And if one star could have planets, then what was to stop others having them?

Scientists long ago decided that these so-called extra-solar planets must exist. But while believing in them is one thing, identifying them is quite another. Today, we know of dozens. But the first to be pinpointed was not what it seemed. It was in 1938 that Peter Van de Kamp, director of Sproul Observatory, near Philadelphia, began studying Barnard’s star, the second closest to our solar system. Although this dwarf sun cannot be seen with the naked eye, it can be photographed through a telescope. And that is what Van de Kamp did - over and over. For a quarter of a century, he and his team measured 2,400 fuzzy photos, hoping to find variations of just 1 micron in the position of each image that would indicate the sort of gravitational wobble associated with an orbiting mass.

And when, in 1963, he announced that the search was over, Barnard’s star planet, as the mass was somewhat unimaginatively called, went straight into all the textbooks, where it remained until 1973, when George Gatewood, at the University of Pittsburgh, took a closer look at the classic extra-solar planet. Using different telescopes and a machine that measured images automatically, he was unable to reproduce Van de Kamp’s findings. The reason was soon apparent.

While studying photographs of another star taken at the Sproul Observatory, astronomer John Hershey found that it apparently wobbled to the same degree and at exactly the same times as Barnard’s star. Either the two suns had an identical planet, or something had happened closer to home.

Checking the Sproul records, Hershey found that the movements seen by Van de Kamp coincided with the fitting of a new telescope casing, a lens adjustment and a change in the photographic emulsion.

Almost overnight, Barnard’s star and its phantom planet vanished from the textbooks.