Jay Quade wears a fragrant sprig of sagebrush in the band of his cowboy hat, replacing it every day while in the desert. He wears leather boots now, but used to do his fieldwork in flip-flops. ("You learn to walk differently," he says; "you place your feet a lot more carefully.") Before that he was a  middle-distance runner at the University of New Mexico, where he once galloped 800 meters -- roughly half a mile -- in 1 minute 49.8 seconds. He spent most of his childhood in Reno; when he was 6, his father, a wealthy securities trader, moved the family there from Los Angeles to start a new career as a self-taught geologist, determined to escape what he called "the pernicious influence of money." Quade hiked avidly, scavenging flecks of gold from streams and abandoned tailings piles. When he was 15, he and his friends found the wreckage of a plane on a remote mountain ridge east of Lake Tahoe.

On a cool May morning, Quade powers effortlessly up the slopes of the Silver Island Range, only occasionally using his hands as he climbs outcrops and slippery gravel chutes. His rock hammer swings at his hip. McGee stays near his side, as do two Ph.D. students who are fortunate to have bodies 25 years younger.

After 30 minutes of scrambling -- and two surprised rattlesnakes -- the group reaches the level platform of a shoreline halfway up the mountain; from a distance, it stood out prominently, but up close it fades into an obscurity of rocks, bushes and sheep dung. "The lake would have been perched here while it was overflowing into the Snake River Basin," says McGee, "just when we think this area was at its wettest."

"It must have been fantastic fishing," says Quade. He gazes across a salt flat that would have been buried under hundreds of feet of water, at a parallel shoreline on another mountain 10 miles away.

The four disappear in different directions in search of promising rocks, the tap-taps of their rock hammers echoing through the still morning air. Just as a fly relies on its proboscis, geologists experience much of the sensory world through their hammers. A rock that has lain for long in the elements will hide beneath a dark rind of oxidation; a whack from a hammer breaks it in half, revealing its layers and its true identity.

The group climbs higher as intriguing rocks are found, GPS readings are taken, and drawings are sketched in notebooks. Gray clouds darken the sky; the slope steepens.

The slanted slabs of the Silver Island Range make for good climbing. The stone grabs hold of your hands as though its cracked, ceramic surface is covered in razor-sharp Velcro.

Now and then you look down at a rock that has sliced open your finger and see its surface jumbled with the broken, cylindrical stems of fossil sea lilies or the scallop-shaped shells of fossil brachiopods. The Hansen Creek dolostone -- the name that geologists give to the rock that these mountains are sculpted from -- coalesced from a shallow ocean floor that teemed with life 450 million years ago. "If you look at this rock microscopically, it's all shells and shit," says Quade. "It's a recycled graveyard. It's passed multiple times through the backside of a worm."

The residue of that past life has dyed the rock in shades of gray. As Quade scrambles up a scree slope, he pauses, cracks a stone with his hammer, and holds the broken half to his nose like a succulent grapefruit. "Pheew!" he exclaims. He hands it over. It exudes the sewer smell of decaying organic matter. "The Hansen Creek dolostone always stinks," he says. "It's full of petroleum."

The much younger algal tufa that clings to the mountain is also chockfull of organic remains; bake the stone in an oven and it darkens from beige to gray like a piece of burning garlic bread. The Silver Island Range has lived many lives, and tracing each stage of each life requires the tenacity of a sleuth.

McGee and Quade occasionally find that the mountain has surreptitiously betrayed the unwritten agreement between geologists and the Earth: the basic principle that the oldest stuff is on the bottom, and each succeeding layer is therefore younger.

On the wall of an overhung grotto, the two men puzzle over an apparent contradiction: A layer of bathtub ring mineral is sandwiched beneath what appear to be older layers of tufa and dolostone. They dig with their rock hammers into the base of the wall, and soon come up with a theory: Layers began to weather and peel away from the wall while it was exposed to open air; the lake then rose and fell multiple times, filling the spaces between the peeling layers with bathtub rings and tufa. The mountain, in other words, has reshuffled its deck of layers midway through the game.

McGee sketches the scene in his notebook and drops some rocks in a sack. Back in Minnesota, he will test the team's interpretation. He'll saw each rock in half and polish it to highlight its wispy layers. He'll grind out samples of each layer using a dental drill and dissolve the powdered rock in acid. He will purify a few thousand rare atoms of uranium and thorium and then measure the age of the layers by counting those atoms, using a machine called a mass spectrometer that accelerates each atom to 450,000 miles per hour and smashes it into a detector.

By measuring the composition of each layer, McGee can see how the lake's salinity changed over time. By comparing amounts of heavy and light oxygen isotopes, he can gauge how the balance between water flowing into the lake and its evaporation shifted whenever the lake had no outflow. He might even be able to sample some of Bonneville's ancient water: Tiny droplets of it are still trapped inside these minerals, in pockets no wider than a few red blood cells. McGee hopes to bore into those pockets, one by one, with a laser.

A new, more detailed version of Bonneville's history will emerge only gradually. Rocks from one site may produce errors from any number of sources -- contamination from groundwater percolating through a cave, for example, or incomplete mixing of Lake Bonneville, leading to different levels of salt or certain oxygen isotopes in different parts of the lake. But do these same analyses at enough places along the lake's hundreds of miles of shoreline, at enough different elevations that the water would have occupied at different times, and the errors will gradually sort themselves out. The relationship of Bonneville's water level to global climate swings will eventually become clear.

Suddenly, the clouds open up. Heavy-bellied raindrops pelt rocks, hats and notebooks. McGee, Quade and the two students scramble down a gravel chute and head for the trucks a mile away.