The Perfect Machine

by Ronald Florence

  It wasn’t that they were too far from the East Coast and the markets. California was no longer a hothouse hybrid. By 1928 the movies had made the once-strange images of Los Angeles an alternative norm of American culture, a mythical world that people knew from the silents even if they had never taken a trip out west. The reality of California was frighteningly close to the cinema images. The movie and aircraft industries fueled an exploding industrial and agricultural economy. Growth seemed to reach in every direction at once. Buildings, neighborhoods, even whole cities seemed to spring up overnight. From nowhere universities leapt into a national and world prominence that had taken East Coast institutions hundreds of years to achieve. Where once a rising young scientist would only have sought a position at one of the famed eastern universities, now there were institutions in California with top-notch faculties, with more money to offer for research and laboratories, and with the incomparable advantage of newness. Perhaps it was only illusion, but to a generation of bright young scholars, the California schools seemed more open to new ideas, less hidebound, than schools elsewhere. They seemed like places where a young scientist or engineer could make his mark unfettered by the constraints of precedent and tradition.

  In 1928 men like J. Robert Oppenheimer and Ernest Lawrence, who had their choice of institutions, gravitated to California, Lawrence to the University of California at Berkeley, and Oppenheimer to a joint appointment at Berkeley and Caltech. Two years later the first meeting of the National Academy of Sciences ever held west of the Mississippi met in Berkeley. At the meeting Lawrence unveiled his first cyclotron, starting a chain of developments that would ultimately surpass even the two-hundred-inch telescope as big science.

  Explosively, in less than a decade, Caltech established a reputation as a first-rate school for science and engineering. Powerful patronage in Pasadena and elsewhere in Southern California came forward with generous endowment and capital support for the new school, which multiplied the drawing power of the prestigious initial faculty appointments like Millikan and Noyes. The academic center of gravity of the United States was still on the East Coast, but in fields like aeronautics, the local aircraft industry in Santa Monica, Burbank, and San Diego made the Southern California location a magnet for top faculty. Engineers like Theodor von Karmann joined Caltech as senior professors or department heads. Their availability as consultants drew the attention and support of the aircraft industry, which in turn attracted bright younger scholars and students who could look forward to jobs at Douglas, Lockheed, or Convair after graduation from Caltech.

  Even before the two-hundred-inch telescope project was announced, rumors of the telescope were enough to draw graduate students to Pasadena. A young Berkeley graduate named Olin Wilson thought about graduate study in astrophysics at Caltech before the school had a single faculty member in the field. Harvard, Princeton, and the University of Chicago were the famous departments for astronomy, but the chance of someday doing research on the big telescopes at Mount Wilson, or the unbuilt two-hundred-inch telescope, was an irresistible drawing card.

  For engineers, too, the chance to work on a once-in-a-lifetime project was a powerful magnet. Engineers and opticians and mechanics sought jobs at the fledgling institution in the hope that somehow they could be part of the instantly famous project. Few of these engineers had any experience with telescopes, but engineers liked to see themselves as problem solvers, men of few words, ready to do battle with their slide rules and graph paper. Men who could do calculations of induced drag on airfoil sections or wind-stress calculations for tall buildings could do the same for an observatory dome, and men who had calculated the bearing loads for a battleship gun turret could also calculate the loads for a telescope mount.

  There wasn’t much of a campus yet at Caltech, and Pasadena, the newcomers were soon to discover, wasn’t the California of the movies. By the late 1920s, motion pictures had become the leading industry in Southern California, with aircraft factories a close second. In Pasadena, the leading industry was clipping coupons from bonds. The assessed wealth in Pasadena in 1929 was $186 million, which would have been high for a city twice its size. Even as Los Angeles grew by leaps and bounds, Pasadena remained an isolated small city, nestled against the foothills of the San Gabriel Mountains.

  Upton Sinclair lived in Pasadena during his losing bid for the governorship of California, but the presence of the old socialist warhorse didn’t change the determinedly conservative politics of the city. Even the relatively staid campus of engineers and scientists at the California Institute seemed a radical intrusion to some. The Reverend Robert B. Schuller, successor to a long line of Southern California evangelists that included the infamous Aimee Semple McPherson, galvanized opposition to the ungodly threat of science at Caltech with revival meeting speeches on topics like “Evolution Unmasked” and “The Mark of the Beast.” Pasadena voted three-to-one in favor of a statewide ballot measure ordering the King James version of the Bible placed in public schools. But for strong opposition by Northern California voters, the measure would have passed statewide.

  Yet by 1928 the evangelists and even Aimee Semple McPherson’s scandalous disappearance and reappearance were but brief interruptions to the steady news of the stock market that dominated headlines everywhere, including Pasadena. Despite periodic “corrections,” the rise in the market seemed inexorable. There were fortunes to be made, the headlines cried. Old money, fearing the challenge of new, rose to the lure of the market. Even a bastion of conservatism like Pasadena wasn’t exempt from the appeals that drew money from fixed-income investments to the rising market, where the mysterious yeasts seemed to renew themselves weekly. Men like Henry Robinson, of the board of trustees of the California Institute, who had personally promised to capitalize an endowment for the two-hundred-inch telescope, were soon heavily committed in the market.

  The astronomers and engineers, in makeshift offices on California Street, at the Santa Barbara Street offices of the Mount Wilson Observatory, and in the library of George Hale’s solar laboratory on Holladay Road, worked on, their future assured by the grant from the IEB. They were too busy for the stock market. They had a perfect machine to build.

  11

  Hope

  Nothing went right in West Lynn.

  It all seemed so promising when they started. Back in March 1928, Elihu Thomson’s assistant, A. L. Ellis, had shown Hale a beautiful piece of hard, transparent fused quartz. Even under magnification it was free of striae that might cause strains or distortions in a mirror. The test results when the material was heated were spectacular. That summer Thomson showed off samples of the clear quartz to a visiting delegation from the AAS, including Harlow Shapley. Before long major observatories in the United States and Canada were asking about quartz blanks for astronomical mirrors. Given Thomson’s reputation as a genius of engineering and production, the mirrors seemed assured.

  Elihu Thomson was addressed as “Professor,” although he was more a shirtsleeves scientist than an academic. He had been offered the presidency of MIT in 1897 and declined. In 1919 he accepted the position, but only for two years as an acting president, before he returned to his beloved laboratories. Like traction power, electric welding, or fused quartz, he saw MIT as problems that needed solutions. Two years was long enough for Professor Thomson to solve most problems.

  He was at home at GE because the company had the same attitude. Although they conducted substantial research programs, the research was directed toward solutions, and specifically toward production. GE had converted the tinkering of geniuses like Thomas Edison and Elihu Thomson into lighting and electrical equipment, trolley traction drives, appliances, industrial diamonds, electric motors, monitoring and metering devices—products that ranged from consumer goods to massive industrial systems. GE’s immense factories, like the River Works in Schenectady, supplied the world. Between the reputation of men like Elihu Thomson, and the sheer vastness of GE’s manufacturing facilities, the company epitomized
the technological optimism of the 1920s. The company, and the public, believed that GE could accomplish anything it tried.

  In 1904 Thomson had already patented a process for molding quartz. He would pour fine quartz sand into a mold in a high-temperature vacuum furnace, seal the furnace, and slowly raise the temperature until at 500° the quartz sand would “explode,” vaporizing any contaminants. At 1400° the quartz would begin to turn viscous. When the furnace reached 1700° (approximately the melting point of platinum) the quartz would fuse into the pattern of the mold. During the process powerful vacuum pumps sucked the air out of the furnace to remove gaseous contaminants and eliminate as many air bubbles as possible from the quartz.

  But even with the most powerful vacuum pumps available at the West Lynn laboratories, the blanks emerged from the mold filled with tiny bubbles. The bubbles did not interfere with the thermal qualities of the quartz or weaken it, and for large telescope mirrors, the lightness of a bubble-filled blank might even be an advantage, reducing the load on the telescope mounting. But the surface of a bubble-filled blank could not be polished to the fine figure required for an optical mirror. True, a pock in the surface would only reduce the light-gathering capacity of the mirror by a tiny percentage of the total area. But there was no way to achieve a fine figure on a pocked surface, because the pocks would ruin the grinding and polishing equipment, and the resulting surface would scatter instead of focus light from faint objects. Thomson’s challenge was to find a way to coat the fused silica that emerged from their molds with a layer of pure silica fine enough to take an optical surface.

  The professor tended to lose patience with experiments once he understood the process, and after a bout with gout and asthma in 1926, he had withdrawn from day-to-day activities at the laboratory. He turned the laboratory experimentation on fused quartz over to Ellis.

  If Ellis wore a jacket and tie and even a vest in the laboratory, this was more a reflection of the formality of business and science in the late 1920s than of personal style. On the project he wasn’t afraid to roll up his sleeves and plunge into dirty work that front office scientists would shun. He was Thomson’s kind of man.

  Ellis started by standing short rods of clear quartz on end on the surface of the molded quartz, and reheating the disk in a furnace. The intense heat of the furnace melted the rods into a clear coating, but the resulting surface was marked with striae in the shape of the rods. When Ellis substituted a mosaic of sheets of pure quartz for the rods, the sheets fused in a patchwork of striae. Using ideas of his own and suggestions from Thomson, Ellis tried variation after variation of the shape of the rods and the pattern of the overlapping sheets. Nothing he tried produced a surface without striae. Ellis wasn’t an optician, but he knew that any inconsistency in the surface material meant strains that would ultimately affect the stability and optical performance of a mirror.

  He was close to exhausting the variations of sheets and rods on the surface of the disk, when an engineer named Niedergasse reminded him of an earlier experiment at the West Lynn laboratory. On a very different project, an early effort to create artificial sapphires and rubies, Elihu Thomson had introduced finely ground refractory substances like alumina into a high-temperature flame. Nothing else seemed to work with the quartz, so with Thomson’s approval, Ellis tried the technique for the fused-quartz project.

  He began by designing a burner that would blow pure crystal quartz powder, ground to the consistency of flour, into an oxyhydrogen flame at 3000°F. In a furnace the burner produced a sleet storm of quartz that fused to the base disk in fine layers, like ice coating trees after a winter storm. The process was painstakingly slow, demanding huge supplies of hydrogen to fire the torch, enormous quantities of super pure ground quartz, and infinite patience. By the end of the summer of 1928, Ellis finally sent an eleven-inch fused-quartz disk to Pasadena for testing.

  The blank looked superb. At the optical laboratory on Santa Barbara Street, Anderson subjected the disk to violent heating “which no glass could possibly bear.” After the tests Hale and Anderson proudly announced: “Its performance has been marvelous, and although this disk was not annealed at all, it has already returned to its original figure, leaving no doubt as to its internal quality.” Enthusiasm was high, but Ellis knew the blank proved little. The surface of the test disk had been built up in a small gas-fired furnace with a single burner operated entirely by hand.

  The troubles began when Ellis and his staff tried to make bigger disks. Ellis switched to a larger, electrically heated glazing furnace and trained a crew in the use of the high-temperature oxyhydrogen torch. The heat and the yellow glow of the molten quartz were so intense that even with heavy insulated suits and dark goggles, the men had to work at a distance from the furnace, using long rods to maneuver the burners. The opaque protective glasses and the awkward long rods made it difficult to judge how much material had been deposited in any one spot as they moved the burners over the disk. The first efforts with the new process, Ellis reported, looked “much like the Rocky Mountains.” He fiddled with where the men stood and how they operated the spraying equipment, to no avail. He couldn’t produce consistent disks. Ellis finally shut down the furnace and went back to the smaller gas-fired furnace for more experiments.

  Spraying the surfaces of the disks by hand wouldn’t work. What Ellis needed was an apparatus that would hold the spraying equipment a constant distance from the molded blank and move it over the surface of the blank uniformly and at a constant speed. He had the lab shop fabricate a burner and piping to carry the fuel, oxidizer, and powdered quartz. The metallurgists experimented until they found a nickel alloy that would work in the nozzle of the burner. To keep the brass of the oxygen and hydrogen pipes, and the burner itself, from melting in the intense heat of the furnace, Ellis added more pipes to his apparatus for cooling water, until the device began to take on the appearance of a Rube Goldberg drawing.

  After months of tinkering he had the workmen fire the gas furnace to spray another disk. The process seemed to work, but the disk emerged from the furnace with the surface pocked with bubbles. It took weeks of sleuthing before Ellis traced the bubbles to iron particles and porcelain contamination in the supposedly pure quartz. More detective work identified the culprit as the ball mill that had been used to crush the raw quartz to a flourlike powder. The pages of the project calendar began turning by months instead of weeks.

  Ellis confidently reported that the problems were isolated and fixable, but by the end of 1928, he and his staff still hadn’t produced a mirror blank that could be ground to an optical surface. In Pasadena the opticians were impatient for disks to grind. Pease and Porter needed to know whether the mirrors could be cast before they went ahead with design plans. Hale and the Observatory Council needed to know whether they could build the telescope they had promised.

  While he waited for the ball mill to be rebuilt, Ellis designed a new mechanical support capable of moving three burners together inside the furnace. The new apparatus was installed in the large furnace, and everything was readied to attempt a second eleven-inch disk. The machinery they were using for this disk, Ellis reported to Pasadena, could be scaled up to build mirror disks large enough for use in the telescope, at least as auxiliary mirrors.

  When everything was finally in place to fire the electric furnace, one of the three workmen trained in the use of the burners sprained his back and was laid up for a week. As soon as the man recovered and returned to the lab, he and his assistants got the flu. Two more weeks were lost. The project seemed cursed.

  Finally, at the beginning of the new year, Ellis fired the large furnace. The roar of the oxyhydrogen burners was deafening. Through the peepholes, the inside of the furnace glowed an intense, hellish yellow. The process consumed tank cars of hydrogen and oxygen at a ferocious rate as the layers of quartz were slowly fused onto the base of the disk. After seventy-two hours of continuous spraying, during which he had gotten little sleep, Ellis realized that the coating
was being deposited unevenly, with lumps of fused quartz in some areas and bald spots in others.

  He tried to adjust the flow to the three individual spray burners, but the only way he could reset the valves was temporarily to stop all work, partially cool the furnace, relieve the vacuum inside, and finally lift the cover of the furnace enough so the workmen could get to the equipment inside. After adjustments the furnace was resealed and fired, and the entire crew waited until the pumps drew down a vacuum before they could begin spraying again. Even the most cynical pessimists hadn’t predicted the process would be this complex or slow.

  The only aspect of the program that was ahead of schedule was the billing. Every piece of new equipment, each delay, and each breakdown increased the cost. When Anderson asked about the huge bills that Ellis was sending to Pasadena, Ellis provided detailed breakdowns, documenting every cent. The lab equipment, all experimental and specially fabricated for this job, was expensive. The process needed lots of men and huge quantities of fuel and supplies. The original agreement with GE, made in the summer of 1928, had been for one year of work. Six months had gone by, expenditures were already approaching the figure projected for the entire series of mirrors, including the two-hundred-inch mirror, and Ellis still hadn’t shipped a mirror that could be used in a telescope.

  Another month passed before Ellis finished a second eleven-inch disk. It was the first complete test run of the molding and surfacing process they hoped to use for the production of telescope mirrors. When the furnace cooled and the workmen lifted the cover, they found that the surface of the disk had cracked, probably from the repeated thermal shocks when the furnace had been partially cooled to service the burners. Despite the bad news, Ellis and Thomson hadn’t lost confidence: “We have not encountered any serious obstacle,” Ellis wrote to Anderson, “and are convinced that the production of the large mirror is merely an engineering problem.”