Krakatoa: The Day the World Exploded

by Simon Winchester

  A crystal of magnetite, usually aligned with the poles along its long axis.

  Our task – which had been set by a small number of curious laboratories around the world, and whose professors some months before had persuaded a variety of grant-bestowing bodies to give us sufficient money to charter the icebreaker and buy our Nansen sleds and our pemmican and hard tack and then to set off – was to collect hundreds of samples of these basalts. If and when we got them home, they would be sent off to the researchers, who would measure the strength and the direction of the relict magnetism of the iron-oxide particles contained within them. In truth the strength of the age-old magnetism only claimed their most cursory interest. It was the direction of magnetism that beguiled them most – and in connection with that we were asked to note very carefully indeed, while taking the samples, the way they were aligned, relative to the North and South Poles of today.

  So we spent hours – day and night, bathed in the Arctic summer's continual sunshine – painstakingly drilling basalt samples out of the sheer walls of the local nunataks. We used a portable power-drill (with a diamond-studded bit that needed to be cooled by dousing it with bucketfuls of snow) and a very accurate sun-compass* to make sure we always knew where each sample lay relative to the present-day poles. The selected cores, eight inches long, two inches in diameter, were then indelibly marked with their geological horizon (how high up each had been found among the other layers of rock) and their sun-compass orientation, wrapped in plastic and set aside in specially made boxes of strong and waterproof fibreboard.

  The science that required the presence of all those basalt cores was elegantly simple and, by the standards of the present day, somewhat mundane. The interested scientists all wanted to work on the samples to satisfy a suspicion that had been growing during the early sixties. This burgeoning belief was rooted in an unsay-able, almost (in some quarters) a heretical view of the making of the modern world, and held that the magnetic alignments of crystals in rocks laid down in the past might, just might, be substantially different from the alignments of the rocks themselves today. This was something that, if it could be demonstrated, would spawn a profound revolution in contemporary geophysical knowledge and thinking. If it were to be proved, then the unsay-able would suddenly become something to preach, and the heresy would, overnight, become dogma.

  And that is just what happened. The rocks we brought back did eventually (though many months later) prove essentially and exactly what everyone wanted them to prove. A comparison of the directions of the spinel compasses (the science of studying such phenomena had a century before been given the hybrid Greek–Latin name of palaeomagnetism) confirmed what a very large number of scientists suspected. There was a very considerable difference between the direction in which the Tertiary magnets were pointing and the position of the poles today. All of the millions of magnets were lined up not towards the present poles at all, but towards a point some fifteen degrees to the east of them. It was a simple realization, but it had a stunning effect.

  This meant one of two things: that either the poles had moved in relation to the rocks or the rocks had shifted in relation to the poles. The first was initially the more tempting possibility, since it seemed quite simple for the North Pole, which was after all an invisible and rather mysterious entity, to have somehow shifted itself around by fifteen degrees. But the scientists who were doing the work on our rocks had an advantage, in that they knew this could not be the case. They were already performing palaeomagnetic studies on the other Arctic rocks of the same age, on samples from Spitsbergen and the Faeroes and Norway. What they found was that the computed polar positions suggested by the spinels in these rocks varied wildly – so much so that it looked not like the poles were wandering, but rather like there were scores of North Poles, all existing at the same time.

  So if there was no polar wandering, then the alternative could be the only explanation. It was an explanation that proved a Eureka moment, a truly life-changing epiphany for many of those researchers back in the sixties. It was the realization from the record of the rocks that the basalts of east Greenland had moved. Somehow the basalts of east Greenland had drifted westward, through fifteen longitude degrees or so, in the thirty million years since they had been extruded from the earth.

  In other words, the long-imagined (but until this moment, generally discounted) phenomenon of continental drift had – and now, moreover, provably and incontrovertibly – occurred. During Tertiary times the sea-floor beneath the Atlantic Ocean had clearly and demonstrably spread open. Now, from the Greenland basalts, there was powerful evidence to suggest that the theories of continental movement – with the world beginning as one supercontinent, Pangaea, which had then broken up and spread itself over the global surface – which Alfred Wegener had advanced so obsessively and which had been so widely dismissed by the scientific establishment during most of the half century beforehand, had at last been more or less decisively confirmed.

  Later in the 1960s, and with a growing amassment of evidence like that from east Greenland now safely in the bag, a great scientific sea-change was in the offing. It first came about unexpectedly. It first came about by accident, and in a much more appropriate piece of geography, off the coast of Java, almost within sight of the island of Krakatoa. And it first came about, as sea-changes are wont to do, at sea.

  There were two initial discoveries. One advanced the evidence of sea-floor spreading and the consequent drift of continents from the merely circumstantial into the happily incontrovertible. The other provided sceptics with something that Wegener had never managed to produce: the model for a mechanism that explained how spreading and drift might work.

  This first came to light after a series of unrelated experiments conducted off the south coast of Java – quite properly for the present story, and particularly so in that the scientist who performed the work was a Dutchman, from Delft Technical University, by the name of Felix Vening Meinesz.

  His initial intellectual interest was quite unrelated to continental drift: Vening Meinesz was solely concerned with the very accurate measurement, at various places on the globe, of the earth's gravity. He was especially concerned with the measurement of gravity in the mysterious world below the very deep oceans. Since gravity is an acceleration, it is extremely difficult to measure from something that is itself moving – like, say, a ship. So Vening Meinesz had to try out a series of home-made and very specialized measuring devices.

  Although the wholesale acceptance of continental drift was not to come about for fully forty years after Wegener's death, Vening Meinesz performed his early work while the unsung pioneer was still alive, between 1923 and 1927. He had taken a crude gravimeter, consisting of a pair of pendulums swinging in opposite directions, and mounted it on gimbals inside the most stable sea-borne vessel he could imagine, a submarine. He then had the Dutch Navy, using submarines with the somewhat unimaginative names of Her Majesty K II and Her Majesty K XIII,* perform a series of shallow dives off Java – and found to his astonishment that about 190 miles off the southern coasts of both Java and Sumatra, there was a dramatic lowering in the strength of the local gravitational field.

  This enormous gravity anomaly coincided precisely with the existence of a tremendously deep and long gash in the sea-floor known as the Java Trench. To appreciate the depth of the Trench, one can imagine someone moving eastwards (driving some seabed crawling machine, if that is not straining credulity too much) two miles beneath the surface of the sea, somewhere off Christmas Island. Without warning the seabed starts to slope downwards at a gradient of 10 per cent or more. It keeps on going down, down, down–until, at the dark and ice-cold bottom of the Trench towards which the slope is heading, the sea is nearly five miles, or 24,440 feet, deep. Then, even more abruptly, the bed begins to climb back up, dipping briefly down, once, and then rising one final time clear up to the continental shelf, the shallows, the fringing reefs and finally the beaches of south Java itself. Within 200 miles of t
he shore are some of the deepest parts of the world ocean – and directly above them, Vening Meinesz discovered, some of the lowest and weakest ravit accelerations to be found anwhere.

  The Dutchman was promptly invited to the United States by a young Princeton scientist named Harry Hess; and, together with two other young men who were to go on to become rising stars in the new field, Maurice Ewing and Teddy (later to be Sir Edward) Bullard, they took off in a boat called the Barracuda to see if the Javan anomalies could be found above the submarine trenches known to exist in the Caribbean. They did, spectacularly so. The four excitedly discussed why this might be – with Harry Hess and Vening Meinesz openly speculating that they were caused by some mysterious force dragging the rocks of the seabed downwards and (as it were) dragging the gravity down with them. Hess wrote a seminal paper in 1939:

  Recently an important new concept concerning the origins of the negative strip of gravity anomalies… has been set forward… It is based on model experiments in which… by means of horizontal rotating cylinders, convection currents were set up in a fluid layer beneath the ‘crust’ and a convection cell was formed. A down-buckle in the crust… was developed where two opposing currents meet and plunge downwards. So long as the currents are in operation, the down-buckle is maintained… the currents would have velocities in nature of one to ten centimetres a year…

  Perhaps, at last, we had a theoretical mechanism. Perhaps there were currents below the solid surface of the planet, currents that dragged the continents along on top of them and that then plunged downwards, dragging the continents down with them. Continents might thus be moving towards or away from one another at between half an inch and four inches every year – a rate that, miraculously, was entirely consonant (once someone hurriedly did the arithmetic) with what Wegener had proposed for the break-up of Gondwanaland more than twenty years before. But this was 1939, and the world was being enveloped in a

  The process of convection inside the earth's mantle, the driving mechanism behind continental drift and plate movement.

  man-made turbulence of a very different kind: Harry Hess and his bold theory would have to wait until the war was over.

  But when it was over, a wholly new and unexpected piece of evidence for crustal movement came to light – also at sea, but this time not off Java, but off the north-western United States. And while the first ground-breaking pre-war study had to do with anomalies in the earth's gravitational field, this next series of experiments had to do with a study of the earth's magnetism. More especially, they were concerned with what was known as the remanent magnetism that might be held in the rocks of the sea-floor, the old magnetic signatures that we later studied in east Greenland.

  The pioneering work on remanent magnetism had been accomplished largely by an ebullient red-headed Mancunian named Keith Runcorn. He was a lateral-thinking geophysicist who wondered out loud if perhaps the magnetic field of the earth somehow varied over time – if it had perhaps varied in its strength or its direction. And if it had varied, then perhaps a record of those variations would be discoverable by examining the remanent magnetism of rocks laid down during the time that was being investigated.*

  In the early 1950s Runcorn and his colleagues, and associates, using a range of devices (including a sphere made of thirty-seven pounds of pure gold, borrowed from a very sceptical British Royal Mint), studied this fossil magnetism in a variety of rocks of a variety of ages across Britain. The conclusion to which they came was published in a paper in 1954. The evidence showed that there were indeed significant variations in the magnetism held in rocks of different geological eras, and this could really be accounted for by only one of two happenings: either the magnetic poles were wandering about relative to the earth's landmasses, or the earth's landmasses were wandering about relative to the poles. The latter was continental drift – and to Keith Runcorn it looked a most temptingly plausible explanation.

  Further compelling evidence to support this notion was soon to come in. If the rocks on other continents showed the same remanent magnetic evidence for a wandering of the poles, then it was likely that yes, the poles themselves had moved. On the other hand, if the results from continents far away from one another were different, then it would suggest that it was not the poles that had moved but the continents. And in 1956 that evidence came in – showing, to the delight of people like Keith Runcorn and Harry Hess, widely different degrees of magnetic variation. The differences could be accounted for only by the continents having drifted apart from one another. The noose was closing. Wegener's theories were being resurrected, and fast.

  The clincher came quite accidentally, in work that began in August 1955, in the cold seas between America's most westerly point, Cape Mendocino in California, and the southern tip of Canada's Queen Charlotte Islands. An English geophysicist named Ron Mason, on sabbatical at Caltech, was vaguely aware of highly classified US Government research into underwater magnetism – classified, it was said, because the US Navy was looking at deep-sea hiding places for its long-range submarines.

  Over coffee one morning he asked whether it might be possible for him to join Project Magnet, as it was called and, without interfering with the government work, to tow a magnetometer behind the project's ship and make his own maps of any magnetic anomalies he might find down on the sea-floor. The project director agreed: and that summer Mason arranged for a long, floating, fish-like object – known formally, in the kind of language heard in science fiction, as an ASQ-3A fluxgate magnetometer* – to be towed behind the US Coast Guard's ship Pioneer as it searched for seemingly more important things on behalf of the Pentagon.

  Ron Mason's instruments up in the Pioneer's operations room, hooked by wire to the magnetometer that was trailing behind and below the ship, recorded variations in the strength and direction of the magnetism of the seabed rocks below. The recording paper unrolling from its drum (which in due course was to be superimposed on a topographical map of the sea-floor and the Pacific coastline) indicated the data by way of an intricate trace of lines – with some of the lines showing rocks that had certain properties, and others indicating rocks that had properties that were precisely the opposite.

  The magnetic ‘zebra stripes’ discovered on the seabed of the north-western Pacific in 1955, which were finally to confirm the idea of sea-floor spreading.

  As the hundreds of miles of ink traces began to plot out on the screens, Mason could hardly believe what he was seeing. At first the marks on the ever rolling drum seemed meaningless, no more than a random mess of indecipherable hieroglyphs. Within a few

  hours, however, the traces quite inexplicably began to display an absolutely regular, absolutely consistent pattern. As the ship cut steadily through the water, and as the floating ASQ-3A behind recorded the magnetic fields in the rocks hundreds of fathoms below, so the traces began to arrange themselves in an unmistakable pattern of parallel and linear stripes, just like those found on the skin of a zebra or tiger.

  The stripes became more and more obvious as, month by month, the little ship worked itself steadily along its recording route. The vessel would scuttle back and forth on its preordained track, the midshipman keeping it at a steady five knots for hour after hour, as it traced and retraced its passage along the specific sector of sea-floor that Ron Mason had decided to measure. And as it did so, so the zebra-stripe pattern from the rocks below steadily built itself up – with the recording paper eventually being covered with long black-and-white patches that alternated from black to white to black and back to white again in an ever more uncannily regular manner.

  All of the stripes, moreover, were not simply regular. After just a few passes of the Pioneer, their arrangement could be seen: they pointed not merely in parallel but essentially only to the north and to the south. Or, to be more accurate, they pointed along the long axis of the ocean in which they were found.

  The ship might go east–west, it might go diagonally; it might reverse its voyage – but no matter. Pass after pass
after pass, the magnetism in the rocks below was recorded as a series of stripes arranged in long north–south trending patterns that made the plot of the seabed look like an acrylic bed-sheet from Frederick's of Hollywood, or a herd of standing zebras or tigers, rather more exotic than it deserved to be.

  And then in a flash everyone realized what it was. The stripes on the paper were noting anomalies that were a record of the reversals that had occurred from time to time in the polarity of the earth's magnetic field.

  This was a curious reality that had been dimly recognized by a Frenchman named Jean Brunhes in the early part of the century, then confirmed (by a Japanese geophysicist named Motonari Matuyama) in the 1920s. He showed that there had been a field reversal during the late Pleistocene, 10,000 years ago. Thirty years later his work had been fully confirmed: basalt layers in Iceland showed a series of back-and-forth field reversals in their remanent magnetism. From then on there was no doubt about it.

  The work in Iceland and elsewhere showed field reversals to be a standard (if stubbornly inexplicable) feature of the earth's magnetic field, just as can happen in any man-made self-exciting dynamo. Over the last 76 million years, the palaeomagneticists soon reckoned, there had been no fewer than 76 such reversals – and they had continued to a measurable degree right back to the early Jurassic, 150 million years ago (except for a long period between 85 and 110 million years ago when, equally inexplicably, there were no reversals at all, during what has since come to be known as the Cretaceous Quiet Zone).

  And now here, at last, was evidence in black and white of reversals in the rocks' magnetic field, and moreover reversals that, when plotted on to a map, seemed to occur along regular lines, following a pattern that was imprinted indelibly on the seabed of the north-eastern Pacific. And, as more and more data was recorded and analysed, something even more astonishing was noticed: that the peculiar intricacies of the stripe pattern on one side of the ocean were almost identical to those that could be seen in the stripe patterns on the other, and that there was a point, or axis, in mid ocean, on which this symmetry seemed to hinge. In an instant the explanation for that became clear too.