ainer Weiss is a dervish, whirling about his office at the speed of sound—his sounds. He's refreshingly direct and rakishly profane. One spring weekend in the midst of LIGO's construction, Weiss is hunched over his desk, worried about a graph displayed on his computer screen. It exhibits the growing density of gases leaking into the newly installed vacuum tubes at LIGO in Louisiana. “This was actually expected,” he says offhandedly. “The same thing happened at the Washington site.” His high-ceilinged office was then situated on the edge of the MIT campus in a ramshackle, three-story building known solely by its numerical designation: Building 20.

If buildings can pass on good fortune, Weiss will have a favorable shot at finding a gravity wave. The wooden structure was built during World War II to carry out military research on a new technology known as radar. After the war, Building 20 witnessed the founding of the modern school of linguistics under Noam Chomsky, the erection of one of

physics' earliest particle accelerators, the refinement of the Bose speaker, and Harold Edgerton's astounding stop-action photography. “The plywood palace,” they called it. Weiss himself punched through its flimsy ceilings and walls to help build the most accurate atomic clocks in the world. Building 20 was supposed to have lasted just through the war. With its stained cedar siding and peeling paint, it remained standing for five decades as field mice and squirrels occasionally ran along the pipes in the corridors. Weiss's research group was the very last to leave before it was finally demolished. The building was too environmentally hazardous to preserve.

Weiss was not eager to pack up. “My whole career was here,” he says, with a sweep of his arm as he walks down the silent hallways with their well-scuffed wooden floors. All the rooms are completely stripped of fixtures and furniture. Only wires and pipes remain dangling from the walls. For the moment, Weiss's pedestrian office, with its worn-out couch and linoleum floor, stands as a lone oasis. Long tables jut out into the room, each tabletop piled high with files stuffed with papers. Surprisingly, Weiss can recite the contents of each stack by heart. He's a talker. Stories, explanations, remembrances all spill out, seemingly simultaneously. He's feisty yet self-deprecating at the same time, a man who doesn't hesitate to tell you the story of how he flunked out of MIT as an undergraduate because he paid more attention to a girl than his studies. Short with thinning gray hair and an assortment of glasses to see at different distances, Weiss is a workaholic who was supposed to cut back after a heart attack (but hasn't).

First and foremost Weiss is an experimentalist. If he hadn't become a physicist, he'd be an electrician. He plainly prefers hanging out with scientists who get their hands dirty rather than those from the chalkboard brigade. It makes his blood boil to hear of experimentalists described as foot soldiers, with the generals—the theorists—looking down from the hill. “That's what those bastards want you to think,” he says, with typical bluntness. “They think they own the field. They don't. Ideas come as much from experimentalists as from theorists. They're not generals. They're jerks like the rest of us. A good book has to be written on general relativity from the experimental approach. Wait until we discover black hole events. To understand what's really hap-

pening there will take more than what the theory can now provide. We're going to turn general relativity into a science, not just a description.” LIGO is a completely different approach to gravity wave detection, extending and advancing the endeavor begun by Weber. Although LIGO is very much a collaborative project, involving dozens of scientists, if forced to name the founding father of the effort, many point to Rainer Weiss.*

Weiss was born in Berlin in 1932. His father, a physician, came from a wealthy Jewish family but rebelled by becoming a Communist and marrying a Protestant, a German actress. Foreseeing the political troubles ahead for Germany, the family moved to Prague. By 1939 they were one of the last Jewish families allowed to emigrate to the United States, largely due to the father's much valued medical degree. Plopped down into the middle of Manhattan, the young Weiss began dabbling in mechanics. “I took things apart all the time,” he says. “Motors, watches, radios. My room was constantly a mess, and I was always getting into trouble for it.” His father, who went on to become a psychoanalyst, was a bit annoyed. As a cosmopolitan German, he was more engaged with the arts and humanities, drama and literature. Weiss's one concession to his father's tastes is an interest in classical music, which he came to love as an adult and expertly plays on the piano.

By high school Weiss began fixing radios for friends and acquaintances, an enterprise that eventually grew into a business. As soon as the war ended, surplus equipment started being sold on the streets of New York. “Equipment was flooding the cities,” says Weiss. “You could go downtown and buy the damnedest stuff. If you knew even a little bit, you could get the most modern technology —transformers and radar sets—for a pittance.” He would either skip school or go down on Saturday and pick up vacuum tubes, capacitors, any electronic component then imaginable. When a huge fire destroyed the Paramount Theater in Brooklyn when he was 16, Weiss salvaged 10 loudspeakers, then the state of the art. He refurbished and sold them. In fact, he became

so successful at making and selling complete audio systems, a novel endeavor at the time, that he almost didn't go to college. His business was highly profitable. But he also had an intellectual quest: he wanted to figure out how to solve the problem of noise, to take the hiss out of a recording as the phonograph needle rubs against the shellac of the record. He specifically chose MIT so he could study to become an audio engineer and learn enough to find a solution. What he didn't count on in 1950 was MIT's strictly ordered universe, far different from his rough and tumble neighborhood in New York where he fixed gang members ' radios to keep from being harassed. All the buildings were known by numbers. All the courses were listed by numbers. “Everything was numbers,” he says. “It was completely wacky. I asked myself, ‘Am I going to survive in this place?' ” He almost didn't.

Bored by his engineering courses, Weiss went into physics, but his record was dismal. Hopelessly in love with a Northwestern University coed, a musician and folk dancer, he barely attended classes his junior year, spending most of his time in Illinois. When she dumped him, Weiss returned to MIT. He passed his exams, but the department flunked him out anyway for nonattendance. Despondent that he might get drafted into the Korean War, Weiss started walking around campus to clear his head. Passing by Building 20, he looked through one of its windows and saw two guys screaming at each other. “One guy was down on the floor looking at a big brass tube. The other guy was perched near the ceiling, and he was adjusting something up there, ” recalls Weiss. “They were trying to find a resonance in a beam of atoms, but it was hopeless.” The two were working in the laboratory of Jerrold Zacharias, the scientist who constructed the first robust atomic clock for commercial use. Listening to the argument, Weiss decided the combative pair needed someone who knew electronics, his specialty. Hired on as a technician, he made himself indispensable. He found a home in Zacharias 's laboratory (which both startled and chagrined his physics professors). That is where Weiss learned to design experiments, to build, and to manage. He got technical training in welding and soldering. With all the lab's high-tech equipment, he felt like a kid in a candy shop. “It was a style of physics that engendered you to do things,” recalls Weiss with obvious fondness.

Eventually, Zacharias bent the rules to get Weiss back into school. As a graduate student, Weiss worked with Zacharias in making more and more accurate atomic timepieces. He specifically worked on a new concept, the atomic fountain, an experiment conducted right in Building 20. The idea was to send a beam of atoms upward. Like balls thrown into the air, the atoms would eventually stop and return to Earth. Once the atoms were slowed down at the turnaround, it would be easier to measure their vibrations, the very crux of an atomic clock. The setup first ranged over one floor. Weiss sent 100 million atoms upward; none were recorded as ever coming down. Weiss soon punched through the ceiling to send the atoms two floors upward, then three. He kept going higher and higher in an attempt to see at least some of the atoms, the least energetic ones, finally stopping and falling back.

Weiss worked on this device for three years, only to discover that all the atoms were moving far more energetically than anyone had anticipated. The atoms were shooting right out of the building entirely. Today, four decades later, a successful version of the atomic fountain has been built using supercooled atoms. It's not three stories tall but rather one inch high. But there was one favorable outcome from Weiss's failed experiment: he caught the gravity bug from Zacharias, who long imagined his atomic clocks being used to test general relativity. Zacharias was hoping to eventually place one of his clocks on a tall mountain in Switzerland and another in a valley some 7,000 feet below in order to measure the gravitational redshift. It would have been an early version of Vessot's experiment, only this time carried out on the Earth's surface rather than in space. Weiss started learning general relativity in anticipation. The experiment was never done, but by then Weiss was hooked on gravity and in 1962 traveled to Princeton, the hub of experimental relativity, for a postdoc with Robert Dicke.

Weiss worked with Dicke on building gravimeters to measure the Earth 's unique resonances, its “ringing” when excited. His detector went off the rails during the devastating Alaska earthquake of 1963. After two years, though, Weiss was eager to return to MIT. He loved its legacy of experimental freedom. “ You could think of something and then do the experiment within a couple of days,” he recalls. “You could do that because there was ‘junk' all around and people who knew how

to use that junk. That's why MIT was such a pleasure back then.” As a newly appointed MIT assistant professor, Weiss decided to measure whether the gravitational constant was changing over time, a project inspired by Dicke's alternate theory of gravity. This spurred him to work on lasers, a necessary piece of equipment for his measurement. At the same time he also began looking into the “tired photon” theory, a hypothesis (now discredited) that cosmic photons lower their frequency by losing energy as they travel through space. This introduced Weiss to the technique of interferometry, the method of choice for such a test.

In the midst of this ongoing research, the educational officer of his department asked him to teach the course in general relativity. “ ‘After all, he said, ‘you ought to know it,' ” recalls Weiss. “I couldn't admit that I didn't know it. I was just one exercise ahead of my students.” He had learned relativity as an experimentalist, not as a theorist, and so taught the course from that perspective. To understand the concept of gravity waves, for instance, he came up with a homework problem. He asked his class to envision three masses suspended above the ground, their orientation forming the shape of an L. One mass would be in the corner of the L, the others at each end. The assignment called for his students to calculate how the distance between the masses would change as a gravity wave passed by. Weiss understood that as a gravity wave moves through space, it doesn't simply compress everything in its path and then, as it passes, expand it again. Rather, it has a multiple effect. It does two things concurrently in different directions. The wave compresses space in one direction—say north-south —while simultaneously expanding it in the perpendicular direction—east-west. Though a gravity wave is a distortion of space-time, it conserves volume. The phenomenon is somewhat akin to the squeezing of a balloon: press in on a balloon's sides and the rubber will immediately bulge out from its top and bottom, in a direction perpendicular to the squeeze. Gravity waves impose a similar effect on space-time. If a gravity wave were to come straight down on a L-shaped setup, passing through the Earth, the masses in one arm would squeeze closer together while the masses in the other arm would move farther apart. This distortion can be visualized by looking at the weave in a piece of

cloth as you pull it along one dimension. The squares of the weave distort in just this way. A millisecond later, as the gravity wave continues onward, this effect would reverse, with the compressed arm expanding and the expanded arm contracting. In the course of working out this homework assignment, Weiss came to see that it could be a doable experiment, especially given his recent work with lasers and interferometry. He figured that laser beams, bouncing back and forth between the masses at each end of the L, could monitor those expand/ contract flutters. Here was a completely different way to detect gravity waves.* What Weiss was imagining was a modified form of the instrument Michelson used in his attempt to detect the ether.

A continuous stream of light from a laser would enter the corner of the system and be split into two beams, each directed down an arm of the L. Mirrors affixed to the center and end masses would then bounce the beams back and forth. (Later, the mirrors themselves became the test masses.) The beams are eventually recombined, at which time they optically “interfere” (hence the term laser interferometry). The beams initially could be set so that their wave patterns are “out of step.” In this way, when added together, the two beams would cancel each other out. The peak of the light wave in one beam is added to the trough of the light wave in the other beam, resulting in a null signal —darkness. But if a gravity wave caused the masses to move, the two laser beams would travel slightly different distances. In that case, when the length of one of the arms changes the tiniest bit, the beams will be more in step and produce some light when combined. The properties of the gravitational wave are hidden within those light changes. Weiss

Laser light enters the system and is reflected in each arm to keep tabs on the length. When no gravity wave is present (left figure), the two arms of a laser interferometer are equal in length. When a gravity wave passes through (right figure), one arm contracts while the other expands, resulting in a signal. The movement here is highly exaggerated. The length changes will be smaller than the width of an atomic particle.

came to see that the light should bounce back and forth many times before being recombined and compared, for the repeated ricochet would increase the total distance traveled and so magnify the difference enough for sensors to detect. The sensitivity would increase even more as the masses were moved farther and farther apart.

Weiss was actually rediscovering an idea that had been in the air but not widely discussed in the gravitational wave community. Several years before Weber even announced his purported signals, two researchers in the Soviet Union suggested using an interferometer instead of a bar to detect gravity waves. But the article, published in a Soviet journal in 1963 by Mikhail E. Gertsenshtein and V. I. Pustovoit, received no notice whatsoever. Others in the field were completely unaware of its existence until Kip Thorne uncovered it many years later. “Gertsenshtein was the ultimate Caspar Milquetoast. He was unbelievably shy and mild. He had a number of seminal ideas that were totally ignored by the world,” says Thorne. Weber independently thought of the idea as well. Although he did not publish it, he did discuss it with his protégé, Robert Forward, who drew a rough sketch of Weber's

scheme in his lab notebook. Weiss later came up with the concept on his own when he was inspired by his classroom exercise. The idea became more and more popular over the years because, as the field evolved, it became quite apparent that bars had a number of limitations. The supercooling, for one, can be tricky. If something goes wrong, it can take several months to warm up the detector, fix it, and cool it back down again. And a bar's size limits the range of signals it can pick up. Fixed in length, a bar antenna can be tuned to only one frequency. If it were an optical telescope, it would be seeing only one color and no other, which limits its view. For many of these reasons, researchers began to focus their attention on the more versatile laser interferometer, which has the flexibility to carry out long-term astronomical investigations. It registers not just one frequency but a whole range.

A NASA astronaut served as the catalyst in getting Weiss's idea out of the classroom and onto the drawing board. In 1967 Philip Chapman had received his Ph.D. at MIT in instrumentation, with a focus on general relativity. Going on to become a scientist-astronaut, he was on the lookout for gravity experiments to conduct in space and consulted Weiss, who had been on his thesis committee. “We were going to the Moon, NASA had plenty of money, and anything seemed possible,” recalls Chapman of that era. Weiss told Chapman about his idea to use laser interferometry to search for gravity waves. Chapman himself had been thinking of approaches other than bars. Enthused by the prospect and looking for further collaborators in industry, he then talked to Forward, who was continuing his bar work at Hughes. (He had three small bars operating up and down the coast of California, one of them in his bedroom closet.) “Phil Chapman put me on to Weiss's interferometer idea,” says Forward, resurrecting the notion he first heard from Weber years earlier. The possibility of a future NASA project got both Forward and Weiss working on the idea independently.

Weiss's role was highly significant in this venture, for he presciently envisioned three decades ago nearly all of the crucial pieces of the laser interferometer observatories presently coming online. Carrying out an extensive design study through 1971 and 1972, the first serious examination of the technique, he identified the fundamental sources of noise with which researchers are now struggling. Moreover, he com-

pletely outlined the approaches needed to control such noises. “I was trying to be like Dicke, who would initiate an experiment by first sitting down and thinking it through completely,” says Weiss. Weiss's thorough analysis, published as an MIT Quarterly Progress Report, is now viewed as a landmark paper, which is still consulted today. It is ironic that his career began with a determination to get rid of the noises in a hi-fi system, only to transfer that interest to reducing the noises that could mask a gravity wave, whose wavelength happens to be in the audio range.

Meanwhile, Forward began to construct a small prototype. He and his colleagues, Gaylord Moss and Larry Miller, spent three years building and enhancing their system. The interferometer was located in a basement room at the Hughes Research Labs in Malibu. Two aluminum pipes, usually used for irrigation in farming, were set at a right angle and served as the laser beam tubes. Each arm was 2 meters long* and aligned to be most sensitive to radiation emanating from the galactic center (where Weber's signals were then supposedly originating). The masses, set in the corners of the L, each weighed a couple of pounds. The entire system was mounted on a granite slab, set on air mounts for cushioning (an earlier setup rested on an inner tube). That was their greatest problem: isolating the instrument from various acoustic and ground noises. It was designed to receive gravity waves over a wide range of frequencies, from 1,000 to 20,000 hertz. As Forward pointed out in a journal article, they hoped that widening the bandwidth would give “significant insight into the nature of the source.”

This tabletop system operated for 150 hours during the nights and weekends from October 4 to December 3, 1972, a time when the center of the galaxy was in prime alignment. The night shift had to be used because of the high levels of noise in the lab during regular business hours. Data collection was fairly tedious and required extraordinary effort. The researchers had to sit nearly motionless for hours at a time as they monitored the interferometer, so as not to introduce any

extraneous noises. The output was recorded on a stereo tape recorder, and they listened in with earphones. “Gaylord Moss and I took turns spending the night ‘observing,' ” notes Forward. “I found it helped to keep my eyes closed and think as if I were part of the apparatus.” One channel recorded the photodetector's output, while the other channel was used to monitor environmental interferences, such as noises in the laser beam, motions of the floor, any clatter in the lab, or audible sounds from the power lines. And in the background, like a rhythmic metronome, was the incessant tic, tic, tic of the National Bureau of Standards time signal, broadcast by radio station WWV. This was to make sure any potential event could be timed to the nearest thousandth of a second.

At times various tones and clicks could be heard rising above a continual white noise hiss. Most of these sounds could be traced to either noises from the laser or thermal and mechanical contractions in the equipment. But occasionally, about once every 10 minutes, there would be a distinct “chirp” from the interferometer that could not be traced to any internal noise or outside disturbance. None of these signals, though, were picked up by bar detectors operating at the same time. “In view of Weber claiming to have seen gravity wave events,” says Forward, “I believed it was worthwhile to operate the interferometer as an antenna for a few months, just to see if there was anything there. I did, and whatever leftover noises I heard on the interferometer were not Weber events.”

Chances were extremely slim that such a small prototype, the first of its kind, would have detected a cosmic signal anyhow. To improve its response, Forward had plans to take his interferometer to a remote site and extend its arms to much longer lengths, possibly a kilometer or more. An optical telescope gains more resolution and sensitivity by going to bigger and bigger mirrors to gather more photons. A laser interferometer gains sensitivity by extending its arms. The expansion and contraction of space-time are simply easier to discern as longer and longer distances are fully measured because the effect is cumulative. If the mirrors are twice as far apart, they will move twice as much relative to each other when a gravity wave passes by. But by the end of the tests at Hughes, Forward had exhausted the funds his company was

willing to spend on a gravity wave telescope. Chapman had left the astronaut corps in 1972, which meant funding from NASA to expand his prototype was not available either. Consequently, the Hughes Laser Interferometer Gravitational Radiation Antenna project came to an end. But others would continue the advancement of this new approach. One of the most innovative was Glasgow physicist Ron Drever.

Drever's interest in gravity was sparked around 1959, just a few years after he received his Ph.D. in nuclear physics from the University of Glasgow. He came up with an intriguing way to test Mach's principle, the suggestion made by Ernst Mach that inertia, the tendency of an object to resist acceleration, arises when a mass interacts with all the other masses in the universe. With that hypothesis in mind, it was plausible to assume that a particle would accelerate differently toward a large collection of mass, such as the center of our galaxy, than at right angles to it, in a direction where mass is more sparse. That's what Drever tested. The particle in his case was a nucleus of lithium. When excited by a magnetic field (for this test the natural magnetic field of the Earth), the lithium could be made to produce an electrical signal at a specific frequency, a distinct spectral line. “I watched that line over a 24-hour period as the Earth rotated. As the axis of the field swung past the center of the galaxy and other directions, I looked for a change,” recalls Drever. A change would indicate that the lithium was indeed being accelerated differently, depending on whether it was directed toward or away from the massive galactic center.

Others had done similar experiments, but Drever, like so many physicists a gadget lover since childhood, did it in a very offbeat way. He put together car batteries and assorted odds and ends in his backyard garden and ran the test from there. It was hardly makeshift, though: his experiment could have detected a shift nearly as small as one part in a trillion trillion. “It beat everyone else who was trying to do it with much fancier stuff,” says Drever. In the end he detected no change at all, at least to the level he could measure. Inertia seems to be the same throughout the universe, no matter where a mass is headed. Such tests in physics are now known as Hughes-Drever experiments. Yale physicist Vernon Hughes independently conducted a similar test at the same time. Afterward, Drever spent a year at Harvard, where he

constructed sensitive radiation detectors for Robert Pound's gravitational redshift experiments.

Through the 1960s Drever built detectors for nuclear physics and other applications. He also dabbled in cosmic-ray physics, studying the light emitted as the cosmic particles raced through the atmosphere. During a visit to southern England to conduct these tests, Drever stopped by Oxford University to hear Joe Weber lecture about his recent claim to have discovered gravity waves. Drever immediately thought, “If he's right, I'm sure we can do better than that,” which brought him into the infant field. He and his group in Glasgow eventually got two bar detectors operating but ultimately saw nothing. “Weber was wrong, and I was very sad,” says Drever. “I was hoping he was right, because then we'd be in business.”

Having no experience in cryogenics, Drever figured he couldn't compete with the supercooled bars then under construction at Stanford and Louisiana State, so he chose a different path. Forward had recently visited Glasgow and had talked with Drever about his pioneering experiment in the Malibu basement. “I thought that the interferometers were likely to be better in the end and also much cheaper,” says Drever, a prime consideration in Scotland where funds for novel projects were scarce. One staff member was good at hunting down local companies that could make things cheaply. The vacuum tanks for their first bars had been made by a firm that manufactured ovens and other food industry equipment. Using old bar equipment and second-hand parts, Drever's group constructed its first interferometer in 1976. The one expensive item was the laser.

Drever quickly learned that laser interferometry was going to be far more difficult than he had initially imagined. The first problem that cropped up was simply light scattering. As light bounced back and forth between the mirrors in the interferometer, with each reflection tracing its own glowing thread as it hit a different part of a mirror, much of that light got lost. It scattered off imperfections in the mirrors. Drever's solution was to switch from a Michelson interferometer to a Fabry-Perot interferometer, a scheme that allowed the light over its many roundtrips to stay as one beam and confine its reflections to a small area of each mirror. This reduced the chance that the light might

ricochet off a “bump” on the mirror and go off in the wrong direction to ruin the measurement. It also increased the light efficiency tremendously. “The big advantage to me as a Scotsman was that this design was much cheaper,” says Drever with a wry smile. That was because the mirrors in this case could be made smaller, as well as the vacuum pipes. That was decisive for the field's advance. The technology did not yet exist to polish particularly big mirrors to the fine levels of smoothness required in this endeavor.

But there was also a downside to this new design: it would not work unless the laser was extremely stable, far steadier than any laser yet available. At that time the wavelength of light put out by the laser would sporadically jitter, which would have made it impossible to keep track of the infinitesimal shivers in the masses mounted in a gravity wave detector. Undeterred, Drever simply invented a means of keeping the wavelength of a laser's light pure and steady. He later discovered that the idea was similar to one Robert Pound had used earlier for microwave cavities. Drever figured a laser could be stabilized—its frequency kept fixed—through a feedback mechanism, locking the laser in a special way to an optical cavity. He visited John Hall, a leading laser expert at the Joint Institute for Laboratory Astrophysics in Colorado, to build such a stablized laser, since the institute had ready access to the necessary parts. Drever and his colleague James Hough also built a cruder version in Glasgow. “The Glasgow one was comical,” says Drever. “It was largely contained in tobacco tins. At that time Jim used to smoke a pipe a lot, and so he had dozens of tins around. They made good screening for the circuitry. The device used about a dozen tobacco tins.”

Just as Drever in Glasgow, Weiss at MIT, and a seminal group in Germany were beginning these investigations into laser interferometry, Kip Thorne was working on the theoretical end of this enterprise. He was making the theory of general relativity “user-friendly” in the search for gravity waves. It was the time when Thorne was a rising star at Caltech, working with his students in making general relativity more testable by generating sets of parameters that experimenters could measure. In the process Caltech was replacing Princeton as a world center in relativity theory. Thorne himself became involved with grav-

ity waves in 1968 when he was introduced to Braginsky. Immediately impressed by the Soviet researcher, Thorne began a collaboration. Until the end of the Cold War, Thorne would spend about a month in Moscow every other year, becoming the unofficial “house theorist” for Braginsky's gravity wave group. From Braginsky, Thorne became convinced of the long-term experimental possibilities of gravity wave detection, although he was skeptical of any short-term success. His wary attitude would soon change, though.

The turning point came at a meeting in the medieval town of Erice, Sicily, a favorite summer spot for physicists to convene at the Ettore Majorana Center for Scientific Culture. Held in an old monastery perched on a cliff overlooking the Tyrrhenian Sea, the 1975 conference had been organized by Weber to take stock of the field and discuss advanced techniques. As a theorist, Thorne had been calculating the gravity wave strengths expected from various astronomical sources. Listening to the experimental presentations, he began to see that researchers had a good chance of getting to the required sensitivities with advanced techniques, such as the use of special materials or going to low temperatures. “I came away from that meeting convinced that the field was very likely to succeed,” he says. “I hadn't had that kind of conviction before. It was because I was seeing the ideas for improvements in the detectors and what one might plausibly expect those ideas to achieve over 10- or 20-year timescales.” As a result, Thorne became the field's most dedicated barnstormer, going around the United States giving talks about the field's promise and the sources that might be detected with the new technologies coming online. “The Weber controversies had left a black mark on the field to some degree,” notes Thorne, “and there was the need to erase that and maintain momentum in the United States.”

Thorne was instrumental in convincing the Caltech faculty and administration to establish a gravity wave detection team at the university, a natural complement to his theory group. Thorne was not wedded to any specific approach at first. “My attitude,” he says, “was to leave it up to whomever we hired to decide the best direction.” Weiss, though, hoped to change his mind. Weiss was chairing a NASA committee in 1975 on relativity experiments that NASA might carry out in

space. He was already chairman of the science working group on another big NASA project, COBE, the Cosmic Background Explorer satellite then being built to measure the vestigial hiss of the Big Bang with exquisite precision. Along with his gravity wave investigations, Weiss has been a major player in ongoing measurements of the microwave background, first with balloons and then from space. He was one of COBE's originators. Because of Thorne's expertise in general relativity, Weiss invited Thorne to come to Washington and speak to his NASA committee. That evening in their hotel they stayed up nearly all night in conversation about gravity wave detection. Thorne at the time didn't hold out much hope for laser interferometers. In a section on gravity wave detectors in his book Gravitation, Thorne had written, “As shown in exercise 37.7, such [laser interferometer] detectors have so low a sensitivity that they are of little experimental interest. ” That night Weiss would begin to convince Thorne that laser interferometry was a contender. (Today, Weiss keeps a copy of that quotation posted on his office door, just to tease Thorne whenever he visits MIT.)

After a committee study, Caltech agreed with Thorne to recruit the world's leading expert in gravity wave detection, someone who would direct construction of a sophisticated prototype, either bar or interferometer, that would allow the university to refine the techniques and hardware necessary for a future gravity wave observatory. Thorne would have liked to have brought in Braginsky, but with the Cold War still in progress it was not possible. Braginsky feared the consequences to his family and colleagues. The transfer would have been viewed as a defection. Weiss was then heavily involved in the early phases of COBE, which diverted his attention. But another name was often at the top of the lists of Thorne's consultants: Ron Drever. “Drever then had the best track record in terms of dealing with technical obstacles. He was preeminent. He had beautiful ideas, ideas that people would pooh-pooh at first and now they're incorporated into LIGO,” says Thorne. Drever's research team in Glasgow was just then beginning to improve its laser interferometer, employing Drever's latest modifications. The size of the instrument was fixed by the length of the available room, an old particle accelerator laboratory. The interferometer arms were 10 meters long. “It was a struggle to make it all work, ” says Drever. It was

in the midst of this start-up that Caltech began to vigorously court Drever, who was ambivalent about moving just as his Glasgow group was making progress. Although Caltech was the “big leagues,” as Drever puts it, he was more attracted to the way in which European universities supported new endeavors. “I was quite happy where I was. You could do a lot with little money. The university employed technicians who could be used on any project. That meant you could try out new ideas without having it tied to some grant,” notes Drever. By 1979, though, Drever finally decided to spend half of his time in California, which gave both him and the university the chance to see if the new Caltech venture was viable. It was. After five years Drever became a full-time faculty member. And with Drever on board, laser interferometry became Caltech's method of choice.

The growing momentum of laser interferometry had already caught the eye of the National Science Foundation. When Richard Isaacson arrived at NSF in 1973 as associate program director for theoretical physics, he recalls his predecessor giving him a bit of parting advice: “ I was visited a few weeks ago by a very clever guy—Rai Weiss,” said Harry Zapolsky. “He has some interesting new ideas about gravity wave detectors. If he comes back, you should pay attention.” Eventually reviewing its national program of gravitational wave detection at the end of the 1970s, the agency decided to expand its funding into this new arena. Also influential was Caltech's stepping forward and investing its own money in the technique. “Physicists tend to follow one another, ” notes Weiss, who had faced delays when he first approached NSF for funding. “Once a large and prestigious university decided to go into it, that gave an extra little nudge.”

With money from both Caltech and NSF, Drever proceeded to set up a full-fledged gravity wave laboratory on the northeast corner of the university's campus. Stan Whitcomb, who became Drever's right-hand man, was brought on board to assist with overseeing its construction. Caltech's aim was to build an interferometer identical to the one in Glasgow, only bigger. This interferometer now resides in a one-story structure, unassuming in its beige tone, that wraps around a corner of the university's engineering shop, forming two long corridors. Only a modest sign on the door reveals the building's purpose. Inside,

the lab's most prominent features are two 40-meter-long steel pipes meeting at right angles. The 40-meter length was not chosen for scientific considerations. Drever would have gone even longer, but a tree was in the way, a tree that no one was eager to cut down. A vacuum chamber stands at the corner of the L as well as at each end of an arm. In each chamber a mirror/test mass is suspended. Each mass is a 5-pound cylinder of fused silica. (When first built, the masses were mounted in glass tanks christened Huey, Dewey, and Louie, after Donald Duck 's nephews—a tip of the hat to nearby Disneyland and typical of the school's humor.) Inside the long pipes the laser beams reflect back and forth—you might say from Dewey to Huey and from Dewey to Louie. Vacuum pumps silently work in the background, keeping the pipes evacuated from stray atoms that could disrupt the light's journey.

To protect the suspended mirrors from such outside disturbances as passing trucks or the seismic tremors that occasionally shake Pasadena, the supports from which the mirrors are hung are cushioned by layers of stainless steel and rubber. When first set up in the early 1980s, toy cars were used for the cushion, a colorful assortment of tiny pink, green, yellow, red, and blue rubbery sedans. It was a clever and cheap rubber source at the time but turned quite troublesome in the end. Outgassing from the toys dirtied the vacuum system. Starting in the early 1990s, the Caltech prototype was completely refurbished, with a new vacuum system and new pipes. The goal was to make it a smaller version of the full-scale observatories then planned for Louisiana and Washington. It is now a test bed for future innovations. A graph on the wall depicts its evolution. When first operating in the early 1980s, the prototype reached a “modest” strain of 10^-15 (able to discern a movement as small as an atomic particle). By 1994 the system reached a strain of 10^-18, a thousandfold improvement. The progress was largely due to a slow but continuing series of technological improvements. Laser power, for example, has increased over the past 20 years, which directly affects the sensitivity of a laser interferometer system. The equipment is also better isolated from seismic disturbances. And, perhaps most crucially, the Caltech detector now uses “supermirrors” as its test masses. Made of layers of dielectric material, they lose only 100 photons for every million reflected.

As soon as he arrived in California, Drever started checking out all the commercial vendors to find out who was making the best mirrors. He heard that Litton was making special mirrors for the military for use in laser gyroscopes. They were not as yet available commercially, but Drever forged a connection and arranged for the company to make a special batch for his new interferometer. “They were fantastic, at least a 100 times better in terms of reflectivity losses,” he says. With such mirrors in hand, Drever was spurred to think of additional improvements. When checking out his new supermirrors, he noticed that the reflected light was strong enough to bounce within the interferometer many times. Given such low light losses, he figured he could “recycle ” the light, have it bounce back and forth over and over again, which essentially boosts the power of the laser and increases the instrument 's sensitivity. It was the development of the “supermirrors,” with their miniscule losses, which allowed Drever to even consider such power recycling. At the time this idea was revolutionary. “ You'd catch the light and send it back in,”says Drever. Now it's standard practice. When a visible-light laser is used, the effect can be spectacular. The laser beam enters the detector and reflects back and forth between the mirrors in each arm some 100 times. It 's as if 100 laser beams are superimposed on one another. If the mirrors are aligned exactly right, so that the beams are in phase, the relatively dim laser beam suddenly brightens within the cavity into a brilliant shaft of light. With those two key improvements—a stabilized laser and power recycling—Drever enabled laser interferometry to turn a corner as a gravitational wave detection system. The technique looked more and more promising in its ability to reach the sensitivities needed to conduct astronomical investigations.

The Caltech prototype has never been a true gravity wave telescope, more a working model to continually improve the instrument design. But that didn't keep Caltech investigators from attempting a test. For 12 days and nights in the winter of 1983, the Caltech interferometer was hastily put on the air after radio astronomers discovered a neutron star spinning what was then a record 642 revolutions per second, possibly jiggling space-time in the process. The 1987 Magellanic supernova was examined, too, though days after the initial burst. In both cases the Caltech detector perceived not a wiggle.

From his previous experience as an experimentalist, Weiss had initially envisioned the field of gravity wave astronomy growing steadily but very, very slowly. He figured that researchers in this arena would have to work on the innumerable technical challenges of laser interferometry before attempting a full-blown observatory. But a series of pressures and frustrations soon changed his assessment. Several years after Weiss returned to MIT's Research Laboratory for Electronics upon completing his postdoctoral fellowship at Princeton, its mission changed. Previously, federal grants to the lab could be applied to whatever ideas the lab was currently working on. Indeed, such funds helped him set up his first laser interferometer, a prototype with 1.5-meter-long arms, and support the graduate students building it. But in the thick of the Vietnam war, a new rule was imposed which required that all research funded by the Defense Department, the lab's major funder, have a direct bearing on the military's needs. As a result, cosmological and gravity-related projects eventually lost their support. At the same time, Weiss was getting little respect from the MIT physics department, which was then more concerned with enhancing its solid-state division. “The faculty gave my students such a terribly hard time,” he says. “They sneered at the extremely low sensitivity of the instrument.” Indeed, in those early days Weiss found it difficult to convince many in the physics community that this new approach to gravity wave detection had the potential to surpass the bars in sensitivity. Some thought the scheme was far too complicated—and perhaps even an erroneous method of detection.

As a last-ditch effort, Weiss arranged with the city of Cambridge to have the road right outside his laboratory—Vassar Street—closed down two weekends in a row at night. Stopping the trucks from rattling down the road would give his students the necessary peace and quiet to run their sensitive tests. Sawhorses were set up in the street to block the traffic. “We were trying to get my students a thesis,” says Weiss. They obtained a strain of 10^-14, decent for a small prototype but impossibly weak for astronomical investigations. “During their oral exams,” continues Weiss, “my colleagues had the temerity to ask these kids what they discovered. ‘Well,' said one student, ‘We didn't see the Sun blow up.' One professor replied, ‘I can look out the window to see

that. What do we need your data for?' They didn't see the technological aspects of it at all. That's when I decided I was never ever going to put a student in that situation again.” With his funding slashed and his colleagues indifferent, Weiss realized he had to get a full-scale observatory under way. He needed to get into the business of astrophysics as soon as possible. And that meant going beyond tabletop detectors or prototypes with extended arms. It meant going big—very big. In 1976 he began to work on an idea that would become the seed of LIGO.

Right off Weiss envisioned a system with two widely separated detectors. He initially set the length of the arms at 10 kilometers, a little over 6 miles. Such a long length was necessary to make the instrument sensitive enough to detect the waves that theorists said were out there. Knowing that such a sizable facility would cost tens of millions of dollars, Weiss reasoned he'd have to call an international meeting and try to turn its construction into a worldwide effort. He never imagined he could obtain funding from the National Science Foundation alone, by then the only U.S. funder for gravity-related research. “I knew right off that this was going to cost over $50 million,” says Weiss. “It was outrageous to contemplate that the NSF would go into this field that (a) had that terrible history with Weber, (b) had no experience with big projects, and (c) had no scientific basis as yet. It was cuckoo.”

But as soon as NSF made its decision to fund the Caltech prototype, a sizable investment for the agency, Weiss decided to be more aggressive and offered his idea to leapfrog to an even bigger detector. Among gravity wave researchers at the time, Weiss was the only person with a firm grounding in major physics projects, being in the thick of building the COBE satellite. From that experience he came to learn it would be wise to build a large facility fairly quickly, and then develop and advance the techniques along the way. “Rai recognized, far more clearly than anyone else, that the only way to get to the required sensitivity was to build something with long arms,” recalls Thorne. NSF gave Weiss the go-ahead to conduct a feasibility study of his ambitious scheme, then called the “Long Baseline Gravitational Wave Antenna System.” Most needed were realistic cost estimates for constructing such a facility. Completed in 1983, in collaboration with Peter Saulson and Paul Linsay, the study (now familiarly known as the “Blue Book”

for the color of its cover) ultimately convinced NSF to initiate research and development. The decision came with one understanding: any major laser interferometer observatory had to be a joint project between Caltech and MIT. There were both political reasons—the combined clout of two prestigious institutions to get the financial support from Congress—and technical reasons. Big interferometers required a much larger effort than you could mount with one professor and a handful of assistants. Weiss had anticipated such a collaboration forming all along.

With this decision, gravity wave astronomy moved toward the big leagues. Weiss, Drever, and Thorne were in charge of overseeing the new collaboration. Given Thorne's Russian connection, the three came to be known as the “troika.” The troika's first order of business was to put together a detailed construction plan for two full-scale interferometers, which had been reduced to having 4-kilometer arms due to budget and engineering considerations. (The available sites, for one, were limited in size.) The NSF's response to their initial ideas, though, was decidedly cool. Their suggested schemes were judged not good enough to be viable, especially in an era when federal budget woes were putting the brakes on other big science projects. “Caltech and MIT were simply not ready,” says NSF's Isaacson. “Their plans were premature.” As a result, the troika went back to the drawing board to rework their plans but continued to get lukewarm critiques. Normally, that would have been the death knell for a proposed science endeavor, but Isaacson and Marcel Bardon, then director of NSF's division of physics, had faith in the idea and pulled every string to keep the project alive. They made sure that money was provided to continue research and development. “Bardon in particular recognized the technological promise,” recalls Isaacson. “The intellectual excitement was overwhelming. But what we had to do was reduce the risks. We wanted to nurture the dream until Caltech and MIT were up to the job of managing such a project.”

“It was a miracle,” says Weiss. “So many things would have killed it, but the National Science Foundation was responsible for keeping it going.” As a result of this favored handling, the proposal became highly visible within the scientific community—attention that brought much

criticism along with it. Richard Garwin, Weber's nemesis, began to loudly question the worth of building a large gravity wave observatory so soon. He did not trust the grand claims being made for it by its supporters. With assistance from Caltech and MIT, NSF answered by assembling a blue-ribbon panel, including Garwin, to advise the agency on “going big.” The panel met for a week in Cambridge, Massachusetts, in the fall of 1986, bringing in the major players in gravity wave detection from around the world to discuss the prospect. Also at the meeting were members from industry to discuss the technical feasibility of making the required optics, lasers, and servo systems. “It was the turning point for the field,” says Thorne. “This hard-nosed committee produced in the end, after much internal debate, a unanimous report. It said that the field had great promise and the correct way to do it was to build two big interferometers right from the beginning, because you can't get any science with one by itself.”

But this support came with a strong caveat: urged on by Weiss and others, the committee recommended that the troika be disbanded and replaced by one project director. There had been disruptive tensions between Weiss and Drever all along, technical differences of opinion that made it difficult for the two institutions to work together effectively. “It was five years of sheer agony for everyone,” says Weiss. Drever has always been driven by an intuitive physical instinct. Weiss is far more analytical. Drever is essentially a loner, while Weiss is experienced on big projects and more realistic about the compromises required in such a setting. Drever preferred fine-tuning the prototype before jumping up in size. Weiss was eager to build big, then tweak. Thorne was caught in the middle. This mismatch in temperaments became a serious problem for the project, holding up some critical decisions. Consequently, NSF demanded that a single director be put in place with authority over the entire project. Caltech and MIT found that director in Rochus Vogt, known to all as Robbie.

Vogt took over in June 1987. He had a distinguished track record. Trained in cosmic-ray physics at the University of Chicago in the 1950s, Vogt served as the first chief scientist at the Jet Propulsion Laboratory in the mid-1970s, making science preeminent at the NASA facility. “He was also one of the best chairmen I have ever seen heading up the

division of physics, math, and astronomy at Caltech,” says Thorne. It was then that Caltech was building its millimeter-wave radio astronomy array in California's Owens Valley, which was in dire trouble during its construction and about to be canceled. Vogt went in and pulled it together on time and on budget. But Vogt is also known as a tough man, who garnered political enemies over his career. As a youth in Nazi Germany, he developed a fervid distaste for wasteful authoritarian bureaucracies. With his short-clipped hair and black-framed glasses, he strikes one as a taller and leaner Henry Kissinger. Vogt was available only because he had just been fired as Caltech's provost, due to conflicts with the university president. Reluctant to accept the directorship at first—he yearned to get back to being a “real scientist who plots data”—Vogt eventually relented. University trustees sold the project to him as Caltech's next Palomar telescope, for many years the most powerful optical telescope in the world. Vogt 's organizational skills turned out to be invaluable to the project, which at last had a name—LIGO. Vogt personally husbanded the final proposal, getting his fingers into all the nitty-gritty details and corralling everyone to get involved. He thought of himself as the “resident psychologist.” Right off, he broke the scientific deadlock between Drever and Weiss, by choosing Drever's Fabry-Perot design over the Michelson interferometer favored by Weiss. He even persuaded Thorne to turn experimentalist and come up with a successful solution to extinguishing stray light in the vacuum pipes. Vogt, now LIGO's most ardent champion, was certain that if the project were terminated, it would kill gravity wave astronomy for a generation.

The final proposal for LIGO got strong reviews at NSF in 1990. An external review panel also gave it a thumbs up. But the requested money was so sizable—an initial $47 million outlay of a total $211 million construction cost—that it had to get Congress's okay. This was a first for NSF. Unlike the Department of Energy, which regularly deals with large projects such as particle accelerators, NSF had never before sponsored a project so large that it required line-item approval in the federal budget. Opposition immediately arose in the astronomical community, which proclaimed that such money would be better spent on telescopes. At the time $211 million was twice the total of the NSF astronomy budget. Astronomers were angered that NSF had chosen to

put the money into a gamble, rather than into already-proven technologies. Was it worth such a high price to find a gravity wave, the critics were asking Congress? Nobel laureate Philip Anderson, a condensed-matter physicist, wondered aloud “if it didn't have Einstein's name on it, would you give a damn?”

LIGO researchers were particularly dismayed when a former member of the gravitational wave detection community, Tony Tyson, testified against the project before a panel of the House of Representatives Committee on Science, Space, and Technology. He stressed to Congress that LIGO “demands a truly phenomenal increase of sensitivity,” up to 100,000 times more sensitivity than the Caltech prototype was then currently attaining, before it could hope to obtain important astronomical data. A LIGO endeavor, to Tyson, was simply “premature,” since too many engineering problems were still unsolved at the time. He preferred a slow-but-sure approach. “Innovation is not necessarily synonymous with a ‘shot in the dark,' ” he concluded.

Berkeley astrophysicist Joseph Silk got a jab in while reviewing a popular book Thorne had written on general relativity. Silk wrote that LIGO “has misleadingly cloaked itself as an ‘observatory.' Despite its name, any astronomy on the symphony of waves from merging black holes in remote galactic nuclei will have to await a greatly refined second-generation detector, and undoubtedly a vastly more expensive undertaking.” Even bar detector scientists liked to point out that bars were cheaper and more developed. Others questioned the importance of gravity wave studies altogether, insisting that scarce government funds would be better spent on projects with lower price tags and far likelier scientific payoffs. LIGO supporters countered that they were after more than sheer detection. They resented it being called a shot in the dark. What they wanted, they said, was to open up a whole new arena for gleaning information from the universe, a method far different than gathering electromagnetic radiation. Including the word “observatory” in LIGO's title was a deliberate choice, an expression of their intention to use LIGO as an ongoing experiment. Moreover, they noted, LIGO's construction funds were independent of the NSF 's regular budget. Rejection of LIGO did not mean the funds would necessarily be applied to other science projects.

Even with blue-ribbon panels strongly supporting the project,

Congress became wary when eminent scientists stepped forward to object. LIGO was stalled for two years. Vogt was a relative novice in his first dealings with Congress. In 1991 he failed to get the go-ahead for construction, although he did get funding for further engineering and design work. Concerns over the federal budget deficit were high. Congressmen questioned whether they were ready to invest such a large sum of money in an unproven facility. The next year Vogt honed his lobbying skills with the help of a consultant, learning to sell the story of gravitation to key legislators. While in Washington, for example, Vogt wrangled a short 20-minute meeting with Louisiana Senator J. Bennett Johnston, who later became an ardent, behind-the-scenes supporter of LIGO, especially when his home state was chosen as one of the two sites. “After I had my 20 minutes,” recalls Vogt, “Johnston's senior staffer looked at his watch and said, ‘Senator, the 20 minutes are up. Let's go.' But the senator responded, ‘Cancel the next appointment.' ” Johnston was so captivated by Vogt's tales of cosmology that he cancelled the following appointment too, as well as the one after that. The two ended up sitting on the floor by the coffee table, while Vogt drew pictures of curved space-time. Einstein's name once again wielded its magic power. In the end Congress appropriated the money.

With the funding came a sudden shift in the tenor of the project. The transition is quite evident at both MIT and Caltech. Blueprints and photographs line the hallways. Memos clutter the desktops. The offices more resemble an industrial corporation than an ivory tower. The initial laser interferometer team, when it first started up, was quite small. There were about a dozen people on each coast, including technicians, scientists, engineers, and administrative personnel. Today, there are more than 150 on staff, two-thirds at the Caltech headquarters and the rest at MIT and the two detector sites. The jump from prototype to mature facility was a canyon-sized leap: the arms going from 40 to 4,000 meters, a factor of 100. “It was a big transition,” says MIT's David Shoemaker, a deputy detector group leader for LIGO. Scientists had to move from their individual lab environments, where they had total control, to a hierarchical facility. Participants now document their every move and deal with myriad outside companies. Once a singular pursuit, gravity wave detection has become a networking of many players, each serving a specific role.

There were repercussions to this change. As with any burgeoning enterprise in science, LIGO has had its share of heated discussions, compromises (both political and scientific), and struggles between pioneering scientists with strong and volatile personalities. New fields often attract the risk takers, whose passions and fervor can be difficult to handle day after day. Some observers even label it hubris, the unwillingness to concede that someone else might see a better way to handle a particular problem. “Most of the conflicts could be blamed on the shift from tabletop physics to big physics,” says Peter Saulson, who has been a LIGO participant since its beginning, now as an independent researcher. “It's a time when you have to step out of your laboratory, set up deadlines, and follow budgets. Many of the original people didn't have this experience. Knowledge had to be brought in from the outside. At first we thought we were uniquely cursed. But I have since found out that it's part of every major science endeavor, where you invent something from scratch and transfer it to a larger arena. ” Construction of the 200-inch Palomar telescope, for decades the largest optical telescope in the world, faced an identical crisis. Ronald Florence, writing on the telescope's long development, said that the scientists' “insistence on exploring, designing, and engineering every step of the project from scratch—doing basic research on subjects as varied as oil bearings, the wind resistance of dome sections, and the chemistry of glass—was for [the chief administrator] a sure route to a quagmire of indecision that would never see the telescope built. . . Men who were building a unique machine, an instrument perfect enough to explore the secrets of the universe, didn't like that attitude.”

A LIGO member with experience in industry saw this history play out once again. Research scientists, long used to laboratory independence —the freedom to change an experiment at will—became upset over LIGO's rigorous schedules and their inability to make last-minute changes in the instrument. Scientific considerations suddenly had to bow to financial and engineering concerns. This meant that certain technologies had to be locked in early, even though advances may have developed later. Some adapted; others left. One unwilling victim was Ron Drever.

Originally brought in to jump-start Caltech's entrance into gravity

wave detection, Drever was unceremoniously ousted in 1992 after growing conflicts with the LIGO team, particularly Vogt. A short portly man with a bulbous nose and warm blue eyes—a Santa Claus without the beard—Drever is always talking and chuckling, ideas constantly bubbling up fast and furiously. And for some this rambling creativity posed a problem. Drever was most comfortable working in his own laboratory, where he could constantly alter his experiments as he developed new schemes. He was reluctant to lock into a final design, if a better method was on the horizon. Drever was a veritable fountain of ideas, as many bad as good, but he would champion them all with equal energy.

And then there was Vogt, whose management skills played a large part in getting LIGO's final approval. He was a man revered by many at Caltech for his organizational talents yet who was also feared for his sporadic outbursts and at times harsh tactics. He was in battle mode as he fought both a skeptical scientific community and a wary Congress for acceptance of LIGO in the early 1990s. You were either friend or foe to the LIGO venture, an uncompromising stance that some found hard to deal with. Ultimately, there was a clash of wills: Drever, always coming up with new approaches, wanting to try them out, versus Vogt, the manager wanting to maintain order and discipline because of a rigid time schedule. In a way the strain between them was also a conflict over ownership. Who got credit for LIGO? Drever, the brilliant experimentalist, whose major breakthroughs were recognized worldwide and enabled a facility such as LIGO to go forward, or Vogt, the expert strategist, who got the controversial project approved by Congress?

A review team, brought in to evaluate the LIGO project in the midst of this internal strife in 1992 and 1993, likened the major players to “a dysfunctional family that needs to be split up.” (One observer ruefully joked that Caltech needed “to put Prozac in its water coolers.”) A university committee eventually agreed that Drever had been inappropriately dismissed, and Caltech provided money for him to continue his research on advanced interferometers independently. Yet even with Drever gone, tensions persisted. Vogt ran a tight ship and kept his decisions close to the vest, for he utilized a managerial approach used at times on covert military projects. “Give me the money and stay out

of my way” was his philosophy, a method he had successfully applied in building Caltech's Owens Valley radio telescope array. Vogt preferred working with a small elite team of scientists and engineers, free of governmental control and costly bureaucratic reviews. He was convinced that only then could LIGO be built for $220 million (the planned construction cost was ever rising). But NSF ultimately decided that LIGO needed to be more open and accountable—its construction activities laid out in meticulous detail and a coherent plan developed for accommodating outside scientists. To do this the agency wanted LIGO's management to shift to a managerial mode long used on high-energy physics projects. Vogt resisted.

In response, after lengthy consultation with both MIT and NSF, Caltech replaced Vogt with physicist Barry Barish as LIGO's principal investigator in February 1994, just a month before groundbreaking took place in Washington state for the first of the two LIGO detectors. Vogt, astute in the politics of science, had shepherded the project for seven years from proposal to funded experiment but was less prepared to deal with managing an evolving large-scale facility that was at last being cast in steel and concrete. Barish, though having no experience in gravity wave research, was tapped for his expertise in handling large physics projects. The cancellation of the Superconducting Super Collider in 1993 made Barish available; he had been heading up one of its detector teams. Barish entered particle physics in its golden era during the 1960s, when new particles were being found fast and furiously. It was a time when universities were abandoning their in-house particle accelerators and starting to use large national facilities. Today, along with his directorship of LIGO, Barish oversees an immense detector in Gran Sasso in Italy, which is on the lookout for monopoles, the hypothetical units of magnetic charge. So operating a facility like LIGO is familiar territory for Barish. The monopole search, a joint Italian-American venture, involves large numbers of detectors spread over an area the size of a soccer field. It is located deep underground, to block out disruptive cosmic rays. Barish understands searching for the will-o'-the-wisps of physics. “In a sense the searches for both magnetic monopoles and gravity waves are very similar,” he says. “But, theoretically, gravity waves are more solid.”

Barish, a longtime member of the Caltech faculty, had watched LIGO undergoing its growing pains from the sidelines. He credits the initial team for performing the necessary conceptual work and pin-pointing the technique's limitations. But, he adds, taking the leap from tabletop physics to a miles-long facility was beyond their level of expertise. They had a certain blindness to this limitation, says Barish, even an arrogance when confronted with reviewers' criticism. They were literally founding a field, yet they hadn't grown up in the environment that particle physicists take for granted. The field of particle physics evolved over decades. With LIGO, gravity wave astronomy was growing up virtually overnight. Its evolution from small to big was occurring at a rapid tempo and that meant certain scientific styles had to give way quickly to new approaches, ones more compatible with a large infrastructure. “In a small lab if you make a mistake, you can go in the next day and fix it,” he explains. “But here, when you are committed to spending a hundred thousand or a million dollars, you can't fix it later. You need to have a system of checks and balances internally. In particle physics that's just part of the structure.” When Barish took over the leadership of LIGO, just months after the superconducting supercollider was closed down, he found a “battle-worn” group recovering from the strains of the Vogt-Drever tempest. He set one simple goal: to build LIGO.