Epicurus, a Greek philosopher in the fourth century B.C., did not explicitly predict the existence of planets around stars other than the Sun, but he believed in an infinity of worlds, meaning other ordered systems beyond the visible universe as it was then conceived. This contrasted to the Earth-centered cosmos of the contemporary Aristotle, whose cosmology prevailed in Western thought for more than two millennia.
Only in the last three years have astronomers established the reality of latter-day Epicurean speculations about a plurality of worlds, which in recent centuries came to mean planets beyond the solar system, some possibly inhabited. But while astronomers tip their hats to Epicurus, they just wish he had advised them how to make sense of the distant planets being detected by their telescopes.
Finding something is not the same as discovering what is found. The more astronomers study the growing evidence of extra-solar planets, the less the planets resemble anything in the one planetary system they had known and had based their theories on: the Sun's family of planets.
At last count, astronomers in the United States and Europe had observed 18 nearby Sunlike stars showing telltale motions from the gravity of large, unseen planets orbiting them, and they fully expect to find more. Yet they suspect they have seen enough to begin rethinking how nature creates and destroys planets and choreographs their orbital minuets.
Nine of the objects hug closer to their parent stars than Mercury is to the Sun, closer than standard theory predicted planets could be; one is so near that it makes a complete revolution -- its full year -- every 3.1 Earth days. The other nine travel unusually elliptical, or oval-shaped, orbits, several of them plunging in relatively close to their stars and then swinging far out again; orbits in the solar system are almost circular. Several extrasolar planets are at least three times as massive as Jupiter, the solar system's giant, and one is estimated to have 11 times the Jovian mass -- raising questions about how massive can a planet be.
Dr. Geoffrey W. Marcy, the astronomer at San Francisco State University who has had a hand in most of the discoveries, is as surprised as anyone. "A trend is now being stamped on these discoveries that we thought, frankly, would go away," he said.
But it has not, and questions pile up. Many stars may have planets, as the discoveries suggest, but is there a typical pattern? Could the solar system be an oddball? If so, does that diminish prospects for intelligent life's existence elsewhere in the universe?
Dr. Alan P. Boss, a theorist of planetary systems at the Carnegie Institution of Washington, thinks astronomers will eventually "find systems that look something like our solar system." He acknowledged that they would also "find many more surprises to make us rethink what we're doing."
Astronomers concede that so far their sampling of extrasolar planets may not be representative, only a reflection of detection capabilities. They have no proof yet of another Sunlike star with more than one planet, or with anything considerably smaller than Jupiter. But it is easier to observe the gravitational effects of Jupiter-class planets, especially those extremely close to the host stars. It takes years of repeated observations to gather reliable evidence for planets traveling the longer orbits at much greater distances from a star. And it is not yet possible to detect in any orbit, near or far, an Earth-size or even Saturn-size planet.
The newest detection, announced last month, was of the smallest extrasolar planet yet examined, one that has less than half of Jupiter's mass and is only 1.4 times more massive than Saturn. The planet, in a tight 3.5-day orbit around the star HD 75289, was found by a team of Swiss astronomers led by Dr. Michel Mayor of the Geneva Observatory, who in October 1995 reported the first confirmed planet around another star like the Sun.
In the next decade, the National Aeronautics and Space Administration expects to fly several space telescopes for a more comprehensive survey of planets around nearby stars. On the drawing board is an advanced satellite called Planet Finder that someday could send back the first pictures of Earth-type extrasolar planets.
"It's a terribly exciting field right now," said Dr. Stephen Lubow, an astrophysicist at the Space Telescope Science Institute in Baltimore. "The discoveries have really opened a new window on the nature of planetary objects in the universe."
The first discoveries of extrasolar planets should have prepared astronomers to expect the unexpected. In 1992, radio astronomers reported the first strong evidence of such objects, but the two planets were not orbiting a normal star. They accompanied a pulsar, the dense remnant of an exploded star and not a neighborhood likely to be hospitable to life.
Dr. Mayor and Dr. Didier Queloz of Switzerland then detected a planet around 51 Pegasi, a solar-type star, and this was soon confirmed by Dr. Marcy and a colleague, Dr. R. Paul Butler. The first accepted planet of an ordinary star excited and puzzled astronomers. Both teams were startled to find that the planet, about half the mass of Jupiter, was in an almost circular orbit less than one-sixth the equivalent distance of Mercury to the Sun.
Ever since then, theorists have been puzzling over how several of the large planets -- dubbed "hot Jupiters" because of their proximity to the intense heat of their stars -- could be where they are. Why were they not out somewhere the equivalent of the Jupiter-Sun distance?
Because known physical laws rule out the formation of large planets so close to a star, theorists think they formed in a more benign environment far out and migrated inward. The unlucky ones probably crashed into their stars. Others somehow settled into cozy orbits at less than one-fourth of an astronomical unit, the standard measure of planetary distances in which one unit is the distance from the Sun to Earth, or 93 million miles.
The migration theory that receives the widest attention was proposed by Dr. Douglas Lin of the University of California at Santa Cruz, Dr. Peter Bodenheimer of the university's Santa Barbara campus and Dr. Derek Richardson of the University of Washington.
Their concept drew on research by Dr. William Ward of the Southwest Research Institute in Boulder, Colo., and ideas developed to explain the interplay of the rings and satellites around Saturn.
According to prevailing theory, a planetary system forms from a disk of gas, dust and chunks of rock that surrounds a newborn star. The star's heat would drive gas out of the inner disk and prevent the huge gaseous planets from forming there. They would instead develop in the gas-rich outer disk.
In the early period of planetary formation, the disk would still be thick with gas and other material. Drag from the disk material and its general inward flow, caused by the young star's gravity, would have drawn many of the large planets out of their original orbits.
Astrophysicists postulate several phenomena that prevent at least some of the planets from plunging all the way into their stars. One is that when the star was young it was spinning more rapidly, creating tidal forces that arrest a planet's migration short of catastrophe. Another idea, suggested by some observations of young stars, is based on the likelihood that the gravity or magnetic forces of newly formed stars soon sweep away disk material from their nearest surroundings, leaving a doughnut hole at the center. Here the migrating planets could settle into parking orbits.
Dr. Lin, pondering why something like this did not happen to Jupiter, decided that previous Jupiters in the solar system probably migrated to their destruction. Jupiter and the other solar planets represent the last generation, created as the planetary disk was dissipating and leaving more stable conditions. Or perhaps the solar system's protoplanetary disk never had enough gas and dust to perturb the orbits of its new planets.
A more recent variation on the migration theory was introduced by a team of theorists at the University of Toronto led by Dr. Norman Murray. In the early formative period, they argue, planets might be orbiting through a disk of planetesimals, small rocky objects colliding with or being ejected by the planets. The destabilizing interactions with the planetesimals could push the newly forming planets toward their stars.
One of the first detections by the Marcy-Butler team, announced in January 1996, introduced another puzzlement. The planet around 70 Virginis was more than seven times the Jovian mass and not as close to its star as many others, but its orbit was highly elliptical. Other planet discoveries revealed similar characteristics. The one around 16 Cygni B has the most elliptical orbit; if it was in the solar system, the massive planet would sweep in as close as Venus and retreat out as far as the asteroid belt between Mars and Jupiter.
Some powerful gravitational forces, astrophysicists said, must have perturbed the planets' orbits. A star passing too close could knock a planet out of its generally circular orbit.
Or the planet's own star could be part of a binary system, one of two stars in gravitational embrace, and the companion star could be unsettling the nearby planets. But not all the planets in elliptical orbits are in binary systems.
Dr. Frederic A. Rasio of the Massachusetts Institute of Technology, working with Dr. Eric Ford, has proposed a concept of gravitational scattering that, he said, "explains very naturally and simply planets in wide eccentric orbits."
The idea involves two or more huge planets orbiting in close proximity so that they generate a kind of gravitational slingshot. The forces might sling one planet off on an elongated orbit to the inner planetary system, while the other might fly off toward the fringes of the system, perhaps escaping into interstellar space.
Such a scenario could also explain why astronomers have yet to find more than one planet around a single star. Any other large ones there were catapulted into deeper, longer orbits and would be undetectable in the brief time astronomers have been looking for evidence of extrasolar planets.
Just one giant planet on an elongated orbit, moreover, would probably spell doom for smaller planets as it crosses their paths time and again, scattering or destroying them in the turbulence of their gravitational wakes.
"If our Jupiter were in an eccentric orbit, the Earth and Mars would likely be gravitationally scattered out of the solar system," Dr. Marcy said. "Thus our existence depends on both Jupiter and Earth being in mutually stable, circular orbits."
The implications are profound for the search for extraterrestrial life. "The big bullies may wipe clean the terrestrial planets in those planetary systems, rendering them void of any Earth analogues," the astronomer said.
Dr. Marcy takes an optimistic view. Of all the Sunlike stars that have been studied so far by planet seekers, he said, only 5 percent have been found to have Jupiter-mass planets in such dangerously eccentric orbits.
That leaves 95 percent of stars that may be free of these wrecking forces and so could harbor habitable planets.
All theories to explain the newly detected extrasolar planets, Dr. Rasio said, remain at "the hand-waving level." Theorists are severely limited by observations, which have yet to reveal more than one planet around a single normal star. The two or three objects around a pulsar offer little insight. Earlier reports of possibly two planets around the star Lalande 21185 have not been confirmed.
"If you only see one companion to a star, you cannot say that this is a planetary system," Dr. Rasio said. "That's going to be the next major breakthrough, finding multiple planets and then putting some constraints on the properties and behaviors of other planetary systems. That's our holy grail."
Planet hunters like Dr. Marcy are looking. They speak of some interesting hints, but nothing yet for the new Epicureans of astronomy to feast on.
