[Roger Jennison, ca. 1963]
Roger Jennison, ca. 1963 (Photo from Sullivan Papers)



NATIONAL RADIO ASTRONOMY OBSERVATORY ARCHIVES

Papers of Woodruff T. Sullivan III: Tapes Series

Interview with Roger C. Jennison
At Grenoble, France
31 August 1976
Interview time: 90 Minutes
Originally transcribed as typescript only by Pamela M. Jernegan (1979), retyped to digitize by Candice Waller (2017)

Note: The interview listed below was originally transcribed as part of Sullivan's research for his book, Cosmic Noise: A History of Early Radio Astronomy (Cambridge University Press, 2009). No release form was obtained by Sullivan for this interview. In preparing Sullivan interviews for Web publication, the NRAO/AUI Archives has made a concerted effort to obtain release forms from interviewees or from their heirs or next of kin. In the case of this interview, we have been unable to find anyone to sign a release. In accordance with our open access policy, we are posting the interview. If you suspect alleged copyright infringement on our site, please email archivist@nrao.edu. Upon request, we will remove material from public view while we address a rights issue. Please contact us if you are able to supply any contact information for Hachenberg's heirs/next of kin.

The original transcription was retyped to digitize in 2017, then reviewed, edited/corrected, and posted to the Web in 2018 by Ellen N. Bouton. Places where we are uncertain about what was said are indicated with parentheses and question mark (?).

We are grateful for the 2011 Herbert C. Pollock Award from Dudley Observatory which funded digitization of the original cassette tapes, and for a 2012 grant from American Institute of Physics, Center for the History of Physics, which funded the work of posting these interviews to the Web. Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event.

Click start to listen to the audio for tape 73B of the Jennison interview.

Begin Tape 73B

Sullivan

Now weíre talking with Roger Jennison on 31st August 1976 at Grenoble. So can you tell me what the first thing you did in radio astronomy was? How did you get involved?

Jennison

I got involved Ė may I just go back one quickly? Just for personal history, before the War, at the very beginning of the War, I was an undergraduate in engineering. I then volunteered for (air crew?); I was in the RAF for five or six years. I was made redundant that group towards the end of that, and I went into ground radar, and finished off my career in the armed forces in ground radar. I came out Ė just before I came out Ė personal anecdote, if you like, I was at the end of my Air career for a short time in charge of a library whilst I was resting between things. And in that library, I came across a biography of Lord Rutherford. I read it because I had nothing else to do, and I decided I should not be an engineer, although incidentally Iím acting as an engineer almost nowadays. But I decided that my real interest was more fundamental, and that I should go and take a degree in physics. Go back to square one. So I went back to square one after the War, took a degree in physics at Manchester Ė started taking a degree in physics at Manchester. The reason for going to Manchester in the end, personal, was that the, I had ample time in the Air Force to go through the various lists of universities in the UK, and I was influenced most by who happened to be in charge of a particular physics department. I turned down Cambridge, curiously enough, because for one thing, I couldnít understand their prospectus; it was in Latin or something like that. But I could not really fathom it out. At Manchester, there was (?) the cavity (?) had been invented at Birmingham, and that certainly interested me. At Manchester, there was a chap called Blackett, who I repeated encountered in reading Rutherfordís biography. Therefore, I decided that Manchester was the place for me, and I went to Manchester and became an undergraduate in physics. Thatís, of course, of more elderly years that a lot of my colleagues. And during my first year there, a friend of mine, who was also an ex-serviceman, encountered a paper by Menzel and Salisbury, maybe in my second year, Menzel and Salisbury, Don Menzel, who is at this conference.

Sullivan

I talked to him at Cambridge, actually.

Jennison

Heís looking very, very old now Ė poor old Don. But still he goes on and on. Anyway, yes 1948, thatís right, because I went to Manchester in 1947, so in 1948 this paper by Menzel and Salisbury came out, and I was rather interested in it. It was concerned with some (?) very low frequencies. So we thought we, I think I did a quick calculation and decided that the frequency concerned, or not quite their frequency, but at a frequency of about 7 cycles a second, it was possible for the ionosphere of the earth and the solid sphere of the earth to form a cavity resonator. So I thought, ďWell, letís have a look at 7 hertz,Ē and we constructed an extraordinary aerial system. It was a bunch of galvanized iron wire. We had to do it very, very cheaply. Blackett, incidentally Ė this is why I mentioned the early part Ė Blackett looked with favor on two of his ex-service undergraduates doing -

End Tape 73B

Click start to listen to the audio for tape 74A of the Jennison interview.

Begin Tape 74A

Sullivan

Continuing with Jennison on 31st August 1976. You were saying that Blackett -

Jennison

Yes, Blackett provided us with the minimal, well, not with minimal, but with sufficient facilities to do the job. Obviously, we were not a full research team, but we were able to use the university van to get it to the local countryside away from manmade interference; and we constructed an aerial system very, very cheaply, indeed. An aerial system, remember, at 7 cycles a second, so this was really a (?) coil with an iron core. Menzel and Salisbury, if I remember rightly, used a big loop of wire very near to (?)

Sullivan

Oh, they had actually done something experimental?

Jennison

They had done something experimental.

Sullivan

I thought they just suggested it. Iíll have to check that.

Jennison

Well, I think that they had actually taken a big loop of wire, and picked up something with it. What I did was use a, I mean, one was already in hertz and dipole region, so it didnít really matter much what you did. And I made a compact, not nowadays, rather like a (?) aerial with packed galvanized iron wire, cut off two feet long, (?) strands which (?) that length, a big coil of wire around it, and then a synthetic receiver. I say it was synthetic because one in those days had not many of the modern techniques. We had a very lossy aerial coil, so I increased its Q by applying regenerative feedback to it. So the Q was obviously increased very high. But then I had to tune it. So to make the capacitor to tune it, I used the Miller effect and synthesized the capacitor. So one had both the inductance was artificially queued up, and the capacitor was artificially expanded in size by good old, well, very classical electronic techniques. And that was a two-circuit, the rest of it doesnít really matter. We used that apparatus in the distant, well, rather distant, parts of Derbyshire by using Blackettís van, and we detected various signals, but we were having trouble. There was still interference. Blackett recommended that we should try it at a place called Jodrell Bank, which was the Botany Department of Manchester University, where one of the staff, who was then a lecturer, perhaps, a senior lecturer, a chap called Lovell, had just started to put some aerials into a field, or had already got some aerials in the field and was doing some experiments on meteor astronomy where the environment was quieter than elsewhere. So we took our apparatus to Jodrell Bank, and carried on our observations there quite independently of Lovellís work or anything else that was going on Ė but with the facilities available to us by grace of Blackett. Well, Lovell through Blackett.

Sullivan

So was there anything published of this? A report or anything?

Jennison

Only a report to Professor Blackett.

Sullivan

Do you have a copy of this per chance?

Jennison

I really donít know. I think I might, actually. Iíve got a feeling that a few years ago in turning over some old papers, I did come across it.

Sullivan

Iíd love to see it if you have it.

Jennison

I can just tell you, briefly, that most of the signals we received were from fairly distant thunderstorms. In fact, very distant thunderstorms, I believe. And there did seem to be possibly due to (?) resonator effect going on, but I, without referring back to that very old paper, I couldnít remember. This was a contact with Jodrell Bank; and as a result of that, my colleague, who was a chap called W. A. S. Murray, and I, decided to continue, if we could be accepted, to continue doing research at Jodrell Bank after graduation. And that, indeed, we did. W.A.S. Murray turned up at Jodrell Bank to start research shortly after I did, a week or two later, so that I had first pick on one of the two research projects which were available. One of the research projects was on lunar radar; the other was concerned with a newfangled interferometer that Hanbury Brown had just thought up.

Sullivan

This would be 1949?

Jennison

This would be 1950. And so I opted for the interferometry. It just intrigued me because I liked techniques one way or another, and Iíd gotten fed up with radar. And it just seemed something different. So I opted for that one, and my colleague, W.A.S. Murray, did the moon radar. Have you, incidentally, just switching quickly, been in contact at all with W.A.S. Murray?

Sullivan

No.

Jennison

Do you have any reference of his work?

Sullivan

I do have references. Where is he now?

Jennison

He is now the head of the department, not the head of the department, but the equivalent. Heís the, whatís it called now, well, itís the head of the physics division at RORDE. Ė Civil Service Royal Ordinance Research and Development Establishment, not far from (?), Fort Halstead, near Seven Oaks in Kent.

Sullivan

Okay. Iíll get that address from you afterwards.

Jennison

He, incidentally, had a lot of trouble on the radar. He was getting very worried. But then everything seemed to fall into place; he found, I think he was the first one to do this Ė Bay had already got lunar radar results, but it was Sandy Murray who, because he couldnít understand why his signals were doing funny things Ė found that it was a Faraday effect, on the effect of the signals coming in.

Sullivan

In the ionospheric?

Jennison

Thatís right. It was a Faraday Effect. It was just the polarization (?)

Sullivan

They were using linear feeds.

Jennison

Yes, they were using linear feeds for the first things. So Sandy put it down to the Faraday effect in the apparatus.

Sullivan

Who was it that really got the lunar radar going at Jodrell?

Jennison

He was the first chap to get the lunar radar results at Jodrell. But he was followed, he was preceded, by a chap called, yes, Evans went to town with it, but it was Sandy Murray who actually got it going.

Sullivan

I see.

Jennison

Yes, Evans went to town with it and refined it, and sophisticated it.

Sullivan

Right. What about this weird interferometer that Hanbury Brown -

Jennison

Well, Hanbury Brownís interferometer. This was fascinating. I was intrigued by the principle, and I was intrigued by the challenge of electronics of it. I was still, to some extent, a bit of a radio engineer. Iíd been a radio ham, incidentally; I think I still was at that time, and I liked the idea of doing things like that. So I got into that. But Hanbury had put it to me, he left this almost entirely to me. One often wonders whether one should leave research students to themselves, or, you know, guide them all the time. Hanbury left this, surprisingly, to me, in the circumstances, in retrospect. He was doing experiments of his own at the time on the 218 foot telescope, and was quite absorbed with those. And so that the actual work for the development of the principle that he enunciated was left almost entirely to myself. And so I constructed a prototype instrument, and we tried it out on the sun. This would be in 1951. 1951 was roughly a sun spot minimum, and we expected the sun to be quite quiet. We Ė I designed a set of aerials for this interferometer which were very large Ė they were 120 feet long and 40 feet wide, and they could be tilted over towards Cygnus, 15 degrees from zenith latitude, or towards Cassiopeia which was 5 degrees the other way. And this big bedstead array had been made in such a way that it could be transported to various sites in the countryside. And this was quite a challenge, a bit of cheap engineering.

Sullivan

Now this was to get variable length?

Jennison

Variable baselines.

Sullivan

And also were you thinking that variable orientations might be valuable at that time?

Jennison

I hadnít thought of that. The variable length baseline was the main thing, at that time. But the main challenge was how the heck do you produce an aerial that size? It had to be that size because, as you probably are well aware, the intensity interferometer depends very much on a strong signal. The signal noise ratio has to be essentially the order of unity or greater, otherwise you start to lose severely (?) square.

Sullivan

And you realized this from the theory even before?

Jennison

Yes, indeed. I think Hanbury may have pointed that one out to me. Anyway, I had to make very big aerials Ė much bigger than those used by Mills, and so on, when he did his extended baseline surveys at around the same time.

Sullivan

Right.

Jennison

And so these very large aerials were constructed in such a way that they could be collapsed. They were, dammit, Iíve forgotten how many, but they were linear broadside arrays of many, many elements Ė 120 megacycles, I believe, and the whole pack of cards could be collapsed and dismantled and packed onto a single (?). And we could move it around the countryside that way, and then reassemble them. Later we were able to reassemble them in about one day, even connecting all the feeders. Initially, I think it took about two days to assemble it. Now we used a quarter of these big aerial systems for the solar measurements to establish whether or not the intensity interferometer principle would really work.

Sullivan

Right. (?) baseline.

Jennison

That baseline was trivial, hundreds of yards, or a hundred yards, that sort of baseline. They were about a hundred yards, if I remember rightly. And we measured the angular diameter of the sun. At least we tried to, but there were sunspots; and even the smallest sunspot gives you an erroneous answer. And so we did it, actually. I had thought we published that result, but we probably didnít. (?) lots of other things; I donít think we published it because somebody else, was it Bateson?, you know, at Cambridge, had been measuring the angular diameter of the sun. But it was awfully early. It was quite a good one.

Sullivan

(?) sunspots, you mean.

Jennison

Well, it was the sun because we waited, and thatís why we took so long; otherwise weíd have been long ahead of the other chaps doing the Cygnus and Cassiopeia measurements. Because we were troubled by these darned sunspots, and we wanted to make quite sure that the fundamental equation of the intensity interferometer was really right. Nowadays, it seems obvious that it was right, but in those days you had to be quite sure, so we really had to make sure that our (?) on the sun really did go down to the zero (?) And this took a lot longer than weíd expected.

Sullivan

As you were working on this, did you have any doubts that the thing would finally work?

Jennison

No. I had trouble initially, the original plan was just to multiply the signals together coming out of the (?) interferometer, and we tried all the principles of multiplication of signals as we had in those days. And we couldnít get a really positive result with that. Then we decided to Ė [long pause] I think you really call it phase switch, yes, a different way to, this was phase switching actually before Martin Ryleís, but itís different. We werenít phase switching at RF. What I did, I put a switch into the lead of the, just before the signals multiplied together, I put a switch, one of these quick make-and-break switches, into the lead of that, and I reversed the sense of the signal. And I spent about two hours one day, whilst the sun was transiting, with my watch in one hand and the switch in the other, flicking the switch every two or three minutes. This way, and then that way, and then this way and then that way, and out came a beautiful record in the form of a square wave.

Sullivan

You were beam switching, really.

Jennison

Not beam switching, phase switching, actually, strictly phase switching Ė but not synchronously rectifying the phase switch. The signals were adding in one sense, subtracting in the other sense, but instead of synchronously detecting, which you do one real phase switching interferometer, and gives you this sort of result.

Sullivan

Oh, I see.

Jennison

I just saw the sun, and the difference superimposed on the bump due to total power. So the bump of the total power now had a big square wave going through it, almost up to the top. And we measured the depth of modulation on my switched wave form (?) the switch one way and another, and the depth of modulation gives you the visibility.

Sullivan

This is very analogous to Ryleís.

Jennison

Yes, itís extraordinary. Now I think about it now, it is the same thing apart from the fact that we didnít put the synchronous detector at the end.

Sullivan

No, no -

Jennison

Not at that stage, we did later.

Sullivan

Was this used, I mean, is this published anywhere, this switching?

Jennison

No, only in my thesis and in Das Guptaís thesis, probably.

Sullivan

What about this fellow Das Gupta Ė you havenít mentioned him yet.

Jennison

Oh, well, yes, sorry I should have mentioned him. I had been on the job for about three months, two months, perhaps, when I was told that an Indian fellow was coming along, a chap by the name of Das Gupta and he would give me a hand on this, and this was Das, and we became very, very firm friends, and he is a very nice chap. And we worked together throughout the whole of this phase on that system. Das, coming from India, loved to go out and work on the aerial systems when it was snowing or freezing cold.

Sullivan

Which you didnít object to.

Jennison

Which I didnít object to. I used to love to go out into the sunshine and work on them then, and spend the rest of the winter fiddling with the electronics inside. And thatís the way we subdivided the operations. Yes, Das and I were together on the whole project.

Sullivan

Well, back to the experiments.

Jennison

Back to the experiments.

Sullivan

Did you get fringes immediately with this thing? I mean, as soon as you got it working to the stage you thought you should get it to?

Jennison

We had no fringes until we used this switch, which we threw. Sorry, the system doesnít produce fringes, no obvious correlation.

Sullivan

Well, okay correlated.

Jennison

No obvious correlation unitl I put in the switch, and then yippee! Ė we saw the thing really working. It was marvelous, really, and there it was. We actually got the intensity interferometer doing its intensity interferomenter thing. And also, we got, well, the whole (?) and so at that stage, then, it was a case of refining.

Sullivan

Right. When was this one first working?

Jennison

It must have been sometime in 1951; Iím sorry, I canít tell you without looking back in the old records, sometime in 1951. And then, so we had the system going in í51 or sometime around then. We couldnít go straight on to Cassiopeia and Cygnus because it had to be proved that it would measure the angular diameter. You had to prove the equation Ė the visibility function is the square of the normal visibility function Ė you had to prove that. We had the system going; we had to then measure that. Thatís where the delay came with the sun. But, immediately after doing this switching business, of course, we then refined it and put a phase-sensitive detector on the end, and we got a bump coming out in the usual, clean, more sophisticated way of ordinary systems nowadays.

Sullivan

Right. Were you aware of Ryleís work when you made that improvement?

Jennison

Not then.

Sullivan

Was it only a matter that when the proceedings of the Royal Society paper came out that you heard about this? Or did you hear about Ryleís phase switching from some other channel?

Jennison

I think he came - Iím awfully sorry, I cannot remember.

Sullivan

It came out in 1952.

Jennison

This was certainly before that. I think it was, one was influenced much more by Dickeís work. All we did was do what Dicke did. Itís the same principle; I donít think it was Ryle at all, he didnít come into it.

Sullivan

That was on microwaves, yes.

Jennison

It was the Dicke system, actually, but we were using it to switch the phase rather than to do that. But we werenít doing, as I say, quite what Martin Ryle did. It still happens that the technique is almost the same thing. We were doing it right at the end of the system, where Martin Ryle did this through the front end of the system, you see.

Sullivan

And all of the solar work that you were doing to check out the principle is not published anywhere but your thesis, is that right?

Jennison

Thatís right.

Sullivan

Iíll have to check it.

Jennison

There was, I think Lovell was a little bit reticent in those days about publishing that sort of stuff, and as I said, Bateson already published (?) the sun and so as far as I recall, we didnít publish it.

Sullivan

What was Lovellís criterion for something that should be published?

Jennison

He was a hard taskmaster on that. I remember I was put off writing papers for many years because he was so critical before papers could be released. Itís very good, actually, but one was held up and this is a little later on, but one tended to, perhaps he went a little too far.

Sullivan

So what youíre saying is that he had very high standards for the (?) and also that it had to be something that was sort of significant?

Jennison

Thatís right. And he would be hesitant, certainly sometimes something was highly original, perhaps in case it might be wrong.

Sullivan

It had to be checked out in many ways?

Jennison

Thatís right.

Sullivan

But what about this paper as an example though, your 1952 paper.

Jennison

Yes, well, letís see now.

Sullivan

Thatís rather highly original, to say the least.

Jennison

Well, he believed it! Thatís all. [laughter] I think thatís all it was Ė he believed it. We spent a lot of time doing the sun two or three times a day, but he just reappraised it because thatís why we were so long getting going on Cassiopeia and Cygnus, which was the main object. But having, oh, in order to do the survey, which will fix the date pretty well for you, we must have done the sun during the summer of 1951 and the autumn or fall of 1951 into the early winter, at least of 1951. The reason I can fix it is very simple, my aerials designs (?) had some tripods made, some wooden tripods, perhaps 6-feet high so that I could put this center point of the (?) 15 degrees one way and 5 degrees the other. I was going to put center points up on these six-foot tripods so that (?) to see the sun.

Sullivan

Right.

Jennison

But as the weeks and months went by, the sun got too low, and I couldnít get my (?) So I think I even put tripods on top of tripods. That ultimately was not sufficient, and I had to dig a trench, or Das Gupta and I. I occasionally recount this to my research students nowadays Ė what they must be prepared to do Ė we dug a trench, ultimately, about six feet deep and three feet wide in the soil on one of the fields of Jodrell Bank. At that time it wasnít our field, it was a field adjoining Jodrell Bank owned by one of the farmers, but we got his permission. We dug this big trench Ė 40-feet long, or thereabouts, six feet deep and two or three feet wide and dropped the (?) into the trench so we could see further along the (?) Well, and then, we (?) runs on that. So that would place it at the end of 1951 when we completed the solar runs Ė I think we might have gone a little bit into the following year. We then had to construct the larger aerial systems, the rest of the aerial systems, to use on Cassiopeia and Cygnus, and so during 1952, we, I had to make, thatís right, radio links. I had to send the signals back by radio beam from the distant side of Cassiopeia and Cygnus.

Sullivan

Oh, Iíd forgotten that that was involved.

Jennison

Yes, of course. Because of the sun, we could do it by cable. Now whether or not I built the radio links so (?) that point and use it over fifty yards, I really donít know. I rather doubt it, but I may have done that. But I certainly had to build that radio link to bring back signals from the proposed Cassiopeia site.

Sullivan

Did this create problems?

Jennison

No Ė Iíd been a radio ham; I loved the job. And I was talking to Bernie Mills only a few days ago, and I was just mentioning something about radio links; I have problems with those members of my staff who from Cambridge who wouldnít touch radio links, and how we used to throw them together in the old days. And Bernie was agreeing with me entirely Ė one just did it, and one would still do it. And thatís the point Ė I did it; it was a nice bit of fun. One just made a radio link. But I thought that I had to go fifty miles or a hundred miles to measure Cassiopeia or Cygnus; that was the thought in those days.

Sullivan

Based on what?

Jennison

Based on the fact that everybody was thinking these were stars Ė these were the radio stars; the word was in use then.

Sullivan

Right.

Jennison

These were radio stars because you couldnít revolve them. Therefore, you had to go at least fifty miles or something like that.

Sullivan

Maybe you didnít know about it, but Smith came out with his precise positions in September of 1951 or so, Summer of 1951 in Nature. Very soon thereafter, Baade and Minkowski got the identifications, which were well, Cygnus was, of course, still a very small thing optically, but Cas was sort of a filamentary thing over the largest region. But you were still thinking that - ?

Jennison

Yes, the first indication we had of Cassiopeia was Dewhirst, Cambridge. It was (?)

Sullivan

Oh, thatís true. Dewhirst was -

Jennison

Dewhirst was the chap, it wasnít Baade or Minkowski.

Sullivan

Well, youíve got -

Jennison

(?) He knew there was something extended there.

Sullivan

Right, but it was all from the radio position and therefore, (?)

Jennison

Well, I think I thought at least that Dewhirst was onto something. If youíll forgive me, I donít remember the date, Graham Smithís positions were not important to our work. They were important to Baade and Minkowski.

Sullivan

Yes.

Jennison

And to Dewhirst.

Sullivan

But the point was that there was an identification with some wispy extended stuff. You must have known about that, and still you were thinking Cas A was a -

Jennison

What I probably meant there was the date, when the Baade and Minkowski paper came out.

Sullivan

Oh, the paper was much later. But the news, September of 1951 is when they -

Jennison

(?)

Sullivan

Yes, they went right on the telescope, the 200 inch in the fall of 1951.

Jennison

Okay. This could well be so, then, sometime in the fall of 1951. Yes, it could be so. But sorry, it doesnít detract from what I was saying before. We had thought that weíd need long distance radio links.

Sullivan

When you were starting out?

Jennison

Yes. And there was no reason to change those plans, really. So the radio link was capable of doing much more. I think in later years, I remember a question of Hanburyís where we did resolve the (?), he said weíd use a steam roller to crack a nut. They had something that was much more powerful in the end, and we were prepared to go further. I donít remember exactly when the news filtered through about the optical observations or how it influenced us.

Sullivan

But you may well, even though you have this filamentary stuff, you may well have said the radio source is still a very small thing, and hereís this optical film thatís around; they donít have to be the same size.

Jennison

Yes, we probably did that. Itís too long ago, Iím sorry. I canít remember. Okay. Well, sometime in 1952, and again, I canít remember when, we went the whole hog Ė we put our first aerial system up in the, big aerial system, in the grounds of Lovellís house at Quinta, which is three and a half kilometers south of Jodrell Bank. And we tried it over that baseline, from a small cornfield, really, I suppose it was, to the site of his house. We set up the aerial system there, and weíd since been sending radio signals back by radio link and we got the whole thing going over that baseline, and on that baseline, let me see now, well, we got Ė what did we get? I think that Cygnus did not appear to be resolved, probably, to any extent, and that Cassiopeia was looking as though it was somewhat resolved. (?) in retrospect, I could probably work it out, but it was three and a half kilometers south. The (?) was that we (?) the diagram in the Ď51 paper, Ď52 paper, that first measurement would be, this one of Cygnus and Cassiopeia quite well down on the ring. And we made that measurement, kicked everything out, and thought, ďWell, okay weíve got something, but is it?Ē I mean, literally, I think we just thought, ďthat was it like other ways?Ē We just directed ourselves that way Ė well this is quite different from what Graham Smith and Mills had done. They had carried on in the same direction.

Sullivan

East-West.

Jennison

Thatís right. We got that measurement and thought, ďWell, whatís the next sensible thing to do?Ē And instead of going along the same line, we turned around. And then we decided to (?) this was three axis of course, 60 degrees. So we then looked for suitable sites at the same space distance, I suppose, three and half kilometers or something, probably it says here. It must have been, I think, the same distance away. And we turned around at the same distance away. The great problem, of course, was to put a bigger aerial system up in the direction in which you wanted at the baseline you wanted when youíd land right on top of some cows or some farmerís field, or whatever it was, that you didnít particularly want to. But we managed to do it and get it sufficiently near for the right positions, and did measurements in these two other directions Ė Cassiopeia made sense, we got this roughly circular distribution. Cygnus didnít make sense. The three measurements did not fit together, but we had confidence in our measurements; we were sure weíd done them right, so the fault wasnít ours Ė it just had to be ďwrongĒ somehow Ė there was something peculiar about this result.

Sullivan

What exactly do you mean was peculiar?

Jennison

Well the fact that the measurements on the three baselines did not fit a single Ė what should I call it? Ė a distribution.

Sullivan

An elliptical shape.

Jennison

An elliptical shape could not be fitted. And in fact, no ordinary, regular singular shape would fit the result. And that, in view of that, at least (?) measurements. Again, I think because one had faith in oneís own measurements. Whereas someone else looking at it might say, ďThey made bad measurements, those donít make sense Ė that oneís too short and the other two canít fit.Ē But the fact that if you take a single object with only one maximum, and you rotate around it, you must always get a projection of that distribution. That wouldnít work with the measurements weíd made; and with the confidence in these measurements, something else had to be the cause. And just going back, one, I notice that in these measurements, they were, for what they are, remarkably correct. The visibility function that one measured was remarkably correct. It just didnít make sense.

Sullivan

Right.

Jennison

Cassiopeia again is quite remarkable. I see we quoted in this paper, ďthe equivalent angular size of the source appears to be about four minutes of arc.Ē I think the current value according to (?) is around about 4.1 minutes of arc. It was extremely close. And it is roughly circular. It was in the later paper that I think I pointed out for the first time these phase-sensitive measurements I did later (?) was limb-brightened, which was long before Martin Ryle was to do his synthesis on this. That was Cassiopeia. Now Cygnus, something funny was wrong there. So, Iíve told this tale before and it sounds silly, but its perfectly true. I took a long bath. I know it sounds like Archimedes, but I took a long bath and I lay back in the bath thinking about this distribution. Well, all of a sudden, it clicked that if this thing, Cygnus, were two blobs instead of one, then I could get this peculiar difference between the projections in different directions.

Sullivan

Right.

Jennison

Now, you see, it seems rather remarkable that one got it from that, whereas nowadays one interprets that from usually such things as height of the second maximum in the transform, etc. But my information was from an entirely different source, just making sense of an irreconcilable distribution in three different directions.

Sullivan

Sure.

Jennison

So I just thought of the model that did it. And then, presumably, having gotten out of the bath, I probably did a quick calculation and tried out what would happen if I took a simple distribution like that, a model fitting it and I found it would predict that way indeed. And then the next thing to do was to see whether I was right. So we decided, where according to these three axes that were taken, where the most likely major axis lay. And we decided it would be around about 113 degrees, I think. Anyway, about 113-120 degrees. As far as I remember, itís right about there.

Sullivan

And then what was the test observation that you were going to do?

Jennison

So then, the thing to do was now find the distribution along that axis. That comes after this first paper, but perhaps at this point, let me just mention my colleagues elsewhere in the world.

Sullivan

Right. Did you know about their work?

Jennison

We knew on the grapevine that Mills was doing something (?) Australia. Oh, Mills is building an interferometer also to look at Cassiopeia and Cygnus. No, to look at Cygnus, Iím sorry, not Cassiopeia; he couldnít see Cassiopeia, he could only see Cygnus. Cambridge had always been very secretive; it was very difficult to find. This has always been so, and will always be so, I think. Good luck to them, but thatís the way they work. We knew that Smith was doing something or other, he had been doing the (?) measurements. It sort of filtered out that he was probably looking at the angular sizes, but on pretty short baselines. The top bit of the transform. I donít remember much more detail, just that we thought that he was doing something. Now, what did happen next?

Sullivan

Well Hanbury Brown has told me that -

Jennison

(?) Australia. He went to Australia.

Sullivan

- the URSI meeting, he set up three papers and it would be nice to publish them together.

Jennison

Yes. What happened, I think, was this. Probably Smith, well, Mills was obviously at the URSI meeting; Smith probably went there. Hanbury certainly went there, and when Hanbury left for Australia, we hadnít got a measurement. We were all ready to take it, but we hadnít actually got the measurement. And so we promised Hanbury that as soon as we got something, weíd send him a cable. And we sent Hanbury a cable about half way through the URSI meeting, I guess it was, quoting what we measured for the two sources.

Sullivan

I see.

Jennison

And he received that in Australia, and then, I think the decision was made at the URSI meeting that all three should be published together. But it just arrived in time at the URSI meeting, by cable. In fact, I remember Hanbury sending back querying, the word Iíd used, which in retrospect was an ill-chosen word. I said, ďEquivalent line diameter of the sources,Ē or something, but I intended that the ďequivalent strip diameter.Ē And I used the word ďlineĒ and Hanbury put that in (?) right or something like that. But that was the state of the art then. Which probably takes us now to Cygnus distribution. Does it now?

Sullivan

Well, I was just wondering how you looked upon these other results when they became available to you in late 1952. Did it all seem like everybody was agreeing?

Jennison

Oh near enough, yes. Bernie Mills had been unfortunate if you like; he jumped too far in his -

Sullivan

Exactly.

Jennison

And -

Sullivan

He almost got it.

Jennison

He almost got it; he went too far and assumed something, a little training from was it (?) I thought myself, but that model (?) jumping to conclusions with plots. Itís very, very easy to think that something is due to an error sometimes, or in your own work, and not to attach significance to something thatís not quite right. As with the Cygnus distribution, itís very, very important sometimes. Iíve seen, Iím talking generally now, Iíve seen chaps miss very important results because they decided that their measurements werenít quite as good as they thought they were, and theyíve drawn the mean line through the points, instead of making that little dip or little peak or whatever it was and having convinced that theyíd done the right measurements. Itís very important.

Sullivan

Supposing Jansky hadnít followed up this little Ďhissí he made.

Jennison

Thatís right. Itís very, very important, the principles of measurements and so on. But if one really believes that one has done the job properly, believes oneself, then accept it and see what it means.

Sullivan

So do you think that may be a component though, is that many people just donít have the confidence that they really have been that careful and therefore -

Jennison

Itís surprising how many people do not have confidence; and over the years, since these early days, the number of times Iíve had a, for example, very good (?) members of the staff who are very good in the sense that they are marvelous analysts, who would never dare, do you understand what I mean, who would never dare to push something because of their own convictions, or whatever it was, because they thought that somebody else must have thought of it first, or you know, and a large number of times, some of the best chaps that do the analytical sort of work, they havenít got that curious flare that says, ďdammit it all, I know Iím right.Ē Youíve got to go that way. However away from the mainstream it may seem to be.

Sullivan

Do you think experimentalists have this more in their blood, so to speak?

Jennison

Yes, but again, some experimentalists, and all those experimentalists, especially. For example, you take radio astronomy, the present day radio astronomy, so much of it is just plugging into a big instrument and plugging out the data. Itís a very, very different world except for the odd little things that are still going on here and there. Thank God. And well, if you like, the pulsar business; they believed in the end that those were real (?), and so on going around in space, same old story. You could just as easily read it as interference, or going back Jupiter. There have been many, many cases and a lot depends, I think, in these cases, who is there at the time. If we (?), I mean, so many (?) some people may have missed these things because theyíre the wrong type of person to appreciate that the signal concerned, the (?) concerned, were really significant.

Sullivan

Yes, theyíre more concerned with getting the data to fit their frequency.

Jennison

Thatís right.

Sullivan

- than really seeing what the data says.

Jennison

Yes, thatís right. And we are drifting completely, put you on (?)

Sullivan

This is very useful information indeed. Well, now you followed this up and went along the expected position angle.

Jennison

Yes, with the intensity interferometer, went along the expected position angle, and indeed, up popped the second maximum, as predicted by my model. And the first minimum we looked at rather carefully, and it did seem to be very low. This was fairly important because one of the main differences between the two and three body problem is the depth of that minimum, but also the symmetry of the source as gathered by the depth of that minimum.

Sullivan

So you were still open to the idea that there might be three or four components?

Jennison

Well, one had to be certain if one was going to publish something about it being two components, that it was two components. And with the intensity interferometer, you could only go as the square of the visibility function, so you could not investigate the minimum so well as you could with an ordinary interferometer. For example, at ten, what would it be now, the square so, if we got down to 1%, we couldnít attempt it (?) interferometer, so one had to measure very carefully there or otherwise it could be checked out by somebody else with a different system that you (?) So, and again, with this Iíve already mentioned how careful one had to be about publishing in those days; one had to be quite sure. So we spent some time thinking for quite sure, as well as we could with that system, and I did something then for the first time, I think. This was a little first in radio astronomy. I couldnít move my big aerial systems very easily over short distances, so I did a trick Ė I changed frequency by a small amount.

Sullivan

Oh, I see.

Jennison

Right? I changed my frequency about 5% either way, and this gave me the equivalent shift. I was talking to somebody at this conference, Iíve just forgotten who it was Ė was it, oh dear, the chap who first discovered these, who is he?

Sullivan

Burke?

Jennison

Yes, Burke, Bernie Burke. He told me he was using two of my things: he was using my three station interferometer system at that moment, and I think he also said he was using the baseline change. It may not have been he, it may have been somebody else. Sorry, it may well have been somebody else using a three station interferometer. Somebody else was using that same trick. It seems so silly, so trivial; it was just because it was so darned difficult to move the aerials that I was forced into changing frequency, which is all right as long as youíve got a well-behaved function of frequency, which it did seem to be. Now at that point, I was probably influenced by Mills and Smith. But as the results didnít differ greatly, it didnít matter very much changing frequency.

Sullivan

(?) spectra of the sources.

Jennison

Yes, yes. But we were given the same order of result, therefore, I could risk changing frequency slightly, and I did this along the visibility function. So instead of getting a single point, I got triplets - or usually triplets, maybe some were double.

Sullivan

Can we just look at -

Jennison

I skipped an (?) gradient. Incidentally, the gradients were very useful.

Sullivan

Can we look at the second paper to see if thereís anything else.

Jennison

Thereís a triplet, and you see the gradient. But donít ask me why we took that zero with that point just above it. Iím sure we had good reasons for it, but sometimes in a scientific paper, one does not publish all the feel one has Ė all the information one gets in oneís ďguts.Ē And we were absolutely certain that that went down. I remember arguing about it with my colleagues at the conference (?), that that was well under the 10% visibility function, then we talked about 1% in the square visibility function. I was quite sure of it. Because one of the chaps who didnít believe this result for years was Bracewell. Bracewell, of course, was very keen on his Fourier transform. I think from entirely the other point-of-view. If you take the Fourier transform of the objects, then you can have an infinity. (?) satisfy our points, whereas this was done the other way around. My model (?) But again, itís one of these curious situations where one is armed with a lot more information which doesnít directly come out; you have the distribution around the source, the first spot that (?) first of all, together with the visibility function, and I could not see that the three body problem which would fill this in or (?) the symmetrical one, would satisfy the results. I was absolutely convinced that it was two bodies.

Sullivan

You werenít trying to do two-dimensional Fourier transforms Ė that was too complex to think about?

Jennison

We thought of a model, analyzed it; you cannot work backwards with those few points. Thereís just too few; you just get too many possibilities. Itís far, far better, I think, to start with a model because the model has to satisfy this visibility function on the major axis as well as satisfying the projection along the three axes, and that restricts the model quite a lot. There are, I remember, (?) Bracewell about a possible (?) and in fact, I think it was at the (?) or some time or other, and Bracewell actually claimed there were no (?) in the middle of Cygnus.

Sullivan

What symposium?

Jennison

May have been the Paris Symposium, 1957 or í58. I remember claiming he was quite sure heíd done some measurements at Stanford on this thing at a higher frequency, mind you, and it could well have been that at a lower frequency that thing was quite different Ė a point Iíve always made. These measurements were made at much lower frequencies (?) But he was, it was, a long battle with Bracewell, I think. It wasnít until the early sixties when so many double sources were discovered, that I think he finally capitulated that this was on.

Sullivan

I have a little note here that you did not put in either of Millsí, I mean Millsí long spacing, or Smithís short spacing in your model fitting. Wouldnít it have been a good idea to have put them in? See how that works?

Jennison

What have we got? Weíve got something by, published by Mills and by Smith, what did we put in?

Sullivan

Well, itís only -

Jennison

Yes, thatís Smith.

Sullivan

But it was only one of Smith.

Jennison

Oh.

Sullivan

Smithís point at 36 wavelengths and Millsí at 3340. So you left out one of Smithís and one of Millsí, and I was just wondering why that was? I actually checked the details to look.

Jennison

Itís a marvelous thing to do. I really donít know. I think the 36 one would not have meant much, would it? Because down here, thatís 500 wavelengths, so youíve had hardly seen it Ė almost on the origin; I think itís feasible to leave that one out. Millsí second point is almost (?) Smith, Millsí point, I donít know except that itís a long way along. Well, I think the reason may be (?) It doesnít on its own so far away, how many wavelengths?

Sullivan

23?

Jennison

Yes. On its own, I think itís too big a jump to necessarily associate with this (?)

Sullivan

But you still want anything you do to be consistent with that point.

Jennison

Still, sorry?

Sullivan

You would still want any model you came up with to be consistent with Millsí (?)

Jennison

Yes. I canít think of any really good reasons, no. By the way, you can see on this how the intensity interferometer wasnít all that good with respect to (?) visibilities. Smith is it? No, whose is it? This is Millsí.

Sullivan

Thatís Millsí point, I see.

Jennison

Thatís Millsí, the old devil. He goofed. Yes, thatís odd to find out. Yes, that came up in the other paper, actually.

Sullivan

So you had this model; Iíd be interested to ask you how much did you believe in your deepest heart of hearts that there really were two sources there, especially considering that had been identified with this half arc minute?

Jennison

That was the fascinating thing. At that point, Minkowski and Baade were talking about a double (?) galaxy object, Cygnus, but their thing was almost microscopic. But the thing Iíd measured -

Sullivan

Right, you had expanding things.

Jennison

I had enormous things. Good Lord! So that it was like it was even more of a (?) thing; we were both getting a double object, but quite different double objects. So obviously, I was not influenced by Baade and Minkowski; I was saying something entirely different Ė very, very different. What I was saying was that out there in apparently empty space, the radiation seems to be coming from these two almost symmetrical blobs. Whatís going on? This is remarkable. There was absolutely nothing where the real signals were coming from, and the thing with Baade and Minkowski, you see, is coinciding beautifully with the ďcenter of gravity,Ē but itís not coming from that.

Sullivan

Why didnít this kill the colliding galaxies thing very quickly Ė it stayed around, as you know, for many years.

Jennison

Well, in those days, the big problem was how it worked. However could one get all that energy, anyway, in the nearest, the best source that one could think of at the time was the colliding galaxies. We didnít have enough energy, that was all.

Sullivan

But it would seem like if you had two things collide, the energy comes out there, not out here.

Jennison

Well, fair enough. But the answer to that is very simple, isnít it? Even at this conference now -

End Tape 74A

Click start to listen to the audio for tape 74B of the Jennison interview.

Begin Tape 74B

Sullivan

This is continuing with Jennison on 31 August 1976. So you were saying that even today - ?

Jennison

Yes, even at this conference, they are still arguing, they still donít know the answers, as to what causes the two blobs on either side of Cygnus or any (?) counterparts in the sky. We donít know the answer. So I donít think you can blame us way back in the early fifties for thinking that the colliding galaxy was as good an argument as any. And I did spend quite a bit of time in those days, trying to think what was going on. And I think even my thesis thereís a dabble of one or two ideas of what might be causing it. I was rather favoring plasma. Synchrotron was just coming on the scene Ė I think I canít quite remember when this really came up, but I was trying to think plasma solutions as to how you could get the radiation. (?) we thought that the sun was probably plasmic a bit, and one came up with some jolly good ideas, but they donít seem to be the right ones. And that didnít get published.

Sullivan

But the ultimate energy source, what were you thinking there?

Jennison

I think then we probably did think that galaxy was colliding; (?) Baade and Minkowski, two very, very great astronomers, were so adamant then, they really were adamant then, that these were colliding galaxies. I think I quoted them myself in a book in a paragraph on Minkowski quite a bit later, 1956 or something. No, it was the Paris Conference, Paris Symposium, thatís right, where Minkowski was absolutely adamant that it be colliding galaxies, and how it was proved that these were colliding and so on.

Sullivan

Yes, the two line systems and so forth.

Jennison

So all right, we had to defer to the experts, these chaps that knew what they were doing.

Sullivan

Were you thinking at that time of these things expanding away from the galaxy somehow as a result of the collision?

Jennison

I think I probably did. There had to be, really, because (?) sort of lifetime as a result of that, which was to, I think, what this implied. Certainly, we did. There seemed to be no other way out.

Sullivan

Any notion of looking for proper motions or anything like that?

Jennison

Not that I recall.

Sullivan

Yes. I think probably that could be ruled out when the calculations (?)

Jennison

We were just intrigued. I remember talking to Professor Blackett at that time, and I remember the time of these early measurements. Blackett was still at Manchester, and he was senior to Lovell in those days. And I mentioned to him the different measurements from Cassiopeia and Cygnus, and the fact that one seemed to be possibly just about the time for the observations to be made, as well, of course, it would have been (?) And Blackett said something Iíve never forgotten. Because there are arguments that the Cygnus source and the Cassiopeia source could have entirely different origins or raisons díetre or whatever you like; and Blackett said, no, I think it was Lovell, or maybe Lovell and Hanbury Brown were reeling off all sorts of things, and Blackett said something that sort of stopped me in my tracks. He said, ďNo chaps, you canít have that Ė too many criminals.Ē Now Iíve remembered that all my scientific life, okay? And so often when members of my staff or others, they get involved in certain work and there seems to be a certain difference between things with a strong common denominator. I use the same expression now myself Ė ďcareful, my chaps, mind you, donít get too many criminals.Ē Because in other words, and I say still, that there are strong common denominators between Cassiopeia and Cygnus. Theyíre not the same all over, but if you like, the synchrotron radiation or something like that Ė those are common denominators. And Iíve just never forgotten those words of Blackettís. A very experienced experimenter coming out with something that sticks with you for the rest of your life. Donít get too many criminals.

Sullivan

Maybe Iím a bit dense, but is the meaning there that usually it takes only one criminal to commit a crime?

Jennison

Thatís right, yes. Why bother to have two criminals if one could do it?

Sullivan

Right. Now, when did you actually finally get your degree?

Jennison

My first degree in 1950 my Ph.D. degree in 1954.

Sullivan

And I see that Jennison and Das Gupta in 1956 in Phil. Mag.

Jennison

(?) was that?

Sullivan

Is where you write up the details of your results.

Jennison

Yes. Incidentally, Das got his degree a year earlier. I didnít entirely please Ratcliffe, who was my external examiner with my first (?). What Iíd done was (?) thinking that Das Gupta and I would get the same external examiner; Das had written up all the experimental results so I avoided the experimental results and wrote up about technique entirely. And the first account, (?) and things like that, and a lot of other interferometers that have since come on the market, you know Ė the single (disc?) multiplier, multiplying grating, and things like that; youíll find in my thesis, but Iíd gotten intrigued with the techniques. But I had to put another chapter in to please Ratcliffe Ė so I was a year after Das getting my actual PhD degree. But thatís quite alright. But you were saying something else.

Sullivan

I was just saying that this was the write up of -

Jennison

Yes, and it was very late because Das Gupta (?) got brought in, I think (?) But these -

Sullivan

These techniques youíre talking about, thatís never published anywhere else other than your thesis?

Jennison

Just in my thesis, yes. And the odd ones, I think, thereís a book I did on radio astronomy Ė some of them are mentioned, not all of them. In those days, one went right round the houses, if one was interested, in such things as thinking up new types of interferometers.

Sullivan

Yes.

Jennison

And the place of interferometers, I thought of necessity, because the measurements were just disgusting on Cygnus, for example. First of all, they were made with the intensity interferometer, which was not very sensitive to low visibilities, and so one wanted to use a Michelson interferometer. But how the heck do you use a Michelson interferometer on those sort of baselines, in those days, at those frequencies. Going ten miles or fifteen or twenty miles ultimately, how could you do that in the circumstances when you could not have a surveyed rail track Ė you could not put cables through, in between the legs of the cows. You had to do something different. And so I somehow or other thought up how to get around that problem and came out with the solution. I think, if one had three apertures, multiplied the three apertures together so you got three different systems coming out, then by something which even today I can hardly credit, actually. I think it was quite remarkable that three (?) interferometer, if you put an error into any one of the channels, it disappears. Itís extraordinary. Itís quite remarkable. You put in a wedge of ionosphere, you put an extra bit of cable, you change the height of one of the aerials, you can do almost anything to it Ė you still get the right answer, which is an extraordinary fact, and I surprise myself with that. It doesnít seem possible sometimes that it will get rid of so much. In fact, very much later than that, oh, I donít know how much later, the late fifties or early sixties, I remember Martin Ryle saying weíd never be able to measure the visibility function on very long baselines because in particular, the ionosphere being different over the stations. And I remember getting up and saying, ďYouíre not quite right, Martin, the three station thing will do it.Ē And he wouldnít believe it, but it is true. You can actually use that principle on very different (atmosphere?) on very different sites.

Sullivan

I think it has been in VLBI since.

Jennison

I think it might have well been. Itís certainly been used in optics. Now the (?) up to eight, which is rather pleasing.

Sullivan

Why did you publish this in Phil. Mag.?

Jennison

Oh yes.

Sullivan

Thatís also where Hanbury Brown and Twis published.

Jennison

Do you know why? You see, being a very impressable research student, one was impressed the most of all by, in those days, by the papers by the great man, Michelson, which were published in the Philosophical Magazine of 1880, or whatever it was. 1890. So one thought the place to publish that was in Philosophical Magazine, not realizing that nobody ever read the Philosophical Magazine.

Sullivan

You were thinking about future generations rather than present.

Jennison

Yes, the obvious place to put it was where Michelson put his.

Sullivan

Is that what influenced Hanbury Brown also?

Jennison

It might well have been, I donít know. I canít speak for Hanbury. It was the reason that I did it.

Sullivan

Okay. Well, what was the next step. You saw that the intensity interferometer had limited applications.

Jennison

Right. So I thought up the three station interferometer which had the advantage that it was very sensitive as well as being the Michelson interferometer, which was much more sensitive that the intensity interferometer, but it was also phase sensitive. So one could then, in principle, measure everything one wanted to about the source. What one could not do with the three station principle was to measure the absolute position of the source. What you get with that system is a reconstruction of any part of the sky without an absolute dating for reference. But the part of the sky that we construct is in itself, intact, you just canít place it as being in any particular direction, unless you have an identifying object included in that reconstruction, which does give you a fix Ė then itís easy. And so then I took on a research student called Latham, Das Gupta had left, back to India, and this lad called Latham I got to help to build up a new interferometer. This was also a radio link interferometer sing three apertures. We tried it out, first of all, on the campus at Jodrell Bank, just between the diesel engine house and my own lab, a distance of about a hundred yards. And then we applied that to the measurement of the source Cygnus, confirmed, indeed, that Cygnus was a doublet, and measured very accurately this time, and this does not appear in the published papers so much as it does in the actual the records we have Ė our personal records Ė measures very accurately the depth of the first minimum. In fact, we sat an aerial actually within yards of that, we moved the aerial literally only a matter of yards and found out exactly where that came.

Sullivan

I see.

Jennison

And we had a remarkable (?), with that big a system we only needed to use a dipole for that. I think we used a dipole for (?) bigger. But we found out very precisely where that came. And we did then the first phase sensitive measurements, I suppose, of any radio source. And therefore, the first synthesis, if you like, unambiguous synthesis up to the value of the points we got of any source.

Sullivan

Yes, in a sense, thatís right.

Jennison

It is true Ė itís the first synthesis actually, using the word in a different sense to Martin Ryleís wonderful synthesis where he actually does a real Ė he takes the results pouring out from his interferometer and then synthesizes, (?), but by the approach of model fitting, to get the same result, but it was a synthesis.

Sullivan

You didnít actually do a Fourier transform -

Jennison

Itís a mixture, itís very difficult to describe it. Itís a matter of feel. You do model fit, you change it a bit, you transform back; itís interplay between the two.

Sullivan

Itís a trial and error, so to speak, until itís consistent with the data.

Jennison

Thatís right.

Sullivan

But you canít always be sure that you might have found some minimum; that there might be several things that are consistent with the data, and youíve just honed in on one.

Jennison

Thatís right. This is where we come back to an earlier point really. That some peculiar thing, je ne sais quoi, I donít know what it is, and you see, the only point I think Iíve made is that this is the first time that, with the intensity interferometer weíd done the very first structure on the source. With the phase sensitive interferometer, we did the first well, phase sensitive structure if you like, the first unambiguous within the number of points.

Sullivan

You didnít have to assume symmetry -

Jennison

Yes, we got, we made sure that the objects were symmetrical. In fact, with that interferometer, we discovered something asymmetrical about Cassiopeia.

Sullivan

Okay, will you tell me?

Jennison

Okay, Iíll tell you. The Cassiopeia source at a fairly low level on our interferometer showed some, showed one or two things. It showed, first of all, limb brightened above a uniformly distributed disc. Fairly markedly so, and certainly quite sufficient for us to publish (?). We just dismissed it in one sentence. We said that we observed that the distribution is more limb brightened than that of a uniformly (?) disc, it was fairly clear cut. And also we observed that there was, apparently, a spur, to one side. Now when we first got that, we got onto Minkowski. I have a vague feeling that we gave some slides, some very early slides, to Hanbury Brown who took them to a conference in America. Maybe (?) or something, I canít remember what it was. And he showed them there, and Minkowski got onto it. Minkowski went on to Mount Palomar dish where (?), and looked carefully at Cassiopeia and found indeed, a spur, on the side of Cassiopeia, just as we said. And he got this thing (?) came right out. And well, Minkowski has written up all about that. But he saw it, anyway. The fascinating thing about all this was that we put Minkowski onto this, and we published it a bit belatedly. I mentioned already the delays in publishing in those days, but we did ultimately publish it. We made our measurements at a fairly low frequency, round about 130 megacycles, weíd made them a long time ago, 1954 I think it was, 1955. And Cassiopeia certainly seemed to show something going out that way. Now all I can really say (?) is that I believe in those early measurements. There was something to the side of Cassiopeia, Iím certain of it, but itís rather fascinating the way that, for example, Minkowski actually found something there. All subsequent work, or subsequent to my knowledge, till very, very recently, has tended to be at very high frequencies, and indeed, the ring, the limb brightening of Cassiopeia has shown up repeatedly in things like aperture synthesis or various (?) or bits of Martin Ryleís work. The extraordinary thing in the first, that one noticed, in the first of the beautiful pictures from Martin Rylesís synthesis was that where I said there was a spur, he had a gap in his ring. Just the same position, there was a gap. But this was the synthesis at 1400 megacycles Ė pretty high frequency.

Sullivan

This was in the mid-sixties then.

Jennison

Yes, sometime then. And all the later ones have shown more and more detailed structure, and well, (?) of course, (?) I was always sure that there was this spur sticking out to one side, so it meant one of two things, which I probably mentioned in the first paper, either that the spur was temporarily variable or spectrally variable relative to the rest of Cassiopeia. Because it was spectrum. And only in the last two years, two American observations have (found the point of this again quoting the early paper?). One, forgive me, Iíve forgotten the authors.

Sullivan

(?) You mean low frequencies?

Jennison

Yes, one was a VLB observation at 111 megacycles/75 megacycles in which the (?) something to the side of Cassiopeia, getting much more pronounced at 75 megacycles than it was at 111. This sounds a bit like a point of some sort. (?) with a different spectra than Cassiopeia. And they gave various, well many ambiguities in the VLB; they gave various possible positions, but they rather favored one which goes (?). The other observation was a very recent one, who was it? Itís been checked out at Cambridge in the last few months, that the spectrum of Cassiopeia at 35 megacycles is very much enhanced compared to that at -

Sullivan

I see. Integrated spectrum.

Jennison

Yes. Which is (?) so it may or may not. The moment, the only bit of radio astronomy Iím personally doing at the moment, is actually going back and using the old three station interferometer at 35 megacycles and to look again at Cassiopeia and seeing if there is a distribution out that way. It neednít be a spur; it can be a point source. And talking of ambiguities in readings, and looking at that paper by Jennison and Latham in the mid-fifties or whenever it was, that the spur is one solution. A point source to one side could give the same first order distribution.

Sullivan

Well, let me ask you about some meetings at this time. You probably were included in the 1953 and 1955 Jodrell Bank symposium, do you have any recollections of either of those meetings? What you might have learned from them or what the atmosphere was?

Jennison

The atmosphere was wonderful Ė you had a lot of people starting off a new science. And certainly as a new boy oneself, one benefited very much indeed from those early symposia. Now, what did we learn? I think probably the same ostensibly so even today, (?) meeting. One learned much more by meeting and talking to other people (?) than one learned necessarily from the formal presentations. I canít really tell you more than that. But I always remembered them.

Sullivan

It seems to me that this might have been a part of the problem between, the antagonism between Sydney and Cambridge, that it had so few opportunities to discuss things over a glass of beer. Would you agree with that?

Jennison

Well, at those meetings, one certainly got contact between the Cambridge people and the Sydney people and the Manchester people, but I think that youíre left with this fact: that Cambridge has always been secretive, historically. This is rather curious. In fact, Iíve got three radio astronomers on my staff at the moment, ex-Martin Ryle chaps, and even in my own laboratory, you can recognize quite clearly this group of three. They are apart. And they are a little group within themselves, itís quite (?) It was a heritage of Cambridge.

Sullivan

Were you implying that this was more the cause of the troubles that the great distance between Australia and England?

Jennison

I think it might be. Cambridge was never terribly anxious to discuss with others, to get help from others, on anything. They wanted to do it themselves and keep it to themselves. Perhaps I shouldnít have said that. No, I think I can. Because I think itís probably acknowledged by everybody. That Cambridge have always sat on things. And very seldom, it has happened occasionally, but only later on, in things usually when theyíll come by, and when theyíve got (?) things like that, or got combinations going on -

Sullivan

Yes, but theyíre only the exceptions.

Jennison

Yes, exceptions. On the whole, Cambridge tends to sit on things Ė thrash them out themselves and then come out with a bang. Good luck to them.

Sullivan

I suppose one could also say this is true because there was no great antagonism between Jodrell and Sydney.

Jennison

No. There was a common denominator. One was Taffy Bowen there, for example, as an ex-radar type and an old colleague of Hanbury Brownís.

Sullivan

(?)

Jennison

(?) radio boys, thatís true, isnít it?

Sullivan

They knew each other during the War, and so forth. So what was the next thing you worked on?

Jennison

Gosh. Got to think. Oh, right. I was interested in interferometry, in particular, the technique side, and for some reason, I think this is, well just a side comment at the moment that is interesting Ė I remember hearing Martin Ryle a long time ago say that he was really a radio engineer or something. One was very often drawn on by the challenge of the techniques. At certain points, you had the challenge of the actual astronomy, but very often it was the technique side that was fascinating.

Sullivan

Yes, Iíve had other people tell me this.

Jennison

Yes, itís rather curious, and still itís so in the fields Iím working in at the moment. But I find it most rewarding to think up a few things (?) At this time, in the early fifties, one had the history of Michelson before you and everything heíd done. A wonderful (?) and now you had a new medium Ė you had radio, and Hanbury thought up the principle, deduced it first if you like, of intensity interferometers, Martin Ryle had come out with the phase switch interferometer, and Iím talking about (?) now.

Sullivan

Right.

Jennison

And what else was there? Iíd come up with the three station interferometer, I think that one was really of necessity because I had to measure sequence somehow so that people like Bracewell would not argue with me, I had to think of some way of doing it, so thatís the way that came about, I think it was a necessity more than anything. But it gave one the challenge. What else can you do with interferometers that Michelson hadnít done? For example, he had the 3 station interferometer; he had the, thanks to (?), optical version that (?) anyway. So one thought of other types, and one I thought up, I still get queries of that because curiously enough, the theoreticians have never liked the results that I claimed to have gotten with it, and I never bothered to do anything more about it. I just thought Iíve got confidence that what I did was right; itís up to them to try it again, but donít just stick to your theory. If you donít believe me, try it. And this was an interferometer which I called a harmonic interferometer.

Sullivan

I see.

Jennison

It was an (?). And what I did there was to take signals coming in through one antenna system, and to put them through the first stages as usual, (?) bit, make it so that the first stage (were all?) linear, by the way, in this case. And, now, I miss out on a lot of the technique (?) but in principle, the signal goes into one antenna and is amplified somewhat, goes into a square law device and you take out the second harmonic. You can actually convert (down to IF, get the signnal down to IF?) but that doesnít matter, this is linear (?) But I pick up the harmonic of that signal, the second harmonic; then I have a second antenna which is tuned to the second harmonic, and picks up signals from the sky at the second harmonic. You bring these two signals together and correlate them. Youíre taking the artificially formed second harmonic of one signal and beating it against the celestially received second harmonic on the other side. And I tried it and I got wonderful results from, for example, emission interference, any (?) I cared to tune in, from (?) because the sensitivity of the system like that to a tiny (percentage?) of harmonics is incredible. It really goes to town as you can well imagine. Then I looked at Cassiopeia, and I looked at Cygnus, there was nothing happening on Cassiopeia and Cygnus of any interest at all. In other words, fairly white noise did nothing to it as you can check from (?) of course, (?) but anything like a Delta function did. Delta function leaves all the signals starting at the same distance of time, therefore the harmonics should phase (?) and come out with a bang. So (?) switch or something like that. Then I looked at the sun, and on the sun, the quiet sun did nothing. But over the period I used it, of about six months, I was doing it at the same time.

Sullivan

When was this now?

Jennison

It was during the time that Latham was there, towards the end of the time that Latham was there, I suppose. So it would be towards the end of the phase sensitive interferometer work, 1956? Thereabouts, roughly. On the sun, there were four events which produced wonderful harmonic correlation. I mean, I had on display, on a (?) recorder, I had the output from Channel A, fundamental, output from Channel B, second harmonic, and output from Channel C, the cross product between these two. And so often you get something that appeared on the first two that wasnít correlated. Four events, though, did appear, correlated on the second, sorry, on the third, then. And these ones did not show a straightforward correlation, that is, they didnít show a single uniphase bump. The phase rotated, in other words, it (?)

Sullivan

I see.

Jennison

So the the second harmonic wasnít quite a second harmonic Ė it was slightly out of tune. Okay, by the odd cycle, (?), or second, whatever, I canít remember, not quite in tune. So all I did with that was just to write it up very briefly in the Observatory saying that I made this type of interferometer, I tried it on various things, didnít say anything, except on the sun, and I quoted these things, quoted also all the events which appeared in the Quarterly Bulletin of Solar Activity. So they were, actually, solar events, it wasnít my imagination, Iíd checked them on our own solar records as well, of course, but they werenít very quotable (?) as real solar events. And those four had produced this correlation. Well, many looking like (?). And one can argue that if you have a type two type event, you (?) then where the second harmonic (?) the second harmonic, of course, is in a different type of index thatís being produced (?) to the fundamental, and is moving through it, so the Doppler shift is not the same from both. So the second harmonic will not be quite a second harmonic, not quite coherent as a second harmonic.

Sullivan

(?)

Jennison

And then Iíve had the odd queries about that recently, but every time there are any queries about it, they say, ďWe gather that the theoreticians donít believe you.Ē And this always amuses me; one day maybe theyíll try it. Just try it.

Sullivan

Nobodyís actually followed it up experimentally.

Jennison

As far as I know nobodyís ever followed it up. Thatís extraordinary. Anyway, that was one of the little things I did then. Then I moved to, there were many other things; I moved to the hydrogen line work, extra-galactic hydrogen line work.

Sullivan

Well, this is (?)

Jennison

Okay, sorry (?)

Sullivan

This is not when the big dish came along -

Jennison

Thatís right; I didnít like this (?) but there was a bit of confusion. What had happened was that, I was really an interferometer man by then, independently or under Hanbury, Henry Palmer had been working on another interferometer system, straightforward (?) interferometer stuff. And Lovell had more or less promised me that when the big dish came along that I could do interferometry on the big dish. I think Henry must have heard of that dish and didnít like it, so he must have prevailed (?) I think, because Lovell then tried to get me to take over the hydrogen line work.

Sullivan

I see.

Jennison

And I didnít really want to do it. I was much too much a lover interferometry. But I did, anyway, I took over the work on that. The main thing was then to, Lilley and McClain had come out right about then with the paper claiming that Cygnus showed extra-galactic absorption.

Sullivan

Right.

Jennison

So the big thing was to see whether Lilley and McClain was right or not. So -

Sullivan

(?)

Jennison

(?) literally hated that work. It took a lot of doing, because -

Sullivan

Did you think they were right? Did the data look good?

Jennison

I donít know. I think I was open minded. Canít remember. Just cant remember. But I didnít like doing it; I didnít like - . Perhaps I think we must have suspected they were wrong. But I never liked having to show that somebody else was wrong; it doesnít appeal. And also if you do it, from oneís own point of view, it means that youíve got to go a hell of a lot further than the other chap who first published. Youíve got to go not just one order of magnitude, but youíve got to do two order of magnitude, which we did. We actually went about a couple more magnitude than Lilley and McClain, and showed that up to that point, on Cygnus at least, there apparently was no Ė we would not claim anyway, there was anything like extragalactic absorption. We did see something and its (?) Cygnus. No, not on Cygnus, on Virgo, very slightly. And I remember arguing with Rod Davies about it, but I donít think we ever published it because Rod Davies claimed that it wasnít significant. It rather amused me in later years that somebody, I think, did find a little bit of absorption of some kind (?) Virgo. But we did (?) It was tricky that this wasnít ideal (?) we were getting much bigger scans than anybody had been before, and technique-wise, it was very tiring.

Sullivan

It just did not fascinate you, apparently.

Jennison

I donít like that epoch in my life, really, I didnít really enjoy it. And I wanted to get out of it, which I ultimately did.

Sullivan

When did you leave Jodrell?

Jennison

In the late fifties, I asked Lovell if there was anything else going on, to get out of the hydrogen line, and he mentioned that Britain was getting seriously interested in space research, would I be interested in doing something with that? And that even though he (?) the meteors and radio astronomy up there were not going to do something, so he left it entirely to me, of course, and at that point, the Americans and the Russians had got very peculiar results in the flux of dust particles, magnetic particles from space vehicles, which seemed to be much higher than one was expecting, so that all the debris coming onto the earth, seemed to be in that part of the spectrum. The particles, they were using microphones, both the Russians and the Americans. I thought up a new technique, very simple, this is not radio astronomy so we wonít go into it. It was entirely different, using perforation of (?)

Sullivan

Why was it that you wanted to change from radio astronomy?

Jennison

I didnít necessarily want to change from radio astronomy, I just didnít want to work on the hydrogen line. There was nothing else going on.

Sullivan

And so in order for you to stay at Jodrell, that was the (?)

Jennison

I didnít really mind, I suppose, if I was staying at Jodrell or not. I just got fed up with the hydrogen line work. And so I did this, and incidentally, that was (?) brought down the flux of those particles from previous (?) by a factor of 1/1000. And now itís the accepted (?) level.

Sullivan

When was this aerial two?

Jennison

1963, thatís right, 1963.

Sullivan

I see that you had a paper in 1960 on the possibility of doing radio astronomy from artificial satellites.

Jennison

Yes. Oh, that created quite a bit of sensation. Because of Sputnik you could get at two things. First of all, I looked upon it as (?) much papers as you (?) papers, some of the technical journals (?) something further, but it went in stages. The first thing I realized was that the ionosphere ought to act as a sort of Lloydís mirror, to put it crudely, (?) on top of it, and therefore, interferometry-wise, it was rather interesting. One might be able to make use of that to study the angular diameters of (?)

Sullivan

Oh, yes.

Jennison

And that was the very first thing about this that, at least from the top, the ionosphere would reflect, and (?). Later on, we got onto it that it wasnít as reflecting as refracting, and that it would focus the radiation and the frequency in the order of a megacycle or so, you can actually get the beam from an aerial hundreds of kilometers across, just from a (?)

Sullivan

From a satellite?

Jennison

Yes.

Sullivan

Looking through the ionosphere at a cosmic source?

Jennison

You set your satellite just above the critical level of the ionosphere, or tune your receiver to just above the critical frequency, okay? When youíre on the top side, right? Now if you now trace out (the ray paths?) youíll find that they go out almost (?) upwards, because the ones that are straight up obviously go straight up, the ones that go out sideways are bent around (?) straight up, but the ones that are reflected downwards, go down a little bit and then up again, again focusing.

Sullivan

Oh, yes.

Jennison

So you get the equivalent of an enormous dish formed by the ionosphere (?) choosing that frequency.

Sullivan

Has this principle been used or the ionosphere?

Jennison

We tried it on a black (?) pocket in the early sixties and then Graham Smith went a tad more on it, wrote a lot about it.

Sullivan

So it has been used? This principle.

Jennison

He tried it; I donít know how he got on with it.

Sullivan

I donít think in the Radio Astronomy Explorer satellites theyíve used anything like this, do they?

Jennison

Probably not.

Sullivan

Straight observations with a long antenna.

Jennison

Yes, yes. The other thing I went on to (?) was using occultation of the earth. I didnít actually apply these satellites with our proposed techniques (?) of the earth with its ionosphere to give you a refractive occultator.

Sullivan

I see.

Jennison

And to look at the sky or anything that might be around in those frequencies. Never did the experiment. I got involved a bit, as you see there in that, but at that time I had also got involved in an experiment to measure the speed of gravity, which led me to something thatís occupied me almost ever since. As a result of conversation with Bondi, who said he didnít know how a ray of light crossed rotating the system, I was so set, more set than Iíve ever been in all my life. And I solved that one and went on to other problems in rotation, which people take for granted, but which are not understood and not known. This led me on from that to inertia and some other things. And so my own real interest these days is astronomy in accelerated systems.

Sullivan

I see.

Jennison

And making sense of them. Which is curiously enough, though we all think itís been done, itís never been done.

Sullivan

Yes, David Edge was telling me about some of these experiments that youíve done. Thereís another meeting that I forgot to mention. That was the 1958 symposium in Paris. What are you impressions of that? You were at that meeting?

Jennison

Yes. I was at that meeting. David Edge quoting something I said at that meeting, I thought it was, I thought it was (?)

Sullivan

(?)

Jennison

I canít remember, (?) It was some comment I remember. The other thing I remember was Bracewell claiming that Cygnus was all wrong; that the three bumps (?) and not (?) Itís just that so often when I meet (?) these days and its tending to get more so, sometimes when a chap is (?) how extraordinarily blind they can be to common sense. Itís a curious business. If you stick too much in theory of something, you (?)

Sullivan

(?) They just donít think that way.

Jennison

Iíll tell you what it is. Itís a different matter of thought. As in again, one meets this very, very much (?) one hears so often that thought experiments are no good, they never teach you anything, a though experiment. Youíll see this written up time and time again by the philosophers of science or this that and the other. Because you only put in what you start with. Itís a lot of absolute nonsense. I agree, you only put in what you start with, but thereís nothing like a thought experiment for clearing the air and making sense of things. In fact, the average experimentalist uses a thought experiment all the time. He thinks up this; he thinks up that; he discards it because of this; he discards it because of that. In the end he gets a beautiful, clear conception of a way heís going to do it. And itís all done by thought experiments.

Sullivan

Yes, thatís right.

Jennison

Without spending penny. And itís a terribly important way of doing it. And you can apply the same principles in theoretical work and do it beautifully; it works marvelously. The trouble is that some people do it wrongly. Thatís why itís got a bad name. But Einsteinís original 1905 paper was a thought experiment. The exploding (?). Pure thought experiment. And so it goes on. They are beautiful things, thought experiments. And Iím all for reinstating them as a very, very valuable part of science. Sorry, Iím pattering.

Sullivan

No, thatís a very interesting comment. You have given me a very thorough going-through of what you worked on in the 1950s. I might ask a more general question: as you look at the development of radio astronomy, up until 1960 or so, do you have any comments?

Jennison

One comment. Can I give you exactly the same comment that I gave to David Edge or David Edgeís minion when they were doing their survey, and which appeared in a write-up I saw on this chapter as the very enigmatic reply. I was the responsor for the very enigmatic reply to the question: ďWhat, in your opinion, were the three greatest discoveries in radio astronomy?Ē My so-called very enigmatic reply was: ďHeyís first three discoveries.Ē

Sullivan

Why is that enigmatic? It seems like a straight answer.

Jennison

I donít know. Seemed to me to be a straight answer. The meteor astronomy stuff, the solar radio astronomy, and Cygnus.

Sullivan

Right.

Jennison

Right? And look whatís come from those.

Sullivan

Right.

Jennison

Iíll still stick to it.

Sullivan

Each one has founded a whole branch.

Jennison

Thatís right. All right, youíve got to take -

Sullivan

Jansky -

Jennison

Youíve got to take Janskyís thought in it, I agree. But if you then, within that spectrum of radio astronomy that was done, the three greatest things, I think, were Heyís three things (?)

Sullivan

I guess I might throw in van de Hulstís discovery along with, the actual discovery, and then those five things sort of encompasses all of radio astronomy.

Jennison

Yes, youíre right. So I donít know why David Edge referred to it as the enigmatic reply, everybody else seemed to come out with something relatively modern. I stick to those. That founded the science.

Sullivan

Well, thank you very much.

Jennison

OK, old chap.

Sullivan

That ends the interview with Jennison on 31 August 76.

End Tape 74B


Modified on Tuesday, 06-Feb-2018 15:44:16 EST by Ellen Bouton, Archivist (Questions or feedback)