The clock in the jeweler’s window was not for sale. It was a public service: trimmed each week against the sun, it told the town what noon meant here, at this longitude, in this season. Forty miles east the answer differed by a couple of minutes, and both answers were correct. Time in 1883 was not a coordinate you could look up; it was a possession, kept tower by tower, and the railroads — which had to write all of it down — were the first industry to find this unbearable.

I.A possession of towns

Nobody at the time considered local time a problem, because it was not one. Noon was a local fact, the way weather is local, and a town’s clock was right the way a town’s church is right — by settlement, not by measurement. City hall kept the reference dial, the jewelers corrected their stock against it, and the factory whistle answered the clock tower with the confidence of scripture. The arrangement had exactly one defect, which was that it was not a system at all — merely a habit that every town kept for itself.

What the century added was speed. A stagecoach absorbed a four-minute disagreement between one county and the next without comment; a railroad at forty miles an hour turned it into a column of arithmetic on every timetable in the land. Each company ran on the time of some principal city — the time its dispatchers had grown up with — and where lines met, dispatchers kept conversion tables the way customs officers keep exchange rates. By the 1870s the chaos had a number, and the number was worse than the public guessed: one count put the American network on fifty-three separate operating times,1 on a continent that reads and writes by a single sun.

II.The man with the guidebook

The man who did the arithmetic was William F. Allen, a civil engineer by training and the editor of the Official Guide of the Railways by trade — which is to say, the man who printed America’s operating times and had therefore read them. As secretary of the General Time Convention, the railroads’ standing truce on scheduling, Allen proposed in 1883 not to invent time zones but to round them. Five hourly standards, each anchored to a meridian fifteen degrees from the next, each named — Eastern, Central, Mountain, Pacific — for the geography it approximated rather than any city it obeyed.2

The average error of the new standards against local sun time was small; the maximum, in towns far from their meridian, ran past half an hour. What Allen’s table offered was not accuracy but symmetry — the same rule for everyone, wrong in the same direction, at the same moment. The plan succeeded precisely because it cheated everyone equally. Fig. 1 tallies the ledger for six cities on the day: New York would surrender four minutes; Washington, which had been keeping time east of its own station, would take eight back; Detroit, holding a meridian of its own, would owe twenty-eight.

5:00 75TH MERIDIAN 6:00 90TH MERIDIAN 7:00 105TH MERIDIAN 8:00 120TH MERIDIAN New York +4 min Washington −8 min Detroit +28 min Chicago +10 min Denver ±0 min San Francisco −10 min Local mean time, 1883 New railroad standard offsets relative to Greenwich
Fig. 1The minute ledger of November 18, 1883. Dots mark each city’s local mean time computed at four minutes per degree of longitude; diamonds mark the new standard for its zone. Deltas show the correction applied at noon — Denver, sitting almost exactly on the 105th meridian, is the convention’s one honest town, accurate by a pure accident of longitude.3

The plan succeeded precisely because it cheated everyone equally.”

III.The Sunday of the two noons

Allen chose a Sunday — November 18, 1883 — when the trains ran light. At noon in each new zone the time services went out over the wires: the Naval Observatory’s signal in the East, Western Union relays beyond it, and station agents, hotel clerks, and jewelers advanced or retarded their clocks by the instructed minute. In towns whose clocks had been running ahead of the new standard the hands went backward, and noon arrived twice; the newspapers needed no better headline than the fact itself, and mostly did not use one. The country has been calling it the Day of Two Noons ever since, though only half the country got two.

The changeover was a ceremony of small embarrassments rather than grand ones. No parade is possible for an event whose entire content is four minutes, and none was attempted. Crowds gathered anyway — at the telegraph offices when the signal came through, at the jewelers’ windows where the reference clocks were corrected in public, beneath town dials whose hands a locksmith walked forward to their new truth. What the crowds had come to watch was arithmetic they could not feel: the same sun, the same shadows, a different noon.

IV.The holdouts

Municipal pride kept its own schedule. City councils that had raised clock towers as civic statements were slow to demote their dials to branches of a national convention, and the new standards leaked across the map town by town rather than zone by zone. Detroit held out longest of the large cities, keeping its own mean time — both times posted at the station, like a bilingual border town — until 1900, when the council finally voted the city onto standard time.4 Elsewhere the pattern repeated in miniature: a year or two of dual time in the newspapers, then a quiet Tuesday when nobody printed the old column.

What settled the argument everywhere was not ordinance but inconvenience. Every pocket watch still on local time was a missed connection waiting to be itemized in a claim, and the railroads’ arithmetic was patient in a way that civic sentiment was not. When Congress finally acted — the Standard Time Act of 1918, thirty-five years after the fact — it was not introducing zones so much as notarizing them: the borders were drawn where the habit had already worn a groove.5

V.The meridian of the world

The year after the railroads, the world took a shorter turn. Twenty-five nations sent delegates to Washington for the International Meridian Conference of 1884, and by a vote of twenty-two to one, with two abstentions, the prime meridian of the world became the one passing through the Royal Observatory at Greenwich — a choice defended less on principle than on arithmetic, since Greenwich already carried most of the world’s shipping charts.6 The delegates adjourned having standardized a line, not a clock. Where the hours began was now settled; everything else was implementation.

The implementation was left to the schedules. Sandford Fleming had been arguing since the 1870s for a planetary system of twenty-four hourly bands, and the world built roughly what he proposed — not by treaty but by timetable, one railway, telegraph office, and post office at a time. Standard time is now infrastructure of the purest kind: an idea whose complete success is that almost nobody can remember there was ever an alternative.

VI.Coda

The reference clocks stayed in the jewelers’ windows a while longer, corrected by habit against a sun nobody consulted. Then the shopkeepers noticed that nobody was checking, and the clocks became ordinary merchandise, sold like everything else in the window. Noon had become something you join rather than something you keep — a subscription your ancestors took out on your behalf, one Sunday, in minutes nobody missed.

Notes

  1. 1The count of fifty-three operating times appears in the convention papers of the early 1880s; Allen himself preferred to say “about fifty,” which was, if anything, generous to the timetable writers.
  2. 2The five standards of 1883: Intercolonial (60th meridian), Eastern (75th), Central (90th), Mountain (105th), and Pacific (120th) — hourly steps of fifteen degrees, each degree worth four minutes of sun.
  3. 3Local offsets in Fig. 1 are local mean time computed at four minutes per degree of longitude from published city longitudes; corrections are rounded to the minute as the railroads rounded them.
  4. 4Detroit’s council ordinance of 1900 ended seventeen years of dual time; the city’s jewelers reportedly took down one set of signs and doubled the price of the rest.
  5. 5The Standard Time Act (1918) gave the Interstate Commerce Commission authority over zone boundaries and, briefly and less durably, over daylight saving time.
  6. 6International Meridian Conference, Washington, October 1884. The Greenwich resolution passed 22–1 with two abstentions; France and Brazil abstained, France holding out on civil time until 1911.