For centuries art experts have attempted to solve the riddle of how Vermeer painted canvasses of such sublime realism that Proust said he wanted to die under one. Now one of the great mysteries of art history has finally been solved - Johannes Vermeer's delicately composed snapshots of Dutch 17th century domestic life look so like photographs because he used a camera.
A study by Philip Steadman, of University College London, has proved beyond doubt that the Dutch master, whose reputation now ranks with that of Rembrandt, used a camera obscura to trace out the scenes he later painted. What is more, like all the best detective stories, Vermeer left one teasing clue of his "guilt" so a clever investigator might one day rumble him.
Professor Steadman, who has spent nearly 30 years poring over his canvasses, said it had long been suspected that Vermeer, who was all but forgotten for 200 years after he died in penury in his home town of Delft in 1672, had been up to something.
Even the Victorians, who were largely responsible for reviving his fame, had noticed something odd, and puzzled about why the paintings displayed similar qualities to their new fangled photographs. "An article in the British Journal of Photography in 1890 picked this up in The Soldier and The Laughing Girl," said Prof Steadman. "It looks like a snapshot grabbed by a modern camera, and this sort of perspective was very unusual in Vermeer's time, an effect which becomes even more intriguing when you consider the precision of the maps in the background. These, I discovered, were real maps which were copied with quite astounding accuracy. The Victorians also noticed that the blurring of some of the objects in his paintings seemed to mimic the out-of-focus areas of photographs."
But no one could quite prove that he had used a camera obscura, primitive forms of which had been around for a century. Even Leonardo da Vinci had toyed with these relatively simple contraptions that used lenses and mirrors to reflect images on to tracing paper inside darkened booths. But, according to Prof Steadman, the images they produced were too dim - even in the sunshine of the Mediterranean, never mind the marsh mists of Holland - to be any good to an artist.
He began his remarkable investigation by making detailed geometric analyses of a series of paintings which he suspects Vermeer painted in the front room of his mother-in-law's house in Delft towards the end of his life. X-rays had already shown that unlike most artists before and since, Vermeer did not bother with under-drawings or sketched outlines and appeared to have painted straight onto the canvas.
The mathematics, using the grid pattern of the black and white marbled titles of his mother-in-law's floor, and the comparisons taken from the chairs Vermeer painted, which Prof Steadman found in the attic of Delft museum, suggested strongly that he was using some sort of lens.
However, in The Music Lesson, where a mirror reflects the image of the part of the room the artist is sitting in, an easel but no painter is shown where the mathematics says the camera obscura should be.
Prof Steadman, who revealed his findings at the Hay-on-Wye festival, believes this was a deliberate tease. "You can see the easel and the canvas, even the foot of the stool where the artist might have sat, but not Vermeer himself. In putting the tools of his trade at the precise point where you would expect to see a camera, I suggest, Vermeer is laying a deliberate false trail."
Prof Steadman reconstructed the room for his new book, Vermeer's Camera, and found that a camera positioned where the artist's camera obscura might have been produced an almost identical picture. "The only major difference was the shadow behind the mirror, which I think Vermeer exaggerated for effect," he said.
It was another mirror in another painting from the The Mother-in-Law's Front Room series - soon to go on show at the Vermeer exhibition at the National Gallery in London - that provided Prof Steadman with the clinching proof. In Allegory of the Faith, much of the left foreground is masked with a curtain.
But close inspection of three blobs reflected in a sphere in the extreme background revealed the reflection of the three windows that appear in the other paintings, the last of which is itself shut tered off as it would have to have been if there was a camera obscura there.
"That is exactly where the camera would have to be, and at precisely that point, Vermeer with his own hand has painted a mysterious black box - his camera obscura."
Further mathematical analysis, he said, showed that the projected image was exactly the same size as the painting, another telltale trait of the camera obscura.
But Prof Steadman is adamant that the discovery does not devalue the worth of any of Vermeer's paintings or show that he was in some way cheating.
"It may seem to be a form of cheating, but what he did was no shortcut by any means." he said. "He would have pricked the outline of what he was doing out with a pin in a piece of paper and then covered it with ground charcoal so it left a stain on the canvas."