[Martin Ryle, 18 August 1976]
Martin Ryle, 18 August 1976 (Photo from NRAO Archives, Kraus Papers)



NATIONAL RADIO ASTRONOMY OBSERVATORY ARCHIVES

Papers of Woodruff T. Sullivan III: Tapes Series

Interview with Martin Ryle
At Cavendish Laboratory, Cambridge, England
19 August 1976
Interview time: 2 hours 10 minutes
Transcribed by Sierra E. 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 E. Smith in 2015. The transcript was reviewed and edited/corrected by Ellen N. Bouton in 2016. Any notes of correction or clarification added in the 2016 reviewing/editing process have been included in brackets. During processing, full names of institutions and people were added in brackets when they first appear. Places where we are uncertain about what was said are highlighted and indicated with parentheses and question mark, plus a notation of the time on the audio e.g. (? 00:50) or (possible text? 10:32). If researchers are able to suggest correct text, please contact the Archivist. Sullivan's notes about each interview are available on Sullivan's interviewee Web page. We are grateful for the 2011 Herbert C. Pollock Award from Dudley Observatory which funded digitization of Sullivan's original cassette tapes.

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 for part 1 of the 1976 interview.

Part 1 | Part 3

Begin Tape 59B

Sullivan

Continuing with Ryle on 19th August í76. So about the other groups, which other groups...?

Ryle

We used to obviously have quite close contact with Jodrell Bank, meeting at RAS [Royal Astronomical Society] meetings and the like. I suppose we probably met up with the French. I forget when the first sort of international meeting was, probably 1950 or so. And from then on there was quite a lot of contact with the French.

Sullivan

Were you reading their stuff in Comptes Rendus?

Ryle

Yes, I think so. Yes. And Iím sure we were aware of the NRL work. í54 I went there.

Sullivan

Oh, by that time, yes, but Iím thinking of in the Ď40s. There wasnít very much but they did a few things. But it sounds like it was not terribly influencing what you were doing.

Ryle

No, I donít think it did. I think we were all going all with what we thought we could do, and just getting on with it. And thatís the way perhaps science often ought to go.

Sullivan

Rather than worrying about what the other fellow is doing.

Ryle

Yes. Jumping on the latest bandwagon -

Sullivan

What about with optical astronomers? Could you tell me what sort of contact you had with them in those days?

Ryle

Well, we had quite a lot of contact with our own, at that time the Harald Von KlŁber, I think he was here then. But anyway there was a strong active solar group in the Observatories here which is trying to correlate the bursts with sun spot and (? 01:38) activity with certainty.

Sullivan

And who were the people there?

Ryle

Iím not sure when Von KlŁber was there, maybe as early as í48. But certainly there were people - [Frank J.M.] Stratton I suppose was the Director then.

Sullivan

I think he was. Thatís alright. I can check. But did useful information ever come out of these correlations of optical and radio?

Ryle

Well, I donít know how sure the correlation between radio bursts and optical was at that time. They probably did help in that. Because the Sun doesnít always shine in Cambridge, but on the occasion when it did then we used to try to compare the observations.

Sullivan

Now once again relative to optical astronomers in April í49, you gave a talk at an RAS meeting talking about a part of what was to become the 1C survey, namely that you had found 23 radio sources and they were unresolved and so forth, positions of Ī1 degree and so forth. How did optical astronomers receive this? You were talking at their Society. But was it just something that was entirely outside anything they could ever work on?

Ryle

Obviously there was interest, but I donít think it really could have been much more than that simply because if you can only tell them where to look to a degree and it isnít something as obvious as the Sun or the Crab Nebula, then what are you to do about it. One degree is a rather big part of the sky. And, of course, it went on a rather long time like this. They didnít, I think, feel that there was anything much they could do about these funny things and it might go away it they waited. Thatís a little unfair.

Sullivan

But still it was a fantastic amount of energy being produced in these cases...

Ryle

But one didnít know anything about energy because you didnít know the distance. They could be quite local.

Sullivan

But even if they were local stars you had to make them be super suns.

Ryle

Well it had to be much more energy than comes out of the Sun in radio but not out of the Sun in total energy. The actual total energy probably wasnít very severe. The Sun happened to be a particularly feeble source. That was the interpretation I think.

Sullivan

Right. It really wasnít until Cygnus A and so forth...

Ryle

That we got the distances, yes. Even the Australian identification of (M87? 4:37) and the Crab, I think everybody, possibly us as well, werenít all that convinced by the identification. I mean, ok, they fitted within a degree or so, and they were unusual objects, rare objects, but if that had been the case perhaps one ought to have seen some other super novae. Although fairly soon it was evident it was right because everyone got those positions, but even at the first announcement of that I think probably the optical astronomers took that with a pinch of salt.

Sullivan

I see. And you say you did also?

Ryle

I think we did originally, yes.

Sullivan

Even for the Crab? Thatís surprising.

Ryle

Well we didnít know how remarkable the Crab was, not the optical astronomers. Ok, nowadays someone said, probably one of the [Geoffrey R. or E. Margaret] Burbidges, there are two astronomies, one of the Crab and one of everything else. I mean, ok, we know that now because we know a lot about the Crab but at that stage we didnít know anything about it except that it was a little fuzzy thing on a photograph.

Sullivan

Well along this same line can I ask you what did you consider that you were doing in the nature of the science. Did you consider you were doing ionospheric physics extended to outside the ionosphere? Or something that we would call radio astronomy or radio physics?

Ryle

I think radio astronomy. Iím not sure, but I think we invented the term incidentally, and certainly I consulted my father about the propriety of mixing up Greek and Latin terms. In 1950 when I wrote the Physical Society, Reports on Progress in Physics, or whatever it was called, that I think was the time that the term radio astronomy was used in a definitive way. And my father said it was alright to mix up Greek and Latin derived terms.

Sullivan

There are probably a lot of other examples. But in any case you did definitely think you were becoming astronomers essentially?

Ryle

Yes.

Sullivan

But what about this term radio physicists which Iíve heard used quite a bit.

Ryle

Well radio physics is a term which has been used for quite a time, I think, which includes ionospheric physics and then tropospheric propagation physics, all these things. Itís a much broader term. The Australians were in a radio physics laboratory because I mean they converted their radar research group into a thing which was studying all these things, which was definitely concerned with tropospheric propagation, and proper scatter, and all those things which developed out of it, as well as ionosphere.

Sullivan

Something else that you just brought to mind, Iíd be very interested in your opinion on why it was that radio astronomy did not take off in the States like it did here and in Australia. You had the same sort of thing, I mean, tremendous radar push during the war and at the end you had all these trained people and all this surplus equipment.

Ryle

Myself, I believe the difference to be the way that radar development happened. It was, I think, a fantastic occurrence to a young chap to get into radar research in England. Now as I say there were maybe 200 of us at the beginning of the War and the number of fields where radar had application was extremely broad. So you found yourself with responsibility for a rather surprisingly large amount of this for oneís age. And even two or three years later, youíd find young people two or three years after, well not even a proper degree because it would only be a one or two year course, running groups with 25 people in them. It was a great field. Now that was a way of growing up, a feeling of responsibility of what was possible, "All by myself, I will build up this great thing with the help of some engineering boys," but a lot of it was your own work, (sewing up a de Graaff? 08:23) tube. And youíd build something, a prototype, and then youíd try it. And then maybe youíd apply it. And this is a way of growing up that really is a most fantastic thing to happen to anybody.

Now in America the war situation was different. You came in in a much more organized way. Most of the work was done in large, existing laboratories and then you had some commercial ones. You had no contact day by day between the operational people, the RAF [Royal Air Force] that was flying your stuff, the operational needs, what was possible. There was no interaction whereby a young chap with a couple of years physics training and a couple of years of messing around and deploying a (de Graaff? 09:00) tube could tell an Air (? 09:09) Marshall, "Thatís a bloody silly thing to do. You want to do it like this." You didnít have that you see. Someone higher up in the Air Force wrote a specification for equipment, you presumably had this from all these places, and then it was put through like an ordinary production run. There was no way in which, that young people had this direct contact with operational flying. And I think it was this contact with, as it were, life in the raw, being able to talk to the people who were going to be flying, who were going to be killed in these machines, that made one grow up fast. And I think there is this feeling that one had a training to give one confidence, that one could do things, rather than saying that one could write a specification for a radio telescope and get a bid for it and it due course it will come. I mean if you want to be nasty youíd say that this is what happened to the 600 foot at Sugar Grove, you see. That was designed by a lot of people, none of whom were young enough.

Sullivan

Very interesting. Although, of course the Sugar Grove thing was not really for science. That was sort of tacked on to placate Congress but there are other examples one could think of. That indeed is very interesting.

Ryle

There was a huge difference in cost too. The only way radio astronomy was going to get off the ground in this country was if a small university group with a few people in it could scrounge enough bits and pieces to do something. Well that never seemed to happen in the United States. Much too soon you lept at the idea of a national observatory, and that was a great mistake.

Sullivan

Even before then though -

Ryle

Yes, and I donít know why it didnít start to get off the ground before then. Hendrik van de Hulst and I were invited over by the NSF [National Science Foundation] to talk about the formation of a national observatory and we both came out stronger. The one conclusion that we reached and put forward most forcefully was they must get cracking now, the small university groups, training people, to decide what was wanted, what was going to be the instruments that mattered. But they didnít ignore it, they said that was not the way they were going to go about it. So I said, "What are you going to do about the radio astronomers when youíve got the equipment?" He said, "Weíll always be able to attract enough people from Europe, you see." Now that was an admission not so much of defeat, that was the cheapest way of doing it. And that shocked both of us actually.

Sullivan

Who was the other person?

Ryle

Hendrick van de Hulst.

Sullivan

I see, yes. Very interesting.

Ryle

(? 11:34)

Sullivan

Well anything that I might want to quote you Iíd certainly ask your permission. Henry Palmer yesterday made the interesting remark to me that he thought science or virtually anything was best done in groups of sort of ten. Five was a bit too small. Twenty was a bit too big, and I guess you would agree with this.

Ryle

I think probably I would put it a bit bigger than that simply because I think itís important to have more than one thing going on. I think one thing is done by five people, but I myself always thought 50 is the right size for a research group. So you have a number of lines going on which can interact and help each other. But the group itself, I agree, is about five. And I think weíre too big here actually. Weíve got 80 or something. I think 50 is about right. And independence, Iím a great disbeliever in national instruments.

Sullivan

Well, also on this independence, it seems to me that Ratcliffe must have given you virtually completely independent hand.

Ryle

Yes, obviously he helped us an enormous amount. He didnít want to interfere with the way we were going. We were very small of course. We were only sort of 20% of the group as a whole.

Sullivan

At the beginning?

Ryle

Yes, yes. He saw that or perhaps saw (? 12:58) -

Sullivan

- leave you on your own. Well, letís go back to science. I think the next thing indeed is that you obviously got fascinated with these radio sources having picked up a couple with this thing that was lashed together. What was the next step?

Ryle

Well I think as soon as we realized that Cygnus wasnít unique, there was another one of them, we then built this Long Michaelson instrument, which now had just about enough resolving power so that you could recognize the central maximum, whatever it was 10 to 1 or something. Well it went in two stages. We doubled after a time because we wanted [more?] gain and we werenít of course able to recognize the central maximum. To go back a step, it was evident on the first morning that what a funny thing. That period is different from that, you see. We hadnít realized that here was a way of measuring declination until it actually appeared. So you can get both coordinates from a system with no resolution in the North-South direction. And, of course, it had advantages because you didnít have moving structures. Itís an incredibly cheap thing. You could measure declinations moderately accurately except at low decs. as with all these things. And therefore you get coordinates in the straight away.

Sullivan

Who was it that actually worked on the building of this? The three of you?

Ryle

A chap called Ryle, a chap called Smith, and a chap called [Bruce] Elsmore as far as I recall. No, we had an assistant. What was his name?

Sullivan

Well, Iím asking just in order to get a feel for what the operational style was. You had an idea that you wanted to do something like this and basically the three of you just went out and built the thing?

Ryle

Yes, virtually. There was quite a pile of angle iron needed so we ordered that from down the road and got into (? 15:11) because you didnít have a power drill. And then we put them into the ground and poured concrete around them. We put the guide poles up. We got ahold of some insulators from an old German radar, posts with dielectric tops to mount the dipoles on.

Sullivan

Now this paper was submitted in August í50 but the observations were carried out when? Over the previous year or two years?

Ryle

Oh no, it wouldnít have been that long. I canít remember when that was built exactly.

Sullivan

Since 1948 May is referring to the first array you had for radio sources.

Ryle

Well yes, but we also did measure the spectrum. We build simple instruments on these three wavelengths but those are much smaller than this. This was a thing with whatever it was, 29 meters.

Sullivan

What I was wondering was about how long did it take to do this survey?

Ryle

It was done in one night you see.

Sullivan

Well in fact is that what you did?

Ryle

Well not in fact. The observations because it covers the whole sky (? 16:39) declination strips. The point really was that the cheapest way of getting gain was to make it long and thin and cover a declination range to avoid the need for rotating it. And as soon as this was evident that we could measure declination of the source from the period there was no need to have resolving power this way. And therefore, (? 17:00) it was the best thing to do to get -

Sullivan

So in fact the observations were done in a very short time is what you are saying.

Ryle

Oh yes. I donít imagine the observations took very long.

Sullivan

And you came out with 50 radio sources. I might ask you how many might you have expected? Did you have any idea? Of course, [John G.] Bolton you knew he had six sources. So you knew there were six at his sensitivity level.

Ryle

Yes, well one didnít expect anything. One didnít know what was going to happen. I mean clearly you were going to get some. But I think one knew enough about things that suggested if you got two or six or whatever up to distance A, if you go four times weaker in flux you should be able to get whatever it is, 9 times as many sources. That sort of sum we could do.

Sullivan

Well, that still is a presumption, though, that things are continuing out in an isotropic manner.

Ryle

Yes. Of course the question whether it was isotropic or not was also... I think fairy soon after this we were also thinking that the galactic background was these things, in which case, of course you wouldnít expect to see any anisotropy. For a long time, there is an awful lot of collective background with these things.

Sullivan

It takes an awful lot of stars.

Ryle

Yes, and we did some sums on what the space density would have to be to...

Sullivan

And what flux density youíd have to get down to before you could detect an anisotropy.

Ryle

Yes, that would have come out as well.

Sullivan

And you werenít expecting it at all here.

Ryle

No, no. I think probably the point was the identification that the Australians had made, the lack of the identification of the brightest two which required a lot stronger, of course, than the ones that they had identified, was one mystery. But clearly if some had been identified then you wanted to see what would happen if you looked at more.

Sullivan

Now you are talking about radio stars, and yet in the paper you make quite a bit of discussion over coincidences with galaxies, so you must have been also thinking along that line.

Ryle

Yes.

Sullivan

How did that come about?

Ryle

Well I mean of course the Australian had got M87 as one of them, and as I said I donít know seriously how we were convinced by it, the accurate positions that were then available.

Sullivan

And Centaurus A also.

Ryle

And Centaurus A, yes. That was one of the earliest -

Sullivan

The three of them came together.

Ryle

They did. Yes, thatís right. And clearly one obviously wants to get as many as one can and see optically what you can find at their positions. At that time it was two extragalactic and one galactic so...

Sullivan

So you really were rather open even though these were being called radio stars at this stage.

Ryle

I donít think we had the connotation of within the Galaxy for the word star so much as here was a thing which was a bright point source distinct from the background radiation. And I think people have read too much into the use of the word star in that context. I mean looking at the radio sky youíve got the great Milky Way and youíve got hot points of light in it. And I think it was just the analogy with optical stars, one called them stars this early. Although there was this other paper, I forget when it was, where we did a sort of limits of row?). Was it ...

Sullivan

I donít think so. That was... maybe it was... [shuffling of papers]

Ryle

No, because there was graph. Maybe it was in...

Sullivan

The 2C?

Ryle

No. In the -

Sullivan

The 1950 review.

Ryle

Thatís the thing Iím thinking of. (? 20:55) should give a straight line if they are uniformly distributed without any falling off or anything. And then I think from this we said how near theyíd have to be if you were going to explain the background with them.

Sullivan

Right. Thatís rather interesting because I noticed that Edgeís thesis, the 3C survey, in 1959 is also called a survey of radio stars. And what you are saying is that there is no meaning in the interpretation of that except radio source? It really means nothing different?

Ryle

I think itís meant to imply a compact radio source as opposed to an extended Milky Way.

Sullivan

But it seems confusing to me that this terminology was kept along, if that were the case because there were, of course, these competing ideas about whether they really were stars or not. And why was there not a neutral term? I guess eventually -

Ryle

Well there was, but I donít think at this time this was really developed, this conflict of views.

Sullivan

Well by Edgeís, by í59 -

Ryle

By í59?

Sullivan

Edgeís thesis Iím talking about. Itís also titled radio stars. I was surprised to see this.

Ryle

We didnít call 3C radio stars, did we?

Sullivan

Not in the paper. But his thesis was titled -

Ryle

Well what he writes means no difference to me. No, I think we did call it sources by the time of the 3C.

Sullivan

Iím not sure but I think thatís probably right.

Ryle

But I donít believe, if he called his thesis that, I donít believe he did it for any particular...

Sullivan

I asked him that and as a matter of fact, he said he didnít. Letís see, about the 1C survey, is there anything else that I wanted to... Well, a general question which I want to ask is that the 1C begins a long chain, which is still continuing. Weíd heard at this conference here about the 6C. It just keeps on going. When you were developing this and saying, "Well, letís do a survey of the sky," did you realize that you would just be wanting to go to fainter and fainter levels assuming that you got some sources here?

Ryle

It was obviously dependent on what came out of this. It also depended on whether anybody was going to give us any money. Because up till not we hadnít had any money. I mean literally it was hundreds of dollars, our annual budget. And all of this was done on that sort of money because weíd got bits of ex-radar receiver, weíd got Würzbergs free. We had to mount them but we got the structures free. Most of it, lots of the electronic components were free. Obviously before we went on to the next stage, which was in fact the 2C survey mixing in the paraboloid structures, there was going to be a lot of money involved. At this stage there was no knowing whether that sort of money, which clearly was not coming from university funds, would ever be available. And, in fact, as it turned out it depends on making a good case from what youíd done with your hundreds of dollars. And we did eventually get whatever it was, 11,000 bucks or something, to build the telesopes that made first 2C and 3C surveys.

Sullivan

Was this sort of a breakthrough, youíre saying, in funding in Britain, to get this amount of money? There were not precedents for...

Ryle

Well there wasnít any money at all. There wasnít any money, full stop. But I forget what time it was when, what is called the Science (? 24:35) now. It was called the Department of Scientific and Industrial Research. They started getting funds for the development of university research.

Sullivan

It must have been the early Ď50s or mid Ď50s.

Ryle

Yes, thereabouts. Early Ď50s I suppose. But you didnít have great ambitions early on. You did with what you had.

Sullivan

Ok, well that was the attitude then but was there any time later on when the 2C results were beginning to come in or the 3C that you began to see sort of a 10, 20 year plan? Looking back at your career one can make a nice case that it all develops very logically, et cetera. Is it fair to say that you saw that beforehand?

Ryle

I think perhaps the greatest discontinuity, as it were, was the identification of Cygnus A because that showed that we were in the cosmology game. As Geoff Burbidge says and will continue to say, "Some peopleís minds are shut." But that I know to me was the point where one said, "Well, this is now something much more interesting than it might have been. Itís much more interesting than them being galactic objects and much more interesting than them being M87s." Here is something, even with the little instruments of that stage, was quite likely seeing that things like Cygnus existed, as well as Cygnus. Quite like seeing things as far as the 200 inch. It was a crummy little instrument, costing 200 bucks you see. That was a very important thought and from then on, certainly my own involvement in the cosmology game started in 1950, í51, which was part of [Rudolph] Minkowskiís 200 inch measurements...

Sullivan

Right, í51, yeah.

Ryle

...based on the two Würzberg interferometer positions that Graham Smith got, which were the first ones to get down on a fraction of a minute of arc.

Sullivan

So you think at that stage it became clear to you that here was something that could really go -

Ryle

It went right out into extragalactic space in a way which nobody could have foreseen until that observation, [Walter] Baade and Minkowskiís.

Sullivan

Then you mentioned that the funding for the 2C was the first time you entered the big time as far as funding went so to speak.

Ryle

Well, it was the first time we got money outside the small university Cavendish grant, yes.

Sullivan

But I donít think you still answered whether you could sort of see in your mindís eye over the next 10-15 years that you would be continually pushing to fainter and fainter levels and more and more sources and so forth.

Ryle

I donít think one thought like that. I think it was clearly important to get more sources than this could see and get them at better precision, because here was something which the further you could go was back in time.

Sullivan

So each decision was made only made on the basis of the previous few yearsí experience for the next few years?

Ryle

I think clearly so, because you didnít know what distance the next instrument after this of this was going to show. And until youíd got that, you clearly couldnít design another instrument. You needed to know what one wanted to do. And this affects things like sensitivity versus resolving power and all these things. You hadnít a clue how to design the instrument, the hard fact of what frequency to work on or anything.

Sullivan

Ok. Well I think we better get some lunch.

[Break for lunch]

Sullivan

So continuing a couple hours later. I see at the August í50 RAS meeting that Bolton, well that's independent, but Bolton did talk about Australian results. But you said some of the fluctuations of the radio sources were ionospheric. Itís still not clear if some of them might be intrinsic?

Ryle

Well that was referring to these things, which we were -

Sullivan

These are night time bursts?

Ryle

Yes, and they didnít seem to be something that could only be interpreted in terms of intrinsic.

Sullivan

And weíve already mentioned your paper with Tony Hewish in Monthly Notices in 1950 on the -

Ryle

Yeah, that was the real proof that it was ionospheric.

Sullivan

And you also measured an actual jitter in the positions of the sources, which I think may have been the first time that was done.

Ryle

Yes, I think...

Sullivan

2 or 3 arc minutes.

Ryle

Yes, yes.

Sullivan

Was that done with the two Würzberg dishes?

Ryle

No I think not. I think that would have been working on 210 MHz. It would have been done at 80, I suspect, or even 38, with a small array which we had.

Sullivan

Well one thing I did want to ask you about, in this article according to abstract, I have actually looked at the article, you talk about interstellar matter being accreted by the Sun may be actually what is disturbing the ionosphere as opposed to what we would now call the solar wind, solar particles. I was wondering why you were thinking that way? What was the objection to solar wind?

Ryle

There wasnít a solar wind at that time. People had talked quite a lot about the accretion of interstellar matter by the Sun as a form of supply of energy to the corona. I think origin of this and probably that was sort of in full flight at the time. This is [Hermann] Bondi I think. And I think the feeling was that you measure 1,000,000 degrees coronal temperature, which, of course, about the hydrostatic value youíd have to have for free escape or free fall in. Itís the same naturally. Therefore the coronal temperature distribution tells you no way which the matter is going or the energy is going. And I believe thatís probably why the reason for that remark.

Sullivan

So in other words the source of the heating for the corona was no understood and this was one possibility that was being talked about at that time?

Ryle

Yes.

Sullivan

Your review paper in Reports in Progress in Physics in 1950 is really the first major review of radio astronomy. I wanted to ask you as you wrote that, if you can try to remember back 25 years, did things seem to be coming together in any sense or did it just become more confusing as sat down?

Ryle

No, I think obviously there were very few problems solved but it was about the right time to write it I think. There were a number of fields becoming apparent that were obviously going to be interesting, which obviously the most interesting at the time was the nature of the radio sources. But I think it was an important time to put things together.

Sullivan

In what sense? Do you think?

Ryle

Enough observations had been made to be able to say something worthwhile about them. Quite a lot of groups had got going by then; quite a lot of techniques had been developed. And the overall picture was becoming interesting. There was clearly a future in this game and it worth giving a title, radio astronomy, which I think I coined.

Sullivan

And it was also worthwhile to sit down and say, "Now where do we go from here?" You were obviously going to stay in the game.

Ryle

Yes, yes.

Sullivan

Something that you did not publish anything with, I donít think, but nevertheless were involved with, of course, were the two Würzberg dishes that Graham Smith used. I donít think you published anything relevant to that.

Ryle

No, that was primarily his work. It was built specifically for, we want to push it to a higher frequency both because of the evident satisfactoriness of ionosphere at 80 Mhz where the regional array was, and because its electrical center was much better defined. You see in this measurement of declination by the periodicity of the interference pattern you are required to know the spacing. Now a phased array of dipoles in phase can have imperfections where the electrical center is displaced along it. And that comes in first order as an error in declination. So (A) the higher frequency gets us to the ionosphere and (B) something that had a much more definite electrical center was the right thing. We had these variable dishes available to us and this seemed the right way of going about making, well, the first astrometric instrument.

Sullivan

Right. Now was this phase switched? I canít remember.

Ryle

Oh, yes.

Sullivan

That was the first instrument to be so.

Ryle

No, I think this long Michaelson thing was -

Sullivan

That was also? Ok, yes.

Ryle

Well, phase switching was written up when, 19...

Sullivan

Yes, that was í52.

Ryle

That was sort of a Royal Society write-up, so it was in use sometime before then I think. This certainly didnít start with phase switching because, you see, this record is not a phase switched one. But Iím pretty sure by the time we built the 210 MHz thing -

Sullivan

Right, alternatively in phase and in anti-phase, right. Can you tell me about how that idea came about?

Ryle

Well basically I think the point was that we realized that you are trying to measure something small in the presence of something large of identical characteristics electrically. And you have just got to do something that changes the thing you want and doesn't change anything you don't want or vice versa. The switching gets a noise source with one such thing where whatever you did didnít affect the receiver noise. Now therefore receiver noise could be eliminated apart from all the ordinary noise it produces. The next part which became evident from the picture in there is that if you are going to get sources that are pretty weak then obviously you are limited by the thing not overloading the system and that limits what you can see on this record. So if you can get rid of that as well, and one clear way is to switch the alternative in and out of phase which will not affect anything which is large compared with the...

Sullivan

Constant off set is what you are referring to, right.

Ryle

Well, itís the gains of things as well you see. Before we had the first order receiver gain and now weíve got all the gain of the post-detector thing still present as a deflection, which can vary by small amounts because it isnít all that much gain but it is still an important variation. But the real thing gets rid of the thing you donít want right at the beginning. Therefore it was a fairly easily modification to make the switch to put in a halfway phase between the two and cut the noise diode.

Sullivan

Really one could say that the original Michaelson interferometer back in í46 was the same sort of reasoning that you were trying to detect something small compared to something large.

Ryle

Yes it was. We knew that the Sun radiation was going to be very small compared with the sunspot radiation. We didnít know how small. In fact, it would have been quite difficult to detect with the telescope we had, you know, only a few dipoles each.

Sullivan

There is a story about the acquisition of the Würzberg dishes which Iíve heard in the hallway so to speak. Iíd like to hear from the horseís mouth, so to speak. Could you tell me that story?

Ryle

Well, I knew these things were brought over at the end of the War and went down the Royal Aircraft establishment to be assessed. One came from Holland, I forget where the other one came from. And we knew of the existence of these, so we got an agreement that we could have then. So we went down to arrange for their transport but found that theyíd unfortunately just been sold to a scrap metal merchant, which was sad. But we drove around to see the scrap metal merchant who was a very nice chap, he was a very nice guy. He asked what we wanted it for. We said for scientific research. He said, "You can have them. I like science." So then we were in a bit of a fix because gifts to the University of Cambridge I think have to be recited to the Senate house in Latin. We were quite sure how we would translate all this stuff about German radar sets. However the problem was solved, because we swapped him for a big German trailer which weíd also picked up and some other stuff. We didnít want the trailer. So we all parted happily. He gave us the two.

Sullivan

I see. Did you finally have to go through the procedures at the University?

Ryle

No, because it was a straight exchange, you see. Weíd given him the trailer so that solved this problem.

Sullivan

I see. Ok, now having the 1C survey and seeing that it worked, it was obvious, you already mentioned, that you had to get some outside funding, could you tell me about how that proceeded?

Ryle

Well, here Ratcliffe was an enormous help in making a strong case for us. He, I think, had really realized probably, though he wasnít in the game, that this was going to be an important field and so he helped make our case to the Department of Scientific and Industrial Research it was then called, now itís the (SRC?). And we put up the case presenting the evidence that weíd got so far on these things and, well, of course before we put the case up weíd designed an instrument which had considerably greater collective power and element resolving power. And weíd decided to go up to a four element system because it was a large area of possibilities for using interferometric techniques in various ways by cross connecting them in a phase switch to make a phase difference and relative amplitude of all four connected in different phasing. There were various possibilities of getting accurate declinations, which clearly the periodicity method was not satisfactory as it fell off at low decs. So we designed that in some detail with a local engineering firm, a maker of electric circuits very cheaply by stretching wires across parabolic frames, which parabolic frames are virtually two dimensional structures with stretch wire filling up the third dimension. And this has very low windage. Stretch wires are very much better than mesh, but you only needed stretch wires for one polarization. So it all seemed a very practical way of making it. So we got the design fairly well developed and costed and put in this application for, I think, it was 6,400 pounds, $14-15,000 at the time. And that was granted, and we went out and built that.

Sullivan

And how long did it take to build?

Ryle

It took about a year to build.

Sullivan

And having built it you began what became the 2C survey.

Ryle

Well, the notorious 2C survey was what came out of it first, and what we did wrong was to try to analyze it too deep. And as a result the [overlapping?] abilities gave too many weak sources. But I think that Iíd like to make it quite clear that we presented the case that the distribution of sources in depth didnít seem compatible with the uniform distribution, we were sufficiently worried about the analysis of these faint sources that we invented this alternative method, the P(D) analysis, which was at the time, as it still is, entirely independent of this overlapping image problem. It's dependent, it has to be realized, on the noise of the system, which is quite small. But it did give an absolutely unambiguous answer. The source distribution was not, in fact, obeying a 3/2 power law. And the first time we ever made any public statement, publication, about the distribution sources, we always said that we had confirmed this effect by the statistic method of analysis. Well unfortunately nobody understood the method, if they even read the papers, which they probably didnít. And as a consequence, 2C is mud and the conclusions from 2C are also mud. And that was not, in fact, really fair because this other method was put forward right from the beginning. We never put the source counts from 2C alone as an argument that there was something funny happening.

Sullivan

But I think it would be fair to say, though, that if one looks at those papers now that the main emphasis was logN-logS and P(D) was sort of back it up rather than the other way around.

Ryle

No, the Haley lecture which was the first time this was talk about, that the two are presented as alternative ways of doing it.

Sullivan

I see. Why do you think it was that people were reluctant to look into P(D)? Or did they look into it and have objections to it?

Ryle

Itís fairly complicated. I mean itís just a fairly difficult statistical problem. You can do it just by a crude Monte Carlo type procedure, which is tedious. Or you can do it analytically, which Peter Scheuer eventually did and proved it all precisely. And then Tony Hewish actually applied it in much more detail than had been done before. Let me just give the paper... That was the source counts themselves and thatís the other thing.

Sullivan

So looking at the Haley Lecture now.

Ryle

Here is the probable distribution of amplitude D over recorded phase. [looking at paper] And these are the experimental curve, the hatched bit, and these are experiment curves with spatial density. "The ordinary (? 43:00) instrument might give us an apparent explanation of sources."

Sullivan

Iíve noted here that you did not explicitly mention the concept of confusion and so forth. You must have been aware of it.

Ryle

We certainly were or we would not have invented a system.

Sullivan

Right, that was the motivation for the P(D)?

Ryle

Well, it doesnít actually explicitly say that, does it?

Sullivan

Right.

Ryle

Well the other paper was not far behind actually.

Sullivan

The Bakerian Lecture? Or you mean the Scheuer...?

Ryle

No it wasnít that one.

Sullivan

Was it Ryle and Scheuer that you are thinking of?

Ryle

What was that called?

Sullivan

That was the one that interpreted 2C in Proceedings of the Royal Society. I donít have the title here. I donít think I have that one. In í55, the same time as the... Here it is. I think that is the one you mean.

End of Tape 59B

Part 1 | Part 3

Modified on Tuesday, 26-Apr-2016 09:03:09 EDT by Ellen Bouton, Archivist (Questions or feedback)