Published on
The metre was born in 1791 and it came out of the Earth: it was defined as one ten-millionth of the distance from the North Pole to the equator along the meridian through Paris. Which means that, by definition, that quadrant of meridian is 10,000 kilometres long. Nobody had measured it yet.
The circle on the map has a radius of 1,075 kilometres and is centred on Dunkirk: its edge runs through Barcelona. That is exactly the stretch of meridian that Jean-Baptiste Delambre and Pierre Méchain set out to triangulate in 1792 in order to find out how long a metre was — one tenth of the quadrant, with the rest extrapolated. It took them seven years, they got it slightly wrong, and that error is still inside the unit today. This article is part of the distance units and where they come from; its opposite number is the mile, which was built the other way round.
Before the metre, France ran on roughly 800 names of units covering, historians reckon, more than 250,000 distinct values: the same word meant a different amount in every town. A pinte in Paris was 0.93 litres; in Saint-Denis, ten kilometres away, it was 1.46. One guild's foot was not the next guild's foot, and the grain measure changed from village to village, almost always in favour of the lord who defined it. Measuring was a way of charging.
All those units shared one flaw: they came from the human body or from human work — the foot, the cubit, the pace, the furrow — and so there was nothing to check them against. The revolutionary idea, in the literal sense, was to look for a standard that sat outside the argument: something anyone, in any country and in any century, could go and measure again. The full story of the older units is in the foot, the league and the other distance units.
In 1790 Talleyrand put the simplest solution to the Assembly: define the unit as the length of a seconds pendulum at 45° of latitude — the one that takes a second per swing. It is an experiment that fits in a room and that anyone can repeat with a string, a weight and a clock.
The Academy of Sciences turned it down for an awkward reason: gravity is not the same everywhere. It varies with latitude and with altitude, so the pendulum would measure slightly different lengths in Paris, in Quito or on a mountain top, and the unit would depend on where you stood. Defining a length through a time also meant fixing the second first, and the second came from the calendar of the old regime.
The entertaining part is how close it was: a seconds pendulum is 993.6 millimetres long under standard gravity. Had Talleyrand's proposal won, the metre would be 6.4 millimetres shorter than it is, and everything built on it — the kilometre, the litre, the kilogram — would be shifted by the same proportion.
The Academy chose the meridian. The metre would be one ten-millionth of the quadrant of the meridian through Paris, from the North Pole to the equator. Why ten million and not some other figure is the least heroic and most practical part of the story: that fraction was known to give a length close to the measuring rods already in daily use, so the new unit would not feel alien. The name came from the Greek metron, “a measure”.
The plan had an obvious hole: nobody could walk from Paris to the North Pole. What could be done was to measure a section of meridian very precisely, work out how many degrees it spanned and extrapolate the rest. The arc chosen ran from Dunkirk to Barcelona: about 1,075 kilometres, one tenth of the quadrant, almost all of it French, crossed by the Paris meridian, with both ends at sea level and straddling the 45th parallel so that the flattening of the Earth would cancel out.
They left Paris in June 1792. Delambre took the northern half, Dunkirk to Rodez, 742.7 kilometres; Méchain the southern one, Rodez to Barcelona, 333 kilometres. The method was triangulation: chaining triangles between steeples, towers and summits, measuring one single baseline by hand with platinum rules and deriving every other distance by trigonometry, without touching the ground again.
The timing could not have been worse. France was in the middle of the Revolution and, soon after, at war with Spain. Two men climbing church towers with brass instruments and signal flags looked, to any villager, like spies. Delambre was stopped again and again and had to produce safe-conducts signed by officials who had since been guillotined; in late 1793 the Committee of Public Safety threw him off the weights and measures commission itself, along with Borda, Laplace, Coulomb and Lavoisier, for being insufficiently republican — Lavoisier, the commission's treasurer, went to the guillotine on 8 May 1794. Méchain was arrested on suspicion that his instruments were weapons and, once war broke out, was interned in Barcelona: he left for Italy and could not return to France until 1795.
They were back in Paris with the data at the end of November 1798. Seven years to measure a thousand kilometres.
Stuck in Barcelona, Méchain used the time to measure the latitude of the castle of Montjuïc with extreme care. Then he measured it again. And the two results did not agree: the gap was a few seconds of arc, invisible to anyone but him, yet enough to make the figure that was about to define the world's unit of length untrustworthy.
He said nothing. He tidied the notebooks, handed over the good number and spent the rest of his life trying to prove to himself that he had not been wrong. He ended up going back to Spain to extend the arc southwards to the Balearics, hoping the new figures would bury the old ones. He died there of yellow fever on 20 September 1804, in Castellón de la Plana. Delambre inherited the notebooks, understood what had happened, published the data and locked the rest away at the Paris Observatory, where it surfaced a century and a half later.
It is worth being clear about this: Méchain's error was not faked results, it was a physical problem nobody yet knew how to handle. The vertical marked by a plumb line does not point exactly at the centre of the Earth when there is a mountain next to it — what is now called deflection of the vertical — and Montjuïc, with the sea on one side and the hills on the other, is a superb place for that to happen.
Modern geodesy knows how long the meridian quadrant really is: about 10,001,966 metres. That is 1,966 metres more than the ten million the definition called for. Spread across ten million, it means every metre came out 0.2 millimetres short (0.1966, to be precise).
Put into things you can see:
The striking part is that the error can no longer be corrected, and not out of laziness. When the platinum standard bar was made in 1799, the unit stopped being “one ten-millionth of the meridian” and became, in practice, “the length of this bar”. Every later redefinition — 1889, 1960, 1983 — was calculated to preserve exactly the length that already existed, because changing it would have invalidated every drawing, every machine and every standard in the world overnight. The metre of the speed of light, the best defined unit in history, carries inside it the error of two eighteenth-century astronomers having a bad day on Montjuïc.
The metre has been redefined four times, and each redefinition does the same thing: take precision away from the object and hand it to nature.
| Year | What a metre was | What it fixed |
|---|---|---|
| 1791 | One ten-millionth of the quadrant of the Paris meridian | A unit that depended on no king |
| 1799 | The length of the mètre des Archives, a platinum bar | Something you could actually compare against |
| 1889 | The international prototype metre, platinum-iridium (90%/10%) with an X-shaped cross section | A stable alloy and a section that does not sag |
| 1960 | 1,650,763.73 wavelengths of an orange line of krypton-86 | Any laboratory could reproduce it without travelling to Paris |
| 1983 | The path light travels in vacuum in 1/299,792,458 of a second | Fixing the speed of light and measuring with lasers |
The 1983 definition, which is the one in force, has a particular elegance: the speed of light is no longer measured. It is assigned the exact value of 299,792,458 metres per second, and the metre is derived from it. Put the other way round, light takes 3.34 nanoseconds to travel one metre, and that is today's definition of the unit. The 2019 revision of the SI left it untouched and came into force, neatly, on 20 May: the 144th anniversary of the Metre Convention.
The metre was not settled by France alone. In 1798 an international congress met in Paris to check the calculations and fix the definitive standards — arguably the first international scientific body in history. Spain, whose coastline had supplied the southern end of the arc, sent two delegates: the Valencian naval officer and mathematician Gabriel Císcar and the Asturian mathematician Agustín de Pedrayes.
The 1799 verdict set the metre at 443.296 lines of the toise of Paris — the 1793 provisional value had been 443.44, so the Delambre and Méchain arc shortened it by about three tenths of a millimetre — and the platinum bar known as the mètre des Archives was deposited in the national archives. That is the metre we inherited.
Britain and the United States stayed out of the metric system, but not out of the club: the Metre Convention, signed in Paris on 20 May 1875 by seventeen states, created the International Bureau of Weights and Measures, and the United States was a founding signatory. Since the International Yard and Pound Agreement of 1959, the yard has been exactly 0.9144 metres — which is to say that the mile, the foot and the inch are all defined in metres.
| Metres | Imperial | What it is |
|---|---|---|
| 1 metre | 3.281 feet · 39.37 inches | The SI base unit |
| 0.3048 m | 1 foot exactly | Defined in metres since 1959 |
| 0.9144 m | 1 yard exactly | The foot and the inch derive from it |
| 1.8288 m | 6 feet | The height of the movies |
| 1,000 m | 0.621 miles | A kilometre |
| 1,852 m | 1 nautical mile | One minute of arc of a meridian |
And a note that explains half of this series: the foot, the inch and the yard are defined in metres. The imperial system has been measuring with the metre, quietly, for nearly seventy years.
Just want the number? The distance converter turns metres into feet, yards, miles or nautical miles. Want to see it? Type a distance into the distances tool and it is drawn to scale over your city, or measure a distance point by point on the map. And if you want the whole story — the mile, the foot, the league and why we still cannot agree — it is in the distance units.