ART MARKET / When a fake is not a Keating: It may be by Samuel Palmer, by a master faker, or by an unknown. Who is the creator of a suspect watercolour at the Royal Academy?

QUESTION: when is a Samuel Palmer not a Samuel Palmer? Answer: when it's done by Tom Keating. Or that is what art historians would nowadays have one believe. 'Did you know there was a Tom Keating at the Royal Academy?' a respected art historian asked me on the phone the other day, after we had discussed something quite different. I didn't. I didn't even know that one of the Palmers in the 'Great Age of British Watercolours' exhibition at the Royal Academy was under suspicion of being a fake.

Samuel Palmer brought an extraordinary mystical vision to landscape painting for about six years around 1830. The inspiration faded and, though he imitated it later in life, he never recaptured his youthful inspiration. The drawings remained in his family and virtually unknown until an exhibition at the Victoria & Albert Museum in 1926 of 'Samuel Palmer and other Disciples of William Blake'. It had a huge impact on British artists and connoisseurs; the artists imitated his work and the connoisseurs bought the watercolours from his son, A H Palmer.

In the early days his drawings were not expensive and no one is known to have started faking his work until Tom Keating took it up in the 1960s and 1970s - when prices had risen. But that is not to say that others did not try making Palmers, either for fun or to earn a few irregular pounds. Indeed, in my opinion the suspect painting at the Royal Academy - a very orange watercolour called Harvesters by Firelight - makes it clear that someone did.

I have a special interest in Tom Keating, since I was the journalist who unmasked him as a picture-faker back in 1976 - on account of his fake Samuel Palmer watercolours. So I rang round the Palmer buffs. 'When was it that your friend started out?' laughed Martin Butlin, former keeper of the British collection at the Tate. 'We know this one was in existence by 1951. Was Keating at work by then?' Keating had just left art school by 1951 and was already copying Old Masters but not, to my knowledge, making Palmers. I found a couple of Keating Palmers tucked away at the back of a drawer. One is reproduced here along with Harvesters: a clear demonstration, in my view, that the watercolour in the R A show is not by Keating. The way the foliage is treated is, perhaps, the most obvious giveaway. Keating indicates leaves by making quantities of individual brush strokes. Whoever did the R A drawing has used a black outline for the shape of the trees and filled it in with wash.

That puts paid to the Keating idea. But having a suspect Palmer in the Royal Academy show - acknowledged as such by the exhibition's organiser - is a very unusual bit of miscalculation. As far as I can make out from the embarrassed participants, no one realised it wasn't by Palmer until the catalogue was written and the show was on the walls.

The drawing belongs to the National Gallery in Washington, which received it as a gift from Paul Mellon in 1986. Mellon, who inherited one of the largest fortunes in America, has given the National Gallery - which was founded by his father - more than 800 pictures. It was not difficult for one fake to sneak in among them.

Mellon himself is a passionate anglophile and has formed the most important private collection of British art anywhere in the world, even rivalling the Tate. It is now mostly housed at the Yale Center for British Art. He bought the 'Palmer' watercolour at Christie's in 1981 for pounds 77,000 through his friend and agent John Baskett, then a Bond Street dealer. 'I've always thought it was a Palmer,' John Baskett told me last week. 'There was no doubt at the time.' Mellon's reaction to my enquiry was 'No comment'.

It was not until 1988 that a catalogue raisonne of Samuel Palmer's work - a catalogue, that is, which lists all known Palmers and discusses them - appeared. Raymond Lister, its author, told me: 'I included Harvesters because I felt it couldn't exactly be rejected as a Palmer - but I'm not as convinced as I was. It could be an original that someone's played about with. I don't think the red figure in front is by Palmer.'

There is an inscription on the back, Lister points out, which is particularly suspicious. It is supposed to be in the hand of Palmer's son, who did inscribe several drawings; if it was not written by him, it must have been written by someone who was consciously trying to turn the picture into a Palmer.

The inscription explains that the wild orange glow over the scene is: 'The reflection of one of the incendiary fires, fires in Kent, I think about 1830 done I think the next day. The building is, I think Ightham Mote. A H P. Subject the harvesters hurrying away the last of the harvest.' Lister wrote in his catalogue: 'The building is not Ightham Mote. Such an indecisive statement is uncharacteristic of A H Palmer. Moreover, it does not make sense: it would have been unnecessary to 'hurry away' the last of the harvest for the 1830 incendiary attacks were aimed against stacks and barns and not against growing crops'.

It is not entirely clear when doubts about the painting began to surface. 'I remember considerable enthusiasm for it when we had it for sale in 1981,' Anthony Browne of Christie's told me. However, both Martin Butlin and his successor at the Tate, Andrew Wilton, tell me they did not think it was by Samuel Palmer at the time.

Andrew Wilton is the curator of the Royal Academy show and says that the watercolour slipped in because the exhibition had to be mounted in such a hurry - he only had six months to put it together, from start to finish.

'The Palmers we wanted were not available,' Wilton said. 'I rather hoped that as the National Gallery in Washington offered it as one of the things they were happy to lend, they had sorted out the attribution.' No such thing: 'I first heard of the doubts when the Royal Academy rang me in Italy two weeks ago,' Andrew Robison, the National Gallery's curator of drawings, told me. 'If it's not by Palmer, we won't have any problem about changing the attribution but I don't yet understand what's wrong.' Piers Rodgers, secretary of the Royal Academy, maintains that the argument over the attribution is not yet resolved. 'Scholars have been discussing it for ages,' he said. 'We knew that when we put it in the show.'

The inclusion of the drawing in the exhibition has been hard luck on the pundits. Simon Jenkins, former editor of the Times, described Palmer's 'idylls of the gloaming' in a column about the Academy show, pointing especially to 'his brilliant Harvesters by Firelight'. Brian Sewell in the Evening Standard illustrated the painting in colour to demonstrate what he thought of English watercolours - but saved his reputation as a connoisseur by describing it as 'a sickening confection of glutinous marmalades coarse cut'.-