Russian and American researchers have come up with a radically new recipe for making oil.

Take some rusty iron, add a soupçon of marble, throw in a bit of water, then heat and squeeze the mixture at temperatures and pressures thought to exist more than 100 kilometres below the surface of the Earth. At the right pressures and temperatures, some of the iron, marble and water turn into methane, ethane, butane, propane and a host of other hydrocarbons.

The successful experiment, reported today in the Proceedings of the National Academy of Sciences, is thought by its authors to be the most conclusive laboratory evidence for what is called the abiotic theory of oil formation.

For the better part of half a century, Russian scientists and a few others have been arguing that a significant part of the petroleum on the planet is naturally generated from within the deep Earth by heat and pressure. This view has placed its proponents in direct opposition to most of the world geological community that has for several hundred years believed that oil, gas and coal are the residue of decaying plant and animal matter.

Beyond its theoretical implications, if the results can be replicated, the experiment provides a physical basis for the belief that oil will be found in many deep Earth locations where existing theory says it shouldn't be. The economic result would be a world that John Kenney, an American physicist who is one of the co-authors of today's article, has previously characterized as a place that isn't running out of oil, but instead is "running into oil."

What the experiment doesn't demonstrate is a new way of producing oil, as scientists who have looked at the paper suggest that the energy required to produce oil from water, marble and rusty iron would be much greater than the energy content of the petroleum.

Even before it was published, the result was evoking strong comment from the scientific community. "It is remarkably ignorant of the literature," says University of California at Berkeley geophysicist Raymond Jeanloz. Not only does it not mention any papers published by Thomas Gold, a Cornell University astronomer who has argued for the deep-Earth hypothesis for a quarter century, but it fails to quote other work that showed when carbon and hydrogen compounds were combined at high temperatures and pressures you ended up not with hydrocarbons, but diamonds and free hydrogen.

Geological Survey of Canada research scientist Lloyd Snowdon says that numerous experiments starting with organic matter "produce a slick that for all the world looks like crude oil." And that if non-organic processes do contribute to the oil supply he believes "only one part in kagillions is abiotic."

Despite this skepticism Peter Bernath, a chemist at the University of Waterloo, also reflects a general consensus when he says "it is hard to argue with the paper's data."

The data say that when the mixture was heated to 1,500 degrees Celsius and subjected to pressures 50,000 times greater than on the surface of the Earth, the petroleum compounds formed were stable after the mixture was cooled and depressurized. Mr. Kenney and his Russian co-authors have previously argued that the oil from great depths could make it to the near surface through an eruptive process similar to that which brings up diamonds from the deep Earth.