Interview with Francis Graham Smith

Francis Graham Smith, 1923- . Interviewed 27 August 1976 in Grenoble, length of interview: 75 minutes.

Papers of Woodruff T. Sullivan III

NRAO/AUI/NSF

Oral History

Sullivan, Woodruff T., III

Graham Smith, Francis

Audio cassette tape

75 minutes (2 tapes)

1976-08-27

Radio source positions and sizes at Cambridge ~1947-53; optical identifications; development of precision interferometric techniques; ionospheric scintillations; solar corona occulations of Tau A; philosophy of Cambridge group; 1950s meetings.

See the Listing of Conferences and Lectures for Graham Smith's 1982 talk on history.

The interview listed below was originally transcribed as part of Sullivan's research for his book, Cosmic Noise: A History of Early Radio Astronomy (Cambridge University Press, 2009). The original transcription was retyped to digitize in 2016, then reviewed, edited/corrected, and posted to the Web in 2016 by Ellen N. Bouton. There is a considerable amount of background noise in the audio of this interview, traffic noise in the first portion and crowd noise in the remainder, making the interview difficult to hear/understand at times. Places where we are uncertain about what was said are indicated with parentheses and question mark (?).

We are grateful for the 2011 Herbert C. Pollock Award from Dudley Observatory which funded digitization of the original cassette tapes, and for a 2012 grant from American Institute of Physics, Center for the History of Physics, which funded the work of posting these interviews to the Web. 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.

Working Files Series

Individuals Unit

8A

Originally transcribed as typescript only by Pamela Jernegan (1979), retyped to digitize by Candice Waller (2016)

Begin tape 68A

Sullivan

This is talking with Graham Smith on 27 August ’76 at Grenoble. So, can you tell me what your background was before you joined the radio astronomy group at Cambridge?

Graham Smith

I was very much of a physicist, a radio physicist at that. Of course, like some of the other radio astronomers, I came from war-time radar research. Telecommunications Research Establishment in Malrern. And I had no particular interest in astronomy – I can’t say that I was an astronomer as a small boy or anything like that. I was definitely a physicist, a scientist without and particular leanings, except that I was interested in radio and I obviously felt myself to be more of a practical type than a theorist. So that it was mainly to get back to Cambridge and to finish my degree, but also with an interest in radio, that took me back to Cambridge after the War. And that, of course, took me on into radio astronomy as a research project.

Sullivan

So it could just as well have been radio work with anything.

Graham Smith

It could well have been. And of course, when I joined the group, radio astronomy didn’t exist by name. And it was only solar work with, of course, some ionospheric work going on – ionospheric work was predominant in the radio group at the time.

Sullivan

Right. How was it decided that you would work with Ryle?

Graham Smith

Well, I knew him, of course, during the war, and I think I picked up and soon shared his enthusiasm for the radio waves from the sun, which was the thing that was going at the time.

Sullivan

Right, but what was the first thing you work on? I don’t think it was solar, was it?

Graham Smith

I worked with Martin Ryle on solar work even before I became a research student, because it so happened that we lived in the same house. His house in Herschel Road, he now lives in, was at that time divided into three flats. And one of the flats was shared between my wife and myself and Derek Vonberg and his wife. So we were all closely associated. The third flat was occupied by a zoologist, who had also been a friend during the War. And when Martin Ryle and Vonberg were working on interferometric measurements of the sun, which is just outside the back gate, and so we used to help them move around the antennae, particularly, at the time I remember July 1946 sun spot thing, when it was first shown that radiation actually came from the sun spot rather than the whole disk.

Sullivan

Right.

Graham Smith

We were helping move the antenna around (?) larger and larger spacings to make an interferometer measurement of the diameter. I think it was really this that caught my imagination.

Sullivan

So you never really were a co-author, but you were quite involved with it.

Graham Smith

Well, really only as a pair of hands, but this is what, I think, sparked my interest.

Sullivan

And also, just for the record, I have learned somewhere that there’s a relationship, by marriage, between you and Ryle.

Graham Smith

That’s correct, yes. I think it’s fair to say that my wife introduced her sister to Martin Ryle.

Sullivan

I see. Now, so you helped with the solar work – what was your main – what were you mainly working on these first couple of years? Was that it?

Graham Smith

I think it was about the time I started when Hey had published the point source in Cygnus, and Bolton, of course, was also working on point sources and was publishing from the southern hemisphere. I think it was Hey’s work on Cygnus that really started us, and I then worked with Martin Ryle, modifying apparatus, which was originally for interferometric work on the sun, so as to work on “radio stars” as they were then called, and in fact, at 81 megahertz, and interferometer which consisted of 4 yagis at each end of the baseline, it was for this that we first had to (?) the Cygnus source, and of course, at the same time to our astonishment, Cassiopeia A came up on the same recording a few hours later, and from then on, we were into the “radio star” business.

Sullivan

Right. Is it true that – I asked Martin Ryle this – in your first paper, Ryle & Smith in 1948, you talk about another source of variable intensity which was perhaps coming through the (?)

Graham Smith

That’s right, that’s right, yes. That stayed in the catalogues for some years.

Sullivan

What about these variations – what did you think they were when you first heard about them?

Graham Smith

The variations in the, you mean in the supposed source, the one that wasn’t real?

Sullivan

No, no, no.

Graham Smith

The ones in Cygnus. I think, perhaps, that’s hard to say precisely, but I think we were perfectly prepared to think it was ionospheric effects. And we, of course, spent a lot of time trying to distinguish between ionospheric effects and intrinsic effects, by using spaced receivers and that sort of thing. That developed quite rapidly.

Sullivan

Well, in fact, I’ve just come across two papers, Little and Lovell and yours published together. Could you tell me about that? That’s one of the few Cambridge/Jodrell collaborations over the years.

Graham Smith

That’s right. In fact, the observations were made, more or less, independently. But, it was obvious that we were after the same thing, and it was obvious that the two papers should be published together. You see, in fact, those were separate papers published at the same time, right?

Sullivan

That’s right. But didn’t you go up to Jodrell?

Graham Smith

Oh, yes, we went up there and talked and made sure that the paper came out together, because we were obviously on to the same thing. Furthermore, we actually did compare some of the recordings made simultaneously at the two places so as to see if there was anything which was obviously intrinsic to the source. And it was, I think, at that time that we showed at least those fluctuations came from the ionosphere.

Sullivan

So it sounds like, from what you’re saying, it is difficult for me to determine from the published records that you all along suspected the ionosphere. It was just a matter of pinning it down. You really felt it was never intrinsic.

Graham Smith

I think we were open-minded, but at the same time the radio group had a very strong ionospheric background because Ratcliffe was responsible for so much of it. So it was natural that we should turn to the ionosphere as the explanation of it. And in the same way, when position finding started out, then it was natural to look for ionospheric refraction, you see, and in fact, I did quite a bit on that. As you know, it led me to a measurement of the total electron content of the ionosphere, which was, I think, again, an indication of our ionospheric interest under Ratcliffe.

Sullivan

Right. Did you know about the Australian work which was going on, more or less, simultaneously with yours on Cygnus A fluctuations?

Graham Smith

Not really, no. I think that at that time, not being astronomers, we weren’t really very conscious of what was going on anywhere else in any kind of astronomy. And we weren’t really very conscious of what was happening in Australia, even in radio astronomy. I think it’s fair to say that the radio astronomers in Australia probably sounded just the same way. We worked really quite independently, and then found that we were producing answers of the same sorts of problems – which produced some interesting situations. [laughter]

Sullivan

At this stage, or at a later stage?

Graham Smith

Well, of course, we very quickly were measuring the positions (?) and (?) the second question and it really very quickly became a competition as to who got an accurate enough answer to be meaningful.

Sullivan

But there were no big conflicts on these early positions that I remember.

Graham Smith

No, there were no great conflicts, but at times we had a more accurate position than they had and so forth, and that was the element of competition.

Sullivan

With the exception of Bolton’s very first positions, which were degrees off.

Graham Smith

Bolton’s were somewhat off at one time.

Sullivan

That got corrected rather quickly.

Graham Smith

That’s right, yes.

Sullivan

What about the identifications, while we’re talking about the Australians versus the British? What about the identifications that came out, tentative ones, of Virgo, Taurus, and the Crab?

Graham Smith

And the Crab.

Sullivan

Right.

Graham Smith

And the Crab as well.

Sullivan

How did you view those?

Graham Smith

Well, again, I think it’s fair to say that we were pretty blinkered and pretty busy on our work and we didn’t have an awful lot of notice of what we happening elsewhere. And so it was some time before it dawned on us that their positions were leading to real identifications. After all, they did get at that time the first tentative identification of the supernova in the Crab and they did get Centaurus A, the first extra-galactic object, and we maybe looked at these and said, “Maybe these positions aren’t good enough for the real identifications.” Looking back on it, of course, they were dead right. And I think we all perhaps shared their enthusiasm for these identifications. And again, maybe we could have said right away to optical astronomers, if we’d been talking to them, “Look, this is what’s happening and it’s very exciting.” Well, I think at that point, our horizons weren’t yet expanded.

Sullivan

Do you think you would have gotten anywhere with the optical astronomers?

Graham Smith

Oh, I think there were one or two people in Cambridge who were interested in what we were doing. Yes.

Sullivan

Dewhirst.

Graham Smith

Dewhirst, in particular, yes. And I think Redman, although he’s a man who some people would say was old-fashioned, in fact, would have taken it very seriously. He really was an encouraging chap to talk to.

Sullivan

But now you did do the 1C survey paper to try to make identifications with many nearby galaxies.

Graham Smith

Yes.

Sullivan

And I don’t quite understand your remark saying you should have been a little more enthusiastic in trying, because you actually went overboard.

Graham Smith

We did, no, I mean – what I meant was – I think we should have taken the Australian results more seriously.

Sullivan

Oh, I see what you’re saying.

Graham Smith

We, I think at that time, we were saying, “This sort of position accuracy isn’t good enough.” And in point of fact, it was, as it turned out. And I think that was maybe a mistake.

Sullivan

Well, at least for objects that unusual.

Graham Smith

That’s right; that’s right. It was, perhaps, a lucky guess. But, it was a very good guess.

Sullivan

One didn’t know how low down you were going to have to go before -

Graham Smith

No, no. That’s right. When we felt that we were getting accurate positions out, because we got this Michelson type interferometer, rather than the Lloyd’s mirror, and we felt that we were getting more accurate positions. And, of course, we didn’t appreciate right from the start how many sources there were going to be, and so we didn’t appreciate the confusion.

Sullivan

Well, we’re jumping ahead now.

Graham Smith

I know, but even so, when you talk about 1C, 1C was, how many sources, most of these were wrong and that was a confusion (?).

Sullivan

I’ve never checked that out; I’ve looked at sort of the five or ten obvious ones and tried to see what they correspond to now, but I’ve never actually gone through the whole list.

Graham Smith

Most of them are (spheres?)

Sullivan

I didn’t realize that.

Graham Smith

I think it was at that stage that we were saying, fortunately not out loud in publication, but we had definitely gotten several of the extra-galactic nebulae, M31, M33, M81 and so forth. And we could prove with statistics were real because you couldn’t have accidental identification of so many sources. In point of fact, I believe those identifications were entirely without foundation because of the confusion problem, even in the 1C days.

Sullivan

I see, well, I’ll have to look at that again more closely. Now, (?) is that you built an aerial steel to do a full survey, having gotten Cygnus and Cas (overwhelming background noise) Can you tell me about that?

Graham Smith

This was really Ryle’s inspiration (?) Mostly there were two directions we went. One, to get accurate positions, and that entailed the interferometer which I used a great deal and the two 27ft. dishes. And at the same time, we were developing the larger arrays, trying to see more sources. And I think the development of the OH was very much a joint effort, although, of course, Martin Ryle was throughout the prime mover. The first idea to get a big array was to build a thing in the form of a bedstead and there we made a flat reflecting surface using stretched piano wires on an angle iron frame with a (?) side array of dipoles and we made one for each end of the experimental range we had, and we used that, and then the next stage was really to build two larger reflector arrays. We had all sorts of ideas of trying to make very large areas, collecting areas, cheaply. And we thought of the arrays of Yagis, and other traveling wave sort of antennas, and so forth. But we ended up building the large spherical cylindrical paraboloids – those are the ones that really going (?) telescopes.

Sullivan

What was your experience working in radar during the War – was it antennas primarily?

Graham Smith

It was largely microwave, they were in those days, coaxial cables so that I found it fairly easy to get on to low-noise amplifiers and coaxial cable techniques, and in particular the things like hybrid arrays and diode switches and things. In fact, I had a paper in the EE about hybrid arrays. That, I would say, was partly sparked off by Bernard Burke who turned up at the Department of Terrestrial Magnetism in Washington when I was there.

Sullivan

This was in 1954 or so?

Graham Smith

This is 1953, I think, yes. But that shows my interest, I think, in coaxial cables and other techniques – antennas came quite early, too.

Sullivan

Well, now going back to the 1C survey, what did you think you could do with this besides the identifications that were attempted?

Graham Smith

I think our objective was just to see what it was to be seen in the sky. I mean it was astonishing that the sun produced anything, even more astonishing that there were a lot of point sources in the sky, and the question then came: “What sort of things can we do?” Well, the first thing is you try and identify, and second, you may decide that you can find something statistically about the sky. On the first fifty sources, I did, in fact, do a log-n/log-s plot. Which that was the first one.

Sullivan

That’s right.

Graham Smith

I remember sort of a flash of inspiration that this would be a good idea, and it’s so funny to look back and realize that optical astronomers had been having all these thoughts years before, but yet they come as an entirely new idea.

Sullivan

Okay. So it was not based at all on the plots that they had been making of magnitude versus number of galaxies -

Graham Smith

Not at all, I hadn’t a clue. I had not been in contact with astronomers.

Sullivan

How were you learning astronomy – just reading a chapter in a book when it became necessary?

Graham Smith

Well, you see, I think I was acting much more as a radio engineer, thinking in terms of radio techniques and doing what we could to pick up radio signals from the sky. Now you find, you bump against astronomy, conventional astronomy, almost by accident. For example, when we came to finding that, well, take the time when Cassiopeia A first appeared on the record. This was the third occasion. I remember turning up at the laboratory, and the thing had been set running and there was an interferometer trace of Cygnus just as it should be – absolutely marvelous. And then three hours later, or whatever it was, another interferometer trace which subsequently turned out to be Cassiopeia A, but nobody had ever detected this radiation before, we didn’t know it was there, it appeared as a fluctuating trace with a different periodicity than Cygnus A. This was a complete surprise. We thought it must be something to do with the way things moved in the sky, but we had no training, neither of us, in celestial coordinates. I went away to think about it, and I remember coming into the Cavendish Lab an hour or two later and saying to Martin Ryle and to Ratcliffe, “I think it’s in a constellation called Cassiopeia.” That, of course, became the problem using the Würzburg dishes, which, incidentally, were largely constructed by another chap, Machin, who appears in these early papers.

Sullivan

Right.

Graham Smith

And became the problem of finding positions with such an interferometer.

Sullivan

Hold it now. How did you get in the constellation of Cassiopeia? After thinking about it, you determined that the fringe rate should be different?

Graham Smith

I decided the fringe rates would be different according to the distance from the pole, so that was my first introduction to spherical astronomy.

Sullivan

Right, and when you say that Machin was the one that built the Würzburg dishes -

Graham Smith

This was largely his. He was also very much engineering inclined, and when it came to the idea that we ought to build an accurate interferometer, where the electrical centers of everything could be accurately determined and so we could find the accurate positions of anything – solar sources as well as other things – he largely did the construction, and I took over and I did most of the electrical (?) on it. I installed coaxial cables and so forth; he joined in on (?). But to give another example of my lack of astronomical background, when it came to using the accurate interferometer, of course, I had to take account of the fact that this was a kind of instrument which was not aligned on an east-west axis, it was 5° off.

Sullivan

Why was that?

Graham Smith

Well, it was built along a hedge of a field.

Sullivan

Oh, I see.

Graham Smith

Precisely, yes. It was roughly east-west as far as we were concerned, but it became rather important that it wasn’t exact east-west – it was 5° off. And it was about 20 minutes of arc off level, you see. And it was also an unknown distance apart, which we had to measure very carefully; we had to find out where there was a standard 100 meters, which was not so easy. We found it in the surveyor’s department in (?) Hall. At any rate, to find out how to correct for this 5°, I had to learn some spherical trigonometry. I was, in fact, (?) in the tripos examination during that summer and I spent the hours of peace one gets in an examination room working out spherical trigonometry from scratch.

Sullivan

I see.

Graham Smith

And I worked out how to correct for the lack of east-west and the lack of level of the instrument, and I’m proud to say I got the right signs for these corrections, and it was after I had completed my calculations. I found the positions of these sources that I was told that a chap called Bessel had done it a few years earlier.

Sullivan

You’re referring to the Bessel?

Graham Smith

Yes, of the Bessel functions. And if you look in the textbooks on spherical astronomy, of course, it’s all set up there. But I wasn’t familiar with these textbooks.

Sullivan

This is not a complete loss, by any means, (?) barely gone through it from scratch.

Graham Smith

Not at all, I’m very pleased to have done it, and I didn’t waste an awful lot of time, in fact.

Sullivan

Well, tell me now, I think we’re coming to the stage where you must have determined that you really needed to get good positions on the sources you could get them for, namely (?). So how did you go about doing that?

Graham Smith

I think one’s motivation can be very simply set out as saying you do the next thing that you know you can do well, and it was obvious that we could get the positons on this. We knew of the Australian “attempts” (?) identify and we thought they were not good enough – we’ve got to get better positions – and we had this accurate interferometer going. So that I spent some time, first of all, I worked at the lower radio frequency – 80 megahertz, using the arrays that we’ve already talked about. From there you have to worry about refraction, ionospheric refraction, which, of course, turned out to be a subject in itself, but as soon as we got the dishes, we went to a higher frequency, which I remember was 214 megahertz, which was again, corresponded to an old war-time radar equipment. Oh, good heavens yes, I remember we did, in fact, use an array, a radar array, which was an array of yagis on a single mount for a time, and then we went over to the Würzburg dishes.

Sullivan

I see.

Graham Smith

And so that somewhat determined the frequency we worked on, 214 megahertz. Now, that’s not much affected by refraction by the standards of those days, and that was really when I did a to concentrate piece of work which was very much all on my own, the position work. And that led to the position accuracy at about a minute of arc on four sources.

Sullivan

This is written up in Nature.

Graham Smith

That’s right.

Sullivan

In 1950 or 1951, one or the other.

Graham Smith

That’s right. I can’t remember the exact one, it would be about 1951.

Sullivan

1951, I think. Now having gotten, well, what else – I think there must have been more troubles than simply worrying about refraction.

Graham Smith

I don’t remember running into any severe troubles. I remember developing techniques, getting amplifiers with good noise performance and so forth. These were war-time radar preamplifiers, disk-sealed triodes, and I spent a little time understanding noise factor of these things and the sources of noise but I understand noise sources in triodes better than I understand transistors. And I think we just made steady progress doing this. There wasn’t a real set-back in it. It was just -

Sullivan

The end result was a quantum improvement by a factor of ten almost.

Graham Smith

Yes, I think that one sometimes strikes lucky in the jobs which stare one in the face, which is what happened at that time.

Sullivan

So given that you had these positions, what does one do with them?

Graham Smith

Well, we did, at that time, know one or two people in the observatory. For one thing, we were interested to know what our latitude was. I had discussed this problem, particularly with Dewhirst and I’d been along there and had talked to him about problems of identification. And he, of course, was the first chap who really tried to identify the Cyngus (?)

Sullivan

I talked to him at Cambridge, and he gave me his side of the thing. And he came up with 11/2 identifications if anyone wants to look at them.

Graham Smith

That’s right. And at that point, I think the advice to write to Baade came definitely from the observatory, whether it came from, (?) or Dewhirst or Redman, I really don’t know.

Sullivan

I see. And also, David Edge has shown me, he didn’t actually show me, but he told me about the correspondence between yourself and Baade, which I would very much like to have copies of if you’re willing.

Graham Smith

Yes, certainly. I don’t know what the best way of getting them, I hope I can put my hands on the copy. But otherwise, there should be copies in the Cavendish. And David Edge, you said, has a copy.

Sullivan

Well, I’m not sure if he’s got copies, or whether… I think, he must have made copies.

Graham Smith

I suspect so. You’re probably directly in correspondence with him. Why don’t you try to get them from him?

Sullivan

Do I have your permission?

Graham Smith

Of course, of course, certainly.

Sullivan

It was very soon, I think, just a couple of months after the Nature article, when Baade got Cygnus and Cas.

Graham Smith

That’s right, yes.

Sullivan

And I guess he wrote to you at that time. And what was your reaction?

Graham Smith

Oh, it was very exciting. I really was, I think, the first feeling that we were really astronomers. Here were real-life astronomers saying that we had done was exciting. Of course, let’s be fair, that had happened on the solar work before there was a lot of interest in the solar work had brought us in contact with astronomers. But here was a completely new field of astrophysics that brought us in contact with the California astronomers for the first time. And that was certainly very exciting, and the thought that we’d gotten objects in which new physical processes were going on was also very exciting.

Sullivan

But now, it’s interesting that you say that the optical solar astronomers had shown some interest, because I’ve just been talking with Christiansen and he said that it was very rare that they -

Graham Smith

All right, all right. It was a rare one, but we knew that we were in astronomy doing solar work… and we… (laughter)

Sullivan

I see.

Graham Smith

I think the (?) work on sources… I think we really still felt that (?) radio engineers finding out what it was that was coming from the sky, and we didn’t feel ourselves as having any sort of place in the community. And I think it was from that time on that we really felt as though we were slotting in. And we would get something from them, and they would get something from us.

Sullivan

So what was the next step? I gather you did not really go along the identification route yourself.

Graham Smith

No, it was definitely our task to carry on the kind of thing that we had done, and obviously we had begun to do quite well. So then you then have to build a larger array and get more sources and get accurate positions from that. So we were all very much involved in the techniques and actual efforts of construction, and, I as you know, we worked very much with our hands. We got very good at wielding sledge hammers and all that sort of thing. So at that point, I think that I kept going on positions with ionospheric work. I can’t remember when would be the last publication of mine on the ionospheric work; it would be about that time, I suppose. And then I think I helped him a lot with the 2C and 3C survey construction work.

Sullivan

If I remember right, you’re not a co-author on the 2C actually.

Graham Smith

No.

Sullivan

But you were working with the (?) -

Graham Smith

We all worked together very much.

Sullivan

Did you get involved in the interpretation of 2C and the definitions of what a radio source was, and so forth? Before it got published, I mean?

Graham Smith

Yes, I think we were still, not so many, but we were in any way differential in than what we did.

Sullivan

And, I can’t refrain from asking, as this went along, did you see any problems developing or did you … as far as too many sources? You’ve already mentioned -

Graham Smith

No, really no. I think it’s a problem which came to us by stages. I mean, each time we realized that we needed more angular resolution; say you build a larger antenna and you get more sensitive, and you see more of the (?) you see. I mean, it isn’t until we really reached the 3-C, and after the 3-C had been published, that there is a full realization of the magnitude of the problem.

Sullivan

But, so the comment that you made about 1C was definitely one in hindsight?

Graham Smith

Oh, it was -

Sullivan

No idea at the time -

Graham Smith

Well -

Sullivan

A large fraction of those might be in trouble.

Graham Smith

I don’t think there was full appreciation of what a large number of those small sources can be. But, of course, there was an appreciation of the fact that sources were not being seen singly, and that you had therefore had to get better angular resolution before you can really be sure of what you’re doing.

Sullivan

The blending, you mean.

Graham Smith

Yes, there was an obvious blending, but the full horror of the situation – that there were a very, very large number of sources to be detected, and much of the blending was a blending of a very large number of small sources – I don’t think that was really appreciated until the analysis of 3C by Hewish and Scheuer. Yes, I would think if there had been a full appreciation, 3C wouldn’t have been published as it was.

Sullivan

2C you mean.

Graham Smith

2C. Well, even 3C had quite a lot in it.

Sullivan

Well, okay, I’m not quite aware; I’ve always thought of 3C as (?) globe-shift problems.

Graham Smith

Well, there’s 3C and 3C revised.

Sullivan

Right.

Graham Smith

And 3C revised is a much sounder catalogue than 3C. And 3C certainly affected not just by lobe shifts, but by some over-optimistic interpretations of individual sources that were not real.

Sullivan

I didn’t realize that. 3CR is -

Graham Smith

3CR is sound.

Sullivan

Right. But what were the differences? Checking for lobe shifts?

Graham Smith

Yes, that’s right. And using the stringent criteria about identifying individual sources. That was done by Bennett. But I had quite a bit to do with that.

Sullivan

Can you be more explicit about (?)

Graham Smith

I think that’s a little hard. Bennett set the task which I think was supervised quite largely, of referring back to the original records and sorting out which sources were quite unequivocally right, which behaved (?) proper beam shape in both coordinates. And being as objective as possible, when one could say there was not likely to be a shift when there was likely of being a shift. I mean there were many fewer sources in 3CR that there are in 3C. And that was a result of the weeding process.

Sullivan

I see. I must admit, I’ve looked at 3C, but I haven’t looked at 3CR. That came out in 1962?

Graham Smith

Yes, approximately then. If you compare them, I think you’ll see that (?)

Sullivan

Okay, well, now if you’ll bear with me… let’s just see if we’ve missed any points. Okay, we’ve got the 1C survey, we talked about the ionospheric, (?) you had a paper in Nature that we mentioned before on radio stars, and then you also had one about trying to measure the annual parallax.

Graham Smith

That’s right.

Sullivan

Or proper motion.

Graham Smith

That’s right. Well, of course, that’s right. This was a long continuing series of observations. I remember observing those four sources throughout a year, without missing many days or nights. And, of course, those four (?) sources came across the meridian at fairly evenly spaced times through the day and night. So I got entirely used to getting up, dashing out and taking observations and going back to bed. And that was because we really hadn’t identified them at the beginning of that, so we didn’t really know whether they were near or far. We had no clear idea of whether these things were even possibly in the solar system. Certainly we had no idea where they were in the galaxy or extra-galactic. So we thought, “Well, we’ll measure parallax and measure proper motion. Let’s just go on the positions.” And, of course, we find the ionosphere is a nuisance in that, so you find yourself investigating the ionosphere, and in point of fact, you don’t measure any parallax or proper motion – that’s the way it goes.

Sullivan

Can I ask you, do you remember sort of what your inclination was … I see, it looks like we are going to get inundated here. Storm from the East.

Graham Smith

It may not rain, it come from the East but it doesn’t rain every night.

Sullivan

I see. Well, we’ll wait. What was your predilection as to the nature of these things?

Graham Smith

The nature of the position shifts?

Sullivan

No, the other sources, radio stars, as they were.

Graham Smith

Certainly when I worked out a log-n/log-s, I was thinking of them as galactic objects. Martin Ryle had some speculations, in a paper in Proc Phys Soc. I remember, as to the nature of things. And I think we were all a bit baffled about the possible radiation mechanism. The idea of the synchrotron radiation was coming out in a paper by Alfvén and Herlofson about that time.

Sullivan

And Kiepenheuer.

Graham Smith

And Kiepenheuer as well, yes. Hadn’t really sort of latched onto this, so we hadn’t got much of a good physical picture as to what we could do.

Sullivan

You bring that up, of course the Russians really carried the ball with this synchrotron radiation.

Graham Smith

Oh, indeed.

Sullivan

But were you ever aware of the Phys. Rev. articles that suggested this in 1950?

Graham Smith

Yes, I think so, but I wouldn’t say that anyone of us saw with a flash that this must be radiation by synchrotron emission, and must therefore, or could therefore, be at a very great distance. I mean, I think one would need to be very perceptive to pick up one of those articles and suddenly see the whole thing as falling into place.

Sullivan

Well, sure it was of course, speculation.

Graham Smith

It’s more gradual than that.

Sullivan

Well, here’s another one, Machin and Smith, “On a new radio method of measuring of electron density in the solar corona.”

Graham Smith

Oh, yes.

Sullivan

Nature, in 1951. Can you tell me about that?

Graham Smith

Oh yes. I can’t remember who had the first ideas on that. I have an idea it might be Machin, but we certainly discussed together the ideas of refraction in the solar corona, and it was obviously part of (?) And the idea is that if you keep on measuring the position of radio source and it gets close to the sun, you may be able to see an extended solar corona. Because of the refraction. And we tried this.

Sullivan

In 1950-51.

Graham Smith

That’s right. Of course, it didn’t work. Because what you see is irregular defraction, rather than a regular refraction. In fact, regular refraction doesn’t seem to (?)

Sullivan

This is telling you something about the nature of the corona.

Graham Smith

Indeed, yes. Yes.

Sullivan

So, did you follow it up at all?

Graham Smith

I was in it for a year or two, but then that very much became Hewish’s work, and it still is.

Sullivan

Right. I talked to him about that in the recent past. I think we’d better move this inside – rain.

[Brief Break]

Sullivan

So an occultation in 1952 was also reported in Nature by you and Machin in 1952.

Graham Smith

That’s right. I think that was probably the last one that I was involved in there.

Sullivan

Okay. Here’s another set of observations, and this was on the angular sizes.

Graham Smith

Ah, well now this is a different subject that I really moved onto at this point. As you know, you get angular diameters by using interferometers of different spacings. And we didn’t have anything on railway lines at that times, we only had an interferometer with a fixed spacing, so I started the idea of using a smaller, portable antenna which I used in conjunction with the two fixed ones to make two baselines that I could use either sequentially or even at the same time. And in that way, I did make the measurements of the diameters of Cygnus A and Cassiopeia A radio sources.

Sullivan

What was the real motivation for getting these angular diameters? To aid in identifications or to learn more of the physics of the sources or what?

Graham Smith

I think identifications first, but again, I think one has to remember that this sort of joy at being able to measure anything and say, “By gosh, we can not only measure the positions, we can also measure the diameters.” And I think, as a physicist, of a practical experimental turn of mind, on immediately recalls Michelson’s work and said, “Well, gosh, we can do the same sort of thing that Michelson did.” And it’s not long before you’re led on to Fourier analysis of the source distribution. In fact, of course, in Australia, Pawsey was onto that with the solar work and he used, I think, he was probably the first person to find out you could use Fourier methods.

Sullivan

Yes, I think so.

Graham Smith

I think that the attitude is the same in both cases; you find a radio physicist who suddenly finds he can apply his interests and techniques into a new field, and you don’t sit around saying, “I wonder what these sources are. If I could measure the diameters, I would be able to tell. How can I measure the diameters?” You work the other way, you find that you’ve got a jolly good technique and you say, “What’s the most interesting ways of applying this technique?” And from that, you get the fact that these radio sources have finite diameters, and then you start thinking, “Well, maybe this tells me something about them.”

Sullivan

Right. Would you say that happens at all today? Or is it more the other way today?

Graham Smith

I think it’s more the other way today because we have very much more sophisticated radio telescopes, which, of course, are capable of measuring very many more things than they ever will measure. In a way, we were working on techniques which had a limited application because all the antennas and so forth were pretty small, and really we’d say, “If we just modified this apparatus or build a slightly better bit of apparatus, we can measure something more.” Whereas today, to measure something more, you can go on using the same apparatus on different objects. And so you naturally say, “What are the most interesting objects to do this measurement on?” And that takes you more in the way of conventional astronomy.

Sullivan

Yes, yes. Was the thrust to measure angular sizes influenced by the concept of radio stars, that you might well expect that the thing could never be resolved.

Graham Smith

Yes, I think so. Of course, as a technique, it was probably influenced by the solar work, where you can tell if it’s a sun spot or the whole corona by measuring its diameter. So that I think there was a clear feeling that you could get into the nature of the source by measuring its diameter.

Sullivan

But for instance, did anyone ever say, “Well, the stars are a few thousandths of an arc-second in diameter and -

Graham Smith

Oh, I think that we thought it couldn’t be the surface of stars that we were looking at. We knew that it would have to be something extended. Oh yes. I mean, I think we thought enough about the nature of the sources to say it can’t just be the surface of the star radiation on a black body curve, we knew that very well.

Sullivan

Right, but still, it might be much smaller than an arc second even.

Graham Smith

Yes, of course, at that time that was fairly unlikely thing, because it was very difficult to conceive of a thing which had a brightness temperature as high as these things would have. I think you’ll find that sort of concept in Ryle’s Proc Phys Soc paper about 1951 or so.

Sullivan

You knew it was not thermal, but still that was a bit (?)

Graham Smith

No, it’s clearly not thermal, but you see, the solar radiation turned out to be thermal, quite largely. The corona was at a million degrees, and we knew that even if the thing is not thermal in the sense of being a black body, still it probably had to have particle energies which corresponded to the brightness temperatures we were seeing. That, in fact, has remained true through most of radio astronomy. It’s only when you get to the pulsars that you really have got coherent radiation and the particle energies do not need to be that high. So it was very unlikely that you’d get particle energies that high, and then, of course, there were -

Sullivan

Well, the masers came before pulsars, to be fair.

Graham Smith

That’s true, the masers did come first. But even if you had particle energies that high, you’ve got to start worrying about a mechanism that doesn’t get optically thick and we already knew of that concept. So that I don’t think we seriously thought that these were stars whose hot surfaces were producing radio waves, no.

Sullivan

Now, I talked with Bernie Mills a couple of days ago about the set of three papers that appeared together in Nature and how they came about at the URSI meeting. What is your version of that URSI meeting and the work in Sydney and so forth?

Graham Smith

Yes -

Sullivan

1952, that was.

Graham Smith

Yes. That was the first big scientific meeting I’d ever been to, and I was, I think, in a fairly bewildered state. I think it’s true that up to that time, I hadn’t had a full appreciation of how much was going on in Australia, and it’s certainly true that we had been working independently, so a whole lot of observations came together and –

End tape 68A

Begin tape 68B