“I really can’t stand it down there,” Westphal remarked from the safety of his office, one floor above. “I just hope none of that stuff doesn’t breed together and start growin’.”
In the case of 4-shooter, it did. Gunn walked into the Wastebasket one day and handed around a flyaway stack of papers containing photocopies of pencil drawings (he hated making blueprints; it bored him). He showed the wizards how 4-shooter would be similar to the Wide Field/Planetary Camera—a pyramid would break a shaft of starlight into four parts and reflect the light into an array of CCD cameras. “Gunn doesn’t toot his horn,” said J. DeVere Smith. “He tells you what he wants to do, and in the back of your mind you are saying, ‘That man is going for broke.’ ” Some of Gunn’s sketches of imagined machines inside 4-shooter were barely legible—he had whipped them off during all-night brainstorms. The camera would not have any control knobs. It would contain robots that would take orders only from a computer. Astronomers are clever people, Gunn figured, and would monkey with knobs. He did not want a knob going down some astronomer’s windpipe.
The wizards studied Gunn’s photocopies. “We have this little problem,” DeVere Smith would later say. “We can’t read Gunn’s writing. It’s not a question of doing what he wants; it’s a matter of interpreting what he wants.” Exercising their option to interpret, the Wizards of the Wastebasket began to build robots that could take orders from a computer. The wizards luxuriated in a certain freedom to improvise. About one third of the machinery inside 4-shooter consists of surplus parts and rehabilitated garbage. It wasn’t that Gunn planned to put used parts inside 4-shooter. They just ended up there. “I actually prefer to work with new parts,” Gunn said. “But the wizards had their own ideas.” The wizards had learned the hard way. They had learned that new parts can fail. For example, they ordered a mini aerospace motor, the size of a roll of quarters, from an aerospace company, at a cost of $125. They plugged the motor into a unit. “It slopped, it grinded, it sounded like hell,” Victor Nenow recalled. “We had ordered the wrong kind of motor.” It can easily take between eight and thirty weeks to hammer an order for one motor through a corporate marketing bureaucracy, but the wizards knew of a better way. Every lunch hour they pay a visit to a warehouse in Pasadena that sells electrical junk—a place called C&H Sales. “We know the good stuff when it hits,” DeVere Smith said. “Sometimes we pick it off the delivery truck before it gets inside the store.” At C&H Sales, the wizards found a smooth little Swiss motor in a bin full of motors. They paid five dollars for it, and it ran like a watch. 4-shooter also needed six high-precision stepper motors, some of which would drive wheels that would emplace filters in front of the cameras. Not wanting to take any chances, the wizards asked Gunn if they could go straight to the surplus bins at C&H Sales. Gunn did not mind. At C&H, the wizards found six surplus motors, for which they paid next to nothing, and the motors worked flawlessly.
Most of the common electronic components inside 4-shooter—resistors, transistors, capacitors—came out of bins either in the Wastebasket or at C&H Sales. “Some of the stuff had lived somewhere before,” Nenow put it. The advanced components for critical circuits—logic chips and gold connectors, for example—Gunn ordered new. 4-shooter needed electricity of different voltages, and so Victor Nenow built a power supply of a type that DeVere Smith describes as “one of Vic’s innovations.” It is stuffed with fuses, wires, resistors, transformers, and a muffin fan; and Nenow built it largely from nuggets that he dug up around the Wastebasket. Incorporating surplus and junk parts, Nenow built a control sensor that monitors the flow of liquid nitrogen through the cameras. It senses the presence of liquid nitrogen with a ten-cent carbon resistor. “We had to order the carbon resistor specially,” Nenow said, “because they are so cheap that nobody sells them.” J. DeVere Smith said, “Vic can take nothing and make a real piece of scientific gear out of it.” As can Smith—he took nothing and built a seismograph from it and installed the seismograph at his home, in order to watch the earthquakes come and go.
The liquid nitrogen tanks on the imaging cameras hang on piano wires that Michael Carr had bought at a warehouse. At one point during 4-shooter’s evolution, Gunn installed plastic drive belts inside a spectrograph that would be attached to 4-shooter. The drive belts cracked. Gunn consulted Nenow, who, after some thought, suggested that the belts be replaced with steel-spring drive cables from a movie projector. At C&H Sales, Nenow found some movie projector cables for fifty cents apiece.
Gunn and the wizards believed that American corporations did not want to deal with them. It seemed that nobody was interested in helping Caltech build a supersensitive camera for the Hale Telescope. In Nenow’s words, “I had a marketing manager laugh and say, ‘We don’t want your business.’ ” Many major American corporations dislike handling small orders from scientists—especially when the scientists are loners, when they have little money to spend, when they are attempting to push a technology to its limits, and who are thus building machines that exceed the comprehension of American corporations. On the customer desirability lists of many companies, gadgeteers from the California Institute of Technology rate somewhere between scrap-iron dealers and the KGB. “Nobody,” Nenow said, “wants to send Caltech any trade magazines.” Gunn said, “Trying to deal with the front-office people in many companies is almost impossible. Most semiconductor companies won’t even send you a catalog. They don’t want to send you a catalog. They know perfectly well that you are one guy in a basement at Caltech and that you are going to order one chip. Nobody is interested in selling you one part. They want to sell you a thousand. Then they take four months to deliver. You get beautiful exceptions, once in a while, largely because someone, somewhere in a company, thinks it is a good idea to deal with scientists.” Often enough, American companies would accept a small order for parts and never deliver the parts at all. Gunn felt that Japanese firms tended to behave in the same way, at least when dealing with American scientists. Without a trace of irony in his voice he said, “I gather that things are very much better in Great Britain. British companies seem much more willing to talk with scientists.”
While 4-shooter was coming together, Gunn’s marriage to Rosemary Wilson deteriorated; another lesson in entropy. He and Rosemary eventually were divorced. Gunn was dating Gillian Knapp. In 1980, she moved to New Jersey to get a job as an astronomer at Princeton University, and Gunn left Caltech to follow her east, where they were married. Jill Knapp and Jim Gunn have become prominent members of the Princeton faculty—they are no longer Calteckers. But a rabbit-warren of rooms in the basement under Jim’s office at Princeton has accumulated a lot of electronic junk that has come straight from the Dumpsters of southern California. “At Caltech,” Gunn said, “I felt myself turning into an engineer, and that was not where I wanted to go. I also wanted, more than anything else, not to lose Jill Knapp.”
Jill Knapp has an interest in the molecular clouds of the Milky Way. She grew up in Dalkeith, Scotland, outside Edinburgh, where her father worked as an industrial chemist. “He gave me a chemistry set,” she told me once, “and I am afraid I was one of those wretched kids who are forever making bangs in the kitchen.” After she found an old cardboard tube and built a telescope from it, the bangs tapered off. She introduced him to opera. She took him to see a performance of La Bohème, and by the last act, the two of them had dissolved into tears. He bought a Sony Walkman, which he carried with him when he traveled to the Hale Telescope. As he gazed at galaxies through the eyepiece of the camera called Pfooey, he listened to Rigoletto, Madame Butterfly, and Don Carlo and ate M&Ms. Soon he discovered Giuseppe Verdi’s Requiem, a Latin Mass sung for a dead novelist, Alessandro Manzoni. Gunn listened to it over and over, until the tapes hissed and crackled.
Sanctus, Sanctus, Sanctus
Dominus, Deus Sabaoth
Pleni sunt coeli et terra gloria tua
Hosanna in excelsis …
(Holy, Holy, Holy
Lord, God Almighty
/>
Heaven and earth are filled with your glory
Hosanna in the highest …)
He could lose himself in the “Libera Me,” at the end of the Mass, when the voices sang,
Requiem aeternum dona eis, Domine,
et lux perpetua luceat eis.
(Give them eternal rest, Lord,
and may the eternal light shine on them.)
The word eternal had no meaning for him—acquainted with the mathematics of the Grand Unified Theories, Gunn felt that nothing was eternal, especially temporary things such as space and time. “The universe,” he would say, “cannot survive forever.” The universe was the final lesson in entropy. The galaxies would disperse, while their stars burned down into fogs of black dwarfs and black holes. After that the basic particles of ordinary matter, the protons, would almost certainly decay into photons and electrons. The black holes would also decay. If the universe kept on expanding, it could conceivably end up as a substance called positronium, some googol of years hence. Positronium is a very cold plasma of electrons and antielectrons, bound into loose orbits with one another, each particle separated by ten million times more space than is contained in the observable universe today. Gunn had reason to believe that even the vacuum itself might be unstable, liable to catastrophic decay. Thus, like a bubble, the universe might one day pop and vanish. Not even nothing could last forever. The universe might not even be unique. Why should there be one universe? Nature never stopped at one of anything! Not when an octillion models might work. This universe might be a bubble drifting on the Planck soup, one of an infinity of other universes that were constantly erupting out of the Planck soup, by random chance. Maybe God plays at dice with universes. But when he was cocooned in prime focus with Giuseppe Verdi and a value-pack of M&Ms, staring into the mirror, where he could see with his own eye a warp in spacetime in the form of a gravitational lens shining like a pair of headlights out of the past while the Big Eye tipped westward through the tremendous unvalving doors of night, Gunn felt that there was a kind of salvation in light.
Jill Knapp worried about Jim up in prime focus. She worried that he might die in prime focus, a victim of hypothermia, sleeplessness, and chocolate shock. She believed that there were worse ways to go.
After he had settled on the East Coast, Gunn flew back to Caltech every few weeks to build a unit or to solder a circuit board himself, while gradually 4-shooter’s white cylinder accreted parts. The Hale Telescope had always been guided in the traditional manner: the astronomer held a paddle while watching a set of crosshairs and a guide star. Gunn thought that a robot could do better than the human hand, so he built a robot to guide the Hale Telescope and tucked it inside 4-shooter. An arm holding a pick-off mirror—it resembles a mirror on a dental tool—reaches into the shaft of starlight coming into 4-shooter and picks off the gleam of a single star: the guide star. The mirror shines the guide star’s light through a miniature telescope—a telescope big enough to hold one star. The telescope focuses the starlight onto a spinning razor blade. The razor blade chops up the light, making the star seem to blink. A small computer, which Gunn built by hand, watches the blinking star. The computer interrogates the guide star. It asks: Is this star drifting? Where?—and orders the Hale Telescope to correct its motion accordingly. The razor blade is a Wilkinson Sword that Gunn bought at a Rexall drugstore in Pasadena. He broke the blade to the right shape with a pair of pliers and fastened it in place with a dab of glue.
Most of the electrical wiring inside 4-shooter is Teflon-coated wire, a type often used in aerospace applications, because it does not easily melt. It is expensive stuff. The wizards bought their Teflon wire at C&H Sales for a half a cent a foot. Since 4-shooter contains something like one mile of wiring, the savings added up. Victor Nenow saved a bit more by stripping Teflon wires out of dead computers. When Nenow had nothing better to do, he broke open dead computers. While he hacksawed a circuit board and nipped wires out of it, he toasted cheese sandwiches in a toaster oven that he had built for himself—another of Vic’s innovations. He showed me how it worked. He slid a cheese sandwich onto a heat sink, under a grill. He turned a dial. The oven glowed. He cranked up the dial. A blinding light flooded from the toaster oven—the heating elements were four movie studio lamps. He put on a pair of sunglasses and said, “That’s only half voltage.” He had marked the dial to avoid damaging his eyes or igniting the cheese. “I used to electrocute hot dogs,” he said. “Put a clip on either end of the hot dog and send a current through it. The hot dog would steam and split. It worked great.”
The administrators of the California Institute of Technology have not the slightest official idea of how to classify the Wizards of the Wastebasket.
“Our job titles?” DeVere Smith wondered. “Vic!” he called out, “what are our job titles?”
Nenow’s answer filtered around a corner: “We’re garbologists.”
J. DeVere Smith, who etched and soldered many of 4-shooter’s circuit boards, is a tall, white-haired man with large, deft hands. In 1930, Smith opened a radio repair shop in Los Angeles that he deemed Advance Radio, although he admits that he did fix the occasional Victrola. Advance Radio also handled meat—delivered it to grocery stores (a profitable sideline). Smith eventually got out of meat and radios and moved into television repair, although he continued to call his shop Advance Radio. He finally sold his shop, and as he puts it, “I retired and came up to Caltech. When they get through with me here, they’ll have to bury me.” Shortly after Smith arrived at Caltech, he and Victor Nenow utilized the science of garbology to build the electronic systems for four mass spectrometers that Caltech geophysicists used to analyze moon rocks gathered during the Apollo lunar landings.
An astronomer walked into the Wastebasket and threaded his way to DeVere Smith’s corner. “Hi, DeVere,” he said. “I need a knob.”
“You’ve come to the right place,” DeVere said. He pushed some things around his workbench until he had found a knob. He said, “How’s that?”
The astronomer inspected it. “DeVere. This is trashy. I want a shiny knob.”
J. DeVere Smith slid open a legal-sized filing drawer by his knee. The drawer was absolutely packed with knobs. “You want two?” he asked.
Smith is a miner of dumpsters. “You’d be amazed,” he said, “at what you can find in dumpsters.” I was amazed. One day I was hanging around the Wastebasket, talking with the wizards, when a geologist walked in.
Nenow held up a bag and pulled a ceramic cylinder out of the bag. He handed the cylinder to the geologist. He said, “I thought maybe you could tell me what this is.”
The geologist turned the object over. “Oh! It’s a proton precession magnetometer.”
“A what?”
“It measures the strength of the earth’s magnetic field. This is a pretty good sensor. Where did you get this?”
“It wound up in a trash bin. DeVere found it.”
“You mean somebody threw it out?”
“Oh, sure.”
“Can I keep it?”
“Sure.”
“Thanks. These things are fifty thousand dollars each.”
Exit geologist, carrying sensor.
Richard Lucinio, the digital wizard, preferred to work at night. He would leave his home in Topanga Canyon in late afternoon, with his two dogs, and drive to Caltech. He designed the logic boards inside 4-shooter. These circuits contain chips used to control both machines and other chips. Lucinio’s logic boards, for example, can order the CCDs to dump their electrons into the amplifiers. Lucinio would collect a robot from one of the other wizards as the wizard was leaving work. He would hook the robot into a logic board and play with it all night, trying to get a motor to start or some wheels to turn. He often worked with Barbara Zimmerman, who wrote the software program that controls 4-shooter. When they could not get a robot to work, they would have a glorious time of finger pointing: “It’s the hardware!” “Naw, it’s the software!” They would try various command
sequences on the robot, cajoling it. Eventually the robot would wake up. Jovanni Chang (who soldered and tied together many of the cable harnesses inside 4-shooter) liked to watch these miracles. As he described it, “We would hear a whee! Chunk! Something would happen—a motor would spin, a trapdoor would open. It was like a flower blooming.”
Gunn built a set of amplifiers, to process electrons coming from the CCD chips. Some of the circuits he soldered himself, and some he gave to DeVere Smith to solder. On a Saturday afternoon in September 1983, Gunn and Michael Carr rented a Ryder truck and stuffed 4-shooter into it. Carr took the wheel. Four years of Carr’s life were in the back of that truck, and he drove the Ryder at thirty-eight miles an hour down Interstate 210. Gunn asked to drive. Carr exchanged places with Gunn, which he immediately realized had been a mistake, because Gunn put the Ryder in the fast lane and mushed the accelerator. (“Gunn was overanxious,” Carr thinks.) When the Ryder headed up the switchbacks of Palomar Mountain, Michael Carr experienced a touch of pure fear. One hour after dusk they had loaded 4-shooter into the Hale Telescope, and 4-shooter’s chips had begun collecting photons. A nearby dwarf star contributed first light to 4-shooter, and then Gunn aimed it at galaxies, running for the deep range. “He had that little Gunn grin on his face,” Carr remembered. “I just hope I never lose sight of Jim Gunn.”
Not long after its installation in the Hale Telescope, 4-shooter entered unexplored spacetime. The camera exposed images of galaxies that were deeper in the universe than any galaxies that had ever been seen before. These galaxies were different from nearby galaxies. They seemed to be bluer in color; saturated with hot, young stars. 4-shooter was seeing galaxies in an earlier cosmic epoch. When 4-shooter was installed in the Hale Telescope, the Hale became at least one hundred times more powerful than it was when it was first built. In order to equal the light-gathering power of 4-shooter coupled to a two-hundred-inch mirror, George Ellery Hale would have had to build a telescope with a mirror at least one hundred and sixty-six feet across—a mirror the size of a parking lot.