Years of tests on painting dismissed as fake uncover Vermeer's 'specific extravagance'

The small painting of a woman had long been dismissed as a fake, written off by academics after a forgery scandal and consigned to obscurity.

One man, an art dealer who fell in love with the work in 1960 and traded some of his finest possessions to secure the canvas, believed it was by the Dutch Old Master Johannes Vermeer. Yesterday, Sotheby's auctioneers vindicated Baron Frederic Rolin's convictions, announcing that a panel of experts had decided the painting, Young Woman Seated at the Virginals, was indeed a genuine Vermeer worth some £3m.

The announcement, after 10 years of research, makes it the 36th acknowledged masterpiece by the painter who died 1675 aged 43, now much beloved and fêted in books and on film this year in Girl with a Pearl Earring. When it is put up for auction on 8 July, it will be the first Vermeer to come on to the market since 1921. It is almost impossible to price, but as the only work by the artist in private hands - every other work is in a museum or gallery collection, plus one in the Royal Collection of the Queen - it is thought likely to make £3m.

Gregory Rubinstein, the Sotheby's expert who has been leading the research and conservation since Baron Rolin first showed the painting to the auction house in 1993, said it had been "the most exciting thing I've ever worked on".

The early history of the painting is obscure, although it may have once belonged to Pieter van Ruijven, Vermeer's most important patron. It was certainly considered a Vermeer at the beginning of the 20th century, when it was owned by Sir Alfred Beit, a distinguished Irish collector.

But a scandal in the 1940s - when a master forger, Han van Meegeren, revealed that he had sold no fewer than seven fake "Vermeers" - raised suspicions and the work was "de-attributed". It languished virtually forgotten until Baron Rolin spotted it at a London dealers in 1960. Only when it was shown to Sotheby's in 1993 did an extraordinary detective story began.

Mr Rubinstein took the painting to the National Gallery in London where curators helpfully took down its two, larger Vermeers, both also showing women at a virginal, a keyboard instrument related to the harpsichord, for comparison. Under a microscope, conservators believed the three works were by the same hand. The art historians were not convinced.

Part of the problem was that the painting a decade ago was not as it appears today, after restoration. "It was in some ways beautiful, but rather puzzling," Mr Rubinstein said. "There were areas that looked wonderful but there were others that didn't seem quite so typical of Vermeer." There was, nonetheless, something "magical" about the work, he said.

It was decided that a scientific analysis was needed and in 1995 work began under Libby Sheldon at University College London, using a whole range of techniques from infra-red examination to X-rays. The findings were striking, though it took several more years before everyone was won over.

What the analysis discovered was evidence of three particular pigments which were rarely used by other painters or which were particularly characteristic of Vermeer.

The first was lead-tin yellow in the shawl, a pigment that was not used after the 17th century, which proved it could not be a later imitation. Next, there was Green Earth in the flesh tones, which was used rarely by other Dutch artists but was regularly found in Vermeer. Finally, and most importantly, there was evidence of the most expensive colour available to 17th century Dutch artists, ultramarine, made from ground lapis lazuli. Its great cost meant that it was used only sparingly, yet Vermeer used it extensively, though invisibly, in the creamy tones of his background walls. "This specific extravagance is something that has never yet been found outside the work of Vermeer," Mr Rubinstein said.

Analysis of the canvas showed that the warp and weft exactly matched the canvas used in a Vermeer of a similar size, The Lacemaker, in The Louvre in Paris, and were likely to have been cut from the same cloth. Even the priming layers matched, suggesting they had been prepared at the same time.

Mr Rubinstein took the painting to New York for comparisons with the Vermeers in the Metropolitan Museum. Walter Liedtke, its curator of European paintings, was so struck by the work, he asked if he could borrow it as a last-minute addition to the Vermeer show which opened in New York in March 2001 and later came to London.

This prompted such debate that it decided to convene a panel of academics and conservators to resolve the matter. They met over 18 months to review the evidence. At the same time, some of the murkier parts of the painting were restored. The effect was transformative. "It was a great moment," Mr Rubinstein said yesterday. "It was absolutely extraordinary and very moving. It was really clear that the cool, calm atmosphere and light and sense of space that we associate with Vermeer really were there."

Sadly, the baron had died. His heirs are selling the work, though Mr Rubinstein thinks the baron would have done so. What mattered to Baron Rolin was that the painting he fell in love with should be recognised for what it really was.