When, in 1950, the 11th Duke of Devonshire was hit with a swingeing ￡7m inheritance tax bill, he did what many stately homeowners do and gazed up at his walls to see what art might be sold. The duke found that he had three Rembrandts, and suggested the nation choose one of them in part payment of the tax. Thus in 1957 a fine, signed Rembrandt, “Old Man in an Armchair” (1652), went on display at the National Gallery in London.

Barely a decade later, however, the picture was relegated to “follower of Rembrandt”. “Imposing as the mood is,” said art historian Horst Gerson in his 1968 catalogue raisonné of Rembrandt’s paintings, “the overall structure of the picture is very weak, even contradictory [with] divergences not to be found in Rembrandt’s autograph portraits from this great period.” Ever since, the picture has been largely ignored. In 2010 it featured in the National Gallery’s mea culpa exhibition Fakes and Mistakes.

And yet, this year, the world’s leading authority on the artist, Dr Ernst van de Wetering of the Rembrandt Research Project (RRP) in the Netherlands, said the picture was, after all, by Rembrandt. “It is a very important painting,” he says, a “painting about painting” that heralds a reinvention in Rembrandt’s technique in the 1650s. Van de Wetering believes Gerson made “a vast mistake”. The picture should not be seen as a mere portrait. I agree with him.