The Clock That Measured the Ocean: John Harrison, the Longitude Problem, and How a Carpenter Saved Navigation

The Hook: An Error of Two Seconds, Measured in Human Lives

Imagine you are on a plane over the Atlantic, and the onboard computer knows exactly how far you have traveled from north to south — but not how far from east to west. It can tell you your latitude, but your east–west position is pure guesswork that grows less reliable with every hour. After a few days you could be off by hundreds of kilometers, and not know it.

This was precisely the situation of every seafarer for more than two thousand years. Latitude (north–south) had been determinable since antiquity: measure the altitude of the sun at noon or of the Pole Star above the horizon, consult a table, and you have your latitude. But longitude (east–west) remained an unsolved riddle. Ships lost their way, ran onto rocks, missed islands, starved, or fell to scurvy because voyages lasted longer than planned. Longitude was no academic puzzle — it was a matter of life and death, of trading empires and of military power.

In the end, this centuries-old problem was solved not by a famous astronomer, not by Newton's heirs at the Royal Society, but by a largely self-taught carpenter's son from Yorkshire who spent his life building a clock that could keep time even aboard a rolling, humid, temperature-swinging ship. His name was John Harrison, and his story is at once a triumph of engineering, a lesson in the stubborn struggle between outsider and establishment — and, as we will see, the direct historical ancestor of GPS and of the distributed systems we work with today.

At its heart lies one of the most beautiful ideas in the history of science: to know where you are, you must know when it is somewhere else. Position from time. And an error of a few seconds translates — through the rotation of the Earth — into a positional error of nautical miles.

Part 1: Why Longitude Was So Hard

Latitude Is Easy, Longitude Is a Time Problem

To understand why longitude was so much harder to determine than latitude, you have to consider the geometry of the Earth. The circles of latitude are pinned to the sky: the celestial equator and the poles are fixed reference points that do not rotate along with us. Measure the angle of the Pole Star above the horizon and you have (in the Northern Hemisphere) your latitude directly. Nature herself provides the ruler.

The meridians of longitude, by contrast, have no natural zero and no marker in the sky. The prime meridian — today running through Greenwich — is pure convention, set arbitrarily. And because the Earth rotates, the entire starry sky sweeps across all meridians over the course of a day. A given star reaches its highest point over London at a different clock time than over New York.

Here lies the key: longitude is essentially a difference in time. The Earth turns 360 degrees in 24 hours. That is exactly 15 degrees per hour or, put another way, 1 degree every four minutes. So if you know your local time (the sun is highest at local noon) and simultaneously the time at a reference location (say your home port), then the difference reveals your longitude. If it is noon where you are while it is already 2 p.m. in Greenwich, you are 2 hours × 15 degrees = 30 degrees west of Greenwich.

The solution therefore sounds almost trivial: carry an accurate clock, set it to home time, and compare it each noon with the local solar time. The problem was not the idea — it had been known since the sixteenth century (the Dutch mathematician Gemma Frisius formulated it as early as 1530). The problem was the clock. Quite simply, none existed that could do the job.

Consider what was demanded of that timekeeper. Pendulum clocks that achieved an accuracy of seconds per day on land were utterly useless at sea: the rolling of the ship disturbed the pendulum, the salty, humid air corroded the metal, temperature swings between tropical heat and arctic cold expanded and shrank every component, and the varying force of gravity on the open ocean upset the rate. And the permissible error was tiny. Let us do the arithmetic: at the equator, 1 degree of longitude equals about 60 nautical miles (roughly 111 kilometers). One degree is four minutes of time. So one minute of time equals about 15 nautical miles and a single second about a quarter of a nautical mile. For a clock to stay accurate to half a degree (30 nautical miles) across a six-week Atlantic crossing, it could be off by at most about two minutes over those six weeks — less than three seconds per day, under conditions that drove every clock of the era to distraction. No wonder even Isaac Newton considered the mechanical route hopeless and bet on astronomy.

The Deadly Art of Dead Reckoning

Without reliable longitude, navigators resorted to dead reckoning. The navigator estimated the ship's speed (measured with the log — a piece of wood thrown overboard on a knotted line, hence the "logbook"), noted the heading with the compass, measured elapsed time with the sandglass, and calculated from these how far and in what direction the ship had traveled since the last known position.

The method was a chain of estimates, and every inaccuracy propagated and grew. Currents, leeway from crosswinds, errors in the speed measurement — all of it accumulated over weeks into substantial deviations. A captain might believe himself safely in open water while in truth drifting toward a reef. Many ships therefore adopted the strategy of first sailing to the latitude of their destination and then doggedly "running down" that parallel due east or west — a cumbersome, slow, and predictable procedure that also delivered ships to pirates and enemies who knew exactly where to lie in wait.

The Scilly Disaster, 1707

How high the price of ignorance could be became clear in one of the worst naval catastrophes in British history. On the night of 22 October 1707 (by the then-current Julian calendar), a British fleet under Admiral Sir Cloudesley Shovell approached the English Channel in stormy weather, homeward bound from the Mediterranean. The officers conferred about their position and agreed that they were safely west of the French island of Ushant, with a clear run into the Channel.

They were fatally mistaken. In reality they were far to the north, right off the Isles of Scilly southwest of England. The flagship Association struck the rocks and sank within minutes; three more ships followed. Estimates of the death toll range between 1,400 and 2,000 sailors, including Admiral Shovell himself. It was longitude — their east–west position in the Channel — that they had miscalculated.

The shock of this catastrophe — four warships and thousands of men lost, not to the enemy but to a navigational error — rattled the nation. Together with petitions from merchants and seamen, it led a few years later to one of the most famous state-sponsored innovation drives in history.

Part 2: The Prize and the Two Camps

The Longitude Act of 1714

In 1714, the British Parliament passed the Longitude Act. It offered a tiered cash prize for a practicable method of determining longitude at sea. The top sum was a staggering £20,000 — by today's purchasing power, several million pounds depending on the calculation. The tiers were tied to accuracy, measured on a voyage to the West Indies:

• £20,000 for a method accurate to within half a degree (about 30 nautical miles, or two minutes of time),
• £15,000 for accuracy within two-thirds of a degree,
• £10,000 for a full degree.

To administer the prize, the Act created the Board of Longitude (the Commissioners of Longitude), staffed with astronomers, mathematicians, naval officers, and parliamentarians — including, at times, names such as Isaac Newton and Edmond Halley. The Board could also make interim payments to encourage promising approaches. It is remarkable that a state already used the instrument of the advertised competition (today we would speak of a "prize challenge" or "inducement prize") to spur technological innovation, rather than commissioning an agency to do the job.

An important note on historical accuracy: the popular narrative of "the Longitude Prize" as a single, clearly defined trophy owed to Harrison and maliciously withheld from him is a simplification. The Act provided for tiered rewards and discretionary judgment, and the precise interpretation — what counted as "practicable" and "generally applicable" — was exactly the point of dispute we will return to.

The Astronomers: The Lunar Distance Method

The Longitude Act prescribed no particular technique — it rewarded the result. And indeed two fundamentally different camps stood opposed, which we might call the astronomical and the mechanical.

The astronomical camp wanted to turn the sky itself into a clock. Its most promising variant was the lunar distance method (or "lunars"). The idea: the Moon moves relatively quickly against the background of the fixed stars — it covers its own diameter in about an hour. The sky thus works like a gigantic clock face on which the Moon is the hand. If you measure precisely, with a sextant, the angular distance between the Moon and a bright star (or the Sun), and look up in a table at what Greenwich time that angle was predicted, you obtain Greenwich time — and from it, compared with local time, your longitude.

The catch was the prediction. You needed extremely accurate tables of the Moon's position, and lunar motion is notoriously irregular (its full mathematical description occupied astronomers for centuries). Only when the German astronomer Tobias Mayer supplied usable lunar tables, and the future Astronomer Royal Nevil Maskelyne turned them in 1766 into the first Nautical Almanac (with the Moon's positions precomputed for 1767, at three-hour intervals of Greenwich time), did the method become practically usable.

Even then it remained laborious: a single determination of longitude required a series of careful angular measurements followed by a long, error-prone calculation — the "clearing" correction for refraction and parallax could easily cost a skilled navigator a good four hours. And the accuracy was realistically about 30 nautical miles — right at the edge of what the Longitude Act demanded. The lunar distance method had one priceless advantage: it needed no expensive clock, only a sextant, tables, and arithmetic, and the sky never "ran out of adjustment." But it was slow, demanding, and usable only under a clear sky with a visible Moon.

Against it stood the mechanical idea: simply carry the reference time with you in a clock. So simple, so impossible — until Harrison came along.

Part 3: The Outsider — John Harrison

A Carpenter Who Built Clocks from Wood

John Harrison was born in 1693 in Yorkshire, the son of a carpenter whose trade he first learned. He had no academic education, no guild apprenticeship as a clockmaker, no connections to London's scientific elite. What he had was an extraordinary mechanical instinct, inexhaustible patience, and a willingness to rethink every problem from the ground up.

Harrison built his first clocks in the 1710s and 1720s almost entirely from wood — more precisely from lignum vitae, a tropical timber so hard and naturally oily that it acted essentially self-lubricating at the bearings and needed no film of lubricating oil that would congeal in the cold. These precision longcase clocks achieved an accuracy, unheard of for the time, of about one second per month — orders of magnitude better than the usual clocks of that era. Already here Harrison's leitmotif emerged, running through all his work: the uncompromising war against friction and against temperature — the two enemies of every accurate clock.

Two Inventions: Metal Against Heat, a Grasshopper Against Friction

Two of Harrison's inventions from this early period became building blocks of every later sea clock.

The first is the gridiron pendulum, which he devised around 1726. The basic problem: a metal pendulum grows longer when warm and therefore swings more slowly — the clock runs slow. Harrison's solution was elegant to the point of beauty. He built the pendulum from alternating rods of iron and brass. Because brass expands more than iron when heated, and the metals were arranged so that their expansions acted in opposite directions, the changes in length canceled each other out. The effective length of the pendulum — and thus the rate — stayed constant across a wide temperature range. It was one of the first practical applications of the principle of bimetallic compensation, which became ubiquitous in later horology.

The second invention is the grasshopper escapement. The escapement is the heart of every mechanical clock: the mechanism that releases the stored energy in even portions to the timekeeper while keeping it in motion. Conventional escapements of the time (the verge escapement) produced much friction and a disruptive recoil, and needed lubricating oil that congealed over time and altered the rate. Harrison's grasshopper escapement worked with light, springy arms that engaged and disengaged in a hopping motion (hence the name) — almost frictionless and without lubrication. No oil, no congealing, no recoil. For a clock meant to run maintenance-free for years aboard a ship, this was a decisive advantage.

With this toolkit — friction avoidance and temperature compensation — Harrison set out on the real task: a clock that works at sea.

Part 4: H1 to H3 — The Great Sea Clocks

H1: The Rocking Dumbbell

Between 1730 and 1735, Harrison built his first sea clock in Barrow-upon-Humber, probably with the help of his brother James — today simply called H1. It is an imposing behemoth of gleaming brass, over half a meter tall and weighing about 34 kilograms.

The ingenious core of H1 was its solution to the rocking problem. A pendulum is worthless on a ship. Harrison therefore replaced it with two large, dumbbell-shaped balances, linked to each other by springs and swinging in opposite directions. The trick: when the ship tips one mass one way, it tips the other the opposite way — the disturbances from the ship's motion cancel each other out. The clock was thereby largely independent of the ship's attitude and movement. This was complemented by the grasshopper escapement, gridiron temperature compensation (here applied to the tension of the balance springs), and low-friction bearings. H1 needed no lubricating oil.

The Lisbon Trial, 1736

In 1735, Harrison proudly brought H1 to London and presented it to the scientific community in the workshop of the esteemed clockmaker George Graham. They were impressed. A trial was arranged.

In May 1736, Harrison went aboard HMS Centurion, bound for Lisbon, with H1. The outbound leg went roughly at first, but on the return (aboard the Orford) came the great moment: as the ship neared the English coast, Harrison contradicted the calculations of the ship's officers. They believed a certain headland was Start Point; Harrison declared, on the evidence of his clock, that it was the Lizard, farther west. Harrison was right — the officers had been off by about 60 miles. The clock had worked.

The Board of Longitude responded with cautious encouragement. On 30 June 1737, the Commissioners granted £500 — £250 of it up front — so that Harrison could build an improved clock. It was the first official recognition. Remarkably, Harrison never claimed the full prize for H1; he was his own harshest critic and convinced he could do better.

H2, H3, and Nineteen Years of Struggle

Harrison moved to London and completed his second sea clock, H2, within the promised two years. But it never went to sea: Harrison himself discovered a fundamental design flaw and discarded it.

From 1740 he began work on H3 — and labored over it for an incredible nineteen years. H3 simply would not reach the required accuracy; Harrison made countless changes. Yet these years were not in vain, for H3 gave birth to two further consequential inventions: the bimetallic strip (two metals rolled together that bend when the temperature changes — still ubiquitous in thermostats today) and the caged roller bearing, an early forerunner of the modern ball bearing. Harrison's stubborn war against friction drove him to invent components that gained importance far beyond horology.

And yet H3 was a dead-end masterpiece. The real breakthrough came from an entirely different direction — from an idea that everyone involved initially thought absurd.

Part 5: H4 — The Clock That Changed Everything

From Behemoth to Pocket Watch

The decisive change of mind was a change of scale. Around 1751–52, Harrison commissioned a London watchmaker named John Jefferys to build a pocket watch with a novel balance. To his surprise, it ran excellently. Harrison realized: the road to seaworthiness was not the great, heavy behemoth but a small, fast-oscillating watch. He built these insights into his fourth timekeeper — H4, completed in 1759 and ready for use in 1761.

At first glance H4 looks like an oversized silver pocket watch, about 13 centimeters in diameter and weighing roughly 1.45 kilograms — a fraction of H1's weight. But appearances deceived. The secret revealed itself in the racing tick: H4 ticked five times per second, with a large balance swinging fast and wide. This high frequency and large amplitude made the watch insensitive to outside disturbances — a moving ship could barely throw it off. Added to this were a sophisticated temperature compensation and jewel bearings of diamond and ruby against wear. No one in the 1750s regarded a pocket watch as a serious precision instrument. H4 proved the opposite.

The Jamaica Voyage, 1761/62

The great test came at the end of 1761. Harrison, by now nearly 70, sent his son William on the voyage with H4. On 18 November 1761, HMS Deptford put to sea, bound for Jamaica. Even en route, William impressed the crew: using H4, he predicted arrival at the island of Madeira more accurately than the ship's navigation did.

When the ship reached Jamaica on 19 January 1762, the moment of truth arrived. H4 was compared with the local time determined from astronomical observation. The result was sensational: after 81 days at sea, H4 had drifted only 5.1 seconds from the correct time. Converted into longitude, that corresponded to an error of less than one nautical mile — an accuracy that exceeded the Longitude Act's requirement (half a degree, 30 nautical miles) by more than thirtyfold.

A carpenter's son had built what Newton had thought impossible. One might have assumed the matter settled and the prize due. But now the real fight began.

Part 6: The Dispute — Earned Yet Unpaid

The Board Changes the Rules

Back in England, the Board of Longitude balked. The Commissioners declared that the Jamaica test was not conclusive enough: the astronomical determination of Jamaica's longitude had itself been uncertain, and a single successful run might have been luck. They demanded further tests — and, this was the real bone of contention, they required Harrison to fully disclose the inner workings of H4 and to demonstrate that other watchmakers could verifiably build comparable clocks.

From the perspective of Harrison and his supporters, this was a malicious moving of the goalposts: he had achieved the accuracy the law required, so the prize was his. His friends launched an outright pamphlet and newspaper campaign against the Board.

I am of the opinion that a fair assessment is more nuanced than the popular hero's tale suggests. From the Board's perspective, the point was not a one-off feat of brilliance but a practicable, generally available method for the entire merchant and war fleet. And for that, a single marvel handcrafted by a lone genius over decades is not yet a solution — as long as no one else can reproduce it. At its core, the dispute was an argument over whether a prototype or a reproducible procedure should be rewarded. This question — a single artifact versus a scalable, documented, reproducible-by-others solution — is strikingly modern and confronts every engineer to this day.

Barbados, Cook, and the King

A second test was arranged in 1764, this time to Barbados, under the supervision of Nevil Maskelyne, who also championed the rival lunar distance method. Barbados's longitude was determined independently via the moons of Jupiter. H4 again convinced with outstanding accuracy. In February 1765, the Board confirmed that H4 had stayed within the strictest limits of the law.

Nevertheless, the Board paid only half. It recommended £10,000 upon disclosure of the mechanism, and the remaining £10,000 only once other watchmakers produced reproducible examples. In practice, Harrison received £7,500 in 1765 (less earlier payments). He then had to take H4 apart and disclose it, and hand it over for further testing at the Royal Observatory in Greenwich, where Maskelyne tested it over ten months — with a result disappointing to Harrison, which reignited the dispute.

That Harrison's approach was in fact reproducible was soon proven by another man: the watchmaker Larcum Kendall built K1, a copy of H4. This watch accompanied Captain James Cook on his second great Pacific voyage (1772–1775). Cook praised it effusively as his "never-failing guide" — the practical proof that the marine chronometer revolutionized navigation.

Harrison, by now old and embittered, finally appealed directly to King George III. The King, himself scientifically inclined, had Harrison's fifth clock (H5) tested at his private observatory at Kew and was outraged at the treatment of the old man. At his instigation, Parliament in 1773 granted Harrison a further £8,750 through a special act. It was never the official "Longitude Prize" — which formally no one was ever paid in full — but it was the recognition he deserved. John Harrison died in 1776 at the age of 83, on his own birthday.

An Honest Assessment

It is worth setting the two competing methods soberly side by side, rather than declaring one "the right" one:

Historically, neither method won alone. For decades they complemented each other: the chronometer supplied the daily, convenient longitude, while lunar distances served as an independent check and a fallback should the clock fail or need resetting. Only when chronometers became cheap and reliable in the nineteenth century — watchmakers such as John Arnold and Thomas Earnshaw drastically simplified Harrison's complex design — did the clock displace astronomy. HMS Beagle, which carried Charles Darwin around the world, took along no fewer than 22 chronometers for safety. And Greenwich became the world's prime meridian in 1884 for no accidental reason — it was the consequence of British dominance in chronometer navigation and the publication of the Nautical Almanac.

Part 7: The Principle Behind It — Position from Time

Why should this story interest you beyond its sheer fascination? Because its core principle did not die with Harrison but is the foundation of modern positioning — and because it is the same conceptual figure that underlies every distributed computer system.

The principle: position is a function of synchronized time. Harrison determined longitude by comparing a local time measurement with a transported reference time; the difference in time was the difference in position. Through the constant of the Earth's rotation (15 degrees per hour), an error in time becomes an error in position.

GPS works in exactly this way, only pushed to the extreme. A GPS satellite is essentially a flying, extremely accurate atomic clock that continuously broadcasts time signals. Your receiver compares the arrival times of the signals from several satellites and calculates from them how far away each satellite is — and from that, your position. The relationship is the same as with Harrison, only with the speed of light as the conversion constant instead of the Earth's rotation: because light travels about 30 centimeters in a nanosecond, a clock error of one millionth of a second already leads to a positional error of roughly 300 meters. Where Harrison needed three seconds over six weeks, we today need nanoseconds — but it is exactly the same equation: time error times propagation constant equals position error. GPS is Harrison's idea, executed with atomic clocks and relativistic corrections.

And here the circle closes to the distributed systems we work with. There too, time is the hidden axis on which everything hangs. A distributed system must order events without possessing a perfect shared clock. Google's database Spanner solves this by — entirely in Harrison's spirit — treating time not as an exact point but with a known uncertainty (see Clocks That Know Their Own Uncertainty: Google Spanner, TrueTime, and Mastering Time in the Cloud). Where Harrison had to minimize the uncertainty of his clock over six weeks, Spanner minimizes the uncertainty interval across a global datacenter network. The deepest level of precision timekeeping — clocks even more accurate than the best atomic clocks — is delivered to us today by pulsars, rotating neutron stars that serve as cosmic metronomes (see The Clocks Made of Neutrons: Pulsars, Pulsar Timing Arrays, and the Hum of Spacetime). From the bronze balance wheel to the atomic clock to the neutron star: it is always the same hunt for a reliable beat.

The Central Takeaway

The longitude problem teaches two things at once — a technical lesson and a human one.

The technical lesson is the loveliest formula of this story: if you want to know where you are, you need reliable knowledge of when it is somewhere else. Location arises from the comparison of synchronized time. This insight is worth not less but more today than in 1762 — it underpins every navigation device, every GPS-based application, and, figuratively, every distributed system that must order events.

The human lesson lies in the dispute with the Board of Longitude. Harrison had achieved the required accuracy — and still met resistance, because the Board demanded a reproducible, buildable-by-others, documented solution, not a single marvel in the hands of a lone genius. As galling as that was for Harrison personally, the concern was legitimate as an engineering virtue: a solution only one person can build and understand is not yet a robust solution for a system.

A concrete call to action: This week, take a critical skill or a system in your own work environment that only a single person masters, or that runs only "on X's machine." Treat it the way the Board treated Harrison's H4: demand reproducibility — of yourself or of the team. Document the inner workings, have a second person rebuild it or at least understand it. Ask plainly: "What if this person is gone tomorrow?" The "bus factor" of your most important knowledge is the modern version of Harrison's dilemma.

A question to reflect on: Harrison spent more than four decades minimizing the uncertainty of a clock, because a small error in time became a deadly error in position. Where in your own work does a seemingly tiny inaccuracy at one point — a few milliseconds, a rounded value, an unchecked assumption — turn, through propagation and leverage, into a large, expensive error at the end of the chain?

Cross-References in the Vault

• Clocks That Know Their Own Uncertainty: Google Spanner, TrueTime, and Mastering Time in the Cloud – The direct modern descendant: Spanner, too, makes time with a known uncertainty manageable; Harrison did it for the ocean, Spanner for global distributed systems.
• The Clocks Made of Neutrons: Pulsars, Pulsar Timing Arrays, and the Hum of Spacetime – The continuation of the hunt for the perfect beat: from the bronze balance wheel through the atomic clock to rotating neutron stars as cosmic clocks.
• The Cosmos in Bronze: The Antikythera Mechanism and Humanity's First Computer – Another masterpiece of precision mechanics that rendered sky and time in gearwheels; a kindred spirit of mechanical computation.
• The Clock in Every Cell: The Molecular Rhythm of Life from Fruit Fly to Chrono-Medicine – Timekeeping from an entirely different side: the biological clock that paces life itself.
• The Voice of the Clay Tablets: How Linear B Was Deciphered and the Oldest Greek Learned to Speak – As with Harrison, here too a stubborn outsider (the architect Ventris) solved a problem on which the established experts had broken their teeth.

Sources and Further Reading

• Royal Museums Greenwich – Longitude found: the story of Harrison's timekeepers: https://www.rmg.co.uk/stories/time/harrisons-clocks-longitude-problem
• Royal Museums Greenwich – Longitude found: Nevil Maskelyne and the lunar method: https://www.rmg.co.uk/stories/time/longitude-found-nevil-maskelyne-lunar-method
• Historic England – How did the sinking of a ship in 1707 lead to the invention of the marine chronometer?: https://historicengland.org.uk/listing/what-is-designation/heritage-highlights/sinking-of-ship-lead-to-invention-of-marine-chronometer/
• MacTutor History of Mathematics – English attack on the Longitude Problem: https://mathshistory.st-andrews.ac.uk/HistTopics/Longitude2/
• Wikipedia – John Harrison (overview, continuously updated): https://en.wikipedia.org/wiki/John_Harrison
• Wikipedia – Board of Longitude: https://en.wikipedia.org/wiki/Board_of_Longitude
• Sobel, Dava – Longitude: The True Story of a Lone Genius Who Solved the Greatest Scientific Problem of His Time (1995) – the classic popular-science standard work on the subject.

Created as part of the daily learning workflow. Field of interest: History. Estimated reading time: ~30 minutes.

← All articles