Accurate Mechanical Clocks
In 1656, Christiaan Huygens, a Dutch scientist, made the first pendulum clock, regulated by a mechanism with a "natural" period of oscillation. (Galileo Galilei is credited with inventing the pendulum-clock concept, and he studied the motion of the pendulum as early as 1582. He even sketched out a design for a pendulum clock, but he never actually constructed one before his death in 1642.) Huygens' early pendulum clock had an error of less than 1 minute a day, the first time such accuracy had been achieved. His later refinements reduced his clock's error to less than 10 seconds a day.
Around 1675, Huygens developed the balance wheel and spring assembly, still found in some of today's wristwatches. This improvement allowed portable 17th century watches to keep time to 10 minutes a day. And in London in 1671, William Clement began building clocks with the new "anchor" or "recoil" escapement, a substantial improvement over the verge because it interferes less with the motion of the pendulum.
In 1721, George Graham improved the pendulum clock's accuracy to 1 second per day by compensating for changes in the pendulum's length due to temperature variations. John Harrison, a carpenter and self-taught clock-maker, refined Graham's temperature compensation techniques and developed new methods for reducing friction. By 1761, he had built a marine chronometer with a spring and balance wheel escapement that won the British government's 1714 prize (worth more than $10,000,000 in today's currency) for a means of determining longitude to within one-half degree after a voyage to the West Indies. It kept time on board a rolling ship to about one-fifth of a second a day, nearly as well as a pendulum clock could do on land, and 10 times better than required to win the prize.
Over the next century, refinements led in 1889 to Siegmund Riefler's clock with a nearly free pendulum, which attained an accuracy of a hundredth of a second a day and became the standard in many astronomical observatories. A true free-pendulum principle was introduced by R.J. Rudd about 1898, stimulating development of several free-pendulum clocks. One of the most famous, the W.H. Shortt clock, was demonstrated in 1921. The Shortt clock almost immediately replaced Riefler's clock as a supreme timekeeper in many observatories. This clock contained two pendulums, one a slave and the other a master. The slave pendulum gave the master pendulum the gentle pushes needed to maintain its motion, and also drove the clock's hands. This allowed the master pendulum to remain free from mechanical tasks that would disturb its regularity.
UPDATE:
Chaucer's Tretise of the Astrelabie, perhaps the first scientific treatise written in the English language, deals with many uses of the instrument including finding the time. [4] He first explains how to calculate the angular length of "the arch of the day, that some folk callen the day artificiall, fro sonne arisyng tyl it go to reste" (Conclusioun 7) and then moves onto "the day vulgar, that is to seyn fro spryng of the day unto verrey night" (Conclusioun 9). To work out the length of the relevant hours in degrees, one divides the "arch of the day" by twelve. Then he moves on to his favoured method of counting hours:
Conclusioun 11: To know the quantite of houres equales.
The quantite of houres equales, that is to seyn the houres of the clokke, ben departed [divided] by 15 degrees alredy in the bordure of thin Astrelaby, as wel by night as by day, generaly for evere. What nedith more declaracioun?
Wherfore whan thou list to knowe hou many houres of the clokke ben passed, or eny part of eny of these houres that ben passed, or ellis how many houres or parties of houres ben to come fro such a tyme by day or by night, know the degre of thy sonne, and ley thy label [alidade] on it. Turne thy ryet [rete] aboute jointly with thy label, and with the poynt of it rekne in the bordure fro the sonne arise unto that same place there thou desirest, by day as by night.