Interview with Peter A. G. Scheuer
Description
Creator
Rights
Type
Interviewer
Interviewee
Location
Original Format of Digital Item
Duration
Interview Date
Interview Topics
Notes
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 and Alan H. Bridle. Any notes of correction or clarification added in the 2015 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 when they first appear. 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.
Series
Unit
Transcription
Sullivan
So this is 27 August ’76 at Grenoble and I’m talking with Peter Scheuer. So before we talk about your first scientific publication and so forth, could you tell me a little bit about your background before you came to Cambridge and then how you got involved with the radio astronomy group?
Scheuer
Well, born Frankfurt am Main 1930. My father was Jewish so we left Germany after the usual difficulties. We fortunately got out just before the War started. And then I went to school and all that. I was an undergraduate at St. John’s College Cambridge. And towards the end of my third year I heard Martin Ryle talk, giving one of the talks that various research groups give to undergraduates. That seemed to be a subject where one could still make progress with fairly straight forward experimental methods. One could understand what the problems were because they were very open. So even though I have no taste whatever for electronics, I thought this would be an interesting thing to go into.
Sullivan
What year was this now?
Scheuer
The first thing that I worked on… sorry?
Sullivan
What year was this?
Scheuer
This was ’51. So the first thing I worked on was trying to get some notion of the width of the radio emission from the galactic plane for two reasons. One is that there had been a suggestion in the literature that the galactic plane got wider and wider at lower frequencies and the suspicion was that this was simply because the beams got wider and wider. And the other was the paper by [Gart] Westerhout and [Jan Hendrik] Oort had appeared in which they discuss the contribution of the H II regions to possible radio emission from the Galaxy. And one should be able to separate this out because the H II regions were very closely concentrated to the galactic plane. So we set up an interferometer at curious angles…
Sullivan
Now when you say we I assume…
Scheuer
This was Martin Ryle and I…
Sullivan
Martin Ryle basically had the idea to do this…
Scheuer
Well Ryle had the idea to do this…
Sullivan
And you were his research student.
Scheuer
I was his research student, yes.
Sullivan
Now can you just describe the experiment and the results?
Scheuer
Well, the experiment was… obviously the methods were very primitive in those days. And so we put up a couple of aerials. And we put them up at an odd angle across the field behind the university football ground where radio astronomy was conducted in those days.
Sullivan
That’s the Grange Road site?
Scheuer
That’s the Grange Road site. And along a line such that we’d catch the galactic plane at the time when the Milky Way formed a vertical great circle. And we just recorded fringes and moved the aerials along a bit. And so it did amplitude spacing.
Sullivan
I thought Ryle was pretty much interested in radio sources at this time yet, this was a bit of a reversion to extended radiation. Or was it the idea that you were trying to understand the relationship between the radio stars and the possibility that the background might just (be integrated?)? Was that Ryle’s real interest do you think?
Scheuer
No, I think it was trying the separate out the component H II regions. And, of course, it did do that. We did this experiment at several frequencies and it showed that the temperature required of the H II region would be well above the conventional 10,000 degrees. It would be more like 25,000 [degrees]. And in the end it turned out that there was a population of radio sources which was also very strongly concentrated towards the galactic plane. And what we were observing was a mixture of the H II regions and these radio sources, supernova remnants.
Sullivan
Which we would call supernova remnants, right. Did you recognize these discrete things?
Scheuer
No, no, no. I mean our interferometer had no resolution along the galactic plane to speak of. It was about 10 degrees resolution along the galactic plane, varying from one frequency to other somewhat.
Sullivan
So about this time [Bernard Y.] Mills came out with his Class 1 and Class 2 sources. I’m trying to understand the relationship. You did not really then come up with a second kind of source in the plane. You were just more worried about what the integrated emission was.
Scheuer
We were more worried about what the integrated emission was. Obviously we were aware of Cassiopeia. The Mills result, I think, didn’t reach us till a little later. It was [Robert] Hanbury-Brown who came up with the notion that all the radiation was from supernova remnants.
Sullivan
I see.
Scheuer
Or at least from a particular class of source, not necessarily from supernova remnants. And, of course, that’s not right either. There was an H II component in what we observed at these low frequencies—well, in the absorption that one saw at low frequencies.
Sullivan
Now did you see any absorption?
Scheuer
Well…
Sullivan
It talks about a bright band, a 2 to 3 degree wide bright band.
Scheuer
Yes, that’s right. Well, there was a very primitive experiment indeed that used rhombic that went down to 38 MHz. And that probably began to see a little absorption. At least the brightness didn’t follow the standard spectrum.
Sullivan
And was it clear to you that from the theory of free-free that…
Scheuer
Oh, yes, yes.
Sullivan
…at some point you were going to see absorption?
Scheuer
Yes. Well, that’s probably enough on that.
Sullivan
But that’s sort of an aberration from most of the rest of your work. What was the next thing that you got involved with?
Scheuer
Well, that was my thesis work essentially, that and a little bit of work on free-free, which I didn’t publish till…
Sullivan
You had an article in 1960…
Scheuer
That’s right.
Sullivan
On the absorption coefficient.
Scheuer
That’s right. But the thing which was the big new instrument was the antenna of four parabolic cylinders which made the 2C survey and later the 3C survey, and still later the 408 MHz survey, which in retrospect would probably have been the right thing to start off with.
Sullivan
Now, hold on, you said 408?
Scheuer
Yes, 408 MHz. There is, in fact, a paper in the literature by [R. J.] Long, [J. B.] Haseler, and [Bruce] Elsmore in some order [Note added 2015: MNRAS, 125, 313 (1963)]. But anyway that doesn’t matter. By the time we got around to that it was unimportant.
Sullivan
Well it’s my understanding that the 2C aerial was built with the intention of doing two frequencies, 81 and 170.
Scheuer
And 170, that’s right.
Sullivan
And I’m more confused now about what you are saying about…
Scheuer
Oh, what I’m saying is that if we had understood the confusion problem well enough, it would have been fairly clear that the right frequency to start off on would have been a much higher frequency than 80 MHz. Because with the wisdom of hindsight one knows that one wants 50 beams per source rather than one beam per source, which is one’s first intuitive order of magnitude estimate. And which, of course, led to all the trouble with the 2C survey.
Sullivan
But even 170…
Scheuer
170 was better but it was still confusion limited.
Sullivan
Yeah, well, it has to be one or the other, confusion or sensitivity limited.
Scheuer
Yes.
Sullivan
But this is an interesting question to me. Apparently there was discussion about which frequency to do first. Low frequency won out. Ryle said there was interference at the higher frequencies and such. I’m just interested if indeed the high frequency survey had been done first, do you think the confusion problems would have still been there at 170?
Scheuer
At 170 there would still have been confusion problems. And what the interpretation would have been, I don’t know. But we’ll get on to that I’m sure. That’s a long story.
Sullivan
Anyway, this is all trying to rewrite history. I just want to look at what did happen. So you were involved with the 2C survey.
Scheuer
No, I was not involved with the 2C survey.
Sullivan
Oh, no, you are not. You’re not a co-author actually. But you were very involved in the analysis of the results.
Scheuer
I was not involved with the results of the 2C survey. No. The thing that I was involved in is what has come to be known as P(D). Now this has a very curious history in that as the 2C survey was progressing, one of the worries was that... Martin was aware of the confusion problem, although obviously it wasn’t very clear what it consisted of. The way he expressed it to me as far as I can remember in the first instance was that if you have a bright source it will conceal a fainter source that is near it. Therefore we shall miss a lot of faint sources. Of course, on going into the problem one realizes in the end that it is the other way up and that the predominant effect is two weak sources, or several weak sources, building up into something that appears rather stronger.
Sullivan
So the initial impetus you’re saying was a bit mistaken, but nevertheless would be a problem?
Scheuer
Nevertheless, I then went to being one of the more mathematically inclined of the students there - which isn’t saying much - there weren’t many of us. I thought, well, one ought to be able to work out what the probability of the distribution of the amplitude on the record is. This is a very defined problem for probability theory. And I felt very pleased with myself when I got an answer after several months of doing what a well-educated student now could do in an afternoon, rediscovering along the way various well-known items in probability theory. [Laughs] And the trouble with this was, I think, threefold. We then went ahead and published an article in which the counts from the 2C sources were published and also some results of the analysis of the records using the probability of the deflection on the record.
Sullivan
This is Ryle and Scheuer ’55 that you are referring to?
Scheuer
That’s right. And the P(D) method did not give the same answer. It did not indicate as large a deviation from the slope of 1.5 as the counts did. And so it was thoroughly unpopular at home for that reason, if one pressed to this quantitative conclusion. And nobody else believed it because, of course, it was well known… well, for one thing there was the great argument with Mills’ group. And typical of the comments on this would be one that was made at the Paris Symposium that it’s obvious that you can’t get anything out of the observations by this alleged piece of black magic. You can’t extract anything out of the observations by mathematics that isn’t there in the first place. And as the observations were no good in the first place, therefore this must be all wrong. And nobody understood it largely because nobody bothered to understand it.
Sullivan
They basically just said, well that the survey is bad, so why should we worry about analysis of sort of a rear guard action to scrape something out of it. Is that what people were thinking?
Scheuer
Well, if you read the proceedings of the Paris Symposium I think you’ll see the remark I’m referring to. That was 1958.
Sullivan
But nevertheless the slope that you got was not 1.5.
Scheuer
No. Well the nearest guess I could make was around about 2. I can’t remember what the figure quoted was. It was around about 2 if one had to make a guess at some mythical slope. Of course, it doesn’t work that way.
Sullivan
Well, I’ve just been talking to Mills actually in the last days and one point he made that he didn’t like, I’d be interested in your response, is that you had to assume an even distribution of sources in this, that everything was sort of…
Scheuer
Random, yes. The sources were uncorrelated with each other. That, of course, was another bone of contention, in that he found a lot of extended sources.
Sullivan
Well, that was one part of it. But also he found in his initial part of his MSH survey that there was some clustering effects. Now they went away later on, but he was arguing at that time that such effects would invalidate the P(D) analysis.
Scheuer
Yes that is quite true, quite true. There is another thing that would invalidate the P(D), all the counts, which is that one didn’t know at that time was the angular diameter distribution was and, of course, anything done with an interferometer would discriminate against large diameter sources. And given the same kind of source, of course, the closer ones would have larger angular diameters. And this was the thing that worried me most about the source counts later on when we understood the confusion problem properly and, of course, I believed the P(D) analysis. [Laughs] So that I believed that there probably was something wrong, even though the statistical significance wasn’t enormous.
Sullivan
Were you willing to look at Mills’ survey to find out what percentage of sources were extended?
Scheuer
Well, we weren’t in a position to do that because his were in the Southern Hemisphere, very largely. His first strip was fairly far south. But we couldn’t look at the angular diameters very much and most of the angular diameters were too small to resolve.
Sullivan
Yes, but what I’m thinking is that you could take his survey just as a statistical sort of thing, and figure out what percentage are in what… what fraction of them are of a certain angular size, at least to his resolution. And this might give you some input into the angular diameter distribution.
Scheuer
Well, how would we do that? I suppose we could take the total power survey and compare that with the interferometer survey and see whether very much more turned up.
Sullivan
This sort of thing, and if…
Scheuer
The tendency was for the total power survey to show a good deal less than the interferometer survey. And the reason was partly that in the interferometer survey you could convince… and this was part of the optimism that was in the original analysis of the 2C survey, that on the interferometer record you could convince yourself that every beam-sized set of wiggles represented a source, if you felt that way inclined, whereas on the total power survey you couldn’t. So at that stage we really weren’t in a position to do it. The obvious way out of that, I was quite amused to see that Bernie Mills proposed this again at last week’s symposium in Cambridge, was that you did surveys with various scaled aerial systems so that you could cancel out any angular diameter effects or clustering effects or whatever they might be, by changing the scale as you go deeper, to lower flux levels. This is the first order way of cancelling out any angular diameter effects.
Sullivan
I wasn’t there went he made that comment. Now did he make this comment also back in the late ‘50s, are you saying?
Scheuer
I don’t know whether he did or not. I don’t remember him doing it in print. I certainly made it privately amongst ourselves. It would involve a large scale construction program for the job. By the time one might have done this, there were better things invented like synthesis.
Sullivan
Well let me ask about the P(D). I’m a little confused as to whether it all along was part of the game plan so to speak for the 2C…
Scheuer
Oh no, no, no. It came up after the observations had started to come through.
Sullivan
Before Mills? Was it a response to Mills...?
Scheuer
No, no. It wasn’t a response to Mills’ thing. No, I did that analysis, I worked all that out in 1954. I didn’t put it into my thesis because it wasn’t really related to that sort of thing. And then, of course, I was caught in the tail end of national service. I spent the next three years away. That’s why it wasn’t written up till 1957.
Sullivan
I see. That explains the three year gap.
Scheuer
That’s right. Which, of course, everybody else knew very well was Cambridge secretiveness.
Sullivan
It was really the Suez Crisis or something like that.
Scheuer
No, no, no, no.
Sullivan
Let me also ask how did it happen to get published in the Proceedings of the Cambridge Philosophical Society?
Scheuer
Because no reputable journal… Sorry, I beg your pardon. We tried various journals like Monthly Notices and they wouldn’t have it because it was tarred with the same brush as the source counts. It was clearly disreputable. So in the end rather reluctantly the editors of Proc.Cam.Phil.Soc. [Proceedings of the Cambridge Philosophical Society] said that well they would take this on. Nobody was willing to have it.
Sullivan
Ok, another question about this whole controversy I’d be interested in your point of view. It seems to me that there was a great confusion, not to be a pun, in issues. That people were arguing back and forth about several different things and often at cross purposes. One was the validity of P(D) and whether it was really an independent check. The second thing was the effect of extended sources.
Scheuer
Well, one of the things about P(D) that I ought to make clear . The sort of comment was that you clearly couldn’t extract information about sources below the one per beam width level. And the thing that took years and years and years and years to sink in was that I had in fact carefully excluded… the was a comment made mostly, of course, by people who hadn’t read my paper, and I don’t complain of this because I don’t read the literature very much either. But I had, in fact, spent two pages explaining precisely why you couldn’t go beyond the one per beam width level. But the P(D) method could give you information down to the one per beam width level. Where the source counts, as such, could only give you information down to the one per twenty or fifty beam width level.
Sullivan
So they were arguing at…
Scheuer
As you said, arguing at cross-purposes. I was merely explaining where the cross purpose arose.
Sullivan
Another thing that was being argued about and, of course, and it influenced all this was which sort of cosmology you were you were inclined towards—the steady state or evolving cosmology. Another thing was what should be counted as a source. I think that this was going on. The net effect was that these issues didn’t get defined, really. Would you agree with this, that if there’d been some debate referee or something that could have kept the issues separate that things might have gone more smoothly?
Scheuer
I didn’t get the impression that one was inclined to any particular kind of cosmology. After all Martin Ryle didn’t have particular prejudices about what sort of cosmology he liked.
Sullivan
Well maybe not before he began the 2C. But once he had the steep slope and he saw that this could say something about cosmology, he certainly had inclinations for the next many years, did he not?
Scheuer
Towards saying that it couldn’t be steady state, yes.
Sullivan
Yes.
Scheuer
But not towards any particular cosmological theory. He’s not a theoretician. He’s not inclined that way.
Sullivan
I’m just talking an evolving cosmology versus steady state. Just that kind of distinction, not what kind of evolving.
Scheuer
Oh, yes. But I think he would have been quite happy with steady state if steady state could have explained those source counts.
Sullivan
OK, I see what you are saying.
Scheuer
I don’t think he had any particular predilection for one cosmology or another.
Sullivan
Ok, I guess that’s right. So in the case of Ryle, anyway, you could say that. So Ryle and the Cambridge group are sticking by their source counts. The theorists like [Fred] Hoyle and [Hermann] Bondi, of course are sticking by their theory.
Scheuer
It’s very natural. Hoyle built the… well, Hoyle, Bondi, and [Thomas] Gold built the steady state theory. Hoyle was one of steady state theory and Ryle built his radio telescope.
Sullivan
And Mills also built his radio telescope.
Scheuer
And Mills built his radio telescope.
Sullivan
OK, I see what you are saying. That makes sense.
Scheuer
It’s what people’s interests are.
Sullivan
Another question is that it seems incredible to me with the amount of talent in the Cambridge group that the confusion problem wasn’t realized. I mean it’s always easy in hindsight, but do you have any thoughts to how this went wrong?
Scheuer
I think it was realized. [Long pause] It was realized in principle. What wasn’t realized probably was just how bad it was. You know, one knew that there was a dimensionless number, the number of beam widths you had to have per source. One’s first assumption was that this is of order unity. And I think what one didn’t realize at first was how big that number was going to be. And the other thing was that having built what was at that time, you know, the biggest aerial in the world and so on and so forth, the temptation to extract as many sources as possible was very great. And I’m sure that played a part. How big a part I don’t know. As I said, I wasn’t directly involved. Obviously we were all very closely involved at that stage because there were so very few of us and everybody knew what everyone else was doing, which is no longer the case. But different people were inclined different ways on this.
Sullivan
As to how many sources should be extracted, you’re saying?
Scheuer
As to how big the confusion problem was. But all I can say now would be speculation, and would probably be tainted with all sorts of consideration since. So I think we would be better if I didn’t put it on your tape.
Sullivan
Ok. Well, with that proviso it still might be useful if it’s your honest opinion as to what was going on then. I mean that’s true of everything one says. It’s after twenty years. So don’t let that stop you. I’d be interested in hearing some more about…
Scheuer
I think it would be so speculative that’d I’d rather not put it on your tape.
Sullivan
If it’s speculative then, right.
Scheuer
You know, one can never rely on one’s memory. One’s memory plays strange tricks.
Sullivan
Now, the 3C survey, as I’ve already mentioned, was not something in reply to Mills’ survey. It was part of the plan all along.
Scheuer
That was part of the plan all along, yes. It really was a question of what frequency one did first.
Sullivan
But I guess it gained an urgency to it because of all this ruckus. And you wanted to get things straightened out, the Cambridge group I mean.
Scheuer
I doubt whether it would have made much difference because after all, once you’ve done a survey of the sky, you’ve done a survey of the sky and that aerial is then obsolete until you change it to a different frequency. So I think that just went ahead. I’d be surprised if it speeded things up.
Sullivan
Now what happened when the 3C came along? You were not involved, I guess, once again…
Scheuer
Again I was not involved in the actual survey, that’s right.
.Sullivan
But certainly the analysis. You’re a co-author in a paper in Monthly Notices and at the Paris Symposium about comparing 3C with 2C and with MSH and P(D) and so forth.
Scheuer
Are you thinking of [David O.] Edge, [John R.] Shakeshaft and myself?
Sullivan
Right.
Scheuer
That was one of the most gruesome papers I’ve ever had to write because I disagreed quite fundamentally with Martin Ryle about what ought to go into that paper. I’m now saying a few of the things that I said I wouldn’t say. I was inclined to be very much more cautious about what one could assert. Largely for the wrong reasons. Because what I said was that we didn’t know what the angular diameters were, and therefore we couldn’t be sure we weren’t cutting out a whole lot of bright sources for that reason.
Sullivan
Which were extended?
Scheuer
Which were extended. And I knew the P(D), which I believed more than anybody else did, of course, gave a somewhat flatter slope, a considerably flatter slope from the source counts, taking it straight.
Sullivan
And what was Ryle’s counter argument? That there weren’t many extended sources?
Scheuer
Possibly that. Possibly. I think that he just thought I was being bloody-minded or pedantic or some combination of the two. In a sense, I was being pedantic. I didn’t believe on the basis of what little information that we had that we knew the answer to the angular diameter problem. And, well, I was no doubt very young and even more tactless than now. We had some jolly good rows. [Laughs] I must say the thing...
End of Tape 65A
Beginning of Tape 65B
Sullivan
So this is continuing with Peter Scheuer on 27 August ’76.Scheuer
Just as the last piece of tape ran out, I was saying that though I had some tremendous scientific arguments with Martin Ryle, I’ve been consistently impressed by his generosity, and that this has never been reflected in anything personal or in things he has done for me. On quite a different subject, something which occurred to me after we finished this morning that perhaps I should say, at the risk of repeating what other people have told you, is that in the early days there was not only resistance to the sorts of measurements we made because of anything to do with the confusion problem or the angular diameter problem. There was a much more basic hang-up so far particularly as the optical astronomers were concerned, which is that they had virtually no experience of or understanding of interferometers. And the standard argument was, well, that the interferometer only gives you a few Fourier components. It doesn’t give you an image and you can’t really deduce anything from just a few Fourier components. So an interferometer doesn’t really tell you anything one can believe. And it took a very long time before any optical astronomers would really take any serious notice of anything that wasn’t produced with something that looked like a telescope. They could understand what a paraboloid was. But an interferometer, again, was the unknown devil.
Sullivan
This comment I would presume goes all the way back to the late ‘40s that you are talking about.
Scheuer
Well, it goes on into the ‘50s, I think, quite comfortably.
Sullivan
I mean it’s not just 2C you’re talking about.
Scheuer
I became a research student in ’51, so it certainly went on into the mid-‘50s and maybe somewhat beyond.
Sullivan
What you care to name any names?
Scheuer
No, I wouldn’t care to name any names, partly because I can’t remember them rightly, but it was very much in one’s consciousness. And it’s something that by now I think one tends to forget because everybody is familiar with interferometers. And it wasn’t merely old fuddy duddies. It was very bright people who had this attitude, even some people in radio astronomy who were not concerned with (continuum?) measurements.
Sullivan
Right because in the earliest days of radio astronomy, the ‘30s and ‘40s, there was a natural reluctance by optical astronomers to understand what this was all about, all these wires and radio technology. But they got over that to some extent. But what you are saying is that only if it was somehow nicely analogous to optical astronomy, namely with a dish. And they still had conceptual problems with the power of the interferometric technique and so forth.
Scheuer
Yes.
Sullivan
Well, just to finish up, we have not actually talked about the Paris Symposium as a meeting, I don’t think.
Scheuer
No.
Sullivan
I’d be interested as you were there--it’s a rather significant meeting as I see it—to what your comments would be about that meeting.
Scheuer
Well, what do you want me to comment on?
Sullivan
Well, the atmosphere of it, whether you learned anything from it or…
Scheuer
We had a session in which there was a rather tense debate about the source count business. And at the end of which I felt, I think, somewhat resentful at not having been allowed to reply to various criticisms, which were just allowed to stand. The other remark that I remember is Fred Hoyle’s remark that the main thing that had happened since whatever the last meeting was that he referred to, had been the hardening of opinion in favor of the synchrotron emission process, implying that it had become established since it was never going to be thought as anything else.
Sullivan
One could almost still say that today.
Scheuer
Yes.
Sullivan
I’m always amazed as how synchrotron seems to pop up everywhere because it’s so useful. But I always wonder if that is really what is going on. Were you involved at all in synchrotron theory? I don’t…
Scheuer
Not at that stage. No, I really got interested in the details of synchrotron theory somewhat later.
Sullivan
But as an explanation, for instance, for the galactic background as opposed to the integral of millions of radio stars, when would you say that this really came to be accepted? Was it between ’55 and ’58?
Scheuer
I find that very hard to remember. It was a gradual process, I think. There was a school of thought which held that the background was really the superposition of lots of supernova remnants, which may in some roundabout, devious sense be true, maybe if all the cosmic ray particles made it into there. But which I don’t think stands up if taken literally. And there was the question of the continuum. Of course, one of the interesting things that happened that I wasn’t particularly concerned with right at the beginning, was the existence of the galactic halo. It was really John Baldwin’s vision. And he proposed this and it was seized on by the cosmic ray astronomers. And then the radio astronomers took it away again and the cosmic ray astronomers haven’t let it go yet. It’s still a matter of lively debate from time to time, how many galaxies have such things.
Sullivan
How did you view the Russian synchrotron theory work as a theorist yourself?
Scheuer
I wasn’t a theorist in those days.
Sullivan
Oh, you weren’t? I see.
Scheuer
No, I was an honest observer.
Sullivan
Well, of course, P(D) I suppose is an area of analysis.
Scheuer
I suppose. Ok, yes, alright.
Sullivan
I see. So you weren’t directly involved. I’m interested in how that became accepted in the west and so forth.
Scheuer
I think gradually, as people learned about it. I mean it’s quite a sizeable, quite a nasty bit of analysis to work one’s way through and one has to be fairly convinced that it is relevant before one goes through that trouble.
Sullivan
At the Paris Symposium, did you come back to Cambridge with any new ideas as to what direction you should take in your own research?
Scheuer
No. I think the directions that one wanted to take in one’s own research were fairly clear.
Sullivan
What were they?
Scheuer
There were a large number of very obvious questions. And what I was particularly interested in getting into was the measurement of angular diameters, lots of them, partly for their own sake as this was clearly the way to go. And partly, which is the way which Cambridge has gone since, was synthesis, which, by the way, in principle was quite clear from the mid ‘50s. It was a matter of building the instruments and developing the techniques. And also it was made possible, it was really made practicable by the advances in computers. It would have been very difficult to do in any other way.
Sullivan
Yeah, I think Ryle has traced it in his notebook back to ’54 or so, supersynthesis. But the angular diameters you were saying.
Scheuer
Yes, the angular diameters, partly for their own sake and partly to clear up this question whether they accounted for the lack of bright sources.
Sullivan
And so did you get involved in that after the Paris Symposium?
Scheuer
Not immediately afterwards. I did later, in one of the scientifically sadder parts of my life. I went to Australia for three years, Sydney ’60-’63. And I did get an engine going for measuring angular diameters of lot of sources. But the problems with calibration and analysis were quite difficult. And the amount of material that was produced, the number of sources that turned up on the records was huge. And, of course, there wasn’t digital data processing in those days. And so I’m afraid that was all but lost in limbo.
Sullivan
It never was published?
Scheuer
Never published.
Sullivan
Now, can you tell me a little bit about what this engine was that you referred to?
Scheuer
Oh, well, this was quite fun in a certain way. The thing that I used was a kind of antenna system which we called barley sugars, which were invented by… now, who did them? Was it Tony Turtle? Or John Baldwin, or both between them? I’m not sure. You’d have to check up on that. Which consisted of a transmission line, which you wound into a helix. So you’ve got a little bit of quadrupole radiation, you see. And, in fact, what is it called now, apodize I think it is the trendy word for it, on the aerial as one liked by widening the thing as it went out to compensate in part for the loss of energy, considering it was a radiating mechanism, as you go towards the ends. The beauty of this that it is exceeding cheap to make rather a lot of it and you can roll it up and hang it up somewhere else on some sort of washing line. And we made quite a lot of this stuff.
Sullivan
I see. What does this name come from?
Scheuer
Because barley sugars often have a twisted pattern through the middle, you know the rock.
Sullivan
I don’t know what barley sugar is.
Scheuer
Well, do you know what rock means?
Sullivan
Rock candy?
Scheuer
Rock candy. Well, there is a kind of it…
Sullivan
Which is twisted, I see. And so what did you actually make then?
Scheuer
Well we made a lot of this stuff using the east-west arm of the Mills Cross. That’s one end of the interferometer. And taking this rolled up thing out to various out stations.
Sullivan
I see.
Scheuer
So I did have a go at this and it didn’t come to anything. Of course, later on there was no point because of the synthesis methods.
Sullivan
What was it that really was the problem?
Scheuer
Well, the problem was that there was too much data to process manually, really.
Sullivan
You got overtaken by the synthesis method, would you say?
Scheuer
Well, partly that, partly that I moved back to Cambridge before it had all been sewn up and the analysis completed.
Sullivan
Now you also worked on the Parkes Dish, I believe, while you were there.
Scheuer
I worked very little bit on the Parkes Dish, just really mainly for the fun of using it, because at that time one could actually push the buttons.
Sullivan
Wasn’t this relevant to the whole source count and confusion business? Wasn’t that what you were trying to straighten out?
Scheuer
No, no. The thing which I did with the Parkes Dish was to look at some of the sources which appeared in Mills’ catalog as extended sources, as large extended sources, to see whether they were in fact blends of smaller sources or whether they really were extended things.
Sullivan
And the result was?
Scheuer
The result was fairly mixed. Well, on the whole I think they broke up into blends of weaker sources. But the difficulty, of course, is that to do this job with the Parkes Dish you had to work at 21 cm and to compare that directly with observations at 85 MHz. This really was a rather large leap so I couldn’t say anything absolutely conclusive, but my impression is that those extended sources had faded away. I don’t think we saw any more of them.
Sullivan
You had to work at 21 to get enough angular resolution?
Scheuer
Yes. I mean that’s the only sort of procedure…
Sullivan
Perhaps it says in your paper but I’m not familiar with it, was there a certain fraction of these that were galactic background features, so to speak, as well as blended?
Scheuer
Oh, low latitude ones certainly partly were, yes. There’s no doubt near the galactic plane but the high latitude ones I think are on the whole. There are, of course, some big sources. They do exist.
Sullivan
It’s rather interesting that you went to the camp of the enemy in 1960. How did that come about?
Scheuer
Oh, because Bernie Mills wrote, or perhaps it was [Joseph L.] Pawsey, and said would you like to come over.
Sullivan
I see. So things had settled down enough…?
Scheuer
I don’t know whether things had settled down. I think the point was that these disputes don’t have as much personal animosity as people often think. In spite of the very heated and sometimes quite bitter scientific disputes. One of the things that amazed me when I got to Australia and met Bernie Mills personally—and you know he was extremely friendly. He came down to the ship to meet me—was that in some ways they have a great deal in common in…
Sullivan
Their style.
Scheuer
Their style. The sort of things they are interested in. Their politics.
Sullivan
And as you know, when two people do have very similar things there sometimes is sort of a grating there.
Scheuer
I enjoyed Australia tremendously.
Sullivan
What was the other question I was going to ask? Oh yes. Mills, in talking to him a few days ago, was telling me about the operational definition of what he called a source, namely anything that had a bump on it, whether it was extended or not. And he was trying to make a contrast between this sort of thing and the Cambridge thing in the 2C anyway, which… what was I going to say… there there was an operational definition also, I would say, anything that fit the overall pattern, if the envelope looked right.
Scheuer
Well that’s much the same, isn’t it?
Sullivan
Yes, it is.
Scheuer
All that happens is that in one case you have fringes within the bump.
Sullivan
Yes, although there was some different he was pointing out and I wanted to get your opinion. I guess the point was that I asked him, “Why didn’t you throw out the extended ones for fear that they might be blends and such?” And he thought that that would be dangerous to some degree. I still don’t see quite why it wouldn’t be a safer way to proceed from his point of view.
Scheuer
Well, then you’d need an impersonal way of drawing the line between what is a source that is worthy of being included in extended sources and what’s worthy of being thrown out. That’s even more difficult.
Sullivan
Yes, but there is always that. Even if you take all the extended ones, there is some signal level at which you decide that it’s not a real thing.
Scheuer
Well, this is a problem which was aired again at the IAU Symposium number whatever it was last week in Cambridge…
Sullivan
74.
Scheuer
People were saying, “Well maybe there are lurking quite a lot more sources like 3C236 which have vast outer components. There are probably some more of these that we don’t know about yet.” Even this problem…
Sullivan
It stays with us. And, of course, the controversy about…
Scheuer
The converse one is also there, and it is very much there for the faint sources. When do you count two dots as being parts of a double and when do you count them as two separate sources? And there are cases where this has practical consequences that we know about now. 3C225, for example, is now thought to be two unrelated sources and so is 3C402. I think it’s 402, which is broken up into two of these low power sources, each with its own associated…
Sullivan
Well, along the same line, it seems to me that that meeting pointed out that the entire LogN-LogS thing, however you interpret it, or whether you should even bother to interpret it, it is still an extremely active issue after 20 years. Would you agree with that?
Scheuer
I doubt whether it’s as active as it was, myself, but this may merely reflect that my interests shifted away from it. That is to say that it was a very gentlemanly meeting last week. And there were various long discussions of marginal results, anisotropies and so on and so forth. But the basic source counts, the basic data seemed to be agreed. How one interprets it is not in very much doubt. And after all the arguments about whether the source counts show evolution or not has become a bit more academic because anybody who doesn’t care to believe the source counts will still very likely believe that the microwave background indicates that we are in an evolving universe. At any rate, there’s no plausible explanation for it in steady state.
Sullivan
But what I was trying to say is that there is still controversy about whether LogN-LogS can be used as additional evidence for an evolving universe. Do you think that is a fair statement?
Scheuer
Marginally. I wouldn’t have thought there was all that much disagreement about it. Because even if you can find enough sources in one way or another to fill up the gap for high flux density, you still run into a lot of trouble because we now know that a lot of these things have high red shifts. And if you believe that large red shifts mean cosmology, then the source counts ought to bend down rather quickly. So it’s just not enough. And not only that, even if you don’t admit that the red shifts are cosmological, you still have a good deal of trouble with the old, old argument that appeared in Martin Ryle’s Bakerian Lecture in ’58, you get too much background with all these red shifts that are ignored. You put them immediately around us.
Sullivan
Yeah, indeed. In closing I’d be interested in your comments about the nature of working in the Cambridge group. You’ve been there now 25 years or so. It seems to me that there was an incredible unity in the group and I’d be interested in how you think that has come about.
Scheuer
It’s come about a very largely, I suppose, because of Martin Ryle. Because he is a great enthusiast and because he had his feet very firmly on the ground, because he has not become an administrator and committee man. Because he is very intimately concerned with the details of the hardware and with everything that goes on with his group. That’s a very important thing. And I’m amazed that it works, because one of the things that I find very disturbing, considered in the abstract, that I would consider terrible if somebody just told me about it, is that the staff members of the Cambridge Group, I think, are all but one people with Cambridge PhDs. They’ve all gone through the system. So it’s a terribly inbred collection. And if I’d had just heard that in the abstract, I would say that such a group must have decayed. It must have grown short of ideas. It must have grown short of steam. And it amazes me that it hasn’t.
Sullivan
I made exactly the same comment to John Shakeshaft last night, that I find it amazing, because I know several of cases where one can point to that were that has happened, a sort of stasis has set in.
Scheuer
I think the reason for people staying there is that Cambridge is very pleasant place to work in. It’s a very good place to be in for astronomy now because there is a variety of different things going on. There’s a very strong group in the institute across the road. And furthermore, with a continual flow of visitors from foreign parts, so that probably plays a part.
Sullivan
Any other comments in closing that you, looking back over the development of radio astronomy, that you’d like to have.
Scheuer
I think nothing that can be accommodated in one turn of the tape. [Laughs]
Sullivan
OK, thank you very much. That ends the interview with Peter Scheuer on 27 August ’76 and this is the end of the tape.