My son asked me, how people knew that an hour is an hour before they invented the clock. Reading through History of timekeeping devices on Wikipedia and also Hour it became apparent to me, that there was hours of different length in the history until "more accurate current clock" was invented.

If I would have to guess, the first attempts at this would not have given the modern measure of hour anyway, surely there were a few iterations until we settled on the Hour as we know it.

So I would like to know, when people started to use modern hour, and how they arrived to this exact definition of hour, and who was the key people making that happen (if known).

If no precise answer can be given, I'll settle for date ranges.

  • Actually, the hour -- and the second -- did become "fixed" in length only when atomic clocks were invented. (GMT defined the second as 1/86400th of a full rotation of the earth, and the rotation time differs, hence the leap seconds in UTC.)
    – DevSolar
    Nov 10, 2016 at 8:47
  • Related: history.stackexchange.com/questions/28957/…
    – SJuan76
    Nov 10, 2016 at 10:47
  • @SJuan76, yeah, it's an excellent answer, also linked in the "Related" sidebar on the right. Nov 10, 2016 at 10:52
  • As I recall, the Romans divided daylight into 12 hours; as a result the hours varied in length from season to season. The French tried decimal hours. I suspect that @Alex nailed it with Maskelyne's Nautical Almanac, specifically with ephemeris time.
    – MCW
    Nov 10, 2016 at 17:59
  • @DevSolar: the hour (and the second) were fixed long before the atomic clock invention: by the Sun and stars motion in the sky. We just found it convenient to fix them DIFFERENTLY.
    – Alex
    Nov 10, 2016 at 22:44

1 Answer 1


This happened gradually and no specific person or date can be stated. Ancients divided day into equal parts and night into equal parts. So day and night hours were unequal. Of course they realized this. To measure time in the ancient world, there were sundials and water clocks. In the Hellenistic times there were sophisticated water clocks showing unequal hours during the day and night. With the invention of mechanical clocks, gradually people started to measure time by mechanical clocks, but this was a slow process. They begin to spread in 14th century, and first ones were tower clocks in the cities. I suppose in the villages, people still used Sun and unequal hours. The standard of clocks was based on the revolutions of Sun (one full revolution=24 hours).

However, astronomers realized very long ago (in the ancient times) that revolutions of Sun are also not equal: even if you divide day+night (one complete revolution of Sun) into equal parts, these hours will be unequal because of the equation of time (see "equation of time" in Wikipedia). So the time unit should be based on the revolutions of stars, which is more uniform.

For people other than astronomers, this was clearly realized and described only in the middle of 18th century, with the publication and use of the Nautical Almanac by Nevil Maskelyne. Again attempts were made to produce sophisticated mechanical clocks which would show "true" solar time, that is unequal hours. Only with the spread of precise pocket watches, people started using "equal hours" defined astronomically, as the "mean solar time", that is the revolutions of the starry sky was taken as the standard.

In the end of 19th century scientists realized that any hours defined astronomically are also slightly unequal. It was J. C. Maxwell who first proposed to base the definition of time units on physical processes (frequency of certain radiation) instead of astronomy. His proposal was adopted in the beginning of 20s century. So the modern definition of hour is not related to the revolutions of the Earth at all, and the hour is defined as 3600 seconds, and the second, the fundamental time unit, is defined as "the duration of 9 192 631 770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the cesium 133 atom". The final refinement of this definition, specifying the conditions under which this radiation has to be measured was established in 1999, if you wish a date.

Of course the corrections made in 20th century are too small to concern our daily life. So for the hour used in daily life, the standard is a watch, and it spread slowly with the spread of non-expensive watches in the beginning of 20s century.

The described process can be summarized as follows: people use some natural process which is thought to be uniform, to measure time. With the development of knowledge, they discover that the process is in face not exactly uniform. They find a new process, which is first used by scientists, but then its use gradually spreads. And this story repeats and lasts as long as our civilization lasts.

  • OP asked specifically about "modern" time measurement. I the end of your first paragraph and the second paragraph answer the question.
    – MCW
    Nov 10, 2016 at 13:46
  • It depends:-) The last change in the standard happened in 1999, so it is rather the third paragraph from the end. All depends on what is really meant by "modern time measurement". In science or in daily life.
    – Alex
    Nov 10, 2016 at 13:54
  • In modern daily life, some of us use mechanical watches, some electronic, and some use time signals transmitted from atomic clock. But the first two are checked and adjusted by atomic clock.
    – Alex
    Nov 10, 2016 at 13:57
  • I think you mean "the last two" - unless your mechanical watch is way more fancy than any I've seen.
    – MCW
    Nov 10, 2016 at 13:58
  • I use a good non-expensive mechanical watch made in 1960s. I adjust it once a week from my computer, which means from the atomic clock signals transmitted by Internet.
    – Alex
    Nov 10, 2016 at 14:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.