"The Mill," the most important and most often questioned Rembrandt landscape in America, has been taken from display at the National Gallery of Art. The last director there, John Walker, knew it was inevitable, but still hoped it wouldn't happen. "The Mill" is being cleaned.

Walker loved it dirty. He knew the moody landscape, cleansed of its patina of old and darkened varnish, would never look the same.

Melancholy, moving "The Mill" is a strange painting. It is both loved and distrusted. It beauty is not questioned, nor is its antiquity, nor its historical importance, but its authorship is in doubt.

Modern Rembrandt scholars are not sure it is by Rembrandt, for they have never really seen its handwriting, its colors, its details of brushwork. The glaze is in their way.

Scholars prefer clarity. "The Mill" relies on mystery. The evening light is fading, the sun's already set. By deepening its shadows, its mood of sublime sadness, that yellowing, translucent shield of old varnish, may have added something to the painting's reputation, J.M.W. Turner, the master landscape painter who revered "The Mill," wrote, in 1808, of "its inestimable gloom."

We know the picture - varnished - changed the art of England.

First praised for its subject (fits picturesque stone rums, the washerwoman in the foreground, its fidelity to Nature), it later was admired for its deep Romantic gloom. Sir Joshua Reynolds studied it as a guide to landscape painting, and John Constable judged it "sufficient to form an epoch in the art." "Probably no single canvas," writes Walker "has so strongly affected English painting."

When the American collector, Peter A.B. Widener, a Philadelphia millionaire, bought it from Lord Lansdowne (for 100,000 pounds) the Fleet Street papers howled.

The Standard of Feb. 28, 1911, mourned the impending sale as "a national misfortune," Sir Charles Holroyd, the director of England's National Gallery, told reporters that the painting was "the first of the great Romantic landscapes - the chief pride of modern European art."

When given its last English showing, it drew huge crowds: 11,000 people viewed it the first day. (Daily attendance at the National Gallery's Tut show here averaged 7,000.)

That the painting was a masterpiece, and that it was by Rembrandt, was not then in doubt.

Its heritage was noble, it had long been owned by princes.The painting is first mentioned - in 1724 - in an inventory of the works of art collected by Philippe, the Duke of Orleans (1674-1723). We don't know if, or when, he bought it. It may have been acquired by his father, also named Philippe, who entered the Palais Royal and assumed his noble title in 1661 when Rembrandt (1606-1669) was still alive.

Gallery curator Arthur K. Wheelock Jr., who has written on "The Mill," its mystique and its history, describes the holdings of the second Duke as "one of the most distinguished princely collections of the 18th cenury." We know the Duke believed his landscape was by Rembrandt; it says so in thee inventory, and great collectors, then as now, did not often purchase pictures without first receiving specialists' advice.

The painting moved to England in 1793 when the third Duke, besieged by creditors and troubled by the French Revolution, sold his works of art en masse. William Smith, a member of Parliament from Norwich, and a friend of Turner's, bought "The Mill" in London for 500 pounds. It had entered the collection of the Marquis of Landsdowne by 1824.

The work may be Rembrandt, but tough-minded scholars today are far from sure. Eighteeth-century documents and 19th-century assumptions will not resolve their doubts. They want to see the painting as it looked whne new.

The more romantic Victorian collectors liked the look of varnish. Old master paintings, they believed, should have the mellow glow of old master violins. Scholars nowadays prefer a harsher light. Before they grant "The Mill" to Rembrandt, they will examine it minutely with microscopes and X rays, they will analyze its materials and compare its details with those of other works by Rembrandt. They want to see it clean.

Using gentle solvents and many tiny swabs (toothpicks tipped with cotton), Kay Silberfeld, the Gallery's conservator of paintings, already has made "windows" in the painting's antique varnish. The sky above the water is no longer golden yellow, but instead a silver gray. Where the woman on the bank sits and washes clothes, the ripples splash bright white. Silberfeld and Wheelock already have discovered that the boat was once complete. Stress patterns in the linen show that the painting, which is now 34 inches high and 41 inches wide, was once at least four inches wider. The right side has been trimmed.

Silberfeld and Wheelock will leave next week for Europe to look at other Rembrandts, study modern X rays, analyze his methods, habits and materials, and the way he painted shadows, distant trees and evening sky. They are proceeding cautiously. They have reached no conclusions. "I am not working on "The Mill" to prove or disapprove authorship," says Silberfeld. "The painting needed work." It has already been relined. The lifting of the varnish may take about a year.

The Gallery, as always, moves slowly in such matters. But doubts about who painted it do hover round "The Mill."

Recent Rembrandt scholars have been stricter than their teachers. Half a century ago, the standard text on Rembrandt, that by Wilhelm Valentiner, listed 714 paintings, "The Complete Edition of the Paintings," which Abraham Bredius published in 1935, cut that number down to just 619. "The Mill" was not among them. Though Bredius, in an article of 1911, had emphatically defended its Rembrandt attribution, he later changed his mind.

Kurt Bauch, a young scholar, was even more restrictive. The volume he published in 1966 gives only 562 paintings to the hand of Rembrandt. And in 1969, when Horst Gerson edited Bredius' "Complete Edition of the Paintings," he cut the canon further. Accepting "only those pictures whose authenticity seems to me beyond all doubt," Gerson listed only 468 Rembrandts. "The Mill" is not among them.

Though other well respected scholars - Jacob Rosenberg, Wolfgang Stechow - still give "The Mill" to Rembrandt, there is no consensus. The issue of its attribution remains unresolved.

"One of the great ironies of Rembrandt literature," writes Wheelock (who will lecture on the painting at the Gallery on Sunday, Sept. 25), "is that this painting, so beloved and so important to the 19-century conception of Rembrandt, has been almost totally ignored by recent 20th-century critics. One suspects that the painting is omitted not only because of the uncertainties of attribution, but also because the rich meanings "The Mill" once held for so many are no longer as compelling or as convincing to contemporary eyes."

We no longer look to Rembrandt for his picturesque effects, his fidelity to nature, or his ability to show us "the inestimable gloom" that Turner so admired. The moody, yellowed painting that changed the course of English landscape art, the darkly glowing picture that John Walker loved to look at, will not be seen again.

"The Mill," once cleansed of its dark varnish may be accepted as a Rembrandt, or as partially by Rembrandt, or is a mid-17-century landscape from some other hand. It is too soon to tell.

"But we have already learned one thing," said Charles Parkhurst, the Gallery's deputy director. "The painting is not brown."