The Perfect Machine

by Ronald Florence

  In 1934 there was still no design for the telescope, only a collection of specifications, ideas, and requirements.

  From the earliest talk about a big telescope, Pease had favored a fork mount, like the mounting on the sixty-inch telescope. The advantage of the fork mount was simplicity: Because the tube of the telescope was held only at one end, the fork permitted the telescope to point to all of the sky overhead. Pease had designed most of the one-hundred-inch telescope. His early drawings and model were the departure point for the project, and the models that were shown at the National Academy of Sciences and brought out for curious reporters were based on his drawings and early models. As late as February 1931, when a reporter asked about the design of the telescope, Hale answered that “the fork type [mount]… seems the most promising.”

  But there were problems with a fork mount. If the telescope tube were supported only at one end, while the other end carried the weight of auxiliary mirrors, mechanisms to interchange the mirrors, and a cage big enough for an astronomer—achieving the required rigidity in the tube became an impossible engineering task. At the tolerances required for the telescope, even a massive tube of steel girders behaves like a hollow noodle cooked al dente. Hold it at one end, and the other end droops. Making the forks long enough to support the tube in the middle would substitute drooping forks for a drooping tube. The fork mount also presented problems for the use of some of the alternate foci astronomers had requested, and the concentration of the entire weight of the telescope on a single set of bearings troubled the engineers.

  There were two alternatives, which pivoted the tube in the middle. Early on, Hale had urged consideration of the so-called “German” mount, made popular by Zeiss in the telescopes they had designed and built. The German mount—also called the “Victoria” mount after a 72-inch telescope in British Columbia—put the tube of the telescope on an extended shaft, with a counterweight on the other end of the shaft. It was popular for small telescopes, but a counterweight would effectively double the load on the bearings for the tube of the telescope, making the mount impractical for a telescope as large as the two-hundred-inch. The other possibility was a form of yoke mount like the one on the one-hundred-inch telescope. The yoke could be supported at both ends, dividing the bearing load for the telescope. The disadvantage of the yoke was that it blocked the tube of the telescope from dipping low enough to view the North Polar region of the sky. Hubble and Humason were already chafing at the inability of the one-hundred-inch telescope to photograph or take spectra of galaxies in the north polar region. No one wanted to build another telescope with that limitation.

  Russell Porter, who had never before worked on a large telescope, had played with some maverick ideas. In the telescopes he had designed for the Stellafane amateur astronomy center, and in drawings he had done for the “Amateur Telescope Maker” column of Scientific American, Porter had come up with telescope designs unlike anything anyone had ever seen before. Some telescopes kept the observer in a fixed location indoors, with the moving parts of the telescope outdoors. One design, his Garden Telescope, looked like a decorative sundial, with parts of the mounting done up in Art Nouveau scrolls.

  As early as 1918 Porter proposed a split ring for the north end of a yoke. The split ring would allow the tube of the telescope to be lowered enough to view the polar region, while still dividing the weight of the telescope between bearings at the two ends of the yoke. Porter patented the idea in 1918, before he began work on the two-hundred-inch telescope. When he got to Pasadena, he began sketching ideas for a mounting for the two-hundred-inch telescope, using variations of his split ring.

  Porter, who was older than the others, and whose hearing difficulties left him the odd man out at meetings, had a hard time convincing people that his ideas deserved attention. He was an outsider, without the formal academic credentials and big-telescope experience of the Mount Wilson and Caltech staff on the project. He had also been brought in by Hale at what others in the project considered a too-high salary. Porter’s initial appointment was for a year, and his official design assignments were small projects like the telescope used to measure the seeing at sites, and architectural details of the Astrophysics Laboratory and Optics Lab on the Caltech campus. Frustrated, Porter threatened to go back to Vermont. Hale encouraged him to stay on. When others ignored Porter, Hale would ask him privately to sketch up various designs. Porter’s memos were pencil sketches and hand printed notes rather than the typed memorandums with eight carbon copies that characterized most of the official communications of the project.

  By 1934 the chief design question was whether the telescope should have a fork mount or some variation of the split-ring design. Porter’s sketches, amalgams of his own ideas and the suggestions from brainstorming sessions, gradually evolved into a giant or horseshoe, with the outside of the horseshoe serving as a bearing surface to support the telescope as it turned in right ascension to match the motion of the earth. Francis Pease, who probably came up with the idea independently, had used a form of split ring in one early sketch he did of the three-hundred-inch telescope that had been talked about in 1921. As the design group settled on the horseshoe, Pease was generally credited with the design. Porter, angry that he was not given proper credit, became even more isolated from the others working on the telescope.

  With the main design agreed on, it was time to move from sketches to engineering drawings. From the earliest days Hale and Anderson had turned to the Caltech engineers for advice and designs. Theodor von Karmann in aeronautical engineering and Romeo Martel in mechanical engineering, stars of their departments, and the mathematical physicists Harry Bateman and Paul Epstein had all had been consulted on the final design for the ribbed back of the glass disks. They were asked to recommend both personnel and ideas for the telescope mounting.

  Romeo Martel recommended a young engineer named Mark Serrurier, who had come to work for Caltech in November 1932. Engineering jobs were tough to find in 1932. The aircraft companies were at a standstill, and public authorities were cutting back to limited maintenance programs. The design program for the telescope soon had its pick of young graduates, eager for the chance to work on a major project. Serrurier had been eager to work in Southern California. The movie industry would have been his first choice, but he was willing to take any work he could get. He did some consulting on the new Golden Gate Bridge in San Francisco. The new telescope seemed an exciting project. Serrurier was too young to realize how impossible the assignment was.

  Even with the fast focal ratio of f/3.3, the main telescope tube for the two-hundred-inch telescope would be fifty-seven feet long and weigh close to 250,000 pounds. Early in the design process Hale wrote: “The first point is to study the tube, which should be extremely rigid and capable of carrying … the heaviest attachments without injurious flexure.” Dr. Ross, the designer of the corrective lens for the prime focus of the telescope, calculated that effective use of the lens required that it be fixed within 0.040 inch (approximately one millimeter) of the optical path of the telescope, and that the position of the lens could not move more than 0.020 inch during an exposure. What this meant in engineering terms was that for the duration of a photographic exposure, typically one hour, the alignment of the tiny lens and a mirror fifty-seven feet away could not vary by more than a half-millimeter—all while the enormous telescope tube that held both in alignment was moving to track an object. To complicate the design, the light path for the Coudé focus required a slit on one side of the telescope tube, so the light could be bounced to an auxiliary mirror.

  In his engineering courses or his work on the Golden Gate Bridge, Serrurier had been given design problems for skyscraper and bridge frames that were allowed to sway by feet. Quick calculations showed that the truss sections and cross-bracing in his engineering texts wouldn’t work for the telescope tube. Nor would the massive riveted-girder construction that had been used for the sixty-and one-hundred-inch telescopes. If the telescope tube were massive
enough to hold its alignment by brute force, the ends would droop under their own weight, and the telescope would be so massive that it would be impossible to design bearings and a drive mechanism that could move it smoothly.

  Day after day Serrurier sat in his office in the basement of the astrophysics building, playing with rulers and pencils to model the stresses on a tube structure. No idea worked, until one day Martel came by and suggested that the tube didn’t really have to be so rigid that the mirror and the lens at the focus would never move in relation to one another; what mattered was that they remain aligned within a half-millimeter. Imagine, Martel suggested, a tube that sags at both ends, but in such a way that the ends remain precisely parallel to one another. The lens at one end would still be perfectly aligned with the mirror at the other. In engineering terms Martel had switched the problem from flexible stress to sheer stress. Serrurier listened, sketched the idea on paper, then went back to work with his pencils and rulers. The solution wasn’t immediately obvious, but at least he was off dead center.

  Sinclair Smith had joined Anderson and Pease as a part-time Caltech employee. Smith was a brash young researcher in physics and astronomy, until he spent a year at the famed Cavendish Laboratories in Cambridge, where James Chadwick and his associates were probing the secrets of the atom during the 1920s. He returned “much matured,” and Anderson recommended him for the lab staff at Mount Wilson. Smith had worked with electronic detectors and controls, so his special project became the drive mechanisms for the telescope.

  Large telescopes like the one-hundred-inch and the sixty-inch were equipped with clock drives that turned the telescope synchronously with the sidereal rotation of the heavens, so that a star would appear to stand still in the telescope during a long exposure. Astronomers had long known that the apparent motion of the heavens is not simple. The rotation of the earth is not perfectly regular, and the apparent motion is also affected by atmospheric effects, gravitational influences from the moon, and other factors which could be modeled in equations. Smitty took on the challenge of duplicating not just the simple rotation of the earth but the other minuscule motions, so the motion of the telescope would come as close as possible to mimicking the apparent motion of the heavens, minimizing the demands on the observers for hand-guiding the exposures.

  He had the assignment of finding out all he could about photosensitive devices, time standards, servo controls, and other devices that might be useful to control the telescope. Like others who had pledged their time to the project, he soon discovered that the telescope could be all-consuming. Although Caltech was paying only half his salary, the project was eating up most of his time, cutting into the time he had free for research. And even as Anderson brought the engineers into the project, some basic and seemingly insoluble problems remained.

  First the rough engineering calculations showed that the bearing for the split-ring horsehoe would have to support a weight of close to 1 million pounds—five hundred tons. The big telescopes on Mount Wilson had used drums of mercury and floats for their polar axis bearings, which produced a smooth motion along with a then-unrecognized danger to the staff from the quantity of mercury. The weight and size of the two-hundred-inch telescope precluded that solution. The consideration of other choices—ball bearings, roller bearings, or a plain bearing—were frustrating enough that discussions of the bearings, Anderson recalled, “usually resulted in a headache.” The bearing for the horseshoe, which would carry most of the 500,000-pound weight of the moving portion of the telescope, would be the largest journal bearing ever built. The load would deform any kind of ball bearing, and while Pease put roller bearings into his drawings, even his own calculations showed that roller bearings would require some 22,000 foot-pounds of torque to overcome the friction and move the telescope. Moving a telescope against that much friction would require huge electric motors; the resultant vibrations would be impossible to isolate from the telescope. The weight of the horseshoe would also ultimately distort the rollers, which would then introduce wobbles into the motion of the telescope.

  Even if a bearing could be built, no one was sure that a structure as large as the horseshoe could be built with the rigidity the telescope required. Pease and Porter weren’t engineers, but their preliminary calculations showed that no matter what material or structure was used to build the horseshoe, the open end would deform slightly as it turned from one extreme to the other. A sag of one-sixteenth inch in the forty-six-foot diameter horseshoe would be enough to leave the telescope axis unacceptably misaligned.

  Still, the horseshoe design solved so many other problems that the unresolved issues were shunted aside. Caltech was a cocky institution. The engineers and scientists who worked on the project, some with little commercial experience or background in the strict hierarchy of an academic department, weren’t easily daunted by challenges that stodgier souls would call “uneconomical,” “impractical,” or “impossible.” And after the apparent success of the mirror casting in Corning, even Hale, Adams, and Pease—who had lived through the birthing pains of other big telescopes—were confident that the two-hundred-inch telescope could be built. Their optimism was contagious for the young scientists and engineers who worked on the project. Getting a job during the depression was an achievement. When the job was at Caltech, working on the biggest scientific project ever undertaken, with a committed budget, it was hard not to feel that there was nothing you couldn’t do, including solving problems that the books and experts said were insoluble.

  John Merriam, at the Carnegie Institution of Washington, had been chafing at the publicity given the project ever since the grant for the two-hundred-inch telescope was announced. Every time a newspaper or radio report appeared based on the GE press releases, Merriam wrote George Hale to protest that proper credit had not been given to the Mount Wilson Observatory for their contributions to the project. Hale or John Anderson would write back, explaining that they had nothing to do with the GE publicity and that they had tried repeatedly to persuade GE not to publicize their work on the project. Merriam brushed the fine points aside. He knew only that adequate credit hadn’t been assigned to his institution, the Mount Wilson Observatory.

  Mount Wilson—still the site of the largest operating telescope in the world—had stayed in the news. Edwin Hubble, with his tall stature, tweedy attire, omnipresent pipe, and affected British accent, was an ideal subject for the newsmagazines. He liked the attention, liked to be photographed with visiting celebrities, from Albert Einstein to movie stars, and the magazines liked to run photographs and stories on what they called Hubble’s “law” of the expanding universe. The favorable publicity wasn’t enough for John Merriam.

  By early 1934, after Lowell Thomas’s broadcast made the telescope a household word again, Merriam was so angry at what he saw as slights of his institution that he appointed a committee “to give study to questions touching cooperation with California Institute in furtherance of the 200-inch telescope project.” Merriam named Walter Adams as chairman; the members were Hubble and Seares from the Mount Wilson staff, Fred Wright from the Carnegie board, and “Dr. Hale if he desires to associate himself with the committee.” Merriam had a way with words that could provoke even a peacemaker like George Hale.

  Merriam pulled other strings as well, sending Arthur Day, the director of the Carnegie Institution Geophysical Labs in Washington and a vice president of the Corning Glass Works, to talk to Max Mason at the Rockefeller Foundation. Day explained that he was concerned that the original spirit of the grant for the telescope was not being observed. The casting of the mirror seemed to increase the confidence of the Caltech people, and Day worried that the telescope might not be open to anyone else. Mason, reluctant to interfere, agreed that overlapping staffs of the Mount Wilson Observatory and Caltech were a good idea—a harmless concession, since Anderson, Sinclair, and Pease were already overlapping—and that Day should keep him informed if it appears that the “spirit” of the grant was not being observed. Mason
marked his memo to the file on the meeting PERSONAL lest it fall into other hands and start a stir.

  Merriam’s committee was a sham. He knew what he wanted: “As the greater weight of authority relative to matters that touch questions of astronomical study and operation lies with the Carnegie Institution, I assume that this contribution by the Institution may at least equal in significance the use of funds available to California Institute and the contribution made by California Institute in the general scientific and engineering sense.” Merriam’s concern was the same as it had been in 1928, when he almost aborted the project: He wanted credit, especially public credit, for the Carnegie Institution.

  George Hale was too tired to fight. His health was faltering. With the mirror apparently successfully cast at Corning, the design process in full gear, the site picked, and negotiations underway for the actual observatory site, he was marshalling his energy, selecting which meetings he could attend, eager to see the project through to completion. He was frequently fatigued, more often than not confined to his dark room, reluctant to waste time on matters that did not contribute to the progress of the telescope. He had begun jotting autobiographical notes, recalling his childhood and early years.