AI-generated from sources

The Earth as a Bad Clock, How the Measure of Time Stopped Following the Turning World

Main authors

Martin P. Nilsson Robert S. Ball George Howard Darwin

For most of human history, telling the time meant watching the world turn. The sun crossing the meridian, a star returning to the same point of the sky: the day was a fact of nature before it became a unit of measure. Yet the observers who knew the heavens best had long understood that this fact was less simple than it looked. As early as 1666, the Philosophical Transactions of the Royal Society warned that the natural day, counted from noon to noon, was not constant, because the Earth's annual journey around the sun adds a varying increment to each rotation ^[1]. The common confidence in equal days rested on a regularity that astronomy had already dissolved.

From this first crack, the measure of time detached itself from the sky in two distinct movements. In the first, clocks abandoned the visible sun for an average, the mean day, which no single noon ever displays ^{[2, 3]}. In the second, slower and more unsettling, astronomers found that the average itself rested on a faltering motion: the Earth's rotation, which had become the ultimate reference of all timekeeping, was being braked by the tides ^{[4, 5]}. The stakes went beyond precision. Once the turning of the world was shown to be irregular, the relation between the standard and the thing it measured had to be reversed, and the question of what a clock actually measures received an answer that none of its early makers had anticipated.

From the Visible Sun to an Averaged Rotation

The inequality of solar days remained a scholarly curiosity for as long as no instrument could make it felt. The astronomer George C. Comstock locates the turning point precisely: when clocks and watches came into general use, the solar time of the sundial ceased to be usable, since no well-regulated mechanism could follow a sun that seems to move faster among the stars in winter than in summer, as the Earth speeds up and slows down along its orbit . The irregularity had not changed; the instrument had. Mechanical regularity turned a known astronomical nuance into a daily contradiction between the dial in the garden and the clock in the house. What had been, for the authors of 1666, a refinement reserved for astronomers became, more than two centuries later, a problem of social coordination .

The writers who recorded the outcome did not agree on how to describe it. A contributor to Popular Science Monthly in 1885 presents the abandonment of solar time as a necessity, and the adoption of an arbitrary standard as the triumph of a principle he states in utilitarian terms, the greatest good of the greatest number, extended to what he calls the civilized world . T. Geoffrey W. Henslow, the author of a book on sundials, describes the same Greenwich mean time as purely artificial and imaginary, matching no natural time at all, while conceding that it serves commerce very well ^[6]. Between these two readings runs the central tension of the first break: mean time could be seen either as rational progress over a capricious nature or as a convenient fiction substituted for it. Both authors nonetheless agree on the essential point, which is that civil time no longer coincided with any phenomenon a person could observe.

Martin P. Nilsson, in his study of primitive time-reckoning, offers a long-range frame for this move. In his account, genuine time-reckoning is born precisely when the calendar is detached from the concrete phenomena of the heavens and acquires a numerical character; convenience of reckoning, he argues, matters more in practice than faithful agreement with celestial motions that are incommensurable with one another ^[7]. Clinging to a natural basis, by contrast, long obstructed any rational system, since the day has no fixed natural limits and sunrise and sunset shift from one day to the next ^[8]. Read in this light, mean time appears less as a modern rupture than as the latest step in a very old tendency. The analogy has limits, since Nilsson writes of calendars and seasons rather than observatories, but it captures the logic of the substitution: the count is preferred to the spectacle.

Yet the first break left the sky less completely than its critics suggested. Harry Chase Brearley, tracing time-telling through the ages, stresses that modern timekeeping still refers all its measures to the apparent revolution of the heavens, which is in fact the motion of the Earth itself, so that the observer at his telescope corrects every small timepiece by the great master clock of the universe ^[9]. An account of observatory routine from 1886 confirms it: the astronomer deduces the errors of his clock from the transits of many stars, and relies on the clock's known rate only when cloudy weather deprives him of the sky ^[10]. William Leitch, in a work of natural theology published in 1867, went further and read the solar system itself as a clock whose disturbances cancel one another, like a pendulum built of metals that offset each other's expansion, so as to yield what he presents as invariable motion ^[11]. What the first break abandoned, then, was the visible sun, not the rotation: averaged, the turning of the Earth became the standard behind every standard.

The Rotation as the Last Natural Standard

By the end of the 19th century, the rotation carried a weight it had never carried before. A Popular Science Monthly article of 1885 ranks the uniformity of the Earth's rotation as a more important and anxious question for the modern astronomer than the exact shape of the globe, precisely because that rotation supplies the fundamental unit of time ^[12]. Two decades later, the same magazine reports a telling substitution: the sidereal day, the duration of one rotation, had replaced the beats of the pendulum as the constant unit of time, by stipulating that any two rotations have the same duration . The planet was enthroned as the definition of time at the very moment its regularity was in doubt, for the same text adds that many astronomers suspected the tides of acting as a check on the globe . The Encyclopaedia Britannica treats the possible variability of the day as a question of fundamental importance, and states without hesitation that tidal retardation must gradually lengthen it .

The conceptual tool for thinking about an imperfect standard already existed in the clockmaker's workshop. Defending his timekeeper against a pamphlet issued under the authority of the Board of Longitude, John Harrison argued that a watch need not complete its revolutions in exactly the time the Earth takes to complete its own; it is enough that it does so invariably, in a known time, so that the constant difference, its rate, can be added or subtracted to recover true time ^[13]. This argument separates two ideas that common sense merges: regularity and agreement. A clock may disagree with its reference as long as it disagrees steadily. The observer of 1886 applied the same logic when he fell back on his clock's rate under a clouded sky . Nothing in this reasoning reserved it for watches. Applied to the Earth, it suggested that the rotation itself might be nothing more than a timekeeper with a rate, and therefore with an error.

For several decades, the smallness of the effect shielded the rotation from that suspicion. James Arthur, in a treatise on the measurement of time, concedes that the mathematicians of astronomy conclude the rotation is slowing, but judges the amount a vanishing quantity compared with the smallest errors of the finest clocks, so that the rotation may be treated as uniform for countless generations ^[14]. Herbert Spencer draws a philosophical lesson from the same apparent stability: the day, he asserts, is not a second shorter than it was two thousand years ago, a sign that the relations of force in the solar system have settled at the end of a long evolution ^[15]. A writer in The Catholic World in 1866–1867 takes the opposite path, acknowledging a slowing scarcely perceptible yet generally recognized, and extrapolating it to the point where the Earth would cease to rotate and turn one face constantly to the sun ^[16]. Practical neglect, cosmological fixity and eschatological extrapolation: three readings of one tiny quantity, each revealing more about its author's framework than about the Earth.

Behind this confidence lay an assumption that only became visible when it collapsed. Aram D'Abro, writing in 1927 on the evolution of scientific thought from Newton to Einstein, observes that classical science took for granted that two planets rotating at the same speed would always beat out the same unit of time, whatever their relative motion, because time was held to be one and universal; the choice of planet was therefore indifferent ^[17]. The rotation could serve as a standard only because time was presumed to exist independently of it. The philosopher Jaime Luciano Balmes exposed the circularity in another register: if a celestial motion is taken as the measure of time, then a revolution performed at half or double its usual velocity would still mark the same time, which he judges absurd ^[18]. A motion can serve to measure time; it cannot be time. That distinction, obvious in the abstract, became an empirical matter as soon as the tides were shown to act on the Earth.

The Tidal Brake and the Irreversible Day

The mechanism of the brake was described in images borrowed from engineering. Sir Oliver Lodge explains that the energy of the tides is continuously dissipated by friction and that all of it is drawn from the Earth's rotation: tidal mills, were engineers to build them, would literally run at the planet's expense, with the moon as fulcrum and the tidal bulge as the handle of a friction brake ^[19]. Andrew Gray, recounting the work of William Thomson, offers a parallel figure: the solid Earth turns beneath a mass of water held roughly fixed by the moon, like a fly-wheel turning under a stationary friction band, and is retarded by this water-brake ^[20]. These metaphors matter. They place the Earth among the machines whose regularity it was supposed to guarantee, subject like them to wear, friction and loss.

Robert S. Ball turns the image into an accounting. Since the energy withdrawn from the Earth–Moon system cannot come from the moon's revolution, it must come from the rotation of the Earth on its axis; the rotation must therefore slow, the day must lengthen, and the share of moment of momentum carried by the Earth must diminish ^[21]. What the Earth loses, the system keeps, transferred to the moon's orbit. Gray reports Thomson's view of the resulting tendency: the moon recedes and moves more slowly in its orbit, while the period of the Earth's rotation drifts toward coincidence with that of the moon's revolution ^[22]. The clock, in other words, is not merely imperfect; it is running down toward a state in which the day would stretch to the length of the month.

This finding broke with a central conviction of gravitational astronomy. Arthur Berry, in his history of the discipline published in 1898, recalls that the great stability theorems showed ordinary perturbations to be self-compensating, so that a motion accelerated at one time would be retarded at another; tidal friction escapes this rule, lengthening both day and month and widening the distance between Earth and moon ^[23]. Those theorems, he notes, had assumed perfectly rigid bodies and so ignored both the ocean tides and the bodily tides of the solid Earth; tidal friction, minute as it is, differs in kind from the classical perturbations because its action is irreversible ^[24]. The image of the solar system as a self-compensating clock, which Leitch had drawn three decades earlier with theological confidence , loses its foundation. A universe of oscillations gives way to a universe with a direction, and the Earth's rotation, the reference of all references, falls on the wrong side of that line.

The magnitudes remained modest, and their transmission imprecise. According to Gray, Thomson estimated from the best astronomical observations that the Earth would fall behind a perfect timekeeper by about twenty seconds in a century . A popular astronomy published in 2020 by Juan Carlos Beamin restates the transfer of energy to the moon's orbit and gives the moon's recession as about four centimeters a year, yet in the same passage says that the day would grow two thousandths of a second shorter within a century, before concluding that it will one day exceed 24 hours ^[25]. The slip is telling: the direction of a minute secular change, plain in principle, remains easy to invert once put into words. It also recalls that the evidence for the slowing was never a direct reading of the Earth's spin, but an inference drawn from other bodies, above all the moon.

When the Standard and the Measured Changed Places

The moon was where the anomaly first showed. David P. Todd recalls that Halley had found the moon's motion around the Earth subject to a slight acceleration, which puzzled Lagrange and Laplace because they had proved such an effect impossible under gravitation alone, unless some other force intervened ^[26]. Ball shows how the puzzle could be resolved without any new force acting on the moon: the moon might keep exactly the same pace while the day lengthened, in which case fewer days would be needed for each lunar revolution, and the moon would seem to accelerate ^[27]. On his account, the observed acceleration combines two causes, one found by Laplace in the indirect action of the planets, though Laplace had overestimated it, and the other lodged in the measuring day itself .

George Howard Darwin, in the 1911 Britannica, states the reversal with the greatest clarity. Tidal friction retards both the moon's angular motion and the Earth's rotation, but the latter more than the former; an observer who treats the Earth as a perfect timekeeper will therefore read a real retardation of the moon as an acceleration ^[28]. Darwin adds that a true acceleration also exists, due to a slow change in the eccentricity of the Earth's orbit, so that the observed figure mixes a genuine motion with an artefact of the clock . Here the second break is complete in principle. The moon, once the object measured against the Earth's rotation, becomes the instrument that checks it; the perfect timekeeper to which Thomson compared the Earth is no longer embodied by any natural motion. The standard has become an abstraction, and the rotation a phenomenon to be measured against it.

This reversal has a philosophical counterpart in the same decades. Bertrand Russell, writing on relativity in 1925, holds that the collapse of a single all-embracing time must eventually affect ideas of cause, evolution and progress, since whether the universe advances may depend on which of several equally good clocks one chooses ^[29]. D'Abro draws the consequence for planetary rotation directly: once time is no longer universal, the unit beaten out by each planet's turning loses the indifference that classical science had granted it . Balmes's argument, framed in purely philosophical terms, pointed the same way : no motion, however majestic, can be identified with time without absurdity. Tidal dynamics and relativity reached this conclusion by entirely different routes, but they converged on one claim, that nature supplies no privileged clock.

The practical outcome appears in a United States government text of 2002, more than a century after the 1885 article that called the rotation the fundamental unit of time . Leap seconds, it explains, allow a clock to keep approximately in step with the sun; but because the rate of the Earth's rotation varies, their occurrence cannot be predicted in detail ^[30]. The roles are now exactly inverted. The clock is the reference, and the Earth is the irregular phenomenon to which an occasional correction is conceded, much as Harrison added a rate to his watch . Yet the inversion is not a severance. That the correction exists at all shows that civil time still refuses to drift away from the sun, and the compromise preserves, in diminished form, the bond that the first break had loosened without cutting .

Across these sources, civil time leaves the sky in two stages of different nature. The first replaced the visible sun with an averaged day but kept the sky as arbiter: the stars still corrected the clocks, and the rotation became the foundation of every standard . The second turned that arbiter into an object of measurement, once tidal friction revealed an irreversible slowing that the stability theorems had excluded . In neither case was a new phenomenon decisive. The inequality of solar days was known as early as 1666, more than two centuries before the rotation was named the fundamental unit of time , and the slowing was admitted as real by the 1860s ; what changed was the precision of instruments and the collapse of the belief that the heavens compensate their own irregularities .

The story nonetheless leaves open the question Nilsson raised about far older societies: whether time-reckoning gains in truth, or only in convenience, as it detaches itself from natural phenomena . Henslow's verdict on mean time, artificial yet serviceable , applies with greater force to a standard that no longer coincides even with the Earth's rotation. Darwin's observer offers perhaps the most durable lesson: whoever treats a body as a perfect clock will mistake that clock's drift for a movement of the world . The measure of time became trustworthy only when it stopped trusting the turning of the world.