Da Vinci fakes proliferate

How many new Leonardos can there be?

I have previously discussed various recent claims that a formerly unknown work by Leonardo da Vinci (1452-1519) has been discovered. Another was announced on September 27.

People may think they should turn out the attic.

There's a long history of "finding" Leonardos but many were not accepted and the last to make it through to general acceptance, the Benois Madonna, was proposed in 1909. Or at least, it was the last until the Salvator Mundi, proposed in 2005, won general acceptance in 2011.

What are the chances of others in the queue?

Also in 2005, a Madonna and Child with St Joseph, owned by the Galleria Borghese, was proposed. In 2008, so was the Lucan Portrait of Leonardo, privately owned in Italy.

In 2010, La Bella Principessa, also known as Portrait of a Young Fiancee, discovered in New York but in private Swiss ownership, was proposed by Martin Kemp. Now, the Isleworth Mona Lisa has been announced. The property of anonymous owners, it's been proposed before but hasn't been seen for 40 years and it's said there's new evidence.

The cases are different.

They're worth reviewing to see what makes acceptance more or less likely.

The Salvator Mundi which progressed so rapidly has long been known but was thought a copy of a work there is clear, early evidence Leonardo painted. This one's history is known since 1650 but it had been badly treated and considered a poor thing.

When it was cleaned and examined pentimenti were found - alterations during the course of making - some seen under infrared light, which was telling. The work also withstood scholarly examination and publication of detailed evidence. Its media and support are characteristic.

The Borghese Madonna was long known but thought to be by another artist. It isn't claimed there is any old record of such a work but it is like known paintings of Leonardo's youth and a fingerprint has been found on it similar to one on his undoubted Lady with the Ermine. The media and support are typical but the investigations have not been published and people remain sceptical.

The Lucan self-portrait, long-known, without historical reference, with characteristic media and support has also been said to have the master's fingerprint. There's been little study or publication and doubt remains.

Portrait of a Young Fiancee faced a very steep hill. Its history was only known since the 1990s. There's no historical reference to such a work.

Its media and support, chalk on vellum, were otherwise unknown in Leonardo's output. It had been seriously examined by Prof Martin Kemp, and the findings published. But it didn't help that a claim it bore a Leonardo fingerprint was soon thoroughly discredited.

The latest discovery, the Isleworth Mona Lisa, has been known since the early 20th-century when an Englishman, Hugh Blaker, claimed it was another autograph version of the Louvre Mona Lisa but made 12 years earlier. It's an oil on canvas, not usual for Leonardo. Its sitter looks younger than the Louvre's and lacks the latter's sfumato - special shading effects - responsible for much of the mysterious appearance of the famous lady. There have been no claims about fingerprints but an attempt has been made to give it historical authority.

Vasari, the renaissance historian of renaissance painting, said Leonardo started a portrait of Mona Lisa in 1503 which he left unfinished. He also referred to a painting of a "certain Florentine Lady" Leonardo had with him near the end of his life.

It is usually supposed there is just one work, that in the Louvre today. The Isleworth's protagonists say there were two and this is the 1503 one. There have been laboratory studies but their detail is unpublished. Prof Kemp has dismissed Isleworth as an honest but unsophisticated copy of the Louvre painting.

What this shows is that it helps if the proposed work has a long history, also if there's a historical reference to such an image, that the media and support have precedent in Leonardo's output and that scientific and other scholarly work is helpful, if it is published and withstands wider scrutiny.

I am not a Leonardo specialist but on broad grounds, I think the Isleworth painting faces an uphill battle. It looks like a copy.

Looks aren't everything as the Salvator case shows. But by contrast Kemp's Fiancee's prospects seem better.

Despite the misadventure with the fingerprint, other things have worked for her.

There's a record Leonardo wanted to get chalks. A book has been found from which it seems the Fiancee was cut which places it in the Milanese court when Leonardo was there. More study and publication are needed but she seems better placed than the latest discovery.

But this is all a work in progress.

• Peter Entwisle is a Dunedin curator, historian and writer.