The Perfect Machine

by Ronald Florence

  As the plans were refined, Porter’s drawings went to the architect G. W. Iser. What emerged was an extraordinary building. The interior was a single room, which would be used for grinding and polishing the big mirror. The building would be windowless, the walls lined with cork, and with a special air-conditioning and ventilation system to hold the temperature and humidity steady and to maintain the air pressure inside the building higher than the pressure outside, to prevent the entry of dust. Additional rooms at the end of the building and in the basement provided laboratories for working on smaller disks and for experimentation with coatings for the mirrors.

  In early March, Ellis finished surfacing the sixty-inch disk with clear quartz. For nine days the disk was sealed in the huge furnace, while the temperature was slowly reduced to anneal the surface of the disk. The spider frame on the big 135,000-ton lift hung overhead, waiting for the unveiling. There were no guidelines for annealing fused quartz, certainly not for a mass as large as the five-foot mirror, but Ellis had extrapolated from their experience with smaller trial disks and concluded that ten days should suffice.

  A few hours before the scheduled annealing time was up, an impatient workman lifted the top of the oven to peek at the disk. Ellis was called over to take a look. He telegraphed the news to Pasadena: EXPERIMENTAL SIXTY INCH MIRROR COMPLETED FOURTEENTH. SURFACE QUALITY NOT GOOD. COOLED TOO RAPIDLY BELOW TWO HUNDRED DEGREES. STARTED CRACK.

  The gloom in West Lynn and in Pasadena, coming on the tail end of the brief euphoria over the successful spraying of the base, was deadly. Harlow Shapley visited the River Works not long after the crack was discovered to discuss a disk for a Harvard telescope. He was “sufficiently disturbed by the expense and the uncertainties of this great experiment” to catalog his qualms for Walter Adams, his former colleague from his days at Mount Wilson.

  Thomson and Ellis, Shapley reported, were “entirely too cocksure about the superiority of quartz.” Shapley believed that the next effort at a sixty-inch disk would probably succeed, and that if the money held out GE could probably go ahead and produce the larger blanks. But he wasn’t convinced that the quartz disks could ever be ground and figured to the precision required in a two-hundred-inch telescope:

  The surface of a 60-inch, if I remember correctly, must be corrected to one-tenth of a wave length. The larger aperture and longer focus of the 200-inch requires correction to a fortieth of a wave length. Who can do that kind of work? With these quartz mirrors laid down with a sprayer working back and forth in one direction, how can we be at all confident that a cylinder will not appear? … Is it not doubtful if we know how to anneal quartz at the present time? Some inquiries of Mr. Ellis the other day elicited the surprising information that he did not know whether or not quartz has two critical points, as all glasses. The whole secret of glass annealing depends upon accurate knowledge of the two critical points. Ellis didn’t seem to know that such existed even for glass.

  Shapley tried to explain his concerns to Ellis—how the pattern the sprayers made might affect the way the quartz reacted to changes of temperature. A piece of rolled glass, he explained, responds to changes in temperature with cylindrical distortions* that reflect the rolling process; if two pieces of rolled glass are fused at right angles to the direction of rolling, the distortions are canceled out. He suggested that the burners for the quartz blanks be moved north-south for one layer, then east-west for the next to produce the same effect. Ellis answered: “Quartz is not glass.”

  Ellis was in charge now. Thomson had lost touch with the fused-quartz project. At one point he had visited the lab and noticed that thousands of mosquitoes were attracted to the electric resistance furnace. He collected mosquito corpses to determine their sex, then consulted with an entomologist at Harvard to find that the mosquitoes were males, and that they had probably been attracted by the sixty-cycle pulses in the resistance circuits. At the time GE didn’t capitalize on the discovery to produce electric mosquito traps, in part because by 1928 Thomson was already interested in yet another project, an effort to recover minerals from a meteor crater in the Southwest. Thomson was on a ship when he got the news that the sixty-inch disk had cracked. Other GE officials assured him by wire that the crack in the disk was “a blessing in disguise” because they had learned important lessons about how to handle the annealing cycle.

  It took only a few weeks before the GE officials turned the latest disaster into positive publicity. Ellis summarized their experiments in a long internal memo that recalls some of the early battlefield reports from World War I. As Russian troops ran roughshod over millions of hectares of the Austro-Hungarian Empire, the Austrian high command triumphantly announced: “Lemberg is still in our hands.” Ellis, conceding that they needed more work on the spraying process and that they didn’t know what would happen to a mass as large as the proposed two-hundred-inch mirror if it were kept at 1100°C for weeks, triumphed that the furnace had not broken and that the fundamental process appeared to be sound.

  On the basis of these last experiments, Ellis announced that they were ready for what he called “production” runs. The only impediment to taking orders for disks from Harvard and other observatories was the need to come up with a formula to amortize the initial expenses that had been paid by the Observatory Council. Assuring Hale and Anderson that he would soon be ready to try for a new sixty-inch disk, Ellis sought Swope’s approval before he sent yet another set of budget figures to Pasadena.

  In Pasadena they read different fortunes from the tea leaves. For the Observatory Council the problems with the disk had reached a crisis point. After three years and more than six hundred thousand dollars of expenditures, all they had received from GE were more revised (and inflated) budget figures, assurances that had begun to sound hollow from repetition, and bits of imperfect quartz for testing. Hale, Millikan, Henry Robinson of the Board of Trustees, Noyes, Adams, and Anderson scheduled a long meeting at the solar lab on Saturday morning when they wouldn’t be disturbed. Ostensibly the agenda item was a review of the new budget figures from GE. The real question was whether they could afford to continue the effort to get mirror disks from GE. If not, what were the alternatives?

  Before they turned to Ellis’s new figures, Adams suggested that he had some test data that might be useful to the council. He had just completed initial tests of a new mirror-support system Pease had installed on the one-hundred-inch telescope at Mount Wilson, based on a design Pease had developed for the two-hundred-inch.

  Instead of using defining arcs and springs to hold the mirror, the new supports mounted the disk on balls to reduce the friction* as the mirror moved. Adams reported that in the first tests with the new supports, the optical changes in the figure of the mirror seemed to be greatly reduced: instead of needing a long period to readjust, the mirror held its shape as it was slewed to different areas in the sky.

  If the mounts worked as well in use as in the preliminary tests, he said, the one-hundred-inch telescope had a new lease on life. Hours or whole evenings that had been lost to waiting for the mirror to adjust to changes in temperature would be minimized or eliminated. The changes in figure of the mirror had limited the resolution of the telescope in certain positions. Now, it seemed, Hubble, Humason, and other observers could extend their reach even farther.

  Hale and his colleagues talked for most of the morning about Adams’s report. If Pease’s new supports made that much difference, it seemed that the decade of problems with the one-hundred-inch telescope had not been caused by the coefficient of expansion of the plate-glass mirror. And if plate glass would work for a one-hundred-inch mirror, maybe the two-hundred-inch mirror could be fabricated from a material less exotic than fused quartz. Plate glass still wouldn’t work for a mirror that large, but one possibility was the new Pyrex borosilicate glass from Corning. Hale had used experimental Pyrex mirrors in some of the solar telescopes on Mount Wilson. He found the material optically stable and reliable, even in an application as demanding as a telescop
e element exposed to the full heat of the sun. He knew that Corning was already melting one hundred tons of Pyrex at a time in its furnaces.

  After the meeting Hale optimistically wrote, “We believe that a Pyrex disk would serve admirably for the 200-inch telescope.” He asked Anderson to quietly get quotes and a timetable on mirrors from Corning.

  Shortly after the meeting Heber Curtis, who had gone from Lick Observatory to the Allegheny Observatory and then to the University of Michigan as chairman of the Astronomy Department, asked Hale for his opinion of the GE process for a large mirror blank for a new telescope for the University of Michigan. Citing the experiments with the one-hundred-inch telescope, Hale recommended that Curtis go to a Pyrex blank.

  Emboldened by the possibility of an alternative source for the disks, the council drafted a telegram to Thomson at GE, questioning Ellis’s newest estimates of $80,000 for another sixty-inch disk, and an additional $1 million for a two-hundred-inch disk. The language of the telegram was strong. Before he sent it Hale telephoned Max Mason to get his views. Mason urged Hale to wait before breaking off all work with GE.

  Mason had reasons for his cautions to Hale. Like every other institution dependent on portfolio income, the Rockefeller Foundation had suffered from the drop in the stock market. The IEB, which had made the grant for the telescope, saw the value of its funds drop so precipitously that it could not make good on the balance due under the grant. With Wickliffe Rose gone, there was no continuing need or support for the fund that had been established at his insistence, so the remaining IEB funds were transferred to the GEB, which took over the financial responsibility for the grant. The lines between the various Rockefeller foundations had grown hazy enough that Mason could personally oversee the telescope project, with the checks being paid out of GEB accounts.

  The drop in the stock market did not directly affect the funding for the telescope, but the telescope project could not avoid the consequences of the depression. The grant had been awarded in an era of unbounded optimism, when Americans trusted, even worshiped, science. The early twentieth century had produced electric lights, automobiles, airplanes, radio, telephones, refrigerators. The following decades, with the wondrous scientific work that was being reported on the radio and in the news weeklies, were supposed to bring even greater prosperity and ease to ever more Americans.

  The depression rattled America’s love affair with science and technology. Breadlines, soup kitchens, and tent cities stood in stark contrast to the buoyant promises of progress and prosperity. The economic critic Stuart Chase told the Women’s City Club of New York that the advent of “talkies” had thrown ten thousand movie house technicians out of work. Men and women who only a few years before had rushed to read about or buy new inventions now turned their backs on machines and science, as Luddite arguments, once confined to a radical fringe, began to appeal to wider audiences.

  Congress joined the reaction. The budgets for science research in the United States were small to start with. With the exception of the grant for the two-hundred-inch telescope, and Lawrence’s successful fund-raising for his cyclotrons at Berkeley, most scientific research got by with the minuscule funds that universities could provide to their faculty. For the 1932 fiscal year Congress cut the appropriations for all science-related agencies drastically: The budget of the Bureau of Standards was cut by 26 percent. The few state-funded scientific agencies were cut even more. Private institutions devoted to science, like the Rockefeller Foundation or the Carnegie Institution, even if they were eager to keep up their funding efforts, felt the effects of the depression on their investments or in their fund-raising efforts. The annual income of the Carnegie Institution of Washington fell by $ 1 million, a substantial sum when the entire annual budget of the Mount Wilson Observatory was $250,000.

  Signs of the depression were everywhere by 1931, even in prosperous towns like Pasadena. The movie business was still hard at work, churning out musicals that invited the world to sing or dance away its cares. The rest of Southern California’s economy collapsed. The aircraft industry stood idle. The new refrigerated-produce industry, which was on the verge of revolutionizing California agriculture, went on hold. As prices for farm commodities fell, the agricultural exchanges tried dumping goods to preserve prices. In the citrus belt east of Pasadena, hundreds of tons of “surplus” oranges were thrown into dry riverbeds and sprayed with oil and tar so the unemployed and relief clients wouldn’t be tempted to salvage from the decaying heaps.

  When Fortune magazine sent a team of reporters to the Caltech campus, the reporters searched in vain for signs of the depression that had struck everywhere else. The reporters dwelled on the exciting discussions of cosmology and nuclear physics, Millikan’s own research on cosmic rays, and the path breaking observational efforts on the one-hundred-inch telescope. What they didn’t know was that the faculty had voluntarily agreed to a 10 percent pay cut, and that Millikan, Hale, and Noyes, the three great lights of the campus, had made up a portion of the school’s deficits from their personal funds.

  All construction on the campus had stopped, with the exception of the astrophysics laboratory and the optical lab, which were funded out of the grant for the two-hundred-inch telescope. The resentment from other departments was hard to ignore. Famous faculty members like Millikan and Noyes could always count on foundation grants to support their work, but the physics and chemistry labs were hard pressed for equipment and supplies to continue modest research programs by junior faculty and graduate students.

  While other faculties scrounged for materials and space, astrophysics had a huge, well-equipped machine shop, its purpose to build parts for a telescope that couldn’t even be designed until there was positive evidence that the great mirror could be successfully cast. An astrophysics laboratory was nearing completion on California Street, to design instruments and observation programs and to analyze the data from the unbuilt telescope. At ten different sites in Southern California and Arizona, telescopes had been set up to accumulate data on local weather and seeing, as part of the site survey to pick the best possible site for an unbuilt telescope.

  The ever-volatile religious movements of Southern California, prospering in inverse correlation with the economy, found ready targets in Caltech, telescopes, and any other manifestation of science. The “work of the devil,” the evangelists were quick to point out, exacted its tolls; the joblessness of the depression was punishment for decades of immorality, including the assaults of science on God’s kingdom. In the early years of the institute, the gathering of scientists had been a ready target for the evangelists. By 1930 Millikan had come out for religion, or at least for God, which made an uneasy truce with the fundamentalists. The graffiti writers had a good time with that one: Under the signs saying JESUS SAVES, they would scratch, BUT MILLIKAN GETS CREDIT. The campus cynics didn’t like Millikan’s soft stance on religion, but it did mean that the fundamentalists found targets other than the telescope for their rallies and demonstrations.

  Once at the center of headlines, the telescope had faded from public attention. The new focus of attention for the public who hadn’t yet lost faith or interest in science was Albert Einstein, who announced that he was leaving Germany permanently.* American institutions competed extravagantly to recruit him. Millikan’s bids from Caltech were too late and too small. Einstein ended up at the Institute for Advanced Studies, in Princeton, a university without students. Photographs of the white-haired scientist who biked around Princeton and played his violin at dinner parties were soon familiar fare in the weeklies.

  The individual members of the Observatory Council were mostly spared the effects of the depression. George Hale’s investments had been conservative. Even after the stock market crash, his income was approximately four times what the highest paid faculty member at Caltech made. The other members of the Observatory Council—Millikan, Noyes, and Adams—were not dependent on their investments for income. The exception was Henry Robinson, an ex-officio membe
r. Robinson had been a booster of the telescope project from the beginning, not only for Caltech, but for Southern California. He had lost a considerable sum in the collapse of the stock market, including the funds he had pledged as an endowment for the telescope. No one, even on the Observatory Council, seemed to notice. Progress on the telescope had been so slow that the endowment for operation seemed a distant concern in 1931.

  Max Mason accompanied Noyes, who was heading east for other business, on a fact-finding mission to West Lynn in April 1931. Ellis and his staff put on a full show, with a presentation on the spraying process and a tour of the facilities. GE had accumulated an enormous array of equipment for the mirror disk program, from banks of transformers to power the electric furnace to the stockpiles of quartz, and the strange expandable steel building that housed the enterprise.

  GE’s show was a faux pas. If they were VIPs on the telescope project, Noyes and Mason were also working scientists who could see that there were fundamental problems with the quartz fabrication process. They brought up Shapley’s suggestion of switching direction of the movements of the spraying head, first north-south, then east-west, to avoid introducing strains in the quartz. Hale had raised the same concerns in a telegram, asking that the tensile strengths of a piece of quartz from the flawed sixty-six-inch disk be tested “both parallel and at right angles to direction of burner motion.” Ellis promised to carry out the tests. Mason also recommended that Ellis try pressing down on the surface of the heated quartz immediately after spraying the disk to minimize the ridges that formed on the surface.