Sixteen Marble Metres

At 36 rue de Vaugirard in Paris, set into a wall, there is a slab of marble with a ruler cut into it and one word, METRE. The plaque above explains it. The National Convention, "in order to generalise the use of the metric system", had sixteen marble standard metres placed "in the busiest places in Paris", installed between February 1796 and December 1797. This is one of the last two that survive, and the only one still where it was put.

It was there so that ordinary Parisians could check a length of cloth or a plank against the new unit with their own eyes. The government did not expect anyone to take the metre on trust, which was the whole point of it.

A unit nobody owned

Before the Revolution, the length of a foot or the size of a bushel in France could change from one town to the next. Whoever set the measure set the terms of trade, and the buyer usually had to take it on trust.

The revolutionaries wanted a unit that belonged to nobody, because it belonged to nature. In 1791 the Paris Academy of Sciences put a report to the National Assembly proposing that the new unit of length be one ten-millionth of the distance from the North Pole to the equator, measured along the meridian through Paris. Anyone, anywhere, could in principle measure the earth and arrive at the same metre. No king's arm or foot came into it.

The survey

Someone had to actually measure the earth. In June 1792, with the Revolution turning violent around them, two astronomers set out in opposite directions. Jean-Baptiste Delambre went north to Dunkirk. Pierre Mechain went south to Barcelona. Between them they were to triangulate the arc of the meridian, chaining together sightings from church towers and hilltops across a country at war, and from it work out the full quarter of the globe.

It took them seven years. In 1799 their results were turned into a platinum bar, the Metre of the Archives, defined as the distance between its two polished ends at a set temperature.

What almost nobody knew was that Mechain had found a problem. While taking latitude measurements in Barcelona in the winter of 1793, he discovered a discrepancy in his data. Afraid he would be blamed, he kept it secret. The historian Ken Alder, who found the evidence in the archives two centuries later, described it as "a secret error" that was "known to the two astronomers and hidden by them from public view."

The metre was slightly wrong from the start. It does not exactly match a ten-millionth of the quadrant, and it never has.

Admitting it didn't matter

The striking thing is what happened next. Nobody changed the bar.

By 1872 the International Commission of the Metre had made the Metre of the Archives the official definition, and, in the words of the American standards laboratory NIST's own history, "it was admitted that its relationship to a quadrant of the earth was tenuous and of little consequence anyway." The point had never really been the earth. The point was a single agreed length that everyone could copy exactly.

On 20 May 1875 seventeen countries, the United States among them, signed the Treaty of the Metre and set up the International Bureau of Weights and Measures outside Paris. In 1889 a new platinum-iridium bar with an X-shaped cross-section became the international prototype, and each member country received copies. America's, Metre No. 27, served as its national standard until 1960 and now sits in a museum in Maryland.

Swapping the bar for light

A metal bar can be damaged, stolen or warped, and it has to be in one place. So the scientists went back to the original idea of a standard anyone could reproduce, only now using physics instead of geography.

On 14 October 1960 the metre was redefined as 1,650,763.73 wavelengths of a particular orange light given off by krypton-86. On 20 October 1983 it was redefined again, as the distance light travels in a vacuum in 1/299,792,458 of a second. That is still the definition. The speed of light is now fixed, by law, at exactly 299,792,458 metres per second, and the metre follows from it.

In every one of these changes, NIST notes, the aim was to make the definition more precise while changing the actual length as little as possible. The metre on the wall in rue de Vaugirard is, for any practical purpose, still the metre.

India's switch

India made the change in two steps, after independence. The coins went first. The Indian Coinage Act was amended in September 1955, and on 1 April 1957 the rupee, which had been divided into 16 annas and 64 pice, was divided into a hundred paise. To avoid confusion with the old pice, the new coins were called naya paisa, new paisa, until 1 June 1964, when the "naya" was dropped. The rupee itself kept its value and its name. Only the way it was cut up changed.

Weights followed from 1958, when the government began replacing the old system of tolas, seers and maunds, which had never been uniform across the country, with the kilogram. It worked. The generation that grew up after it has no idea what a seer weighs.

The countries that held out

Today almost every country in the world uses the metric system for nearly everything. The obvious exception is the United States, which legalised metric units by act of Congress in 1866 but never made them obligatory, and still measures roads in miles and people in pounds.

The cost of that split is usually small and occasionally spectacular. On 23 September 1999 NASA lost the Mars Climate Orbiter as it arrived at Mars. The investigation board found that a piece of ground-based navigation software reported the force of the spacecraft's thruster firings in pound-seconds, an imperial unit, when the navigation team's software expected newton-seconds, the metric one. The spacecraft's real path drifted from the calculated one until it came in far too low and was lost. The root cause, in the board's words, was a failed translation of English units into metric units.

The marble metre in Paris was put on the street so that people could see the length for themselves and stop arguing about it. More than two centuries later, the arguing is mostly over.

Sources

• William B. Penzes, "Time Line for the Definition of the Meter", National Institute of Standards and Technology (the 1791 Academy report to the National Assembly; the meridian from Dunkirk to Barcelona and the platinum Metre of the Archives of 1799; the 1866 US act legalising metric use without making it obligatory; the 1872 admission that the link to the earth's quadrant was "tenuous and of little consequence anyway"; the Treaty of the Metre signed by seventeen countries on 20 May 1875; the 1889 platinum-iridium prototype and US Metre No. 27; the krypton definition of 14 October 1960; the speed-of-light definition of 20 October 1983; the aim of changing the length as little as possible).
• International Bureau of Weights and Measures, "Metre" (the current definition of the metre by the fixed value of the speed of light).
• NASA Goddard Engineering Colloquium, "The Measure of All Things", lecture by Ken Alder, 2004 (the secret error known to the two astronomers and hidden from public view).
• Smithsonian Libraries, "The measure of all things: the seven-year odyssey and hidden error that transformed the world" (Ken Alder's history of the survey: Delambre going north to Dunkirk and Mechain south to Barcelona in June 1792, and Mechain's concealed discrepancy from Barcelona in 1793).
• Reserve Bank of India, "Republic India Coinage" (the amendment of the Indian Coinage Act in September 1955; decimal coinage from 1 April 1957; the rupee's earlier division into annas and pice; the rupee keeping its value and name; the "naya" dropped on 1 June 1964).
• The Quint, "The History of India's Weights, From a Seed To Kilogram" (the move to a kilogram-based system in 1958 to end the confusion of older weights).
• NASA, "Mars Climate Orbiter Mishap Investigation Board Phase I Report", November 1999 (the loss of the spacecraft and the root cause in thruster data delivered in pound-seconds rather than newton-seconds).
• Wikimedia Commons, "Le metre en marbre de la rue Vaugirard a Paris" (the plaque text: sixteen marble metres placed by the National Convention between February 1796 and December 1797, of which this is the only one still on its original site).