But there was a complication with Gilbert’s theory that magnetism was spread evenly through the Earth and that was declination. Gilbert was well aware of the phenomenon, since by now deviations of the compass needle from true north had been noted all over the explored world. If Earth’s magnetism were truly identical to that of a terrella and symmetrical about the rotation axis, there would be no declination and the compass needle would point due north wherever it was placed on Earth.
By 1600 Gilbert had amassed a sizeable collection of declination observations. He knew that starting from the Azore Islands and moving east, declination increased only as far as Plymouth on the south coast of England, where it reached a maximum of 13°24′ E. Further east the declination decreased again; at Helmshud in Finmark (a region in the extreme north of Norway and about 30 degrees east of Plymouth), the compass needle again pointed due north.
Furthermore, Gilbert found that although the declination at Corvo, an island in the Azores, might be zero, the same was not true along the whole meridian of the Azores. Nor, in general, was the declination constant along any meridian:
… still along the entire meridian of the island of Corvo the compass does by no means point due north. Neither in the whole meridian of Plymouth at other places is the variation 13 degrees 24 minutes, nor in other parts of the meridian of Helmshud does the needle point to the true pole.
A chart of magnetic declination around 1600, compiled from observations that would have been known to William Gilbert. Compare this with Halley’s chart of declination over the Atlantic Ocean a century later on page 58.
He therefore rejected Porta’s assertion that declination was related simply to longitude as “vain hope … baseless theory … false as false can be.”
Gilbert was unable to see any overall symmetry or regular pattern in his collection of declination measurements. He came to the conclusion that while the compass was influenced mainly by the magnetism of the globe as a whole—which he still believed to be symmetrical about the axis of rotation—superimposed on this was a secondary pull toward the magnetic rocks that made up the continents. This pull would, he thought, be strongest at the edge of the continents, so declination there would reach a maximum, as at Plymouth. It would decrease to zero in the middle of either the continents or the oceans, where the effects to east and west would cancel each other out, leaving the compass pointing due north.
Gilbert carried out a number of experiments designed to show this was a feasible explanation. In one, he gouged out an area on the surface of his terrella to represent the basin of the Atlantic Ocean. In another, he added lumpy blobs of lodestone to resemble the raised areas of continental land masses. He then repeated his observations, this time with two barleycorn-long iron wire “compasses.” These behaved as he predicted: those in the middle of the “oceans” and “continents” pointed towards the pole of the terrella, while the others deviated in the direction of the simulated continents.
Gilbert went on to maintain that the continents were permanently fixed in position and so declination should remain constant at any given place on Earth:
… forever unchanging, save there should be a great break-up of a continent … as of the region of Atlantis whereof Plato and ancient writers tell.
There was, of course, no way he could have foreseen the twentieth-century discovery of continental drift and plate tectonics. Even so, it seems curiously out of character that this pragmatic man should have given credence to the legend of Atlantis. Whatever the case, Gilbert was soon to be proved wrong: the notion of a static and unchanging magnetic Earth would be gone forever.
Meanwhile, in March 1603 Elizabeth I died, presumably through no negligence on the part of Gilbert, who continued as physician to her successor, James I. Indeed, it is said that the queen left just one personal legacy, a bequest to enable Gilbert to continue his scientific experiments. Unfortunately, though, the great scientist had little chance to use it: he succumbed to the plague and died just eight months later.
Gilbert’s explanation of declination meant the only way to get a complete picture was to go and make compass measurements over the entire globe. Until this was done, the compass and the angle of declination could provide little help in solving the longitude problem. By now, however, finding an answer to longitude had become something of a holy grail, and the idea that declination might provide a simple solution was too tempting to discard readily.
Hot on the heels of De Magnete, and not long before Gilbert’s death, an eccentric French aristocrat, Guillaume de Nautonnier, Sieur de Castelfranc-sur-Lot, had thrown his hat into the ring. De Nautonnier, Geographer Royal in the court of Henri IV, was a mathematician and astronomer. Although he had an interest in Earth’s magnetism, he evidently did not share Gilbert’s dedication to the experimental method. Instead, he preferred to stay with the Greek tradition of geometric calculations based on pure hypothesis.
In 1602 he published his own mammoth tome, Mécometrie de l’Eymant, which, loosely translated, means Determination of Longitude by Means of the Lodestone. In it he claimed to have independently deduced that the Earth is a great magnet. However, without experimentation his claim lacked the validation and authority of De Magnete.
De Nautonnier took issue with Gilbert’s explanation of declination, pointing out inconsistencies in his arguments. There was some truth in this. In Book I of De Magnete Gilbert had noted how insignificant the topography of the Earth was in comparison with the planet’s size. This was true: as we now know, while Earth’s radius is some 6400 kilometers, the highest mountains rise only ten kilometers and the deepest ocean trench lies only about ten kilometers below sea level. However, in Book IV Gilbert had used this same topography to account for quite significant angles of declination. Instead of developing a new explanation of declination, however, De Nautonnier tried to revive Mercator’s idea of an offset between the magnetic and geographic poles, but now he argued that it was the magnetization of the whole Earth that was tilted with respect to the rotation axis. The idea was simple, but having a physical explanation—a uniformly magnetized Earth— did not improve the fit of the actual declination observations at all. De Nautonnier’s model ran into all the same problems that Mercator had encountered in trying to calculate the locations of his magnetic poles. It was just too simple.
However, in 1602 de Nautonnier was still adamant that a simple declination pattern must exist. He was convinced that, once found, it would lead to a method of determining longitude and so he doggedly pursued the concept of a tilted magnetization. It is not difficult to see how tilting the magnetization and offsetting the magnetic poles produces a longitude-dependent declination pattern. Picture Gilbert’s geocentric axial dipole, with its poles on Earth’s axis of rotation. The situation is completely symmetrical: wherever a compass needle is placed on Earth, it will point due north towards the pole along a meridian, or line of longitude.
Now imagine tilting the magnetic axis so the magnetic poles move to opposite points in Siberia and Antarctica—de Nautonnier calculated the positions to be 67° N, 150° E and 67° S, 30° W. A compass needle will still point directly towards the magnetic north pole.
Imagine a new set of “magnetic meridians” outlining the alignments of compass needles over the globe—a network of great circles connecting the south magnetic pole to the north magnetic pole. These new magnetic meridians form a symmetrical pattern about the new magnetic axis. Two—the ones that pass through the geographic poles as well as the magnetic poles—coincide with geographic meridians, and along these the compass needle points due north. In other words, along these meridians the declination is zero (or 180 degrees along the short segments between the geographic and magnetic poles). In de Nautonnier’s model these meridians are at longitudes of 330° E and 150° E.
At any given point on Earth’s surface a compass needle points in the direction of the magnetic meridian. The solid line depicts meridians due to a geocentric axial dipole—essentially a magnet at the center of Ear
th aligned with the rotation axis—as proposed by William Gilbert. The dotted line depicts meridians resulting from a tilted geocentric dipole, as suggested in 1602 by Guillaume de Nautonnier, a geographer in the court of French king Henri IV, to account for observations of declination and as a possible means of determining longitude. The angle between a magnetic meridian and true north is equal to the declination.
All other magnetic meridians cut through the geographic meridians at angles that depend on location. These angles correspond to declination. Imagine taking a trip eastwards around the equator starting from 330° E. The declination will start at zero and steadily increase. After you have traveled through 90 degrees of longitude, it will reach a maximum 23 degrees at longitude 60° E. It will then decrease back to zero at 150° E, the longitude of de Nautonnier’s north magnetic pole. When you move north or south from the equator, declination will increase with latitude, eventually becoming close to 180 degrees at locations between the geographic and magnetic poles.
Most of de Nautonnier’s massive book was taken up with calculations and tabulations from which a mariner, armed with knowledge of his latitude and a measurement of declination, should have been able to work out his longitude. The latitude could be obtained by the traditional astronomical methods. Alternatively, a measurement of magnetic inclination could be used to calculate magnetic latitude, which itself is a function of latitude, longitude and the tilt of de Nautonnier’s dipole.
All this was without doubt a very clever piece of mathematics, and it would have been an elegant solution to the longitude problem —if only the premise on which it was founded, the tilted geocentric dipole, had been completely accurate. Even today the tilted dipole is a reasonable approximation to the most significant features of Earth’s magnetism but, like Gilbert, de Nautonnier was unwittingly up against the “non-dipole” component of Earth’s magnetism—the small fraction that does not fit the symmetrical pattern of a dipole, whatever its orientation.
It would be twenty-seven years before an Italian monk would finally demonstrate experimentally that Gilbert’s explanation of declination did not work either.
The Jesuits, a Roman Catholic order founded in 1534, had always placed a high value on education and scholarship. Although they maintained a traditional belief in an immobile Earth at the center of a perfect Creation, they were not averse to scientific experimentation to demonstrate the inner workings of Creation.
Niccolò Cabeo, a monk in the order, was interested in Gilbert’s work on magnetism, and in 1629 he published a critique under the title Philosophia Magnetica. Supposing Gilbert’s terrella to be about ten centimeters in diameter, Cabeo estimated the elevation of the continents and the depths of the oceans—constructed to scale—to be no more than one-tenth of a millimeter. He then showed that the effect of such tiny relief on the compass was negligible.
A new explanation was needed for the irregular pattern of declination over the globe. Although he had proved that Gilbert’s suggestion of the magnetization of continental rocks was insufficient, Cabeo was more interested in defending Aristotelian metaphysics than in explaining irregular features of Creation. It would be some considerable time before anyone came up with a better explanation of this “non-dipole” component of Earth’s magnetism.
Interestingly, Cabeo may never have seen De Magnete’ s Book VI: it had been removed from certain editions destined for Catholic Europe. In it Gilbert had eventually turned his mind to the question of Earth’s place in the universe and what role magnetism might play. He saw little option but to side with Copernicus. The distance to the stars must, he reasoned, be enormous. Therefore, to explain their apparent daily revolution around an immobile Earth, the stars above the equator had to be going at unimaginably high speeds. At the same time, those close to the pole—the Pole Star for example—must be hardly moving at all.
This was, Gilbert argued, highly improbable: “there cannot be diurnal motion of infinity, or of an infinite body.” Instead, the Earth must be simply spinning on its axis beneath, or rather within, the whole universe of planets and stars, completing one rotation each day. As to what drove and maintained this rotation, Gilbert supposed it to be an interaction between Earth’s magnetism and some sort of universal magnetic background.
In Catholic Europe, Gilbert’s explanation would certainly have been considered heresy. In 1600, the year that De Magnete was published, an Italian philosopher and former priest called Giordano Bruno was burned at the stake in Rome for holding similar beliefs. Fortunately for Gilbert, England was more liberal. Nevertheless it was still deemed prudent to spare continental Europeans from the doctrine of Book VI.
The Wandering Compass Needle
There is yet a further difficultie … great changes in the Needles direction within this last Century of years, not only at London, but almost all over the Globe of Earth … the effect of a great and permanent motion.
—EDMOND HALLEY, 1683
In 1634 Europe finally caught up with the secret that had lain hidden for centuries in the layout of Chinese village streets—that Earth’s magnetism was not static and unvarying, but declination, in particular, changed with time. In 1580 Gilbert’s colleague William Borough had measured the declination in London as 11°20′ E. In 1622, however, Edmund Gunter, professor of astronomy at London’s recently founded Gresham College, recorded only 6°0′. Gunter was surprised by this decrease of more than five degrees in scarcely forty years and, failing to come up with an explanation, doubted the accuracy of both his and Borough’s measurements. However, in 1634 Henry Gellibrand, Gunter’s successor at Gresham College, recorded 4°6′ E. It was clear that in London the declination was steadily decreasing.
Subsequent records would show this “secular variation” continuing. By the end of the seventeenth century a compass needle in London would point seven degrees west of north, while in the southern part of North America the declination would be to the east of north. In the space of one hundred years the pattern of declination across the Atlantic Ocean would change beyond recognition, completely invalidating Gilbert’s theory that declination arises because the compass needle is attracted towards extra magnetic material in the continents.
Scarcely fifty years after Gilbert’s death the challenge of understanding Earth’s magnetism attracted the attention of another young Englishman, Edmond Halley. Although he is now remembered principally for the comet that bears his name, Halley would make long-lasting contributions not only to astronomy but also to geomagnetism.
The scientific environment had changed a great deal since Gilbert’s time. Born in 1656, Halley was a contemporary of Isaac Newton, Newton’s friend Christopher Wren and his adversary Robert Hooke, all early members of the Royal Society. Founded in 1660, and initially based at Gresham College, itself an independent institution for the promotion of learning, the “Royal Society of London for Improving Natural Knowledge” set out to enhance communication and collaboration between scientists. It had built on Gilbert’s practice of regular meetings at which scientists could present and share ideas, solicited support for the work of its members and, sometimes controversially, fended off or sought to mediate in disputes of a scientific nature.
The society’s principal publication, Philosophical Transactions, first appeared in 1663 and continues in an unbroken series to this day. It was one of the first scientific journals to be made available on the Internet, with all issues back to the first now accessible at the click of a mouse. Similar societies soon appeared in other major cities across Europe, where scientists such as Descartes, Leibniz and Huygens were at work.
More people, albeit almost exclusively men, were taking an active interest in science as it became clear that, more than just an academic curiosity, science had valuable practical applications. Scientific language and techniques were developing rapidly. Terminology was becoming more precise, enabling scientists to formulate their questions more clearly. Experimental methods and instrumentation were improving, and new mathematical tools, s
uch as calculus, that naturally lend themselves to well-formulated physical problems, were emerging.
Edmond Halley was the son of a wealthy London merchant. His passion for both astronomy and geomagnetism were evident from his early days at the prestigious St. Paul’s School. In 1672, at the age of sixteen, he measured the magnetic declination in London as 2°30′ W, and noted that the compass needle now pointed slightly to the west of north, not east of north as it had earlier in the century. Secular variation was continuing to draw the compass needle further towards the west.
From St. Paul’s, Halley went up to Oxford, but in 1676 he cut short his studies when he was given the opportunity to join an astronomical expedition to St. Helena and the Ascension Islands, the goal of which was to map the stars of the southern hemisphere skies. As was to become the case in all his travels, Halley took his compass with him, and he made frequent measurements of declination. On his return, following the publication of his southern hemisphere star chart, Halley was elected a Fellow of the Royal Society, and at the bequest of the king was awarded the degree of M.A. by Oxford University. His scientific career was assured.
If 1600 had been a year of achievement for Gilbert, 1682 was an annus mirabilis for Halley. In August astronomers noticed a bright comet close to the sun. Halley studied its course closely and deduced that it must be the same comet as that seen in 1531 and again in 1607. Its elliptical orbit about the sun had, in the intervening years, taken it to the far reaches of the solar system. He calculated the period of the comet’s orbit as a little over seventy-six years, and predicted that its next appearance would be in 1759.
North Pole, South Pole: The Epic Quest to Solve the Great Mystery of Earth's Magnetism Page 5
