John Harrison - The First Three Marine Timekeepers

The First Three Marine Timekeepers

The English clockmaker Henry Sully had already invented a marine clock to determine longitude accurately, this was in the form of a small clock with a large balance wheel vertically mounted on friction rollers and impulsed by a frictional rest Debaufre type escapement. Very unconventionally, the balance oscillations were controlled not by a spring but by a weight at the end of a pivoted horizontal lever attached to the balance by a cord, a solution which rather neatly avoided temperature error, a problem which affects steel balance springs. However Sully's clock was not successful as a marine timekeeper. It only kept accurate time in calm weather since the balance oscillations could be affected by the pitching and rolling on board ship. However Sully's clocks are of importance as they were amongst the first serious attempts to find the Longitude by using a clock. It is perhaps not coincidental that Harrison's machines which though much larger, are of similar layout. For example his third sea clock (H3), this also has a vertically mounted balance wheel but linked to another the same size, thus eliminating the problems arising from pitching and rolling motions.

He presented a first Montre de la Mer in 1716 to the French Académie des Sciences. and in 1726 published Une Horloge inventée et executée par M. Sulli.

In 1730 Harrison created a description and drawings for a proposed marine clock to compete for the Longitude Prize and went to London seeking financial assistance. He presented his ideas to Edmond Halley, the Astronomer Royal. Halley referred him to George Graham, the country's foremost clockmaker. He must have been impressed by Harrison, for Graham personally loaned Harrison money to build a model of his 'Sea clock' as Harrison called it.

This clock was in essence, an attempt to make a sea going version of his wooden pendulum clocks which had performed so exceptionally well. Thus the use of wooden wheels, roller pinions and a version of the 'grasshopper' escapement. Instead of a pendulum he used two dumbbell balances linked together,

It took Harrison five years to build Harrison Number One or H1. He demonstrated it to members of the Royal Society who spoke on his behalf to the Board of Longitude. The clock was the first proposal that the Board considered to be worthy of a sea trial. In 1736, Harrison sailed to Lisbon on HMS Centurion and returned on HMS Orford. On their return, both the captain and the sailing master of the Orford praised the design. The master noted that his own calculations had placed the ship sixty miles east of its true landfall which had been correctly predicted by Harrison using H1.

This was not the transatlantic voyage demanded by the Board of Longitude, but the Board was impressed enough to grant Harrison £500 for further development. Harrison moved on to develop H2, a more compact and rugged version. In 1741, after three years of building and two of on-land testing, H2 was ready, but by then Britain was at war with Spain in the War of Austrian Succession and the mechanism was deemed too important to risk falling into Spanish hands. In any event, Harrison suddenly abandoned all work on this second machine when he discovered a serious design flaw in the concept of the bar balances. This was that the period of oscillation could be affected by the pitching action on board a ship, this led to his adoption of circular balances in the Third Sea Clock (H3). The Board granted him another £500, and while waiting for the war to end, he proceeded to work on H3.

Harrison spent seventeen years working on this third 'sea clock' but despite every effort it seems not to have performed exactly as he would have wished. Despite this, it had proved a very valuable experiment. Certainly in this machine Harrison left the world two enduring legacies – the bimetallic strip and the caged roller bearing. The failure of the sea clocks 1,2 and 3 were due mainly to the fact that their balances though large, did not vibrate quickly enough to confer the property of stability on the timekeeping. Around 1750 Harrison had also come to this conclusion and abandoned the idea of the 'Sea clock' as a timekeeper, realizing that a watch sized timekeeper would be more successful as it could incorporate a balance which though smaller, oscillated at a much higher speed. A watch would also be more practicable, and indeed this was a stipulation required by the Longitude Act of 1714.