The Perfect Machine

by Ronald Florence

  Before electronic instruments took over, the original Cassegrain cage, behind the mirror, had a swinging seat for the observer. An observation session, as the telescope slewed to new positions, felt like sitting in a gimballed drink holder on a boat.* The gimballed chair was replaced in 1965 with a new cage, large enough to hold the huge new electronic detectors that had replaced photographic plates. The refrigerator-size instruments are packed with auxiliary lenses, CCDs, dispersion gratings, cooling dewars, and a mass of wiring and electronics. There was a time when astronomers studied optics and built their own instruments. Now, the “wizards” who build and maintain the instruments in the basement warren of offices in Robinson Hall practice magic that most astronomers don’t even try to understand. Fred Harris, who does much of the critical work on the CCDs, calls it a black art, part skill and experience, part luck. He eats a Mexican lunch “to steady [him]self” before he does critical work on a CCD, then does tricks with the sensitive chips that match Walter Baade or Milt Humason’s tricks with their photographic plates. Connections have to be soldered to connectors thousandths of an inch apart. Extreme vacuums and baths in oxygen are the equivalents of the plate-hypering an early generation used to increase the sensitivity of their photographic emulsions. Harris is the first to admit that he sometimes doesn’t know why the complex detectors choose or refuse to work. The early CCD detectors were mazes of wires and tubes for vacuum pumps, held together with miles of “Palomar glue,” duct and strapping tape. After years of refinements, the instruments are reliable. The ones not in use wait in individual pens at the perimeter of the dome, labeled with names like “the four-shooter.”

  Astronomers rarely discuss religion in the Monastery, but privately even veterans of years of observing runs on the big telescope talk of the awe-inspiring mystery of deep space observations, the inescapable feeling that every night on a big telescope is a unique voyage to an uncharted corner of the heavens. It is a magic moment when the night assistant flips switches on his console to start the telescope. From within the bowels of the observatory, a low rumble starts. There is no vibration, just the sound of the pumps for the oil bearings. The night assistant may say nothing more than, “The telescope’s ready,” but it is hard to shake off the feeling of starting on a vast voyage. Under the canopy of sky the huge dome and telescope feel tiny. It seems the height of audacity to think that this machine, five hundred tons of glass and steel, can reach out to the edge of the cosmos, that from a mountaintop on a planet circling around an ordinary star, one of billions in a not very special galaxy, we are about to reach into the secret depths of the universe.

  When the telescope is ready, the night assistant types the coordinates of the first target of the evening into his console and follows with the command “Go.” As the telescope slews across the sky, images appear on the video display in front of the observer—stars, the fuzzy glow of a distant galaxy, the clear spiral of a near galaxy, the stellar gridlock of a cluster. Some observers talk of a new age of fully digital telescopes, when the observer won’t need to go to the observatory at all and can sit comfortably at a workstation at his home institution, typing commands and waiting for sample images to appear on his display. If the observer is halfway around the globe, even the time changes of nighttime observing disappear. The wee hours at Palomar are the working day in Europe or Asia.

  Most observers today still work at the telescope. Few can resist going outside, into the dome, to watch as the telescope slews from one target to another, marveling that so huge a machine can move so smoothly. From the balcony of the dome, an observer today feels the same deceptive sensation that captivated the workmen who first tried the dome, and the visitors on dedication day. The dome motion is so eerily smooth that it seems to stand still. Only a look outside convinces a visitor that the observatory underneath is not turning.

  Today, the position of the dome is tracked with scanners that pick up barcodes placed around the dome. But the railroad carriages, the motor drives, and the ground rails are the same units that Byron Hill and his crew labored over more than a half-century ago. The phantom telescope that Sinclair Smith designed is still in place under the cabinet at the head of the stairs to the Coudé room. The tiny telescope tracks every motion of the telescope and dome. The probe that represents the telescope holds its place in the middle of the slot as the tiny dome mirrors the motions of the great overhead dome. The phantom telescope always brings a smile; it is a comprehensible machine, a far cry from the black boxes of contemporary technology.

  Maintenance has kept away the ravages of age, but the telescope has the quirks of maturity. When the mirror is realuminized, volatile elements evaporate from the lubricants in the support mechanisms, leaving the bearings stiff. Observers discover the mechanical arthritis when the images the telescope produces degrade from pinpoints to blobs of light. It took years to diagnose the problem. Now the night assistants all know the prescription for stiff joints: exercise. When the images degrade, the night assistants slew the telescope across the maximum range of its movement, forcing the support weights awake. After a few “reps,” the mechanisms limber up, and the images snap back to pinpoints.

  Much of the telescope is unchanged. The edge supports that John Anderson introduced to correct the astigmatism required no service for forty years. The spring scales that Bowen put on the back of the mirror are still there, a reminder of human limits, like the scar an Arab craftsman puts in an artistic masterpiece lest Allah be offended by a human effort at perfection. The springs came out once, when someone thought they didn’t belong. Observers immediately protested that the images of faint stars had changed from periods to commas. Bowen’s springs went back on.

  The two-hundred-inch telescope isn’t perfect. There is a residual astigmatism in the mirror in declinations above sixty-five degrees that has resisted every effort to tune the mirror supports. Despite twelve years of grinding and polishing, the mirror is flawed, with tiny fractures in the surface that couldn’t be ground or polished out. Minuscule holes have been drilled at the ends of the fractures to keep them from growing, and pitch fills the hairline cracks. A workman—no one remembers when—accidentally dropped a wrench on the mirror, dinging the surface. It hurts to see the scars, but the effect on the performance of the mirror is negligible. Starlight strikes all parts of the mirror at once. The scars cut down on the overall light-gathering ability of the mirror by a fraction of a percent, but the telescope won’t miss a distant object because of the imperfections.*

  Several times a year the mirror is removed from the telescope for a wash. The procedure has been rehearsed and practiced—the technicians did it dozens of times during the final figuring—but the great mirror is still cradled like a newborn. Hardhats are required for the initial stages, when the mirror cell is unbolted from the telescope. Once the mirror is exposed, all hardhats come off to protect the mirror. The reason for the regulation is obvious, but the bare heads still seem an act of respect as the staff surrounds the mirror to wash it with Orvus soap, a fragrance-free industrial version of Ivory soap. The final rinse is with distilled water. For periodic realuminizing, the surface is stripped and cleaned with powerful solvents. A few bottles of Wildroot Cream Oil are still on the shelf of the cabinet that holds the chemicals, a reminder of the old days.

  Astronomers, even veterans with hundreds of nights on the big telescopes, are guests at Palomar. They come up for an observing run, usually two to four nights, then go back to their home institutions. The staff, entrusted with the instruments, stays on—engineers, technicians, night assistants, groundskeepers, mechanics. There are job titles and specializations, but the mountain is a small world. Some jobs need every person on the mountain. The secretary to the superintendent sometimes tends the gift shop in the museum; she also helps wash the mirror of the two-hundred during maintenance runs.

  Byron Hill was the first superintendent of the observatory. Hill’s strong ideas didn’t always make him popular, and a few of his stern rules, like the
absolute ban of liquor on the mountain, were hard to enforce. One night assistant was famed for providing 150-proof rum to ease the frustration of cloudy nights. A groundskeeper clearing the brush around the dome of the forty-eight-inch Schmidt camera once discovered a cache of empty whiskey bottles that Humason had hidden during observing sessions. Hill was less tolerant of some foibles. Once he threw an English observer off the mountain for wearing shorts to lunch. Another observer was asked to leave the mountain when a maid reported that he had masturbated in his bed in the Monastery. Most astronomers believed Hill when he said, “We could run this mountain a hell of a lot better if the astronomers wouldn’t come up here.”

  For those who stay year round, life on the mountain isn’t easy. There’s an elementary school at the observatory, but for high school, children usually live with relatives down in Escondido. Winters are hard. Only one regular staff member tries the daily commute to the mountain.

  From the earliest days an unwritten policy at the observatory has required that the engineers, mechanics, and technicians who work at Palomar not be amateur astronomers lest they be distracted from their duties on the machines. But men and women who choose to spend their lives on remote mountaintops acquire an awe of the sky that much of America knew in a simpler era, before street lights, urban sprawl, and highways took away the dark skies and their wonders. For all the hardships and loneliness, the mountain can be compelling. When Byron Hill retired he moved to a trailer on a mountaintop near Sonora, with an unbroken view as far as Yosemite. Ben Traxler retired to a mountaintop in Northern California. The telescope makes romantics of cynics.

  On one engineering run Fred Harris, the wizard of CCDs, was at Palomar working on a device at the prime focus, which was more crowded than usual because a portion of the observer’s cage was taken up with a video display. Engineering runs are normally scheduled during the “light” of the month, when the moon is up, but Harris realized that an eclipse of the moon would darken the skies, if only briefly. He hurriedly phoned his girlfriend in San Diego, told her to drive up to the mountain, and took her up into the prime-focus cage with him. He said nothing until he had turned the telescope toward Lyra, the famous ring nebula. In an amateur telescope the ring nebula is tiny; on the two-hundred the beautiful, mysterious ring filled the screen of the display. There he asked his wife-to-be to marry him, offering her a ring like none on earth.

  The success of the two-hundred was a model: The ribbed Pyrex mirror, massive equatorial mount, horseshoe bearing, Serrurier truss, passive supports to correct the shape of the mirror, oil pressure bearings, a prime-focus cage for the observer, “fast” optics with corrective lenses to broaden the good field—all design elements that had been pathbreakers at Palomar—set a norm from which few dared depart. For more than two decades after the telescope went into operation, the few large telescope projects, like the 120-inch at the Lick Observatory and an 84-inch telescope for Kitt Peak in Arizona in 1958, borrowed wholesale from the two-hundred. The Lick telescope used the 120-inch practice disk that had been cast before the two-hundred-inch disk. George McCauley came out of retirement in Corning to supervise the casting of a disk for Kitt Peak.

  Many of the new telescopes had birthing problems. The Serrurier truss that had worked so well on the two-hundred-inch looked elegant on the longer tube of the f/5 telescope at Lick, but they were long enough to flex. Some suggested that the two-hundred-inch telescope had set an impossible standard, that a design that could only be built in the depression, when glass workers were paid fifty-four cents per hour and Westinghouse would build huge mounting structures for thirty-seven cents per pound, impeded new advances in telescope design.

  Only gradually, in the 1970s, did new ideas appear. George Ritchey’s optical design, the Ritchey-Chrétien telescope, which used a complex secondary mirror to produce a wide field at the Cassegrain focus, replaced the paraboloid mirror and correcting-lens combination of the 200-inch telescope in many new telescopes. In the Soviet Union, Russian engineers, waging another battle in the war for space that had begun with Sputnik, built a six-meter telescope in the Caucasus, with a primary mirror a full meter larger than the two-hundred-inch. The Russian design eliminated the massive equatorial mount in favor of a simple alt-azimuth mount: an oil pressure bearing turns the telescope on a table; the tube of the telescope rises and falls in a short fork. The complex motions to translate the altitude and azimuth motions to match the sidereal motion of the stars are controlled by a computer.

  By the late 1970s a new wave of telescope building was on. Four-meter telescopes went up in Arizona and Australia. Some design points from the two-hundred were inescapable. Mark Serrurier kept a list of the telescopes that used his truss in their tube designs; it is a listing of every large telescope built since the two-hundred. Designers tried lighter telescope tubes, pivoting the tube in the horseshoe bearing and dispensing with the massive yoke. Pyrex disks and support systems became simpler as glass foundries copied McCauley’s newest procedures, “sagging” mirrors from chunks of pure borosilicate glass.

  Some innovations failed. The Russian six-meter telescope has never rivaled the performance of the two-hundred. The first try at a mirror developed cracks on the grinding machine in Leningrad (St. Petersburg). A second mirror, designed as a monolithic mass, was poorly figured and subject to thermal shocks. The relatively slow f/4 optics required a long tube and a big dome with poor thermal characteristics. The location in the Caucasus had few nights of good weather or seeing. Even with a third mirror, figured to higher standards than the earlier efforts, the telescope is not a productive instrument.

  It was only in the 1980s and 1990s that new telescopes came along with designs that finally broke away from the standards set by the two-hundred. Ideas that had been considered and set aside fifty years before reappeared in new guises. Corning began producing mirror blanks from fused quartz, using high-temperature casting techniques that finally solved the problems that had defeated GE. Designs appeared with super thin meniscus mirrors, supported by computer-controlled plungers; George Hale had once hoped to do the same thing with a cushion of air or water. Honeycomb mirrors, like those Ritchey had tried to fabricate in France, provided light weight and quick thermal response. Mirrors were cast in spinning ovens, resulting in a dished shape that cuts the time needed for rough-grinding of the mirror to a fraction of the years required on the two-hundred. Hale had asked McCauley to do the same trick.

  At the new Keck Observatory in Hawaii, a telescope with a primary mirror twice the diameter of the two-hundred is now in service. The mirror is made of individual segments, each constantly reshaped by computer-controlled supports. On some telescopes, including the venerable sixty-inch reflector on Mount Wilson, experiments are under way with “adaptive optics,” computer-controlled optics that use continuous measurements of artificial stars created with a laser to adapt the optics of the telescope to the changing atmospheric conditions, so that bad or marginal seeing no longer limits the telescope. When the systems are refined, the quality of images at the telescope will rival those of a telescope freed from atmospheric limitations, as if it were in outer space.

  The Hubble Space Telescope goes a step further, putting a 96-inch telescope, with a variety of detectors and instrumentation, in orbit above the atmosphere. It was a frightfully expensive project, and with the many delays, it rivaled the two-hundred-inch in the total years it took from conception to launch. The infamous error in the primary mirror of the space telescope would have popped out to John Anderson and Marcus Brown on one of their Saturday tests. As a measure of scale, which needs some adjustment for inflation, the cost of the repair mission for the Hubble Space Telescope was almost exactly one hundred times the total cost of designing and building the Hale telescope.

  The two-hundred-inch telescope is no longer the biggest working telescope in the world. The mirror is no longer the most perfect optical surface ever polished. Stressed-lap polishing, using a machine with a polishing surface that continuousl
y adjusts itself as it turns over the disk, enables opticians to grind deeper, faster mirrors, to finer tolerances, in a fraction of the time. Instead of the twelve years it took to figure the two-hundred-inch, a large mirror can be figured in less than two years. Alt-azimuth mounts have become the mark of newer computers, permitting a simpler, lighter, and more compact instrument. New dome designs, some looking like futuristic angular architectural experiments, offer better thermal adjustment and insulation for the telescopes inside.

  But the two-hundred endures. If some newer telescopes are bigger, and in sites with superior seeing and without the light pollution of Los Angeles and San Diego, for some work the two-hundred is peerless. The scarcity of observing time on big telescopes puts a premium on reliability and versatility, and the two-hundred is the “rock” of big telescopes. Newer telescopes with their alt-azimuth mountings look more modern than the massive amount of the two-hundred, but they have the disadvantage that to match the sidereal motion of the heavens, instruments have to be rotated constantly, requiring complex software and slip-ring connections. In the interest of lightness, which generally translates to a saving in cost, some new telescopes have limited space and rigidity to support big instruments. Although the size and weight of modern instrumentation wasn’t contemplated in the original design, the two-hundred was overbuilt, constructed to last a century. Pease, Porter, Serrurier, Kroon, and others produced a tube and mount rigid enough to maintain the critical optical alignment with two tons of instruments hung behind the mirror. That stability and reliability, and the long record of successful observations from Palomar, make the telescope an ideal testbed for new instrumentation and a prime candidate for new innovations, like adaptive optics.