A persistent and universal symbol in the mythology of virtually every culture is that of a bottomless pit or an engulfing whirlpool. It was the maw of the abyss, and those venturing too close were dragged inward toward Chaos by an irresistible force. Socrates talked of a chasm that pierced the world straight through from side to side; Ulysses encountered it, as did a mythical Cherokee who escaped, but not before he was drawn down to the narrowest circle of the maelstrom where he could see into the nether world of the dead. And so it goes through Polynesian, Indian, Icelandic, Egyptian, and Sumerian mythology. Tradition is ambiguous as to the location of the whirlpool in the mythical landscape. Usually it was just beyond the boundaries of the known universe on the high seas, but some mythologies located it - properly, many astronomers would say -in the heavens, in Sagittarius in the Milky Way. That is, in the center of our galaxy.
In a sense, the search for a solution to one of astronomy's most persistent and perplexing riddles, the source of violent activity in galaxies and quasars, could be viewed as a continuation of the search for the whirlpool that is the maw of the abyss, a depth our telescopes cannot reach, and from which nothing returns. What is incredible to contemplate, and what sets us apart from the ancients, is that not only do we think we can locate engulfing maelstroms in the form of black holes in binary star systems and the nuclei of galaxies, we also think we have a fair idea as to how they are formed, how large they are, and so forth. A combination of theory and observation has led to the growing suspicion among many astrophysicists that the nucleus of virtually every galaxy harbors a massive black hole.
To an observer, the nucleus of a galaxy can be defined as a small region of high luminosity in the center of a galaxy. Optical photographs of most galaxies show such a bright center. The smallest size that can be detected is limited by atmospheric scattering to about half an arc second. This corresponds to a size of about 5 light years at the distance of the Andromeda Galaxy, about 100 light years for a galaxy at a distance of 50 million light years, and about 2000 light years for a galaxy a billion light years away.
Radio astronomers, using very long baseline interferometry (radio telescopes in West Virginia and California, for example, are phased together so that the ability to detect structure is roughly equivalent to a radio telescope 3000 miles across) have found galactic nuclear features 1000 times smaller than those detected with optical telescopes. These features....
 ....strongly suggest that the nuclei of galaxies are the sites of repetitive explosive activity in which clouds of high energy particles are ejected, always along the same axis of symmetry.
Over 50 of these jets have been observed and studied by the radio telescopes in New Mexico known as the Very Large Array. When these jets are large enough, they show up at optical and X-ray wavelengths. The HEAO 2 has detected X-ray jets in two violently active galaxies and one quasar. In each case, the jets extend out from the nucleus and point toward much larger bubbles, or lobes of high energy particles. The indications are that these objects, and presumably many others, have been undergoing explosive activity for a million years or more. The alignment of the jets of different size indicate that the explosive ejection of high energy particles has been channeled in the same direction, to a precision of a few percent, throughout the million years or so of violent activity.
There is further evidence of explosive activity. Large flares in radio, optical, and X-radiation are common in galactic nuclei and quasars. A gamma ray flare from the nucleus of our galaxy has also been observed (see Chapter 19). The flares occur at intervals ranging from years down to several hours. These variations can be used to set a limit on the size of the flaring region. Unless every particle in the region is heating up or cooling....
X-ray image, obtained from the Einstein Observatory, shows
an X-ray jet extending from the center to the upper left, in line
with the radio lobes. The jet indicates that energy in the form of
high energy particles is continually pouring forth from the nucleus
along a preferred axis. (Smithsonian Institution Photo No.
 ....down simultaneously, a very unlikely situation, or unless we are seeing some sort of reflection phenomenon, the stimulus for the variation must be able to travel across the nucleus in less than the observed time scale for the variation. In other words, the size of the flaring region cannot be so large that the time it takes a signal to travel across the region is longer than the observed time scale for the flaring. Otherwise, the variation would have been smeared out.
Since no signal can travel faster than the speed of light, it follows that a flaring source must have a size less than a light year if the flare occurred in a year, or a size of a light hour if the flare occurred on a time scale of an hour. This important result is central to the theory of quasars and galactic nuclei, because it limits the size of the source much more severely than do direct observations of the size. The HEAO X-ray observations have been most valuable in this regard. In half a dozen cases intense flaring has been observed on a scale of less than a day. This limits the size of the region to a light day, or roughly the size of our solar system. In three cases, the observed flares occur on a time scale of a thousand seconds or less. That means that, in the most extreme cases, an energy output of 10 billion suns is pouring forth from a region whose diameter is about that of Earth's orbit around the Sun! Conventional stellar energy generation processes, that is, thermonuclear reactions, will not suffice. We are driven to consider extreme conditions such as can be found in the vicinity of massive black holes.
One of the most important findings in astronomy over the past two decades is that violent activity is a widespread phenomenon that occurs in virtually every galactic nucleus. The HEAO observations have provided the latest and perhaps the most convincing evidence of this. Well over 100 galaxies have been detected as X-ray sources. They exhibit a continuous range of power generation in galactic nuclei, from normal galaxies such as our own and Andromeda, to quasars. It all suggests that the only difference between normal galaxies, the so-called active galaxies, and quasars, is the power level of the central source. A quasar seems to be an ordinary galaxy with an extraordinarily active powerhouse in its nucleus. What is the source of the violent activity in galactic nuclei and quasars? The observations indicate that two conditions must be met. In the words of astrophysicist James Gunn, an active galaxy must have a "monster" lurking in the nucleus and a supply of "food" for the monster. The monster is presumed to be a massive black hole that is fed by the infalling gas.
There are two basic reasons for the popularity of the black hole model for active galactic nuclei. First, there is observational evidence for the existence of dark massive compact objects in the nuclei of some galaxies. Second, plausible models involving the accretion of matter by a massive black hole appear to be capable of explaining many of the observed phenomena.
Infrared and radio observation of the nucleus of our galaxy indicate that a dark, nonstellar concentration of matter is needed to explain the motions of gas in the nucleus. The mass required is on the order of 5 million solar masses. The gamma ray observations from HEAO 3 provide further evidence. The observation of radiation from the annihilation of matter and antimatter seems to require conditions that are likely to be found only in the vicinity of a massive black hole (see Chapter 20). Observations of the nucleus of the Andromeda Galaxy imply a mass concentration of 10 to 100 million solar masses there. Optical observations of the nucleus of M87, a giant elliptical galaxy in the constellation of Virgo, show that the average velocity of stars in the nucleus rises rapidly inside the core. This is just the opposite of what would be expected if all the mass was in the form of stars. A nonstellar mass concentration in the nucleus of some 5 billion solar masses is needed to explain their results. The gravitational radius of such a black hole would be about the size of our solar system. This fits well with the time variations in the X-ray flux from a number of active galaxies, which require a power source of about this size or smaller.
None of these observations proves that there are black holes in the nuclei of galaxies. Other peculiar concentrations of mass, such as a superdense swarm of neutron stars and white dwarfs, or a single massive superstar, might work. But there is nothing to guarantee that these entities would not collapse to form a massive black hole. Indeed, it seems probable...
....that they would. The inevitability of gravitational collapse as the final state is a strong point in favor of black hole models-that and the cheap source of energy they provide.
At first sight, a black hole, the ultimate sink, would not appear to be a promising source of energy. The key to this paradox is the resistance of matter to compression. As the particles fall toward the black hole, they gain energy and crowd together. As a result, they radiate strongly in the ultraviolet, X-ray, or gamma ray band, depending on the size of the black hole and the conditions of the accretion. Under quite general conditions, the heating produced by accretion into a black hole can be 100 times as effective at producing radiation as the thermonuclear fusion reactions responsible for the energy generation in the Sun and most other stars.
Normally the gas would not be expected to fall straight into the black hole. Rather, it would spiral inward as part of a gaseous disk. If the radiation from the disk becomes too intense, it will exert a back pressure that prevents much of the matter from falling into the black hole. This effect together with the buildup of magnetic stresses due to the twisting of the magnetic field in the gas, could lead to huge flares and the ejection of clouds of gas at very high velocities. The direction of the ejection would be along the rotational axis of the disk, that is, perpendicular to the plane of the disk. This provides a natural explanation for the observed ejection of matter from active galactic nuclei along a preferred direction.
HEAO observations of the spectrum and variability of the radiation from active galactic nuclei, in combination with observations at other wavelengths, can provide important checks on the theory sketched above  and allow us to study the physical state of matter near a massive black hole. Already the observations tell us that if a hot, churning, bubbling cloud of gas around a black hole is to explain the conditions in the center of our galaxy, then gas must be falling into the black hole at a rate of a few Earth masses per year. In the case of M87, the number would be more like several hundred Earth masses per year. In many active galactic nuclei, several thousand Earth masses per year would have to fall toward the black hole. For some of the most luminous quasars, roughly a million Earth masses, that is, the mass of a few solar type stars, would have to be accreted every year.
How are large black holes formed in the centers of galaxies? Nothing close to a consensus of opinion exists here, but one good possibility is that the formation of a central black hole naturally and inevitably results as globular star clusters spiral toward the center of the galaxy. Globular clusters are spherical clusters of a hundred thousand to a million stars. They are believed to be some of the first stellar units created during the initial formation of galaxies. These clusters, like individual stars, orbit the galactic nucleus in ponderous orbits that take millions of years. When they pass through the dense central regions of their parent galaxy, they suffer many grazing collisions with individual "field" stars. The cumulative effect of these collisions exerts a drag on the clusters, just as the atmosphere exerts a drag on satellites. As a result, the clusters spiral inward to the center of the galaxy until they are torn asunder by tidal forces.
In the course of billions of years, millions of stars may be concentrated into a region only a few light years across. These stars collide with one another, sometimes at high speeds, in which case they are partially or completely torn apart. Some of the collisions are gentle bumps, and the stars stick together, or coalesce. This new star presents a larger target, and more than likely it will eventually coalesce with other stars. The bigger the star gets, the more likely it is to be hit again and the faster it grows until it reaches instability, collapses on itself, and forms a black hole.
Once formed, a central "seed" black hole grows mainly through the accretion of gas accumulated in the nucleus. This gas could come from stars disrupted by collisions with other stars, from supernova explosions, from stars torn apart by the gravitational field of the black hole, or from gas falling into the galaxy from outside. In this connection it is intriguing that many active galaxies appear to have suffered a collision or near collision with another galaxy. Such encounters could result in the transfer of large amounts of gas from one galaxy to another, providing a plentiful supply of gas for the black hole.
When most of the stars and gas in the core of a galaxy have been swallowed up or ejected from the galaxy by the black hole, the nucleus of the galaxy settles down to a relatively quiet existence. This is probably the state of the nucleus of our galaxy. Every hundred million years or so it may flare up to a brightness 100 times its present level, when a globular cluster or especially large gas cloud spirals into the nucleus. These flaring episodes  may correspond to what we are presently observing in some relatively nearby active galaxies.
This scenario explains the near universality of activity in galactic nuclei. It is all a variation on the same theme. Taken literally, it suggests that every normal galaxy was once a quasar, and every quasar will eventually become a normal galaxy. This is of course an over-simplification, but it may represent a rough approximation of the truth. This requires that the quasar phase be short lived on a cosmologic time scale, perhaps only a few million years. Exceptionally dense nuclei might produce exceptionally bright sources and exceptionally massive black holes, but normally, we might expect the quasar to shut off when the mass of the central black hole grew to a mass of several million suns. Normal galaxies are about 10 billion years old; if the quasar phase lasts only a few million years, or a few ten thousandths of this time, then we would expect to see only a few ten thousandths as many quasars as we do normal galaxies. This is about what is observed.
It would be foolhardy at this point to say that the enigma of quasars and active galactic nuclei has been solved. However, the progress in the last few years has been encouraging. The Space Telescope and advanced orbiting X-ray facilities planned for the 1980s should allow us to rule out alternative models, to find more dark, massive central cores in other galaxies, and hopefully to settle once and for all the age-old question as to the existence of an abyss in spacetime from which there is no return.
Until then, quoting Robert Frost,