As a child, I was always intrigued by the idea of the standard metre bar held in Paris. But things have changed dramatically, as I learnt from Professor Ian Mills, president of the Consultative Committee for Units at the International Bureau of Weights and Measures, south west of Paris, which is the centre of an international collaboration. That is where the units are defined, and prototypes such as the kilogram are kept - in a beautiful chateau overlooking the Seine, specially given for that purpose by the French Government in 1875.
It started in 1793 with Napoleon, who wanted to define the metre more accurately than the metal metre bar in Paris. Scientists decided on one- ten-millionth of the distance between the Equator and the North Pole. After three years of surveying, going from Dunkirk to Barcelona and measuring the angle of the Pole Star, they came up with the "platinum standard metre" - but the length was never independently checked.
Instead, in 1875 the French called an international conference, which led to the establishment of the agreed international prototype metre bar. This definition of the metre lasted until 1960, when we entered the atomic era.
The modern definitions of the units are based, whenever possible, on the properties of atoms. The wavelength of the light from a krypton atom can be very accurately measured, and in 1960 the metre was redefined as a specified number of wavelengths of this light.
This was the basis for the metre until 1983, when it was again redefined in terms of the speed of light. Since light travels at 299,792,458 metres per second, the metre is now defined as the distance that light travels in that fraction of a second. Length is defined by time.
Thus the definition of the metre depends on defining the unit of time, the second. The second had for a long time been defined as one-86,400th of the 24-hour day. Harrison's clocks from the 18th century were extremely reliable, varying by only a few seconds per month, as told in the book Longitude by Dava Sobell.
It was not until the Thirties, when the first quartz crystal clocks were developed, that timekeeping was further improved - and then it was discovered that the length of the average day is not constant, but gets longer as the years go by. This is due to the braking effect of the tides on the Earth's speed of rotation.
Quartz crystal clocks, like modern quartz watches, make use of the natural frequency of oscillation of a thin quartz crystal, maintained by an oscillating electric field. However, since each crystal has a different frequency of oscillation, it was not until the Fifties, when the much more accurate caesium clock was developed, that the second was redefined. A caesium atom behaves like a perfect tiny clock, oscillating always at the same frequency, this oscillation being due to the interaction of the magnetic fields due to the spinning electron and the spinning nucleus. All caesium atoms oscillate at exactly the same frequency, which is independent of temperature; this oscillation is the basis of the caesium clock, just as the oscillation of the balance wheel is the basis of an old-fashioned wrist-watch.
In 1968 the second was redefined as the time corresponding to 9,192,631,770 oscillations of the caesium atom. Caesium clocks are now reliable to about one second in 10 million years. An even more reliable atomic clock is under development; it will keep time accurately to about one second in the entire age of the Earth - some four billion years.
There remains the unit of mass - the kilogram. This was originally defined in the 18th century as the weight of a litre of water at 4C. Then, in 1889, it was redefined as the mass of the international prototype kilogram, made of platinum iridium, kept in the chateau at Sevres, where it still sits in a locked room that can be opened only in the presence of the director himself and two other grand officials.
However, balances are now extremely accurate, and it has been found that the kilogram standard is gaining weight, due to surface chemistry causing material to bind to its surface. Thus a new standard needs to be found.
The current approach is to redefine the kilogram in terms of a specified number of carbon atoms and this requires, in turn, the mass of a single atom of carbon to be measured with greater precision than we have yet achieved.
The units of time, length and mass are clearly of international importance, and they must have values that even the most sceptical of scientists and engineers can trust with confidence.
The writer is professor of biology as applied to medicine at University College LondonReuse content