How long is a second

Sounds easy to answer at first: A second is of course the 60th part of a minute, and that is the 60th part of an hour, and that is the 24th part of a day, the duration of one rotation of the earth around its axis.

But that's not enough for physicists. For them, the day is not a good measure of the unit of time, because the earth does not rotate as evenly as one thinks. The duration of the "mean sunny day" becomes longer and longer when viewed over the centuries; this is also superimposed on periodic (seasonal) and non-periodic fluctuations.

That is why in 1967 the second - as one of the basic units in the international SI system of units - was given an atomic physical definition:

"The second is 9 192 631 770 times the period of the radiation corresponding to the transition between the two hyperfine levels of the ground state of atoms of the nuclide Cs-133."

In 2018, the 26th International Conference on Weights and Measures changed the definition of some units as well as the wording of otherwise unchanged definitions. The definition given here is taken from Directive 2019/1258 of the European Commission. This is an official German translation of the new definitions from the 9th edition of the SI brochure of the International Bureau of Weights and Measures (BIPM). The definitions came into effect on May 20, 2019, the date of the annual World Metrology Day.

“The second, symbol s, is the SI unit of time. It is defined by setting the numerical value 9 192 631 770 for the cesium frequency Δν, the frequency of the undisturbed hyperfine transition of the ground state of the cesium atom 133, expressed in the unit Hz, which equals s−1 is. "

This second also roughly corresponds to the 86,400th part of the mean sunny day. But with the help of atomic clocks, every second can be realized for exactly the same length - much more precisely than before.

You can find more about the new SI and the seven basic units under "Research on the new SI "


Time and frequency, 4.4

PD Dr. Ekkehard Peik

Phone: (0531) 592-4400
ekkehard.peik (at)


Birgit Römer

Telephone: (0531) 592-4401
birgit.roemer (at)


Physikalisch-Technische Bundesanstalt
Bundesallee 100
38116 Braunschweig