he Pythagoreans were an ancient brotherhood founded by the Greek philosopher Pythagoras, who is best remembered for his famous theorem concerning right triangles: The square of the hypotenuse of a right triangle is equal to the sum of the squares of the other adjacent sides. In the fifth and sixth centuries B.C. Pythagoreans were devoted to such examples of mathematical beauty and extended this fervor into their contemplation of the cosmos. They intently believed that celestial spheres ascended from Earth to heaven like the rungs or a ladder, which majestically carried the planets around and gave forth harmonic tones that created a wondrous music of the spheres. One version of this system had each planet intone a note higher than the one before it, starting with the Moon and working outward to the fixed stars. The seventeenth-century astronomer Johannes Kepler was a devotee of this idea and even wrote down some celestial tunes, audible to God alone, that he associated with various

planets in their orbital journeys. Little did he realize that his musical vision would find substance in an entirely different astronomical arena.

For most of astronomy's history, the universe was studied with one means and one means only: collection of its electromagnetic radiation. For the astronomers who came before Kepler, it was with their eyes alone. Later, lenses and mirrors focused and magnified the visible light. By the middle of the twentieth century, astronomy expanded its arsenal of instruments to collect photons from other regions of the electromagnetic spectrum —radio, infrared, ultraviolet, x rays, and gamma rays. At the same time, particles such as cosmic rays and neutrinos began to be gathered from space. With each new technique, notes Penn State theorist Sam Finn, astronomers found something they didn't expect. Radio astronomers found pulsars, quasars, and massive molecular clouds dotting the celestial landscape. X-ray astronomers were surprised by the power of x-ray binaries, which strongly hinted at the existence of black holes. The lesson to be learned, says Finn wryly, is that “astronomers have no imagination.”

Given that understanding, who knows what gravity waves might uncover, for they will be offering a far more radical method of gathering information about the universe. That fact is quite apparent when comparing electromagnetic radiation to gravity waves. Electromagnetic waves are emitted by individual atoms and elementary particles. Gravity waves, on the other hand, are generated by the bulk motions of matter. The frequency of the wave is directly related to the frequency of the massive movement generating the gravity wave. At the moment, for example, the two neutron stars in the Hulse-Taylor binary are emitting gravity waves at a frequency of 10^-4 hertz as they circle one another every eight hours. As they get closer together over the millennia, though, and their orbital velocities speed up, the gravity wave frequency will increase. LIGO will detect the waves from such systems when they match (coincidentally) the audio-frequency range (20 to 20,000 hertz).

The strongest gravity waves of all are created when matter approaches the speed of light, which would occur in supernovas or when black holes collide. This offers the opportunity to explore the most violent phenomena in the cosmos. It would allow astronomers, for

instance, to peer into the very heart of an exploding star. This is because gravity waves pass through matter as if it were not there, unlike most electromagnetic radiation, which can be absorbed or scattered by matter as the radiation proceeds on its journey. Electromagnetic waves travel through space-time, while gravity waves are actual jiggles of space-time itself. With gravity wave detectors, we will be learning about the cosmos from its space-time vibrations. It will be like adding sound to the silent pictures that have been constructed up until this time. Gravity wave astronomers will be listening to the modern-day version of Pythagoras's music of the spheres.

As Weber and other experimentalists made their first forays into gravity wave detection, theorists were active as well. As the number of detectors grew, specialists in general relativity launched their own effort to figure out what “tunes” the detectors might be receiving. First they had to learn how to manipulate the equations of relativity in such a way that they could even address the problem; then they went on to determine the types of waves that might be emitted by various astrophysical events. At the vanguard of this effort was Caltech's Kip Thorne.

Rai Weiss, admittedly no friend of theorists, makes an exception with Thorne. Introducing him one day at an MIT seminar, he noted that “Thorne is one of the most approachable theorists, a physicist who championed LIGO rather than sitting back and doing idle calculations. ” Thorne even contributed to the instrumentation. It was his idea to serrate the edges of the baffles inside LIGO's beam tubes, to make sure that any stray laser light was properly dispersed. As a theorist, Thorne wears many hats. He has looked into the origins of classical space and time and explored the physics of a black hole. In the public's eye he is most notorious for his work on “wormholes,” hypothetical cosmic tunnels through hyperspace that provide shortcuts to both other reaches of the universe and other times. They sound like science fiction but are founded in genuine solutions to Einstein's field equations. Thorne started studying these weird entities in the mid-1980s because of a friend's request. Astronomer Carl Sagan, then working on his novel Contact, had asked Thorne whether there was a scientifically legitimate way for his characters to dart about the cosmos with

ease. Thorne and his students came up with a solution that utilized wormholes. Their result, in a paper entitled “Wormholes, Time Machines and the Weak Energy Condition,” was published in the prestigious Physical Review Letters in 1988. But Thorne's most massive theoretical endeavor in recent years, in collaboration with an armada of graduate students and postdocs, has been carrying out the theoretical needs for LIGO. His Caltech group has been modeling potential cosmic sources of gravitational radiation and estimating the characteristics of the various waveforms.

Born in 1940, Thorne grew up in Logan, Utah, then a small college town of 16,000. Although his parents were Mormons (their ancestors had moved west with Brigham Young), they didn't fit the group's typical conservative profile. Far more liberal, his father was an eminent soil chemist at Utah State University. His mother, who held a Ph.D. in economics, initiated the women's studies program at Utah State and participated in anti-Vietnam war marches. The oldest of five children, Thorne caught the science bug early: “When I was eight, my mother took me to a lecture on the solar system given by a geology professor at the university. I was immediately fascinated. That was my first introduction to astronomy. Before that I wanted to be a snowplow driver. For a little boy growing up in a town in the mountains, where you have snow banks that are six or eight feet high, snowplow drivers are the most powerful people in the world.” By the time he was a teenager, a book by physicist George Gamow, One, Two, Three, Infinity, hooked Thorne on relativity. Geometry became a passion. He spent many summer hours working on problems in four dimensions. In high school, Thorne was a “cocky kid,” as he puts it. Starting in the ninth grade, he sat in on college classes, including geology, world history, and mathematics. If bored in a high school class, he would just get up and leave. “They presumed I was just going off to the university, ” says Thorne.

Thorne's rebellion continued at Caltech, which he entered as an undergraduate in 1958. For three summers he worked at the Thiokol Chemical Corporation helping design rocket engines for the Minuteman missile program. When asked in his fourth summer to sign a loyalty oath, a legacy of the McCarthy era, he refused. He lost a presti-

gious National Science Foundation graduate school fellowship for the same reason (although he later received one when the loyalty oath requirement was finally dropped).

His choice of graduate school in 1962 was almost predetermined. Browsing through the physics journals, Thorne immediately recognized that the most interesting work in general relativity was being done at Princeton University, home of John Wheeler. In their first meeting, Thorne intently listened as Wheeler spoke for two hours outlining the outstanding problems of the time. Thorne chose to immerse himself in black hole physics, although that name wouldn't be used for five more years. Joe Weber was also a presence on campus, as he regularly shuttled between Maryland and Princeton to talk with Wheeler, Dicke, and Dyson about the construction of his first bar detector. Thorne finished his Ph.D. in a speedy three years after writing a dissertation on hypothetical relativistic objects that were long, thin, and cylindrical. To his and everyone's surprise, he found that some of these unusual objects would be stable. Today, it's more than an academic exercise. Theorists wonder if the early universe, in the first fraction of a second of its existence, generated similar bodies, now called cosmic strings.

Thorne returned to Caltech as a postdoc in 1965, just as his skills in general relativity were most needed. Two years earlier quasars had been discovered, and some suspected black holes were involved. But others at Caltech, particularly William Fowler and Fred Hoyle, wondered whether supermassive stars were the source of a quasar' s power. Fowler funneled students over to Thorne to help him look into such questions. When Fowler gave up his NSF grant in relativistic astrophysics, because of some new duties he took on, Thorne essentially inherited the stipend, which has allowed him to supervise and support nearly 40 Ph.D. students and another 36 postdocs over the past few decades. As a result, Caltech eventually supplanted Princeton as the new mecca for pursuing general relativity. Thorne became well-known on campus for his bohemian flair—shoulder length hair, full beard, colorful shirts, and sandals.

Always Thorne made sure that his work touched bases with the real world. “There was a great richness of things to be done in bring-

ing relativity in contact with the rest of physics,” says Thorne. It was the legacy of Thorne's studies at Princeton, where he not only studied under Wheeler but also regularly dropped by Robert Dicke's group to keep in touch with its experimental work. When Thorne was driving across the country with his family to make his move from Princeton back to Caltech, he recalls dropping by the University of Chicago to consult with the noted astrophysicist Subrahmanyan Chandrasekhar (a 1983 Nobel laureate for whom the Chandra X-ray Observatory is named). They talked about neutron stars and gravitationally collapsed objects, which at the time were still conjectures. “They seemed so far from any observation,” says Thorne. “But Chandra expressed a confidence that neutron stars and what came to be known as black holes would, in some moderate number of years, be found. That had a big impact on me. I only wanted to work in areas that had an observational backup.”

Pulsars were soon discovered in 1967, giving Thorne added confidence that his work on exotic objects would no longer be purely theoretical. But while neutron stars were gaining credence, black holes were still suspect. Many astronomers still clung to the view that nature would somehow find a way to prevent stellar cores—cores heavier than neutron stars—from collapsing into singularities (the upper limit on neutron stars is believed to be anywhere from 1½ to 3 solar masses). Stars lose mass as they age, and some figured that they would always lose enough to drive them under the black hole limit. Thorne was working in an intellectual climate of great skepticism. “Not unlike the skepticism we've seen in recent years about gravitational wave detection, whether waves are emitted sufficiently strong enough so that we will see them,” he says. “The intellectual ambiance of that era was very much one that relativity was a beautiful subject but that it didn't have much to do with the real world. It was just a mathematical subject to be pursued for its own intellectual interest. This attitude pervaded physics from the mid-1930s into the 1970s. I found myself following Wheeler's footsteps as an advocate for these things being truly relevant to the astrophysical universe.” Thorne's famous textbook Gravitation, published in 1973 with Misner and Wheeler, was to a large extent designed to promote the contact between relativity and the rest of phys-

ics. “It was in some sense a propaganda piece, as was my public lecturing. I was trying to convince the community that this was a field that was relevant to the rest of physics,” says Thorne. What ultimately turned things around were the observations arriving fast and furiously from new arenas of astronomy, especially x-ray astronomy. The “final clincher,” according to Thorne, was the close examination of an exceptionally bright x-ray source located in the direction of the Cygnus constellation. Cygnus X-1's powerful x rays come from a double star system consisting of a giant blue star and a dark invisible companion whose measured mass of some 10 to 20 solar masses strongly suggests it is a black hole. The x rays are generated as matter, drawn away from the supergiant star, spirals inward toward the black hole.

Starting three decades ago Thorne carved out a special niche for himself in the general relativity community as he and his students began to examine the stability of neutron stars and black holes. “All of this was done in the context that Weber was working on his experiment, and we were trying to understand the potential sources for his bar,” says Thorne. “His experiments were very much on my mind.” At the time there were suspicions that if a black hole got spun up by accretion to a high speed—by stealing mass from a nearby companion—it would start vibrating and tear itself apart. That was one way to avoid having this ugly awkward creature in the universe. But three of Thorne's students, Richard Price, Bill Press, and Saul Teukolsky looked at the pulsations of a black hole and proved that if you perturb a black hole the disturbance will quickly dampen as the energy radiates away as a set of gravity waves. In the end a black hole remains and very much intact. Thorne showed that the same would be true for a neutron star. Thorne had his students visit the math department to bring back new techniques for analyzing the radiation of gravity waves from stars and other systems. In test after test they and their colleagues at other universities came to the same conclusion: The formation of a black hole was inevitable if enough mass was around. Vibrations alone would not stop it. Even if one black hole were thrown at another black hole—one of the most cataclysmic cosmic events imaginable—what results is simply a bigger black hole that is perfectly stable.

In this way Thorne came to specialize in gravitational wave phys-

ics. “I have an aversion to working in areas where other people are working because I prefer to do something that is unique. I don't like to be in the position of worrying that if I don't do something today a competitor will solve the problem tomorrow,” he explains. Thorne was making a calculated bet. Most people in black hole research at the time didn't have high expectations that the technology would be good enough to see gravitational radiation anytime in the near future and so didn't care to examine its physics deeply. “But I thought there was a shot at it,” he says today, still comfortably attired in a loose cotton shirt but his once-long hair now shorn and graying.

It's a tricky business to determine just how much gravitational radiation might be bathing the Earth daily. It strongly depends on the theoretical models for determining how much gravitational energy might escape an event, such as a supernova or a black hole collision. The models are complex and frequently change as differing solutions go in and out of fashion. Thorne has long maintained a chart of the possibilities, with the uppermost line marking off his “cherished belief” boundary, the strongest waves possible without violating any conventional beliefs about the nature of gravity. The potential sources that his Caltech team and other groups around the world have cataloged so far are many and varied.

LIGO will be after big game. The sources will weigh at least the mass of our Sun but likely heavier. Their movements will be a sizable fraction of the velocity of light, anywhere from a tenth to nearly full speed. The most exciting find by far would be the collision of two black holes. Such a sighting would finally christen black holes as bona fide denizens of the universe. Up to this point, evidence for their existence has been circumstantial. X-ray telescopes regularly pick up signals from remote orbiting bodies, which astronomers interpret as the high-energy radiation released right before a black hole permanently swallows the matter pulled off a companion star. Yet the black hole itself remains unseen. But if two black holes should be orbiting each other, they would at a certain moment reveal themselves. They would

eventually spiral into one another, releasing an unmistakable set of gravitational waves that preserves a record of the fateful collision. It would be a cosmic signature unique to black holes.

Picture two black holes slowly circling one another, like a pair of sumo wrestlers warily checking each other out in the ring. Tens of millions of years earlier these two black holes were simply stars, until they exhausted all their fusible fuel and collapsed to the most compact state imaginable. More than mere indentations in space-time, black holes are fathomless pits. No bits of light or matter can climb out of these deep gravitational abysses. That's why ordinary telescopes can't see them, and theorists can do no more than imagine them. Only a gravity wave telescope has a chance of detecting them.

The sighting would occur at one decisive moment, after the two black holes have been slowly orbiting one another, perhaps over millions of years. During that time the pair would have been emitting a steady stream of very weak gravity waves, a wake that continually spreads outward along the canvas of space-time, like the spiraling pattern of a spinning pinwheel, as the black holes circle about. Gradually losing energy in this fashion, the two black holes relentlessly draw together as the years go by. And the closer they get, the faster they orbit one another.

In the final minute of this fateful dance, the gravity waves being emitted become strong enough to be detectable. Instruments on Earth would register a sort of whine, a series of waves that rapidly rise in pitch, like the sound of an ambulance siren that is swiftly approaching. These black holes should not be thought of as masses, points out Thorne. Rather, they should be envisioned as whirling tornadoes of space-time, which are both dragging space-time around them as they orbit one another. “It's like two tornadoes encased in a third tornado, all coming together,” he says. As the twirling black holes are about to meet, spiraling inward faster and faster at speeds close to that of light, the whine turns into a “chirp,” a birdlike trill that races up the scales in a matter of seconds. A cymbal-like crash, a mere millisecond in length, heralds the final collision and merger. The two black holes become one. A “ringdown,” akin to the diminishing tone of a struck gong, follows as the new entity, a pit in space-time that swirls around like the

The type of waveform that gravity wave astronomers expect to see when two black holes collide. As the two holes spiral in, the waves increase in frequency. In the end, there is a final “ringdown.” No one yet knows what the signal will look like at the moment of collision.

fearsome tornado in The Wizard of Oz, wobbles a bit and then settles down. The masses of the two black holes can be determined from the total duration of their phenomenal coupling: the heavier the holes, the greater their attraction to one another, and the faster the merger.

For many years theorists generally assumed that gravity wave telescopes would register a few of these black hole collisions a year, as soon as they were sensitive enough to detect the signals arriving from as far away as the dense Virgo cluster of galaxies, some 50 million light-years distant. More recent calculations by Simon Portegeis Zwart and Stephen McMillan tentatively suggest that there may be more black hole binaries in the Milky Way and other galaxies than earlier suspected, perhaps a thousand times more. Globular clusters, tightly packed groupings where stars are separated by light-minutes rather than light-years, could act as the incubators. More than a hundred of these dense celestial balls, many containing hundreds of thousands of stars, are scattered above and below the plane of our Milky Way (as in other galaxies). Any single black hole produced in such a cluster would eventually settle toward the cluster's core, where it would likely meet

other black holes. Gravitationally attracted to each other, the black holes would pair off and form a binary. Circulating within the cluster, some of these pairs might gain enough speed to escape the cluster altogether. Then undisturbed outside the cluster, the black holes would gradually spiral in and merge. It is the one incontrovertible way that physicists can finally clinch the existence of a black hole, nature's strangest star. The black hole would give itself away by the melody of its gravity wave “song,” the distinctive ripples of space-time curvature transmitted throughout the heavens. The National Research Council, in its 1999 report Gravitational Physics, stated that such a finding “would be the capstone of one of the most remarkable discoveries in the history of science.” Listening to these tunes, astronomers might even discern supermassive black holes, each containing the mass of a million or more suns, being constructed in the centers of far-off galaxies, as the holes gobble up and swallow their celestial victims. Gravity wave astronomy will ultimately make black holes seem ordinary, says Thorne.

While the initial whine and chirp phases of the inspiraling are well modeled from numerical relativity, the final crash of the black hole/ black hole collision is bringing theorists to the cutting edge of general relativity's challenges. That particular phase of the event is a problem not yet solved in its entirety. Steps toward a solution were carried out in recent years during the “Binary Black Hole Grand Challenge,” one of a series of problems solvable only by vast computations. Funded by NSF, the other scientific challenges included the formation of galaxies, the synthesis of images from radio astronomy data, the behavior of quark-gluon plasmas in elementary particle physics, and quantum mechanical simulations of exotic materials, such as high-temperature superconductors. In the black hole challenge, a team of more than 30 scientists at eight universities was organized to both study the nature of black holes and predict the gravitational waves produced when black holes collide. Such predictions will be incredibly useful for validating and interpreting the waveforms detected by LIGO.

“Einstein's equations describe gravity via elegant but complicated nonlinear partial differential equations,” says Richard Matzner, director of the Center for Relativity at the University of Texas in Austin and

leader of the Grand Challenge team. Such equations cannot be solved by pencil and paper alone but rather require brute computation on the world's fastest and most powerful supercomputers. When Einstein 's elegant equations are recast in this numerical mode, any one part can involve thousands of terms, which requires special software to handle. Because the solution is so complicated, the team tackled it in steps. First, team members handled the “simple” case of two nonspinning black holes approaching one another head-on, meeting, and then merging. At each stage—from far encounter to near touching—the gravitational landscape changed, an evolution they captured step by step. But as far as nature goes, that's an unrealistic simulation. Like all other stellar objects, black holes spin. And when two black holes are involved, they will also be orbiting one another, which adds more complexity to the numerical simulations. The game in the challenge was to develop the tools needed to first model the process and then optimize the calculations. This optimization depends on both the way the calculations are set up and the power of the computer. In the early 1990s scientists would have needed 100,000 hours of computer processing time (more than 11 years!) to perform an accurate three-dimensional simulation of two black holes spiraling into one another, which made the endeavor wishful thinking. Today, using algorithms optimized for parallel processing machines, that time has been cut to 1,000 hours, a much more doable prospect. But computer memory still needs to be increased a hundredfold before the full solution of a blackhole/blackhole collision—finding out what happens during the last few orbits and the ultimate coalescence —can be obtained. Sam Finn calls it “the final uncharted frontier” in black hole physics.

One phenomenon that theorists believe is almost guaranteed will be the resounding crash that occurs when two neutron stars, paired together in a binary system, spiral into each other as their orbital dance decays. In 1963, four years before the first neutron star was even discovered, physicist Freeman Dyson estimated the gravitational radiation expected from such a neutron star pair. “It would seem worth-

while to maintain a watch for events of this kind, using Weber's equipment or some suitable modification of it,” he wrote at the time. It was a prescient thought. Gravity wave astronomers suspect such events may turn out to be the bread and butter of their trade. As noted earlier, the two compact balls of matter in the famous Hulse-Taylor binary now emit gravitational waves around 10^-4 hertz. Only in its last 15 minutes of life, when the two neutron stars have finally drawn quite close to merge, will the waves sweep from 10 to 1,000 hertz, setting off gravity wave detectors on Earth. But the Hulse-Taylor binary won't be colliding for several hundred million years. “So we must reach outside the galaxy to give our graduate students a thesis,” jokes Thorne. As soon as gravity wave detectors are sensitive enough to see beyond the Milky Way, they will likely pick up binary neutron star bursts in other galaxies. The length of the LIGO arms was chosen, in fact, ultimately to see such events out to a few billion light-years and so obtain a good population of sources.

How often might this happen? That depends on both the statistics and the latest theoretical models. Theorists take the number of neutron star binaries known to exist in our own Milky Way galaxy and then extend that outward to encompass the volume of space that will be observable to LIGO, eventually a vast region spanning hundreds of millions of light-years. The event rate for LIGO I is rather low, about 10 per century at best. But with upgrades a LIGO II might see one a day.

These neutron star binaries will broadcast their own distinctive sets of whines and chirps. More lightweight than black holes, a pair of neutron stars will take longer to merge, so the final recordable signal will last minutes instead of seconds. Gravity wave telescopes would register a sinusoidal wave that sweeps to higher and higher frequencies as the two mountain-sized balls spiral into one another. Five to ten minutes before their lethal meeting, the two neutron stars are about 500 miles apart and orbiting one another about ten times each second, at nearly a tenth the speed of light. In the final moments they are severely stretched by tidal forces and revolving around each other as much as 1,000 times a second, dragging space-time around with them. These waveforms—the “chirps”—will contain within them a wealth of

information for those who know how to look, such characteristics as the density and composition of the compact star's nuclear matter. As soon as they touch, the two stars are shredded to pieces, possibly releasing a burst of gamma rays.

What happens afterward? No one knows for sure. An ongoing NASA-sponsored study, the Binary Neutron Star Grand Challenge, may help decide. The remnants might coalesce into a new, more massive neutron star. Or if heavy enough, they might condense to utter invisibility, forging a black hole. Only a gravity wave telescope will be able to reveal the final outcome. But once these signals are detected, they would be a boon to cosmologists, who have long been arguing over the universe 's size. Current measurements of distance rely on such yardsticks as the luminosity of stars and the apparent size of galaxies, but astronomers continue to quibble over the interpretation of those standard candles. Until recently, estimates of distance varied by factors of two, which was as vexing to astronomers as if geographers could only estimate the distance between New York and Los Angeles as somewhere between 2,000 and 4,000 miles. But by knowing the amount of gravitational energy emitted by inspiraling neutron star pairs about to collide and comparing these estimates with the strength of the waves when they arrive on Earth, astronomers could calculate how far the waves had to travel to reach our planet. This, in turn, could provide a measuring tape directly out to the galaxies, without the worry of the intervening steps that can plague other methods.

A collision between two neutron stars would also offer insights into the nature of nuclear matter. A neutron star is, in some sense, one big atomic nucleus. Only in this case the nucleus contains 10⁵⁷ neutrons. “Somewhat larger than physicists are normally accustomed to,” notes Thorne. So to study this particular form of nuclear material, physicists can't rely on particle accelerators to push the matter to near the speed of light. Fortunately, though, neutron stars in a binary system do it themselves due to their mutual gravitational attraction. Here would be a means of finding out whether some of the mysterious gamma-ray bursts in the sky are the result of two neutron stars colliding, as some suspect. Their distribution and intensity are in the right

range. Roughly once a day a burst of gamma rays appears from some random direction on the celestial sky. On average, the flash lasts some 10 seconds. Such bursts were first noticed by U.S. Air Force Vela satellites, launched in the 1960s to monitor nuclear explosions on Earth. While evidence suggests that most bursts originate from far outside our galaxy, their exact origin is not fully known. One of the brightest bursts to date, uncovered by the Italian/Dutch satellite BeppoSAX, is an example. This particular burst was traced to a faint galaxy situated some 10 billion to 12 billion light-years away, nearly to the edge of the visible universe. Over a matter of seconds, it appeared to release several hundred times more energy than a supernova. For that one moment it was as luminous as the rest of the universe. As soon as the burst was spotted in December 1997, an armada of ground-based and space telescopes—optical, radio, x-ray, and infrared—aimed their instruments in that direction and recorded the afterglow of this mighty celestial fireball. It's difficult to explain the tremendous energies; they might involve neutron star or black hole collisions or perhaps the collapse of a massive star to form a black hole. Maybe both. A LIGO detection will help sort it all out.

There could be another type of signal in the gravity wave sky as well, although it would be far less frequent. A solitary tsunami of a wave might hit our shores every once in a while, generated at the very moment a star explodes in our galaxy as a brilliant supernova, its core crumpling up to form a dense neutron star. This collapse triggers a shock wave that blows off the star's outer mantle of gases, which we see as a supernova. Astronomers routinely spot these explosions in other galaxies with the aid of telescopes. For laypersons using only their eyes, it requires a little more patience. The last supernova visible to the naked eye occurred in 1987. The explosion was sighted in the southern sky within the Large Magellanic Cloud, a satellite companion to the Milky Way. The last visible supernova in our galactic neighborhood before that was spotted by Kepler in 1604.

Sighting a supernova with gravity wave detectors is not assured, though. A lot will depend on the explosion's dynamics. If the collapse of the remnant core is perfectly smooth and symmetrical, gravity wave astronomers will not hear even a whimper; gravity waves emitted symmetrically tend to cancel each other out, much the way out-of-phase light waves do. At the same time that one part of the wave is causing space to stretch, another part is causing it to contract; the net result is no change at all. Gravity waves would be emitted only if the collapse is a messy affair, with the newborn neutron star squishing down like a pancake and then stretching out before settling down. As a result, a gravity signal with waves extending hundreds of miles from peak to peak will be sent out. If the core is spinning madly at the end of its life, it could even flatten and be turned into a barlike configuration, spinning end over end like a football. In that case the collapsing core would send out very strong waves, perhaps being seen far beyond the Virgo cluster. An advanced LIGO detector, with more sensitive equipment, might see several a year, which would make it one of LIGO's more dependable sources.

There is some evidence that supernova explosions can be imbalanced. Astronomers have seen individual pulsars speeding through the galaxy at velocities greater than 100 miles per second. It is suspected that the pulsar got shot out by an asymmetric explosion, an extra kick on one side more than the other. Supernovas occur within our galaxy only two or three times every century on average, but their signal strength will be spectacular. It is estimated that the supernova seen in 1987 had a signal 100 times stronger than LIGO is capable of detecting when it first turns on. The unstable neutron star should even “boil” vigorously for the first second of its life. During this boiling, high-temperature nuclear matter (some trillion degrees) rises to the surface, where it cools and is swept back downward. Such boiling could send off a series of 100-hertz waves, strong enough to be seen by a LIGO-type detector out to 100,000 light-years.

All the while, playing in the background of this gravity wave symphony, could be ongoing rhythms, a steady beat. When a neutron star

forms, for instance, it might briefly vibrate and develop a bump on its surface, an inch-high “mountain” that grows and freezes into place. And as the neutron star feverishly whirls around, this deformation, jutting out like a finger, would send out a periodic gravity wave as it continually “scrapes” the space around it. For a neutron star rotating once every thousandth of a second, its equator ends up spinning at about 20 percent the speed of light. Lumpy neutron stars could serve as gravity wave lighthouses scattered over the heavens, each blinking away until its lump smoothes out. The only interruption might be an occasional gravity wave burp, released whenever the neutron star undergoes a “starquake.” This could happen when the outer crust of the neutron star slips at times over its superfluid core. The signal would be extremely weak, too weak to see right away. In this case the interferometer would have to gather the data for weeks and add it up in order to have the signal emerge from the background noise.

More recently, theorists have been excited by a calculation that suggests that a neutron star's dense nuclear matter might actually “slosh around” soon after the star forms, fed by the star's rapid spin. Such oscillations occur in our own oceans, creating circulation patterns. In the case of a neutron star, gravity waves are generated. More interesting, the gravity waves increase the sloshing, which increases the production of gravity waves. What starts out as a quantum jiggle can amplify quickly. How large does the sloshing get? Theorists don't know but suspect other forces, such as friction and magnetic fields, step in at some point to cut it off. Until that happens, though, the newborn neutron star might release a unique gravity wave cry for up to a year as it cools and settles down.

And beneath the chirps, pops, and beats emanating from the gravity wave sky, there could be an underlying murmur—constant, unvarying, and as delicate as a whisper. This buzz would be the faint reverberation of our universe's creation, its remnant thunder echoing down the passages of time, similar to the residual microwave heat already detected from the Big Bang. But those microwaves started their journey half a million years after the Big Bang—the time when atoms

first formed and light could at last travel through the universe, unimpeded by a jumble of particles. If we attempt to look farther back in time, we perceive only a fog.

Primordial gravity waves, on the other hand, would cut right through that fog. They would be fossils from the very instant of creation, tiny jiggles in space pumped up by an explosive burst of expansion that took place a scant 10^-43 second into the universe's birth. No other signal survives from that era. These relic waves would bring us the closest ever to our origins, perhaps verifying that the universe emerged as a sort of quantum fluctuation out of nothingness. At the same time they might tell us how fast the universe expanded over the eons and whether there is enough matter in the heavens to bring this cosmic marathon to a halt in the far, far future.

It's possible that scientists have already registered the imprint of these primordial waves. In 1993 George Smoot of the Lawrence Berkeley Laboratory and Paul Steinhardt of the University of Pennsylvania suggested that Big Bang gravity waves may have “ruffled” the cosmic microwave background, which had been mapped so spectacularly by the COBE satellite. COBE discerned tiny fluctuations in the smooth sea of microwaves bathing the universe. Theory suggests that these fluctuations were quantum disturbances that grew and expanded in time. But some of those ripples, say Smoot and Steinhardt, might be attributed to primordial gravity waves. To separate the quantum from the gravitational requires comparing the COBE data to other measurements of the microwave background. COBE measured fluctuations on a relatively large angular scale. Other instruments, such as balloonborne detectors and ground-based instruments set up at the South Pole, can measure smaller scales. On those finer scales any contribution from gravity waves should fade away. The background would look a bit less bumpy. Scientists are avidly checking out this possibility.

Perhaps more exciting is the prospect of encountering the unanticipated. The exact form of the gravity wave signal, for one, might disagree with the predictions of general relativity. This may indicate that Einstein's equations have to be amended when dealing with sources that involve a horrifically strong gravitational field. Gravity

wave astronomy's findings could usher in new physics of gravity, akin to the way Einstein supplanted Newton. It might even provide clues as to how theorists could forge a “theory of everything” that unites general relativity with quantum mechanics. If not, there is still the chance that strange new celestial creatures could greet us when gravity wave astronomers make their first discoveries. Not until astronomers scanned the heavens with radio telescopes did they discover pulsars and quasars: neutron stars had been contemplated, but not as pulsing radio beacons; quasars were never even fantasized. What else might be skulking about in the darkness of space as yet unseen? Pulsars and quasars may turn out to be commonplace in comparison to the exotic astrophysical events that gravity wave astronomy reveals. Some theorists already wonder whether there might be relics from the early universe, highly energetic “defects” that were generated as the cosmos cooled down over its first second of existence. They include pointlike monopoles, one-dimensional cosmic strings, and what are called domain walls.

Cosmic strings are one of the more interesting defects hypothesized. One might think of them as extremely thin tubes of space-time, skinnier than an atomic particle, in which the energetic conditions of the primeval fireball still prevail. Any strings that survived to this day would be either exceptionally long (spanning the entire width of the universe) or bent back on themselves, creating closed loops that continually lose mass-energy by vibrating at velocities approaching the speed of light. If these potent strings truly exist, astronomers may not be too desirous to observe them close-up. While such a string, anorectically thin, could actually whiz through your body without bumping into one atom, its peculiar gravitational field would wreak havoc nonetheless: if this string sliced through you, your head and feet would proceed to rush toward one another at 10,000 miles per hour. Because of the tremendous tension in a string, it would wiggle around like a rubber band, producing lots of gravity waves. Such gravitational radiation emanating throughout the universe could very well affect the timing of radio pulsars (astronomers are checking). Massive cosmic strings would also be excellent candidates for gravitational lensing.

Meanwhile, x-ray astronomy serves as reconnaissance. Celestial

x rays suggest that there's a rich landscape yet to be mined by gravity wave detectors. Take the case of the x rays being emitted by flows of gas close to the center of the active galaxy MCG 6 –30 –15. This gas is traveling at near the speed of light. At the moment the only explanation for such a high velocity is that the gas is caught in the whirlpool swirling around a massive black hole. But only gravity wave telescopes will be able to tell for sure. They will be able to cut right through the gaseous fog and catch objects falling into the hole directly.

Mounted along the wall of a lengthy hallway directly across from Kip Thorne's office at Caltech is a row of documents, 10 in all, each set in a black frame. Each letter records a bet made by Thorne with some prominent astronomer or physicist, including a few with Stephen Hawking. The challenges are varied, involving either the nature of black holes, the existence of “naked singularities” (a black hole stripped of its event horizon), or some predicted property of the universe at large. One, handwritten on Caltech stationery, concerns the detection of gravity waves. Bruno Bertotti bet a free dinner that gravity waves wouldn't be discovered by midnight May 5, 1988, ten years to the day after the letter was signed. Thorne obviously lost. “Conceded with sad regret,” he wrote at the bottom of the letter. But ever the optimist, he had already placed another wager on May 6, 1981, with astrophysicist Jeremiah Ostriker of Princeton that a detection would be imminent:

Whereas both Jeremiah P. Ostriker and Kip S. Thorne believe that Einstein's equations are valid.

And both are convinced that these equations predict the existence of gravity waves.

And both are confident that Nature will provide what physical law predicts.

And both have faith that scientists can ultimately observe whatever Nature does supply.

Nevertheless they differ on the likely strengths of natural sources and on the possibility of a near-future and verifiable detection.

Therefore they agree to wage one case of good red wine.

Ostriker wagered a case of French wine. Thorne went for Californian. The detection had to occur by January 1, 2000. Thorne lost again but is eager to place another bet with any ready taker.