The Discovery of Radio Galaxies and Quasars The following is reproduced from the Editors' Introduction to the book "Quasistellar Sources and Gravitational Collapse", Proceedings of the First Texas Symposium on Relativistic Astrophysics, edited by Ivor Robinson, Alfred Schild, and E. L. Schucking, published by The University of Chicago Press (Chicago and London), 1965. It is an excellent story. Read on ... -------------------------------------------------------------------------- To listen to Cristoforo Colombo, sitting in a dimly lit favorite tavern with a glass of port, recounting his adventures and describing vistas of far lands never seen before, must have made an exciting, unforgettable evening. To hear the reports of the astronauts who set foot on our Moon and the moons of Mars may be an equally rewarding experience in the future. But one might doubt if this could ever match the strange fascination of an evening with the late Walter Baade from the Mount Wilson and Palomar Observatories. For more than a quarter of a century he had worked with the biggest optical telescopes on Earth. When he told what he had seen and discovered in careful scanning of thousands of plates, the incredible grandeur of the cosmic realm, the galactic and extragalactic world, began to unfold behind numbers, pictures, and astronomical gossip. This man in his dark blue tie and gray suit, wearing brown shoes of enormous size, was absolutely fascinated by his research. Gesticulating, incessantly smoking, with carefully parted thin white hair, white somewhat bushy eyebrows, protruding hawk nose, Baade saw the mysteries of the universe as the greatest of all detective stories in which he was one of the principal sleuths. Sparkling with ideas, confident of the latest numbers and results, modest but full of ferocious criticism, independent and amiable, he told us one evening the story of Cygnus A: "In 1951, at a seminar talk that Minkowski gave in Pasadena on the theories of radio sources, I got mad. I had just published the theory of colliding galaxies in clusters and identified the Cygnus A source with such a pair in collision. Nobody would believe that there were extragalactic radio sources. Minkowski reviewed all the other theories first; and then, at the end of the seminar, as if he were lifting a hideous bug with a pair of pincers, he presented my theory. He said something like: 'We all know this situation: people make a theory, and then, astonishingly, they find the evidence for it. Baade and Spitzer invented the collision theory; and now Baade finds the evidence for it in Cygnus A.' "I was angry [said Baade] and I said to him 'I bet a thousand dollars that Cygnus A is a collision.' Minkowski said he could not afford that; he had just bought a house. Then I suggested a case of whiskey, but he would not agree to that either. We finally settled for a bottle, and agreed on the evidence for collision-emission lines of high excitation. I forgot about the thing until, several months later, Minkowski walked into mv office and asked 'Which brand?' He showed me the spectrum of Cygnus A. It had neon-five in emission, and thirty-seven-twenty-seven, and many other emission lines. I said to Minkowski: 'I would like a bottle of Hudson Bav's Best Procurable,' that is the strong stuff the fur hunters drink in Labrador. "But that was not everything. For me, a bottle is a quart; but what Minkowski brought was a hip flask. I did not drink it. I took the flask home as a trophy. But that's not the end of the story. Two days later, it was a Monday, Minkowski visited me in order to show me something-he saw the bottle and emptied it." Baade chuckled: "Isn't it a shame that you get no returns when the horse you bet on is a dead-sure thing?" That was six years ago, and the horse is dead. Most of the experts would agree now that Minkowski had a right to consume the whiskey because Baade had not won his bet. But in one thing he was right. He had put Cygnus A where it belonged, deep out in space time, some five hundred million light-years away. And with this distance determination, based on the redshift of the galaxy, the major discovery of contemporary astronomy emerged: the existence of violent events in galaxies, with tremendous release of energy, in the range of 10^60 erg. Cygnus A is the brightest radio source in the constellation Cygnus or Swan (hence the "A"), and the second brightest in the whole sky. Another current notation for this source is 3C 405 (No. 405 in the third Cambridge Catalogue of Radio Sources). The first to "see" Cygnus A was Grote Reber, an amateur astronomer (like Sir William Herschel). With a $2000 homemade radio telescope in his back yard in Wheaton, Illinois, Reber made a radio map of the Milky Way, which was published in 1944. Several patches of high intensity appeared on his chart: One of them was Cygnus A. The resolution, however, was poor. It is given by the ratio of wavelength to aperture. Reber, observing waves about a million times as long as~hose of visible light, would have needed a mirror some 600 meters in diameter to attain the same accuracy as an optical astrono- mer using his unaided eye. In fact, his mirror was 10 meters wide; and it was not clear whether the Cygnus A radio source was many times the angular size of the full moon, or ten thousand times smaller. The first to discover the discrete nature of the Cygnus source was the British physicist Stanley Hey, one of the co-discoverers of the Sun in the radio range. Hey and his two co-workers, S. J. Parsons and J. W. Phillips, used modified 64 MHz anti-aircraft radar sets after World War II. In the region of Cygnus they found a rapidly fluctuating radio source. This result they published in Nature in 1946. Their paper, entitled "Fluctuations in Cosmic Radiation at Radio Frequencies," came to the conclusion: "It appears that such marked variation could only originate from a small number of discrete sources." Within two years the Australian radio astronomers, John G. Bolton and G. T. Stanley, succeeded in giving a much more accurate position of the Cygnus source by using an ingeniously devised interferometer. They showed that the angular diameter of the source was less than 8 minutes of arc. Their paper in Nature bore the title: "Variable Source of Radio Frequency Radiation in the Constellation of Cygnus." The argument that Cygnus A was a point source (a "radio star") seemed undefeatable: the intensity fluctuations were considerable (up to 15 per cent) and noticeable within seconds. Such an object could, therefore, only have stellar dimensions, dimensions of the order of light-seconds (the Sun is about 5 light-seconds across). But the argument was completely wrong. When the Australians and their colleagues in Cambridge and Jodrell Bank recorded waves of the same frequency simultaneously at two different places, they found little correlation in their Sty fluctuations. Thus it was established that the fluctuations come about at a place many billion times nearer than expected: in the F-layer of our own ionosphere. The phenomenon is the analogue of the optical scintillation of starlight. It might have seemed that the evidence for the discrete nature of Cygnus A was exploded. Not at all. Every schoolboy knows the simple difference between stars and planets (pointand disklike objects) on the night sky: stars twinkle, planets do not. The conclusion of the radio astronomers was unshaken. Cygnus A had to be a point source because it twinkled. The next decisive step in the mystery story of Cygnus A came in 1951. F. Graham Smith in Cambridge got the first accurate position of this source. With a precision never before obtained in radio astronomy, he measured the location at R.A. l9h57m45.38 + 18, declination +40:35' +/1' (coordinates for 1950). He airmailed these precious numbers at once to Walter Baade in Pasadena. Baade received the letter near the end of August, 1951. "I really became interested," he says. "Up to then I had refused to be drawn into attempts to identify the Cygnus source. The positions had not been accurate enough. But I knew that with the Cambridge data something could be done. They are still unsurpassed." He determined to observe the position of the source during his next trip to Mount Palomar, if it could possibly be worked into the schedule. A night on Palomar at the 200-inch telescope is always a nerve-racking experience for Baade. Although he has been observing for years, he still feels excited every time he operates the tele- scope. Furthermore, each night is a gamble, since seeing conditions are unpredictable. The entire night may be lost because of poor weather, and information obtained by the giant instrument is so valuable that every available second is scheduled in advance for definite observations -and every available second must be used. Baade usually spends ten to twelve consecutive nights on Mount Palomar, and he approaches the work with the tenseness of an athlete before a big game. He loses about ten pounds during every session. Baade's next session started September 4. It turned out to be a good night. He took sets of photographs of the Andromeda galaxy and of certain bright nebulae in the Milky Way-these observations were part of continuing long-term studies-and before midnight he found time to squeeze in the extra project. He focused the 200-inch telescope on the position indicated in Smith's letter, and took two photographs of Cygnus, one in blue and one in yellow light. He developed the photographs himself the next afternoon. "I knew something was unusual the moment I examined the negatives. There were galaxies all over the plate, more than two hundred of them, and the brightest was at the center. It showed signs of tidal distortion, gravitational pull between the two nuclei. I had never seen anything like it before. It was so much on my mind that while I was driving home for supper, I had to stop the car and think." Baade does not remember exactly when the correct idea occurred to him, but the time was somewhere between the stopping of the car and the end of supper. He interpreted an irregular gray-black blob on a photographic plate as the mark of an event that had never been observed before in the long history of astronomy. The chances against observing this particular event are conservatively estimated at a hundred million to one. Baade decided it was a traffic accident of truly cosmic proportions. Two spiral galaxies-flattened island universes, each containing billions of stars were colliding head-on, face to face, like a pair of cymbals. That evening, before entering the observing cage of the telescope, Baade examined the plate again briefly, "to take one last look and hold the Cygnus pictures in my mind." In the cage he spent a large part of the night trying to find flaws in his interpretation. But there was something about the appearance of the two nuclei, something that gave the feeling of two overlapping or merging disks, and the notion of colliding galaxies seemed inescapable. It is doubtful that the idea would have occurred to any other astronomer in the world, but Baade's experience over the past decade or so had trained him specifically to recognize that image on the photographic plates. Interpreting difficult astronomical records, and di'.ficult records in all branches of science, depends in part on what the investigator has seen in the past, and on his ingenuity in noting significant differences and similarities. It depends in a subtle way on what the investigator is ready to see. Because of his previous research, Baade was uniquely ready to understand the meaning of the evidence before him. [Reported in Pfeiffer (1956).] Then came the famous bet. Minkowski was converted to the collision theory, and with him nearly all of the astronomers in the West. In 1954 Baade and Minkowski gave an extensive discussion of their findings in the Astrophysical Journal. The idea of colliding galaxies seemed to have an irresistible impact. Velikovski's modest idea of "Worlds in Collision" seemed to have attained reality on a grand scale. The flux measurements for Cygnus -A indicated a total energy output in the radio frequency range of about 10^44 erg/sec. If the collision lasted for some 3 million years (10^44 sec) this gave a total energy output of 10^58 erg in the form of radio-frequency radiation. The collision theory seemed to be able to provide energies of that order. According to the ideas of Baade and Spitzer, a collision of galaxies in a rich cluster would take the following form: the stars of two colliding galaxies would not collide, only the gas in their disks. Assuming a relative velocity of 1000 kilometers per second (10^8 cm/sec), a gas content of 1 per cent, a mass of 10~ solar masses (2 X 10^44 am) one would obtain a kinetic energy of 10^58 erg. Assuming higher gas content, higher masses, and higher relative velocities one could perhaps push this estimate up by two or three orders of magnitude. But still there was one uncomfortable fact about the collision theory right from the beginning: the high efficiency of the conversion of kinetic energy into radio-frequency radiation. The efficiency factor seemed to be close to 100 per cent. If one assumed the highest collision energy one might get the efficiency down to 0.10 per cent, still uncomfortably high if Cygnus A was not the artifact of a superb cosmic radio engineer. There is now hardly any doubt among the experts that collisions of galaxies in rich clusters occur. The question is: Do such collisions lead to the formation of strong radio sources? Four years later came another decisive step. It was the summer of 1958 and the scene was the Cite Universitaire of Paris. One hundred and sixty-two radio astronomers from seventeen countries came together to discuss the status of their new art. Tremendous progress had been achieved in the few years that had gone before. Observers in several countries had found convincing evidence for the spectacular theory of Vitali Lazarevich Ginzburg, of the Lebedev Physical Institute in Moscow, and Iosif Samuelovich Shklovski, of the Shternberg Astronomical Institute in Moscow. The brilliant young theoreticians, both of whom were unable to attend the conference, had independently suggested that highly relativistic electrons exist in strong non-thermal radio sources and that the radio emission was synchrotron radiation. Shklovski went further and suggested that the continuum portion of the visible light from the Crab Nebula (the supernova of 1054 in our Galaxy) had the same origin. This process occurs on a very modest terrestrial scale in the largest particle accelerators. It had already been proposed as an explanation of the radio outbursts from the Sun and possibly other stars. Shklovski and Ginzburg, however, were the first to imagine that this phenomenon might take place in clouds of particles measuring light-years or even thousands of light-years across. Their idea, published in 1953, created a revolution in astrophysics. It was confirmed immediately by the discovery of the predicted polarizations in the light of the Crab Nebula (M1) and the galaxy M87. But as Geoffrey Burbidge showed in his introductory lecture in Paris, the new breakthrough in understanding made the mystery of Cygnus A and other strong radio sources still more puzzling than before. The British astrophysicist computed from the theory of synchrotron radiation the minimum energy of Cygnus A, contained in relativistic electrons and in the magnetic field, as 2.8 X 10^59 erg. He also put forward (independently of Ginzburg) that a much larger amount of energy might still be present in relativistic protons. He thus arrived for Cygnus A at a minimum total energy contained in protons and magnetic field of 3.9 X 10^60 erg, corre- sponding to a mass equivalent of two million times the mass of the Sun. These staggering numbers were clearly not in favor of the collision theory or any other explanation. Among the many papers presented in Paris there was another one that proved important for the discussion of Cygnus A: Jennison from Jodrell Bank reported his measurements on the structure of this radio source. His astonishing result was that the radio emission came from two bright sources 82 seconds of arc apart, several times the diameter of the Galaxy. Such a structure, now known to be quite frequent among strong radio sources, was also no evidence to support the collision theory. The identification of other strong radio sources reported in Paris were not favorable to Baade's idea. Multiple-centered elliptical galaxies (which usually contain no considerable amount of gas) were identified, and also field galaxies outside of clusters, where the probability of a collision was extremely small. Thus after the Paris Symposium the enigma of the strong radio sources had begun to emerge in its true and colossal proportions: Mysterious energy sources exist in the universe able to supply more than 106° erg, or perhaps even many times more than that, in the form of relativistic particles. Viktor Amazapovich Ambartsumian had already refuted the collision theory for Cygnus A in 1958 at the Solvay Conference in Brussels. But he had not been able to propose a physically feasible mechanism for the energy sources. And so, since no workable alternatives showed up, the collision theory was not generally discarded. In 1960, when Otto Struve treated it as superseded in one of his Sky and Telescope articles, Rudolph Minkowski, who had himself once attacked the theory, wrote a long letter to Sky and Telescope correcting Struve. At that time Minkowski had just finished his beautiful work on 3C 295 identifying the most distant object in the universe (this record was broken only recently by Maarten Schmidt's measurement of the redshift of 3C 147). Minkowski's radio source 3C 295 showed up exactly where it should have been according to the collision theory: in a rich cluster of galaxies. It was another spectacular discovery, made in 1963, that again directed the attention of astrophysicists to the mysterious, huge energy sources: the discovery of the quasistellar radio sources. The history of their discovery was a case of mistaken identity, strangely similar to that of Cygnus A. A radio source like Cygnus A would be observable even at far greater distances. If the diameter of such a source could be measured it would provide a very interesting test for models of the universe. Fred Hoyle was the first to discuss this cosmological test at the Paris Symposium. He pointed out that the Einstein-de Sitter universe, for instance, would lead to a minimum diameter of radio sources in contrast to the steady-state model. Hoyle's test became a strong incentive to search for radio sources of smallest diameters. Mainly through the work of Palmer and his collaborators at Manchester the structure of many radio sources, and especially their diameters, became measurable. Among the many sources measured was one that could not be resolved: 3C 48. A huge interferometer with a maximal base length of 61000 wavelengths, limited in size essentially by the width of England, gave the result that the radio source 3C 48 had a diameter of less than 4 seconds of arc. The Californian radio astronomers in Owens Valley used their 90-foot twin dishes for a positioning of 3C 48 that was six times better than the Cambridge measurements. With this position from Thomas A. Matthews, Allan R. Sandage of the Mount Wilson and Palomar Observatories, the world's leading cosmologist, identified 3C 48. On his plates, taken with the 200-inch optical telescope at Mount Palomar, he found at the position of 3C 48 a sixteenth-magnitude star of a strange blue color and with a faint wisp. Sandage made his discovery known at the 107th Meeting of the American Astronomical Society, held December 28-31, 1960, in New York City. His unscheduled paper announced the discovery of the first true radio star. An account of this paper in Sky and Telescope says: "Since the distance of 3C 48 is unknown, there is a remote possibility that it may be a very distant galaxy of stars; but there is general agreement among the astronomers concerned that it is a relatively nearby star with most peculiar properties." And so it stayed for more than two years. Sandage found variations in brightness which seemed to give definite proof of the stellar nature of 3C 48. The mere thought that a galaxy, a system of 10'~ stars (distributed over a huge volume of space) could vary simultaneously in brightness seemed so utterly ludicrous that it could apparently be discarded instantly. But still the spectrum of this strange radio star in the constellation Triangulum was very difficult to interpret. It was years before Sandage and Matthews finally sent the results of their observations and calculations to the Astrophysical Journal (see reprint in this volume). But then a note at the end of their paper reveals the dramatic changes that took place at the beginning of 1963 through the identification of 3C 273. The British radio astronomer Cyril Hazard had found an ingenious way to obtain positions and dimensions of discrete radio sources with an unsurpassed accuracy. In observing occultations of the radio source by the Moon, he was able to find radio positions that were as Rood as optical ones. Together with his co-workers M. B. Mackey and A. J. Shimmins he used the 210-foot dish at Parkes in Australia to apply his method to 3C 273. This source turned out to consist of two components, called A and B. separated by 20 seconds of arc. The B-source had a radio diameter of less than a half-second of arc and coincided with the image of a thirteenth-magnitude star. Never before was the identification of a radio source achieved with the same degree of certainty (see reprint in Appendix I to this volume). At first it seemed that another "radio star" had been found, i.e., another object of the 3C 48 class, now called "quasi-stellar radio sources," "quasi-stellar objects," or more shortly, "quasi-stellars" (the term "quasars" coined by Chin has not found general acceptance among astronomers). The optical spectrum of 3C 273B showed bright emission lines, but it seemed impossible to identify them. They belonged apparently to forms of matter never found in the universe before. It was the Dutch astronomer, Maarten Schmidt, from California Institute of Technology, who found the correct solution: all wavelengths are shifted to the red by an amount of 16 per cent. Now the lines belonged to familiar elements like hydrogen (see reprint in Appendix I to this volume). J. B. Oke in Pasadena checked Schmidt's interpretation by finding the redshifted Ha line in the infrared spectrum of 3C 273B with a spectrum scanner (see reprint in Appendix I to this volume). What was the nature of the redshift? Was it a gravitational redshift of a superdense star or the Doppler shift of a distant receding galaxy? The decision was not easy. A galaxy like 3C 273B had never been seen before. It actually had to be some kind of a superstar outshining the biggest galaxies, an object about a million million times brighter than the Sun. However, the alternate hypothesis of a nearby superdense star posed even greater difficulties and had to be discarded: 3C 273 was a distant galaxy. Greenstein and Matthews tried to reinterpret Sandage's spectrum of 3C 48, allowing for an unknown redshift (see reprint in Appendix I to this volume). This was possible and the redshift came out even larger than that of 3C 273B. As in the case of Cygnus A the radio-star hypothesis had been disproved. The distance of 3C 48 had been underestimated by a factor of a million. Again there were now the indications of tremendous energy requirements. The Acomponent of 3C 273 had been identified optically with a small nebula. If both were of common origin and had separated with the velocity of light, the minimum age of 3C 273 would be a hundred thousand years. Assuming a constant output of optical energy over this time would lead to an energy estimate of about 1058 erg. The fabulous nature of this energy source attained a new aspect when Allan Sandage (see reprint in this volume of his article from the Astrophysical Journal) and Harlan Smith and Dorrit Hoffleit discovered the variability of 3C 273B (see reprint in Appendix I to this volume). Variability of such an object was equivalent to switching on and oR the light of hundred thousand millions of suns within short times. A large energy release over a short time is called an explosion. Already in 1943 the astronomer C. K. Seyfert of Vanderbilt University had found evidence for violent events in some galaxies that made a supernova explosion of a star appear as a firecracker com- pared to a big bomb. The most amazing photographs of an exploding galaxy are those of the nearby system M82. The beautiful pictures in the light of the Ha line and the spectroscopic measurements published in 1963 were clear-cut proof for the existence of such super-super-explosions which are not necessarily connected with strong radio emission (see reprints of papers by Lynds~and Sandage, and by the-Burbidges and Sandage in this volume). The explosions seem to take place near the center of the galaxies. The frequency of these occurrences indicates that such events of gigantic violence might even happen repeatedly in many galaxies. There is now some preliminary evidence for a former explosion m the center of our own Galaxy. When the structure of extragalactic radio sources became known in some detail, the picture of single or repeated explosions at the center of radio galaxies became very suggestive. Ambartsumian, Bolton, and Hanbury Brown stressed the point of view that the central regions of galaxies might be the seat of this mysterious activity. The double nature of many strong radio sources might then be due to the fact that the debris of the explosion, relativistic particles contained in magnetic fields and emitting synchrotron radiation, had been thrown out from the center in opposite directions along the axis of rotation. What energy sources for these violent events could possibly tie together strong radio sources, quasi-stellars, medium-strong radio sources like Met, and Seyfert galaxies? Many "wild" ideas had been proposed, for example, matter-antimatter annihilation or a chain reaction of supernova explosions in the dense central regions of a galaxy, but none of these won general acclaim. If the number 10^60 erg for strong radio sources had to be taken seriously, something still "wilder" was being called for. Fred Hoyle and William Fowler supplied it: gravitational collapse of a massive superstar in the center of a galaxy. A first exploration of this fascinating proposal was one of the main topics of the Dallas Symposium. Here we shall end this sketchy, non-technical account of the events that led to the holding of the Symposium in Dallas. The reader will find additional pre-Dallas information in an article by Jesse L. Greenstein in the December, 1963, issue of Scientip;c American. A report on the Symposium can be found in chapter i of this volume. A careful reader of this volume might find a few discrepancies between Dr. Chiu's report and the original papers. But we feel that this does not detract materially from the value of this excellent article that tries to sum up what experts in many fields had to say who were united only by their diverse interests in quasi-stellar sources. We regret that we were not able to include more reprints of relevant recent papers or some of the older ones. In many cases the reader will have no difficulty finding them through the references given in this volume. Only some aspects of the great energy events were discussed at Dallas. The bearing of these events on high-energy astronomy and cosmology will be discussed at the second Texas Symposium on Relativistic Astrophysics at Austin, December 15-19, 1964. REFERENCE Pfeiffer, J. 1956, The Changing Universe (London: Victor Gollancz Ltd.), pp. 107-109.