skip to content

Cavendish Astrophysics


A Brief History of Radio Astronomy in Cambridge

The beginnings

In 1933, Karl Jansky, working at the Bell Telephone Laboratories at Holmdel, New Jersey, discovered the radio emission from the Galaxy, a discovery confirmed by Grote Reber a few years later.  These observations attracted little attention from professional astronomers and the nature of the emission was unknown.

In 1939, Martin Ryle joined Ratcliffe’s ionospheric research group at the Cavendish Laboratory.  On the outbreak of the Second World War, Ratcliffe joined the Air Ministry Research Establishment, later to become the Telecommunications Research Establishment (TRE), and  Ryle followed in May 1940.  Ratcliffe supervised the activities of Martin Ryle, Bernard Lovell and Antony Hewish, later to become the leaders of UK radio astronomy.

The development of radar during the Second World War had two consequences for radio astronomy. Firstly, sources of radio interference which might confuse radar location had to be identified. Hey and his colleagues continued to improve the sensitivities of the receivers and discovered that the background radio emission from the Galaxy itself limited the sensitivity of the telescope system.  At the end of hostilities, Hey and his colleagues began mapping the sky at 5m wavelength and in 1946 they discovered the first discrete source of radio emission, which lay in the constellation of Cygnus - the source became known as Cygnus A.

The second consequence was that the extraordinary research efforts to design powerful radio transmitters, sensitive receivers and improved radio antennae resulted in new technologies which were to be exploited by the pioneers of radio astronomy, all of whom came from a background in radar – the three main groups were to be headed by Martin Ryle at Cambridge, Bernard Lovell at Manchester and Joseph Pawsey at Sydney.

Ryle returned to Cambridge supported by an Imperial Chemical Industries (ICI) fellowship and joined Derek Vonberg. Their first project was to measure the properties of the radio emission from the Sun. There was scarcely any money for equipment, but they were able to buy considerable amounts of surplus War electronics very cheaply and also acquire large amounts of high quality German radar equipment which had been requisitioned after the War. They took away five truckloads of surplus equipment from the Royal Aircraft Establishment (RAE) at Farnborough, including several 3m and 7.5m steerable Wurzburg radio antennae which were to be used for many years.

The Cavendish Laboratory had already made major contributions to radio studies of the ionosphere through Appleton’s pioneering researches and had a large body of experience relevant to the radar programme.   Ratcliffe returned to Cambridge and Ryle soon joined him.  Ryle preferred to carry out research on the extraterrestrial radio signals which had been discovered during the War rather than ionospheric research.   Ratcliffe split the Radio Group into two parts, the Radio Ionosphere Section, which continued radio studies of the atmosphere and ionosphere, while Ryle became head of the Radio Astronomy Section, a discipline which barely existed at the time.

The angular resolving power of the radio antennae available at that time was not sufficient to resolve the disk of the Sun. Ryle and Vonberg adapted the surplus radar equipment and developed new receiver techniques for metre wavelengths to create a radio interferometer, the antennae being separated by several hundred metres. Only later was it realized that they had invented the radio equivalent of the Michelson interferometer. A massive sunspot occurred in July 1946, and their observations showed conclusively that the radio emission originated from a region on the surface of the Sun similar in size to that of the sunspot region.

Early instruments – discrete radio sources

Graham Smith joined Ryle as a graduate student in 1946 and they realised that the interferometer could be modified to study the source Cygnus A. The interferometer traces revealed not only strong signals from Cygnus A, but also another source, which was located in the constellation of Cassiopaeia – it became known as Cassiopaeia A.  It was realised that interferometric techniques could be used to measure not only accurate positions, but also to analyse the brightness distribution of the radio emission of the sources.

In 1949 further discrete sources of radio emission were discovered by the Australian radio astronomers John Bolton, Gordon Stanley and Bruce Slee, who succeeded in associating three of them with remarkable nearby astronomical objects.  Ryle began the first of a series of radio telescopes designed to make surveys of the Northern Sky. What eventually became known as the 1C (first Cambridge) survey was carried out with what was referred to as a ‘long Michelson’ interferometer which was built on the rifle range site.   One of the key inventions made by Ryle early in the history of the new discipline was that of the phase shift receiver.   The receiver detected only compact radio sources and this arrangement soon became the standard technique for all radio interferometers.  The 1C catalogue contained about 50 discrete radio sources, but the positional accuracy was too poor to allow them to be identified securely with known astronomical objects.  Ryle and his colleagues favoured the view that the majority of the sources were ‘some hitherto unobserved type of stellar body, distributed widely throughout the Galaxy’. The interferometer system had cost about £350.

In 1951, Graham Smith measured interferometrically the positions of the four brightest sources in the northern sky with an accuracy of about 1 arcmin.  The observations of Cygnus A and Cassiopeia A led to their optical identification by Walter Baade and Rudolph Minkowski . Cassiopeia A with a young supernova remnant in our own Galaxy, while Cygnus A was associated with a faint, distant galaxy with a redshift z = 0.0561.  Ryle quickly changed his view on the nature of the radio source population. Cygnus A, the second brightest radio source in the Northern Sky, was associated with a faint galaxy at cosmological distances and so fainter radio sources must lie even further away and could be used for cosmological investigations.

Antony Hewish joined the Radio Astronomy Section in 1948 and his research included investigations of the nature of radio source scintillations. The cause of the radio scintillations was the deflection of radio rays when they pass through irregularities in the ionospheric plasma. In 1951, the theory of the process of scintillation as applied to observations of unresolved radio sources was worked out in detail by Hewish. The same concepts and techniques could be applied to the physics of fluctuations due to ionospheric, interplanetary, and interstellar electron density fluctuations. Throughout his career, Hewish was to pursue these studies which involved the observation of the short time-scale variability of radio sources – these were to result in an unexpected key discovery for all astronomy, the radio pulsars, in 1967.

Two further discoveries of the early 1950s were to be crucial for the development of radio astronomy. The nature of the Galactic radio emission was the synchrotron radiation of high energy electrons gyrating in magnetic fields.  A consequence was that the energy requirements of some of the most luminous radio emitters such as Cygnus A must be enormous, about 108 times more luminous as a radio emitter than our own Galaxy.  The second discovery was that the radio emission of Cygnus A did not originate from the body of the galaxy. In 1953, Roger Jennison and Mrinal Kumar Das Gupta at Jodrell Bank used interferometric techniques to show that the radio emission originated from two huge radio lobes located on either side of the radio galaxy.  These observations and their interpretation in terms the properties of relativistic plasmas and magnetic fields provided a powerful stimulus for the new discipline of High Energy Astrophysics.

As Sullivan points out, the period 1953-54 marked a turning point in the development of radio astronomy. The succeeding generations of radio telescope were to be expensive instruments, as the discipline changed from being a cheap ‘string-and-sealing-wax’ activity to a major branch of astronomy and a ‘big science’ with correspondingly much greater budgets. 

The controversy over the number counts of radio sources

Over a twenty-five-year period starting in about 1950, Ryle and his colleagues developed a series of radio interferometers of increasing complexity and ingenuity that made it possible to carry out surveys of the sky and unravel the structures and nature of the radio sources. This program involved a great deal of innovative electronics, such as the phase-switching interferometer, which enabled the full amplitude and phase information to be recovered.  The practical development of aperture synthesis was a virtuoso technical achievement that involved a considerable team of researchers and support staff. Ryle put together a tightly knit group of physicists, including Francis Graham-Smith, Antony Hewish, John Baldwin, John Shakeshaft, Bruce Elsmore, Peter Scheuer, Paul Scott and Sidney Kenderdine, as well as a strong support team of technicians and research students, many of whom were later to become leaders in radio astrophysics.

Once Ryle was converted to the idea that the ‘radio stars’ observed in directions away from the plane of our Galaxy were in fact distant extragalactic objects, he and Hewish designed and constructed a large four element interferometer to carry out a new survey of the northern sky at 81.5 MHz.  The Second Cambridge (2C) survey of radio sources was completed in 1954. Ryle and his colleagues found the startling result that the small diameter radio sources were uniformly distributed over the sky and that the numbers of sources increased enormously as the survey extended to fainter and fainter flux densities.  As Ryle expressed it in his Halley Lecture in Oxford in 1955 (Ryle, 1955),

‘This is a most remarkable and important result, but if we accept the conclusion that most of the radio stars are external to the Galaxy, and this conclusion seems hard to avoid, then there seems no way in which the observations can be explained in terms of a Steady-State theory.’

A rather acrimonious dispute arose between the proponents of the Steady State Cosmology and the standard Friedman world models, the former having been promoted enthusiastically by Fred Hoyle, Thomas Gold and Hermann Bondi.   

The Sydney group led by Bernard Mills was carrying out similar radio surveys of the southern sky at about the same time with the Mills Cross and they found that the source counts could be represented by the relation N(≥ S ) ∝ S −1.65, which they argued was not significantly different from the expectation of uniform world models. In 1957 Mills and Bruce Slee stated:

‘We therefore conclude that discrepancies, in the main, reflect errors in the Cambridge catalogue, and accordingly deductions of cosmological interest derived from its analysis are without foundation. An analysis of our results shows that there is no clear evidence for any effect of cosmological importance in the source counts.’

The controversy involved not only the proponents of the standard Big Bang and Steady State cosmologies, but also between the Cambridge and Sydney radio astronomers. The problem with the Cambridge number counts was that they extended to surface densities of radio sources such that the flux densities of faint sources were overestimated because of the presence of fainter sources in the beam of the telescope, a phenomenon known as confusion. Scheuer devised a statistical procedure for deriving the number counts of sources from the survey records themselves without the need to identify individual sources. He showed that the slope of the source counts was actually −1.8.

The Mullard Radio Astronomy Observagtory

Despite the ongoing controversy, it was clear that radio astronomy had the potential to provide new types of astrophysical and cosmological information. Ryle already had plans for the next generation survey telescope which would overcome the problems of the 2C survey.  The space available at the Rifle Range was too small for the necessarily longer baselines and so, in 1956, the radio observatory moved to a disused wartime Air Ministry bomb store at Lord’s Bridge about 10 km to the south-east of Cambridge.  In 1955 Mott and Ratcliffe persuaded the Mullard Company to make an endowment of £100,000 for the new telescope and associated facilities. This was supplemented by a grant of £40,000 from the Department of Scientific and Industrial Research (DSIR) and £40,000 from the University Development Fund to cover the costs of construction and operations for 10 years. The observatory was named the Mullard Radio Astronomy Observatory and opened by Sir Edward Appleton on 27 July 1957.

2C interferometer was upgraded with receivers for the higher frequency of 159 MHz, thus doubling the angular resolution and so decreasing the importance of confusion. The resulting Third Cambridge (3C) Catalogue contained 471 radio sources.  Meanwhile, the large 4C antenna was coming into operation at Lord’s Bridge, using the full power of the principles of aperture synthesis to create a much higher angular resolution and higher sensitivity survey of the northern sky.  The Fourth Cambridge (4C) Catalogue of radio sources, containing almost 5,000 radio sources at 178 MHz was published in two parts in 1965. The radio source counts derived from the 4C catalogues showed an excess over the expectations of Euclidean world models, the slope of the integral counts being −1.8, exactly the result found by Scheuer.  By the mid-1960s, the evidence was compelling that there was indeed a very large excess of sources at large redshifts and this was at variance with the expectations of the steady state theory.

The identification of extragalactic radio sources with very distant galaxies required radio positions better than 1 arcmin and this could now be achieved for the brighter sources with the elements of the 4C interferometer. The results of these efforts was the compilation of the revised 3C (3CR) catalogue, which contained 328 radio and became the standard low-frequency catalogue of radio sources in the northern sky.

The radio astronomical discoveries of the 1950s stimulated a great deal of astrophysical interest and led to major investments in the construction of radio telescope systems.  Among the extragalactic radio sources which could be securely identified, most were found to be associated with some of the most massive galaxies known which are very luminous and so could be observed to large redshifts.

By 1962, Thomas Matthews and Allan Sandage had identified three of the brightest radio sources, 3C 48, 3C 196 and 3C 286, with ‘stars’ of an unknown type with strange optical spectra. The breakthrough came in 1962 when 3C 273 was associated with what appeared to be a 13th magnitude star. Maarten Schmidt discovered that this was no ordinary star but the first, and brightest, of this class of hyperactive galactic nuclei. Its optical luminosity was about 1000 times greater than the luminosity of a galaxy such as our own. In addition the enormous optical luminosity of 3C 273 varied on a time-scale of years.  These sources were termed quasi-stellar radio sources and within a year this term had been contracted to the word quasar. With the improved positions available from early observations with the 4C array, it became possible to search for the counterparts of many more of the 3CR radio source. Those located away from the Galactic Plane were either very faint galaxies or quasars.  By 1965, there were sufficient data to demonstrate clearly the very large evolutionary changes needed to account for the steepness of the radio source counts.

Earth Rotation Aperture Synthesis

The next step was to develop the technique of aperture synthesis so that a particular region of the sky could be tracked. In the early 1960s, Ryle and Ann Neville carried out the most ambitious experiment to date – the use of the rotation of the Earth to carry telescopes at fixed points on the Earth about each other as observed from a point on the celestial sphere.   The result was a synthesised radio map about the North Pole of radius 7.5◦. The angular resolution of the survey was 4.5 arcmin and the sensitivity 8 times greater than that of the original antenna system. This experiment demonstrated convincingly the remarkable possibilities opened up by the technique of Earth-rotation aperture synthesis.  The succeeding generations of aperture synthesis arrays employed fully steerable antennae so that specific regions of the sky could be tracked and high angular resolution images of individual sources obtained.

Ryle was fortunate in bidding for the resources for the One-Mile Telescope, and subsequently for the 5-km telescope, during an era when the resources available for pure scientific research were increasing at a rate of 5% per year in real terms. The One-Mile Telescope was completed in 1964 and consisted of three 120-ton, 18-metre diameter antennae. The first radio maps made by the One-Mile Telescope revealed the power of the technique of Earth-rotation aperture synthesis for revealing the structures of Galactic and extragalactic radio sources and led to new astrophysical challenges concerning the origin of the enormous fluxes of relativistic electrons and magnetic fields present in these sources.

Some of the most important programmes of the One-Mile Telescope were the deep surveys of small regions of sky to extend the number counts of radio sources to very much lower limiting flux densities than the 4C surveys.  The results of the 5C2 survey were published by Ryle and Guy Pooley in 1968 and extended to limiting flux densities of 10 mJy. These observations showed clearly the convergence of the number counts of radio sources at low flux densities. From these data it was shown that the most luminous members of the radio source population had maximum comoving space density at a redshift z ∼ 2 and decreased at later cosmological epochs.

With the controversy over the number counts of radio sources resolved, the emphasis shifted to the study of the astrophysics of the radio sources. The success of the One-Mile Telescope encouraged Ryle to plan for a yet more powerful synthesis radio telescope, the Five-Kilometre Telescope, subsequently named the Ryle Telescope. The Lord’s Bridge site was not large enough for this length of baseline, but the Cambridge to Bedford railway line had been closed and it so happened that it ran more or less exactly East-West along the northern boundary of the Lord’s Bridge site. With a grant of £2.1M from the Science Research Council, Ryle built an eight-element interferometer operating at 15 GHz (2cm wavelength) which would result in an angular resolution of about 2 arcsec, comparable with that of an optical telescope.

Both the One-Mile and 5-km radio telescopes were far ahead of the capabilities of any other radio telescope system in the world when they came into operation. Some measure of Ryle’s achievement is the fact that, over a 25 year period, the sensitivity of radio astronomical observations increased by a factor of about one million and the imaging capability of the telescope system improved from several degrees to a few arcseconds

Ryle was knighted in 1966, appointed Astronomer Royal in 1972 and jointly was awarded the Nobel Prize in Physics with Hewish in 1974 ,

‘for their pioneering research in radio astrophysics: Ryle for his observations and inventions, inparticular of the aperture synthesis technique, and Hewish for his decisive role in the discovery of pulsars.’

The Discovery of Pulsars

In 1954, Hewish had remarked in his notebooks that, if the angular sizes of the radio sources were small enough, they would illuminate the solar corona with a coherent radio signal and so give rise to rapid time variations in their intensities.  By 1964, a number of radio quasars were known and some of these radio sources had small angular sizes. With Paul Scott and Derek Wills, Hewish showed that the radio scintillations were due to the scattering of the radio waves by inhomogeneities in the ionised plasma flowing out from the Sun, what is known as the solar wind.

Hewish realised that a large, low-frequency array dedicated to the measurement of the scintillations of compact radio sources would provide a new approach to the study of three important astronomical areas: (a) it would enable many more quasars to be discovered; (b) their angular sizes could be estimated; and (c) the structure and velocity of the solar wind could be determined. In 1965, he designed a large array to undertake these studies and was awarded a grant of £17,286 by the U.K. Department of Scientific and Industrial Research to construct it, as well as outstations for measuring the velocity of the solar wind.  Jocelyn Bell joined the 4.5 acre array project as a graduate student in October 1965. The telescope was commissioned during July 1967 with the objective of mapping the whole sky once a week so that the variation of the scintillations of the sources with solar elongation could be studied. A key aspect of the array was that it had to be possible to measure the fractional scintillations of the radio sources in real time.

While the array was being constructed, Leslie Little and Hewish carried out a theoretical investigation of the strength of the scintillations as a function of heliospheric coordinates . The key point was that the scintillations decrease to very small values when observed at large angles from the Sun.  The commissioning of the 4.5 acre array proceeded through the summer of 1967.

The discovery of the rapidly pulsating radio source, or pulsar, CP 1919 was made by Bell on 6 August 1967. The remarkable feature was that the source scintillated at roughly the 100% level in the anti-solar direction, quite contrary to the expectations of the scintillation models of Little and Hewish. Furthermore, the source was highly variable and not always present. The source was not observed again until 28 November 1968 when the pulses were detected separately for the first time. The paper, Observation of a Rapidly Pulsating Radio Source, was published on 24 February 1968. Three other pulsars were reported, one of them turning out to have a pulse period of only 0.25 s, which was so short that it excluded the possibility that a white dwarf star could be the parent body of that pulsar.  Within a few months, Thomas Gold convincingly associated the pulsars with magnetised, rotating neutron stars

The Cosmic Microwave Background Radiation

In 1965, the Cosmic Microwave Background Radiation was discovered more or less by accident by Arno Penzias and Robert Wilson. Penzias and Wilson had joined the Bell Telephone Laboratories in the early 1960s and had the responsibility of calibrating the antenna for use at these frequencies, for which they had access to a 7.35 cm cooled maser receiver. Wherever they pointed the telescope on the sky, they found an excess antenna temperature, which could not be accounted for by noise sources in the telescope or receiver system. There remained about 3.5 ±1 K excess noise contribution.  At almost exactly the same time, Robert Dicke’s group in Princeton were preparing exactly the same type of experiment to detect the cooled remnant of the Big Bang. Within a few months, the Princeton group had measured a background temperature of 3.0 ± 0.5 K at a wavelength of 3.2 cm, confirming the black-body nature of the background in the Raleigh-Jeans region of the spectrum.

The Cavendish Group had had a long involvement with studies of the diffuse background radiation. John Shakeshaft and Timothy Howell took up the challenge of attempting to measure the low frequency, long wavelength spectrum tail of the apparently thermal emission of the Cosmic Background Radiation at 1.4 GHz. They concluded that the background radiation had radiation temperature 2.8 ± 0.6 K.  They followed this up with even more challenging measurements at 610 and 408 MHz, with the result that their measurements were consistent with the presence of a background component of emission with radiation temperature 3.7 ± 1.2 K between 408 and 610 MHz.   

The Cavendish Radio Astronomy Group had little further involvement with the Cosmic Background Radiation, until determined efforts were made to measure temperature fluctuations in the spatial distribution of the radiation and the detection of decrements in its radiation temperature due to the Sunyaev-Zeldovich effect in the directions of clusters of galaxies.

The construction of the One-Mile and 5-km (Ryle) Telescopes opened up the study of the astrophysics of Galactic and extragalactic radio sources.   The great success of the One-Mile and 5-km Telescopes gave the Cavendish Laboratory a world-leading position in radio astronomy and it was to be some years before other radio observatories were able to compete with and eventually surpass the capabilities of these telescopes. This began to take place in the late 1970s with the commissioning of the Very Large Array in New Mexico in the USA.

New Directions

An important development was the publication of two papers by John Baldwin and Peter Warner in 1977 and 1978 on phaseless aperture synthesis. They demonstrated how aperture synthesis images could be reconstructed from the cross-correlation functions between pairs of telescopes without using the phase information, provided the flux density of one of the point sources far exceeded that of the sum of all the fainter sources in the field.  The procedure was generalized in the second paper to the case in which there was no dominant single source or in which the sources had complex structures. The further development of these procedures led to closure phase and self-calibration techniques, which have become standard data reduction techniques in radio astronomy.   They were also to be central to the development of optical–infrared interferometry in the 1980s and 1990s.

Another burgeoning discipline was millimetre and submillimetre astronomy. Ryle and his colleagues were faced with the decision as to whether or not they wanted to take up the challenges of the quite different radio astronomy of molecules and dust.  Richard Hills returned from Germany and the USA as the project scientist for the 15-metre millimetre/submillimetre telescope which was to be named the James Clerk Maxwell Telescope. 

With the oil crisis of the early 1970s, Ryle became a passionate advocate of wind power as a viable alternative source of renewable energy. This matched another of his passions, sailing and the design of sailing boats – he had already built innovative trimarans with his own hands.

Ryle was joined in his experimental programme by Paul Scott and Donald Wilson from the Radio Astronomy Group and by Alan Metherell from Laboratory Astrophysics Group, as well as by a number of research students and post-docs. The main thrust of the research was to achieve efficiencies as close as possible to the theoretical maximum of 59% and to understand all the technical and materials issues which went along with the new concepts.  The programme was prescient and successful, well ahead of the time when wind turbines became commercially viable. 



This summary of the history of radio astronomy in Cambridge is taken from the forthcoming history of the Cavendish Laboratory by Malcolm Longair.  The book will be entitled "Maxwell's Enduring Legacy: A Scientific History of the Cavendish Laboratory (2016)".  The book contains many more details the history and its context.