Mirror Earth (6 page)

During the 1970s, it had also became apparent that the universe was pervaded by some sort of mysterious, invisible substance, at first known as the “missing mass” and later called
“dark matter” (that mystery still hasn't been solved). The dark matter would have had a powerful influence on how the cosmos evolved, so astronomers wanted to understand how galaxies were spread throughout the universe. Were they sprinkled evenly or were they assembled into some sort of pattern that hinted at how much dark matter there was and how it was distributed? (We now know they're in patterns that resemble Swiss cheese, which helps to rule out some models of cosmic evolution.)

Latham was involved in some of the early surveys that tried to understand galaxy distribution. The galaxy surveys, operated out of the Smithsonian's telescopes on Mt. Hopkins, near Tucson, Arizona, used spectrographs. But galaxies are too faint to see when the Moon is bright in the sky. Latham hated to see the instruments sit idle, so when the Moon sidelined the galaxy surveys, he began looking at stars in the Milky Way to see if they wobbled. He wasn't interested in planets at this point, but in binary stars—pairs of stars that orbit each other. These are actually more common than single stars in the Milky Way. The Sun is unusual in wandering through the galaxy alone. “We were looking to see how frequent binaries were,” said Latham, “and what their characteristics were. By the eighties we were mass-producing radial velocities [the technical term for motion toward and away from the observer]. We had our errors down to maybe five hundred meters per second,” which is about a thousand miles per hour.

Then, in 1984, an Israeli astronomer named Tsevi Mazeh contacted Latham. Mazeh was interested in radial-velocity measurements too. He had just been out in Santa Cruz, California,
consulting with Steve Vogt. Geoff Marcy had barely begun thinking about planets at this point. Mazeh was on his way back to Tel Aviv, but he stopped in Massachusetts en route. “He had this idea,” said Latham, “to use my radial-velocity instrument to search for giant planets.” Latham patiently explained to Mazeh why this wouldn't work, that even a planet like Jupiter would pull on its star too weakly to show up, and that it would take more than a decade to see a single orbit in any case. “No,” said Mazeh, “I mean giant planets with very short periods.”

You could certainly find those: A short orbital period means you can watch several orbits go by without waiting forever. The planet Mercury orbits the Sun once every eighty-eight days. If a giant planet hugged its star the way Mercury does the Sun, you could see four full orbits in just under a year. Beyond that, a planet's gravitational effect on its star is more powerful the closer it orbits—it has more leverage—so the observations don't have to be nearly as delicate. That was all true, acknowledged Latham, but giant planets don't orbit that close to stars. They certainly don't in our solar system, and most planetary theorists were confident they couldn't exist at all.

“Maybe,” said Mazeh, “the theorists are wrong.”

“He seemed like a nice fellow,” recalled Latham, “so I agreed to work with him.” To boost their chances further, Mazeh suggested they look at M-dwarfs, red stars that are at most half as massive as the Sun. M-dwarfs make up about 70 percent of the stars in the Milky Way, but they're too dim to see with the naked eye. The fact that they're such lightweights means that
a giant planet orbiting an M-dwarf would make the star wobble even more, making it that much easier to detect. “We also decided to look at some Sun-like stars,” said Latham. “On the night of March 31, 1988,” he continued, “I was working up the observations on one of them, a star called HD 114762.”

At this point, readers should be warned that the naming conventions for stars are complicated enough to make your head hurt. The HD in the name of the star Dave Latham was looking at signals that it appears in the Henry Draper catalog, a list of more than 350,000 of the brightest stars visible from Earth, classified according to features in their spectra. (Draper, a medical doctor and amateur astronomer, took some of the earliest photographs of a star's spectrum. The catalog named in his honor was assembled by Harvard astronomer Edward Pickering, using funds donated by Draper's widow in the 1880s.)

But that's just one of many star catalogs. Another is the Gliese catalog of the closest, rather than the brightest, stars. The German astronomer Wilhelm Gliese put it together in 1957. “Gliese names,” Geoff Marcy told me, “are kind of nice, because they tell you right away it's a nearby star even if you don't know anything else. But the HD catalog is more widely used.” And then there are the Hipparcos and the Tycho catalogs, gathered by the European Space Agency's Hipparcos satellite in the 1990s (the satellite's mission was to map the positions of millions of stars; the Hipparcos catalog has very high precision, the Tycho catalog is based on somewhat less careful measurements with the same satellite).

It doesn't end there. Many stars are also identified by the constellation they're part of. Alpha Centauri is the brightest
star in the southern constellation Centaurus (
alpha
is the first letter in the Greek alphabet). If a constellation has more than twenty-two stars, you run out of Greek letters, so you go to numbers. The constellation Pegasus has a star known as Alpha Pegasi, and it also has one called 1 Pegasi. Then there are stars with given names—Vega, Sirius, Aldebaran, Betelgeuse, Polaris, Arcturus, and more. These stars are so bright and prominent in the night sky that ancient Greek and Arab sky watchers thought of them as old friends, with distinct personalities. Finally, when astronomers are conducting an organized search, they'll sometimes create their own catalogs—the Kepler catalog, the Wide Angle Search for Planets (WASP) catalog, the Hungarian Automated Telescope (HAT) catalog, and so on. According to one astronomer I spoke with, this is partly to make sure the credit for a planet's discovery goes prominently to the search team.

The one catalog astronomers don't pay attention to is the one compiled by the International Star Registry. This is the company that lets you “name a star after someone.” The ads say, quite truthfully, that the names will be recorded in book form in the Library of Congress. But that's true for any book, whether it's an erudite volume of history or a trashy romance novel. The company makes clear on its website that astronomers don't consult the book.

Even when you're talking about stars identified by their numbers in a single catalog—in this case, the Henry Draper catalog—the names begin to swim before your eyes. Unless you're a planet hunter, that is. Then they're as distinctive as the names of your children. “Exactly!” said Debra Fischer, a
Yale astronomer, when I proposed this analogy. “You don't know HD 209458? These names are burned into my memory. Someday I will have Alzheimer's, but I will remember these stars.” For the record, Fischer tends to rely on the Hipparcos catalog, because it has more than two million stars; because it notes their positions and distance from Earth with exquisite accuracy; and because all of the information is online, making it easy to access.

So Latham was working on HD 114762, a star in the Henry Draper catalog. His calculations suggested a wobble with a period of about eighty-four days, about the same as Mercury's. The magnitude of the wobble suggested an object with a mass about eleven times that of Jupiter—or rather, that was the
minimum
possible mass. This was a crucial point. Since the object pulling on the star was itself invisible, Latham couldn't be certain its orbit was truly edge-on. If it was, the mass was eleven times that of Jupiter. If the orbit had been at right angles to Latham's telescope, you wouldn't see any motion at all toward or away from the telescope; it would all be side to side. But if the orbital plane was tilted somewhere in between those extremes, you'd see something between the full effect and zero. Any motion toward or away from the telescope would reflect
part
of the tugging. A much bigger object, pulling at an angle, could mimic an eleven-Jupiter-mass body seen directly edge-on.

Latham sent an e-mail to Mazeh, with a copy to Michel Mayor, a Swiss astronomer who was also looking for wobbling stars. Mayor wasn't looking for planets either; he was looking for brown dwarfs, objects bigger than planets but smaller than
stars. A brown dwarf could be as much as eighty times as massive as Jupiter before it would start fusing hydrogen into helium in its core—the same reaction that powers an H-bomb, and the one that makes stars shine. At this point, brown dwarfs were purely theoretical (they've since been shown to exist), but looking for them had made Mayor an expert on radial-velocity measurements too. The e-mail said, in part, “This is interesting—the minimum mass is well under the stellar limit. It could even be a giant planet.” By the time Latham hit “send,” the clock had ticked past midnight. It was now April 1, and he was a little bit worried that Mazeh and Mayor might think it was an April Fool's joke.

But Mayor went out and did his own measurements, and got the same results. Mayor also determined that whatever this object was, its orbit was eccentric—it was somewhat oval rather than nearly circular. That ruled out a planet, because, as Latham said, “everyone knew that giant planets had to have circular orbits.” Everyone also knew you couldn't have planets bigger than about twice the mass of Jupiter. And with a “year” just eighty-four days long, well, said Latham, “that was just a killer. Three strikes, you're out! Tsevi and I had a bet—we still do, in fact—I said it was a small star, he insisted it was a big planet.” When Latham, Mayor, Mazeh, and two others reported the discovery in
Nature
on May 4, 1989, they wrote: “The companion is probably a brown dwarf, and may even be a giant planet.” Latham told me Mayor “wasn't too happy with that wording.” He was on Mazeh's side.

In any case, Latham was too busy with other things to keep looking for objects like this one, but Mayor, he says, “picked
up [the project] and ran with it.” Like Marcy and Butler, he and a French instrument builder named Andre Baran began beating down the errors in their own spectrograph. Marcy and Butler had chosen to do it with iodine cells and horrifically complex software. Mayor and Baran chose instead to make their spectrograph as utterly stable as they could. They used a reference spectrum from a lamp outside the telescope, but they piped it into the spectrograph with fiber-optic cables in such a way that it was as undistorted as it could possibly be.

When they'd done everything they could think of, Mayor told me at a conference on the Isle of Capri in 1996, they had beaten down their errors to thirteen meters per second, or about 30 mph. They couldn't find a Jupiter like the one in our solar system, but then, they weren't looking for one. Mayor still cared mostly about brown dwarfs, and the spectrograph, far more sensitive than Dave Latham's, could detect them easily. In early 1994, Mayor and a graduate student named Didier Queloz began taking measurements of wobbly stars. By now, Marcy and Butler had been at it for half a dozen years.

Mayor and Queloz had put more than a hundred stars on their observing list to maximize their chances of finding something. Brown dwarfs might be relatively rare, after all, and their orbits would have to be nearly edge-on for the astronomers to make a strong detection. Not all of them would be, of course. So the European astronomers began methodically ticking through their list. Within a few months, they noticed something very odd. A star named 51 Pegasi (the fifty-first brightest star in the constellation Pegasus) seemed to be wobbling, but
in an impossible way. It was moving back and forth, not with a Jupiter-like rhythm of 11 years, not with an Earth-like rhythm of 365 days, not even with a cadence of 84 days, like HD 114762. This star was moving toward Mayor's telescope and dancing and advancing again
once every four days
. If this motion were truly caused by an orbiting body, it was hugging its star an absurd ten times closer than Mercury hugs the Sun.

Not only that, but based on how hard it was yanking on the star, this was no brown dwarf: It was only half as massive as Jupiter. Mayor's spectrograph wasn't sensitive enough to find a Jupiter in a Jupiter-like orbit, but something this close in had a huge amount of leverage on the star. It looked just the way a giant planet should look—if a planet could exist in this location. Theorists said it couldn't. But as Tsevi Mazeh had said ten years earlier, “maybe the theorists are wrong.”

In fact, not all theorists had said such a close-in planet couldn't exist. Douglas Lin, of the University of California, Santa Cruz, had proposed back in 1992 that giant planets might migrate inward from where they originally formed. He figured they'd just spiral all the way into the star and be destroyed—but for a while, they could take up an orbit like the one Mayor was describing. And forty years before that, the legendary theorist Otto Struve had written a paper for the October 1952 issue of a journal called
The Observatory
titled “Proposal for a Project of High-precision Stellar Radial Velocity Work,” in which he foreshadowed not only Mayor's and Marcy's and Dave Latham's work, but Bill Borucki's as well. In part, Struve wrote:

We know that
stellar
companions can exist at very small distances. It is not unreasonable that a planet might exist at a distance of 1/50 astronomical unit, or about 3,000,000 km. Its period around a star of solar mass would then be about 1 day. If the mass of this planet were equal to that of Jupiter … [it] might be just detectable … There would, of course, also be eclipses … This, too, should be ascertainable by modern photoelectric methods.

Struve's idea had been largely forgotten, however, and Lin's work was considered highly speculative. If an observer is wrong as much as 10 percent of the time, goes the astronomical rule of thumb, he or she is a pretty careless observer. But if a theorist is wrong as little as 10 percent of the time, he or she isn't taking enough creative risks. Lin was, and remains, a very good, creative theorist, so his colleagues took his predictions with a grain of salt.