[Bracewell touching up signature pier on spectroheliograph dish]
Bracewell touching up signature pier on spectroheliograph dish. (FBOA image)


[Bracewell, 21 November 2004]
Bracewell writing, 21 Nov. 2004. (FBOA image)


[Bracewell, Swarup, and Pawsey, 1958]
Bracewell and Govind Swarup watch Joe Pawsey carve his name on telescope pier, 1958. (NRAO/AUI image)


[Bracewell circa 1955]
Bracewell, circa 1955. (Stanford University image)


NATIONAL RADIO ASTRONOMY OBSERVATORY ARCHIVES

Papers of Woodruff T. Sullivan III: Tapes Series

Interview with Ronald N. Bracewell
At AAAS Meeting, San Francisco CA
8 January 1980
Interview time: 185 minutes
Transcribed by Sierra Smith

Note: The interview listed below was conducted as part of Sullivan's research for his book, Cosmic Noise: A History of Early Radio Astronomy (Cambridge University Press, 2009) and was transcribed for the NRAO Archives by Sierra Smith in 2014. The transcript was reviewed and edited/corrected by Ellen N. Bouton and Kenneth I. Kellermann. Any notes of correction or clarification added in the 2014 reviewing/editing process have been included in brackets; places where we are uncertain about what was said are indicated with parentheses and a question mark, e.g. (?) or (possible text?). Sullivan's notes about each interview are available on Sullivan's interviewee Web page. During processing, full names of institutions and people were added in brackets and if especially long the interview was split into parts reflecting the sides of the original audio cassette tapes. We are grateful for the 2011 Herbert C. Pollock Award from Dudley Observatory which funded digitization of the original cassette tapes, and for support from Associated Universities, Inc., which funded transcription of this interview.

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.

Part 1 | Part 3

Begin tape 131B

Sullivan

Continuing with Ron Bracewell on 8 January 1980. Well, in í52 when by necessity you had to change your course, you told me that Pawsey asked you to take on administrative tasks, basically.

Bracewell

Yes, to organize the 10th General Assembly of URSI, as organizing Secretary. The Committee in charge consisted of Bowen, Pawsey, D.F. Martyn, and Sir John Madsen as chairman, and maybe one other person. They would hold regular meetings well ahead of time and decide what was to be done and I had to execute it. And they said to me, ďYou have all our resources. Just tell us what you want done but you do it.Ē And it worked out that way. I had appointed dozens of committees and was at the head of a huge pyramid. That was a very interesting experience, it used up about six months. I designed the URSI flag, which is still flown at General Assemblies to this day and was first flown in Sydney in 1952. We invented the colors, which are gold and blue, and the flag was made by a committee I appointed.

Sullivan

Was that where the diamond shape came from?

Bracewell

No, no. The flag doesnít have that. And it was first hoisted by [Edward] Appleton outside the Peter Nickel Russell School of Engineering in 1952. And there is a photo of it being hoisted. So that was very exciting because all of a sudden the names that had become familiar to us from papers from many countries suddenly materialized as faces. And we saw people from France. Jean-Louis Steinberg was there and later to become head of the radio astronomy at Nançay and then to move on to head of their space physics. And [Marius] Laffineur, who was the chief moving force at that time.

Sullivan

The Reber of France, I think of him as.

Bracewell

Reber of France, yes. And we met a number of Americans such as John Hagen and Ė I donít believe [Fred T.] Haddock was there. Two other Americans were Larry Manning and Bob Helliwell, who moved on to capitalize on the discovery of whistlers and convert that into a whole new field of science, which has been amazing. And assorted people from strange countries in small numbers and then a contingent from England of the figures that are still well known. So it was a very interesting meeting. There were no parallel sessions that I can recall. Everyone took in everything and there was a very nice central assembly room with free coffee. So I think it was a great success and was instrumental in lubricating contact between the major groups.

Sullivan

Did you write more freely to other groups, do you think, after that?

Bracewell

I donít think there is any doubt. It helped enormously. Graham Smith was there. Ratcliffe was there. Ryle didnít come.

Sullivan

Did it change what you and others decided to do at Radiophysics, which experiments to do?

Bracewell

It probably meant that considerations were taken into account that we were previously ignorant of.

Sullivan

What was your next task after the URSI?

Bracewell

Pawsey wanted to write a book. And he said, ďThe time is ripe. We have reached the point where there is a kind of plateau and we are re-gathering our forces. Lots of discoveries have been made in several fields, and the next round of discoveries will require bigger instruments, and theyíll be breathing space while this happens. So letís fill in the time writing a book. Now is the right moment.Ē Well, of course, it took two years to write. And I donít know whether there ever was a plateau, but still, that was that. Now he didnít want to pull the leaders of the principal groups off their work for that purpose such as [J. Paul] Wild or Mills or [Wilbur Norman ďChrisĒ] Christiansen, who would have been perfectly rational choices in terms of their writing ability. I see now looking back on it he had a slightly delicate job because there may have been some feeling of resentment that a person like myself who had been working mainly in the ionosphere should now be working in radio astronomy. But the logic of it was quite strong. My work in the ionospheric research was running down and a six month gap had intervened and Iíd lost my momentum, as well as losing a lot of my observations. And so it made sense and I enjoy writing. So Pawsey sold that choice to the rest of the group without any apparent ripple and we worked for about two years assembling chapter after chapter and revising, and revising, and sending these chapters around the lab for someone or other to read who would be an appropriate choice.

Sullivan

And did little else for those two years?

Bracewell

I essentially did nothing else. No, thatís exactly right.

Sullivan

This obviously gave you a chance to really review the status of radio astronomy.

Bracewell

Well, I read everything that had been written. I already read everything that had ever been written on the Sun before I began that. I had read all the journals that published solar observations. In fact, I read several journals backwards starting from the present and just read them backwards into time. So I was thoroughly familiar with that side but I had to learn a lot of other subjects. And so I was reading voraciously and at the time writing. You may recall that in that book there are some introductory chapters which purport to give the status of the subject at the time. There would be some astronomical background and that thing. And in some cases that material was rather hard to come by. We didnít have any one to consult. On interstellar matter for instance I donít believe there was anybody in Australia who could tell you about interstellar matter who had worked in to or knew about it. So I found myself reading some strange things. Fortunately I could read some other European languages so that I wasnít particularly held up by strange references. But thatís mostly what I did.

Sullivan

One thing along that same line is that Iíve been struck that there is no mention of the synchrotron mechanism in your book which speaks to the fact that it was just not being considered from your point of view or probably in world radio astronomy as anything of importance; even though by late í53 or mid í53, whenever you sent the final version of the book off, there had been many books published in Russia.

Bracewell

Well I was totally ignorant of anything that was happening in Russia. I donít recall that we had any Russian periodicals around, except maybe some standard things in physics. But I donít even recall that.

Sullivan

It could very well be.

Bracewell

Iím not entirely sure of that but I certainly donít recall it. I was studying Russian, and so had there been material I needed to read I would just have incorporated that into my interest in studying Russian. But I was quite unaware of anything happening in the Russian language. Except for we did know about an expedition that the Russians had sent to Brazil. But that was easy to find because it was of a radio astronomical context. But anything that might have been written connected with plasma physics, Iím not sure how we would have keyed into it, and as we heard [Jan Hendrik] Oort brought it to the attention of other Europeans.

Sullivan

Well, this may have been part of the motivation to start the whole Radiophysics Abstracts business, which did consider the Russian literature once they got started. They get started in í53 or í54.

Bracewell

Well I donít when it began. I know that we had the Cornell Abstracts, which Martha Stahr Carpenter was preparing very ably. The Radiophysics Abstracts were appearing simultaneously with her for some period of time but I forget the details.

Sullivan

Well í54 is when the Radiophysics Abstracts got going in a big way. There was some earlierÖ

Bracewell

Well, she ran into troubles and then Radiophysics decided on the basis of some international symposium that they would take over that duty.

Sullivan

No, it went on for quite a while in parallel. Do you have any particular thoughts on radio astronomy at that time, 1953, since you did pause and review the whole deal? What stage was it at? Where was it going?

Bracewell

Let me just make another comment about synchrotron radiation. Pawsey and I both knew about magneto-ionic theory so gyro-resonance or cyclotron-resonance or anything that was connected with magneto-ionic theory was something we could conceive of. Now we also knew about accelerators, now that I come to think of it. As a matter of fact, Mills had built a linear accelerator and he did that in the same room as myself and we used to wonder whether weíd get cancer and things like that. And Pulley [Owen O. Pulley], who you may not have run across as a figure but was one of the contemporaries of Pawsey, who was a significant figure at Radiophysics, Owen Pulley, he knew about accelerators. But Iím not aware of ever thinking about synchrotron emission. We may have known about synchrotrons and betatrons. But synchrotron emission as it is now expounded, in a very elementary way, of course, I have just no recollection of it.

Sullivan

It had been suggested in 1950 by [Hannes] Alfven and [Nicolai] Herlofson and by [Karl Otto] Kiepenheuer in Phys Rev. but those articles were largely ignored except by the Soviets. [Note added 2014: Phys. Rev. 78, 616, 1950; Phys. Rev. 79, 738, 1950]

Bracewell

Yes. Well, we had different sorts of responsibilities. Pawsey had a global responsibility for determining which would be productive lines to go in. Now, generally speaking that meant, he would go tearing off on some observational project involving the development of instruments. Theoreticians were not very highly regarded around Radiophysics. People did it by their own inclinations but they had to be pretty good to cut much ice with Pawsey. They had to come up with something really brilliant that he could understand or he wasnít terribly excited. So we had plenty of amateur theorists. Now, people generally speaking had very specific duties. The work was carved up among the available people. We would discuss who should do what, what was within the range of what could be done, and the timing with which results might come out. And then we would bury ourselves for months at a time, wiring things up, you know, stringing antennas up and doing that. So when you have a hierarchy of that kind the underlings are not likely to think in a global fashion. There would be one or two people doing that and then one or two others who just think of something original anyhow. Now Pawsey was very amenable to taking off in any direction that was profitable if you could convince him.

Sullivan

No, I see what you are saying. But what about back to an overview of radio astronomy?

Bracewell

Well, it was a hodgepodge thing. There must have been six or seven totally unrelated subjects. The Sun had nothing to do with the galactic radiation. And the discrete sources had very little to do with the galactic. And then there were a number of other things which were important for a time. Lunar radiation was quite important, and the study of meteors. Then the hydrogen line came along. All of these things were totally unrelated then. So when in 1954 I went to Berkeley to spend a year there giving lectures, I was able to lecture for a straight year on an incredible mixture of things. I didnít have to stop onceÖ

Sullivan

It was just radio astronomy?

Bracewell

It was radio astronomy and nothing else.

Sullivan

And laterÖ

Bracewell

Well, I wouldnítÖ

Sullivan

The meteorÖ

Bracewell

I may have mentioned meteors. I forget. I didnít know anything about that except the theory of reflection which I thought was sort of beautiful. So it was all disconnected and, of course, disconnected from optical astronomy. We knew solar astronomers.

Sullivan

At Radiophysics, you mean?

Bracewell

Yeah, the people at Radiophysics were in contact with Mount Stromlo, with people who observed the Sun. I spent several weeks myself peering through a spectroheliograph. So I found that congenial.

Sullivan

Who would be the main people at Stromlo who were interested in this kind of contact?

Bracewell

Well, David Martyn was there. He had been chief of Radiophysics and Cla Allen who wrote the famous Astrophysical Constants. The other important figure there whose name is not coming back to me for the moment but who stammers was there also [Note added 2014: S.C.B. ďBenĒ Gascoigne]. There was another man whose name Iíve forgotten. Anyhow there was no lack of contact.

Sullivan

They were all interested in the radio stuff? Didnít look upon it as being a bunch of engineers who donít know anything about astronomy?

Bracewell

I donít think it was anything like that. You see, we werenít engineers, to a large extent. Those people were physicists and they regarded themselves as physicists. I had a Ph.D. in physics and my passport said ďphysicist.Ē Matter of fact, I think it still does. So we did indeed have people who were engineers and regarded themselves as engineers but the pecking order definitely had the physicists running the place. Most of the people had degrees in physics.

Sullivan

This strikes me. This is a difference with the U.S. where the early radio astronomers were pretty much all engineers and therefore it may have been a bit harder to bridge that optical/radio gap in the U.S. Iíll have to think about that.

Bracewell

Iím only referring to, well at the moment, to solar but now that Iíve come to think of it the connections with the Magellanic Clouds were quite good too. And that one time you know [Gerard] de Vaucouleurs lived in Stromlo for a time and he was frequently seen at Radiophysics with his beautiful wife. And I donít think there were any great difficulties but there were not an enormous amount of optical astronomy taking place in Australia. There were no great telescopes there at the time. And the quantity of staff was rather small. The Observatory at Stromlo for instance was in charge of time keeping for Australia. That would have been a quite noticeable fraction of the whole set up.

Sullivan

So having finished the book, you say you went to California.

Bracewell

Yeah, [Otto] Struve arrived to see what all this was going on in Australia and he asked Pawsey to send someone to give lectures for a year. As a matter of fact, he invited Pawsey to come and give lectures for a year. Now Pawsey didnít wish to do that and I was an obvious choice because by that point I knew more radio astronomy than any single individual except for Pawsey. I knew all the other subjects, you see. We had a bunch of experts and I was a generalist at that point. So I was the obvious choice. So thatís how I came to set foot in America.

Sullivan

I see. And did you while here visit any of the radio astronomy groups undoubtedly at NRL and letís see what else in í55, not much.

Bracewell

Iím pretty sure I went to the Carnegie Institution and met [Merle] Tuve and [Kenneth L.] Franklin and whatís his name?

Sullivan

Bernie Burke?

Bracewell

Bernie Burke. Donít tell him I couldnít think of his name. And John Firor. I remember all those people. And Howard Tatel. And they were working busily on the hydrogen line and assorted other things including of course Jupiter. But well, Iím not sure exactly when I went there but it was either in 1955 or í56 because I was back again in í56, you see. So I definitely went to NRL at a very early point.

Sullivan

What was your impression of NRL, the work they were doing?

Bracewell

Technically that was a pretty high caliber. First of all the 50 foot dish was a great tour de force, which was very impressive. Not appreciated by everybody but it is a very impressive thing. And the electronic techniques that they could muster there were of the highest caliber and the many great experts. Just seeing them making a matched load was pretty impressive. I understood all these things from having worked with microwave equipment. When I saw how they made matched loads, which I thought I knew an awful lot about, I saw they knew a thing or two. So the level of technology there was very high. And they were also doing something right in respect to their science. Although there has been a certain amount of turbulence there and complaints from the workers about the management, mostly Haddock complaining about Hagen, they were doing something right. We could get a whole long list of bungles that were made but that was a very productive operation.

Sullivan

Iíd be interested on your opinion. I imagine it wasnít the Cavendish style of doing things.

Bracewell

That was nothing like the way that the Cavendish did things.

Sullivan

Because there was much more money available I suspect.

Bracewell

Well there were resources and facilities. If you wanted something manufactured you could have it drawn up by a draftsman who knew what he was doing and could design things as he drew. And then you could have it put into the machine shop and someone who really knew how to fabricate things would make it with some years of experience because you had continuity of teams that had worked together for some years. Thatís a tremendous resource and that was never broken up.

Sullivan

Thatís true at Radiophysics too.

Bracewell

That was also true at Radiophysics. The atmosphere at NRL was not unlike what I was accustomed to at Radiophysics. So they could really move and on sizable projects, with all the backup existing. Now the Cambridge situation was quite different and Ryle operated in a completely different way. He would select out one option among those he could choose from which would require technology in favor of one where he could get to the result more directly. He was following the standard Cavendish tradition there, which I learnt a lot about. Of course, I already had that before I got there because I got it from Pawsey. So I appreciated both those ways of going about things and you take your choice, you think out the whole problem before you move.

Sullivan

Were there other places in the US that I missed that you visited? Iíd be interested in your impressions.

Bracewell

Well, I went to Ann Arbor, Pontiac [Michigan] in 1955, about July or August. When the school year finished at Berkeley, I went traveling around and I went to Pontiac. And I stayed there overnight in a hotel and I got there by bus from Columbus, where I had just been to Indiana and see, Frank, whatís his nameÖ?

Sullivan

Edmundson

Bracewell

Edmundson, who somehow I had gotten to know at Berkeley. And while I was sitting in my hotel room in Pontiac, having come by bus, I heard on the radio advertising that you could get a used car for $50. So I found that it was on the same street as the hotel and I had $50. I walked out and it was the middle of summer so it was still quite light. I walked up the street and sure enough hereís this place. I went in and said, "I came in to see about those $50 cars." The man said, "What $50 cars?" I said, "I just heard about it on the radio." He said, "We sold it, but weíve got one for $75." So I bought a Pontiac in Pontiac. And then the next morning I went to see the solar installations, McMath-Hulburt installation. I was already acquainted with Helen Dodson by having written to her many times in connection with the sudden phase anomalies Iíd observed in Cambridge. I was delighted to meet her and Ruth Hedeman. And [Robert] McMath was there and he was treated as a very important man. Orren Mohler was also there. So I was ushered into McMathís office. I felt as though I was visited General MacArthur. It was quite an eerie atmosphere of very big boss. Well, I found him a very impressive fellow, and he asked me a few questions and then he said, "Are you going to Ann Arbor." And I said, "No, Iím going to Annapolis." And it was in the opposite direction, you understand. So he said, "I advise you to go to Ann Arbor and see Leo Goldberg. It might be to your advantage." Very cryptic statement, very brief, "It might be to your advantage." Being an obedient person, by golly I did. And it wasnít to my advantage at all except for the pleasure of meeting Leo Goldberg for the first time. But it turned out that Goldberg was looking for an appointee to a job, which was given to Fred Haddock. That enabled Fred to escape from NRL and led to the development of very significant radio astronomy at the University of Michigan.

Sullivan

Did you see John Boltonís operation in this first year or so when he came.

Bracewell

Yes, I visited Palomar and he had set up an antenna there. And I arrived within a few hours of him making his first transit observation of some source through his antenna, 30 foot antenna [Note added 2014: actually 32 foot], I would say roughly. And the exciting news was that the beam width was narrower than theoretical. As a matter of fact, I have a copy of that record. He must have been so proud of it. I have a copy, but it has no scales or date on it, but I remember what it was.

Sullivan

Now theoretical being λ/D or 0.22λ/D?

Bracewell

With Bolton and [Gordon J.] Stanley you never knew but they always had something surprising to tell you. And the news was, "This is narrower than theoretical."

Sullivan

They told you, I see.

Bracewell

And you would open your mouth in amazement and before you could recover theyíd amaze you with something else. They had their own peculiar kind of one up-manship. And they didnít care whether they damaged their reputation while telling you this.

Sullivan

You havenít mentioned, it occurs to me, Boltonís name in all of this talking about Radiophysics.

Bracewell

Well, Bolton was in Sydney when I arrived back in 1950. He was not there when I left. So he must have arrived about í47 or í48.

Sullivan

It was í47 I think.

Bracewell

So he was working at the station at Dover Heights and had already obtained records of several sources. And all that I can recall now on my first visit there was being in this concrete block house and seeing a chart record pouring off, which was from a receiver of such good sensitivity that the galaxy was continually pushing it off scale. And then you would reset the recorder or it would reset itself automatically with a switch, and switch in a little voltage and go back to the bottom of the chart. And after it had run off scale several times along would come one of these fascinating sources. And this was being done with the antenna above the sea and therefore it would show up as a fringe pattern. And you could see indications of very faint ones, which if you turned the gain up no doubt would also be sources. Now of course as you know the switched interferometer had been thought of in Cambridge to get rid of that embarrassing background. We donít always remember that now, but it was a result of high sensitivity and the galaxy pushing you off scale.

Sullivan

Too much dynamic range.

Bracewell

So they were just working on schemes on suppressing the background. Now Bolton I think worked pretty hard out there. He always liked to dig holes himself and mix concrete. And we didnít see him much in the lab, very rarely in fact. He just used to go straight to his little concrete blockhouse and work on antennas and receivers very hard, putting things together. So I didnít have very much contact with him, certainly during í50 and í51. All I can say is that I didnít see him all that much.

Sullivan

Ok, going back to his comment about narrower than theoretical, this was said to you, of course, because of your work at the timeÖ

Bracewell

Oh, they probably said it to everyone. I donít know how they did the theory.

Sullivan

But that does anyway get us on to your tremendous amount of work in the late Ď50s, primarily on beam sharpening and resolution and strip integration. How did you get off on this general stream to begin with?

Bracewell

I can answer that quite clearly. Pawsey and I had to write an early chapter that explained that different source distributions could look the same when scanned by an antenna. One was acutely aware that a double source, if the components were close enough, would look the same as a single wide source. And by calculation of artificial records you could convince anybody of this beyond a shadow of a doubt. So the question was, "If you make a certain observation, whatís the most you can say about the source." So I embarked on some calculations to see what I could find out about that. I calculated a few empirical cases. And then it suddenly struck me that the antenna is a low-pass filter. Now that is so common place today that we even say that an antenna has a transfer function. We forget that that term was borrowed from filter theory. Now we know that the mathematics is identical but the mind doesnít work that way. A waveform which goes through a filter is something that is a function of time, voltage is a function of time. And the mind doesnít deal with that in the same bin as images, which are two dimensional, spatial things. But nevertheless the analogy is there and all of a sudden we hit on it. And for weeks and weeks we used to tell anyone in the tea room that would listen that what the antenna is doing is being a low-pass filter. Well everybody knew about the pattern of an antenna. That was the only thing we knew about it. So the question is, "What is the filter function? Whatís the filter characteristics, or what we would now call the transfer function?" And we found out that that was the autocorrelation of the aperture distributions. We knew about aperture distributions because of the Fourier relation with the pattern. But nobody knew about filter functions. So we got that into the book, and I think that was very influential in alerting everyone. And it is so clear and well explained that I donít think anybody recalls that it was actually invented at one point. As far as I know, that description in the text book is whereÖ

Sullivan

What you are saying is that your mistake was in not making it erudite and obscure enough in your initial exposition that it was clear it was new.

Bracewell

Well there were lots of things like that in the book. In fact, in the lunar chapter there is new material that was never published in any other way. But itís never referred to; people who read it in a book assume that it was taken from some previous paper.

Sullivan

Thatís right. So it was basically writing the book that got you off on thinking about all these issues.

Bracewell

Yes. In 1954, I was assigned the duty Ė Pawsey was going on a trip abroad and before he did so he had a big organizational meeting and assigned duties to everybody and appointed a committee to be in charge while he was away. And I discovered the other day that I was a member of the radio astronomy committee in his absence but I must have been the youngest of that group because Frank Kerr, who is older than I am, and by quite a margin I suspect, maybe five years, which was almost infinite in those days, he was almost a contemporary of Pawseyís, you see, and he was Pawseyís lieutenant in many circumstances. But on this occasion Pawsey left a written record of who the committee was and which things they were responsible for. Also in that document everybody was assigned some duty and my duty consisted of just four words, "Aerial smoothing, strip integration." That was 1954. In 1954 Iíd already finished the first paper arising out of the book, a paper that was joint with Roberts, and I was beginning to look at two dimensional aerial smoothing and the inversion of strip integration, which had become a serious problem for Christiansenís observations. And the reduction was being carried out by [Govind] Swarup and I used to look over his shoulder and see him laboriously trying to invert strip integration by guessing the answer and then he would compute the strip integrals by hand, adding up numbers and compare that with the scans. And in one or two stages of fiddling, you could do that inversion. Well, the theory turned out to be fascinating and I had written a couple of internal reports. I remember writing an internal report called Abel Transform. It was a gadget, a computer, which took Abel transforms, and it could be made to convert Abel transforms by a simple feedback system. So the idea was you put in the waveform and the output waveform would be the inverse Abel transform. Well, thatís for circular symmetry. It was a kind of curious concept. No one ever built the computer to do that. Thereíd be a few extra quirks to it in order for it to work. But the general strip integration problem turned out to be absolutely fascinating. And I finished writing that paper while I was in Berkeley and it was submitted at the time I got back to Sydney from Berkeley. [Note added 2014: Bracewell, R.N. Strip Integration in Radio Astronomy. Aust. J. Phys., 9, 198, 1956.]

Sullivan

Well one thing that usually gets a bad press I think in your work is the business about sharpening up resolutions and so forth. The general idea is that there was a lot of enthusiasm in the late Ď50s that one could do a lot better than nature really allows and youíre often named as the culprit. Is this a fair thing? Was there a bit too much optimism or people misusing what you were saying?

Bracewell

I donít think thatís an accurate story. The first paper was one by Bolton and [Kevin C.] Westfold in which they proved by matrix algebra that you could invert the aerial smoothing problem and they gave the solution. [Note added 2014: Bolton, J.G.; Westfold, K.C. Galactic Radiation at Radio Frequencies. I. 100 Mc/s. Survey. Austr. J.Sci. Res. A, 3, 19, 1950] Now not everyone could read all this matrix stuff, and Iím quite sure Bolton couldnít, so we may assume Westfold wrote it. And Westfold knew nothing about physics. Now they are all friends of mine so Iím sure they will recognize that those are true statements. Well, Westfold may know something about physics now, but he was a pure mathematician at that point. So here they purported both in words and in symbols to have inverted the problem. Well I found a mistake in that, and so my discoveryÖ

Sullivan

And what error now are you saying? Their thing was in í50 and yoursÖ

Bracewell

That is wrong you see.

Sullivan

When did you publish yours?

Bracewell

In connection with writing the aerial smoothing section of the book.

Sullivan

í53 probably.

Bracewell

Well, it would have been í52 or í53, probably í52. It was probably written quite early on. It was probably í52. And I discovered some frequencies were wiped out entirely. Nobody knew that. People knew there was a smoothing effect but they hadnít thought of it as a filtering effect in terms of spatial frequencies. What I discovered was two things: a band of frequencies is entirely wiped out and another band of frequencies has the relative strength upset. So once you knew that you could see that Bolton and Westfold couldnít possibly be correct because you canít get back things that arenít in the data. But there is one thing open to you and that is to restore the balance of the thing whose balance has been upset. And if you do that you get a solution, which I called the principal solution, and a term which is very generally used now. But we also very soon discovered that the principal solution is not acceptable for many purposes. Very often it oscillates and goes negative for example. But I also discovered that thatís not the only solution, that there are many solutions. Now I never personally sharpened up or restored any observations that I ever took so I donít think you have an accurate report of my role in aerial smoothing. I never sharpen up any observations I took.

Sullivan

Would you say that your papers never gave anyone else any false optimism about restoration? I mean you canít be held responsible forÖ

Bracewell

Well after youíve explained the whole theory, and Iíve just explained it to you in words very clearly, itís very simple you see. There is a band that disappears. There is a band that is upset. You can restore the equilibrium. You can put in invisible distributions. You can get physically unacceptable things. That the whole story. Thatís my contribution to it.

Sullivan

Let me do it this way. As you look at the work that people were doing in the late Ď50s, did you see violence being done to data in the name of Fourier theory or antenna beam restorations, these types of things?

Bracewell

There must have been one or two papers. But look: Hey, Parsons, and Phillips did a sky survey [Note added 2014: Hey, J.S.; Parsons, S.J.; Phillips, J.W. Fluctuations in Cosmic Radiation at Radio-Frequencies. Nature 158, 234, 1946], which they restored, and I donít think there was anything wrong with that restoration. They didnít have the theory of it, so they may have overdone it, but I suspect not because they only put in one correction term. And we know they would have done partial restorations, so they sharpened it up a little bit. Now to my mind that was perfectly alright. Now there may have been one or two papers but I canít recall who. There certainly was an air of optimism around in some circles. I agree that I remember that. In fact there were fanatics who thought you could get the original distribution back. But thatís not my fault, you see, because it had already been claimed that you could. As a matter of fact Hey, Parsons, and Phillips in their text probably implied that their operation, if continued for two or three terms or so, would produce a true distribution. Iím not sure that they do say that but they may very well imply it. It sounds plausible, you see. You find something which, when you smooth it with your antenna, reproduces what you observed. It sounds quite plausible. But the fact that the series converges was not proved by them. I proved that it converges. And furthermore I found out what it converges to, which is not the true distribution. So my contributions were entirely theoretical, and I accept no blame for the false optimism of those utilizing the information provided.

I do have a further thing to add though. I found out another thing in that theory which didnít come out until 1958. And that is that is the optimal restoration when you know the statistics of the noise. Now I believe I discovered that restoration requires a knowledge of the statistics of the noise. That was a new contribution. And the correlation between the noise and the signal, something that nobody had ever thought of, enters into it too. However the computation required to do that sort of thing in 1958 was unreasonable, and as far as I know was never applied. But the situation has now changed where we are in a position to make use of that 1958 result because a computer could do it for us. But I am not aware that anybody has ever made the necessary statistical analysis of their noise or of the correlation betweenÖ

End of Tape 131B

Part 1 | Part 3

Modified on Wednesday, 17-Dec-2014 08:30:54 EST by Ellen Bouton, Archivist (Questions or feedback)