TWO LOS ANGELES police investigators, a police artist and a witness paid a visit in January, 1974, to Martin Reiser, the department's staff psychologist. The witness, a middle-aged woman, had been in her boyfriend's apartment on Dec. 17 at 1:30 a.m. when he was murdered. But because she had been drinking and popping pills, she couldn't remember anything except seeing another man and hearing five shots being fired outside the front door. Her boyfriend's body was found on the doorstep.

The investigators wondered if hypnosis might help the witness remember something about the suspect. The psychologist, however, was skeptical - he felt that her original perceptions would be too fuzzy. But because the woman wanted desperately to help the police find the killer, Reiser agreed to the procedure.

After the witness was put in a hypnotic state, she was told to see herself in her favorite rocking chair in front of her television. She was about to watch a special TV movie, she was told, one that could slow down, stop and zero in on scenes or people. She would stay calm and detached, she was assured, because this was, of course, only a TV show.

The woman was asked to see herself getting up and turning on the TV set and then going back to her rocking chair. While she was waiting for the imagined set to warm up, she was told that tonight's motion picture was a documentary about what happened at her boyfriend's palce on Dec. 17 from midnight to 2 in the morning. After being informed that the "television screen" was now bright, she was asked to described what was happening on the tube.

In a low voice, the woman slowly began telling about how she and her boyfriend had been talking, dancing and carrying on when a man knocked at the door. Her friend let him in.

The imaginary film was stopped for a closeup of the intruder. The woman had no trouble describing him as a tall, thin man in his early twenties. Then, with the "camera" zoomed in on his face, she clicked off his race, the shape and size of his head, nose and lips, and his hair style, including what his sideburns, mustache and beard looked like. She also gave an exact account of what he had on, from the stripes in his pants to the dots on his tie.

The woman was told that the "camera" would go in slow motion through the entire 20 minutes that it took the crime to unfold. Obediently, she recalled how her boyfriend had silently let the man in, as if he knew him, and how her boyfriend went out to the kitchen and came back with a gun in his back pocket. She remembered that both men then went outside on the front step, out of view, and that a little while later she heard the five shots.

A few months later, a man was identified in a police lineup as a dead ringer for the composite drawing. But, as it turned out, he proved not to be the murderer. Two accomplices had been waiting outside and one of them had shot the boyfriend. Both were subsequently apprehended.

If you liken conscious perception to a camera, her lenses were all fogged up with booze and pills, and she wasn't aware of any auditory or visual input," says Reiser. "But cerebrally - down at a subconcious level - she was recording this stuff, and we got it back through hypnosis."

Martin Reiser has been the Los Angeles Police Department's psychologist for the past 9 years.

"I began using hypnosis in about 1970 myself, kind of serendipitously, on request of a couple of investigators in a homicide case who had wondered if hypnosis might be useful in enhancing the recall of the witness," he says. "I was interested in that. I hadn't done that before, although I had used hypnosis before. And we got some good information from this witness. A good description of the suspect."

The word got around amond detectives and, all of a sudden, Reiser had more referrals than he could handle. He decided that the best solution was to teach seasoned police investigators how to hypnotize witnesses and victims to elicit information.

After finishing a 48-hour training course, 13 hypno-investigators, operating in pairs, went to work on 67 criminal cases. "And what we found was that in 77 percent of those cases, important information was elicited from witnesses and victims that was not available by routine interrogarion," says Reiser. "We also found that 16 percent of those cases were solved because of the use of hypnosis that likely would not have been solved at all without it, since no leads were available prior to hypnosis."

In the fall of 1976, Reiser started the Law Enforcement Hypnosis Institute to teach criminal justice people outside the police department about investigative hypnosis. Some 280 detectives, judges, district attorneys and public defenders from 21 states, the District of Columbia and Canada have completed the institute's 4-day training program. This year, three more basic seminars are on tap. And, for the first time, an advanced seminar for those who have taken the basic program and have gone back to their agencies will be held this June.

"There's a growing, tremendous interest in this area, and the reason is that it works," say Reiser. "It's been used in many cases - the Chowchilla [school bus kidnapping] case up north was a good example, where a license number was elicited from the bus driver, a key factor in solving that case."

Reiser likens the brain sensory system to a giant cybernetic system that processes information better than a dozen IBM computers. It does this at two levels: at a conscious, cortical level and at a subconscious, cerebral level. Consequently, a witness can't be eliminated as a possible information giver just because he or she doesn't consciously remember seeing or hearing anything. This is why the woman had no trouble describing her boyfriend's killer, even though she had been really out of it.

Reiser believes that trauma is a key to hypnosis investigation. Witnesses and victim are so emotionally affected by the crime that many repress information they've mentally recorded. Because of this, the psychologist says hypnosis works best on homicide, kidnap and rape cases. The Risk of "Implanted Memories"

NOT EVERYONE, however, believes that using hypnosis to jog the memories of witnesses and victims is the best police-science tool to come down the pike since patrol cars.

In December 1977, attorneys for a Santa Barbara murderer - backed by an association of criminal defense lawyers and two world-renowned authorities on hypnosis - asked the state supreme court to curb the use of hypnosis in criminal investigations. The California Attorneys for Criminal Justice, an organization of some 1,500 defense lawyers, didn't want the technique to be used except under court order and only if it was medically supervised and videotaped. (The California Supreme Court, by one vote, refused to hear the case of John Philip Quaglino. Quaglino - who was convicted of having run down his estranged wife, Dyanne, on Sept. 10, 1975 - had been identified by a witness who had undergone hypnosi.)

Two experts, Martin Orne and Ernest Hilgard, supported the attorneys' position with affidavits. Both agreed that what a hypnotized subject thinks he or she remembers may be completely inaccurate. "There are many instances where subsequently verified accurate license numbers were recalled in hypnosis by individuals who previously could not remember them," said Orne, who is president of the International Society of Hypnosis.

"By the same token, however, a good many license numbers which witnesses recalled turned out to belong to individuals where neither they nor their cars could possibly have beem implicated. Unfortunately, a witness who is uncertain about his recall of a particular set of events can, with hypnosis, be helped to have absolute subjective conviction about what had happened - though the certainty can as easily relate to a confabulation [the making up of details to fill in memory gaps] as to an actual memory."

Hilgard, director of Stanford University's Laboratory of Hypnosis Research, added: "It is well-known that hypnotists may implant memories, so that the hypnotized person accepts them as his own."

None of this fazes Martin Reiser. "We know by corroboration when we're getting facts and when we're not" he snaps. "What I'm saying is that in 80 to 90 percent of ourcases we are able to get subsequent corroboration of a composite drawing of suspects, license numbers, vehicle descriptions. So this charge is patently ridiculous."

Reiser says the academicians don't know what they're taling about. "The fact that they have worked in therapeutic hypnosis doesn't make them experts in investigative hypnosis," he argues. Unlike the therapist, the hyno-investigator isn't interested in underlying conflicts and unconscious material. Like Joe Firday, he just wants the facts - what has been perceived and recorded by the senses in a certain situation. Never Work With Suspects

REISER ADMITS that confabulation might happen. Hyponosis isn't a lie-detector technique. People under it can tell the biggest cock-and bull stories with straight faecs. "But keep in mind that we're dealing with volunteers, motivated witnesses and victims," the police psychologist points out. "We're not after confessions or convictions of them. ANd so those people really have very little motivation to lie."

Reiser and his students never work with suspects or possible suspects. "We don't coerce people into remembering things against their wills that could be harmful to them," he says. "They are their own pacers of what they feel comfatable in remembering."

The Los Angeles police have never had a report of a subject suffering mental or emotional damage because he or she was hypontized. In fact, witnesses and victims - especially rape victims - have said that after their sessions they slept better.

Most of the data gleaned by the police department through hypnosis are only used by investigators as leads to further information. Seldom, if ever, are they brought into court as evidence - even though Reiser says they could be. He says the 9th U.S. Circuit Court of Appeals has indicated that the question isn't one of the admissibility of testimony elicited under hypnosis, but how much weight should be given to that testimony. And that's up to a jury to decide.