Now the situation is reversed. Scarcely a month seems to pass without a new discovery being reported - and confirmed by other astronomers. At a recent international conference in Capri, Paul Butler from the Lick Observatory in California showed details of yet another planet, which showed up on his computer analysis only four hours before he flew from the US to Italy.
This new discovery, circling the star Upsilon Andromedae, is the fifth planet found by Butler and his colleague Geoff Marcy. To this total we can add a planetary system announced in June by George Gatewood, of the University of Pittsburgh, and half a dozen new results from the Swiss astronomer Michel Mayor, who last year found the first planet beyond the Solar System.
None of these researchers has actually seen a new planet. They measure how the star wobbles as an orbiting planet pulls on it. But this technique only finds the heaviest of planets. Astronomers investigating our Solar System from afar would detect how giant Jupiter is making the Sun wobble, but not discern the puny effect of the Earth. The assumption is that where we find massive planets smaller worlds are likely too.
The first "new planet" was like Jupiter, but circling its star - 51 Pegasi - much closer in than Mercury is to the Sun. Next was 70 Virginis, which seems to have a planet much heavier than Jupiter in a distinctly oval orbit.
But are these all really planets? In Capri Michel Mayor suggested - drawing on evidence of more than a dozen systems - that where the smaller companion to a star is very massive and follows an oval orbit, it is not. Instead, it is a "brown dwarf". Such a body is born like a star, out of a lump of gas that collapses under its own gravity. But it is not massive enough to begin to shine. All it can emit is invisible infra-red radiation: hence the rather tongue-in-cheek adjective "brown" for this type of dwarf star.
Until now, astronomers have found very few brown dwarfs. The best-studied, orbiting a star called Gliese 229, has 40 times Jupiter's mass. Mayor adds five brown dwarfs from his collection of suspected planets, ranging in mass from 10 to 35 Jupiters. He also reclassifies one of the first "new planets" from the Lick Observatory, the companion to 70 Virginis, as the smallest of the brown dwarfs.
The other newly found companions are all less massive than five Jupiters, and follow circular orbits. These are almost certainly true planets, condensed from the remnants of the matter that made up their stars. Two of these planetary systems are reminiscent of our Solar System. The star 47 Ursae Majoris has a planet twice as massive as Jupiter, in an orbit that would, in the Solar System, place it between Mars and Jupiter. Lalande 21185 has a Jupiter-like planet at about the same distance, and a similar planet at Saturn's distance from the Sun.
But, to astronomers' surprise, they are outnumbered by an entirely unexpected kind of planet: the "hot Jupiter". The first discovery, 51 Pegasi, seemed to be a planet like Jupiter, but so close to its parent star that it would be literally red-hot. Originally, many astronomers thought there might be some other way to explain the observations, or at least that this was a one-off mutant of a planetary system.
But Swiss and American teams have now turned up three more "hot Jupiters". Although evidently common, their origin is a complete puzzle. Theory suggests that only small planets should condense from the original dusty disc this close to the parent star - just as we find the small planets Mercury, Venus and Earth closest to the Sun.
The hot Jupiters were probably born much further out, but spiralled inwards as they ploughed their way through the remaining dust and gas in the original disc. But, in that case, they must have moved 99 per cent of the way from the original orbit to the central star, and then - inexplicably - stopped with just 1 per cent of the journey to complete.
If so, that bodes ill for finding planets similar to the Earth: the massive planets would have driven any smaller worlds ahead of them into the central star. Perhaps we should look elsewhere for little green men - not on a planet at all, but on the surface of a rocky moon orbiting a hot Jupiter. The images of Jupiter's moons now coming back from the Galileo spacecraft might be our nearest glimpses of the landscapes that other species in the Universe call home.
The night sky in August
Brilliant Jupiter dominates the evening sky, low down in the constellation Sagittarius, which rather resembles a teapot in shape. To the right lies Antares, a red giant star marking the heart of Scorpius (the scorpion).
Saturn rises in the east around 10pm. Although fainter than Jupiter, Saturn is unmistakable as it currently lies in a region of dim stars. Venus, the brightest of all planets, is rising about 2am, as the Morning Star. It reaches its greatest brilliance on 17 August. If you watch carefully you may see it right through sunrise, and get the rare opportunity of seeing a planet in the bright blue daytime sky.
Around the middle of August, look out for the annual shower of shooting stars that seems to rain down from the direction of the constellation Perseus. These Perseid meteors are debris from a comet, burning up. This is a particularly good year for observing the Perseids, because around the time of maximum the moon will not be around to drown out the fainter meteors.
You'll see most meteors after midnight on 11 August, but there'll be plenty of Perseids for a few days either side
Diary (all times BST)
August 6 6.26am moon at last quarter.
11-12 Maximum level of the shower of Perseid meteors .
14 8.34am new moon.
20 Venus at greatest western elongation.
21 Mercury at greatest eastern elongation.
22 4.37am moon at first quarter.
28 6.53pm full moon.Reuse content