Keith Runcorn continued to be bugged by such questions, and began to consider mechanisms for the generation of Earth’s magnetic field and for continental drift. Meanwhile, Ted Irving— who had now moved to Canberra and set up his own laboratory at the Australian National University—and Ken Creer addressed the obvious need to construct polar wander paths (or apparent polar wander paths, as they soon became known) for the other continents.
While British scientists were now convinced about polarity reversals, there were still doubters. As late as 1963, when the question was put to the vote at a meeting of twenty-eight eminent paleomagnetists in Munich, four reckoned that all instances of reversed magnetization were caused by self-reversal, fourteen thought that although self-reversal did occur there was sufficient evidence to believe that some records represented field reversal, and ten could not make up their minds. Not one voted unerringly for polarity reversals.
The problem was that while there was evidence for self-reversal—the Haruna lavas had a reversed natural magnetization and clearly showed self-reversing properties in the laboratory— there was no such direct proof of field reversal. Rather, the evidence was circumstantial. To unambiguously reject the field reversal theory, it would be necessary to demonstrate that all rocks carrying reversed magnetizations were capable of self-reversal, while none of the normally magnetized ones were. The accumulating evidence indicated that roughly half of all rocks studied were reversely magnetized, but examples of the peculiar minerals and crystal structures necessary for self-reversal remained extremely few and far between.
Meanwhile, indirect evidence for field reversals was gaining strength. First, although dating rocks was not easy—for sediments, it was traditionally based on identification of fossils, and for lavas on estimates of the frequency of eruptions—there was growing evidence that all across the world rocks of the same age recorded the same polarity. A timescale of geomagnetic polarity reversals was beginning to emerge. It was hard to come up with a reason for self-reversing and non-self-reversing rocks to occur in chronologically controlled sequences.
Secondly, there were a growing number of cases in which the directions of magnetization of lavas and the baked clays underlying them were the same. A rather special example was reported in 1962 by Rod Wilson, a geologist at Liverpool University’s Department of Earth Sciences. Wilson had found an example of sediment that had been baked not once, but twice. It had first been overlain by a lava flow and its upper margin remagnetized in a reversed direction, which was also recorded in the lava. Subsequently, a dyke had pushed up through both the sediment and the lava. The dyke was also reversely magnetized, but in a direction that, because of secular variation, differed measurably from that of the lava. The parts of the sediment immediately adjacent to the dyke had been heated by it and remagnetized. Some parts that were close to both the lava and the dyke had been heated and remagnetized twice, but not necessarily to the same extent each time.
By carefully heating his samples to successively increasing temperatures, cooling them in the absence of a magnetic field, and then measuring their remaining magnetization, Wilson had been able to separate out the components of magnetization carried by grains of differing stability. This, in principle, would unravel the entire magnetic history of his rocks.
This study was too much for almost all the remaining adherents of self-reversal. Even if one in a thousand rocks was capable of self-reversal, the chance of finding one self-reversing sediment, one self-reversing lava and one self-reversing dyke in such close proximity was one in a thousand million—not quite impossible, but very nearly.
The case for polarity reversals was now almost proven. In just a few years, even more compelling evidence was to appear for both polarity reversals and continental drift—and from another surprising source.
The Story on the Seafloor
From this moment I never had any doubts about the concept. It locked three theories together in a mutually supporting way: continental drift, ocean floor spreading, and the periodic reversing of the Earth’s field.
—LAWRENCE MORLEY, 1986
Attempts to survey the depths of the world’s oceans go back to the 1850s, when the first transatlantic telegraph cable was laid between Ireland and Newfoundland. Contrary to previous assumptions that the seafloor must be flat and featureless, it was found the cable had to traverse a range of underwater mountains that rose 2000 meters above the seabed. These original bathymetric surveys were conducted by dropping a lead sinker down to the seabed on the end of a sounding line. Such work was laborious, tedious and prone to errors. To chart the entire ocean floor in this way would have been totally impractical. Clearly a more efficient method was needed.
During World Wars I and II the Allied nations were terrorized by German submarines. Winston Churchill is said to have admitted that the only thing that really frightened him was the U-boat threat. Not surprisingly, then, considerable British and American efforts went into devising methods of detecting the steel submarine hulls—efforts that resulted first in ultrasonic echo-sounding techniques, and later in airborne and rugged seagoing magnetometers.
Sonar detection technology evolved during World War I from the pioneering work of, among others, a French physicist, Paul Langevin. This technology works by continuously timing the reflection of high-frequency sound waves from the seafloor or other objects: since the speed at which the waves travel in water is known, the distance to the seabed or object can be calculated. By the 1950s a detailed picture of the ocean floors was emerging from sonar surveys. It was discovered that the mid-Atlantic Ridge, traversed by the first telegraph cable, was just part of a 40,000-kilometer-long chain of underwater mountains running around the globe, rather like the seam around a tennis ball. The chain snakes down the Atlantic, through the southern oceans, and up the eastern part of the Pacific. The average height of the mountains above the surrounding sea floor is about 4500 meters, and their flanks spread out to a width of 1000 kilometers or more.
The ridges—or rises as they are sometimes called—had several notable features, clues to the mechanism of continental drift. First, most ridges had a distinct trough or valley along the top. Second, close to these ridges scientists found an unusually high rate of heat flowing up from the Earth’s interior. Third, when drill cores were retrieved from the ocean floor the covering layer of sediment was found to be much thinner than expected, particularly near the ridges.
Fourth, and most surprisingly, the whole of the seafloor was found to be geologically very young. Radiometric methods of dating rocks were in their infancy, but early results showed that whereas the oldest continental rocks were several billion years old, nothing on the seafloor was older than about 150 million years: the oldest seafloor had existed for only a tiny fraction of the lifetime of the Earth.
And lastly, the mid-ocean ridges frequently jinked to the right or left along well-defined lines, or fracture zones, also delineated by submarine mountains.
Furthermore, in other places far removed from the ridges, most notably around the margins of the Pacific Ocean, there were deep trenches that also extended many thousands of kilometers. Seismologists were beginning to notice that a high proportion of the earthquakes they recorded, as well as much of the world’s volcanic activity, was concentrated along these ocean trenches and ridges. Clearly both played an important role in the process of continental drift—but how?
By 1962 two men had come up with the seed of an explanation. Harold Hess was a Princeton University geology professor who had served in the United States navy during World War II and Robert Dietz was a scientist in the navy’s electronics laboratory. The anomalously high heat flow at the mid-ocean ridges had set both of them wondering about the source of that heat deep within the Earth, and they had recalled the idea of convection—heat carried by the movement of fluid material.
The notion of convection in the Earth went back to the 1830s, when Cambridge mathematician William Hopkins had considered the possibility of fluid layers wit
hin the planet. Hopkins was thinking of the sort of rapid convective circulation commonly seen in a pot of soup on a stove. Hot material at the bottom of the pot expands, becomes buoyant with respect to the cooler, denser fluid above, and therefore wells up. When it reaches the surface it loses heat and flows sideways, until it becomes cool enough and dense enough to sink again, thus completing a cycle, or “convection cell.”
Convection in Earth’s interior. As the mantle is not quite solid, hot material rises up under mid-ocean ridges and cooler material sinks at ocean trenches. Harold Hess and Robert Dietz suggested that, carried along by this convection, the oceanic crust spreads symmetrically away from the ocean ridges and is eventually drawn underneath the abutting continental crust.
In 1944 the idea of very slow convection in a much more viscous mantle had been revived by a Scottish geologist, Arthur Holmes, in an early attempt to explain continental drift. But at the time Holmes had found himself a lone “Drifter,” a small voice in a crowd of “Fixists.”
Hess and Dietz now suggested that the upwelling from this slow mantle convection reached the Earth’s surface at the midocean ridges. Where the measured heat flow was highest, they said, magma must be rising through the mantle, erupting and forming new crust. As new crust formed at the ridges, older crust moved away in both directions, riding on the thermally driven convective motion of the upper mantle. Furthermore, beyond the edges of the oceans, the continental crust was also riding high on the denser mantle.
Finally, the formation of new crust at the ocean ridges was being accompanied by the sucking down of old crust back into the mantle at the ocean trenches. Earthquake locations, Hess and Dietz suggested, delineated a downward-sloping surface at each ocean trench, on which the old seafloor slid beneath the abutting continent, creating the downgoing part of the mantle convection cell. Dietz introduced the term “seafloor spreading,” which has been used ever since.
Although Hess and Dietz’s ideas were, by the men’s own admission, simply intuitive starting points, they fitted well with the geophysical observations and provided a credible mechanism for continental drift. Thirty years earlier Harold Jeffreys had complained that Alfred Wegener’s theory of continental drift would have the continents “plough” through Earth’s rigid lithosphere— the crust plus the uppermost mantle. He had finally been silenced.
The next breakthrough, which would tie geomagnetism into the seafloor spreading story, followed the development and deployment of the fluxgate magnetometer in World War II. The first fluxgate had been constructed in Germany in 1936, but it was its use by the Allies that would lead to its post-war geophysical applications.
The fluxgate magnetometer is an electromagnetic instrument that capitalizes on the strong, easily induced magnetization of ferrite materials, such as those Néel had been developing. Flown at low level over the sea, it had proved Churchill’s best means of detecting the dreaded U-boats. After the war, fluxgate sensors began to be routinely deployed in watertight streamlined “fish,” which were towed some 200 to 300 meters behind survey ships.
From the late 1950s, they were gradually superseded at sea by another new class of instrument, proton precession magneto-meters, which use the principle of nuclear magnetic resonance to measure the intensity of the prevailing magnetic field. These were well suited for use on boats: their operation was independent of orientation, something always difficult to establish with precision at sea, and they did not require the regular calibration checks essential for fluxgates.
Because of these advances, high-resolution magnetic field intensity measurements began to be carried out around the oceans of the world alongside detailed bathymetric profiling. At first, measurements were made along lines or profiles, but with the beginning of systematic mapping—surveys consisting of series of parallel profiles—startling patterns began to appear. Charts of “magnetic anomalies” revealed long, alternating bands of high and low field-strength running parallel to the ridge systems, with steep gradients in between.
In the first detailed survey, which took place off the west coast of North America, this pattern was found to repeat again and again, over huge areas of seafloor. The regular alternations were completely different from the almost random pattern of magnetic anomalies that scientists were used to seeing in aeromagnetic surveys over land, in which a magnetometer was carried by an airplane. When positive anomalies were colored black and negative anomalies white, the marine magnetic charts resembled the stripes on a zebra’s back.
The common opinion among geomagnetists was that the patterns were the result of magnetization induced in rocks on the seafloor by Earth’s magnetic field. Black (positive) anomalies were supposed to indicate highly susceptible, easily magnetized rocks, and white more weakly magnetizable rocks. But why did strongly and weakly susceptible rocks alternate in this way? Why should the magnetic composition of seafloor rocks change so rapidly and regularly? And why should the strongest anomalies always occur right over the mid-ocean ridges, broken only by sharp offsets that mirrored the fracture zones?
A marine magnetic survey off the west coast of the United States and Canada, originally published in 1961. Black shading shows areas where the magnetic field is stronger than expected, and white shading where it is weaker. The significance of the striped patterns was not recognized at the time. Only later were the segments of the East Pacific Rise—the Gorda, Juan de Fuca and Explorer Ridges—recognized, and the magnetic anomalies interpreted in terms of seafloor spreading and reversals of Earth’s magnetic field.
In the summer of 1962, Drummond Matthews, a marine geophysicist returned to Cambridge University from a scientific cruise in the Indian Ocean to find his new research student, Fred Vine, eagerly awaiting instructions. Matthews set Vine to work on the magnetic, bathymetric and gravity data he had just collected over the Carlsberg Ridge of the northwest Indian Ocean. At the same time, on the other side of the Atlantic, Lawrence Morley, chief of the Canadian Geological Survey’s Geophysics Division, was becoming increasingly distracted by the extraordinary marine anomaly records coming in from the east Pacific Ocean.
Vine, Matthews and Morley were all familiar with the reversed remanent magnetization now found increasingly frequently in rocks. All three were adherents of the field-reversal theory, so not surprisingly they put two and two together at the same time. Seafloor spreading, combined with geomagnetic polarity reversals and remanent (rather than induced) magnetization could, they realized, explain the seafloor stripes.
Vine and Matthews wrote to Nature magazine putting forward this theory—and so did Morley. In a faux pas of the scientific refereeing system, Vine and Matthews’ letter appeared in Nature on September 7, 1963 but Morley’s was turned down. Morley’s manuscript was then rejected again, this time by the Journal of Geophysical Research with the comment “interesting” but “more appropriately discussed at a cocktail party than published in a serious scientific journal.” It was over a year before his paper was eventually published by the Royal Society of Canada, and later still that the theory became known as the Vine-Matthews-Morley hypothesis.
This hypothesis proposed that magma erupting at a mid-ocean ridge would, when it cooled below the Curie point of its magnetic minerals, become stably magnetized parallel to Earth’s magnetic field. It would then begin to spread in both directions on the conveyor belt of the mantle convection cell. Closest to the ridge you could expect to find rock magnetized in the present “normal” field direction, but further away you could expect to find older rock, which had erupted and cooled before the last polarity reversal, magnetized in the opposite direction. The seafloor would, therefore, be acting like a double tape recorder, advancing in both directions from the ridge, producing strips of rock magnetized alternately in “normal” and “reversed” directions, symmetrically arranged about the ridge; the width of each strip would be determined by the length of the polarity interval and the rate of seafloor spreading.
As the seafloor spreads away from mid-ocean ridges, new
magma rises and solidifies. As this magma cools it becomes magnetized in the direction of the prevailing magnetic field: the seafloor therefore becomes a record of polarity flips. In this “barcode,” black represents normal polarity, when compasses would have pointed north, and white reversed polarity, when they would have pointed south.
Morley’s reasoning was purely conceptual, but Vine and Matthews tested theirs by calculating the anomalous magnetic field that such alternately magnetized blocks would produce, and comparing the results with the data they had collected from the Carlsberg Ridge in 1962. Lucky enough to have access to a computer at the Cambridge University Mathematical Laboratory, they programed it to model magnetic anomaly patterns corresponding to the Pacific Rise at a magnetic latitude of about 40°, the mid-Atlantic Ridge at a latitude of about 47°, and the Carlsberg Ridge. Their models turned out to be a far better fit with the observations than models based on the earlier idea of normal induced magnetizations of varying strength.
In 1963, when Vine and Matthews’ results were published, the field-reversal theory was still controversial—as seen in Munich, when half of the twenty-eight assembled “experts” had either rejected the idea or been unprepared to commit themselves to it. Equally, seafloor spreading was still an intuitive theory not fully proven. Vine and Matthews’ original calculations had sidestepped details such as the rate of spreading and the dates of polarity reversals by assuming twenty-kilometer wide strips of seafloor that were alternately normally and reversely magnetized.
Morley had gone on to do some rough calculations based on early estimates of the frequency of reversals and the length of magnetic anomaly patterns, and had shown the concept was consistent with a rate of seafloor spreading, or mantle convection, of a few centimeters a year. This was a credible figure, but it still did not constitute hard proof.
North Pole, South Pole: The Epic Quest to Solve the Great Mystery of Earth's Magnetism Page 17