Teddington, Middlesex

The National Physical Laboratory (NPL) is in Teddington, Richmond upon Thames, TW11, near Hampton Court.

In 1895, Douglas Galtonb's Presidential Address to the BAAS urged the establishment of a National Physical Laboratory. The BAAS set up a committee with Galton as Chairman and including Rayleigh. A deputation, including Rayleigh, presented their case to the Prime Minister, Lord Salisbury (Rayleigh's uncle-in-law) and the Government agreed to set up a committee of enquiry with Rayleigh as chair. On 6 July 1898, the committee reported setting up NPL under the control of the Royal Society. Parliament approved. Rayleigh was Vice Chairman of the General Board and Chairman of the Executive Committee. On 5 Jul 1899, Richard Tetley Glazebrook was appointed first Director, commencing on 1 Jan 1900. He continued until 1919. NPL is located near Bushy House, which was given to the Laboratory in 1900. Bushy House was used as the Director's residence and the administrative offices. The Laboratory opened in Mar 1902. [A.T. Humphrey, "Lord Rayleigh - the last of the great Victorian polymaths", Bull. Inst. Math. Appl. 31:7/8 (Jul/Aug 1995) 113-120]

It is often said that Bushy House was left to Newton's niece, Catherine Barton Conduitt, by George Montague, first Earl Halifax and Keeper of Bushy Park [Craig, p.28], but [White & Foster, p.17, 22-28] gives a detailed history showing that Halifax and Barton lived at Upper Lodge, which was replaced by 1840 and was on the site used by the Admiralty Research laboratory in 1945-1994. Bushy House was built in 1663 by a courtier of Charles II, possibly as a place for the king to meet his mistresses. It was embellished by George Montagu, first Earl of Halifax of the second creation, nephew of Charles Montagu, who succeeded his uncle as Keeper, and the House was generally the Keeper's residence from then on. In 1797, the Keepership was given to William, Duke of Clarence, third son of George III, who lived there with his mistress, Dorothy Jordan, the leading comedy actress of the time. They already had three children and they had seven more at Bushy House, resulting in appropriate alterations. When it became apparent that William might succeed to the kingdom, he dumped Mrs Jordan in c1812, who had supported him for twenty years - she had to flee her creditors and died in poverty near Paris in 1816. William then married Princess Adelaide of Saxe Meiningen and brought her to Bushy House and the house was then expanded considerably in 1820-1832. In the mid nineteenth century, the Keepership appears to have disappeared. In 1896, the occupier of Bushy House died and it reverted to the Crown. Victoria exchanged it for two government houses in Pall Mall and the government then assigned it to the new National Physical Laboratory. [White & Foster, pp.29-41]

Glazebrook had been Maxwell's assistant at the Cavendish Laboratory and worked with him on electrical standards. He was Assistant Director of the Cavendish in 1891-1899. He brought the electrical standards work with him. The Laboratory was originally supervised by the Royal Society and Kew Observatory was a part of it from 1900 to 1912. NPL was transferred in 1918 to the Department of Scientific and Industrial Research. Later it was part of the Ministry of Technology. Kelvin and Rayleigh were early supporters and consultants. There are roads and buildings named for Glazebrook, Kelvin and Rayleigh.

Lewis Fry Richardson was here c1903-c1908, NPL resisted carrying out experiments based on Lanchester's theory of aerodynamic lift, which consequently was not developed until the work of Prandtl in the 1920s [G.R. Taylor, pp.41-42].

Robert Alexander Watson-Watt (1892 1973) was superintendent of the Radio Department at NPL in 1934 when the Committee for the Scientific Survey of Air Defence was formed in November and asked his opinion whether a 'death-ray' was feasible. He said there was no way enough energy could be projected, but that a plane ought to reflect enough energy that it could be detected and, using more than one station, even located - this was the practical invention of radar. Hertz had observed reflections of radio waves c1888; Richardson had patented the idea of echo-location, covering both sound and radio, in 1912?; Baird had patented the idea in 1926 and detected a plane in 1927; Kahnold had detected ships in 1933. [Weber, pp.62-63; Storer, pp.118-121.] See also Seaview. Watson-Watt later coined the term 'operational research'.

W. L. Bragg was Director in 1937.

About 1938, J.S. Clarke used a reversible pendulum to determine g as 9.81183 m/sec2 to probably better than one part in a million.

C.G. Darwin was Director 1938-1949.

Olga Taussky was here, 1940-1947.

J. R. Wormersley was head of the new Mathematics Division from 1944. It was then located in Cromer House, but later moved to Teddington Hall.

Turing was here 1945-1948 and produced the proposal for the ACE (Automatic Computing Engine) but no action was taken until after Turing had left for Manchester. The Pilot ACE was then built and it ran in 1950.

J.H. Wilkinson was here for many years, beginning as Turing's assistant on the ACE in 1946.

Leslie Fox was also here in 1945-1957.

Donald Watts Davies (1924-2000) joined NPL in 1947 to work with Turing on the Pilot ACE after hearing Turing lecture about it. He remained until retirement in 1984. In 1965, he conceived of packet switching as the method for computer to computer communications. He picked the name 'packet' with care to be sure there were related words in many languages. In Oct 1967, the idea was presented at a US conference and ARPANET soon was redesigned to use it. Paul Baran, at Rand Corporation, independently developed the same idea, perhaps earlier, but originally in the context of telephone networks where it was not very suitable.

Louis Essen joined NPL, c1930. In the 1930s, he developed the quartz ring clock, the first clock accurate enough to routinely determine variations in the speed of the earth's rotation, which are several parts in 108. In the war he developed the cavity resonance wavemeter, later widely used. He then adapted the devices to measure the speed of light to greater accuracy than ever done before, getting 299,792 km/sec in 1946 and 299,792.5 km/sec in 1950 (the current value is 299,792.458 km/sec and is now used to define the metre). He became Deputy Director of NPL in charge of frequency standards. In the early 1950s, he visited the US and saw early but unsatisfactory caesium clocks. Returning to NPL, he and J.V.L. Parry designed a built a much better version. A paper read to the Royal Society in Dec 1956 described this and showed it was accurate to one part in 1010, i.e. 1/3 second per century. This clock is now in the Science Museum [F.A.B. Ward (2), p.104]. By the time the paper was published, the accuracy was improved to one part in 1012. This clock became the basis of the present coordinated universal time (UTC) adopted on 1 Jan 1972. International Atomic Time (TAI) had been running since 1956, but the relation to astronomical time had not been thought out and used both frequency adjustments and insertion of tenths of seconds - UTC provides for whole 'leap seconds' to be inserted when necessary to keep UTC coordinated with astronomical time (UT2) - see below. Essen's booklet, see below, describes how observations of time signals over three years (1955-1958 ?) determined atomic time in terms of Universal Time at the US Naval Observatory in Washington which was then related to Ephemeris Time, leading to the definition of the second by the 1967 General Conference of Weights and Measures in Paris as "9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium-133 atom." By 1973, time standards accurate to about 1 part in 1012 were readily available. The first version of the caesium clock is now in the Science Museum, while the second version is in the NPL Museum. Essen also carried out modern versions of the Michelson-Morley experiment and other basic experiments relating to relativity, leading to some criticisms of Einstein's work. OBE, 1959. FRS, 1960. Retired from NPL in 1972. [Anthony Tucker, "Split-second decisions", The Guardian (4 Sep 1997) 15; Louis Essen, The Measurement of Frequency and Time Interval; HMSO for National Physical Laboratory, 1973]

As of early 1998, 240 atomic clocks were involved in UTC, in 35 laboratories in 24 countries, coordinated by the International Earth Rotation Service at the Observatory in Paris. NPL has 7 of these clocks. The clocks mostly use vibrations of caesium-133 atoms. The earth is found to be running a bit slow by about one second every 400 days, leading to the occasional introduction of 'leap-seconds', which have been added at the end of December or June. 21 leap-seconds were added during 1972-1997, but they were not regularly spaced as the earth's motion is not uniform. All observatories, radio stations, etc., now use UTC, but I have read that GMT is still the legal time!

NPL presently maintains caesium clocks with an accuracy of one part in 1013. Lengths are determined by a He-Ne laser stabilised with Iodine to about 3 parts in 1011. The British copy of the standard kilogram (No. 18) is kept at NPL. NPL's scales are accurate to about one part in 109. The triple point temperature of water, 273.16 K, can be reproduced with an accuracy of 1 mK. Angles can be measured with an accuracy of 0.1".

NPL is not normally open to the public, but they have some open days for schools each year and other interested parties can attend - telephone 020-8943 6400 to find out when the open days occur. There is also a museum - telephone 020-8943 6054 to arrange a visit. My thanks to Sue Osborne for showing me around.

One of the present buildings is the Glazebrook Aerodynamics Building, but I think it no longer has wind tunnels. There is a Hooke Building. The Laboratory stretches over 82 acres and has about 700 staff.

To see an Ordnance Survey map click at THIS LINK

Gazetteer Index Main MacTutor index

An extract from The Mathematical Gazetteer of the British Isles created by David Singmaster

The original site is at THIS LINK