Tide-predicting machine

See also: Tide, Theory of tides, and Tidal force
10-component tide-predicting machine of 1872-73, conceived by Sir William Thomson (Lord Kelvin), and designed by Thomson and collaborators, at the Science Museum, South Kensington, London

A tide-predicting machine was a special-purpose mechanical analog computer of the late 19th and early 20th centuries, constructed and set up to predict the ebb and flow of sea tides and the irregular variations in their heights – which change in mixtures of rhythms, that never (in the aggregate) repeat themselves exactly.[1] Its purpose was to shorten the laborious and error-prone computations of tide-prediction. Such machines usually provided predictions valid from hour to hour and day to day for a year or more ahead.

The first tide-predicting machine, designed and built in 1872–73, and followed by two larger machines on similar principles in 1876 and 1879, was conceived by Sir William Thomson. Thomson had introduced the method of harmonic analysis of tidal patterns in the 1860s and the first machine was designed by Thomson with the collaboration of Edward Roberts (1845-1933, assistant at the UK HM Nautical Almanac Office), and of Alexander Légé, who constructed it.[2]

In the US, another tide-predicting machine on a different pattern was designed by William Ferrel and built in 1881–82.[3] Developments and improvements continued in the UK, US and Germany through the first half of the 20th century. The machines became widely used for constructing official tidal predictions for general marine navigation. They came to be regarded as of military strategic importance during World War I,[4] and again during the Second World War, when the US No.2 Tide Predicting Machine, described below, was classified, along with the data that it produced, and used to predict tides for the D-Day Normandy landings and all the island landings in the Pacific War.[5] Military interest in such machines continued even for some time afterwards.[6] They were made obsolete by digital electronic computers that can be programmed to carry out similar computations, but the tide-predicting machines continued in use until the 1960s and 1970s.[7]

Several examples of tide-predicting machines remain on display as museum-pieces, occasionally put into operation for demonstration purposes, monuments to the mathematical and mechanical ingenuity of their creators.

  1. ^ See American Mathematical Society (2009) II.2, showing how combinations of waves in non-commensurable frequencies cannot repeat their resultant patterns exactly.
  2. ^ The Proceedings of the Inst.C.E. (1881) contains minutes of a somewhat disputatious discussion that took place in 1881 over who had contributed what details. Thomson acknowledged previous work of the 1840s relating to the general mechanical solution of equations, plus a specific suggestion he had from Beauchamp Tower to use a device of pulleys and a chain once used by Wheatstone; Thomson also credited Roberts with calculating the astronomical ratios embodied in the machine, and Légé with design of the drive gear details; Roberts claimed further credit for selecting other parts of the mechanical design.
  3. ^ Ferrel (1883).
  4. ^ During World War I, Germany built its first tide-predicting machine in 1915-16 when it could no longer obtain British hydrographic data (see Deutsches Museum exhibit, online), and when it specially needed accurate and independently-sourced tide data for conducting the U-boat campaign (see German Maritime Museum exhibit, online).
  5. ^ See Ehret (2008) at page 44).
  6. ^ During the 'Cold War', East Germany built its own tide-predicting machine in 1953-55 "at unbelievable expense", see German Maritime Museum (online exhibit).
  7. ^ The US No.2 machine was retired in the 1960s, see Ehret (2008); the machine used in Norway continued in use until the 1970s (see Norway online exhibit).