The 4 Percent Universe: Dark Matter, Dark Energy, and the Race to Discover the Rest of Reality - Richard Panek (2011)

Part I. More Than Meets the Eye

Chapter 2. What's Out There

WHAT THE UNIVERSE could be, or should be, didn't much concern her. She wasn't a theorist. She was an astronomer—an observer. The universe was what it was. And what it was, everywhere you looked, was in motion.

Well, she wasn't an astronomer yet. She'd never actually observed, except as a child, using a telescope that her father had helped her build out of a lens she'd ordered through the mail and a cardboard linoleum tube she'd gotten free from a store in downtown Washington, D.C. And that telescope didn't even work properly; she couldn't take pictures of the stars with it, because it couldn't track their motions—or, more accurately, their apparent motions, since it's the turning of the Earth that gives stars the illusion of arcing across the night sky.

She should have known that the camera wouldn't work. The motions of the stars were part of what got her interested in astronomy. Her second-floor bedroom window—the one right above her bed—faced north, and around the age of ten she noticed that the stars appeared to be slowly circling a point in the northern sky, and that over the seasons the stars themselves changed. Ever since, she found that she would rather track the motions of the night sky than sleep. She memorized the paths of meteors, then registered them in a notebook in the morning. Later, in high school, whenever she had to write a research paper the topic she chose was invariably something to do with astronomy—reflecting telescopes (the kind with mirrors) or refracting telescopes (the kind with lenses). At a certain point in the evening her mother might call up the stairs, "Vera, don't spend the whole night with your head out the window!" But she did, and her parents didn't seem to mind, not really.

Hers was, in a way, a Newtonian view of the universe: matter in motion; predictable patterns; celestial objects (and the Earth was one, too, if you thought about it) that, for all their peregrinations, invariably wound up back where they started. But Vera Cooper was born in 1928, three years after Edwin Hubble announced that our Milky Way galaxy was hardly singular, and one year before he presented evidence that the galaxies seemed to be receding from one another—the farther apart, the faster. The only universe she'd known was full of galaxies, and those galaxies were in motion.

And so, as a graduate student at Cornell, when she had to think about a topic for her master's thesis, she tried to update the old clockwork view of the cosmos for the new expanding universe. She reasoned that since the Earth rotated on its axis, and the solar system rotated, and the galaxy rotated, then maybe the universe had an axis too. Maybe the whole universe rotated.

The premise seemed reasonable. Her husband, Robert Rubin, a doctoral candidate in physics at Cornell, had shown her a brief, speculative article by George Gamow in the journal Nature, "Rotating Universe?" Then she heard that Kurt Godel, at Princeton, was working on a theory of a rotating universe.

Her approach also seemed reasonable. She gathered data on the 108 galaxies for which astronomers had managed to measure a redshift. Then she separated out the motions that were due to the expansion of the universe—what astronomers call recessional motions. Did the motions that remained—the peculiar motions—exhibit a pattern? She plotted them on a sphere and thought they did. In December 1950, at the age of twenty-two, still half a year shy of getting her master's degree, Vera Cooper Rubin presented her thesis at an American Astronomical Society meeting in Haverford, Pennsylvania.

Rubin had never suffered from a lack of confidence. When an admissions officer at Swarthmore College told her that because astronomy was her profession of choice and painting was one of her favorite hobbies, she might want to consider a career as a painter of astronomical scenes, she laughed and applied to Vassar. When she got a scholarship to Vassar and a high school teacher told her, "As long as you stay away from science, you should do okay," she shrugged and pursued a BA in astronomy (with a heavy load of philosophy of science on the side). When a Cornell professor told her that because she had a one-month-old son he would have to take her place at the Haverford AAS meeting and present her paper in his own name, she said, "Oh, I can go," and, nursing newborn and all, she went.

The response from the AAS crowd when she concluded her presentation was nearly unanimous: The premtse was odd, the data weak, the conclusion unconvincing. The criticism continued until the astronomer Martin Schwarzschild kindly signaled an end to the discussion by rising and saying, in a high-pitched voice, "This is a very interesting thing to have attempted." The moderator called a coffee break, and Rubin left the meeting.

She herself hadn't thought her paper was extraordinary; it was a master's thesis, after all. Still, she thought that as master's theses went, it was fine. She had taken a pile of numbers and handled them in the most careful fashion she knew, and she thought the result was worth reporting. She thought that she'd given a good talk, and that she'd given it as well as she could. She reminded herself that she had never been to an AAS meeting before, and that she hadn't even met many professional astronomers. Maybe this was just how astronomers behaved. She decided she would file these criticisms in the same category as the comments from the admissions officer and her teachers. The next day her hometown paper the Washington Post ran an article under the headline "Young Mother Figures Center of Creation by Star Motions." So she could console herself that real astronomers would at least know who she was (or, because of a typo, who "Vera Hubin" was).

Still, the experience did teach Rubin an important lesson: She was such a novice that she didn't know how far out of the mainstream her work was. She didn't know that Gamow was nearly alone among astronomers, and Godel among theorists, in finding the question of a rotating universe worthy of serious consideration. Gamow had admitted, in the Nature paper, that the idea of a rotational universe was "at first sight fantastic"—which, at first sight, it was. But what if you didn't trust first sight? First sight—the evidence of the senses, unaided by technology—tells you that the Earth is stationary, that the Sun revolves around the Earth, that Jupiter is moonless and Saturn ringless and the stars motionless, and that the stars are as far as there is. The point Gamow was trying to make was that astronomers needed to go beyond first sight, because now they had a new scale of the universe to consider.

Saying that all the billions of stars we see are part of our galaxy and that billions of galaxies lie beyond our own doesn't do justice to the scale of the universe. Just as our eyes didn't need to evolve to see radio waves in order for us to survive, maybe our minds didn't need to evolve to understand the numbers that astronomers were now trying to incorporate into their thinking. Like cultures that count "One, two, three, more," we tend to regard the scale of the universe—to the extent that we regard it at all—as "Earth, planets, Sun, far."

Consider: How long would it take you to count to a million at the "one Mississippi" rate of one second per number? Eleven days—or, to be exact, 11 days, 13 hours, 46 minutes, and 40 seconds. How long would it take you to count to a billion at the same rate? A billion is a thousand million—that is, a million one thousand times over. So you would have to count a million Mississippis—eleven days of counting—a thousand times. That's 31 years, 8% months. To reach a trillion, you'd have to count to a billion a thousand times—31 years a thousand times, or 31,000 years. A light-year—the distance light travels in a year—is about six trillion miles. To count to six trillion, you would need six sets of 31,000 years, or 186,000 years.

Earlier generations of astronomers had to learn to adjust their thinking to accommodate successive discoveries about new scales of the universe: that the Sun is 93 million miles distant; that the nearest star after the Sun is 4.3 light-years, or 25 trillion miles, away (that's 186,000 years of Mississippis 4.3 times, or about 800,000 years); that our "island universe" consists of billions of stars at similar distances from one another; and that the diameter of this island universe, from one end of its spiral disk to the other, is about 100,000 light-years (186,000 years of Mississippis a hundred thousand times, or more than eighteen billion years of counting, a number you couldn't appreciate without first appreciating the meaning of "billion").

In this context, however unfathomable and even ludicrous, the term "billions of galaxies" at least begins to suggest the difference in scale between the island universe Hubble inherited and the universe he bequeathed to the next generation. His universe was saturated with galaxies as far as the "eye" could see—whether the "eye" was the one he used, the behemoth 100-inch Hooker telescope atop Mount Wilson, or its successor, the 200-inch Hale telescope atop Mount Palomar, which saw first light in 1949, promising astronomers access to galaxies at greater and greater redshifts. Who knew where this emerging reconception of the universe might lead? Astronomers of the mid-twentieth century who wanted to work in Hubble's universe would have to engage with its hist ory and structure on the grandest scale imaginable. They'd have to do cosmology.

Not that Rubin thought of herself as a cosmologist. She didn't even think of herself as an astronomer, and not just because she'd never looked through a professional telescope. Six months after that AAS meeting, she had her master's degree and her husband had his doctorate, and they had moved to the D.C. area to be near his job. Their son wasn't yet one, and they were planning to have another child, and even though her husband kept encouraging her to pursue her doctorate, she felt that life was complicated enough at the moment. So it was her choice not to become an astronomer just yet, and to wonder every day whether she would ever become one. Even so, she felt that nothing had prepared her for this life: living in a suburb, staying home with her son while her husband went to work, crying whenever an issue of the Astrophysical Journal —she'd kept the subscription—came into the house.

Then one day the phone rang. It was George Gamow.

She was standing at the window in her apartment. The phone rested on a table. The sofa was elsewhere. There was no place to sit. Did the cord stretch? No matter. It was the kind of conversation you wanted to have while standing. So she stood and stared out the window and listened as George Gamow asked Vera Rubin about her research.

Her husband shared an office with Ralph Alpher at Johns Hopkins University's Applied Physics Laboratory in Silver Spring, Maryland. Robert Herman had an office down the hall. Gamow did some consulting work for the laboratory, and Alpher and Herman often collaborated with him. From them Gamow had heard about her thesis. He told her that he wanted to know about her work on the rotation of the universe for a talk he was giving at the lab. (She wouldn't be able to attend: No wives allowed.)

Robert Rubin had taken the job at the Applied Physics Laboratory because the proximity to Washington might give his wife educational or professional options in astronomy—options she hadn't yet explored. After the phone call from Gamow, she started taking ApJ to the sandbox, and by February 1952, pregnant with her second child, she was attending classes at Georgetown University, the only school in the D.C. area to offer a doctorate in astronomy. There, by special arrangement with George Washington University, she would be working under the supervision of Professor Gamow.

That spring she met Gamow for the first time. He had suggested they meet in the library at the Carnegie Institution's Department of Terrestrial Magnetism, a modest campus on a hilly, wooded outskirt of Rock Creek Park in northwest Washington. DTM was an unassuming brick building in an unlikely setting. At the top of a hill at the end of a long, curving driveway in a residential neighborhood, it could have been a hospital or a retirement home. Instead, beginning in the first decade of the century, it had been the headquarters of worldwide expeditions to chart the Earth's magnetic field; once that mission was completed, in 1929, DTM had adopted a looser interpretation of investigating the nature of our planet, and began research on nuclear physics and the geology of other planets.

Rubin had visited the sylvan setting on one earlier occasion—a talk of some sort, probably. Now she found herself returning just about every month. The entrance to the library was to the right of the stairs on the second floor. To get from the door of the library to the reading room you had to squeeze through a narrow passage between two sets of bookshelves. Every time she visited she had to hesitate on the threshold, assessing one more challenge to becoming an astronomer. The passage was, perhaps, two feet wide. Pregnant with her second child, she was, perhaps, wider.

George Gamow turned out not to be the sort of person she might have hoped. When they didn't meet at the quiet, wood-paneled DTM library, they met at his home in Chevy Chase, Maryland. There he would invariably be shouting abuse at his wife in some distant part of the house. Where were his papers? What had she done with his papers? Why was she always going through his papers? Whether Gamow's wife was ever actually there, Rubin couldn't be sure. In the summer of 1953, Rubin and her husband paid their own way to an astronomy workshop in Michigan; Gamow was there too, and his behavior embarrassed her. He dozed during talks, and when he woke up he asked questions that had already been answered. During afternoon discussions, just her and him and the great Mount Wilson astronomer Walter Baade, Gamow would down half a bottle of liquor. During his own lecture, he perspired alcohol.

Rubin was beginning to realize that there are two kinds of geniuses. There's the kind we would all be if we were extremely smart and knew what we were doing. And then there's the kind you can only watch, knowing that your mind could never work that way. That's the kind of genius Gamow was. He may have dozed during lectures and asked redundant questions, but he also answered questions that nobody else could answer. Whatever Gamow's personal failings, when he spoke, you listened.

"Is there a scale length in the distribution of galaxies?" he said to her at one of their first meetings. He was suggesting that she think not just about the overall motions of galaxies, as she had in her master's thesis, but instead about the overall result of those motions: the arrangement of galaxies.

Was the distribution of galaxies throughout the universe random and uniform, as most astronomers assumed? Hubble himself had thought so. "On a large scale the distribution is approximately uniform," he had written in his highly influential 1936 book, The Realm of the Nebulae. "Everywhere and in all directions, the observable region is much the same." In a sense, he was simply reiterating the two assumptions of modern cosmology, homogeneity and isotropy, in layman's terms. But the way he was framing the issue was also reminiscent of the premodern island-universe thinking—emphasis on "island." In Hubble's view, and therefore the view of a generation of astronomers, the galaxy clusters that astronomers had observed would be accidents of nature, or perhaps a sort of cosmic optical illusion arising from multiple galaxies falling along our line of sight. But Gamow was thinking on a different scale. Maybe the peculiar motions of galaxies—the motions that were separate from a straightforward outward expansion—weren't random, as most astronomers assumed. Maybe the gravitational interactions among galaxies, even across previously unthinkable distances, were sometimes strong enough to counteract the expansion on a local level. Maybe no—or at least not every—galaxy is an island.

The premise seemed respectable to Rubin, and not just because the visionary George Gamow was suggesting it. Shortly after that first phone call from Gamow, she had received a letter from Gérard de Vaucouleurs—a French astronomer then working in Australia—and then she heard from him again, and then again. She found the correspondence relentless; she always seemed to owe him a letter. But she couldn't complain. As was the case with Gamow, de Vaucouleurs wanted to discuss her master's thesis. He wrote to her that he had noticed a pattern among the galaxies similar to the one she had possibly detected, and in February 1953, midway through her doctoral work, her patience with the persistent de Vaucouleurs paid off. He began an article in the Astronomical Journal with a citation from her work: "From an analysis of the radial velocities of about a hundred galaxies within 4 megaparsecs Mrs. V. Cooper Rubin recently found evidence for a differential rotation of the inner metagalaxy." To de Vaucouleurs, however, her evidence seemed to suggest not the rotation of the universe but the motion of a cluster of galaxy clusters—a supercluster. Even so, his argument was a variation on the theme Gamow was now asking Rubin to consider: Did galaxies cluster, and if so, why?

Once again she marshaled the data that was already out there, available to anyone, this time galaxy counts from Harvard. And once again she applied a conceptually straightforward analysis, this time plotting the galaxies in three dimensions by comparing locations in the sky with distances suggested by their redshifts.

She learned to balance becoming an astronomer and being a mother, sometimes literally: a thick German textbook in one hand, the handle of a baby carriage in the other. Two or three evenings a week she attended classes in the observatory at Georgetown. She worked on her thesis at night, after the children were in bed. She finished her studies in two years, and her thesis, "Fluctuations in the Space Distribution of the Galaxies," appeared in the July 15, 1954, issue of Proceedings of the National Academy of Sciences. Her conclusion: Galaxies don't just bump and clump arbitrarily; they gather for a reason, and that reason is gravity.

This time she didn't receive a drubbing, as she had at Haverford. The reaction was worse: silence.

During an AAS meeting in Tucson in 1963, Vera Rubin went on a tour of the Kitt Peak National Observatory, in the desert mountains fifty-five miles southwest of the city. Rubin was by then the mother of four and an assistant professor in astronomy at Georgetown, but she was still not a practicing astronomer. "Galaxies may be pretty remarkable," she liked to explain, "but to watch a child from zero to two is just incredible." Her youngest, however, was now three.

She secured time at Kitt Peak later that year. With her students at Georgetown she had been studying the motions of 888 relatively nearby stars, following the same methodology she'd used in her master's and doctoral theses—by consulting catalogues.* Now she would be continuing that work, only she would be using a telescope and collecting the evidence herself.

While most astronomers at the time were studying the motions of stars in the interior of the Milky Way galaxy, she went in the other direction—the galactic anticenter, astronomers call it: the stars that lie at greater distance from the galaxy's central bulge than our own star, the Sun. The following year she received an invitation to become the first woman to observe at Mount Palomar, in the mountains northeast of San Diego. She decided that the time had come to do what she couldn't do as an assistant professor using the increasingly limited resources at Georgetown: become a full-time astronomer.

By now she was living in a quiet, leafy neighborhood near the Department of Terrestrial Magnetism, where she used to meet Gamow in the library. Occasionally she would make the fifteen-minute walk there to visit her friend Bernard Burke and discuss the radio-astronomy analysis he was performing on the rotation of the Milky Way. In December 1964, however, she visited for a different purpose. Even though DTM, founded in 1904, had never had a woman staff member, she walked into Burke's office and asked for a job.

"He couldn't have looked more surprised if I'd asked him to marry me," she told her husband that night.

Once Burke had recovered his composure, he took her to the communal lunchroom and introduced her to his colleagues; she was impressed that one of them, W. Kent Ford, had recently returned from Mount Wilson. Someone encouraged her to go to the blackboard—it was that kind of lunchroom—and talk about her latest work. Before Rubin left DTM that afternoon, Merle Tuve, DTM director and a longtime (since the 1920s) staff scientist, gave her a two-inch-by-two-inch photographic plate and asked her to perform a spectroscopic analysis. After she returned the plate along with her analysis, Tuve phoned to set up an appointment.

She said she could be there in ten minutes.

He said the following week was fine.

She said she could be there in ten minutes.

As was the case at all the Washington labs of the Carnegie Institution, the responsibilities of a staff scientist at DTM included no teaching, no tenure, and infrequent, if any, grant-writing. All it required was the ability to maintain a collegial atmosphere with colleagues and produce meaningful science. In Rubin's case, she had the choice of sharing an office with her friend Bernie Burke* or with W. Kent Ford. Burke was an on-staff radio astronomer, as were Tuve and Kenneth Turner; she noticed that the radio astronomers had commandeered a room on the first floor, thoroughly blanketing a large table with geological layers of charts and other paperwork. Rubin didn't want to immerse herself in their world. She wanted a world of her own, and she figured she was more likely to find it in Ford's office.

Ford was an instrumentalist. He had recently built an image-tube spectrograph—a variation on the standard instrument that records the electromagnetic spectrum from a source of light. His version, however, didn't photograph the light from a distant object. Instead, it converted those faint photons into a fountain of electrons, which in turn sprayed onto a phosphorescent screen, which in turn gave off a vivid glow—and that was what his instrument photographed with all the clarity of a "normal" camera. The intensity of the image compensated for the dimness of the distant source of light. As a result, the instrument reduced the exposure time to one-tenth that of an unaided photographic plate. In Ford's new spectrograph—officially the Carnegie Image Tube Spectrograph—Rubin saw the chance to join the hunt for what was then astronomy's hottest prey.

Quasars—short for quasi-stellar radio sources—were extraordinarily powerful pointlike signals, possibly from the farthest depths of space. Their discovery in 1963 provided breathtaking evidence for astronomers that the universe visible in radio waves is not the universe we see with our eyes. And the quasar work that Rubin and Ford did with the new image-tube spectrograph was not unrewarding. Only months after they'd published one of their findings, Jim Peebles was using their data to advance a theoretical exploration of the early universe. Rubin was thrilled. Her research, she marveled, was contributing to a subject she had never even thought to investigate.

On the whole, though, the two years she spent chasing quasars were burdensome. The field was too crowded, competition for time on the big telescopes favored more established astronomers from more mainstream institutions, and the pressure to provide data to her non-image-tube-blessed colleagues was crushing. Constantly they insisted on answers even though she wasn't yet sure her answers were right.

This wasn't the way she wanted to do astronomy. She already had enough personal pressures in her life; she didn't need professional pressures, too. So she quit quasars. She was beginning to realize that in her earlier work she hadn't known what the mainstream was because she herself wasn't working at mainstream institutions. Cornell was no Harvard or Caltech when it came to astronomy; Gamow and de Vaucouleurs weren't the masters of Mount Wilson or Mount Palomar. In those days, though, her outsider status had been inadvertent. Not this time. At least now, she told herself, she knew what the mainstream was; she knew what she was leaving behind. She would shape her observing program accordingly.

She needed to find a subject she could explore with small telescopes, the kind that generally would be more available to someone of her relatively junior status. She wanted a research program that nobody would care about while she was doing it. But she also wanted it to be work that the community would eventually be glad someone had done.

She found it next door, cosmically speaking: Andromeda, the nearest galaxy that resembles our own.

"Within a galaxy, everything moves," Rubin would write. "In the universe, all galaxies are in motion." Every two minutes "the earth has moved 2500 miles as it orbits the sun; the sun has moved 20,000 miles as it orbits the distant center of our galaxy. In a 70-year lifespan, the sun moves 300,000,000,000 miles. Yet, this vast path is only a tiny arc of a single orbit: it takes 200,000,000 years for the sun to orbit once about the galaxy."

Yet such is the scale of the universe that astronomers don't see galaxies actually rotating. If observers in Andromeda were studying our galaxy—a scenario that Rubin enjoyed imagining—they would see an apparently motionless spiral. So do we when we look at Andromeda. The spectrograph, however, would tell a different story: how much the light from Andromeda had shifted toward the blue or the red end of the electromagnetic spectrum—how fast it was advancing toward or receding from Vera Rubin.

In effect, she had inverted her earlier approach to the image-tube spectrograph. She would still be looking at fainter and fainter objects. But rather than pushing deeper and deeper into space, she would be looking at subtler and subtler details close to home. And she would be doing it in record time.

When the American astronomer Francis G. Pease studied that same galaxy in 1916, he needed eighty-four hours of exposure time over a three-month period to record a spectrum along one axis of the galaxy; the following year he needed seventy-nine hours over a three-month period to record a spectrum along the other axis. Instruments had improved since then, but even by the mid-1960s obtaining a single spectrum of a galaxy still took tens of hours over several nights (assuming that you could even guide the telescope precisely enough and keep the spectrograph stable enough for such a long period, always iffy propositions). Ford's new instrument, however, could reduce the exposure time by 90 percent. Obtaining four to six spectra in one night was routine. In Ford's instrument Rubin saw the potential to measure the rotation motions of Andromeda farther from its central bulge than any astronomer had ever measured on any galaxy before.

Again and again Ford and Rubin made the trip to the two main observatories in Arizona. On occasion their families joined them— Ford had three children, and the two families socialized in Washington—but mostly they went alone. In the dark of the dome, Rubin and Ford would sometimes bump heads—literally—as they each tried to be the one to guide the instrument. In general, though, their competitiveness was restricted to who would spot the first saguaro on the drive south from the Lowell Observatory in Flagstaff. For part of the way on the three-hundred-mile drive—through Phoenix and Tucson to Kitt Peak, with Ford's image tube safely tucked in back— they talked about their children. But mostly, during those days and nights out west, they talked about science.

At the December 1968 AAS meeting Rubin announced that she and Ford had achieved their goal. They had gone farther from the center in their observations of Andromeda than any other astronomers had gone in observing a galaxy. After Rubin's talk, Rudolph Minkowski, one of the most eminent astronomers of the era, asked her when she and Ford were publishing their paper.

"There are hundreds more regions that we could observe," she said, referring to Andromeda alone. She could gather this sort of data forever. It was beautiful. It was clean. It was unobjectionable. It was what it was.

Sternly, emphatically, Minkowski addressed her. "I think you should publish the paper now."

So she and Ford did. But they knew that before they could submit a formal paper on their research, they would have to address a problem that had bothered them from almost the first night of observing Andromeda.

Going into the darkroom, Rubin had expected to detect the pattern that holds for the planets in our solar system: the farther the planet from the Sun, the slower the orbit—just as Newton's universal law of gravitation predicted. A planet four times as far from the Sun as another planet would be moving at half the velocity. A planet nine times as distant would be moving at one-third the velocity. Pluto is one hundred times as far from the Sun as Mercury, so it should be moving—and does move—at one-tenth the velocity of Mercury. If you plotted this relationship between distance and velocity on a graph—the farther the distance, the slower the velocity—you would get a gradual falling-off, a downward curve.

That's what Rubin and Ford had assumed they would see in plotting the relationship between distance and velocity in the different parts of a galaxy: The farther the stars were from the center of the galaxy, the slower their velocity would be. That's what astronomers had always done—assumed they would get a downward curve, as if the great mass of stars making up the central bulge in the galaxy affected the wispiest tendrils in the same way that the great mass of the Sun in our solar system affected the wimpiest planet. But those astronomers hadn't actually made those observations because, without the benefit of Ford's spectrograph, they couldn't have. Instead, they drew their assumption as a dotted line. Rubin and Ford, however, had pushed the observations farther than ever, as far as the image-tube spectrograph would allow them, to the farthest edges of the spiral. But they couldn't help noticing that the outermost stars and gas seemed to be whipping around the center of the galaxy at the same rate as the innermost stars and gas. It was as if Pluto were moving at the same speed as Mercury. Plot the rotation curve of Andromeda, and it wasn't a "curve" at all.

Maybe the gas was interacting with the stars in some way Rubin couldn't imagine. Maybe Andromeda was just an oddball galaxy. Maybe a theorist could supply a logical explanation. They submitted their paper to the Astrophysical Journal in the summer of 1969, and in it Ford and Rubin declared that "extrapolation beyond that distance is clearly a matter of taste." Her taste, Rubin would say in private, was that plotting data that didn't exist was "offensive." So she and Ford agreed that they would plot only what they got. It was what it was.

And what it was, was a flat line.

Shortly after Rubin finished her work on Andromeda, her good friend Morton Roberts, at the National Radio Astronomy Observatory in Charlottesville, Virginia, called to say he was driving over. He had something he wanted to show her.

They met in a basement conference room at DTM, along with a group of three or four other DTM astronomers. Roberts, too, had been studying the rotation curve of Andromeda, except his observations were at radio wavelengths. He placed a copy of the Hubble Atlas of Galaxies on the table and opened it to a photograph of Andromeda. Then he laid the plot of his radio observations on the photograph. He had pushed far past the familiar cyclone of stars and gas, far past the point that Ford and Rubin had managed to reach with their optical probes, into a ring of hydrogen gas clouds. But a graduate student from Harvard who was spending some time at DTM, Sandra Faber, seemed unimpressed.

"There's nothing new in this," she said. "It's all part of the same problem. Velocity has never made sense."

She was right. As Rubin herself had shown, velocities of galaxies varied all over the map of the heavens. But for Faber the problem was a given. Unlike Rubin, she'd come of age in a universe that was in motion in more ways than anyone had ever imagined.

"Don't you understand?" Roberts said. "The galaxy has ended, but the velocities are flat." He gestured at the points he'd plotted. "What is the mass out there? What is the matter? There's got to be matter there."

They all stared at the photograph. Here was this beautiful swirl of billions of stars—the kind of majestic image that had captivated astronomers for more than half a century—though that's not where they were looking. They were looking beyond it. Beyond the bulge, beyond the stars, beyond the gas of the spiral arms—beyond all of the light, whether optical or radio. And even though there was nothing to see there, the small group of astronomers understood that they were nonetheless looking at the Andromeda galaxy.

It was what it wasn't.