Books > A Grand and Bold Thing >

Excerpts

Chapter 1
Stakes Worth Playing For

I earned my spurs by doing other stuff, but instruments are what I enjoy.

—Jim Gunn, Princeton University

JAMES EDWARD GUNN is thin and pale and below medium height; his hands are outsized and look remarkably competent. He has mild, dark eyes and a look of remote sweetness. Sometimes he goes into trances and you can’t tell whether he’s concentrating on something he’s looking at, or thinking about something else, or just in a state of sleep and energy deprivation. For some reason, he almost always wears Hawaiian shirts.

With nonastronomers, he is patient with ignorance and polite to the point of deference. He’s patient and polite with astronomers too, though now and then he gets wrought up at what seems to him to be willful stupidity or inattention to exactness. Once wrought up, he continues to talk politely but with increasing intensity, and some words are in aural italics or, if he’s writing, in capital letters. Sometimes he ends up saying something he shouldn’t. When he’s not traveling, he spends most of his time in the basement of Prince-ton University’s Peyton Hall, in a crowded but highly organized electronics workshop. In the middle of the workshop is a circular six-foot vacuum chamber that has been obsolete for years but is too big to be moved out through the door; Jim put his desk behind it, where, from the door, he’s invisible.

Jim is famous. He has a named chair in one of the world’s best astronomy departments and has won nearly every prize available to an astronomer, including a MacArthur and the National Medal of Science. Astronomers admire him for his ability to build astronomical instruments, use them to make observations, and explain the observations with theory. Theorists claim Jim as a theorist and observers claim him as an observer. He’s that good at everything.

Jim was born in 1938 in Livingston, a small town in the oil fields of east Texas. His father was an exploration geologist who moved to new oil fields every six months or so; Jim grew up in Mississippi, Alabama, Georgia, Arkansas, Louisiana, Oklahoma, Texas, and Florida. With all that moving and because he was an only child, he was pretty much a loner. His closest friend was his father, who had set up a machine shop in a trailer that moved with them—during the war, when parts were hard to get, his father had to make them himself. From his father, Jim learned the construction of intricate things.

Jim was five years old when he read The Stars for Sam, which, while written for children, was not written for five-year-olds. The book said that all the stars in the sky were in our galaxy, the Milky Way, and that the Milky Way was only one of many galaxies, all filled with their own stars and scattered throughout space. Each one of those galaxies could be called an island universe. All together the universe held 30 million of them, and they could be classified by shape—spiral, elliptical, etc.—and cataloged as in a museum. When Jim was six, he and his father sent away for lenses and built a telescope. The next year, Jim found his father’s old undergraduate astronomy textbooks and read them, fascinated by the idea that the universe had a beginning.

When he was twelve, his father died suddenly of heart disease and Jim felt part of his life had emptied out permanently; he didn’t like to think about it. So he focused his energies on power tools. He designed and built twenty or thirty model airplanes, a couple of telescopes, furniture, a hi-fi, some rockets, and played around with high explosives.

High school was mostly in Beeville, Texas, where he was astonishingly good at science and math, and where he read Frontiers of Astronomy by the British astronomer Fred Hoyle. Hoyle hadn’t liked the theory that the universe was created in an explosion and had named it, snippily, the Big Bang theory: “an explosive creation of the Universe is not subject to analysis,” he said with some justification. Hoyle’s own theory, laid out in his book, was that the universe lives, and has always lived, in a balanced and steady state—an excellent idea, Jim thought. Hoyle went on to say that everything in the universe was related: “From the vast expanding system of galaxies down to the humblest planet, and to the creatures that may live on it, there seems to be a strongly forged chain of cause and effect.” All astronomical evidence could fit into one framework, Hoyle wrote, and that framework was the laws of physics. Since those laws were comprehensible, the universe was comprehensible too.

For a high school student, this was powerful stuff. The laws of physics are a set of rules you can believe, you can build on, and they hold for every place in the entire universe. And all the pieces of the universe—the earth, the planets, the sun, the lives of all kinds of stars, the galaxies, the universe’s expansion—were intimately related, and so they should click together into one coherent picture. But Hoyle warned that lovely as this picture is, we shouldn’t believe it until we have tested it in every way. The picture, the theory, must be exposed to observational attack from every direction and still endure. Only then, wrote Hoyle, would we be in a position to obtain a complete understanding of the universe as a single, interlocking thing. “The stakes are high,” he wrote, “and win or lose, are worth playing for.” Jim decided to be an astronomer.

For college, Jim stayed in Texas and went to Rice University in Houston, which was the best university in that part of the world. He’d been accepted at other schools, but Rice charged no tuition for local students and Jim’s family—his mother had remarried—was not well-off. To cover living expenses, he won scholarships and achievement prizes. Rice didn’t offer a major in astronomy, but Jim decided that was fine: instead, he’d major in astronomy’s foundations, math and physics. Besides, he thought, while astronomy is easy to learn, physics is hard and you can’t pick it up on your own. As it turned out, he found he was flat-out good at it. But he hadn’t stopped reading astronomy books, and the summer before he went to Rice, he’d built a little 8-inch reflector telescope for which he’d ground and polished the mirror; while he was at Rice, he added a motor to drive the telescope, then a camera to take pictures of what the telescope saw. He finished it during his senior year and wrote it all up for the observers’ column in Sky & Telescope.

In 1961, the year he graduated summa cum laude, Time ran a feature called “Top of the Heap” about some “extraordinary” college graduates, and Jim was one of them. Jim studied ten hours a day, Time said, and his only extracurricular activity was astronomy club. His physics professor was quoted, saying, “I’ve never been able really to determine the limits of his ability. I’ve never been able to ask him an exam question that he can’t give a perfect answer to.” Because scientists have to get doctoral degrees—the philosophiæ doctor, the PhD—and because Jim was now fascinated by physics, he applied to and was accepted at the astronomy department at the California Institute of Technology to study Einstein’s general relativity, the physics that was the basis of all studies of the universe.

Caltech looks like a garden—jewel-green grass in brown, dry Southern California, flowers and flowering trees, winding paths, little fountains and pools and waterfalls—set among pale stuccoed buildings with red tile roofs, connected with arcades. It’s light and graceful and silent. At Caltech, Jim felt that he’d landed in the midst of an intense community of scholars who knew the answers to his questions, with whom he had no trouble being heard, who treated him as one of their kind. He felt almost light-headed with joy.

Happily, Caltech was also rich. It owned several telescopes, which were situated about 100 miles south of campus on Palomar Mountain, and one of them, the 200-inch Hale Telescope, was the biggest in the world. With a bigger telescope, you can see things that are fainter and farther away; the farther away, the farther back in time. The early universe had for years been the province of theorists only; until recently nobody had actually seen it. When Jim got to Caltech, Caltech’s astronomical observers had just found the earliest and most distant things anyone had ever seen.

These distant things happened to be sources of intense radio waves—Caltech also owned a radio telescope. And because astronomers have a human prejudice toward the optical wavelengths in which they see, Caltech observers had double-checked the radio sources with optical telescopes. The radio sources were outright odd. They looked like pinpricks the way stars do, but the other information about them had been unintelligible. A Caltech observer named Maarten Schmidt had just figured out that the radio sources made sense only if these starlike things were at highly unstarlike distances.

An object’s distance can be read from its velocity—how fast it’s moving with the universal expansion—and its velocity can be read from its spectrum, its light spread by a prism into a rainbow. Superimposed on the rainbow are specific features—they can look like the lines of a barcode and are caused by the behavior of the galaxy’s atoms—at specific colors, that is, specific wavelengths. When the galaxy is moving away, its light takes longer and longer to get to us, and these features in the spectrum shift down toward longer, redder wavelengths; astronomers call this redshift. The larger a galaxy’s redshift, the faster it’s moving away from us and the more distant it is. Schmidt had already found a regular galaxy with a redshift of 0.46, meaning that it was about 5 billion light-years away.

Then Schmidt noticed that one of these radio sources, called 3C 273—with a redshift of 0.16, only about 2 billion light-years away—was enormously brighter than the redshift 0.46 galaxy. So Schmidt and his colleagues looked at the other odd radio sources and found more single starlike things, each forty times brighter than the biggest galaxies full of 10 billion stars. They named the starlike things quasi-stellar radio sources; later they shortened the name to quasars. In the next few years, they found a handful of quasars, each one farther away than the one before. No one had any idea what quasars were or what made them shine so brightly.

In 1965, when Jim was a fourth-year graduate student and hardly aware of quasars, he went to a talk Schmidt gave. Schmidt began with his first quasar and walked the audience through the rest one by one: 3C 254, redshift 0.73; 3C 245, redshift 1.02; 3C 287, redshift 1.05; 3C 9, redshift 2.01—10 billion light-years away, two-thirds of the way back to the beginning of the universe. These redshifts, these distances, these ages, were unheard of. Schmidt himself, who wasn’t yet forty years old, had only dreamed of getting such redshifts by the end of his career. Jim thought it was the most exciting single astronomy talk he’d heard.

Jim had gone to the talk with a fellow graduate student, Bruce Peterson, and sitting there listening, they noticed an anomaly in the spectrum of the quasar 3C 9. 3C 9 was in the early universe; the universe between 3C 9 and Caltech was full of hydrogen atoms, and hydrogen atoms absorb light at certain ultraviolet wavelengths. At those wavelengths, then, 3C 9’s spectrum should have been flat and dark, but it wasn’t; it was full of ultraviolet light. That light would be there, Jim and Bruce thought, only if something had changed all those hydrogen atoms—if their electrons had been ripped off, or ionized, by something violently energetic or hot. If so, that violence must have been coming from something like giant stars lighting up or quasars, whatever they were, doing whatever they did.

The dark place in the spectrum that should have been there became known as the Gunn-Peterson trough or the Gunn-Peterson effect. Finding an object whose spectrum showed the trough would mean finding an object from the time when the hydrogen atoms were still intact, before the universe created the first energetic, shining objects, when the universe was an infant. But the trough wasn’t there in 3C 9 at redshift 2.01, and no one knew how much farther back you’d have to go to see it—certainly farther than their instruments could now see. Gunn and Peterson published their idea in 1965 and because it was so clever, they became mildly famous for it. Jim was happy to have a trough named after him and wished someone would find it, but in the meantime he went on to other things until the day when instruments improved.

After Jim graduated from Caltech in 1966, he spent the next few years in the army, building instruments at NASA’s Jet Propulsion Lab—Caltech’s astronomers had connections that could keep their star graduate students around telescopes—and then took a position as an assistant professor at Princeton. For a while, he worked with Princeton astronomer Jeremiah Ostriker on a theory about an unusual kind of regularly flashing star just discovered, called a pulsar. But Princeton didn’t have a telescope, and Jim felt he needed one, so in 1970 he returned to Caltech.

For the next ten years he stayed at Caltech, working meticulously and painstakingly on one research project after another, heading off on one subject, then digressing onto another—usually without giving up the first—and then digressing yet again. His style of research was uncommon. Most astronomers stake out claims on certain kinds of objects or certain parts of the universe: for example, Maarten Schmidt had spent much of his career studying quasars. But Jim worked on different populations of stars; the evolution of stars; the Milky Way; gravitational lenses; the particles of dark matter; binary stars; local galaxies; rare and peculiar stars; globular clusters; supernovae; quasars; and clusters of galaxies. Some of these subjects are related—classifying stars into populations can mean tracking the early, middle, and late stages of their evolution—but most are not. And all the while he was inventing and building cameras and spectrographs. He obviously had an enormous intellectual range and apparently the attention span of a housefly.

In fact, he seemed to have taken Hoyle’s interconnected universe to heart, studying each of the universe’s pieces to see if somehow they’d click together into one whole history. Toward that end, he began two long-term surveys of the universe’s opposite ends. Maybe if you could connect the near and far, the present and past ends of the universe, you could see how it changed with time and you could figure out its history. He started with the far end, the earliest things seen, the quasars.

Maarten Schmidt had been racking up higher and higher numbers of quasars, and the highest numbers seemed to be at the greatest distances. Other quasar hunters had joined him and found the same. Then, somewhere around redshift 3.0 or 4.0, the numbers dropped dramatically; quasar hunters called the drop the quasar cut-off. It seemed to imply that somewhere out beyond redshift 3.0, at even higher redshifts even earlier in the universe, quasars first appeared.

Before believing the cut-off, though, you’d want to expose it to Hoyle’s observational attack and systematically count quasars at increasing redshifts. The quasar hunters had collected a few hundred quasars, but believability required better statistics, and those wouldn’t be achieved by collecting quasars one by one. One night in 1977, Maarten Schmidt and Jim were having dinner at Palomar, and Schmidt asked Jim whether the new CCD camera Jim had just built, a prototype for WFPC, could be rigged to do a wholesale spectroscopic survey of quasars.

Within a year of dinner with Schmidt, Jim had built a CCD camera that was sensitive to the faintest light. He called it the Prime Focus Universal Extragalactic Instrument, or PFUEI (pronounced the way it looks, “phooey”). PFUEI was mounted at prime focus, near the top of the telescope, and to use it Jim sat in a little cage next to it, 100 feet off the ground. Sitting up there was cold but glorious, and just that 100 feet closer to the stars. He’d put PFUEI on the Palomar 200-inch, point it at the sky, lock it in place, and let the earth’s rotation move the sky past the CCDs—a maneuver called drift scanning. In the early 1970s, getting the spectrum and therefore the redshift of one quasar the traditional way could take most of the night. In the same amount of time, PFUEI could find the redshifts of twenty to thirty quasars.

Jim, Schmidt, and Don Schneider, who had been Jim’s graduate student and was then Schmidt’s postdoc, didn’t actually begin the survey with PFUEI until the early 1980s. They found no quasars with redshifts over 3.0, and they had expected to find many—an interesting contradiction, but as scientists like to say, absence of evidence is not evidence of absence, so they had no proof of a cut-off. In the mid 1980s they expanded the survey, this time with a second camera Jim built with four CCDs arranged in a square, which Jim named Four-Shooter. It could see much more sky and was much more sensitive. Over the next few years, the Four-Shooter survey found one hundred high-redshift quasars, including a record setter at redshift 4.04, then one at 4.73, then another at 4.9. Schmidt, Schneider, and Gunn considered they had a reasonable suggestion of a cut-off and published: beginning nearby, the numbers of quasars rose until, around redshift 3.0, only 2 billion years after the Big Bang, they peaked, a hundred times more quasars than now. Even farther back than that, their numbers seemed to fall rapidly, or at least get harder to see and count, but in any case, quasars turned on no later than redshift 4.9. They still hadn’t seen the Gunn-Peterson trough; if it existed, it was farther back than the farthest quasar.

During much of the time Jim was surveying quasars in the distant universe, he was also working with another Caltech colleague, John Beverly Oke, on the other survey, the one of the nearby universe. They were looking specifically for local evidence of cosmology’s two numbers, the universe’s density and its rate of expansion. The exact rate at which the universe expands had been cosmologists’ problem child since 1929, when Edwin Hubble measured the distances of some nearby galaxies—he called them nebulae—and found that the farther the galaxies, the faster they were moving away from us, as though they were painted on the skin of an inflating balloon, as though they were riding expanding space. The expansion rate became known as the Hubble constant; Hubble thought it was 500 kilometers per second every 3.26 million light-years, that is, every megaparsec.

Ever since, observational astronomers had been measuring and remeasuring the expansion and disagreeing wildly with one another’s numbers. One camp said the constant was 50, the other said it was 100, and Jim didn’t believe either one. In any case, he was more interested in the universe’s density, in whether the amount of its gravitating matter could slow the expansion enough that the universe would collapse in on itself, or whether it would just coast on, expanding forever.

Back in 1974, Jim and three other young cosmologists—Richard Gott of Princeton, David Schramm of the University of Chicago, and Beatrice Tinsley of the University of Texas—had put together all the physics, observations, interpretations, and arguments about the known state of the universe, assessed them all for believability, and in a scholarly paper called “An Unbound Universe?” announced that the universe, though slowing, would never stop expanding. Their paper became known for being succinct, exhaustive, and brave, and it was eventually referred to just by names of the authors: Gott, Gunn, Schramm, and Tinsley. The question was, were they right?

At the time, astronomers were mostly measuring expansion, and to do that, they balanced velocity—that is, redshift—and distance. Redshifts could by now be measured with fair precision, repeatability, and lack of controversy. Measuring distances, however, was a hornet’s nest and the reason the Hubble constant fight was so bitter. Distance can be extrapolated from brightness: the more distant an object is, the dimmer it appears. This measure works only with certain families of objects whose brightness is so standardized that astronomers call them standard candles. Brightness falls off proportional to the distance squared: so if standard candle 1 is a hundredth as bright as standard candle 2, then standard candle 1 is reliably ten times more distant.

For years, astronomers tried one standard candle after another and continued to disagree wildly. Their measuring errors were large, between 10 and 30 percent. The candles that were reliably standard were too nearby to say much about the expansion of the entire universe, and the ones far enough away couldn’t be trusted to be standard.

The standard candle Jim wanted to work on, and the one most popular in the early 1970s, was called the brightest cluster galaxy. In clusters of galaxies, the galaxy in the center is usually the biggest and brightest. Astronomers called the big, bright, central galaxy a brightest cluster galaxy, or just a BCG. The ones nearby all had about the same absolute brightness and should make good standard candles. So when Jim and Bev Oke began measuring the universe’s expansion, they surveyed first for clusters, and then within the clusters for the BCGs. By the mid-1970s, they’d found enough BCGs to estimate that the universe’s expansion was in the same range as Gott, Gunn, Schramm, and Tinsley had predicted.

But Jim didn’t trust the results. Beatrice Tinsley had just given a talk at Princeton saying that BCGs were unreliable: while nearby BCGs were one kind of standard candle, the distant ones almost certainly had to be another kind. Galaxies are made of stars. Stars evolve; as they age, they change color and brightness. Obviously galaxies must evolve too, changing color and brightness, and obviously they’re useless as standard candles. Jim was at Tinsley’s talk with two other cosmologists, Jeremiah Ostriker (with whom Jim had worked on pulsars) and Scott Tremaine, who were now inspired to make the uncertainty even worse.

Ostriker and Tremaine calculated the paths of all the galaxies within a given cluster. Inside clusters, galaxies buzz around like flies, colliding, merging, generally swapping around stars. At the same time, they are orbiting the BCG in the center, gradually slowing, spiraling in until the BCG captures them, one after another. So not only are the stars in the cluster galaxies evolving, the cluster galaxies themselves are merging with one another and with the BCGs, changing the BCGs’ brightness. BCGs once and for all lost the case for being standard candles. Jim got disenchanted. He’d thought the problem would be simple, or at least straightforward, and here was Nature one-upping him, saying prissily, “Everything is a great deal more complicated than you think.”

But Jim and Bev Oke had this perfectly good cluster survey, and they didn’t want to give up on it. If BCGs in the clusters couldn’t be used to find the universe’s fate, maybe the whole clusters themselves could. Cosmology’s two numbers, expansion and density, are linked—high density always means slowed expansion, and vice versa—and by measuring one of them, you get the other for free. So forget expansion and standard candles; try the other number, try density and gravity instead.

Clusters of galaxies are the largest objects in the universe that are held together by gravity, and the amount of time gravity would take to pull them together would depend on the overall density of the universe. In a high-density universe, clusters trying to pull themselves together will have more competition from surrounding matter and will take longer to form. So at a higher redshift—the earlier, more distant part of the universe—clusters should be fewer in number. Jim and Oke and a variety of collaborators decided to continue their cluster survey, counting the number of clusters at greater and greater distances. By the mid-1980s, they’d found enough clusters between redshifts 0.5 and 1.0 that they could again support Gott, Gunn, Schramm, and Tinsley’s vote for universe with low density and slow and infinite expansion.

And again, Jim couldn’t quite trust the observations, and again, the reason was that no one knew quite how cluster galaxies changed with time. They were counting clusters down to some minimum brightness. But maybe the galaxies in the clusters had been brighter in the past; if so, they should be able to see more of them farther back. And high numbers of clusters in the early universe wouldn’t reliably imply the universe’s density after all.

The one outcome of the cluster survey that Jim did trust implied the general course of the universe’s history. Jim and collaborator Alan Dressler studied several clusters in detail, comparing the younger, distant ones with older, nearby ones. The farthest clusters, at redshift 0.5, seemed to have a high fraction of galaxies that were either bursting with newly forming stars or clearly just poststarburst. The nearest clusters, redshifts of 0.02 or so, seemed full of galaxies that were dead, whose stars had gone from infant blue through adult yellow to old-age red. The universe seemed to be turning from hot, young, and blue to cool, old, and red—an unverified and broad-brush history, but a history nonetheless. The universe had been a much livelier place when it was young.

About now Jim began to feel that he was spinning his wheels. He’d done an extraordinary amount of work on instruments, theory, and observations in an extraordinary number of fields. But he worried that he’d begun operating on inertia. He was in a complex state of mind. The quasar and cluster surveys had been important, and his colleagues were right in continuing them, but in his own mind, the surveys had gone about as far as PFUEI and Four-Shooter could take them, and further observations were hitting diminishing returns. Both surveys covered a few square degrees—a few Moons’ worth—and had a few hundred objects chosen because they were bright enough to be seen, and neither the smallish area nor the fewish objects nor the bias toward brightness inspired confidence. Cosmologists should always doubt their observations, Jim thought, and the surveys’ uncertainties were no worse than any other astronomer’s observations. But even so, too much of what he’d done didn’t appear to him to be secure enough to publish, let alone to believe.

In fact, too much of late-twentieth-century cosmology was still not quite believable, more theory than observation. Observers did the best they could, but the observations themselves were time-consuming, finicky, and had measuring errors that could be the size of the measurements themselves. An eminent observer, Vera Rubin of the Carnegie Observatories, said that observers needed to be incredibly devious and take ten objects and infer the rest, but for her to believe anything about the arrangement of galaxies on the large scale, she said, she’d want the redshifts of a million galaxies; the number known at the time was just over ten thousand. An equally eminent theorist, James Peebles of Princeton, said that when his latest theory of how galaxies formed became accepted, he got nervous: “I could think of lots of other ways to make galaxies,” he said. Jim told the writer Alan Lightman in 1988, “Cosmological observations are always right at the hairy edge of the possible.” As a result, he said, observers overinterpreted their observations and theorists overinterpreted them even more: “The correlation of what is really true about the universe and the set of notions that we think are true about this universe I think is not very high at the moment.” Peebles said, “There was a lot to do in the late eighties.”

Jim thought they had to start with galaxies. His surveys had made it clear that if you wanted to understand the universe, you had to understand galaxies. Galaxies are the basic units of the universe; all the stars are in galaxies. At the far end of the universe, quasars—astronomers now knew—were things that happened in galaxies, and to understand quasars, you had to understand galaxies. And at our near end of the universe, to understand clusters, you had to understand galaxies. You couldn’t find the expansion, density, or fate of the universe and you couldn’t map its history at all without understanding galaxies. And so little was known about them. Galaxies seemed to come in a whole zoo: blue, red, faint, bright, spiral, elliptical; galaxies that looked like messes, or that were forming stars actively, or that were just sitting there; massive galaxies, wispy little ones, galaxies living in clusters or scattered loosely around in empty fields.

In particular, Jim thought, cosmologists’ affection for the distant universe meant that no one had a reliable census of the galaxies nearby. How could you build a history if you studied only the past and never found out what the present looks like?

In fact, if the universe was a single thing whose parts were all related and which evolved with time, then cosmologists weren’t going to see it. Cosmologists had, from technological necessity, broken the universe into different, apparently unrelated problems: expansion and density and fate, quasars and the earliest universe, galaxy clusters and galaxies. Jim’s career so far was a one-man incarnation of how the whole field operated: pick a piece of the universe, study it until you hit diminishing returns, then put it down and pick up another one. The thing itself, the coherent universe, seemed to be lying in fragments all over the place.

Jim didn’t think he’d been wasting his time, but he suspected he was casting about, looking for something to get passionate about, something with enough intellectual heft to be irresistible.

In early 1987, Kitt Peak National Observatory had a mirror it had commissioned as an experiment and now didn’t know what to do with. To brainstorm options, it held a meeting in Tucson on February 18, 19, and 20, 1987, and Jim went. On the first day, Jim gave a talk in which he got on his high horse about understanding the universe from the galaxies up and said that the most exciting use for a telescope with the Kitt Peak mirror was something that had been “simply impossible before, namely the direct study of the evolution of galaxies.”

Then other astronomers gave talks outlining other ideas, and while Jim sat there listening, he stopped thinking about galaxies and thought about technology instead. The Kitt Peak mirror might be a good match for the newest CCDs. Just recently Jim had been at Caltech and run into an engineer named Morley Blouke who had opened a polyethylene pizza box and showed Jim the latest CCD. WFPC’s state-of-the-art CCDs, which had been made by Blouke’s company, were 800 pixels on a side, 6,400 pixels altogether, meaning a field of view of 10 arc minutes, a third of a moon, across. The CCD in the pizza box had 2,048 pixels on a side, 4 million pixels altogether. Looking at those enormous CCDs, Jim felt they were the wave of some future he wanted to be part of.

Jim, listening in Tucson, thought he could arrange the new CCDs into a mosaic and build a camera with a field of view of 120 arc minutes, that is, 2 degrees or four moons across. Moreover, Jim thought, he could use the same CCDs in a spectrograph, and if the CCDs’ pixels were fed by the new optical fibers that happened to be a good match to the size of the individual pixel, the spectrograph could take the spectra of hundreds of galaxies at once. The spectrum of a galaxy shows much more than its redshift/velocity/distance. The amount of light an object has at the different wavelengths also shows its inner workings, its physics: what elements it’s made of, how fast it’s rotating, how hot it is, and how old it is. Images and spectra together tell you everything about an object that, from a distance, is possible to know.

He could put this camera and this spectrograph on a telescope with the extra Kitt Peak mirror and let the telescope drift scan along one strip of the sky after another. All those strips could be stitched together into a survey, which if run for five years would give a highly informative three-dimensional picture of a large fraction of the nearby universe. It could take the spectra of hundreds of thousands of galaxies—the same survey with current spectrographs would take several thousand years—and the images of tens of millions of galaxies. It would specify and track the whole galaxy zoo. It would be digital: images and spectra of wide fields of the sky, debugged and cleaned up and analyzed and dumped into your computer.

Jim talked about his idea in the meeting’s hallways and over dinner, and by the end of the meeting, everyone had discussed it. They saw no big technical problems with it and in fact seemed excited. When asked whether they’d be willing to work on a proposal to build what they were now calling the digital telescope, three-quarters of them raised their hands.

After the meeting, Jim worked out that the mirror that most suited the size of the CCDs and the size of the optical fibers was smaller than Kitt Peak’s. A little 2.5-meter mirror, he said, would be about right. Then he went back to the Hubble Space Telescope and WFPC and the quasar and cluster surveys and dropped the subject. The talks given at the Tucson meeting were written up and printed out and bound with a nice plastic spiral, and then everybody else dropped the subject too.

© 2010 Ann Finkbeiner