WHEN DEMANDS are heard for a longer school day or a longer school year, the usual retort is that schools don't make enough use of the time they already have. The truth of this criticism varies with the school, but it underlines the need to take a close look at experiments in individual schools that seem to use students' and teachers' time better. One such example is the experiment now in its third year at Richardson Elementary School in the District, where a teacher, Johnny Brinson, has been teaching one group of youngsters ever since they were first-graders and means to "follow" them all the way through sixth grade if he can.

The appeal of this model is easy to appreciate. Equally clear are the complications that might result if it became the across-the-board policy of a school system or even a school. The continuing attention of one teacher fosters both efficiency and stability -- no need to learn students' names or test for their ability levels in the fall; no need to start from scratch every year earning their trust or building bridges to their sometimes diffident parents. The children in the Richardson experimental class (there are just 23 of them) come mostly from the sort of homes where stability is lacking and a continuing, reassuring adult presence can help a lot. (In European countries such as West Germany and Italy, where multi-year teacher assignments are common, they are often used to smooth the path of disadvantaged children or those from ethnic minorities.) Mr. Brinson's bond with the youngsters was already established when, after they finished first grade, he suggested the experiment -- a harbinger of what so far seems to be its success.

But the scheme's possible drawbacks will be equally obvious to any parent who remembers the September scramble to make sure his or her kids don't get stuck for a whole year with the "wrong" teacher or just one with whom the child doesn't get along. Extended from one year to three or six, classroom and home-room assignments can come to resemble an inadvertent form of ability tracking; worse, within one group they can lock kids into a hierarchy of supposed strengths and weaknesses established early on. These troubles are greatly minimized if the multi-year setup is less than rigid or if it exists as one option among many.

Some provocative (and paradoxical) research in recent years has pointed to the strength of the old one-room schoolhouse that still exists in some tiny towns; students at least work in a stable environment over the years, and they avoid the impersonality of big, ever-shifting public school classes. Multi-year teaching assignments can preserve the stability of that old system without the accompanying chaos of many ages being taught together; and it makes sense to include it as yet another possible tool in the arsenal.