A recent book advocates the theory that changes in climate owe more to solar activity than to greenhouse gases. While the `greenhouse theory' is more strongly supported by science, the rival solar theory has many statistics in its favour.

I have just been reading The Manic Sun (Pilkington Press, pounds 24.95), Nigel Calder's recent book on the role of solar activity in changing the Earth's climate. Calder, it should be remembered, was the author, in 1974, of The Weather Machine, one of the most lucidly written popular science books of its day, and one of the most readable explanations ever of the physical mechanisms behind climate and weather.

The only thing wrong with The Weather Machine was the basic warning running through the book that the Earth was facing an imminent ice age, caused, at least in part, by man's irresponsibility in burning fossil fuels and squirting aerosols with abandon. That, however, was the fashionable theory of the age, and Calder was only one of many eminent scientists caught out by a downturn in global temperatures that suggested a change in climate.

In his latest book, Calder has come out firmly against the establishment view. He dismisses theories of global warming caused by greenhouse gases as nothing more than fashionable political correctness. In fact, he dismisses these theories on all possible fronts: greenhouse gases are not a major factor in creating global warming; the Earth is not getting hotter anyway; and even if it were, it would not be such a big deal. "We are the first inhabitants of a temperate zone in the history of the world who think that warming is bad," he told me yesterday. "It's cold that kills."

The thesis advocated in The Manic Sun is that solar activity is what counts in determining changes in the Earth's climate. In particular, there is a strong correlation between sunspot activity and the Earth's temperature. This is by no means a new idea. Even in the 19th century several scientists proposed a connection between sunspot activity and the Earth's weather, and more recently correlations have been claimed between sunspot numbers and temperatures on Earth. The years of the worst droughts in farming land in the west of the US, for example, fit well into the 111/2-year sunspot cycle: 1842, 1866, 1890, 1912, 1934, 1953, 1976 - one drought every two sunspot cycles, with another one due this year or next.

Calder, however, supports a more sophisticated theory of three Danish scientists, Eigil Friis-Christensen, Knud Lassen and Henrik Svensmark, who have plotted a convincing graph of the Earth's temperature against the varying length of solar cycles. While the average sunspot cycle is just over 11 years, it may vary between seven and 17. On a short time- scale, the Earth may follow the sunspots, but on a longer one, it is the length of the cycles that matter. Longer cycle-length brings low temperatures.

So why hasn't the scientific establishment been convinced by all this? Calder would suggest that the scientific agenda has been hijacked by politicians. The problem of causality, however, also plays a part. Sunspot activity can affect the amount of solar energy reaching the Earth by only 0.1 per cent - scarcely enough to make a measurable difference. Calder maintains that cosmic rays, which increase and decrease with sunspot activity, affect the Earth's cloud cover, and that in turn has considerable effects on climate. Piers Corbyn, however, another prominent advocate of the sunspot theory, claims that magnifying effects in the atmosphere can hugely boost the 0.1 per cent. Without consensus on causality, the sunspot theory, however statistically convincing, may continue to be shunned by the scientific establishment.