Krakatoa: The Day the World Exploded: August 27, 1883 - Simon Winchester (2003)

Chapter 8. THE PAROXYSM, THE FLOOD AND THE CRACK OF DOOM

And I thought: I would give all these people's lives, once more, to see something so beautiful again.

- Dutch pilot in Anjer, quoted in ‘Krakatau’, a short story by the novelist Jim Shepard, 1996

1. The Event

The death throes of Krakatoa lasted for exactly twenty hours and fifty-six minutes, culminating in the gigantic explosion that all observers now agree happened at two minutes past ten on the Monday morning, 27 August 1883. The observers, as is often the way of such things, agree on precious little else. Thousands of people, far and wide, suddenly became aware of the events in the Sunda Strait - but their accounts of it, like the accounts of any monstrous and traumatizing event, present today a morass of conflict and confusion.

The countdown to the final hours of the mountain's existence properly began at six minutes past one in the afternoon of the previous day, the Sunday. All across the colony, Dutchmen and Javanese alike were looking innocently forward to the long lazy stretch of the afternoon of a much needed day of rest.

The first indications that all was not right became apparent more or less simultaneously to a number of people near by. They were at the time almost all completing the last moments of their familiar Sunday ritual: pushing their chairs away from the luncheon-table, folding their napkins, draining the last dregs of the coffee, standing up and stretching their legs, picking up their cigars, dogs and wives for the Dutch tradition of the afternoon family walk.

In Anjer, from where most of those very early reports originated, the relaxed mood of the afternoon must have seemed peculiarly suited to the place. Anjer was a sedate, pretty little port-town, as pleasant a posting for a visiting Dutchman as it might be possible to find. It was situated in a shallow bowl in the volcanic coastal range, a place where the hills dipped steeply down to the sea and formed a cosily protective natural harbour. The beaches were wide and white, the fringing palm trees leaned into the trade winds, there were flowers and banyan trees and a paradise of birds, and everywhere a heavenly scent of spices.

The local people lived in kampongs of small thatched cottages, the colonials in neat white stucco houses with red roofs. The finer of these mansions - some, like that of the Assistant Resident,* sporting a Dutch flag flying from a staff in the middle of its magnificent lawns and a private dock with the official's impeccably maintained launch - could be seen to best advantage from out at sea, where they seemed separated from one another by acres of deep-green jungle. There was usually a flag-signal flying from the pilot-station that spelt out: CALL HERE FOR MAIL; and the inbound ocean-going ships would indeed invariably call, giving with their presence an air of vibrancy to the town; and, because they were stopping for orders and not for stevedores, Anjer was free of the slums and squalor of a cargo port.

The Europeans, in their white tropical suits and topis, would leave their houses - servants had their Sundays off - and saunter on an afternoon along broad seafront avenues, under groves of tamarind trees. There would be crowds of local Javanese, children running everywhere, pye-dogs sleeping under upturned boxes, chickens, pigs, goats, creaking bullock-carts, insistent pavement salesmen - all of the carefree magic of an Eastern street, in other words, out for entertainment and fun on what seemed likely to be an easy, lazy summer Sunday.

Relations between colonizers and colonized in the East Indies were less than perfect - indeed much less than perfect, for the Dutch were not very kindly in the ways they wielded their imperial powers, and they are consequently remembered with much less affection today than are most of those other Europeans who ruled far-flung territories around the globe. Yet, by all accounts, on this particular Sunday any feelings of antipathy were soothed and muted by the holiday mood. People smiled at one another, Dutchman muttered cordial greetings to Javan, everyone sauntered contentedly along in the broad heat of the early afternoon.

And then, without warning, from out to sea in the west - a sudden sound.

The first two accounts to be written were of a kind that was to be repeated, in essence, many dozens of times that day. ‘We plainly heard,’ wrote one, ‘the rumbling of an earthquake in the distance.’ ‘We didn't take much notice at first,’ wrote a second, ‘until the reports got very loud.’

The newly appointed Anjer telegraph-master, Mr Schruit, was once again idling on the veranda of the Anjer Hotel, which was owned by his new friend, the confusingly named Mr Schuit, the local Lloyd's agent. It was clearly Schruit's preferred place to spend his Sunday mornings. He was a young man, and cut rather a lonely figure. His wife and children were still in Batavia, and, after weeks of searching, he had only now found them a suitable house in Anjer, and was at long last looking forward to a family reunion.

But until then, as for the past few months, his holiday Sundays were spent on the veranda, puffing on a cheroot, gazing out at the view. Three months earlier he had been there too, trying under Schuit's tutelage to decipher the flag message that was flying from the yards of the German warship Elisabeth, then sailing down the Strait - a message which almost certainly told of the first eruption. He did nothing about it then: he was new to the area, he was not on duty, and in any case the local Lloyd's agent had the matter in hand, was well able to read the Elisabeth's flags, and was already sending his historic dispatch about what the telegraph operators would misread as ‘Krakatan’ spewing out its great clouds of smoke and ash.

But on this occasion Schruit was on duty, and much more observant than before. He remembered later seeing a fully rigged barque heading north, a billow of white sails gliding along the blue mirror of the sea. Then he spied the much less pretty steamer Gouverneur-Generaal Loudon, the locally familiar government-chartered vessel that had taken the eighty-six tourists to Krakatoa back in May, heading into Anjer port.

(On this occasion she was undertaking more mundane and customary tasks: first picking up a hundred coolies who had been hired to help build a local lighthouse and ferrying them across the Strait; then going on up the eastern Sumatran coast to the troublesome region of Aceh, of the north, and there delivering, among others, some 300 miscreants, all members of a chain gang destined to be put to work on a variety of government building sites.)

It was at the very moment when Schruit was watching the Loudon steaming towards the safety of port that there came the first roar of an explosion.

It was an extraordinary sound, he thought: far, far louder than anything he recalled from before. He looked sharply over to his left and saw, instantly, the unforgettable sight of a tremendous eruption. To judge from the billows of white smoke that were now tumbling up from the mountain - ‘as if thousands of white balloons had been released from the crater' - it was a far, far larger eruption than anything he had witnessed when he was standing at this exact spot back in May.

Moreover, whatever was happening on the mountain was also having an immediate effect on the sea. It was rising and falling, strongly, irregularly, in bursts of sudden up-and-down movements of the sea-water that seemed immediately unnatural and sinister. It wasn't tide, or wave, or wash: it was some terrific disturbance, and the water was slopping up and down, dangerous and unpredictable.

He ran down to the beach, where he had spied his deputy telegraphist watching the eruption, transfixed. He said he too was utterly perplexed by the movement of the sea. Perhaps the tide was on the turn, he said. But, as he spoke, a furious rush of water roared towards them, sending them scurrying back up to the roadway. The two men ran to the small white stone building that was the Anjer telegraph office - and, as they did so, the enormous cloud from the volcano began to drift down on them. Within moments all Anjer was enveloped in dust and cloud and became strangely dark.

Some remember the cloud as black; others, like Schruit, are equally certain it was white. One of the pilots waiting for orders at the Anjer pilot-station, a Mr de Vries, swore it alternated in colour, from white (when presumably it was largely made up of steam) to black (when it was composed largely of eruptive smoke). But no matter: it was so thick and heavy that within moments an artificial night had descended on Anjer port, and the two men who groped their way to the cable office then found they had to light lanterns, in the middle of the afternoon, in order to be able to send their first message. They timed it at 2 p.m. Krakatoa, they tapped out in urgent Morse to their head office in Batavia, was beginning a major eruption. It was ‘vomiting fire and smoke’. It was so dark in town that it was now no longer possible to see one's hand before one's eyes. What were the instructions?

Batavia replied, with equal urgency. Yes, they had already become aware that something was taking place. The governor-general himself, his sabbath ease disturbed, had been inquiring. People were milling about in the streets, worried. Chinese traders in particular seemed to have a peculiar sense of unease: the cable office reported hearing wailing. So it was important for Anjer to keep its telegraph station open and the information coming in, the operator tapped out. For the next six hours the operators did just that, giving the Dutch officials in Weltevreden (the name, ‘Well Contented’, must have prompted on this occasion a sardonic smile or two) a moment-by-moment chronicle, in the staccato language of the telegraph, of the unfolding events. ‘Detonations increasing in loudness.’ ‘Hails of pumice.’ ‘Rain of coarse ash.’ ‘First flooding.’ ‘Vessels breaking loose in harbour.’ ‘Unusual darkness.’ ‘Gathering gloom.’

The Anjer harbour-master, who was by now aware that the crisis was frightening many of his friends and colleagues - ‘the Day of Judgement has come' was a common belief - tried to collect as many of the local expatriates together as he could, to reassure them. How he imagined he might accomplish this is left unsaid. But he did manage to assemble a fair number of the colonial establishment - the Assistant Resident, the public works supervisor, the lighthouse keeper, the registrar, the town clerk, a local doctor and a prominent local widow - and tried to tell them that what they were witnessing would soon blow over, that it was, in his considered and experienced view, nothing to worry about. He could hardly have been more wrong.

At 2.45 p.m. the Loudon, all passengers aboard, set off for the forty-mile journey to the port of Telok Betong, at the head of Lampong Bay across in Sumatra. Her master, Captain Lindeman, steamed well to the east of the exploding island, trying as best he could to avoid the showers of rock and ash cascading down from the plumes of smoke. One British ship in the vicinity, the Medea, estimated that by mid-afternoon the column had risen to a height of seventeen miles, more than three times the height of Mount Everest; the Medea's Captain Thomson said there were ‘electrical displays’ in the cloud, and explosions every few minutes were shaking his ship - even though he was at the time at anchor off Batavia, more than eighty miles to the east.

In the centre of the capital, meanwhile, people were very rapidly becoming aware that matters were getting out of hand. Two seasoned observers of Krakatoa's earlier throat-clearings - Dr J. P. van der Stok down at the Observatory in Batavia and the mining engineer Dr Rogier Verbeek up in the hills above town - had already telegraphed one another to find out what was going on. Van der Stok - the man whose wife had lost her delft dinner-plate in the May eruption and who himself had noted with great precision the time of the very beginning of the earlier events - once more swung into official Observatory mode, even though yet again this was a Sunday. He checked his watch at the very moment he heard the first loud rumblings, dashed from his house to the Observatory buildings and wrote the time down in the official log: 1.06 p.m. That time remains today, etched in official records, as the one known certain commencement of Krakatoa's final phase.

*

Confirmation of these figures then came from a totally unexpected source: the city's gasworks, to the south of Batavia old town. They proved, quite uncannily, to be of the greatest use to scientists who later studied the eruption. And they did so, quite simply, because of the way they had been built.

The most visible parts of any plant that produces gas from coke are the tall drum-shaped metal containers for the flammable gas - containers that are in essence telescopic, which ‘float’ on enormous ponds of water or mercury and which grow taller or shorter, higher or lower, depending on the amount of gas pumped from the works to be stored inside them. (Today's gasometers tend to stand tall against a city skyline in the morning, fall gradually over the day as the gas inside them is consumed, and then are replenished as more gas is manufactured overnight. In those cities like Batavia in the 1880s that had gas street lamps, the profile of variations in pressure, and thus the times of the varying heights of these gas containers, would necessarily be rather different. They would stand tall in the early evening and fall away during the dark, during the gas-illumined hours.)

What superintendents of gasworks have long known is that these storage containers - commonly, but in fact wrongly, called gasometers - also act as gigantic barometers. The pressure inside them, and in the lines leading from them, goes up and down by infinitesimal amounts according to the rise and fall of atmospheric pressure outside them. Normally one would never notice these small amounts of movement, since they would be superimposed on the much larger movements resulting from the consumption of gas. But an eagle-eyed superintendent, in charge of arranging the pressure in the gas lines according to demand, could indeed notice. Moreover, a paper register of the gasometer pressure is always produced (and was, for most of the time, in Batavia), which would show all the movements with great precision.

The pressure records are being taken constantly, but they can in fact record the minute fluctuations in the atmospheric pressure

The invisible and inaudible pressure wave from Krakatoa's cataclysmic final explosion, measured - until it blows off scale - at the Jakarta Gasworks.

(and hence also the fluctuations caused by an event like Krakatoa) only when the base pressure is low enough for the recording meter to be affected by them. The gasworks superintendent would increase this pressure each evening, when the street lights were illuminated; he would keep it high until the middle of the evening; and he would then lower it in hourly stages until dawn. So the best recordings of changes in the atmospheric pressure would be made during daylight hours, when the pressure in the gas lines was kept low because of the low demand.

Which is exactly what happened - after a hiccup. For when the very first explosions occurred at lunchtime on Sunday there is no record, mysteriously. But then whatever problem existed vanished, and the recording trace begins properly at 3.34 p.m. Batavia time (which, since this was still some while before the formal international establishment of time zones, was a little more than five minutes ahead of Krakatoa time).* From that moment on until dusk, and then throughout all of Monday morning after dawn, the Batavia gasworks pressure gauge provides an incredibly accurate, minute-by-minute record of the massive air-pressure waves that radiated out from the volcano, each and every time it erupted. The paroxysmal eruption itself at 10.02 on Monday morning blew right off the scale: it caused a pressure spike of more than two and a half inches of mercury, unheard of in any other circumstance.

By five o'clock on Sunday evening, when in normal circumstances ordinary civil twilight would be only an hour away, it was, in fact, nearly totally dark up and down the entire west Java coast, and was becoming similarly so in the capital. It was at this point that enormous chunks of pumice began to rain down from the sky.

There were three European ships inside the narrowest part of the Sunda Strait at the time - the Loudon with Captain Lindeman, which, because of the bucking sea, failed to reach Telok Betong and so anchored well out in Lampong Bay instead; the Danish salt-carrying barque Marie, which also rode out the enormous and ever-growing waves in the same Sumatran bay; and the cargo-carrying barque Charles Bal, reaching the end of her long voyage from Belfast to Hong Kong. All three vessels were deluged with pumice: dangerous, heavy, sharp, fast-moving masses of rock, the larger pieces still warm to the touch.

The captain of the Charles Bal, W. J. Watson, found himself in a more perilous situation yet - somehow embayed, horribly pinioned in the sudden dark rain of rocks and compelled to beat around purposelessly, navigationally blinded. He was for a long while during the Sunday night just ten miles away from Krakatoa, closer to it than anyone else who survived. It enabled him to leave a record that was vivid in the extreme - except that chronologically (since it is thought that in the confusion he forgot to set his bridge chronometer to Batavia time) it was different by a single hour from everyone else's.

We first encounter Captain Watson* when he was beating northwards, with Java Head and First Point, Welcome Bay and Pepper Bay to his starboard, the great mountains of Sumatra to port, and the islands in the narrows of the Sunda Strait directly ahead. Then suddenly, at what he incorrectly records as 2.30 p.m. (it was in fact only 1.30 p.m.):

... we noticed some agitation about the point of Krakatoa, clouds or something being propelled from the northeast point with great velocity. At 3.30 we heard above us and about the island a strange sound as of a mighty crackling fire, or the discharge of heavy artillery at one or two seconds' interval.

At five the roaring noise continued and was increasing [wind moderate from the SSW, notes Captain Watson here, his mariners' routines never quite deserting him]; darkness spread over the sky, and a hail of pumice-stone fell on us, of which many pieces were of a considerable size and quite warm. We were obliged to cover up the skylights to save the glass, while our feet and our heads had to be protected with boots and sou‘westers.

… we sailed on our course, until at 7 p.m. we got what we thought was a sight of Fourth Point light; then brought the ship to the wind, SW, as we could not see any distance, and knew not what might be in the Strait.

The night was a fearful one; the blinding fall of sand and stones, the intense blackness above and around us, broken only by the incessant glare of varied kinds of lightning, and the continued explosive roars of Krakatoa made our situation a truly awful one.

At 11 p.m…. the island became visible. Chains of fire appeared to ascend and descend between it and the sky, while on the SW end there seemed to be a continued roll of balls of white fire. The wind, though strong, was hot and choking, sulphurous, with a smell as of burning cinders, some of the pieces falling on us being like iron cinders. The lead came up from the bottom at thirty fathoms, quite warm.

From midnight to 4 a.m. of the 27th… the same impenetrable darkness continued, while the roaring of Krakatoa less continuous, but more explosive in sound; the sky one second intensely black, the next a blaze of light. The mast-head and yard-arms were studded with corposants* and a peculiar pink flame came from fleecy clouds which seemed to touch the mast-head and the yard-arms.

At 6 a.m., being able to make out the Java shore, set sail, and passed the Fourth Point lighthouse. At 8 a.m., hoisted our signal letter, but got no answer. At 8.30 passed Anjer with our name still hoisted, and close enough in to make out the houses, but could see no movement of any kind; in fact, through the whole Strait we did not see a single moving thing of any kind on sea or land.

At 10.15 a.m. we passed the Button Island, one half to three quarters of a mile off; the sea being like glass all around it, and the weather much finer looking, with no ash or cinders falling; wind light, at SE.

At 11.15 a.m. there was a fearful explosion in the direction of Krakatoa, then over 30 miles distant. We saw a wave rush right on to the Button Island, apparently sweeping entirely over the southern part…

... by 11.30 we were enclosed in a darkness that might almost be felt, and then commenced a downpour of mud, sand, and I know not what… we set two men on the lookout for‘ard, the mate and the second mate on either quarter, and one man washing the mud from the binnacle glass. We had seen two vessels to the N and NW of us before the sky closed in, adding not a little to the anxiety of our position.

At noon the darkness was so intense that we had to grope our way about the decks, and although speaking to each other on the poop, yet we could not see each other. This horrible state and the downpour of mud and debris continued until 1.30 p.m., the roaring of the volcano and the lightning from the volcano being something fearful. By 2 p.m. we could see some of the yards aloft, and the fall of mud ceased; by 5 p.m. the horizon showed out to the northward and eastward, and we saw West Island bearing E by N, just visible. Up to midnight the sky hung dark and heavy, a little sand falling at times, and the roaring of the volcano very distinct, although we were fully 75 miles from Krakatoa. Such darkness and such a time in general few would conceive and many, I daresay, would disbelieve. The ship, from truck to water-line, was as if cemented: spars, sails, blocks and ropes were in a terrible mess; but thank God!, nobody hurt nor was the ship damaged. But think of Anjer, Merak and other little villages on the Java coast!

Other ships, more distant, experienced even more drama. The Berbice, a German paraffin-carrier bound from New York under the command of a Glaswegian, William Logan, found herself in a peculiarly exposed situation. When Logan saw the towering black clouds and lightning flashes ahead of him, from his position in the Strait's western approaches, he supposed it to be no more than a tropical storm. But as soon as flaming ashes began to fall on deck - a wooden deck that was only inches from his highly flammable cargo - he recognized what was going on, understood the perils of his position and promptly hove to in the lee of a protective island. He huddled there for the next two days - even though the island, by all accounts, afforded him precious little protection:

The lightning and thunder became worse and worse. Lightning flashes shot around the ship. Fireballs continually fell on deck and burst into sparks… The man at the rudder received heavy shocks on one arm. The copper sheathing of the rudder became glowing hot from the electric discharges.

The electricity in the air proved an even more serious problem, as Logan later recalled in an interview with an Australian newspaper:

Now and then when any sailor complained that he had been struck, I did my best to set his mind at ease, and endeavoured to talk the idea out of his head until I myself, holding fast to the rigging with one hand, and bending my head out of reach of a blinding ash shower which swept past my face, had to let go my hold, owing to a severe electric shock in the arm. I was unable to move the limb for several minutes afterwards.

Logan's crew were gripped with terror, volcanic dust covered the ship ‘at least eight English thumbs deep’, his masts and sails were alive with fire and sparks, his barometer fell impossibly low, all the ships' chronometers mysteriously stopped, and the world beyond him was concealed in a frequently impenetrable miasma of whirling dust and smoke. His account does, however, conceal one note of incongruous optimism.

As well as his thousands of gallons of paraffin, Captain Logan carried in his cabin a small package wrapped in brown paper and tied with string, addressed to the curator of the Botanical Gardens at Buitenzorg. Inside were five specimen seedlings of a variety of spurge found in the forests of the Amazon, known as Hevea brasiliensis - wild rubber. There had already been numberless plans to harvest rubber commercially from such trees where they grew wild in Brazil, but, for a variety of reasons,* all of the schemes failed: the Berbice was now bringing Amazonian plants to the Indies from which, it was hoped, plantation rubber could in due course be grown. The conditions of weather and soil made it likely that it would grow well in the East, the botanists predicted: dozens of related plants, like tapioca, castor bean and poinsettia already flourished there.

And this particular small story does have a happy ending. Despite the rigours of his passage through the Strait, Captain Logan, together with his ship, his cargo of paraffin and his infant rubber trees, all survived Krakatoa; and the parent-plants of what are now some of the most economically important rubber plantations in the world remain today in the Buitenzorg Botanical Garden, duly and safely delivered.

But otherwise the story of that long Sunday night makes for grisly reading. ‘Everything became worse,’ wrote an elderly Dutch pilot at Anjer. ‘The reports were deafening, the natives cowered panic-stricken, a red fiery glare was visible above the burning mountain.’

Hevea brasiliensis, Brazilian rubber.

At 6 p.m. the cable linking Anjer and Batavia finally broke - the line going dead at the very moment Telegraph-Master Schruit was telling government officials that yes, the eruption was continuing and indeed intensifying still. Schruit, tapping frantically at his Morse key, found he could not even make contact with the small town of Merak, seven miles up the coast. With his assistant telegraph operator in tow, he promptly dashed out into the gloom, ran through the old Dutch fort, fully intending to press on up the coast road to find and repair the rupture. He found it soon enough, just as he reached the drawbridge at the mouth of the harbour:

... there, a fearful sight met my eyes: a schooner and twenty-five or thirty prahus were being carried up and down between the drawbridge and the ordinary bridge as the water rose and fell, and nothing remained unbroken, including the telegraph wires which had been snapped by the schooner's mast.

But we felt no alarm as the water did not overflow its banks. Not entertaining any idea of danger, I sat down to table at about half-past 8. Of course I had made the necessary arrangements for beginning the repair of the broken line the first thing in the morning.

It was not to be. Anjer would not speak to Batavia again for the duration of this crisis - and Batavia would thus be wholly unaware of the terrible fate that would soon befall the town, and would befall all its neighbour villages up and down both the Java coast and across on the far side, in Sumatra.

The astronomical logs for 26 August note that civil twilight in Anjer port began that night at 6.22 p.m., half an hour after the sun had set, when artificial lights were first needed in the street; nautical twilight, the time when the horizon ceases to be the sharply delineated line that a navigator deems essential for working with his sextant, began at 6.47. Both periods would in normal circumstances endure for thirty minutes. This night there was no such thing. It had been dark in Anjer since mid-afternoon, and when the invisible sun did set, the darkness was Stygian indeed the air a hot, ashy breath, filled with grit and sulphur, disorienting, confusing and poisonous.

By late evening it was the turn of the ocean to take over as the more terrifying manifestation of Krakatoa's gathering power. As the great volcanic engine pumped and stoked more and more explosive energy into the atmosphere, so the sea surrounding the dying mountain became progressively more and more disturbed - and communities that were already huddling, frightened, along the low coastline of the Strait began to experience ever greater waves, ever more dangerous seas.

The reports from after sunset speak continually of smashed boats and inundations of low-lying land, of ruined houses and of bystanders pulled off their feet into the raging waters. The first and most melancholy accounts were those later given to a number of newspapers in Java by the colonial contrôleur in the south Sumatran town of Ketimbang, Willem Beyerinck - the man who had given the first official news of the impending eruption back in mid May, when he telegraphed to the Resident of Lampong to say he had felt an outbreak of ominous tremors.

From that first Sunday afternoon, Mr and Mrs Beyerinck and their three children were to endure a week of the most exquisite agony - much of which they remembered well; and by so doing they provided one of the more reliable chronicles of this very complicated series of events.

The Sunday had begun innocently enough, with the opening of a new village market. There was the ritual slaughter of a baby buffalo, the playing of a gamelan orchestra, perhaps a wayang kulit puppet show - the kind of ceremonial that the Beyerincks had seen countless times during their tour. But these were strange times. Krakatoa had begun to rumble again that afternoon, and it somehow cast a shadow over an opening that was not, in consequence, an entirely happy affair. There had been a local outbreak of cholera too - a housemaid had just died, and Mrs Beyerinck was worried about the health of her children. The children's ayah had seemed agitated for other reasons - complaining among other things that the birds that normally flocked around the family house had lately seemed restless, and that the auguries were not good. And Mrs Beyerinck, warily watching the smoke roiling about the summit of Rakata, was wise enough and prescient enough to accept that one ignored the superstitions of the local people at one's peril.

When they returned from the market, Mrs Beyerinck made what at the time seemed a strange request - that the family not go home, but make right away for a tiny village in the hills, where they rented a holiday cottage. Her husband, however, wouldn't at first hear of it. The locals, he said, would wreak havoc in Ketimbang town if he did so; and the wilder elements who had only lately been hired to pick the summer pepper harvest up in the highlands would soon hear that the Dutch controleur had run away, and would descend on the town in short order. No, he declared; the family would stay. Mrs Beyerinck went off to her room to sulk, only remembering later the distant tinklings of the gamelan and the rhythm of a great drum sounding what she thought was a threnody.

But then everything changed, very suddenly. Her husband sauntered down to the shore to see what effects the new eruption might be having - and came on a scene that astonished him. While in the distance the mountain was roaring and boiling from behind an immense pillar of clouds, here enormous waves were breaking on the beach, and the level of the sea was piling up, rising and falling alarmingly, crashing with a weird randomness against anything solid on the shore. There was no wind, no storm. But the surface of the sea had a terrible, writhing, coiling awfulness about it.

He could see the Loudon, with its cargo of Anjer coolies bound for the pepper fields of Telok Betong, beating up Lampong Bay towards him, then trying desperately to dock. It was being tossed every which way, one moment corkscrewing high on the crest of a huge mass of water, the next being twisted as if by an unseen hand and plunged deep down into a trough. The master then evidently lost his nerve and gave up the struggle, for, as Beyerinck watched in horror, the boat, now looking so vulnerable and fragile, suddenly turned away, presumably to try to ride out the fury in mid channel.

The contrôleur stared, now momentarily dumbstruck, as the thrashing waters rose higher and higher up the shore, soon reaching the outbuildings of his own Residency. Water began crashing against the stucco, breaking hard against what looked like increasingly insubstantial structures. It was this very sight that finally made up his mind for him. His earlier decision had been wrong, he told his servants: Mrs Beyerinck and the children should leave right away. Everyone should flee to their summer cottage up in the hills.

For a few moments, it seemed unlikely that they would make it. For at 8 p.m., as a hail of pumice began to rain down, the waves began their first orgy of destruction. They were eventually to reach well over a hundred feet in height, and right from the start even the precursors of the mighty waves, even the first tentacle-feelers of water, did the most amazing damage. In an instant Beyerinck's office suddenly came crashing down, along with a clutch of outbuildings. The family and their servants escaped drowning only by shinnying up coconut palms and waiting until the waters receded for a few moments' respite. Then they climbed back down, gathered up valuables, set the horses and their other animals loose and ran, as far as they could manage, inland.

Their flight was the stuff of the cinema epic. There was a dreadful roaring behind them as they stumbled, half blind, frightened, soaking wet, through miles of paddy, sinking into thick mud while trying desperately to outrun the ever pursuing monster. At one stage Mrs Beyerinck, by now covered with mud from head to toe, tried to shout, but her throat was horribly sore and she couldn't utter a sound. She felt her neck - it was thick with a collar of leeches. They ran on and on, getting lost, from time to time joining forces with other local people, who themselves were fleeing in great crowds from the thundering, roaring floods behind them. Pieces of pumice hurtled down from the sky, burning fiercely like jagged meteorites.

The family and such servants as had run with them reached the hilltop cottage at midnight. They broke out supplies, fed their terrified children, settled them to some kind of fitful sleep. The adults then knelt on the slatted floor and peered through the window towards the raging volcano, which they could see distinctly through the fog of falling rock. Outside the hut lay thousands of local people, all crying and wailing in desperation. Some of the more sober were praying to Allah for relief from the nightmare.

But it was not to end for some hours yet. One of Beyerinck's servants arrived just before dawn, saying that the entire Residency had been ripped from its foundations by a gigantic wave at about 2 a.m. All the signs suggested the waters were getting higher and higher, and the entire town of Ketimbang would likely go under. And indeed when the contrôleur sent scouts downhill at dawn to see the damage, it had been destroyed, totally. An enormous series of waves had flooded over every rooftop at about 6 a.m.: nothing was left standing.

Up at Telok Betong, where the Resident later said water had come within ten yards of his house, which was perched at the top of a hill 120 feet high, there was a mass of destruction. The harbour-master, waiting in vain for the Loudon, said he was swept off his feet eight times before he ran for his life. From up on the hills surrounding the town he watched as the Dutch Navy's well-armed paddle-steamer Berouw strained mightily at her mooring buoy. It looked to him as though not only might the chains give and the vessel founder, along with her crew of twenty-eight, but that the buoy might free itself too and be hurled about the harbour, destroying all other ships there including the barque the Marie, which was waiting in the roads. Mass destruction seemed to be waiting in the wings.

And on the Loudon herself Captain Lindeman found himself dealing with the unimaginable waves, the close presence of land and a host of other ships near by (a perilous combination that sailors fear most in a storm, denying the vessel vitally necessary sea-room) - as well as a terrified crew. The locally hired men in particular were petrified by the eruption of St Elmo's fire in the riggings, and they left their posts in droves to try desperately to beat them out, to extinguish what was in fact inextinguishable. The phosphorescence was, they insisted, evidence of ghostly spirits: if these phenomena found their way below they would eat their way through the hull and the vessel would sink like a stone.

There was an almost endless succession of other, very similar reports. There were those from nine other nearby ships - like the magnificent American barque W. H Besse, forty miles north of Krakatoa; the British vessels Sir Robert Sale and the Norham Castle, both of which were hove to off Sumatra; the Norwegian

The Royal Dutch Navy's armed paddle-steamer Berouw about to be picked up by one of the giant tsunamis generated by the eruption.

barque Borjild, eighty miles north-east of the volcano; the Welsh cargo vessel Bay of Naples, under way to Singapore and 120 miles south of Java Head; the Rotterdam Lloyd's steamship Batavia, well to the south-west of the entrance to the Strait; the steamer Prins Frederik which had passed Krakatoa on 25 August and was by the time of the explosion also well out into the Indian Ocean; the Annerley, southbound, and at the time of the eruption standing to the north of the Strait; and the British-flagged Medea, whose Captain Thomson had managed to measure with some accuracy the height of the volcanic cloud at the outset of the eruption, at seventeen miles.

There were all manner of curious survivals. One man had fallen asleep at home and awoke to find that the wave had lifted him and his bed to the top of a hill, and deposited him there in perfect safety. Another grabbed on to the corpse of a cow and floated to high ground. More bizarre still - and barely credible - was the man who reportedly found himself being swept inland next to a crocodile: he clambered on to its back and hung on for grim death with his thumbs dug deep into the creature's eye-sockets.

And there were in addition statements and newspaper interviews and private letters home from lighthouse keepers and Residents, Assistant Residents, contrôleurs, agents of Lloyd's, telegraph operators, harbour-masters and sharp-eyed civilians of all stripes, as well as an enterprising member of the Catholic clergy named Julian Tenison-Woods* who wrote an exceptionally long letter about the events to the Sydney Morning Herald. From this mass of information a broad summary can be distilled.

Krakatoa's final twenty hours and fifty-six minutes were marked by a number of phases. First, from early afternoon on Sunday until about 7 p.m. there was a series of explosions and eruptions of steadily increasing frequency and vigour. From early evening the ash falls and the deluge of pumice began. By 8 p.m. the water had become the next medium of transmission of the volcanic energy, and as night fell the temper of the sea in the Sunda Strait became one of unbridled ferocity.

Then, just before midnight, a series of air waves - fast-moving, low-frequency shocks sent out invisibly and inaudibly by the detonations - began arriving in Batavia. The time-ball on the astronomical clock down at Batavia's harbour stopped dead at eighteen seconds after 11.32 p.m. because of the ceaseless vibrations. Audible evidence of the explosions began to radiate outwards too, and there was a report from Singapore and Penang* that thudding sounds could be heard at about the same time. In Batavia a large number of people, kept awake by the explosions and for want of something better to do, were walking around the Koningsplein; they noticed that the gas lanterns suddenly dimmed at about 1.55 a.m. Along Rijswijk, the main shopping street, several shop windows suddenly and inexplicably shattered at about the same time.

Then at about 4 a.m. the nature of the explosions reportedly changed, very slightly, becoming less continuous but more explosive. Someone described the sounds as like a steam-engine, emitting full-throated whoomphs as it gathered speed. At about 4.56 a.m. an enormously powerful air wave was detected at the Batavia gasworks - suggesting, if travel time over the ninety miles to the volcano is allowed for, that something else had just happened deep within Krakatoa's heart. The culminating explosion - though no one on the ground at the time knew it - was soon about to happen.

There were four gigantic explosions still to come. The first was noted at 5.30 a.m. The Sumatran town of Ketimbang was then destroyed at 6.15 a.m., and Anjer, her Javan sister-port across the Strait - according to the few who survived to tell the tale - was inundated and wrecked very shortly thereafter. The second mighty explosion came at 6.44 a.m. - forty-one minutes after a dawn that, to those in all of western Java, never arrived that day. Ashes began to fall on Batavia at 7 a.m. - although Oscar Hatfield, the American consul in Batavia, reported seeing them falling in the consulate grounds two hours later.* At 8.20 a.m. a third, quite terrible explosion was felt in Batavia, and many of the buildings started to make what were described as ‘crackling’ noises. And then finally, at 10.02 a.m., came the culminating, terrifying majesty of it all.

Two minutes to go and, according to simultaneous reports: the sky was completely darkened in all of southern Sumatra; the Loudon was weathering heavy ash falls in Lampong Bay; the nearby Marie reported ‘three heavy seas came after each other; at once a fearful detonation; sky in fire; damp’. The Annerley lit all her lights, noted that it was raining pumice stone, that the barometer was rising and falling half an inch a minute. In Batavia it became eerily dark again, and - most significantly - it started to get cold. From 10 a.m. the temperature began to fall - as many as fifteen degrees Fahrenheit over the coming four noontide hours.

Explosions like a battery of guns are heard across in Telok Betong. Lightning strikes the lighthouse at Vlakke Hoek in southern Sumatra. The lighthouse at Fourth Point, just to the south of Anjer, is hit by a vast wave and destroyed, ripped off at its base, leaving only an amputated stump of jagged masonry. An immense wave then leaves Krakatoa at almost exactly 10 a.m. - and then two minutes later, according to all the instruments that record it, came the fourth and greatest explosion of them all, a detonation that was heard thousands of miles away and that is still said to be the most violent explosion ever recorded and experienced by modern man. The cloud of gas and white-hot pumice and fire and smoke is believed to have risen - been hurled, more probably, blasted as though from a gigantic cannon - as many as twenty-four miles into the air.

‘A fearful explosion.’ ‘A frightful sound.’ Captain Sampson of the British vessel Norham Castle wrote simply in his official log: ‘I am writing this blind in pitch darkness. We are under a continual rain of pumice-stone and dust. So violent are the explosions that the ear-drums of over half my crew have been shattered. My last thoughts are with my dear wife. I am convinced that the Day of Judgement has come.’

The British consul in Batavia at the time was one Alexander Patrick Cameron; and five days later he sat down in his study and had his confidential clerk write out, in the usual impeccable sweeps of fine Victorian copperplate, his summary of what he then knew of the disaster. The document remains today in the Public Record Office in London, largely unread and unconsulted because of a confusion that has led those who have chronicled the Krakatoa events to believe the British consul was in fact a man named Henry George Kennedy.

The error is understandable. Kennedy had in fact been consul in Sumatra, and was called in to replace Cameron when the latter asked for leave in November 1883. Kennedy wrote a summary of the terrible events for the Royal Society in September 1883. His name is known to what might be called the Krakatoa community today as a result, and most indexes of most books will have a reference or two to him. Alexander Cameron, on the other hand, remains forgotten and unsung. What he wrote, though, seems today a model of diplomatic felicity, as perfect a summary of the events as could be imagined, considering the awful circumstances of the moment.

The elegant copperplate and exquisitely courteous tone of Consul Cameron's lengthy Krakatoa dispatch to Lord Granville, in London.

His report is dated Batavia, 1 September 1883, and is addressed to Gladstone's foreign secretary, the Earl Granville:

My Lord:

Enclosed I have the honour to hand Your Lordship a copy of my telegram of yesterday, giving notice of the volcanic disturbances which have lately taken place in the neighbourhood of my Consular district.

The spot where the subterranean forces have found vent is the island of Krakatau* lying in Longitude 105°27'E, Latitude 6°7'S, at the southern entrance to the Straits of Sunda. This island was the scene of a volcanic eruption of less importance on the 20th May last which, although on that occasion an entirely new crater was formed, had no such disastrous results to life and property as have attended the explosions which commenced on the 27th inst.

The present outburst commenced on Sunday last, and on that night the inhabitants of nearly the whole of Java and Sumatra were alarmed by loud noises resembling the reports of heavy artillery, which continued throughout the night and at rarer intervals during Monday 28th inst. It soon became known that these noises were produced by a fresh eruption of Krakatau and since Monday intelligence has been slowly reaching Batavia from various quarters apprising us of the extent of damage done, and proving by the loss of life and property that this is one of the greatest calamities of this century.

The residencies of Bantam and Batavia were darkened throughout the early hours of last Monday by a thick cloud of grey ashes, the light diminishing gradually, as the cloud progressed from west to east, from twilight to almost total darkness at midday, and a continuous shower of ash fell during the forenoon giving the ground an appearance as if covered by snow. At about 11.30 a.m. at Batavia and at earlier periods of the day in the more immediate vicinity of Krakatau the sea suddenly rose, presumably owing to the subsidence of part of Krakatau and other islands or to a submarine upheaval, and a wave of considerable height advanced with great rapidity on the shores of western Java and southern Sumatra, causing greater or less damage according to its distance from the centre of disturbance. A second wave higher than the previous one followed the first at an interval of about an hour with even more serious results. It is now reported that part of Krakatau island, the island of Poeloe Temposa and other small islands in Sunda Straits have disappeared, and that a reef has been formed between Krakatau and Sibesie islands, the channel usually taken by steamers. Dwars-in-den-weg/Thwart-the-Way, an island at the northern entrance to the Straits, is reported split into five pieces, while numerous small islands are said to have been raised which had no existence previously.

These reports however still require verification and with a view to ascertaining the extent and nature of the changes caused by the volcanic action a Government survey-steamer has been dispatched to the neighbourhood to take a new survey of the Straits.

The destruction caused by the waves on shore both to life and property, although known from reports already to hand to be very widespread, can hardly yet be estimated with any degree of certainty, as owing to the action of the sea and the heavy rain of ashes, telegraph and road communication has been either entirely interrupted or is much delayed.

It appears beyond a doubt however that the whole of the southeastern coast of Sumatra must have suffered severely from the effects of the sudden influx of the sea, and thousands of natives inhabiting the villages on the coast must have almost certainly perished.

The west coast of Java from Merak to Tjeringin [has] been laid waste. Anjer, the port where vessels bound for the Java and China Seas call for orders and a thriving town of several thousand inhabitants (natives), no longer exists, its former site now being a swamp.

The lighthouse at Anjer (Java's Fourth Point) has also been much damaged.

Many Europeans, including numerous officials, and many thousands of natives have been drowned, in the district of Tjeringin alone on the southeast coast of Java it is reported that no less than ten thousand persons have lost their lives. The result to agriculture in west Java [is] not yet officially known. The fact however that owing to the covering of ashes which spreads over the whole country, the cattle are deprived of their ordinary nourishment, is in itself a very serious consideration and measures have already been taken to supply the afflicted districts with food for man and beast. It is to be feared that the natives will be greatly impoverished by the damage done to fruit and palm trees which form a source of wealth, while coffee and tea gardens and standing crops of all descriptions must have suffered severely.

With a view to rendering safe the navigation of the Sunda Straits the Rear Admiral, Commander in Chief of the Netherlands Indian Navy, has stationed one man-of-war to cruise off the southern and another to cruise off the northern entrance to the Straits to warn vessels to proceed with caution.

In view of the quantity of shipping (principally British) which daily passes through Sunda Straits and the important nature of the circumstances above related I have thought it my duty to dispatch the telegrams mentioned in the accompanying memorandum, and trust my action will meet Your Lordship's approval.

I have the honour to be,

My Lord,

Your Lordship's Most obedient, Humble Servant,

A. P. Cameron

Her Britannic Majesty's Consul, Batavia

The island of Krakatoa, meanwhile, had in essence disappeared. Six cubic miles of rock had been blasted out of existence, had been turned into pumice and ash and uncountable billions of particles of dust. The rumblings and roarings continued for some while, then on Monday afternoon became ever fainter. By dawn on Tuesday they had stopped completely. That last great detonation at two minutes past ten on that Monday morning had blown the island apart, and sent most of it to kingdom come.

Now it was time for those who could, together with those whose duty it was, to venture out to see just what damage the eruption had caused.

Admiralty charts show the islands of Krakatoa before and after the 1883 disaster.

2. The Effects

It was just before dawn on the Monday, and an elderly Dutch harbour-pilot, one of those stationed in Anjer to guide ships to and from the Batavia roads, was walking on the beach. He couldn't sleep; besides, staying inside was perilous, not least because the intermittent hails of pumice stones, many of them too hot to touch, threatened either to set ablaze the atap thatch with which his house was roofed, or to smash holes in it and wreak who knows what damage inside. Much the better, he thought, to watch the great events from the comparative safety of the shore.

There was not much visible through the gloom. The clocks said that the sun ought to be ready to rise; but the falling, swirling ash had effectively dimmed the view for more than a few yards in any direction. Krakatoa herself, thundering away angrily to the west, was quite invisible - except that there was a dark-orange glare to the ash clouds in the mountain's direction: it was like the view of a very distant furnace glimpsed, only half seen, through the dark clouds of its smoke.

But then, all of a sudden, the image shifted. Suddenly the old pilot, who had spent a lifetime guiding vessels through dangerous and unpredictable waters, became aware of something that was just barely visible, something that shouldn't have been there at all. He related to the Reverend Philip Neale, the British chaplain in Batavia who later in the year set about collecting the stories of eyewitnesses, exactly what it was:

Looking out to sea, I noticed a dark black object through the gloom, travelling towards the shore. At first sight it seemed like a low range of hills rising out of the water - but I knew there was nothing of the kind in that part of the Sunda Strait. A second glance - and a very hurried one at that - convinced me that it was a lofty ridge of water many feet high…

In the aftermath of Krakatoa's eruption, 165 villages were devastated, 36,417 people died, and uncountable thousands were injured - and almost all of them, villages and inhabitants, were victims not of the eruption directly but of the immense sea-waves* that were propelled outwards from the volcano by that last night of detonations.

It was in this one respect - the production of a number of massive and highly destructive sea-waves - that Krakatoa was then and remains today so very unlike almost all of the other of the world's great volcanic disasters. Its scale was phenomenal. The number it killed was unimaginably vast. But it was the way that it killed all those people that still sets Krakatoa apart.

Other volcanic eruptions around the world kill people in more direct and predictable ways - and they kill and injure, it should be remembered, a not insignificant number of people, since one in ten of the world's population is currently reckoned to live near volcanoes that are either active or have the potential to become so. So far as volcanoes are concerned, there are a great number of people - in the Philippines, in Mexico, on Java, in Italy even - who are currently living in harm's way.

The types of hazards to which such people are likely to fall prey, or to which their forebears fell victim in the past, are many and manifest. Erupted boulders and lumps of partly congealed lava - generally known by the term tephra, from the Greek word for ash - scream back down from the skies and flatten anything in their path. Perhaps a relatively small number of people, fewer than a thousand, died in this way from the Krakatoa eruption. All of them were in southern Sumatra, in the path of the prevailing wind: the hot ash that burned them alive had sped westwards from Krakatoa on top of a cushion of superheated steam.

Most of the other means by which volcanoes kill their victims were not experienced here. In other eruptions lava flows surround and trap victims, and sear them to death. Earthquakes associated with volcanoes destroy buildings, and huge seismically caused cracks in the earth swallow people and the buildings in which they live. The terrifyingly fast-moving clouds of hot lava, ash pumice and incandescent volcanic gases, known to the French as nuées ardentes and to the rest of the world as pyroclastic flows, sweep people up and incinerate them in seconds - as with, for example, almost every one of the 28,000 inhabitants of St Pierre, in Martinique,* who in May 1902 had been persuaded to stay in town for a supposedly important election, but were burned and suffocated by the sudden pyroclastic flows coursing down from the eruption of Mount Pelée.

Clouds of sulphur-dioxide gas, usually released during eruptions, choke and poison their victims. Clouds of carbon dioxide suffocate them. Clouds of hydrochloric acid gnaw away at their lungs. The torrents of volcanic mud and water slurry that course down the sides of certain volcanoes and that have the Javanese name lahars (since there are so many such flows running down the sides of Javanese volcanoes - though not, as it happens, on Krakatoa) carry victims miles away, and drown and bury them.

Sometimes secondary events can prove fatal. In 1985 a small eruption of a Colombian volcano called Nevado del Ruiz melted a glacier near the summit: the resulting river coursed down a valley that was quite unused to such huge flows, and the mud sea that was eventually created drowned an entire village below, killing 23,000 people. There are still more obscure risks: for example, volcanoes that erupt beneath glaciers - which tend not to have too many people living near them - produce sudden floods of melting ice, which have recently been given the exotic Icelandic name jökulhlaups. These can also prove fatal.

However, of all the victims whose deaths can be attributed directly to volcanic activity during the last 250 years, fully a quarter are now believed to have died - drowned or smashed to pieces - as a result of the gigantic waves that were created by the eruptions. The entire Minoan civilization on Crete was supposedly wiped out in 1648 BC when volcanic tephra from the eruption of Santorini - or, much more probably, the tsunamis thrown up by the eruption - destroyed the palaces at Knossos. More than 10,000 people died in 1782 in the waves that were created by an avalanche of volcanic debris that hurtled into the sea from Japan's Mount Unzen. In 1815 a similar number of Javanese died when Tambora exploded, sending pyroclastic flows raging into the ocean, with tsunamis radiating out in all directions and inundating the coast.

Careful study of the records for the last two and a half centuries has come up with a total of some ninety tsunamis for which volcanoes alone can be held responsible - and the greatest of these by far was the 1883 eruption of Krakatoa. About 35,500 men, women and children died as victims of the two gigantic waves that accompanied or were caused by the death throes of this island-mountain, and they account for more than half of all those in the world who are known ever to have died from waves caused by an erupting volcano. So this should be remembered well: it was neither fire nor gas nor flowing lava that killed most of the victims of Krakatoa. All but the thousand who were burned in Sumatra by the immolating heat of newly made ash and pumice and scalding gases died by the primary agency of water.

During the eruptive days back in late May the state of the sea was certainly noticed, but was never once reported to be the cause of any undue alarm. The hopper Samarang noticed a swell powerful enough to lift her screw out of the water; the lighthouse keeper saw the surface of the Strait turn suddenly white; the rudder of the Bintaing, another small hopper, swung around and hit her own hull with a mighty clang when a freak wave caught her. But that was about all: the eruption in its opening stages was about ash falls and noise and that seven-mile-high column of coiling smoke. The ocean seemed to prefer not to become involved.

But three months later matters were very different. The way that Krakatoa's immense outpouring of thermal energy was converted into mechanical energy - for this conversion is what essentially determines both the immensity and the enormity of any volcanic eruption - was altered. The noise was there, on an extraordinary scale. The expulsion of material high into the sky went on, both in gigantic amounts and for a very long time. But most of the mechanical energy went into the enormously difficult task of moving the ocean - movement that, once started and given additional shoves from behind, can become one of the most powerful natural forces imaginable.

In August the state of the sea was something noticed by all. Right from the beginning, when Telegraph-Master Schruit took his lunch and strolled out on to the hotel veranda to first see the column of smoke, it was the strangely erratic motion of the sea that most alarmed him. On the far side of the Strait, in Ketimbang, Monsieur Contrôleur Beyerinck too was astonished by the punishment his town's little dock was having to take from the curiously restless waters. The ships out in the Strait - the Loudon, the

A classic wall-of-water tsunami generated by only the most moderate of earthquakes on Krakatoa.

Marie, the Charles Bal - all reported on the state of the sea. For them it was not so serious, as waves at sea are less dangerous to a ship than waves close to land. The electricity in the air and the rain of flaming rocks from the sky were quite dangerous enough.

As darkness fell, so the sea became ever more furious. At 7 p.m. on the Sunday, Beyerinck saw small boats being tossed about. At the same time on the Javan side, Schruit found that his telegraph cable had been snapped by the mast of a schooner tossing on the waves. Between 7 p.m. and 9 p.m. several houses close to the seafront in the small town of Tyringin, well to the south of Anjer, were reported destroyed and swept away.

At about 7.30 p.m. a quarry near Merak, where dozens of Chinese labourers were hewing out stone* for the new Batavia docks, was then inundated, and the camp where the workers slept was washed away: they may have been the first casualties of what was to be a long and mortally expensive night. But now there was a lull: though a village five miles out of Anjer was reported to be submerged at 10 p.m., by midnight the sea was glass-smooth once again; and at 1 a.m. on the Monday morning Schruit, still trying feverishly to repair his severed cable (he eventually failed), noticed only small oscillations in the surface of the sea close to where the Anjer canal debouched into it.

Then at 1.30 a.m. one almighty wave is reported to have rushed up the long funnel of Lampong Bay to Telok Betong, where it ripped through and ruined several houses. Although it was clearly highly destructive, and though the time of its occurrence seems to be accurate, having been cross-checked with other witnesses (not least the servants of Beyerinck, sheltering with his family in their hilltop cottage), this one wave appears to be something of an aberration - far larger than its predecessors but unrelated to any particular event at the volcano. It was indicative, however, of what was to come.

The greatest and most terrifying volumes of water began moving in concert with Krakatoa's four culminating explosions - the first eruptive paroxysm being timed, as we have seen, at 5.30 a.m. It was as though then something deep within the mountain had begun a series of low-frequency pulsations, the sea moving back and forth in time with each pulse, and the amplitude of these movements becoming greater and greater, the volume, as it were, of the waves becoming stronger and stronger with each sequence of pulses. The four major tsunamis that were caused by, or were coincident with, these giant volcanic explosions then hit the shores like planet-sized wrecking balls, the effects all unimaginably and fatally destructive.

The destructive capacity of a great wave can be calculated, with difficulty, from a mess of competing and combining features, including the configuration of the shoreline, the funnelling effect of cliffs and headlands, and the depth of the coastal waters. It seems from the various eyewitness reports that what was most impressive about the waves that struck the shores of Java and Sumatra that morning was their sheer size - the high and unstoppable moving walls, the majestic volume of hundreds of billions of tons of roiling, thundering, foaming green water.

The last four great explosions of Krakatoa's life took place at 5.30 a.m., 6.44 a.m., 8.20 a.m. and, finally and most terrifically of all, at 10.02 a.m. - all of these well-chronicled moments being recorded in Krakatoa Time, which (because each local Dutch administrator still set his official watch according to when the sun rose and set and reached its noontime peak in his own district) was in those days 5 minutes and 42 seconds behind what the capital's civil servants regarded as Batavia Standard Time. The energy that was released in these eruptions was transformed into a variety of violent effects. There were massive expulsions of rock and ash and gas. There were torrents of heat, searing and welding together everything around them. There were sounds - bangs, cracks, thunderous roars, shattering low- and high-frequency noises - that were so loud they could be heard thousands of miles away. Seismic shocks were triggered that caused buildings 500 miles away to rock on their foundations.

And the eruptions also produced two kinds of shock waves. One was a wave that passed invisibly through the air, a sudden burst of pressure that bounced around the world, and was recorded as doing so, moreover, a remarkable seven times. These air waves - which recorded as pressure spikes at the Batavia gasworks, ninety miles to the east - radiated outwards from Krakatoa very fast, at what was an easily calculated velocity of about 675 mph. They were recorded as reaching Batavia at 5.43 a.m., 6.57 a.m. and (there seems, curiously, to be no firm record of any air wave resulting from the third explosion) at 10.15 a.m. respectively, Batavia Standard Time. (These events took place before the invention of time zones, either in the East Indies or anywhere else in the world. This, taken together with the dubious accuracy of many of the mechanical clocks of the day, the absence of the coordinating abilities of radio, which had of course not quite yet been invented, and the wide range of anecdotal reports from frequently panicky eyewitnesses, makes it tricky, though not entirely impossible, to construct a firm chronology of what took place in the aftermath of the eruptions.)

The other shocks, considerably more ‘complex in the way they moved, of much shorter duration but of equally extraordinary geographical spread, involved the disruption of the surrounding sea-water. Sea-borne waves in general move much more slowly: in the relatively shallow waters of the Sunda Strait probably at an average speed of about 60 mph.* However the Krakatoa tsunamis were forged, it would take one of them about thirty minutes from the moment of eruption to travel to the closest point on the mainland.† And it would be thirty-seven minutes before the wave was close enough to the town of Anjer for the people there to see it, to recognize just what it was that was bearing down on them and to start - a fairly, but not entirely futile, gesture - to try to outrun it.

It would take a further fifteen minutes for a great wave like this to seek out and destroy the quarries in Merak in the north, and drown all the Chinese workers there (as it did). It would take seven minutes fewer to flatten and wipe out all of Tyringin in the south (as it did also). And it would be one hour and one minute* before the same wave, slowing itself down but building itself up all the while, would reach all the way up to the head of Lampong Bay and, as it was equally sure to do, wreak havoc in the attractive little south Sumatran town of Telok Betong.

The coastlines of Sumatra and Java are, like any coastlines, made hugely complex by all the inlets and island-shadowed estuaries, bays and peninsulas, rocks and reefs. The way that an inrushing wave behaves as it courses towards the shore is only vaguely understood - making it somewhat challenging to try to work out from the survivors' tales which wave actually struck and destroyed each affected town, village, kampong and home on the edges of the two great East Indian islands.

Which wave was it that killed the vast majority of those 36,000 who were lost?

Was it ‘the low range of hills rising out of the sea' that was seen, chillingly, by that elderly Dutch pilot in Anjer at dawn? ‘The sight of those receding waters haunts me still,’ he was to write later, since for him this was the killer wave, without doubt. ‘As I clung to the palm tree… there floated past the dead bodies of many a friend and neighbour. Only a mere handful of the population escaped. Houses and trees were completely destroyed, and scarcely a trace remains of where the once busy, thriving town originally stood.’

Or was it the climax of all that terrible agitation of the sea that compelled Mrs Beyerinck across in Ketimbang to demand that her husband and family flee for the hills and the safety of high ground? Was it the ‘giant black wall of water’ that roared into Telok Betong at 7.45 a.m., picked up the gunboat Berouw as though it were a child's bathroom toy and dropped it in the middle of the Chinese quarter of town? The same wave that stranded the government's revenue cutter, and smashed all the local prahus and scattered the fragments of their hulls about like so much confetti?

Could it have been one of the ‘four waves’ supposedly seen that morning by an engineer named R. A. van Sandick? He was a passenger aboard the Gouverneur-Generaal Loudon - the steamer which, it will be remembered, was unable to dock at any of the quays in Lampong Bay because of the raging surf. The waves, which came in at tremendous speed sometime between 7.30 and 8.30 a.m.

… destroyed all of Telok Betong before our eyes. The light tower could be seen to tumble; the houses disappeared; the steamer Berouw was lifted and got stuck, apparently at the height of the cocoanut trees; and everything had become sea in front of our eyes, where a few minutes before Telok Betong beach had been. The impressiveness of this scene is difficult to describe. The unexpectedness of what is seen and the tremendous dimensions of the destruction, in front of one's eyes, make it difficult to describe what has been viewed. The best comparison is a sudden change of scenery, which in a fairy tale occurs by a fairy's magic wand, but on a colossal scale and with the conscious knowledge that it is reality and that thousands of people have perished

The tide-meter at Jakarta registers a sudden swell at 12.36 p.m., two and a half hours after the eruption - showing how relatively slowly tides move, compared to the fast-spreading barometric pressure wave recorded at the gasworks.

in an indivisible moment, that destruction without equal has been wrought…

Or was it perhaps the wave that struck Merak at 9 a.m. - the wave that drowned all but two of the town's 2,700 inhabitants? An accountant named Pechler who somehow survived by running before it, climbing further and further uphill until he was beyond range, certainly would imagine this tsunami to be immeasurably vast: it destroyed stone buildings that stood on top of a hill later measured at 115 feet high; it drowned all thirteen Europeans who lived there and who had had good reason to feel secure, surrounded as they were by walls of heavy masonry on the summit of a good high hill. But the wave displayed all the insouciance of its great power; and at the time it roared over, submerged and then wrecked these mansions it was towering above them by a good twenty feet meaning that whether what Pechler saw was the wave or not, it was at least 135 feet high, formidable in its terror. It drowned everyone in the town below, and when the waters receded almost everything in the town was either smashed beyond recognition or swept clear away.

Or yet again, might the great wave have been the one that was recorded when Merak was savaged once more, at 10.30 a.m.? A Dutch contrôleur named Abell, on the road to Batavia with his wedono* to report to his superiors details of tragic happenings yet further down the coast, looked around to see ‘a colossal wave’ roaring up the shore. It was, he said later, taller than the tallest palm tree he could see - a wall of water that no one caught by it could possibly have survived, something so dreadful it was quite beyond nightmares. Might this have been the one?

The answer on this occasion is probably yes. In fact, almost without a doubt, however compelling and awful the accounts of eyewitnesses to the other tsunamis of that dreadful morning may be, this last was indeed the one, the real killer wave. It happened at what seems to have been the correct time - with a travel speed of 60 mph, its arrival at Merak at 10.30 would put its time of origin at Krakatoa at almost exactly ten o'clock, which is the moment of the culminating, self-destroying explosion.

Most crucially, this one wave is recorded as having hit with extraordinary destructive power, a short while either before or later than the Merak 10.30 a.m. arrival, at all of the population centres of the west Java and south Sumatra coast. ‘An immense wave inundated the whole of the foreshores of Java and Sumatra bordering the Strait of Sunda,’ reported a contemporary study, ‘and carried away the remaining portions of the towns of Tjirin-gin, Merak and Telok Betong, as well as many other hamlets and villages near the shore.’

Its arrival was also recorded on the well-armoured tide-meter in Batavian Harbour - at 12.36 p.m. A wave so powerful as to give an almighty jolt to that tide-recorder would have to have been enormous indeed. It would also have had, if travelling at 60 mph, to have begun its journey some two and a half hours beforehand. This means, in other words, that it would have originated at a few minutes past ten in the morning. Undoubtedly, from all the evidence, this too was Krakatoa's most colossal wave, the biggest consequence of the biggest and final explosion. It was a wave so enormous and so powerful that it turned out to be the grimmest of grim reapers, the terrible climax to a long and deadly day.

‘Everyone was frozen with horror,’ wrote the Resident of Lampong, Mr Altheer, of the moment when he heard the explosion, just after ten on that Monday morning. He well knew, from what had already happened disastrously three or four times before during the previous twenty hours, just what to expect: another tidal wave, probably much larger than before since this was so great an explosion, would now come racing out from the island, and it would arrive within minutes. That is, of course, had there been an island: Altheer had no means of knowing that Krakatoa was no more, having just been blown to oblivion.

In the event the wave reached Telok Betong at 11.03. One anonymous European, writing some days later in a Batavia newspaper, was down on the town's shore, helping the local people who had already had their houses wrecked by the morning's earlier onslaughts from the sea. He was just lifting a huge wooden beam from on top of a trapped man, when he heard a scream. He looked up and saw a tall front of water rearing up and rushing towards him at a barely believable speed. There was a thunderous noise as it hit the beach and began rushing, crashing upwards through the town.

At this point the man's specific recollections become confused, his sudden panicked flight jumbling all conscious experience together into one amorphous mass. He is not alone in his confusion. The event must have been unforgettably dreadful, but in its details liable to the highly selective amnesia of those caught up by it.

Each of those snared by the Telok Betong wave speaks of running, wildly, panicked, trying madly to stay ahead of the wave, following natives running wildly too; and, in the particular case of the anonymous European writing in the Java Bode, of running behind a woman who stumbled and dropped her baby and could not abandon it and so was swept away, of running behind another woman who was somewhat incredibly, it must be said - in the very process of delivery as she raced on, screaming and bloody, of seeing a man desperately trying to avoid the wall of water by climbing up as high as possible, by running up every slope that he found, of snatching hurried looks behind him to see, horrible in its immensity, the ever pursuing wall, which from time to time smashed against some obstacle and broke, disintegrating into huge and dirty grey piles of spray and wreckage-filled foam, but then regrouping and following him always with a roaring relentlessness, with an unstoppable energy, with a dogged and seemingly murderous resolution such that he could only continue to run, despite being so leaden-legged and air-starved and exhausted, run ever onwards, always impelled by the frenzied gale that howled ahead of the wave, and by the certain knowledge that if he stopped or took a wrong turn that set him downhill rather than up he would be brought down drowned and his body crushed and hurled against the broken walls and jagged edges of spars and smashed glass and masonry that was rising up all around him.

Any doubts about the power of this single awful wave would be dispelled later by the discovery of a single compelling piece of evidence: the position of the Dutch steam gunship, the Berouw. The brief fame of this doughty little craft - four guns, a draught of six feet, a thirty-horsepower reciprocating steam engine, paddle-wheels and a crew of four European officers and twenty-four

The Berouw, well and truly stranded - but very little damaged - a mile and a half up the Koeripan River. Hunks of rusting iron remained in the jungle until the 1980s.

native ratings - provides a singular measure of the ocean's ferocity.

The captain of the Berouw had been the first to warn the mate of the Loudon that too strong a sea was running for him to risk a landing. That was at about 6 p.m. on the Sunday. The Telok Betong harbour-master then spotted her in difficulties about five hours later, in the middle of the night she was well lit, and her plight was clearly visible through the gloom. Exceptionally strong waves were breaking about her, and the official feared that not only might she break her mooring lines, but that the heavy chains holding down the two-ton conical steel mooring buoy to which she was attached might snap as well.

Then early next morning, disaster struck. There were two eyewitnesses: both the anonymous European in Telok Betong and the Loudon passenger R. A. van Sandick saw her lifted up on high by one of the 7.45 a.m. waves, then saw the mooring springs part one by one. The ship broke free of her buoy and was transported high on the crest of the mighty wall of green water. She was swept westwards for a quarter of a mile until, as the wave broke, she was crashed down precipitously on the shore, at the mouth of the Koeripan River.

It is thought that this fearful crash - in which the vessel remained upright - killed all of the crew. But it was not the end of the ship's own nightmare. When the great wave of 11.03 a.m. hit, the ship was picked up once again and carried westward a further two miles. She was driven all the way up the Koeripan River valley, along which the tsunami sped, and crashed down when the wave was spent, about sixty feet above the level of the sea from which she had been plucked. She lay askew across the river, forming a bridge. She was upright once again, a macabre tomb for the twenty-eight members of the crew.

She was found and inspected the following month by the crew of a rescue ship: ‘she lies almost completely intact, only the front of the ship is twisted a little to port, the back of the ship a little to starboard. The engine room is full of mud and ash. The engines themselves were not damaged very much, but the flywheels were bent by the repeated shocks. It might be possible to float her once again.’

But whether possible to float her or no, no one ever seemed interested in trying to slide her back all that way to sea. This was no Fitzcarraldo. And so the Berouw remained where she was thrown, lying athwart the river for the better part of the next century, picked apart by scavengers over the years, like a carcass, or rotting quietly in the steam and sun.

The hulk was more or less intact when it was visited in 1939: it was rusting and swathed in vines, and had become home to a colony of monkeys. Pieces of her were last seen in the 1980s. Nowadays she is all gone. The Koeripan River trickles uninterrupted past where she lay, and the only memorial is her great mooring buoy, sitting on a plinth at the site where it was washed up, two miles from where it had last floated, and fifty feet higher than the level of the sea. The name Berouw is the Dutch word for ‘remorse’.

The devastation in Sumatra was fully matched by that across the Strait in Java. The tales from the survivors are every bit as memorable and dismaying. In the main towns - Anjer especially - the ruination was near total. There were only momentary lapses into levity: a telegram received in Batavia reported tersely Fish dizzy and Caught with Glee by Natives. For the rest, all was melancholy. And for the most melancholy memorial of all, a symbol on the scale of the Berouw over on the far side, one need only look at the great granite lighthouse on what was called Java's Fourth Point, a little way to the south of Anjer.

It survived the first onslaught, as had the gunship; it survived the wave that drove the Berouw up on to the beach; but when the wave that hit Telok Betong at 11.03 struck Anjer - about fifteen minutes earlier, since Anjer is closer to the volcano - it picked up an immense piece of coral rock, weighing perhaps six hundred tons, and dashed it against the column. Despite its iron cage of reinforcing ribs, the light crashed down, extinguishing one of the most important navigation beacons of the entire Sunda Strait. And although his wife and child were drowned, the keeper himself survived. With the phlegmatic way of both the well-trained lighthouse keeper and the fatalistic acceptance of a true Javanese, he returned to his duties as soon as was physically possible, and had a temporary light erected, and lit, within a matter of hours.

The stone stump of his lighthouse can still be seen, standing like an old and rotten tooth, rising no more than ten feet above the ever grinding waves of today's more peaceful sea. A replacement, built by the Dutch government three years after the eruption, is close beside it - except that it has been placed a prudent distance, probably about a hundred feet, back from the shore. And it has been constructed entirely of iron, just in case.

The stump of the Fourth Point light remains. And the carcass of the paddle-steamer gunship high in the river valley. But that is about all. The town of Anjer did not survive. Nor did Ketimbang. Nor Telok Betong. Nor Merak. Nor Tyringin. The Anjer Hotel, from whose veranda the first signs had been spotted weeks before, was no more than foundations and twisted banyan roots. The massive Dutch fort walls, which had survived the depredations of centuries, were cracked and tumbled into no more than a shapeless mass of weathered stone. Railway tracks were twisted and scattered across the ground like so many yards of iron ribbon. Iron gearwheels, shards of broken iron and fractured lumps of machinery seemed to be everywhere. Giant boulders sat in entirely improbable places, picked up and smashed down as though they had been pebbles. Thousands upon thousands of houses and settlements up and down the coasts of Java and Sumatra were ruined, flattened, everyone in or near them crushed or drowned or never to be found again.

And what also did not remain was the volcano that had caused it all. To everyone's astonishment, it was seen, once the dust had cleared and the gloom had been swept from the sky, to have totally vanished. Krakatoa, after the final majestic concatenation of seismic and tectonic climaxes that occurred just after ten on the Monday morning, had simply and finally exploded herself out of existence.

Lloyd's agent in Batavia, the Scotsman Mr McColl, was able to send the following message within the week to his colleagues back in London, as concise a summary of the reality as that from his diplomatic colleague Consul Cameron down the road, and only a little less elegant:

We shall probably not be in possession of full particulars for some days yet, as telegraph lines are damaged and roads destroyed, but so far we can give the following particulars. The island of Krakatoa, the summit of which peak was 2,600 feet above water level, has totally disappeared beneath the sea, and the neighbouring island of Dwaisin-deweg* is split in five parts. Sixteen new volcanic islands have been formed between Krakatoa and Sibesie, † and the sea bottom in the Straits of Sunda has completely changed. In fact the Admiral Commanding-in-Chief has issued a circular stating that till new soundings have been taken the navigation of the Straits of Sunda is likely to be extremely dangerous. Anjer and lighthouse and the other lights of southwest Java have all been destroyed. The subsidences and upheavals we have alluded to caused a large wave about 100 feet in height to sweep down on the southwest coast of Java and south of Sumatra. This was swept in for a great distance, thereby doing great injury both to life and property. We are here only twelve miles away from one of the points on which the wave spent its fury. The whole coastline to the southwest has changed its configuration. The inhabitants of the island of Onrust were only saved from the flood which swept over the island by taking refuge on board two steamers. At Merak government establishment the inhabitants took refuge on a knoll, fifty feet high, but were all swept off and drowned, with the exception of one European and two Malays, who were saved. Mauk and Kramat, on the west side of Batavia roads, have been laid waste, and about 300 lives lost. In Tjeringin only one house has been left standing. Both the native and European officials have perished. A rain of mud also fell at the above place, which is situated opposite to where Krakatoa once lay.

Anjer seems to have been completely destroyed. Lloyd's sub-agent there wires from Serang: ‘All gone. Plenty lives lost.’

3. The Experiences

Not much of great excitement seems ever to have happened on the island of Rodriguez, which in the late nineteenth century was one of Britain's more remotely idyllic tropical possessions. According to the 1881 census some 5,000 people lived there, contentedly farming the forty square miles of agreeable farmland and happily fishing the 200 miles of sandy coastline (the relic of an old volcano itself) that had been set down in a lonely corner of the western Indian Ocean. Mauritius, the notional mother-ship of which Rodriguez was then and still remains a dependency, lies 350 miles away to the west. There is ‘a regular steamer service’ today, and an occasional plane; in the latter part of the nineteenth century there were infrequent supply calls by a chartered sailing ship. A telegraph cable connecting the capital village of Port Mathurin with the capital of Mauritius, Port Louis, was not built until the beginning of the twentieth century.

The people of Rodriguez, who were Creole speaking and (according to a short book written in 1923 by a civil servant who claimed ‘no pretence of fine writing’) possessed of ‘a deep brown velvety skin… hair of a very deep black, woolly and curly… protruding thick red lips… and magnificent snow-white teeth’, were descendants of slaves imported by the French to cultivate their sugar plantations. They had been left behind when the French were thrown out by the British at the end of the Napoleonic Wars.

Their lives were run under the genial superintendency of four British imperial administrators, a quartet who might well have found their way into a Gilbert and Sullivan operetta - a magistrate, a medical officer, a chief of police and a ‘First Class’ priest (the last paid an annual government stipend of a thousand Mauritian rupees to remind the woolly-haired local people that God was, most naturally, an Englishman and, in this corner of the world, a Catholic Englishman to boot).

Placidly unexciting though Rodriguez may have been through its three centuries of inhabited existence, it did make an appearance in the history books, courtesy of the volcano in the far-off East Indies. In August 1883 the chief of police on Rodriguez was a man named James Wallis, and in his official report of the dependency for the month he noted:

On Sunday the 26th the weather was stormy, with heavy rain and squalls; the wind was from SE, blowing with a force of 7 to 10, Beaufort scale. Several times during the night (26th-27th) reports were heard coming from the eastward, like the distant roar of heavy guns. These reports continued at intervals of between three and four hours, until 3 p.m. on the 27th, and the last two were heard in the directions of Oyster Bay and Port Mathurie [sic].

This was not the roar of heavy guns, however. It was the sound of Krakatoa - busily destroying itself fully 2,968 miles away to the east. By hearing it that night and day, and by noting it down as any good public servant should, Chief Wallis was unknowingly making for himself two quite separate entries in the record books of the future. For Rodriguez Island was the place furthest from Krakatoa where its eruptions could be clearly heard. And the 2,968-mile span that separates Krakatoa and Rodriguez remains to this day the most prodigious distance recorded between the place where unamplified and electrically unenhanced natural sound was heard, and the place where that same sound originated.

A popular Victorian science writer, Eugene Murray Aaron,* explained to his readers why this figure of 2,968 miles should amaze them one and all:

If a man were to meet a resident of Philadelphia and tell him that he had heard an explosion in Trenton [NJ], thirty miles away, he might be believed, although there would be some doubt as to his powers of imagination. If however he should make the same assertion of an explosion in Wheeling, West Virginia, three hundred miles away, all doubts of his accuracy would vanish. But if, with every sign of sincerity and a desire to be believed, he should earnestly insist upon his having heard an explosion in San Francisco, three thousand miles away, he would receive a pitying smile, and his listener would silently walk away.

Yet just this last marvelous thing was true of those… on the island of Rodriguez…

It was heard in a score of other equally exotic places besides. No sound was heard in Rodriguez or anywhere else before Sunday, the 26th, nor any after the night of the 27th. And there is general agreement (although among the usual welter of confusions, not the least of them caused by the same lack of time zones that frustrates attempts to make a chronology of the eruption and the sea-waves) that the loudest sounds occurred everywhere in the middle of the day on the Monday, suggesting that they originated on Java some short while before noon.

So, for example, the detonations were clearly heard in early afternoon, local time, on what is now the notorious British-owned American base-island of Diego Garcia.* In those days it was also a Mauritian dependency, where local farmers pressed palm oil and made copra, and where there was a coaling station for steamers crossing the Indian Ocean. The plantation supervisors plainly heard the explosion while they were taking their lunch. ‘Nous avons cru telement à l'appel d'un navire en détresse, they later reported (‘We thought it was a ship in distress firing its guns’).

There were scores of broadly similar reports. The explosions were heard in Saigon and Bangkok, Manila and Perth, and at a lonely cable station south of Darwin called Daly Waters.* From Port Blair, the capital of the Indian prison-islands of the Andamans, came news that someone heard a sound ‘as of a distant signal-gun’. No fewer than eighteen different sets of witnesses in what was then Ceylon came forward with stories (‘Captain Walker and Mr Fielder were puzzled at various times… by hearing noises as if blasting was going on’, ‘Sounds as of firing of cannon at Trincomalee’, ‘Mr Christie of the Public Works department… presumed some man-of-war was practising with her big guns, out of sight of land, as he could see no ships’).

His Highness the raja of Salwatty Island, in New Guinea, said he had heard strange sounds and demanded of a local doctor why the white men were firing their cannons. Stockmen driving cattle across the Hammersley Range† in western Australia heard what they thought was artillery fire to the north-west. In Aceh, at the northern tip of Sumatra, the location then (as still now) of a fierce pro-independence rebellion, the Dutch garrison commander assumed a local fort had been blown up by insurgents, and ordered all his men to battle stations. The Honourable Foley Vereker, who, despite belonging to one of Ireland's leading military families, had been cast out to the remoter colonies as commander of HMS Magpie, off Banquey Island, near Borneo, recorded in his log how he and his crew all heard the sound. And close by them, those Dayak islanders who had recently murdered a local official named Francis Witti (and, according to lore, had eaten his torso and limbs and had shrunk his head as a keepsake) heard the extraordinary sound too, assumed it to be the authorities coming to get them and fled deep into the jungle.

Ships were launched from scores of places by eager would-be rescuers and salvors, convinced there was an unseen vessel in trouble. A pair of fast ships set out from Macassar,* for instance, because they assumed another was in dire straits; two more set out from Singapore; a government boat went out to search off Timor; and when the sounds continued in Port Blair, the British authorities on the Andamans sent out a lifeboat too. In Singapore it became impossible, on one set of telephone lines, to hear yourself speak, since ‘a perfect roar, as of a waterfall, was heard, and by shouting at the top of one's voice the clerk at the other end heard the voice, but not a single sentence was understood. The same noise… was noticed on every line here.’

There was one aberrant report - from a man named Foley on Cayman Brac, now a small nodule of highly expensive Caribbean real estate, but then a forlorn sliver of a tropical sandspit, a day's sailing south of Cuba. Mr Foley insisted that he heard the banging cannonades, and that they happened in rapid succession on the Sunday. But there is neither corroboration nor a common-sense explanation. There was no eruption anywhere in the Caribbean at the time (it would be another nineteen years before Mount Pelée exploded with classically Plinian intensity); and though freak atmospheric phenomena might be able to explain away this report of the explosion being heard twelve thousand miles away, the fact that Mr Foley also claimed to have heard it twelve hours before Krakatoa exploded suggests his memory, or hearing, was indeed at fault.

Yet there were some oddities about the dissemination of the sounds - not the least being that a great number of people in Batavia, Buitenzorg and west Java generally heard nothing. Others simply felt strangely deaf, or heard a curious buzzing in their ears, or else were aware of wild swings in the pressure all about them, as if they had been caught in some silent atmospheric hypertension. Experts in the field have taken many a stab at explaining why this might be. Some point out that the behaviour of sound waves in the upper and lower atmospheres is very different, and that in the speeding up and slowing down that any wave experiences on passing through them there may be a focus-like phenomenon brought into play that would make one locality receive a lot of sound, another very little. Others, less sophisticated, explain it by reminding us how snowfall muffles sound - and that ash falls, which covered Batavia and its suburbs at the time, are likely to do much the same.

None the less, the overall conclusion remains inescapable, both to historical anecdote and to science: the sound that was generated by the explosion of Krakatoa was enormous, almost certainly the greatest sound ever experienced by man on the face of the earth. No man-made explosion, certainly, can begin to rival the sound of Krakatoa - not even those made at the height of the Cold War's atomic testing years. Those other volcanoes that have exploded catastrophically in the years since decibel-meters were invented - Mount St Helens, Pinatubo, Unzen, Mayon - have not come close: no one suggests that the explosion of Mount St Helens in May 1980 was heard much beyond the very mountain ranges in which it was sited.

Dr Verbeek, with the modest assurance of one who had seen and heard and experienced the vastness of the eruption, stated in his report of 1885 that ‘the exceptionally loud noises require our attention… the large explosions that in loudness have far exceeded all known noises. At no earlier event was a noise heard over such a large part of the earth's surface.’ Under the impact of Krakatoa's explosion, 13 per cent of the earth's surface vibrated audibly, and millions who lived there heard it, and when told what it was were amazed.

The inaudible waves, it was soon discovered, had ventured even further afield. The thousands of Europeans and Americans who noticed and recorded them without in most cases ever realizing just what they were - did so, all around the world, at more or less the same time. The fact that they did so at all points up, quite unexpectedly, one of the newly formed habits of middle-class Victorian society - a set of habits that never anticipated such a catastrophe as Krakatoa, but none the less in due course took full advantage of its effects.

The new-fangled habits of the late nineteenth century included, among the time's multitudes of other scientific advances, the development of increasingly precise means for forecasting the weather. And while cost and complication limited popular access to most other sciences, it did become rapidly possible - and indeed rather popular - for people to buy and use scientific instruments to help them understand the daily fluctuations in their climate. Consequently, as Victorian homes, clubs and hotels filled up their halls and vestibules with the ever newer and more handsome barometers, recording thermometers, sun-gauges and rain-gauges, so the middle classes became an unwitting army of amateur meteorologists, faithfully tapping the glass each day the better to predict whether it would be Fine or Stormy, Changeable or Fair.

The most costly and sophisticated of these instruments was the recording barograph. Because of its price, it was reserved most commonly for the mantelshelf in the club, rather than for the hall at home. The task of this small machine is to record, with an ink trace on a sheet of graph paper secured around the circumference of a clockwork-driven drum, the slight hourly variations in atmospheric pressure over the week that it takes for the drum to rotate a single time.

A well-made barograph is a joy to behold - an elegant confection of brass and steel, mahogany and glass, its mechanicals visible inside its crystal case, its clockwork heart ticking away happily as it takes the pulse of the day. And the ink trace - flowing gently up and down, sometimes more steeply downwards if bad weather is on the way, arching up again if the storm passes over at speed - has a seductively sinuous beauty to it also: the records of the passing week's atmospheric alterations are stored in a small drawer under the instrument, to be studied later, whenever the weather becomes the subject for reminiscence or chatter.

One very noticeable aspect of a barograph's recorded trace is, however, just how smoothly the weather changes. The line invariably curves up or down slowly, steadily. It does not jump erratically, like a seismic record during an earthquake or a lie detector revealing a falsehood. Atmospheric pressure changes with measured deliberation - a feature that the curves on a barograph reflect, with their steady and considered moves across and along the ever unrolling snake of recording paper.

As scores of people around the world came to change their paper for the week that ended on Sunday, 2 September 1883, and as they smoothed the records out to place them away in the drawer, they all, almost simultaneously, noticed something. On the trace for the Monday of the week just gone, the 27th of August, there was a sudden and unanticipated jerk. A hiccup. A notch, an interruption - an altogether most puzzling thing.

The pen that had been so smoothly and seamlessly noting the pressure on the instrument's vacuum chamber had suddenly been flicked up, and then equally violently snapped down again. When looked at more closely, the oscillation was even more peculiar than that: first there was a sudden recorded rise in pressure, then two or three minor oscillations, then a very deep depression, followed by a less steep rise, then more small oscillations, and then finally, after an interruption that lasted for the better part of two hours, back to the smooth and gently changing trace of normal times. In summary, it seemed as though, and for some inexplicable reason, there had been the quite impossible occurrence of an earthquake in the air.

It took only a matter of hours of excited discussion between observatories and weather-fascinated members of the lay public to draw the inescapable conclusion. This, faraway though it may have been, was all Krakatoa's doing.

As soon as news of the extraordinary degree of the eruption had become widely known, it remained a simple matter to check the times of the strange two-hour-long blips on the barograph traces, to make an allowance for the probable approximate-speed-of-sound rate of travel of the shock wave, and to figure in the time difference between Krakatoa and the various club mantelshelves around the world. And lo! they all matched. The eruption that had sent out flame and ash and tidal waves and an incredible explosive sound had also sent an invisible, inaudible shock wave that passed cleanly through the atmosphere, and had been recorded, quite unexpectedly, on scores of machines designed for the much more prosaic task of suggesting to middle-class Victorian gentlemen in Birmingham and Boston and beyond whether they should take their umbrellas to luncheon.

Except that it was, as is much that relates to Krakatoa's eruption, much more complicated than it seemed at first sight. When the traces were examined more closely and compared with other barograph traces from more distant cities around Europe and then around the world, it appeared that the shock wave from Krakatoa's final cataclysmic explosion had travelled around the earth not once but seven times.

The barographs, barometers and weather stations all recorded the signature two-hour wave, with the amplitude of the oscillations diminishing at each pass; it had apparently reverberated, flown back and forth and around the planet in a manner that seemed quite out of proportion to the magnitude of the original event itself.

All this provoked much fluttering in the scientific dovecotes. Every weather expert in the world suddenly wanted to know what was going on - why a pressure wave like this would behave in this peculiar way. In scientific London the degree of interest was particularly intense - and it prompted what was an entirely unanticipated response: that, despite Krakatoa being on Dutch sovereign territory, it would be best if it were to be left to a distinguished and entirely British body to investigate its eruption.

The high-handedness, seen from today's perspective, quite boggles the mind. Perhaps it was the vast reach of British influence of the day that made this seem desirable to some; perhaps, more specifically, it was the existence of all those records from all those British-owned and British-designed barographs (for all of those instruments, in places as far-flung as Melbourne and Mauritius and Bombay, turned out to have been manufactured in England) that provoked this entirely unwarranted - but it has to be said, quite unresented - example of British imperial busybodyness.

The timetable went like this. The explosion occurred in late August. In early September the British barograph paper records - initially from the leather-bound and entirely masculine fastnesses of London clubland, but later from weather observatories at Greenwich, Kew, Stonyhurst, Glasgow, Aberdeen, Oxford and Falmouth - were changed, and the blips noticed. Suspecting that something of absorbing scientific interest was afoot, a senior British government official named Robert Scott - the secretary to the Meteorological Council - promptly sent telegrams to his colleagues in observatories around Europe, asking if they would examine their traces too.

And back came the reports, from Vienna and Berlin, Leipzig, Magdeburg, Rome, Paris, Brussels, Coimbra, Lisbon, Modena and Palermo and other places besides, confirming in all respects what Scott suspected: that a wave of sudden pressure had swept around and around the planet from its birthplace in the Sunda Strait; that the passage of the wave had been a remarkable event; and that, moreover, it had lasted, echoing around the globe for no fewer than fifteen days following the eruption.

Scott found this quite extraordinary, and told his superior, an old India hand and engineer named General Richard Strachey. In December - only four months after the eruption, and so with unusual dispatch - the pair presented a brief paper to the Royal Society, that most estimable and ancient of British scientific institutions. It was entitled ‘Notes on a Series of Barometrical Disturbances which Passed over Europe between the 27th and 31st of August 1883’. It caused an immediate sensation.

And it did so because here was one of the first provable instances in which a natural event occurring in one corner of the planet had effects that spread over the entire world (or what would be the entire world, if further records could be sought from the Americas and Asia and elsewhere, for they would show the same evidence). Here was the event that presaged all the debates that continue to this day: about global warming, greenhouse gases, acid rain, ecological interdependence. Few in Victorian times had begun to think truly globally - even though exploration was proceeding apace, the previously unknown interiors of continents were being opened for inspection, and the developing telegraph system, allowing people to communicate globally, was having its effects. Krakatoa, however, began to change all that.

The world was now suddenly seen to be much more than an immense collection of unrelated peoples and isolated happenings: it was, rather, an almost infinitely large association of interconnected individuals and perpetually intersecting events. Krakatoa, an event that intersected so much and affected so many, seemed all of a sudden to be an example of this newly recognized phenomenon. And so it was up to a British scientific society - most decidedly a British one, given the imperial mood of the day, like it or not - to investigate it.

A decision to do just this was taken in January, following the reading of two more short papers presented to the Royal Society, both describing the scene on the ground in and around the Sunda Strait. The first paper was by the British consul in Batavia - now a Mr Kennedy from Sumatra, since Consul Cameron had fallen ill - and the other by the socially well-connected Captain Vereker of HMS Magpie, who reported from Borneo. Both papers amply confirmed the view in London that this extraordinary event off Java had been so huge and so world-affecting that a body had to be set up right away to investigate it. And so on 12 February 1884 an advertisement, in the form of a ‘Letter to the Editor’, was placed in The Times:

THE KRAKATOA ERUPTION

Sir - The Council of the Royal Society has appointed a committee for the purpose of collecting the various accounts of the volcanic eruption at Krakatoa, and attendant phenomena, in such form as shall best provide for their preservation and promote their usefulness. The committee invite the communication of authenticated facts respecting the fall of pumice and dust, the position and extent of floating pumice, the date of exceptional quantities of pumice reaching various shores, observations of unusual disturbances of barometric pressure and of sea level, the presence of sulphurous vapours, the distances at which the explosions were heard, and exceptional effects of light and colour in the atmosphere. The committee will be glad to receive also copies of published papers, articles and letters bearing upon the subject. Correspondents are asked to be very particular in giving the date, exact time (stating whether Greenwich or local), and position whence all recorded facts were observed. The greatest practicable precision in all these respects is essential. All communications are to be addressed to -

Your obedient servant,

G. J. Symons, Chairman, Krakatoa Committee*

Royal Society, Burlington House

It took five years for what at first might have seemed a some-what pumice-obsessed Committee to study all of the complex and new types of information that flowed from the event. Their final report, issued in 1888, had 494 pages of text, as well as countless drawings, graphs and exquisite coloured prints. It stood as a lasting masterpiece of determined study, elegantly composed style and splendid Victorian brio - and it collated, among other things, all of the pressure-wave observations that had led to the establishment of the Committee in the first place.

And what was found out about them was simple but remarkably beautiful: that the shock waves from Krakatoa had radiated across the world like ripples on the surface of a pond - the surface of a vast and slightly flattened sphere in the case of the earth, of course, rather than the merely circular surface one would find on a pond. They had radiated outwards at a speed approximating that of audible sound: between 674 and 726 miles per hour.

The waves had travelled outwards, expanding their width until they were halfway to their goal and then contracting again until they reached their precise antipode - which, in the case of Krakatoa at 6°06'S, 105°25'E, is a point (at 6°06'N, 105°25‘W) in the Pacific Ocean off Botoga, Colombia. Then, having taken

The sound of the great explosion could be heard as far off as Rodriguez Island, nearly 3,000 miles from its origin. Records suggest that people living within the shaded area were well able to hear the rumbling and banging, most thinking it the noise of a naval bombardment.

nineteen hours to reach this nameless and watery antipode, they headed offback again and returned to Krakatoa (with the passage noted on all the barographs in between, at a bewildering variety of places that included St Petersburg, Toronto, the Antarctic island of South Georgia and a village that is now a pretty New York suburb called Hastings-on-Hudson).

Every time the wave was noticed, it was found that the time of its passing agreed with the eruption time that had been marked by observers back in the Sunda Strait. The wave's passage above the Greenwich Observatory, for example, was recorded on all the barographs' registers - with the sharp upward tic of pressure, the minor ruffles on the record, the sudden downward blip, further ripples, the slow upwards rise to the resumption of normal state - at 1.23 p.m. on the Monday. Krakatoa Time was seven hours ahead of London - meaning that at 10.02 a.m., the local time when the volcano exploded, it was 3.02 a.m. in Greenwich. Subtracting that from the time that the Observatory barographs recorded the blip gives a figure for the wave's travel time of ten hours, twenty-one minutes - the precise figure that could be calculated for inaudible shock waves moving across the 7,220 miles of a Great Circle separating London from Krakatoa.

The Greenwich Observatory corroborated the figures for the first pass - and for the six further passes of the shock as it moved back and forth across the capital before it had weakened to the point where it could no longer be detected. And as it did so, two things became even more clear: that the eruption time was exactly right, 10.02 a.m., and that the world now knew a great deal more about the transmission of atmospheric shock waves than ever before. Meteorology in general profited mightily from the findings; and in the mid twentieth century, when large atmospheric explosive tests were conducted during the Cold War, the way in which shock waves were propagated through the atmosphere was well understood also. The Krakatoa Committee had, even if only in these respects, fully justified its existence.

The tsunamis that killed so many on the shores of Java and Sumatra crossed the world as well. It could be seen from the beginning that up close to the volcano the waves were enormous, and killed thousands. That they then became ever smaller in proportion to their distance from Krakatoa was to be expected. But the discovery that they were in fact so deeply powerful, and radiated away from the volcano so aggressively that they could still be detected in the sea as far away as the English Channel, was the cause of general astonishment.

The first to receive firm news of the appearance of some faraway Krakatoa-induced waves was Charles Darwin's son George, shortly after he was elected (by the narrowest of votes) the professor of astronomy at Cambridge.* His friend Major A. W. Baird, who had the post of chief of the Tidal Survey of India, wrote to him in Cambridge to say that ‘the wave caused by the volcanic eruption at Java is distinctly traceable on all the tidal diagrams hitherto received, and I am informed of a great tidal disturbance at Aden on August 27; but the daily reports are always meagre in information. Kurrachee and Bombay also show the disturbance, and as far as I have examined the wave reached halfway up to Calcutta on the Hooghly.'

This last - the news that waters had rushed up the Hooghly River almost to the city that was then the capital of British India - did it. The Royal Society promptly asked for an immediate report. Major Baird swiftly obliged with a six-page summary of his Survey stations' reports from across that immense swathe of imperial territory that stretched between Aden and Rangoon. And the Krakatoa Committee, recognizing a significance that went beyond the simple threat to Calcutta, immediately commissioned a senior Royal Navy captain to investigate the phenomenon, worldwide.

Records from tide-gauges at ports all across the world were speedily gathered in. An early analysis showed a fascinating, but not entirely unanticipated, trend: almost all of the stations that recorded sudden and unexpected waves that could be positively linked to the eruption (by their timing and their type) lay to the west and south of the island.

Almost all: not Batavia herself, which lies eighty-three miles away to the east as the crow fines, and considerably further so far as a tidal wave might pass. Despite where she lay, the capital city did indeed witness what even the Royal Society saw fit to call ‘a wall of water’, when the wave hit her gauge at 12.36 p.m. on the Monday afternoon - two hours and thirty-four minutes after the explosion. According to the Reverend Neale, the water rushed into the Batavian canal system, rising suddenly by several feet, forcing hundreds of merchants and residents to flee for their lives.

The day - unusually cold, half dark and drab, the air still filled with grey, gritty ash that got in the hair and the eyes and the teeth - had begun, surprisingly, with a fair sense of stoic normality. The steam-trams were filled with people setting off for work, the markets were thronged, the private horse-drawn carriages were trotting around the Koningsplein, their occupants talking excitedly about the events of the night before, confident that the worst was over. Then came the arrival of what was swiftly understood to be the huge relic of the great tsunami - the remains of a wave that somewhere had been much, much worse - and it made all these good burghers of Batavia realize, very suddenly, that in fact the worst was still to come.

The maximum height of this bore (the needle on the Batavia tide-meter shot up vertically, clear off the scale) was at least seven feet and six inches - a fraction of the height of the devastating waves that destroyed Anjer and Telok Betong maybe, but an impressive enough display. The waters promptly fell back again, to ten feet below normal sea-level, and then rose back up again, then sloshed back down - oscillating a total of fourteen times over the next twenty-eight and a half hours, the height of the successive waves diminishing all the while. Finally, after what was no more than a three-inch ripple hit the Batavia tide-meter at 5.05 on the afternoon of the next day, the Tuesday, they vanished clear away.

But Batavia alone in the area experienced the great wave. Almost no other places to the north and east of the volcano experienced anything at all - Singapore's tide-meters registered nothing, nor was there any discernible blip in the records of Hong Kong, Yokohama or Shanghai; and even at Surabaya, at the eastern end of Java, the disturbance that was picked up on the port's three tide-gauges was only ten inches, ‘too insignificant to be otherwise noticed’. There is a very simple reason for the lack of any dramatic effects on this side of the volcano, as a glance at the map will show.

To the east of Krakatoa the two sides of the Sunda Strait pinch inwards like the jaws of a nutcracker. There are islands blocking the way too - Thwart-the-Way Island being one such, notorious in its nuisance value - and before a wave has any chance of touching Batavia port herself it reaches long fingers of shallows and sandbanks and further inlets and reefs, all of them conspiring to slow down and frustrate the eastward movement of any wave. Nothing would stand in the way of a sound wave or a shock wave; but, faced with the dissipating influence of the shoals and headlands, a water wave would essentially not move eastwards at all, as the recorders everywhere confirm.

However, to the west of Krakatoa, barring the presence of a small headland called Vlakke Hoek in southern Sumatra that acts as a small chicane on a westbound wave's right-hand side, there is only the wide-open sea of the Indian Ocean. Any tsunami moving out from the eruption in this direction would be free to go wheresoever it wished, without maritime hindrance or interruption. And in August 1883 the great ten o'clock wave did indeed fan out westward entirely untrammelled, and managed to go just about anywhere and everywhere it wanted. Two types of wave were detected: what were called long waves, which reverberated back and forth at periods of as much as two hours; and the short waves, which were steeper and with a much less regular and more frequent repeat.

The old Dutch port of Galle, close to the southern tip of the island of Ceylon, is where the arrival of these short waves - or more precisely, a sequence of fourteen waves, each separated by just a few minutes - was first noticed. The Ceylon Observer correspondent filed on 27 August that

... an extraordinary occurrence was witnessed at the wharf at about 1.30 p.m. today. The sea receded as far as the landing stage on the jetty. The boats and canoes moored along the shore were left high and dry for about three minutes. A great number of prawns and fishes were taken up by the coolies and stragglers about the place before the water returned.

A woman was killed at the port of Panama - still in Ceylon, not on the isthmus - when she was swept from the harbour bar by an immense influx of water. Both the Panama harbour-master and the local ruler, the splendidly titled Ratamahatmaya, said later that ships had suddenly sunk downwards and were then drawn backwards to be left stuck in the drying mud, their anchors exposed - and just as suddenly were borne up by an inrushing surge of water. The local streams, with hitherto sweet water, all promptly turned salty for at least a mile and a half upriver. The woman who died, from the injuries she sustained in falling while she was carrying a sheaf of paddy from the fields, is thought to have been the most distant casualty of the eruption that took place nearly 2,000 miles away.

At Hambantota, further south still, Ceylon government officials estimated the height of the wave to be twelve feet, and said that, like at Panama, its currents were irresistible, taking small craft back out to sea, and then sweeping them back and dashing them to pieces on shore. But however violent, there were no casualties here; nor were there any further afield.

The long waves tended to be the ones that were recorded by the

Tiny oscillations in the tides were noticed nearly 11,000 miles away from Krakatoa, such as here at Socoa, a small port near the French resort of Biarritz.

automatic tide-gauges around the world, and it is these that make up the bulk of the formal record; the short waves tended to be more the stuff of anecdote, and, because they oscillated so swiftly, rarely made an appearance on the recorders. By the time the long waves reached India they were diminishing, fast - fourteen inches high in Madras, a series of ten or so six-inchers at Calcutta, a foot high in Karachi, half that in Aden. They spread south-westwards towards the African coast as well: they broke a hawser of a boat moored in Port Louis, Mauritius; and in the rarely visited Indian Ocean reef harbour known as Cargados Carajos the captain of the Evelina reported huge, smooth oscillations in the sea, breaking only when they came into contact with coral heads. Already the wave was 2,662 miles from its point of origin, and racing steadily along at a calculated 370 miles each hour.

A four-foot-high wave was noticed at Port Elizabeth, on the bleak and generally unlit east coast of South Africa,* and undulations were picked up at Cape Town. A visiting German South Polar expedition (which failed to reach its goal) on the island of South Georgia saw the icebergs and brash in the harbour of the whaling station at Grytviken lifted fifteen inches in a series of a dozen recorded and remarkable swells.

And so the waters progressed ever outwards - with numerous waves of different types and styles, heights and frequencies, and deriving from what oceanographers now reckon were a number of different causes too. Eventually they ran out of steam and reached as far as they could go, in the further recesses of northwestern Europe. By the time they reached the North Atlantic, and then the Bay of Biscay, the oscillations were small indeed - such that the tide-diagrams had to be photographed and blown up in order to be able to measure the fluctuations in the record.

But they are there all right, tiny but still distinct. In Socoa, a tiny French harbour near the more celebrated resort of Biarritz, and 10,729 sea-miles from Krakatoa, there were seven undulations, each of them three inches high - barely enough to be noticed by promenaders on the beach, though I like to fancy pomaded young men and their lady friends skipping amusingly out of the way when the unusual small swells suddenly threatened their boots and their turn-ups. Further north at Rochefort, a town on the Charente a little north of the Gironde and Bordeaux, the magnifying power of the estuary pushed the waves up to five inches high - they had sped there from the volcano at a calculated (and now slightly faster) rate of 414 miles per hour, barely slowing as they did so.

And finally, turning the corner into the English Channel, the slightest trace only is noticeable. There is a ripple recorded at Cherbourg, another at Le Havre and an irregular but discernible series of undulations at Devonport, close to where all Royal Navy cadets are now trained to be officers. But closer in - at Portland, Plymouth and Dover, nothing. Maybe there is just the vaguest sign of it on the gauge on the inner side of the Portland breakwater - though the Royal Society report admits, ‘the indications of disturbance are not very conclusive, as no regularity of period is traceable in the small indentations which do appear’.

I confess I would have derived no small pleasure from discovering that on the Tuesday morning, when waves were seen in France and in the West Country, the tide-gauge in Dover port also suddenly startled its keeper, with a splash, or a slop, or a curious and inexplicable swell. But there was to be no such luck. With Dover being no less than 11,800 sea-miles from the Sunda Strait, and lying at the distant end of a channel in which the water shallows to a few hundred feet and less, which would kill or dampen the kind of long waves that were produced by the eruption, it is perhaps not surprising to learn that the gauges here showed no deflection at all. The air wave may have found its way to Greenwich, seven times; but the sea-wave was brought up short, 500 miles from home, and never arrived near the British capital, even once.

Yet other, quite different effects of the eruption were before long visible in London, New York and other northern capitals, and many of them were aesthetically and dramatically quite memorable. Art was born out of the after-effects of this volcano - art that was then quite unexpected. Though not so popular today, it is not entirely forgotten. For the millions of tons of dust that were hurled into the upper air in the East Indies disseminated themselves around the world for many years and caused all manner of extraordinary phenomena - not the least of which were sunsets. These were seen all over the world decked out in the most lurid rainbow of colours, and they attracted the interest of a great number of suddenly excited painters.*

One of the more prominent of these was Frederic Edwin Church, a member of what came to be known as the Hudson River School of American nineteenth-century landscape painters. He turned out to be precisely the kind of artist who would benefit from the atmospheric consequences of Krakatoa.

Frederic Church specialized in highly dramatic landscapes and highly coloured skyscapes - he had a predilection for the grand (his gigantic Niagara presents an astonishing image of the raw power of falling water) and the excessively radiant (his Twilight in the Wilderness has an unforgettable richness about its evening colour). Both of these artistic preferences came together in December 1883 when Church - supposedly well aware of the remarkable effects of Krakatoa's spreading trail of dust on the world's sunsets - travelled north from his ornate Moorish castle in the riverside town of Hudson, to the very tip of upstate New York, on the Canadian border. There he would try to capture an image of what he suspected would be an especially livid northern twilight.

He chose Chaumont Bay, at the very eastern end of Lake Ontario. The early-winter ice would be piling up in the westerly winds; there would be newly undraped trees on the wispy peninsulas; and above all there would be the vast expanse of the lake into which the setting sun would appear to sink. He chose to render the image in watercolour. The resulting picture - not surprisingly called Sunset over the Ice on Chaumont Bay, Lake Ontario - has a range of colours in the sky, a gently intermingling play of pinks and mauves and orange and salmon and purple that is quite astonishing, and unusual, and suggestive of something happening, something inexplicable, high in the evening atmosphere. It is the only major painting to be created in the immediate aftermath of Krakatoa: and even if Church had not set out with the deliberate intention of making post-eruptive art, the piece stands now as vivid documentary testimony to the giant volcano's effects.*

Lesser artists had a field-day. The most notable was William Ascroft, who lived beside the River Thames in Chelsea. Early in September, two weeks after the eruption, he noticed that London was suddenly being gifted with a series of memorable setting suns and, more interesting, unusually strong afterglows. He painted them at a furious rate - creating a total of no fewer than 533 watercolours over the months of his fascination.

On especially gorgeous evenings he might make several paintings, one every few minutes, creating as he did so a kind of time-lapse image of the entire process. On 26 November, for example, he painted the aftermath of the sun's disappearance once every ten minutes between 4.10 p.m. and 5.15 p.m., catching a sequence of fiery purples and oranges with all the fleet-footed accuracy of a film camera. He wrote lengthy notes and analyses of what he saw - ‘Blood Afterglows' and ‘Amber Afterglows’ among them - and examples of the beaming coronas that often surrounded the setting sun itself and which were named, after the Hawaiian naturalist* who first spotted them, ‘Bishop's Rings’. All 500 of his paintings later went on show at an exhibition in a museum in South Kensington. They remain today inside what is now the Natural History Museum, locked away and half forgotten.

Dust of all grades and compositions was thrown into the air by the eruption. Much of it, too heavy to be kept on high for long, fell down as drifting veils of grey, and was widely reported as having done so. Ships at sea experienced dust falls for a fortnight after the eruption: the Brani and the British Empire came under a slow rain of a white ash that one master said ‘looked like Portland cement’, when they were sailing in the Indian Ocean within a 2,000-mile range of the volcano; the Scotia experienced falling dust until 8 September, when she was off the Horn of Africa, 3,700 miles away.

But the lighter material, the finest particles of all, were thrown up right through the troposphere until, almost defying the pull of gravity, they were caught up for months in the lower reaches of the stratosphere† itself. Modern estimates suggest the Krakatoa eruption hurled material at least 120,000 feet into the air - some say 160,000 feet, or thirty miles. Tests have since shown that up so high, material can hover in a kind of weightless stasis.

A dust particle with a diameter of a micron - whether it is an aerosol droplet or a tiny fragment of volcanic silicate mineral makes no difference - has recently been shown to take many weeks to descend vertically through just half a mile of stratosphere. A particle of half a micron in diameter will take many months to fall, so slight is the tug of gravity upon it. Yet horizontal movement was apparently no problem for the Krakatoa material. The strong globe-encircling winds spread them far and wide. And the rains that would tend to flush out those particles that might have been tempted to stay in the lower reaches were of course non-existent up high.

And so they stayed, undisturbed, for very long periods. And by refracting and filtering and in a myriad other ways altering so vividly the colours of the sunlight that passed by them, and by staining the crepuscular skies with vermilions and passion-fruits and carmines and royal mauves, so they ensured more potently than any other effect that Krakatoa would soon become the most famous volcano in world history.

Krakatoa killed more people than any other eruption, and for that it remains notorious: but it became much more widely known to hundreds of millions of people around the world for a more benign and beautiful reason, one that all could readily see for themselves each time they looked westwards at eventide.

The poets were inspired in much the same way as the painters. Tennyson is widely believed to have been thinking of Krakatoa when in his almost wholly forgotten epic poem St Telemachus, published nine years after the eruption, he wondered out loud:

Had the fierce ashes of some fiery peak

Been hurl'd so high they ranged about the globe?

For day by day, thro' many a blood-red eve,…

The wrathful sunset glared…

The Royal Society's Krakatoa Committee, displaying the almost obsessive need for the complete and the comprehensive that was so much a signature of Victorian studies like this, invited responses from the general public. They received wagon-loads of material,* and painstakingly catalogued every single report of every atmospheric phenomenon, no matter how trivial, of which they were made aware. Two thirds of the Society's eventual 494-page report is devoted to ‘the unusual optical phenomena of the atmosphere, 1883-1886, including twilight effects, coronal appearances, sky haze, coloured suns, moons &c’.

Forty-eight further pages give the details of the 800 places where unusual phenomena had been seen, organized into chronological order. The list actually starts some months before the cataclysm, presumably to give an observational base-line from which the deviations that follow can be measured. So the first report comes from a town called Graaff-Reinet, in the centre of the South African tablelands, beginning during the southern winter - with someone noticing the presence of ‘fine sunsets, gradually increasing from February to June’. The very first eruption of the final paroxysmal cycle, it will be recalled, took place towards the end of May.

And thereafter came a cascade of responses from people who had seen or had been told about the Royal Society's announcements in the newspapers. A seemingly endless string of reports is included, from a bewildering variety of places and people and ships and lighthouses - all with news of strange phenomena, most of them seen in the skies. A blue moon was spotted by a Mr Haughton in Kokkulai, Ceylon. The lighthouse keeper at the Chinese city then called Chefoo - now Yantai - saw a pale-red glow, like fire, in the west. There was a blue sun seen by a Dr Earl Flint in Rivas, Nicaragua. Captain Faircloth of the Caribbean Signal Service saw ‘a lurid glare’ emanating from the sun setting over Nassau in the Bahamas. The scientific journal Nature - which collected dozens of such reports in tandem with the Society - had first a report of a green sun seen in Colombo, and then a letter from a Reverend W. R. Manley, a missionary in Ongole, south India, who saw ‘splendid twilight glows… deep red more than one hour after sunset’.

And Gerard Manley Hopkins, the Victorian poet and Jesuit priest, then teaching classics at Stonyhurst College, wrote a lengthy essay for Nature, outlining in detail what he had seen - in a style unusual for the normally arid magazine, but easily recognizable to those who know their Wreck of the Deutschland:

… the glowing vapour above this was as yet colourless; then this took a beautiful olive or celadon green; not so vivid as the previous day's, and delicately fluted; the green belt was broader than the orange, and pressed down on and contracted it. Above the green in turn appeared a red glow, broader and burlier in make; it was softly brindled, and in the ribs or bars the colour was rosier, in the channels where the sky shone through it was a mallow colour…

There were four main kinds of phenomena, all amply evident from these pages and pages of reports. There were the sunsets themselves; there were the vivid and highly unusual colorations of the moon (often blue,* sometimes green), the occasional colorations of the sun and, very rarely, of some of the larger planets; the whitish solar coronas that were frequently seen just before sunset; and the monstrously flaming afterglows.

The afterglows in fact turned out generally to be a more dominant and widely noticed feature than the actual sunsets themselves. They are brilliantly hot-looking glows that appear from time to time some way up in the sky above where the sun has set some minutes before. The geometry of their origin is well known - having been studied assiduously after the eruption. They are caused when the rays of the sun, at an ever increasing distance from the observer, pass tangentially to an optically unusual layer in the atmosphere. In the case of the afterglow sightings that were made after Krakatoa, the optically unusual layer was the drift of ash - the ash particles absorbing and reflecting and causing a fiery reddening of the red light that is the last to be bent from the source, before, thanks to the disappearance of the sun and all of its light now coming from well below the horizon, they eventually vanish altogether.

A pattern also emerged swiftly from the Royal Society's massive catalogue of observations. There could be no doubt that the immense cloud of stratospheric ash that spread out from Krakatoa wound itself around the planet in a westerly direction - as one might expect, with the world turning eastwards underneath it - and spread both northwards and southwards as it did so. At first, in other words, the blue suns and moons and the fiery afterglows and the Bishop's Rings and the extraordinary sunsets appeared in the low latitudes, close to the latitude of the volcano: during all of September in places no further north than Honolulu, no further south than Santiago in Chile.

But then the cloud spread itself further afield. By the beginning of October the phenomena were seen in the Gulf of Mexico, then Nashville, Buenos Aires, the Canary Islands, Shanghai. By now, six weeks after the eruption, the particles were refracting and reflecting and dissipating and dispersing light over fully sixty-two degrees of latitude and had spread, quite literally, half a world away from their birthplace.

And so the trend continued: by late October the amazing sunsets were causing bystanders to gasp and write poetry, to send letters to newspaper editors, and to paint vivid pictures in places like Tasmania, South Africa and the southern cities of Chile. After which the cloud of volcanic aerosols, still wafting and widening north and south, performed a strange shiver and apparently began to move backwards and outwards at the same time, so that sometime around 23 November, after touching western Canada and California, it became apparent not in Alaska, the Aleutians or Hawaii but in England, Denmark, Turkey, Russia and (coming from the west) Siberia.

Examination of all of the northern hemisphere records - including the 500 sunsets of the Chelsea artist Mr Ascroft, who started to paint in earnest in November what he saw each evening on the Thames - appears to show that countries lying to the east were to be affected later. It looks, in other words, as though the cloud that had in its early days been sliding ever westwards was now moving in the opposite direction, as though it were trying to describe a long and lazy spiral across the surface of the earth.

(Close study of these upper-air movements later turned out to be quite crucial for modern meteorology. Analysis of the effects of the Krakatoa eruption - which still continues today, particularly at universities in Hawaii, Rhode Island, Oxford, Auckland and Melbourne - has informed very many areas of science, but in particular has quite revolutionized the business of weather-forecasting, helping to nudge it from its earlier and rather dubious standing as a mere drawing-room fancy into something that approaches the modern science of today. Each time, for instance, that one sees a map of the unfolding patterns of the jetstream, it is well worth remembering that it was the study of the stratospheric movement of the Krakatoa aerosols that led to the understanding of this particular weather-making phenomenon. Reverend Bishop in Hawaii seems to have been the first to notice, and called the initial spread of ash the Equatorial Smoke Stream.)

The eruption particles continued to move ‘at about 73 mph’ - such again was the scientists' exactitude - until, by the end of the year, they settled wherever they were and stayed, moving downwards under gravity's pull with an infinite slowness. And so they seemed to hover permanently in the sky and, though having effects not quite as radiantly beautiful as in those first months, gave the greatest of pleasure to all below who saw them, for the next two or three years. Mr Ascroft chronicled the entire long life of these sunsets from his Chelsea studio: they did not, he wrote, ‘entirely fade from view until the early part of 1886’.

And the spectacle was only the half of it. The cloud reached the environs of New York in December. In the city itself they saw one thing. ‘The entire island seemed ablaze,’ said the World.* ‘Great tongues of flame shot up from across the water, reddening the southwestern sky and tinting the Jersey shore with delicate shell colors.’ The Times, then a less restrained newspaper than today, was only marginally more circumspect: ‘The clouds gradually deepened to a bloody red hue, and a sanguinary flush was on the sea; the brilliant colours finally faded to a soft roseate [Peter Mark Roget's Thesaurus had been available since 1852], then into pale pink, and finally died away upon the darkening horizon.’

Out in the countryside, the results were very different, and the florid sunsets caused no end of confusion. In Poughkeepsie, seventy miles north along the Hudson valley, the Daily Eagle noted with wry pleasure on Wednesday, 28 November, Thanks-giving eve, that:

Po'keepsie firemen have always been noted for their zeal and promptitude in going to fires, and their efficiency in putting them out, but the effort last night was a little too much for them. The light of an immense conflagration was visible, the bells pealed out vehemently and the boys ran vigorously down Market to Montgomery, down Montgomery to Riverview Academy they rushed - and when that point was reached it became evident that the fire was on the other side of the river, and a few moments of cooler reflection convinced them it was too far to reach, although it could probably last until they got there. Still, there was no means of transport available, and no likelihood of water. The fire being located 91,000,000 miles off, it would have taken the boys somewhere in the neighborhood of seven or eight million years to get there, let them run their best, and their terms would have expired before getting back. The sun is rather a large customer to tackle in the ways of fire, it beats a barn to death. We think the boys did well not to run out of hose.

Various causes are assigned for the peculiar light in the sky, the main one being the reflection of the rays of the declining sun upon the haze in the horizon. For several days past also there has been a peculiar state of atmosphere, which may or may not have something to do with it.

The no doubt embarrassed ‘boys' who were duped so nicely by the sunset belonged to the town's Hose Company No. 6 (of seven), better known as the Young America Hose Company, and they were based close enough to the river to be invariably the first called out to deal with a fire in the western end of town. A few days later, trying to assuage their pain, a meteorologist wrote helpfully to a neighbour paper, the Rochester Democrat and Chronicle, to say that the same confusing light had been experienced at his end of the state as well; and in his opinion whatever it was that had prompted the volunteers down in Poughkeepsie to dash off on their horse-drawn pumping engines was caused by ‘a stratum of decomposed vapours in the upper atmosphere’.

The Poughkeepsie Sunday Courier went further, suggesting with some prescience that ‘the apparent reflection of a large conflagration, which called out our fire department' was caused by the sun's rays bending through ‘extremely small water droplets… perhaps mixed with dust and smoke, sifting out the blue and yellow rays’. But another correspondent, probably taking down the office copy of Old Moore, offered only that the light show, all lumière and no son, was simply an augury of fine, anticyclonic weather in the days to come. Given the coming coldness of the season - made much worse by the dust - he could hardly have been more wrong.

Certainly the month of November 1883 was a time of unforgettable light shows throughout the northern world. In addition to the fire-spotters believing they had seen a blaze in the western suburbs of Poughkeepsie, there were reports that the fire-engines from New Haven, Connecticut, had been sent out too, for much the same reason. There was for a while a curious, half-panicked mood about people who had to see these ghastly skies night after night: to some they seemed almost apocalyptic, often unnerving; and it was only when they were explained away as being caused by dust from a distant volcano that people began to relax, and to bask in the simple sight of a terrible beauty they would long remember. But it was not to be a permanent fixture, there or anywhere.

And then there was the matter of temperature. Batavians noticed the chill immediately. At dawn on the very Monday morning of the eruption - though it was less a dawn, more a vague lightening of the drab ash-filled gloom - it was colder than records had shown for years - 65°, fifteen Fahrenheit degrees lower than normal. People were seen shivering in the streets - though perhaps as much from fear as from a need to keep warm. Dense clouds hung in the air for days afterwards, enveloping the city and an area perhaps 150 miles in diameter in a grey shroud through which the sun's rays could not penetrate.

Logic would suggest that a veil of dust particles spreading around the world would in time produce much the same effect, with those places that languished under the sunset-inspiring dust clouds feeling chillier than normal - though on a lesser scale than places near by. And to an extent they did. Oddly, the Royal Society, so assiduous in cataloguing the sound and light shows put out by Krakatoa, never bothered to consider the idea that the world might be cooled down by all the particles in the upper air. The Society's editors present a catalogue of the world's barometric pressures but none of the world's ambient temperatures. Those studies had to wait until much, much later - a rather puzzling omission, considering the time's enthusiasm for climate-related studies, and one never satisfactorily explained.

And when these studies of temperature came - the first carried out in 1913, the second in 1982 both found that as expected there had indeed been a worldwide drop in temperature. It had amounted on average to about one Fahrenheit degree and it had occurred, according to all the surviving records, at a time that appeared to be coincident with the eruption of Krakatoa. What has not been established, and what still concerns the scientific community, is which came first: did the eruption lower the world's temperature? Or did a lowering of the world temperature because of some other reason perhaps - unthinkable though it seems - somehow prompt the crust to undergo stress and strain and crack, and a rash of volcanoes to explode?

There is no doubt that there is a correlation, a definite link. Benjamin Franklin* was the first to notice it, and told an audience at the Manchester Literary and Philosophical Society that the ‘dry fogs' that had seemed to cool the European summer of 1783 and made the ensuing winter exceptionally bitter were almost certainly the work of dust in the air. He had examined the volcanic records and found a fissure volcano called Lakagíar, or Hekla, in Iceland that had erupted earlier in the year (curiously, exactly a century before Krakatoa). It had produced great clouds of dust, for week after week, which tumbled up high into the atmosphere. This, he declared, must surely be the culprit.†

The infamous eruption in 1815 of Tambora, on the Indonesian island of Sumbawa, 700 miles east of Krakatoa, ejected twice the volume of material into the atmosphere (eleven cubic miles of rock, ash and dust, compared with Krakatoa's six). The devastation it caused locally was profound - supposedly 50,000 dead, an entire language (Tambora) extinguished, an entire island rendered uninhabitable for years. But its climatic effects were astounding too. For it lowered the world's temperature by almost one Centigrade degree, on average: for every day when the normal temperature might be 33 °F, just above freezing, the temperature in the year after Tambora would be 31°, and ice would have formed on every pond and, more fatally, in every newborn crop, flower and hatching egg.

So in New England the farmers claimed that 1816 was ‘the year without summer’. There were frosts as far south as New Jersey in late May, in upper New England in June and July, and the growing season was slashed from the usual 160 days to seventy. Soup kitchens opened in Manhattan. Livestock had to be fed on fish carried over from the Atlantic seaports - 1816 is also still remembered as ‘the mackerel year’. There were crop failures - ‘the last great subsistence crisis of the Western world' - and, as a result, there was emigration to the western states. No small number of today's Californians can rightly lay responsibility for their being Californians squarely at the door of the proximate cause of that year's ruinous cold - Tambora, a volcano unknown to most of them, and 10,000 miles away. (Although there was migration into California from Europe, in Newfoundland the reverse took place: migrants were sent back east across the ocean, because there was not enough for them to eat.)

And yet back in Europe it was quite as bad. The weather for 1816 is the worst recorded, with low temperatures stretching as far south as Tunisia. French grapes could not be harvested until November. The German wheat crop failed entirely, and prices for flour had doubled in a year. In some places there were reports of famine, and in others there were riots and mass migrations. The diaries and newspapers of the day presenta litany of miseries. It is said that Byron composed his most miserable poem, ‘Darkness’ - Morn came and went -and came, and brought no day - the influence of that dismal year; and Mary Shelley may have composed Frankenstein while gripped by a similarly unseasonable melancholy.

Nowadays sophisticated instruments and the measurements they take have usurped the role of anecdote and diary. Ice-cores show minute layers of ash, or increases in sulphuric acid, as indicators of eruptive material in the atmosphere. And there is stunting of tree rings - the creation of a ‘frost ring’ in trees that have suffered through an exceptionally cold winter. The examinations of deep ice-cores, and of such trees as were living in the nineteenth century, have confirmed what the stories had long suggested: that eruptions of any of the world's larger volcanoes tend to coincide with periods of a cooling of the earth, some of the periods longer and with a very much lowered temperature, others shorter and with less of a fall in the mercury (the precise decisive factors are still not quite agreed). Tambora's eruption of clouds of ash coincided with a cooling of the world in 1815; and so, in 1883, did the appearance of the clouds of ash from Krakatoa.

The most tragic of cargoes to move out from the volcano happened also to be the slowest. The audible sounds and the unhearable shock waves may have sped away at more than 700 mph, and the dust may have wandered across the globe at more than seventy. The immense rafts of floating pumice that drifted away from where they splashed into the seas around Krakatoa made it as far as the south-east coast of Africa - but they did not make landfall for more than a year, travelling half a mile each hour, at best.

And when they did arrive, and were found washed up on the shores, they were discovered in some horrifying cases to have transported skeletons along with them, bringing as passengers the unidentifiable remains of some of the unfortunate thousands of Javanese and Sumatrans, Dutchmen and Chinese, who had perished.

Pumice is one of the better known by-products of vulcanism. The most widely used products are, of course, volcanically produced rocks, used for building stone. These range in type from the dark basalts and gabbros, such as were used for making the spires of Cologne Cathedral, to the paler andesitic rocks, as are to be found in the temple lanterns all around Kyoto. Volcanic soils, especially a group called andosols, are uncommonly rich in minerals, and, in the attractive phrasing of the Encyclopedia of Volcanoes, ‘have nourished many ancient civilizations’. And scores of the most ordinary, everyday items have incorporated volcanic products - particularly a highly absorbent mineral derived from weather ash called bentonite - in their manufacture: batteries, surfboards, refrigerators and air conditioners generally make use of ash invisibly, while the cheap building material known as breeze block makes no secret of its origins: crushed-up coke, cinders, furnace clinker and, in places where it is readily available, finely ground pumice and volcanic ash.

Pumice resides also in many an old-fashioned bathroom, sitting beside the loofah and the scrubbing brush. It is pleasantly abrasive and, because of the high proportion of gas bubbles that are caught in it before it solidifies in flight, it is of such low density that it floats quite readily. Makers of distressed denim fabrics also like to use it in their giant washing machines, where it brushes softly against the cloth, whitening and ageing it in a way that finds favour with youngsters.

But these more benign uses of pumice tend to shroud the awful truth about the immense tonnage of it that was released by Krakatoa. The headmistress of a mission school in Zanzibar,* off the east coast of Africa, wrote to the Royal Society in response to the appeal, to report that

… about the third week in July 1884, the boys… were much amused by finding on the beach stones which would float, evidently pumice-stone. The lady who was with them… also noticed that there were a quantity of human skulls and bones ‘all along the beach at high water-mark’; these were quite clean and had no flesh remaining on them, and were found at intervals of a few yards, two or three lying close together.

The closer to the volcano, the thicker the rafts of pumice, of course. On the west coast of the island of Kosrae, in what is now Pacific Micronesia, huge plates of pumice sixteen inches thick were hauled from the beach early in 1884: they were covered with barnacles, and many were accompanied by the roots of huge trees, with extra pumice lumps caught up in their roots, helping them stay afloat. These trees, torn up and floated 3,000 miles to the east, were presumably parts of Krakatoa's old forests - the same forests that had been noted and painted by Captain Cook's homebound expedition in 1780, and those that the Sundanese boat-builders had been cutting when they were forced to flee for their lives from the first eruptions of May.

The crews of ships moving through fields of pumice - such as those ‘acres in extent' that were encountered by one vessel coming into the Sunda Strait from Australia in January - were struck by the peculiar sound of the bow slicing through the rock. There was no real noise, ‘just a soft sort of crushing sound‘. And all the passing ships did their best to avoid desecrating the terrible cargo the pumice rafts all too often carried. A crewman on the vessel Samoa, which was heading south-westwards, off into the Indian Ocean, wrote of the nightmarish unreality of such encounters:

For two days after passing Anjer we passed through masses of dead bodies, hundreds and hundreds of them striking the ships on both sides - groups of 50 and 100 all packed together, most of them naked. We passed a great deal of wreckage, but of course we cannot tell if any vessels were lost. We also passed bedding chests and a number of

Rafts of pumice, many laden with the remains of victims, drifted as far away as Zanzibar.

white bodies, all dressed like sailors, with sheath knives on them. For ten days, we went through fields of pumice stone.

What the seamen witnessed on the Samoa, and on the Bothwell Castle, and on the Loudon and the Berbice and the Charles Bal and the Kedirie and a score of other ships beside that scoured the Sunda Strait during those weeks in late August, September and October does not bear too much repeating, so awful is what they have to say. Most reports were much more dreadful than the following account, which was published in a letter to The Times, from a correspondent writing from Batavia in October:

The British ship Bay of Naples had called at these islands and had reported that on the same day, when 120 miles from Java's First Point, during the volcanic disturbances, she encountered carcasses of animals including even those of tigers, and about 150 human corpses, of which 40 were those of Europeans, besides enormous trunks of trees borne along by the current.

Yet it is what they did not see that remains, to take the longer view, of greater significance. For what they did not see was this: the half-mile-high pointed peak of Krakatoa. They did not see it smouldering menacingly in the aftermath of what it had done, as most volcanoes are prone to do.

Usually - whether named Vesuvius, or St Helens, or Pinatubo, or Unzen or Etna - a volcano explodes, and in so doing causes manifold devastation and death. Then it simply stands there, raging and smoking ever less fiercely, presiding with a titanic smugness over the ruin it has so lately made. But, uniquely in much of volcanic history, Krakatoa did no such thing.

For Krakatoa had gone. Six cubic miles of rock of her, most of the island's great bulk, had just vanished, either blown into the sky or collapsed into the sea, and with the most thunderous roar and the greatest loss of life ever recorded in certain history.

For so long before that Krakatoa had been an island of no consequence. It had been little more than a genial, easily recognized sailors' companion spotted off the bow of an approaching ship, a sea-mark that would always help guide any navigator who was feeling his way up or down this most vital waterway between Java and Sumatra. It was that old ‘island with a pointed mountain’, and nothing more.

Now in the middle of the summer of 1883 it had suddenly and without much warning gone totally berserk, sent the sea berserk as well and then had, essentially, vanished. Not vanished in name, perhaps; not vanished from memory either; and recently, and in much the same form, it has been reborn (as we shall see). But in August 1883 this inconsequential little island went mad and disappeared. The reasons just why this happened - as it did, and when it did - have occupied the minds of a great community of geologists around the world for all the long years since.

4. The Explanations

Why did Krakatoa happen? Why, indeed and more generally, do volcanoes do what they do? Why does the terra firma upon which we so confidently and innocently secure all our lives, sometimes and so capriciously tear itself open and cause such fearful destruction as it does so?

To those caught up in such a moment of appalling terror, such as the thousands whose lives were wrecked in 1883, it all must seem a most monstrous injustice, a terrible cheek perpetrated by the earth and its presiding deities. Krakatoa is a stark reminder of the truth of Will Durant's famous aphorism ‘Civilization exists by geologic consent, subject to change without notice.’ Yet geology, which is an unemotional and rational science, allows us to step back from our shock and dismay at such events, to accept a longer view - and to be awed by something rather different: that despite her seemingly cruel caprices, this planet in fact enjoys by and large an extraordinarily fortunate situation.

The simple, very obvious features of the earth - its location in space, its size, the processes that led to its creation, processes that include the very volcanic events that took all the lives west of Java - happen to have been suited perfectly, when taking the long view, to the sustenance and maintenance of organic life.

To victims of a volcanic eruption like this, of course, the very reverse must seem true. But consider location, for instance. Planet earth is sited just close enough to the star around which it orbits to derive only benefits from the latter's infernal solar heat. It is neither so close as to risk the boiling of its oceans and the loss of its water into outer space by photo-dissociation in the upper atmosphere, nor so far away that all its present liquid water remains uselessly and inconsumably frozen.

The size of the earth is spot on too. Thanks to its moderate size its gravitational pull is just right. It is strong enough to overcome in particular the escape velocities of the molecules both of water and carbon dioxide, which means that we have a sheltering canopy - a benevolently situated greenhouse, even though this is a word with more negative associations today - that first allowed life's building blocks to be assembled, and then ensured that the fragile living entities so made could be cosseted against the perilous radiations from outer space.

And then there are the volcanoes - just the right number, of just the right size, for our own good. The deep heat reservoir inside the earth is not so hot, for instance, as to cause ceaseless and unbearable volcanic activity on the surface. The amount of heat and thermal decay within the earth happens to be just perfect for allowing convection currents to form and to turn over and over in the earth's mantle, and for the solid continents that lie above them to slide about according to the complicated and beautiful mechanisms of plate tectonics.

Plate movement and convection and the volcanic activity that is their constant handmaid may not seem, to the victims of eruptions and tidal waves, to be in any way benign, or to be good for the planet as a whole. And yet, taking the long view once again, they most certainly are: the water, carbon dioxide, carbon and sulphur that are so central to the making and maintenance of organic life are all being constantly recycled by the world's volcanoes - which were also the probable origins of the earth's atmosphere in the very first place. It is not merely that volcanoes bring fertile volcanic soils or useful minerals to the surface; what is more crucial is their role in the process of bringing from the secret storehouses of the inner earth the elements that allow the outer earth, the biosphere and the lithosphere, to be so vibrantly alive.

Almost all our neighbour planets are, so far as is known, volcanically lifeless. They are also, on all the available evidence, more or less biologically lifeless - and that is quite probably at least in part because they are so volcanically dead. Only Io, one of Jupiter's many moons, seems to sport a significant number of volcanoes: spectacular sulphur-rich fountains of magma have been seen spouting on its surface. But there is no suggestion of plates or of any movement of the solid crust, either on Io or on any planet or moon known to exist between Mars and Pluto. The vigorous business of plate movement apparently does not occur on planets that are hotter than our own; nor does it on those that are much more frozen and more deeply dead.

But it is the movement of the plates, and the internal storms that rage below and cause them to slip beneath or alongside one another or tear themselves apart along their suture-lines, that is the driving force behind our earth's highly unusual degree of vulcanism. Plate movement, as well as shaping the planet's topography, also creates most of the very vulcanism that is central to its life. Plate tectonics, in other words, is the key to it all - and any examination of just why Krakatoa happened as it did, and how it did, must inevitably now refer to this newly minted catalogue of knowledge about the workings of the earth.

It was of course not always so. In the distant past, whenever the earth behaved with terrible and unanticipated violence, mankind could do little more than wonder, horror-struck, at the sheer effrontery of it. In very early times this wonderment was answered, inevitably, mainly by religion and the making of myths. Volcanoes were hills occupied by temperamental gods: they could be appeased by frequent sacrifice. The appeasing flesh could be that of a young human (a small child thrown every twenty-five years into the crater of a particular Nicaraguan volcano, for instance, would guarantee its quietude) or an animal (Javanese today toss chickens into the crater of Mount Bromo - superstition plays an important role in East Indian attitudes towards their volcanoes still).

The ancient Greeks and Romans then hammered some kind of order into their beliefs, as might be expected: the idea of the existence of Hades, the nature of such gods as Pluto and Vulcan, the character of Titanic monsters like the fearsome, wild-eyed and flaming-tongued Typhon were all connected with the wayward behaviour of an earth that all then knew had a terrible and dangerously hot interior. It was no coincidence that the gateway to Hades - believed by the Ancients to be in the earth's centre - was the Romans' most notorious local volcano, Mount Etna, with its gas-belching vents known as solfataras, and the phrase ‘sailing to Sicily' was for a while a euphemism for entering the fiery furnaces of the Devil's domain.

The seers of the classical world were on rather shakier ground when it came to deciding just why, other than for divine reasons, there was just so much heat inside the earth. The Greeks - the philosophers Anaxagoras and Aristotle in particular - favoured the human analogy of trapped wind, with the friction of the escaping wind causing the generation of heat, a sort of volcanic vindaloo. The Romans, on the other hand, and among them most notably Lucius Seneca, favoured the notion that the heat came from the combustion of a vast inner-earth storehouse of sulphur - and in some Roman poetry of the time this idea extended to the burning of deeply buried reservoirs of alum, coal and tar.

This idea, that volcanoes were the consequence of the steady burning of a finite store of earthly combustibles, exerted a grip on the scientific mind for centuries. Then, as chemistry developed as a science, so its innumerable secrets offered themselves as the favoured sources for all the necessary heat, and were widely accepted as doing so. During the seventeenth and eighteenth centuries a great many seers - Isaac Newton among them - believed that so-called exothermic chemical reactions were the answer. By 1807, when the Geological Society of London, the world's oldest such body, was founded, the oxidation of newly discovered alkaline metals, such as sodium and potassium, was thought to be an answer.

Even as late as the 1920s there were two now notoriously blinkered scientists who clung to what might seem today quite fatuous chemical theories. One of them, Arthur Louis Day, proposed in 1925 that volcanic heat was due to a series of complex chemical reactions between gases, and he won support from the redoubtable and influential Sir Harold Jeffreys,* while at the same time dismissing vulcanism generally as a phenomenon that was merely ‘local and occasional, not perpetual and worldwide’.

However, in tandem with all those chemists and physicists who for so long had such an influence on geophysical thinking, there were also other natural philosophers - René Descartes most notable among them - who started out on what would prove to be the right track. In the mid seventeenth century Descartes - better known for his cogito, ergo sum, and for his legacy of Cartesian coordinates - came up with a quite revolutionary idea: that the earth originated by way of gravitational attraction and gaseous condensation, that heat was an essential primordial component of this process, and that its slow decay resulted in the earth having three internal concentric parts: a highly dense and incandescent liquid core, a half-cooled plastic central region, and a cold, solid and comparatively light crust. Moreover, there was ample primordial heat left over from the creation process to power all known volcanoes for a very, very long time.

The subsequent advent of the science of field geology, the furious debates that went on between Neptunists, who believed all rocks to have precipitated from a primeval ocean, and Plutonists, who saw countless of them as having their origins in melting and magma, belongs to another story, temptingly diverting though the various interlocking sagas may be. In essence, though, the mystery that occupied most minds for most of the late nineteenth and early twentieth centuries was simply why rocks melt - what combination of physics and chemistry, of depth, of heat and of the presence or absence of water in the mix of minerals would lead a rock to become plastic and mobile and molten, and then to emerge on the surface and cool and harden and solidify back into rock once again.

The chemist and the chemistry that sought to answer questions about the make-up of the earth in earlier times had been overtaken by the physics and the physicists seeking to do much the same in recent years; and though the physics answered much of the detail involved, many of the fundamental questions remained doggedly, in essence, unresolved.

Or at least they did until that memorable July day in 1965 when, as I have explained in an earlier chapter, the soft-spoken and self-effacing Canadian geologist J. Tuzo Wilson managed to combine both the chemistry and the physics of the earth into one, inaugurating the science of plate tectonics. In doing so he launched a brand-new and all-encompassing global theory that would offer the answer to almost everything volcanic that had ever been wondered at.

Uniquely in the solar system this planet sports a crust that is, by virtue of this process, being constantly destroyed and regenerated - an ever mobile chemical factory where materials that exist in solid, liquid and gaseous states are being recycled endlessly. They are burst out from the middle of oceanic plates by a process, newly understood, that allows upwelling materials to melt without heat being added to them - to melt simply because the pressure on them is relieved by their being convected upwards and outwards towards the atmosphere.* There are volcanoes here in these mid-crustal ridges, big but not especially explosive volcanoes, mountains that ooze basalt, like those in Hawaii and Iceland, the Azores and the rift valleys of East Africa.† They are the stuff of research and fascination in their own right. But they are the distaff side of the volcano of this account, the Alpha to Krakatoa's Omega, the mid-plate reciprocal to all that goes on at the plates' edges, the other side of the story.

For the materials that rise up in the middle, along with whatever they sweep before them on their way, are in due course swept down again at the peripheries of plates. They are swept down by the process that is most crucial of all, which, though an essential part of earthly regeneration, also leads directly to the making of highly explosive, dramatic and deadly arc volcanoes like Krakatoa. Colloquially the phenomenon that exists at these plate edges is known among geophysicists and vulcanologists as the subduction factory - and Krakatoa stands front and centre in one of the largest and most complicated of these extraordinary, world-shaping entities.

The factories and the subduction zones that underpin them are essentially coextensive. It probably bears repeating that each of the zones is where one of the world's many heavy oceanic plates slowly collides with one of the many lighter and thicker continental plates and slides, buckling as it does so, underneath. The zones are very long, and very thin. If unravelled they would extend for about 19,000 miles. But they are rarely more than sixty miles wide. The total area of the subducting world's assembly-lines amounts thus to about a million square miles - about the size of Greenland, or the American Confederate South, or Argentina.

And enclosed within the zones, and formed, allowed to grow, and then destroyed or mutated or otherwise dramatically affected by the processes going on inside them, are about 1,400 of the world's 1,500 historically active land volcanoes. Of all visible volcanoes, 94 per cent, in other words, stand within subduction zones. A mere handful of countries - Indonesia, Japan, America, Russia, Chile, the Philippines, New Guinea, New Zealand, Nicaragua chief among them, and in that order, play host to most of them: these nine countries are home to more than nine out of every ten volcanoes that are liable to erupt today or have done so in recent history.

The most readily recognizable subduction zones are those that enfold the Pacific Ocean. As a reasonably familiar example, consider that which runs along the western edge of South America, and which has created the chain we know as the Andes.* This is where the heavy basaltic Nazca Plate collides with the lighter granitic-and-sedimentary-rock South American Plate. (It does so simply because it is at the same time splitting itself away from its neighbour, the Pacific Plate, along what is called the East Pacific Rise - close to where the Isla de Pascua, Easter Island, is hoisted above the ocean surface.)

The subduction factory that results is a classic of its kind, creating dozens of volcanoes running from the Andean peaks of Ruiz and Galeras in the north, in Colombia, via Chacana, Cotopaxi and Sangay in Ecuador, Huaynaputina† in Peru, Lascar in Chile, Llaima and Villarica on the frontier between Argentina and Chile, and, at the southern tip of the continent, Monte Burney and Cerro Hudson, this last volcano erupting massively in 1991. All told there are sixty-seven volcanoes that have been manufactured by the processes of this one subduction zone - and since there are 4,000 miles separating northern Colombia from southern Chile, and since there is a sort of serrated regularity to the Andes, that means there is more or less one volcano piercing the sky every sixty miles.

Much the same number of volcanoes, with similar intervals between them, is to be found at the other subduction factories around the Pacific - in Alaska and the Aleutian Islands, in the Kamchatka Peninsula, in Japan and the Kurile Islands; and an even greater number is to be found in the most volcanic part of the world, the great subduction zone that stretches 3,000 miles from the northern tip of Sumatra to what is called the Bird's Head on the north-western tip (the West Irian side) of the island of New Guinea.

In this immense factory, there are at least eighty-seven volcanoes that make up much of the archipelago that politics has lately chosen to call Indonesia and the Philippines. Indonesia itself has and has had more volcanoes and more volcanic activity than any other political entity on the earth, in all recorded history. It is a country that is defined by its place at the heart of a subduction zone and is essentially made up of volcanoes and precious little else. On the island of Java alone today there are twenty-one volcanoes that remain fully active. Their eruptions are invariably spectacular and terribly dangerous. And because a very great many people live and work near the volcanoes (not least because the volcanic soil, thanks to the recycling mentioned earlier, is nutritious and ideally suited for farming), they are the cause of a dismaying number of deaths.

The earth fashioned three of the five greatest volcanoes of historic time in this one gigantic factory. The largest the world has ever known was made there: Mount Toba, which erupted 74,000 years ago in what is now northern Sumatra. It had a Volcanic Explosivity Index, or VEI, of 8 - the highest on a scale that is now universally used to classify all eruptions (save for those which merely ooze lava, without exploding). Toba's humongous

caption
The world and its pattern of tectonic plates. Where oceanic and continental plates meet, there is volcanic and seismic activity in great - and often terrible - abundance.

explosion - the curious adjective is now officially used to describe giant volcanoes, the equivalent of the cyclonic sea-state that is these days termed phenomenal left behind an immense lake, fifty miles long and fifteen wide, with the sheer caldera cliffs rising 800 feet straight out of the water. The eruption left layers of dust eighteen inches thick on the ocean floor 1,500 miles away, and must have placed a severe crimp on the development of such Ur-humans as were struggling for existence in those times: it must have lowered the ambient temperature by many degrees, and made even more harsh a climate that was already in the midst of changing into yet another Ice Age.

The eruption of Tambora in 1815, in this same subduction zone, is reckoned the second greatest in history, with an Explosivity Index of 7. (This index, which was first created at the Smithsonian Institution in Washington, is based on two features: the amount of material that is ejected in an explosion, and the height to which it is hurled through the atmosphere. These two factors are clearly observable in modern eruptions; they are also deducible from the records of the past. Even though there were no literate eyewitnesses to Toba, and very few to Tambora - which has to be the principal reason neither has lingered in the public consciousness, while Krakatoa clearly has - the total mass that was ejected from each eruption can be calculated with some precision from an examination of the local geological record, and the distribution of the ash falls far away on the seabed can suggest with fair accuracy the heights to which the columns rose into the sky.)

Third in the list is Taupo in New Zealand, which erupted in AD 180, hugely; and fourth is Novarupta in Alaska - better known as Katmai, on the landward end of the Aleutian chain - which did so in 1912. This last was the largest recent eruption on the North American mainland, but, because of its remoteness, it was little noticed except by what - in terms of calderas and domes and frozen lakes - it left behind.

And then, fifth in the list of all volcanoes known, with a VEI of 6.5, with more than six cubic miles of rock and ash and pumice and dust hurled dozens of miles into the lower stratosphere, with sounds heard 3,000 miles away, with tidal waves of enormous force and height, with shock waves that ran four times to the far side of the world and almost three times back, and with more people killed and more livelihoods ruined than by any other eruption in world history, comes Krakatoa.

Seven weeks after the eruption, when the dust had cleared, the Netherlands government ordered Dr Verbeek and his colleagues to investigate exactly what had happened. The team of four took off in the government hopper-barge the Kedirie, on 11 October, and spent the following two weeks examining every possible aspect of the now seemingly dead remnants of the mountain. The orgy of self-destruction was more than amply evident. It was, Verbeek wrote later, ‘the most interesting eruption witnessed by the human race until now’. But very little of the original mountain was left to see.

The southern quarter of the island remained but was sliced open, as if with a vertical carving-knife - so that the original peak of Rakata looked, from the south, almost the same, but with everything to its north missing. The exposed northern face of Rakata was almost perfectly vertical, and in cross-section, when viewed from the north, quite perfectly triangular; it was pierced with vertical lines and radiating systems of lava-filled dikes and sills and plugs of newly formed rock, all covered with several feet of grey pumice dust, so that it looked from afar exactly as it was - a perfect cross-section, as one might see in a teaching-chart, of a one-time volcano that had been blasted in half and into oblivion.

But at least the summit of Rakata was, more or less, still there. The two northern peaks of Krakatoa, the summits known as Danan and Perboewatan, were, to the expedition's awestruck fascination, no longer anywhere to be seen. Nor was the little skerry of andesite known (because of its shape) as the Polish Hat: it had quite vanished, presumably vaporized in that one paroxysmal instant.

The very opposite had happened to the two small islands, Lang and Verlaten, which had once enfolded Krakatoa like a pair of parentheses. Rather than vanishing into thin air, these two now appeared to be very much larger than they had before. Their beaches, it turned out, had since the eruption become choked and swollen with enormous amounts of stranded pumice. Larger they may have been - but the essential difference about their appearance as maritime parentheses was that now they were enfolding and bracketing nothing - between them was just a huge expanse of empty, lifeless sea, with the immense broken fang of Rakata peak rising alone straight out of the ocean as a reminder of what had once been there.

The sea to the immediate north of the cliff was very deepnearly a thousand feet. Clearly an enormous new caldera had been created - almost the entire volcano had collapsed into an immense void below, and the cliff of Rakata, neatly bisected where it had been shorn off, was all that remained to the south of the collapse. Off to the north-east two entirely new little islands had risen from the waves, and were christened Steers and Calmeyer Islands; because they were composed of little more than stranded rafts of soft pumice they were eroded back to sea-level in very short order; on today's charts there are just warnings of ‘patches of discoloured water’, fifteen feet deep, suggesting where they used to be.

Back in 1885, when Verbeek wrote his official report, there were only the vaguest explanations of just why all of this might have happened. It was easy enough to describe what had happened - the science of descriptive vulcanology was in any case well advanced, and had been for many years. But when the vulcanologists of the day came to explain the reasons for the violent behaviour of their charges - as true for every volcano in the world as it was for Krakatoa - there was very little understanding of the processes of the world to offer them a basis for coming up with a theory.

After all, only a few decades before, many believed that basalt and flows of lava were simply precipitates from the sea. Until 1857 many geologists thought that volcanoes were caused by the bulging upwards of horizontal flows of lava - and not that they had been built vertically by the discharge of their own products. And at the time of Krakatoa's eruption Alfred Wegener - who first came up with the idea of continental drift, which was to lead to the theory of plate tectonics, and who might well have set the bewildered community of vulcanologists off in the right direction - was only three years old.

And so while in all the official reports and learned papers about the event there was plenty of description of the ruin and dismay that Krakatoa had caused, and though there was much speculation about why the volcano exploded with the violence it displayed, there was next to nothing by way of sensible wondering about the larger mechanisms that triggered it.

This was true for Verbeek in his report, for instance. He spent countless pages describing in detail clogged pipes, steam vents and collapses of central parts of the main volcano. What he concluded did display a remarkable prescience: he said that a good deal of the vanished volcano had foundered into the sea, and had not been blasted into the atmosphere. He suggested that the Plinian violence of the explosion was a result of sea-water mixing suddenly with the magma, and flashing over, turning into superheated steam, in a gigantic and uncontrollable explosion that is these days given the somewhat less than attractive name of a phreatomagmatic eruption. But he never tried to step back and wonder why Krakatoa was where it was, and why it did what it did in the first place.

The same was true also for John Judd, president of the Geological Society of London and author in 1881 of a then classic work, Volcanoes. He too wrote eloquently of the way in which hot magma and sea-water mixed, and of how pumice was created by a lowering of the magma's melting point by the addition of water - but, once again, he missed the central point. He never even tried to grapple with the central issue: why Krakatoa?

The last popular book on the subject* was written in 1964. Even then, still lacking any solid theory that might account for the world's inner processes, the author could only really describe what a volcano was (‘a hole in the ground through which hot gas, molten material and fragmentary products rise to the surface’), say where volcanoes were to be found and name the kinds of material that came from them. And when he arrived at the specific case of Krakatoa, which was described over many pages and with a quite magical literary skill, the author turned desperate. The book begins to speak in terms of ‘The Demon’ going in to ‘press the attack’, his ‘searching fingers boring into the defences’, and the pent-up energies of time' and ‘primeval forces’ readying themselves to do battle. One can hardly blame him. Neither he nor anyone else had an inkling of what really caused Krakatoa. And that was hardly his fault: he was simply a very few years too early.

But once plate tectonic theory was in place, all that changed. Nowadays there is a ready explanation for what happened and why. Essentially the same explanation accounts for the eruption of Toba at the north-western end of the subduction zone, and for Tambora's at its eastern end, and for those of all the other volcanoes in between.

Krakatoa erupted because of what happens when two plates collide - specifically, because of what happens when the northbound Australian Oceanic Plate collides, as it has been doing for many millions of years past and as it continues to do today, with that part of the Asian Plate that, for the sake of simplicity, we will call by the name it enjoys today, Sumatra.

The oceanic plate is cold and made of the heavier, darker, less acidic suite of rocks that underlie all oceans, so as it hits it begins to sink below the warmer and lighter rocks of which Sumatra and all other continents are made. As it sinks it takes with it, downwards, the small, wedge-shaped sliver of continental rock that it either scratched or smeared off the Sumatran edge; it also takes along with it some of the sands and clays that had accumulated on the Sumatran coast, the water that was trapped chemically in these, and a fair amount of atmospheric air and sea-water besides. This entire geological cocktail - an amalgam of cold and heavy basalt from the plate; granite-like rocks from the Sumatran crust; sands, clays, limestones and vast quantities of air and water - then plummets. And, as it does so, everything suddenly changes.

Water is the crucial ingredient in this process. Not only does it lubricate the motion of the plates and help the subduction continue, but, even in very tiny amounts, its presence lowers the temperature at which the rocks of the mantle will begin to melt. And since the water also lowers the density of the wedge-shaped mélange that is being swept and smeared downwards too, the molten rock that is being created below it finds that the rocks above it have suddenly become (thanks to the water) less dense, less rigid, less strong. They have become, in other words, a perfect exit route for the partly melted rock below, enabling it to rush upwards, to melt even further because of the decompression mentioned earlier. Then, with the dissolved carbon dioxide and water vapour suddenly turning back into gas and frothing out of solution, the whole mass rushes up and out as a torrent of phenomenal explosivity into the unsuspecting open air: as a gigantic and classical subduction-zone volcano.

That is why Krakatoa exploded. As to why it exploded so

A cross-section showing the basic elements of an oceanic plate - upwelling new material from the centre, spreading outwards and then slid-ing beneath the lighter continental plate it then encounters. Volcanoes and earthquakes are an inevitable feature of this last process of subduction.

powerfully and so very noisily - this entirely different debate continues apace today.

There are some clues. The geography of the islands indicates, for example, that there was once an ancient super-Krakatoa, and that at some unspecified time in the past it exploded and collapsed into itself, leaving behind a caldera; the parenthesis-islands of Lang and Verlaten were clearly the caldera walls, the cliffs at the edge of the old volcano. The volcano that succeeded this then had three distinct peaks - Rakata, Danan and Perboewatan. Each was an exit passage for a gigantic magma chamber that clearly existed deep below the region.

So there can be little doubt from this simple evidence alone that the 1883 Krakatoa existed above a large chamber of magma, and that with the three exit-pipes above it weakening its roof, it had a propensity, in times of violent stress, to collapse. The question that has occupied the minds of many specialists in recent years is this: was it the fact that sea-water managed to get into this chamber at the moment of collapse that was the primary cause of the deafening explosion and the tidal waves? Or was it simply one contributory factor, with some other process also at work to make things even more dramatic?

A series of experiments performed in pressure vessels in a laboratory in Australia has suggested that other factors were indeed at work - but that they were complex and subtle. They suggest that a pulse of fresh basalt from deeper in the earth may have been unexpectedly injected into the base of the magma chamber; that this new pulse heated the existing magma above it, causing a violent convection current and the sudden frothing of even more gas - and the sudden breach of the chamber roof. This idea of magma mixing has lately taken hold: processes going on even more deeply within the subduction zone may perhaps have contributed to the might of the Krakatoa event.

And then, finally, there is the overall siting of Krakatoa, halfway between Java and Sumatra. It lies directly above what might be called a hinge point around which the two islands are slowly swinging, the Strait ever widening, the islands turning like the pages of a northward-closing book - Sumatra moving to the north-east, Java to the north, Krakatoa in the middle.

There is certainly a complex network of faults in the Sunda Strait. Their existence is one of the reasons why there is a strait there, an absence of island mass, in the first place. Slowly, very slowly, science is trying to make some sense out of the complexity of it all. Geophysicists in Troy, New York, have spent recent years placing a rash of Global Positioning System receivers on the nearby islands, and have found that all manner of tiny movements are taking place - that the main subduction continues to creep away, as it has done for millions of years, but that tiny little sideways jogs are taking place too, small weakenings, creations of a blizzard of tiny faults that make the region into the most remarkable of geological laboratories - a fascinating study, even if Krakatoa had never existed.

The basic tectonic structure of the region: the Australian Plate is moving north to collide with the Asian Plate, while all manner of stresses and faults build up along the collision zones by Java and Sumatra.

But it does exist, and will play its tricks on the world once again, and before very much longer. The processes that led to the events of August 1883 are unstoppable. There is a subduction factory of monumental proportions to the south and east of Sumatra. It is uniquely sited around and beneath the small island that lies on the hinge-point between Sumatra and Java. The island is surrounded by the volatile waters of the sea, water that causes mayhem if it gets within a mile of boiling magma. The island itself is surrounded by countless small faults and zones of weakness, as well as by a raft of components of basic rock - acidic rock, sedimentary rock - that is twisting and turning every which way under a barrage of stresses and strains that exist more notably here than anywhere else on earth. Small wonder, indeed, that there is only one Krakatoa. The place where it blew itself apart is so geologically dangerous that one can almost imagine there being room for a dozen more.

*

They counted their dead, and they buried them where they could, which was usually where they found them. The Dutch officials reacted with commendable speed, burying bodies at the rate of several hundred a day, drenching the swamps with carbolic acid, pulling down wreckage, setting cleansing fires. The king back in Holland opened a fund. Dutch mothers sent blankets, tents, food. A flotilla of ships travelled east, to see what could be done. The Great World Circus staged benefit performances in Batavia, before packing up and going home, taking their distressed little elephant with them. The process of rebuilding began, with the Fourth Point lighthouse at Anjer, remade with iron plates, hurried back into operation to ensure the safety of the commercial shipping lanes, a symbolic beginning of the rebirth. The cable lines to and from west Java and south Sumatra were repaired. The aid workers came into town. The charities set up shop. The scientists fanned out to investigate, to report, to recommend.

But in time they all went home again, to deal with other problems and to answer fresher questions. They left the coastal people of Java and Sumatra, and those island shore-dwellers known as the Bantenese, among their patched-up ruins, and in time they forgot all about them. They did not stop to wonder where these people might eventually look for sustenance and succour.

Perhaps they should have done. For it turned out that not a few of these unhappy, dispossessed and traumatized people eventually looked to the west, to Mecca, and to the benevolent power of their Islamic religion to answer their needs. This was a political and religious consequence of the disaster - a consequence entirely unanticipated by the ruling Dutch colonists - that was to have the most profound and longest-lasting fallout, for the Indies, for Europe, and beyond.