What's in a wine bottle and what's on the label? Not always the same things. To cite the most publicized example, an American wine carrying the name of the grape variety, say cabernet sauvignon, may legally contain as little as 51 per cent of that wine. Yet cabernet sauvignon is the principal grape used in the great Medoc wines of Bordeaux. It has status. It takes more room in a vineyard and yields much less than a number of other red wine grapes. It is expensive.

Therefore, wouldn't it be fairer to the consumer, and the producer, to either require a greater percentage of cabernet sauvignon in the wine to at least list the percentage used?

Another bone of contention. A wine label may tell the consumer the contents are form the Napa Valley. The Napa Valley, a prime grape growing are in Northern California, has status. Land there is expensive. But it is quite possible, and perfectly legal, for a goodly percentage of the wine in the bottle to have been grown elsewhere.

Therefore, wouldn't it be fairer to the consumer, and the producer who uses only grapes from the Napa, to restrict the use of the name? For that matter, wouldn't it be fairer to define the limits of the Napa Valley, something that - for grapes growers at least - has not been done?

The answers to these questions may seem easy, but officials at the Treasury Department's Bureau of Alcohol, Tobacco and Firearms - the arm of the government that sets federal wine policy - can tell you otherwise.

Patient and attentive, Director Rex Davis and a panel of six assistants sat through two days of hearings here last week on the Bureau's proposal to create a new system of wine classification. Earlier in the mouth there had been three days of hearings in San Francisco. Even before the smoke cleared, it became obvious they would have to go back to the drawing boards. The proposed "ATF seal" category of wine, subject to special controls including an 85 per cent minimum for varietals, was opposed by industry and consumers alike.

Nearly a year ago, the bureau's proposal to turn the task of defining growing areas or appelations of origin over to the states met heated opposition at similar hearings. The states lacked expertise and personnel, it was argued. They would be too vulnerable to political pressure. BATF took the jot potato back, juggled it, and came up with the seal concept last November.

The problem with the seal, argued the Wine Institute, a trade association that speaks for most California wine makers,is that it would be taken as a mark of approval, a sanction of quality. The problem with the seal, argued several consumers, was that it would be taken as a symbol of quality. In their view the proposal not only perpetuated some avenues to deception, it opened others. The seal, they argued, would confuse the consumer without offering guarantees of quality.

Two telling examples were put forth by Robert Benson, a law professor, during testimony in San Francisco. He pointed out that a California wine that was judged first in a tasting against French wines in Paris would not be eligible for a BATF seal under the proposal because the grapes used in the wine came from two separate areas within the state.On the other hand, a Chablis made from 85 per cent table grapes and 15 per cent raisin grapes grown in the Central Valley - an area considered too hot to he hospitable to top quality wine grapes - could claim a seal.

According to Davis, as quoted in the release announcing the hearings, "Outdated and confusing wine label regulations should be replaced by incomplicated rules which tell more about what's in a wine and where it originated. This will benefit consumers and industry."

Not so, argued more than a few witnesses. One problem is that wine, unlike such beverages as orange juice and milk, is steeped in romance. The industry is American, but much of its traditions and language has been transported from Europe. Those who market and advertise wines have a vested interest in keeping alive as much of that romance, mystery and language as possible. So, it seems, do many consumers who consider themselves oenophiles and want to keep a leg up on the uninitiated.

The Bureau, an unlikely and unwilling court for finite judgements of taste and quality, is nonetheless aware of the political temper of the times. Wine regulations are outdated and confusing because, until recently, the sine industry wasn't significant enough to attract much attention, because consumer activists were fighting other battles and because no one - or at least not enough some ones - really cared.

There is concern now and it won't cease. Californians want to export wines, but U.S. regulations fail to satisfy entry standards of the Common Market and countries elsewhere. Thus there are international implications as well. On a separate front, the Bureau was slapped down last year by the Food and Drug Administration for not imposing ingredients labelling requirements on wine, beer and soirits. Thar fight has gone to the federal courts.

There is a bond of sympathy between a frustrated industry lawyer at the hearings and a consumer in the audience. The lawyer claimed "varietal percentage (on a label) is not necessary, advisable or any value as far as the wine buying public is concerned." Earlier the consumer had whispered "This is nonsence! Why do you care what's in it as long as you like it?"

From this camp the consumer is seen as an ignorant savage who should not be confused with too much knowledge or information. "I'm surprised at the assumption there is something wrong with the 51 per cent minimum," testified Abraham Buchman, who represents makers of kosher wines. He does not object to "others listing any additional information they so desire" on labels. C. Frederic Schroeder of the Finger Lakes Wine Growers Association testified that wine quality "lies in the integrity and reputation of the winery."

To anyone without a vested interest, however, some of the humming of praise and good intention was at least slightly off key. It doesn't make any difference what's blended as long as the customer is satisfied with the wine," proclaimed one witness.

In testimony at the San Francisco hearings, the spokesman for the Wine Institute worried that it would be "seriously compromising the artistic quality" of labels to add the word "brand" to a pseudo vineyard's name; that it wouldn't hurt to call a non-existent "vineyard" a vineyard as long as the name isn't actually a recognizable wine area. It was suggested that label falsehoods already on the market place continue uncorrected. Political boundaries, state the country lines, were claimed. Yet in the interest of quality such limits might more fairly be judges by similarity of climate and soil.

But wine descussion of the seal proposal did not dominate testimony at the hearings here, instead, a counterproposal made by the Wine Institute grabbed the limelight. In opposing the seal, the Institute suggested increasing the minimum percentage for all varietals from 51 per cent to 75 per cent. This is less than the 85 per cent proposed by BATF for seal wines, but it would apply to all varietals. On the other hand, the seal program would be voluntary. Those who did not choose to seek the seal would be under no pressure to increase the varietal percentage. "Instead of 85 per cent in a few wines, we'd have 75 per cent in all of them," a Bureau official said last week.

He also said, "We'd like to get this over and done with."

Yet if anything became clear during the hearings, it was realization that the debate is nowhere near over and done with. There will be a 60-day period for written comment on the current proposals. If the Bureau substantially revises them - and acceptance of the Wine Institute's 75 per cent proposal would be considered a substantial revision, it is almost certain anothet set of hearings will be demamded. If the Bureau withdraws them, what might be proposed next is anybody's guess .