ALTERNATE JURORS often must feel like the wallflowers of the judicial system.

They are selected at the beginning of a trial to sit in the wings. They have to take the same amount of time off from work as regular jurors to listen to testimony, but they step in only if one of the 12 "real" jurors becomes ill or has to be dismissed from the case. When the jury's deliberations begin, the alternates are told "thank you" and are sent home.

They know from the trial's first day that they are only understudies, and for years lawyers have complained that alternate jurors react to that role by paying little, if any, attention to the cases they are hearing.

In a case now in progress in a federal court in Florida, though, a technique is being used to try to fight that problem.

In that case, a complex medical malpractice action being presided over by senior U.S. District Judge Walter E. Hoffman, the 10 jurors (the normal six in a federal civil case, plus four alternates) have not been told whether they are alternates or regular jurors. They won't know, in fact, until the time comes for deliberations to begin.

It is not the first time such an approach has been used -- Judge Hoffman apparently has used it for some time -- but several judges, attorneys and others in the legal system here believe it is a good idea that possibly should be tried in Washington.

Denver Graham, a personal injury lawyer who specialized in medical malpractice cases and whose litigation schedule may be one of the busiest in the District, says he has never heard of the practice. But "it is one of the new innovations" in the traditional jury system that wouldn't bother him, he said. "It doesn't do violence with my idea of justice."

"It has been my feeling that alternates do not pay that much attention," said one federal judge, who asked not to be identified. Instead, he said he tries to give the alternates regular reminders that they must pay attention because they might be called upon at any time the place of a juror who became ill.

"Maybe a practice such as the one used by Judge Hoffman would be useful in keeping alternates on their toes," the judge said.

Arthur Cohen, an attorney for the plaintiffs in the malpractice case before Hoffman, said he thinks "it's a swell idea. We all agreed because it'll keep alternates awake. I think in the future I will suggest it."

The lawyers in the malpractice case know which jurors are alternates and which are not. There is probably a strong argument that it is unnecessary to play games by withholding such information from the affected jurors, and that when the persons are ultimately told that they are alternates, they might be disappointed or angry when they don't get to join in the deliberations.

Until the approach, is used more often, though, there will always be a lingering doubt about the dedication of the extra two (or more) alternates in the jury box.

The three key concepts of the H-bomb secret are: separate stages, compression and radiation coupling.

Nine months ago, a judge wrote an important legal opinion about those concepts. You could not read the opinion then and I could not write about it. It was secret, one of that handful of instances in which a legal opinion is kept from the eyes of the public for any period of time.

The case: U.S.A. vs. The Progressive. A magazine was going to publish an article about the H-bomb secret and the government successfully blocked its publication for several months. It was prior restraint of an unheard-of magnitude in this country and the subject of much debate and First Amendment concern.

U.S. District Judge Robert W. Warren accepted the government's argument that his ruling must be made in secret. The American Civil Liberties Union argued at the time that the opinion should have been released in its entirety because the weaponry concepts had been published elsewhere, but that even if the discussion of the concepts had to be omitted (about 150 words of a seven-page opinion) there was nothing to prevent the publication of the rest of the ruling.

Two weeks ago, after the case was moot because the article had been published, Judge Warren quietly released his opinion to the public.

One reading of that opinion would be that the government cried "wolf" once again and that the ACLU had a strong argument for having the opinion -- or at least most of it -- made public in the first place.

Much of the opinion dealt with the background of the case and legal arguments. In addition, it shows the judge really ruled in the government's favor only because he thought one of the three key concepts -- not all three -- had been discussed publicly before.

"Throughout the litigation, however, the government continued to claim that all three concepts were secret, a claim the government could not so cavalierly have made had the opinion itself not been secret," said ACLU legal director Bruce Ennis.

Meanwhile, the ACLU and the government are still negotiating over the release of further information filed secretly in the case.

There was an interesting article in last week's Legal Times here that was written by "David Chase," identified as a "frequent observer of the legal scene."

The article commented on alleged inaccuracies in a recent survey in another, unnamed publication (actually a Times competitor, the National Law Journal) and suggested at one point that persons taken in by the alleged inaccuracies should complain to the people who publish such polls.

My sources, inadmissible in court as their evidence may be, tell me "David Chase" is in reality better known as Stephen A. Glasser, publisher of the Legal Times.