Strange stories have long circulated around the coelacanth, the ‘living fossil’ fish discovered off the coast of South Africa in the 1930s.

The latest are claims by French researchers who say they were the first scientists to find the Indonesian variety, in 1995. But the photograph they sent to Nature to support their claims has been denounced by another researcher as a fake.

Until recently, coelacanths had only been found on the west side of the Indian Ocean. The first recorded Indonesian sample was discovered by a US biologist, Mark Erdmann of the University of California at Berkeley, who published news of his find in Nature (see Nature 395, 335; 1998).

In a recent submission to Nature, the French team — Bernard Séret, Laurent Pouyaud and Georges Serre — say they were unable to register their specimen in 1995 because it failed to reach the museum to which it had been sent. They say Serre photographed the fish at the time, then lost the picture (their only other evidence) while moving house, and only found it again this year.

Vive la différence? The French fish (right) bears a striking resemblance to one that appeared inNature.

But Nature staff noticed that the fish in the new photograph appears virtually identical to the one caught by Erdmann. When contacted, Roy Caldwell, a co-author of the 1998 paper, scrutinized the photograph via a picture-editing computer program and said, “I am 100% certain the image is a fake”.

When Erdmann spotted a coelacanth off the northeast coast of Sulawesi in 1998, his find expanded the geographical distribution of the fish by roughly 10,000 km. But Georges Serre, a consultant for what is now the French Institut de Recherche pour le Développement (IRD), has long claimed that a 10-kg specimen was caught in 1995 in the Bay of Pangandaran, in Southwest Java.

Serre says he gave the specimen to a fisherman to hand over to the Indonesian fishery service. But the man gave it to a museum, from which it was stolen. Only recently, according to Serre, did Pouyaud, an IRD geneticist in Jakarta, track down the coelacanth to a private collection, whose owner refuses access to the fish.

The French team's finding, if confirmed, would further extend the distribution of this elusive creature, as it was caught more than 2,000 km from the spot where Erdmann found his 1998 specimen, suggesting a large distribution in the Indo-West Pacific region.

The whole issue is already shrouded in controversy. After analysing the specimen that Erdmann had given to the Indonesian authorities, Pouyaud and his colleagues at the Indonesian Institute of Science named it as a distinct species, Latimeria menadoensis — to the chagrin of Erdmann, who had been analysing tissue samples independently.

Caldwell says that coelacanths have highly individual spot patterns; the pattern in the two photographs is virtually identical. The bulky fish in the 1998 photo was swimming in the sea, yet is seen in the identical position, in the new photo, while lying on a slab. Shadows and other details seen when the image is magnified add to Caldwell's impression that the new photograph has been manipulated.

Serre still claims that the photograph is authentic, though he now says it was taken by a friend who later died and whose widow gave it to Serre before moving abroad.

Séret, who is an ichthyologist at the Muséum National d'Histoire Naturelle de Paris, admits that the two photographs do appear to show the same fish. “This is very embarrassing,” he says.