The Perfect Machine (35 page)

When the surface of the molten glass leveled, the overhead crane picked up the annealing kiln. The crane order had been for a special “slow moving” model, but it still raced across the track, the kiln cover hanging at a rakish angle. A glassmaker jumped onto the kiln to balance it, then had to hang on as his weight sent the hoist rolling down its track, high over the floor. It took three husky millwrights finally to stop the swinging load and stabilize it. Plenty of hearts missed beats, but the Keystone Kops antics miraculously didn’t harm the disk.

When the kiln cover was in place, the disk was consigned to two months of controlled cooling. It emerged in September 1932 intact, and with residual stresses as low as McCauley’s calculations had predicted. The custom shop at Corning was able to grind the back to correct the rib pattern that had been marred by the core that had broken free during the molding, and the Observatory Council accepted the disk. There were no plans or facilities to grind and figure the disk in Pasadena—Anderson’s plan was to grind and figure the auxiliary mirrors after the bulk of work had been completed on the two-hundred-inch primary mirror—so McCauley was told to store it at Corning while they went ahead on the next step, the 120-inch disk that could be used as both a practice run for casting the primary mirror of the telescope, and could also be ground as a flat mirror for use in testing the two-hundred-inch mirror.

In less than one year of work, Corning had gotten further than GE had in three. Optimism was so high that McCauley persuaded the Observatory Council that they could save money and time if they skipped a few steps and built the remaining equipment at Corning large enough for both the 120-inch and the two-hundred-inch disks, rather than going through the stages to build both a 120-inch and a two-hundred-inch annealing oven. With budget approval from Pasadena, McCauley got permission to have one of the large melting tanks in the Corning factory dismantled in November 1932. The site of the former tank was then covered with a steel floor, at the height of workbenches at the adjoining tanks. Two circular holes, each twenty feet in diameter, were cut in the floor on a north-south line, and rails were installed in the cave underneath the floor. A heated igloo, big enough for a two-hundred-inch disk, was suspended over one hole, with an annealing kiln over the other. Beneath the floor, a table large enough to hold the
mold for a two-hundred-inch mirror was constructed on a 60-ton locomotive screw hoist that could travel on the rails.

After watching the heavy kiln swing out of control over a disk, everyone had concluded that moving the disks on rails was safer than swinging heavy equipment over the freshly molded disk. The screw hoist was slow (two inches per minute) but reliable. The new arrangement also simplified the hookup of electrical connections to the igloo and the annealing oven. A separate room for the electrical transformers was constructed on the floor above and just west of the annealing kiln. To complete the arrangement, McCauley had a draftsman design trolleys for the casting oven to move it from its position over the mold, so that mold makers could have complete access to the mold before the casting began.

The arrangements looked good on paper. They took months of work by draftsmen, engineers, millwrights, carpenters, electricians, masons, and the outside suppliers who were called on for the structural steel, insulating brick, electrical controls, and new ladles that would hold 750 pounds of molten glass each. The depression economy had hit much of the glass industry especially hard, so suppliers were quick to produce the needed materials, and Corning was able to assign dozens of men to work on the facilities McCauley needed. It looked as if they would move to the next step of the casting program ahead of schedule.

Then disaster struck.

Harrison Hood ran Corning’s laboratory of glass scientists, who had the job of testing glass formulas under extreme conditions, and concocting new glasses to meet new challenges. A chemist in Hood’s lab, Dr. M. E. Nordberg, had gotten hold of a chip of the glass that had been used for the successful sixty-inch mirror. One test in particular intrigued him.

Nordberg found that in 1923 a researcher named R. D. Smith had noted in a laboratory logbook that annealed 702-EJ glass was more soluble than unannealed ware. Nordberg decided to follow up on Smith’s finding with a series of experiments. When he tested pieces of Pyrex that were heat treated, then rapidly cooled, the samples were fine in solubility tests in water or acids.

But when Nordberg put the chip from the annealed sixty-inch disk into water, it was far more absorbant than any of the test materials. The period of long annealing and slow cooling clearly had affected the glass in ways they hadn’t anticipated. He tried more tests and concluded that the surface probably was too unstable for polishing with moistened or water-flushed rouge or for frequent mirroring with a solution of silver nitrate. And that disk had only been heat treated for two months in the annealing oven. What could they anticipate for the surface of a two-hundred-inch mirror that would be annealed for ten or twelve months?

McCauley went over the test results and could find no flaws with Nordberg’s results. The 702-EJ glass he had been using was not suitable for telescope mirrors. The 60-inch mirror made from the material was a sponge. It was close to Christmas 1932; dozens of workers had dismantled a major melting tank and were constructing the facilities to mold the 120-inch and two-hundred-inch mirrors. He had just sent a revised and slightly accelerated schedule to Pasadena.

He couldn’t bring himself to admit to the Observatory Council that the glass formula was no good and that Corning didn’t have a glass with the proper characteristics for big telescope mirrors.

McCauley began the new year at a dead end.

16
Good News

The good news from Corning about the sixty-inch disk, and the successful tests of the trial Pyrex disk at the optical labs, inspired a flurry of work on the telescope design. Hale and his colleagues were confident that Corning could cast a Pyrex mirror disk, with a cored and ribbed back. The disk would be supported on ball-bearing supports, like those Pease had engineered for the one-hundred-inch telescope. The geometric pockets in the back of the disk would house a system of compensating supports to maintain the shape of the mirror as the telescope moved. Within those basic constraints, Porter and Pease were busy sketching designs for the telescope. The astronomers on the advisory committee, and others with strong opinions, were busy imposing requirements on the design.

The astronomers who focused their research on galaxies and other deep-space objects, like Hubble and Humason, wanted no limits on the orientation of the telescope. The one-hundred-inch, because of its mounting design, could not point to the circumpolar region around the North Star. Studies with the sixty-inch telescope showed this region of the sky to be particularly rich in galaxies. The deep-space astronomers also wanted the best possible facilities for spectroscopy and direct images at the prime focus of the telescope, which would have the greatest light-gathering ability for faint imaging.

The astronomers whose research centered on stars were interested in spectroscopy, which demanded an extremely stable temperature- and vibration-proof room at the long Coudé focus, where they could make detailed spectrograms of individual stars. In brainstorming sessions the astronomers came up with other ideas for the use of the telescope. The Cassegrain focus behind the main mirror would be ideal for some kinds of imaging and spectrograms, and someone suggested that instruments could be mounted at the Nasmyth foci on either side of the hollow declination axis. Each position offered advantages and
disadvantages; in the interest of long-term versatility, no one was willing to abandon his favorite.

Normally it would take hours or even days to change the instruments in use on a big telescope to a different focus position. If the telescope had been used for spectrograms of the red shifts of distant galaxies at the Newtonian (prime) focus, the change would require moving new mirrors into position and realigning them before the telescope could be used at the Coudé or Cassegrain focus. In setting the research programs for the big telescopes, the time-allocation committees would try to string together programs that used the same focus and equipment: deep-space research during the dark of the month, when the moon was down; spectrographic studies on nearby stars when the moon was up.

But weather and the general seeing conditions didn’t always pay attention to the plans of the astronomers and the allocation committee. The telescope and observers could be prepared for an evening of deep-space observing at the prime focus only to discover that the seeing wasn’t good enough to make use of the light-gathering ability and speed of the telescope. The evening might be good enough for less demanding use at a different focus, but if it took six hours to change the instruments and focus of the telescope, the evening would be lost.

In the brainstorming sessions the astronomers asked if the telescope could be switched from one focus point to another in minutes rather than hours, so the balance of the night could be put to profitable use. What if the various auxiliary mirrors that would bounce the light path to the different foci could be mounted so they could be flipped into place with remote controls and still be rigid enough and well-enough aligned to preserve the optical quality of the telescope? The ideas sounded like a fantasy list. No one yet had an idea how to maintain the rigidity of the tube and mount of the basic telescope.

Depression pessimism actually encouraged the pie-in-the-sky ideas. If the two-hundred-inch telescope had once seemed like another step in a progression of telescopes, it was now beginning to seem more and more like the final step, the last big telescope for a very long time. This one had to do it all. No one used the term, but with so many demands to fulfill, the big telescope had to be a
perfect
machine.

George Hale was probably the only one who could have mediated the conflicting demands and requirements of the astronomers and engineers. When he was well he was lucid, quick, and judicious. His experience and prestige as the father of large telescopes were often enough to settle disputes. Unfortunately his nervous condition and its manifestations had grown more intense with the years. The periods when he was free of his demons were shorter and rarer, and the instances when Miss Gianetti would excuse him from meetings were so common that significant decisions were made without him.

John Anderson, as executive director of the project, was a member of each of the committees. As the work at Corning progressed, he was
involved more and more in the day-to-day decisions on the planning for the mirror, and in the construction of the optical lab where the mirror would be ground. Anderson was an easygoing manager. He didn’t fire off memos the way George Hale had, and he made fewer demands on the other staff members. The lack of a single strong guiding force may have been a blessing. Porter, Pease, and some of the young engineers on the Caltech faculty were free to play with ideas that might have been dismissed as too radical by a more rigorous manager.

One obvious question was how the observers would get to the various observation positions on the telescope. The Newtonian focus, used for deep space work on the sixty- and one-hundred-inch telescopes, had always been a problem. The eyepiece and instruments at the open end of the telescope tube were high off the ground and rotated with the telescope so that observers were constantly fighting fatigue and cramps as they scrambled from one position to another. The Cassegrain focus at the bottom of the mirror was less of a problem, but the sheer size of the two-hundred-inch telescope meant that the Cassegrain focus could be high off the ground when the telescope was pointed at an area of the sky close to the horizon. The Coudé focus, in a room beneath the telescope, was easy for observers, but the telescope had to provide for a light path that would reach the room no matter what the orientation of the telescope. If the telescope could dip low enough to see the polar region, it would need not one but two different paths of mirrors to bounce the light down into the fixed room.

The telescope was so large that even the secondary mirrors that would change the light path to the various instruments were heavy and unwieldy. Moving heavy mirrors in and out of the telescope tube each time the light path was changed would be cumbersome and dangerous. No one was sanguine about the idea of regularly suspending heavy equipment by crane over the priceless primary mirror of the telescope.

Gradually Anderson and Porter began looking at the scale of the telescope—the latest estimates were that the primary mirror would weigh close to twenty tons and the mounting more than five hundred—as an advantage. The size of the telescope tube and the stability of the immense mounting meant that the auxiliary mirrors for the Cassegrain and Coudé foci could be mounted in a permanent cell in the middle of the telescope tube, with gear-driven mechanisms to swing and lock them in place as needed. The mirrors and cell, approximately six feet in diameter, would rob the central portion of the two-hundred-inch disk of only one-ninth of its area, an acceptable sacrifice. And since the telescope could tolerate a six-foot cell, by extending the cell on the other side of the compartment that held the swinging mirrors, they had room for an observer to ride
inside
the telescope to use the prime focus.

Putting the observer inside the telescope had obvious advantages. With the observer at the prime focus, they would need no diagonal Newtonian mirror or the awkward platforms for an observer outside the tube.
The corrective lens and plate holder for the prime focus would be in one cell, separated physically and mounted independently of the cell that held the observer, so that vibrations from the observer’s movements would not affect photographic plates during exposures. The observer’s cell would ride circular rails and be adjustable at 22.5° intervals so the observer would always be upright even as the telescope turned. There would be room for a set of controls for the telescope and for telephone equipment to allow the observer to speak with the night assistant far below him on the observatory floor. Astronomers who saw the early plans were fascinated by the idea, although more than a few raised the obvious questions: How would they get up there? Those who remembered the cold nights and aching bladders of long winter exposures worried even more how they would get back down.