Who Decided How Long a Metre and How Heavy a Kilogram Is, and Why Did the Definitions Change?

Before the Metric System, the Same Unit Name Could Mean Different Amounts in Different Places

Imagine buying cloth in one town and selling it in the next, where the “same” unit of length is a different length. Before the French Revolution that was ordinary. One historical estimate says that in 1789 France used around eight hundred units of measure, with up to a quarter of a million different local definitions between them [5].

That is a nightmare for trade, for tax collectors and for anyone building something to a standard size. The revolutionary government wanted a system that belonged to no lord, town or king. So the real question was not just “how long is a metre?” but “what could a metre be tied to that nobody owns?”

In 1791 the Metre Was Tied to the Size of the Earth Itself

The answer chosen in 1791 was the planet. The French Academy of Sciences recommended a metre equal to one ten-millionth of the distance from the North Pole to the equator, measured along a meridian through Paris, and the National Assembly approved the proposal a week later [3]. A length taken from the Earth, the reasoning went, would belong to everyone.

The kilogram was tied to water: the mass of a litre of water, which is the water in a cube 10 centimetres on each side [5]. A volume of water can in principle be reproduced anywhere, so this too avoided any single owner.

The idea was beautiful. But to use it, someone had to measure part of the Earth.

Measuring the Earth's Curve From Dunkirk to Barcelona Was Slow, and One Measurement Went Wrong

Two astronomers, Jean-Baptiste Delambre and Pierre Méchain, set out in 1792 to survey the arc of the meridian between Dunkirk, on the northern coast of France, and Barcelona, in Spain. Delambre took the northern part and Méchain the southern. They worked by triangulation: measuring angles between distant landmarks and building a chain of triangles across the country [3].

Doing this in the middle of a revolution was hard. Delambre was arrested twice in September 1792, the second time because local people thought his strange instruments meant he was a spy [3]. Méchain, in the south, found that war between France and Spain stopped him from continuing at his first observing site [4].

In 1794 Méchain noticed that two sets of latitude measurements he had made around Barcelona did not agree, and he could not go back to check. He told no one. Instead he adjusted his recorded data so the scatter looked smaller, without changing the average. Delambre later discovered this. Worried that admitting it would damage trust in the whole metric project, he kept the matter private and treated Méchain’s data as the careful work of a careful observer [4].

This is a useful reminder that even a “perfect” definition depends on people with instruments, deadlines and worries. It also shows why later generations wanted definitions that did not rest on one hard-to-repeat survey.

In 1799 and 1889 the Definitions Became Platinum Objects That Anyone Could Compare Against

Surveying the whole arc again every time someone wanted a metre stick was impossible. In June 1799 France made platinum standards from the measured results: a bar for the metre and a weight called the kilogram of the archives [5][2]. In practice, the metal now stood for the unit, not the Earth. Wikipedia notes that the final metre of the archives came out a little shorter than the early provisional value, a sign of the small errors in the survey [5].

As trade and science became international, each country kept its own standards, and they did not always match. On 20 May 1875, 17 countries, including the United States, signed the Metre Convention to create shared international standards [2]. In 1889 the first General Conference on Weights and Measures approved an international prototype metre and an international prototype kilogram, both made of platinum-iridium, kept at the International Bureau of Weights and Measures (BIPM) near Paris, with copies for member countries [1][6].

From then on, the 1889 definition of the kilogram was simply the mass of that one cylinder [1]. Very simple, and very risky.

Comparing the Copies With the Original Showed Their Masses Drifting Apart

If the cylinder is the kilogram, then by definition it is always exactly one kilogram. But you can still compare the copies with it. The BIPM describes three verifications of national prototypes. The second, in 1946, showed the prototype masses diverging from the international one on average. The third, in 1989–1991, found a median difference of about 25 micrograms (25 millionths of a gram) for the original set of prototypes approved in 1889 [1].

That is tiny, but it matters in precise science. And the comparison creates a puzzle: did the copies change, or did the master change? With only one reference object, there is no independent way to say. The BIPM page I read does not name a proven cause.

Since 2019 the Kilogram Is Defined by a Constant of Nature, Not by an Object

The fix was to stop using an object. In 2018 the 26th General Conference on Weights and Measures adopted a new definition of the kilogram based on the value of a fundamental constant, the Planck constant [1]. The new definitions came into force on 20 May 2019, the anniversary of the Metre Convention [7].

In plain terms: the number for the Planck constant, written in SI units, is now fixed by agreement. Because that constant links mass to other quantities that can be measured very precisely, laboratories can realize a kilogram without the cylinder in Paris. The cylinder is still kept, but it is no longer the definition.

The Metre Left the Earth and Metal Too: It Is Now Defined by the Speed of Light

The metre went through a similar journey earlier. After the 1889 platinum-iridium prototype, the 11th General Conference in 1960 defined the metre by the wavelength of a particular radiation from krypton-86 atoms. In 1983 the 17th General Conference redefined it as the distance light travels in a vacuum in a specified interval of time, which rests on the fixed value of the speed of light, 299,792,458 metres per second [6]. In 2018 the wording was revised to make clearer that the metre depends on that fixed value [6].

So the same story repeats: a definition tied to the Earth, then to an object, then to something that is the same everywhere and cannot be damaged. Each step answered the question that began in 1791: what can a unit be tied to so that nobody owns it?

Try It Yourself: See the Units Agree on Your Kitchen Scale and Ruler

You only need things you already have at home: a ruler, a tape measure, and a kitchen scale or bathroom scale.

1. Compare two rulers. Lay a ruler beside a tape measure and look at the centimetre marks. If they were made by different companies in different places, they still agree. That is the agreement this note is about.
2. Look for the letters. Find where “g”, “kg” or “cm” are printed on your scale and rulers.
3. Measure the old way. Measure a table with your hand-spans, then ask someone else to do it. You will probably get different numbers, which shows why people needed something better.

Sources

Update history