Science may be on the way to repairing injured brains.

Four scientists report that they have dramatically improved the learning ability of rats with severely damaged brains by implanting fresh brain tissue from rat fetuses.

Someday, said one of the experimenters, Dr. Donald Stein of Clark University in Worcester, Mass., "we may be able to clone fetal brain cells to grow them in plentiful supply, then insert these healthy cells" into a damaged brain.

Scientists in laboratories in the United States and Europe are in the midst of a growing number of experiments to hasten that day.

"This is a very fast-moving and exciting field," Stein said. "It has captured the imagination of a great many people."

Randy Labbe, Arthur Firl Jr. and Stein at Clark University's Brain Research Center and Dr. Elliott Mufson of Boston's Beth Israel Hospital tell of the latest step forward in the current issue of the journal Science.

They said that they destroyed part of the frontal cortex, part of the brain's topmost layer or "thinking cap," in 21 adult male rats. The section removed governs the way animals, including humans, learn to do tasks in the proper order.

"If this part of your brain were damaged it would impair your ability to recognize right from left or orient yourself in space," Stein said. "If I told you I wanted you to do A, then B, then C, you couldn't put the sequence together. You might do B, then C or A. Or you might never get past A."

Seven days after their initial surgery, eight of the rats were given implants of frontal cortex tissue from rat fetuses. Six others were given transplants from a different part of the brain, the cerebellum, and seven received no implants.

Four days after the transplant surgery, the rats began training in a maze. They had to enter the base of a 'T,' then go to the top and turn left or right to get a drink of water.

"What the rat has to learn," Stein said, "is that if he gets a drink one time by turning left, he'll get one next time by turning right. He has to learn to alternate."

To pass the test a rat has to be right 90 percent of the time, Stein said.

Rats with no brain damage can learn this in about 2 1/2 days, given 10 test runs a day.

The test rats with damaged and unrepaired brains took 17 or 18 days, on average, to learn it. The rats who received the cerebellum transplant, the inappropriate repair, learned it in 18 days. But the rats with the cortex transplants, the correct repair, took just 8 1/2 days.

"The treated animal doesn't perform as well as a completely healthy, intact animal, but he certainly does better than the untreated one," Stein said.

Earlier this year, Steven Dunnett of Cambridge University in England also reported improving brain-damaged rats' maze-learning by an implant of fetal brain tissue. Whereas the Clark University scientists began maze-training four days after the brain transplants, Dunnett waited seven months.

The Clark scientists thus believe that the transplants start enhancing behavioral recovery almost immediately, with continued effect for at least a large part of a year and perhaps for the rest of the animal's life.

Doctors and scientists elsewhere are making similar brain transplants to try to cure other brain, nerve and vision defects in experimental animals. They are also injecting chemicals to try to reverse spinal cord injuries.

"The most important thing is that we are learning to understand some of the processes involved in brain and nerve damage," Stein concluded. "We're learning much more about the growth properties of cells. That's what's so exciting."