On the night of 22 October 1707, Admiral Sir Cloudesley Shovell was leading twenty-one ships of the Royal Navy home from Gibraltar when his flagship HMS Association struck the Western Rocks of the Scilly Isles and went down in under four minutes. Three more ships followed her onto the rocks within the hour. Somewhere between 1,400 and 2,000 sailors drowned in the dark water off Cornwall, in what the Royal Museums Greenwich still lists as one of the worst maritime disasters in British history.
The fleet had been lost for twelve days.
Not lost in the sense of missing — they knew roughly where they were on the north-south axis, because latitude could be read off the sun at noon with a backstaff. What they could not read, and what no ship in 1707 could reliably read, was longitude. Their east-west position was a guess built on dead reckoning: speed times time, corrected for currents nobody had mapped, in weather nobody could forecast. Shovell’s navigators thought the fleet was safely west of Ushant, off the coast of Brittany. They were about 100 miles north of that, aimed straight at the Scillies.
A problem older than the compass
Sailors had known the shape of the longitude problem for two centuries. The Earth turns 360 degrees in 24 hours, which means every hour of time difference between two points on the globe equals fifteen degrees of longitude. If you knew the exact time at a reference port — say, London — and you also knew the exact local time on board your ship (easy enough, from the sun), the difference in hours would tell you exactly how far east or west you had travelled.
The catch was the clock. Pendulum clocks, the most accurate instruments on land in the seventeenth century, were useless at sea. The pitch and roll of a ship threw the pendulum off. Temperature changes made the metal parts expand and contract. Humidity rusted the mechanisms. A clock that lost or gained even three seconds a day would, over a six-week Atlantic crossing, put a ship more than fifty miles off course — enough to miss an island, or hit one.
Spain had offered a prize for a solution in 1567. The Dutch States General had offered another in 1636. Galileo had proposed using the moons of Jupiter as a celestial clock, which worked beautifully on a stable observatory floor and terribly on a rocking deck. Nobody had cracked it.
Parliament writes a cheque
The Scilly disaster changed the political weight of the problem. Within seven years, Parliament had passed the Longitude Act of 1714, which offered a tiered reward for any method of determining longitude at sea: £10,000 for a method accurate to one degree, £15,000 for two-thirds of a degree, and £20,000 — a fortune equivalent to several million pounds today — for a method accurate to half a degree, or about thirty nautical miles after a six-week voyage to the West Indies.
A Board of Longitude was set up to judge the entries. Its commissioners included the Astronomer Royal, the President of the Royal Society, and the First Lord of the Admiralty. They were expecting an astronomer to win. Most of the serious money was on the lunar distance method, which used the moon’s position against the fixed stars as a kind of celestial clock.
Instead, the letters started arriving from a village in Lincolnshire, from a self-taught clockmaker who had trained as a carpenter.
The clockmaker from Barrow
John Harrison was born in 1693 in Foulby, Yorkshire, and grew up in Barrow upon Humber, Lincolnshire, where his father worked as a carpenter. He built his first pendulum clock at twenty, almost entirely out of wood — oak wheels, lignum vitae bearings that lubricated themselves through the wood’s natural oils. The clocks he made in his twenties, some of which still run at the Worshipful Company of Clockmakers’ collection at the Science Museum in London, kept time to within a second a month. That was better than most metal clocks made by trained horologists in London.
Harrison went to see the Astronomer Royal, Edmond Halley, around 1730 with drawings for a sea clock. Halley sent him on to George Graham, the finest instrument-maker in England. Graham reportedly talked to him for ten hours, gave him an interest-free loan, and told him to build it.
The first sea clock, H1, took Harrison five years to finish. It weighed 34 kilograms, stood about two feet high, and used counter-oscillating balances connected by springs so that the motion of the ship cancelled itself out. It had no pendulum. It needed no lubrication. In 1736 it was tested on a voyage to Lisbon and back, and on the return leg Harrison correctly identified that the ship’s officers had miscalculated their position by about sixty miles. The Board gave him £500 to keep going.
Forty-six years of one problem
H1 was a triumph, but Harrison was not satisfied with it. He built H2 between 1737 and 1740, then abandoned it before it was even tested because he had thought of a better design. H3 took him nineteen years. He worked on it from 1740 to 1759, inventing along the way the bimetallic strip — two metals bonded together that bend predictably with temperature change, still used today in thermostats — and the caged roller bearing, an ancestor of the ball bearings in every modern machine.
Then he abandoned H3 too.
The insight that broke the problem open was that a sea clock did not have to be big. Harrison had spent decades assuming that a marine timekeeper needed the mass and stability of a large mechanism. In the late 1750s he watched a pocket watch made for him by John Jefferys and realised a small, fast-beating balance wheel could be made more accurate than any of his huge machines. H4, finished in 1759, looked like an oversized pocket watch, about 13 centimetres across, weighing 1.45 kilograms. It ticked five times a second.
On its 1761–62 sea trial from Portsmouth to Jamaica, H4 lost 5.1 seconds over 81 days at sea. That was an error of about 1.25 minutes of longitude, well inside the half-degree threshold. Harrison had won. On paper.
The Board that would not pay
The Board of Longitude did not want to give the prize to a carpenter’s son. They demanded a second trial, then a third. They insisted Harrison hand over the watch, then that he build copies, then that he explain every part in writing so the mechanism could be reproduced by others. The Astronomer Royal Nevil Maskelyne, who was himself a serious proponent of the lunar distance method, sat on the Board and pushed for delay after delay.
Harrison eventually appealed directly to King George III, who reportedly intervened on his behalf. Parliament finally awarded him the balance of the money in 1773, when Harrison was eighty. He died three years later. He had been working on the longitude problem, in one form or another, since roughly 1727 — depending on how the years are counted, somewhere between forty-four and forty-nine years of a single life spent on a single question.
What kept him going through nearly five decades of one problem is the kind of thing psychologists now try to model. Research on motivational dynamics in goal pursuit describes a goal-gradient effect — the tendency for effort to intensify as one draws nearer to a desired outcome — and a phenomenon called choice perseverance, in which the act of repeatedly choosing a difficult path builds tolerance for setbacks along it. Harrison kept choosing the clock. Every year for forty-six years.
What a clock changed
Captain James Cook took a copy of H4, made by Larcum Kendall and known as K1, on his second voyage in 1772. He wrote favorably about the watch’s performance in his journal, noting its reliability across different climates. By the 1780s, marine chronometers based on Harrison’s principles were being built in London workshops for around £65 each — expensive, but no longer impossible. By 1815 the Royal Navy had roughly a thousand chronometers in circulation. The Beagle carried twenty-two of them when Darwin sailed on her in 1831, partly because the ship’s mission was, on paper, to survey the coasts of South America with unprecedented longitudinal precision.
The knock-on effects were enormous. Accurate longitude made accurate sea charts possible. Accurate charts made global trade routes safer, cheaper, and more predictable. It made the coordinated timekeeping of the nineteenth century — railway time, Greenwich Mean Time, time zones — practical. Every GPS satellite in orbit today solves a version of Harrison’s problem, using atomic clocks accurate to a nanosecond and radio signals travelling at the speed of light. The principle is identical: to know where you are, you need to know what time it is somewhere else.
Silicon Canals has looked at how a Soviet probe photographed the far side of the Moon using scavenged American spy-balloon film, and how Cambridge computer scientists built the first webcam to check a coffee pot. Both stories share the same shape as Harrison’s: a small, stubborn, practical fix to a specific problem that ends up rewiring how the world works.
What is left of him
Four of Harrison’s sea clocks — H1, H2, H3, and H4 — are on display today at the Royal Observatory in Greenwich, the same institution whose meridian defines zero degrees longitude for the entire planet. H1, H2, and H3 have been restored to working order and can be seen ticking under glass. H4 is kept stopped, because running it wears out original components that are now nearly 270 years old.
The Scilly Isles still take ships. In 1997 a memorial was placed on the small island of St Agnes to the sailors of the 1707 wreck; divers had recovered ship’s bells, cannons, and Admiral Shovell’s silver plate from the reef by then. The wreck sites are protected under UK law. On calm days, when the Atlantic is flat and the light is right, the outlines of the Association and the Eagle can still be seen through the water, seventy feet down, on the same rocks that four hundred years of maps had never quite been able to place.
