The world before standard time
For most of human history, time was local. A town's noon was the moment the sun crossed its meridian, and the church bell or town clock was set accordingly. A traveler riding from Boston to New York in 1840 would pass through dozens of slightly different "noons," and nobody minded, because nobody could travel fast enough for the difference to matter. The railway changed that. By the 1870s a passenger could cross hundreds of miles in a few hours, and a station master trying to publish a timetable had to choose between the local time of his own town and the local time of the line's headquarters.
Sandford Fleming and the worldwide proposal
The Canadian engineer Sir Sandford Fleming is usually credited with the first detailed proposal for a global system of standard time zones. After missing a train in Bandoran, Ireland in 1876 because of a printed-timetable ambiguity, Fleming began advocating a single 24-hour worldwide clock divided into hourly zones, anchored to a prime meridian. By 1883 the major North American railroads had adopted a five-zone system covering the continent.
November 18, 1883: "the day of two noons"
On Sunday November 18, 1883, the U.S. and Canadian railroads switched to standard railway time. In cities east of the new zone boundaries the clocks were turned back; in some cities clocks chimed noon twice. Newspapers called it "the day of two noons." Standard time was not yet law — Congress did not formally adopt it until the Standard Time Act of 1918 — but in practice the railroads had forced the country onto a uniform grid.
The 1884 International Meridian Conference
In October 1884, delegates from 25 nations gathered in Washington, D.C. for the International Meridian Conference. They agreed, by a vote of 22 to 1, that the meridian passing through the Royal Observatory at Greenwich, England would be the prime meridian of the world. The choice was practical: roughly two-thirds of the world's shipping already used charts based on Greenwich. France abstained and continued to use Paris Mean Time, with a fudge factor, until 1911.
Why 24 zones became 38
In a perfectly geometric world, the Earth would be divided into 24 zones, each one hour wide and exactly 15 degrees of longitude across. Reality is messier. China, which spans roughly five geographic zones, uses a single time (UTC+08:00) imposed by Beijing. India and Sri Lanka use UTC+05:30. Nepal uses UTC+05:45. Newfoundland uses UTC−03:30. The Chatham Islands use UTC+12:45. Counting all of these, there are about 38 distinct standard offsets currently in use.
The IANA tz database
Software does not actually store time zones as offsets, because offsets change. Instead, almost every operating system in the world uses the IANA Time Zone Database, an open-source project maintained since 1986. It records every known historical transition for every named zone — for example, "America/New_York", "Europe/Paris", "Asia/Kolkata" — and is updated several times a year as governments announce new rules.
UTC: not GMT, not quite
Coordinated Universal Time (UTC) is the basis of civil time worldwide. It is often used interchangeably with Greenwich Mean Time, but the two are not identical. GMT is a time zone based on astronomical observation at Greenwich; UTC is an atomic time scale, defined by an ensemble of about 450 cesium and hydrogen-maser atomic clocks at roughly 80 laboratories around the world, coordinated by the BIPM in Sèvres, France. Because Earth's rotation is slowing, UTC is occasionally adjusted by a "leap second" to keep it within 0.9 seconds of UT1.
Daylight saving time
Daylight saving time (DST) was first proposed seriously by the New Zealand entomologist George Hudson in 1895 and independently by the British builder William Willett in 1907. It was first adopted nationally by the German Empire in April 1916 to save coal during the First World War, and the UK followed three weeks later. DST is observed by about 70 countries today, mostly in Europe and North America.
How a clock on your computer knows what time it is
A modern computer keeps time by combining a hardware clock, a software clock incremented from the Unix epoch (January 1, 1970 UTC), and synchronization with external time servers using the Network Time Protocol (NTP). NTP can synchronize a desktop computer to within tens of milliseconds of UTC. For applications that need tighter accuracy, the Precision Time Protocol (IEEE 1588) can reach sub-microsecond accuracy.
Further reading
- Bartky, Ian R. Selling the True Time: Nineteenth-Century Timekeeping in America (Stanford University Press, 2000).
- Blaise, Clark. Time Lord: Sir Sandford Fleming and the Creation of Standard Time (Pantheon, 2000).
- Galison, Peter. Einstein's Clocks, Poincaré's Maps: Empires of Time (W. W. Norton, 2003).