In 1792, a French astronomer named Pierre-François Méchain set out to measure the Earth. His job was to survey the meridian arc running from Dunkirk to Barcelona, so scientists could figure out the exact distance from the North Pole to the equator. One ten-millionth of that distance would become the new metre, a unit of length based on something nobody controlled: the planet itself.
Méchain made a small measurement error at Barcelona. Worse, when he realized it, he tried to hide it rather than admit the mistake. That flawed data went into the official calculation anyway. As a result, the very first metre, a platinum bar locked away in Paris in 1799, was about 0.2 millimeters shorter than it should have been.
Here's the twist: nobody ever "fixed" that error.
The moment the French Academy accepted that platinum bar in 1799, something interesting happened. The metre stopped being "a fraction of the Earth's size" and became "the length of this specific piece of metal." Even though the bar didn't perfectly match the Earth's true measurements, it didn't matter anymore, the bar itself was now the definition. Being slightly wrong about the Earth was irrelevant, because the metre was no longer about the Earth.
This happened again in 1889, when a new prototype bar became the standard, and again in 1960, when scientists switched to a specific colour of light given off by krypton atoms. Each time, the goal wasn't to recalculate the Earth's true size. It was to match the length of the original bar more and more precisely, using better tools. Méchain's mistake just quietly rode along, baked into the system, technically "wrong" forever but functionally irrelevant.
1983: The Metre Stops Being About Space at All
Then in 1983, something bigger happened. Scientists made a decision that sounds almost backwards: instead of measuring how fast light travels using a ruler, they decided to define the speed of light as an exact number: 299,792,458 metres per second, no more, no less, and use that to define the metre.
So now, a metre isn't a bar, a bone, or a fraction of a planet. A metre is defined as: the distance light travels in 1/299,792,458th of a second.
This flips the whole relationship upside down. Distance used to be the fixed thing, and speed was something you measured. Now speed is fixed, and distance is something you calculate from time. Time itself is measured using something even more reliable than a metal bar: the vibrations of a caesium atom, which ticks at a fantastically stable, fixed frequency no matter who's watching or where they are.
Wait — Isn't That Circular?
Here's the sharp question worth sitting with: to know the speed of light before 1983, scientists had to measure both a distance and a time. So doesn't "distance equals speed of light times time" quietly depend on already knowing distance in the first place? Isn't that like trying to weigh yourself using a scale that's balanced on your own head?
This is where it gets genuinely interesting, and where it connects to one of the strangest unsolved puzzles in physics: the one-way speed of light problem.
Here's that problem in a nutshell. It's easy to measure how fast light travels on a round trip: bounce a beam off a mirror and back, time the whole trip, divide by two. Scientists have done this with incredible precision, and light's round-trip speed really does seem to be the same in every direction. But measuring the speed in just one direction: say, from New York to Tokyo, one way only, requires two synchronized clocks, one at each end. And how do you synchronize two distant clocks? By sending a signal between them and correcting for the travel time. But you don't know the travel time... because that depends on the speed of light, which is the very thing you were trying to measure. It's a genuine logical loop, and physicists still argue about whether it can ever be truly broken, or whether it's just a convention we all agree to use.
So, is the 1983 metre stuck in the same loop?
The trick is to separate two things that sound similar but aren't:
1. What the metre means. After 1983, the metre isn't defined in terms of itself. It's defined using the second (measured independently, by counting atomic vibrations, no distance required) and a fixed number for the speed of light. That's not circular; it's just choosing new starting ingredients. It's the same move scientists made in 2019 when they redefined the kilogram using a fixed value of Planck's constant instead of an actual metal weight in a vault.
2. How the metre gets used in practice. This is where a little bit of looping really does creep back in. To compare atomic clocks in different countries, which is how the "official" second is actually maintained around the world, scientists send signals between them (often via satellite) and have to account for how long those signals take to travel. That does lean on the physics of light.
But it only ever leans on the round-trip speed of light, the one part of this whole puzzle that isn't up for debate. The round-trip speed of light has been tested over and over, in every direction, with staggering precision, and it always comes out the same. It's the one-way speed that's the genuinely murky, maybe-unanswerable philosophical puzzle, and the modern metre never actually needs to know that one-way number to work.
So What Was Méchain's Mistake, Really?
Méchain's error belonged to an older way of thinking, where the Earth was the ultimate ruler and everything else had to measure up to it correctly. The 1983 redefinition didn't correct that error, it made the whole question irrelevant by changing what "correct" even meant. Length is no longer a fact about a planet or a bar of metal. It's a fact about time and a number everyone has agreed to hold fixed.
There's still a small, honest kind of circularity underneath all of this: every measurement system in science has to start somewhere, with something taken as given. That's not a flaw unique to the metre. It's true of any system of measurement you could ever build, including whatever eventually replaces this one. The real achievement of 1983 wasn't eliminating that starting point. It was being completely upfront about where it is.
https://www.americanscientist.org/article/the-error-of-all-things