The dock at Mountain Lake is everything a dock should be—whitewashed clapboard, punctuated by an airy pavilion with a red roof—but if you jumped off it, all you’d hit is earth. There is no water here. No puddles even. Just soil, sandstone, milkweed, sassafras, and the occasional pine sapling. Skirting the periphery of the lakebed, a belt of rhododendron holds back the woods. Jutting into a meadow as it does, the dock resembles a hitchhiker’s thumb. Well, I’m not needed here anymore. I might as well move on. 

Mountain Lake is the only natural lake in the southern Appalachians, the highest natural lake east of the Mississippi (3,874 feet above sea level), and one of only two natural lakes in Virginia. (Lake Drummond in the Dismal Swamp is the other.) It is a geologic anomaly: it comes and goes, drains and fills. In 1751, a British frontiersman went looking for what locals described as a mystical lake that periodically disappeared and reappeared. He found it and noted “a lake or pond about 3/4 of a mile long . . . and 1/4 of a mile wide,” the lake’s first mention in the written record. A few years later it disappeared. A few decades after that it returned. Core samples indicate it has drained completely only four times over the past six thousand years, most recently in 2008; exactly how and why captivates scientists. They say it’s the only lake like it in the world. Its most recent vanishing act has also frustrated property owners, puzzled engineers, and disappointed fans of the particular movie that made this lake famous. 

Thirty-two years ago, the lake was so large and pristine that Hollywood came calling and filmed a scene that would go down in cinematic history as one of the greatest wet t-shirt moments of all time. In Dirty Dancing, Johnny Castle and Baby Houseman took to this lake to practice their famous lift against the backdrop of an Edenic forest. And it wasn’t just the lake that the movie showcased. An imposing stone lodge, built in the 1930s, is situated on the bank to maximize what were once spectacular water views. Watch the first twenty minutes of the film: The Mountain Lake Lodge plays Kellerman’s Resort in the Catskills. 

When the lake drained completely—and unexpectedly—ten years ago, there were many unknowns, but one thing was clear: water was escaping through a constellation of holes in the rock at the far end of the lakebed. The holes were sucking water into the earth like a body gasping for air. With no lake at Mountain Lake, business at the resort tanked. A contractor came in and pushed dirt around to try to fill the holes. This was both controversial and pointless: They’d fill a hole, the lake would find another. The project failed. Management, scientists, neighbors—nobody was happy. Nobody is happy. 

The lake is still dry. The only remaining water is a small stream that trickles through the lakebed. It has collected at the far end, a half mile distant from the lodge, in a muddy pit.

Nestled in the saddle of Salt Pond Mountain, Mountain Lake once offered vacationers a summer respite from Virginia’s brutal heat and humidity. The hotel has a hundred units (forty-three rooms in the main lodge), which were routinely at capacity during the high season. Even off-season it wasn’t uncommon to have a comfortably full lodge. Today, sepia-toned photographs of these better times line the halls so that a visitor can see it as it was then: a site for boating expeditions, girls’ camps, picnics, fishing, and tête-à-têtes on the gazebo at the water’s edge. Looking at the pictures, you can practically hear footsteps ascending the dock, oars hitting water, the pages of beach reads turning. The lake was the main event, everything else a sideshow. The photographs also speak to the duration of these traditions, populated as they are by women in long white dresses and men in one-piece swimsuits that cover their torsos like undershirts. The faces in these photographs, too, are white, evidence of a social segregation unacknowledged in the book sold at the front desk, Remembering Mountain Lake, that almost instructs you upon check-in to channel history. Or at least parts of it. There are more recent color photographs and paintings of the lake as well. These speak to how quickly the seemingly immutable can change, even vanish. 

Shortly before I first saw Mountain Lake, I had begun thinking about water—water as an idea—on a flight from my home in Virginia across the country to California. On a cloudless day from my window seat, the sight of the parched West caught me off guard. I’d been hearing about the region’s water scarcity on the news, but then, somewhere over Arizona, in the midst of a vast expanse of the tans and browns of dust and dry dirt, I saw, far below me, a green triangle. It was unmistakable: a baseball diamond in the desert. Could there be, I wondered, anything more American? Back east, we never seemed to pay much attention to water. 

I am an intensely visual person, and the baseball diamond sparked in me a particular fascination with water-related images. When I heard about Mountain Lake, I knew I had to see it. When I finally did, it was the image of the dock to nowhere jutting into the dry lakebed that I found arresting. In the image of a lake filling and refilling, I saw metaphors. In the lake’s absence, I saw a snapshot of a drought-plagued future. In its cycles, I saw dangerous confidence in what is scarce always returning eventually. These metaphors now register as facile. Water, after all, is not an idea. Water is not a metaphor. 

A scientist’s natural wonder can be a hotelier’s natural nightmare. Betty Massey helms the Mary Moody Northern Endowment, which oversees the Mountain Lake resort. She lives in Galveston, Texas, and from there negotiates a balance between a for-profit enterprise and a conservancy that encompasses the property around Mountain Lake. In 2008, the country was in the midst of the worst financial crisis since the Great Depression. Then, in September, Hurricane Ike blew into Galveston without mercy. Massey lost her house. The following month, Mountain Lake was swallowed into the earth. 

Starting in 2005, lake levels had been erratic but, given how rarely full drainage had occurred in the past six millennia, no one thought it would happen again anytime soon. And then it did. Over the next few years, losses were measured in millions, whether in dollars or gallons of lake water. 

Massey, together with new on-site management of the lodge and advised by a board, decided to undertake a “fruit basket” approach: a hodgepodge of rebranding and an effort to engineer an end to the water drainage. “It’s hard to be Mountain Lake when you have no lake,” Massey said. The problem was that water was gushing through the holes at a rate exceeding water entering elsewhere via precipitation, groundwater, and runoff; thus if the holes could be successfully plugged, the lake might fill again. They hired Skip Watts and George Stephenson, geologists from nearby Radford University who had been studying the lake for decades, to investigate the likelihood of coaxing Mother Nature along. Their challenge was to reverse this balance without disrupting the natural hydrology developed over thousands of years. Watts and Stephenson knew that eventually sediment would plug the holes, like hairballs in drains, and the water would start to rise again, but this would take longer than Massey and the resort’s management felt they had the luxury to wait. 

Among the possible options laid out in their subsequent report was gradual redistribution of natural sediment around the lakebed to reduce the leaks. Massey and the Mountain Lake management decided to try it. A contractor came in and bulldozing began. While Watts and Stephenson had requested to be kept apprised of the project’s progress, their specific contract was up; as the contractor moved in, the geologists were cut out of the process. No other scientists were present to oversee implementation of the project. The contractors used some non-native materials; they scraped rock off the sides to pack the bottom, reshaping the organic lakebed. Making matters worse, the project was delayed waiting for the permits, so, instead of starting when it was warm and dry, it began as rain and snow washed away sediment. The contractors pumped that sediment out of the lake and into the watershed, upsetting neighbors as far as a mile and a half downstream where the mud plume met the Little Stoney. Researchers found that the pH of the lake reached levels unsustainable for animal life. Lake levels rose temporarily, but within months the water found other holes and went on its merry way. The project was a bust.

This is how I found myself at the National Air and Space Museum in Washington, D.C., killing time wandering through an exhibit called “Looking at Earth” devoted to aerial photography. What do we see when we look down at ourselves? The revelations are general—weather patterns, mountain ranges, eroding shorelines—and specific: the hole in the ozone that opens annually over Antarctica; smoke from the California Thomas Fire. Airplanes, rockets, and satellites hang suspended from the ceiling in the main atrium, but these images inspired in me fewer thoughts about the glories of human ingenuity than its potential costs. 

I was there to talk with biologist Jon Cawley and geologist Ross Irwin, who work as researchers at the museum and know Mountain Lake well. We convened by the Star Trek Enterprise studio model. Cawley, in his mid-fifties, is a bit ruddy, a bit rumpled, and emotive, whereas Irwin, in his mid-forties, is fair skinned, starched, and circumspect. Together, we took the elevator to Irwin’s office, where the hallway wall was filled from floor to ceiling with thick black binders containing original photographs from the Apollo missions. Opposite, dominating the far wall, a window looked out to the National Mall. Irwin positioned his computer screen against this scene and pulled up Mountain Lake on Google Earth. The landscape framed against the cityscape was a striking juxtaposition. 

“There’s a reason why there’s nothing else like it,” Irwin said. “It’s a very unusual setting, a very unusual set of circumstances, and so it makes it a hard problem. . . . You can’t just open up a book and say, ‘This type of thing is all over the place. Here is your textbook model for how it works.’” 

He was explaining why, in a new paper he and Cawley co-authored, they argue against an accepted theory that the lake was formed by either one large landslide or a series of smaller ones. Landslides were part of it, they theorize, but did not tell the whole story. They make the case that when the lake drains, it’s because the water slips through a narrow, yet deep gorge lying beneath the bedrock like a crack in a floor. In other words, the pressure of the water on the sediment that has accrued in the gorge sufficiently to have created a plug (of sorts) becomes so great that the water breaks through that barrier and washes down into the earth. Moreover, they say, the conventional bathtub analogy of drainage at the lake is erroneous because the sides of the lake are less like porcelain than cheesecloth; the water loss has more than one cause. Irwin’s point was that if you don’t understand the complexities of a natural system, you can’t responsibly tamper with its hydrology. Had Mountain Lake’s cycle been left to its own geologic schedule, things today might be different, he said. While wary of judging too harshly, they’re both concerned that the hole-plugging project was handled rashly and that there could be long-term consequences. They are worried that the breathing lake may have stopped breathing. 

“We’re in a country that has property rights, where you can do to your land not what science suggests should be done but what your business decision is, and there are a lot of things that go into that decision,” Cawley said. As Cawley shot me meaningful looks for emphasis, Irwin sat poker-faced beneath a giant rainbow-colored photograph of the surface of Mars, then leaned forward and pointed to the far end of the lake where the holes are clustered, returning purposefully to the science. 

“I think the particular approach that they took was unlikely to work and could even be counterproductive,” he said. “It may actually make the basin less likely to ever fill to full again. . . . My feeling is that if this lake fills up again it will not be because of what they did.” 

If. That is the crux of it, whether the natural cycles have been so disturbed as to prevent refilling—at least in our lifespans. Geologic time is the final arbiter. This corner of southwestern Virginia sits atop one of the most active seismic regions of the eastern United States, and though it may take centuries, the Earth will shake things into the shape it wants them.

While the contractor tried plugging the holes, the resort began marketing its other attributes—the views, the walking trails. They renovated buildings and built a ropes course. They reframed the lakebed as a boon, a once-in-a-lifetime opportunity. Business has improved, and it got a bump in 2017 with the thirtieth anniversary of Dirty Dancing. Massey also pointed out that it would have been more financially viable to tear down the original buildings and rebuild, but they preserved the original structures. The resort’s stone lodge is especially magnificent, a relic of an earlier era of American vacations. The shells of the grand old camps this era produced are scattered over the hills of the Catskills and the Poconos, where people came to convalesce from tuberculosis and to escape the petri dishes of urban centers, or for honeymoons or just some fresh mountain air. 

Mountain Lake is the kind of place that still hosts regular Fish Fridays, for which locals get dressed up and trek up the mountain. It’s the kind of place that might sprinkle bacon bits on salad and still call it vegetarian. Cell phone service is poor to nonexistent. There are no televisions in the rooms. Internet is extra. 

Some evenings bartender Mike Richardson plays Dirty Dancing on the television above the hard liquor. He knows every line and is prone to recitation five seconds ahead of the actors. I don’t think that you ever had any intention of telling him. Ever, says Richardson, mimicking Patrick Swayze as Johnny Castle. He also shouts out the bloopers. If you inquire, he’ll pull a full script from under the bar and pass it around. He says the shower curtain in Swayze’s room has been stolen thirty times and that the locals who were extras were supposed to look Jewish or Italian, a comment which made me pause and wonder whether I should mention my own Jewish ancestry or ask if he could see it in my face. He says tourists ask why they don’t fix the lake, that they say, Bring it back! They suggest pouring concrete and drilling holes. He says he tells them they ain’t got Hilton money.

Dirty Dancing is a movie about a lot of things, abortion and segregation among them. It is also about people from different classes converging around—even in—a lake. During the 2016 election, Giles County went seventy-two percent for Donald Trump. It is also ninety-six percent white. The sight of the empty lakebed in this political climate struck me as another potential—and again facile—metaphor. Economically, Giles County, Virginia, has fared better than some counties in the southwestern pocket of the state, but that is only because it has had moderate success promoting tourism. Mountain Lake is (was? will be?) the acknowledged scenic gem of the region. It is hard sometimes to know what to save: today’s jobs or tomorrow’s land. The either-or is a cliché, but it’s one that appears to have gotten Mountain Lake tangled up in its own priorities. Like Dirty Dancing, Mountain Lake itself raises larger questions, among them issues of ownership. While there are the competing claims for the best course of action to safeguard the lake’s cycles, there are also competing claims on the Dirty Dancing history. Some scenes—notably the staircase scene—were filmed at North Carolina’s Lake Lure, which also capitalizes on that association. 

Despite Ross Irwin’s assertions of there being no textbook examples to explain Mountain Lake, “the very first exercise in the hydrogeology [textbook] is exactly the same thing,” Skip Watts told me. “It’s trying to evaluate what happens if you increase or decrease any part of the big budget.” 

Water budgets are the algebra of water coming into versus leaving a lake, and the equilibriums that result. At Mountain Lake, he said, it’s this math that’s wrong. Input has slowed. Moreover, the slowing coincided with the construction of a conference center in 2001, which diverted surface water that otherwise would have supplied the lake. 

Watts and his colleague, George Stephenson, the Radford engineering geologists, study the impact of the manmade world on the natural one, and assess risks. They also resent the guff they have taken for the Mountain Lake project. They say they don’t take sides. If a project—lake filling, dam building, highway construction, mining, drilling—is going forward, their charge is to provide the information to do the job as safely as possible. Their recommendations may or may not be taken. That is not their prerogative. But neither of them was happy with the bulldozers, either. In proposing sediment redistribution, they did not mean willy-nilly. 

Americans engineer everything. Water, especially. We straighten its natural paths. We divert it. We pump it. We levee it and build breakwaters. We dig canals, drain wetlands. We build dams—more than eighty thousand of them in the United States to date. The question of whether or not a landscape is “natural” is contrived to begin with. Nevertheless, it is fraught. If Mountain Lake’s idiosyncratic cycles have been engineered out of existence, does that now change the only “natural” lake in the southern Appalachians officially to “man made.” It raises a broader question of what we lose when humans don’t want to—or cannot—wait, when our schedules do not align with nature’s. 

Massey maintains that the lake is entirely natural, a state that cannot be taken away from it. Bruce Parker, a professor emeritus at Virginia Tech who studied the lake, practically spit out his coffee when I asked him his opinion. For him there’s no debate: The lake is ruined. The breathing lake no longer breathes. “It’s a man-made pond,” he said. 

Cawley and Irwin’s assessments of Mountain Lake’s current status fall somewhere in the middle of this “natural” to “unnatural” spectrum. 

“How purist do you want to be?” asked Cawley. “I would point out that it’s got some scars that it didn’t have a few years ago and those scars might not have been good management.” 

When I met Watts at Mountain Lake on an unseasonably warm February day, we drove down to a cottage at the far end that now overlooks the giant mud puddle some sixty feet below. This was the resort’s most popular rental back when water came to its threshold, creating a houseboat effect. Now it serves as a de facto research base for Radford and Virginia Tech scientists, which means it’s where they go to charge batteries for their gadgets. When you peek inside, it’s all torn furniture and dust. Watts says there used to be a beautiful aerial photograph of the full lake hanging above the fireplace that he had intended to salvage, but one day it just disappeared. 

“I don’t like conflict,” he said as we hiked the area where the holes are clustered. “I sympathize too much with the people I’m working with. I sympathize with the situations people find themselves in with nature.” 

Hiking up from the lakebed, we headed for a boulder field called the Garden of the Gods which, he said, was incontrovertible evidence of a landslide formation. The boulders were enormous, some flat enough to host small theatrical performances, some shaped like giant refrigerators. Looming above them, two stories up, is a rocky cliff a quarter of a mile long from which these fell sometime long, long ago. Watts explained that these boulders went who-knows-how-deep beneath us, and that through and between them is where the water goes. 

Heading back to the lodge, Watts said, “When I get really, really rich, I’m going to build a whole new cottage here. Oh, and somehow bring back the lake, too.” 

About water budgets, Butch Brodie—an evolutionary biologist who directs the University of Virginia’s Mountain Lake Biological Station, which sits adjacent to the resort property and whose scientists have full access to the land—mentioned something people aren’t talking about: “I don’t think that any of us have looked at water use levels,” he said. 

The UVA Research Station and the resort both pull from the aquifer, and the research station recently had to redraw its well and go much deeper into the mountain, suggesting that the water table is dropping. If both the resort and the research station are taking water from a declining water table, lake levels would be affected. Unlike western states, Virginia has no definitive map of its aquifer, in part because Virginia never thought water scarcity would be a problem. Each time a well is dug in a rural area, pieces to the puzzle are filled in, but in places with low population density, as here, data can be scarce. With mountains, you can usually look and see the drainage pattern: down. In the case of Mountain Lake, however, Brodie said that “it’s just a morass of crevasses and cavities and . . . so we don’t know what the aquifer is like in the mountain.” 

His implication is that if we understood the structure of the aquifer better, it might be a little easier to understand relatively simple things like where we should be taking water from and where we should be dumping it. Regardless, the shifting status of the water table demands we interrogate what we take for granted: the rain, the aquifer, and the lake’s return.

Walking through the Garden of the Gods, Watts had wanted to show me fossils of wormholes that dated back millions of years, but the pickings had been slim because, among their other missteps, the contractors had disposed of some “museum quality” rocks he and his colleagues had wanted to collect and put on display. Thus, the rocks in the walls of the lodge were the best places to find fossils. Sure enough, when we walked the exterior arcade, they were there: evidence in stone of prehistoric wormholes, water currents, soft pebbles. 

“What about fish?” I asked. 

“What kind of fish?” 

“Marine, I guess.” 

“Absolutely. In the quarry.” 

So we went to the quarry just up the hill, and there we turned over stones. I asked what I was looking for, and Watts told me I’d know when I saw it but that sometimes it required a chisel. He was wielding one and demonstrated how it worked, how you could split open a stone and find inside it the ghostly fingerprints of prehistoric life. He then continued to walk around, picking up rocks and splitting them open. It didn’t take long for him to find an example to show me. At first I saw nothing, just dirt, loose gravel, errant roots. He brushed these off. Then they emerged like a Rorschach test: the delicate scalloped edges of bivalves imprinted into sand forever ago. They weren’t fish, but they were unmistakably marine, and that was good enough for me. Bending down to examine the details, I could almost hear an ocean lapping at the shores we stood on.

Enjoy this story? Subscribe to the Oxford American.