The Abyss of Time

In 1788 James Hutton took two companions by boat to Siccar Point on the Berwickshire coast to show them a rock formation.

What they saw was this: nearly vertical layers of greywacke, cut off flat, with horizontal beds of red sandstone lying on top. Hutton read the sequence. Sediment laid down in horizontal layers. Buried, compressed to rock. Tilted to vertical by forces from below. Raised above sea level. Eroded flat, which takes a very long time. Submerged again. Buried under new horizontal sediment. Raised again, for him to stand on.

Each of these processes was slow. Together they required a span of time that nobody in 1788 had a framework for. John Playfair, who was there, wrote that the mind seemed to grow giddy by looking so far into the abyss of time.

Hutton's principle (that the processes visible today, operating over enough time, explain the geological record) is called uniformitarianism. It has been moderated since, because catastrophic events do happen: impacts, megafloods, enormous volcanic episodes. What survives is that ordinary processes plus enormous durations explain most of what you see, and that invoking unique catastrophes should be a last resort rather than a first one.

Before radioactivity, everything about the history of the Earth was relative rather than absolute, and the tool was superposition: in undisturbed layers, deeper is older. Steno stated it in the seventeenth century, and the crucial word is undisturbed, because folding can turn a sequence upside down entirely. Recognising an overturned bed from ripple marks or graded bedding is one of the basic skills of field geology.

Relative dating is more powerful than it sounds. The entire sequence of the history of life (invertebrates, then fish, then amphibians, then reptiles, then mammals) was established from superposition and fossil correlation before anybody could attach a number to any of it.

The numbers arrived with radioactive decay.

An unstable isotope decays at a rate that is fixed and unaffected by temperature, pressure, chemistry or anything else anybody has been able to do to it. That is what makes it a clock. Measure the ratio of the remaining parent isotope to its decay product, know the half-life, and the elapsed time follows.

Different pairs suit different ranges. Carbon-14 has a half-life of about 5,700 years and is useful to perhaps 50,000. Uranium-lead reaches billions. The clock starts when a mineral crystallises and locks the parent isotope into its structure, which means the method dates crystallisation rather than anything that happened to the rock afterwards: a distinction that has caused plenty of confusion.

Which brings up how the Earth's age was actually established, and it is a better story than the number suggests.

You cannot date the Earth from the Earth, because the surface is continuously recycled by tectonics. The oldest surviving rocks are around four billion years old and the original crust is long gone.

Clair Patterson's solution in 1953 was to date meteorites, which formed at the same time as the solar system and have not been reprocessed. He got 4.55 billion years, which has barely moved since.

Getting there required eliminating lead contamination from his samples, and the contamination turned out to be everywhere: in the air, the dust, the equipment, the water. Patterson had to build the first ultra-clean laboratory to get a usable measurement, and in the process he worked out how much industrial lead was in the modern environment, and how recently it had arrived. He spent much of the rest of his career campaigning to get it out of petrol, against sustained opposition from the industry. The dating was the side effect.

One last case, because it shows the evidence working the way it should.

In 1980 Walter and Luis Alvarez found a thin layer of clay at the boundary between Cretaceous and Paleogene rocks, and it was rich in iridium: an element rare in the Earth's crust and common in asteroids. The same layer turned up worldwide, at exactly the same stratigraphic position.

Their proposal, that a large impact had caused the extinction, was resisted for a decade. It required a crater, and nobody had one. Then the Chicxulub structure off the Yucatán coast was identified, 180 kilometres across, dated to the right moment.

The honest current picture is that the Deccan Traps volcanism was also in progress and the ecosystem was already under stress, and that the impact was decisive. Which is roughly where the uniformitarian and catastrophist traditions ended up: slow processes explain most of the record, and occasionally something falls out of the sky.