Less than a month after Rubens's Massacre of the Innocents sold for a record £49.5m at Sotheby's in London, questions have been raised about whether it is by the Flemish master after all.
Scientific dating of the picture suggests that it may have been painted years after the dates 1609 to 1611 ascribed to it by art historians in the auction house's sales catalogue.
If this were the case, the painting's attribution would be cast into doubt, as its purist classical style is radically different from the work Rubens is known to have been producing by 1615.
News of the disparity has been greeted with astonishment by senior academics including James Beck, Professor of Art History at Columbia University, who described it as "potentially devastating".
However, Sotheby's insist that the painting, credited to Jan van der Hoecke until it was re-attributed after being discovered in an Austrian monastery, was authenticated as a Rubens by the world's leading experts on the artist.
The new debate arises from a complex analysis of the dating of one of the painting's wooden panels included in a detailed technical report commissioned by Sotheby's.
In it, Dr Peter Klein, one of the world's foremost dendrochronologists, states that the earliest possible felling date for the tree from which the wood was taken is 1607. However, he goes on to write: "More plausible is a felling date of 1611 ... 1613... 1617 plus."
Dr Klein, who had to use probability theory to judge the number of outer sapwood rings that had disappeared from the Baltic oak timbers before their incorporation into the painting, adds: "With a minimum of two years for seasoning, an earliest creation of the painting is possible from 1609 upwards. Under the assumption of a median of 15 sapwood rings and two years for seasoning, a creation is plausible from 1615 upwards."
Dr Klein told The Independent on Sunday: "The statistical numbers are only a help for further discussion. I am not an art historian. I cannot say that it is a Rubens or not."
Rubens did not sign his work. While many of his later paintings have been securely attributed with the help of accompanying documentation, no such written records exist for his earlier pieces.
Dr Klein said: "The entire attribution is only based on an idea of what Rubens's style probably would have been after he got back [to Antwerp] from Italy. That seems to me quite shabby."
Critics point to several further findings in the document, including an analysis of the picture's colour scheme by Dr Nicholas Eastaugh, in which he highlights a pigment, orpiment, that has never before been found in a Rubens.
The painting's new owner, Lord Thomson of Fleet, who was bought it by his son, David Thomson, the billionaire newspaper tycoon, was unavailable for comment.
A Sotheby's spokeswoman said the study of growth rings on wooden artefacts was a matter of "if, but and maybe".
"The art historians we consulted were all agreed that this was a Rubens and that the most likely date for it was between 1609 and 1611. We didn't just think to ourselves, 'This looks like a Rubens.'"