The painting was bought by a UK dealer for €10,000 (£8000) at Christie’s in Amsterdam in June 2015 when it was catalogued as ‘after Sir Anthony Van Dyck’.
Restoration work commissioned by the purchaser removed strips of canvas and overpaint that had been added in the 19th century, to reveal a smaller 22in (56cm) wide oil sketch. It was offered at Sotheby’s in New York on 25 January 2017 as Sir Peter Paul Rubens, Study Of A Horse With A Rider.
Sotheby’s catalogue states that it is a “newly discovered work [and] a rare example of a large-scale animal study by Rubens”.
However, Christie’s stands by its own view. In a statement it said: “We remain confident in the cataloguing from the Amsterdam auction and understand there is no consensus of opinion from leading scholars and experts in the field on the new attribution of the work.”
Christie’s did not give extra details of those who had questioned the picture.
‘Cannot be completely sure’
Both Ben van Beneden, director of the Rubenshuis Museum, and Professor Arnout Balis, chairman of the Centrum Rubenianum in Antwerp, were credited in the Sotheby’s catalogue as having confirmed “the attribution to Rubens after first-hand inspection”.
Balis has now said he would like further research to be carried out before being “completely sure” it was by Rubens.
Speaking to ATG, Balis said: “When I was presented with this painting I said it might be a Rubens. But there is a technical anomaly: the background is beige and usually the background would be greyish.”
“At this point we cannot be completely sure. It could be a Rubens but it needs further investigation. Let’s wait and see, it needs further study.”
Van Beneden, director of the Rubenshuis Museum, told ATG: “It does not surprise me that opinions may be divided.
[With] Rubens this is often the case… The scholarly debate is open and ongoing.
“I am happy to confirm that I endorse the attribution. Style and technique are at ease with what we know from Rubens; horse and rider are in perfect balance and the foreshortening is impressive.
“In my view this technical oddity does not necessarily pose a problem as regards the attribution, but it is a conundrum worthy of further thought.”