299,792,458

For most of history, people assumed light was instant. You open your eyes, and the world is simply there.

Then, in the 1670s, a young Danish astronomer named Ole Rømer was working at the Paris Observatory, timing the little moon Io as it slipped in and out of Jupiter's shadow.

Io should have disappeared like clockwork. But it didn't.

When Earth was moving away from Jupiter, the eclipses came late. When Earth was moving towards it, they came early.

In 1676, Rømer made a bold prediction: an upcoming eclipse of Io would be about ten minutes late.

His explanation was simple and astonishing. The clock in the sky was fine. It was the light from Jupiter that took time to reach us - and when Earth was farther away, the light had farther to travel.

Light, it turned out, had a speed.

For the next eighty years, one man became almost obsessed with this number.

Albert Michelson began measuring the speed of light as a young instructor at the US Naval Academy, publishing his first result in 1879. He refined the method again and again for the rest of his life.

In 1907, he became the first American to win a Nobel Prize in the sciences, for his precision optical instruments and the measurements he made with them.

In the 1920s, he fired light between two mountain tops in California - from Mount Wilson to Mount San Antonio, about 35 kilometres away - and measured its speed at 299,796 kilometres per second, give or take 4.

That was less than 4 kilometres per second away from the value we use today.

By the 20th century, a strange problem was creeping up on scientists.

The speed of light was being measured in metres per second. But what, exactly, is a metre?

For a long time it was a metal bar kept in a vault near Paris. Later it was defined using the wavelength of light given off by krypton atoms.

Then, in 1972, a team at the US National Bureau of Standards in Boulder, Colorado, led by Kenneth Evenson, used a laser to measure the speed of light more precisely than ever: 299,792,456.2 metres per second, uncertain by only about one metre per second.

Their measurement was so precise that the weak link was no longer light. It was the metre itself. The krypton definition was too fuzzy to keep up.

Scientists were measuring light with a ruler that was blurrier than the thing they were measuring.

Why 299,792,458, and not a neat round 300,000,000?

Because the new metre had to be the same length as the old one.

Every ruler, every map, every building plan and every running track in the world was already measured in metres. If the value had been rounded to 300 million, every metre would have shrunk by about 0.7 millimetres. That sounds tiny, but over a 100-metre sprint it adds up to nearly 7 centimetres - enough to throw off everything from engineering to athletics records.

So scientists chose the number that best matched the old, krypton-based metre and the best laser measurements of the day. The 1972 measurement had found 299,792,456.2, give or take about a metre per second; the value fixed for the future was 299,792,458.

The number looks untidy. That untidiness is the fingerprint of history - a promise that the metre you use today is the same metre your grandparents used.

Most numbers in science describe something we measured.

299,792,458 is different. It began as a measurement - the prize of three centuries of patient chasing, from a late moon, to a spinning wheel, to mountain tops, to lasers.

Then it became a foundation.

Today, every length you measure - every GPS position, every map, every centimetre on a tape measure - ultimately rests on light and a very precise clock.

It is a quietly beautiful idea: we did not invent our ruler. We borrowed it from the universe, and built the metre from the speed of light.