'I read about that,' said van Maarten. 'The Canary Islands authorities say the theory is dubious.'

'Dubious?' Frost thundered, like the trumpets of Jericho. 'If anything's dubious, it's their failure to address the problem in any of their statements. All they care about is not worrying the tourists. But this problem isn't going to go away. The world's already experienced similar disasters, albeit on a smaller scale. In 1741 Oshima-Oshima erupted in Japan, triggering thirty-metre-high waves. More waves were generated by the collapse of Ritter Island in 1888 in New Guinea, but the amount of falling rock was barely one per cent of the landslide that could take place here. A GPS network has been continuously monitoring Kilauea volcano in Hawaii, looking for any sign of movement, and it sure as hell is moving. The south-eastern flank is slipping seaward at an annual rate often centimetres, and God help us if it starts to gain momentum. The consequences are too dire to imagine. Volcanic islands have a tendency to get steeper with age. Eventually a section breaks off. The authorities on La Palma don't want to face the truth. It's not a question of if it will happen, it's a question of when. In a hundred years? A thousand? The only thing we can't be sure of is the timing. The volcanoes here don't give much warning.'

'So what would happen if half of the island fell into the sea?' asked the executive.

'The mass of rock would displace vast quantities of water,' said Bohrmann. 'A dome would form on the surface of the ocean. According to our estimates, we'd be looking at a speed of impact of three hundred and fifty kilometres per hour. The fallen debris would extend sixty kilometres over the seabed, stopping water flowing back over the landslide, and creating an air cavity that would displace far more water than the volume of the rock. There's some debate about what happens next, but none of the scenarios are especially comforting. The landslide would create a mega-wave off the coast of La Palma, with a probable height of six to nine hundred metres. The wave would set off across the Atlantic at a thousand kilometres per hour. Unlike earthquakes, landslides and slope failures are point events, which means the wave's energy dissipates as it radiates across the ocean. The further it travels from its source, the flatter it becomes.'

'At least that's something,' said the technology specialist.

'Not really. The Canary Islands would be wiped out in a flash, then an hour later, a hundred-metre-high tsunami would wash over the northwest African coast. Think of it this way: the European tsunami reached a height of forty metres in the fjords, and we all know what happened there. Six to eight hours after the eruption, a fifty-metre wave would sweep over the Caribbean, laying waste to the Antilles and flooding the east coast of America from New York to Miami. Soon afterwards the wave would hit Brazil with similar force. Smaller waves would travel as far as Spain, Portugal and the British Isles. The consequences would he devastating, even in central Europe. The European economy would collapse.'

The representatives from De Beers paled. Frost grinned at them. 'I don't suppose you've seen Deep Impact?'

You mean the movie? That wave was a lot higher,' said the executive. 'It measured hundreds of metres.'

'Fifty would be enough to flatten New York. The impact of the wave would release more energy than the United States uses in a year. It doesn't matter how tall a building is – it's the base that takes the force of the tsunami. The rest of the building collapses, regardless of how many storeys there are. And we won't have Bruce Willis to save us.' He gestured towards the edge of the ridge. 'There are two ways of destabilising the western flank of the island: either Cumbre Vieja erupts, or there's an underwater avalanche. The worms are working on a landslide – a kind of miniature version of what they did in Europe, although the force will be enough to detach a segment of volcano. The rock will sink into the depths, and that in turn will prompt a minor earthquake and destabilise the Cumbre ridge. The earthquake might even trigger an eruption, but in any event the western flank will detach. It's going to happen either way – and we'll have a disaster on our hands. The worms off the coast of Norway took a few weeks to finish the job. Things could move even faster here.'

'How long have we got?'

'It's almost too late already. Those worms are pretty cunning, and they've gone to work in spots that are hard for us to reach. The whole scheme depends on the power of mega-waves to propagate on open water. They've already scored one hit in the North Sea, but that was relatively minor by comparison. If this harmless-looking little island collapses into the ocean, human civilisation is going to see just how tough things can get.'

Van Maarten rubbed his chin. 'We've already produced a prototype for the suction pipe. It's operational at depths of up to three hundred metres. We haven't tried it any deeper yet, but…'

'We could extend the pipe,' suggested the executive.

'We'd have to figure something out pretty quick, but if we stopped work on everything else… What worries me is the support vessel.'

'A vessel won't be large enough,' said Bohrmann. 'A colony of a billion or so worms – that's a huge biomass. You'd have to find somewhere to pump it.'

'That's not the real problem – we can always set up some kind of relay. No, I was thinking about a command ship with the control desk for the pipe. If we extend the length to four or five hundred metres, we'll need a vessel big enough to transport it. That's half a kilometre of pipe! It'll weigh God knows how much, and it's a damn sight thicker than deep-sea cable. We won't be able to coil it up and stick it in the hold. Besides, we need the boat to stay stable while we're steering the pipe. I don't think we need to worry about an attack, but the hydrostatics is going to be tricky. A pipe of that length can't be left dangling from the side of the vessel without affecting its balance in the water.'

'How about a dredge?'

'Wouldn't be big enough.' The man thought for a moment. 'A drillship, maybe. No… We'd be better off with a floating platform. We're familiar with those already. We need a kind of pontoon system, ideally a semi-submersible construction like the type they use in the offshore industry – except we wouldn't want to anchor it in position. We'll have it travelling across the water like a normal boat. It has to be manoeuvrable.' He moved away from the others, muttering something about resonant frequencies and swell variations. Then he rejoined them. 'A semi-submersible should do the trick. It's stable, mobile, and provides an ideal base for the boom, which, let's not forget, will have to take a lot of weight. There's a semi-submersible in Namibia that would adapt quite easily. It's got two propellers, each with a six-thousand-horsepower engine. We can get it fitted up with some additional thrusters, if we think we might need them.'

'The Heerema?' asked the executive.

'Right.'

'I thought we wanted to get rid of her.'

'She's not ready for the scrapyard yet. She has two main rudders, and the deck's supported by six huge columns – it's just what we need. OK, it was built in 1978, but it'll do the job. It's the simplest solution. We won't need a derrick, we'll have two cranes instead. We can use one to lower the pipe. Pumping up the worms won't be a problem. And we'll be able to moor the vessels before we fill them with worms.'

'Sounds good,' said Frost. 'When will she be ready?'

'Under normal circumstances it would take six months.'

'And in these circumstances?'

'I can't promise anything. Six to eight weeks, if we start right away.' The man looked at him. 'We'll do everything in our power to have it ready as soon as we can. We're pretty good at that kind of thing. But if we get it done in time, you should see it as a miracle.'