Marine Chronometer

A marine timekeeper about 35 by 25 by 25 centimetres in its gimballed outer box, of wood, steel, brass and glass, built to keep time at sea accurately enough to fix a ship's longitude by comparison with a known reference time. It is not a large or visually imposing object. Its importance is entirely inside the movement.

Inside that movement Le Roy combined, for the first time in one working machine, a detached detent escapement of his own design, a balance compensated for changes of temperature by a bimetallic mechanism, and a balance spring whose length he had determined, by experiment, to bring its coils to isochronism. Each of the three had been attempted separately by other makers. Le Roy built the machine that used all three together, with a going barrel in place of the fusee-and-chain drive common in English work, four wheels and their pinions, and no jewelling at all, since the process was not yet in use on the Continent. The pivots turn in plain brass holes.

Pierre Le Roy was born in Paris in 1717, the eldest son of Julien Le Roy, himself a celebrated clockmaker who held the post of horloger du roi with lodgings in the royal galleries and who died in 1754, at which point Pierre succeeded him in the post. Three of Pierre's brothers went into other branches of the Enlightenment sciences: Jean-Baptiste Le Roy became a physicist and a fellow of the Royal Society, Julien-David Le Roy an architect and antiquarian, and Charles Le Roy a physician. The family produced clockmakers for four generations before Pierre and remained active in the trade after him; a short-lived manufactory of marine timekeepers was founded under royal patronage in 1787, two years after Pierre's death, with the cooperation of several of his relatives.

Le Roy communicated his invention of a detached escapement to the Academie des Sciences as early as 1748, and by his own account it was not entirely without precedent: he acknowledged that the elder Dutertre had shown a similar device years before, and that the germ of the idea existed earlier still. He continued to refine the escapement, the compensation balance and the isochronous spring through the 1750s and early 1760s, presenting a memoir on his balance-spring findings to the Academie in 1763. He is recorded receiving the English writer Samuel Johnson at his workshop in Paris in 1775; Johnson's travel notes call him 'a man of character in his business... A decent man,' who showed his visitor a small clock made to find the longitude.

Le Roy's other lifelong subject was Ferdinand Berthoud, a Swiss clockmaker working in Paris who pursued the same problem by a different route: steady, incremental trial and error across a long series of machines, rather than Le Roy's single scientifically argued design. The two men were bitter and public rivals. Both claimed priority on the isochronous balance spring; Le Roy published his findings first, in 1763, while Berthoud's first marine timekeeper, described in print in 1763, was by Le Roy's own account still built on the crude lines of Harrison's early machines, the same machines Le Roy had personally examined in London in 1738.

Le Roy pressed his case in print. In 1768 he published a comparison of his own work against Harrison's fourth timekeeper, titled Expose succinct des travaux de MM. Harrison et Le Roy, closing it with a tribute of respect to Harrison as a venerable old man and a wish that his own labour might one day receive similar recognition. It never did in his lifetime. Berthoud, for all that his machines never beat Le Roy's in a single Academie trial, built and sold vastly more of them, more than fifty timekeepers across some eighty voyages by one tally, and in 1770 received the coveted title Horloger de la Marine, the naval clockmaking appointment Le Roy never held. Berthoud's greater output and greater public reputation left Le Roy, by most accounts, disillusioned; he appears to have scaled back further development work after the early 1770s. The Conservatoire national des arts et metiers today holds work by both men, including, by one nineteenth-century count, thirteen of Berthoud's own Horloges Marines.

The Academie royale des sciences administered a bequest from the Comte de Meslay, established in 1714, whose smaller prize was reserved for the best method of finding longitude at sea. Le Roy entered his 1766 machine, along with a duplicate of the same design, for the prize; the Academie postponed its decision until the machines could be tested at sea. Francois-Cesar Le Tellier, Marquis de Courtanvaux, a member of the Academie, funded the construction of a corvette, the Aurore, purely for that purpose, designed by the naval architect Nicolas Ozanne and built at Le Havre.

When the trial came, in the spring and summer of 1767, Le Roy's two machines were the only ones actually produced for testing. Other entrants withdrew: Romilly's machine was accidentally damaged during a preliminary shore test by a member of the Academie, and Tavernier withdrew his own entry in sympathy. Le Roy travelled to Le Havre with his machines in a post-chaise, and the shocks of the road journey broke the fine wire supporting the balance of his first machine; he had it repaired overnight at his inn and reached the ship with both instruments running. The Aurore then made a roughly three-and-a-half-month cruise along the coasts of France, Flanders and Holland, putting into port repeatedly so the watches' rates could be checked against astronomical observation.

Courtanvaux's own report to the Academie, as later translated by Rupert Gould, records the first watch's error at 38 seconds across 46 days of comparison, and the second, more recently completed watch's error at only 7.5 seconds across the same period. A separate figure circulating in later secondary accounts gives the first watch an error of 4 minutes 52 seconds over a 52-day outbound leg and 51 seconds on the shorter return leg. This record has not been able to reconcile the two figures from the sources consulted and presents both rather than choosing between them.

The Academie tested Le Roy's machines a second time in 1769, aboard the frigate L'Enjouee, on a longer voyage from Le Havre to Newfoundland, on to Cadiz and back to Brest, lasting some twenty-four weeks. The results held to the same standard as the first trial, and the Academie awarded Le Roy its double Meslay prize, covering both the timekeeper and his written memoir describing it. The Academie's own report was not uncomplicated praise: it noted that in one trial the watch had accelerated suddenly by eleven or twelve seconds a day even on dry land, and concluded plainly that the desired degree of perfection had not yet been reached.

A further Academie prize was offered in 1771, tested aboard the frigate La Flore against a weight-driven machine by Arsandeaux and a pendulum clock by Biesta. Le Roy's machines won again, though less decisively; during the voyage the ship struck Wilmington rock off Antigua, and the resulting shock broke a thermometer on his first machine, disabling it as a timekeeper for the rest of the trial. After this, Le Roy appears to have judged his design essentially complete. He wrote that he had tried, almost daily since finishing his last marine watch, to find some advantageous change to make, and concluded that nothing remained but to execute the machines well. He later began work on a further, unrecorded machine in his final years, of which no details survive. He died on 25 or 26 August 1785; sources consulted give both dates and this record does not resolve the discrepancy.

The popular account of who solved the longitude problem centres John Harrison, most influentially through Dava Sobel's bestselling 1995 book Longitude, which tells Harrison's story as a lone struggle against an obstructive scientific establishment, resolved by his fourth timekeeper, H4, and its 1761-64 trials. That story is accurate as far as it goes. It is not the whole picture of how the marine chronometer, as a type of instrument, actually came to exist.

Harrison solved the specific engineering problem in front of him by building an extraordinarily dense, extraordinarily skilled watch: a remontoire, a maintaining-power mechanism, an automatic regulator, diamond pallets, unusually high tooth counts in its wheels and pinions, and lavish jewelling throughout. No other maker duplicated H4's mechanism afterward, in France or in England; it worked, once, by mastery rather than by a method others could repeat. Rupert Gould, examining both machines for his 1923 history, put the comparison in terms he did not soften: 'Harrison built a wonderful house on the sand; but Le Roy dug down to the rock.' Le Roy, working from an entirely different premise and without any knowledge of H4's internal mechanism, which was not published until 1767, a year after Le Roy presented his own machine to Louis XV, isolated the sources of error instead of compensating for them piece by piece: a detached escapement that barely touched the balance, a balance whose temperature response was built into its own construction rather than adjusted afterward, and a spring whose isochronism came from getting its length right in the first place. Gould's verdict, after describing the mechanism in detail, was unequivocal: 'the credit of having designed and constructed the first modern chronometer is entirely his, and his alone,' and, in Gould's closing assessment of Le Roy's career, 'he stands alone, the father of the chronometer as we know it.'

The historical argument for this is not French chauvinism; it is a question of lineage. The English makers who actually built the trade that followed, John Arnold and Thomas Earnshaw, did not copy Harrison's remontoire-and-jewelling approach. They refined Le Roy's detached detent escapement into the simpler spring-detent form that became, in Earnshaw's hands, the standard chronometer escapement used for the next century and a half; his 1943 obituarist among French hydrographic writers found the subject of Le Roy's originality still worth a dedicated review nearly two centuries after the fact. Harrison's H4 is rightly celebrated as the machine that proved the longitude problem could be solved by a clock at all. It is not the machine whose internal logic the trade then built on. That machine is this one.

None of this diminishes what Harrison achieved first: a working proof, years before Le Roy's machine existed, that a sufficiently accurate timekeeper could survive a transatlantic voyage and hold its rate. Le Roy had the advantage of building after that proof existed in the world, even though he could not see how Harrison had done it. The Academie des Sciences' own prize structure, funded since 1714, predates the British Parliament's 1714 Longitude Act only by coincidence of date, not by priority of achievement, and it was Harrison, not Le Roy or Berthoud, who first collected a comparable national reward, roughly 20,000 pounds sterling from the British government, against the far smaller medal Le Roy received in France for what French commentators, including a later reviewer writing for M. Gros, called a comparatively greater sacrifice of fortune and twenty years of his life. Crediting Le Roy with the surviving mechanism is not the same claim as crediting him with being first; both claims are documented, and they are not the same claim.