The new scheme would retain Pluto previously threatened with ejection from the club but it would also admit several new members, including the former asteroid Ceres, and even Charon, which until now has been classed as Pluto's moon. Watch a a simulated flyby of all the new planets (mpeg, courtesy of IAU/Martin Kornmesser).

The fuss over Pluto's planethood started a few years ago when astronomers began to discover large, icy bodies in the outer solar system, some of them not much smaller than Pluto.

Then in 2005 came the killer: the object designated 2003 UB313, popularly known as Xena, is 2400 kilometres across, making it slightly bigger than Pluto. Should it be called the tenth planet? If so, how many more planets might there be? Nobody had a clear definition of the term.

A committee of 19 astronomers set up by the IAU in early 2004 failed to reach a decision. The nearest they came was a suggestion that planets have to be more than 2000 kilometres across but eight out of the 19 members rejected that scheme as being too arbitrary (see Xena reignites a planet-sized debate).

Radical redefinition
Now the seven members of a new planet definition committee, comprised of astronomers, writers and historians, have reached a unanimous decision.

"We came to the conclusion that we should base our definition as strongly on the science as possible. The word 'planet' was invented millennia ago, and modern science tells us so much more," says committee member Richard Binzel, an astronomer at MIT, US.

The result is radical. "If gravity can make it round, it's a planet," Binzel told New Scientist. That is not quite the whole story planets also have to orbit a star, and not be either stars themselves or satellites of other planets but the new part of the definition is roundness.

Whereas very small heavenly bodies tend to be irregular rocks, larger ones are crushed by their own gravity into a spherical shape. The committee decided that the threshold of roundness should distinguish planet from non-planet. "We let nature decide," says Binzel.

The threshold size depends a bit on the material, but for ordinary ice or rock it turns out to be well below 1000 kilometres. So the new definition includes Ceres, previously classed as the largest asteroid, which is 950 kilometres across.

Round asteroids
Ceres would become the fifth planet from the Sun and the outermost of the five rocky planets, orbiting between Mars and Jupiter. "It was considered a planet once before, after it was discovered in 1801," says Binzel. "It is again."

Ceres might not be the only asteroid to be promoted. If the definition passes, the IAU will also have to consider the claims of Vesta, Pallas and Hygiea, as all of them are round-ish. The smallest, Hygiea, has a diameter of only about 400 kilometres.

Perhaps an even more surprising planet is Pluto's companion Charon. At more than 1000 kilometres across, it is certainly large enough to have settled into a sphere; and according to the new definition, Charon is not a satellite of Pluto.

Charon is so heavy that the centre of gravity of the two bodies is in space, between the two objects (whereas the Earth and Moon orbit each other about a point deep inside the Earth, making the Moon a true satellite). Pluto and Charon would become the first double planet in the solar system.

Plentiful plutons
A new term has been invented for Pluto, Charon, Xena and any other distant planet with an orbit that takes more than 200 years: They will be called "plutons" (though geologists might be a little peeved that their word for an igneous intrusion has been appropriated). Most new planets would be plutons; there are already several candidates, and perhaps a few dozen will be found altogether.

The new definition is intended to apply to planets outside our solar system as well. It clearly excludes "planemos", objects of planetary mass that have been discovered wandering in space, orbiting no star.

But so far it says nothing about how to distinguish giant planets from brown dwarfs, which are "failed" stars not massive enough to sustain nuclear fusion. "That is a matter for later discussion," says Binzel.

"Mixed feelings"
"I think it's about time that we finally have a definition for the word 'planet'," says Neil deGrasse Tyson, director of the Hayden Planetarium in New York, US. "The committee have accomplished their goal, and defined 'planet' unambiguously. But I have mixed feelings."

"Roundness may be the only definition where Pluto ends up in the same class as Jupiter," he continues. "If you like roundness, then sure they belong together. But any other scheme would distinguish Pluto from the other large objects in the solar system."

When the resolution is debated at a meeting in Prague next week, there may be more objections. Do we want so many planets, or such small ones? The definition also makes material strength a factor.

An object made of exceptionally weak material might collapse into a sphere, and therefore be called a planet, despite being smaller than other irregular non-planets. The gathered astronomers will vote on the resolution on 24 August.