National Physical Laboratory

What is the history of weighing? (FAQ - Mass & Density)

Weighing

Weighing - by which we mean using a balance to measure the weight of an object or to compare the weight of two objects - has been undertaken for thousands of years. Images from the earliest civilisations in the Middle East show items being weighed using a beam balance and the process continued almost unchanged until the twentieth century. The Science Museum in London has an example of an early Egyptian quartz beam about 80 mm long with leather support thongs. This, it is believed, was used to weigh gold found in the sands of the river Euphrates and used as a barter medium. The only way of quantifying the gold was to compare its weight, using a beam balance, with a 'standard' weight - the Beqa, equivalent to 200 grains of barley corn. This was the 'kilogram' of the day (not literally!), 5000 years ago, and is thought to be both the origin of the use of gold as a currency and the beginning of the gold standard.

As recently as medieval times, when wealth was measured in land (and most of the land belonged to the monarch) even money was measured by weighing . The only coin was the silver penny, chiselled into two pieces for a halfpenny and four for a quarter-penny or farthing (fourthing). The pennyweight was 1/240 of the Pound of Troye (named after the city in France which was the centre of mercantile trade in the 12th century). This pound was 5 760 grains - barleycorn grains selected in a precise manner (with a consistent mass and length, so that they were used as a basis for small units of length measurement as well). The Troy weight system (20 pennyweights to the Troy Ounce and 12 ounces to the pound) was used for all precious materials and also for bread, the price of which was controlled by law from the 12th until the 19th century. In 1280 a farthing loaf weighed about 3½ pounds.

A later system of weights introduced the haber de peyse or Avoirdupois pound, which was eventually fixed at 7 000 grains and divided into 16 ounces each of 16 drams. The only connection between the values of the Troy and Avoirdupois systems was the grain of barleycorn!

After land, precious metals and stones, the next most important measure of wealth was wool, which by the 15th century accounted for over half of England's export trade. Larger denominations of weight were needed, and the smallest value in common use was the stone - which finally settled at 14 Avoirdupois pounds. Two stones made a quarter, and four quarters made a hundredweight (112 pounds - in commercial measurement a hundred was not necessarily ten times ten). Three hundredweights made a sack.

Fine weighings were needed for medicines and the Apothecary's measure was based on the Troy ounce - 20 grains to the Scruple, three Scruples to the Drachm and eight Drachms to the ounce. This measuring system remained in use until very recently.

When dealing with all these different systems of weight measurement the benefit of a single-universally accepted system soon became apparent.

The birth of the metric system and the kilogram

In 1790 the French National Assembly obtained Louis XVI's assent to commission the country's leading scientists to recommend a consistent system for weights and measures. (It is said that Louis wrote to the King of England suggesting that French and English scientists should collaborate on this, but the Assembly did not wait for a reply before ordering the French scientists to work!)

The report which the French scientists Lalande, Laplace, Borda, Monge and Concordet presented to the Academy of Sciences on 19 March 1791 recommended a system based on a unit of length, the metre, equal to one ten-millionth part of the distance from the Earth's pole to the equator. The unit of mass would be equal to the mass of a defined volume of water at its freezing point. Legislation, authorizing construction of this new system of units, was passed on 26 March 1791 and the Metric System of measurement was born.

The kilogram

After the 1791 report, measurements were made to decide an appropriate volume of water for the standard of mass. In 1799 it was agreed that the unit should be the mass of one cubic decimetre of water at a temperature of 4 °C, which would be called a kilogram (kg). The mass of one cubic centimetre of water would be called a gram (g).

Brass weights were made with mass equal to the new unit, the kilogram, then later a weight of platinum was made and adjusted to the value for the new unit (18827.15 French grains. Note: the French grain was a sub-unit derived from an old Parisian bullion weight). The platinum weight became known as the kilogram des archives and effectively became the standard of mass for most of Europe.

International interest in measurement standards grew quickly and in 1870, and again in 1872, the French Government called meetings to discuss the construction and distribution of new metric standards. At the third meeting, in 1875, eighteen countries subscribed to a treaty called the Convention du Metre. At the same time the Comite International des Poids et Mesures (International Committee on Weights and Measures - CIPM) and the Bureau International des Poids et Mesures (International bureau of weights and measures - BIPM) were set up to be responsible for the custody and verification of metric standards. The convention agreed that a new kilogram weight (artefact, not definition) should be made using an alloy of 90% platinum and 10% iridium because of its stability and ability to withstand handling.

After many attempts in France, a successful casting of the alloy was made by George Matthey of Johnson, Matthey and Co of London and in 1879 three cylindrical pieces of the alloy were delivered to the metallurgist St-Claire Deville in France. The cylinders were hammered in a press, and then polished and adjusted and finally compared with the kilogram des archives by M Collot, a maker of weights and balances. By 1883 the CIPM were convinced that one of the cylinders "was indistinguishable in mass from that of the kilogram des archives" and this weight was chosen as the international prototype of the kilogram and called K

A further 40 one-kilogram weights were ordered from Johnson, Matthey & Co in 1882 and delivered in 1884; after re-melting and hammering, to increase their density, these were adjusted - had material removed by polishing - to be close in mass to the selected international prototype. In 1889 the signatories of Convention du Metre (by now including 20 countries) were each allocated one of the weights.

The allocation was made by lot and the UK was given Copy No 18. The certificate which accompanied 'Kilogram 18' on its first journey to the UK gave its mass as 1 kg + 0.070 mg, with an uncertainty of ±0.002 mg, and its volume at 0 °C as 46.414 ml.

To this day Kilogram 18 is the cornerstone of measurements of mass in the UK. It has been back to the BIPM and intercompared with K on numerous occasions and in 1991 it took part in what was called the Third Periodic Verification - a period during which all copies of the Kilogram were systematically compared. At that time its mass was found to be 1 kg + 0.053 mg ±0.002 3 mg, and its volume at 0 °C (calculated from the original value in millilitre) as 46.414 9 cm3.

The measurement uncertainty of ±0.002 3 mg associated with Kilogram 18 following the Third Period Verification of all official copies is marginally larger than the uncertainty of ±0.002 mg quoted back in 1889. But all other comparisons in the intervening century-and-a-bit only managed to achieve uncertainties that were significantly higher (worse); the reason is interesting and surprisingly common in metrology, if little appreciated.

As time passes, new research-derived knowledge tends to highlight a growing list of previously unknown subtle factors that influence a particular type of measurement - mostly to its detriment. With such new knowledge there grows a corresponding list of additional measurement uncertainties and their sum total gets bigger, not smaller. In the case of mass metrology at the highest level, it has taken over a century for improved techniques to compensate for the extra sources of uncertainty that have been discovered since the advent of the kilogram. The values of uncertainty quoted in 1889 and 1991 may be similar but there is no doubt that the more recent one represents a genuinely more accurate set of measurements. Despite best endeavours in 1889, the values were too optimistic. This problem is not confined to the history of weighing; it applies just as much today and in all areas of metrology.

(See a related everyday example of the same problem is a measurement uncertainty of, say, 0.095% meaningfully better than 0.10%?)

The Unit of Mass in the SI system

The unit of mass in the International System of Units (SI) is the kilogram, which is abbreviated to kg. The value of the kilogram is defined as being equal to the mass of the international prototype of the kilogram. The kilogram is unique in the SI, being the only unit whose definition is based on a physical artefact. The pound is still used in the UK for legal metrology and is defined in terms of the kilogram by the Weights and Measures Act.

Also see:

Last Updated: 17 Feb 2012
Created: 8 Oct 2007

Registration

Please note that the information will not be divulged to third parties, or used without your permission

Login