I'm relieved. Dr. James T. Guines, acting superintendent of the D.C. public schools, says he isn't chickening out on the school board's decision to stop promoting elementary school children to the next grade before they have mastered the one they're in.

All he has in mind -- tentatively, at that -- is a modest modification of the plan.

Instead of retaining first- or second-graders who fail to learn the reading and math skills appropriate to their grade level, he would make the end of third grade the first critical checkpoint.

It's helpful, he said in a Monday interview, to think of the first three years as a "block" -- a period for learning the basic reading and math concepts that, beginning in fourth grade, the children will then learn to apply.

Viewed this way, he says, there's really not that much difference in the things that first-and second -graders are expected to learn: basic reading and math skills. If the children don't learn these things in first grade, they get another shot in second. And if they learn them superlatively well, they can move to more advanced work without the necessity of midterm promotions. in any case, they must acquire the basic skills before being promoted to fourth grade.

What Guines is talking about is a little like the old track system and a lot like the 1950s notion of ungraded primaries. A lot of educators will tell you it makes sense.

But some of us who have been watching the public schools for a few years get nervous at the first hint of promoting youngsters who have not mastered what they are supposed to have been taught.

We remember the Northeast school whose principal used the ungraded primary idea for her entire elementary school. Instead of assigning children to, say, third grade during their third year at her school, each child was assigned to a reading or math group based on the level of work he was able to perform at the time.

That one made sense, too, and the children at this school -- particularly the third-graders -- routinely tested at well above the city-wide norms on the annual achievement test.

Then it was revealed that things weren't quite what they seemed. When the time came for the annual standardized tests, this principal grouped her children according to the work they were actually doing, no matter how many years they had been in school. A fifth-year child doing second -grade work was tested as a second-grader. Naturally the scores looked good.

The crunch came only at the end, when the children were ready to move into junior high. Only then was it revealed that some of her high-scoring children were doing no better than their counterparts in some of the poorest-performing schools in the city. And by then it was too late.

Jim Guines knows about that situation and insists that they system he is thinking about won't turn out that way.

Under his modification, the schools would go right on testing to see if the children had mastered the reading and math skills appropriate to their grade level. Teachers would still use whatever resources they could muster -- including tutors -- to bring up to par those students who failed to master the skills in time.

But promotion would not become an issue until it was time to move to fourth grade. Thus, he said, the schools would be able to take into account the fact that some children take a little longer than others to master the basic skills. In addition, he said, there would be less of the psychic damage caused by labeling small children as failures.

Parents, meanwhile, would be undere no illusion. They would be on notice that there was catch-up work to be done.

And that, he insists, is the key point: The skills would still have to be acquired.

"I have no intention of backing away from the strict new standards we have established," he said. "after all, I'm the same guy who fought for putting the scores out there so the public could see what was going on. I'm the same guy who opposed measuring our children against the (lower) big-city norms rather than the norms for all the children in the country. I'm still for standards."

It doesn't sound unreasonable the way Guines describes it. But it does warrant some careful watching.