The hunt for the oldest records of life is not a simple task," wrote the eminent paleontologist J. William Schopf in his 1999 book "Cradle of Life: The Discovery of the Earth's Earliest Fossils." "Like trying to solve a really complicated mystery when only a few clues have been revealed, it's easy to make mistakes. These can be worse than embarrassing, major blunders that set back the search for knowledge." To Schopf's eyes, the most serious such blunder of recent times is the claim that a meteorite found in Antarctica, ALH 84001, held the fossils of Martian microbes. From the moment it was unveiled at a NASA press conference in 1996, Schopf has been a trenchant critic of the purported evidence of Martian life. Now a team of scientists is pointing to a similar blunder considerably closer to home. According to a paper in last week's issue of the journal Nature, microscopic fossils found in Australia aren't fossils at all. And to give the whole story a personal irony, the scientist who devoted much of his life to studying the Australian fossils, and who proclaimed them to be the earliest evidence of life on earth, was none other than Schopf himself.

Schopf was a professor at UCLA in the 1980s when he found what he took to be fossilized microbes in a chert--a rock made of grains of quartz--from a site called Chinaman's Creek in Western Australia. These microscopic fossils lack the drama of dinosaurs, or the charm of trilobites. But in the 4 billion or so years before these creatures arose (the so-called Precambrian era), microbes were the only sort of life earth had to offer. Since the Apex chert was 3.5 billion years old, it appeared that Schopf's fossils came from far earlier in the Precambrian than any other life form ever found.

Not only were the fossils very old, but they were also surprisingly complex. Most scientists, Schopf included, had expected fossils from that early on, if there were any at all, to be very simple, primitive things. But some of Schopf's fossils seemed to be filaments made up of rows of separate cells, "like garden hoses with partitions"; Schopf argued that they were photosynthetic blue-green algae. If so, the Apex fossils were not just evidence that life had evolved early--they were evidence it had evolved quickly, as well. All subsequent stories about the origin and evolution of life have had to take that speed into account.

Now Martin Brasier of the University of Oxford is reinterpreting Schopf's evidence. He starts not with the fossils, but with a careful study of the rocks around them. "People go in and grab the fossils and hope the context will come out later," he explains, "and that's precisely where they go wrong." As a result of these studies, where Schopf sees the chert as a sedimentary rock laid down near a seashore, Brasier and his colleagues have come to see it as something created in a hydrothermal vent. They point to evidence from minerals that no oxygen-producing photosynthesis could have occurred there, and from the makeup of the rock they infer that the temperature at which the chert was formed was above 200 degrees Celsius--too hot for life.

In early versions of their critique, Brasier and his colleagues doubted that there was any organic matter in the fossils at all. Since then Schopf has used a new technique to show that the structures identified as fossils are rich in organic molecules. But "organic," a chemical term that applies more or less to all molecules with carbon in them, isn't the same as "biological." Most organic molecules are produced by living creatures, but that need not always have been the case. Brasier argues that the hydrothermal vent could have produced organic molecules from carbon monoxide and hydrogen in a process similar to the way the South African government made gasoline substitutes while under sanctions. Schopf is scornful of the idea, seeing it as "way, way out in left field," and "exceedingly implausible." "Nonbiological organic matter has never been found in the geological record ever," Schopf says. "I'd be delighted if there were such a thing, but you do have to have some criteria to say this is biological and this is not. And they don't have any criteria."

What about the filament-like structures in which the organic molecules were found? Brasier sees them as part of the fabric of the quartz. He argues that the structures come in many different forms, only some of which look like algae. Some look like no bacteria ever seen, and some appear to be branched--a no-no for primitive blue-green algae. Schopf denies that there's any branching: he says that Brasier has mistaken folded filaments (like an AIDS-awareness ribbon) for branched ones (like a Y). He puts the confusion down to a lack of experience with the microscopes used to study such structures. But not everyone who has seen the samples agrees with Schopf's interpretation. Andrew Knoll of Harvard, an expert on Precambrian fossils, agrees with Brasier that the structures identified as microfossils are just a subset of a wide variety of similar structures in the rock, many of which are clearly not biological: "Brasier has a case that needs to be taken seriously," he says.

Schopf remains adamant that the organic material in the structures is a clincher. His analysis shows it to be much the same as the organic material found in a wide variety of undisputed fossils, and so if the organics in the Apex structures were formed by some strange new process, all the others would need explaining the same way. "If the Apex things are wrong... it would mean that all Precambrian fossils were not fossils, and I'm not willing to accept that."

There is another possible explanation for the organics, though geochemistry suggests that there was already microbial life on earth as long as 3.8 billion years ago, and those microbes would have left behind residues of organic matter even if they weren't preserved as recognizable fossils. Thus the organic matter in the disputed filaments could come from microbes even if the filaments are not fossils of microbes. According to Knoll, Precambrian cherts often contain organic matter derived from living things and then concentrated into clumps by the growth of crystals and the heating of the rocks. But those clumps aren't fossils.

While Schopf and his critics go back to the rocks for more and more studies, it's hard to imagine the supporters of the case for the Martian meteorite fossils being so noble as to avoid a faint glow of schadenfreude as their foremost critic gets criticized himself. But if anything, the new dispute about ancient fossils on earth just makes the Martian case even harder to prosecute. If people can't agree about rocks from a known location that can be returned to and compared to a range of analogous sites elsewhere, how are they ever going to agree about fossils in a one-off meteorite? If recognizing life from earth, on earth, is so hard, recognizing life from elsewhere will surely be harder still.