Longitude

The Book

In November 1993, William J.H. Andrewes organized a three-day Longitude Symposium at Harvard, bringing together historians, scientists, and horologists to present different facets of the centuries-long quest to determine position at sea. Dava Sobel attended as a science journalist and recognized what the academics had not: this was a story that could hold a general audience. Longitude was published in 1995, went through 29 hardcover printings, and was translated into 24 languages. It reached number one on the Times bestseller list in England and essentially invented the genre of short, single-subject narrative nonfiction about the history of science.

The problem it describes is deceptively simple. Sailors had been able to determine latitude — north-south position — for centuries by measuring the angle of the sun or Polaris above the horizon. Longitude — east-west position — was effectively impossible. The theoretical solution was known: if you could carry the exact time at a reference point (London) while simultaneously knowing the local time aboard ship, the difference gives you degrees of longitude. One hour equals fifteen degrees. But no clock existed that could keep accurate time aboard a rolling, pitching, salt-sprayed, temperature-fluctuating ship for months at sea.

Without longitude, navigation was dead reckoning — estimating position from heading, speed, and elapsed time. Over long voyages, errors compounded. Ships missed island landfalls, ran aground on reefs they didn't know were there, and sailed hundreds of miles off course until provisions ran out. Thousands of ships were lost. The economic and strategic consequences for maritime empires were enormous.

The Scilly Disaster

On the night of October 22, 1707, a fleet of Royal Navy warships under Admiral Sir Cloudesley Shovell, returning from the siege of Toulon, ran aground on the rocks of the Isles of Scilly off Cornwall. Shovell and his navigators believed they were safely west of Brittany. They were a hundred miles off. Four warships went down. Between 1,450 and 2,000 sailors drowned, including Shovell himself, whose body washed ashore on St. Mary's island.

This was not an obscure merchant vessel but the Commander-in-Chief of the British Fleets, with an entire squadron. The humiliation contributed to Parliament passing the Longitude Act of 1714, establishing the Board of Longitude and offering a top prize of £20,000 — equivalent to millions today — for a method accurate to within half a degree (30 nautical miles). Britain was not the first nation to offer such a reward; Portugal, Spain, and the Netherlands had tried since 1598. But the British prize was the largest and most formalized.

John Harrison

Harrison (1693–1776) was a self-taught Yorkshire carpenter who became the most consequential horologist in history. He followed his father into woodworking but developed an early obsession with clockmaking. By his twenties he was building longcase clocks from lignum vitae — a dense, self-lubricating tropical hardwood — that required no oil and kept time to within a second per month, rivaling the best London-made instruments.

Around 1730, Harrison conceived a marine clock and traveled to London. Edmond Halley referred him to George Graham, the country's foremost clockmaker. Graham was so impressed he personally lent Harrison money to build his first sea clock — an extraordinary act of generosity from the leading figure in his field toward a provincial stranger.

H1 through H3: The Sea Clocks

Harrison spent five years building H1 (completed 1735), a large brass mechanism with twin interconnected dumbbell balances that cancelled out the ship's motion, a grasshopper escapement of his own invention requiring no lubrication, and temperature compensation using alternating brass and steel rods. Tested on a voyage to Lisbon in 1736, H1 correctly identified landfall as the Lizard when the ship's master had the fleet sixty miles east — a potentially fatal error.

H2 (1739) refined the design but was never sea-trialed: Harrison discovered the dumbbell balances were susceptible to yaw errors, and Britain was at war with Spain. H3 consumed nineteen years. It produced two inventions of permanent importance — the bimetallic strip (now ubiquitous in thermostats) and the caged roller bearing (ancestor of the ball bearing in every machine on Earth) — but Harrison could never get it accurate enough.

H4: The Breakthrough

While struggling with H3, Harrison had a pocket watch made to his own specifications by John Jefferys. Its performance astonished him. He realized a small, high-frequency oscillator — a balance wheel beating five times per second — was inherently more stable at sea than the large mechanisms of H1–H3. He abandoned the sea-clock approach entirely.

H4 (completed 1759) resembled a large pocket watch: 5.2 inches in diameter, 3.2 pounds. It featured diamond pallets in the escapement, a refined remontoire, and jeweled bearings. On its Jamaica trial in 1761–62, carried by Harrison's son William, H4 lost only 5.1 seconds over 81 days — an error of roughly 1.25 nautical miles, well within the prize threshold of 30. William correctly predicted an earlier-than-expected landfall at Madeira, impressing the ship's captain.

The Battle with the Board

Despite two triumphant sea trials, Harrison did not receive his prize. The Board of Longitude, dominated by astronomers championing the rival lunar distance method, demanded H4's surrender, full disclosure of its workings, and proof that others could replicate it. Nevil Maskelyne, Astronomer Royal and a Board Commissioner with a personal stake in the astronomical approach, took custody of H4 and tested it at Greenwich, where it performed suspiciously poorly.

Harrison was in his late seventies before King George III personally intervened, having H5 — Harrison's final timekeeper — tested at his own observatory at Kew. Over ten weeks, H5 lost one-third of one second per day. The King reportedly told Harrison: "By God, Harrison, I'll see you righted!" and threatened to appear before Parliament himself. Harrison received his last payment in 1773, at age 80. He never received the official prize. He died in 1776.

The Astronomers' Method

The principal alternative was the lunar distance method: measuring the angular distance between the Moon and reference stars, then comparing to pre-calculated tables to determine Greenwich time. It required a skilled observer, a sextant, clear skies, and three to four hours of calculation for each observation.

Sobel frames the competition as a class conflict. Harrison was a provincial carpenter — a manual craftsman, an outsider to London's intellectual elite. The Board of Longitude was staffed by gentlemen astronomers, Fellows of the Royal Society, who believed a problem so grand should yield to theoretical astronomy and mathematics, not to gears and grease. The Royal Observatory at Greenwich existed specifically to map the heavens for navigation; an astronomical solution validated their institutional mission.

Modern historians have complicated this narrative. The Board's demand for replicability was arguably reasonable — a unique, irreplaceable watch was of limited use to the fleet. Maskelyne's Nautical Almanac (1767), which enabled the lunar method, served the Navy well for decades. Both approaches proved complementary in practice. But the human story — the craftsman versus the establishment — is what gives the book its force.

Selected Quotes

"The zero-degree parallel of latitude is fixed by the laws of nature, while the zero-degree meridian of longitude shifts like the sands of time. This difference makes finding latitude child's play, and turns the determination of longitude, especially at sea, into an adult dilemma — one that stumped the wisest minds of the world for the better part of human history."

— On the fundamental asymmetry of the problem

"Time is to clock as mind is to brain. The clock or watch somehow contains the time. And yet time refuses to be bottled up like a genie stuffed in a lamp... The most we can hope a watch to do is mark that progress. And since time sets its own tempo, like a heartbeat or an ebb tide, timepieces don't really keep time. They just keep up with it, if they're able."

— On the nature of time and timekeeping

"With his marine clocks, John Harrison tested the waters of space-time. He succeeded, against all odds, in using the fourth — temporal — dimension to link points on the three-dimensional globe. He wrested the world's whereabouts from the stars, and locked the secret in a pocket watch."

— On Harrison's achievement

Where We Are Now

Harrison's insight was that position is a function of time. Carry precise time with you, compare it to a local reference, and you know where you are. Every navigation system built since rests on this same principle. The instruments have changed beyond recognition. The idea has not changed at all.

The Chronometer After Harrison

Harrison proved the concept; John Arnold and Thomas Earnshaw made it a product. Arnold introduced the term "chronometer" and built the first production models in the 1780s. Earnshaw developed the spring detent escapement that became the universal standard — keeping time to within one to two seconds per day. By the early nineteenth century, prices dropped from a significant fraction of a ship's value to something a merchant captain could afford. Marine chronometers remained the primary method of oceanic navigation until radio time signals in the early twentieth century.

The precision progression since Harrison tells the story of engineering ambition:

In 1967, the definition of the second was changed from astronomical (a fraction of the Earth's rotation) to atomic (9,192,631,770 oscillations of cesium-133). Harrison demonstrated that a man-made mechanism could keep better time than the heavens. The 1967 redefinition made it official.

Inertial Navigation: Carrying Position Without the Stars

Harrison's problem resurfaces in a different form for intercontinental ballistic missiles. An ICBM in flight cannot rely on radio signals (which an enemy could jam), landmarks, or any external aid. Like a ship in the open ocean, it must carry its own knowledge of position from launch to impact.

Charles Stark Draper at MIT solved this with inertial navigation. The principle: gyroscopes maintain a stable reference frame (they remember which direction is which), accelerometers measure every change in velocity, and a computer continuously integrates — acceleration integrated once yields velocity; velocity integrated once yields position. Starting from a known launch point, the system accumulates increments to maintain a running position estimate. No external references needed.

The early Atlas ICBMs (late 1950s) used radio-inertial hybrid guidance — the missile's onboard sensors transmitted data to a ground station, which sent course corrections back. This was a vulnerability: the ground station could be attacked, the radio link jammed. The Titan II and Minuteman series switched to fully autonomous all-inertial guidance, eliminating any dependence on external signals.

The Peacekeeper missile (1986–2005) carried the Advanced Inertial Reference Sphere (AIRS) — widely considered the most accurate INS ever built. A beryllium sphere containing three accelerometers and three gyroscopes floats freely in fluorocarbon fluid with no gimbals, eliminating mechanical friction entirely. AIRS achieved a CEP (circular error probable) of approximately 90 meters at intercontinental range — placing a warhead within a football field after traveling 8,000 miles, using nothing but internal measurements.

The parallel to Harrison is direct. Both problems are fundamentally about self-contained position knowledge — carrying your reference with you rather than depending on external landmarks. Both are limited by the precision of their instruments: Harrison by his springs and balance wheels, an ICBM by its gyroscopes and accelerometers.

Space Shuttle Navigation

The Shuttle carried three redundant Inertial Measurement Units, each containing gyroscopes and accelerometers on a gimbaled platform. Before launch, the IMUs were calibrated and aligned to the precisely surveyed position and orientation of the pad, establishing the initial conditions for all subsequent dead-reckoning navigation.

Two star trackers compared observed star positions against an onboard catalogue of roughly 100 bright stars, using the discrepancy to correct IMU drift — a direct descendant of celestial navigation, automated and precise to fractions of a degree. A backup Crewman Optical Alignment Sight let crew members manually sight on stars if the trackers failed.

During reentry, the Shuttle entered a plasma sheath that blocked all radio communications — no GPS, no ground contact, no beacons. For roughly twelve minutes, the vehicle relied entirely on IMU dead reckoning and drag-altitude estimation, deriving altitude from sensed deceleration via a Kalman filter. The transition from pure internal navigation through blackout to aided navigation for final approach was one of the most critical phases of every mission.

The avionics ran on five IBM AP-101 General-Purpose Computers. Four operated in synchronized lockstep running identical software; if one disagreed, it was voted out. The fifth ran independently written backup code — an entirely separate software implementation as the ultimate safeguard against a common-mode bug.

GPS: Harrison's Method from Orbit

GPS is the longitude problem solved from space. The connection could not be more direct.

Each GPS satellite carries atomic clocks (rubidium and cesium standards) and broadcasts the precise time of transmission along with its orbital position. A receiver on the ground measures the time-of-flight of each signal. Because radio waves travel at the speed of light, time-of-flight gives distance. Four satellites yield four equations for four unknowns: latitude, longitude, altitude, and the receiver's clock error.

Harrison demonstrated that if you carry an accurate clock and compare it to local solar time, you can determine your position. GPS does the same thing in reverse and in three dimensions: atomic clocks in orbit, radio signals replacing astronomical observations, and the comparison happening at the speed of light rather than at local noon.

The system requires relativistic corrections that Harrison could not have imagined. GPS satellites orbit at 20,200 km altitude, where weaker gravity causes their clocks to run 45 microseconds per day faster (general relativity). Their orbital velocity of 14,000 km/h causes clocks to run 7 microseconds per day slower (special relativity). Net effect: satellite clocks run 38 microseconds per day fast. Light travels 11.4 km in 38 microseconds. Without correction, GPS would accumulate 10 km of error per day — useless within two minutes. The fix: satellite clocks are slightly detuned before launch so they appear correct from the ground.

Until May 1, 2000, the U.S. government deliberately degraded the civilian signal through Selective Availability, introducing errors of up to 100 meters. President Clinton ordered it turned off at midnight. Accuracy jumped from ~100 meters to ~20 meters overnight. The newest GPS III satellites have SA physically removed — it cannot be reinstated.

When GPS Fails: The Return to First Principles

GPS has a vulnerability Harrison would understand intuitively: it depends on an external signal. Jam or spoof that signal and the navigator is blind. Russia has routinely jammed GPS in Ukraine since 2014 and spoofs signals around Moscow. Both sides in the current conflict extensively jam and spoof GPS, disrupting drones, precision-guided munitions, and communications.

The response has been a return to self-contained navigation — Harrison's approach, updated with twenty-first-century physics:

• Chip-Scale Atomic Clocks (CSACs) — DARPA-funded devices the size of a sugar cube that maintain atomic-clock-grade timing for hours after losing GPS, allowing INS systems to continue providing accurate position data.
• Quantum inertial sensors — Atom interferometry uses the wave nature of ultracold atoms to measure acceleration and rotation with extraordinary precision. In 2025, Infleqtion and the Royal Navy demonstrated a quantum optical atomic clock on an underwater autonomous submarine — precision timing in an environment GPS cannot reach.
• Celestial navigation revival — In 2016, the U.S. Naval Academy reintroduced mandatory celestial navigation training after a decade-long hiatus. The ancient art of navigating by the stars, updated with modern computation, cannot be jammed, spoofed, or denied.

Modern submarines use the AN/WSN-7 ring-laser-gyro inertial navigation system as their primary sensor, supplemented by GPS fixes when at periscope depth, electronic bottom-contour matching, and the STELLA celestial navigation computer — which processes periscope observations of the sun through code developed by the U.S. Naval Observatory.

From a Yorkshire carpenter's workshop in the 1730s to quantum optical clocks on submarine drones in 2025, the thread is unbroken. The problem is the same. Carry precise time with you, and you can always find where you are.

Verdict

Longitude is 175 pages and reads in an afternoon. Sobel turns what could be a dry history of timekeeping into a genuine page-turner — part technical marvel, part institutional drama, part underdog story. The Kirkus review called it "a science book as enjoyable as any novel." A tenth-anniversary edition added a foreword by Neil Armstrong.

Some criticism is deserved. Sobel's portrait of Maskelyne as villain has been complicated by historians who note the Board's demands for replicability were reasonable, and that both the chronometer and lunar distance methods ultimately proved complementary. The narrative sacrifices some nuance for dramatic arc. But the core story — a self-taught craftsman solving a problem the scientific establishment insisted was theirs — is irresistible, and the engineering details of Harrison's timekeepers remain genuinely awe-inspiring.

The 2000 television adaptation, starring Michael Gambon as Harrison and Jeremy Irons as Rupert T. Gould (the Royal Navy officer who painstakingly restored Harrison's neglected clocks in the 1920s), won five BAFTAs. Harrison's original timekeepers are on permanent display at the Royal Observatory, Greenwich. H1, H2, and H3 are still in running order.