Humankind's efforts to tell time have helped drive the evolution of our technology and science throughout history. The need to gauge the divisions of the day and night led the ancient Egyptians, Greeks and Romans to create sundials, water clocks and other early chronometric tools. Western Europeans adopted these technologies, but by the 13th century, demand for a dependable timekeeping instrument led medieval artisans to invent the mechanical clock. Although this new device satisfied the requirements of monastic and urban communities, it was too inaccurate and unreliable for scientific application until the pendulum was employed to govern its operation. The precision timekeepers that were subsequently developed resolved the critical problem of finding a ship's position at sea and went on to play key roles in the industrial revolution and the advance of Western civilization.
Today highly accurate timekeeping instruments set the beat for most of our electronic devices. Nearly all computers, for example, contain a quartz-crystal clock to regulate their operation. Moreover, not only do time signals beamed down from Global Positioning System satellites calibrate the functions of precision navigation equipment, they do so as well for cellular telephones, instant stock-trading systems and nationwide power-distribution grids. So integral have these time-based technologies become to our day-to-day lives that we recognize our dependency on them only when they fail to work.
ACCORDING TO archaeological evidence, the Babylonians and Egyptians began to measure time at least 5,000 years ago, introducing calendars to organize and coordinate communal activities and public events, to schedule the shipment of goods and, in particular, to regulate cycles of planting and harvesting. They based their calendars on three natural cycles: the solar day, marked by the successive periods of light and darkness as the earth rotates on its axis; the lunar month, following the phases of the moon as it orbits the earth; and the solar year, defined by the changing seasons that accompany our planet's revolution around the sun.
Before the invention of artificial light, the moon had greater social impact. And, for those living near the equator in particular, its waxing and waning was more conspicuous than the passing of the seasons. Hence, the calendars developed at the lower latitudes were influenced more by the lunar cycle than by the solar year. In more northern climes, however, where seasonal agriculture was important, the solar year became more crucial. As the Roman Empire expanded northward, it organized its calendar for the most part around the solar year. Today's Gregorian calendar derives from the Babylonian, Egyptian, Jewish and Roman calendars.
The Egyptians formulated a civil calendar having 12 months of 30 days, with five days added to approximate the solar year. Each period of 10 days was marked by the appearance of special star groups (constellations) called decans. At the rise of the star Sirius just before sunrise, which occurred around the all-important annual flooding of the Nile, 12 decans could be seen spanning the heavens. The cosmic significance the Egyptians placed in the 12 decans led them to develop a system in which each interval of darkness (and later, each interval of daylight) was divided into a dozen equal parts. These periods became known as temporal hours because their duration varied according to the changing length of days and nights with the passing of the seasons. Summer hours were long, winter ones short; only at the spring and autumn equinoxes were the hours of daylight and darkness equal. Temporal hours, which were adopted by the Greeks and then the Romans (who spread them throughout Europe), remained in use for more than 2,500 years.
Inventors created sundials, which indicate time by the length or direction of the sun's shadow, to track temporal hours during the day. The sundial's nocturnal counterpart, the water clock, was designed to measure temporal hours at night. One of the first water clocks was a basin with a small hole near the bottom through which the water dripped out. The falling water level denoted the passing hour as it dipped below hour lines inscribed on the inner surface. Although these devices performed satisfactorily around the Mediterranean, they could not always be depended on in the cloudy and often freezing weather of northern Europe.
The Pulse of Time
THE EARLIEST RECORDED weight-driven mechanical clock was installed in 1283 at Dunstable Priory in Bedfordshire, England. That the Roman Catholic Church should have played a major role in the invention and development of clock technology is not surprising: the strict observance of prayer times by monastic orders occasioned the need for a more reliable instrument of time measurement. Further, the Church not only controlled education but also possessed the wherewithal to employ the most skillful craftsmen. Additionally, the growth of urban mercantile populations in Europe during the second half of the 13th century created demand for improved timekeeping devices. By 1300 artisans were building clocks for churches and cathedrals in France and Italy. Because the initial examples indicated the time by striking a bell (thereby alerting the surrounding community to its daily duties), the name for this new machine was adopted from the Latin word for bell, clocca.
The revolutionary aspect of this new timekeeper was neither the descending weight that provided its motive force nor the gear wheels (which had been around for at least 1,300 years) that transferred the power; it was the part called the escapement. This device controlled the wheels' rotation and transmitted the power required to maintain the motion of the oscillator, the part that regulated the speed at which the timekeeper operated [for an explanation of early clockworks, see box on pages 50 and 51]. The inventor of the clock escapement is unknown.
ALTHOUGH THE MECHANICAL CLOCK could be adjusted to maintain temporal hours, it was naturally suited to keeping equal ones. With uniform hours, however, arose the question of when to begin counting them, and so, in the early 14th century, a number of systems evolved. The schemes that divided the day into 24 equal parts varied according to the start of the count: Italian hours began at sunset, Babylonian hours at sunrise, astronomical hours at midday and great clock hours (used for some large public clocks in Germany) at midnight. Eventually these and competing systems were superseded by small clock, or French, hours, which split the day, as we currently do, into two 12-hour periods commencing at midnight.
During the 1580s clockmakers received commissions for timekeepers showing minutes and seconds, but their mechanisms were insufficiently accurate for these fractions to be included on dials until the 1660s, when the pendulum clock was developed. Minutes and seconds derive from the sexagesimal partitions of the degree introduced by Babylonian astronomers. The word minute has its origins in the Latin prima minuta, the first small division; second comes from secunda minuta, the second small division. The sectioning of the day into 24 hours and of hours and minutes into 60 parts became so well established in Western culture that all efforts to change this arrangement failed. The most notable attempt took place in revolutionary France in the 1790s, when the government adopted the decimal system. Although the French successfully introduced the meter, liter and other base-10 measures, the bid to break the day into 10 hours, each consisting of 100 minutes split into 100 seconds, lasted only 16 months.
FOR CENTURIES after the invention of the mechanical clock, the periodic tolling of the bell in the town church or clock tower was enough to demarcate the day for most people. But by the 15th century, a growing number of clocks were being made for domestic use. Those who could afford the luxury of owning a clock found it convenient to have one that could be moved from place to place. Innovators accomplished portability by replacing the weight with a coiled spring. The tension of a spring, however, is greater after it is wound. The contrivance that overcame this problem, known as a fusee (from fusus, the Latin term for spindle), was invented by an unknown mechanical genius probably between 1400 and 1450 [see illustration in box on page 50]. This cone-shaped device was connected by a cord to the barrel housing the spring: when the clock was wound, drawing the cord from the barrel onto the fusee, the diminishing diameter of the spiral of the fusee compensated for the increasing pull of the spring. Thus, the fusee equalized the force of the spring on the wheels of the timekeeper.
The importance of the fusee should not be underestimated: it made possible the development of the portable clock as well as the subsequent evolution of the pocket watch. Many high-grade, spring-driven timepieces, such as marine chronometers, continued to incorporate this device until after World War II.
Pendulums Get into the Swing
IN THE 16TH CENTURY Danish astronomer Tycho Brahe and his contemporaries tried to use clocks for scientific purposes, yet even the best ones were still too unreliable. Astronomers in particular needed a better tool for timing the transit of stars and thereby creating more accurate maps of the heavens. The pendulum proved to be the key to boosting the accuracy and dependability of timekeepers. Galileo Galilei, the Italian physicist and astronomer, and others before him experimented with pendulums, but a young Dutch astronomer and mathematician named Christiaan Huygens devised the first pendulum clock on Christmas Day in 1656. Huygens recognized the commercial as well as the scientific significance of his invention immediately, and within six months a local maker in the Hague had been granted a license to manufacture pendulum clocks.
Huygens saw that a pendulum traversing a circular arc completed small oscillations faster than large ones. Therefore, any variation in the extent of the pendulum's swing would cause the clock to gain or lose time. Realizing that maintaining a constant amplitude (amount of travel) from swing to swing was impossible, Huygens devised a pendulum suspension that caused the bob to move in a cycloid-shaped arc rather than a circular one. This enabled it to oscillate in the same time regardless of its amplitude [see illustration in box on next page]. Pendulum clocks were about 100 times as accurate as their predecessors, reducing a typical gain or loss of 15 minutes a day to about a minute a week. News of the invention spread rapidly, and by 1660 English and French artisans were developing their own versions of this new timekeeper.
The advent of the pendulum not only heightened demand for clocks but also resulted in their development as furniture. National styles soon began to emerge: English makers designed the case to fit around the clock movement; in contrast, the French placed greater emphasis on the shape and decoration of the case. Huygens, however, had little interest in these fashions, devoting much of his time to improving the device both for astronomical use and for solving the problem of finding longitude at sea.
IN 1675 HUYGENS devised his next major improvement, the spiral balance spring. Just as gravity controls the swinging oscillation of a pendulum in clocks, this spring regulates the rotary oscillation of a balance wheel in portable timepieces. A balance wheel is a finely balanced disk that rotates fully one way and then the other, repeating the cycle over and over [see illustration in box at left]. The spiral balance spring revolutionized the accuracy of watches, enabling them to keep time to within a minute a day. This advance sparked an almost immediate rise in the market for watches, which were now no longer typically worn on a chain around the neck but were carried in a pocket, a wholly new fashion in clothing.
At about the same time, Huygens heard of an important English invention. The anchor escapement, unlike the verge escapement he had been using in his pendulum clocks, allowed the pendulum to swing in such a small arc that maintaining a cycloidal pathway became unnecessary. Moreover, this escapement made practical the use of a long, seconds-beating pendulum and thus led to the development of a new case design. The longcase clock, commonly known since 1876 as the grandfather clock (after a song by American Henry Clay Work), began to emerge as one of the most popular English styles. Longcase clocks with anchor escapements and long pendulums can keep time to within a few seconds a week. The celebrated English clockmaker Thomas Tompion and his successor, George Graham, later modified the anchor escapement to operate without recoil. This enhanced design, called the deadbeat escapement, became the most widespread type used in precision timekeeping for the next 150 years.
Solving the Longitude Problem
WHEN THE ROYAL OBSERVATORY at Greenwich, England, was founded in 1675, part of its charter was to find the so-much-desired longitude of places. The first Astronomer Royal, John Flamsteed, used clocks fitted with anchor escapements to time the exact moments that stars crossed the celestial meridian, an imaginary line that connects the poles of the celestial sphere and defines the due-south point in the night sky. This allowed him to gather more accurate information on star positions than had hitherto been possible by making angular measurements with sextants or quadrants alone.
Although navigators could find their latitude (their position north or south of the equator) at sea by gauging the altitude of the sun or the polestar, the heavens did not provide such a straightforward solution for finding longitude. Storms and currents often confounded attempts to keep track of distance and direction traveled across oceans. The resulting navigational errors cost seafaring nations dearly, not only in prolonged voyages but also in loss of lives, ships and cargo. The severity of this predicament was brought home to the British government in 1707, when an admiral of the fleet and more than 1,600 sailors perished in the wrecks of four Royal Navy ships off the coast of the Scilly Isles. Thus, in 1714, through an act of Parliament, Britain offered substantial prizes for practical solutions to finding longitude at sea. The largest prize, 20,000 (which is equivalent to about 12 million today), would be given to the inventor of an instrument that could determine a ship's longitude to within half a degree, or 30 nautical miles, when reckoned at the end of a voyage to a port in the West Indies, whose longitude could be accurately ascertained using proved land-based methods.
The great reward attracted a deluge of harebrained schemes. Hence, the Board of Longitude, the committee appointed to review promising ideas, held no meetings for more than 20 years. Two approaches, however, had long been known to be theoretically sound. The first, called the lunar-distance method, involved precise observations of the moon's position in relation to the stars to determine the time at a reference point from which longitude could be measured; the other required a very accurate clock to make the same determination. Because the earth rotates every 24 hours, or 15 degrees in an hour, a two-hour time difference represents a 30-degree difference in longitude. The seemingly overwhelming obstacles to keeping accurate time at sea--among them the often violent motions of ships, extreme changes in temperature, and variations in gravity at different latitudes--led English physicist Isaac Newton and his followers to believe that the lunar-distance method, though problematic, was the only viable solution.
Newton was wrong, however. In 1737 the board finally met for the first time to discuss the work of a most unlikely candidate, a Yorkshire carpenter named John Harrison. Harrison's bulky longitude timekeeper had been used on a voyage to Lisbon and on the return trip had proved its worth by correcting the navigator's dead reckoning of the ship's longitude by 68 miles. Its maker, however, was dissatisfied. Instead of asking the board for a West Indies trial, he requested and received financial support to construct an improved machine. After two years of work, still displeased with his second effort, Harrison embarked on a third, laboring on it for 19 years. But by the time it was ready for testing, he realized that his fourth marine timekeeper, a five-inch-diameter watch he had been developing simultaneously, was better. On a voyage to Jamaica in 1761, Harrison's oversize watch performed well enough to win the prize, but the board refused to give him his due without further proof. A second sea trial in 1764 confirmed his success. Harrison was reluctantly granted 10,000. Only when King George III intervened in 1773 did he receive the remaining prize money. Harrison's breakthrough inspired further developments. By 1790 the marine chronometer was so refined that its fundamental design never needed to be changed.
AT THE TURN of the 19th century, clocks and watches were relatively accurate, but they remained expensive. Recognizing the potential market for a low-cost timekeeper, two investors in Waterbury, Conn., took action. In 1807 they gave Eli Terry, a clockmaker in nearby Plymouth, a three-year contract to manufacture 4,000 longcase clock movements from wood. A substantial down payment made it possible for Terry to devote the first year to fabricating machinery for mass production. By manufacturing interchangeable parts, he completed the work within the terms of the contract.
A few years later Terry designed a wooden-movement shelf clock using the same volume-production techniques. Unlike the longcase design, which required the buyer to purchase a case separately, Terry's shelf clock was completely self-contained. The customer needed only to place it on a level shelf and wind it up. For the relatively modest sum of 15, many average people could now afford a clock. This achievement led to the establishment of what was to become the renowned Connecticut clockmaking industry.
Before the expansion of railroads in the 19th century, towns in the U.S. and Europe used the sun to determine local time. For example, because noon occurs in Boston about three minutes before it does in Worcester, Mass., Boston's clocks were set about three minutes ahead of those in Worcester. The expanding railroad network, however, needed a uniform time standard for all the stations along the line. Astronomical observatories began to distribute the precise time to the railroad companies by telegraph. The first public time service, introduced in 1851, was based on clock beats wired from the Harvard College Observatory in Cambridge, Mass. The Royal Observatory introduced its time service the next year, creating a single standard time for Great Britain.
The U.S. established four time zones in 1883. By the next year the governments of all nations had recognized the benefits of a worldwide standard of time for navigation and trade. At the 1884 International Meridian Conference in Washington, D.C., the globe was divided into 24 time zones. Signatories chose the Royal Observatory as the prime meridian (zero degrees longitude, the line from which all other longitudes are measured) in part because two thirds of the world's shipping already used Greenwich time for navigation.
Watches for the Masses
MANY CLOCKMAKERS of this era realized that the market for watches would far exceed that for clocks if production costs could be reduced. The problem of mass-fabricating interchangeable parts for watches, however, was considerably more complicated because the precision demanded in making the necessary miniaturized components was so much greater. Although improvements in quantity manufacture had been instituted in Europe since the late 18th century, European watchmakers' fears of saturating the market and threatening their workers' jobs by abandoning traditional practices stifled most thoughts of introducing machinery for the production of interchangeable watch parts.
Disturbed that American watchmakers seemed unable to compete with their counterparts in Europe, which controlled the market in the late 1840s, a watchmaker in Maine named Aaron L. Dennison met with Edward Howard, the operator of a clock factory in Roxbury, Mass., to discuss mass-production methods for watches. Howard and his partner gave Dennison space to experiment and develop machinery for the project. By the fall of 1852, 20 watches had been completed under Dennison's supervision. His workmen finished 100 watches by the following spring, and 1,000 more were produced a year later. By that time the manufacturing facilities in Roxbury were proving too small, so the newly named Boston Watch Company moved to Waltham, Mass., where by the end of 1854 it was assembling 36 watches a week.
The American Waltham Watch Company, as it eventually became known, benefited greatly from a huge demand for watches during the Civil War, when Union Army forces used them to synchronize operations. Improvements in fabrication techniques further boosted output and cut prices. Meanwhile other U.S. companies formed in the hope of capturing part of the burgeoning trade. The Swiss, who had previously dominated the industry, grew concerned when their exports plummeted in the 1870s. The investigator they sent to Massachusetts discovered that not only was productivity higher at the Waltham factory but production costs were less. Even some of the lower-grade American watches could be expected to keep reasonably good time. The watch was at last a commodity accessible to the masses.
Because women had worn bracelet watches in the 19th century, wristwatches were long considered feminine accoutrements. During World War I, however, the pocket watch was modified so that it could be strapped to the wrist, where it could be viewed more readily on the battlefield. With the help of a substantial marketing campaign, the masculine fashion for wristwatches caught on after the war. Self-winding mechanical wristwatches made their appearance during the 1920s.
AT THE END of the 19th century, Siegmund Riefler of Munich developed a radical new design of regulator--a highly accurate timekeeper that served as a standard for controlling others. Housed in a partial vacuum to minimize the effects of barometric pressure and equipped with a pendulum largely unaffected by temperature variations, Riefler's regulators attained an accuracy of a tenth of a second a day and were thus adopted by nearly every astronomical observatory.
Further progress came several decades later, when English railroad engineer William H. Shortt designed a so-called free pendulum clock that reputedly kept time to within about a second a year. Shortt's system incorporated two pendulum clocks, one a master (housed in an evacuated tank) and the other a slave (which contained the time dials). Every 30 seconds the slave clock gave an electromagnetic impulse to, and was in turn regulated by, the master clock pendulum, which was thus nearly free from mechanical disturbances.
Although Shortt clocks began to displace Rieflers as observatory regulators during the 1920s, their superiority was short-lived. In 1928 Warren A. Marrison, an engineer at Bell Laboratories in New York, discovered an extremely uniform and reliable frequency source that was as revolutionary for timekeeping as the pendulum had been 272 years earlier. Developed originally for use in radio broadcasting, the quartz crystal vibrates at a highly regular rate when excited by an electric current [see illustration in box on opposite page]. The first quartz clocks installed at the Royal Observatory in 1939 varied by only two thousandths of a second a day. By the end of World War II, this accuracy had improved to the equivalent of a second every 30 years.
Quartz-crystal technology did not remain the premier frequency standard for long either, however. By 1948 Harold Lyons and his associates at the National Bureau of Standards in Washington, D.C., had based the first atomic clock on a far more precise and stable source of timekeeping; an atom's natural resonant frequency, the periodic oscillation between two of its energy states [see illustration in box on opposite page]. Subsequent experiments in both the U.S. and England in the 1950s led to the development of the cesium-beam atomic clock. Today the averaged times of cesium clocks in various parts of the world provide the standard frequency for Coordinated Universal Time, which has an accuracy of better than one nanosecond a day.
Up to the mid-20th century, the sidereal day, the period of the earth's rotation on its axis in relation to the stars, was used to determine standard time. This practice had been retained even though it had been suspected since the late 18th century that our planet's axial rotation was not entirely constant. The rise of cesium clocks capable of measuring discrepancies in the earth's spin, however, meant that a change was necessary. A new definition of the second, based on the resonant frequency of the cesium atom, was adopted as the new standard unit of time in 1967.
The precise measurement of time is of such fundamental importance to science that the search for even greater accuracy continues. Current and coming generations of atomic clocks, such as the hydrogen maser (a frequency oscillator), the cesium fountain and, in particular, the optical clock (both frequency discriminators), are expected to deliver an accuracy (more precisely, a stability) of 100 femtoseconds (100 quadrillionths of a second) over a day [see Ultimate Clocks, by W. Wayt Gibbs, on page 56].
Although our ability to measure time will surely improve in the future, nothing will change the fact that it is the one thing of which we will never have enough.
William J. H. Andrewes is a museum consultant and maker of precision sundials who has specialized in the history of time measurement for more than 30 years. He has worked at several scholarly institutions, including Harvard University. In addition to writing articles for popular and academic journals, Andrewes edited The Quest for Longitude and co-wrote The Illustrated Longitude with Dava Sobel. His past exhibitions include The Art of the Timekeeper at the Frick Collection in New York City.