The first time scales were based on the rotation of the Earth. However, as clocks became more precise, it became clear that the rotation of the Earth was not constant, as Flaamsted thought he had shown (Kant thought it wasn't). By 1950, improvements in clocks made possible a precision greater than the variations in the rotation of the earth. A second found by subdividing the day was no longer adequate, as such a measurement would vary depending on when it was made.

To get away from the variations in the earth's rotation, a conference on Fundamental Constants in Astronomy (Paris, 1950) recommended that the time standard
be based on the revolution of the Earth around the sun, instead of its rotation
on its axis. The second would be defined as a fraction of one particular year, rather than
a fraction of the mean solar day. The length of the year was based on
Simon Newcomb's classic study *Tables of the Sun* (1895).

The International Astronomical Union introduced such a scale, Ephemeris Time, in 1952.

The new scale required a new definition of the second. In 1954 the 10th CGPM proposed that the ephemeris second be "the fraction 1/31 556 925.975 of the length of the tropical year for 1900.0" This was considered insufficiently precise, and in 1956 the CIPM defined the ephemeris second as “the fraction 1/31 556 925.9747 of the tropical year for 1900 January 0d 12h ephemeris time.”¹ In 1960 the 11th CGPM abandoned the definition of the second as a fraction of the mean solar day in favor of the ephemeris second.

The epoch of the time scale was defined in 1958 when the IAU General Assembly resolved: “Ephemeris Time (ET), or Temps
des Ephémérides (TE), is reckoned from the instant, near the
beginning of the calendar year A.D. 1900, when the geometric mean longitude of
the sun was 279° 41′ 48″.04, at which instant the measure of
Ephemeris Time was 1900 January 0^{d} 12^{h} precisely.”²

Ephemeris Time had two shortcomings.

One was that the ephemeris second was based on a standard that could never be measured again. Current years are not the same length as the year in 1900, since years are getting shorter by about half a second per century.

The other was that for some users, a time scale in which every second is just the same as every other, and hence that is not based on the Earth's rotation, is not ideal. Navigators, for example, in crude terms need the sun to be overhead at noon. In the United States, for example, the National Bureau of Standards stations, WWV and WWVB, broadcast a time based on Ephemeris seconds. Meanwhile, the U.S. Naval Observatory station (NSS) broadcast time based on the Earth's rotation, for the use of navigators.

What really did in Ephemeris Time, however, was the development of atomic clocks.

The astronomers stuck with Ephemeris Time until 1979,
when they defined two new time scales
that used the atomic second and that took into account relativity (velocity
affects time). From 1 January 1984, these scales replaced ephemeris time in
national ephemerides like the *Nautical Almanac*.

1.
*Procès-verbeaux des Séances*.

Comité Internationale des Poids et Mesures, 2e série, Tome xxv, session de
1956.

Paris, 1957.

Page 77.

2.
D. H. Sadler, editor.

Proceedings of the 10th General Assembly, Moscow, 1958.

*Transactions of the International Astronomical Union*. Volume X.

New York: Cambridge University Press, 1960.

Copyright © 2000 Sizes, Inc. All rights reserved.

Last revised: 25 March 2004.