The Hour Has No Good Reason to Exist
Of all the units of measurement humans have invented, time units are the strangest. Length can be grounded in the body — the foot, the cubit, the hand. Mass can be grounded in physical objects — a grain of barley, a block of metal. Even temperature can be grounded in observable physical events — the freezing and boiling of water. But time has no physical object to anchor it, no body part to borrow its scale from, no substance whose behaviour could serve as a reference. Time is something we experience but cannot hold, and every unit we have invented to measure it is, at some level, an act of intellectual improvisation.
The second is defined by the vibration of a caesium atom — a number so precise it staggers comprehension. The minute is sixty seconds because ancient Babylonian mathematicians preferred arithmetic in base sixty. The hour is sixty minutes for the same reason. The day exists because the Earth rotates, but the Earth's rotation is slowing, irregularly, unpredictably, which means the day itself is not a fixed quantity. The week has seven days because of a Mesopotamian belief about the planets. The month is a ghost of the lunar cycle, imperfectly preserved in a solar calendar. The year is the time the Earth takes to orbit the Sun, but we cannot divide it cleanly into days, which is why we have leap years, and leap seconds, and the slow accumulated grief of calendars that refuse to stay aligned with the sky.
Every time unit carries this history inside it: the collision of astronomical reality with human arithmetic, the friction between how the universe actually moves and how people need to organise their lives. This is the story of how we invented the units of time — and how strange it is that any of them work at all.
Babylon and the Number Sixty: Why We Still Count the Way They Did
To understand why an hour has sixty minutes and a minute has sixty seconds, you have to go back approximately four thousand years to the river valleys of Mesopotamia, where Babylonian mathematicians developed one of the most sophisticated numerical systems the ancient world ever produced — and chose to base it not on ten, but on sixty.
The choice of sixty was not arbitrary. Sixty is what mathematicians call a highly composite number: it is evenly divisible by 1, 2, 3, 4, 5, 6, 10, 12, 15, 20, and 30. No number smaller than sixty has as many divisors. This made base-sixty — which historians call the sexagesimal system — extraordinarily convenient for practical arithmetic. If you need to divide a quantity into halves, thirds, quarters, fifths, sixths, tenths, or twelfths, sixty divides cleanly every time. In an era before calculators, when merchants and astronomers were doing arithmetic by hand or with counting tokens, this property was genuinely valuable.
The Babylonians used sexagesimal notation for astronomy in particular, tracking the movements of the Moon, the planets, and the Sun with remarkable precision. They divided the sky into 360 degrees — 6 times 60, another number chosen for its divisibility — and subdivided each degree into 60 minutes of arc and each minute into 60 seconds of arc. When Greek astronomers inherited Babylonian astronomical knowledge, they kept the sexagesimal system intact. When medieval scholars in the Islamic world and later in Europe translated and extended Greek astronomy, the sixty-based angular divisions came with it.
The transfer from angular measurement to time measurement was a natural step. Ancient astronomers tracked time by observing the positions of celestial objects, so the angular system and the time system were always intertwined. The Babylonians had already used sexagesimal fractions to express fractions of a day. By the time medieval European scholars were using water clocks and early mechanical clocks to divide the day into smaller units, the framework of sixty was already so deeply embedded in the mathematics of timekeeping that nobody seriously considered an alternative. The hour had sixty minutes. The minute had sixty seconds. The decision had been made four thousand years earlier by people who needed to divide the sky.
This is why your phone, your microwave, your car dashboard, and every other timekeeping device on Earth still counts in units inherited from the Babylonian sexagesimal system. It is one of the most durable intellectual decisions in human history, and it was made before writing was fully developed, before the wheel was common, before bronze had replaced stone as the dominant material for tools.
The Hour Itself: A Unit That Was Never Fixed
We tend to think of the hour as a fixed quantity — sixty minutes, 3,600 seconds, an exact and reliable division of the day. But the hour spent most of its history as something far more fluid, and the idea that all hours are equal is a surprisingly recent invention.
The ancient world divided the day into twelve hours of daylight and twelve hours of night, a system borrowed from Egypt and transmitted through Greece and Rome into medieval Europe. The problem with this system — obvious once stated — is that daylight hours and night hours are only equal to each other at the equinoxes, twice a year. At midsummer in Rome, a daylight hour was nearly ninety minutes long by our reckoning, while a night hour was barely forty-five. At midwinter, the proportions reversed. These were called temporal hours or seasonal hours, and for a civilisation without mechanical clocks, they made good practical sense: the day was divided into twelve equal parts of daylight, regardless of how long the daylight lasted, and that gave a consistent social framework even if the hours themselves varied in length.
The seasonal hour persisted for centuries because it matched human experience. People went to bed at the same hour of night in summer and winter, rose at the same hour of morning, attended religious offices at the same hours, even though those hours represented different absolute durations. The Roman system of scheduling public events, Christian canonical hours, and daily social rhythms all operated on seasonal time without apparent discomfort.
Mechanical clocks changed everything. The first mechanical clocks in medieval Europe, appearing in the late 13th century, could not adjust their rate to match the varying length of seasonal hours. They beat at a fixed frequency, indifferent to the calendar. This forced a choice: either redesign the clock to accommodate seasonal hours, which was mechanically very difficult, or redefine the hour to be a fixed duration, which would require reorganising social life around clock time rather than daylight time. Over the course of the 14th and 15th centuries, European society gradually made the second choice. The equal hour — one twenty-fourth of a full day, fixed regardless of season — became the standard, and the clock tower became the new authority over daily life.
This was a profound cultural shift. For the first time, the primary social experience of time was mediated by a mechanical device rather than by the sky. The clock did not tell you when the sun was at its zenith or how much daylight remained; it told you a number, abstracted from direct celestial observation. Time became a quantity to be measured rather than a phenomenon to be observed.
The French Revolutionary Calendar: When Ten Tried to Replace Sixty
If the hour's history demonstrates how hard it is to change a time unit once established, the French Revolutionary calendar demonstrates what happens when you try anyway.
The French Revolution of 1789 was not content to transform politics and society. It wanted to transform the very framework of daily life, including time itself. In 1793, the National Convention adopted a new Republican Calendar that replaced the Gregorian calendar's months with twelve new months of thirty days each, named after natural phenomena and agricultural seasons: Vendémiaire (vintage), Brumaire (mist), Frimaire (frost), and so on through the year. The five or six remaining days were national holidays called sans-culottides.
More radically, the French also introduced decimal time: each day was divided into ten decimal hours, each decimal hour into one hundred decimal minutes, each decimal minute into one hundred decimal seconds. This made a decimal second approximately 0.864 of a conventional second and a decimal hour approximately 2.4 conventional hours. New decimal clocks were manufactured. Official documents were to use decimal time. The entire framework of sixty-based timekeeping was to be swept away in favour of the rational clarity of ten.
It lasted about two years. The practical difficulties were immense. Decimal clocks were expensive and rare. Workers and merchants who had organised their lives around the conventional hours refused to adapt. Religious rhythms, which ran on a seven-day week rather than the ten-day décade that the new calendar substituted, created constant friction. Most critically, France did not exist in isolation: trade, correspondence, and navigation with the rest of Europe required constant conversion between decimal and conventional time, which defeated much of the purpose of the reform. Napoleon officially abolished decimal time in 1805, and it was never seriously revived.
The failure of decimal time is instructive. It demonstrated that time units are not merely technical specifications but social agreements, and social agreements cannot be changed by government decree unless the entire society switches simultaneously. The sixty-based hour was not just a number stored somewhere. It was embedded in the rhythms of work, worship, commerce, and conversation for everyone in Europe simultaneously, and the cost of replacing it all at once was too high for any government to bear. The Babylonian decision, made four thousand years earlier, proved more durable than the French Republic.
The Earth Is an Unreliable Clock
Lurking beneath all of this history of invented time units is a deeper problem: the Earth itself is a poor timekeeper.
We define the day as the time it takes the Earth to rotate once on its axis. We define the year as the time it takes the Earth to complete one orbit of the Sun. These seem like stable, natural quantities — and for most of human history, they were reliable enough. But in the 19th and 20th centuries, as timekeeping became precise enough to detect tiny variations, it became clear that neither the day nor the year is fixed.
The Earth's rotation is gradually slowing due to tidal friction from the Moon. Over geological timescales, the effect is dramatic: 500 million years ago, the Earth rotated in about 21 hours, and a year contained more than 400 days. The slowing is too gradual to notice in a human lifetime — the day lengthens by about 1.4 milliseconds per century — but it is real and measurable, and it means that the second, if defined as a fraction of the day, would itself grow longer over time.
More immediately disruptive is that the Earth's rotation is not just slow — it is irregular. The rate of rotation fluctuates slightly from year to year, influenced by the movement of the liquid outer core, the redistribution of mass in the atmosphere and oceans, large earthquakes, and even the seasonal melting of ice. These fluctuations are small — fractions of a millisecond — but they are enough to cause the astronomical day to drift relative to a fixed-rate clock. Before precise atomic timekeeping existed, this was undetectable. Once atomic clocks arrived, it became a problem that needed a name and a solution.
The Atomic Second: Detaching Time From the Sky
The solution came in 1967, in a decision that was scientifically revolutionary and philosophically momentous: the second was redefined in terms of atomic physics rather than astronomy.
The new definition — still in force today — states that one second is exactly 9,192,631,770 periods of the radiation corresponding to the transition between two hyperfine levels of the ground state of the caesium-133 atom. This sentence is worth pausing over. The second is no longer defined as a fraction of a day. It is defined by a quantum mechanical process inside a specific type of atom, a process that is the same in every caesium atom everywhere in the universe, regardless of what the Earth happens to be doing.
The caesium number — 9,192,631,770 — was chosen specifically so that the new atomic second would match the existing astronomical second as precisely as possible at the moment of transition. The definition was calibrated to preserve continuity with history. But the philosophical break was total. Time was no longer measured by the sky. It was measured by atoms.
The precision this enabled is difficult to comprehend. Modern caesium atomic clocks lose or gain less than one second over 300 million years. The best optical lattice clocks — the current frontier of timekeeping technology — are accurate to one second over 15 billion years, which is longer than the current age of the universe. These clocks are so precise that they must correct for the effects of general relativity: a clock at a higher altitude runs slightly faster than one at sea level, because it sits in a slightly weaker gravitational field, and modern clocks are precise enough to detect the difference in elevation between a clock placed on a table and one placed on the floor.
The Leap Second: Keeping Atoms and the Sky in Agreement
The redefinition of the second in 1967 created an immediate practical problem. The atomic second was now fixed, but the astronomical day was not — and the day defined by Earth's rotation kept drifting relative to atomic time. If left uncorrected, atomic clocks would slowly diverge from solar time, and eventually noon on the clock would not correspond to noon in the sky.
The solution, introduced in 1972, was the leap second: an occasional extra second inserted into the official time scale to keep atomic and astronomical time within 0.9 seconds of each other. Unlike the familiar leap day added to February every four years on a predictable schedule, leap seconds are irregular, announced only six months in advance by the International Earth Rotation and Reference Systems Service, which monitors the actual rotation of the Earth and decides when an adjustment is needed. Between 1972 and 2016, 27 leap seconds were inserted into the official time scale, almost always at midnight on December 31 or June 30.
The leap second has proven enormously controversial in the modern era — not among the general public, who mostly ignore it, but among computer systems engineers and network administrators. The internet's infrastructure is built around the assumption that time is a smooth, continuously increasing quantity. A leap second inserts an extra second where the clocks read 23:59:60 — a moment that does not exist in ordinary timekeeping — and this has caused significant software failures. In 2012, a leap second caused crashes across hundreds of major websites including Reddit, Yelp, LinkedIn, and Gawker. The problem is not that the extra second is hard to handle in principle, but that time-handling code is extremely subtle and the failure modes of getting it wrong are often silent until they suddenly are not.
In 2022, the General Conference on Weights and Measures voted to abolish the leap second by 2035, allowing atomic and astronomical time to diverge freely. This decision effectively concedes that the abstraction has won: for practical purposes, the official second belongs to the atom, not the sky.
The Week: Seven Days for Seven Planets
While the second, minute, and hour trace their lineage to Babylonian astronomy and medieval mechanics, the week is a different kind of invention entirely. It has no astronomical basis in the same sense as the day, month, or year. It does not correspond to any observable celestial cycle. It is purely a human construct, and its seven-day structure reflects a cosmological belief that most people in the modern world have forgotten they are carrying.
The seven-day week originated in ancient Mesopotamia and spread through Jewish tradition, Roman culture, and eventually Christianity and Islam to become the nearly universal standard it is today. The Babylonians considered the numbers one through seven to be significant, associated with the seven celestial objects visible to the naked eye: the Sun, the Moon, Mars, Mercury, Jupiter, Venus, and Saturn. These seven objects governed days in rotation, and the day governed by each one took on the character of that object's associated deity.
The modern names of the days in English make this inheritance visible if you know where to look. Saturday is Saturn's day. Sunday is the Sun's day. Monday is the Moon's day. Tuesday is Tiw's day (Tiw being the Norse equivalent of Mars). Wednesday is Woden's day (Woden being the Norse equivalent of Mercury). Thursday is Thor's day (Thor being the Norse equivalent of Jupiter). Friday is Frigg's day (Frigg being the Norse equivalent of Venus). The English week is a Babylonian astronomical system refracted through Roman and then Norse religion, all of it still ticking along in the words we use for the most ordinary divisions of our schedule.
The seven-day week has proved even more resistant to reform than the sixty-based hour. The French Revolutionary calendar replaced it with a ten-day décade, as noted above, and this was one of the most deeply resisted aspects of the reform. The Soviet Union tried a five-day week in 1929 and a six-day week in 1931 before abandoning both and returning to seven days in 1940. The practical resistance of the week to change is partly cultural and partly economic — a shared week is necessary for coordination, and coordinating an entire society to switch simultaneously is effectively impossible — but it is also, perhaps, something deeper. The seven-day rhythm has been embedded in human life for roughly three thousand years. It structures not just schedules but language, religion, and the way people experience the passage of time.
The Month and the Year: An Imperfect Marriage
The most fundamental tension in calendar design is the one between the Moon and the Sun. The Moon's cycle — from new moon to new moon — takes approximately 29.53 days. The Earth's orbit around the Sun takes approximately 365.2422 days. Neither number divides cleanly into the other, and this irreconcilable mismatch has generated an enormous amount of ingenious human effort over thousands of years.
A pure lunar calendar — like the Islamic Hijri calendar — tracks twelve lunar months, giving a year of approximately 354 days. This is 11 days shorter than the solar year, which means the lunar calendar completes a full cycle relative to the seasons every 33 years. Ramadan, observed in the ninth month of the Hijri calendar, gradually rotates through all the seasons, falling sometimes in winter and sometimes at the height of summer over a person's lifetime. There is no attempt to stay aligned with the agricultural or astronomical year; the lunar cycle is the primary reference.
A pure solar calendar — like the ancient Egyptian civil calendar — ignores the Moon entirely and divides the year into months that bear no relationship to lunar cycles. The Egyptian calendar had twelve months of thirty days each, plus five additional days at the end of the year. The months were regular and arithmetically convenient, but they had nothing to do with the actual Moon.
The Gregorian calendar that most of the world uses today is a solar calendar with vestigial lunar influence. The months have irregular lengths (28, 29, 30, or 31 days) that reflect the messy history of Roman calendar reform rather than any astronomical logic. The year is kept aligned with the solar cycle by the addition of a leap day every four years, with corrections for century years, producing an average year of 365.2425 days — accurate to within one day per 3,030 years relative to the actual solar year of 365.2422 days. It will be the year 4909 before the Gregorian calendar is a full day out of step with the actual orbit of the Earth.
What Time Actually Is
Underneath all the Babylonian arithmetic, the medieval clock towers, the French revolutionary failures, the caesium atoms, and the leap seconds, there is a question that timekeeping can answer only partially: what is time itself?
Physics offers a strange answer. In Newtonian mechanics, time was absolute — a universal backdrop against which events occurred, flowing at the same rate everywhere and for everyone. Einstein's relativity shattered this. Time runs at different rates depending on how fast you are moving and how strong the gravitational field around you is. There is no single universal "now" that all observers share. Time is not a backdrop but a dimension, woven together with space into a four-dimensional fabric that curves in the presence of mass and energy. The clocks on GPS satellites must correct for relativistic effects — both the slowing of time from their speed and the speeding of time from their reduced gravity — or they would drift by ten kilometres per day relative to positions on the ground.
But even setting relativity aside, time has a property that no other physical dimension shares: it only goes one way. You can move freely in any direction through space. You cannot move freely through time. The past is fixed and inaccessible; the future is open but unreached. This asymmetry — called the arrow of time — is one of the deepest unsolved problems in physics. Nothing in the fundamental equations of physics requires time to move in one direction; the laws work equally well forwards and backwards. And yet time moves relentlessly in one direction for everyone, everywhere, always. The reason is still not fully understood.
The units we use to measure time — the second, the minute, the hour, the day, the week, the month, the year — are human responses to this mystery. They impose a grid on something that resists gridding, carve regularity from something that is only approximately regular, and anchor the most abstract of human experiences to the vibration of an atom, the rotation of a planet, and the counting of a Babylonian astronomer working in base sixty four thousand years ago.
The hour has no good reason to exist. And yet it has never been more precisely defined.
A Final Note on the Numbers
Every time you glance at a clock showing 14:37:22, you are reading a document written in layers of history. The 24-hour format divides the day as ancient Egyptians divided it, into two twelve-hour halves. The 37 minutes is a Babylonian fraction of the hour, expressed in the base-sixty system that came from the river valleys of ancient Mesopotamia. The 22 seconds is a further Babylonian subdivision, maintained across four millennia, now anchored to a quantum transition in caesium-133. And the whole thing is coordinated by leap seconds that compensate for the irregular wobble of the Earth, updated by an international bureau that monitors the planet's rotation to within fractions of a millisecond.
The most ordinary act of telling the time is a four-thousand-year collaboration between Babylonian astronomers, medieval clockmakers, French revolutionary reformers, 20th-century atomic physicists, and the slowly spinning, gently wobbling, entirely indifferent Earth.