In the first volume of Geomagnetism, Chapman and Bartels appeared to accept this interpretation, but in the second they seemed to hedge their bets, writing:
Certain geological evidence that suggests a possible complete reversal of the Earth’s field during past ages is usually disregarded as unreliable—a view which is perhaps too dogmatic.
Chapman and Bartels’ book would be one of the last Anglo-German collaborations for some time. The outbreak of World War II would affect scientific research in Britain and much of Europe. Non-strategic research ground to a virtual standstill, while technology, including the development of materials and instruments that might have a military application, accelerated beyond all imagination. Ironically, this would have positive spin-offs for geomagnetism: the production of rugged electronic devices, and of materials that are easily magnetized and extremely stable, would have an enormous impact in the post-war years.
In France, a keen young physicist named Louis Néel now took the first steps towards an explanation of how and why rocks and baked clays retain remarkably strong and stable magnetizations over hundreds of millions of years of geological time. Néel’s studies had begun in 1928 at the University of Strasbourg’s Physics Institute, which had been established by Pierre-Ernest Weiss to focus on magnetic research. Weiss and his fellow physicists Pierre Curie and Paul Langevin had already begun to rewrite the physical theory of magnetism. Their pioneering work had dovetailed with revolutionary late nineteenth- and early twentieth-century discoveries in atomic and quantum physics. It would form the basis of Néel’s theory of rock magnetism.
Materials that can hold on to a stable, long-term or “remanent” magnetization are generally called ferromagnets. Iron and lodestone (magnetite) are common examples, but other natural minerals such as the rusty red iron oxide hematite and some iron sulfides are also ferromagnetic.
Faraday had suggested that most other materials should be classified as “paramagnetic” or “diamagnetic” because they acquired weak “induced” magnetization, either parallel or opposite to an ambient magnetic field, and lost it as soon as they were removed from the field. Pierre Curie had found that in both diamagnetic and paramagnetic materials the strength of the induced magnetization rose with the external magnetic field, and in the case of paramagnetism also decreased with temperature. However, the strong remanent magnetization of a ferromagnet was very different. Curie envisaged there was some sort of interaction between the atoms that caused spontaneous magnetic alignment and resulted in an overall magnetization.
It had long been known that ferromagnets such as iron and lodestone lost their magnetization when heated in a fire. Curie showed the temperature at which this occurred—now known as the Curie temperature—was characteristic of the particular material, and that above this temperature ferromagnetic materials behaved much like paramagnetic ones.
Curie’s work was published in 1895, just before the announcement by J.J. Thompson, Cavendish professor of physics at Cambridge, of the long anticipated discovery of the electron—the tiny, negatively charged particle present in electric currents and electric discharges. By 1905 atoms and electrons were all the rage in physics. Coulomb and Ampère had theorized that the basis of magnetism lay at the atomic level, but without a good picture of the atom they had been unable to develop the idea further. Now Langevin came up with an explanation of diamagnetism and paramagnetism by supposing that the electrons in an atom were themselves elementary magnetic dipoles. If the electronic dipoles in an atom cancelled one another out, a material made of such atoms was diamagnetic. If the elementary dipoles did not cancel one another out, the atom possessed a residual “dipole moment” and the material was paramagnetic. Using this model and the latest statistical theory, Langevin was able to account for Curie’s experimental observations of paramagnetic behavior.
Weiss’s contribution had been to explain the strong, stable magnetization of ferromagnetic materials by combining Curie’s suggestion of interatomic interactions with Langevin’s statistics. However, at the time (1907) there was still no good picture of the inner structure of the atom, and although it was suspected that the laws of physics might work a little differently within an atom, Weiss was in no position to even guess at the details. Instead, he assumed that the overall effect would be equivalent to an intense internal magnetic field, which he called the “molecular field,” and which would act on each atom over and above any externally applied magnetic field. Using this idea, he predicted there should be a critical temperature above which a ferromagnetic material would become paramagnetic. In other words, his theory predicted the Curie temperature.
It was only after the studies of Curie, Langevin and Weiss that the New Zealand-born physicist Ernest Rutherford, experimenting at the University of Manchester’s physics laboratory between 1909 and 1911, presented his famous picture of the atom as a tiny, central, positively charged nucleus surrounded by virtually empty space, occupied only by even tinier, negatively charged electrons. In 1920 Rutherford would name the nucleus’s positively charged particles protons, and in 1932 James Chadwick identified neutrons, the uncharged particles that make up the remaining mass in the nucleus.
Hot on the heels of Rutherford’s atom came an explosion of new ideas in physics. Scientists such as Niels Bohr, Erwin Schrödinger and Werner Heisenberg burst on to the scene, and quantum physics swept away many entrenched ideas of nineteenth-century classical physics. In particular, Heisenberg invoked the quantum mechanics of the electrons orbiting the nucleus of the iron atom to give a clear and lucid explanation of the interatomic interactions and molecular field that Curie and Weiss had deemed responsible for ferromagnetism.
French physicist Louis Néel, whose theoretical work showed that rocks can retain a stable record of the Earth’s magnetic field, sometimes for hundreds of millions of years.
This, then, was the state of play in 1928 when Louis Néel joined Weiss’s research group at Strasbourg. Immediately he noticed certain things that Weiss and Heisenberg had not, and which their theory could not explain. In particular some materials were not diamagnetic, paramagnetic or ferromagnetic: in the presence of a magnetic field they were as strongly magnetic as a ferromagnet, but removed from the field they showed no sign of a remanent magnetization. Néel hypothesized that interactions between atoms in these materials caused a spontaneous alignment in which the atomic dipoles perfectly cancelled; he coined the term “antiferromagnetic” to describe them.
In the years leading up to World War II, Néel began to study the magnetic properties of alloys. These are solid-state mixtures of two or more metals, in which the atoms are regularly arranged in a crystal structure. Magnetically, these alloys seemed to lie somewhere between ferromagnets and antiferromagnets, as if within them were two opposite but unequal magnetic alignments. For these Néel introduced a fifth class: ferrimagnetism, with the alloys becoming known as ferrites.
Néel had always had an underlying interest in geomagnetism. At one time in the mid 1930s he had considered taking up the old positions of Brunhes at Clermont-Ferrand but the opportunity had fallen through for lack of funding—alloys were a surer financial bet in hard economic times. In 1939 many of the Strasbourg group did relocate to Clermont-Ferrand, but Néel moved instead to Grenoble, where he would later establish a major international research institution, Centre d’Études Nucléaires de Grenoble (Grenoble Center for Nuclear Studies), and spend the remainder of his long and illustrious career. Early in the war he devised a method to demagnetize the hulls of ships and so protect them from the enemy’s magnetic mines, and personally supervised its application to all major vessels in the French navy. As the war progressed and many other French cities were suffering miserably under German occupation, Néel took several Jewish scientists into his Grenoble laboratory. They would later prove to be invaluable collaborators.
The war over, Néel’s interests returned to ferrimagnetism, particularly the “ferrites,” which by then were being widely investigated with a view to
producing stronger and more stable permanent magnets. This provided the basis for his work on naturally occurring magnetic minerals. The puzzle was how the magnetization gained by lava as it cooled in Earth’s magnetic field came to be so strong and stable compared with the magnetization taken on by a rock that was simply left in a magnetic field of the same strength at room temperature.
Néel’s first discovery was that he could explain the properties of many natural magnetic minerals—magnetite, for example—in the same way as he had his two-metal alloys. The crystal structure of magnetite meant that the iron ions were arranged in two networks of opposite but unequal magnetization. Like the ferrite alloys, magnetite was ferrimagnetic: it carried a strong, stable natural magnetization, and had a characteristic Curie temperature of 585°C.
Néel thought of a rock as containing a small percentage of ferrimagnetic grains, each uniformly magnetized. These grains were randomly distributed, and far enough apart that they did not interact magnetically. Each grain became magnetized when it first cooled through the Curie temperature of its mineral—about 500° to 600°C—but, as Néel showed, permanent, absolutely stable magnetization did not occur.
The stability of a grain’s magnetization depended on the grain’s size and shape, chemical composition, and the temperatures to which it was subsequently subjected. Néel’s calculations showed, for example, that while iron grains up to twelve nanometers in diameter (a nanometer being a millionth of a millimeter) were unstable at room temperature, grains of 16 nanometers or more were, on the average, extremely stable. Collections of such grains might retain their magnetization for hundreds of years: even if the local magnetic field direction changed, the original magnetization would remain.
For natural minerals such as magnetite the figures were even more impressive. They clearly showed that fine-grained volcanic rocks and sediments might easily retain a stable magnetization for hundreds of millions of years—provided they remained at a sufficiently low temperature. If a rock was reheated close to its Curie temperature, the magnetization would become less stable, and likely to be affected by changes in the local magnetic field. The challenge for paleomagnetists would be to gain a deeper understanding of this trade-off between temperature and magnetic stability and to exploit this in unraveling the complexities of Earth’s ancient magnetic field, hidden in its rocks.
Néel’s discoveries laid the cornerstone of rock magnetism and paleomagnetism. Although only eight of his more than 200 publications, almost all in French, were directly concerned with the magnetization of rocks, the most important ones, which appeared between 1949 and 1952, were reviewed in English in 1955 and have since been cited in almost every work of paleomagnetism.
Rock magnetism was just one facet of a long career devoted to understanding magnetism and magnetic materials. When Néel received the Nobel Prize for Physics in 1970, it was noted that his pioneering work on the materials he called “ferrites,” and his discovery of the properties of antiferromagnetism and ferrimagnetism, had been central to the fortunes of all the world’s major electronic companies.
Poles Flipped, Continents Adrift
Measurements of the magnetism of English rocks indicate that in the last one hundred and fifty million years the country as a whole has turned through some forty degrees.
—THE MANCHESTER GUARDIAN, SEPTEMBER 9, 1954
By the late 1940s, research into the history of Earth’s magnetism was regaining momentum in Europe, and active research groups soon sprang up in London, Manchester and Cambridge. Britain’s first contribution to the polarity reversal debate came when a New Zealander, Edwin (Eddy) Robertson, arrived at London’s Imperial College wanting to test a new magnetometer he had designed for field surveys. Robertson headed to the Isle of Mull in western Scotland, from where a swarm of thirty-million-year-old dykes radiate south and east, rather like the Pilanesberg dykes in South Africa.
The British dykes are up to 180 miles in length, but do not outcrop continuously: long sections lie hidden beneath surface rocks and soil. A magnetic survey is often useful in mapping such covered intrusions because they are usually strongly magnetized. To Robertson’s great surprise, although his magnetometer recorded a strong signal it was exactly the opposite of what he had expected: the dykes were reversely magnetized. To be absolutely certain, he and John Bruckshaw, a colleague at Imperial College, collected samples to study in the laboratory. Their experiments confirmed that the dykes were indeed magnetized in the opposite direction to the magnetic field in the north of Britain.
In 1949 Horace Manley, a third member of the Imperial College team who had also become interested in the magnetization of natural materials, and of dykes in particular, wrote an amazingly perceptive review of the first half-century of paleomagnetism. Curiously, although he discussed early work on secular variation at length, he did not mention Brunhes’ discovery of reversely magnetized rocks, nor the later studies of Mercanton or Matyama. However, he was in no doubt as to the cause of the “inversely” magnetized dykes:
Sufficient experiments have now been made to allow only one plausible explanation of this “inverted” magnetization—that the Earth’s magnetic field was itself reversed at the period when the rocks were formed.
There would, he forecast, soon be enough data to document the “history of the terrestrial magnetic field since rocks were solid at the Earth’s surface.”
Manley was well ahead of his time, but his article fired the imagination of a young Dutch scientist who was about to begin research at Cambridge University. Jan Hospers had been assigned the task of correlating Icelandic lava flows according to the strength of their magnetization. This would prove near impossible for all sorts of reasons, but the directions of the magnetizations caught Hospers’ attention. In two separate sequences of lava flows, he found that the younger flows, which he called the “gray phase,” were magnetized in a normal direction, while the underlying older flows were reversed. In a third sequence there were even older lava flows that were again normally magnetized. Although each group of directions was quite scattered, it was obvious that the average directions were opposite one another: roughly parallel and anti-parallel to the present-day magnetic field.
Back at Cambridge, Hospers met up with two people who were to influence not just his studies but his future research in the entire fields of paleo- and geomagnetism. Keith Runcorn had recently arrived from Manchester, where he had just completed a PhD with the physicist Patrick Blackett. Blackett had won the Nobel Prize for Physics in 1948 for his development of the “cloud chamber” and his study of cosmic rays and so was best known for his work in this area of physics, but he also had an interest in Earth’s magnetic field. He firmly believed that planetary magnetism was intrinsically related to rotation.
This had been the subject of Runcorn’s PhD research, but he had failed to prove a link and was looking for fresh avenues of research. He immediately grasped the significance of Hospers’ results and began to provide him with much-needed support, even though he was not Hospers’ supervisor. In particular, he encouraged him to resample the Icelandic lava flows in order to replicate his results, and so validate the sequence of normal and reversed polarity intervals. In no time Runcorn became the nucleus of a young and energetic geomagnetism research group at Cambridge.
The second influence on Hospers was Ronald Aylmer Fisher. Fisher was a renowned statistician, a professor of genetics, and a Fellow of Gonville and Caius College, where he often dined with Runcorn. Several years earlier he had devised a statistical method of analyzing directional datasets, but his work still lay in a drawer, almost forgotten for want of a meaningful application. Now, during discussions with Runcorn over the dining table, Fisher recognized that Hospers’ scattered paleomagnetic data offered a real-life use for his method. He got hold of the data and personally calculated the average directions and the angular equivalents of standard deviation and confidence limits, conferring on them the seal of scientific authenticity.
Hospers gave se
rious consideration to various possible causes of his reversely magnetized lava flows, including the idea that by some quirk of nature certain rocks might become magnetized opposite to the local magnetic field. At the end of the day he rejected all explanations but one: that the polarity of Earth’s magnetic field had flipped, not once but at least twice—from normal to reversed and back to its current normal state.
From the rough estimates of age he had at his disposal, Hospers calculated that about half a million years had passed between polarity changes. However, he could find no intermediate directions. He therefore concluded that the reversal process itself must happen quite rapidly, perhaps taking less than ten thousand years.
Armed with Fisher’s statistics, he went on to argue that the average magnetization directions of his lavas were just what you would expect from Gilbert’s uniformly magnetized Earth or, equivalently, from a dipole at the center of the Earth aligned with the rotation axis—that is, a geocentric axial dipole. Statistically, the average direction of the normally magnetized lavas was indistinguishable from that of a geocentric dipole with its south pole “upwards,” as is the case for the present-day field, while the average direction of the reversely magnetized lavas was indistinguishable from that of a geocentric dipole with its north pole “upwards.”
This “geocentric axial dipole hypothesis”—the proposition that, over a long enough period of time, the average positions of the geomagnetic poles coincide with the geographic poles—was to become a central tenet of geomagnetism. However, its importance and implications were not fully appreciated immediately, any more than was the evidence for polarity reversals.
North Pole, South Pole: The Epic Quest to Solve the Great Mystery of Earth's Magnetism Page 15