This article appeared in Science 80, the consumer sister of the journal Science, in October 1982. It helped us win that year's National Magazine Award for General Excellence.

To Catch a Comet

When the prize is the first glimpse of Halley, astronomers are willing to gamble.

By Terry Dunkle

Last February, when everyone else had suspended the search for Halley's comet, Harlan Smith, of the University of Texas, quietly shuffled the schedule at McDonald Observatory to allow his team another attempt.

"Oh, they won't find it," Brian Marsden, the Harvard-Smithsonian astronomer who verifies comet discoveries, told me on the phone. "It's too faint."

Nobody had seen the famous comet since the summer of 1911, when—after hanging like a bright, arrested meteor in the heavens the year before—it had shrunk to a pinpoint and receded beyond grasp of the largest telescopes. It was supposed to have climbed out beyond Neptune by the end of World War II, turned around, and begun falling toward its 1986 return. In February it was a billion miles from Earth. Marsden believed it to be 26th magnitude—fainter than a candle flame at 50,000 miles. "Nobody is going to pick it up for at least a year and a half," he said.

Because other experts had said the comet might be 24th magnitude—six times brighter—many astronomers had been trying anyway. Attempts had been made at Mount Palomar in California, Kitt Peak in Arizona, Mauna Kea in Hawaii, Siding Springs in Australia, and many other sites where astronomers were sneaking plates while supposedly photographing other objects. "I don't know anybody who isn't looking for Halley," one told me.

The most serious contenders had hoped to succeed before spring, when the comet would disappear behind the sun. When it emerged, in October, it would be 70 percent brighter and likely to draw more competition.

By January, a few had reached 24th magnitude but found nothing. Most of them considered February their last chance. February was a bad month, however. The comet lay in a part of the sky overlaid with thousands of minor planets, called asteroids, which can make dots on photographic plates that look exactly like Halley. The only way to tell the difference is by taking a succession of plates and measuring how fast and in what direction the dots move. In February, the difference would be slight.

Even more discouraging was the news from Mount Palomar. In December, the largest telescope in the free world, focusing light from a 200-inch mirror into a new camera that was sensitive enough to record individual photons of light, had gone a full magnitude deeper than the 24th and still failed. Hearing this, most Halley hunters though it wiser to skip February and put their money on fall.

"They're probably right," said Smith on the phone a few days before his attempt, "but we still think it's worth a try. That Palomar camera had a field of view less than a hundredth the size of the full moon. It's fine if we know Halley's location precisely, but we don't. Comets, you see, are jet-propelled. Besides falling around their orbits, as other bodies do, they can speed up, slow down, or wobble by expelling gas. We're prepared to believe Halley is a few days off schedule, and we happen to have a camera that covers a bigger piece of sky: the Griboval Electrographic."

The Griboval is very sensitive," said J. Derral Mulholland on Thursday, February 18, the first night of the attempt. "It can make a 30-inch telescope go deeper in one hour than the 200-inch goes in a whole night with a conventional camera. You can imagine what it will do on this instrument."

We were riding the elevator up into the dome of McDonald's largest telescope, a 107-inch. The observatory sits on a 6800-foot peak in the Davis Mountains, in west Texas. Mulholland, a short man wearing bifocals and a trim gray beard, had tucked his logbook under his arm and was pulling on his gloves, which had the fingers cut off. "Easier to write with," he explained.

The elevator lights went off, and the doors opened. We stepped out into a darkness full of the sound of electric fans. A woman's voice echoed faintly above the noise. A red light bloomed up ahead, revealing a pair of warmly dressed figures standing beneath a telescope as big as a silo. It climbed three stories to look out a vertical slit, where Orion the Hunter, a bright rectangle with a belt of blue stars slanting across the middle, stood frozen in a moonless sky. Outside, it was so dark that one couldn't actually see the Davis Mountains; they were merely a starless, wavy-edged void at our feet.

"How goes it?" asked Mulholland when we had climbed the metal stairs to the observing platform.

"We have a super photocathode," said Elizabeth Bozyan, her breath making a red cloud in the lamplight. She was chunky and round-faced and wore a purple ski cap whose tassel jiggled when she talked. "We could get 25th magnitude in an hour if it weren't for the damned sky."

"Transparency problems," said Edwin Barker, the team leader. He had a boyish face and drooping eyelids. The sky, he explained, was filmed with cirrus clouds. Unless it cleared, they would have to spend their first night photographing brighter comets.

Because Smith had remained in Austin on business, Barker was calling all the shots this time. A native of New Mexico, Barker had attended a one-room school before earning degrees at the New Mexico, Kansas, and Texas state universities. During more than a decade on the McDonald staff, he had earned a reputation as a quiet, level-headed astronomer who was good at spectroscopy—reading the chemistry and physics of celestial bodies by analyzing their colors. He had used it in the early 1970s to measure the atmospheric pressure of Mars so that NASA could design a suitable airfoil for the Viking landers. Another of his specialties was faint comets. He was 41.

Mulholland, 47, was not really on the team. He was there to photograph Pluto at the end of each night, after Halley set. But his experience would come in handy if they found any dots on their plates. He had spent more than 20 years studying the motions of small bodies in the solar system, including comets and asteroids.

Bozyan, 39, was a graduate student who had returned to school for a Ph.D. in astronomy after raising two children. Her specialty was blue galaxies, which are a mystery since they appear younger than they really are. Because she had photographed a great many of them with the Griboval, she was an expert on the camera's workings. She had assisted Barker on all three previous attempts on Halley.

"You want to do Swift-Gehrels next, or Grigg-Skjellerup?" she asked Barker.

"I guess Swift-Gehrels."

Barker went downstairs to get the position of Comet Swift-Gehrels from the computer. The rest of us stood underneath the telescope, looking at the Griboval. It was a stainless steel canister, two feet in diameter, with a long metal rod jutting down from its center. From its sides hung a trio of vacuum pumps, a red lamp, and a cable that snaked across the floor to a rack of instruments for monitoring its insides. It had cost a quarter of a million dollars and taken its inventor, Paul Griboval, 15 years to perfect. It was whirring and hissing.

"Ordinary cameras make pictures directly from light," Bozyan told me. "The Griboval uses electrons. Its secret is the photocathode, up here." She rapped the canister near the top. The photocathode was a wafer of pure silicon. A photon of light hitting its front surface would kick an electron out the back. The electron, accelerated by a 50,000-volt field, would zip down through the canister and crash into the film, where it would flip a few grains of silver halide. "The effect on the film is the same as in ordinary photography," said Bozyan, "but it's 10 or 20 times faster because you're flipping the silver with bullets instead of light."

"Liz?" Barker's voice filled the dome. "Are you ready?"

Bozyan stepped to a control panel at the edge of the platform and spoke into an intercom: "Go ahead, Ed." While Barker recited numbers, she began punching buttons. Suddenly, we were engulfed in a bright beeping, like the warning signal of a backing garbage truck. The telescope began tilting. There was a rumbling overhead, and the dome began rotating. Then a loud clunk, and the platform began sinking. Finally, everything stopped.

"That should be Swift-Gehrels," said Bozyan.

Through the eyepiece, the comet looked like a glowing ball of lint caught among the stars. It had no tail. That would sprout later, if at all.

A comet begins as a "dirty snowball" a few miles in diameter, orbiting far beyond Pluto. Pulled out of that orbit, perhaps by the influence of nearby stars, it falls toward the sun. For thousands of years, nobody on Earth knows it is coming, but eventually it reaches a point where the sun begins to vaporize the delicate ice on its surface, and the tiny nucleus throws up a cloud of dust and gases thousands of miles thick. Suddenly brighter, the comet is noticed by some astronomer, often an amateur with a homemade telescope, who wires Marsden at Harvard to claim the discovery. As it falls nearer, the pressure of sunlight may blow some of the vapor cloud backwards, forming a tail. Missing the sun, the comet whips around it and sails off in the direction whence it came. Usually it is gone forever, but a passing planet may pull it into a long, skinny obit that brings if back again and again. Halley's comet is named for Edmond Halley, the British astronomer who first realized it was periodic. It returns every 74 to 79 years, losing a few yards of its surface each time.

The nuclei of comets are believed to be the oldest and least altered fossils in the solar system. They are remnants of the huge cloud from which the sun and planets were born, 4.6 billion years ago. Because they are small—Halley's is thought to be less than five miles in diameter—they have escaped most of the crushing, scorching, and scouring that have eroded so much history from the planets. Written in their chemistry may be clues to the questions: How did the sun end up with a planet close enough, large enough, and wet enough to develop life? Could other stars have similar planets?

Unfortunately, no one had ever seen a comet's nucleus. A typical nucleus was too small for detection by ordinary cameras until the comet lay very near the Earth, and by then its obscuring vapor cloud would have erupted. There was a slim chance, however, that the Griboval would pick up Halley early enough to reveal its nucleus.

Stepping to the eyepiece, Bozyan touched a few buttons on a control paddle hanging beside it, to center the comet. After dimming the lights, she pulled the shutter, a simple metal slide, out of the Griboval. Five minutes later, she pushed the shutter home and detached the metal rod from the canister, along with a box at its root containing the film. She carried the assembly across the platform to a table where, under an amber safelight, she removed the exposed plate: a disk of sheet film the size of a silver dollar. It went into a safe for development later. "Now for Grigg-Skjellerup," she said.

"Uh-oh," said Mulholland, looking upward.

A bright star near the bottom of the slit was fading. Then another, and another.

Behind us, footsteps were booming on the stairs. "Humidity's up," Barker called from the end of the platform. "We're getting clouds."

More stars were fading.

"We're losing it," said Bozyan. "Close the damned slit."

The telescope time for the first evening had cost $1450, and Smith had given his team three more nights at the same rate. With salaries, lodging, and travel expenses, the bill for his February attempt would reach $10,000. Some of his colleagues believed the scientific payoff wouldn't be worth it. Even if he succeeded, they said, the comet would remain too faint for spectroscopic analysis until long afterwards. But science was only part of Smith's motive.

In 1969, when it was completed, the 107-inch telescope was the third largest in the world. Now it is 13th. Smith wanted to build a bigger one—a 300-inch that would go twice as deep as the giant at Palomar and make McDonald the greatest observatory on Earth. All he needed was $40 million.

Smith knew he couldn't get that kind of money from the government these days, so he aimed to raise it privately. His first half-million had come easily in 1981, but his goal for 1982 was $5 million. As of mid February, he had raised exactly none of it.

"Let's just say," he had told me on the phone, "that we hope the publicity generated by a Halley recovery will draw positive attention to our goals."

When the astronomers awoke on Friday afternoon, they found the window in the observatory's dining room blank white with fog.

"It was supposed to be clear!" said Mulholland, throwing a backhand at the window. "I don't believe the Weather Service anymore. I think they've gone down the tube, just like the Postal Service." He stalked out of the room with his calculator holster slapping at his side.

Half an hour later, Barker, Bozyan, and five other astronomers were sitting near the window eating dinner, talking shop, and telling weather jokes.

"Hey, Ed," said Bozyan, "did I ever tell you Tom Moffett's theory on clouds?"

Barker shook his head.

"He says there's this cloud, just a little bitty cloud, and it wanders the Earth every night and hovers over whatever observatory has the most important work scheduled. Isn't that the truth?"

Barker smiled.

"Good news, folks," said Mulholland, hopping up the stairs from the gameroom. "The TV informs us we've been having a beautiful summer day."

The astronomers snickered.

"Did they show the satellite map yet?" asked Barker.

"I didn't know they had one."

"Oh, sure," said a voice across the room. "It ought to be on right now."

There was a squawking of chairs, and the astronomers ran down the stairs. Mulholland got the best seat, an overstuffed chair near the screen, and sat down with his hands folded on his lap, jokingly twiddling his thumbs. A commercial was on. When it ended, a panting dog appeared on the screen. The dog barked, and a yong man with perfect hair and a red blazer said, "Now, folks, here's our Weather Watchdog with tonight's forecast. This beautiful sunshine we've been having all day should continue into Saturday. But watch out for this high-pressure center over in east Texas. That'll be causing a counterclockwise flow of air—"

"We know that," said Mulholland. "Let's see the satellite map."

At that moment, the satellite map flashed onto the screen. Laughter filled the room. Mulholland threw back his head and delivered a scream at the ceiling.

The map showed that west Texas was completely cloudless except for a neat little oval that covered the Davis Mountains.

Up in the dome that evening, Bozyan gave the Griboval a "dark-current" test. After loading it with a fresh plate, she dimmed the lamp until the camera was hardly visible, then raised the photocathode voltage to 50,000 and let the plate sit without opening the shutter. The test was to make sure the Griboval wasn't seeing things that weren't there. Flaws in the photocathode, for example could allow electrons to jump out without the prompting of light, producing dots on the film that could be mistaken for a real object. "We don't want any Kodak comets," she said.

After 45 minutes, she took the plate downstairs to a darkroom and developed it. She was hoping it would come out transparent.

"I knew it!" she exclaimed.

The plate was smoky. Her magnifier showed millions of black dots, which, because the plate was a negative, represented a field fogged with light. In a background like that, Halley's image would be lost.

Barker came in. "What's wrong?"

Bozyan, sitting on a packing crate near the sink, made no reply. She blew a raspberry at the plate as Barker lifted it out of the wash water for inspection.

"Maybe there's light leaking in around the shutter," he said.

He hung the plate on a line over the sink and led the way upstairs. There, they exposed a second plate—this time in pitch darkness. It, too, came out gray. A third, a fourth, and a fifth plate followed, each after some new adjustment. By five in the morning, the line over the sink was a string of gray polka dots undulating in midair.

"I could use a good stiff drink about now," said Bozyan near dawn. She and Barker were standing on a catwalk outside the dome, five stories off the ground. Clouds were racing overhead. She took a drag on her cigarette and let the wind peel the smoke from her mouth. "I'm beginning to think it's a resistor or something overheating inside the vacuum chamber," she said. "And if that's true..."

It would mean hauling the camera back to Austin and forefeiting the remaining two nights.

"Maybe Paul will know something," said Barker, rubbing his eyes. He meant Paul Griboval.

"Maybe," said Bozyan. She decided to stay up and phone him.

On Saturday the weather began improving. The fog vanished, the cloud brightened, and suddenly the line of smooth brown peaks on the northwest horizon was skewered with a streak of blue.

"Paul says we might have contaminants inside the photocathode," said Bozyan when she got off the phone. "If we run the camera all day at 60,000 volts instead of 50,000, it might gradually pump them out."

"Good," said Barker.

The sky that evening was remarkably clear, with thousands of stars shining steadily against the frozen, luminous smoke of the Milky Way. From the catwalk, not a single street or house light could be seen. Blackout curtains had been closed in all the staff dwellings at sundown.

To see whether Griboval's advice had worked, the team first made a 15-minute dark-current plate. They were hoping it would come out transparent enough to permit comet exposures a full hour long. Only then could they easily reach the 24th magnitude.

The team also had to prepare for tracking the comet. During each exposure the comet would, of course, be moving. Besides coasting westward along with the stars, it would be moving among them. Unless the telescope followed both motions precisely, Halley's image would drift on the plate, spraying the electrons along a line instead of piling them into a point. The team wanted their guiding error to be less than 0.15 arc seconds—the width of a pea at a range of five miles.

In theory, they could do that by computer. Downstairs was an IBM 1800 that knew where the comet was supposed to be at every minute. But lately the little wheels that turned against the telescope's main bearings, to tell the computer how fast the telescope was moving, had been slipping as much as a full arc second per hour. It would be safer, said Bozyan, to guide by hand.

I asked how they would do that when they couldn't see their target.

Barker brought out a special eyepiece with a thumbwheel on the side, marked off in fifths of a turn. He slipped the eyepiece into the telescope and said, "Look."

Red crosshairs glowed against the sky.

"Now spin the wheel."

The crosshairs crept sideways.

The trick, he explained, was to guide the crosshairs on a star near Halley while turning the thumbwheel at such a rate that the telescope was actually following the comet. This would fix the comet's image on the plate while letting the stars trail. Halley's speed among the stars was known; all that remained was to convert it into a thumbwheel rate.

While Barker and Bozyan were doing the conversion, Allan Butcher, a 20-year-old night assistant, took the dark-current plate downstairs and developed it for them. The plate came out slightly gray. Looking at it, nobody was sure whether the Halley plates could go a full hour or not.

"Let's make the first plate a half-hour," said Barker to Bozyan. "Then you can develop it while Derral and I start another one. The first one will tell us whether we can go any longer on the second."

"Fine," said Bozyan.

They explained that although a half-hour plate wouldn't reach 24th magnitude, two plates might do it if they were combined later by computer.

The exposure was made with great care and cooperation. While Bozyan went downstairs to the control room, Barker and Mulholland stood at the telescope. Barker, glued to the eyepiece, kept his guide star centered by pushing buttons on the control paddle. Every 55 seconds—the interval they had gotten from the conversion—Bozyan called over the intercom, "Turn!", signaling Barker to advance the thumbwheel a fifth of a turn. To make sure Barker didn't miss any, she added, "That should be another complete turn!" every fifth time and waited for Mulholland to relay Barker's confirmation by calling back, "Check!"

Afterwards, in the darkroom, the plate turned out to be as smoky as those on the previous night. "We'll take a look at it with a magnifier when it's dry," Bozyan told me, "but I don't expect to find anything." They stopped the second exposure at 30 minutes and gave up on Halley for the night.

At dinner the next evening, Bozyan said she and Barker had found something on the plates after all. "It's awfully faint," she said. "We can't even be sure it's real until we see the prints. And it's off the track by a quarter of a day. So we're going to call it an asteroid for now. But we're definitely interested in getting more plates tonight."

"Let's not get excited," said Barker. The object couldn't be Halley, he said, unless the comet's vapor cloud had erupted—and that was unlikely.

"You're right," said Bozyan. "It's probably an asteroid. But it looks so nice and fuzzy and round. If we're guiding on Halley, then everything that isn't Halley should be drawn out into a line."

"Let's wait and see the prints," said Barker.

On the camera problem, they had gotten fresh advice from Paul Griboval: turn the voltage down. They decided to begin the evening by making dark-current plates at 40,000 and 45,000 volts, to see which worked better.

The sky on the final night was even better than Saturday's. Steadier air caused the stars to twinkle less, concentrating their light into finer points that would register more quickly on the plates.

While Barker and Bozyan were exposing the dark-current plates, Allan Butcher began printing the Saturday-night plates. He projected the first plate onto an eight-by-ten sheet of high-contrast paper, slipped the paper into a tray of developer, and gently rocked it under the safelight. Slowly the black sky faded in, leaving dozens of star trails. Among them was a fuzzy white ball that looked remarkably like a comet.

Butcher slapped the print into a fixing bath and ran upstairs. The observing platform was high in the air. He shouted up at it, "Ed!"

Barker and Mulholland poked their faces over the edge.

"There's definitely something there," said Butcher, "and it's definitely round, not trailed."

"Hmm," said Mulholland.

"Hmm!" said Barker.

While the two astronomers hurried to the print room, Butcher ran downstairs and told the news to Bozyan, who was developing the dark-current plates.

"Wow!" she said. "Aw right!"

Butcher told her he couldn't wait to print the second plate. It would show whether the object had moved.

In walked Barker and Mulholland, fresh from seeing the wet print. "Interesting, it is!" said Mulholland.

Bozyan held up the 40,000-volt plate. It was virtually transparent.

"That's clean, kid," said Barker. "Let's go!"

I just love it when it's like this," Bozyan said to me four hours later. "It really makes me glad I came back to school. Turn! That should be another complete turn!"

She crossed the hatchmarks on her notepad and shifted her eyes back to the clock above the desk.

"Mr. Barker confirms that it was indeed another complete turn," said Mulholland's voice on the intercom. "Hey, Liz, could we have synonyms? This 'Turn!' business is getting a bit tiresome."

"You want synonyms, you got 'em," said Bozyan.

The Halley team was in the middle of the third and final exposure of the night. It was going a full hour. Meanwhile, Butcher was down in the print room blowing up the second Saturday plate.

"Rotate!" said Bozyan. She made another hatchmark. "School is really a long haul when you're going part-time," she said to me. "I've been at it six years. A lot of people thought I was crazy for even starting. At least I haven't been sitting around doing nothing. I can't stand that." She took a puff on her cigarette. "You know, Ed's really excited about that print," she confided. "I really think we might have something this time. Roll!"

Suddenly, the door opened and Butcher was standing over us, smiling. "There's definitely a second image," he said. "And it's moved."

Grinning triumphantly, Bozyan pounded the desk with her fist.

Later, the prints were examined by Derral Hulholland in the dining room. He stared at the tabletop and slid the first print before his eyes, then the second. The fuzzy object appeared to move, all right. But its motion was relative to the star-trail pattern, which itself had moved because the two exposures had been aimed differently. Most significant: the object appeared not to move at all in relation to the edge of the plate. It seemed to be glued to the camera's field of view like a speck of dirt on a windshield. He picked up a teaspoon and measured the prints with its handle, to make sure. "That's a cathode flaw!" he said disgustedly, throwing the spoon down the table.

"Oh, well," said Bozyan. "We still have tonight's plates—and they're probably a lot deeper."

The rest of the search was Ed Barker's. On Monday morning he took the five Halley plates back to Austin where, during the next few days, he scanned them on a machine that converted them into television pictures. Playing them on a 19-inch screen, he cranked the contrast even higher than Butcher had been able to go in the printroom. Round, fuzzy objects popped out all over the place. But a computer analysis showed that none of them moved at the right rate. "I'm afraid all we have is asteroids," he told me on the phone on March 3.

"Oh, we're not dead here yet," Harlan Smith said when I called. "Ed still has a lot of measuring to do. There are plenty of fainter candidates running around near the plate limit. It'll take us literally weeks to squeeze all the information out of those plates." He promised to send me a brochure on the 300-inch telescope.

For Barker, the next two weeks were a continual frustration. Every time he found a dot sequence that moved correctly from Saturday night to Sunday, it turned out to be moving too swiftly from one plate to another on the same night. "I don't know what's going on," he told me on March 17.

The next morning, on a hunch, he went upstairs to Bozyan's office and pulled her logbook off the shelf. Inside was the thumbwheel conversion they had done on Saturday night before guiding their first exposure. Halfway through, he found a mistake: Bozyan had written a figure in "turns per arc second" where she meant "arc seconds per turn."

"I was watching her at the time," Barker told me, "and I didn't catch it, either."

The upside-down fraction had led them to guide everything 50-percent too fast, trailing not only the stars but also the comet. "It cut our reach by a magnitude, maybe two," said Barker. "Depth-wise, we're not even in the ballpark."

We probably wouldn't have gotten the comet anyway," Smith told me. "You know, some people are saying the darned thing is 27th magnitude, which means nobody in the world has the equipment to pick it up now."

The mistake was nothing unusual, he said. "Things like that happen all the time. You just never hear about them." He told stories of astronomers photographing the wrong piece of sky, ruining plates by developing them in the wrong solution, and guiding three-hour exposures without opening the shutter. "When you work in darkness, with complicated equipment, and you're exhausted all the time because of the sudden shift to a night schedule, you learn to expect such things," he said.

Smith said his team would be making another attempt this fall—"probably in October, although we may be able to sneak in a few plates in late September. We'll see."

Later, I phoned Brian Marsden to tell him about the failure. He chuckled, declared it was too bad, and repeated his claim that nobody is going to find Halley before the fall of 1983. But, he added, if any team could prove him wrong, it might well be Smith's. "They do seem to be getting a lot of practice."Top