By that time, we had moved out to the Mullard [Radio Astronomy] Observatory and, as I said, I… I spent a lot of time designing feeder arrays for Martin Ryle's telescopes but, at the same time, I was operating these… these interferometers. And then when we discovered that we could detect scintillation when we looked at quasars – and that was a total surprise, actually – it came from the very first surveys that Martin Ryle was carrying out. He, I mean… quasars, then it just wasn't known that you could have radio galaxies that had these intense, tiny bright components. We thought that all radio galaxies were like Cas A [Cassiopeia A] and… and the Crab [Crab Nebula], rather… rather large astronomical objects in the sense that they subtended a large angle in the sky, I mean, minutes of arc. And in order to get scintillation you need… you need sources which are a fraction of an arc second, hundreds of times, a hundred times smaller at least. And I thought about this, but… but in the… there's… there's in my notebook I've got… I've got a comment that you wouldn't, you couldn't get scintillation on radio sources other than the ionospheric because they were much too large. I mean, all… all radio sources then known were a minute of arc across or several minutes of arc across and they just wouldn't do it. But during one of Martin Ryle's surveys fluctuations began to be seen on one or two sources; this was the work done by Phil Williams, who got into Welsh politics later. He and his… his research student got onto this but they didn't know what it was. And we used to have great discussions, group discussions, when Martin Ryle and his research team would just, as it were, settled down on a Saturday morning and just discuss the work in general and… and other people's publications and so on, group discussions, and Martin Ryle looked at me and said, 'Well, why couldn't it be scintillation?', you see. And at that time they just detected 3C48 as having an angular size of whatever it is, an arc second or two, and I worked out that, yes, you could, you would get scintillation in the Sun's atmosphere with something that… that small. And that was… was a revelation; I mean, that… that came in in the early 1960s. In fact, we could have discovered quasars in the… 10 years earlier if… if we'd bothered to do the experiment because we had the apparatus to do it. The telescopes were sensitive enough. But I'd ruled it out because obviously radio galaxies couldn't twinkle, but they did because there was something there which we didn't know about.

And so it was, when we first saw… saw the twinkling, I said, hey, I'll measure the speed of the solar wind, just like I did with the ionosphere. And that would be good because the first spacecraft, it was called Lunik – it was one of the Russian… early Russian spacecraft – detected the solar wind which had been predicted theoretically a few years before by… by an American whose name I forget at the moment. You can remind me, perhaps. The solar wind had been predicted but not measured and the Russians found it, and I decided that with radio astronomy we could measure the solar wind on the ground, just as I'd measured ionospheric winds.

And in 1964 we made this discovery of… that… that certain radio galaxies, the quasars, twinkle and I decided to use those to measure the speed of the solar wind by tracking the cloud pattern, defraction pattern, as it moved… as it moved over the ground. But it's moving very fast. It's moving… the solar wind goes typically at 300–600km per second; that's… that's over a million miles per hour, so if you're going to measure a time-lag you need a very big baseline to do it. I mean, it crosses England in a fraction of a second. So I had to build interferometers which were at least 50 miles apart, preferably more, but I started off by putting in antennas using a Cambridge antenna and building one on the… on the way to Newmarket, the other side of Newmarket on the way… on the way to Thetford. And I put one on the East Coast at Clacton, so I had a triangle of very simple radio telescopes, and I could just measure time delays as the solar wind swept across. And so I was able to measure solar wind on the ground and that – we're now talking about the mid 1960s, 1964, '65, that sort of thing – and we did something which was interesting and that the spacecraft couldn't do. Because when you look at quasars, you can see quasars all round the Sun and so you can measure above and below the solar poles, as well in the solar equator. Now, if you're… measuring the solar wind in spacecraft, which was then becoming quite a popular thing to do, you can only launch spacecraft into the plane of the ecliptic – that's the plane in which the planets move – because to get a spacecraft away from the Earth's atmosphere out into far space, you have to use the Earth's orbit velocity as a… as a launch tool. In other words, if you're going to launch… use the spacecraft to measure the solar wind and get clear of the Earth, you can only launch it, an artificial satellite space probe using the Earth's orbital speed, which keeps you into the… into the plane of the ecliptic.

But I was looking at radio galaxies and was able… was able to measure the solar wind anyway, and what we found was that the solar wind coming from the poles was actually about twice the speed as the solar wind coming from the… from the equator. I didn't follow this up too much, but we published it and it took until the 1990s for spacecraft, American spacecraft, to actually confirm that measurement, that the solar wind really was about twice as fast coming from the poles of the Sun. And I thought that was a nice thing to have done, to have been the first to do that, I mean, 30 years earlier, using simple radio astronomy on the ground with antennas that you could build for £10,000 or so, whereas the spacecraft, to… to do that with spacecraft you need to launch a spacecraft out to the planet Jupiter and then use Jupiter's gravitational field to direct the spacecraft out of the plane of the ecliptic and over the pole of the Sun. And that didn't… that didn't happen until the 1990s, the things like the Helios. I think Ulysses spacecraft was the first one to do that. And the Russians, lo and behold, found the same thing that we did and were able to study it, of course, much more closely and seriously than we were able to do. But we at least discovered that and got that published and that's… that's the kind of research which interests me because, you know, you've done something with your own hands and made a brand new result and you've done that with your own graduate students and you hadn't… it didn't cost the earth. I mean, you built quite simple telescopes to do it. And it was that sort of background then that led on to the construction of the very large array which I then built to study, to detect quasars and to map them over… over the sky. And the good fortune there was that it was exactly the right instrument to pick up the fluctuations that… that the pulsars radiate. And that's quite a long story but I suppose it's the thing that still keeps me awake at night, actually.