Interview with W. Bruce McAdam

Sullivan: 00:00

This is talking with Bruce McAdam on 17 March 78 at Radiophysics in Sydney where he's visiting from the University of Sydney. Now, could you tell me how you first came into contact with radio astronomy?

McAdam: 00:14

Right [unclear]. In Auckland I had been doing the MSc work under Francis Farley, Cambridge graduate but in nuclear physics, and he had been working as a research supervisor to most of the nuclear physics group in Auckland. I'd been doing some radioactivity. I got a reasonably good MSc there, and in vacations I'd been working on an underwater sound project with echo [unclear] there at that time two different fields of experience. I applied and got a commission in the New Zealand Defense Scientific Corps, a thing which then existed to make up [leeway?] in science and the new service interfaces. Now in that year going to England and going to Cambridge there was a Grahame Fraser who did ionospheric work with Ratcliffe, and Mark Barber who did electrical engineering, and myself.  There were a couple of others, Roy Kerr, a mathematician, but he was not in the Defense Science.  The organization of getting us placed into various English universities to gain overseas experience was done partly by our own wishes, partly by the Army, or Air Force, or Navy's idea of what they wanted, and then in relation to Armstrong and the scientific liaison staff in London, backed by a network of who had been successful before.  Now Ratcliffe, with his association with Rutherford, had already had a number of New Zealanders through the Cavendish. Jack Maudsley was one, Colin Franklin, and while I was joining, the understanding was it would be to do some radio research in either Manchester or Cambridge, those were the two things I wanted. But until I was fairly close to leaving in September 55, I didn’t know which one was having me. And the position with the radio astronomy group rather than the ionospheric group [unclear], at Cambridge was decided in England. Now this was done through Ratcliffe because until then Ratcliffe had insisted that Ryle had only Cambridge PhDs, and that 1955 October intake of research students was the first that had any overseas people in it at all.  That is why it was spread fairly widely, Carman Costain from Canada, Jan Högbom from Sweden, myself from New Zealand, and then David Edge was the homegrown Cambridge undergraduate to come through. They made the four intake.

Sullivan: 03:21

For 55.

McAdam: 03:22

Yes.

Sullivan: 03:23

I see.

McAdam: 03:23

Now the early ones you'll have one check out, Baldwin, Thomson, [unclear – crosstalk] for earlier still and they were all [unclear] I think correctly had been feeding Martin with the best of the associates through that period to get him started.

Sullivan: 03:42

So when you got to Cambridge, how is it decided what you were going to work on?

McAdam: 03:47

Well, I can document fairly [reasonably?] because I came in with the firm undertaking of writing three month reports back to the Army in New Zealand, and so kept these documents [of what I was doing?]. Graham-Smith was appointed my supervisor. So October 26th I made a change from [thinking?] around with matching of 45 megahertz IF type receivers, noise figures, and trying to get inputs matched and my bandwidth. Let's see. September the 26th I arrived in Southampton. Had a meeting on 27th September and we went to Cambridge. Immediately [the first?] [crosstalk]--

Sullivan: 04:41

[crosstalk] detail from anyone else.

McAdam: 04:44

Well, the first thing that happened is we put out onto the Rifle Range painting the aerial [crosstalk] spot [crosstalk]--

Sullivan: 04:51

Right. [Good practical experience?].

McAdam: 04:52

--climb out in there. And it was a [inaudible] so I had 3rd to the 7th down with a cold. 14th to the 25th of October under Graham I started that 45 Meg work. And it was obvious I think then that I wasn't doing very much. Each Saturday morning they had a group discussion on what was going on. And some of the things were interesting. I think it was Robin Conway who was wanting to improve the integration time on some high frequency 400 Meg [inaudible] he was [inaudible]. And instead of a time constant on the typical Cambridge interferometer [inaudible] frequency around about one second [inaudible] and they had about three second time constant sitting on that. They wanted to get time a constant more akin to the band-pass. And we're talking of all sorts of radio [inaudible] RC networks [crosstalk]--

Sullivan: 05:45

Conway was in this article, I mean.  Or do you mean the Cambridge group was?

McAdam: 05:49

Conway and the technical group in the small, then 12 people total in the group from Martin right through to four new students. [inaudible]. So Robin Conway was doing the work. He had the need and the general Saturday discussion would be roaming over any problems.

Sullivan: 06:15

I thought that that Saturday morning thing was for reviewing current literature.

McAdam: 06:19

That was the other half [crosstalk]--

Sullivan: 06:20

I see, it was divided in half--

McAdam: 06:21

From 9 until 11:30-ish. And [from?] 9:00 to 10:00 or 10:30 it was literature I think. And then there was the general organization, who would be doing what? Who [crosstalk]--?

Sullivan: 06:35

Oh, I haven't heard about that aspect of it. Can you tell me did Ryle sort of run these meetings and just sort of get progress reports from different groups or was it more or less just giving out assignments?

McAdam: 06:48

As I recall there was more a general round table, who was going to be where? What students would meet with the governing body if at all that week?

Sullivan: 07:00

Who was the governing body?

McAdam: 07:02

Martin, Tony, Graham--

Sullivan: 07:04

Hewish, Smith, Ryle.

McAdam: 07:07

Yes. Ryle. Sorry. I'm going to get myself [inaudible] think [inaudible] there because it was a small group. It was very informal. [It was?] Ryle is still about 33, 34. [inaudible] everyone else was younger so the age range was mighty small in comparison to the most recent times.

Sullivan: 07:30

Right. And this was every Saturday morning?

McAdam: 07:32

Yes. Yes.

Sullivan: 07:33

And would the last week's work be discussed also and decide [what to do next?]--?

McAdam: 07:38

Maybe if they brought up some problems. This was the way the top trio kept in touch with what was going on. I would normally see people around since the afternoon tea at the Cavendish tended always to be a time of meeting people, talking. But it was possible for a research student to be working at home or at nighttime [unclear] not see anyone else much. It was always expected into Saturday morning.

Sullivan: 08:10

anyway. I sidetracked you, you were talking about this [longer integration time?]--

McAdam: 08:14

I had experience playing with ionization chambers with very high, 10 to the 12, 10 to the 14 ohm input impedance inputting of electronic tubes back in the complete valve era. And so I knew what sort of input impedance, grid currents, grid current noise, shot noise, valve. And so spoke up [about] that and found out October 26, which is just about a month after I arrived I got a comment, “switched to the output unit receiving system quadrature switching and smoothing unit” where the idea is to put a relay switching and a filter on.

Sullivan: 09:01

I see.

McAdam: 09:02

No. 2nd November “leakage grid current work.” 15th November “build a unit and test.” 28th “revised test, get a nice chart comparison with the old record.” So by December there I would start a design of permanent gear and put a revised unit on a four-day trial. Now, that idea of switching meant you had an in-phase and a quadrature phase switching, which gave you a sine and a cosine type outputs which had to be separately integrated. I started thinking how that could be recombined instead of having two chart outputs which were pretty heavy to work out. And to add two such waveforms in quadrature and get the sum of the squares, the amplitude, and trying ideas of thermistors, and things, and it was very slow response.

Sullivan: 09:56

Were you working pretty much by yourself on all this?

McAdam: 09:58

At this stage, yes. I think I had got the switch and fairly simple RC stage for Conway to go, and that would be mentioned in paper on some of this 400 Meg stuff they published by me. I gave up attempts at recombining these two, and about April in that next year, '56, realized that it was just as easy to combine them in quadrature in time as attempt it with voltage squared to power any other system. And by rechopping each of these outputs and adding or subtracting appropriately, filtering, one out of the whole algorithm, found all the nasty terms exactly canceled. And you could get a true band pass filter. That then followed right through to the band pass filter I made for the 3C survey and the chopping time interweaving modes for the 38 Meg Costain interferometer. So what I'm saying from the one month after the work of start on the receiver, I was then working on the tail end [inaudible] the data recording and equipment hardware outputs for the whole thing.

Sullivan: 11:22

Well, one thing I'm interested in, was there a course of lectures for the research students or was it simply a matter [crosstalk]--?

McAdam: 11:29

There's no formal lecture course whatsoever. When some lectures started, there was no examinations on them. I think we're into our second year, Reddish or Redman, the Prof in Astronomy gave a course in basic astronomical things for the radio astronomy group. Which was useful since I don't think any of us had come in from the background of astronomy, not even as a hobby in my case. And this covered the galaxy, extragalactic stuff, [crosstalk] population one and two, and population two and one, supernova

Sullivan: 12:10

Over a full year, was this?

McAdam: 12:11

That was over probably only one term. Imagine, by memory, something like 15 or 20 lectures.

Sullivan: 12:19

But nevertheless, that's not all that much.

McAdam: 12:21

No. There was another course that started. It was given every second year by Ratcliffe on transform methods, his ionospheric stuff; which led on, I think, later to Bracewell stuff and Budden's work on diffraction.

Sullivan: 12:44

Well, Bracewell had been in Cavendish around 1950 or so [crosstalk]--

McAdam: 12:48

For 15 years before, yes. I didn’t overlap at all.

Sullivan: 12:53

So it seems like, then, there were only a few courses that were taken and basically, you just learned what you had to.

McAdam: 13:01

You learned by apprenticeship system, which I think probably still carries on there. They have more lectures, more colloquia, seminars, than I remember from subsequent visits there in '70 and '77. But there’s still no formal course exam requirements before you hit your thesis type, essentially.

Sullivan: 13:23

So it's just a matter of the - what was the term? - the directory or whatever, the--

McAdam: 13:30

Interaction with the rest of the group.

Sullivan: 13:31

They just decide that you're ready for thesis research? The governing body?

McAdam: 13:36

Oh, no. You'd start that as soon as you arrive. But you might get pushed across from one topic to another after.

Sullivan: 13:44

So it's only when you’ve accumulate a large enough project or [enough] to support a large enough project and then you can say, "Well, it's my thesis."

McAdam: 13:54

Right.

Sullivan: 13:56

There's no admittance to working on a thesis or anything like that?

McAdam: 13:59

No. Once you're admitted as a research student, you are on your thesis then.

Sullivan: 14:03

I see.

McAdam: 14:04

This is, of course, the same system we're using here in Sydney and seeing it from the other side, one is always looking for moderately compact ideas; thesis topics that are likely to come to fruition inside two years, which means a student has a chance, inside four-- has a chance of being separately identified and allowing to work somewhat on his own. The requirement is to show initiative and analytic ability. An MSc has just to show competence in handling things, but PhD requires that independent initiative.

Sullivan: 14:44

Also, I'm struck by the fact that you were basically working on an electronic problem. Were you worrying at all about the astronomy that was going on?

McAdam: 14:56

One regret is I didn't worry enough through that period. I knew the work of the others, of course, what was going on currently in radio astronomy, but I didn't read extensively in the library to know much what was going on in other labs, which I think is the part which is most interestingly [inaudible]. Right until the end, I had the dual role of getting some astronomy out of my 3C work and simultaneously - I think it was December '57 - I was constructing the switching and output then for what would be the 4C, as well as a separate one for the 38 Meg, Carman Costain synthesis. And those were tested and running, and some form of a manual roughly put together by the time I left. But then through the middle of '58, it had to take a bit of second place to getting the analysis done on the observations-- somewhere about mid '56, August and September '56, I did the observations of the lower culmination 3C survey and was analyzing them, therefore, early '58, mid-’58, and then writing up the thesis.

Sullivan: 16:24

That's what became his thesis, then.

McAdam: 16:25

Yes. I was unusual in having a deadline. I was told the end of October, 3 years. That's when I went back to New Zealand. So thesis just had to be finished then or else I paid my own way back for the viva, as did one lazy man, John Barge in engineering.

Sullivan: 16:42

I see. So within a year of arrival, you're saying that you had taken the data for your thesis, although it wasn't analyzed by any means?

McAdam: 16:52

Yes. Yes. This meant, for much of the aerial design, of course, and instructions on matching and monitoring and gain, was dependent on John Shakeshaft. He was, I think, the effective supervisor for David Edge who was doing the main 3C part. David's observations were around that same period and--

Sullivan: 17:18

I've talked to him about his own work.

McAdam: 17:21

All right. Well, the case of putting a long couple of meter, slotted coax line onto a wheelbarrow and taking it out to a field to plug up to check what's matching and things like this.  That was all done. I never quite understood what stages of matching I should do. I could match one point and was told go and do it. But John Shakeshaft then was the one who we relied on about aerial technology.

Sullivan: 17:55

Who was making the decisions as to how the surveys were going to be split up amongst the research students?

McAdam: 18:05

That is one of the hard ones. David Edge's main theme was--

Sullivan: 18:10

In his book--

McAdam: 18:11

--observations of 3C [crosstalk]--

Sullivan: 18:12

Hold on. What do you mean? His thesis?

McAdam: 18:14

Yes. Plus, he held detailed observations of things, so that he had the upper culmination then. When this band switching type thing started to show that it could reduce the noise on the output-- frankly, I am not sure who got the idea of it, but of course, since the declination of a radio source altered its periodicity, then such a band pass filter would discriminate against things at the wrong declination.

Sullivan: 18:44

So you could actually-- it's a declination filter in essence.

McAdam: 18:49

Yes. And this would have been useful for cutting off sidelobe responses and very strong signals (Cas A) well off the Dec[lination]. It must have been a governing body decision I think, that it had potentials to clear at very high declinations some of the side lobes or the fringes inside the main lobe area. And then by working to a lower culmination where the interferometer spacing projection is very much smaller, the aerials were very nearly obstructing each other, shadowing each other. The main lobe had only three fringes, the center one and two side fringes. And it was possible to build a filter and chop drastically on the two sidelobes. Therefore, we got rid of all declination with lobe ambiguity there. And because it was at a three or so degree offset in the alignment of the Grange Road aerial system - the Rifle Range - it meant the lower culmination observations crossed the sky at the opposite angle to the upper culmination. They had two intersecting things that would yield quite a lot of information about the right ascension fringe.

Sullivan: 20:05

This three-degree offset, was it purposeful?

McAdam: 20:08

No, that was just the way, the original Michelson—the 1C aerial was booked and then they-- this is the alignment of the block of land that we had available there. And so all those Grange Road sites were built pretty much on that slightly offset angle.

Sullivan: 20:27

I see.

McAdam: 20:28

Two Wurzburgs that Graham Smith got his Cygnus anchor position with. Then John Blythe aerial for the original synthesis. The dual-twist [inaudible] line of yagis were used for another synthesis there. All of these tended to be aligned with the boundaries and property, which was a long thin--

Sullivan: 20:50

I see. Well, I'd like to get your impressions of the general atmosphere as a research student in the radio astronomy group at that time.

McAdam: 21:04

Now, just the personal interactions from top to bottom?

Sullivan: 21:06

Right.

McAdam: 21:08

Well, I've said the age group was narrow, and this meant--

Sullivan: 21:11

This must've meant a lot of camaraderie.

McAdam: 21:12

No great restrictions or inhibitions on the first few months. Research students talking to the top governing trio. Ratcliffe was regarded as somewhat above that level. He was always Mr. Ratcliffe. Jackie Saville, his secretary, was always front line boundary between and one didn't go popping in-- but one did go popping in or out for minute or two with a query to Ryle, Smith, Hewish.

Sullivan: 21:49

Right. And I guess Ryle actually got his-- no, he got his chair after you left.

McAdam: 21:55

Quite a lot after. I can't remember-- just then he had his FRS. He got his chair '59 just after I had returned.

Sullivan: 22:07

Quite a lot should have changed?

McAdam: 22:09

I don't think so. It probably did for the research students just coming in. But it didn't for us who had been, so each subsequent visit I've been able to find

Sullivan: 22:24

but even now I know his office is the Cavendish is not buffeted by a secretary.

McAdam: 22:27

Right, and I think you’ll find that was policy. All the radio group.

Sullivan: 22:33

So anyway, you were saying that there was this direct access to anyone in the group.

McAdam: 22:39

Yes. The other thing which is important is there was extremely good liaison between each level of PhD student, the ones writing up, ones who had just finished, and this then would include Peter Scheuer was theoretician, John Baldwin, John Shakeshaft, came at a level just after things with Tony Hewish. Then there was a range of John Thompson, George Whitfield, John Blythe, most of the senior students, they were the ones who were then finishing, and tending to get married that stage. John B. John S. Had already written up theses before I arrived.

Sullivan: 23:30

Right. But now what is your point about these different--

McAdam: 23:35

They were just as capable of acting as effective supervisors to us when we needed, problems to know about parts of the system that hadn't met.

Sullivan: 23:46

So what you're saying is really that virtually anyone with a bit more experience than you was useful, you didn't have to only go to the governing body or something like that right.

McAdam: 23:54

Or they would come down to us, to me in particular on this input thing, if they had an inquiry there.

Sullivan: 24:05

Yeah, well, that's a good point. What about if you sort of dawdled along? Would you be getting--

McAdam: 24:12

No, a specific comment on that one, there would be intense pressure for a couple of weeks if you haven't turned up, or nothing much seemed to be happening. They did keep almost a week by week watch. More out of personal contact, [to recording]. Nothing formal about it. But they would be aware if somebody didn't seem to be making any progress and after a couple of weeks you'd start getting a prod. Three, four weeks, you'd be fairly intense pressure. Maybe two months and then it would drop off. There'd be no more and thet’re general reaction was they're not there to push the ones who aren't really interested.

Sullivan: 24:54

In other words. They'd really give it up at that stage.

McAdam: 24:57

No. If the person showed interest, well, he came back again.

Sullivan: 25:02

Right, but they weren't going, they're just going to let him stay and maybe they let them do it on their own project. If it was an important one.

McAdam: 25:11

That I don't know. They would eventually, but it would take a year, a year and a half perhaps, before anything like that would happen. One case. This will be private won’t it? Use the information. But honestly, Simon Archer was a case where he had started. I think, with the full intention of doing a degree, and he didn't mix a great deal. He tended to be a loner. Worked for a long time away from the lab, and it was getting more and more difficult as his work needed to be blended in with the rest of the 3C information to know what to do with it. And I think that's where John Baldwin came back into the 3C efforts, to try and push Simon. Not because they wanted Simon to get a degree, but because they wanted the information, the whole outlook on the source structure. He was doing the close spaced north-south interferometer type mode in the 3C, as opposed to the interferometer between the east pair and the west pair. This was essential to cover the chance of missing extended sources which is one of the bones of contention between Mills and Ryle.

Sullivan: 26:34

Well, just the next topic I was going to bring up. You've already mentioned that you are not as aware of work at other observatories as you

McAdam: 26:43

[Thinking of Jodrell Bank?] because in my three years I only made one visit to Jodrell Bank, that was getting late in 58 when I went up with Carman Costain. John Thompson was already there. I remember, we went around looking at the work that was there, the 220-

Sullivan: 27:01

But it took you three years, though, to get up there.

McAdam: 27:05

If I had been, say [my present mood here in Sydney and some major observatory research students would have been freaked?] and it would have been colloquia from people at Manchester.

Sullivan: 27:19

So you didn't see people coming down either from Jodrell?

McAdam: 27:21

Not much.

Sullivan: 27:24

But what I was going to ask you about is, this, of course, is the period of great contention over the 2C survey versus the Mills survey. And from the research to this point of view, are you always hearing about this?

McAdam: 27:37

You're not always hearing, no. It was fairly much agreed inside the group that there had been a great overestimate of the 2C. This was by August, September '55, remember? And I think the criticisms and reassessment has been done by then. But there's equally a firm conviction that Bernie Mills was going too far in his criticism, and a lot of my work and orientation in the thesis was designed to show that with four aerials involved the final resolution is not that of just the Polar diagram. One of them that inside the interferometer north-south and east-west there is independent information. And it was one of the regrets the hand type analysis of the 2C system which used exactly the same aerials but at 80 Meg.

Sullivan: 28:35

Which type analysis, did you say?

McAdam: 28:36

The hand chart record or template. There were no computers around in those days. EDSAC was still a mathematical development which came up eventually the end of 58 or something. No, everything was done off paper chart at the 2C and all the 3Cs were chart analysis, too.

Sullivan: 28:58

Anyway, you were going to say that limited the amount of analysis-

McAdam: 29:03

That had limited the amount of information in the aerial pattern that was used in developing the 2C. They hadn't used the phase or timing information implicit in the interferometer system except to judge which was the center and get a fairly accurate timing of that center lobe. Hence the lobe ambiguities when you weren't quite sure among noise which was the center. But the idea inside a ripple like that attempting to find whether there were interfering sources nearby blending to make the composite record. Did the phase, for instance, shift across the apparent polar diagram worth of fringes that were on the record? I don't think that would be used at all in the 2C analysis. And yet the phase and amplitude were equal information sure, they lost the amplitude one.

Sullivan: 30:00

But you say this, quite early-- well, when you arrived, there seemed to be [general?] agreement that a major error had been made [crosstalk] in the 2C survey-

McAdam: 30:06

[Unclear] Certainly by the time of the analysis stage, late 57, 58, in the comparisons with the upper colimation [unclear] 3C were good news, comparisons within the earlier catalog, the 2C, was just no interest. It had to be done out of curiosity since this was a published existing thing, to see really how bad the 2C was at inventing sources. But no one took it as being publishable [inaudible] interest.

Sullivan: 30:50

You mean because 2C had already been written off?

McAdam: 30:52

2C had already been written off, I think by, certainly, March '56.

Sullivan: 30:57

Yes, but the thing that I have difficulty in understanding is, obviously the incredibly clever group of people there who know the ins and outs and invented the interferometry to a large extent. How did they go wrong in this? Where do you think is the fatal flaw in their thinking?

McAdam: 31:21

That would be the supervision of the earlier work, I think. The records of the interferometer system, did look as though it came down to a fairly small residual thing and then came up as a source. You could plot a fairly reasonable profile of large small sources scattered around.

Sullivan: 31:45

So what you're saying is the bumps all look pretty good?

McAdam: 31:47

They looked reasonable.

Sullivan: 31:48

And they were reproducible?

McAdam: 31:49

Certainly, they were reproducible, yes. That is, noise or interference at times [it’s a nuisance?], but go back over and the vector error for noise was extremely small compared to the amplitude of the smaller- This is well known. Both 2C and 3C were confusion limited rather than noise limited.

Sullivan: 32:15

Right. But, now you were saying the lack of supervision?

McAdam: 32:21

In getting out a list from the records- I was not there. I imagine now that it would have been the research student, John Shakeshaft, himself doing that. He would compile the lists and then these would be perhaps edited a little bit, but not with a keen analytic eye of say Tony Hewish or Martin. Looking at the record itself and the list to try and decide what was being put in.

Sullivan: 32:55

Well, I agreed it seems like something like that must have happened, but yet that's hard to understand because--

McAdam: 32:59

It's [?] being with the idea?

Sullivan: 33:01

No, I mean, that implies that they did not care that much about the survey to spend the time doing that. And yet, of course, it was extremely important.

McAdam: 33:13

You remember, every source that was listed there was something permanently on the record. So that level of cross checking.

Sullivan: 33:21

Okay, so what you're saying is that they really did miss the idea of confusion and how many sources per beamwidth--

McAdam: 33:28

Is really reliable, yes. I don't think anyone really knew much then. To some extent Mills had the benefit of looking at his and saying this is crazy, this doesn't agree. Why not?

Sullivan: 33:40

Yes. Oh, sure.

McAdam: 33:41

And it's easier to answer the question, why not, once you know [inaudible].

Sullivan: 33:45

Sure.

McAdam: 33:46

But then the difficulty is someone in Cambridge then believing that without seeing the records, Bernie Mills’ records, we really did prove that there was nothing around that spot. They weren't as firmly convinced [the thing was nonsense?]. And when you're looking at your own records, which you trust and see something published, you're much more convinced that there's a contradiction, the thing published is wrong, than if you only hear somebody else has records to say you're wrong. So for a while, there would not be the same conviction in Cambridge that 2C was at fault.

Sullivan: 34:26

Until they got the preprint of Mills and Slee’s- ?

McAdam: 34:31

Yeah. They -

Sullivan: 34:32

First comparisons--

McAdam: 34:33

Australia, pretty much confirmed that the thing was nonsense. I would mean, too--

Sullivan: 34:38

Am I getting you wrong? Are you saying that there was rumors, or perhaps more, that something was awry even before they got the preprint of Mills and Slee?

McAdam: 34:48

No, I’m not saying that. But that there would be reluctant to believe that, where they had not access to Mills’ records. So it was a slow awareness that they had maybe--

Sullivan: 35:05

Right.

McAdam: 35:06

And like most people, I suppose. [Unclear - blow away?] and not being too serious. But most of us recognized then, it put a lot more pressure on the 3C.

Sullivan: 35:18

Yes.

McAdam: 35:18

That had to be right.

Sullivan: 35:19

Right. Did that change any of the parameters of 3C? David Edge just told me about this, was it going to be a lower frequency?

McAdam: 35:30

No, I think it had to be twice the frequency. And the survey was pretty much underway. When I arrived, the aerials had all been, feeds put up [unclear] for the 159 MHz, I guess. It's just twice the frequency.

Sullivan: 35:50

But I think I remember that there was some definite motivation that 3C was going to be such that this-- so a criticism of confusion could not be levied against it.

McAdam: 36:00

Right.

Sullivan: 36:01

And so--

McAdam: 36:02

This was at the analysis, or the repeating scans, across [?].

Sullivan: 36:08

But you've got to design the aerial in the proper manner.

McAdam: 36:12

No, no, I don't think so.

Sullivan: 36:15

Well, okay.

McAdam: 36:18

Not aware then, is this your [crosstalk]--

Sullivan: 36:20

-- can't cut off at a higher flux. You know?

McAdam: 36:24

You made much more use of the phasing-

Sullivan: 36:27

Yeah, okay, yeah.

McAdam: 36:29

total information. So this comparison against the lower-upper culmination is one which was in that zone, anyway.

Sullivan: 36:40

So it's not correct to say that 3C, 2C having been pretty much established to be quite bad, but then Ryle and cohort said, "All right, let's do 3C and we'll design it this way. And this will be a good survey."

McAdam: 36:57

I think the decision to move to twice the frequency was a valid one, much akin to our Molonglo decision these last years that 408 is now coming to an end, what can we do, what's all that effect [inaudible] let's go up. Instead of going to 1400 [S band?] a year back, they're going to 8-4-3. That same sort of jump, they can use existing hardware as being implicit in all observatories.

Sullivan: 37:21

So--

McAdam: 37:23

It's the same thing as putting [crosstalk] on Parkes dish or on--

Sullivan: 37:24

But it was only by chance that that greatly reduced the confusion problem.

McAdam: 37:29

Oh.

Sullivan: 37:30

That's what I'm getting at.

McAdam: 37:32

No, there's perhaps a chance of going up in frequency, which was the scientifically optimal thing then, got the solid angle down to a quarter and gave you a chance to get into much higher source densities.

Sullivan: 37:50

Well, gave you a much higher-- much weaker source. Many fewer sources, honestly. Well, of course, the two things play off of each other, don't they? You've got to be careful.

McAdam: 38:02

The sensitivity was still adequate to get everything confusion-limited. Let's say it gave four times the resolution. Twice in each [inaudible]. And the determination when [the 3C?] was influenced by [? 2C?] scanning as much as the analysis scanning, it's more robust than-, you are quite correct that there are a lot fewer sources in the end despite the solid angle cutting to a quarter.

Sullivan: 38:33

About a factor of five or so fewer, if you are at 400 versus 2,000. Did you go to the Paris symposium, by any chance?

McAdam: 38:40

No. If you look at the date of the Paris symposium, I would have dearly loved a week over there.

Sullivan: 38:47

It was just when you were going back.

McAdam: 38:48

I was writing my thesis. I got married on October the 11th and I sailed-- flew from England, rather, on November 3rd, I think. My thesis had to be finished. It was at the binders while I was on my honeymoon. I submitted it when I got back. No, there was no way I could spare a time for it.

Sullivan: 39:10

Well, one final question I'd like to ask you is about what are the key ingredients to the success of the Ryle’s group, would you say, over the years? You saw one section of it, but -

McAdam: 39:23

Well, firstly, they, like most of these, grew out of radar groups and they were the brightest of the development engineers and scientists there. That's one ingredient. The second is the immense amount of cooperation inside the group, sharing day to day problems with anyone in the group.

Sullivan: 39:47

Not much internal wrangling of any kind.

McAdam: 39:49

I don't know of any that had went on, no.

Sullivan: 39:55

Which almost all organizations seem to have a healthy share of.

McAdam: 39:58

There must have been some rivalry in different parts.

Sullivan: 40:01

Right. But that's been my general impression, though. They're--

McAdam: 40:03

There seem to be extremely harmonious relations. Mind you, 12 is a number, in my period, below the Parkinson limit. It says by the time it's up to 20, 25, you're having difficulty getting that feeling of coherence.

Sullivan: 40:19

12 was the size of the staff at that time?

McAdam: 40:20

It was something, yes. I'm pulling that out of memory without counting out.

Sullivan: 40:26

Yeah, we can check that.

McAdam: 40:27

The third thing would be this deliberate fostering of the liaison on a timetable business. The Saturday morning sessions, the afternoon tea things which started as a subsidy by I’d say Rutherford, but I'm not sure. Either that or Thompson. But it got washed out with inflation, but the idea that the Cavendish afternoon tea was a place where people met to talk, not just to sit and have tea, was very strong. Every student would meet, not his supervisor but the governing body, about once every three weeks, probably, in that period. Now this meant, then, the student was meant to give a very close report on what he had been up to and what problems he had. But again, if he had problems, then somebody else would be brought in from the sidelines, the other students or whatever else, the post-grad fellows, to try to clear it up. Mechanical work or anything, then everyone, regardless, joined in. The technicians and the groundsmen, the research students and the university lecturers all helped pull wires through.

Sullivan: 41:52

So it seems like you're saying that, besides the obvious innate ability of the chief and the staff, that it was a real group effort and that this developed a certain atmosphere.

McAdam: 42:07

The ideas grew at any of these meetings. It became difficult, even half an hour later, to jot down who had started an idea or what contribution anyone had done. It was more of the brain tank concept--

Sullivan: 42:25

Think tank.

McAdam: 42:25

--although no one ever put that into words. I think that's what it was in each of the after literature discussions.

Sullivan: 42:34

Interesting.

McAdam: 42:36

And there's no doubt about it, at that time my respect for Martin Ryle was developed, that he could get an idea from anyone. And he seemed to follow it through, find out what its [call?] was, and then push it along a bit further. He would then perhaps toss the idea to someone else to mull over for a while. Among the team, he did seem to be the fastest fellow we had. Not that the others were slow. I imagine in any of the main university centers around the world the same ability would be there. So any difference in performance or any type of output that went on in Cambridge must be something other than the innate mental agility and background experience.

Sullivan: 43:29

Well, I think that's right. It depends on various styles.

McAdam: 43:31

It did not happen, as far as I've been able to tell from subsequent discussions at Jodrell. There were much smaller groups and it tended to become two or three people here, two or three there, in somewhat isolation. Three is below the critical number too.

Sullivan: 43:48

Although you have to be careful, because that was the way it was here at Radiophysics too. And here, they were actually very separated sites, and yet it worked down here in terms of getting good research results.

McAdam: 43:58

I would have thought you'd learned already why it went here.

Sullivan: 44:01

Because of Pawsey?  Pawsey moving around and coordinating things?

McAdam: 44:03

Yes. Pawsey was acting as the three man trio governing body in Cambridge.

Sullivan: 44:12

Yes, but nevertheless there was not the sort of interaction of everyone getting together, tea, on Saturday mornings, drinks--

McAdam: 44:19

Well, that I don't know. Although [there was a little there?]--

Sullivan: 44:21

Well, that's what I've learned. Often people didn't see other people much at all. There were some fortnightly meetings, at which some things were done, but it was not like a Cavendish [experience?].

McAdam: 44:34

No, we would have had perhaps fortnightly meetings with some of the INS group as well, who were very much separated.

Sullivan: 44:41

That ends the interview with Bruce McAdam on 17 March '78.