He drew diagrams of how this should appear – how its oppositeness-from-normal should manifest itself. He created paper-and-crayon diagrams of the model, cut them out and kept them in his wallet, showing them to anyone who would give him the time of day. This friendly Canadian gentleman – scrupulously dressed, nicely mannered – with his cut-out papers in his pockets soon became a familiar figure on more than a few university campuses in the early 1960s.
And, just like John Dewey, everyone who saw Wilson's demonstration knew in an instant that he had it right there, the proof. All that was needed was the test. And for this, the Cold War – of
blessed memory; at least blessed to many scientists – came promptly to the rescue. Just as the US Navy had been happy to tow magnetometers behind the Pioneer in the fifties, so in the sixties it was content to let scientists make use of a worldwide seismographic network it had deployed to listen for Soviet and Chinese nuclear weapons tests. It was a system with scores of highly sensitive pieces of seismic equipment scattered around the planet. It was ideally positioned to find out, in addition to the sites of bomb detonations, something that it had never been designed to discern: the direction of slippage along some of the great fault-line gashes in the mid Pacific and the mid Atlantic.
The system worked its magic. The Pentagon agreed the result should be free from military classification, which had so slowed earlier research. And so in 1967, from a laboratory at Columbia University that had examined the numbers, the results came tumbling out: the great faults in the middle of both oceans were indeed transform faults, their slip directions were the exact opposite of what would be expected had they been standard transcurrent faults. Tuzo Wilson's ideas had been vindicated yet again: first with the evidence from the conveyor-belt island chain of Hawaii, and now with his supposition about the world in general, and what was going on in the unseen middle of its oceans.
And of that, there was now no doubt at all. The rocks underpinning the oceans were most assuredly splitting themselves apart. The academic battle that had been fought for the previous fifty years was now over, for good. The stabilists* – as were called those who believed, as most once had, that the world and its continents had always been in approximately the same place – had finally to yield. The day belonged to the mobilists, who had since Wegener's time argued that the continents wandered, with what are now known to be dramatic and highly visible effects – such as the creation of the modern map of the world.
It was, for science, an incredibly exciting time. An odd time too – for geologists soon began to realize they had been studying their earth backwards. It was as though an entomologist had begun studying a bee by looking not at the bee as a whole but at the microscopic structure of the yellow hairs on its abdomen; or as though a botanist had tried to get to know an oak tree by first using an electron microscope to peer at a cross-section of an acorn. It is all a consequence of man's own insignificant size compared to what he is studying, of course. But the fact is, when plate tectonics came along, it was realized that geology had been spending its previous two millennia as a major science looking in great detail at sandstones and gneisses and rift valleys and ammonites – but had never been able to stand back and look at the planet as a whole and then to work out the details, as happens with most sciences where man is generally bigger than whatever it is he is studying.
Plate tectonics offered for the first time an intellectual mechanism for taking the earth and looking at it as an entity – and the fact that its emergence as a brand-new science coincided so nicely with the development of satellites that could look at the planet as a whole was fortuitous, to say the least. One might say that all this meant that, for the first time, geologists were able to begin looking at things right side up.
And what a vision was now laid out before them! The oceans were coming apart at the seams. The crust and the upper mantle beneath them were spreading out across the deep-ocean floors, moving in opposite directions on each flank of the mid-ocean ridges. Then, when the mobile material reached the edge of the ocean the edge of the plate, to use Tuzo Wilson's now universally accepted term – it plunged down below whatever it met, readying itself to be recycled in the deepest recesses of the planet, to return to the half-molten underneath of the world and in due course to thrust itself up once more through the oceanic centres and begin the long circular process all over again.
Plate tectonics is, in essence, the way by which the world deals with its steady loss of heat. A vast amount of heat accumulated during the formation of the planet a little over 4,500 million years ago – and natural radioactivity, particularly the decay of isotopes of potassium, uranium and thorium, served only to add to the ferocity of the internal fire. But that heat is now ebbing away, and the means by which it is transferred from the deep interior to the surface tends to be by way of convection currents, just like those one sees working in a vat of vegetable soup simmering on the hob.
The rate of movement of these convection currents is very, very small, generally measured in only a certain number of millimetres each year. Currents of material of the inner earth rise up from the red-hot region, maybe a thousand miles down, pass right through a swimmy, hot, weakly plastic region called the asthenosphere, and then, when they reach a region above this, anywhere from four to twenty miles of the surface, slow, stop and eventually – in classic convection-current, vegetable-soup fashion – turn back downwards once again.
The currents, with the soft, pliable and plastic rocks of which they are composed, turn tail at this point on their upward journey, because they encounter an inconveniently brittle and rigid layer of the earth: the upper part of the mantle and the entirety of the crust, which is today called the lithosphere. And it is in the lithosphere that the tectonic plates themselves exist.
In the oceans the lithosphere is thin – maybe four miles thick – and it is young, no more than 200 million years old. The lithosphere at the continents, with all their long-ago-made limestones and granites and shales and gabbros and schists and gneisses piled on top, can on the other hand be as much as twenty miles thick and much, much older – typically, in fact, as much as two billion years old. It is from these two kinds of lithosphere that the tectonic plates are made; the convection process that is going on below them, which is slowly cooling the earth to its ultimate frigid darkness, is what drives them to move and shuffle and shunt and bang their way around the surface of the earth. It is a deeply complex subject, the stuff of mathematical modelling and the employment of banks of supercomputers. It is still far from being fully understood.
But out of the early unravelling and understanding of the process was born in those early years yet another new term: a subduction zone. This, it seemed, was where the real business of the world was done. This was the comparatively narrow strip of territory below which, at the edges of the plates, the moving material collided, and one plate slipped beneath the other and began to head back downward, to balance the creation of new lithospheric material at the oceanic seams. Most of the plates that converge do in fact display subduction. And so by finding out just where the world's subduction zones were, science could find the boundaries of most of the planet's tectonic plates and, by identifying them, work out what was happening to them, where, how fast and why.
Ever since the beginning of the 1970s – the phrase subduction zone appears to have been first used in Nature (once again) in the issue of 14 November 1970 – a huge proportion of the world's geophysicists have been engaged in studying their bewildering complexities and their seminal influence on the arrangements of the earth. What happens deep within their hot and roiling mysteries is crucial to an understanding of the processes of the earth's making. All the greatest dramas of the solid world's evolution happen within them.
A subduction zone, like that responsible for the Sunda Strait eruption – and, indeed, for 95 per cent of the world's most violent volcanoes.
And dramas that also include the realization, which was published in an otherwise obscur
e book in 1980, that it is
... subduction along the Java Trench, where the Indo-Australian Plate is moving under the Indonesian island chain… that fuelled the 1883 eruption of Krakatoa.
The earth's surface appears to be armoured with between six and thirty-six of these rigid plates, depending upon how they are defined and counted. Many of the boundaries between plates have been drawn on the basis only of educated guesswork; what happens deep in the interior of China, for instance, where one plate is thought to meet another, has not yet been fully explained. Nor is it known what goes on around what is called the Scotia Arc in the chilly South Atlantic, east of the Falkland Islands. But, generally speaking, modern tectonics accepts the undisputed existence of about a dozen major plates. The boundaries between these, where the mechanics of the planet's architecture are best exhibited, are by now very well drawn.
Exactly what happens at the boundaries depends on a number of factors. If the plate is entirely of oceanic material – that is, basically basalt – and it collides more or less head-on with another plate of the same type, then there is an almost randomly decided subduction, with one plate indeed sliding beneath the other. An arc-shaped arrangement of small volcanic islands is usually created, the arc being shaped according to the way the plates move relative to one another. The chain of islands stretching along the International Date Line south of Tonga, for example, is a classic illustration of what happens when two oceanic plates bump into one another. For the purposes of this account, they are best overlooked.
If both plates are, on the other hand, composed mainly of continental material – which can best be thought of as a light scum of ancient granites and other sediments and metamorphic rocks lying on top of the oceanic basalts – then their collision may not result in a subduction at all. As the continental material is made up of lighter rocks, it generally resists being thrust down into the earth's subsurface; both plates may instead stay on top in line with one another, and buckle and crumple, forming a chain of mountains that is in most places not volcanic at all. India and Asia are a pair of continental plates that are colliding, for example – with the effect that the earth's crust where they smash into one another has crumpled itself up and more than doubled in thickness, to form the world's highest mountains, the Himalaya.
The Africa Plate is similarly moving slowly northwards and colliding with the Eurasian Plate. The highly unstable mountains of the Caucasus, the earthquake-prone hills of Turkey, the ever restless fracture zone of the Balkans, the ski-slopes of Iran's Mount Damavand are all a consequence of this entirely continental plate collision. But where there is some small amount of subduction, as near Vesuvius and Etna, there are volcanoes. Shallow-focus, very dangerous earthquakes are to be found everywhere along a continental collision zone: volcanoes are more likely where there is subduction.
It is also possible, given the complicated geometry of the earth, that two plates may well not hit each other head-on but slide alongside one another, like a tyre sliding past the kerb during a lame attempt at parking. The world's best-known example of this kind of meeting is the San Andreas Fault in California, notorious for what humankind regards as its capricious behaviour (a geologist would not think it capricious at all). It causes earthquakes in abundance and has enabled Hollywood to come up with some heroically awful films about volcanoes in Pasadena, tsunamis – lethally massive sea-waves – scooping surfers off Pebble Beach and faraway politicians uttering such droll slogans as the Coast is toast.
The Fault is a classic example of what is known as a conservative plate boundary, where there is no collision, no crumpling, no subduction. The immense Pacific Plate is here simply moving northward relative to the North American Plate, sliding along it at about half an inch a year. Extraordinary stresses can build up if this movement is somehow hindered – by friction, for example. And when this stress becomes overwhelming and is suddenly released, terrible earthquakes can and do occur. During the infamous San Francisco Earthquake of 1906 parts of the fault, which in recent years has accelerated its rate of sliding to almost four inches a year, shot past each other a total of twenty-one feet in a matter of twenty seconds!
Finally, and most importantly for the understanding of what happened at Krakatoa, there are those events that occur when two plates of different types collide – when one plate that is all oceanic basalt smashes itself into another plate that is loaded up with continental crust. These, so far as this story is concerned, are the important ones.
*
What takes place when plates of different composition run into one another, as might be expected (but was actually not, until that sudden realization made in 1967), is that the heavier of the two, the oceanic basalt plate, dips itself underneath the edge of the lighter continent-laden plate. Immediately it has started to do so, matters directly beneath the point of collision – matters within the classically formed subduction zone, that is – become complicated beyond belief and, to the geophysicist, endlessly fascinating. What takes place within the zone also has a profound effect on the making of the region above and on the lives of the human and animal inhabitants there – not least because, intermittently, the results of the plates' meeting and the making of a subduction zone can be very dangerous indeed.
In essence the process at a subduction zone begins when the colliding oceanic basalt plate begins to head off down back into the earth. It drags some of the continental plate down with it and, in doing so, causes a pinch – a trench – in the bed of the sea. In the case of Java, the five-mile-deep Java Trench, 200 miles off the coast, provides vivid evidence of the oceanic basalt plate as it begins barrelling its way downward. Behind the plate – the coral reefs of Cocos (Keeling) Island, Christmas Island, the flat blue Indian Ocean and, in the far distance, the coast of Australia – all is placid, paradisal, serene. In front of the down-racing plate is the Trench, a line of offshore islands, and then Java and Sumatra – one of the most volcanically unstable pairs of islands ever known.
They are unstable because of a well-understood, if wonderfully complicated, mechanism. The down-racing plate heads into the heat, dragging with it billions upon billions of tons of additional material – and, most crucially of all, water. This it gets from the huge thickness of the waterlogged sediments that are being dragged down from the seabed. Once this waterlogged material is stirred into the mix and reaches a critical depth, quite unexpectedly to those who first noticed the phenomenon, it begins to melt – doing so because the addition of this water has lowered the melting temperature of the mix.
And so what starts off beneath the seas as cold and solid now moves down towards the hot mantle, and turns viscous and runny; its fluid components begin to ‘sweat out’ – suddenly bubbling and frothing and coursing and, because they are light and volatile, so rising back up again, passing into the solid mantle through which all the downsliding ingredients had passed. To make matters more complicated still, as they course upwards through this material, they begin to melt that too.
Suddenly a Hadean nightmare is created miles beneath the subducted continental crust: immense volumes of boiling, gaseous, white-hot magma, alive with bubbles and energy and restless muscle, seethe in vaults and chambers of unimaginable size and temperature. The promethean material searches ceaselessly for some weakened spot in the crust above it. Every so often it finds one, a crack, crevice or fault, and then forces its way up into a holding chamber. Before long, the accumulating pressure of the uprushing material becomes too great, and the temperature too high, and the proportion of dissolved gas becomes too large, and it explodes out into the open air in a vicious cannonade of destruction. A type of volcano, which, because of its position at the edge of a subduction zone, is far more explosive and dangerous than any other of the world's many different kinds of volcano, suddenly – and if there are people around, invariably terrifyingly – erupts.
Two points of simple geography remain. The first is a matter of tectonic trivia. The Indo-Australian Plate, the culprit that creates
all of the volcanic tamasha so notoriously present on Java, on Sumatra and on the tiny island that once lay between them in the Sunda Strait, is moving because the sea-floor south of Australia is spreading open. It is possible to calculate both the speed of that opening and the pole of rotation around which this spreading is taking place.
The speed of the sea-floor spreading, the rate at which the Indo-Australian Plate is moving northwards and colliding with the Eurasian Plate, seems to be about four inches a year, a hundred yards every millennium. Put another way: when Java Man – the Homo erectus who first came to these parts from Africa – was living in central Java about 1.7 million years ago, Australia and Asia were rather more than a hundred miles further apart than they are today. They have been moving towards each other ever since, and doing so in a direction that, unusual though it may seem, turns around a pole of rotation that is located a few miles to the south and east of the Egyptian capital, Cairo.
And secondly: there is one delightful symmetry that is noticeable, though imperfect at best and complex in the extreme, as more and more details become known. The tracks of the subduction zones that enfold the islands of Indonesia, the zones that create all the trenches and island arcs and volcanoes and mayhem that goes on among the islands of the archipelago, more or less follow the invisible line of biology and botany that was first hinted at by Philip Sclater in 1857 and drawn a year later by Alfred Russel Wallace.
On one side: Australia, cassowaries, emus and kangaroos. On the other: cows, monkeys, thrushes and elephants. On one side: the Indo-Australian Plate; on the other: the Eurasian Plate. The middle, where the two plates meet, and where they come together very slowly but with immense and unthinkable raw power, is in consequence a serried line of the world's greatest, most dangerous and most predictably unpredictable volcanoes – including, lurking just on the Asian side of an imagined extrapolation of the Wallace Line, the most demonstrably dangerous of them all, the once-great island of Krakatoa. It is an island that has exploded many more times than on the one occasion for which it is now so notorious.