First Light: The Search for the Edge of the Universe

by Richard Preston

  “Oh, some stars,” he said.

  One by one the astronomers and engineers sat in the lift chair and went up to the eyepiece. When Byron Hill got his chance to look, as he would remember, “I had never seen so many stars in my life. It was like pollen on a fish pond.” The sight, he said, “made me feel pretty good.”

  They all knew that much tweaking and polishing still remained to be done. First light on a large telescope is the beginning of a process of adjustment that may continue for years. Although glass is brittle, it is actually a supercooled liquid. Glass is physically similar to Jell-O. Glass can flop, tremble, and shudder. As a large mirror moves through varying angles, it buckles and droops. The Hale Telescope’s mirror is rubbery. You could push down firmly on it with your thumb and throw the stars out of focus.

  Today, pressure pads controlled by computers push and warp large telescope mirrors to keep them in shape. When Hale first proposed a two-hundred-inch mirror, he sensed that the problem of supporting a lake of glass to a tolerance of four millionths of an inch across two hundred and nine square feet of surface area might be impossible to achieve with existing technology. He decided to hope that the technology would come along. In the early 1930s, an engineering team designed and built thirty-six mirror-support machines, weighted with lead. When the glass disk arrived in the Caltech optical shop, the machines were plugged into pockets in the back of the disk. An engineer named Bruce Rule then tested the glass for signs of slumpage and found that the glass behaved somewhat in the manner of uncured latex rubber—when the opticians leaned the mirror at an angle, the glass would droop and not return to normal shape for quite a while. The mirror-support machines were failing to compensate for slumpage in the glass. During the summer of 1948—six months after first light—Bruce Rule extracted the mirror-support machines from their pockets in the glass and rebuilt the machines. Rule’s thirty-six mirror-support machines work passively, by means of levers and lead weights. The levers barely move, yet they exert three-dimensional forces throughout the glass, which, in places, reach stresses of up to twelve hundred pounds.

  Bruce Rule was a tall, white-haired man who wore thick glasses and spoke in a soft, measured voice, and who was widely regarded around Caltech as a genius, which is a reputation not easy to get in a place like Caltech, where the geniuses do not generally refer to each other as such. I visited Bruce Rule one day at his home in Pasadena. “I wouldn’t call them machines,” Rule said. “I would call them compound support units.” Each unit, which resembles a piston inserted in the glass, contains an uncounted number of parts. Rule said, “I think that between six hundred and one thousand parts in each unit is a reasonable number.” Since there are thirty-six mirror-support units, that would mean that the Hale mirror is held up by as many as thirty-six thousand pieces of metal, most of which move, if only slightly. Now we see why Bruce Rule was considered a genius. Rule said, “That estimate depends on how you want to count parts. If you want to count all the little parts inside ball bearings, then the number would be larger.” The support units are, in fact, mechanical computers. They react to forces in the mirror and apply corrective action. Rule said, “I never recommended that this type of system be tried again.” Virtually everybody at Caltech understands electronic computers, but nobody at Caltech understands mechanical computers, and consequently nobody dares to monkey with Bruce Rule’s support units. Since 1948, there has been one attempt to oil them. It was not much of a success. The lead weights on the units are adjustable, but nobody wants to adjust them. Once or twice a year an engineer walks around the cage at the base of the telescope and reaches up inside the mirror cell. He takes hold of the weights and wiggles each in turn, in order to give the units a bit of exercise; but the feeling around Caltech is that anybody who tries to open Rule’s units to see what is inside them will get himself fired. Rule did not worry about his units. “We didn’t give ninety-day guarantees,” he said. “We built for life.”

  Once in a while these days, the stars on the video screen turn into hollow triangles—the support units have become stuck. The astronomer turns to Juan Carrasco and says, “The mirror needs exercise.” Juan then slews the telescope from horizon to horizon, from north to south, from east to west, until the stars turn back into points. The nightmare of the engineers who take care of the Hale Telescope is that one night the stars will turn into triangles, Juan will exercise the mirror, and the triangles will get bigger. In that event, the engineers would have to search the Caltech archives for microfilm of Rule’s blueprints for the support units, although no Caltecker is sure that he would understand the blueprints. During the summer of 1948, when he was designing the units, Bruce Rule liked to go to the beach for a weekend, where he would lie on the sand and hear the surf and see shapes in his mind’s eye—levers and pistons and rippling glass. “I could keep a crew of thirty draftsmen going,” Rule said.

  About the time that the mirror went into the telescope, the chief optician, Marcus Brown, retired. An astronomer named Ira Bowen was named director of the observatory, and Bowen took personal control over what is known as the final figuring of the mirror. In the spring of 1949, the opticians stripped the aluminum from the glass and went to work on the glass with small polishing tools. Under Ira Bowen’s gaze an optician named Don Hendrix did much of the polishing; Melvin Johnson assisted Hendrix. They would mount the glass in the telescope. Bowen would look at a bright star, which he could see reflected in the bare glass, and take measurements of the star, while Hendrix or Johnson would outline any defective zones in the glass, using a grease pencil. These measurements would require from one to three nights to complete. Finally, at dawn, they would remove the glass from the telescope and rest it on a cradle. Hendrix and Johnson would rub the glass in one or two spots, using their polishing tools. Hendrix’s favorite tool was an aluminum disk faced with black pitch, the size of a Thin Mint. Sometimes they used a piece of cork. The smallest defective zones were an inch or two across and circled with a grease pencil. To polish these zones Mel Johnson would dip a watercolor brush in water mixed with a silt of polishing compound called Barnesite. He would paint the zone with Barnesite, then rub the zone with his thumb. “There are no sharp edges on your thumb,” he said. “Your thumb flows into the zone.” Johnson liked to use his thumb because he could feel the temperature of the glass change as he rubbed it. Each stroke removed about two hundredths of a millionth of an inch of glass, but heat from the rubbing swelled the glass by more than that. They would polish a little here, a little there, until they sensed that they had swelled the glass. Then they had to let the entire two-hundred-inch glass cool for the rest of the day, to let the swelling go down, before they could see what they had done to it. They would mount the glass in the telescope and test it on a star. This process continued throughout the summer and fall of 1949. “All we were asking for,” Johnson said, “was a hard point of light.” In the end they polished the mirror to a mathematical formality. If the Hale mirror were expanded to the size of the United States, it would exhibit no hill higher than four inches. That is not counting pits left by bubbles in the glass, which the opticians plugged with pitch.

  Bowen’s final tests revealed an astigmatism in the glass—the mirror was slightly warped. The opticians could have polished the glass for another three years, but they solved the problem with a kludge. They purchased four fisherman’s scales at a dime store and hooked the scales to the back of the glass, where their springs each tugged at the mirror with about seven ounces of pull, just enough to open the mirror by a few millionths of an inch and flatten the warp. In 1981, a Caltech engineer, poking around nooks in the back of the mirror cell, asked himself what in the world these fisherman’s scales were doing in there, and removed them. When the astronomers complained that the telescope had gone out of focus, the fisherman’s scales went back on the mirror in a hurry.

  There is a saying among those who polish astronomical mirrors for a living that an optician never finishes a mirror—you take it away fr
om him. By the autumn of 1949, Hendrix and Johnson were still rubbing the glass with pitch, cork, and thumbs. They kept telling the astronomers, “We’d like another week.” That would provoke a stormy meeting behind closed doors, because the astronomers were hot to use the telescope. “Ira Bowen just kicked all the astronomers in the teeth,” Byron Hill said. Mel Johnson said that, for his part, he could have lived happily with that glass for another two years. The astronomers began to go wild. Bowen kicked them in the teeth again, and they came back like dogs for more. Bowen finally gave up. He took the mirror away from Hendrix and Johnson—he declared the mirror finished. He ordered Hendrix to put a coating of aluminum on the mirror and to put it in the telescope. In November 1949, the Hale Telescope went into regular use.

  George Ellery Hale never knew that his telescope would be called the Hale Telescope, for the naming of the instrument did not occur until a dedication ceremony in the dome on June 3, 1948, more than ten years after Hale had died. James R. Page, chairman of Caltech’s board of trustees, opened the dedication with a speech in which he said, “This telescope is the lengthened shadow of man at his best.” One wonders what the elf thought of that remark. Later Bruce Rule slewed the telescope back and forth over the heads of a mass of spectators, while he prayed to heaven that the telescope would not drop any nuts or oil into the crowd.

  Every six months Palomar engineers remove the mirror from the telescope and wash it with natural sponges, using Procter & Gamble Orvus soap, which is Ivory Soap without fragrance. Once a year the engineers strip the aluminum off the glass. The superintendent of the observatory is a stocky man with a mustache, named Robert Thicksten. After the aluminum is stripped off, Thicksten stands on a little platform in the hole at the center of the glass and inspects the bare glass. The glass, Thicksten said, reminds him of a jewel. George McCauley’s flame Pyrex holds lapidary colors that change under different lighting. Sometimes the glass seems to be topaz yellow, or pale green, or amber, like maple syrup. Under bright lights, the glass discloses a ring of blue brilliant haze, caused by an unknown contaminant in the Pyrex, encircling the hole in the center, as if the glass had a blue iris. From above, the waffle structure in the back of the glass is clearly visible—a network of triangles and hexagons that gives the glass the frightening appearance of an insectiform eye. Backlighting reveals dark shapes caught in the glass—chunks of firebrick that had broken away from the oven and dropped into the melt. The glass sparkles with silvery air bubbles and is swirled with liquescent folds, eddies, and stria, which, in one area, erupt at the surface of the glass into a spiderweb of cracks. The cracks hold pockets of red jeweler’s rouge that worked its way downward into the glass. The opticians had drilled small holes at the ends of the cracks, to arrest the cleavage through the glass, and plugged the holes with pitch. The cleverest human craft had barely harnessed the natural world’s random currents in the two-hundred-inch mirror, a surface for gathering light, intermediate between land and sky, that had risen in a dream before the eyes of a frail dreamer, and which the work of almost anonymous human hands had made into a physical thing. The inspection complete, Bob Thicksten and his engineers clean the glass with solvents and put it in a vacuum tank and vaporize aluminum over it, which turns it back into a mirror. Thicksten has elected to stop massaging the glass with Wildroot Cream. “That was a black art,” Thicksten believes.

  Virtually all of the Hale’s builders are gone. John Anderson died of a heart attack in 1959. George McCauley, Russell Porter, Marcus Brown, Ira Bowen, and Don Hendrix are dead. Over the years the collaboration that George Ellery Hale had negotiated between the Mount Wilson Observatory and Caltech became more delicate, more formal, more punctilious, and exploded in a cat-fight—the sort of thing that happens all the time around telescopes. In 1979, Caltech assumed administrative and financial control over the telescopes on Palomar Mountain, including the Hale. The Hale glided through the divorce. It had been so beautifully engineered that it seemed to run according to a stubborn will of its own, aloof from human frailty, but if it ever broke down in a major, unforeseen way, Bob Thicksten believed that about six people on earth would know how to repair it. Or might know. On summer nights Thicksten would stand on the catwalk of the dome, listening to the nocturnal sounds of the Big Eye, asking himself if its gears were humming on the right note. “We know when certain things work,” Thicksten once remarked to me. “But the worry is, we don’t know how they work.” The Big Eye had outlived its creators.

  The Hale’s builders equipped it with a number of smaller mirrors so that light can be directed from the main mirror to various observing stations. One station is a small room at the top of the telescope called the prime focus cage. An observer can sit in prime focus and look through an eyepiece directly down into the main mirror, where the observer sees a reflection of the deep. When I consider examples of the power of the Hale Telescope, what comes to mind is a story that Don Schneider told me about something that happened to him one night when he was working in prime focus. Near dawn he had a few minutes of extra time on his hands. He had never seen Venus through the Big Eye. “Point me to Venus,” he said to Juan Carrasco over the intercom. The prime focus cage tilted as Juan slewed the telescope down and east, toward the horizon, where Venus was. “We are there,” Juan said. “You have two hundred inches on Venus.” Don started to look into the eyepiece. A stab of pain hit him in the eye. He pulled his head back, and a pencil-thin shaft of white light came out of the eyepiece. The light was too bright to look into, and it reminded him of a movie projector beam. It was the light of Venus that had fallen on 209 square feet of mirror and had been condensed into the eyepiece. He could see dust motes dancing in the light of the morning star.

  George Ellery Hale’s greatest telescope is a time machine. It reimages lost time. Light from the sun takes eight minutes to reach the earth. The sun is eight minutes into lookback time. Photons from Venus take anywhere from two minutes to fourteen minutes to reach the earth, depending on where Venus is in its orbit in relation to the earth. The planet Saturn is one light-hour away. Proxima Centauri, the nearest star other than the sun, is currently about four light-years and three light-months away from the earth. (Proxima Centauri is moving, and someday it will be nowhere near us at all, because stars are lone voyagers through the galaxy.) A few dozen known stars now drift near the sun, bearing names such as Epsilon Indi, Tau Ceti, Kruger 60, Kapteyn’s star, and Procyon. Distant, giant stars—Rigel, Aldebaran, Betelgeuse, Antares—are hundreds of light-years away. The mist in the Milky Way consists of stellar images that are thousands of years into lookback time, because those stars are, on average, a few thousand light-years away from the earth. The Milky Way appears to be a band of light encircling the sky, because the Milky Way is a spiral galaxy shaped like a disk, and we are within the disk, looking outward. A spiral galaxy is a rotating cloud of matter that contains much gas and dust, a prodigious quantity of a stuff known as the dark matter, which astronomers admit they know almost nothing about, and about one hundred billion stars. A number like one hundred billion is not easy to imagine. If you placed that many ten-dollar bills end to end, they would form a line of bills that would go eight times around the earth, and then out to the moon, back to the earth, and out to the moon again. (We need to elect some astronomers to Congress.) Two square miles of growing wheat contain roughly one hundred billion grains of wheat. A star is to a galaxy what a grain of wheat is to a farm in Kansas.

  Consider a sun the size of the dot over this i. On that scale the earth would be the size of a one-celled microorganism, located about two inches away from the sun. On that same scale the nearby star Proxima Centauri would be about nine miles away—and the center of the Milky Way would be about fifty thousand miles distant. If something were to happen to the earth, it would not be missed. Man is dispensable. So is the earth. The one hundred billion stars in the Milky Way, including the sun, revolve around the Milky Way’s center, just as the earth revolves around the sun. The sun and the earth
take about 250 million years to make one orbit around the galactic center—a period of time known as one galactic year. The sun and the earth have existed for about eighteen galactic years—they have traveled about eighteen times around the galaxy since they were formed. Some kind of extremely heavy, compact object is sitting at the rotational center of our galaxy and giving off radio waves. The radio signals from the galactic core that we are now picking up began to travel toward us in about 23,000 B.C., around the time that Upper Paleolithic hunters were painting handprints on the walls of caves in the Pyrenees.

  Not too far from the Milky Way float other galaxies—the Clouds of Magellan, the Draco dwarf, the Fornax dwarf, Andromeda, the Pinwheel, the Whirlpool, Centaurus A, the Sombrero, the Zwicky Antennae, Stephan’s Quintet. Andromeda, a near neighbor, is a spiral galaxy about two million light-years away. If the Milky Way were the size of a dime, then the Andromeda galaxy would be another dime about two feet away. Certain mysterious forces, which are not well understood, cause galaxies to evolve into extraordinary shapes: barred spirals, globes, footballs, rings, fuzz balls trailing rattails, thin smooth disks, and chaotic patches. Galaxies prefer company. They like to cluster in knots. A small knot such as the Local Group contains about a dozen galaxies, most of them dwarf galaxies, such as the Clouds of Magellan. A so-called poor cluster contains about a hundred galaxies. A rich cluster contains a few thousand inters warming galaxies.