That the story of the floods is now accepted as fact is due to four decades of cantankerous persistence by a geologist named J Harlen Bretz -- "J" being his first name in its entirety. A Michigan native, Bretz moved to Seattle in 1907 to teach high school. Between classes on history and physiography, he liked to peruse topographical maps, and in 1909, he came across a recently published U.S. Geological Survey map of the Quincy Basin. What he saw perplexed him: large coulees, or dry canyons, some of them as much as five miles wide, without any clear indication of what forces formed them.

In the early 1900s, geology was governed by the principle of Uniformitarianism, which held that the processes observable today -- rivers eroding rock and so on -- were likely the same ones that operated in the past. Thus, dramatic landscapes -- canyons and gorges -- were understood to have been patiently etched into the Earth over eons. But what Bretz saw in eastern Washington wasn't consistent with that principle. No watercourses led to or away from those coulees, and they were the wrong shape to be glacial valleys. Something else seemed to be at work. To find out what, Bretz left teaching in 1913 to get his Ph.D. in geology at the University of Chicago. In 1922, as a member of the geology faculty there, he returned to the Northwest, trundling around the Quincy Basin and its environs in an old Dodge, studying curious landforms: the coulees, the dry cataracts, the hanging valleys, the deep potholes, and, sprinkled liberally around the landscape, the large granite boulders, known as erratics, that had no place among the local basalts left over from the lava flows that had coated the region 17 million years before.

Bretz proposed a daring hypothesis: The Channeled Scablands of the Columbia Plateau had been gouged out by an enormous flood. In two papers published in 1923, he outlined his theory, turning Uniformitarianism on its head. Slow, almost dignified processes did not explain the Channeled Scablands, he argued; the landforms had to be the result of some hitherto inconceivable catastrophe. "It was a debacle which swept the Columbia Plateau," he wrote, with special emphasis. But despite marshaling what he felt was compelling evidence, he failed to persuade his colleagues. Still bruised by their efforts to convince a skeptical public that the Earth was, in fact, much more than 6,000 years old, none of them were willing to cast aside the uniform in favor of Bretz's Catastrophism. To them, the theory reeked of the biblical. Why couldn't more acceptable mechanisms explain those features? they asked. And where did all that water come from in the first place? Bretz's attempts to answer these questions -- perhaps volcanic activity under the ice sheets caused them to melt rapidly, or the glaciers might have, well, done something -- weren't persuasive. His temperament didn't help, either. He was characterized by what he himself called "a recurrent earthiness," and he found the occasional feud bracing. Soft spots in his theories, while irksome, were hardly fatal. The landscape spoke for itself, he declared.

But apparently not loudly enough. On Jan. 12, 1927, Bretz was invited to speak to a gathering of distinguished geologists, many of whom worked for the U.S. Geological Survey, at the Cosmos Club in Washington, D.C. It didn't go well. "They were all loaded for me," Bretz later recalled, "and after letting me speak for two hours they opened fire."

Kleinknecht, 58, is not a geologist by training. He is head of the Social Studies Department at Kamiakin High School in Kennewick. He did take one geology course in college, and he happily admits that he got a D. An eastern Washington native, he first became interested in geology when he moved back to the area from South Carolina in 1979. "It opened my eyes," he says. "There was a lot I'd taken for granted before." He started reading, started hanging out with geologists who had made a hobby of the floods, and in 2002 started the Wenatchee chapter of the Ice Age Floods Institute. He has been the Institute's president for the past five years. "We're an older bunch," he says of his constituents. Interest in the floods has increased over the years, and the organization now has around 700 members. "Now we need two buses for our field trips," he says.

We are at Trinidad, Wash., near the Trading Post and Shell Station, standing on a bluff above the Columbia River and the Crescent Bar Resort. Looming over the resort is a sheer basalt cliff, the top of which, planed flat by the floods, is called Babcock Bench. Below the Bench, boaters and waterskiers cut tracks across the Columbia. From our vantage, we can hear the mosquito whine of motors, and it's hard not to see this frenetic recreation as kind of silly and insignificant against the general attitude of lithic sternness. (Geology sometimes makes me dour.)

Kleinknecht has brought me here to see something else, though -- what counts in geology as a smoking gun. Across the river from the resort, the landscape rumples in a series of short, unremarkable hills. When Kleinknecht was younger, he and his father used to hunt among them and ride motorcycles over them. In his current capacity, he appreciates those hills differently. "They were the crucial piece of evidence," he says.