TWO years ago the solar system lost a planet. Pluto was deemed too insignificant to rank alongside Mars, Jupiter and the rest, and was demoted to dwarf planet status. Pluto's fall from favour left us with only eight bona fide planets. But what the solar system has lost, Patryk Lykawka now hopes to replace.

Lykawka, an astronomer at Kobe University in Japan, suspects a ninth planet as large as Earth is hiding beyond Pluto. So far, this frigid "super-Pluto" has escaped detection. But not for much longer, Lykawka hopes. "Within five years or so, we will know for sure if it exists."

Lykawka has become convinced of the existence of this planet thanks to a number of puzzling features in the Kuiper belt , a ring of icy debris in the outer solar system, of which Pluto is one of the largest members.

He is not alone in thinking there is another planet out there. "There are similar proposals in the literature," says Renu Malhotra at the University of Arizona in Tucson, "but Lykawka has done a more comprehensive job. I think his idea should be given fair attention."

Conjuring up a new planet is a tried and tested way of explaining puzzling observations. In the 19th century, Neptune's existence was predicted on the basis of irregularities in the orbit of Uranus. Much later, American astronomer Percival Lowell thought that some further glitches in the orbits of Uranus and Neptune might be caused by what he dubbed Planet X. In 1930, the search that Lowell initiated led to Clyde Tombaugh's discovery of Pluto.

When Pluto turned out to be much too small to tug strongly enough at the two giant planets, the search began for other unknown planets roaming the pitch-black outer regions of the solar system. Increasingly sophisticated observations have since revealed that there is in fact nothing anomalous in the orbits of Uranus and Neptune, but notions of distant, unseen planets still seem irresistible. Larger telescopes, better detectors and more comprehensive surveys have so far come up empty-handed, but none can rule out the possibility that there is an elusive planet just beyond observational reach. The hunt for Planet X is still on.

The evidence for Planet X lies in the region just beyond Neptune, which orbits the sun about 30 times as far away as Earth. This is the beginning of the Kuiper belt, named after planetary scientist Gerard Kuiper, who speculated in 1950 that this region ought to contain a belt of debris left over from the formation of the solar system. The first object in this region was discovered in 1992 by Dave Jewitt and Jane Luu, working at the Mauna Kea Observatory in Hawaii, and well over 1000 have been spotted since.

While most of the known Kuiper belt objects (KBOs) are little more than icy clumps a few hundred kilometres across, some are as big as 1000 kilometres. The largest identified so far, called Eris, is 2400 kilometres across and is 27 per cent more massive than Pluto. It was its discovery in 2003 by Mike Brown , Chad Trujillo and David Rabinowitz working at the Palomar Observatory in California that prompted the International Astronomical Union to define the term "planet" - leading to .

Unspectacular as they are, it was clues from this multitude of frozen chunks that put planet hunters back on the trail. The first clue comes from the unexpectedly sharp outer edge of the Kuiper belt, some 50 astronomical units from the sun (1 AU being the distance between the sun and the Earth; about 150 million kilometres). In this region, known as the Kuiper cliff, the number of KBOs drops dramatically. The second clue is that the belt itself contains different populations of icy rocks with at least three very distinct orbits. Something must have sculpted it, says Lykawka, and that something might well be Planet X.

We know from Saturn's rings that when a ring of small orbiting objects has a sharp, well-defined outer edge, the cause is likely to be the gravitational effect of a large object orbiting further out. Could a similar phenomenon, on a larger scale, have created the Kuiper cliff? Mario Melita at Queen Mary University of London and Adrian Brunini at the National University of La Plata in Argentina argued in 2002 that it could. They proposed the existence of a Planet X at least as massive as Mars and some 60 AU from the sun . But two years later, when Melita looked more closely, he discovered a problem. Working with astronomers at the University of London and Queen's University Belfast, he found that his proposed Planet X could not explain all of the Kuiper belt's intricate features.

Lykawka has now followed this up with a computer simulation which shows that a massive planet so close to the Kuiper cliff would create more of a disturbance amongst many other objects in the Kuiper belt than is in fact the case.

That's not all. Some of the icy debris in the Kuiper belt goes around the sun in step with Neptune. Pluto is a prime example: in the time it takes Neptune to orbit the sun three times, Pluto has gone around twice - it is "in resonance" with the planet. Lykawka has shown that the additional gravitational tug of a Planet X of the kind suggested by Melita and Brunini would destroy this delicate orbital stability and would leave far fewer resonant KBOs than the number we actually see.

"My simulations ruled out many other Planet X proposals too," says Lykawka, who finished his PhD research at Kobe University last year. "None of them is compatible with what we know about the dynamics of the Kuiper belt."

Nevertheless, the lure of an unseen outer planet proved too strong for Lykawka to abandon the chase. A more distant Planet X might yet explain other strange features in the Kuiper belt, including a group of objects within the main part of the belt that have highly elongated orbits and loop round the sun at an unruly angle.

The odd behaviour of some bodies even further out than the main Kuiper belt also needs explaining. Take Sedna, an object over 1000 kilometres across, whose stretched-out orbit takes it 975 AU from the sun before swooping back in to 76 AU. Sedna is not the only "detached" object - one that never comes close to Neptune at all - so could a single Planet X have set them on their peculiar paths and also account for the other Kuiper belt oddities?

Lykawka teamed up with his colleague Tadashi Mukai to find out. "I thought it would be easy," he says, "but it wasn't." Using large-scale computer simulations, the pair worked out the path that Planet X would need to have taken to produce all the known properties of the Kuiper belt. Leading theories of the early days of the solar system suggest that dozens of embryonic planets formed much closer to the sun, from the colliding and clumping of many smaller bodies. Most of these Mars or Earth-sized objects further coalesced into the giant planets Jupiter, Saturn, Uranus and Neptune, which ultimately migrated away from their birthplace around the sun. Gravitational interactions with the fledgling giants would have flung others - including Lykawka and Mukai's Planet X - into distant orbits.

According to their model, Planet X was ejected by a young Neptune into an elongated orbit in the outer reaches of the solar system. Once there, its gravity stirred up the Kuiper belt and swept part of it clean of debris, creating the Kuiper cliff. The second stage in the orbital history that Lykawka and Mukai propose for their Planet X also has its roots in established theories of planetary migration. Tens of millions of years after Neptune formed, the giant planet's gravitational interaction with debris in the outer solar system caused the planet to drift slowly outwards. As it migrated, it captured and swept up KBOs into resonant orbits. This mechanism is generally considered to be the best explanation for the existence of large populations of KBOs, including Pluto, that are resonant with Neptune.

The X factor

According to Lykawka, Neptune's migration pushed Planet X into a distant, resonant orbit. By settling into an average orbit of between 100 and 170 AU from the sun, Planet X was far enough away to leave most other objects in resonant orbits undisturbed, yet close enough for its gravity to create the detached population of objects like Sedna.

Finally, Lykawka and Mukai believe the same subtle gravitational interactions that shape the orbits of small moons around planets played a big part in the evolution of Planet X's orbit. Such interactions were found in 1962 by the Japanese astronomer Yoshihide Kozai as he looked at the orbits of asteroids. He showed that a group of large objects all orbiting in the same plane can tilt the path of a smaller object and make it more circular. The same effect would tilt Planet X's orbit away from the plane of the Kuiper belt.

Today Lykawka's Planet X would take anywhere between 1000 and 2500 years to complete one orbit of the sun, compared with Pluto's 248 years. It would never get any closer to the sun than 80 AU, and its orbital inclination could be as much as 40 degrees from the plane occupied by the major planets.

So could Lykawka and Mukai's planet explain away the Kuiper belt's architecture? Maybe. "It is plausible, from the dynamical point of view," says Malhotra, an expert on planetary migration. "Their proposal is not entirely free of problems, but it has some significant strengths. I am very sympathetic to this idea."

Jewitt also thinks the idea is plausible, though he has some reservations. "The trouble is we are so ignorant of the outer solar system that many things seem plausible, even if they are not true."

Other dynamicists are more critical. When Lykawka presented his idea at the annual meeting of the American Astronomical Society's Division of Planetary Sciences last October, Alessandro Morbidelli from the Côte d'Azur Observatory in Nice, France, dismissed the whole idea as contrived. "Lykawka didn't mention that he is forcing the behaviour of the planet to match what he needs."

Morbidelli's colleague Hal Levison from the Southwest Research Institute in Boulder, Colorado, agrees. "Lykawka sculpts the Kuiper belt by moving and pushing his planet around by hand. I don't believe that it could have happened in the way he describes."

Morbidelli, Levison and two of their colleagues have developed a different theory for the solar system's early history. Called the Nice model, after the French town where it originated, it envisages Jupiter, Saturn, Uranus and Neptune being formed much closer together than they are today. Their orbits gradually changed over time, eventually triggering a series of dramatic disturbances and violent collisions that permanently altered the orbits of the four large planets along with countless asteroids and ice dwarfs.

They claim it successfully explains the orbits of the giant planets, the existence of asteroids that share Jupiter's path around the sun, and a violent epoch called the "late heavy bombardment" that afflicted the inner solar system some 700 million years after it formed .

In a new, as yet unpublished paper, Levison, Morbidelli and their colleagues also invoke the Nice model to explain the properties of the Kuiper belt without the need for a Planet X. According to their theory, the disc of gas and dust from which the first icy planetesimals formed was small and had a sharp outer edge, possibly the result of a passing star sweeping up material. The Kuiper belt formed and then later moved outwards and expanded thanks to the gravitational effects of the migrating giant planets, they say.

Levison and Morbidelli admit that their model has its own problems. It predicts that most objects in the Kuiper belt should have orbits stretched out far further than observations show, and it has difficulty accounting for known objects with extremely tilted orbits. The team consider these minor points, however. "We think that the list of successes of our model outweighs the problems that remain open," they say in their paper.

Levison dismisses as "unphysical" the idea that a planet half as massive as Earth sculpted the Kuiper belt. He argues that there would have been a back reaction from the giant ring of icy debris which would have pulled Planet X from the orbit that Lykawka and Mukai predict into a closer one that would have been detected by now. "The planet should end up in a circular orbit at less than 7 billion kilometres (50 AU) from the sun," Levison says.

Lykawka is undeterred by the criticism. "Several of their comments are prematurely too critical and unbalanced," he says. "I'm planning even more realistic simulations to study this in greater detail."

So does Planet X really exist? Brown points out that his hunt for large KBOs is far from complete. "Could Lykawka's planet be out there and have been missed? Easily," he says. Jewitt agrees. "Given the spottiness and the paucity of published deep sky surveys, one could put almost anything in the outer solar system and it would have escaped detection until now."

More comprehensive surveys on the horizon should eliminate these uncertainties. Sensitive, wide-field telescopes such as Pan-STARRS in Hawaii, the Discovery Channel Telescope in Arizona, and the Large Synoptic Survey Telescope in Chile will soon sweep the skies, leaving no star or space rock unturned.

"I'm always a big fan of theories like Lykawka's," says Brown, "simply because they continue to provide the hope that we might find something big somewhere out there. But if big surveys like Pan-STARRS don't find such a thing, I think it will have to be abandoned."

As ever, it will be astronomers who deliver the final verdict. "When we've done a suitable all-sky survey, we'll know whether or not this Mars-like body exists in the type of orbit Lykawka describes," says Jewitt. "And that will be that - end of story."

Planet X: what to look for

Govert Schilling

Patryk Lykawka and Tadashi Mukai's Planet X is a "super-Pluto" made up of ice and rock. It has a mass between 30 and 70 per cent that of Earth, but will be much less dense, giving it a similar diameter to our own planet's - probably between 10,000 and 15,000 kilometres.

At 100 to 170 AU from the sun, the amount of heat and light falling on the planet would be exceedingly small, so the planet would be completely frozen. Life would be impossible on such a frigid world.

Would Planet X qualify as a bona fide planet, according to the rules of the International Astronomical Union? Probably. It orbits the sun, it would be big enough to be spherical under its own gravity, and it almost certainly has enough mass to have swept debris from the neighbourhood around its orbit. The solar system may have nine planets after all.

Govert Schilling's new book The Hunt for Planet X will be published in July by Springer