Declination

The magnetic compass existed in China back as far as the fourth century BC. It was used as much for feng shui as for navigation on land. It was not until good steel needles could be forged that compasses were used for navigation at sea; before that, they could not retain their magnetism for long. The existence of magnetic declination, the difference between magnetic north and true north, was first recognized by Shen Kuo in 1088.^[3]

The first mention of a compass in Europe was in 1190 AD by Alexander Neckam. He described it as a common navigational aid for sailors, so the compass must have been introduced to Europe some time earlier. Whether the knowledge came from China to Europe, or was invented separately, is not clear. If the knowledge was transmitted, the most likely intermediary was Arab merchants, but Arabic literature does not mention the compass until after Neckam. There is also a difference in convention: Chinese compasses point south while European compasses point north.^[1]

In 1269, Pierre de Maricourt (commonly referred to as Petrus Peregrinus) wrote a letter to a friend in which he described two kinds of compass, one in which an oval lodestone floated in a bowl of water, and the first dry compass with the needle mounted on a pivot. He also was the first to write about experiments with magnetism and describe the laws of attraction. An example is the experiment where a magnet is broken into two pieces and the two pieces can attract and repel each other (in modern terms, they both have north and south poles).^[7] This letter, generally referred to as Epistola de Magnete, was a landmark in the history of science.^[1][2]

Petrus Peregrinus assumed that compasses point towards true north. While his contemporary Roger Bacon is reputed to observe that compasses deviated from true north, the idea of magnetic declination was only gradually accepted. At first it was thought that the declination must be the result of systematic error. However, by the middle of the fifteenth century, sundials in Germany were oriented using corrections for declination.^[8]

Inclination

A compass must be balanced to counter the tendency of the needle to dip in the direction of the Earth's field. Otherwise, it will not spin freely. Often, compasses that are balanced for one latitude do not work as well at a different latitude. This problem was first reported by Georg Hartmann, a vicar in Nuremberg, in 1544. The English mariner Robert Norman was the first to recognize that this occurs because the Earth's field itself is tilted from the vertical. In his book The Newe Attractive,^[9] Norman called inclination "a newe discouered secret and subtil propertie concernyng the Declinyng of the Needle." He created a compass in which the needle was floated in a goblet of water, attached to a cork to make it neutrally buoyant. The needle could orient itself in any direction, so it dipped to align itself with the Earth's field. Norman also created a dip circle, a compass needle pivoted about a horizontal axis, to measure the effect.^[4][8]

William Gilbert

Magnus magnes ipse est globus terrestris. (The Earth itself is a great magnet.)

— William Gilbert, De Magnete

1600 was a notable year for William Gilbert. He became president of the Royal College of Physicians of London, was appointed personal physician for Queen Elizabeth I, and wrote De Magnete, one of the books that mark the beginning of modern science. De Magnete is most famous for introducing (or at least popularizing) an experimental approach to science and deducing that the Earth is a great magnet.^[10]

Gilbert's book is divided into six chapters. The first is an introduction in which he discusses the importance of experiment and various facts about the Earth, including the insignificance of surface topography compared to the radius of the Earth. He also announces his deduction that the Earth is a great magnet. In book 2, Gilbert deals with "coition", or the laws of attraction. Gilbert distinguishes between magnetism and static electricity (the latter being induced by rubbing amber) and reports many experiments with both (some dating back to Peregrinus). One involves breaking a magnet in two and showing that both parts have a north and south pole.^[7] He also dismisses the idea of perpetual motion. The third book has a general description of magnetic directions along with details on how to magnetize a needle. He also introduces his terella, or "little Earth". This is a magnetized sphere that he uses to model the magnetic properties of the Earth. In chapters 4 and 5 he goes into more detail about the two components of the direction, declination and inclination.^[11][12]

In the late 1590s Henry Briggs, a professor of geometry at Gresham College in London, had published a table of magnetic inclination with latitude for the earth. It agreed well with the inclinations that Gilbert measured around the circumference of his terrella. Gilbert deduced that the Earth's magnetic field is equivalent to that of a uniformly magnetized sphere, magnetized parallel to the axis of rotation (in modern terms, a geocentric axial dipole). However, he was aware that declinations were not consistent with this model. Based on the declinations that were known at the time, he proposed that the continents, because of their raised topography, formed centers of attraction that made compass needles deviate. He even demonstrated this effect by gouging out some topography on his terella and measuring the effect on declinations. A Jesuit monk, Niccolò Cabeo, later took a leaf from Gilbert's book and showed that, if the topography was on the correct scale for the Earth, the differences between the highs and lows would only be about one tenth of a millimeter. Therefore, the continents could not noticeably affect the declination.^[11][12]

The sixth book of de Magnete was devoted to cosmology. He dismissed the prevailing Ptolemaic model of the universe, in which the planets and stars are organized in a series of concentric shells rotating about the Earth, on the grounds that the speeds involved would be absurdly large ("there cannot be diurnal motion of infinity").^[12] Instead, the Earth was rotating about its own axis. In place of the concentric shells, he proposed that the heavenly bodies interacted with each other and Earth through magnetic forces. Magnetism maintained the Earths position and made it rotate, while the magnetic attraction of the Moon drove the tides. Some obscure reasoning led to the peculiar conclusion that a terella, if freely suspended, would orient itself in the same direction as the Earth and rotate daily. Both Kepler and Galileo would adopt Gilbert's idea of magnetic attraction between heavenly bodies, but Newton's law of universal gravitation would render it obsolete.^[11]

Guillaume le Nautonier

In about 1603, the Frenchman Guillaume le Nautonier (William the Navigator),^[13] Sieur de Castelfranc, published a rival theory of the Earth's field in his book Mecometrie de l'eymant (Measurement of longitude with a magnet). Le Nautonier was a mathematician, astronomer and Royal Geographer in the court of Henry IV. He disagreed with Gilbert's assumption that the Earth had to be magnetized parallel to the rotational axis, and instead produced a model in which the magnetic moment was tilted by 22.5° – in effect, the first tilted dipole model. The last 196 pages of his book were taken up with tables of latitudes and longitudes with declination and inclination for use by mariners. If his model had been accurate, it could have been used to determine both latitude and longitude using a combination of magnetic declination and astronomical observations.^[2][7][12]

Le Nautonier tried to sell his model to Henry IV, and his son to the English leader Oliver Cromwell, both without success. It was widely criticized, with Didier Dounot concluding that the work was based on "unfounded assumptions, errors in calculation and data manipulation". However, the geophysicist Jean-Paul Poirier examined the works of both le Nautonier and Dounot, and found that the error was in Dounot's reasoning.^[7]

Temporal variation

One of Gilbert's conclusions was that the Earth's field could not vary in time. This was soon to be proved false by a series of measurements in London. In 1580, William Borough measured the declination and found it to be 111⁄4° NE. In 1622, Edmund Gunter found it to be 5° 56' NE. He noted the difference from Borough's result but concluded that Borough must have made a measurement error. In 1633, Henry Gellibrand measured the declination in the same location and found it to be 4° 05' NE. Because of the care with which Gunther had made his measurements, Gellibrand was confident that the changes were real. In 1635 he published A Discourse Mathematical on the Variation of the Magneticall Needle stating that the declination had changed by more than 7° in 54 years. The reality of geomagnetic secular variation was rapidly accepted in England, where Gellibrand had a high reputation, but in other countries it was met with skepticism until it was confirmed by further measurements.^[2][14]

The observations of Gellibrand inspired extensive efforts to determine the nature of variation - global or local, predictable or erratic. It also inspired new models for the origin of the field. Henry Bond Senior gained notoriety by successfully predicting in 1639 that the declination would be zero in London in 1657. His model, which involved a precessing dipole, was strongly criticized by a royal commission, but it continued to be published in navigational instruction manuals for decades. Dynamic models involving multiple poles were also proposed by Peter Perkins (1680) and Edmond Halley (1683, 1692), among others. In Halley's model, the Earth consisted of concentric spheres. Two magnetic poles were on a fixed outer sphere and two more were on an inner sphere that rotated westwards, giving rise to a "westward drift". Halley was so proud of this theory that a portrait of him at the age of eighty included a diagram of it (above right).^[15]