In 1993 Elizabeth Whelan, the president of the American Council on Science and Health, wrote Toxic Terror , in which she dispelled the many alarms being sounded about organochlorine compounds, such as DDT, dioxins and PCBs, which were being blamed for causing cancer. Joe Thornton discusses the links of such chemicals to cancer in Pandora's Poison , yet he ignores her arguments.

He also ignores What Risk ? and Environmental Health , which have been published by the European Science and Environment Forum. These books assess the claims that minute traces of substances like organochlorines damage the environment, but find little proof that they do. Their chapters are written by senior academics, but are likewise not considered worthy of note. Perhaps none of this is surprising because, in his introduction, Thornton eschews scientific risk assessment and places his faith in the somewhat vague "precautionary principle".

This book is a sustained attack on the chloralkali chemical industry by an opponent who writes well and argues skilfully - if not convincingly. It brings together a lot of factual data, and as such could be a useful source book, if only one did not feel slightly uneasy with its careful selection of data. For example, Thornton glosses over the Peruvian cholera outbreak, which began in 1990, when the country suspended the chlorination of water supplies. Greenpeace claimed that banning this use of chlorine would prevent the formation of cancer-causing organochlorines in drinking water. The result was an epidemic that affected a million people and killed 10,000.

Pandora's Poison paints a picture of a world awash with the organochlorines formerly used as pesticides, herbicides, insulating oils and dry-cleaning solvents. With hindsight we were wrong to use some of these chemicals, and we can only regret that much of what was manufactured still lingers in the environment and will be with us for decades to come, until microbes and sunlight destroy them.

ÁñÁ«ÊÓƵ

Was it a road to hell that was paved with good intentions? DDT boosted agricultural production after the second world war, and it is still keeping Ecuador malaria-free. The dry-cleaning solvent perchloroethylene is non-flammable and non-toxic and replaced dangerous chemicals. Because it is volatile, tiny amounts of this chemical can be detected everywhere. Of course, we are talking about only parts per trillion, but that is thanks to the skill of analytical chemists.

And there is the rub. Thornton can only talk about organochlorines being detected in everything on the planet because chemists have the ability to measure them in truly minuscule amounts. Were we able to detect arsenic at the same levels, then Pandora's Poison might have been about this cancer-causing element, which we know is present in breast milk, blood, birds' eggs and baby seals.

ÁñÁ«ÊÓƵ

Pandora's Poison is in three parts: part one states the problem (organochlorines); two identifies the cause (the chemical industry); and three offers a solution (ban all chlorine production). In the final assessment, Thornton assesses the cost of giving up chlorine, but he fails to take into account the accumulated capital of a century of lives saved by chlorination.

Of course, there is a story to tell of our misuse of organochlorines. Carbon tetrachloride was a toxic dry-cleaning solvent, chloroform was a dangerous anaesthetic, PCB oils and dioxins caused severe acne, DDT was grossly overused, and CFCs damage the ozone layer. But these are hazards that have been recognised and dealt with. So why should we deny ourselves other benefits that chlorine brings, not least of which is PVC with its energy-saving and life-saving uses? Are we really that stupid?

John Emsley is science writer in residence in the department of chemistry, University of Cambridge.