TO EVERYONES relief, the "conservation war" is over at the National Gallery of Art. Polemics have subsided, procedures have been tightened, and the workers in the lab there, once again at work, have been cleared of any taint of professional malpractive. But the cleaning of old pictures remains, as it has always been, an extremely tricky business.

Consider, for example, a tale of two paintings, one of a Rubens, one a Rembrandt, in the gallery's collection. Both of them were hidden once behind layers of old varnish. With differing results - some encouraging, some worrisome - both have since been cleaned.

Old paintings are not suits of clothes. They are slowly fading, slowly changing, living bits of history, and their cleaning presents problems both subtle and complex. Esthetics are involved, of course - what should the picture look like? - but so are scholarship, detective work, technology and luck.

"Conservation, and particularly the removal of old varnish," says Gallery Director J. Carter Brown, "has always been a no-win situation. You can't even win by doing nothing. And you can't determine dates, or secure attributions, if you can't see the paint.

"Then there is the 'gestalt effect.' When Malraux cleaned the grime from the old buildings in Paris, there was a great hue and cry. When you see a cleaned picture for the first time, it does look very bright. And very different.

"Polemics have no place in conservation. There really is no 'war,' no division into camps. With almost every picture you have to make adjustments. The differences are almost always differences of nuance."

But while the experts at the Gallery are delighted with "The Mill" - the 17th-century Rembrandt landscape they have found beneath its yellowed coatings - they are much less happy with "The Gerbier Family," the 1629 group portrait by Rubens that is there on display.

In England, in the 19the century, both pictures were famous; both were often copied, studies and discussed. The Gerbier portrait was, in 1860, the most expensive Rubens ever sold at auction ("It was put up at 1,000 guineas, and advanced 500 guineas at each bidding, till it reached the enormous amount of 7,500 guineas," The Times reported. "Great applause followed the adjudication.") The departure of "The Mill" 71 years later was regarded by the British as a national misfortune.

The questionable condition in which both paintings reached this country may well have contributed to their 19th-century fame.

"Those were Brown Decades,' remember," observes Brown. "Collectors of the time like old yellowed varnish. Old Masters were admired for their golden glow.'"

But not by everyone, and not all the time. It is curious to consider now the contratictory statements made about these paintings by critic Roger Fry. (Fry had been a star of the Bloomsbury set; he was a lover of Vanessa bell and his biography was written by Virginia Wolf. Kenneth Clark remembers him as "the most bewitching lecturer I have ever heard.") The strong opinions voiced by Fry did not always jibe.

In 1911, he wrote this about "The Mill": "Supremely beautiful as it is even in its present state, one can but surmise what its effect would be if it could be relieved of the disagreeable mess of yellow-tinted varnish with which it has at one time been smeared, presumably to give it an air of fictious mellowness and venerability . . . Whoever the fortunate owner of this should prove to be, one hopes he may have the necessary work of restitution piously performed."

Despite differing opinions on the varnish of "The Mill" - John Walker, Brown's predecessor, prefered to leave it as it was - few qualms were expressed about the cleaning of the Rubens. When purchased by the Gallery in 1971 (with monies from the Andrew W. Mellon Fund) "it was so darkly toned by reddish varnish," Brown recalls, "that no one would have objected to its cleaning." No one except, strangely, the same Roger Fry, who, in 1927, saw it still uncleaned and wrote: "As color it is one of the strongest and most unexpected of Ruben's creations . . . Here some accident of the moment started him off of the idea of a color harmony built entirely on coppery reds and reddish browns, and an example of Rubens' power of color orchestration."

The trouble was, as cleaning proved, those "coppery reds and reddish browns" praised by Fry were not the work of Rubens, but of some 19th-century vanisher. Though varnishes will yellow as they age, most once were water-clear. But the varnish on various passages of the Rubens had been stained - intentionally - a peculiar deep red-brown. That coating furthermore contained artificial ultramarine, a pigment first marketed in 1828 when the Rubens portrait was 200 years old.

It is not surprisingly that, once the colored coating had been carefully removed, the Rubens seemed radically changed. Mrs. Gerbier's skirt, once thought reddish brown, was now seen as a bright, brocaded green.

In 1971, before the Gallery's conservation lab had been established, the Rubens was cleaned by the late Richard Buck of Oberlin, Ohio. His rediscovery of the skirt's bright green might seem to be a victory, but his removal of the varnish also revealed much that does not please the eye.

Look, for instance at the mother's throat and also at the chin of her eldest daughter. Not only aged varnishes, but the old paints beneath them, too, change eerily with time. Some colors fade, some darken; some paints become opaque, others grow transparent. The cold gray underpaint with which Rubens sketched his faces, and which he later covered over, is now visible again so that the eldest daughter looks as if she has not shaved. The white highlight that one sees on the mother's earring has not lost its brightness, but the daughter's identical earring now seems oddly dull.

Rubens was a maste renowned for his flesh tones, but here the grays and reds he used seem often disharmonious. The pointing finger of the baby looks as if it had been dipped in catsup. That velvet curtain, too, is disturbing: Behind the mother's head, it is a rich, dark purple; but it changes to light pink in the upper left-hand corner. Now cleaned of its coating of admittedly distorting, intentionally toned varnish, the picture has become discordant. What now should be done?

Scholars at the Gallery are pondering that question. They refuse to fake it, to change the picture's colors to make it better fit our preconceptions of a Rubens. Museum paintings, after all, are records of art history, not wall decorations. The pink curtain in the corner might have been producted by some studio assistant: to tone it down in order to make it match the style of the master's hand might make the painting lie. Some corrective restoration may be called for in this case; that has not yet been decided. No one now would wish to add red tones to the Rubens, yet no one at the Gallery finds the paintings as it is entirely satisfactory. Sometime in the future it may be taken to the lab; but once you start adjusting, quieting its discords, unifying draperies, and changing surgace varnishes, where do you draw the line?

Conservators' decisions are rarely cut and dried. "Consider the extremes," says Brown. "Should we 'restore' the Venus de Milo, polychrome her hair, give her back her arms?" The appearance of the Rubens has been changed by cleaning for either good or ill, and, perhaps, for both. The green skirt's reappearance and the fact the Rubens' paint can, at last, be seen, may be considered positive results. But both time and abrasion have left marks on the picture. It is not what it once was.

The Rubens had been authenticated, prior to its purchase, by British scholar Michael Jaffe. And it had been sold by Sir Geoffrey Agnew, the well-known London dealer. When they saw the painting following its cleaning they were startled - and upset. This month's ART-news reports that the two men felt "the picture had been ruined." It is believed that the complaints they voiced to Gallery President Paul mellon led him to impose, in June, the conservation moratiorium that was only lifted earlier this month.

In the treatment of Old Masters there are no rigid rules. Each picture presents different problems. The Gallery's conservators did not know what they would find when they began to clean what Roger Fry described as that "disagreeable mess" of varnish from the surface of "The Mill."

Gallery curator Arthur K. Wheelock, who has watched its treatment closely, says he is relieved and pleased by what has been discovered. Though the paint applied by Rubens was uniformly thin (and delicate in its color harmonies and easily abraded), that employed by Rembrandt was painted more thickly. Beneath its varnish, Wheelock says, "it is in amazingly good consition." He also feels that "The Mill" will be more securely attributed to Rembrandt now that its colors and its brushstrokes may, at last, be clearly seen. The evening sky behind the mill is bright with silver light, and new details may be seen in the shadowed foreground of the picture.

Brown admits dissatisfaction with passages of the Rubens. "What should we do with the gray ghosts of his underpainting? Let them leap out at you? Paint them over? Or find some median ground that quiets the disharmony without hiding what is there? My analogy is to literature. Shakespeare scholars may fill their tomes with footnotes, but no one expects Hamlet to read those footnotes on the stage."

The Rubens, despite its flaws, is now on display. "The Mill," still in the lab, is not available to the public or the press. Brown says "The Mill" is "in surprisingly better shape than we expected. I'm pleased with it." No date has been set for its return to exhibition.