So how can you be sure the auction houses are intercepting any dodgy Dalís before they make it onto the rostrum? Only last year, a former art teacher, Rizvan Rahman, was jailed for 18 months for selling UK galleries some £180,000 worth of fakes, including a £35,000 Lowry lookalike.

Naturally, there’s a fair amount of caveat emptor when you’re buying at an auction. None the less, the experienced London art-verifier Bendor Grosvenor says auction houses are keen to avoid a stain on their good name. “If it turns out there’s any kind of justification for questioning a work’s authenticity, I can’t envisage an auction house doing anything other than refunding the money,” he says. “Fighting the case just isn’t worth the potential damage to their reputation.”

Absolutely right, says Julian Roup, head of press at Bonham’s Auctioneers. Though he won’t discuss the Vekselberg case, he can assure customers that the company takes allegations of fakery most seriously.

If there’s any question over authenticity, “an immediate investigation is launched,” he declares. “We bend over backwards to establish the facts.”

And it’s amazing what a bit of forensic work can find out. Visit the web-archives of Freemanart, and you discover the tiny giveaway clues that can mean the difference between a painting securing you a fortune – or a spell in prison. Take the seemingly genuine Gainsborough, for example, where microscopic examination showed that the artist’s signature had been traced in pencil first for the forger to copy. Or the otherwise perfect Gauguin, given away by the paper it was painted on: the corners were straight, when they should have been round.

But while science’s principal contribution is to downgrade by proving that paintings are fakes, it can sometimes work the other way. Not just by showing that the Mona Lisa is holding a shawl (she is, though it’s invisible to the naked eye), but by upgrading a previously disregarded work.

For example, thanks to the miracle of dendrochronology (the science of dating objects using tree bark), a painting of Mary Queen of Scots, thought to have been an 18th-century copy, has been promoted by the National Portrait Gallery to the status of 16th-century original (tree ring analysis suggests 1560-1592).

Meanwhile, a Gloucestershire art collector, Frank Faryab, has, after five years and a lot of consultants’ fees, gathered sufficient scientific evidence to convince the art world that his oil painting of a distant ship is not the doodle of some old sea dog, but a lost Turner masterpiece, worth as much as £4 million.

Which is heartening news for all of us who dream of stumbling upon a Leonardo da Vinci in the lumber room. But that’s not to say, though, that the art world is always ready to welcome new paintings with open arms, just as it doesn’t like to see authenticated works discredited.

“For a previously disregarded work to be declared authentic, or a previously accepted work to be downgraded, a lot of people have to be proved wrong,” says Dr Grosvenor. “And if there’s one thing people in the art world don’t like, it’s being proved wrong.”