Mind, Machines and Evolution Part 19

Li walked over to the window and gazed out at Peking's soaring panorama of towers, bridges, terraces, and arches, extending away all around, above, and for hundreds of meters below. "How did Gravmas start?" he asked his father.

"Hmph!" Xiang snorted as he moved to stand alongside the boy. "Now isn't that typical of young people today. Too wrapped up in relativistic quantum chromodynamics and multidimensional function s.p.a.ces to know anything about where it came from or what it means. It's this newfangled liberal education that's to blame. They don't teach natural philosophy any more, the way we had to learn it."

"Well, that kind of thing does seem a bit quaint these days," Li said. "I suppose it's okay for little old ladies and people who-"

"They don't even recite the laws of motion in school every morning. Standards aren't what they used to be. It'll mean the end of civilization, you mark my words."

"You were going to tell me about Gravmas . . ."

"Oh, yes. Well, I presume you've heard of Newton?"

"Of course. A newton is the force which, acting on a ma.s.s of one kilogram, produces an acceleration of one meter per second per second."

"Not a newton. The Newton. You didn't know that Newton was somebody's name?"

"You mean it was a person?"

Xiang sighed. "My word. You see-you don't know anything. Yes, Newton was the messiah who lived two thousand years ago, who came to save us all from irrationality. Today is his birthday."

Li looked impressed. "Say, what do you know! Where did this happen?"

"In a quasi-stable, in a little town called Cambridge, which was somewhere in Britain."

"That's in Europe, isn't it?" Li said.

"Oh, so you do know something."

"My friend Shao was in Europe last year," Li went on distantly. "His parents took him on a trip there to see the ruins. He said it was very dirty everywhere, with the streets full of beggars. And you can't drink the water. It sounds like a strange place for a civilization like ours to have started from."

"Strange things happen . . ." Xiang thought for a while. "Actually, according to legend, it didn't really start there."

"What?"

"Gravmas."

"How do you mean?"

"Supposedly it was already a holiday that some ancient Western barbarian culture celebrated before then, and we stole it. It was easier to let people carry on with the customs they'd grown used to, you see.

. . . At least, that's how the story goes."

"I wonder what the barbarian culture was like," Li mused.

"n.o.body's quite sure," Xiang said. "But from the fragments that have been put together, it seems to have had something to do with wors.h.i.+ping crosses and fishes, eating holly, and building pyramids. It was all such a long time ago now that-"

"Look!" Li interrupted, pointed excitedly. Outside the window, a levitation platform was rising into view, bearing several dozen happy-looking, colorfully dressed people with musical instruments. The strains of amplified voices floated in from outside. "Carol singers!" Li exclaimed.

Xiang smiled and spoke a command for the household communications controller to relay his voice to the outside. "Good morning!" it boomed from above the window as the platform came level.

The people on board saw the figures in the window and waved. "Merry Gravmas," a voice replied.

"Merry Gravmas to you," Xiang returned.

"May the Force be proportional to your acceleration."

"Are you going to sing us a carol?" Xiang inquired.

"But of course. Do you have a request?"

"No, I'll leave it to you."

"Very well."

There was an introductory bar, and then,

"We three laws of orbiting are, Ruling trajectories local and far.

Collisions billiard, Particles myriad, Planet and moon and star.

O-ooo . . ."

KNOWLEDGE IS A.

MIND-ALTERING DRUG.

I sometimes suspect that one of the reason writers write is that it gives them an excuse to do the research. In an age when people are constantly being urged to be goal-oriented and efficiency-conscious, and get sent by their firms to seminars to learn how to manage their time, it's easy to develop a guilt complex over reading anything other than a company procedure manual. But I find the most enjoyable reading is that which is purely for fun or out of curiosity-with no conceivable relevance to making money, furthering one's career, or with any other such redeeming quality whatsoever. One solution to any residual guilt from company indoctrination is to be a writer. Then it becomes possible to relax and enjoy whatever one pleases, rationalizing it by the thought that "Who knows? I might need if for a book one day."

I remember once, when I was in my teens, a friend accused me of never being bored by anything-which can be an unforgivable aberration among teenagers. I had never thought about that, but it seemed worth investigating. I resolved, therefore, that to test the allegation, I would force myself to read for one hour on the dullest subject I could think of. I couldn't think of anything that sounded more dull than Greek architecture, and so, when I was next in the public library, I took down a couple of formidable-looking tomes on the subject and steeled myself. It turned out to be fascinating, and I ended up staying until closing time.

It's easy to get carried away, sometimes. Voyage From Yesteryear featured a huge, fusion-powered s.p.a.cecraft, with a population of tens of thousands, that traveled to Alpha Centauri, the nearest star to us.

The entire structure rotated to simulate gravity, and at one point I was writing a part of the story that included a conversation between a boy who was born during the voyage, with no experience of planetary gravity, and his father, who grew up on Earth. To the boy it was self-evident that a thrown baseball moves in a straight line, and the hand of the catcher is carried in a curve by the spin of the s.h.i.+p to intercept it; but the notion of something going up, reversing, and coming down again made no intuitive sense at all. Then, having written that much, I found it difficult to convince myself that a baseball trajectory as seen by somebody on the inside of a spinning structure would look like the curve of one thrown up from the ground on Earth. I spend a whole week deriving from first principles a set of equations to transform curves from fixed to rotating coordinates and drawing graphs of the results-the simulation of "real" gravity turned out to be surprisingly close. Then, of course, I had to write the procedure into computer programs "in case I need to do it again some day" (I never have). That took another week. And after all that, when I finally edited the draft, I deleted the paragraphs in which the conversation took place, because by that time it didn't seem so important.

The Proteus Operation required a lot of research into modern history and World War II, the beginnings of nuclear physics and the Manhattan Project, and on the biographies of the several real-life people who appeared in the story. There was a six-month gap between my writing the prologue and Chapter One-a result of getting carried away again.

When I was writing Inherit the Stars back in England, I received a lot of help with background material from my customers. One of them was a physicist at Sheffield University, called Dr. Grenville Turner, who used one of our computers to a.n.a.lyze moonrock samples from the Apollo missions. On one occasion, while we were eating lunch on the campus lawns, I mentioned the idea of having the moon captured by the earth, as is described in the book. Gren though for a while as he munched a sandwich, and then said suddenly, "You're dead! It won't work."

"Why not?" I asked him.

"Stromatolites."

"Never heard of them."

Stromatolites turned out to be a kind of fossil coral found in Australia that preserves records of the tides from hundreds of millions of years ago. Stromatolites show that lunar tides have existed since the beginnings of Earth's history, and therefore the moon couldn't have arrived comparatively recently in the way the book said. I eventually managed to fudge that around in such a way that it actually became supporting evidence for the capture theory, but the reason I mention it here is that it led me off into a new line of research on the ancient Earth and the processes that have shaped it into what it is today. I believe that much of this kind of thing is taught in schools these days, but it was all new and fascinating to me, because, as you may recall, the curriculum that I took hadn't been updated since the days of King James the whichever. Some other readers may not have met this in school, either. So, for them-or maybe anyone interested in finding out if they can be bored, if it sounds like that kind of subject; but they may get a surprise-here is a distillation from the notes I compiled. They're not doing anyone much good in the bottom of my filing cabinet.

EARTH MODELS-ON A.

PLATE.

What do the coastlines of South America and Africa, an Australian aborigine's fireplace, and earthquakes have to do with the world's oldest magnetic tape recorder?

The answer to this question provides a good example of how science works, following up every clue, fitting together sc.r.a.ps of seemingly unrelated information, and sometimes turning conventional wisdoms around completely in order to arrive at the right question. It also provides a fascinating insight into some of the processes that have shaped the world on which we live.

The story begins back in the days when it first became possible to construct accurate maps. People began noticing the similarity in shape between the coastlines of South America and Africa, and couldn't avoid the feeling that the pieces "ought," somehow, to fit together. The first recorded thought of this kind was voiced by the English philosopher and scientist, Francis Bacon, in 1620. This soon led the more curious among mankind-which of course includes just about every scientist-to ask, "Were South America and Africa ever joined together?" and, "If they were, why aren't they now?"

A clear answer to the first question required a precise definition, expressible in numbers in the way scientists like, of just what, "joined together" means. The problem is that the coastlines that happen to exist today aren't really the thing to go by. As a glance at any physical map will show, coastlines are fairly arbitrary and can change beyond recognition with even minor variations in sea level. If all the ice in the Earth's polar caps and glaciers were to melt, sea level would rise by about two hundred feet, and the large areas that we find plastered with fossil remains of shallow-water animals and plants on what today is dry land testify that this has indeed happened on numerous occasions. Conversely, estimates of the amount of ice that must have existed during the recent Pleistocene ice age suggest that sea level then must have been seven hundred feet lower than today. Hence, we need a more meaningful boundary than contemporary coastlines to define the continents.

For the most part, continents are surprisingly flat on top-which is why small changes in sea level can make such big differences to coastlines-and slope gently out into the oceans to the edge of the continental shelf, which off New York, for example, lies about five hundred feet deep and one hundred miles into the Atlantic. Beyond the shelf, the seabed falls steeply and levels out again at an average depth of ten thousand feet to become the floor of the deep-ocean basins. Somewhere down this "continental slope," therefore, lies the "true" edge of what const.i.tutes a continent.

A number of best-fit-by-eye attempts were made to match South America and Africa at various depths, and although yielding similar results, they all rested ultimately on the somewhat vague a.s.sertion that "it looks about right," which lacks testable precision and is the kind of thing guaranteed to make scientists uncomfortable. So, in 1955, Sir Edwin Bullard, professor of geophysics at the University of Cambridge, and a research student called Everett, developed a computer program that would calculate a best fit as the solution that gave the minimum area of gaps and overlaps between two corresponding contours anywhere down the continental slopes. The fit that finally met this criterion occurred at the five-hundred fathom (three-thousand-foot) contour, at which depth the average misfit was between fifty and sixty miles. Three small overlap areas occurred: one where the northeast corner of Brazil intruded upon the projecting area of the river Niger, and two submerged areas that lay farther south. All in all, the result was considered excellent. To achieve this fit, South America needed to be rotated through an angle of 57 degrees about the point on the earth's surface at lat.i.tude 40 degrees north, longitude 30.6 degrees west, which is in the Atlantic, close to the western Azores. Try it on a map and see.

What this says is that at this depth the continents match in a way that seems too close to be explained away by coincidence. The operative word here, however, is "seems"-coincidences do happen. To elevate the idea above pure speculation, we need some corroborating evidence. The problem is similar to that of deciding if the two doors of a closet have been cut from the same piece of timber, in which case the obvious thing to look at first is the grain pattern across the join. It's important, of course, to distinguish between similarities that mean something and others that don't, such as scratches made after the doors were hung in place. In the case of continents, an example of the latter would be the evidence of extensive rain forests which geologists a few million years from now will observe in the basins of the Amazon and the Congo. Obviously they would be wrong if they inferred from this that the two places had been joined in the twentieth century-the match would be simply the result of similar environments working on similar raw materials.

The true "grain" of continents consists of the mountain belts and old, deep-lying rock strata. When these significant geological formations were distinguished and compared for age and structural characteristics, they showed an extraordinarily good correspondence across the best-fit join selected by the computer a.n.a.lysis. Furthermore, nothing comparable to them appears in the structure of the ocean floor separating the two regions today. Now if somebody observed that the two closet doors matched but the fixed upright strip between them was different, he would conclude that the doors had been cut from the same piece of wood at some time, and the upright inserted between them afterward-it's the simplest explanation that fits the facts. The same conclusion followed, too, for the continents. Data on the ages of the relevant rock formations indicated that the join had persisted until at least five hundred million years ago. All this was not especially new. As long ago as 1912, the German meteorologist Alfred Wegener based his then revolutionary proposal of continental drift on exactly this kind of information, and his work was subsequently expanded by many investigators.

So, we have five hundred million years as an upper limit for the date of separation. How about a lower limit? It turned out that the areas of overlap in the best-fit solution enable a date to be fixed for this, too.

Any parts of the continents that overlap must have formed after there was a gap for them to form in. One of these regions was the Niger Delta. The sediments that make up this formation are all younger than fifty million years and extend well over a hundred miles into the Gulf of Guinea. This says that the separation must have been at least this much by that time. To narrow these limits down further, we need to introduce glaciation and ice sheets into our story.

Between 350 and 250 million years ago, the continents of the southern hemisphere were covered extensively by ice. Rocks can be carried over enormous distances by moving ice, and are dumped wherever they happen to have got to when the ice melts-for example, Norwegian rocks carried by the glaciers of the most recent ice age are quite common in parts of Britain. Huge glacial deposits exist today all over eastern Brazil, which appear to have resulted not from the melting of glaciers localized in valleys, but of vast sheets of ice. In some places these deposits are more than two thousand feet thick-about ten times the depth of the deposits left by the recent ice age in Europe. The thrust patterns in the wrinkles and folds of the underlying rock indicate that the ice moved from southeast to northwest, i.e., from somewhere in the direction of the Atlantic. Well, where did all that material come from?

Let's be scientific and consider the alternatives before jumping to conclusions. Is it possible, for instance, that a large landma.s.s once existed between today's coast and the continental shelf, which was shoveled up by the ice and carried inland? Not really. All the studies that have been made of glaciation indicate that ice sheets don't work that way. They scratch and polish existing terrain, and carry away the looser debris, but they don't grind whole slabs of continent down to nothing. And besides, although one hundred million years sounds a long time, it isn't anywhere near long enough for that kind of major surgery. Very well, could there have been another continent offsh.o.r.e in what today is the South Atlantic-an "Australantis?" No. This conjecture runs into trouble, too, for despite romantic legends to the contrary, continents don't sink beneath the sea. Oceanic crust-the material that forms the floors of the ocean basins-is entirely different from continental crust. The floor of the South Atlantic is perfectly normal, which means that any continent that once existed there would somehow have had to transform itself from twenty-five-mile-thick continental crust into five miles of oceanic crust plus twenty miles of upper mantle (the deeper layer that lies beneath the crust all over the earth), or else have disappeared without trace.

So let's take the simple way out again, and go back to our original idea. If Africa was joined to Brazil at one time, we have a ready-made source for all those Brazilian glacial deposits. What's more, investigations in western Africa revealed widespread evidence of glacial erosion in an east-west direction-i.e., out into the Atlantic-but very little in the way of subsequent deposits left by melting. And as a clincher, the Brazilian deposits include many erratic blocks of such rock as quartzite, dolomite, and chert, which resemble none of the structures that make up Brazil, but which are common in southwest Africa. Quant.i.tative a.n.a.lysis of the glacial evidence brings the upper limit for the date of separation down from five hundred million years to two hundred million. We still have fifty million years as the lower limit.

During the period that lasted from 135 million to 100 million years ago (Lower Cretaceous), the strips which today form the South American and African coastal regions both consisted of chains of sedimentary basins-low-lying flooded areas where successive layers of rocks were laid down, the types differing as depth and other environmental factors changed. The sequences of the sediments found on both sides of today's ocean are similar, and bear no resemblance to the basin floor between the continental slopes. And this, of course, is just what would be predicted if both sequences were in fact formed as parts of the same process at a time when the intervening basin didn't exist. The case seems to be getting stronger. For a better idea of what kind of process this was, we need to turn to biology.

Fossil remains of fish and other organisms preserved in these sediments, particularly the lower layers, include many freshwater species that could never have survived in seawater. This implies that at up to about the same time on both edges of what is today the Atlantic, the water that was laying down the sediments was not ocean. Then, above the freshwater deposits, we find layers of minerals and salts of the kinds left by evaporating seawater, dating from between 110 million and 100 million years ago and again corresponding on both sides. But above the salts the two sequences begin to diverge; and the more recent they get, the more p.r.o.nounced the differences become.

This is all consistent with the suggestion of the two continents fracturing from each other at about this time and moving apart. The rift valley opening up between them would give us the chain of low-lying basins, with freshwater runoff from the surrounding highlands, accounting for the earlier sedimentation and fossil record. The sea's eventual penetration of the rift explains the salt deposits, and the progressively diverging layers above the salts testify to gradually differing sequences of events taking place in what were becoming gradually different places.

Thus, the conclusion seems pretty inescapable that not only do continents break up and move, but their voyages can take them vast distances. In fact, the courses they have followed can be charted with surprising accuracy, which brings us to the next set of clues in our scientific detective story. They concern mysteries that other researchers were finding in the natural magnetism of rocks.

The Earth possesses a magnetic field, which behaves as if there were an enormous bar magnet buried beneath the surface, roughly aligned with the rotational axis. From demonstrations at school with iron filings on a sheet of paper, most people are familiar with the "lines of force" that are said to surround a magnet. A typical line of force from the imaginary terrestrial magnet would emerge from its "north" pole, located not far from the geographic pole, curve up out of the atmosphere like an ICBM to flatten out near the equator, and plunge back down again to reenter the surface and terminate at the magnetic south pole. Now, anybody on the Earth's surface who decides to measure the characteristics of the force lines in his vicinity (our ever-curious scientist, for example) will discover two things: first, they always point the same way, north-south; and second, they intersect the Earth's surface at an angle that tilts up or down depending which side of the equator he's on, and which gets larger as he moves farther away from it-hence the "dip" of a compa.s.s needle. Thus a compa.s.s will tell him not only which direction is north, but also how far north or south of the equator he happens to be. (It won't tell him anything about how far east or west of anywhere he is; to know that, he must first get curious about clocks.) Many rocks contain grains of substances that are naturally magnetic. When these rocks form, either by cooling from volcanic lavas or by settling as sediments, these grains tend to align with the Earth's field.

Hence their tiny individual fields all line up the same way and reinforce, giving the rock as a whole a weak but distinct "fossil magnetism." Measurement of this can determine how far north or south of the equator a rock sample was when it formed, and which way it was lying at the time, while the time itself can be fixed by atomic dating. The mystery was that the early results of such experiments seemed to indicate that the Earth's magnetic poles had wandered all over the planet-in fact the poles do move to some degree, but the amounts suggested by the new figures were unheard of. Things got worse when comparisons of results seemed to indicate that the poles must have been in different places at the same time!

Then a group of British investigators looked at the problem the other way round: the same results would have been produced if the poles stayed where they were, but the rocks moved around while they were being formed. Interpreting data from their own country in this fas.h.i.+on, they postulated that in the last two hundred million years, Britain has rotated clockwise through thirty degrees and at the same time traveled a considerable distance northward-which tied in well with depictions of a subtropical prehistoric Britain that had emerged from other work. Other scientists found that a similar interpretation could explain data collected in Australia, and before long everyone was doing it.

South America and Africa turned out to have followed a zigzag course which 350 million years ago would have put the South Pole somewhere inside South Africa. This, of course, explains the huge ice sheets that we met earlier. The line splits at around 100 million years ago to mark the point at which the two continents went separate ways, and provides an independent corroboration of the date arrived at previously from other evidence. The continuing movement northward into warmer climatic zones also explains how the older glacial features of Africa came to lie today beneath layers of coa.r.s.e rock formed from windblown desert sands.

Applying the same process to the other continents leads to the conclusion that they all originated from the breakup of a single supercontinent, which geologists have christened "Pangea." In Pangea, the east coast of North America fitted against North Africa; Greenland and Newfoundland closed up and fitted with Europe; and Australia, Antarctica, and India were bunched together along the eastern side of Africa.

The group of southern continents was separated from Asia by a V-shaped ocean called the Tethys, whose apex lay in the region of the present eastern Mediterranean. The Tethys was squeezed out of existence when India, Arabia, and part of the Middle East hinged northward like a scissor blade and drove underneath Asia to lift up the Tibetan Plateau (although some people think the Black Sea could be a remnant of it.) The dinosaurs could have marched along in an ancient chain of mountains from Poland and Germany, into the Ardennes of northern France, through Cornwall, Brittany, and Ireland, and down the Appalachians without getting a foot wet. These mountains were formed during an even earlier sequence of events that resulted in the coming together of Pangea itself from previously existing continents.

So, what does the aborigine's fireplace have to do with all this?

When certain kinds of rock are heated to the right temperature, they will, upon cooling again, take on a weak magnetism aligned with the Earth's field, just like naturally forming rocks. In 1969 a team of geologists and archaeologists from the Australian National University were collecting samples of the hearthstones of ancient aborigine cooking fires. The problem with one particular set of samples, taken from the sh.o.r.e of a dried-up lake called Lake Mungo, was that the direction of magnetization was the wrong way round: what should have been north was south, and vice versa.

The phenomenon of reversed rock magnetism was not in itself new-in fact, the observation that naturally formed rocks were frequently found to be magnetized the wrong way round was one of the things that had aroused interest in rock magnetism in the first place. Attempts to devise various mechanisms of self-reversal to account for this all came to nothing, which led to general acceptance of the only alternative explanation available: that the Earth's field itself must have reversed. The picture finally put together from the data indicated that the Earth's field has been reversing itself at irregular intervals and for erratic periods for as far back as reliable measurements can be made-about ten million years-and there's no reason to suppose that the same thing wasn't going on before that. Some "flips" lasted for only a few thousand years, while others persisted for over half a million; on average, the Earth's field seems to have spend about half its time being "right" and the other half "wrong"-which agrees with the observed fact that about half the rock samples studied have reversed magnetism. The significance of the Lake Mungo results was that they show the reversals to have continued into modern times: the samples date from about thirty thousand years ago, which to geologists is only yesterday.

So what?

Well, it leads up to the answer to the second question we started out with which was: If the continents were joined together once, why aren't they now? In other words, What makes them move apart? To link this back to rock magnetism, we have to move our investigation off the continents and down into the deep-ocean basins.

Nineteenth century sailors were well aware of the Mid-Atlantic Ridge, a submerged mountain chain, miles high, that runs all the way down the middle of the North and South Atlantic, from far above Iceland (Iceland is a part of it that protrudes above the surface) to the Antarctic. Later surveys revealed that this formation is just part of a network of ocean ridges that encircle the globe and run through every ocean-although not always down the middle. They can be seen clearly on any world map that shows submarine contours.

In 1960, Professor Harry Hess of Princeton University developed the idea that these ridges were the crests of upflowing currents of fluid rock rising and solidifying from deep below the crust. Material from the underlying mantle was flowing upward to form the ridges, and then spreading out sideways, cooling and sinking as it became more dense, to form newly created ocean floors-visualize two Niagaras face to face and flowing backward. Hess came up with a figure of several centimeters a year for the rate at which the ocean floors are spreading away from the ridges. This doesn't sound much, but it turns out to be sufficient for all of today's ocean floor to have been created within the last 200 million years, which is less than 5 percent of geological time. This accounted well for other facts that had been coming to light about ocean beds, and which up until then no one had been able to explain. For one thing, the sediment layers on the ocean floors were far thinner than they should have been if the sediments had been acc.u.mulating throughout Earth's history. And for another, no sample of sediment from the seabeds had ever been recovered that was older than 100 million years. Both these results accorded well with the idea of seafloor spreading . . . but it was all very new, and many scientists felt uneasy about it. Something more conclusive was needed.

It came from other scientists who were sailing research s.h.i.+ps around the oceans while the debate was in progress, measuring the magnetic fields above seabeds. The results they were getting were peculiar. They discovered distinct zones of strong and weak fields all over the oceans, and the change from one kind to another could occur over a few miles. Furthermore, the zones were not scattered randomly, but formed well-defined stripes, hundreds of miles long, of alternating field strength. In 1963, two researchers at Cambridge University, Fred Vine and Drummond Matthews, connected them with Hess's ideas and published a short paper that was to mark the beginning of a new era in the Earth sciences.

The Vine-Matthews theory pointed out that if the ocean floors were indeed spreading sideways from the ridges and solidifying from a molten state, the kind of magnetic variations observed would be just what should be expected from the periodic reversals of the Earth's field. A strip of material emerging along the crestline of a ridge at a time when the Earth's field was in the same direction as it is today would be magnetized in the "right" direction. At some later time, this strip will have been pushed a distance away from the ridge-although still lying roughly parallel to it-by newer material appearing behind. If the Earth's field had reversed during this interval, the material currently forming along the crest would be magnetized in the opposite direction. Eventually a whole series of strips would appear, alternately magnetized to either add to or subtract from the Earth's field, depending on its direction at a given time, and this would result in precisely the kind of pattern found by the survey s.h.i.+ps. The pattern would, in effect, const.i.tute a series of time markers written magnetically onto the ocean floors, becoming progressively older with increasing distance from the ridge crests.

If this theory is correct, the sequence of field reversals read from the ocean-floor stripes should tally with that deduced from continental rocks. Subsequent comparisons confirmed this to be the case. Also, the patterns of stripes on opposite sides of the ridges are found to be symmetrical about the crestlines, as would be expected from outward spreading in both directions-in the Atlantic, for example, the pattern east of the ridge runs north-south and is a mirror image of the pattern west of it. And finally, rock and fossil samples brought up from the ocean beds by drilling s.h.i.+ps tell a continuous story of the ages of the Earth, just like the exposed strata of a cliff face, such as the Grand Canyon; the difference, however, is that the record on the seabed reads sideways, with the youngest rocks lying closest to the ridges and the oldest thousands of miles away-again, just as the theory predicts.

So there you have it-the world's oldest magnetic tape recorder. Admittedly its specification might look a little odd to anybody familiar with conventional computer magtapes. A regular magtape, for example, moves at a speed of 45 inches per second against this one's 0.000000037 inches per second (3 centimeters per year), writes at a data density of 1,600 bits per inch, compared to 0.0000032 (1 bit every five miles), and writes at a speed of 7,200 bits per second, versus 0.00000000000013 (1 bit per 250,000 years). That's really pretty bad as performance standards go, and I can't see IBM selling very many. On the other hand, though, we shouldn't forget its high reliability (at least 200 million years without a breakdown, and almost certainly a lot longer), superb rugged construction (unaffected by burial beneath a few miles of seawater and several billion tons of mud), and state-of-the-art technology (obsoletes all earlier models). Any offers?

This all seems to add up to a fairly satisfying answer to the questions we began with. But as tends to be the case in science, one question answered raises others that weren't there to begin with. Specifically, in this instance, now that we've established the continual creation of new crust at the ocean ridges, we find ourselves forced to ask what happens to it afterward. If it remains in existence indefinitely, the Earth would have to be getting bigger to make room for it. But there are many reasons for believing that the Earth isn't expanding. Therefore the appearance of new crust at the ridges must be balanced by the disappearance of old crust somewhere else. Where, then, and how, is the old crust disappearing? The answers this time lie in the realm of earthquakes and volcanoes.

Earthquake and volcanic activity is mainly confined to a network of narrow belts encompa.s.sing the globe-the familiar ones that run across dry land, plus the ocean ridges, whose activity qualifies as volcanic. These belts divide the surface of the earth into about a dozen irregularly shaped "plates,"

ranging in size from the Pacific Plate, almost as large as the ocean, down to the Turkish Plate, not much bigger than Florida, which fit together like an enormous piece of spherical crazy paving. The new science of "plate tectonics" (from a Greek word meaning builder) sees changes in the Earth's surface features as the results of continuous plate movements. The plates are about forty miles thick and carry the continents with them like lumps of solidified low-density slag that have floated to the top of the denser material beneath. Wegener's original concept of continental drift is not really accurate in this light; it is the plates that "drift," and the plate margins that separate them bear no resemblance to the outlines of continents.

The two plates that meet along a plate margin can be moving in any of three ways: away from each other, sideways past each other, or toward each other. Gaps between plates don't happen; when plates move apart, new material from the mantle below flows upward to form new crust which welds itself onto the trailing edges-this is the process taking place at the ocean ridges. These are called "constructive"

margins. Margins that consist of plates slipping past each other are called "transform faults" and account for many of the world's earthquake belts. The pressure across a transform fault is enormous and causes the plates to weld at the edges, with the result that they become progressively more deformed as the rest of each plate moves on regardless. Eventually something has to give. When the stress exceeds the strength of the rock, the weld fractures and the sides of the fault spring past each other in opposite directions to catch up with their respective plates. The movement may be only a few feet, but it can represent the explosive release in a matter of minutes of stress energy that has been building up for centuries-often, of course, with devastating results. Once the stress has been relieved, the pressure creates a new weld and starts the cycle over again. The San Andreas Fault through San Francisco is a margin of this type-between the Pacific Plate, which is moving northwest, and the American Plate, which isn't.

The third type of margin occurs where two plates are moving toward each other, and again something has to give. What happens in this situation is that one plate is deflected downward beneath the other and plunges back into the molten material of the mantle whence it came. That's how old crust is destroyed.

The deepest parts of the oceans are the long, narrow "trenches," which can extend down to thirty-thousand feet. The trenches mark the margins between colliding plates and result from the bending downward of the surface of one plate as it is forced beneath the other. Examples are the j.a.pan Trench and Marianas Trench in the western Pacific, the Tonga Trench north of New Zealand, the Peru-Chile Trench off the west coast of South America, and the Puerto Rico Trench east of the Caribbean. That's where old crust is being destroyed. These are called "destructive" margins.

As a map will show, these deep-ocean trenches always lie alongside chains of islands. The islands are formed from the acc.u.mulated sediments on the upper surface of the descending plate, which are sc.r.a.ped off as it slides beneath the overriding plate. The friction between the two plates generates sufficient heat to melt the rocks involved in the process, and much of this molten material finds its way back to the surface, making them volcanic island chains. And finally, the plate movements do not occur smoothly, but as alternations of sticking and slipping as was the case with transforms faults; hence, we end up with earthquake-p.r.o.ne volcanic island chains. In fact it was the a.n.a.lysis of earthquake shockwaves and of the heat-flow patterns obtained in the ocean trench regions that enabled this process to be understood to the point that clear profiles can be reconstructed of descending cold slabs of oceanic crust material which today are melting back into the mantle hundreds of miles beneath the surface.

Sometimes two plates meet head-on that happen to be carrying continents on their backs. As we have already seen, continental crust is too light to sink down into the mantle with the descending plate. Instead, it remains on the surface and collides with the other continent to build mountains. Mountains form in long, thin chains because plates collide along long, thin margins. So when India drove underneath Tibet, the continental crust being carried from both directions piled up to form the Himalayas. Africa's voyage northward drove Italy into Europe like a battering ram to create the Alps, and shocks radiated outward from the impact point to produce effects visible far away today-for example, the shallow folds that form the "Downs" of southern England. Hence, contrary to some ideas that were in vogue early in the century, it doesn't look as if our planet is destined to be eroded down into an enormous billiard ball. New mountains are being built as quickly as the old ones are being worn away, and they'll continue to be built for as long as the energy source that drives the plates lasts out.

So, what does drive the plates?

This brings us right up to date and into the area of much of the research going on today. Hess's original proposal was that the upflowing material at the ridges pushes the plates apart, but later findings have made this improbable, or at least, insufficient in itself. A more recent idea is that deep-seated convection flows in the mantle give rise to horizontal currents beneath the crust that drag the plates with them.

Another theory attributes the movements to the release of gravitational energy as the plates slide downward along the gradients between the ridges and trenches. No conclusive choice seems to warrant being singled out at present, but the causes are undoubtedly complex and could turn out to involve all three of these mechanisms and maybe more, perhaps operating in varying degrees in different places.