Arthur Range is dissolving. Rainwater, which is weakly acidic, runs over the marble and erodes it, fetches up in channels, trickles down fissures and carves them deeper, wider, creating tubes of rushing water that plunge into the mountain’s core. The tubes connect with other tributaries, trickling over avens, tumbling over waterfalls, a great subterranean plumbing system growing ever larger and more complex.

And so it has been for eons; the ‘Marble Mountains’ in Kahurangi National Park west of Nelson are now shot through with holes—some of them pinholes where water seeps, others chambers of cathedral proportions lifted so high above the water table by geological forces that they are as dry as dust, aged and crumbling.

The entire labyrinth is shrouded in utter darkness. Light has never penetrated these sanctuaries, and many, if not most, remain unknown to humans.

Since cavers discovered an aperture among nettles above the Pearse River in 1969, the Nettlebed system—encompassing some 24 kilometres of passages—has become a mecca of ‘big caving’. Three of the biggest cave systems in New Zealand are found in Kahurangi National Park—Nettlebed, Ellis Basin (the deepest) and Bulmer Cavern (the longest). While thousands pour through the glowworm caves at Waitomo and hundreds have explored the caverns of Waipu, explorations of the very large systems in the Arthur Range are the preserve of just the most-experienced and best-prepared cavers, numbering perhaps 50.

Of this elite group, a half dozen banded together in 2009 around a core of cavers, Troy Watson, Aaron Gillespie, Neil Silverwood, Chris Whitehouse and Kieran Mckay, possibly New Zealand’s most ambitious caver. They called themselves The Extreme Caving Team (an audacious title they hoped would attract media attention and funding) and since then have been consistently extending the map of New Zealand’s subterranean world.

Their objectives are clear—to go deep, and to go where no one has been before.

But if this sounds a little heroic, the team would be the first to admit that the reality is very different. Caving is not a normal sport. Five people live together in a five-by-ten-metre space without a toilet, without light, at an ambient temperature of just 5ºC—one degree warmer than a refrigerator. When they get wet, the ambient humidity (usually greater than 90 per cent) means it takes hours to dry again. As supplies are hauled in from the outside over waterfalls and through squeezes, food is rationed to the bare essentials. All waste is carried out the same way .

For cavers exploring new systems, every day underground is spent searching through boulders and rockfalls and freezing meltwater and organ-crushingly tight squeezes for connecting passages that may not be there at all—all the while subject to flash flooding, risk of perilous falls, injuries and cave-ins—for weeks straight, with no outside contact. The demands are extreme, the prospects of success witheringly slim.

The team set their sights first on the Arthur Range and mounted a series of expeditions looking for new systems. In December 2010, Mckay scoured the upper flanks of the Arthur Range, plumbing dozens of shafts that ran out to nothing and finding other more promising leads. He was joined by the rest  of the team in January, and, while sheltering from a storm, Gillespie and Silverwood discovered a massive shaft they called Stormy Pot. (The team returned to their camp to find a tent had blown off the mountain in 100 km/h gales, and their remaining tent flipped over during the night, with all of them in it.)

Over the next few weeks, the team found an alternative entrance to the system, and abseiled down a series of shafts and pitches deeper than Auckland’s Sky Tower is tall to a streamway which bottlenecked in a choking 350-millimetrewide slot they called the Gates of Troy.

For Silverwood, especially, squeezes pose a psychological challenge. “I’m quite claustrophobic,” he says. “When I was 16, I got stuck in a squeeze in a cave on the West Coast for about 8 hours. You feel a rising panic, you can’t move, and it gets worse and worse.”

A fit adult should be able to squeeze through a gap little bigger than the width of this page. “You have to relax your body totally, relax your muscles,” says Silverwood. “Every squeeze has a formula. At the Gates of Troy you have to go in head first, on your left side, with your right arm forward and left arm back. You have to breathe out when you go through and really push with your feet to force your body through. There’s huge pressure on your sternum.”

Like a key in a lock, there’s no other way through the slot, and even when practised correctly, cavers can get stuck, as happened to Silverwood, who had to be dragged out backwards.

The team pressed on from the Gates of Troy through another 400 metres of the streamway, only to be thwarted by a series of waterfalls 10 metres high. They wistfully named the spot Summer’s End.

In March, they were back again, with extra gear and renewed determination. They successfully abseiled over the falls and discovered a dry, comfortable campsite deep in the belly of the system. They named it the Chocolate Room. Well above the water table, it was everything a caver looks for in a camp: elevated, dry, level and about as commodious as a space can be hundreds of metres  underground. This ten-by-five-metre chamber would become the base camp for all future expeditions in the system, their sanctuary from the relentless cold, and the location of every great debate the team would endure over What To Do Next.

Beyond the camp was the main streamway of the system, a deep, searingly cold 1200-metre gauntlet of water called the River of Clowns, after Silverwood forded it wearing nothing but a helmet and a pair of gumboots.

The team returned in June, making their way back down the shaft, through the squeeze, over the waterfalls, through the streamway to their campsite, where they scoured the new  system for leads for a week. Late in the expedition, after 14 continuous hours of scrambling over boulders, along muddy passages and up a complex series of climbs, they found themselves in comprehensive darkness. Their headtorches were not strong enough to light the walls, nor the roof.

The chamber was 50 metres wide and some 80 metres high— bigger than the nave of Notre Dame and twice as high. It was the terminus of the Stormy Pot system, and the five men who discovered it that day are the only five who have ever set foot there. For the caving community, this was the discovery of the decade.

That a system of this extent could exist cheek-by-jowl with Nettlebed, the centre of New Zealand’s exploratory caving activity, was exciting and unexpected. But it was only after drawing up the surveys from that trip and laying them upon a topographical map that Mckay realised the full significance of the discovery.

The team had scrambled and climbed, writhed and waded through passages and streamways clean through Mt Arthur to finish within a stone’s throw of the enormous Nettlebed system. The two surveys terminated within 80 metres of each other.

“Surely,” Mckay thought, “there’s a connection.”

[Chapter Break]

It was an  idea that churned in Mckay’s mind for months. He could think of little else but the possibility of connecting Stormy Pot, with its entrance near the summit of Mt Arthur, to Nettlebed, which flows out of the Pearse Resurgence 1200 metres below. A connection between the two would establish the system as the deepest cave in New Zealand, the second-deepest through-trip in the world (second only to Lamprechtsofen in Austria), and the deepest through-trip in the southern hemisphere. At 38 kilometres, it would be the second-longest cave in New Zealand.

These assumptions, however, were based on Mckay’s cave survey, conducted from the entrance using tape measures, a compass bearing, an inclinometer—a system of dead reckoning which is sketched as a line plot as the caver proceeds. When dealing with complex caves, and particularly caves as long as 15 kilometres, even a two per cent measurement error can result in an enormous margin at the terminus of the cave. When searching for a connection between two systems surveyed in this way, the errors can compound.

But Mckay remains adamant that the two systems are aligned, for he’s seen it written on the walls.

Caves form along geological structures: joints, faults, watercourses. Nettlebed and Stormy Pot are separated by a fault line which is clearly visible in the interior of the cave—one side has moved to the left, the other to the right—and in between, the ancient hydrological system that joined the two is a mangled pile of boulders perhaps 80 metres wide. These features, visible at the ends of both systems, appear to line up neatly with Mckay’s survey.

But does he start the search by digging in this pile of rocks for six hours, or that pile? And how far to dig? The connection might lie under one rock, or 80 metres of rock. It might be a remnant of the original hydrological system, or a new passage formed after the faulting that split the mountain, scoured by water trickling down weaknesses in the fault, enlarging the passage slowly, relentlessly, just as it formed the rest of the labyrinth beneath Mt Arthur.

Around the fringes of the fault there are passages, some metres wide, others mere shafts that taper and bottleneck. Some of these are squeezes through which cavers can contort, penetrate, pour their bodies from one space to another through a gap about the size of a helmet.

In the many highways and byways of a labyrinth, the connection could be around the next corner, and the difference between success and failure the choice between a left turn or a right.

“Everyone knows that the highest mountain in the country is Mt Cook,” says Silverwood. “But no one knows, or will ever know, what our deepest cave is.”

The possibilities are almost endless, and the never-ending discussions debating the pros and cons of each strategy become the long narrative of nights in the Chocolate Room, inevitably after a crushing defeat as a lead ran out to rubble. These can be particularly dark nights in the cave.

“There are times when I ask myself, ‘Why am I doing this?’” says Mckay. “At this time of year, other people are out in their boats, fishing in the sun. And I’m chasing the wind, under 650 metres of rock, dark and wet and frustrated.

“But only a handful of people have ever walked into these caves. They’ve never been explored before. They have never seen light, and we’re lighting them up for the first time. It’s an amazing thing to be the first person to put your footprint in a  place. It makes you feel like Shackleton, or Hillary, or any of those great explorers who went where no one has been before and few have been since.”

In April 2010, the same team made a historic connection from a passage high in the Ellis Basin known as EK3010. After ten days descending through 430 metres of the cave—and on the same day as they had run out of food, stove gas and rope— Mckay, Gillespie and Watson squeezed through a crack and into a chamber. There were footprints on the ground.

“I looked around and recognised the chamber instantly,” recalls Mckay. “We were standing in the Ellis Basin system. We had made the connection that the caving community had  been seeking for 40 years. It was a pretty special feeling.”

The connection created New Zealand’s first 1000-metredeep cave, and put this country on the map of international cavers.

“Going underground doesn’t compete well with kayaking in the sun,” says Mckay. “People want things now, they’re not used to working hard for things. These achievements stay with you for months. It’s not like the temporary thrill of a bungy jump. I’ve spent more than 40 days underground looking for this connection, just about all of my life’s savings. It’s not like there’s a pot of gold down there or something. We do it because we want to keep exploring.”

[Chapter Break]

Caving is not without risk, and climbers decide for themselves the level of risk they’re willing to accommodate. Every anchor, every pitch, every weather forecast analysed is an assessment of risk and reward.

In 1998, deep in Bulmer Cavern, New Zealand’s longest cave system, Mckay himself faced just such a decision. He was running short of rope, and rearranged his rock anchor so that it was around just one rock. The rock was the size of a small table, with no obvious cracks. It looked solid, and Mckay pushed off with due confidence. But as his weight came on to the rope, he felt the sensation of his body tilting backward, out over a 15-metre-deep hole in the pure and certain grip of gravity.

[sidebar-1]

Flailing, he managed to grab at a second rope hanging in the shaft, but the table-sized rock that had come away above him struck his helmet and shoulder, knocking out his light, and he was falling again, in blackness, in silence but for the wind whistling in his ears.

He bounced off an outcrop and landed at the foot of the 15-metre shaft. In a sequence of barely conscious movements.

Mckay stood up, turned on his light, took his abseil rack off the rope and stumbled away from the bottom of the pitch to collapse and slip into unconsciousness. He had broken his arm, his leg could barely move and he was losing blood through gashes on his jaw. He was 500 metres underground, and between him and the entrance was six kilometres of pitches, squeezes, climbs, narrow streams and boulders.

His two colleagues stemmed the flow of blood, but over the next four days he would require the assistance of a further 80 cavers, including Silverwood, to get out of the cavern. (The doctor who assisted was himself rescued by Mckay from a cave in Takaka three years later.)

As bad as the Bulmer fall was, Mckay is aware that a similar accident in Stormy Pot might not have such a successful outcome.

“The hazards are real. We’ll be a long way from the entrance of a cave that only five people in the world know. Stormy Pot is a much harder, much deeper cave than Bulmer, and it would take many more than 80 people to mount a rescue if something went wrong. It’s a big thing, and it’s in the back of my mind all the time. We all want to come out in one piece.”

This experience, and others, changed Mckay dramatically. He began to emphasise the requirements and safety of the whole of the team over just the goals of the expedition.

Silverwood, who admits to being cautious, has witnessed a  number of caving accidents. In 1994, he was caving in a group down a streamway in Mangawhitikau cave, Waitomo. One of the group, Carey Philips, jumped off a small waterfall and got recirculated—sucked down by the force of the waterfall, then pulled back upstream into the torrent, over and over again. Silverwood and others had to drag his lifeless body out of the cave. On another occasion, Silverwood witnessed a colleague fall 37 metres down a shaft when a rope failed. The caver survived.

Both experiences have made Silverwood acutely aware of the need to manage risk underground. But even within the team the appetite for risk can vary dramatically, and there is constant tension between the goals of the expedition and the safety of the climbers. Often, it’s Silverwood and Mckay, the most experienced cavers in the group, who have the most firmly held opinions.

“There are different attitudes to risk. I like a little risk, Kieran likes a lot,” says Silverwood. Though, on other occasions, it has been Mckay who has objected to a course of action he feels too risky, while Silverwood has wanted to press on. In an environment with so many variables and hazards, there are many ways to interpret the same circumstances, and all get hauled out in the long exchanges afforded underground.

“Sometimes it’s like being married,” jokes Silverwood. “Everybody brings something to the team, and we wouldn’t be here if it wasn’t for Kieran. His drive and his enthusiasm are infectious.”

[Chapter Break]

In January 2012, the team returned to Stormy Pot and pushed hard to find connecting passages.

“I guess we thought we were just going to go back into Stormy Pot and walk through into Nettlebed. But we were wrong,” says Mckay. “Everywhere we looked there were piles of rocks. It was really disheartening.”

Stormy Pot is a cave that can break a man. It’s unusually deep. It can take ten hours just to get to the base camp.

“Each cave has a personality,” says Mckay. “Some caves are friendly, fun, dry, easy.

But in Stormy Pot, it feels like you’re fighting the cave all the time. It’s really hard going, and after being underground for a while, you get the impression that the cave is out to get you.”

Nevertheless, the team climbed into the Nettlebed system in May and June that year, to attack the problem from a different angle. They were acting on a rumour three decades old that a caver, Stu Ulrich, had found a hidden passage from the great chamber of Goodbye Yellow Brick Road, in the Nettlebed system, heading in the direction of Stormy Pot—though that system was not known at the time. Could the connection be found from the Nettlebed side?

Mckay’s team searched for days, in vain. Later, they realised their mistake. They had been looking for a passage forking off Goodbye Yellow Brick Road, as a street stems from an avenue. But caves are three-dimensional labyrinths; water travels vertically just as it travels horizontally along weaknesses in the rock. They had forgotten to look down; the passage lay under their feet, through the cave floor.

It made sense. Stormy Pot’s massive terminal chamber, dubbed Neverland, lay underneath the Goodbye Yellow Brick Road, not alongside it. And while the floors of the caves were separated by some 80 metres, the ceiling of Neverland was more than 50 metres high, perhaps substantially higher in the peaks of the rock vaulting. The real distance between the caves could be a pile of rockfall, or perhaps the chambers were  connected by an oculus, through the floor of one, to the roof of another. The idea was tantilisingly simple. Mckay might have walked over the connection already.

[Chapter Break]

Mckay, Silverwood, Whitehouse, Watson and Gillespie returned to the Marble Mountains in January this year, this time with the support of New Zealand Geographic. They had Ulrich’s passage to investigate in Nettlebed, and a promising lead in the Stormy Pot system. Between the Chocolate Room campsite and the mighty vault of the Neverland chamber lies another great cavern, the Black Dwarf, with a slender branch named Coralation Street. The passage is festooned for at least half its length in fine cave coral—lichen-like structures of pure calcium carbonate growing upward, like hoar frost.

Coralation Street heads off in the direction of the Nettlebed system and, auspiciously, a gentle breeze blows down the passage.

Breeze is a good sign for cavers seeking a connection. From the top of Mt Arthur air sinks, travelling through the cave system and out the bottom. So cavers simply need to follow these draughts from the top entrance of Stormy Pot to the lower entrance of Nettlebed. But it is harder than it sounds. Mckay tracks the wind, marking its direction on his cave map as he goes, following zephyrs that are the passage of air from one system to the next, all the while keeping a weather eye on his barometer, because in a storm cycle the draught will change direction. The lower pressure that results from an incoming storm on the westerly side of the mountain will draw air back up the system. The change in direction is noticed immediately by the cavers, a sure sign of approaching bad weather on the surface, and a risk of sudden flooding in the cave

Water is both friend and foe to the cave team. It created the caves and it points the way forever downhill towards Nettlebed. But water can also entrap, and soak, and drown.

In Stormy Pot’s main streamway there is a 400-millimetre gap between the normal water level and the cave roof—plenty of space for cavers, but a small margin of comfort as water tumbles into the system from a downpour above. Should a caver find himself in the River of Clowns during a storm, he would likely never return. Should he be deeper in the cave, he will be  trapped until the storm abates and the water subsides. This is why the exploration of Stormy Pot has been attempted only in high summer, though this is no guarantee of fine weather.

On this expedition, a front moving in from the Tasman Sea dumped 100 millimetres of rain in 24 hours.

“We could hear vibrations in the streamway—woompf, woompf, woompf—as water poured down the avens, setting waves against air pockets in the cave ceiling,” recalls Silverwood. “It was really eerie.”

Pretty soon, the team were trapped underground.

“Many places were flooded up to the roof,” says Mckay. They were safe enough, but he was nursing a chest infection, and in the cold and humidity of the cave it slowly wore away at his zeal. The inclement weather on the surface plagued them, too. The system flooded twice, trapping the cavers each time and frustrating efforts to push forward.

Yet, near the end of the expedition, the team woke to another dark morning in the Chocolate Room, and made their way through the River of Clowns to Black Dwarf. Travelling down Coralation Street they found a new lead heading towards Nettlebed. From a tight squeeze they broke through into a chamber, then a series of chambers connected by tight passages.

“We thought, this is it, we’re just walking into Nettlebed,” says Mckay. With only 80 metres of distance to close on the survey, they were entitled to be encouraged.

The passage turned into a tube that they slithered through, feet first, into a narrow stream, running through a slot canyon barely a body width across. Wind was piping through it.

Here was a choice: left or right, upstream or downstream, success or failure.

The team headed upstream, following the wind, which logic dictates should have been blowing down from the top of Mt Arthur through Stormy Pot and into the Nettlebed system. They squeezed through the tight passage metre by metre, like teasing bread from a toaster, until they were stopped at the base of a sixmetre- high waterfall. It was impassable. It was 10pm, they had been caving for 14 hours, and still had to return to camp. There was no time to investigate the downstream option.

Back in the Chocolate Room, they debated their options.

They had been underground for a week, and the long-rage forecast they had received before entering the cave had  indicated that the weather on the surface would take a turn for the worse any day now. The barometer was low and oscillating, and the unwritten protocol was that the team wouldn’t enter the streamway if the pressure was falling, to avoid being trapped. (There was a stash of food and essentials cached in the Black Dwarf chamber for such an eventuality.)

Had communication with the surface been possible, they would have known that the forecast had changed, the weather would improve, and they could head back down Coralation Street again. But the team had only the wavering needle of the barometer to inform them.

“We were shattered,” says Mckay. Silverwood, Watson and Whitehouse could have pressed on, but Mckay’s chest infection had all but broken him.

“We got beaten. We got beaten by floods, we got beaten by sickness, we got beaten by everything.”

It was only days later that the obvious occurred to Mckay— they should have turned downstream at the juncture on Coralation Street. There is a similar passage in Nettlebed heading towards Stormy Pot with the water running down it.

“All we need do now,” enthuses Mckay, with a resurgence of trademark optimism, “is return to Nettlebed’s top entrance, drop in, and travel up that stream against the flow of water.” Soon enough, he reckons, they would cross their earlier path, walk up Coralation Street and out of Stormy Pot, the first cavers to traverse the deepest through-cave in New Zealand.

“I’m more confident than I’ve ever been. The distance between the systems is about 80 metres, and we’ve just discovered 200 metres of new passage. Once we make the connection we can realign both maps, reduce the error, and potentially find lots of other connection points.

“This should be open and closed. There’s a connection here, there just has to be.”

Silverwood is more guarded. “I think there’s less than a 50 per cent chance of finding a connection from Nettlebed,” he says, promoting the sort of circumspection that must go down like a lead balloon in the Chocolate Room.

Even Mckay admits that the streamway in Nettlebed is “very tight”. “It’s metre by metre, very hard going. But now that we’ve found a similar streamway in Stormy Pot, the connection is possible.

“It’s not a classy achievement. It’s not out there in the public eye. We’re crawling around under the ground. But finding a connection will be like discovering an unclimbed mountain peak at the end of the road, that no one knew was there.”

POST PRESS: In February 2014 Mckay and the team made a last-ditch attempt to find a connection. They followed a passage of breeze from Nettlebed, sorted through five metres of rock, found a crack about two inches wide and hammered it out for an hour or so. Then Mckay breathed out hard, and squeezed through to find himself right next to the team’s last camp in the Stormy Pot system. They had been so close for so long. In fact, Mckay had emerged at the bottom of a shaft they had been using as a latrine. It was cruel and perfect irony. The following pictures are from this last expedition: