In 1937, the Dutch art historian Abraham Bredius examined a canvas from an undisclosed private collection: a biblical supper scene signed with Johannes Vermeer's monogram, unlike any of the three dozen paintings already attributed to the Delft master. Bredius, the leading Vermeer authority of his generation, pronounced it genuine. The Museum Boijmans Van Beuningen in Rotterdam acquired The Supper at Emmaus soon after, ahead of the Rijksmuseum in Amsterdam, and displayed it as a lost early work.
During the German occupation, the same painter produced another "lost" Vermeer, Christ with the Adulteress. A banker and art dealer in the occupied Netherlands sold it on to Hermann Göring. When Allied investigators traced the sale back to its source after liberation, the man who made it was arrested in Amsterdam on suspicion of collaboration: selling a national treasure to the occupier, a charge punishable by death.
His defense was that there was no treasure to sell. He said he had painted the Vermeer himself, and several others besides. A confession made to escape a capital charge proves nothing by itself: a man facing execution has every reason to say whatever keeps him alive, and the same connoisseurs who had judged the paintings genuine once could not simply announce they had been wrong and be believed either.
So in June 1946 the Amsterdam court moved the question out of the hands of connoisseurship, appointing an international commission of chemists and conservators, led by the Belgian chemist Paul Coremans and the Dutch justice-ministry chemist Wiebo Froentjes, to test eight disputed canvases against the physical record and Van Meegeren's own account of his method. X-rayed, several showed two separate networks of surface cracking, one buried under the visible paint and matching nothing above it: proof the visible cracking had been staged, not three centuries old. The paint itself failed a test real seventeenth-century oil always passes, shrugging off prolonged exposure to a caustic-soda solution that dissolves genuinely old paint rapidly.
A chemical color test found phenol and formaldehyde in the paint itself. Heated, resin seized from the painter's own studio gave off the smell of both, and its refractive index, close to 1.59, matched the paint on the canvases and sat nowhere near the 1.50 of dried linseed oil or the 1.51 to 1.54 of a real period resin. The compound, a synthetic resin the chemist Leo Baekeland had first made in 1907, had been dissolved in lavender oil, worked into the pigment, hardened by several hours of heat kept below about 105 degrees Celsius, then cracked on purpose by rolling the stiffened canvas over a rod. None of that chemistry knew or cared what the answer would cost the man who mixed it. The commission filed its report with the Amsterdam court in January 1947; two of its members, in a 1949 account of the investigation, put the finding as plainly as a panel of chemists could manage: not Vermeer's hand, but very possibly the forger's own.
With the forgery established, the collaboration charge fell away: there was no stolen treasure, only a very good fake. Van Meegeren was tried in 1947 for forgery and fraud instead, convicted, and sentenced to a year. He died of a heart attack within weeks, before serving a day of it. A signature can flatter whoever is looking at it. A molecule invented in 1907 has no opinion on who needs the painting to be older.