The picture was not quite complete, however. The seismologists had one more surprise up their sleeves.
Inge Lehmann was an exceptional woman. Born in 1888, she came from an unusually liberal background that prepared her to hold her own in the male-dominated scientific world of the early twentieth century. As a child in Denmark, Lehmann had attended a remarkably progressive coeducational school run by Hanna Adler, an aunt of the famous physicist Niels Bohr. Here all pupils, boys and girls, were taught together and studied the same subjects: reading, writing, arithmetic, football, cookery and needlework. There was no segregation or discrimination.
In 1925, several years after graduating from the University of Copenhagen with a master’s degree in mathematics, Lehmann had begun work on earthquakes and established a network of seismological stations in Denmark and Greenland. By 1928 she was chief seismologist at the Royal Danish Geodetic Institute and had access to data from stations all over northern Europe.
In 1929 a magnitude 7.8 earthquake hit Murchison, a tiny settlement in the South Island of New Zealand. Lehmann’s network of observatories covered a large part of the shadow zone, where no seismic waves were expected. Nevertheless, Lehmann detected faint P wave arrivals from Murchison. How could this be? There was no going back on the conclusion that the core was liquid—more and more evidence was accumulating from disciplines other than seismology. The only possible answer was that there was something inside the liquid core—something with a higher seismic wave speed—that had reflected back the P waves so they eventually emerged within the shadow zone.
Matching Weichert’s boldness, Lehmann now proposed that Earth had an additional inner core. She went further. Assuming uniform P wave speeds of ten kilometers per second in the mantle and eight kilometers in the outer core, and adjusting the radius and the P wave speed in her inner core to fit the observed arrival times, she estimated that the inner core had a radius of 1400 kilometers and a P wave speed of ten kilometers per second. She named these core waves P′ (P -prime), and her original paper on their discovery is entitled simply “P′.”
The discovery that not only did Earth have an iron-rich core, but the outer part of this core was liquid, and so both mobile and electrically conductive, opened up a new era of geomagnetism. Suddenly, permanent magnetism was no longer needed to explain Earth’s magnetic field: electric currents flowing in the metallic core might be generating the field. But if this were the case, why would there be electric currents in the core in the first place?
Seismic ray paths through the Earth. The detection of weak P wave arrivals in the shadow zone led Lehmann to conclude there is an inner core in which the seismic waves travel faster so they are eventually refracted into the shadow zone.
The main features of Earth’s internal structure: a solid iron-rich inner core, a liquid iron-rich outer core, a rocky mantle, and a rigid but brittle crust.
Faraday’s electromagnetic induction provided a clue. If, for some reason, the conducting core fluid moved through a magnetic field, electric currents would be induced in it, in exactly the same way as Faraday had induced currents in his early experiments. These electric currents would, in turn, produce magnetic fields. In the right circumstances, these secondary, induced fields might reinforce the original field. In other words, as long as energy was available to drive the original fluid motion the process could be self-sustaining.
To Gilbert and Halley, even to Faraday and Maxwell, the idea of such a “magnetohydrodynamic” process operating inside the Earth would have been totally unimaginable. By the early twentieth century it was just barely credible. However, by the time the dynamo theorists got seriously down to work a few decades later, several amazing new discoveries about the history of Earth’s magnetic field were to make their task even harder.
Reading the Rocks
Intuitively … one regards the Earth’s magnetic field as an inherent property of such grandeur that reversals of its sense are difficult to grant.
—JOHN GRAHAM, 1952
Scientists continually seek new and better observations so they can create ever more detailed and accurate explanations of the natural phenomena they study. This is never truer than of geomagnetists. Over time, as more observations accumulated of Earth’s magnetic field, a pattern of increasingly complex variations was added to Gilbert’s original concept of a geocentric axial dipole.
By 1900 it had become apparent that secular variation—the gradual changes in direction and intensity of Earth’s magnetic field occurring year by year, decade by decade—was a persistent feature. The 300 years of declination and inclination measurements collected at London and Paris seemed to trace out two-thirds of a loop. It was tempting to speculate that the loop would complete itself in another 150 years, and perhaps even repeat itself again in the next 400 to 500 years. Was there a regular cycle to this?
Scientists are also notoriously impatient, and 150 years was far too long to wait for an answer. What was needed was a prehistoric record so they could find out what had happened in the past.
It was Joseph Fournet, the inaugural professor of geology at the University of Lyons in France, who first suggested that magnetized rocks might hold the answer, but the beginning of paleomagnetism is usually credited to his contemporary Achille Delesse, who held the chair of mineralogy at nearby Besançon. In 1849 Delesse reported that certain volcanic lava flows were magnetized—not strongly and randomly like von Humboldt’s lightning-stricken rocks, but regularly, if more weakly, in the direction of Earth’s magnetic field. Even when removed from their natural position and orientation, samples of the lavas retained a stable magnetization.
A few years later an Italian physicist called Macedonio Melloni showed that lava flows from Mount Vesuvius and the Campi Flegrei (Phlegraean Fields) area west of Naples were also magnetized parallel to Earth’s magnetic field. Melloni took samples of the lavas back to his laboratory, heated them until they glowed red and then cooled them in the laboratory’s magnetic field. He found they lost their original magnetization on heating, but became remagnetized parallel to the laboratory field when cooled again. This confirmed his supposition that the lavas’ original magnetization had been acquired during their initial cooling in Earth’s magnetic field following the eruption of Mount Vesuvius.
This discovery of fossil magnetic records preserved in lava flows seems to have then gone largely unnoticed for nearly fifty years. In the meantime another Italian, Giuseppe Folgheraiter, began to wonder whether archaeological artifacts such as pottery and bricks, which had been fired by humans rather than nature and had cooled in the local magnetic field, might carry a stable magnetization. If so, they could be just what was needed to extend the record of secular variation backwards in time. One feature that attracted Folgerhaiter to such artifacts was that they could often be dated quite accurately from their style and decoration, and sometimes from documentary records associated with them.
Folgheraiter was not the first to recognize the magnetization of bricks and clay objects. As early as 1690 Robert Boyle, a contemporary of Newton and Halley, had found that a brick heated in a fire and then cooled became magnetized in the direction of the surrounding magnetic field, while in the 1860s another Italian, Gheradi, had made a systematic study of Egyptian and Italian pots and demonstrated the stability of their magnetization. Gheradi had also shown that the magnetization of certain building bricks could affect sensitive magnetic measurements, and warned against the use of such bricks in the construction of geomagnetic observatories.
Folgheraiter’s work, published in 1894, involved sampling figured red-clay Greek vases and urns that dated back to the sixth century BC. Folgheraiter reasoned that if the orientation of a vase in a firing kiln were known, the magnetization it had acquired on cooling should give a record of the geomagnetic field at the time of firing. He sampled and examined vases of different ages, and drew up an inclination record for the period 800 BC to 100 AD.
Intriguingly, the earliest of Folgheraiter’s results s
howed a negative inclination, suggesting an anomalous or curious field direction at the time of the vases’ magnetization. Scientists still disagree as to whether Folgheraiter mistook the orientation in which the pots were fired and so got an invalid result, or whether the geomagnetic field really was anomalous at the time and location of their firing.
Whatever the truth, these pioneering archaeomagnetic and paleomagnetic studies set the scene for some startling discoveries in the twentieth century. Almost immediately, Folgheraiter’s success on man-made clay artifacts inspired Bernard Brunhes, a brilliant young French physics professor, to search out clays that had been baked by natural processes. In the early 1900s he and his assistant, Pierre David, went to the Massif Central region of France looking for layers of clay that lay directly under ancient lava flows. They reckoned that not only would the clay have been baked by the hot lava flow during the eruption, but both clay and lava should have then cooled together and so carry the same record of the geomagnetic field direction at the time.
Brunhes and David sampled a number of lava flows and their underlying baked clays. For the most part the results met their expectations. However, one result really caught their attention. It came from a site at Pontfarein, near the town of Saint-Flour, where road-building authorities had made a hundred-meter-long road cut. Brunhes and David had collected several samples of the baked clay from there, but only two of the basalt lava flow as it had been particularly hard and difficult to cut. One of these lava samples turned out to have been struck by lightning and so was useless, but the magnetic direction of the other agreed very well with the average result from the baked clay samples.
However, there was something very strange. The declinations Brunhes and David obtained from the clays and lava were 148° and 154° respectively, while the inclinations were −74° and −76°. The magnetizations of the rocks pointed approximately to the southeast and angled upwards, while the geomagnetic field at Pontfarein, as at most locations in the northern hemisphere, was roughly northwards and inclined down. In other words, the rocks were magnetized in almost the opposite direction to the local magnetic field.
Brunhes and David were astounded. They could hardly believe their results, but the consistency was too good to occur by chance. In 1906 Brunhes nervously announced to the world:
Bernard Brunhes, who, with his assistant Pierre David, discovered rocks at Pontfarein that were magnetized in the opposite direction to the local magnetic field. They deduced that central France had once been closer to the south magnetic pole than to the north—a concept that was to confound the scientific world for the next fifty years.
…qu’en un moment de l’époque Miocène aux environs de Saint-Flour, le pôle Nord était dirigé vers le haut: c’est le pôle Sud de la Terre qui était le plus voisin de la France centrale.
(… at a moment in the Miocene epoch, in the region of Saint-Flour, the north pole was directed upwards; it is the Earth’s south pole that was closest to central France.)
The obvious interpretation was that at some “moment” about six million years ago Earth’s magnetic field had been upside-down. The idea was so surprising that fifty years later scientists would still be struggling to understand if such a thing could happen.
Curiously, neither Brunhes nor David published any further on this subject. As well as his job as professor of physics at the nearby University of Clermont-Ferrand, Brunhes’ time was taken up rebuilding the observatory at Puy-de-Dôme, where he had been director since 1900. He expanded the observatory considerably, and extended its original meteorological portfolio to include both geomagnetic and seismological observations. One night in 1910 the local gendarmerie came across a man lying unconscious on a street in Clermont. Thinking he was the victim of an assault but finding no obvious injuries on his body, they searched him, and on discovering his identity carried him home. Bernard Brunhes died the next day, probably from a brain hemorrhage. He was just forty-three.
Almost a hundred years after Brunhes’ discovery of reversely magnetized rocks, three French paleomagnetists—Carlo Laj (supposedly a distant relative of Brunhes), Catherine Kissel and Hervé Guillou—would relocate and resample the Pontfarein section. Using modern laboratory techniques, they would obtain magnetization directions for both the lava flow and the baked clay that were indistinguishable from Brunhes’ results. In addition, they would date the lava at six million years.
Despite Brunhes’ stunning results, it would take a long time for the geophysical community to be convinced that the polarity of the Earth’s magnetic field had reversed. French physicist Paul Mercanton was the first to continue Brunhes’ work. Mercanton reasoned that if the main, dipole part of the geomagnetic field really had reversed its polarity, reversely magnetized rocks should be found all round the world, and he set out to look for them. Beginning in 1910, he published a series of papers in which he described both “normal” and “reversed” directions of magnetization from lava flows estimated to be up to sixty-five million years old in Spitsbergen, Greenland, Iceland and the Scottish islands. Mercanton cautiously noted that his results supported the theory of geomagnetic field reversals, and urged other scientists to continue the search for reversely magnetized rocks, particularly in the southern hemisphere. He went to the extent of acquiring some lava samples from Australia, but the results were disappointingly inconclusive. Nevertheless he ended his 1926 paper with an idea that would prove to be both inspired and inspiring:
… if a link exists between the rotation axis of our globe and its magnetic axis, the considerable displacements of the magnetic axis which our researches would discover would unexpectedly corroborate the large displacements of the axis of rotation argued for by A. Wegener.
From this time on, the controversy over geomagnetic reversals would become entwined with another equally contentious theory. As early as 1596 a Dutch mapmaker named Abraham Ortelius had observed that there was a remarkable fit between the coastlines of Africa and South America. Europe and Africa, he suggested, had been “torn away” from America by great earthquakes and floods.
The idea resurfaced in the early nineteenth century, courtesy of Alexander von Humboldt among others, but it was not until 1912 that a German meteorologist, Alfred Wegener, formally put forward the theory of what he called “continental drift.” In addition to the way the coastlines matched, Wegener pointed out that various fossil species were to be found on both sides of the oceans, as were similar rock formations and geological structures. There was also evidence of huge-scale climate change: fossils of tropical plants had been found in Antarctica and evidence of glaciation could be seen in South Africa.
Wegener extrapolated back to a time around 200 million years ago when, he proposed, all the continents had been joined together in one mega-continent. He named this Pangaea from pangaia, a Greek word meaning “all Earth.” In 1937 a South African geologist, Alexander Du Toit, further refined the thesis by suggesting that Pangaea had first split in two, forming Gondwana in the south and Laurasia in the north, and that after drifting apart these two super-continents had further broken up into the modern continents.
These twin theories—geomagnetic reversals and continental drift— now sat simmering side by side on the back-burners of science for several decades. Each slowly gathered support but also suffered setbacks. The problem was that no one could come up with viable explanations as to why either phenomenon should occur. And given that both theories were wildly at odds with conventional wisdom, it was hard for many people to take them seriously at all.
Mercanton’s work was eventually noticed and followed up by a Japanese paleomagnetist. Motonori Matuyama studied more than one hundred lava flows from Japan, Korea and Manchuria, and made a significant finding: all his youngest lava flows were normally magnetized, while his reversely magnetized samples all came from underlying, older flows.
Two streams of paleomagnetic research continued through the Depression years of the 1930s but neither made much progress on the mystifying polarity
reversals. Instead they aimed at getting a better picture of secular variation. In France, Émile and Odette Thellier concentrated on fired archaeological artifacts, trying to extract information on the intensity of the paleomagnetic field, as well as its direction. By showing that the strength of the magnetization acquired by an artifact as it cooled after firing depended directly on the strength of the local magnetic field, they were able to estimate changes in the strength of the ancient magnetic field.
Meanwhile, in the United States the pace was also picking up. The Carnegie Institution’s Department of Terrestrial Magnetism in Washington, DC was already well known for its study of historical observations of the magnetic field. Now a paleomagnetism group emerged, and members busied themselves investigating whether sediments that had gradually accumulated on the seafloor might carry a continuous record of changes in the direction of the geomagnetic field—something like a natural tape recording.
In 1940 Sydney Chapman of Imperial College, London and Julius Bartels of the Potsdam Geophysical Institute published a landmark two-volume textbook, Geomagnetism. Curiously, though, paleomagnetism rated scarcely a page: the focus was almost entirely on direct observations of Earth’s magnetic field and their interpretation, and there was no mention of the work of Brunhes, Mercanton or Matuyama. However, Chapman and Bartels did describe in detail the case of the Pilanesberg dyke system in the Transvaal region of South Africa.
Dykes are intrusions of lava that have squeezed into fissures or faults in older rocks. Fourteen or more almost vertical dykes, each about 160 kilometers long and reckoned to be a hundred million years old, radiated from a point about 120 kilometers northwest of Pretoria. When a magnetic field survey had been conducted over each dyke, a strong positive anomaly had been recorded. A positive anomaly could only mean that the dykes were permanently magnetized in the opposite direction to the geomagnetic field.
North Pole, South Pole: The Epic Quest to Solve the Great Mystery of Earth's Magnetism Page 14