– the intergovernmental organization through which Member States act together
     on matters related to measurement science and measurement standards.
Search facility:

| Site map | News | Contact us
50th anniversary of the adoption of the atomic definition of the second

On 13 October 1967, the General Conference on Weights and Measures (CGPM) took the decision to adopt a new definition of the unit of time, the second, based on an atomic transition (Resolution 1 of the 13th CGPM (1967)). The Consultative Committee for Time and Frequency (CCTF) celebrated the 50th anniversary of this fundamental decision by inviting Dr Dennis McCarthy to its 2017 meeting to give a lecture on the history of the second.

Dr McCarthy, a former Director of the Directorate of Time at the United States Naval Observatory (USNO) and former President of the International Astronomical Union (IAU) commissions on Rotation of the Earth and Time, noted during his lecture that a redefinition of the second is very likely to happen. Progress in physics and metrology has allowed the development of new standards in the optical frequency domain, some of which have already been recommended by the CIPM as secondary representations of the second. The best of the optical standards today are about 2 orders of magnitude better than the best of the caesium standards, and 8 orders of magnitude more accurate than the atomic clock developed by Louis Essen in the 1950s.

Time had traditionally been a matter decided by astronomers. The correlation of human activities with the solar rhythm made the 24-hour cycle of the Earth's rotation the natural phenomenon for measuring time, and thus time was defined as a fraction of the day, or to be more precise, of the "mean solar day", representing the apparent motion of the Sun moving around the equator at a constant velocity throughout the year. One mean solar day contained exactly 86 400 seconds.

The second was not among the base units of the metric system adopted in France in 1799; it was Gauss who in 1832 proposed to use the second as the base unit of time in his metre-milligram-second system of units. Thirty years later the British Association for the Advancement of Science (BAAS) agreed to use the second of "mean solar time" as the unit of time. With this decision, the CGS and MKS systems had the same second as the base unit of time. This was the situation prior to the signature of the Metre Convention in 1875.

In the 1930s the variability of the Earth's rotation was discovered. Atmospheric changes as well as solid and Earth tides provoke fluctuations of the angular velocity of rotation that change the duration of the mean solar day, thus rendering it difficult to use as the basis for the definition of the second. From ancient astronomical records the Earth shows a slowing trend of about 2.5 milliseconds per century, and measured in uniform units, the day is becoming longer.

With the intention of improving the definition of the unit of time, the International Astronomical Union considered that the orbital motion of the Earth around the Sun could give access to a more accurate unit of time, and proposed, in 1952, the second defined as a fraction of the tropical year of 1900. The International Committee for Weights and Measures (CIPM) decided to adopt this definition and the CGPM ratified the decision in 1960. Contrary to logic, this definition was supported on a method based on a lower frequency of events, demanding averaging over many years to achieve the required accuracy. The associated time scale, "Ephemeris Time" was difficult to realize and consequently had a short life; it was only used in a few astronomical applications and rotational time continued in use as the practical time scale.

In the 1950s, Louis Essen at the National Physical Laboratory (UK) developed a frequency standard based on caesium-133, and demonstrated that it could be 200 times better than astronomical times, achieving an accuracy of 2 parts in 1010. The era of atomic time was born. William Markowitz at the USNO undertook a world-wide programme to express the caesium frequency in terms of the ephemeris second. The value υCs = (9 192 631 770 ± 30) Hz was found, the uncertainty on the frequency being almost entirely due to Ephemeris Time. This value was finally adopted for the definition of the SI second by Resolution 1 of the 13th CGPM (1967).

This steep evolution in accuracy over fifty years since the adoption of the atomic definition of the second recalls Resolution 2 of the 13th CGPM (1967), which considered that there were good perspectives for realizing other frequency standards of better quality than the caesium to serve to define the second, and invited experts to pursue studies in this direction. This resolution has possibly been the driving force behind such an outstanding improvement.

Select a topic:
BIPM highlights
BIPM CBKT programme
Member States and Associates
Metrology events
Revision of the SI
Staff and recruitments