[Bernard Mills]
Bernard Mills, 2006 (Photo courtesy Chrys Mills)



NATIONAL RADIO ASTRONOMY OBSERVATORY ARCHIVES

Papers of Woodruff T. Sullivan III: Tapes Series

Interview with Bernard Y. Mills
At Grenoble
25-26 August 1976
Interview time: 165 minutes (parts 1 and 2 combined)
Originally transcribed as typescript only by Bonnie Jacobs (1977), retyped to digitize by Candice Waller (2016)

Note: 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. 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.

Click start to listen to the audio of tape 61A of the 1976 interview.

Begin Tape 61A

Sullivan

This is talking with Professor Bernie Mills on 25th August 1976 at Grenoble, sitting in the grass, with the cars going by. Could you tell me a little bit about your educational background and what work you were doing before you got into radio astronomy?

[Indecipherable interchange, perhaps about traffic noise]

Mills

Perhaps I can start with graduating from Sydney University as an Electrical Engineer in 1942, end of 1942, when I joined CSIRO, Division of Radio Physics, to work, along with everyone else, on radar. And I became involved particularly in the receiver section, receiver development, and in electronics and display work generally. I worked on several systems and became increasingly interested in the basic design of electronic systems and developed quite a lot of some of the radar systems which were being made in Australia at that time – the receive and display sections. At the end of the War we diversified efforts quite a bit and I became interested in applying radar techniques to linear accelerators. And, in fact, at this time, some initial experimental work had been done just hooking a magnetron into a resonant cavity to see what could happen. A few hundred keυ were obtained. So I joined the Valve Lab which had been formed then in an attempt to develop this idea into something that could be more practical than was this initial experiment.

Sullivan

This was CSIRO?

Mills

Yes. The aim was, in fact, to provide a working, pulsed X-Ray tube. And I got that going, getting about a million volts. This, I think, was the first practical application of the idea, applying a magnetron to a resonant cavity, and I developed a lot of the basic ideas involving stabilization of resonant cavities in a pulsed mode.

Sullivan

Did you know about the radio astronomy group that was forming at that time?

Mills

At that time, it wasn’t. This was immediately after the War, and I became interested then in the physics of things more. I began to take an interest in high energy physics with other people in the group who were also interested in this. And this was chopped off at the end of ’46. It was obviously too expensive for Australia. So I was at a loose end. I started out with a thought of getting into digital computing which was just starting in those days. F. C. Williams had produced a digital computer in England and we were playing with the idea in Australia. So I, since physics was my background – so I played with that for a while. But then I was picked up as a TB case and put off for 6 months. I overcame it completely, but then that left me with a free field. And in this time, radio observations in radio astronomy had really got going and John Bolton had found some point sources – not identified then.

Sullivan

This was ’47 or so?

Mills

Yes, that’s right. Of course, Pawsey had observed bursts from the sun. And it all looked very interesting, so that was my introduction.

Sullivan

Let me just ask you before we go on any further; did you hear about Hey’s discovery of solar emission or any other discoveries of solar radio noise during the War?

Mills

Yes. We’d heard of the bursts of radiation from the sun which caused radar interference. And Pawsey was, in fact, quite interested in this and had done an experiment during those early days. Someone else may have told you.

Sullivan

During the War?

Mills

During the War where he stuck a simple parabolic reflector out of the window and looked at the sun and saw nothing because it was at 10cm. (laughter) But that was the stimulation, I think. Pawsey became interested and then at the end of the War he started things off.

Sullivan

I gather that there was just no time to try to do it a second time or follow it up?

Mills

Oh, no. He just looked and saw nothing.

Sullivan

Because, as you know, Southworth did detect the sun, but he probably had more sensitive receivers. Would that be true; do you think?

Mills

I don’t remember. What frequency?

Sullivan

Oh, was at 10, 3, and 1 cm. He had state-of-the-art receivers.

Mills

Well, I think they would have been comparable. We were in very close contact with the Americans in our equipment and all the microwave stuff came direct from America.

Sullivan

Do you know of any possible record of that experiment? Is it mentioned in any –

Mills

I don’t think so. Just a few of us knew about it. Frank Kerr might remember.

Sullivan

I think he did mention it to me – it’s been several years since I interviewed him. Ok, having come back in the field fresh. It wasn’t called radio astronomy then, what was it called?

Mills

Oh, I can’t really remember. I think we were a cosmic noise group and a solar noise group, if I remember correctly – the whole field. Everyone was thinking very hard what they could do. There was very little instrumentation immediately available. There were lots of questions. And I first became involved on the Sun by observing a solar eclipse. You’ve probably got the paper there.

Sullivan

That’s Christiansen, Yabsley, and Mills ’49.

Mills

Yes. You could get more details about that from Christiansen probably, because he was the main person to organize it.

Sullivan

Well, I haven’t talked with him yet, so maybe you could just tell me about what you did.

Mills

Well, I there was a partial eclipse in the southern part of Australia extending down to Tasmania, and what we did was to have – I can’t remember if there were 2 or 3 stations now – I looked after the local station and Christiansen and Yabsley went off. And whether they had 1 or 2 stations, I’m not sure. Tracking it from a different place further down south, so that we got intersecting arcs. By that mean, as the sun was occulted, we were able to get the separately placed arcs on it and locate the positions of radio emission, and we did find they corresponded to sunspots.

Sullivan

Now was this before the McCready, Pawsey, and Payne-Scott paper?

Mills

No, that was after it. But they had detected this big burst from a sunspot. But the question was what happened at high frequency, this was quite a high frequency -

Sullivan

50cm.

Mills

Yes. That’s right. Well, in those days that was a pretty high frequency. And the question was what happened there. We found that we had quite steady emission at this frequency from the sunspots. I think that was quite an advance at the time.

Sullivan

But the abstract said that you did find a few bright spots which had an effective brightness temperature of 5 million kelvins, so there must have been a little bit of activity.

Mills

There must have been, as we know it now. In those days, of course, we had no idea of what was going on.

Sullivan

Now things like this partial eclipse – was the group sophisticated enough, I’m not sure that is the right word, to be looking ahead to find out when there were eclipses? Or was it just someone reading the newspaper and saying -

Mills

I don’t know.

End Tape 61A

Click start to listen to the audio for Tape 61B of the 1976 interview.

Begin Tape 61B

Sullivan

This is continuing with Christiansen on 26 August 1976.

Mills

Mills! (laughter)

Sullivan

Continuing with Mills! You were saying that you had just joined Christiansen’s group and so you weren’t part of the planning period.

Mills

Well, I hadn’t really joined his group. I joined the whole radio astronomy group and Christiansen, along with some others, was planning this eclipse work, and he asked me to join forces with him to look after the local station while he went off.

Sullivan

So I should just ask him.

Mills

Ask him, yes.

Sullivan

So what was the next step after this experiment?

Mills

Well, it rather whetted my appetite, and the question was what were the interesting problems at that time. I remember discussing this at some length with Pawsey, and I think it came down to two. One was a search for the hydrogen line. He was very interested in it at the time. And the other was trying to locate very precisely the positions of radio sources. And it was a difficult decision to make. I eventually chose the precise positioning because some of the techniques I was more familiar with, and it looked as if it was something that would lead to an immediate result, whereas the other was extremely speculative – whether anything would be seen at all.

Sullivan

You didn’t realize it was going to take you 10 or 15 years to get those precise positions.

Mills

Oh, actually we got some quite precise ones very quickly.

Sullivan

That’s true for the strong sources. But I’m interested about this looking for the hydrogen line. Of course, it had been predicted by Van de Hulst in his famous paper, and Shklovsky also had a paper about it which I suppose you knew about at that time, ’48 in the Astronomicheskii Zhurnal.

Mills

I don’t think at that time we did know about that Shklovsky paper. Only the Van de Hulst.

Sullivan

But what I’ve heard from various people is that, in fact, the Australian group was not racing with the other two groups. It was only once they’d gotten the word that it had been detected in Harvard and Holland that they then slapped together a receiver to -

Mills

Yes, that’s perfectly true. That’s what happened simply because I think no one really felt strongly enough about it to put the work into what we felt, or I felt, might be a rather speculative sort of thing. There was no certainty of the positive results. There were lots of other things to be done.

Sullivan

You knew that radio sources were there.

Mills

Yes.

Sullivan

Were the technical difficulties forbidding for the hydrogen line thing?

Mills

Well they appeared rather forbidding. One knew one had to get right down to the absolute maximum theoretical sensitivity because the thing was going to be faint probably. As I said, after some discussion it wasn’t clear that it would be above the detectable limit. So it was a speculative project from that point of view. And one knew that the Dutch were doing it too.

Sullivan

Oh I see.

Mills

I’m pretty sure we did know that.

Sullivan

Van de Hulst’s paper itself said this is only a possibility. He had not come out and said it’s a sure bet. Now the radio sources, what was the reason that you wanted better positions at that time?

Mills

Well, it was a matter of identification. I don’t think at that time there had been any clear positive identifications. I’m not even sure there’d been suggestions, but Bolton may have suggested the Crab. I’m not sure about that – it was just about that time that the first identifications were suggested. And it was clear that to make any progress we had to get onto these and get very accurate positions. I felt that the cliff technique, rising above the sea, was not a way to get accurate positions.

Sullivan

Because of the problems of the refraction?

Mills

Yes, that’s right. So at that time Little and Payne-Scott were constructing an interferometer, a very simple one with a couple of Yagis for observing the bursts in the sun and motions of bursts. In fact, they were the first, I think, to observe these bursts flying out from the sun. And this equipment was available and it appeared to me that with fairly simple modification, it could be used for determining the position of the strongest source we could see, Cygnus. And so I set about doing that as my first independent job in radio astronomy.

Sullivan

I see. Just before going any further, I’m interested in this talk with Pawsey. You listed two things that he suggested as possibilities. Was the idea that solar radio astronomy was important, but that was being taken care of?

Mills

Yes. Paul Wild was looking after that at that time. But in the other part obviously there were a lot of things which could be done and John Bolton was really the only one working on it at the time.

Sullivan

And also since I obviously can’t talk to Pawsey, I’m also interested in how he was thinking at that time. Is it possible for you to remember at all?

Mills

He was intensely interested in nature as such, and physics. He was one of the most dedicated people I think I have ever struck. As I said, he had become interested when he’d heard of the interference during the War, and this had stuck with him and he of course knew of Reber’s work by then. He was just interested in getting at the physics of all this, and finding out exactly just what was going on.

Sullivan

So he was much more on the physics side than the radio engineering side?

Mills

Yes. He was, really. He was a very competent radio engineer, particularly on antennas. But perhaps not quite so good on electronics. He was very competent indeed. But his real interest was in the physics of things.

Sullivan

So your first independent project, you said, was working on the position of Cygnus. And you and Thomas published a paper in ’51.

Mills

Yes. A long time after the results had been obtained, I’d say.

Sullivan

Any particular reason for the delay?

Mills

No, I don’t think so. It was all tied up with doing a lot of other things at the same time, and both of us were quite inexperienced in doing basic physical research – we were both engineers by training. And this took a bit of time to get used to.

Sullivan

Just how to write things up?

Mills

Yes. And we still wasted – wasted I think is the word – a lot of time on the on the ionospheric effects which were of considerable interest at that time. We did develop a sort of qualitative theory involving irregularities in the ionosphere which is, I think, quite sound as far as the qualitative theory worked, but we couldn’t get it on to a proper quantitative basis. I think that was one of the main reasons for the delay, and actually getting a paper written up.

Sullivan

Yeah, in the abstract it says that the fluctuations correlate with the F-region, the fluctuations in Cygnus correlate with the F-region.

Mills

Yes. So we did find that, but were also interested in actually constructing a theory of how it worked. And as I said, we were not eventually successful in producing anything decently quantitative.

Sullivan

So it seems like you were as willing to work on the ionosphere as on what we call radio astronomy now, anyway.

Mills

Well, both problems. Yes, I guess I was.

Sullivan

Maybe the distinction wasn’t so clear then.

Mills

It wasn’t. It wasn’t even clear that time when I started that the fluctuations were ionospheric in origin.

Sullivan

Right, well I’d like to hear more about that.

Mills

There was a big argument that they were inherent in the source. Of course this had been argued that this was a source of radiation that was fluctuating rapidly and therefore it was a point source.

Sullivan

That was in Hey’s original paper.

Mills

Yes. But people had realized, and I can’t say who or how, I know I had the (?), that the same argument applied whether the fluctuations were inherent or imposed by the ionosphere. It still must be a point source. We did some experiments. I remember we had some with John Bolton – these were spaced experiments to look for correlations at different spacings – and found none. I don’t think this was even written up anywhere – John Bolton may have, but it certainly showed that the –

Sullivan

It’s only sort of a footnote here and there, that I’ve been able to find on it.

Mills

Yes. It showed that it was clearly not in the source, but -

Sullivan

Were you aware of the work going on in England at this time on the same sort of questions?

Mills

Yes. Most of the publications were in Nature, of course, and we knew all those as soon as they came out. We’d also had some contacts with the Cavendish group.

Sullivan

Were you aware of the work going on before it got published?

Mills

Not at that particular time, although we were later. Bracewell was doing a PhD at the Cavendish at that time. He, like me, joined immediately on graduating with a Bachelor’s Degree, but he went off later to get his PhD in Cambridge. And so he kept us informed, you know, what was going on, and similarly, I think he kept them informed of what we were doing – he was one of the sources, and Pawsey himself had contacts. He was a very good friend of Ratcliffe, for instance, and there was a lot of correspondence going on there.

Sullivan

Do you know if Pawsey has any sort of archives or his widow or anything like this?

Mills

No, I don’t really know what happened to his things. Possibly in the Academy Library.

Sullivan

Well, let’s just look at the basic results of Mills and Thomas ’51. You got the position to 3 arc minutes, the flux measured to within 20%, less than 3% polarized, less than 10 arc minutes in size. Although that had also been established, I think before that time, by Bolton’s sea interferometer. Is that true, that is was less than 7 arc minutes in size or something?

Mills

Yes. I think that had been established.

Sullivan

And you say the parallax was less than 1 arc minute, therefore it must be outside the solar system, so this was apparently a distinct possibility the thing was a comet or something like that?

Mills

Yeah. Or before that it could have been anything (laughter).

Sullivan

Right. Now did these results – what sort of light did they shed?

Mills

There was one other interesting result there in that we pointed out the position of a galaxy which corresponded with this.

Sullivan

That’s right. This is the first overlay, isn’t it, in the history of radio astronomy?

Mills

Yeah. It also turned out to be the so-called colliding galaxy, but I didn’t positively identify it. I actually wrote Minkowski about this, pointing out the coincidence and in typical style Minkowski wrote back saying, “This was an ordinary 17th or 18th magnitude galaxy. It couldn’t possibly be the radio source.” And we just mentioned this in the paper.

Sullivan

I remember now. In fact, I’ve gone through this rather carefully. I don’t know if you’ve ever done this, but your position was actually off a bit more than your errors in R.A., I think. But your overlay also was a bit incorrect.

Mills

Yes, it was.

Sullivan

And this conspired such that it was still within your error box.

Mills

Well, in those days we used to use probable errors, anyway, which, of course, made things a bit easier to get coincidences. It was, in fact, I think, only some 2 or 3 probable errors off, but still a reasonable identification. It was pity, really, we hadn’t taken it more seriously and taken a proper photograph at that time.

Sullivan

That would have been a year or so before, I guess, the Smith position and so forth?

Mills

Yes.

Sullivan

So it was still a complete mystery as to what this Cygnus A thing was?

Mills

Yes. We felt that, and I think Minkowski helped here, it was probably unlikely to be a galaxy, so distant, because it was the strongest source and it was in the galactic plane. And these two made us think: “Well, maybe this was just a coincidence.” And I am sure this is what Minkowski thought about it too. It looked as if it was in fact a galactic source because of its position in the plane.

Sullivan

I get the impression from reading papers at that time that most people were thinking that these were galactic sources. For instance, in the identification paper, Bolton, Stanley, and Slee, they don’t seem to want to believe that [NGC] 5128 is a galaxy – they say it’s been called a galaxy, but so and so has said that . . . Is that right that people are fighting against the idea of these things?

Mills

The term “radio stars” was in common use. I think actually the Cambridge people were responsible for this concept. It caught on very strongly and there was a very strong school of belief that these were in fact stars. The small size – well the strongest, Cassiopeia and Cygnus, were out on the galactic plane – and things like this – made a lot of people think, in fact, they were radio stars. And I think we were also affected to some extent. I think I had a comment about it being unlikely that an object so large as a galaxy should be so unusual as to emit enormous quantities of radio emission, something like this.

Sullivan

Although, were you at the stage at this time when you were actually figuring out what the intrinsic radio luminosity of such an object would be? Were you that much into the astronomy of it, so to speak, or were you just qualitatively saying, “Well, it would be so far away that it - .”

Mills

No, I think it was more of a qualitative argument that it was a very, very large object consisting of very many stars, and it seemed in those days that everyone was thinking in terms of local physics, fairly small objects like stars, and it just seemed inconceivable that a large object could emit a great deal. We knew the Galaxy emitted radiation, but then the Cygnus one was obviously millions of times more bright than the Galaxy. It wasn’t clear - at all.

Sullivan

You were thinking of a nucleus or something.

Mills

Not at all.

Sullivan

This concept of “radio star” interests me. I was interested, for instance, to notice when I was at Cambridge looking through the theses that Edge’s 1959 thesis, which was the 3C survey basically, was titled “A Survey of Radio Stars,” even in 1959.

Mills

That was the dogma in those days.

Sullivan

But I mean then it was clearly recognized. Ryle himself, in the Bakerian lecture, had argued that these things were extragalactic, very strongly.

Mills

No, not in that. Wait a bit. I must get the dates. Oh, yes, by then, yes, that’s true.

Sullivan

And in ’55 he was talking about cosmology.

Mills

He had a complete reversal about ’55.

Sullivan

That’s right. But that’s what I’m wondering, yet the terminology, did the term linger on even though people realized they weren’t stars?

Mills

I would suspect only in Cambridge. We certainly didn’t. Right from the beginning, I think, I don’t think I called them stars. Ryle was weighted towards that belief. When I first joined the group, I sort of always had niggling suspicion that they weren’t, particularly in view of the possible identifications at that time with M87 and [NGC] 5128. Well, M87 particularly. I think that was one of the possible in the early days.

Sullivan

Oh yeah, there were three of them. Centaurus, Virgo, and Taurus. Let me ask about those three identifications, as to the Australian group’s feelings, as to their certitude, so to speak. Did it seem pretty definite that these were right?

Mills

Well, the Crab looked pretty good. Of course it was such a peculiar galactic object anyway, and it tied in with violent events. It looked as if it was probably good. The others, I think, were always a bit, we always felt in the back of our minds that they may not be exactly right. This is one of the reasons I continued on that idea and built bigger antennas with the idea of getting more accurate positions with a larger number of sources.

Sullivan

To pin it down?

Mills

Yes.

Sullivan

Now M87 wasn’t known to have, at least by you anyway, to have this peculiar jet.

Mills

No.

Sullivan

I think there actually was a photograph way back in 1918 that some optical astronomers knew of it. But I haven’t heard any radio person that knew that at that time. Continuing with Cygnus A, before we get to your first survey, there’s this famous set of three papers in Nature in 1952 by the Jodrell, Cambridge, and your group, on the positions and sizes of radio sources. Well, really more sizes.

Mills

I think that paper was on the size at that time.

Sullivan

Right, and you have Cygnus less than 1 arc minute, Taurus less than 4, Virgo less than 5, and Graham Smith has similar sort of things for Cas and Cygnus. I wanted to ask you, what did you think when these three papers came out about this asymmetry that Hanbury Brown and Jennison had with their entirely different technique? They came out strongly saying that the thing is very elongated.

Mills

Did they at that time? I don’t remember that.

Sullivan

Yes. They didn’t have the double quite yet, but they said that is was 4 by 1 or something like that.

Mills

I see. I’d forgotten that one. I can’t remember too clearly the details of that. We had just got our radio link interferometer going, and these were very much preliminary results with it. And the interesting this, as far as I was concerned, was that they all resolved. This said they were not stars. And I remember, in fact, at that time bringing this to Piddington’s attention, because he was still continuing to think in terms of stars. And I remember commenting: “Well these were quite clearly not stars.” We had to think in terms of galaxies rather than stars. That really convinced me.

Sullivan

And I guess Bracewell was back by then, or was he? Because he has this nice little paper in Observatory, “Radio Stars or Radio Nebulae.”

Mills

Oh yes, he’d come back and was discussing this with us at the time. And of course we had also measured angular sizes of some other sources, such as Centaurus A which we found to be very large.

Sullivan

And Fornax?

Mills

Fornax, too. This had been done with our other interferometer which we had constructed for the survey. Smith had built a three element interferometer for his measurement of Cygnus, but we also had this three element interferometer going, which measured these large sources. So we had a general picture, and I think this was what struck Bracewell when he came back, that our concept of the Universe was different from the Cambridge one. They thought in terms of point sources uniformly scattered over the sky, radio stars, in fact. And we had in the south, we had measured all these large angular sizes, and it was obviously much more complex than this. It has been one of the sources of the difference between us in our outlook, I think, that just by chance that the northern ones are smaller than the southern. (laughter)

Sullivan

A conspiracy of nature! But nevertheless, there were Centaurus and Cygnus, and Taurus, which Graham Smith himself measured to have finite sizes.

Mills

Cygnus and Cassiopeia. Centaurus is a southern object.

Sullivan

I said Taurus. I forget if he measured all three of those. I think he measured Taurus and Cas, it may have been.

Mills

And Cygnus.

Sullivan

Probably Cygnus.

Mills

No, he - I can’t remember, we’ll have to look that up. I know he measured Cygnus and Cas.

Sullivan

Anyway, is it your view that these were sort of looked upon as aberrations by the Cambridge group?

Mills

I don’t know. I can’t remember at that time. I think Cas and the Crab were definitely thought of as being galactic things, and I’m not sure about what they thought about Cygnus.

Sullivan

Well, I am going to try to get Graham Smith at this meeting too.

Mills

Yes, he may be able to help you on that one.

Sullivan

Well, what about this new intensity interferometer technique? Was that your first knowledge of it when you saw this one-page thing in Nature?

Mills

Oh, no. All those came out together in Nature. We’d heard about it quite some time ago when they were first getting going, when people had been traveling around and bringing back the news of what they were doing. So we knew they were doing this particular job. During the ’52 URSI meeting in Sydney, both Graham Smith and Hanbury Brown were there and we got together and talked about all these things.

Sullivan

These papers came out of that meeting?

Mills

Yes, that’s why they’re all published together. We organized a joint publication. We were not seriously concerned about the difference in sizes or shapes that we got, since it was obvious that we had only done a fairly coarse sampling of the stuff.

Sullivan

Although, I think that once again, if I had the paper here I could verify it, this is the paper where you had several spacings, didn’t you?

Mills

Yes.

Sullivan

And if you had put a couple more in, you could have gotten the double nature!

Mills

Yes, I could have. It was unfortunate.

Sullivan

Famous looking back, retroactively.

Mills

Yes, I did all my interpretations on the basis of some sort of Gaussian distribution.

Sullivan

Right. Well, if you had no other evidence, sure. But still, you haven’t said what you thought about this intensity interferometer. Did it make sense, to you, that this worked?

Mills

Oh, yes. I think we did understand roughly how it worked. I mean physically we knew it must work. I think we were a bit unhappy about sensitivities. Again, Hanbury Brown and Twiss had made were quite clear on this, it wasn’t a very sensitive instrument, they could only see the bright thing. We had discussed it during the URSI meetings, and there were no worries about it working at all. It was only the quantum physicists who were worried about it working, I think. The radio engineers were perfectly happy (laughter).

Sullivan

Right, there was quite a debate later on, as you know.

Mills

Yes.

Sullivan

Since you brought it up, let me just ask you about that ’52 URSI meeting in Sydney. Is it true to say this was held in Sydney in recognition of the Australian radio astronomy, or was it radio and ionospheric work?

Mills

Oh, I think I was too young and obscure in those days to know the basis of the meeting. You see, it would have been decided a few years before to hold it in Australia. But I think the ionospheric work would have weighed heavily, too, because Australia did have a pretty strong ionospheric school in those days. I think both together.

Sullivan

Do you have any recollections of that meeting as to the sort of atmosphere was there? It must have been an exciting time. Things were popping all over.

Mills

Yes. It was exciting. Oh, I have quite a few odd flashbacks here and there. I don’t know that they’d be of any great use. I remember discussions with Graham Smith and Hanbury Brown about some of their work. About some of the ionospheric details, ionospheric fluctuations and how they worked. I remember Hanbury visiting our angular size interferometer, radio link interferometer, having to be convinced that it would work all right, give the effects of very large separations between the antennas and the delays which had to be corrected for, and things like this. It was a very pleasant time. The main thing that I got out of it, I think, was meeting, for the first time, a lot of these people. Of course, we’d been very isolated in Australia and corresponded with a few of them, but not actually met them before.

Sullivan

Well, that’s an important point, I think. You’ve mentioned a couple of people like Pawsey and Bracewell that had mixed, but most of the Australians had not travelled. I guess Frank Kerr is another case of one who had gone to the outside world, so to speak. He went to Harvard in ’51 for a year. He was actually there when the hydrogen line was discovered.

Mills

I’d forgotten that.

Sullivan

Would you say, that this, perhaps, contributed to Australian radio astronomy developing rather independently? If it had been a European country, it might have - ?

Mills

Yes, I think it did. We did have, I think, quite an independent outlook really. We had no famous names to tell us what we should believe, and we just went ahead following our noses to some extent. And we did have quite a good technique. This was the main thing, it got things going. Of course this is a pretty obvious statement. Lots of people have made it, I think. Our radio techniques were well up to anyone else, because our radar as a result of the War was an international subject in which practically all countries, were not exactly equivalent, were at least moderately so.

Sullivan

Right, although radio astronomy did not take off in the U.S., where there was as much technique and much more money.

Mills

Oh, it could have.

Sullivan

It could have, right. Do you have any ideas as to why it didn’t?

Mills

I think you’ll have to ask the Americans (laughter). I have no idea. I know why it did in Australia; that was because we had the Radio Physics Laboratory which had now its raison d’être was finished. It was no longer required for radar, at least for war-time radar. We had to find something to do in a hurry. I think these reasons led Dr. Bowen, who was then head, to press strongly for radio astronomy once Pawsey had got it going. Because this was a field – you see, we might have got sucked into doing standards for television or something horrible like that. We had a lab which was well set up for doing radio work of all kinds, and it was just a matter of trying all the things which we could. You see, this linear accelerator I mentioned earlier was one of the things we had a go at as being a possible good line of development. Again, Bowen pushed this very hard. And it was only when it was realized that the expense would be too much for us that we dropped it.

Sullivan

Although, of course, you had the Radiation Lab at MIT. If that had kept going I suppose -

Mills

Yes.

Sullivan

I suppose it was some governmental decision that this would – it was disbanded incredible quickly. Maybe that could be blamed in some part.

Mills

I think that probably did have a large part to do with America not getting onto the astronomy so quickly.

Sullivan

Let me pick you up on that comment about you didn’t have experts to tell you what to think. Because it seems to me that there were as many experts in radio astronomy in Australia as there were in England. Is this not true?

Mills

Oh no, no. I was talking about experts in astrophysics and astronomy.

Sullivan

Oh, in astronomy. Well, who exactly are you referring to in astronomy?

Mills

I think the classical astronomers generally had rather tended to hold back radio astronomy. And this is what I had in mind.

Sullivan

Oh, I see.

Mills

Because the general idea was that they knew what was in the sky and you shouldn’t try and look for things which are not there.

Sullivan

Right, and if you can only tell me where they are within a degree, don’t bother me.

Mills

Yes. It’s not worth their doing it. Though that was a very common outlook. But we were less afflicted by it than many other groups, I think.

Sullivan

Although, obviously some groups just didn’t let this bother them.

Mills

Oh yes. Well, they were the groups that forged ahead.

Sullivan

Well, let me ask about the whole radio link technique, as I think you were the first to develop this, were you not, in -

Mills

I think so, yes.

Sullivan

Interferometry – can you tell me how this came about?

Mills

Following on our measurements on Cygnus, which were done with a couple of Yagis, or rather three Yagis now that I come to think of it, we did have three. There was one with a short spacing for resolving ambiguities. We decided we had to have larger area and so I constructed the next system. At this time Thomas, who had been working with me, left for England. Arthur Watkinson, as a technician, came in to help me on this. We got some land quite a long way away. We had to abandon the field station we had because it wasn’t large enough and there was too much interference. So we went out to Badgery’s Creek, which is a little bit further than our eventual Fleurs Station which we still have now.

Sullivan

What was the name of the first one you were at?

Mills

Potts Hill Reservoir. This was about 30 miles west of Sydney – the nearest we could find. And the reason we chose it, basically, was that it was on CSIRO property, on one of the experimental farms. So we set up there with a much larger, but still by modern standards quite small, antenna system – a system with a few arrays of dipoles – and a fairly complex calibration system, which again followed on the technique we had been using before, which had originally been developed by Payne-Scott and Little. We did our surveys there plus position measurements. But while we were doing this we found these larger angular size sources and the question arose naturally that there’s undoubtedly (?) distribution of sizes, and we knew of possible identifications and we thought things several minutes of arc would be the sort of sizes we’d be interested in. A “minute of arc up” sort of thing. There was this debate about whether they were radio stars or radio galaxies, and quite obviously the way to determine this was to build an interferometer which could distinguish whether things were smaller or greater than about a minute of arc. This was the sort of thing I was going to do. Of course, when you look at the spacing – we were still thinking in terms of low frequencies which was all we could use in those days because of the sensitivity questions – obviously you needed spacings up to about 10 km.

Sullivan

You were becoming aware of the steep spectrum, I suppose, at this time?

Mills

Yes. We had already – Bolton and Slee had done some measurements on spectra. Well, although they disagreed again with the Cambridge one, I think, they showed a steep slope and we believed those. So it was clear one had to keep at low frequencies and you had to go to a very large spacing. And it was equally clear we couldn’t afford the cable to do it. I was quite familiar with the techniques necessary to make a radio link, so I just looked into the problems and decided that the worst problem was that of getting a stable compensating delay for the long path links to the distant antenna. There my experience in digital computing was very useful because our computer, following on the original British design, had mercury delay lines for store. So I just borrowed a couple of these delay lines and used them as a variable delay for compensation. So everything sort of fitted together on that one.

Sullivan

Were you using these computers at all in your reduction of data?

Mills

Oh, no. They were purely experimental. They’re very slow by modern techniques.

Sullivan

But still faster than your hand.

Mills

Oh, very much faster, yes.

Sullivan

So, do I have it right then? It was just a matter of expense, really, that made you go to the microwave link? Not a matter of too much attenuation over such a long distance?

Mills

Well, it was also the practical convenience of it. If we wanted to try different spacings and different places, then obviously you couldn’t go coiling and uncoiling miles of cables. The radio link was the obvious way of doing it. As I said, I knew the techniques were adequate and the only real thing I was worried about was getting this compensating delay which I realized was necessary. Even that wasn’t too certain in those days, because I remember looking at how narrow the bandwidth had to be in order to get fringes over the spacing, and decided just too narrow. So it occurred to me to put in this delay.

Sullivan

These sorts of things were not written in textbooks?

Mills

No, never. Or in papers. It had to be thought out right from scratch.

Sullivan

So what was the first thing that was done with this radio link interferometer?

Mills

Our first measurement was a very short baseline just to see if it would work. On all the stronger sources we got it going on a short baseline and we stepped out, first, I think to a kilometer, then, I think, to 5 km., noting that three stronger ones, the Crab, M87, and [NGC] 5128 had all dropped off. They were dropping off at 1. They were completely invisible at 5. I’m not sure how much they’d dropped off at 1m., I’d have to look at the paper to remember that. And then at 5 km., Cygnus A dropped off quite a bit, whereas at 1 km there’d been no drop off. So, it had a constant visibility to 1 km. – these are just orders of magnitude, of course on Cygnus A – and then a big drop by the time we got to 5 km., and by then the others had been dropping off much more rapidly. So that was it.

Sullivan

What were you using to calibrate your fringe amplitude?

Mills

We injected noise, if I remember correctly. I can’t remember the exact calibration technique used in the first system.

Sullivan

It seems to me this would be a problem.

Mills

Yes. Well, yes and no. You see, we were not after an accurate measurement.

Sullivan

Well you still want to know whether it was .7 or .9 or -

Mills

No. that wasn’t necessary at all. What you want to do is take it out until it drops off to zero or drops off a long way. And that gives you the size. Now, we had the original Michelson concept on this. It was a matter of fringe visibility and if one waits until it goes right down, then it doesn’t matter very much about the calibration. I know we were not worried about it and I don’t think we had anything very elaborate.

Sullivan

Ok. And that was the paper – the first thing that was published on it was one of those three papers in Nature in ’52?

Mills

Yes.

Sullivan

And the more accurate, or the more complete, work was also in ’52 in the Australian Journal of Scientific Research, I suspect that’s right.

Mills

No, it was ’53 when the main radio link one came out, I think.

Sullivan

Six accurate positions, was that from the survey?

Mills

That was from the survey.

Sullivan

Ok, I’m confused then. Well, to keep things in chronological order here, let’s go to your first survey then.

Mills

Yes, well that, in fact, had been done before the angular size work. That was done around about 1950.

Sullivan

And published about the same time as the first?

Mills

Yes. You see, that first angular size work wouldn’t have come out at all at that time but for the stimulation of the ’52 URSI meeting and our decision to publish all our results together. And they were very preliminary. It would have been this ’53 one alone which we published otherwise.

Sullivan

Let’s go back to the survey with the Michelson interferometer, which was the first southern survey, basically. Well, is that right?

Mills

No, Bolton and Slee.

Sullivan

Bolton and Slee, that’s right. They’d done – they had how many sources roughly – 22 or something like that?

Mills

I can’t remember the number.

Sullivan

But you had many more anyway – 77.

Mills

Yes, that’s right.

Sullivan

Well, I guess the first thing to ask is what was the difference between your two surveys?

Mills

There was very little contact between Bolton’s group and mine. We each felt rather strongly our own technique was the best. Although we saw each other sometimes, Bolton lived out at Dover Heights and didn’t come into the Lab very often and I spent most of my time out at Badgery’s Creek. So we didn’t have very much contact actually. And there were quite a few arguments about interpretation of the results and things like that.

Sullivan

I see. And Pawsey was quite willing to let both these go on and see?

Mills

Oh, yes. He thought it was a very healthy situation, I think. So this was quite independent of anything Bolton had done. I can’t even remember when he did his and when he published it.

Sullivan

It was about the same time.

Mills

Yes.

Sullivan

So it was about the same frequency, 100 MHz?

Mills

Yes, I think about the same frequency.

Sullivan

And neither of you did – I can’t remember if you did in print anyway – a comparison of these surveys?

Mills

No. He may have. I know we had written ours up well before he had written his up. At least I think we did. At least before his were published. The survey results were prepared independently, basically.

Sullivan

Ok. Well maybe I’ll have to do that comparison.

Mills

The thing which we did agree on was the fact that these sources had angular sizes. We in fact found this first at Badgery’s Creek, [NGC] 5128, and then Bolton found the same sort of thing on his sea interferometer and we agreed at least on that part of it. My memory is that we did our survey and got our results out and largely written up before Bolton had got back onto his. That could be wrong, I wouldn’t stick by it.

Sullivan

Now what were the main results of this survey?

Mills

Two things: the positions, which I’ll talk about later, and the actual results of the survey itself. And there, I think, the important things were that we recognized that there must be two classes of source because of the strong sources being distributed along the plane of the galaxy and the weak ones isotropically. Now at that time we couldn’t be sure that the isotropic sources were extragalactic, but we felt pretty confident that they were. And I think if you read the paper on that survey that I wrote, you’ll find that I was being rather careful not to commit myself completely to one or the other. I was leaning towards the other because of the fact of these identifications. I think that was the important thing that came out of the survey itself. And I know it was debated quite a lot in England by people who had gone overseas – Pawsey was often touring around, for instance. At that time the extragalactic-galactic argument was building up, getting quite strong and heated at times.

Sullivan

Now these were the class 1 and class 2 sources that you refer to?

Mills

Yes.

Sullivan

These two distributions. This terminology, I think, was used throughout the ‘50’s, wasn’t it?

Mills

Yes, it was.

Sullivan

Now the 1C survey came out a bit before yours? You probably saw that.

Mills

Ah yes, that was one of the early ones -

Sullivan

It was published in 1950.

Mills

Was it? We may have had the results a bit earlier – I think we did actually. I know that about the time I was working on this, there was the Ryle survey – Ryle and Elsmore, I think it was.

Sullivan

That’s the one, right.

Mills

Yes. And there was Bolton and Slee’s initial survey – both were available.

Sullivan

But now, Ryle, Smith and Elsmore, I think it was, came out with an isotropic distribution. Why did they not come up with - ?

Mills

Well, that actually is one of the sources of our whole difference in outlook. They had missed the strong sources near the plane because they were using wide-spaced interferometers. So therefore, they divided the sky into a slowly varying component and point sources. Whereas we, by having our close spacing, were able to see these -

End Tape 61B

Click start to listen to the audio for Tape 62A of the 1976 interview.

Begin Tape 62A

Sullivan

This is continuing with Mills on 26 August 1976. So you were saying that having the small spacings you were able to pick up -

Mills

Yes. It was obvious that along the galactic plane there was a lot of complex structure of quite small scale, but obviously not as small as would be picked up by the wide-spaced interferometer which had a resolution of a fraction of a degree. This was the basic cause of our arguments at that time with Cambridge. In fact, I understand that the Cambridge people denied the existence of extended sources along the galactic plane – you should talk to the Jodrell Bank people about this probably. They know more about it than I do. Anyhow, that’s the situation.

Sullivan

So you’re saying that even before the debate was most intense, namely the Mills Cross survey vs. the Cambridge survey – even before the Mills Cross, that there was this difference in philosophy on extended sources?

Mills

There was an argument then on completely different grounds, because the Cambridge people then believed that the majority of the isotropic sources which they’d found were radio stars at that time, and that there were no other strong sources along the plane. We were very open-minded about this and believed rather that the isotropic ones were extragalactic and that there was a very strong component of extended sources along the plane. And this was a considerable source of debate. And a completely different subject, though from the next one.

Sullivan

Well, the point I’m trying to make is, do you think this is correct, that in both debates that you were concerned with the short spacings, whereas, Cambridge was neglecting them essentially. So you got more extended -

Mills

Yes. I think that’s right. This survey was actually the basis for the Cross because – I realized that it was necessary in any survey to have an instrument which would respond to close spacings and large angular size structure. Otherwise, one would simple miss it and miss a lot of the information available in the sky. And it was as a result of this survey that I thought of the Cross as being the sort of thing on must use. One must use pencil beams for survey. That was the basic idea I had in mind.

Sullivan

We won’t have time to get into the Cross today, but let me finish up with this paper which was an extension of your survey in which you had six accurate positions in ’52. You confirmed the three I.D.’s that had been suggested a few years before, and also, had Centaurus and Fornax as being the first large angular sized objects, I think, that were detected. Were there any surprises in this? Were you surprised to find these half-degree size things?

Mills

Well, in a sense, as soon as we switched on our close spaced interferometer, we realized we had them. It was the first thing that hit you in the eye. We’d look at the same part of the sky on a close spacing and we’d see an enormous interference pattern and in the wide spacing a small one. You couldn’t help seeing it.

Sullivan

But how could Centaurus be something that’s a half-degree or degree, and you’re trying to identify it with this one arc minute Galaxy?

Mills

No, no, Centaurus A, the visible Galaxy, was about 6 or 7 minutes of arc. And we did know that things extended out beyond it. Although we didn’t know that -

Sullivan

Optical, you mean?

Mills

Optically, yes. I think at that time de Vaucouleurs was in Sydney and we had discussed with him quite a lot this question of optical extensions. What I didn’t know then was, in fact, that this 6 or 7 degrees, or even more across – these two big blobs on either side – existed. Our interferometer only gave about 20 minutes of arc equivalent size.

Sullivan

So, you were still lacking the shortest spacings that were needed?

Mills

Yes, I was.

Sullivan

So it wasn’t a matter of sensitivity. It was a matter of missing –

Mills

Missing Fourier components. Yes.

Sullivan

So then from this, there still was no conflict with the problem, which is still with us today. Namely, why is the radio emission out here when the optical object is here?

Mills

No, it wasn’t at that time.

Sullivan

That really didn’t come, I guess, until would you say, Jennison and Das Gupta’s double thing in Cygnus was really the thing?

Mills

Yes, that was probably the first time that that became obvious.

Sullivan

That was ’53 in Nature.

Mills

No, I think, before that it was clear that the emission in Cygnus A must extend quite a lot longer than the sort of 1 minute of arc which I had obtained.

Sullivan

Well, that’s true. But then, of course, you might expect at least that it would be concentrated on the optical object.

Mills

Yes. Jennison and Das Gupta, as far as I can remember, was the first indication that you could have a main concentration of emission way out from the Galaxy.

Sullivan

Although I was talking with someone, I can’t quite remember who now, who said even though that result was around from ’53 on, there were a lot of people who just did not believe it. Were you one of these?

Mills

No. Around about that time we realized that this sort of thing was observationally reasonable and it did fit my results, too. All I had done was to miss the zero point.

Sullivan

Those missing spacings you were talking about - ?

Mills

Yes, the zero point in the middle. I myself was quite happy with it and saw no reason to doubt it whatsoever.

Sullivan

But you weren’t particularly worried about radio physics of galaxies?

Mills

Oh, no. Well it was obviously exceedingly interesting as to how this happened. I can’t exactly remember what time I began really to understand what was happening. I think it was about that time. My interests were first, on the theoretical side, stimulated by Fermi’s paper on his accelerating mechanism of cosmic rays, and I had that in the back of my mind since probably before 1950 when I read that. But it wasn’t until Shklovsky’s paper came out that everything made sense and fell into place.

Sullivan

Yeah, there’s one of your later papers that talks about abnormal galaxies as radio sources, Observatory ’54. And you say that 5128 probably is not colliding galaxies, or it seems unreasonable. About this time then, you must have been thinking about the physics.

Mills

Yeah.

Sullivan

This is a good place to break, I think. We’ll have to continue. That finishes the first session with Bernie Mills on 26 August ’76.

End Tape 62A, end Part 1

Part 2

Modified on [an error occurred while processing this directive] by Ellen Bouton, Archivist (Questions or feedback)