THE chemicals that are replacing DDT nowadays have the advantage of breaking down more quickly in the environment, but they still hit a broad spectrum of insects-often beneficial ones.

Most of the 40 chemical companies now looking for safer, more specific insecticides simply screen thousands of chemicals each year for their possible insect-killing value. But university and government researchers, by and large, are investigating more elegant biological controls such as introducing parasites, sterile males, specific diseases or natural hormones.

For the last 15 years, a team of Harvard scientists led by Carroll M. Williams has spearheaded an international search for specific insecticides based on insects' own hormones.

WILLIAMS believes he is hot on the trail of "third generation pesticides" that are far superior to the lead arsenate and DDT that epitomize the two earlier generations.

These chemicals, called juvenile hormones, are made in two small glands in insects' heads. Juvenile hormone must be secreted at certain stages in an insect's life and not secreted at other times.

"The periods when the hormone must be absent are the Achilles' heel of insects," Williams said.

If a modicum of the hormone lands on the eggs or late larvae, the eggs fail to hatch and the larvae die without reproducing.

In 1956, Williams isolated this hormone from silkworms, and Elias J. Corey, Sheldon Emery Professor of Organic Chemistry, later synthesized it. But like many artificial versions of juvenile hormone, this silkworm hormone effected many different types of insects besides the target moth.

If hormonal insecticides were to succeed, they would have to attack only one pest, while leaving the species' natural insect predators unharmed. Otherwise, hormones would be merely a more potent variety of DDT from an ecologist's point of view.

Two findings have pointed out ways to overcome this objection.

Lynn M. Riddiford, assistant professor of Biology, and a Czech scientist discovered that juvenile hormone must be absent from insect eggs for normal hatching to take place. This finding led to the possibility of releasing males with juvenile hormone on their genitalia. Every wild female that mates with one of these males would become sterile. If enough sterilizing males are let loose, the target insect's population would drop, but the hormone would effect no other species.

The other discovery was essentially an accident. Williams' group found that a European bug would grow fairly well in the Bio Labs up until the last larval stage. Then instead of becoming an adult, it would become a giant larva and eventually die without reproducing.

This syndrome suggested that some extra juvenile hormone had hit the bugs, and sure enough, Williams' group eventually pinpointed the hormone source as the paper towels lining the insects' rearing jars.

IT TURNS out that the balsam fir used to make the paper towels produce a compound very similar to the bug's juvenile hormone. But unlike the silkworm hormone, it hits only a single family of bugs. If nature can provide such highly specific insecticides through evolution, Williams argues, man should be able to develop some himself.

With the essential principles of hormonal insecticides worked out, the third generation insecticide may be just a few years away. Field tests are set for this summer in the U.S., Switzerland, and Germany.