The Perfect Machine

by Ronald Florence

  The process was tricky. The aluminum had to be vaporized over the disk in a bell jar by heating it with tungsten coils. To produce a smooth, even layer on the mirror, the aluminum molecules evaporating from the coil had to travel to the surface of the disk in a straight line, which meant they could not hit another molecule. A collision-free path required that the bell jar be evacuated to a vacuum of one ten-thousandth of a millimeter of mercury, in a contained space so free of leaks that if it were evacuated and sealed off, it wouldn’t reach one-half of atmospheric pressure for fifteen years. Strong’s process required that the optician duplicate the near emptiness of outer space—no easy task in a laboratory.

  George Hale, who kept his fingers on any technology that affected big telescopes, had eagerly followed Strong’s work. When Strong aluminized the mirror of the Crossley reflector at the Lick Observatory, the telescope Heber Curtis had used for his surveys of “island universes,” it improved the performance of the telescope so much that he was recruited to aluminize the mirror of the one-hundred-inch Hooker telescope on Mount Wilson. The results were spectacular. Before, the tiny companion star of the bright star Sirius had been difficult to photograph with the one-hundred-inch telescope, because the fine scratches that were inevitable on a silvered surface as it was burnished—no matter how fine the rouge used—would scatter the light. With Strong’s new aluminum coating the companion star was easily resolved in plates. There was no question that they would try for an aluminum coating on the two-hundred-inch disk.

  It had taken some tinkering to get Strong’s temperamental process to work on a mirror as large as the one-hundred-inch. The coating was only one thousand atoms thick, less than one-thousandth of a millimeter. For optimum effectiveness the thickness had to be uniform to within 4 or 5 percent. The slightest contamination on the surface would cause the coating to fail. Strong constantly experimented with new techniques for cleaning the surface before depositing the aluminum.

  In the spring of 1947 Anderson invited Strong to return to Caltech from Johns Hopkins University to supervise the aluminization of the two-hundred inch disk. Strong was delighted to work on the big telescope. He also had no illusions that it would be a tough job. Before he came to Pasadena he urged that whatever vacuum pumps Caltech planned to use, they should have an auxiliary pump system ready to supplement the original pumps.

  The mirror cell of the two-hundred-inch disk, which held the support systems and the edge levers, complicated the plans for aluminizing the disk. The engineers had designed a wheeled platform, with screw elevators on each corner that could raise and lower the mirror and cell to the telescope. The platform served both as a carriage to remove and reinstall the disk in the telescope, and as the bottom half of the vacuum chamber. A steel bell was designed fit over the platform, with a rubber gasket fitted around the edge of the disk to isolate the high-vacuum area from the back of the mirror with its delicate support systems. Without the gasket, drawing down a vacuum to the level required by aluminizing would suck the oil and grease out of the bearings of the support assemblies, contaminating the vacuum and forcing a tricky relubrication of the mechanisms.

  Strong had cleaned smaller disks by hand-burnishing with virgin chamois and extrafine rouge. The success rate was spotty. The worst contaminant, he discovered, was microscopic traces of oil from the human skin. With opticians hand-burnishing a two-hundred-inch disk, removing the traces of oil would be a Sisyphean task. So for the two-hundred-inch disk, he planned a new cleaning technique: he would first coat the surface of the disk with a “special fatty acid compound and precipitated chalk powder.” The precipitated chalk would be wiped off with virgin felt pads, leaving the fatty acid on the surface of the disk. He would then burn the residue of fatty acid off with an oxygen glow, leaving a pristine surface for the aluminizing. Strong’s procedure sounded as if it would work. The opticians eagerly awaited his arrival at Palomar with the special materials.

  When Strong arrived, the “special fatty acid compound” turned out to be cases of Wildroot Cream Oil hair tonic. Radio listeners all across the United States knew what the magic ingredient was:

  You’d better get Wildroot Cream Oil, Charlie;

  It keeps your hair in trim,

  Because its non-alcoholic, Charlie;

  It’s made with soothing lanolin.

  “In order to get glass clean,” Strong told the opticians and astronomers, as he unpacked cases of hair tonic, “you first have to get it properly dirty.”

  Strong set to work on the two-hundred-inch disk with his Wildroot Cream Oil treatment. He and the opticians wiped the surface clean just before the overhead crane lowered the bell onto the base of the aluminizing chamber. When it was sealed in place, Byron Hill flipped a switch, and the oxygen glow of the heating coils burned the lanolin residue off the surface of the disk, along with every trace of human body oil. Then Hill flicked another set of switches to start the big oil-diffusion pumps to evacuate the bell to a high vacuum.

  The pumps ran for days. Every hour someone would check the vacuum gauges and jot figures in the notebook in which Don Hendrix was recording each step of the aluminizing process. Progress was slow. Technicians drifted off to meals, naps, other work. The staff took turns checking the gauges and jotting notes in Hendrix’s log. On the second day Ben Traxler, annoyed by the constant drone of the pumps and the funereal atmosphere of the watch over the process, wrote in the log that he had let air into the chamber to quiet the pumps. Hendrix saw the note and blew up. He had just been named optician for the two-hundred-inch mirror and wasn’t ready for jokes.

  After four days the pumps had pulled down the best vacuum they could achieve. Hendrix fired the tungsten coils that held the pure aluminum, releasing a rain of aluminum molecules. There were no windows into the chamber. They had to lift the bell jar to see the results.

  The crane let off a soul-satisfying growl from the weight of the bell as Ben Traxler lifted it off. The disk was not soul satisfying. The surface was a blotchy mess.

  The problem was obvious. The pumps hadn’t achieved the required vacuum. Hendrix and Strong tried again, running the pumps longer. While the pumps labored, Strong ran around with a tin of shellac, trying to dope up the leaks in the bell jar. Again the process failed. They tried running the pumps for only twelve hours and firing the aluminum in a “dirty” vacuum. They tried firing the aluminum on one-quarter of the mirror at a time. Each time Byron Hill’s notes recorded that the mirror emerged “dark in quarters of old firing. N.G.”

  Nothing they tried worked. The bell jar, with the tricky gasket around the edge of the mirror, leaked. Even without leaks high vacuums were a laboratory challenge. The early cyclotrons at Berkeley were covered with gobs of red sealing wax on the brass chamber to close leaks. There was no quick and dirty fix for this jar: Sealing wax wouldn’t hold the vacuum the aluminization required. The only answer was pumps large enough to pull a high vacuum despite the leaks. The engineers searched the catalogs and found that no company stocked larger pumps. Even if they had found a supplier, there were no funds in the depleted telescope budget to buy larger pumps. Twenty years of work was at a standstill.

  The bad news got to Max Mason, who made some phone calls. Under Mason’s tenure as president, the Rockefeller Foundation had made the first grants supporting Lawrence’s work on cyclotrons at Berkeley. Between his contacts in Berkeley and his friendship with Vannevar Bush, Mason got the old boys busy. One phone call led to another, and before long an official from the Manhattan Project located some huge vacuum pumps at Oak Ridge that weren’t in use. The pumps were rushed to Palomar. The huge pumps were powerful enough to draw down the needed vacuum despite the leaks in the jar.

  One more Wildroot Cream Oil treatment, and the aluminization succeeded.

  First light was three nights before Christmas 1947. John Anderson led the delegation of astronomers and engineers who watched Byron Hill’s workmen install the mirror. Anderson had been waiting twenty years for this moment. He had heade
d the project from the beginning, supervised eleven years of grinding and polishing the disk. Only a heart condition had forced him to relinquish the responsibility for final testing and figuring of the disk on the mountain. This was his moment.

  Byron Hill’s workmen raised the disk and cell into position with the electric screw hoists and the workmen went to work on the circle of bolts that held the mirror cell in position on the telescope tube. Suddenly a bang and a hideous squeal echoed through the dome. Eyes flashed toward Anderson. Had the disk cracked? Would Anderson’s heart survive that?

  No one moved. After a long silence Don Hendrix said, “You ever seen a one-million-dollar bolt snap?” The mirror hadn’t cracked. The securing bolt hadn’t snapped either, only squealed in protest. Anderson and his mirror were fine.

  The telescope was far from finished. Bruce Rule was still working on portions of the drive mechanism. The mirror had a turned-up edge, and the support mechanisms were untested. Still, someone had to have a first look. The honor went to Anderson. There were no provisions for visual observations with the telescope, so he used a small reading glass as an eyepiece. Anderson sat in a lift chair that raised him up to the Cassegrain focus and gazed for a while at the Milky Way. When he finally came down from the eyepiece, everyone wanted to know what he had seen. He answered laconically, “Oh, some stars.”

  Anderson didn’t say more. Others—Bowen, Porter, Brownie, and visiting Dutch astronomer J. H. Oort—followed him. When Hill took his turn, he was amazed. He had never seen so many stars in his life. They looked like pollen on a fish pond. Forty years later he remembered that after so many years of work, he felt “pretty good.”

  Byron Hill wasn’t an astronomer or an optician. When Ira Bowen looked, he understood Anderson’s silence. The images were worse than disappointing. The mirror appeared to have a staggering astigmatism of perhaps twenty wavelengths, which meant a very serious problem. Bowen ended the test session with a curt “What the hell!”

  In the morning the workmen discovered that one of the three points of fixation for the mirror was out of position. The mirror cell had been forced back against an earthquake safety block.

  For the next trial Edwin Hubble showed up. Hubble peeked through the eyepiece and pronounced the star images “good,” just as the mirror was. He went back to Pasadena with his favorable report, and Anderson told Mason, “You can be as optimistic as you please about the mirror.” Anderson’s guarded words were halfway between circumlocution and euphemism.

  While Hubble’s pronouncement leaked to the public, Bowen did some tests with a Hartman screen, a full-size cover at the top of the tube, with regularly spaced apertures admitting light to selected regions of the mirror. The tests were disappointing, with zonal problems, unanticipated problems with the outer edge of the disk, problems with the alignment of the mirror, and a severe vibration in the mount that made testing difficult. Bowen knew then that it was going to be a long time and much work before the telescope would be ready. After Walter Adams’s warnings about what the press would do with even a hint of bad news about the telescope, Bowen kept only handwritten records of his mirror tests. Even Mason’s reports to Weaver at the Rockefeller Foundation were in handwritten letters, avoiding the dangers of talkative stenographers and carbon-copied memorandums.

  It was a disappointing start.

  Mason couldn’t avoid the reports to the Rockefeller Foundation. As of January 1, 1948, the balance of funds available to the project was effectively zero.

  For Mason, who had headed the Rockefeller Foundation for so many years, the bankruptcy of the project was a serious embarrassment. At the end of the war he had requested a supplementary grant of $250,000, assuring the foundation that with that additional sum in hand and the funds from the government for the use of optical and machine shop space and equipment during the war, Caltech would finish the telescope. Mason’s request seemed reasonable, the explanation of the delays and extra expenditures of the war years made sense, and the board and officers of the foundation trusted their former president. The supplementary grant was promptly approved. Two years later, with much work still remaining to be completed, the $250,000 was gone.

  Weaver, Mason’s protégé at the Rockefeller Foundation, had been in charge of overseeing the telescope project since Mason left in 1936. He didn’t welcome the task of reporting that the project he had watched over for so long had exhausted its funds.

  What had gone wrong? Partly it was Murphy’s law. The polishing of the mirror had taken much longer than even the most pessimistic expectations. Deflation had hit portions of the American economy, but the cost of retooling from war production inflated the prices of precision instruments and tools needed to complete the telescope by 50 percent.

  When a committee of astronomers convened in early fall of 1947 to compile a list of essential instrumentation for the telescope, their list of additional instruments needed at the different observing stations of the telescope, and supplementary measuring instruments needed for the reduction of observations, added another $36,500 of unanticipated costs. In round terms the minimum Caltech needed to finish the telescope was another $250,000 and even that sum would omit some instrumentation that would permit full utilization of the telescope.

  Weaver had too much experience with scientists who were longer on ideas than on precise budgets to be surprised. The era of big science, ushered in by projects like the telescope, would see a new generation of scientists who accepted project proposals and detailed budgets as part of the work of science. But Palomar had begun as the project of gentlemen scientists who wrote their budgets in round numbers, like Hale’s original budget of $6 million. “The project is a sort of scientific heroic poem,” Weaver said in a handwritten memo to his boss R. B. Fosdick, the president of the Rockefeller Foundation. “And like all truly great poetry, it contains elements of both joy and sorrow.” The joy was that the biggest telescope in the world was almost finished. The sorrow was that Max Mason’s colleagues at Caltech had not been realistic about the cost and timetable for the telescope. They were dreamers, not accountants: “We must remember that a group with the scientific power of imagination to create this massive and precise instrument is not necessarily composed of persons who would also do the most systematic and orthodox job of accounting and budgeting.” Mason himself was the perfect example: “Almost anyone else would have done a more careful and reliable and patient job on the estimates & the accounting: but if he [Mason] had not put his great immaginative [sic] ability to work on the problem of the elastic deformation of the mirror, it might very well have never been brought to a successful figure.”

  Fosdick and the board of the Rockefeller Foundation accepted Weaver’s recommendation that the foundation grant an additional $300,000, $50,000 more than Mason had requested, to support the completion of the project and to ensure that there was no delay in getting the telescope functioning with a full scientific program. The grant would bring the total Rockefeller commitment to $6,550,000. From the time of the original grant, in 1928, the budget had increased by only 9 percent and had absorbed not only the unsuccessful efforts to develop fused quartz at General Electric but the cost of the 48-inch Schmidt camera, itself a major research instrument, the largest of its kind in the world. The final grant would be the first with conditions attached beyond the vague original specification that the telescope was to be as “perfect” as possible. Caltech was to agree that major construction would terminate on April 1, 1948, that Caltech and the Carnegie Institution would assume the regular operational budget of the observatory from that date, and that any expenditures from Rockefeller funds after that date would be strictly accounted.

  With the disk out of the optical shop, the shop was shut down. The big mirror would never return from the mountain, and future optical work on smaller mirrors and instruments could all be done at the Mount Wilson optical labs on Santa Barbara Street. Lick Observatory agreed to buy the 120-inch disk and grinding machine for $75,000, once it was demonstrate
d that the support mechanisms of the two-hundred-inch mirror actually worked. Because the 120-inch disk had originally been designed for use as a flat to test the two-hundred-inch disk, it wasn’t thick enough to grind into a deep shape for a fast mirror. The Lick Observatory had to settle on a slow f/5 design for their big telescope, which would soon be the second largest in the world. The Griffith Observatory in Los Angeles bought the 40-inch-diameter plug that had been used in the center of the disk and put it on display. Across the country, Corning already had the first, flawed two-hundred-inch disk on display.

  Marcus Brown wrote from his farm that he wanted to buy the grinding machine for the two-hundred-inch disk, for sentimental reasons, but couldn’t afford it. Brownie, his task done, seemed lost. He dabbled briefly in a jewelry store in Pasadena, but his legs were gone, done in by too many hours at the machines. The parts of the machine weren’t worth the cost of dismantling. It was sold for salvage.

  Wickliffe Rose’s original requirement that Caltech provide endowment for the maintenance and operation of the observatory had long been forgotten. Caltech did not have the $30,000 in annual maintenance costs that they were obligated to provide, an expense that was expected to increase, substantially, year by year. The president of Caltech and his fund-raising staff had their work cut out.

  At the observatory Ira Bowen also had his work cut out.

  The most immediate problem was the unexpected vibration in the mount. When it happened the telescope would tremble for as long as one minute, ruining any test in progress. Sometimes the vibration seemed to start spontaneously. It also started when an observer moved from one side of the prime-focus cage to the other. Byron Hill blamed the tiny round seat the designers had provided for the observers, so uncomfortable that an observer couldn’t help but wiggle. He tore the old seat out and substituted the seat from an old hayrake that had been left around from the days when the Beech family farmed on the mountain. It was more comfortable—what observers call the tractor seat is still there—but the new seat didn’t cure the vibration.