Coming of Age in the Milky Way - Timothy Ferris (2003)



In Tahiti … the women are possessed of a delicate organization, a sprightly turn of mind, a lively, fanciful imagination, a wonderful quickness of parts and sensibility, a sweetness of temper, and a desire to please.

—Johann Georg Forster, 1778

           The conception of the solar system that the Western world had attained by the beginning of the eighteenth century was accurate in its proportions but indeterminate in scale. Thanks principally to the theoretical work of Copernicus and Kepler and to the observations of Tycho and Galileo, it had been established beyond dispute that the earth was one of five known planets moving in elliptical orbits around the sun. And, thanks to Newton, these motions could be interpreted and predicted in terms of a mathematically cogent dynamical scheme that embraced terrestrial as well as extraterrestrial physics. But, though the relative distances of the sun and planets were understood, their absolute distances were not.

Copernicus had measured the proportions of the solar system to within 5 percent of the correct values, and Kepler had come closer still. These relative distances customarily were expressed in terms of the distance from the earth to the sun, a quantity known as the astronomical unit. But nobody knew what the distance to the sun might be; in other words, the value of the astronomical unit had not been determined. Here was a clear challenge. Since the proportions of the system already were known, if the distance to the sun or to any one planet could be ascertained, the distances of all the other planets would follow. And, since the apparent diameters of the planets could by now be measured rather well, by using a micrometer eyepiece attached to a good telescope, the sizes of the planets could be ascertained as soon as their distances had been measured. Beyond that lay the exciting prospect that, by using the astronomical unit as a baseline, it might be possible to triangulate nearby stars and measure their distances as well. Accomplishing this feat constituted one of the heroic endeavors of eighteenth-century astronomy.

Traditional estimates of the distance from the earth to the sun were of little help. Beginning with Hipparchus in the second century B.C. and ranging down through Ptolemy, Copernicus, and Tycho, astronomers had assumed as a rule of thumb that the astronomical unit was equal to about twelve hundred times the radius of the earth—in modern figures, some 4.8 million miles. Such a distance seemed appropriately vast; to borrow a conceit from the thirteenth century, had Adam started walking on the day of the creation (usually set at 4004 B.C.) he would have required six hundred years to reach the sun, and would have arrived, footsore, at the planet Jupiter in the twentieth century. Nevertheless, an astronomical unit of twelve hundred earth radii was twenty times smaller than the real distance. Kepler and later observers suspected that it was an underestimate—Kepler guessed that the value was more like thirty-five hundred earth radii, nearly three times the previous estimates—but these early observers lacked observational instruments adequate to test their hunches.

Two ways of obtaining distance data were available. One, micrometry, was theoretically crude but practically accessible. The other, triangulation, was perfect in theory but difficult to accomplish in practice.

Micrometry consisted of using a micrometer—an eyepiece equipped with an adjustable knife blade—to measure the apparent diameter of a planet as seen through a telescope. The astronomer then estimated the distance of a planet by comparing its apparent diameter with what he guessed to be its actual diameter. Obviously, the result could be no better than the guess as to the planet’s diameter. A few astronomers guessed very well indeed: Christian Huygens in 1659 assumed that the diameter of Mars was about 60 percent that of the earth (the correct figure is 53 percent), then measured the apparent size of the disk of Mars through a telescope and calculated a value for the astronomical unit of one hundred million miles. This came astonishingly close to the truth—the mean distance separating the earth from the sun is ninety-three million miles—but it depended entirely upon the accuracy of Huygens’s hunch about the size of Mars, and that, as Huygens himself was the first to concede, was “a slippery basis” upon which to base so important a result.1 The issue was not who made the luckiest guess, but who could obtain observational data that would establish the value of the astronomical unit to everyone’s satisfaction. This micrometry alone could not do.

Triangulation, called parallax (from the Greek parallaxis, for the value of an angle), was the sounder method. If a planet were observed simultaneously by two observers stationed thousands of miles apart—one in France, say, and the other in Mexico—its position against the background stars would appear to be slightly altered for the astronomer in Mexico as compared to the one in France, owing to their different perspectives on it. If both this angle and the baseline distance separating the two astronomers could be measured, the distance to the planet could be calculated through the straightforward application of euclidean geometry.

That triangulation was theoretically sound had been appreciated since ancient times. The difficulty lay in execution. First, one had to know the exact distance between two widely separated observers; this required reasonably accurate intercontinental maps. Second, the observations had to be carried out at the same time, to avoid errors introduced by the motions of the planets and by the rotation of the earth on its axis; this required accurate clocks and a way of synchronizing them. Third, the position of the planet against the stars had to be plotted precisely, because any triangle drawn between a planet and two points on Earth is going to be a very long, thin triangle indeed. Still, the thing could be done, given sufficient exactitude in the measurement of terrestrial space and time.

The parallax of Mars was first obtained by simultaneously observing the planet’s apparent location from two widely separated places on Earth. The difference in perspective made it possible to measure the value of angle X, which yields the distance from Earth to Mars. The angle, however, is small: Were the earth the size depicted in this illustration, Mars would be five hundred feet away.

The parallax of stars can best be measured by using as a baseline not the earth but the earth’s orbit around the sun. Even with so large a baseline, however, angle X is extremely small: Were the earth’s orbit the size here depicted, the nearest star would be more than two miles away.

Fortunately for science, rapid progress was being made in both cartography and chronometry. The agency responsible, however, was less the pursuit of pure knowledge than the accumulation of the booty of empire. The wealth of the world flowed into eighteenth-century Europe in ships: From their holds came the Indian rosewood of the dining tables where Newton and Halley were entertained, the African gold inlay on the plates, the turkey with corn they were served as the main course, the chocolate for dessert, and the tobacco they smoked afterward. But blue-water navigation was as hazardous as it was inexact, and sailors who ventured far beyond the sight of land were forever groping their way in the unknown—they were “at sea,” as we still say today—with results that ranged from delay to disaster. Many a cargo of silver, sugar, or hardwood had been conveyed across the Atlantic or Indian oceans only to be dashed against the rocks of Land’s End or the Cape of Good Hope. The situation had improved little in the century that had passed since the geographer Richard Hakluyt wrote of navigators that “no kind of men in any profession in the commonwealth pass their years in so great and continual hazard of life. … Of so many, so few grow to gray hairs.”2 The definitive catastrophe came in 1707, when Sir Cloudesley Shovell, four ships of his fleet, and fully two thousand of his men were lost on the rocks of the Scilly Islands of southwest England, this on a night when his navigators had reckoned that the fleet was in safe waters hundreds of miles to the west. Clearly, something had to be done.

The problem had to do with the determination of longitude. It had long been possible for a ship’s navigator to find his latitude—his location in a north-south direction—by measuring the altitude above the horizon of the pole star or of the sun at noon. The instrument employed for this purpose was the astrolabe (from the Greek for “to take a star”), a disk made of copper or tin, five to seven inches in diameter, fitted with a movable sighting arm. At local noon on any clear day aboard a ship of the line, three officers could be seen helping to shoot the sun—one holding the astrolabe steady, another sighting it, and a third reading the elevation—while deckhands stood by to catch the navigator when he fell or to retrieve the astrolabe if it were dropped and went scuttling across the rolling deck. The efficiency of the astrolabe had been improving, through the endeavors of Newton, Halley, John Hadley, Thomas Godfrey, and others, who made the instrument less cumbersome by reducing it first to a quarter of a circle (the “quadrant”), then to a sixth (the “sextant”), by employing mirrors to fold its optics so that the observer could see sun and horizon superimposed, and by adding filters and a telescope for greater accuracy. But, although these improvements helped navigators refine their calculations of latitude, they did not help them determine their longitude—their position in the east-west direction. Here the question was as much one of time as of space.

As the earth turns, the stars troop across the sky at a rate of fifteen degrees per hour. This means that if you know the time, the sky will tell you where you are. But knowledge of the exact time was just what navigators of Newton’s day lacked. On land, time was kept by pendulum clocks, but pendulums do not work at sea; the rolling of the boat wrecks their performance. A typical ship’s clock in the early eighteenth century was accurate to no better than five or ten minutes per day, which translated into a miscalculation of fully five hundred miles in longitude after only ten days at sea. It was just such an error that had dashed Cloudesley Shovell’s fleet on the rocks of the Scilly Islands.

The problem of determining longitude at sea had defied resolution for so long that many regarded it as unsolvable. The mathematician in Cervantes’s The Dog’s Dialog muses crazily that he has “spent twenty-two years searching for the fixed point”— el punto fijo, the correct longitude—“and here it leaves me, and there I have it, and when it seems I really have it and it cannot possibly escape me, then, when I am not looking, I find myself so far away again that I am astonished. The same thing happens with squaring the circle.”3 Sebastian Cabot on his deathbed claimed that God had revealed the answer to him, but added, alas, that He had also sworn him to secrecy.

Still, the longitude problem was obviously imperative, and more than a few inventors took it on, encouraged by the large cash prizes proffered by the governments of seafaring states like Spain, Portugal, Venice, Holland, and England. The richest of these was a prize of twenty thousand pounds, offered by the British Board of Longitude to anyone who could devise a practicable method of determining longitude on a transatlantic crossing to within one-half a degree, which equals sixty-three nautical miles at the latitude of London. John Harrison, an uneducated carpenter turned clock-maker, pursued the prize for much of his working life. He constructed a succession of “watches” (the term, meaning a portable clock, comes from the shipboard practice of dividing up the day into six watches of four hours each) of increasingly subtle and rugged design, checking them for accuracy by observing the disappearance of designated stars behind a neighbor’s chimney each night. His masterpiece, a marine chronometer that took him nineteen years to complete, was transported to Port Royal, Jamaica, aboard H.M.S. Deptford in 1761–1762, was there tested against sightings of the sun, and was found to have lost only 5.1 seconds in eighty days—a performance that many of today’s timepieces could not match. Nonetheless, it took Harrison years of lobbying to collect a portion of the prize, and he never got it all; twenty thousand pounds was a lot of money.

The astronomers and geographers, however, did not have to wait as long as did the mariners to improve their measurements of earthly space and time. Maps were constantly improving: Although pendulum clocks were not yet reliable at sea, they could be synchronized on land, by observing transits and eclipses of the satellites of Jupiter. (The Dutch had awarded Galileo a gold chain for proposing this ingenious idea, though they could not make it work on board ship, since any telescopic magnification sufficient to resolve Jupiter’s moons also magnified the rocking of the boat too much for the planet to be kept in view.) In France, cartographers led by Giovanni Cassini and Jean Picard employed Galileo’s method to cage the continent in a cat’s cradle of surveyor’s triangles, producing an accurate map that enabled Picard to determine the circumference of the earth to within 126 miles of the correct value.*

Equipped with better maps and clocks, astronomers tried to triangulate the neighboring planets Mars and Venus. In 1672, an international expedition led by the young French astronomer Jean Richer sailed to Cayenne, on the South American seacoast three hundred miles north of the equator. There he observed Mars during its closest approach to Earth at the same time that his colleagues, their clocks synchronized to Richer’s, sighted Mars from their post at the French Academy. Cassini sorted through the data and derived a value for the astronomical unit of eighty-seven million miles. This approximated the correct figure of ninety-three million miles, but given the many residual inaccuracies of the instruments and techniques of the time, Cassini’s like Huygens’s earlier estimate necessarily was regarded as but an educated guess.

Venus comes closer to Earth than does Mars, and so should be still more accessible to triangulation, but when closest it is lost in the glare of the sun. Twice in a long while, however, in pairs of events separated by just over a century, Venus passes directly in front of the sun. During these transits, as they are called, the planet appears as a black circle silhouetted against the blazing solar disk. Edmond Halley, who had observed a transit of Mercury during his expedition to St. Helena, realized that the distance to Venus might be determined by timing, from widely separated stations, exactly when the planet appeared and disappeared from the face of the sun. The edge of the sun would serve as a clearly defined backdrop, the planet as a kind of surveyor’s stake out in space.

Halley knew that he would not live to observe a transit of Venus. There had been a pair of transits in 1631 and 1639, a generation before he was born; the next pair were due in 1761 and 1769, by which time he would have been over a hundred years old.* (Halley must have been getting used to this sort of thing; he didn’t live to see the return of Halley’s Comet, either.) And so it was with the insistence of a man striving to project his words beyond the grave that Halley, in a paper published in 1716 “which,” he wrote, “I prophecy will be immortal,” outlined the procedure for the benefit of astronomers yet unborn:

We therefore recommend again and again, to the curious investigators of the stars to whom, when our lives are over, these observations are entrusted, that they, mindful of our advice, apply themselves to the undertaking of these observations vigorously. And for them we desire and pray for all good luck, especially that they not be deprived of this coveted spectacle by the unfortunate obscuration of cloudy heavens, and that the immensities of the celestial spheres, compelled to more precise boundaries, may at last yield to their glory and eternal fame.4

Previous observations of transits had been rare and rather haphazard. Pierre Gassendi in Paris managed to observe a transit of Mercury in 1631 that Kepler had predicted; he stamped on the floor to alert his young assistant to measure the altitude of the sun, but the boy, growing impatient after three days of waiting for the great event, had wandered off. Gassendi’s solitary published observation was useless for triangulation, though it did reveal that the disk of Mercury was much smaller than had been thought: “I could hardly be persuaded that it was Mercury, so much was I preoccupied by the expectation of a greater size,”5 Gassendi wrote. This supported the contention of Galileo that the solar system was considerably larger than had been estimated by Ptolemy and the other geocentrists.

As for Venus, its transit on December 6–7, 1631, was visible only from the New World and appears to have been viewed by not a single human being, and the transit of November 24, 1639, was observed by but two people, the English astronomer and clergyman Jeremiah Horrocks and his friend William Crabtree. Alarmingly for Horrocks, who was a clergyman, the transit occurred on a Sunday, when he was obliged to preach two sermons. He rushed home from church, peered through his telescope at 3:15 P.M., and saw Venus, “the object of my most sanguine wishes … just wholly entered upon the Sun’s disk.”6 Venus, like Mercury, looked smaller than had been predicted—Kepler thought Venus would cover one quarter of the sun, an enormous overestimate—and so to behold its tiny apparent size helped improve human appreciation of interplanetary distances. But Horrocks had no way to measure the apparent diameter of the disk precisely, and, since he was but one observer, he could not have triangulated Venus even if he had possessed an accurate clock. Crabtree, for his part, was so overwhelmed by the sight of an entire world dwarfed by the sun that he made no coherent notes at all, prompting Horrocks to protest that “we astronomers have a certain … disposition [to be] distractedly delighted with light and trifling circumstances.”7

But the world had changed by the time the transits of Venus of 1761 and 1769 came due. Astronomy had become an organized science, conducted by professionals, sponsored by scientific societies, and supported by government funds. Now at last, it was felt, science had the resources to sound the dimensions of the solar system. Halley’s implorations were remembered, and the transits were scrutinized by scores of observers equipped with micrometers, accurate clocks, and brass telescopes mounted on hardwood tripods at sites as far away as Siberia, South Africa, Mexico, and the South Pacific.

And, to an extent, the transit observers succeeded, though not without suffering sufficient tribulations to remind them that while the motions of the planets may be sublime the affairs of this world are marbled with chaos. The astronomer Charles Mason and the surveyor Jeremiah Dixon, later of the Mason-Dixon Line, were attacked by a French frigate while making their way to Africa (this was during the Seven Years’ War) with a loss of eleven dead and thirty-seven wounded; they reached Cape Town under military escort and observed the 1761 transit, only to find that they differed by many seconds in their estimate of the time when Venus had entered and left the disk of the sun. William Wales timed the transit from Hudson Bay, Canada, after enduring mosquitoes, horseflies, and a winter sufficiently severe that, as he noted with empirical exactitude, a half-pint of brandy left unattended iced over in only five minutes. Jean-Baptiste Chappe d’Auteroche, dispatched by the French Academy into the depths of Russia, raced across the frozen Volga and through Siberian forests in horse-drawn sleds, arrived at Tobolsk six days prior to the transit, posted guards to repel angry mobs who blamed him for causing spring floods by interfering with the sun, and managed to observe the transit. He died eight years later in Baja California after timing the 1769 transit, of an epidemic that spared but one member of his party, who dutifully returned his data to Paris. Alexandre-Gui Pingré was rained out for most of the transit in Madagascar, lost his ship to the British and was returned to Lisbon under British guns; a humanist as well as a scientist, he took comfort in the ship’s rations of spirits: “Liquor,” he wrote, “gives us the necessary strength for determining the distance of … the sun.”8

Least fortunate of all was Guillaume le Gentil, who sailed from France on March 26, 1760, planning to observe the transit the following year from the east coast of India. Monsoons blew his ship off course, and transit day found him becalmed in the middle of the Indian Ocean, unable to make any useful observations. Determined to redeem the expedition by observing the second transit, Le Gentil booked passage to India, built an observatory atop an obsolete powder magazine in Pondicherry, and waited. The sky remained marvelously clear throughout May, only to cloud over on June 4, the morning of the transit, then clear again as soon as the transit was over. Wrote Le Gentil:

I was more than two weeks in a singular dejection and almost did not have the courage to take up my pen to continue my journal; and several times it fell from my hands, when the moment came to report to France the fate of my operations…. This is the fate which often awaits astronomers. I had gone more than ten thousand leagues; it seemed that I had crossed such a great expanse of seas, exiling myself from my native land, only to be the spectator of a fatal cloud which came to place itself before the sun at the precise moment of my observation, to carry off from me the fruits of my pains and of my fatigues.9

Worse lay ahead. Stricken with dysentery, Le Gentil remained in India for another nine months, bedridden. He then booked passage home aboard a Spanish warship that was demasted in a hurricane off the Cape of Good Hope and blown off course north of the Azores before finally limping into port at Cadiz. Le Gentil crossed the Pyrenees and at last set foot on French soil, after eleven years, six months, and thirteen days of absence. Upon his return to Paris he learned that he had been declared dead, his estate looted, and its remains divided up among his heirs and creditors. He renounced astronomy, married, and retired to write his memoirs. Cassini, eulogizing Le Gentil, praised his character but allowed that “in his sea voyages he had contracted a little unsociability and brusqueness.”10

The most elaborate of the transit expeditions, mounted by the Royal Society, departed aboard the ninety-eight-foot bark H.M.S. Endeavour from Plymouth on August 26, 1768, with a deputation of scientists led by Joseph Banks, a wealthy botanist and future president of the Royal Society. Endeavour was equipped with crates full of clocks, telescopes, and meteorological equipment, as well as a barrel of nails for trading with the Tahitians, who had a passion for anything made of metal. The commander was Captain James Cook, an expert navigator, marine surveyor, and mathematician who had taught himself astronomy so well that, by observing the solar eclipse of 1766, he had been able to determine his longitude in Newfoundland to within two nautical miles. An empiricist in the social as well as the physical sciences, Cook found by experimenting with diet that he could ward off scurvy by feeding his men sauerkraut, which he shrewdly popularized among the hands by at first restricting it to the officers’ mess. The voyage was uneventful by the standards of the day: The expeditionaries took on three thousand gallons of wine and a thousand pounds of onions at Madeira, were fired upon in the Falklands by a half-mad viceroy who understood the transit to involve “the North Star passing through the South Pole,” and lost four men—a veteran seaman who drowned, a young marine who jumped overboard in shame after having stolen a bit of sealskin, and Banks’s two servants, who got drunk in a snowstorm in Tierra del Fuego and froze to death. After seven and a half months Endeavour reached Tahiti, then as now a synonym for paradise.

Cook issued strict orders to his men against unauthorized trading of metal objects with Tahitian females, who adorned their thighs with intricate tattoos of arrows and stars and saw nothing wrong in trading sexual favors for a nail or two. Cook recalled with concern that the crew of an earlier ship to reach Tahiti, the Dolphin, had in their enthusiasm for the Tahitian girls extracted so many nails from the ship that they nearly pulled it apart. When two of Cook’s marines deserted, married Tahitians, and fled to the mountains, Cook had them brought back and clapped in irons; he was a humane man, but he intended to return to England. His orders notwithstanding, though, nails and other metal objects kept vanishing from the ship.

Under Cook’s and Banks’s direction a fortress observatory was erected on Tahiti at what has ever since been known as Point Venus, and from there the transit of June 3, 1769, was observed under clear blue skies.

Timing the transit, however, proved difficult. The trouble was that Venus has a thick atmosphere, which refracts and diffuses the sunlight passing through it. As a result the disk of the planet, rather than snapping crisply into view as does the disk of airless Mercury when it is in transit, seems instead to adhere to the edge of the sun, like a raindrop hanging from a branch. “We very distinctly saw an Atmosphere or dusky shade round the body of the Planet which very much disturbed the times of the Contacts,” Cook noted in his journal.11 Consequently, Cook and astronomer Charles Green, observing through identical telescopes, differed in their estimates of the entry and exit times of Venus’ disk by as much as twenty seconds.

But despite these difficulties, the data gathered by Cook’s and the other scientific expeditions yielded estimates of the distance from the earth to the sun that came within 10 percent of the correct value. The astronomical unit subsequently was measured even more accurately by scientists who drew imaginary triangles, ever more refined, to Venus during its nineteenth-century transits, to Mars when it was in opposition in 1877, and to dozens of asteroids (or “minor planets”) as these previously useless chunks of rock drifted past the earth.

The immensity of the solar system, nearly a hundred times the Ptolemaic estimate of the size of the entire universe, now stood revealed, and scientists could with assurance turn their attention to the depths of interstellar space, and take on the still more ambitious task of measuring the distances of stars.

Here, too, some ground had been cleared by educated guesswork. One early approach to measuring stellar distances consisted of assuming that a given star was intrinsically just as bright as the sun, then measuring its apparent brightness (or magnitude) and estimating its distance by applying the law, known since Kepler’s day, that the apparent brightness of any object in space diminishes by the square of its distance. (This was analogous to earlier attempts to approximate the distances of planets by assuming that they were roughly the same size as the earth.) In the late seventeenth century, Christian Huygens observed the sun from a darkened room through pinholes of various sizes until he obtained an image that seemed equal in brightness to that of Sirius, the brightest star. Since the appropriate pinhole admitted 1 part in 27,664 of the sun’s light, Huygens concluded that Sirius was 27,664 times farther away than the sun—an underestimate by some twenty times, but an enormous distance nonetheless. A somewhat more refined approach, proposed by James Gregory in 1668 and detailed by Isaac Newton in a draft of the Principia, was to use Saturn, the outermost known planet, as a sort of reflecting mirror to gauge the intensity of sunlight. By guessing at Saturn’s reflectivity and assuming the stars to be of similar brightness to the sun, Newton concluded that the brightest stars are (to convert his figures into modern terms) about sixteen light-years away. The flaw here was that stars differ tremendously in their intrinsic luminosity; most of the bright stars we see in the sky are tens of times more luminous than the sun, and are, therefore, much more distant than we would guess by assuming that they resemble the sun.

The more promising strategy was to triangulate the stars. This could be accomplished by using, not the earth, but the orbit of the earth, as the baseline. The idea was to chart the position of a nearby star on two evenings six months apart, when the earth was at opposite extremities of its orbit, then look for a change in position produced by the change in our angle of sight on the nearby star against the more distant stars in the background. This method, known as stellar parallax, became theoretically practicable once the radius of the earth’s orbit—the astronomical unit—had been measured. Before it could be employed successfully, however, some of the subtleties of the earth’s motion had to be better understood.

The hero of this dry but vital business was the British astronomer James Bradley, Halley’s successor as Astronomer Royal. Raised on parallax, Bradley had triangulated Mars while still in his twenties, in the company of his uncle the amateur astronomer James Pound and Halley himself. Their observations indicated that the astronomical unit was equal to some 93 million to 125 million miles.

Eight years later, in 1725, Bradley and another amateur astronomer, Samuel Molyneux, installed a precision telescope in the chimney of Molyneux’s home. This “zenith telescope” pointed straight up, to the part of the sky where distortions of starlight induced by the earth’s atmosphere are at a minimum. It was used to observe but a single star, Gamma Draconis, which passed through the zenith at London’s latitude. Bradley and Molyneux reasoned that as the months went by the apparent position of Gamma Draconis would slowly shift, owing to the changing perspective introduced by the earth’s motion. The extent of this shift was to be measured by means of a plumb bob that would indicate how much the telescope’s aim had to be altered to bring the star back into the crosshairs. (Hooke, in the previous century, had used a zenith telescope to observe the same star, but the crudity of his instruments prevented his reaching any useful conclusion.)

The new assault on the parallax of Gamma Draconis proved more successful, but in an entirely unexpected way. As the months passed and Bradley’s observations of the star accumulated, he was surprised to find that the largest variation in its position occurred not annually, but daily. Intrigued, Bradley installed a second telescope, one capable of greater latitude of motion and, therefore, of observing more stars, and mounted it on the roof of his aunt’s home. (She obligingly permitted holes to be cut through the floors so that the measuring instruments could be placed in the cool, stable air of the basement.) By 1728, Bradley had observed more than two hundred stars and had found, to his amazement, that every one of them behaved in the same way: Each seemed to crawl slightly northward, then southward, every twenty-four hours. Bradley had no idea why.

Estimates of the size of Earth’s orbit, A.D. 100–1769

As often happens, the answer came to him not while he was at work in his observatory but while he was relaxing. While on a boat in the Thames, Bradley found himself gazing at a wind vane mounted atop the mast. It pointed into the wind and therefore seemed to change direction whenever the boat turned. What was changing, of course, was the orientation, not of the wind, but of the boat.

It occurred to Bradley that the earth is like a boat adrift in winds of starlight—that, as the earth moves through the starlight, its motion alters the apparent positions of the stars. Think of the earth as a woman walking briskly through the rain; her motion makes the raindrops seem to slant toward her, so she tilts her umbrella forward to compensate. Similarly, the earth’s motion makes starlight seem to slant, altering the apparent position of the stars hour by hour. Bradley had discovered what is called the aberration of starlight.

Twenty years later Bradley detected another subtlety in the earth’s motion, a nutation, or wobble, in the direction of its axis of rotation. These complications frustrated his efforts to detect the parallax of Gamma Draconis, but they paved the way for future parallax measurements—and, not incidentally, provided direct proof of the old Copernican hypothesis that the earth spins on its axis and orbits the sun.

But, since the stars are very far away, their triangulation called for instruments more precise than were available in Bradley’s day. Were the earth’s orbit represented by a serving platter one foot in diameter, a triangle drawn from the edges of the plate to the very nearest star would be twenty-six miles long, and its sides would be almost indistinguishable from parallel lines. The job facing the parallax astronomers was to detect the angle of convergence of such a triangle, and much thinner ones as well, and to measure the angles precisely enough to say where the lines would meet, for at that point stood the location of the star in three-dimensional space.

Bradley did not live to see the day when so great a degree of exactitude became attainable. But telescopes and their mountings kept improving, and in December 1838, Friedrich Wilhelm Bessel, a mathematician and astronomer who worked at the observatory of Königsberg with a precision telescope constructed by the master optician Joseph Fraunhofer of Munich, announced that after eighteen months of observations he had succeeded in measuring the parallax of the star 61 Cygni. Bessel’s measurement yielded a distance to 61 Cygni that came within 10 percent of the modern value of 10.9 light-years. Soon thereafter Thomas Henderson at the Cape of Good Hope obtained the parallax of Alpha Centauri, and Friedrich Struve in Russia found the parallax of the bright blue star Vega.

The angles, as expected, were tiny. The parallax of Alpha Centauri, which is the nearest star to the sun and therefore has the largest parallax, is only 0.3 seconds of arc, or one ten-thousandth of a degree. Clearly, interstellar space is built on an almost inconceivably gigantic scale. Light from our neighbor Alpha Centauri, traveling at 186,000 miles per second, takes four years and fifteen weeks to reach us (which is to say that Alpha Centauri is 4.3 light-years away), while 61 Cygni, the inconspicuous star scrutinized by Bessel, lies 11 light-years from the earth. But the vastness of the distances, which had long been inferred from the supposition that the stars are suns, made less of an impression than did the fact that such distances actually could be measured by human beings. Triangles born in the mind of Aristarchus of Samos had been extended out into the previously soundless depths of interstellar space, throwing back the conceptual horizons of cosmological thought. The sky was no longer the limit.

And yet, the more that came to be understood about the distant stars, the more intimate they seemed, as connections were identified linking the earth and the stars. One such insight in particular would have interested Captain Cook. It has to do with the iron that made the nails that the Tahitians found so alluring.

When the nuclear chemistry that powers the stars began to be deciphered by twentieth-century astrophysicists, it emerged that iron plays a central role in the evolution of stars. Stars burn by fusing the nuclei of the light atoms of hydrogen, the nucleus of which consists of but a single proton, and helium, which consists of two protons and two neutrons. In doing so stars release energy, which is how they shine, but they also build heavier atoms out of the lighter ones. As the process continues, each star forges atoms of carbon, oxygen, neon, sodium, magnesium and silicon, then nickel, cobalt, and, finally, iron. At iron the building stops; a normal, first-generation star lacks the energy required to make any heavier nuclei. The Sumerian name for iron, which means “metal from heaven,” is literally true: Iron is a working star’s proudest product.

When a star runs out of fuel, it can become unstable and explode, spewing much of its substance, now rich in iron and other heavy elements, into space. As time passes, this expanding bubble of gas is intermixed with passing interstellar clouds. The sun and its planets congealed from one such cloud. Time passed, human beings appeared, miners in the north of England dug the iron from the earth, and ironmongers pounded it into nails that longshoremen loaded in barrels into the holds of H.M.S Endeavour. Off the nails went to Tahiti, continuing a journey that had begun in the bowels of stars that died before the sun was born. The nails that Cook’s men traded with the Tahitian dancing girls, while on an expedition to measure the distance of the sun, were, themselves, the shards of ancient suns.

*The map revealed that France was smaller than had been thought, prompting the Sun King to remark that the scholars of the French Academy of Sciences had cost him more territory than had been lost to all France’s enemies in war.

*The most recent transits of Venus were in 1874 and 1882; the next will occur on June 7, 2004, and June 5, 2012.