The Emergence of a Fraud: A Portrait of Rembrandt
In 1939, with no painting by Rembrandt van Rijn in the collection, the National Gallery of Canada bought this work as a self-portrait by the artist. The painting was virtually unknown and the dealer presented it as a discovery, claiming that its true appearance had been disguised by extensive later repainting. Encouraged by evidence seen in x-radiography, he had the picture cleaned, removing the repaint from the cap, hair, hands and elements of the costume. A photograph shows it during this process. The restoration left the painting closer to another little-known portrait at Woburn Abbey, England. The dealer argued that the Ottawa work was the original, and the painting at Woburn a copy. Spurred on by this sense of discovery and reassured by the technical evidence and positive testimonials by experts, the Gallery felt comfortable purchasing the painting at what was described as a moderate price.
The first opportunity to compare the two versions arose in 1958, when the Woburn painting was exhibited at Morgan’s department store in Montreal. As doubts had arisen about the Ottawa version, the Gallery took its painting for comparison. Seeing the works side by side, Gallery staff found the Woburn version more convincing.
A decade later, when the two versions were shown together in Toronto and Montreal as part of a major exhibition on Rembrandt and his pupils, the Gallery labelled its work a “school” piece. This imprecise designation can encompass pupils, assistants and those working in the style of an artist, perhaps even decades later; in this case, it means only that they knew the work was not by Rembrandt and could offer little else as a suggestion. This reflected broad scholarly opinion, which also had doubts about the Woburn version.
Establishing the authorship of paintings attributed to Rembrandt had long been a problem, and the 20th century had seen the publication of scholarly catalogues with differing opinions about Rembrandt’s paintings. In 1968, the Dutch government funded the Rembrandt Research Project (RRP) with the intention of creating an authoritative catalogue. This group was mostly comprised of art historians, with little background in technical study, which was still a developing field at the time.
Agreeing with the scholarly consensus, the RRP decided that the Ottawa painting was probably a late 17th- or early 18th-century copy of a lost original. The Project believed that the Woburn version was closer to Rembrandt; they thought it could have been made by a contemporary, but one with no direct connection to the artist, and therefore also a copy. The x-radiograph of the Woburn version, however, indicates that this is not the case, since it shows significant alterations to the angle of the sitter’s body made during painting – the kind of changes not seen in a copy. As well, the handling of the painting of the face shows that the artist was very attentive to Rembrandt’s distinctive, late technique – while the general appearance is similar to his works from about 1640, the touch is influenced by paintings from the 1650s–1660s. There is no need therefore to posit a lost original: the artist of the Woburn painting presumably based his work on Rembrandt’s numerous self-portraits in painted and printed form.
While it is difficult to be certain, it seems likely that the Ottawa painting was made in the first half of the 19th century. Although the palette is limited and there are no datable pigments, the nature of the paint is consistent with commercial manufacture at that time. While it likely began life as a faithful copy of the Woburn portrait, it developed into something else entirely.
At a certain point in its genesis, the Ottawa painting was altered: the medallion seen in the Woburn version was made into an overt cross; his hands were modified; the background was made darker and more architectural; his hair made longer; and the hat made more extravagant. (Remnants of these changes can be seen in the photograph of the partly cleaned work.) These changes may have been made to allow the painting to better complement an elaborate portrait thought to be of Saskia, Rembrandt’s wife. These two portraits were first mentioned in 1854, when hung as a pair in the collection of the Earl of Listowel.
It is very likely that the embellishments that transformed the copy into an independent work were added by the copyist. The 1939 restoration by the dealer, seeking to return the painting to a more “authentic” state, actually left it a hybrid, with elements of its earliest and final stages. Since all layers of the paint would have been of a similar character, it should have been clear to the dealer – who also may have been the restorer – that this was not a simple case of removing later repaint from an authentic Rembrandt.
The painting Portrait of Rembrant is currently on view in C205 at the National Gallery of Canada. Share this article and subscribe to our newsletters to stay up-to-date on the latest articles, Gallery exhibitions, news and events, and to learn more about art in Canada.