James Hutton made his fortune out of soot. He and a friend worked out how to cook the black residue scraped from Edinburgh's chimneys into a pale crystalline salt that dyers and metalworkers would pay for, and which had until then come by ship from Egypt. It was a good business. It left him free to spend the rest of his life doing what he actually wanted to do, which was to climb into ditches and look at rocks.

Everyone agreed the world was about six thousand years old. This was not superstition. It was arithmetic, done carefully, from the only continuous historical document anyone had. Hutton, draining a wet Berwickshire farm in the 1750s, noticed how much soil a single winter carried off, multiplied by the thickness of the rock under his boots, and arrived at a number so large that he declined to write it down.

He never did write it down. He said instead that it was in vain to seek any computation of the time, and then closed his great paper with the sentence that made him immortal and got him accused of atheism: that in the history of the Earth we find no vestige of a beginning, no prospect of an end.

This book takes that sentence seriously, which almost nobody does. It is not a claim that the world is eternal. It is a claim about the limits of a record that destroys its own evidence—because rock, as Hutton was the first to see clearly, is made of the wreckage of older rock, and every cycle erases what came before. The stone that survives to be read is not a fair sample of what happened. It is what happened to last.