not line of sight. If you were going from one island in the Caribbean to another which was around the curve of the Earth, it is a pretty effective way of communicating. It was widely used in Europe by NATO after World War II for the same purpose. Instead of having a repeater station every 50 miles, you could have one every 200 miles. So you could span larger distances. Well, thatís the beginning of it. Thatís a form of scattering and often called forward scatter. It works quite well in the troposphere, and it was a commercial success as well as a technical success. We also looked at the idea: Would there be scattering from the stratosphere? Higher heights than the troposphere, and therefore longer paths. The answer there was quite marginal. You could receive signals, but it took great big antennas and lots of power, and it really wasnít practical. So stratospheric forward scatter, while it existed, had no direct benefit to anybody, even the military. On the other hand, Booker on his own did forward scatter from the ionosphere. This was a highly successful form of longer distance communications, distances like a thousand kilometers.
Gordon:It was the same thing as Iíve described for the troposphere. You send out a signal. It travels into the ionosphere. The ionosphere is turbulent. Some of the signal is scattered. The scattering laws are a little different now, because itís an ionized medium, and the frequencies are down where theyíre more sensitive to the ionization. But you could receive signals a thousand or more kilometers away from the transmitter, and that was widely used in the DEW [Defense Early Warning] Line in the Arctic. They had radars all across North America, and they had to get the information back to someplace where it could be used. So having dependable communication paths of the order of a thousand kilometers was significant to the military.
Gordon:It was never used commercially. Bell Telephone (AT&T in those days) never found a way to use it in any practical sense. And they didnít need it then. But thatís scattering. Now, these things developed somewhere between about 1952 and 1955. Ionospheric forward scatter, incidentally, occurs in the lower ionosphere. So the next obvious question was: What about the upper ionosphere? Could you have forward scatter from the higher heights in the ionosphere, say, two or three hundred kilometers, instead of 90 or 100 kilometers? The answer was something that I stewed about. I burned the midnight oil, and one morning I came into the Cornell campus all excited, because I had recalculated the electron scattering of radio waves. I mean, this is really getting down to basics now. Nothing is correlated; itís all simply single electrons scattering and adding up the powers from a bunch of incoherent scatterers. ďYou know, when I was a grad student at Cambridge, I calculated this in some electromagnetic theory course.Ē He said, ďSo thereís nothing new about that calculation.Ē But what I had added to it was the following: I said hereís the scattering. If you were to shine a radio signal on it, what would come back? What would come back is very small, of course. So there was no point in talking about a communication link from one place on the Earth to another, because it was going to require a lot of power and big antennas. So I said: Suppose I just wanted to measure how many electrons were up there, and I built radar instead of a communication system? Iíd point the antenna straight up. How big an antenna would I need? How powerful are the current transmitters? How good are the current receivers? How far away are we talking about? All of those things are known. So it was a practical solution to a problem. But the question that had to be answered was: How big an antenna would it take? And that was enormous. That turned out to be a dish a thousand feet in diameter.
Gordon:You can argue whether thatís practical, and lots of people thought it was a pipe dream. Lots of people didnít believe that the electrons would scatter anything, although they were wrong. Even [J. J.] Thomson knew about electron scattering a hundred or more years ago. At any rate, if you could build a 1,000-foot antenna (thatís a big if), what could you do? You could measure how many electrons there were at a particular height. You would have some kind of thermal spreading of the frequency you illuminated the electrons with, so you would have some idea of the temperature of the electrons. If the electrons were moving toward you or away from you with any speed, that would shift the frequency, you could have a measure of the wind. In effect, you would have a measure of the weather if you think of it that way, at heights where data were not available by other means. Somebody might fire a rocket, although those heights are hard to reach with a rocket. This was early 1958, before Sputnik, and nobody was up there yet with satellites. It was an exciting prospect to be able to measure the atmospheric properties. It turned out there are many properties once you get into it seriously. But even the first ones, the idea of the weather up there, were enough to say, well, maybe itís worth doing. And if it is, all we have to do is build a 1,000-foot dish. So, could you build a 1,000-foot dish? I was at Cornell in those days in the Electrical Engineering Department. So I went around to see some friends in civil engineering, George Winter and Bill McGuire, and said, ďI have an idea about making some measurements that would be interesting to make, providing you could build for me a 1,000-foot dish.Ē They gulped and said, ďWell, here we go again.Ē At the time, Cornell was actually involved in building the Fermi Lab in Batavia, and theyíd long since built a synchrotron on campus. So they were used to having tough questions thrown at them, and they looked at them not as, well, this is some crazy thing, but looked at them as a challenge. It wasnít very long before they began to think, well, maybe you could build a 1,000- foot dish. But you needed to tell them the specifications: How smooth did it have to be? What direction was it going to look? It was going to look up as far as I was concerned to see the ionosphere. Where were you going to put it? Did it have to move? The answer was, no. They werenít going to build a dish that moved at that size. But at the same time that we were engaged in the beginnings of Arecibo, Jim Trexler, who was at NRL [Naval Research Laboratory], was in the beginnings of trying to build at Sugar Grove in West Virginia a 600-foot steerable antenna.
Butrica:You were aware that the Sugar Grove antenna was going up?
Gordon:We were not only aware of it, but lots of people were telling us, ďWhy do we need a 1,000-foot dish? We already have a 600-foot steerable.Ē Those people who werenít telling us that were telling us, well, the idea of incoherent scatter isnít going to work, anyway. There were lots of people telling us why we were wrong and not many people willing to say maybe it was right. But that really didnít stop us. So we decided weíd look for some money. Itís hard to put the story in the right sequence. We knew we wanted a big antenna. We went to the civil engineers again and said, ďLook, we canít build that up in the air and move it around, so we ought to build it in a hole in the ground. Where can you find us a hole in the ground that would accommodate a 1,000-foot antenna?Ē With a certain depth, a couple hundred feet maybe. We went to a very clever civil engineer. His name was Prof. Don Belcher. Don said, ďYou come back tomorrow, and Iíll give you a list of sites where we can find holes like this.Ē As soon as I left the room, the rascal walked over to his bookshelf and took down a book that listed areas of karst topography. Now, he knew what karst was, and I didnít. An electrical engineer wasnít supposed to know those things, but a civil engineer was. Karst topography is limestone sinkholes. Limestone sinkholes are formed when coral builds under the sea, and itís raised up above the water level, and youíre left with the coral and the limestone. Rain falls on the limestone, and it partially dissolves. The water more or less percolates through it, carrying a little of the limestone with it. Pretty soon, it forms streams underground by eating away at the limestone. Eventually, the stream makes caves, and the caves get so big that youíve got portions of the limestone collapsing. These are the sinkholes, wherever thereís a collapse. You can find the sinkholes in all shapes and sizes. What Belcher said to me was: ďThere are sinkholes in Mexico, in the Bahamas, in Cuba, in Puerto Rico, and Hawaii. Iíll look in the other hemisphere, the other half of the world, if you want me to.Ē I said, ďNo, youíve got a good start.Ē We were looking in the tropics because by then we realized that if we had a great big antenna and a powerful radar (that was the real point, a powerful radar), we could not only easily bounce echoes off the Moon (people had been doing that for quite a while, since World War II), but off planets, too. Planets hadnít been done. Why not set the radar at a place where planets came substantially overhead. Thatís essentially the tropics. From the ionosphere point of view, since we didnít know much about the ionosphere anyplace in the world to speak of, it didnít make much difference where we started. So, the first step was to put it in the tropic; thatís why we got the list of areas with karst topography. Out of that, eventually, Don Belcher and I walked over some of Puerto Rico looking at sinkholes, getting lost, and finally finding one that seemed to do the job.
Butrica:Yes. The other sites were eliminated for various reasons. Cuba was for political reasons?
Gordon:Yes. The one place we eliminated not for any good reason, since it was in 1958, was Cuba. Cuba was close to the U.S. Itís easy to get there. Batista was in charge, not Castro. So we could have easily said: Letís go to Cuba, and that really would have fixed it.Ē You could have gone to any of the other places, too. Why did we go to Puerto Rico? Booker had a graduate student from the engineering part of the University of Puerto Rico in MayagŁez, instead of out in Rio Piedras outside of San Juan. This fellowís name was Braulio Dueno. Dueno was a fortyish-year-old professor at Mayaguez who got into some program where they would support him in the States somewhere for two years to get a Ph.D. He came to Cornell, and he must have come in like 1956. He worked with Booker for a while, then Booker went on a sabbatical, and I helped him finish his Ph.D. So we knew him quite well. We knew something about Puerto Rico from Dueno. So there was some reason to pick Puerto Rico over some of the other places. Puerto Rico was a freely associated state. It was in some sense part of the U.S.; it was a commonwealth and still is. Probably always will be. It seemed to be an attractive place. It was a great deal more prosperous than the neighboring islands, some of which are pitiful in terms of living conditions. So Puerto Rico was picked. Let me digress a minute to talk about financing.
Butrica:Okay. Thatís not a digression; thatís one of the key areas I wanted to talk to you about.
Gordon:As I say, the order may not be very good, but itís also interesting. When we first had the idea in the spring of 1958, thatís when the midnight oil burned, and all the excitement began to be generated. In the spring, like March, we decided that we wanted to put some people to work that summer to see whether they could sketch out supports for a big antenna and to send somebody to Puerto Rico to see how the radio noise level was in certain areas. So we needed a little money. Cornell wasnít going to support us doing this. So I called up Arnold Shostak. He was at ONR [Office of Naval Research], and I had a contract with Arnold to do some radio measurements of the Sun. There was some solar radio astronomy going on at Cornell. Arnold could tell we were excited and that we wanted to do this more than anything else. He said to me in March: ďI canít get you a contract that quickly for the summer, but donít you have some money left in your solar radio astronomy contract?Ē I said, ďSure.Ē He said, ďUse it, and do what you want to do.Ē That was marvelous: a program officer who would be that flexible. I donít know how many rules he broke when he did it, and I didnít ask. Iíve always appreciated his help. That, however, got us little bits and pieces of money. It gave us something to do that summer: the civil engineers could start figuring out how the antenna might be built, and the radio engineers could figure out whether the background noises in some of these areas in Puerto Rico we were thinking about were tolerable. That was the beginning of the full-scale money. It was clear that we needed a lot more than small-scale money, if we were going to build a big antenna, big radar, because the transmitter wasnít cheap either. The question was: Where are we going to get money measured more like a million dollars than a couple tens of thousands? The NSF [National Science Foundation] budget in 1958 might have been a few tens of millions, but it wasnít much more than that. While the NSF maybe was the agency of choice, in the sense of funding curiosity-driven, basic research, they were not a viable or a possible supporter, because it was too big a chunk of their budget. At that time, DOD [Department of Defense] was starting up a new agency called ARPA, Advanced Research Projects Agency. It was just getting going. ARPAís mission was to support, as I see it, far-out or wild ideas that might have some military use. I canít tell you exactly how we discovered it, but we realized it existed, and a few phone calls got me to talk to someone in Washington who I then had to convince over and over that incoherent scatter was for real and that if he put some money in it, it would do some good. So from the summer of 1958 until the summer of 1960, it was mostly a commute from Ithaca to Washington to sell them again on why they should do this. We talked, among others, to a man named Ward Low. Ward Low made a very important contribution, aside from finally agreeing to support us after lots of doubts. He would say to me on a visit, ďIím not really sure that what youíre telling me is completely crazy or whether it makes sense. Why donít you go to Princeton and talk to Prof. Boehm and explain it to him. Then have him give me a call to tell me whether he thinks it makes sense.Ē So there was this kind of hurdle, and we did this several times. I donít know how many times anymore, but a lot of times. What Ward Low contributed in a very positive way was to say, ďIf you build a big antenna and just point it straight up in the air,Ē which is what we were talking about, like a parabola sitting in a hole in the ground, ďyouíre not going to be able to see anything but directly overhead. You canít move the beam more than a beam width or so before it completely de-focuses.Ē He was right. You have a beam width which maybe is a fraction of a degree, and you sweep out less than a degree of sky that you can actually make any measurements on. He said, ďI want you to go to the Air Force Cambridge Research Lab up in Boston, Hanscom Field, and talk to the people there, because they are involved in spherical antennas. They apparently have a ten-foot spherical dish with an appropriate feed.Ē
Gordon:Is your background engineering?
Gordon:History, okay. Well, let me just say a couple of words about a parabola and a sphere. A parabola has a focus, a point. From that focal point, if youíre sending a signal out, it all goes in one direction. If you move the sending point appreciably from the focus, youíll still send the signal out, but pretty soon the beam breaks up. So instead of having the nice beam that you had in the proper direction, you donít have much left once youíre off a beam width or two. Thatís a parabola. Now, a sphere doesnít have a focal point. If thereís some special point in the sphere, itís the center of the sphere. But thatís not a focus. The sphere has a focal line instead of a focal point. The focal line is on a radius, any and every radius. Sit on one radius. Start at the center and go halfway down to the sphere. Thatís where the focus begins. It starts there, and it continues toward the sphere. It depends on the geometry how far you have to go, but itís a matter of quite a distance. If youíre transmitting, you have to arrange to transmit the signal with the proper phase and in the proper direction towards the sphere along that focal line. As you move down the line, the phase and the direction angle change. If you do it all properly, you form a beam, a searchlight beam. Now, if you can make this happen on a particular radius, you could do it on any radius. Or you can move your feed from one radius to another. The beam is aimed in the direction of the radius. The radiation goes to the dish and reflects back, so it goes out into space in a direction that is determined by the radius. If you could move from one radius to another radius, thatís the equivalent of moving the beam from one angle in the sky to another angle. Thatís what Ward Low was trying to talk us into adding to our proposal: That the beam ought to be scannable, movable. So we went up to Air Force Cambridge Lab and told them what we were up to, and we asked them what they knew about spherical reflectors. They said: ďOh, well, we have a ten-foot spherical reflector operating, and it has a feed that works,Ē meaning that it had a feed distributed along the focal line. They could move it from one radius to another, as I have just described to you, and make the beam move about. This was fine. The only problem was that the scale of the sphere was ten feet. We were talking about a thousand feet. Would everything work if you scaled it up by a factor of 100? Well, nobody thought it wouldnít, and most people thought it would. The same geometric optics would work regardless of the scale, and you calculate the properties of the feed from geometric optics. Itís a simple bit of geometry that you have to do. Itís pretty straightforward. Thereís nothing different, whether the radius happens to be 500 feet or 5 feet, in terms of how the geometry works. There was a lot of interaction between Air Force Cambridge and Cornell and the ARPA. The net result of all this was that we changed our ideas from using a parabola to using a sphere. That meant we needed a feed which corrected for the so-called spherical aberration, which led us to use a line feed instead of a point feed. We learned a lot from the Cambridge people and from the people that they had contracted with to build their ten-foot dish. Ward Low had added a nice new dimension to the antenna by making it steerable. After some studies and compromises, we decided on a steer ability of 20į from the zenith. So, we had a cone of 40į available to us looking up into the sky where we could point a radar beam. If youíre going to use the antenna just as a receiver, it means you can receive from any place in that 40į cone, which means that when youíre sitting on the Earth and itís rotating, youíre going to sweep out 40į of declination in the sky, and youíre going to be able to see at some time everything in that 40?
Butrica:The addition of the steer ability in order to obtain sufficient sky coverage was when the suspended platform was added.
Gordon:Thatís right. Youíve got it exactly. If you have a focal point, what you have to do is support something at that point. That means you could have sort of a triangular support structure, or a three-legged structure, that supports the feed where you want it. If youíre going to move the feed from one radius to another, then you need to build a structure which will give you the chance to rotate in azimuth or to move in elevation angle. Do you have a picture of Arecibo?
Butrica:Oh, yes. Iíve been there.
Gordon:Well, then you know. The structure is elegant in geometric terms. The triangle which supports everything is held by cables from three towers. Thereís a circle hanging on the triangle, which supports the long arm underneath and permits the arm to be rotated in azimuth, i.e., rotating about the vertical axis. So youíre going to have an azimuth motion by rotating on the track. On the bottom of the arm, thereís another arc, and that arc has the same center as the dish. So, anything that moves along that arc is carried from one radius to another. The feed is a long line roughly 100 feet long. So the idea is to move that 100 feet feed from one radius to another. The combination of moving along the bottom of the arc and rotating the arm gives you the two angular motions that you want: elevation and azimuth. Thatís all clear?
Gordon:Okay. I want to make sure you stay with me. That got us to even more exciting prospects than we had originally. The first idea for Arecibo originated in, letís say, March 1958. In July 1960, there were five Cornell families in Arecibo, Puerto Rico, beginning to build the Arecibo Observatory. Two years and a couple of months. That was in 1958 with a new agency, ARPA, providing the lubrication. These days if you tried to do something major, it would take two years to get a committee formed to examine whether anybody ought to even think about it? The idea was at the right time. Now, youíre a historian. If the time isnít right, a good idea isnít very usable. The idea can be before its time or after its time, and neither of those is much good. The idea has to occur at the right time. In some sense, it has to occur at the right place. I think Cornell and Engineering at Cornell was a very lively group in those days. Iím sure it still is, but itís in different ways. It was the interaction among the EEís [Electrical Engineers] and the CEís [Civil Engineers] and the MEís [Mechanical Engineers] and the administration. The dean, S. C. Hollister, was behind it 100 percent right from the beginning, and so was the rest of the administration. Dale Corson, who I think in the beginning may still have been in the Physics Department, became Dean of Engineering then provost. He became President of the university in time. He personally took a big interest in what we were doing. In that sense, it was the right place. And ARPA was born at the right time, so it could respond. Now, ARPA responded to much bigger requests than we were making to them for various things. It gave us a source when NSF probably would have been the rational choice, but didnít have the money and was unable to act.
Butrica:Let me ask about two other possible sources of funding: One would be the Bureau of Standards, which built a number of ionospheric antennas. And then in 1958, why not ask NASA?
Gordon:Yes. NASA was too busy trying to get the little grapefruits flying, trying to catch up with the Russians and Sputnik. In effect, NASA decided that they were going to do science in space, not on the ground. I think that more or less shut us out from NASA. Now, the Bureau of Standards, thatís a good question. In the 1950s at Cornell, we had a graduate student named Ken Bowles. Have you come across him?
Butrica:Iíve come across his name.
Gordon:Okay. Ken Bowles was a student of Henry Booker. He did a thesis on radio scattering from aurora, and it worked out very well. Bowles graduated from Cornell in 1957, give or take a year or so. He was a student when we were revving up for Arecibo. He went to the Bureau of Standards in Boulder, and he knew what we were doing. He knew about the big antennas that the Bureau of Standards had, one of them being in Havana, Illinois. It was a big antenna they were using for communication, I think ionospheric forward scatter between Illinois and Sterling, Virginia. Bowles thought hard and long about what we were doing and how he could influence it. He was a major player in the game, even though he was at the Bureau. He kept up with what we were doing. The antenna at Havana was pointed more or less toward the horizon in the east. He decided that he could rearrange the feed system to make it point upwards. It had a big transmitter, and he could arrange for pulses somehow. It was sort of a jury-rig radar of some substantial power and some substantial antenna size. The frequency was probably about 50 MHz, which was okay. Itís not the ideal frequency, but it was fine. Bowles talked his bosses at the Bureau into letting him do this. One of the most exciting papers I ever gave at a scientific meeting was given at Penn State in the fall of 1958. It was one of these URSI meetings. URSI had a Washington meeting in the spring and a fall meeting that moved around the country. In the fall of 1958, it was at Penn State. I had a paper on the program, with something about the incoherent scatter theory. But it really had to do with saying we were going to try to build radar that would measure this weak signal from the ionosphere. By the fall of 1958, Bowles more or less had his antenna rewired and was ready to crank his radar up. Bowles and I were in close touch. I told him when I was giving the paper. It was an afternoon session, and Iíll never forget it. It was the first paper in the afternoon on some particular day, and I said to Ken, ďYouíre close to making measurements, see if you canít do something before I give the paper, so that you can make some input.Ē He said he would. I said, ďIíll call you at lunchtime on the day of the paper, and you tell me what you have.Ē Ken had succeeded in measuring a return from the ionosphere, which was what we were thinking of as being incoherent scatter. It had succeeded. So I was able to give my paper and say, ďI want to tell you about building big radar that can measure the properties of the upper atmosphere. And at the end of the paper, I want to tell you about a telephone call I just had.Ē That really got their attention. Some of them were aware of what was going on. The paper ended with my quoting from Bowlesís telephone conversation that the scattering really worked. Here was the first evidence that there was something to it. Now, Bowles did more than that. Because he not only said it worked, but he said the bandwidths that youíre predicting for the signal are much narrower than you think they should be. I was predicting bandwidths that were based on thermal motion of the electrons. The bandwidths that Bowles measured were based on thermal motion of the ions, and they are much slower. Therefore, the frequency is less spread, and, therefore, the signal is, in a sense, stronger, i.e., itís in a narrower band. So, if you measure signal-to-noise, you donít have to open up the receiver to the wide band of the electron spread but only to the ion spread. Thatís a big advantage. Thatís an advantage of 1800 or so. Ken really produced two things: One, there is some scatter. It seems to be controlled by the ions, even though the ions donít scatter very much. The electrons, in effect, were producing what he was seeing. But the motion of the electrons was essentially being controlled by this cloud of ions that were surrounding them. He was seeing plasma or an ionized medium, which, like jello, was rocking at the rate associated with the ions instead of at the rate associated with the electrons. It was an important addition that Ken made, and it corrected, essentially, an error that I had made in my projections. Fortunately, in the right direction. It gave us more flexibility than we thought we might have. That was the beginning of the Bureau of Standards activity. When Ken succeeded in Havana, Illinois, with this, his bosses decided, ďWe ought to push this.Ē Ken proposed that he build in Jicamarca, Peru, which is on the magnetic equator, incoherent scatter radar looking vertically upward. And they supported it.
Butrica:So basically the National Bureau of Standards was building your antenna, but in Peru.
Gordon:They were building powerful radar, but not a 1,000-foot dish. They not only built the powerful radar, they operated it for a while. Ken was the first director. Don Farley was a director. Tor Hagfors was a director. It still exists, with Ron Woodman as director. The Bureau pulled out of it at some point, because they didnít want to continue doing basic studies in Peru. There are various problems with working there. But it was a very successful venture. What Ken built there was a field of dipoles phased together in the right way that he could look up, and he could look off at a few angles by changing the phasing. He got some powerful transmitters and good receivers, and it just all worked very well. Iíve forgotten when he got on the air. We dedicated Arecibo in 1963, and Iíve forgotten whether Ken was on by then. If he wasnít on by then; it was shortly thereafter. Almost the same time. So there was big radar in Arecibo, and a big one in Jicamarca, outside Lima.
Butrica:So the Arecibo antenna was built with ARPA money then dedicated.
Gordon:The important thing to say is that it was built to study the ionosphere, the conditions in the upper atmosphere. ARPA knew missiles were flying in that medium, and they had thought man eventually would be flying in that medium and they wanted to know something about it. That was the thinking, the rationale.
Butrica:It was built for Project Defender. How did the Arecibo Observatory fit into Project Defender?
Gordon:I donít know what Project Defender is. So I donít know whether we fit or we didnít. The work at Arecibo was never classified. I had some clearances, and occasionally I would be asked to come up to the Pentagon and talk to people about what we were doing and what it meant and so on. They had some military objectives they were trying to take care of. I wasnít privy, really, to the details. There was no point. They could grill me, and they didnít have to tell me what they were up to. I agreed to do that. I didnít care where the money came from. In the typical ARPA sense, they didnít monitor the contract. They handed it off to Cambridge, which is normally what they did, I guess. They asked someone else to monitor the project and take care of the accounting and whatnot, and they asked Cambridge to do it.
Butrica:Was Ward Low still your contact person at ARPA during the construction of the Arecibo Observatory?
Gordon:Yes. Ward Low was involved for quite a while. Then, sometime after it was built, he left ARPA, and we had a series of contacts. If they called, I always listened and tried to respond. But formally, Cornell was a contractor of Air Force Cambridge.
Butrica:The antenna and radar were designed to do ionospheric work primarily and lunar and planetary astronomy secondarily. How did radio astronomy come in? Who pushed that?
Gordon:The radar astronomy really came in first because we were building a radar, a powerful radar. It was clear that the radar would be able to bounce echoes off the Moon with much greater ease than anybody had done before and therefore with much better resolution. And if you were going to bounce echoes off the Moon, you could bounce echoes off the nearby planets, Venus, Mars. What was the limit? Well, itís surprising. You couldnít get out to Jupiter, because the Earth rotated your antenna too rapidly. It took more time for the radar signal to go to Jupiter and back than you were able to look on a given day at Jupiter. You remember I said we had a 40° scan angle?
Gordon:If you aimed at Jupiter when it first came in sight, the antenna would be way over on one edge. Youíd send the signals out, and by the time the signals got back, you could no longer receive from Jupiter. The 40° scan angle translates into two hours and something of travel time. Well, if you know the distance to Jupiter and the speed of light, you can decide how many minutes you need.
Gordon:Jupiter was close, but I donít think there was ever anything successfully received. The story is quite different on Venus or Mars, Venus in particular. Pettengill and others did a lot of very good measurements on the surface of Venus, through the clouds of Venus from the Earth, much like NASA satellites have since done by circling Venus with the side-looking radars.
Butrica:Once the radar was built, who was in charge of Arecibo?
Gordon:Originally, the observatory was called the Arecibo Ionospheric Observatory. That was its first name. I was in charge at Arecibo, and I was in Arecibo from 1960 to 1965. Gold had come to Cornell, Iíve forgotten exactly when, around 1960. He had set up the CRSR [Center for Radio physics and Space Research]. At that stage, Corson was Dean of Engineering. So in a sense, Corson had some overall responsibility. It was Corson who finally, in 1965, asked me to come back to Cornell. In a sense he was saying, ďCome back and teach. Let the students share some of this excitement.Ē Have you talked to Corson?
Butrica:No. Youíre the first person whoís mentioned him.
Gordon:You should. Corson can probably give you the least biased version of what happened of anybody. If you ask me, I was mad at the time, and whatever I tell you has some personal bias built in.
Butrica:Iím looking for your personal bias.
Gordon:I thought I was removed from a job that I deserved to have. Corson came to Arecibo and spent some time. We talked about it, and I was very unhappy with it all. I admire Corson as a person.
Butrica:It sounds like this was more than a suggestion, maybe an ultimatum.
Gordon:He and Nellie, his wife, came and visited us in Arecibo and spent at least a long weekend or days. I know he came there with the idea of seeing that I came back to Cornell. Iím sure that Gold had talked him into it, and I think that Corson probably views that as one of the worst decisions he ever made at Cornell. That you need to determine from him, however. Gold is a very bright guy. Heís cunning. He has lots of good ideas. Wait a minute. He has lots of ideas. Some of the ideas are good. But the people around him are snowed trying to sort out the good ideas from the chaff. I donít know how well you know Gold...
Butrica:I interviewed him.
Gordon:I think thatís a fair description. I donít think that has a lot of bias in it.
Butrica:What was the nature of the conflict between you and Gold?
Gordon:It was a matter of who was in control, I guess. Iíll give you one example. Sometime in 1963 or 1964, at Arecibo, we had worked out with the people at Cambridge the idea of building a Gregorian reflector, which is what theyíre now spending millions putting in. The latest upgrading is to add a Gregorian reflector which gives the antenna flexibility in the domain of frequency. One thing I didnít tell you about the line feeds is that theyíre rather narrow band. For each frequency band you want to use, you essentially have to build a line feed tuned to that frequency. If youíve seen the antenna, youíve probably seen it with lots of feeds hanging down on the underside of the arm, each one maybe being one percent wide in frequency and operating on the band for which it was built. But the Gregorian essentially corrects for spherical aberration by bringing all of the rays to a point by using multiple reflectors. You have to use at least one in addition to the main dish. Thatís what we were proposing. The feed they now have is an offset Gregorian, so it has a couple of auxiliary reflectors before you get to the focal point. If you bring it to a focal point, then the width of the band that youíre able to receive on is determined by the sensor that you put at that point. You can put a very broad band sensor, and your chances are much better of doing it at a point than along the line. Instead of having 1 percent bandwidth, you can have 10 percent, and thatís a big gain. Itís a little bit like going from black-and-white pictures to colored pictures. Thatís the right kind of analogy. In terms of radio astronomy, it gives them a chance to simultaneously look across a big band. Now, you were asking about the conflict. One of the conflicts was with the Cambridge contract monitor. We worked out with Phil Blacksmith, who was our contractor monitor at Cambridge, the idea of building a Gregorian reflector based on some material that was in the literature from someone in Australia, whose name escapes me. It must have been in the literature before 1965 certainly.
Butrica:Was this Ron Bracewell?
Gordon:No, it wasnít Bracewell. I know Bracewell very well. I might be able to find it somewhere, but I canít produce it for you this minute. At any rate, it doesnít matter. What he had done was to propose a Gregorian reflector for a radio antenna. What the Gregorian looks like is an eggshell. If you take a sharp knife and cut one end of the egg off, the shell thatís left will more or less have the right surface, and at the right place in the spherical reflector, it will give you a point feed. We worked out with Phil Blacksmith a proposal to build a Gregorian reflector based on the material in the literature. We sent the proposal up to Cornell, because it would have been an amendment to a contract. We waited a while. It finally turned out, to make it short; the proposal was refused at Cornell, which I thought was crazy. The source of the refusal was Gold. Now that didnít sit very well.
Butrica:Why did he refuse?
Gordon:I have no idea. Ask him. He probably wonít even remember that he turned it down.
Butrica:No, he doesnít. That I can tell you. He doesnít remember that.
Gordon:Well, I remember it very well, because it was something that we thought in Arecibo at least would have been a great addition. Adding a Gregorian to a spherical reflector is a great addition.
Butrica:Tor Hagfors tells a different story about the Gregorian system. He claims that Frank Drake was behind it and failed to get the money necessary to put together a proper team to design the Gregorian.
Gordon:I think youíre talking about a later time than the one Iím talking about.
Gordon:The one Iím talking about has to be while I was still in Arecibo, so it was before 1965. It had to be like 1964. The antenna was up and running and the feed wasnít very efficient. So there were efforts to improve the efficiency. In the process, we had an agreement with Blacksmith, and Blacksmith was prepared to come up with the money to build the thing.
Butrica:I have a study here, ďTheoretical Study of Gregorian Radio Telescopes with Applications to the Arecibo Ionospheric Observatory.Ē It was a report from 1964 by a fellow named Pierluissi.
Gordon:Yes. He was a Cornell graduate student. That probably refers to what Iím talking about. I donít remember the details exactly, but the date is right. I think youíre right, though, that later on Frank did try to do something with Gregorian, and I donít know why it didnít work. Let me quickly admit: What we were talking about was a simple eggshell. It was not the offset sophisticated Gregorian thatís being added now. If you offset it, you donít get in your own way. Our Gregorian would have had, probably, a big hole in the middle of the pattern which wouldnít have been very good. But it would have had the frequency flexibility, which was what we were trying to achieve. The one theyíre doing now, by adding two extra reflectors and getting them essentially out of the line of sight of the center of the antenna are a big improvement. I think itís important to say that. But what we were talking about was the simplest version of a Gregorian, and it would have produced something of a shadow, which the line feed triangle does. But the effects arenít big enough that you shouldnít use the antenna. You use it and accept what you have.
Butrica:The real crux of the friction between you and Gold, then, was the improvement of the antenna?
Gordon:Well, that was the beginning. There were lots of, I think, little things that were annoying. He was flexing his muscles at Cornell. He had some sort of a North-South alliance with a man in Australia, Harry Messel. Now, before I forget it, though, let me go back to radar astronomy for a minute. I think the job that people like Pettengill, and Dyce, and Don Campbell were able to do on mapping the Moon, using a range-Doppler grid was really quite a remarkable achievement. They were able to do it on the Moon so effectively because of the big gain that they had. You could get the resolution down into rather smallish squares, small enough that their Earth-based maps lasted for a long time. They were doing similar things on Venus. It was farther away, and the resolution wasnít nearly as good. But the ideas were good, and the fact that they had some information about the surface of Venus, that wasnít available to the optical astronomers because of the clouds, was significant. That was all radar astronomy. Now, I agree that currently the use of the facility for things like pulsars and quasars and other exotic radio astronomical objects has contributed enormously. The first Arecibo Nobel Prize went to a professor and his students for looking at pulsars. The pulsars were not discovered at Arecibo. Tony Hewish and one of his students at Cambridge discovered them. Once they were discovered, Arecibo was ideal, because of the big collecting area, for measuring these weak signals and the structure of the signals which finally had all the information. Honestly, the Observatory was built to study the upper atmosphere. Radar astronomy and radio astronomy were essentially fringe benefits in terms of the building. In terms of today, while atmospheric scientists still use the facility (I donít know what the fraction of time is), ití s much smaller than it was in the past, and theyíve added some different kinds of auxiliary facilities around Arecibo. But the radio astronomers, I think, are the most productive of the groups that are there.
Butrica:The National Science Foundation came into the picture in the late sixties. You know where Iím going with this. Tom Jones?
Gordon:Well, my view of NSF was as long as we could stay with ARPA funding, we would conserve the NSF funding. ARPA had a lot of money; NSF didnít. ARPA got us going. They did want to get rid of us, in the sense that they didnít want to pay forever for the operation. So we stalled in every way we could to let ARPA off the hook. They werenít all that anxious to get off the hook in the beginning. We added an ionospheric heater back in the late sixties. The first ionospheric heater was, in fact, a feed that was put into the main dish. Every time we had to put that feed up, people sweated because if anything fell, it would affect the dish. Then eventually, the heater was built separately at Islote, ten miles north of the Observatory. Islote is almost on the shore just east of the town of Arecibo. It has a big HF facility for heating the ionosphere.
Butrica:Continuing with Tom Jones and the NSF: The story there was the claim that ionospheric research was being short-changed in favor of radio astronomy.
Gordon:Yes. Once we got into NSF, and NSF assigned it to the Astronomy Section to look after, then you might expect some trouble about how things were divided up. There were always some problems inside. Iím not very familiar with them, because most of these are after I was associated directly with it.
Butrica:Tom Jones was heavily involved in the accusation that ionospheric work was not getting its fair share. At the time, after you left Arecibo, I think you were involved in the NSF.
Gordon:Yes. What year was that weíre now talking about?
Butrica:The late sixties.
Gordon:Okay. In the late sixties, I was here [Rice University].
Gordon:Yes. Well, I was here. I came to Rice in 1966. I went back to Cornell for one year from Arecibo. I had a nice offer from Rice, and I decided it would be more fun than trying to work around Gold at Cornell.
Butrica:Jones wrote about the political in-fighting at Cornell and stated that it resulted in the departures of Gordon and Booker.
Gordon:Yes, thatís true. Booker had a great opportunity at La Jolla, University of California at San Diego. He was invited to come out and form a department. And he did and built a great department there.
Butrica:Your reason for going to Rice?
Gordon:I had a good offer from Rice University to come and be one of two deans on the campus, Dean of Engineering and Science, and work with Ken Pitzer, who was the President at Rice. As far as Arecibo is concerned, I then became one of the external users. Until I retired at Rice, Iíve invariably had one or two students doing some thesis work at Arecibo.
Butrica:Returning to the problems with ARPA. From about 1965, ARPA gradually reduced the operating budget of Arecibo. Do you remember what that was about?
Gordon:Well, I expect what was happening was that ARPA may not have been very happy with Goldís running it. But thatís a guess. I was not involved after 1966 in the operation. Iíve been on the Arecibo Advisory Board a couple of times, but thatís spread over almost 30 years now. I see it from the outside, and I still have lots of friends at Cornell.
Butrica:My impression is that ARPA felt that the spherical dish and line feed were so inefficient that they were not going to dump any more money into the facility.
Gordon:When was this?
Butrica:During the early sixties, while you were there.
Gordon:In the early sixties?
Gordon:Well, Iíve forgotten when they built the heater, but they paid for the heater, and that was not the early sixties. That was the late sixties. They paid for two major different versions of the heater, one co-located and one separate. The co-located one, I was at Arecibo at the time. The separate location was built after I left Arecibo, but it was built with ARPA money. Now, what their internal view was as to whether it was inefficient, I have no idea. I must say I always felt my relationships with the ARPA people were very good and open. Iím surprised that there is something in the record like that. Well, Iím surprised that I didnít know we were part of Project Defender.
Butrica:Yes. Now you know.
Gordon:Itís a different mind-set.
Butrica:Are you involved in the current upgrade of the Arecibo Observatory?
Gordon:No. I know whatís going on, because I recently was on their advisory board, and we were briefed on all of the details. We gave them some advice. They have some management problems. I know Hal Kraft very well. I know the people involved very well. I know what theyíre trying to do, and theyíre spoiling our elegant geometry by adding all kinds of extra cables.
Butrica:Theyíre finally getting the Gregorian reflector.
Gordon:Apparently the Gregorian is a heavier weight than the structure was intended to support. Originally, we knew we would be adding things from time to time. When the thing was first built, we had tons of lead weights up there with the idea that, if we added a feed or added a heavy something or other, we would take off some of the ballast. At some point, we ran out of ballast.
Gordon:Theyíve gotten it orders of magnitude better over time for all kinds of reasons. All the elements have been improved in ways that arenít 10 percent but more like an order of magnitude. The transmitters are better, the receivers are better, the antenna is smoother, it goes to much higher frequencies than it did initially.
Gordon:Yes. Iím sure at 6 cm it will work with a very high efficiency, and it may work at 1 cm. Iím not sure. Theyíll know better when theyíre finished. Wavelength gets you sensitivity in a hurry. For the radio astronomers, itís the collecting area. For the radar, if itís a point target, you get that to the fourth power. You get wavelength improvements to the fourth power. The other thing to look at is the comparison with Bonn. When they built the half-scale Gregorian to test out the ideas, it was better than Bonn as a microwave radio telescope. So imagine what itíll be with the full-scale Gregorian!
Gordon:To think that you can take something that seems to be stretching the state of the art when you do it, and 30 years later it has improved as though you had nothing to begin with, I donít know how many orders of magnitude, but there are many, depending on what youíre doing. That strikes me as maybe, from an engineering point of view, the most remarkable characteristic.
Butrica:Yes. Whatís also interesting is how flexible this instrument is. It is now undergoing a second upgrade. What in the world is the lifetime of these antennas?
Gordon:When we first talked to engineers here about building this that was one of the first questions they wanted to know: Whatís the lifetime? We batted that one around and around and finally said, ten years. They were thinking of bridge cables. Once they even removed one cable just to look at it. Letís see, starting in 1963, in 1993 itís 30 years old. Itís like a cat. Itís already had three of its nine lives. In many ways Cornell has taken good care of it. Bob Matyas, who used to be Vice President for Buildings and Grounds (Iíve forgotten what they call it, but it was essentially that), kept a close eye on Arecibo, how it was functioning, and whether it was going to give him a problem like they had at West Virginia, where one of the telescopes just collapsed because some welds got tired. That created some activity. Robert Matyas lives in Ithaca. Heís retired. Very, very sharp. As I say, he was Vice President for Buildings and Grounds for all of Cornell, but he took a special interest in Arecibo. You want to get some different points of view on how it was viewed. Have you run into the names of Merle Lalonde or George Peters?
Butrica:Yes. L. Merle Lalonde did a lot of the engineering design work on the system.
Gordon:Thatís right. He was the one who built a lot of the feeds after 1965. George Peters was one of the early technicians.
Butrica:I ran into that name, too.
Gordon:Thomas Talpey was an engineer we borrowed from Bell Labs. He helped us with receivers in the beginning.
Butrica:Do you remember the people from Cornell Civil Engineering, like Bill McGuire?
Gordon:Oh, yes. Heís worth talking to. Heís retired, but still in Ithaca. Don Belcher, I think, is still around. Belcher is an interesting character. Heís the one that helped us find the site. But maybe the most famous site-finding he did was to help the government of Brazil find Brasilia. That was one of his contracting-consulting activities.
Butrica:How about Winter?
Gordon:George Winter was a great friend. He was the more senior of the engineers in the structural business, and he was a great help and a great advisor. There were a bunch of EEís [electrical engineers] that we havenít named, too. Booker was my mentor and partner in a lot of this. There was Ben Nichols and Don Farley, names that are still around, and that were involved in one way or another in the process. How about Ken Bowles?
Butrica:No, I havenít tracked him down yet.
Gordon:Ken Bowles is in California. For a while, he was with Henry Booker at the University of California at San Diego in La Jolla. Then he went into a computing business, and heís gone through a couple of successes in the computing field. He still lives somewhere north of La Jolla. Where did you get the reference to the Bureau of Standards, which reminded me of some of this?
Butrica:In doing this history of planetary radar astronomy, Iíve come upon the NBS [National Bureau of Standards] many, many times.
Oh, okay. Well, Bowles was the sparkplug in Jicamarca.
. J. Pierluissi, A Theoretical Study of Gregorian Radio Telescopes with Applications to the Arecibo Ionospheric Observatory, Research Report RS 57 (Ithaca: CRSR, 1 April 1964).