If you want to spur action, you need a disaster - as the case of ozone-destroying CFCs shows.

After his death on Saturday, much will be written about chemist Sherwood Rowland's triumph in getting chlorofluorocarbons (CFCs) and other ozone-destroying chemicals banned. The truth about Rowland's story is a little less inspiring than the legend, however.

Rowland and his colleague Mario Molina published the first paper on the threat to the ozone layer in June 1974. It took thirteen years before the Montreal Protocol, limiting the industrial production of those chemicals, was finally ratified.

Those were extremely painful years for Rowland. His colleagues shunned him for his activism in support of a ban. Almost no university chemistry departments would have him come and speak for nearly a decade -- unthinkable for a chemist of his calibre. Twelve years passed without him being invited to speak to industry groups. James Lovelock, now practically a saint, thought Rowland was going too far: he called for a "bit of British caution" in the face of Rowland and Molina's "missionary" zeal for a ban on CFCs.

If the science establishment doesn't come off too well in that era, Rowland was not without fault either. It may have been in response to pressure from his colleagues, but part of the reason the ban took so long to achieve was that, at a crucial time in the debate, Rowland announced results that cast doubt on the case against CFCs before checking them thoroughly or offering them up for review by others.

In 1976, CFCs' defenders had suggested that the ozone-attacking chemicals might get mopped up by nitrogen in the atmosphere. They would then be rendered safe. Rowland entertained the idea and declared that his estimates of likely ozone depletion by CFCs had been between 20 and 30 percent too pessimistic.

The pronouncement threw the whole issue into confusion at an extremely delicate time. The US National Academy had been about to issue a report into what should be done about CFCs; now they said they needed more time. The Observer declared that the "Aerosol scare 'may be over'". Chaos ensued, and the scientists fell upon each other.

Two months later, Rowland had discovered a mistake in his calculations, but the damage was already done. Because of the confusion, the furore and the persistence of doubts, the National Academy eventually issued its report with significantly weakened conclusions -- so weak in fact, that the following day's New York Times reported the Academy as recommending a curb on aerosols, while the headline of the Washington Post screamed out "Aerosol Ban Opposed by Science Unit".

In the end, it wasn't the carefully-honed arguments of scientists that got CFCs banned. In 1985, scientists announced they had discovered an enormous hole in the ozone layer over the Antarctic. There was a public outcry and the politicians leapt to their feet. The Montreal Protocol was signed two years later. If there's a lesson to be learned from Sherwood Rowland's work, it's that science isn't enough. If you want to spur action, you need a disaster.

In fact, the scientists carried on debating CFCs long after the politicians had moved on. In 1992, five years after Montreal, a group of MIT scientists organised a scientific forum ahead of the environmental summit in Rio de Janeiro. They invited Mario Molina to give a talk. But they scheduled a Brazilian meteorologist to talk first; to Molina's shock, the Brazilian declared to the assembly that the ozone depletion theory was a sham. If there was any depletion, he said, it was due to chlorine from sea spray and volcanoes.

In many ways, the cautious nature of science is its trump card, its ace in the hole. We trust science precisely because it has got things wrong in the past, gives ear to corrective viewpoints and slowly put itself right. But when something is important, we can't wait for all the scientific arguments to be resolved -- because, as the case of Sherwood Rowland shows, that can take longer than any of us can afford.

Michael Brooks's "Free Radicals: the Secret Anarchy of Science" is published by Profile Books (£12.99)