Precision Regulators Astronomy was always in the forefront of the demand for more precise and accurate timekeepers: Observatory "regulators" were the most discriminating clocks available, the standard for all others. In fact, the precision regulators made in the late 1800's and early 1900's were the most precise mechanical clocks ever developed. The following section discusses what constituted precision in earlier times, and outlines the developments that led up to the regulators of the late 1800s. The History Leading Up to Precision Clocks The earliest mechanical clocks, dating from the late thirteenth century, used verge escapements and had foliot instead of pendulums. They typically ran for a day, giving their owner a chance to reset them - since they were only accurate to about fifteen minutes per day. It was not until the 1400's that better metallurgy and improved gear cutting techniques justified the addition of minute hands on tower and chamber clocks. It wasn't until around 1670 and the development of the pendulum and the anchor escapement that it was possible to keep time to within 10 to 15 seconds a day. It was about this time that commercial navigation provided the impetus to determine the longitude (east-west location) of ships at sea. In the ideal the search for longitude was perceived as a project of intellectual and humanitarian concern transcending national interests. By the early 1700's it was a competition between countries. France was more populous and richer, hence a stronger land power, than England. But England had the advantage on the oceans. For both countries the eighteenth century was a period of rapid growth of trade in and competition for what were known as the colonial wares: Sugar, coffee, tea, and tobacco. These commodities were the source of fabulous fortunes and windfall revenues. Their profitability depended on the efficiency and productivity of their merchant marines. The Longitude Competition and Graham's Deadbeat Escapement This is the setting for the British Act of Parliament in 1713-14 that created the longitude competition, with a prize of 20,000 pounds for finding the longitude to within 1 degree. Holland and France had similar competitions. Harrison is credited with making the winning timepiece in 1759. His H4 timepiece, in the form of a large pocket watch, ultimately earned him the 20,000 pound prize. I will not go into the relevant features of this watch since it doesn't impact the development of precision regulators. The next important development (for those of us interested in precision regulators) came in 1715 when George Graham used a dead-beat escapement to achieve an accuracy of a second a day. This escapement became the preferred escapement for scientific observation for the next 200 years and was ultimately used in the precision regulators of the late 1800's. It is also the standard escapement for Vienna regulators. The Accuracy of Precision Regulators: Escapement + Pendulum In my mind, precision regulators rely on two primary components for their accuracy - an escapement and a pendulum. As shown above, Graham solved the escapement problem in 1715. What remained was to develop pendulums that are not impacted by changes in temperature. Variation in the length of pendulum rods must be eliminated to achieve precise timekeeping. A clock with a steel pendulum rod which runs correctly at 60 degrees F would run slow by 5 seconds a day at 80 degrees F. While a properly sealed (against the impact of changes in relative humidity) wooden pendulum rod is a big improvement over a steel, brass, or other common metal rods, it did not provide the accuracy demanded in the early 1700's. The first notable increase in accuracy came when George Graham demonstrated in 1715 a method of compensating for the change in length of a steel rod through the use of mercury. The lengthening of a steel pendulum rod with increased temperature lowers the pendulum's center or gravity or "vibrating point". However, in a mercury-compensating pendulum, the mercury, as it fills a larger part of the glass container due to its expansion, lifts its center of gravity, thereby compensating for the lowering due to the lengthening of the pendulum rod. 1756 and the mid-1800's Next came John Harrison in 1756 with his "gridiron" pendulum. In it he used dissimilar metals, such as brass and steel, or zinc and steel to compensate for changes in the length of the rod. A simple explanation of a gridiron pendulum is that while a steel rod expands downward, a compensating brass rod expands upward, negating the expansion of the steel rod. While it is a bit more complex than this (as you can tell by looking at a true gridiron pendulum) that is the basic concept. It wasn't until the end of the 1800's that the next significant improvement was made to pendulum designs. In 1899 Charles E. Gallium is credited with inventing Invar, a nickel-steel alloy that has a very low temperature coefficient of expansion. Riefler, in his effort to make extremely accurate precision regulators, used Invar to produce clocks that are accurate to a tenth of a second a day, or 3 seconds a month. Late 1800's/Early 1900's The quest for precision in the late 1800's included the use of vacuum containers and electrical impulsing to achieve accuracy in the range of 1/100 to 1/1000 of a second a day. The final pendulum development for precision regulators occurred in the early 1900s. To quote Karl Satori from his 1913 patent: "Producing the pendulum rod of clock pendulums and comparable items out of quartz with its low coefficient of thermal expansion has the advantage of obtaining compensation for thermal expansion in a simple and absolutely perfect manner." Finally a material that did not require extensive compensation! While it was less than 20 years later that a completely different application of quartz rendered Satori's pendulum redundant, it is none the less a fascinating foot-note in the pursuit of precision. To my knowledge Satori produced pendulums with three different bobs. Three Satori pendulums from private collections are shown below. 1920's The accuracy of the pendulum-regulated precision regulator was rendered obsolete in the late 1920's with the development of the first quartz clock. Early ones were accurate to two thousandths of a second a day (pretty much the limit of the very best of the precision regulators) though later ones achieved 1 second in 30 years. And, in due course, quartz was superseded by atomic clocks that, in 1955 achieved the accuracy of 1 second in 1,000 years. By the end of the twentieth century atomic clocks have achieved the almost unnerving accuracy of 1 second in 20 million years.