Here's the best-kept secret of Yosemite Valley: It has a twin.

This little brother, as the late naturalist John Muir called it, has a thundering waterfall named Wapama, a feathery cascade named Tueeulala and a towering peak called Kolana. Below Kolana, a valley snakes between granite walls for eight miles to reach a staircase of rock known as the Grand Canyon of the Tuolumne.

Yosemite's little brother has a name. It is called Hetch Hetchy, derived from the Indian name for its native meadow grasses. But despite its grandeur and its presence in a park that is a national treasure, few people know Hetch Hetchy exists and few visit it.

There is a reason for this remarkable obscurity. Hetch Hetchy is underwater.

Since 1923, a dam that supplies water to the San Francisco Bay Area has submerged the valley's roughly three square miles. An act of Congress in 1913 gave San Francisco control of the valley, a precious resource that belonged to the entire nation.

No wonder, then, that Hetch Hetchy is today the least visited natural feature in the 1,189-square-mile Yosemite National Park. In one survey of Yosemite's popular sites, Hetch Hetchy finished last, below "other." No other national park has such a centerpiece jewel that is locked away from the public, both by the ranger's key at 9 p.m. every day and by 300 feet of sparkling, clear Sierra water.

Yosemite serves nearly 4 million visitors a year. Someday soon it will run out of room for the public. When that day comes, the choice will be stark: Ration the chance to experience the glories of the Yosemite Valley or create, literally, more valley.

Such an expansion is possible if an idea once considered fanciful, even quixotic, gains legitimacy: Drain Hetch Hetchy - an enlarged hole at the dam's base would do the job - and let nature begin to reclaim this spectacular setting.

That may sound simple, but it isn't. It would require some changes to the Bay Area's water system and a consensus among major holders of Tuolumne River water rights. But if the notion is complicated, it is not out of the realm of the possible and is well worth discussing. An upcoming replumbing of San Francisco's Hetch Hetchy system and a convincing restoration proposal generated by a new computer program at the University of California, Davis, make this an appropriate time for the conversation to begin.

Any debate about piercing the dam at Hetch Hetchy is sure to be heated. Debates about Hetch Hetchy always are.

The debate that led to the construction of the dam embroiled the U.S. Senate for a week. It ended near the stroke of midnight on Dec. 6, 1913, when senators weighed environmental and development values and made their decision. The vote was 43 to 25. The dam in Yosemite would be built. The Hetch Hetchy Valley would be inundated. And San Francisco would have the use of the water.

San Francisco first set its sights on this river for water in 1901. The city's leaders and residents would understandably be nervous and resistant to change today. Water and electricity are still precious commodities. Hetch Hetchy provides nearly 85 percent of the city's water and about a sixth of its electricity. It also supplies a large portion of the water for Alameda, Santa Clara and San Mateo counties.

So any debate over Hetch Hetchy today would involve more players than in 1913 and even more factors to consider, such as climate changes in the Sierra. But a debate today could lead to a new conclusion because the Tuolumne River watershed and the world have changed so much.

Ninety years ago, the senators' collective clairvoyance was spotty. They had no way to anticipate that in 1971 the New Don Pedro Dam, creating a reservoir more than five times the size of Hetch Hetchy's, would be built downstream. They had no way to know that an invention called the computer would reveal to UC Davis researchers that the big downstream dam could do the work that Hetch Hetchy does now. They had no way to know, in other words, that they were making a decision that might someday be undone.

By design, dams are meant to be solid and permanent. Perhaps that is why their engineering so often defines conventional wisdom and the universe of the possible. The structures are seen as unchangeable features of the landscape, by politicians, by engineers and even by newspapers. As recently as 1987, these pages pooh-poohed the idea of draining Hetch Hetchy.

But Hetch Hetchy today is truly an unusual case and Californians can dare to regard the dam in a new way. If they look carefully at water and electricity options, they may just find the dam more expendable than the lost valley below. It is possible to imagine a different future, one that restores the glories of Hetch Hetchy to the public while satisfying the legitimate municipal demands on this river.

As coming editorials will explain, San Francisco doesn't have to lose water for Hetch Hetchy to be reclaimed. But Hetch Hetchy's restoration will involve more than San Francisco's interests. It cannot occur as an isolated political act. There would have to be a water package to address the needs of every interest. The many public purposes of the Tuolumne River - its spectacular Yosemite watershed, the downstream water demands of San Francisco, electricity, Modesto flood control, Turlock agriculture - all are pieces of an intricate puzzle. The upcoming challenge is to fit them together - for the benefit of Californians and, where Yosemite National Park is concerned, for the benefit of all Americans.

In short, Californians don't have to be prisoners of a 90-year-old debate. Change is coming to the river. As part of that evolution, it is no longer unthinkable to imagine reuniting Yosemite's twin valleys. Something magnificent and unexpected could actually happen. A river could be allowed to run free through a glacial valley, just as it did before Congress locked it away nine decades ago.