The most surprising feature of exoplanets so far, Marcy says one day in his office on the Berkeley campus, is their unusual orbits. In the classic "overhead view" diagram of our solar system, the planets (except for oddball Pluto, recently demoted to a dwarf planet) trace nifty concentric circles around the sun. Marcy reaches behind his neat desk and takes out an orrery, a mechanical model of our solar system. Metal balls at the ends of spindly arms swivel around the sun. "We all expected to see these phonograph-groove circular orbits," Marcy says. "That's what the textbooks said about planetary systems. So when we first started seeing eccentric orbits in 1996, people said they can't be planets. But they turned out to be a harbinger of things to come."
Just after midnight at the Lick Observatory, the astronomers are making good progress on the night's checklist of 40 stars. Their targets usually aren't the major stars of the constellations, but, even so, many are bright enough to see with the naked eye. "When I'm out with my friends, I can point to a couple of stars that we know have planets," Howard Isaacson says. One particularly bright star in the Andromeda constellation has three.
McCarthy offers to reveal the secret of the team's success at spying exoplanets. We walk into the dark dome and pass under the telescope, with its ten-foot-wide mirror that collects and focuses the faint rays of light from distant stars. I had seen the massive telescope during daytime tours, but at night it looks much more vital, its thick metal struts angled like the legs of a tall praying mantis looking up at the heavens. McCarthy leads me to a cramped room beneath the dome's floor, where starlight concentrated by the telescope's mirror is streaming into a cylinder smaller than a soda can. It's wrapped in blue foam, with glass on both ends. It looks empty inside, but I'm told it's full of iodine gas heated to 122 degrees Fahrenheit.
This iodine cell was developed by Marcy and his former student Paul Butler, now an astronomer at the Carnegie Institution in Washington, D.C. When light from a star passes through the hot gas, iodine molecules absorb certain wavelengths of light. The remaining light is spread out into a rainbow by an instrument that acts like a prism. Because the iodine has subtracted bits of light, dark lines are scattered across the spectrum like a long supermarket bar code. Each star carries its own signature of wavelengths of light that have been absorbed by the star's atmosphere. These wavelengths shift slightly when a star moves toward or away from us. The astronomers compare the star's own signature of dark lines with the stable iodine lines from one night to the next, and from month to month and year to year. Because there are so many fine lines, it's possible to detect even minute shifts. "It's like holding the star up to a piece of graph paper," McCarthy says. "The iodine lines never move. So if the star moves, we use the iodine lines as a ruler against which to measure that motion."
For something as big as a star, the only things that can cause a regular, repeating shift are the gravitational tugs of another star—which astronomers could detect easily because of a companion star's own light signature and its hefty mass—or a hidden planet orbiting around it. The iodine cell can track a star moving as slowly as several feet per second—human walking speed—across the vast emptiness of trillions of miles of space. This sensitivity is why many planet-hunting teams use the iodine cell.
I peer inside it and see some crinkled foil and heating wires snaking through the blue foam. Strips of duct tape appear to hold parts of it together. After we return to the control room, McCarthy chuckles and points out the slogan on Keith Baker's sweat shirt: "When the going gets tough, the tough use duct tape."
The more oddly shaped and oddly spaced orbits that astronomers find, the more they realize that the natural process of planet formation invites chaos and disorder. "It became clear that our solar system, with its beautiful dynamics and architecture, was much more stable than those around other stars," says theoretical astrophysicist Greg Laughlin of the University of California at Santa Cruz, who collaborates with Marcy and Butler's team. Trying to figure out how new planets acquired their weird paths has been a daunting task. Laughlin designs computer models of exoplanet orbits to try to re-create the planets' histories and predict their fates. He focuses on the role of gravity in wreaking havoc. For instance, when a big planet moves on an eccentric orbit, its gravity can act like a slingshot and fling smaller nearby worlds. "In some of these systems," Laughlin says, "if you insert an Earth-like planet in a habitable orbit, it can literally be ejected within weeks."
Interactions among planets may be common in the cosmos, Laughlin and his colleagues say. Almost 20 stars are known to have more than one planet orbiting around them, and some of these sibling exoplanets are locked into a dance called a "resonance." For instance, one planet circling a star called Gliese 876 takes 30 days to orbit, while another planet takes almost exactly twice as long. Laughlin's calculations show that their mutual gravitational pull preserves a stable, clocklike arrangement between the two planets.
Resonances are strong clues that the planets migrated far from their birthplaces. The disk of dust and gas that spawns embryonic planets has a gravity of its own. The disk drags on the planets, gradually pulling them inward toward the star or, in some cases, forcing them outward. As this migration goes on for hundreds of thousands of years, some exoplanets become trapped in resonances with their neighbors. When big planets end up in close quarters, they whip each other around and create some of the eccentric orbits seen by the team. At least, that's the current best guess.
Other planets are not long for this world. Laughlin's computer models suggest that some of the planets closest to their stars will plunge into them as more distant planets bully their way into smaller orbits, perhaps in a matter of hundreds of thousands of years. This research into distant solar systems has raised a fascinating scenario about our own solar system. Some astronomers theorize that Venus, Earth and Mars are "second-generation" planets, successors to earlier bodies that were born closer to the sun and migrated inward until they were consumed.