The Bimini Road

Beachrock in 18 feet of water, sold as Atlantis since 1968

The Bimini Road lies in about 5.5 metres of clear water off the northwest coast of North Bimini in the Bahamas: a line of roughly rectangular limestone blocks running 0.8 km northeast to southwest, with a pronounced hook at its southwest end. The blocks reach 3 to 4 metres across, 2 to 3 on average, their corners rounded so that they look like giant loaves of bread. Two shorter, narrower lines of smaller blocks, about 50 and 60 metres long, lie parallel to it and closer to shore.

On 2 September 1968 the zoologist Joseph Manson Valentine and the divers Jacques Mayol and Robert Angove swam over it and called it a pavement. It became a road, a wall, a pier, a breakwater, the work of Olmecs or of Atlantis. Valentine, the writer Charles Berlitz and others argued it was man-made; one expedition, sponsored by North American Rockwell and led by the astronaut Edgar Mitchell, took along psychics from the Edgar Cayce Foundation.

The geologists looked at it too. As early as 1971, a short paper in Nature by W. Harrison argued that the blocks were remains of an originally coherent bed of coquina limestone misread as prehistoric stonework, and that the marble and cement cylinders found nearby, hailed as pillar fragments, had a more mundane explanation. In 1980 the anthropologist Marshall McKusick and Eugene Shinn of the U.S. Geological Survey, who studied Bimini at his own expense, reconsidered the Bahamian Atlantis in the same journal. The stone is beachrock, shell hash cemented by carbonate, which forms naturally in the Bahamas. Detailed surveys by Gifford and Ball found no second course of blocks, no block set squarely on another, no foundations cut into the bedrock and no regular grooves that could be tool marks, and the weathering is so deep that any original surface would long since have gone.

Dating agrees. In 1978 the University of Miami's radiocarbon laboratory dated a core Shinn had taken in 1977. The shells came out at about 3,500 radiocarbon years and the cement binding them at about 2,800, the order expected if loose sediment was cemented into rock some 1,300 years after it gathered. Davaud and Strasser concluded that the beachrock formed beneath the surface of the island and was exposed by coastal erosion only about 1,900 to 2,000 years ago.

Believers still dispute every point. Meanwhile scientists taking cores and the countless visitors who have chipped off souvenirs have left modern tool marks of their own for the next investigator to puzzle over.