Vine was not finished yet. He next pieced together magnetic anomaly profiles extending 3600 kilometers across the Pacific Ocean at a latitude of about 41° N, from the crest of the ridge at 127.5° W right out to 170° W, and estimated they represented a total of eighty million years of seafloor history, and an eighty-million-year record of changes in the polarity of Earth’s magnetic field.
The stage was now set for the next big step. Who was going to take up the challenge? It would not be Vine. Working alongside Hess at Princeton, his interests were moving towards the geological implications of Wilson’s plate tectonics. As chance had it, it would be Heirtzler—Pitman and Herron’s mentor at the Lamont Geological Observatory—who stepped forward, but not before the seafloor had turned up yet another surprise to corroborate the evidence of the marine magnetic anomaly records.
As well as its potential length, the marine magnetic anomaly record had another advantage over lava flows dated through radioactive decay—continuity. With discrete, individually dated lava flows there was always the possibility of errors in ordering, and often gaps remained in the timescale where no lavas had been found. For example, the Gauss-Matuyama reversal was poorly dated for years, simply because no lavas of the right age had been studied. The seafloor timescale filled such gaps and also identified several short polarity “events” that had not been found in lava-flow studies.
The mid 1960s saw the launch of a coordinated international program of deep-sea drilling, and the first long cores of sediments from the seafloor began to come ashore. The process by which sediments become magnetized is quite different from the way magnetization is locked into volcanic rocks as they cool. With sediments, the tiny grains are already magnetized when they are washed into the ocean, and as they fall through the water they align, like a compass needle, with Earth’s magnetic field. Eventually their orientation becomes locked by the surrounding grains as the sediment compacts.
Chris Harrison, a student at Cambridge, and Neil Opdyke, working along the corridor from Heirtzler at the Lamont Geological Observatory, were among the first paleomagnetists to study the records of magnetization in these cores. In 1966 Opdyke published results from seven cores of deep-sea sediment, each between four and twelve meters long, which had been collected around Antarctica. All were normally magnetized at the top, and reversely magnetized below a depth that varied from one to five meters. Some of the cores showed one or two normal events in this reversed interval, while all but one were predominantly normally magnetized again at their base. Marine microfossils, the remains of tiny bugs that lived on or near the seafloor, were used to align the cores with each other chronologically. As these bugs had evolved relatively rapidly, they provided a much better indication of age than the macrofossils traditionally used by hard-rock geologists to date older consolidated rocks.
When the cores were correlated on the basis of the microfossil changes, the correspondence of the polarity epochs and events jumped out immediately: despite differences in sedimentation rate, the cores carried the same continuous magnetic records through the Brunhes, Matuyama and part of the Gauss epochs. These sedimentary records provided valuable independent validation of the geomagnetic polarity timescale of Vine, Pitman and Heirtzler.
By now Heirtzler was an avid supporter of the Vine-Matthews-Morley hypothesis—and he was in an ideal position to extend the marine magnetic anomaly record to the very edges of the oceans, and the polarity timescale with it. Over the next two years he and his coworkers sailed the world collecting new long, high-quality magnetic anomaly records and decoded the history of the seafloor. In 1968 their results appeared in four papers published in a single issue of the Journal of Geophysical Research. Each of the new profiles was between 1500 and 3000 kilometers long and showed the now familiar sequences of wiggles, greatest closest to the spreading ridges, and gradually diminishing but remaining recognizable through their entire lengths.
The records showed that field reversals had been going on for at least the past eighty million years. To start with, Heirtzler and his team matched the first few anomalies of each profile to the first three-and-a-half million years of Doell and Dalrymple’s polarity timescale—back to the Gilbert-Gauss transition. From this match they calculated a spreading rate for each profile. They then took a blind leap in the dark—the only real option open to them. On the assumption that each ridge system had spread at a constant rate throughout its entire history, they worked out the sequence of polarity reversals that best fitted each observed profile.
Astonishingly, the pattern of reversals—the black-and-white magnetic barcode—was the same in every case: the Pacific, Atlantic and Indian Ocean profiles yielded virtually identical patterns of normally and reversely magnetized seafloor. The reversal process was certainly global—only reversals of Gilbert’s “magnusmagnes” could cause simultaneous changes at all these sites.
Of course, there were some discrepancies. Extrapolation is a notoriously hazardous procedure, even when dealing with well-studied physical processes, and the mechanism of seafloor spreading was neither well studied nor well understood. There was little reason to suppose spreading rates were constant in the long term. Nonetheless, only five years after the publication of Cox’s first polarity timescale—which consisted of three polarity intervals over the past three million years, and was based on just nine dated lava flows—Heirtzler and his coworkers had produced a continuous history of reversals going back eighty million years. This was still less than one-fiftieth of the age of the Earth, but there was a wealth of new information to challenge the theorists.
In the next forty years there would be big improvements in methods of dating rocks and sediments, and the accuracy and precision of the geomagnetic polarity timescale would undergo significant fine-tuning. Nevertheless, the basic features would remain unchanged.
First, the eighty-million-year-long “barcode” showed there had been 183 polarity reversals. In other words, Earth’s magnetic poles had flipped, on average, every 430,000 years. Hospers” original estimate of half a million years had been remarkably close to the mark.
Second, the field’s polarity had been surprisingly evenly divided between normal (49 percent of the time) and reversed (51 percent). However, the frequency of reversals had been anything but uniform: polarity intervals had ranged from 10,000 years to five million, with most being less than 500,000 years. The median length was just 230,000 years.
A modern version of the geomagnetic polarity timescale for the past 80 million years, based on the 1995 work of Steve Cande and Dennis Kent at Columbia University’s Lamont-Doherty Geological Observatory. The essential features are very similar to those of the original timescale that James Heirtzler and his colleagues deduced from seafloor magnetic anomalies and published in 1968. The conventional division into “chrons” and intervals of the geological timescale are shown to the right.
There had been many more short intervals than long ones. Only twenty-one of the 183 intervals had lasted more than one million years; only twenty-five had been longer than the present Brunhes epoch. Paleomagnetists now suspect that much longer intervals of stable polarity, for which there is little or no seafloor data, occurred earlier in the planet’s history. Other studies have shown, for example, that for most of the early Cretaceous period, from 83 to 118 million years ago, Earth’s magnetic field was locked in a normal configuration, and it is suspected that during the Permian—225 to 280 million years ago—the polarity remained reversed for at least fifty million years, the so-called Kiaman Magnetic Interval.
It also seems that over a long timescale the frequency of reversals varies. Between forty and eighty million years ago there were only forty reversals—one per million years on average. However, in the past forty million years there have been 143 reversals, more than three times as many. This could be an artificial feature of the recording or preservation of the magnetization on the seafloor, or of the resolution of the measurements as the ship taking them moved away from the mid-ocean ridge int
o deeper water. However, geomagnetists don’t think so: they believe it reflects some deep-seated change in the process by which the geomagnetic field is generated.
When it became clear that there had been a huge number of reversals, Cox’s naming system based on names and locations, although nostalgic, seemed totally inadequate. In 1965, Walter Pitman had started to number the marine magnetic anomalies, and eventually the governing organization, the International Association for Geomagnetism and Aeronomy, recommended a new system based on an extension of this. The main polarity intervals were to be called “chrons.” They would be numbered sequentially backwards from the present, and each would have a normal (N) and reversed (R) part. Thus, the Brunhes epoch became C1N, the upper part of the Matuyama C1R and so on. Shorter events, such as the Jaramillo, became “subchrons,” and very long intervals of stable polarity, like the Kiaman, became “superchrons.”
Was there any way of telling what happened during the process when Earth’s magnetism reversed? Did the dipole simply tumble from one orientation to the other, or was the whole affair more complicated? The seafloor had provided few clues, other than that the process was too fast to be deciphered from the existing generation of marine records. Could the answer lie in deep-sea sediment cores, exposed sequences of sediments, or piles of lava flows? All were amenable to detailed sampling and the samples could be measured in the laboratory.
The first major finding from these studies was that the strength of the magnetic field dropped markedly—to about ten percent of its usual intensity—before its direction started to change. The declination and inclination then took about 5000 years to reverse before the field’s strength recovered. The whole process seemed to take about 10,000 years.
It would prove harder to get consensus on how the direction changed. Because intensity was low in the transition periods, the magnetization, especially of sediments, was weak and difficult to measure precisely. Nevertheless, differences in the records of deepsea cores coming from the various ocean basins soon showed that a simple tumbling dipole, however weak, was not the answer. When Earth’s stable field weakened, the field that remained during the next five thousand years of a transition was a much more complicated one. As the theorists continued to grapple with ideas about the overall source of the field, these discoveries of the marine geophysicists and paleomagnetists had made their task even more complex.
In Search of a Solution
The mechanisms behind the magnetic field and behind the reversals are still really mysterious … one of the grand intellectual challenges … in all of the physical sciences.
—RAYMOND JEANLOZ, 1996
At the beginning of the twentieth century, despite hundreds of years of thought, speculation and experimentation, there was still no good idea of the origin of Earth’s magnetic field. Even Albert Einstein reportedly declared it one of the greatest unsolved problems in physics.
In 1269 when Petrus Peregrinus had penned his Epistola and introduced the concept of the magnetic poles, he was not concerned whether the source of their “virtue” lay within the Earth, on its surface, or in the heavens: it was enough that the magnetic poles and celestial poles shared a common axis.
A little over three centuries later William Gilbert had declared the Earth to be magnus magnes —a great magnet—and so firmly rooted the origin of the magnetic field within the planet itself. He imagined that Earth carried a static, permanent magnetization “innate in all parts thereof and diffused throughout.” Gauss had then used brilliant mathematical analysis to describe the main field in detail. He had clearly demonstrated that its source was internal, but had not probed much further into its origins.
If it were true that the Earth was uniformly magnetized throughout, it took only a simple calculation to show that the strength of that magnetization had to be some ten times that of the most strongly magnetized rocks found at the surface. By the nine - teenth century, when William Hopkins and others were arguing about the make-up of the planet’s interior, it was well known that temperature rose rapidly with depth. Below about twenty-five kilometers the temperature was, in fact, above the Curie temperatures of all known natural magnetic minerals, and so permanent magnetization must be all but impossible. But if only the outermost twenty-five kilometers of Earth were magnetized, the intensity of this magnetization would be unbelievable—greater than the strongest of today’s super-magnets.
On top of this, secular variation, the gradual changes in Earth’s magnetic field, had been discovered shortly after Gilbert’s publication of De Magnete, and it was difficult to incorporate this into any theory based on permanent magnetization. Clearly then some other, dynamic source was at work.
To make things even more complicated, by the beginning of the twentieth century considerable effort had gone into studying the effects of the sun on a compass needle. It appeared the sun played some role in both the regular daily fluctuations—the tiny, rapid changes first noticed by London instrument-maker George Graham—and in the occasional magnetic storms that had intrigued Alexander von Humboldt in the early nineteenth century.
The confusion of ideas was aptly described in a 1903 textbook, The Realm of Nature, by H.R. Mill, a lecturer at Oxford University:
Thus it appears that while the Earth’s magnetism resides in the massive rocks of its crust, and is probably produced and maintained by the Earth’s rotation, the sun’s energy exercises a regulating or disturbing influence upon it.
The role of the sun, superimposed on the bigger picture of the main, internal field, was certainly a complication. However, it would be the discovery at the beginning of the twentieth century that the sun, too, had a magnetic field that would precipitate the next generation of ideas about Earth’s magnetism.
Following James Clerk Maxwell’s revelation that light was an electromagnetic wave, many physicists had studied the wavelengths or “lines” in the spectra of light emitted by the atoms of certain gases, and recognized the same wavelengths in the light radiated by the sun and other stars. A Dutch physicist, Pieter Zeeman, had shown that in a magnetic field a characteristic spectral line might be split into two or more lines of slightly different wavelengths. In 1908 George Ellery Hale, an American astronomer, would spot this “Zeeman splitting” in light emitted by hydrogen atoms in the atmosphere of the sun—particularly in light coming from sunspots which, being darker, were assumed to be cooler than the rest of the sun. It seemed that sunspots, as well as being cool, were also strongly magnetic.
Further studies would show that the sunspots occurred in pairs of opposite polarity: the magnetic field lines left the surface of the sun through one member of a pair and re-entered through the other.
In 1848 the German pharmacist and astronomer Samuel Heinrich Schwabe had discovered that sunspots varied in number in eleven-year cycles. It now seemed that the eleven-year sunspot cycle involved not just the number of the sunspots, but also their magnetic polarities. With this discovery, the magnetism of the sun suddenly became a hot topic of research. For a time it would almost eclipse the problem of Earth’s magnetic field.
By now two sources of magnetism were known—permanent magnets, which had been recognized for at least two millennia, and electric currents, the magnetic properties of which had been discovered by Ørsted the previous century. But a new idea was emerging among atomic physicists: perhaps magnetism was intrinsically associated with rotation.
In 1912 Arthur Schuster, a physicist who had worked under Maxwell in the Cavendish Laboratory at Cambridge and later become professor of physics at Manchester University, reviewed all three as possible sources of Earth’s magnetic field. Surprisingly, Schuster was reluctant to completely rule out permanent magnetism, arguing that although magnetization of the Earth’s crust alone could not account for the observed field, it was not known what happened to Curie temperatures at the extreme conditions likely in the core, and so he kept open the possibility of more deep-seated permanent magnetization.
He was, however, quick to reject
electric currents as a possible cause. For electric currents to exist in the Earth, he pointed out, they would have to be either maintained by permanent electromotive forces, or be the decaying remnant of currents that had existed since the Earth’s formation. He could conceive of no driving mechanism for a permanent electromotive force, and argued that for the remnant of a primordial current to still exist the current must originally have been enormous. The origin of such a current, he said, did not bear examination, adding that whereas “we must keep our minds open to the possibility that the iron contained within the Earth is magnetizable … the difficulties which stand in the way of basing terrestrial magnetism on electric currents inside the Earth are insurmountable.”
So what of the third option? Quantum physicists had recently discovered a connection between the spin, or angular momentum, of an elementary particle such as an electron, and its magnetic moment. The same was true for an atomic nucleus. From this had sprung the idea that magnetism might be intrinsically associated with rotation. Like the force of gravity, such magnetism might be difficult to detect on an everyday scale, but in bodies of astronomical proportions could it be significant? Schuster considered various ways this might come about but reached no firm conclusion.
The idea of rotation as the underlying cause of Earth’s magnetic field languished. However, a new idea was about to burst on to the scene. Its instigator, Joseph Larmor, had been born in Northern Ireland and studied at Queen’s University, Belfast and St. John’s College, Cambridge before becoming Lucasian professor of mathematics at Cambridge in 1903. In his studies, Larmor was to forge a curious link between the classical physics of the nineteenth century and the quantum physics and relativity that sprang up at the beginning of the twentieth.
North Pole, South Pole: The Epic Quest to Solve the Great Mystery of Earth's Magnetism Page 19