in which astronomers are forced to conclude,
without any prior predictions,
that black holes a millionfold heavier than the Sun
inhabit the cores of galaxies (probably)
Radio Galaxies
If, in 1962 (when theoretical physicists were just beginning to accept the concept of a black hole), anyone had asserted that the Universe contains gigantic black holes, millions or billions of times heavier than the Sun, astronomers would have laughed. Nevertheless, astronomers unknowingly had been observing such gigantic holes since 1939, using radio waves. Or so we strongly suspect today.
Radio waves are the opposite extreme to X-rays. X-rays are electromagnetic waves with extremely short wavelengths, typically 10,000 times shorter than the wavelength of light (Figure P.2 in the Prologue). Radio waves are also electromagnetic, but they have long wavelengths, typically a few meters from wave crest to wave crest, which is a million times longer than the wavelength of light. X-rays and radio waves are also opposites in terms of wave/particle duality (Box 4.1)—the propensity of electromagnetic waves to behave sometimes like a wave and sometimes like a particle (a photon). X-rays typically behave like high-energy particles (photons) and thus are most easily detected with Geiger counters in which the X-ray photons hit atoms, knocking electrons off them (Chapter 8). Radio waves almost always behave like waves of electric and magnetic force, and thus are most easily detected with wire or metal antennas in which the waves’ oscillating electric force pushes electrons up and down, thereby creating oscillating signals in a radio receiver attached to the antenna.
Cosmic radio waves (radio waves coming from outside the Earth) were discovered serendipitously in 1932 by Karl Jansky, a radio engineer at the Bell Telephone Laboratories in Holmdel, New Jersey. Fresh out of college, Jansky had been assigned the task of identifying the noise that plagued telephone calls to Europe. In those days, telephone calls crossed the Atlantic by radio transmission, so Jansky constructed a special radio antenna, made of a long array of metal pipes, to search for sources of radio static (Figure 9.1a). Most of the static, he soon discovered, came from thunderstorms, but when the storms were gone, there remained a faint, hissing static. By 1935 he had identified the source of the hiss; it was coming, mostly, from the central regions of our Milky Way galaxy. When the central regions were overhead, the hiss was strong; when they sank below the horizon, the hiss weakened but did not entirely disappear.
This was an amazing discovery. Anyone who had ever thought about cosmic radio waves had expected the Sun to be the brightest source of radio waves in the sky, just as it is the brightest source of light. After all, the Sun is a billion (109) times closer to us than most other stars in the Milky Way, so its radio waves ought to be roughly 109 × 109 = 1018 times brighter than those from other stars. Since there are only 1012 stars in our galaxy, the Sun should be brighter than all the others put together by a factor of roughly 1018/1012 = 106 (a million). How could this argument fail? How could the radio waves from the distant central regions of the Milky Way be so much brighter than those from the nearby Sun?
As amazing as this mystery might be, it is even more amazing, in retrospect, that astronomers paid almost no attention to the mystery. In fact, despite extensive publicity by the Bell Telephone Company, only two astronomers seem to have taken any interest at all in Jansky’s discovery. It was doomed to near oblivion by the same astronomical conservatism that Chandrasekhar was encountering with his claims that no white dwarf can be heavier than 1.4 Suns (Chapter 4).
The two exceptions to this general lack of interest were a graduate student, Jesse Greenstein, and a lecturer, Fred Whipple, in Harvard University’s astronomy department. Greenstein and Whipple, pondering Jansky’s discovery, showed that, if the then-current ideas about how cosmic radio waves might be generated were correct, it was impossible for our Milky Way galaxy to produce radio waves as strong as Jansky was seeing. Despite this apparent impossibility, Greenstein and Whipple believed Jansky’s observations; they were sure the problem lay with astrophysical theory, not with Jansky. But with no hints as to where the theory was going wrong, and since, as Greenstein recalls, “I never met anybody else [in the 1930s] who had any interest in the subject, not one astronomer,” they turned their attention elsewhere.
By 1935 (about the time that Zwicky was inventing the concept of a neutron star; Chapter 5), Jansky had learned everything about the galactic hiss that his primitive antenna would allow him to discover. In a quest to learn more, he proposed to Bell Telephone Laboratories the construction of the world’s first real radio telescope: a huge metal bowl, 100 feet (30 meters) in diameter, which would reflect incoming radio waves up to a radio antenna and receiver in much the same way that an optical reflecting telescope reflects light from its mirror up to an eyepiece or a photographic plate. The Bell bureaucracy rejected the proposal; there was no profit in it. Jansky, ever the good employee, acquiesced. He abandoned his study of the sky, and in the shadow of the approach of World War 11, turned his efforts toward radio-wave communication at shorter wavelengths.
So uninterested were professional scientists in Jansky’s discovery that the only person to build a radio telescope during the next decade was Grote Reber, an eccentric bachelor and ham radio operator in Wheaton, Illinois, call number W9GFZ. Having read of Jansky’s radio hiss in the magazine Popular Astronomy, Reber set out to study its details. Reber had a very poor education in science, but that was unimportant. What mattered was his good training in engineering and his strong practical streak. Using enormous ingenuity and his own modest savings, he designed and constructed with his own hands, in his mother’s backyard, the world’s first radio telescope, a 30-foot (that is, 9-meter)diameter dish (Figure 9.1c); and with it, he made radio maps of the sky (Figure 9.1d). In his maps one can see clearly not only the central region of our Milky Way galaxy, but also two other radio sources, later called Cyg A and Cas A—A for the “brightest radio sources,” Cyg and Cas for “in the constellations Cygnus and Cassiopeia.” Four decades of detective work would ultimately show, with high probability, that Cyg A and many other radio sources discovered in the ensuing years are powered by gigantic black holes.
9.1 (a) Karl Jansky and the antenna with which he discovered, in 1932, cosmic radio waves from our galaxy. (b) Grote Reber, ca 1940. (c) The world’s first radio telescope, constructed by Reber in his mother’s backyard in Wheaton, Illinois. (d) A map of radio waves from the sky constructed by Reber with his backyard radio telescope. [(a) Photo by Bell Telephone Laboratories, courtesy AIP Emilio Segrè Visual Archives; (b) and (c) courtesy Grote Reber; (d) is adapted from Reber (1944).]
The story of this detective work will be the central thread of this chapter. I have chosen to devote a whole chapter to the story for several reasons:
First, this story illustrates a mode of astronomical discovery quite different from that illustrated in Chapter 8. In Chapter 8, Zel’dovich and Novikov proposed a concrete method to search for black holes; experimental physicists, astronomers, and astrophysicists implemented that method; and it paid off. In this chapter, gigantic black holes are already being observed by Reber in 1939, long before anyone ever thought to look for them, but it will take forty years for the mounting observational evidence to force astronomers to the conclusion that black holes are what they are seeing.
Second, Chapter 8 illustrated the powers of astrophysicists and relativists; this chapter shows their limitations. The types of black holes discovered in Chapter 8 were predicted to exist a quarter century before anyone ever went searching for them. They were the Oppenheimer–Snyder holes: a few times heavier than the Sun and created by the implosion of heavy stars. The gigantic black holes of this chapter, by contrast, were never predicted to exist by any theorist. They are thousands or millions of times heavier than any star that any astronomer has ever seen in the sky, so they cannot possibly be created by the implosion of such stars. Any theorist predicting these gigantic holes would have tarnished his or her scientific reputation. The discovery of these holes was serendipity in its purest form.
Third, this chapter’s story of discovery will illustrate, even more clearly than Chapter 8, the complex interactions and interdependencies of four communities of scientists: relativists, astrophysicists, astronomers, and experimental physicists.
Fourth, it will turn out, late in this chapter, that the spin and the rotational energy of gigantic black holes play central roles in explaining the observed radio waves. By contrast, a hole’s spin was of no importance for the observed properties of the modest-sized holes in Chapter 8.
In 1940, having made his first radio scans of the sky, Reber carefully wrote up a technical description of his telescope, his measurements, and his map, and mailed it to Subrahmanyan Chandrasekhar, who was now the editor of the Astrophysical Journal at the University of Chicago’s Yerkes Observatory, on the shore of Lake Geneva in Wisconsin. Chandrasekhar circulated Reber’s remarkable manuscript among the Yerkes astronomers. Bemused by the manuscript and skeptical of this completely unknown amateur, several of the astronomers drove down to Wheaton, Illinois, to look at his instrument. They returned, impressed. Chandrasekhar approved the paper for publication.
Jesse Greenstein, who had become an astronomer at Yerkes after completing his Harvard graduate studies, made a number of trips down to Wheaton over the next few years and became a close friend of Reber’s. Greenstein describes Reber as “the ideal American inventor. If he had not been interested in radio astronomy, he would have made a million dollars.”
Enthusiastic about Reber’s research, Greenstein tried, after a few years, to move him to the University of Chicago. “The University didn’t want to spend a dime on radio astronomy,” Greenstein recalls. But Otto Struve, the director of the University’s Yerkes Observatory, agreed to a research appointment provided the money to pay Reber and support his research came from Washington. Reber, however, “was an independent cuss,” Greenstein says. He refused to explain to the bureaucrats in any detail how the money for new telescopes would be spent. The deal fell through.
In the meantime, World War II had ended, and scientists who had done technical work in the war effort were looking for new challenges. Among them were experimental physicists who had developed radar for tracking enemy aircraft during the war. Since radar is nothing but radio waves that are sent out from a radio-telescope-like transmitter, bounce off an airplane, and return back to the transmitter, these experimental physicists were ideally poised to give life to the new field of radio astronomy—and some of them were eager to do so; the technical challenges were great, and the intellectual payoffs promising. Of the many who tried their hand at it, three teams quickly came to dominate the field: Bernard Lovell’s team at Jodrell Bank/Manchester University in England; Martin Ryle’s team at Cambridge University in England; and a team put together by J. L. Pawsey and John Bolton in Australia. In America there was little effort of note; Grote Reber continued his radio astronomy research virtually alone.
Optical astronomers (astronomers who study the sky with light,1 the only kind of astronomer that existed in those days) paid little attention to the experimental physicists’ feverish activity. They would remain uninterested until radio telescopes could measure a source’s position on the sky accurately enough to determine which light-emitting object was responsible for the radio waves. This would require a 100-fold improvement in resolution over that achieved by Reber, that is, a 100-fold improvement in the accuracy with which the positions, sizes, and shapes of the radio sources were measured.
Such an improvement was a tall order. An optical telescope, or even a naked human eye, can achieve a high resolution with ease, because the waves it works with (light) have very short wavelengths, less than 10−6 meter. By contrast, human ears cannot distinguish very accurately the direction from which a sound comes because sound waves have wavelengths that are long, roughly a meter. Similarly, radio waves, with their meter-sized wavelengths, give poor resolution—unless the telescope one uses is enormously larger than a meter. Reber’s telescope was only modestly larger; hence, its modest resolution. To achieve a 100-fold improvement in resolution would require a telescope 100 times larger, roughly a kilometer in size, and/or the use of shorter wavelength radio waves, for example, a few centimeters instead of one meter.
The experimental physicists actually achieved this l00-fold improvement in 1949, not by brute force, but by cleverness. The key to their cleverness can be understood by analogy with something very simple and familiar. (This is just an analogy; it in fact is a slight cheat, but it gives an impression of the general idea.) We humans can see the three-dimensionality of the world around us using just two eyes, not more. The left eye sees around an object a little bit on the left side, and the right eye sees around it a bit on the right side. If we turn our heads over on their sides we can see around the top of the object a bit and around the bottom of the object a bit; and if we were to move our eyes farther apart (as in effect is done with the pair of cameras that make 3-D movies with exaggerated three-dimensionality), we would see somewhat farther around the object. However, our three-dimensional vision would not be improved enormously by having a huge number of eyes, covering the entire fronts of our faces. We would see things far more brightly with all those extra eyes (we would have a higher sensitivity), but we would gain only modestly in three-dimensional resolution.
Now a huge, 1-kilometer radio telescope (left half of Figure 9.2) would be somewhat like our face covered with eyes. The telescope would consist of a 1-kilometer-sized bowl covered with metal that reflects and focuses the radio waves up to a wire antenna and radio receiver. If we were to remove the metal everywhere except for a few spots widely scattered over the bowl, it would be like removing most of those extra eyes from our face, and keeping only a few. In both cases, there is a modest loss of resolution, but a large loss of sensitivity. What the experimental physicists wanted most was an improved resolution (they wanted to find out where the radio waves were coming from and what the shapes of the radio sources were), not an improved sensitivity (not an ability to see more, dimmer radio sources-at least not for now). Therefore, they needed only a spotty bowl, not a fully covered bowl.
9.2 The principle of a radio interferometer. Left: In order to achieve good angular resolution, one would like to have a huge, say, l-kilometer, telescope. However, it would be sufficient if only a few spots (solid) on the radio-wave-reflecting bowl are actually covered with metal and reflect. Right: It is not necessary for the radio waves reflected from those spots to be focused to an antenna and radio receiver at the huge bowl’s center. Rather, each spot can focus its waves to its own antenna and receiver, and the resulting radio signals can then be carried by wire from all the receivers to a central receiving station, where they are combined in the same manner as they would have been at the huge telescope’s receiver. The result is a network of small radio telescopes with linked and combined outputs, a radio interferometer.
A practical way to make such a spotty bowl was by constructing a network of small radio telescopes connected by wires to a central radio receiving station (right half of Figure 9.2). Each small telescope was like a spot of metal on the big bowl, the wires carrying each small telescope’s radio signal were like radio beams reflected from the big bowl’s spots, and the central receiving station which combines the signals from the wires was like the big bowl’s antenna and receiver, which combine the beams from the bowl’s spots. Such networks of small telescopes, the centerpieces of the experimental physicists’ efforts, were called radio interferometers, because the principle behind their operation was interferometry: By “interfering” the outputs of the small telescopes with each other in a manner we shall meet in Box 10.3 of Chapter 10, the central receiving station constructs a radio map or picture of the sky.
Through the late 1940s, the 1950s, and into the 1960s, the three teams of experimental physicists (Jodrell Bank, Cambridge, and Australia) competed with each other in building ever larger and more sophisticated radio interferometers, with ever improving resolutions. The first crucial benchmark, the 100-fold improvement necessary to begin to stir an interest among optical astronomers, came in 1949, when John Bolton, Gordon Stanley, and Bruce Slee of the Australian team produced 10-arc-minute-sized error boxes for the positions of a number of radio sources; that is, when they identified 10-arc-minutesized regions on the sky in which the radio sources must lie. (Ten arc minutes is one-third the diameter of the Sun as seen from Earth, and thus much poorer resolution than the human eye can achieve with light, but it is a remarkably good resolution when working with radio waves.) When the error boxes were examined with optical telescopes, some, including Cyg A, showed nothing bright of special note; finer radio resolution would be needed to reveal which of the plethora of optically dim objects in these error boxes might be the true sources of the radio waves. In three of the error boxes, however, there was an unusually bright optical object: one remnant of an ancient supernova, and two distant galaxies.
As difficult as it may have been for astrophysicists to explain the radio waves that Jansky had discovered emanating from our own galaxy, it was even more difficult to understand how distant galaxies could emit such strong radio signals. That some of the brightest radio sources in the sky might be objects so extremely distant was too incredible for belief (though it ultimately would turn out to be true). Therefore, it seemed a good bet (but those who made the bet would lose) that each error box’s radio signals were coming not from the distant galaxy, but rather from one of the plethora of optically dim but nearby stars in the error box. Only better resolution could tell for sure. The experimental physicists pushed forward, and a few optical astronomers began to watch with half an eye, mildly interested.
By summer 1951, Ryle’s team at Cambridge had achieved a further 10-fold improvement of resolution, and Graham Smith, a graduate student of Ryle’s, used it to produce a i-are-minute error box for Cyg A—a box small enough that it could contain only a hundred or so optical objects (objects seen with light). Smith airmailed his best-guess position and its error box to the famous optical astronomer Walter Baade at the Carnegie Institution in Pasadena. (Baade was the man who seventeen years earlier, with Zwicky, had identified supernovae and proposed that neutron stars power them—Chapter 5.) The Carnegie Institution owned the 2.5-meter (100-inch) optical telescope on Mount Wilson, until recently the world’s largest; Caltech, down the street in Pasadena, had just finished building the larger 5-meter (200-inch) telescope on Palomar Mountain; and the Carnegie and Caltech astronomers shared their telescopes with each other. At his next scheduled observing session on the Palomar 5-meter (Figure 9.3a), Baade photographed the error box on the sky where Smith said Cyg A lies. (This spot on the sky, like most spots, had never before been examined through a large optical telescope.) When Baade developed the photograph, he could hardly believe his eyes. There, in the error box, was an object unlike any ever before seen. It appeared to be two galaxies colliding with each other (center of Figure 9.3d). (We now know, thanks to observations with infrared telescopes in the 1980s, that the galaxy collision was an optical illusion. Cyg A is actually a single galaxy with a band of dust running across its face. The dust absorbs light in just such a way as to make the single galaxy look like two galaxies in collision.) The whole system, central galaxy plus radio source, would later come to be called a radio galaxy.
Astronomers were convinced for two years that the radio waves were being produced by a galactic collision. Then, in 1953, came another surprise. R. C. Jennison and M. K. Das Gupta of Lovell’s Jodrell Bank team studied Cyg A using a new interferometer consisting of two telescopes, one fixed to the ground and the other moving around the countryside on a truck so as to cover, one after another, a number of “spots” on the “bowl” of an imaginary 4-kilometer-square telescope (see left half of Figure 9.2). With this new interferometer (Figures 9.3b, c), they discovered that the Cyg A radio waves were not coming from the “colliding galaxies,” but rather from two giant, roughly rectangular regions of space, about 200,000 light-years in size and 200,000 light-years apart, on opposite sides of the “colliding galaxies.” These radio-emitting regions, or lobes as they are called, are shown as rectangles in Figure 9.3d, together with Baade’s optical photograph of the “colliding galaxies.” Also shown in the figure is a more detailed map of the lobes’ radio emission, constructed sixteen years later using more sophisticated interferometers; this map is shown as thin lined contours that exhibit the brightness of the radio emission in the same way as the contours of a topographic map exhibit the height of the land. These contours confirm the 1953 conclusion that the radio waves come from gigantic lobes of gas on either side of the “colliding galaxies.” How both of these enormous lobes can be powered by a single, gigantic black hole will become a major issue later in this chapter.
9.3 The discovery that Cyg A is a distant radio galaxy: (a) The 5-meter optical telescope used in 1951 by Baade to discover that Cyg A is connected with what appeared to be two colliding galaxies. (b) The radio interferometer at Jodrell Bank used in 1953 by Jennison and Das Gupta to show that the radio waves are coming from two giant lobes outside the colliding galaxies. The interferometer’s two antennas (each an array of wires on a wooden framework) are shown here side by side. In the measurements, one was put on a truck and moved around the countryside, while the other remained behind, at rest on the ground. (c) Jennison and Das Gupta, inspecting the radio data in the control room of their interferometer. (d) The two giant lobes of radio emission (rectangles) as revealed in the 1953 measurements, shown together with Baade’s optical photograph of the “colliding galaxies.” Also shown in (d) is a high-resolution contour map of the lobes’ radio emission (thin solid contours), produced in 1969 by Ryle’s group at Cambridge. [(a) Courtesy Palomar Observatory/California Institute of Technology; (b) and (c) courtesy Nuffield Radio Astronomy Laboratories, University of Manchester; (d) adapted from Mitton and Ryle (1969), Baade and Minkowski (1954), Jennison and Das Gupta (1953).]
These discoveries were startling enough to generate, at long last, strong interest among optical astronomers. Jesse Greenstein was no longer the only one paying serious attention.
For Greenstein himself, these discoveries were the final straw. Having failed to push into radio work right after the war, Americans were now bystanders in the greatest revolution to hit astronomy since Galileo invented the optical telescope. The rewards of the revolution were being reaped in Britain and Australia, and not in America.
Greenstein was now a professor at Caltech. He had been brought there from Yerkes to build an astronomy program around the new 5-meter optical telescope, so naturally, he now went to Lee DuBridge, the Caltech president, and urged that Caltech build a radio interferometer to be used hand in hand with the 5-meter in exploring distant galaxies. DuBridge, having been director of the American radar effort during the war, was sympathetic, but cautious. To swing DuBridge into action, Greenstein organized an international conference on the future of radio astronomy in Washington, D.C., on 5 and 6 January 1954.
In Washington, after the representatives from the great British and Australian radio observatories had described their remarkable discoveries, Greenstein posed his question: Must the United States continue as a radio astronomy wasteland? The answer was obvious.
With strong backing from the National Science Foundation, American physicists, engineers, and astronomers embarked on a crash program to construct a National Radio Astronomy Observatory in Greenbank, West Virginia; and DuBridge approved Greenstein’s proposal for a state-of-the-art Caltech radio inteferometer, to be built in Owens Valley, California, just southeast of Yosemite National Park. Since nobody at Caltech had the expertise to build such an instrument, Greenstein lured John Bolton from Australia to spearhead the effort.
Quasars
By the late 1950s, the Americans were competitive. Radio telescopes at Greenbank were coming into operation, and at Caltech, Tom Mathews, Per Eugen Maltby, and Alan Moffett on the new Owens Valley radio interferometer were working hand in hand with Baade, Greenstein, and others on the Palomar 5-meter optical telescope to discover and study large numbers of radio galaxies.
In 1960 this effort brought another surprise: Tom Mathews at Caltech received word from Henry Palmer that, according to Jodrell Bank measurements, a radio source named 3C48 (the 48th source in the third version of a catalog constructed by Ryle’s group at Cambridge) was extremely small, no more than 1 arc second in diameter (1/2,000 of the angular size of the Sun). So tiny a source would be something quite new. However, Palmer and his Jodrell Bank colleagues could not provide a tight error box for the source’s location. Mathews, in exquisitely beautiful work with Caltech’s new radio interferometer, produced an error box just 5 seconds of arc in size, and gave it to Allan Sandage, an optical astronomer at the Carnegie Institution in Pasadena. On his next observing run on the 5-meter optical telescope, Sandage took a photograph centered on Mathews’s error box and found, to his great surprise, not a galaxy, but a single, blue point of light; it looked like a star. “I took a spectrum the next night and it was the weirdest spectrum I’d ever seen,” Sandage recalls. The wavelengths of the spectral lines were not at all like those of stars or of any hot gas ever manufactured on Earth; they were unlike anything ever before encountered by astronomers or physicists. Sandage could not make any sense at all out of this weird object.
Over the next two years a half-dozen similar objects were discovered by the same route, each as puzzling as 3C48. All the optical astronomers at Caltech and Carnegie began photographing them, taking spectra, struggling to understand their nature. The answer should have been obvious, but it was not. A mental block held sway. These weird objects looked so much like stars that the astronomers kept trying to interpret them as a type of star in our own galaxy that had never before been seen, but the interpretations were horrendously contorted, not really believable.
The mental block was broken by Maarten Schmidt, a thirty-two-year-old Dutch astronomer who had recently joined the Caltech faculty. For months he had struggled to understand a spectrum he had taken of 3C273, one of the weird objects. On 5 February 1963, as he sat in his Caltech office carefully sketching the spectrum for inclusion in a manuscript he was writing, the answer suddenly hit him. The four brightest lines in the spectrum were the four standard “Balmer lines” produced by hydrogen gas—the most famous of all spectral lines, the first lines that college physics students learned about in their courses on quantum mechanics. However, these four lines did not have their usual wavelengths. Each was shifted to the red by 16 percent. 3C273 must be an object containing a massive amount of hydrogen gas and moving away from the Earth at 16 percent of the speed of light—enormously faster than any star that any astronomer had ever seen.
Schmidt flew out into the hall, ran into Greenstein, and excitedly described his discovery. Greenstein turned, headed back to his office, pulled out his spectrum of 3C48, and stared at it for a while. Balmer lines were not present at any redshift; but lines emitted by magnesium, oxygen, and neon were there staring him in the face, and they had a redshift of 37 percent. 3C48 was, at least in part, a massive amount of gas containing magnesium, oxygen, and neon, and moving away from Earth at about 37 percent of the speed of light.
What was producing these high speeds? If, as everyone had thought, these weird objects (which would later be named quasars) were some type of star in our own Milky Way galaxy, then they must have been ejected from somewhere, perhaps the Milky Way’s central nucleus, with enormous force. This was too incredible to believe, and a close examination of the quasars’ spectra made it seem extremely unlikely. The only reasonable alternative, Greenstein and Schmidt argued (correctly), was that these quasars are very far away in our Universe, and move away from Earth at high speed as a result of the Universe’s expansion.
Left: Jesse L. Greenstein with a drawing of the Palomar 5-meter optical telescope, ca. 1955. Right: Maarten Schmidt, with an instrument for measuring spectra made by the 5-meter telescope, ca. 1963. [Courtesy the Archives, California Institute of Technology.]
Recall that the expansion of the Universe is like the expansion of the surface of a balloon that is being blown up. If a number of ants are standing on the balloon’s surface, each ant will see all the other ants move away from him as a result of the balloon’s expansion. The farther away another ant is, the faster the first ant will see it move. Similarly, the farther away a distant object is from Earth, the faster we on Earth will see it move as a result of the Universe’s expansion. In other words, the object’s speed is proportional to its distance. Therefore, from the speeds of 3C273 and 3C48, Schmidt and Greenstein could infer their distances: 2 billion light-years and 4.5 billion light-years, respectively.
These were enormous distances, nearly the largest distances ever yet recorded. This meant that, in order for 3C273 and 3C48 to be as bright as they appear in the 5-meter telescope, they had to radiate enormous amounts of power: 100 times more power than the most luminous galaxies ever seen.
3C273, in fact, was so bright that, along with many other objects near it on the sky, it had been photographed more than 2000 times since 1895 using modest-sized telescopes. Upon learning of Schmidt’s discovery, Harlan Smith of the University of Texas organized a close examination of this treasure trove of photographs, archived largely at Harvard, and discovered that 3C273 had been fluctuating in brightness during the past seventy years. Its light output had changed substantially within periods as short as a month. This means that a large portion of the light from 3C273 must come from a region smaller than the distance light travels in a month, that is, smaller than 1 “light-month.” (If the region were larger, then there would be no way that any force, traveling, of course, at a speed less than or equal to that of light, could make the emitting gas all brighten up or dim out simultaneously to within an accuracy of a month.)
The implications were extremely hard to believe. This weird quasar, this 3C273, was shining 100 times more brightly than the brightest galaxies in the Universe; but whereas galaxies produce their light in regions 100,000 light-years in size, 3C273 produces its light in a region at least a million times smaller in diameter and 1018 times smaller in volume: just a light-month or less. The light must come from a massive, compact, gaseous object that is heated by an enormously powerful engine. The engine would ultimately turn out to be, with high but not complete confidence, a gigantic black hole, but strong evidence for this was still fifteen years into the future.
If explaining Jansky’s radio waves from our own Milky Way galaxy was difficult, and explaining the radio waves from distant radio galaxies was even more difficult, then the explanation for radio waves from these superdistant quasars would have to be superdifficult.
The difficulty, it turned out, was an extreme mental block. Jesse Greenstein, Fred Whipple, and all other astronomers of the 1930s and 1940s had presumed that cosmic radio waves, like light from stars, are emitted by the heat-induced jiggling of atoms, molecules, and electrons. Astronomers of the thirties and forties could not conceive of any other way for nature to create the observed radio waves, even though their calculations showed unequivocally that this way can’t work.
Another way, however, had been known to physicists since the early twentieth century: When an electron, traveling at high speed, encounters a magnetic field, the field’s magnetic force twists the electron’s motion into a spiral. The electron is forced to spiral around and around the magnetic field lines (Figure 9.4), and as it spirals, it emits electromagnetic radiation. Physicists in the 1940s began to call this radiation synchrotron radiation, because it is produced by spiraling electrons in the particle accelerators called “synchrotrons” that they were then building. Remarkably, in the 1940s, despite physicists’ considerable interest in synchrotron radiation, astronomers paid no attention to it. The astronomers’ mental block held sway.
9.4 Cosmic radio waves are produced by near-light-speed electrons that spiral around and around in magnetic fields. The magnetic field forces an electron to spiral instead of moving on a straight line, and the electron’s spiraling motion produces the radio waves.
In 1950 Karl Otto Kiepenheuer in Chicago and Vitaly Lazarevich Ginzburg in Moscow (the same Ginzburg who had invented the LiD fuel for the Soviet hydrogen bomb, and who had discovered the first hint that black holes have no hair2) broke the mental block. Building on seminal ideas of Hans Alfven and Nicolai Herlofson, Kiepenheuer and Ginzburg proposed (correctly) that Jansky’s radio waves from our own galaxy are synchrotron radiation produced by near-light-speed electrons spiraling around magnetic field lines that fill interstellar space (Figure 9.4).
A few years later, when the giant radio-emitting lobes of radio galaxies and then quasars were discovered, it was natural (and correct) to conclude that their radio waves were also produced by electrons spiraling around magnetic field lines. From the physical laws governing such spiraling and the properties of the observed radio waves, Geoffrey Burbidge at the University of California in San Diego computed how much energy the lobes’ magnetic field and high-speed electrons must have. His startling answer: In the most extreme cases, the radio-emitting lobes must have about as much magnetic energy and high-speed (kinetic) energy as one would get by converting all the mass of 10 million (107) Suns into pure energy with 100 percent efficiency.
These energy requirements of quasars and radio galaxies were so staggering that they forced astrophysicists, in 1963, to examine all conceivable sources of power in search of an explanation.
Chemical power (the burning of gasoline, oil, coal, or dynamite), which is the basis of human civilization, was clearly inadequate. The chemical efficiency for converting mass into energy is no larger than 1 part in a billion (1 part in 109). To energize a quasar’s radio-emitting gas would therefore require 109 × 107 = 1016 solar masses of chemical fuel—100,000 times more fuel than is contained in our entire Milky Way galaxy. This seemed totally unreasonable.
Nuclear power, the basis of the hydrogen bomb and of the Sun’s heat and light, looked only marginal as a way to energize a quasar. Nuclear fuel’s efficiency for mass-to-energy conversion is roughly 1 percent (1 part in 102), so a quasar would need 102 × 107 = 109 (1 billion) solar masses of nuclear fuel to energize its radio-emitting lobes. And this 1 billion solar masses would be adequate only if the nuclear fuel were burned completely and the resulting energy were converted completely into magnetic fields and kinetic energy of high-speed electrons. Complete burning and complete energy conversion seemed highly unlikely. Even with carefully contrived machines, humans rarely achieve better than a few percent conversion of fuel energy into useful energy, and nature without careful designs might well do worse. Thus, 10 billion or 100 billion solar masses of nuclear fuel seemed more reasonable. Now, this is less than the mass of a giant galaxy, but not a lot less, and how nature might achieve the conversion of the fuel’s nuclear energy into magnetic and kinetic energy was very unclear. Thus, nuclear fuel was a possibility, but not a likely one.
The annihilation of matter with antimatter3 could give 100 percent conversion of mass to energy, so 5 million solar masses of antimatter annihilating with 5 million solar masses of matter could satisfy a quasar’s energy needs. However, there is no evidence that any antimatter exists in our Universe, except tiny bits created artifically by humans in particle accelerators and tiny bits created by nature in collisions between matter particles. Moreover, even if so much matter and antimatter were to annihilate in a quasar, their annihilation energy would go into very high energy gamma rays, and not into magnetic energy and electron kinetic energy. Thus, matter/antimatter annihilation appeared to be a very unsatisfactory way to energize a quasar.
One other possibility remained: gravity. The implosion of a normal star to form a neutron star or a black hole might, conceivably, convert 10 percent of the star’s mass into magnetic and kinetic energy—though precisely how was unclear. If it managed to do so, then the implosions of 10 × 107 = 108 (100 million) normal stars might provide a quasar’s energy, as would the implosion of a single, hypothetical, supermassive star 100 million times heavier than the Sun. [The correct idea, that the gigantic black hole produced by the implosion of such a supermassive star might itself be the engine that powers the quasar, did not occur to anybody in 1963. Black holes were but poorly understood. Wheeler had not yet coined the phrase “black hole” (Chapter 6). Salpeter and Zel’dovich had not yet realized that gas falling toward a black hole could heat and radiate with high efficiency (Chapter 8). Penrose had not yet discovered that a black hole can store up to 29 percent of its mass as rotational energy, and release it (Chapter 7). The golden age of black-hole research had not yet begun.]
The idea that the implosion of a star to form a black hole might energize quasars was a radical departure from tradition. This was the first time in history that astronomers and astrophysicists had felt a need to appeal to effects of general relativity to explain an object that was being observed. Previously, relativists had lived in one world and astronomers and astrophysicists in another, hardly communicating. Their insularity was about to end.
To foster dialogue between the relativists and the astronomers and astrophysicists, and to catalyze progress in the study of quasars, a conference of three hundred scientists was held on 16–18 December 1963, in Dallas, Texas. In an after-dinner speech at this First Texas Symposium on Relativistic Astrophysics, Thomas Gold of Cornell University described the situation, only partially with tongue in cheek: “[The mystery of the quasars] allows one to suggest that the relativists with their sophisticated work are not only magnificent cultural ornaments but might actually be useful to science! Everyone is pleased: the relativists who feel they are being appreciated and are experts in a field they hardly knew existed, the astrophysicists for having enlarged their domain, their empire, by the annexation of another subject—general relativity. It is all very pleasing, so let us all hope that it is right. What a shame it would be if we had to go and dismiss all the relativists again.”
Lectures went on almost continuously from 8:30 in the morning until 6 in the evening with an hour out for lunch, plus 6 P.M. until typically 2 A.M. for informal discussions and arguments. Slipped in among the lectures was a short, ten-minute presentation by a young New Zealander mathematician, Roy Kerr, who was unknown to the other participants. Kerr had just discovered his solution of the Einstein field equation—the solution which, one decade later, would turn out to describe all properties of spinning black holes, including their storage and release of rotational energy (Chapters 7 and 11); the solution which, as we shall see below, would ultimately become a foundation for explaining the quasars’ energy. However, in 1963 Kerr’s solution seemed to most scientists only a mathematical curiosity; nobody even knew it described a black hole—though Kerr speculated it might somehow give insight into the implosion of rotating stars.
The astronomers and astrophysicists had come to Dallas to discuss quasars; they were not at all interested in Kerr’s esoteric mathematical topic. So, as Kerr got up to speak, many slipped out of the lecture hall and into the foyer to argue with each other about their favorite theories of quasars. Others, less polite, remained seated in the hall and argued in whispers. Many of the rest catnapped in a fruitless effort to remedy their sleep deficits from late-night science. Only a handful of relativists listened, with rapt attention.
This was more than Achilles Papapetrou, one of the world’s leading relativists, could stand. As Kerr finished, Papapetrou demanded the floor, stood up, and with deep feeling explained the importance of Kerr’s feat. He, Papapetrou, had been trying for thirty years to find such a solution of Einstein’s equation, and had failed, as had many other relativists. The astronomers and astrophysicists nodded politely, and then, as the next speaker began to hold forth on a theory of quasars, they refocused their attention, and the meeting picked up pace.
The 1960s marked a turning point in the study of radio sources. Previously the study was totally dominated by observational astron?mers—that is, optical astronomers and the radio-observing experimental physicists, who were now being integrated into the astronomical community and called radio astronomers. Theoretical astrophysicists, by contrast, had contributed little, because the radio observations were not yet detailed enough to guide their theorizing very much. Their only contributions had been the realization that the radio waves are produced by high-speed electrons spiraling around magnetic field lines in the giant radio-emitting lobes, and their calculation of how much magnetic and kinetic energy this entails.
In the 1960s, as the resolutions of radio telescopes continued to improve and optical observations began to reveal new features of the radio sources (for example, the tiny sizes of the light-emitting cores of quasars), this growing body of information became grist for the minds of astrophysicists. From this rich information, the astrophysicists generated dozens of detailed models to explain radio galaxies and quasars, and then one by one their models were disproved by accumulating observational data. This, at last, was how science was supposed to work!
One key piece of information was the radio astronomers’ discovery that radio galaxies emit radio waves not only from their giant double lobes, one on each side of the central galaxy, but also from the core of the central galaxy itself. In 1971, this suggested to Martin Rees, a recent student of Dennis Sciama’s in Cambridge, a radically new idea about the powering of the double lobes. Perhaps a single engine in the galaxy’s core was responsible for all the galaxy’s radio waves. Perhaps this engine was directly energizing the core’s radio-emitting electrons and magnetic fields, perhaps it was also beaming up power to the giant lobes, to energize their electrons and fields, and perhaps this engine in the cores of radio galaxies was of the same sort (whatever that might be) that powers quasars.
Rees initially suspected that the beams that carry power from the core to the lobes were made of ultra-Iow-frequency electromagnetic waves. However, theoretical calculations soon made it clear that such electromagnetic beams cannot penetrate through the galaxy’s interstellar gas, no matter how hard they try.
As is often the case, Rees’s not quite correct idea stimulated a correct one. Malcolm Longair, Martin Ryle, and Peter Scheuer in Cambridge took the idea and modified it in a simple way: They kept Rees’s beams, but made them of hot, magnetized gas rather than electromagnetic waves. Rees quickly agreed that this kind of gas jet would do the job, and with his student Roger Blandford he computed the properties that the gas jets should have.
A few years later, this prediction, that the radio-emitting lobes are powered by jets of gas emerging from a central engine, was spectacularly confirmed using huge new radio interferometers in Britain, Holland, and America—most notably the American VLA (very large array) on the plains of St. Augustin in New Mexico (Figure 9.5). The interferometers saw the jets, and the jets had just the predicted properties. They reached from the galaxy’s core to the two lobes, and they could even be seen ramming into gas in the lobes and being slowed to a halt.
The VLA uses the same “spots on the bowl” technique as the radio interferometers of the 1940s and 1950s (Figure 9.2), but its bowl is much larger and it uses many more spots (many more linked radio telescopes). It achieves resolutions as good as 1 arc second, about the same as the world’s best optical telescopes—a tremendous achievement when one contemplates the crudeness of Jansky’s and Reber’s original instruments forty years earlier. But the improvements did not stop there. By the early 1980s, pictures of the cores of radio galaxies and quasars, with resolutions 1000 times better than optical telescopes, were being produced by very long baseline inteiferometers (VLBIs) composed of radio telescopes on opposite sides of a continent or the world. (The output of each telescope in a VLBI is recorded on magnetic tape, along with time markings from an atomic clock, and the tapes from all the telescopes are then played into a computer where they are “interfered” with each other to make the pictures.)
9.5 The VLA radio interferometer on the plains of St. Augustin in New Mexico. [courtesy NRAO/AUI.]
A picture of the radio emission from the radio galaxy Cygnus A made with the VLA by R. A. Perley, J.W. Dreyer, and JJ. Cowan. The jet that feeds the right-hand radio lobe is quite clear; the jet feeding the left lobe is much fainter. Notice the enormous improvement in resolution of this radio-wave picture compared with Reber’s 1944 contour map which did not show the double lobes at all (Figure 9.1d), and with Jennison and Das Gupta’s 1953 radio map which barely revealed the existence of the lobes (two rectangles in Figure 9.3d), and with Ryle’s 1969 contour map (Figure 9.3d). [courtesy NRAO/AUI.]
These VLBI pictures showed, in the early 1980s, that the jets extend right into the innermost few light-years of the core of a galaxy or quasar—the very region in which resides, in the case of some quasars such as 3C273, a brilliantly luminous, light-emitting object no larger than a light-month in size. Presumably the central engine is inside the light-emitting object, and it is powering not only that object, but also the jets, which then feed the radio lobes.
The jets gave yet another clue to the nature of the central engine. Some jets were absolutely straight over distances of a million light-years or more. If the source of such jets were turning, then, like a rotating water nozzle on a sprinkler, it would produce bent jets. The observed jets’ straightness thus meant that the central engine had been firing its jets in precisely the same direction for a very long time. How long? Since the jets’ gas cannot move faster than the speed of light, and since some straight jets were longer than a million light-years, the firing direction must -have been steady for more than a million years. To achieve such steadiness, the engine’s “nozzles,” which eject the jets, must be attached to a superbly steady object—a long-lived gyroscope of some sort. (Recall that a gyroscope is a rapidly spinning object that holds the direction of its spin axis steadily fixed over a very long time. Such gyroscopes are key components of inertial navigation systems for airplanes and missiles.)
Of the dozens of ideas that had been proposed by the early 1980s to explain the central engine, only one entailed a superb gyroscope with a long life, a size less than a light-month, and an ability to generate powerful jets. That unique idea was a gigantic, spinning black hole.
The idea that gigantic black holes might power quasars and radio galaxies was conceived by Edwin Salpeter and Yakov Borisovich Zel’dovich in 1964 (the first year of the golden age—Chapter 7). This idea was an obvious application of the Salpeter–Zel’dovich discovery that gas streams, falling toward a black hole, should collide and radiate (see Figure 8.4).
A more complete and realistic description of the fall of gas streams toward a black hole was devised in 1969 by Donald Lynden-Bell, a British astrophysicist in Cambridge. Lynden-Bell argued, convincingly, that after the gas streams collide, they will join together, and then centrifugal forces will make them spiral around and around the hole many times before falling in; and as they spiral, they will form a disk-shaped object, much like the rings around the planet Saturn—san accretion disk Lynden-Bell called it, since the gas is “accreting” onto the hole. (The right half of Figure 8.7 shows an artist’s conception of such an accretion disk around the modest-sized hole in Cygnus X-l.) In the accretion disk, adjacent gas streams will rub against each other, and intense friction from that rubbing will heat the disk to high temperatures.
In the 1980s, astrophysicists realized that the brilliant light-emitting object at the center of 3C273, the object 1 light-month or less in size, was probably Lynden-Bell’s friction-heated accretion disk.
We normally think of friction as a poor source of heat. Recall the unfortunate Boy Scout who tries to start a fire by rubbing two sticks together! However, the Boy Scout is limited by his meager muscle power, while an accretion disk’s friction feeds off gravitational energy. Since the gravitational energy is enormous, far larger than nuclear energy, the friction is easily up to the task of heating the disk and making it shine 100 times more brightly than the most luminous galaxies.
How can a black hole act as a gyroscope? James Bardeen and Jacobus Petterson of Yale University realized the answer in 1975: If the black hole spins rapidly, then it behaves precisely like a gyroscope. Its spin direction remains always firmly fixed and unchanging, and the swirl of space near the hole created by the spin (Figure 7.7) remains always firmly oriented in the same direction. Bardeen and Petterson showed by a mathematical calculation that this near-hole swirl of space must grab the inner part of the accretion disk and hold it firmly in the hole’s equatorial plane—and must do so no matter how the disk is oriented far from the hole (Figure 9.6). As new gas from interstellar space is captured into the distant part of the disk, it may change the distant disk’s orientation, but it can never change the disk’s orientation near the hole. The hole’s gyroscopic action prevents it. Near the hole the disk remains always in the hole’s equatorial plane.
Without Kerr’s solution to the Einstein field equation, this gyroscopic action would have been unknown, and it might have been impossible to explain quasars. With Kerr’s solution in hand, astrophysicists in the mid-1970s were arriving at a clear and elegant explanation. For the first time, the concept of a black hole as a dynamical body, more than just a “hole in space,” was playing a central role in explaining astronomers’ observations.
How strong will the swirl of space be near the gigantic hole? In other words, how fast will gigantic holes spin? James Bardeen deduced the answer: He showed mathematically that gas accreting into the hole from its disk should gradually make the hole spin faster and faster. By the time the hole has swallowed enough inspiraling gas to double its mass, the hole should be spinning at nearly its maximum possible rate—the rate beyond which centrifugal forces prevent any further speedup (Chapter 7). Thus, gigantic holes should typically have near-maximal spins.
9.6 The spin of a black hole produces a swirl of space around the hole, and that swirl holds the inner part of the accretion disk in the hole’s equatorial plane.
How can a black hole and its disk produce two oppositely pointed jets? Amazingly easily, Blandford, Rees, and Lynden-Bell at Cambridge University recognized in the mid-1970s. There are four possible ways to produce jets; anyone of them might do the job.
First, Blandford and Rees realized, the disk may be surrounded by a cool gas cloud (Figure 9.7a). A wind blowing off the upper and lower faces of the disk (analogous to the wind that blows off the Sun’s surface) may create a bubble of hot gas inside the cool cloud. The hot gas may then punch orifices in the cool cloud’s upper and lower faces and flow out of them. Just as a nozzle on a garden hose collimates outflowing water to form a fast, thin stream, so the orifices in the cool cloud should collimate the outflowing hot gas to form thin jets. The directions of the jets will depend on the locations of the orifices. The most likely locations, if the cool cloud spins about the same axis as the black hole, are along the common spin axis, that is, perpendicular to the plane of the inner part of the accretion disk—and the orifices at these locations will produce jets whose direction is anchored to the black hole’s gyroscopic spin.
Second, because the disk is so hot, its internal pressure is very high, and this pressure might puff the disk up until it becomes very thick (Figure 9.7b). In this case, Lynden-Bell pointed out, the orbital motion of the disk’s gas will produce centrifugal forces that create whirlpool-like funnels in the top and bottom faces of the disk. These funnels are precisely analogous to the vortex that sometimes forms when water swirls down the drainhole of a bathtub. The black hole is like the drainhole, and the disk’s gas is like the water. The faces of the vortex-like funnels should be so hot, because of friction in the gas, that they blow a strong wind off themselves, and the funnels might then collimate this wind into jets, Lynden-Bell reasoned. The jets’ directions will be the same as the funnels’, which in turn are firmly anchored to the hole’s gyroscopic spin axis.
Third, Blandford realized, magnetic field lines anchored in the disk and sticking out of it will be forced, by the disk’s orbital motion, to spin around and around (Figure 9.7c). The spinning field lines will assume an outward and upward (or outward and downward) spiraling shape. Electrical forces should anchor hot gas (plasma) onto the spinning field lines; the plasma can slide along the field lines but not across them. As the field lines spin, centrifugal forces should fling the plasma outward along them to form two magnetized jets, one shooting outward and upward, the other outward and downward. Again the jets’ directions will be firmly anchored to the hole’s spin.
9.7 Four methods by which a black hole or its accretion disk could power twin jets. (a) A wind from the disk blows a bubble in a surrounding, spinning gas cloud; the bubble’s hot gas punches orifices through the cloud, along its spin axis; and jets of hot gas shoot out the orifices. (b) The disk is puffed up by the pressure of its great internal heat, and the surface of the puffed, rotating disk forms two funnels that collimate the disk’s wind into two jets. (c) Magnetic field lines anchored in the disk are forced to spin by the disk’s orbital rotation; as they spin, the field lines fling plasma upward and downward, and the plasma, sliding along the field lines, forms two magnetized jets. (d) Magnetic field lines threading through the black hole are forced to spin by the swirl of the hole’s space, and as they spin, the field lines fling plasma upward and downward to form two magnetized jets.
The fourth method of producing jets is more interesting than the others and requires more explanation. In this fourth method, the hole is threaded by magnetic field lines as shown in Figure 9.7d. As the hole spins, it drags the field lines around and around, causing them to fling plasma upward and downward in much the same manner as the third method, to form two jets. The jets shoot out along the hole’s spin axis and their direction thus is firmly anchored to the hole’s gyroscopic spin. This method was conceived of by Blandford soon after he received his Ph.D. in Cambridge, together with a Cambridge graduate student, Roman Znajek, and it thus is called the Blandford–Znajek process.
The Blandford–Znajek process is especially interesting, because the power that goes into the jets comes from the hole’s enormous rotational energy. (This should be obvious since it is the hole’s spin that causes space to swirl, and the swirl of space that causes the magnetic field lines to rotate, and the field lines’ rotation that flings plasma outward.)
How is it possible, in this Blandford–Znajek process, for the hole’s horizon to be threaded by magnetic field lines? Such field lines would be a form of “hair” that can be converted into electromagnetic radiation and be radiated away, and therefore, according to Price’s theorem (Chapter 7), they must be radiated away. In fact, Price’s theorem is correct only if the black hole is sitting alone, far from all other objects. The hole we are discussing, however, is not alone; it is surrounded by an accretion disk. If the field lines of Figure 9.7d pop off the hole, the lines going out the hole’s northern hemisphere and those going out its southern hemisphere will turn out to be continuations of each other, and the only way these lines can then escape is by pushing their way out through the accretion disk’s hot gas. But the hot gas will not let the field lines through; it confines them firmly into the region of space inside the disk’s inner face, and since most of that region is occupied by the hole, most of the confined field lines thread through the hole.
Where do these magnetic field lines come from? From the disk itself. All gas in the Universe is magnetized, at least a little bit, and the disk’s gas is no exception.4 As, bit by bit, the disk’s gas accretes into the hole, it carries its magnetic field lines with it. Upon nearing the hole, each bit of gas slides down its magnetic field lines and through the horizon, leaving the field lines behind, sticking out of the horizon and threading it in the manner of Figure 9.7d. These threading field lines, firmly confined by the surrounding disk, should then extract the hole’s rotational energy by the Blandford–Znajek process.
All four methods of producing jets (orifices in a gas cloud, wind from a funnel, whirling field lines anchored in a disk, and the Blandford–Znajek process) probably operate, to varying degrees, in quasars, in radio galaxies, and in the peculiar cores of some other types of galaxies (cores that are called active galactic nuclei).
If quasars and radio galaxies are powered by the same kind of black-hole engine, what makes them look so different? Why does the light of a quasar appear to come from an intensely luminous, star-like object, 1 light-month in size or less, while the light of a radio galaxy comes from a Milky Way–like assemblage of stars, 100,000 light-years in size?
It seems almost certain that quasars are not much different from radio galaxies; their central engines are also surrounded by a 100,000-light-year-sized galaxy of stars. However, in a quasar, the central black hole is fueled at an especially high rate by accreting gas (Figure 9.8), and frictional heating in the disk is correspondingly high. This huge heating makes the disk shine so strongly that its optical brilliance is hundreds or thousands of times greater than that of all the stars in the surrounding galaxy put together. Astronomers, blinded by the brilliance of the disk, cannot see the galaxy’s stars, and thus the object looks “quasi-stellar” (that is, star-like; like a tiny, intense point of light) instead of looking like a galaxy.5
The innermost region of the disk is so hot that it emits X-rays; a little farther out, the disk is cooler and emits ultraviolet radiation; still farther out it is cooler still and emits optical radiation (light); and in its outermost region it is even cooler and emits infrared radiation. The light-emitting region is typically about a light-year in size, though in some cases such as 3C273 it can be a light-month or smaller and thus can vary in brightness over periods as short as a month. Much of the X-ray radiation and ultraviolet light pouring out of the innermost region hits and heats gas clouds several light-years from the disk; it is those heated clouds that emit the spectral lines by which the quasars were first discovered. A magnetized wind blowing off the disk, in some quasars but not all, will be strong enough and well enough collimated to produce radio-emitting jets.
9.8 Our best present understanding of the structures of quasars and radio galaxies. This detailed model, based on all the observational data, has been developed by Sterl Phinney of Caltech and others.
In a radio galaxy, by contrast with a quasar, the central accretion disk presumably is rather quiescent. Quiescence means small friction in the disk, and thus small heating and low luminosity, so that the disk shines much less brightly than the rest of the galaxy. Astronomers thus see the galaxy and not the disk through their optical telescopes. However, the disk, the spinning hole, and magnetic fields threading through the hole together produce strong jets, probably in the manner of Figure 9.7d (the Blandford–Znajek process), and those jets shoot out through the galaxy and into intergalactic space, where they feed energy into the galaxy’s huge radio-emitting lobes.
These black-hole-based explanations for quasars and radio galaxies are so successful that it is tempting to assert they must be right, and a galaxy’s jets must be a unique signature crying out to us “I come from a black hole!” However, astrophysicists are a bit cautious. They would like a more ironclad case. It is still possible to explain all the observed properties of radio galaxies and quasars using an alternative, non–black-hole engine: a rapidly spinning, magnetized, supermassive star, one weighing millions or billions of times as much as the Sun—a type of star that has never been seen by astronomers, but that theory suggests might form at the centers of galaxies. Such a supermassive star would behave much like a hole’s accretion disk. By contracting to a small size (but a size still larger than its critical circumference), it could release a huge amount of gravitational energy; that energy, by way of friction, could heat the star so it shines brightly like an accretion disk; and magnetic field lines anchored in the star could spin and fling out plasma in jets.
It might be that some radio galaxies or quasars are powered by such supermassive stars. However, the laws of physics insist that such a star should gradually contract to a smaller and smaller size, and then, as it nears its critical circumference, should implode to form a black hole. The star’s total lifetime before implosion should be much less than the age of the Universe. This suggests that, although the youngest of radio galaxies and quasars might be powered by supermassive stars, older ones are almost certainly powered, instead, by gigantic holes—almost certainly, but not absolutely certainly. These arguments are not ironclad.
How common are gigantic black holes? Evidence, gradually accumulated during the 1980s, suggests that such holes inhabit not only the cores of most quasars and radio galaxies, but also the cores of most large, normal (non-radio) galaxies such as the Milky Way and Andromeda, and even the cores of some small galaxies such as Andromeda’s dwarf companion, M32. In normal galaxies (the Milky Way, Andromeda, M32) the black hole presumably is surrounded by no accretion disk at all, or by only a tenuous disk that pours out only modest amounts of energy.
The evidence for such a hole in our own Milky Way galaxy (as of 1993) is suggestive, but far from firm. One key bit of evidence comes from the orbital motions of gas clouds near the center of the galaxy. Infrared observations of those clouds, by Charles Townes and colleagues at the University of California at Berkeley, show that they are orbiting around an object which weighs about 3 million times as much as the Sun, and radio observations reveal a very peculiar, though not strong, radio source at the position of the central object—a radio source amazingly small, no larger than our solar system. These are the types of observations one might expect from a quiescent, 3-million-solar-mass black hole with only a tenuous accretion disk; but they are also readily explained in other ways.
The possibility that gigantic black holes might exist and inhabit the cores of galaxies came as a tremendous surprise to astronomers. In retrospect, however, it is easy to understand how such holes might form in a galactic core.
In any galaxy, whenever two stars pass near each other, their gravitational forces swing them around each other and then fling them off in directions different from their original paths. (This same kind of swing and fling changes the orbits of NASA’s spacecraft when they encounter planets such as Jupiter.) In the swing and fling, one of the stars typically gets flung inward, toward the galaxy’s center, while the other gets flung outward, away from the center. The cumulative effect of many such swings and flings is to drive some of the galaxy’s stars deep down into the galaxy’s core. Similarly, it turns out, the cumulative effect of friction in the galaxy’s interstellar gas is to drive much of the gas down into the galaxy’s core.
As more and more gas and stars accumulate in the core, the gravity of the agglomerate they form should become stronger and stronger. Ultimately, the agglomerate’s gravity may become so strong as to overwhelm its internal pressure, and the agglomerate may implode to form a gigantic hole. Alternatively, massive stars in the agglomerate may implode to form small holes, and those small holes may collide with each other and with stars and gas to form ever larger and larger holes, until a single gigantic hole dominates the core. Estimates of the time required for such implosions, collisions, and coalescences make it seem plausible (though not compelling) that most galaxies will have grown gigantic black holes in their cores long before now.
If astronomical observations did not strongly suggest that the cores of galaxies are inhabited by gigantic black holes, astrophysicists even today, in the 1990s, would probably not predict it. However, since the observations do suggest gigantic holes, astrophysicists easily accommodate themselves to the suggestion. This is indicative of our poor understanding of what really goes on in the cores of galaxies.
What of the future? Need we worry that the gigantic hole in our Milky Way galaxy might swallow the Earth? A few numbers set one’s mind at ease. Our galaxy’s central hole (if it indeed exists) weighs about 3 million times what the Sun weighs, and thus has a circumference of about 50 million kilometers, or 200 light-seconds—about onetenth the circumference of the Earth’s orbit around the Sun. This is tiny by comparison with the size of the galaxy itself. Our Earth, along with the Sun, is orbiting around the galaxy’s center on an orbit with a circumference of 200,000 light-years—about 30 billion times larger than the circumference of the hole. If the hole were ultimately to swallow most of the mass of the galaxy, its circumference would expand only to about 1 light-year, still 200,000 times smaller than the circumference of our orbit.
Of course, in the roughly 1018 years (100 million times the Universe’s present age) that it will require for our central hole to swallow a large fraction of the mass of our galaxy, the orbit of the Earth and Sun will change substantially. It is not possible to predict the details of those changes, since we do not know well enough the locations and motions of all the other stars that the Sun and Earth may encounter during 1018 years. Thus, we cannot predict whether the Earth and Sun will wind up, ultimately, inside the galaxy’s central hole, or will be flung out of the galaxy. However, we can be confident that, if the Earth ultimately gets swallowed, its demise is roughly 1018 years in the future—so far off that many other catastrophes will almost certainly befall the Earth and humanity in the meantime.
1. By light, I always mean in this book the type of electromagnetic waves that the human eye can see; that is, optical radiation.
2. See Figure 7.3. Ginzburg is best known not for these discoveries, but for yet another: his development, with Lev Landau, of the “Ginzburg–Landau theory” of superconductivity (that is, an explanation for how it is that some metals, when made very cold, lose all their resistance to the flow of electricity). Ginzburg is one of the world’s few true “Renaissance physicists,” a man who has contributed significantly to almost all branches of theoretical physics.
3. For background, see the entry “antimatter” in the glossary, and Footnote 2 in Chapter 5.
4. The magnetic fields have been built up continually over the life of the Universe by the motions of interstellar and stellar gas, and once generated, the magnetic fields are extremely hard to get rid of. When interstellar gas accumulates into the accretion disk, it carries its magnetic fields with itself.
5. The word “quasar” is shorthand for “quasi-stellar.”
